CHEMICAL THERMODYNAMICS OF
NICKEL
Heinz GAMSJÄGER (Chairman)
Lehrstuhl für Physikalische Chemie
Montanuniversität Leoben
Leoben (Austria)
Jerzy BUGAJSKI
Tamás GAJDA
Lehrstuhl für Physikalische Chemie
Montanuniversität Leoben
Leoben (Austria)
Department of Inorganic
and Analytical Chemistry
University of Szeged
Szeged (Hungary)
Robert J. LEMIRE
Wolfgang PREIS
Deep River
ON K0J 1P0
(Canada)
Lehrstuhl für Physikalische Chemie
Montanuniversität Leoben
Leoben (Austria)
Edited by
Federico J. MOMPEAN (Series Editor and Project Co-ordinator)
Myriam ILLEMASSÈNE (Volume Editor) and Jane PERRONE
OECD Nuclear Energy Agency, Data Bank
Issy-les-Moulineaux (France)
i
Federico J. Mompean
TDB Project Coordinator (E-Mail: tdb at nea.fr)
OECD NEA Data Bank
November 2005
Special thanks go to Prof. Heinz Gamsjäger (Leoben), Prof. Lars-Olof Öhman (Umeå), Dr. Malcolm Rand (Abingdon) and Dr. Lian Wang (Mol) for
their efforts to ensure the consistency of text, tables and data files.
Errata are primarily listed by page number but several entries corresponding to the same system have been grouped together for the sake of clarity.
Authors and readers have pointed out several errata in the 2005 Review on Nickel. These errata are corrected below.
Gamsjäger, H., Bugajski, J., Gajda, T., Lemire, R. J., Preis, W., Chemical thermodynamics of Nickel, Nuclear
Energy Agency Data Bank, Organisation for Economic Co-operation and Development, Ed., vol. 6, Chemical
Thermodynamics, North Holland Elsevier Science Publishers B. V., Amsterdam, The Netherlands, (2005).
Errata for the 2005 Review on the Chemical Thermodynamics of Nickel
Reads
vol. 6
584, line 4
vol. 7
ε(PuCl2+, ClO −4 ) = Delete entry.
(0.39 ± 0.16) See accompanying
notes.
426, entry for PuCl2+ in
Table B-4.
[92GRE/FUG]
Delete duplicated
(second) entry for
HSO −4
i = – 5.30393E+04
[92GRE/FUG]$
––
i = 5.30393E+04
71, entry for Si(OH)4(aq)
58, second entry for HSO −4
52, entry for NiO(cr) for
298.15 K < T < 519 K
78, Table V-1 last row
c = – 4.45300E–06
Typographical error.
The auxiliary formation values for this species have been adopted from
CODATA Key values at the time of the first NEA TDB Review,
[92GRE/FUG].
Typographical error.
Typographical error.
No
Typographical error.
For modeling An(III) in chloride solutions, a suggested procedure is to
use the SIT ε coefficients based on ε(Nd3+, Cl–) = ε(Am3+, Cl–) =
(0.23 ± 0.02).
Yes
The 2003 Update [2003GUI/FAN] did not retain the selection made by
(SIT interaction [2001LEM/FUG] for log10 β ο for the reaction Pu3+ + Cl– U PuCl2+ for
the reasons given in section 11.3.1.1, page 322 of [2003GUI/FAN].
coefficients)
No
No
Yes
(Heat capacity
tables)
Yes
(Heat capacity
tables)
Typographical error.
Affects selected Notes
values?
Add in Participating
No
Organisations
NAGRA, Switzerland
Should read /
corrective action
Errata for the 2005 Review on the Chemical Thermodynamics of Nickel
52, entry for Ni(cr) for 690
c = 4.45300E–06
K < T < 1728 K
x
Page and position in text
2/2
Preface
This volume is the sixth of the series "Chemical Thermodynamics" edited by the OECD
Nuclear Energy Agency (NEA). It is a critical review of the Thermodynamic Data Base
of nickel and its compounds initiated by the Management Board of the NEA Thermochemical Database Project Phase II (NEA TDB II).
The TDB Ni review team first met at the University of Leoben, Austria in
January 1999. Three subsequent plenary meetings were held at NEA Headquarters at
Issy-les-Moulineaux (France) in January 2001, October 2001 and in November 2002,
two were held at Leoben in May 2002 and in November 2003. Smaller working subgroups met at Luleå, Sweden in October 1999 and in April 2001. The Executive Group
of the Management Board provided scientific assistance in the implementation of the
NEA TDB Project Guidelines. Hans Wanner participated in meetings of the Review
Team as the designated member of the Executive Group. At the NEA Data Bank the
responsibility for the overall co-ordination of the Project was placed with Eric Östhols
(from its initiation in 1999 to February 2000), with Stina Lundberg (from March 2000
to September 2000) and with Federico Mompean (since September 2000). Federico
Mompean was in charge of the preparation of the successive drafts, updating the NEA
thermodynamic database and editing the book to its present final form, with assistance
of Myriam Illemassène, Jane Perrone, Katy Ben Said and Cristina Domènech-Ortí.
Originally Willis Forsling, Lars Gunneriusson (Luleå University of Technology, Sweden), Jerzy Bugajski, Heinz Gamsjäger, and Wolfgang Preis (University of
Leoben, Austria) participated in the nickel project. In October 2001 Robert Lemire
joined the Review Team. As member, chairman and peer reviewer of previous NEA
TDB projects he contributed invaluable expertise to seeing the Ni review through to
completion. In August 2002 time constraints and the pressure of other commitments
forced Willis Forsling and Lars Gunneriusson to resign from the Review Team and
Tamás Gajda, University of Szeged, Hungary, joined as a new member.
Although almost all of the members of the final review team contributed text
and comments to several chapters, primary responsibility for the different chapters was
divided as follows. Jerzy Bugajski prepared the sections on elemental nickel, silicon
compounds and complexes, germanium compounds and complexes and, with the
chairman, the section on boron compounds and complexes. Tamás Gajda drafted the
sections for halide and pseudohalide complexes, the section for hydroxo complexes, and
the sections on nitrato and thiocyanato complexes. Robert Lemire prepared Chapter VI,
Discussion of auxiliary data selection, the sections on halates, sulphates, phosphorus
compounds and complexes, arsenic compounds and complexes and, with the chairman,
the section on solid halides (he also extensively reviewed several of the other sections,
and revised the English throughout the text). Wolfgang Preis prepared the sections on
the oxide, the sulphides and the hydrogensulphido complexes and, with the chairman,
v
vi
PREFACE
the section on Ni(III,IV) hydroxides (for the Leoben group he also implemented a data
base for ionic strength corrections and calculations ensuring internal consistency of the
selected values). The chairman drafted the sections on aqua ions, hydroxides and carbonates.
The initial contributions from Willis Forsling and Lars Gunneriusson on halide
and pseudohalide complexes and (together with Wolfgang Preis) on hydroxo complexes
are gratefully acknowledged as they constituted the starting point for subsequent discussions within the Review Team.
While there is no need to repeat the general purpose of this review a side effect
deserves mentioning. The selection of key values, e.g., for Ni2+(aq), revealed gaps in our
knowledge which may stimulate rewarding projects on the experimental thermodynamics of nickel compounds.
Leoben, Austria, October 2004
Heinz Gamsjäger, Chairman
Acknowledgements
For the preparation of this book, the authors have received financial support from the
NEA TDB Phase II Project. The following organisations take part in the Project:
ANSTO, Australia
ONDRAF/NIRAS, Belgium
RAWRA, Czech Republic
POSIVA, Finland
ANDRA, France
IPSN (now IRSN), France
FZK, Germany
JNC, Japan
ENRESA, Spain
SKB, Sweden
SKI, Sweden
HSK, Switzerland
NAGRA, Switzerland
PSI, Switzerland
BNFL, UK
NIREX, UK
DoE, USA
Jerzy Bugajski, Heinz Gamsjäger and Wolfgang Preis would like to express
their gratitude to the Institut für Physikalische Chemie of the Montanuniversität Leoben
for having provided the infrastructure necessary for their contributions to this project.
Tamás Gajda gratefully acknowledges the technical support of the Department
of Inorganic and Analytical Chemistry, University of Szeged.
Robert Lemire wishes to thank Atomic Energy of Canada Limited (AECL) for
allowing him to participate in this project, and for allowing him to use the library and
other AECL facilities at the Chalk River Laboratories.
The authors also thank Werner Sitte for his help with and comments on the
section of silicon compounds and complexes. Heinz Gamsjäger thanks the library staff
at the University of Leoben and the Austrian Central Library for Physics for their literature searches and assistance in obtaining copies of many references, he also appreciates
the help of Malcolm Rand, Alan Dinsdale, Heiko Kleykamp and Herbert Ipser for their
vii
viii
ACKNOWLEDGEMENTS
advice concerning papers difficult to come by.
The unceasing efforts of Federico Mompean, coordinator for the TDB project
during the time the main part of the draft of this book was assembled, edited for peer
review and finally prepared for the press, is greatly appreciated. His work built on the
earlier efforts of Erik Östhols and Stina Lundberg. The authors are indebted to Myriam
Illemassène, Jane Perrone and Katy Ben Said who, with admirable alertness and scientific competence, transformed contributions of five authors, prepared in many different
formats, into a consistent text with correctly numbered tables and figures. At the NEA
Data Bank, Pierre Nagel and Eric Lacroix have provided excellent software and advice,
which have eased the editorial and database work. Cynthia Picot, Solange Quarmeau
and Amanda Costa from NEA Publications have provided considerable help in editing
the present series. Their contributions and the support of many NEA staff members are
highly appreciated.
The entire manuscript of this book has undergone a peer review by an independent international group of reviewers, according to the procedures in the TDB-6
Guideline, available from the NEA. The peer reviewers have seen and approved the
modifications made by the authors in response to their comments. The peer review
comment records may be obtained on request from the OECD Nuclear Energy Agency.
The peer reviewers were:
Dr. Donald A. Palmer, Batelle, Oak Ridge National Laboratory, USA
and
Prof. Wolfgang Voigt, University of Freiberg, Germany.
Their contributions are gratefully acknowledged.
Note from the Chairman of the
NEA TDB Project Phase II
The need to make available a comprehensive, internationally recognised and qualityassured chemical thermodynamic database that meets the modelling requirements for
the safety assessment of radioactive waste disposal systems prompted the Radioactive
Waste Management Committee (RWMC) of the OECD Nuclear Energy Agency (NEA)
to launch in 1984 the Thermochemical Database Project (NEA TDB) and to foster its
continuation as a semi-autonomous project known as NEA TDB Phase II in 1998.
The RWMC assigned a high priority to the critical review of relevant chemical
thermodynamic data of inorganic species and compounds of the actinides uranium,
neptunium, plutonium and americium, as well as the fission product technetium. The
first four books in this series on the chemical thermodynamics of uranium, americium,
neptunium and plutonium, and technetium originated from this initiative.
The organisation of Phase II of the TDB Project reflects the interest in many
OECD/NEA member countries for a timely compilation of the thermochemical data that
would meet the specific requirements of their developing national waste disposal programmes.
The NEA TDB Phase II Review Teams, comprising internationally recognised
experts in the field of chemical thermodynamics, exercise their scientific judgement in
an independent way during the preparation of the review reports. The work of these
Review Teams has also been subjected to further independent peer review.
Phase II of the TDB Project consisted of: (i) updating the existing, CODATAcompatible database for inorganic species and compounds of uranium, neptunium, plutonium, americium and technetium, (ii) extending it to include selected data on inorganic species and compounds of nickel, selenium and zirconium, (iii) and further adding
data on organic complexes of citrate, oxalate, EDTA and iso-saccharinic acid (ISA)
with uranium, neptunium, plutonium, americium, technetium, nickel, selenium, zirconium and some other competing cations.
The NEA TDB Phase II objectives were formulated by the 17 participating organisations coming from the fields of radioactive waste management and nuclear regulation. The TDB Management Board is assisted for technical matters by an Executive
Group of experts in chemical thermodynamics. In this second phase of the Project, the
ix
x
NOTE FROM THE CHAIRMAN OF THE NEA TDB PROJECT PHASE II
NEA acts as coordinator, ensuring the application of the Project Guidelines and liaising
with the Review Teams.
The present volume is the second one published within the scope of NEA TDB
Phase II and contains a database for inorganic species and compounds of nickel. We
trust that the efforts of the reviewers, the peer reviewers and the NEA Data Bank staff
merit the same high recognition from the broader scientific community as received for
previous volumes of this series.
Mehdi Askarieh
United Kingdom Nirex limited
Chairman of TDB Project Phase II Management Board
On behalf of the NEA TDB Project Phase II Participating Organisations:
ANSTO, Australia
ONDRAF/NIRAS, Belgium
RAWRA, Czech Republic
POSIVA, Finland
ANDRA, France
IPSN (now IRSN), France
FZK, Germany
JNC, Japan
ENRESA, Spain
SKB, Sweden
SKI, Sweden
HSK, Switzerland
PSI, Switzerland
BNFL, UK
Nirex, UK
DoE, USA
Editor’s note
This is the sixth volume of a series of expert reviews of the chemical thermodynamics
of key chemical elements in nuclear technology and waste management. This volume is
devoted to the inorganic species and compounds of nickel. The tables contained in
Chapters III and IV list the currently selected thermodynamic values within the NEA
TDB Project. The database system developed at the NEA Data Bank, see Section II.6,
assures consistency among all the selected and auxiliary data sets.
The recommended thermodynamic data are the result of a critical assessment
of published information.. The values in the auxiliary data set, see Tables IV-1 and IV-2
have been adopted from CODATA key values or have been critically reviewed in this or
earlier volumes of the series.
xi
xii
How to contact the NEA TDB Project
Information on the NEA and the TDB Project, on-line access to selected data
and computer programs, as well as many documents in electronic format are available at
www.nea.fr.
To contact the TDB project coordinator and the authors of the review reports,
send comments on the TDB reviews, or to request further information, please send
e−mail to [email protected]. If this is not possible, write to:
TDB project coordinator
OECD Nuclear Energy Agency, Data Bank
Le Seine-St. Germain
12, boulevard des Îles
F-92130 Issy-les-Moulineaux
FRANCE
The NEA Data Bank provides a number of services that may be useful to the
reader of this book.
•
The recommended data can be obtained via internet directly from the
NEA Data Bank.
•
The NEA Data Bank maintains a library of computer programs in various areas. This includes geochemical codes such as PHREEQE, EQ3/6,
MINEQL, MINTEQ and PHRQPITZ, in which chemical thermodynamic data like those presented in this book are required as the basic input data. These computer codes can be obtained on request from the
NEA Data Bank.
Contents
Preface
v
Acknowledgement
vii
Note from the chairman of the NEA TDB Project Phase II
ix
Editor’s note
xi
Part I
1
Introductory material
I INTRODUCTION
3
I.1
Background .........................................................................................................3
I.2
Focus of the review .............................................................................................5
I.3
Review procedure and results............................................................................6
II STANDARDS, CONVENTIONS, AND CONTENTS OF THE TABLES
II.1
11
Symbols, terminology and nomenclature .......................................................11
II.1.1 Abbreviations ..............................................................................................11
II.1.2 Symbols and terminology............................................................................13
II.1.3 Chemical formulae and nomenclature .........................................................15
II.1.4 Phase designators.........................................................................................15
II.1.5 Processes .....................................................................................................17
II.1.6 Equilibrium constants ..................................................................................18
II.1.6.1
Protonation of a ligand...........................................................................18
II.1.6.2
Formation of metal ion complexes.........................................................19
II.1.6.3
Solubility constants................................................................................20
II.1.6.4
Equilibria involving the addition of a gaseous ligand............................21
II.1.6.5
Redox equilibria.....................................................................................22
II.1.7 pH ................................................................................................................25
II.1.8 Order of formulae ........................................................................................27
II.1.9 Reference codes...........................................................................................28
II.2 Units and conversion factors............................................................................28
II.3
Standard and reference conditions .................................................................31
II.3.1
II.3.2
II.3.3
Standard state ..............................................................................................31
Standard state pressure ................................................................................32
Reference temperature.................................................................................35
xiii
xiv
CONTENTS
II.4
Fundamental physical constants......................................................................35
II.5
Uncertainty estimates .......................................................................................36
II.6
The NEA-TDB system ......................................................................................36
II.7
Presentation of the selected data .....................................................................38
Part II
Tables of selected data
41
III SELECTED NICKEL DATA ...............................................................................43
IV SELECTED AUXILIARY DATA ........................................................................53
Part III Discussion of data selection
V
73
DISCUSSION OF DATA SELECTION FOR NICKEL ..................................75
V.1
Elemental nickel........................................................................................................... 75
V.1.1 Nickel gas ....................................................................................................75
V.1.2 Nickel crystal...............................................................................................76
V.1.3 Nickel liquid ................................................................................................78
V.2 Simple nickel aqua ions ....................................................................................79
V.2.1 Ni2+ ..............................................................................................................79
V.2.1.1
Gibbs energy of formation of Ni2+ .........................................................80
V.2.1.1.1 Temperature coefficient of the standard electrode potential Ni2+ | Ni
..........................................................................................................82
V.2.1.2
Enthalpy of formation of Ni2+ ................................................................82
V.2.1.3
Partial molar entropy of Ni2+ .................................................................84
V.2.1.3.1 NiSO4·7H2O(cr) cycle.......................................................................84
V.2.1.3.2 NiCl2·6H2O(cr) cycle ........................................................................85
V.2.1.3.3 Entropy of Ni2+ from its ∆ f Gmο and ∆ f H mο ......................................86
V.2.1.4
Heat capacity of Ni2+ .............................................................................86
V.2.2 Other oxidation states ..................................................................................88
V.2.2.1
Ni3+ ........................................................................................................88
V.3 Oxygen and hydrogen compounds and complexes ........................................90
V.3.1 Aqueous nickel(II) hydroxo complexes.......................................................90
V.3.1.1
Hydroxo complexes in acidic or neutral solutions .................................90
V.3.1.2
Hydroxo complexes in alkaline solutions ..............................................99
V.3.2 Solid nickel oxides and hydroxides ...........................................................102
V.3.2.1
Ni(II) oxide ..........................................................................................102
V.3.2.1.1 Solubility measurements.................................................................107
V.3.2.2
Ni(II) hydroxides, Ni(OH)2(cr) ............................................................108
V.3.2.2.1 Crystallography and mineralogy of nickel hydroxide.....................108
CONTENTS
xv
V.3.2.2.1.1 β-Ni(OH)2 ...................................................................................108
V.3.2.2.1.2 α-Ni(OH)2 ...................................................................................109
V.3.2.2.2 Heat capacity and entropy of Ni(OH)2(cr) ......................................109
V.3.2.2.3 Thermodynamic analysis of solubility data of Ni(OH)2(s) .............111
V.3.2.3
Ni(III, IV) hydroxides..........................................................................115
V.3.2.3.1 β-NiOOH ........................................................................................116
V.3.2.3.2 γ-NiOOH.........................................................................................118
V.4 Group 17 (halogen) compounds and complexes...........................................119
V.4.1 Nickel halide compounds ..........................................................................119
V.4.1.1
Solid nickel fluoride, NiF2(cr) .............................................................119
V.4.1.1.1 Enthalpy of formation of NiF2(cr) ..................................................120
V.4.1.2
Solid nickel chloride, NiCl2(cr) ...........................................................122
V.4.1.2.1 Low-temperature heat capacity and entropy of NiCl2(cr) ...............122
V.4.1.2.2 The high-temperature heat capacity of NiCl2(cr)............................124
V.4.1.2.3 Enthalpy of formation of NiCl2(cr).................................................124
V.4.1.2.4 Gibbs energy of formation of NiCl2(cr) ..........................................127
V.4.1.3
Hydrated NiCl2 solids ..........................................................................128
V.4.1.3.1 Gibbs energy of formation of NiCl2·6H2O......................................128
V.4.1.3.2 Enthalpy of formation of NiCl2·6H2O.............................................128
V.4.1.3.3 The entropy and heat capacity of NiCl2·6H2O(cr) ..........................129
V.4.1.3.4 Gibbs energy and enthalpy of formation of nickel chloride
tetrahydrate .....................................................................................130
V.4.1.3.5 The entropy and heat capacity of nickel chloride dihydrate ...........131
V.4.1.3.6 Gibbs energy and enthalpy of formation of nickel chloride dihydrate
........................................................................................................131
V.4.1.4
Solid nickel bromide, NiBr2(cr) ...........................................................132
V.4.1.4.1 Low-temperature heat capacity and entropy of NiBr2(cr)...............132
V.4.1.4.2 The high-temperature heat capacity of NiBr2(cr)............................134
V.4.1.4.3 Enthalpy of formation of NiBr2(cr).................................................134
V.4.1.5
Hydrated solid, NiBr2·xH2O.................................................................135
V.4.1.6
Solid nickel iodide, NiI2(cr) .................................................................136
V.4.1.6.1 Entropy of NiI2(cr)..........................................................................136
V.4.1.6.2 Heat capacity of NiI2(cr).................................................................136
V.4.1.6.3 Enthalpy of formation of NiI2(cr) ...................................................137
V.4.2.7
Solid nickel iodate compounds ............................................................137
V.4.2 Aqueous nickel halide complexes .............................................................140
V.4.2.1
Introduction..........................................................................................140
V.4.2.2
Solution structural studies....................................................................141
V.4.2.3
Aqueous Ni(II) - fluoro complexes......................................................142
V.4.2.4
Aqueous Ni(II) - chloro complexes .....................................................146
V.4.2.5
Aqueous Ni(II) – bromo complexes.....................................................153
V.4.2.5.1 Determination of the Ni2+ – Br– ion interaction coefficient ............156
xvi
CONTENTS
V.4.2.6
Aqueous Ni(II) – iodine complexes .....................................................157
V.4.2.6.1 Aqueous Ni(II) – iodo complexes...................................................157
V.4.3 Determination of the Ni2+ – ClO −4 ion interaction coefficient..................158
V.5 Group 16 (chalcogen) compounds and complexes .......................................161
V.5.1 Sulphur compounds and complexes ..........................................................161
V.5.1.1
Nickel sulphides...................................................................................161
V.5.1.1.1 Solid nickel sulphides .....................................................................161
V.5.1.1.1.1 Crystallography and mineralogy of nickel sulphides ..................161
V.5.1.1.1.1.1
Heazlewoodite, Ni3S2(cr) ...................................................161
V.5.1.1.1.1.2
α-Ni7S6 ...............................................................................161
V.5.1.1.1.1.3
Pentlandite, (Fe, Ni)9S8(cr) ................................................161
V.5.1.1.1.1.4
Godlevskite, Ni9S8(cr)........................................................162
V.5.1.1.1.1.5
Millerite, β-NiS ..................................................................162
V.5.1.1.1.1.6
Polydymite, Ni3S4(cr).........................................................162
V.5.1.1.1.1.7
Vaesite, NiS2(cr) ................................................................163
V.5.1.1.1.2 Ni – S phase diagram ..................................................................163
V.5.1.1.1.2.1
Ni – S melt .........................................................................166
V.5.1.1.1.2.2
β1-Ni3S2 and β2-Ni4S3 ........................................................167
V.5.1.1.1.2.3
Ni3S2(cr) (heazlewoodite) ..................................................167
V.5.1.1.1.2.4
Ni7S6(cr) and Ni9S8(cr).......................................................168
V.5.1.1.1.2.5
NiS(α) and NiS(β) .............................................................170
V.5.1.1.1.2.6
NiS2(cr), vaesite .................................................................173
V.5.1.1.1.2.7
Ni3S4(cr), polydymite.........................................................174
V.5.1.1.1.2.8
Ni4.5Fe4.5S8(cr), pentlandite ................................................174
V.5.1.1.1.3 Discussion of selected thermodynamic properties of nickel
sulphides......................................................................................174
V.5.1.1.2 Solubility of NiS(s) and aqueous nickel hydrogen sulphido species
........................................................................................................177
V.5.1.2
Nickel sulphates...................................................................................181
V.5.1.2.1 Aqueous nickel (II) sulphato complexes.........................................181
V.5.1.2.1.1 Enthalpy of formation of NiSO4(aq) ...........................................188
V.5.1.2.1.2 Heat capacity values....................................................................190
V.5.1.2.2 Solid nickel sulphates .....................................................................190
V.5.1.2.2.1 NiSO4·7H2O(cr)...........................................................................190
V.5.1.2.2.1.1
α-NiSO4·6H2O ...................................................................191
V.5.1.2.2.1.2
Other hydrated nickel sulphate solids ................................192
V.5.1.2.2.1.3
NiSO4(cr) ...........................................................................193
V.5.2 Selenium compounds and complexes........................................................195
V.5.3 Tellurium compounds and complexes .......................................................195
V.5.3.1
Nickel tellurides...................................................................................195
V.5.3.1.1 NiTe0.775(γ1).....................................................................................195
V.5.3.1.2 NiTe0.85(γ2)......................................................................................195
CONTENTS
V.6
xvii
Group 15 compounds and complexes ............................................................196
V.6.1 Nitrogen compounds and complexes.........................................................196
V.6.1.1
Solid nickel nitrates .............................................................................196
V.6.1.1.1 Ni(NO3)2·9H2O(cr) .........................................................................196
V.6.1.1.2 Ni(NO3)2·6H2O(cr) .........................................................................196
V.6.1.1.3 Ni(NO3)2·4H2O(cr), Ni(NO3)2·3H2O(cr), Ni(NO3)2·2H2O(cr) ........198
V.6.1.1.4 Ni(NO3)2(anhydrous) ......................................................................200
V.6.1.2
Aqueous Ni(II)-nitrato complexes .......................................................200
V.6.1.2.1 Determination of the Ni2+ – NO3− ion interaction coefficient........203
V.6.2 Phosphorus compounds and complexes ....................................................204
V.6.2.1
Solid nickel phosphorus compounds....................................................204
V.6.2.1.1 Nickel phosphides...........................................................................204
V.6.2.1.2 Nickel phosphates ...........................................................................204
V.6.2.2
Aqueous nickel phosphorus species.....................................................205
V.6.2.2.1 Simple nickel phosphato complexes ...............................................205
V.6.2.2.2 Aqueous nickel diphosphato complexes .........................................207
V.6.3 Arsenic compounds ...................................................................................210
V.6.3.1
Nickel arsenides...................................................................................210
V.6.3.1.1 NiAs(cr) ..........................................................................................210
V.6.3.1.2 Ni11As8(cr) ......................................................................................211
V.6.3.1.3 Ni5As2(cr) .......................................................................................212
V.6.3.1.4 NiAs2(cr).........................................................................................212
V.6.3.2
Nickel arsenates ...................................................................................212
V.6.3.2.1 Solid nickel arsenates......................................................................212
V.6.3.2.2 Aqueous nickel arsenato complexes ...............................................213
V.6.3.3
Nickel arsenites....................................................................................214
V.6.3.3.1 Solid nickel arsenites ......................................................................214
V.7 Group 14 compounds and complexes ............................................................214
V.7.1 Carbon compounds and complexes ...........................................................214
V.7.1.1
Nickel carbonates.................................................................................214
V.7.1.1.1 Solid nickel carbonates ...................................................................214
V.7.1.1.1.1 NiCO3(cr) ....................................................................................214
V.7.1.1.1.1.1
Crystallography and mineralogy of nickel carbonate.........214
V.7.1.1.1.1.2
Heat capacity and entropy of NiCO3(cr) ............................216
V.7.1.1.1.1.3
Thermodynamic analysis of solubility data on gaspéite ....217
V.7.1.1.1.2 NiCO3·5.5H2O(cr) ....................................................................................... 218
Crystallography and mineralogy of nickel carbonate hydrate .
V.7.1.1.1.2.1
........................................................................... 218
V.7.1.1.1.2.2
Heat capacity and entropy of NiCO3·5.5H2O(cr) ............ 219
V.7.1.1.1.2.3
Thermodynamic analysis of solubility data on hellyerite ..219
V.7.1.1.2 Aqueous nickel carbonato complexes.............................................223
xviii
CONTENTS
V.7.1.2
Nickel cyanides....................................................................................226
V.7.1.2.1 Aqueous Ni(II) cyano complexes ...................................................226
V.7.1.2.1.1 Complexes in neutral and alkaline solution.................................226
V.7.1.2.1.2 Protonated complexes .................................................................231
V.7.1.3
Nickel thiocyanates..............................................................................231
V.7.1.3.1 Aqueous nickel thiocyanato complexes ..........................................231
V.7.2 Silicon compounds and complexes............................................................238
V.7.2.1
Solid nickel silicates ............................................................................238
V.7.2.1.1 Nickel orthosilicate Ni2SiO4(cr) .....................................................238
V.7.2.1.1.1 Crystal structure and phase transitions ........................................238
V.7.2.1.1.2 Crystallography and mineralogy of nickel orthosilicate..............239
V.7.2.1.1.3 Heat capacity and standard entropy.............................................239
V.7.2.1.1.4 Enthalpy of formation .................................................................241
V.7.2.1.1.5 Gibbs energy of formation ..........................................................243
V.7.3 Germanium compounds and complexes ....................................................245
V.7.3.1
Solid nickel germanates .......................................................................245
V.7.3.1.1 Nickel orthogermanate Ni2GeO4(cr) ...............................................245
V.7.3.1.1.1 Crystal structure of nickel orthogermanate .................................245
V.7.3.1.1.2 Enthalpy of formation from NiO(cr) and GeO2(cr).....................245
V.8 Group 13 compounds and complexes ............................................................245
V.8.1 Boron compounds and complexes .............................................................245
V.8.1.1
Solid nickel borates..............................................................................245
V.8.1.1.1 Crystal structure..............................................................................245
V.8.1.1.2 Thermodynamic data for NiO·2B2O3(s)..........................................246
V.8.1.1.3 Solubility of Ni(BO2)2·4H2O(s) ......................................................246
V.8.1.2
Aqueous nickel borato complexes .......................................................247
VI
DISCUSSION OF AUXILIARY DATA SELECTION ...............................249
VI.1 Group 15 auxiliary species .............................................................................249
VI.1.1 Additional consistent auxiliary data for As compounds ............................249
VI.2 Other auxiliary species ...................................................................................249
VI.2.1
Auxiliary data for KCl(cr), KBr(cr) and KI(cr).........................................249
Part IV Appendices
251
A
Discussion of selected references .......................................................................253
B
Ionic strength corrections ..................................................................................443
B.1
The specific ion interaction equations ...........................................................445
B.1.1
Background ...............................................................................................445
CONTENTS
xix
B.1.2 Ionic strength corrections at temperatures other than 298.15 K ................451
B.1.3 Estimation of ion interaction coefficients..................................................453
B.1.3.1
Estimation from mean activity coefficient data ...................................453
B.1.3.2
Estimations based on experimental values of equilibrium constants at
different ionic strengths .......................................................................453
B.1.4 On the magnitude of ion interaction coefficients.......................................457
B.2 Ion interaction coefficients versus equilibrium constants for ion pairs .....458
B.3
Tables of ion interaction coefficients .............................................................458
Assigned uncertainties ......................................................................................471
C
C.1
The general problem.......................................................................................471
C.2
Uncertainty estimates in the selected thermodynamic data. .......................473
C.3
One source datum ...........................................................................................474
C.4
Two or more independent source data .........................................................475
C.4.1 Discrepancies.............................................................................................477
C.5 Several data at different ionic strengths .......................................................479
C.5.1 Discrepancies or insufficient number of data points..................................481
C.6 Procedures for data handling ........................................................................483
C.6.1
C.6.2
C.6.3
C.6.4
Correction to zero ionic strength ...............................................................483
Propagation of errors .................................................................................485
Rounding ...................................................................................................486
Significant digits........................................................................................487
Bibliography .............................................................................................................. 489
List of authors ............................................................................................................ 585
List of Figures
Figure II-1:
Standard order of arrangement of the elements and compounds based on
the periodic classification of the elements ...........................................27
Figure V-1:
The standard molar heat capacity of nickel as a function of temperature.
..............................................................................................................77
Figure V-2:
Standard electrode potential of Ni | NiSO4 | Hg2SO4 | Hg at 25°C. .....81
Figure V-3:
Estimation of the Ni3+ | Ni2+ standard electrode potential. ..................89
Figure V-4:
Extrapolation to Im = 0 of the experimental data for Reaction (V.23) in
NaClO4 media.......................................................................................96
Figure V-5:
SIT analysis of the experimental data for Reaction (V.23) in chloride
media, including log10 *b1,1ο ((V.23), 298.15 K) obtained in NaClO4....96
Figure V-6:
Extrapolation to Im = 0 of the experimental data for Reaction (V.25) in
perchlorate media. ................................................................................98
Figure V-7:
Solubility of Ni(OH)2(s) in alkaline solutions at 298.15 K. ..............100
Figure V-8:
Experimental results of heat capacity measurements on nickel oxide
from 3.2 to 1809.7 K. ........................................................................104
Figure V-9:
Comparison between the temperature dependence of the Gibbs energy
of formation for NiO obtained from electrochemical as well as chemical
reduction/oxidation equilibrium measurements and the prediction based
on the present selection of thermodynamic properties for NiO..........106
Figure V-10:
The heat capacity of Ni(OH)2(cr) as a function of temperature .........110
Figure V-11:
Solubility constant of Ni(OH)2 as a function of temperature. ............114
Figure V-12:
β-Ni(OH)2 | β-NiOOH system. Variation of reversible potential with the
oxidation state of Ni at 298.15 K and I = 0. ......................................116
Figure V-13:
α-Ni(OH)2 | γ-NiOOH system. Variation of reversible potential with the
oxidation state of Ni at 298.15 K and I = 0. ......................................118
Figure V-14:
Molar heat capacity of anhydrous NiCl2 between 2 and 30 K............123
Figure V-15:
Heat capacity of anhydrous NiCl2(cr) ................................................123
Figure V-16:
∆ r Gmο –function of Reaction (V.59)....................................................125
Figure V-17:
Third-law analysis of data for Reaction (V.59). . ...............................126
Figure V-18:
Gibbs energy of formation of NiCl2(cr) versus temperature. ............127
Figure V-19:
Values from the heat capacity studies of Stuve et al. [78STU/FER] and
White and Staveley [82WHI/STA]. ...................................................133
xxii
LIST OF FIGURES
Figure V-20:
Extrapolation to Im = 0 of the experimental data for Reaction (V.70) in
NaClO4 media. ...................................................................................143
Figure V-21:
Extrapolation to Im = 0 of the experimental enthalpies for Reaction
(V.70) in NaClO4 media. ....................................................................145
Figure V-22:
Extrapolation to I = 0 of the log10 b1 values reported in [75LIB/TIA]
for Reaction (V.71) in Ni(II) perchlorate solution. ............................149
Figure V-23:
Extrapolation to I = 0 of the accepted experimental data for Reaction
(V.71) in NaClO4 media. ....................................................................150
Figure V-24:
Extrapolation to I = 0 of the log10 b1 values reported in [78LIB/KOW]
for Reaction (V.73) in Ni(II) perchlorate media.................................155
Figure V-25:
Plot of log10 γ ± versus molality of aqueous NiBr2 solutions at 298.15 K
............................................................................................................157
Figure V-26:
Plot of osmotic coefficient versus molality of aqueous NiBr2 solutions at
298.15 K. ...........................................................................................158
Figure V-27:
Osmotic coefficient plotted as a function of molality of aqueous
Ni(ClO4)2 solutions at 298.15 K. ........................................................160
Figure V-28:
Plot of log10 γ ± versus molality of aqueous Ni(ClO4)2 solutions at
298.15 K.............................................................................................160
Figure V-29:
Phase diagram for the binary system Ni – S with T plotted versus mole
fraction of sulphur. ............................................................................164
Figure V-30:
Section of the calculated phase diagram for the composition range
0.30 < xS < 0.55. . ...............................................................................165
Figure V-31:
Gibbs energy of reaction for (V.79) plotted versus temperature ........168
Figure V-32:
Partial pressure of sulphur for various phase equilibria in the binary
system Ni – S plotted against temperature. .......................................170
Figure V-33:
Gibbs energy function for the stoichiometric composition of NiAs-type
Ni(II) sulphide, α-NiS, plotted versus temperature. ..........................173
Figure V-34:
Plot of equilibrium partial pressure of sulphur versus temperature ....176
Figure V-35:
Distribution of nickel sulphide complexes as a function of total molality
of nickel (II) in aqueous solutions saturated with H2S, T = 298.15 K, I =
1.00 mol·kg–1 NaCl. ...........................................................................179
Figure V-36:
Solubility of Ni(II) sulphides in aqueous solutions saturated with H2S as
a function of initial concentration of hydrochloric acid at 298.15 K. .181
Figure V-37:
The variation of molar conductances of aqueous nickel sulphate
solutions with concentration at 25°C..................................................186
Figure V-38:
Values of log10 K1ο for nickel sulphate association at different
temperatures .......................................................................................189
LIST OF FIGURES
xxiii
Figure V-39:
Measured water vapour pressures over hydrated Ni(NO3)2·6H2O(cr)
[37SAN], [72AUF/CAR] and Ni(NO3)2·4H2O(cr) [72AUF/CAR2]..199
Figure V-40:
Extrapolation to I = 0 of the constants for Reaction (V.96) in lithium
perchlorate media based on the data in [73FED/SHM]. ....................202
Figure V-41:
Plot of log10 γ ± versus molality of aqueous Ni(NO3)2 solutions at
298.15 K.............................................................................................203
Figure V-42:
Osmotic coefficient plotted as a function of molality of aqueous
Ni(NO3)2 solutions at 298.15 K. ........................................................204
Figure V-43:
Temperature dependence of hellyerite solubility. ..............................220
Figure V-44:
Ionic strength dependence of hellyerite solubility..............................221
Figure V-45:
Thermodynamic analysis of hellyerite solubility. ..............................222
Figure V-46:
Three-dimensional predominance diagram for the system
Ni2+–CO2–H2O. ..................................................................................223
Figure V-47:
calculated solubility of NiCO3(cr) in dilute sodium carbonate solutions
under atmospheric conditions.............................................................225
Figure V-48:
Extrapolation to I = 0 of experimental formation constants of
Ni(CN)24 − (V.113) determined in NaClO4 medium (insert shows the
corresponding plot of data reported for KClO4 medium). ..................229
Figure V-49:
Extrapolation to I = 0 of experimental equilibrium constants for
Reaction (V.114) determined in NaClO4 medium..............................229
Figure V-50:
Extrapolation to Im = 0 of the accepted formation constants reported for
the formation of NiSCN+ in perchlorate media. .................................235
Figure V-51:
Extrapolation to Im = 0 of the accepted formation constants reported for
the formation of Ni(SCN)2(aq) in perchlorate media. ........................236
Figure V-52:
Extrapolation to Im = 0 of the accepted formation constants reported for
the formation of Ni(SCN)3− in perchlorate media..............................236
Figure V-53:
The standard molar heat capacity of Ni2SiO4-olivine as a function of
temperature.........................................................................................240
Figure V-54:
Standard Gibbs energy of Reaction (V.121) as a function of
temperature.........................................................................................244
Figure A-1:
Solubility of precipitated nickel carbonate.........................................268
Figure A-2:
High temperature enthalpy of anhydrous NiCl2. ................................274
Figure A-3:
Plot of E (vs. SHE) versus log10([NiSO4]/m)......................................276
Figure A-4:
Extrapolation to Im = 0 (using the SIT) of the formation constants for the
species NiSCN+ as reported in [58YAT/KOR] . ................................290
Figure A-5:
Extrapolation of activity coefficients to saturated solution at 25°C. .295
Figure A-6:
Extrapolation of osmotic coefficients to saturated solution at 25°C...295
xxiv
LIST OF FIGURES
Figure A-7:
Plot of log10 b1 (A.17) reported in [61LIS/WIL] vs. 1/T...................298
Figure A-8:
log10 b1 – Im plot obtained from the potentiometric data (t = 35°C)
reported for the Ni2+ – SCN– system in [61MOH/DAS]. ...................299
Figure A-9:
Solubility product of Ni(BO2)2·4H2O(s) at 22°C from data reported in
[61SHC]. ............................................................................................300
Figure A-10:
Nickel and cobalt borate complexes from data in [61SHC]. ..............301
Figure A-11:
Temperature dependence of log10 *b1,1 values derived from
[63BOL/JAU].....................................................................................306
Figure A-12:
Temperature dependence of log10 *b 4,4 values derived from
[63BOL/JAU].....................................................................................307
Figure A-13:
Extrapolation to I = 0 of the equilibrium constants log10 *b1,1 (A.23)
reported in [64PER] at 20°C, using the SIT. ......................................313
Figure A-14:
Plot of ln *b1,1ο (A.23) recalculated from [64PER] vs. 1/T...................314
Figure A-15:
Temperature dependence of the association constant for the reaction
Ni(CN)24 − + CN– U Ni(CN)35− reported in [60MCC/JON]. ...........320
Figure A-16:
Molar enthalpy of solution of NiSO4·6H2O(cr) at 25°C. ...................325
Figure A-17:
Temperature dependence of log10 b1 values of NiSCN+ complex
reported in [68MAL/TUR]. ................................................................333
Figure A-18:
Heats of reaction determined from the calorimetric data of [69IZA/EAT]
by using the 298.15 K values of log10 K1 from [73KAT]. ................336
Figure A-19:
Extrapolation to I = 0 of the formation constant of NiCl+ using Equation
(A.40) and the experimental data published in [70HAL/VAN]. ........343
Figure A-20:
Temperature dependence of log10 b1 and log10 b 2 values for the
Ni(II) – (SCN)– system (I = 0.2 M) as reported in [71TUR/MAL]. ...351
Figure A-21:
Heats of dilution of hydrated nickel nitrate salts based on the
experimental heats of solution and an assumed heat of dilution to
infinite dilution of 520 J·mol–1 for a solution with a solvent:salt ratio of
8000:1.................................................................................................352
Figure A-22:
Experimental [73FED/SHM] and calculated solubility of nickel(II) in
the aqueous phase with increasing nitrate concentrations. .................358
Figure A-23:
Experimental [73FED/SHM] and calculated values of the solubility of
Ni(IO3)2·3H2O in aqueous LiClO4 solutions.......................................358
Figure A-24:
Variation of formation constants from Perlmutter-Hayman and Secco
[73PER/SEC] with ionic strength.......................................................363
Figure A-25:
Temperature dependence of log10 K 2 for the reaction Ni(CN)24 − + H+
U Ni(HCN)(CN)3− at different ionic strengths using NaCl. .............367
LIST OF FIGURES
xxv
Figure A-26:
Extrapolation to I = 0 of experimental equilibrium constants for the
reaction Ni(HCN)(CN)3− + H+ U Ni(HCN) 2 (CN) 2 (aq) at different
temperatures.. .....................................................................................367
Figure A-27:
∆ r H mο (298 K) of reaction of B2O3(cr) with MO or M2O. ..................370
Figure A-28:
Heat capacity of NiS2 plotted versus temperature. .............................382
Figure A-29:
SIT-plot of recalculated data from [80MIL/BUG], [65BUR/LIL2] and
ο
[71OHT/BIE], and the selected log10 *b 4,4
((A.70), 298.15 K) ..........391
Figure A-30:
Solubility of NiO in acidic solutions as a function of initial molality of
HCl. . ..................................................................................................394
Figure A-31:
Solubility of NiO in basic solutions as a function of initial molality of
NaOH. ...............................................................................................395
Figure A-32:
Values of C p ,m (NiBr2, cr) as a function of temperature as reported by
White and Staveley [82WHI/STA]. ...................................................400
Figure A-33:
Plot of logarithm of the oxygen partial pressure versus temperature for
the phase equilibria α−NiS / β2-Ni4S3 and β2-Ni4S3 / NiO, respectively,
at pSO = 1 atm. .................................................................................407
Figure A-34:
Logarithm of the oxygen partial pressure plotted as a function of
temperature for the equilibrium NiS(α) / NiO at p(SO2) = 1 atm.......408
Figure A-35:
Plot of logarithm of the oxygen partial pressure versus temperature for
the phase equilibria NiS(α) / β2-Ni4S3 and β2-Ni4S3 / NiO, respectively,
at pSO = 1 atm ...................................................................................419
Figure A-36:
Heat capacity data for Ni3S2 as a function of temperature..................421
Figure A-37:
Heat capacity of “Ni7S6” consisting of a mixture of Ni3S2 and Ni9S8 as a
function of temperature. .....................................................................427
Figure A-38:
Heat capacity of NiS as a function of temperature .............................429
Figure A-39:
Logarithm of stability constants of nickel bisulphide complexes in
seawater (NaCl solutions) plus the Debye-Hückel term for ionic strength
correction plotted as a function of ionic strength, T = 298.15 K.. ......432
Figure A-40:
Solubility of theophrastite as determined by the pH variation method.. ..
............................................................................................................440
Figure A-41:
Ionic strength dependence of theophrastite solubility. .......................440
Figure B-1:
Plot of log10 b1, m + 4 D versus Im for reaction (B.12), at 25°C
and 1 bar.............................................................................................455
2
2
List of Tables
Table II-1:
Abbreviations for experimental methods .................................................11
Table II-2:
Symbols and terminology. .......................................................................13
Table II-3:
Abbreviations used as subscripts of ∆ to denote the type of chemical
process. ....................................................................................................17
Table II-4:
Unit conversion factors ............................................................................28
Table II-5:
Factors Š for the conversion of molarity, cB, to molality, mB, of a substance
B, in various media at 298.15 K ..............................................................30
Table II-6:
Reference states for some elements at the reference temperature of
298.15 K and standard pressure of 0.1 MPa. ...........................................31
Table II-7:
Fundamental physical constants...............................................................36
Table III-1: Selected thermodynamic data for nickel compounds and complexes ......44
Table III-2: Selected thermodynamic data for reaction involving nickel compounds and
complexes ................................................................................................48
Table III-3: Selected temperature coefficients for heat capacities of nickel ...............52
Table IV-1: Selected thermodynamic data for auxiliary compounds and complexes..55
Table IV-2: Selected thermodynamic data for reactions involving auxiliary compounds
and complexes..........................................................................................68
Table V-1:
The temperature coefficients of the heat capacity function for Ni(cr) .....78
Table V-2:
Enthalpy of Reaction (V.3) ......................................................................83
Table V-3:
Partial molar heat capacity values for Ni(II) salts in aqueous solution at
298.15 K...................................................................................................88
Table V-4:
Experimental equilibrium data (logarithmic values) for the hydrolysis of
Ni(II) in acidic or neutral solutions. .........................................................93
Table V-5:
Smoothed C οp ,m data and calculated values for the entropy of NiO. .....103
Table V-6:
Literature data for log10 *K sο,0 (Ni(OH)2)................................................113
Table V-7:
Thermodynamic properties of Ni(OH)2(cr)............................................115
Table V-8:
Experimental stability constants (logarithmic values) of the species NiF+. .
...............................................................................................................142
Table V-9:
Experimental enthalpy values for the Reaction (V.70). .........................144
xxvii
xxviii
LIST OF TABLES
Table V-10: Experimental formation constants for the NiCl+ species. ......................147
Table V-11: Experimental equilibrium constants for the Reaction (V.72).................152
Table V-12: Reaction enthalpies reported for Reaction (V.71)..................................153
Table V-13: Experimental formation constants of the NiBr+ complex. .....................154
Table V-14: Reaction enthalpies reported for Reaction (V.73)..................................156
Table V-15: Gibbs energy functions for the binary system Ni – S relative to metallic
nickel, Ni(s), and gaseous sulphur, S2(g). ..............................................165
Table V-16: Temperatures and enthalpies of transition for β-NiS U α-NiS. ...........172
Table V-17: Temperatures for invariant three-phase equilibria in the binary system
Ni - S......................................................................................................175
Table V-18: Heat capacity functions of nickel sulphides. ..........................................176
Table V-19: Standard enthalpy of formation of nickel sulphides at 298.15 K. ..........177
Table V-20: Standard entropy of nickel sulphides at 298.15 K..................................177
Table V-21: Heat capacity of nickel sulphides at 298.15 K. ......................................177
Table V-22: Recalculated stability constants and SIT parameters for nickel sulphide
complexes at T = 298.15 K and I = 0. ....................................................178
Table V-23: Solubility constants for Ni(II) sulphide at T = 298.15 K and I = 0. .......180
Table V-24: Experimental values for the association constants and other
thermodynamic quantities for nickel sulphate. ......................................183
Table V-25: Results of reanalysis of NiSO4(aq) conductance data using a form of the
Lee-Wheaton equation that is compatible with the SIT procedure. .......185
Table V-26: Experimental values for the first association constant of nickel sulphate
determined in the presence of supporting electrolytes ..........................187
Table V-27: Formation constants of the NiNO3+ complex (ion pair)........................201
Table V-28: Experimental values for the association constants and other
thermodynamic quantities for nickel phosphate complexes...................206
Table V-29: Experimental values for the association constants and other
thermodynamic quantities for nickel pyrophosphate complexes. ..........209
Table V-30: Experimental equilibrium data for the Ni(II) cyanide system................228
Table V-31: Experimental enthalpy values for the Reaction (V.113) and (V.114). ...230
Table V-32: Experimental equilibrium data for the Ni(II) – thiocyanate system. ......233
Table V-33: Experimental reaction enthalpy values for the formation of Ni(II)thiocyanate complexes. ..........................................................................237
Table VI-1: Enthalpy of formation values for KCl(cr), KBr(cr), and KI(cr) at
298.15 K.................................................................................................250
LIST OF TABLES
xxix
Table A-1:
Equilibrium constants of Reaction (A.3) [21BER/CRU], [24CRU]. .....258
Table A-2:
Equilibrium constants of NiCl2(cr) + H2(g) U Ni(cr) + 2 HCl(g).........260
Table A-3:
Electrode potential of the cell: Ni|NiSO4(m)|Hg2SO4|Hg.......................262
Table A-4:
Standard electrode potential of Ni2+|Ni at 25°C. ....................................276
Table A-5:
Equilibrium constants of Reaction (A.9)................................................278
Table A-6:
Equilibrium constants of Reaction (A.9) [53SAN] ................................279
Table A-7:
Enthalpy of formation values calculated from equilibrium gas
compositions. .........................................................................................282
Table A-8:
Equilibrium constants of Reaction (A.10) [54SHC/TOL] .....................282
Table A-9:
Enthalpy of solution of NiCl2(cr) in H2O(l) at 298.15 K [58MUL]. ......289
Table A-10: log10 b1 versus Im. .................................................................................304
Table A-11: Recalculated constants. ..........................................................................306
Table A-12: Comparison between experimental data [63LIN/LAF] and calculated
values of log10[p(H2S)/p(H2)] for the coexistence of Ni3+xS2 with metallic
nickel......................................................................................................308
Table A-13: Reported and recalculated log10 *b1,1ο ....................................................314
Table A-14: Dataset. ..................................................................................................342
Table A-15: Experimental results for the enthalpy of formation for various nickel
sulphides determined by calorimetric measurements.............................345
Table A-16: Gibbs energy functions for various nickel sulphides. ............................385
Table A-17: Auxiliary data used in the recalculation of NaBr heat of formation values
with the data from [78STU/FER]...........................................................386
Table A-18: Corrected formation constants. ..............................................................390
Table A-19: Stability constants for the complexes NiOH+, Ni(OH)2 and Ni(OH)3− .393
ο
Table A-20: Stability constants log10 *b 3,1
for Ni(OH)3− , I = 0.................................395
Table A-21: Enthalpies and entropies of reaction from Skeaff et al.. ........................404
Table A-22: Enthalpies of formation for solid nickel sulphides at 298.15 K. ............406
Table A-23: Enthalpy of solution of NiCl2 in H2O at 298.15 K [90EFI/FUR]...........417
Table A-24: ∆ f H mο (NiCO3, 298.15 K) from decomposition data [91TAR/FAZ]......422
Table A-25: Enthalpy of solution of NiCl2 in H2O at 298.15 K [95MAN/KOR].......430
Table A-26: Recommended formation constants ( *b nο, m ) and reaction enthalpies for the
hydroxo complexes formed in the acidic region, at 298.15 K, mNi2+ +
nH2O(l) U Ni m (OH) n2 m − n + nH+ .........................................................435
Table A-27: Recommended equilibrium constants ( *K sο, n , m ) and reaction enthalpies for
the hydroxo complexes formed in the alkaline region, at 298.15 K,
mNi(OH)2(cr) + nOH– U Ni m (OH) −2 mn + n ............................................435
xxx
LIST OF TABLES
Table A-28: Recommended ion interaction coefficients ε(j, k). ................................436
Table A-29: Standard thermodynamic functions of the aqueous species. ..................436
Table B-1:
Water activities aH2 O at 2968.15 K for the most common ionic media at
various concentrations applying Pitzer’s ion interaction approach and the
interaction parameters given in [91PIT]. ...............................................448
Table B-2:
Debye–Hückel constants as a function of temperature at a pressure of 1
bar below 100°C and at the steam saturated pressure for t ≥ 100°C.....452
Table B-3:
The preparation of the experimental equilibrium constants for the
extrapolation to I = 0 with the specific ion interaction method at 25°C and
1 bar, according to reaction (B.12).........................................................454
Table B-4:
Ion interaction coefficients ε( j , k ) (kg·mol–1) for cations j with k = Cl–,
ClO −4 and NO3− , taken from Ciavatta [80CIA], [88CIA] unless indicated
otherwise................................................................................................460
Table B-5:
Ion interaction coefficients ε( j , k ) (kg·mol–1) for cations j with k = Li, Na
and K, taken from Ciavatta [80CIA], [88CIA] unless indicated otherwise.
...............................................................................................................467
Table B-6:
Ion interaction coefficients ε(1, j , k ) and ε(2, j , k ) for cations j with k =
Cl–, ClO −4 and NO3− (first part), and for anions j with k = Li+, Na+ and K+
(second part), according to the relationship ε = ε1 + ε2 log10Im. The data are
taken from Ciavatta [80CIA], [88CIA]..................................................470
Table B-7:
SIT interaction coefficient ε(j,k) kg · mol–1 for neutral species, j, with k,
electroneutral combination of ions.........................................................470
Table C-1:
Details of the calculation of equilibrium constant corrected to I = 0, using
(C.19) .....................................................................................................483
Chapter I
I Introduction
I.1 Background
The modelling of the behaviour of hazardous materials under environmental conditions
is among the most important applications of natural and technical sciences for the protection of the environment. In order to assess, for example, the safety of a waste deposit,
it is essential to be able to predict the eventual dispersion of its hazardous components
in the environment (geosphere, biosphere). For hazardous materials stored in the ground
or in geological formations, the most probable transport medium is the aqueous phase.
An important factor is therefore the quantitative prediction of the reactions that are
likely to occur between hazardous waste dissolved or suspended in ground water, and
the surrounding rock material, in order to estimate the quantities of waste that can be
transported in the aqueous phase. It is thus essential to know the relative stabilities of
the compounds and complexes that may form under the relevant conditions. This information is often provided by speciation calculations using chemical thermodynamic data.
The local conditions, such as ground water and rock composition or temperature, may
not be constant along the migration paths of hazardous materials, and fundamental
thermodynamic data are the indispensable basis for dynamic modelling of the chemical
behaviour of hazardous waste components.
In the field of radioactive waste management, the hazardous material consists
to a large extent of actinides and fission and activation products. Two isotopes of nickel
(59Ni and 63Ni) pertain to the latter category. The scientific literature on thermodynamic
data, mainly on equilibrium constants and redox potentials in aqueous solution, has
been contradictory in a number of cases, especially in actinide chemistry. A critical and
comprehensive review of the available literature is necessary in order to establish a reliable thermochemical database that fulfils the requirements for rigorous modelling of the
behaviour of the actinides, fission and activation products in the environment.
3
4
1 Introduction
The International Atomic Energy Agency (IAEA) in Vienna published special
issues with compilations of physicochemical properties of compounds and alloys of
elements important in reactor technology: Pu, Nb, Ta, Be, Th, Zr, Mo, Hf and Ti between 1966 and 1983. In 1976, IAEA also started the publication of the series “The
Chemical Thermodynamics of Actinide Elements and Compounds”, oriented towards
nuclear engineers and scientists. This international effort has resulted in the publication
of several volumes, each concerning the thermodynamic properties of a given type of
compounds for the entire actinide series. These reviews cover the literature approximately up to 1984. The latest volume in this series appeared in 1992, under Part 12: The
Actinide Aqueous Inorganic Complexes [92FUG/KHO]. Unfortunately, data of importance for radioactive waste management (for example, Part 10: The Actinide Oxides, or
data on fission and activation products) is lacking in the IAEA series.
The Radioactive Waste Management Committee (RWMC) of the OECD Nuclear Energy Agency recognised the need for an internationally acknowledged, highquality thermochemical database for application in the safety assessment of radioactive
waste disposal, and undertook the development of the NEA Thermochemical Data Base
(TDB) project [85MUL], [88WAN], [91WAN]. The RWMC assigned a high priority to
the critical review of relevant chemical thermodynamic data of compounds and complexes for this area containing the actinides uranium, neptunium, plutonium and americium, as well as the fission product technetium. The first four books in this series on the
chemical thermodynamics of uranium [92GRE/FUG], americium [95SIL/BID], technetium [99RAR/RAN] and neptunium and plutonium [2001LEM/FUG] originated from
this initiative. Simultaneously with the NEA’s TDB project, other reviews on the physical and chemical properties of actinides appeared, including the book by Cordfunke et
al. [90COR/KON2], the series edited by Freeman et al. [84FRE/LAN], [85FRE/LAN],
[85FRE/KEL], [86FRE/KEL], [87FRE/LAN], [91FRE/KEL], the two volumes edited
by Katz et al. [86KAT/SEA], and Part 12 by Fuger et al. [92FUG/KHO] within the
IAEA review series mentioned above.
In 1998, Phase II of the TDB Project (TDB-II) was started to provide for the
further needs of the radioactive waste management programs by updating the existing
database and applying the TDB review methodology to other elements (occurring in
waste as fission or activation products) and to simple organic complexes. In TDB-II the
overall objectives are set by a Management Board, integrated by representatives of 17
organisations from the field of radioactive waste management. These participating organisations, together with the NEA, provide financial support for TDB-II. The TDB-II
Management Board is assisted in technical matters by a group of experts in chemical
thermodynamics (the Executive Group). The NEA acts in this phase as Project Coordinator ensuring the implementation of the Project Guidelines and liaising with the
Review Teams. The present volume, the sixth in the series, is the second one to be published within this second phase of the TDB Project. A previous volume dealt with an
update on the chemical thermodynamics of uranium, neptunium, plutonium, americium
I.2 Focus of the review
5
and technetium [2003GUI/FAN]. Three additional volumes are in preparation as this
book goes to press dealing with the inorganic species and compounds of selenium, the
inorganic species and compounds of zirconium and with compounds and complexes
formed by the organic ligands oxalate, citrate, EDTA and iso-saccharinic acid with U,
Np, Pu, Am, Tc, Zr, Ni and Se.
I.2 Focus of the review
The first and most important step in the modelling of chemical reactions is to decide
whether they are controlled by chemical thermodynamics or kinetics, or possibly by a
combination of the two. This also applies to the modelling of more complex chemical
systems and processes, such as waste repositories of various kinds, the processes describing transport of toxic materials in ground and surface water systems, the global
geochemical cycles, etc.
As outlined in the previous section, the focus of the critical review presented in
this report is on the thermodynamic data of nickel relevant to the safety assessment of
radioactive waste repositories in the geosphere. This includes the release of waste components from the repository into the geosphere (i.e., its interaction with the waste container and the other near-field materials) and their migration through the geological
formations and the various compartments of the biosphere. As ground waters and pore
waters are the transport media for the waste components, the knowledge of the thermodynamics of the corresponding elements in waters of various compositions is of fundamental importance.
The present review therefore puts much weight on the assessment of the lowtemperature thermodynamics of nickel in aqueous solution and makes independent
analyses of the available literature in this area. The standard method used for the analysis of ionic interactions between components dissolved in water (see Appendix B) allows the general and consistent use of the selected data for modelling purposes, regardless of the type and composition of the ground water, within the ionic strength limits
given by the experimental data used for the data analyses in the present review.
The interactions between solid compounds, such as the rock materials, and the
aqueous solution and its components are as important as the interactions within the
aqueous solution, because the solid materials in the geosphere control the chemistry of
the ground water, and they also contribute to the overall solubilities of key elements.
The present review therefore also considers the chemical behaviour of solid compounds
that are likely to occur, or to be formed, under geological and environmental conditions.
It is, however, difficult to assess the relative importance of the solid phases for performance assessment purposes, particularly since their interactions with the aqueous phase
are in many cases known to be subject to quantitatively unknown kinetic constraints.
Furthermore, in some circumstances sorption of aqueous ions at mineral-water interfaces may be a more important factor in determining migration of nickel than dissolu-
6
1 Introduction
tion and precipitation phenomena.
This book contains a summary and a critical review of the thermodynamic data
on compounds and complexes containing nickel (cf. Chapter V), as reported in the
available chemical literature up to mid-2002, but a few more recent references are also
included. A comparatively large number of primary references are discussed separately
in Appendix A.
Although the focus of this review is on nickel, it is necessary to use data on a
number of other species during the evaluation process that lead to the recommended
data. These so-called auxiliary data are taken both from the publication of CODATA
Key Values [89COX/WAG] and from the evaluation of additional auxiliary data in the
uranium and other volumes of this series [92GRE/FUG], [99RAR/RAN],
[2003GUI/FAN], [2005OLI/NOL] and their use is recommended by this review. Care
has been taken that all the selected thermodynamic data at standard conditions (cf. Section II.3) and 298.15 K are internally consistent. For this purpose, special software has
been developed at the NEA Data Bank that is operational in conjunction with the NEATDB data base system, cf. Section II.6. In order to maintain consistency in the application of the values selected by this review, it is essential to use these auxiliary data when
calculating equilibrium constants involving nickel compounds and complexes.
This review does not include any compounds and complexes containing organic ligands or species in non-aqueous solvents. Nickel alloy systems have not been
reviewed and treatment of gaseous species is very limited.
I.3 Review procedure and results
The objective of the present review is the critical compilation of a database for the inorganic species of nickel. This aim is achieved by an assessment of all sources of thermodynamic data published until mid-2002. This assessment is performed in order to decide
on the most reliable values that can be recommended. Experimental measurements published in the scientific literature are the main source for the selection of recommended
data. Previous reviews are not neglected, but form a valuable source of critical information on the quality of primary publications.
When necessary, experimental source data are re-evaluated by using chemical
models that are either found to be more realistic than those used by the original author,
or are consistent with side-reactions discussed in another section of the review (for example, data on carbonate and hydroxide solubilities might need to be re-interpreted to
take into account crystal structure and particle size of the phases actually investigated).
Re-evaluation of literature values might be also necessary to correct for known systematic errors (for example, if the junction potentials are neglected in the original publication) or to make extrapolations to standard state conditions (I = 0) by using the specific
ion interaction (SIT) equations (cf. Appendix B). For convenience, these SIT equations
are referred to in some places in the text as “the SIT”. In order to ensure that consistent
I.3 Review procedure and results
7
procedures are used for the evaluation of primary data, a number of guidelines have
been developed. They have been updated and improved since 1987, and their most recent versions are available at the NEA [99WAN], [99WAN/OST], [2000OST/WAN],
[2000GRE/WAN], [2000WAN/OST]. Some of these procedures are also outlined in
this volume, cf. Chapter II, Appendix B, and Appendix C. Parts of these sections, which
were also published in earlier volumes [92GRE/FUG], [95SIL/BID], [99RAR/RAN],
[2001LEM/FUG], [2003GUI/FAN] have been revised in this review.
Once the critical review process in the NEA TDB project is completed, the
resulting manuscript is reviewed independently by qualified experts nominated by the
NEA. The independent peer review is performed according to the procedures outlined in
the TDB-6 guideline [99WAN]. The purpose of the additional peer review is to receive
an independent view of the judgements and assessments made by the primary reviewers,
to verify assumptions, results and conclusions, and to check whether the relevant literature has been exhaustively considered. The independent peer review is performed by
persons having technical expertise in the subject matter to be reviewed, to a degree at
least equivalent to that needed for the original review.
The thermodynamic data selected in the present review (see Chapters III and
IV) refer to the reference temperature of 298.15 K and to standard conditions, cf. Section II.3. For the modelling of real systems it is, in general, necessary to recalculate the
standard thermodynamic data to non-standard state conditions. For aqueous species a
procedure for the calculation of the activity factors is thus required. This review uses the
approximate specific ion interaction method (SIT) for the extrapolation of experimental
data to the standard state in the data evaluation process, and in some cases this requires
the re-evaluation of original experimental values (solubilities, emf data, etc.). For
maximum consistency, this method, as described in Appendix B, should always be used
in conjunction with the selected data presented in this review.
The thermodynamic data selected in this review are provided with uncertainties
representing the 95% confidence level. As discussed in Appendix C, there is no unique
way to assign uncertainties, and the assignments made in this review are to a large extent based on the subjective choice by the reviewers, supported by their scientific and
technical experience in the corresponding area.
The quality of thermodynamic models cannot be better than the quality of the
data on which they are based. The quality aspect includes both the numerical values of
the thermodynamic data used in the model and the “completeness” of the chemical
model used, e.g., the inclusion of all the relevant dissolved chemical species and solid
phases. For the user it is important to consider that the selected data set presented in this
review (Chapters III and IV) is certainly not “complete” with respect to all the conceivable systems and conditions; there are gaps in the information. The gaps are pointed out
in the various sections of Part III, and this information may be used as a basis for the
8
1 Introduction
assignment of research priorities. The following example for a rewarding project in experimental thermodynamics must suffice:
Precise measurements of the enthalpies of dissolution in aqueous media according to the reactions:
NiX2 (cr) U Ni2+ + 2 X–
(where X = Cl, Br, I)
and
NiCl2·6H2O (cr) U Ni2+ + 2 Cl– + 6 H2O(l)
as well as low temperature heat capacity measurements on NiCl2·6H2O(cr) would lead
to the important key values ∆ f H mο (Ni2+, 298.15K) and S mο (Ni2+, 298.15K) by thermodynamic cycles.
The results of experimental work done on formation of weak complexes are often ambiguous. In the TDB assessments, the intent is to provide an interpretation that is
consistent with the experimental data. This is not necessarily a unique interpretation, but
must be one that can be used within the SIT framework. It can be an interpretation with
or without an association constant, but it must be clear which is to be used. We have
attempted to state the assumptions at each stage so that a reader can:
• reconstruct the arguments,
• reconstruct the calculations and results,
• with the final selected values (Tables III-1 and III-2) and interaction coefficients reported in Tables B-4 and B-5, be able to regenerate the original experimental data within the uncertainties of the selected values.
In some cases, an association constant and related interaction coefficients have
been used where it might have been possible to “make do” with a different set of interaction coefficients. We make no apologies for this so long as the calculation procedure
is well defined and consistent.
For example, in several hydrolysis studies, allowance is made for complexation
of Ni2+ with chloride, using an association constant, the calculated activity of water. An
interaction coefficient of NiOH+ with Cl- is then derived as part of the same calculation
that defines a value for the first hydrolysis constant for Ni2+. Conversely, association
between Ni2+ and Cl– could have been neglected, an interaction coefficient for Ni2+ with
Cl– then would have been specified, along with a different interaction coefficient for
NiOH+ with Cl–. In theory, there should be no effect on the value of the selected hydrolysis constant, K1ο , at I = 0.
I.3 Review procedure and results
9
Our inherent assumptions about weak complexes can be stated as follows:
• A weak complex forms between Ni2+ and some singly charged anion X–,
• It is assumed that there is no complexation between Ni2+ and ClO −4 ,
• If a weak complex forms between Ni2+ and X–, the interaction coefficient
ε(Ni2+, X–) is arbitrarily set to a “reasonable value”, in some cases a value
similar or identical to the interaction coefficient ε(Ni2+, ClO −4 ),
• All other interaction between Ni2+ and X– is accounted for by the association
constant ( K ο = a(NiX+) / a(Ni2+) a(X–)) and the interaction coefficient
ε(NiX+, X–) (and, if necessary, ε(NiX+, ClO −4 ) also is specified).
The ε(Ni2+,X–) values are unlikely to be identical, and estimated uncertainties
have been increased for values obtained from experiments in which a considerable part
of the background electrolyte was replaced by the complex-forming anion. In some
cases it has been necessary to make very arbitrary assumptions in generation of association constants for weak complexes. If used consistently, these should have little or no
impact on geochemical modelling using the TDB database, or on regeneration of the
experimental values.
When the experimental data are sparse, or the experiments were not well designed, the estimated uncertainties in some association constant values are large. If we
omitted these association constants, and only generated interaction coefficients, the
propagated uncertainties would not be reduced. They would simply reappear in a different guise.
In most cases this does not matter whatsoever, and a good argument can be
made for eliminating the extraneous constants. However, modellers have found that
when using a properly assembled database, there is only one thing worse than a species
that has been included with a large uncertainty, and that is a species, for which there is
qualitative or semi-quantitative evidence, but for which the database compiler has declined to propose a value. Therefore, there is a tendency for modellers to augment any
database with extra values. In our view, it is better to supply association constants (with
their inherent large uncertainties) than to allow the introduction of inconsistent “secondary” values.
Chapter II
II Standards, Conventions, and
Contents of the TablesEquation Section 2
Equation Section 2
This chapter outlines and lists the symbols, terminology and nomenclature, the units and
conversion factors, the order of formulae, the standard conditions, and the fundamental
physical constants used in this volume. They are derived from international standards
and have been specially adjusted for the TDB publications.
II.1 Symbols, terminology and nomenclature
II.1.1 Abbreviations
Abbreviations are mainly used in tables where space is limited. Abbreviations for methods of measurement are listed in Table II-1.
Table II-1: Abbreviations for experimental methods
AIX
Anion exchange
AES
Atomic Emission Spectroscopy
CAL
Calorimetry
CHR
Chromatography
CIX
Cation exchange
COL
Colorimetry
CON
Conductivity
COR
Corrected
COU
Coulometry
CRY
Cryoscopy
DIS
Distribution between two phases
DSC
Differential Scanning Calorimetry
DTA
Differential Thermal Analysis
EDS
Energy Dispersive Spectroscopy
EM
Electromigration
(Continued on next page)
11
12
II. Standards, Conventions, and Contents of the Tables
Table II-1: (continued)
EMF
Electromotive force, not specified
EPMA
Electron Probe Micro Analysis
EXAFS
Extended X-ray Absorption Fine Structure
FTIR
Fourier Transform Infra Red
IDMS
Isotope Dilution Mass-Spectroscopy
IR
Infrared
GL
Glass electrode
ISE-X
Ion selective electrode with ion X stated
IX
Ion exchange
KIN
Rate of reaction
LIBD
Laser Induced Break Down
MVD
Molar Volume Determination
NMR
Nuclear Magnetic Resonance
PAS
Photo Acoustic Spectroscopy
POL
Polarography
POT
Potentiometry
PRX
Proton relaxation
QH
Quinhydrone electrode
RED
Emf with redox electrode
REV
Review
SEM
Scanning Electron Microscopy
SP
Spectrophotometry
SOL
Solubility
TC
Transient Conductivity
TGA
Thermo Gravimetric Analysis
TLS
Thermal Lensing Spectrophotometry
TRLFS
Time Resolved Laser Fluorescence Spectroscopy
UV
Ultraviolet
VLT
Voltammetry
XANES
X-ray Absorption Near Edge Structure
XRD
X-ray Diffraction
?
Method unknown to the reviewers
Other abbreviations may also be used in tables, such as SHE for the standard
hydrogen electrode or SCE for the saturated calomel electrode. The abbreviation NHE
has been widely used for the “normal hydrogen electrode”, which is by definition identical to the SHE. It should nevertheless be noted that NHE customarily refers to a standard state pressure of 1 atm, whereas SHE always refers to a standard state pressure of
0.1 MPa (1 bar) in this review.
II.1 Symbols, terminology and nomenclature
13
II.1.2 Symbols and terminology
The symbols for physical and chemical quantities used in the TDB review follow the
recommendations of the International Union of Pure and Applied Chemistry, IUPAC
[79WHI], [93MIL/CVI]. They are summarised in Table II-2.
Table II-2: Symbols and terminology.
Symbols and terminology
length
l
height
h
radius
r
diameter
d
volume
V
mass
m
density (mass divided by volume)
ρ
time
t
frequency
ν
wavelength
internal transmittance (transmittance of the medium itself, disregarding boundary
or container influence)
Ti
internal transmission density, (decadic absorbance): log10(1/Ti)
A
molar (decadic) absorption coefficient: A / cB l
ε
relaxation time
λ
τ
Avogadro constant
NA
relative molecular mass of a substance(a)
Mr
thermodynamic temperature, absolute temperature
T
Celsius temperature
t
(molar) gas constant
R
Boltzmann constant
k
Faraday constant
F
(molar) entropy
Sm
(molar) heat capacity at constant pressure
C p ,m
(molar) enthalpy
Hm
(molar) Gibbs energy
Gm
chemical potential of substance B
µB
pressure
p
partial pressure of substance B: xB p
pB
fugacity of substance B
fB
(Continued next page)
14
II. Standards, Conventions, and Contents of the Tables
Table II-2: (continued)
Symbols and terminology
γf,B
fugacity coefficient: fB/pB
amount of substance
(b)
n
mole fraction of substance B
xB
molarity or concentration of a solute substance B (amount of B divided by the
cB, [B]
volume of the solution) (c)
molality of a solute substance B (amount of B divided by the mass of the
mB
solvent) (d)
factor for the conversion of molarity to molality of a solution:
( ν+ + ν− )
mean ionic molality (e), m±
ν
ν
m /c
B
B
= m+ + m− −
Š
m±
activity of substance B
aB
activity coefficient, molality basis: aB / mB
γB
activity coefficient, concentration basis: aB / cB
yB
= a B = a+ + a− −
a±
( ν+ + ν− )
mean ionic activity (e), a±
(e)
mean ionic activity coefficient ,
ν
(ν +ν )
γ± + −
ν
=
ν ν
γ ++ γ −−
γ±
osmotic coefficient, molality basis
2
2
ionic strength: I m = 1 ∑ i mi zi or I c = 1 ∑ i ci zi
2
2
SIT ion interaction coefficient between substance B1 and substance B2
stoichiometric coefficient of substance B (negative for reactants, positive for
φ
I
ε(B1, B2)
νB
products)
0 = ∑ B νBB
general equation for a chemical reaction
equilibrium constant
(f)
K
charge number of an ion B (positive for cations, negative for anions)
zB
charge number of a cell reaction
n
electromotive force
E
pH= − log10 [ aH /(mol ⋅ kg )]
−1
pH
+
κ
electrolytic conductivity
superscript for standard state(g)
°
1
12
of the mass of an atom of nuclide 12C.
(a)
ratio of the average mass per formula unit of a substance to
(b)
cf. Sections 1.2 and 3.6 of the IUPAC manual [79WHI].
(c)
This quantity is called “amount-of-substance concentration” in the IUPAC manual [79WHI]. A solution with a concentration equal to 0.1 mol ⋅ dm
−3
is called a 0.1 molar solution or a 0.1 M solution.
−1
(d)
A solution having a molality equal to 0.1 mol ⋅ kg
(e)
For an electrolyte N ν X ν which dissociates into ν ± ( = ν + + ν − ) ions, in an aqueous solution with
+
−
molality m, the individual cationic molality and activity coefficient are m+ ( = ν + m ) and
is called a 0.1 molal solution or a 0.1 m solution.
γ + ( = a+ / m+ ) . A similar definition is used for the anionic symbols. Electrical neutrality requires that
ν + z+ = ν − z− .
(f)
Special notations for equilibrium constants are outlined in Section II.1.6. In some cases, K c is used to
(g)
See Section II.3.1.
indicate a constant in molar units, and K m a constant in molal units.
II.1 Symbols, terminology and nomenclature
15
II.1.3 Chemical formulae and nomenclature
This review follows the recommendations made by IUPAC [71JEN], [77FER], [90LEI]
on the nomenclature of inorganic compounds and complexes, except for the following
items:
•
The formulae of coordination compounds and complexes are not enclosed in
square brackets [71JEN] (Rule 7.21). Exceptions are made in cases where
square brackets are required to distinguish between coordinated and uncoordinated ligands.
•
The prefixes “oxy–” and “hydroxy–” are retained if used in a general way, e.g.,
“gaseous uranium oxyfluorides”. For specific formula names, however, the
IUPAC recommended citation [71JEN] (Rule 6.42) is used, e.g., “uranium(IV)
difluoride oxide” for UF2 O(cr).
An IUPAC rule that is often not followed by many authors [71JEN] (Rules
2.163 and 7.21) is recalled here: the order of arranging ligands in coordination compounds and complexes is the following: central atom first, followed by ionic ligands and
then by the neutral ligands. If there is more than one ionic or neutral ligand, the alphabetical order of the symbols of the ligating atoms determines the sequence of the ligands. For example, (UO 2 ) 2 CO3 (OH)3− is standard, (UO 2 ) 2 (OH)3 CO3− is non-standard
and is not used.
Abbreviations of names for organic ligands appear sometimes in formulae.
Following the recommendations by IUPAC, lower case letters are used, and if necessary, the ligand abbreviation is enclosed within parentheses. Hydrogen atoms that can
be replaced by the metal atom are shown in the abbreviation with an upper case “H”, for
example: H 3 edta − , Am(Hedta)(s) (where edta stands for ethylenediaminetetraacetate).
II.1.4 Phase designators
Chemical formulae may refer to different chemical species and are often required to be
specified more clearly in order to avoid ambiguities. For example, UF4 occurs as a gas,
a solid, and an aqueous complex. The distinction between the different phases is made
by phase designators that immediately follow the chemical formula and appear in parentheses. The only formulae that are not provided with a phase designator are aqueous
ions. They are the only charged species in this review since charged gases are not considered. The use of the phase designators is described below.
•
The designator (l) is used for pure liquid substances, e.g., H 2 O(l) .
•
The designator (aq) is used for undissociated, uncharged aqueous species,
e.g., U(OH) 4 (aq) , CO 2 (aq) . Since ionic gases are not considered in this review, all ions may be assumed to be aqueous and are not designed with (aq). If
16
II. Standards, Conventions, and Contents of the Tables
a chemical reaction refers to a medium other than H 2 O (e.g., D 2 O , 90% ethanol/10% H 2 O ), then (aq) is replaced by a more explicit designator, e.g., “(in
D 2 O )” or “(sln)”. In the case of (sln), the composition of the solution is described in the text.
•
The designator (sln) is used for substances in solution without specifying the
actual equilibrium composition of the substance in the solution. Note the difference in the designation of H 2 O in Eqs.(II.2) and (II.3). H 2 O(l) in Reaction
(II.2) indicates that H 2 O is present as a pure liquid, i.e., no solutes are present,
whereas Reaction (II.3) involves a HCl solution, in which the thermodynamic
properties of H 2 O(sln) may not be the same as those of the pure liquid
H 2 O(l) . In dilute solutions, however, this difference in the thermodynamic
properties of H 2 O can be neglected, and H 2 O(sln) may be regarded as pure
H 2 O(l) .
Example:
UO 2 Cl 2 (cr) + 2 HBr(sln) U UOBr2 (cr) + 2HCl(sln)
(II.1)
UO 2 Cl 2 ⋅ 3H 2 O(cr) U UO 2 Cl2 ⋅ H 2 O(cr) + 2 H 2 O(l)
(II.2)
UO3 (γ) + 2 HCl(sln) U UO 2 Cl 2 (cr) + H 2 O(sln)
(II.3)
•
The designators (cr), (am), (vit), and (s) are used for solid substances. (cr) is
used when it is known that the compound is crystalline, (am) when it is known
that it is amorphous, and (vit) for glassy substances. Otherwise, (s) is used.
•
In some cases, more than one crystalline form of the same chemical composition may exist. In such a case, the different forms are distinguished by separate
designators that describe the forms more precisely. If the crystal has a mineral
name, the designator (cr) is replaced by the first four characters of the mineral
name in parentheses, e.g., SiO 2 (quar) for quartz and SiO 2 (chal) for chalcedony. If there is no mineral name, the designator (cr) is replaced by a Greek letter preceding the formula and indicating the structural phase, e.g., α-UF5,
β-UF5.
Phase designators are also used in conjunction with thermodynamic symbols to
define the state of aggregation of a compound to which a thermodynamic quantity refers. The notation is in this case the same as outlined above. In an extended notation (cf.
[82LAF]) the reference temperature is usually given in addition to the state of aggregation of the composition of a mixture.
II.1 Symbols, terminology and nomenclature
17
Example:
∆ f Gmο (Na + , 298.15 K)
standard molar Gibbs energy of formation
of aqueous Na + at 298.15 K
S mο (UO 2 (SO 4 ) ⋅ 2.5H 2 O, cr, 298.15 K)
standard molar entropy of
UO 2 (SO 4 ) ⋅ 2.5H 2 O(cr) at 298.15 K
C pο,m (UO3 , α, 298.15 K)
standard molar heat capacity of α-UO3 at
298.15 K
∆ f H m (HF, sln, HF ⋅ 7.8H 2 O)
enthalpy of formation of HF diluted 1:7.8
with water.
II.1.5 Processes
Chemical processes are denoted by the operator ∆ , written before the symbol for a
property, as recommended by IUPAC [82LAF]. An exception to this rule is the equilibrium constant, cf. Section II.1.6. The nature of the process is denoted by annotation of
the ∆, e.g., the Gibbs energy of formation, ∆ f Gm , the enthalpy of sublimation, ∆ sub H m ,
etc. The abbreviations of chemical processes are summarised in Table II-3.
Table II-3: Abbreviations used as subscripts of ∆ to denote the type of chemical process.
Subscript of
∆
Chemical process
at
separation of a substance into its constituent gaseous atoms (atomisation)
dehyd
elimination of water of hydration (dehydration)
dil
dilution of a solution
f
formation of a compound from its constituent elements
fus
melting (fusion) of a solid
hyd
addition of water of hydration to an unhydrated compound
mix
mixing of fluids
r
chemical reaction (general)
sol
process of dissolution
sub
sublimation (evaporation) of a solid
tr
transfer from one solution or liquid phase to another
trs
transition of one solid phase to another
vap
vaporisation (evaporation) of a liquid
The most frequently used symbols for processes are ∆ f G and ∆ f H , the Gibbs
energy and the enthalpy of formation of a compound or complex from the elements in
their reference states (cf. Table II-6).
18
II. Standards, Conventions, and Contents of the Tables
II.1.6 Equilibrium constants
The IUPAC has not explicitly defined the symbols and terminology for equilibrium
constants of reactions in aqueous solution. The NEA has therefore adopted the conventions that have been used in the work Stability Constants of Metal ion Complexes by
Sillén and Martell [64SIL/MAR], [71SIL/MAR]. An outline is given in the paragraphs
below. Note that, for some simple reactions, there may be different correct ways to index an equilibrium constant. It may sometimes be preferable to indicate the number of
the reaction to which the data refer, especially in cases where several ligands are discussed that might be confused. For example, for the equilibrium:
m M + q L U M m Lq
(II.4)
both b q , m and b (II.4) would be appropriate, and b q , m (II.4) is accepted, too. Note that,
in general, K is used for the consecutive or stepwise formation constant, and β is used
for the cumulative or overall formation constant. In the following outline, charges are
only given for actual chemical species, but are omitted for species containing general
symbols (M, L).
II.1.6.1 Protonation of a ligand
H + + H r −1L U H r L
r H+ + L U Hr L
K1,r =
b1,r =
[H r L]
[H ][H r −1L]
+
[H r L]
[H + ]r [L]
(II.5)
(II.6)
This notation has been proposed and used by Sillén and Martell [64SIL/MAR], but it
has been simplified later by the same authors [71SIL/MAR] from K1,r to K r . This review retains, for the sake of consistency, cf. Eqs.(II.7) and (II.8), the older formulation
of K1,r .
For the addition of a ligand, the notation shown in Eq.(II.7) is used.
HL q −1 + L U HL q
Kq =
[HL q ]
[HL q −1 ][L]
(II.7).
Eq.(II.8) refers to the overall formation constant of the species H r L q .
r H+ + q L U Hr Lq
b q,r =
[H r L q ]
[H + ]r [L]q
(II.8).
In Eqs.(II.5), (II.6) and (II.8), the second subscript r can be omitted if r = 1, as
shown in Eq.(II.7).
II.1 Symbols, terminology and nomenclature
19
Example:
H + + PO34− U HPO 42 −
β1,1 = β1 =
2 H + + PO34− U H 2 PO −4
β1,2 =
[HPO 24− ]
[H + ][PO34− ]
[H 2 PO 4− ]
[H + ]2 [PO34− ]
II.1.6.2 Formation of metal ion complexes
ML q −1 + L U ML q
Kq =
M + q L U ML q
βq =
[ML q ]
[ML q −1 ][L]
[ML q ]
(II.9)
(II.10)
[ M ][ L]
q
For the addition of a metal ion, i.e., the formation of polynuclear complexes, the following notation is used, analogous to Eq.(II.5):
M + M m −1L U M m L
K1, m =
[ M m L]
[ M ][ M m −1L]
(II.11).
Eq.(II.12) refers to the overall formation constant of a complex M m L q .
m M + q L U M m Lq
β q,m =
[M m L q ]
[ M ] [ L]
m
q
(II.12)
The second index can be omitted if it is equal to 1, i.e., b q , m becomes β q if
m = 1. The formation constants of mixed ligand complexes are not indexed. In this case,
it is necessary to list the chemical reactions considered and to refer the constants to the
corresponding reaction numbers.
It has sometimes been customary to use negative values for the indices of the
protons to indicate complexation with hydroxide ions, OH − . This practice is not
adopted in this review. If OH − occurs as a reactant in the notation of the equilibrium, it
is treated like a normal ligand L, but in general formulae the index variable n is used
instead of q. If H 2 O occurs as a reactant to form hydroxide complexes, H 2 O is considered as a protonated ligand, HL, so that the reaction is treated as described below in
Eqs.(II.13) to (II.15) using n as the index variable. For convenience, no general form is
used for the stepwise constants for the formation of the complex MmLqHr. In many ex-
20
II. Standards, Conventions, and Contents of the Tables
periments, the formation constants of metal ion complexes are determined by adding a
ligand in its protonated form to a metal ion solution. The complex formation reactions
thus involve a deprotonation reaction of the ligand. If this is the case, the equilibrium
constant is supplied with an asterisk, as shown in Eqs.(II.13) and (II.14) for mononuclear and in Eq.(II.15) for polynuclear complexes.
ML q −1 + HL U ML q + H +
*
+
*
M + q HL U ML q + qH
 ML q   H + 
Kq = 
 ML q −1   HL 
 ML q   H + 
bq =
q
[ M ][ HL]
q
(II.14)
 M m L q   H 
b q ,m = 
m
q
+
*
m M + q HL U M m L q + qH +
UO
3 UO
+ HF(aq) U UO2 F + H
+
2+
2
*
+
+
5
+ 5 H 2 O(l) U (UO 2 )3 (OH) + 5 H
q
[ M ] [ HL]
Example:
2+
2
(II.13)
+
(II.15)
 UO 2 F+   H + 
K1 = b1 =
 UO 2+


2   HF (aq) 
*
*
b 5,3
 (UO 2 )3 (OH)5+   H + 
= 
3
 UO 2+

2 
5
Note that an asterisk is only assigned to the formation constant if the protonated ligand that is added is deprotonated during the reaction. If a protonated ligand is
added and coordinated as such to the metal ion, the asterisk is to be omitted, as shown in
Eq.(II.16).
 M(H r L) q 
bq = 
q
M + q H r L U M(H r L) q
[ M ][ H r L]
(II.16)
Example:
UO
2+
2
+ 3 H 2 PO
−
4
−
4 3
U UO 2 (H 2 PO )
 UO 2 (H 2 PO 4 )3− 
b3 =
− 3
 UO 2+


2   H 2 PO 4 
II.1.6.3 Solubility constants
Conventionally, equilibrium constants involving a solid compound are denoted as “solubility constants” rather than as formation constants of the solid. An index “s” to the
equilibrium constant indicates that the constant refers to a solubility process, as shown
in Eqs.(II.17) to (II.19).
II.1 Symbols, terminology and nomenclature
21
K s ,0 = [ M ] [ L ]
a
M a Lb (s) U a M + b L
b
(II.17).
K s ,0 is the conventional solubility product, and the subscript “0” indicates that
the equilibrium reaction involves only uncomplexed aqueous species. If the solubility
constant includes the formation of aqueous complexes, a notation analogous to that of
Eq.(II.12) is used:
m
 mb

M a Lb (s) U M m L q + 
− qL
a
 a

K s , q , m =  M m L q   L 
(
mb
−q)
a
(II.18).
Example:
K s ,1,1 = K s ,1 =  UO 2 F+   F−  .
UO 2 F2 (cr) U UO 2 F+ + F−
Similarly, an asterisk is added to the solubility constant if it simultaneously involves a protonation equilibrium:
m
 mb

 mb

− q  H+ U M m Lq + 
− q  HL
M a Lb (s) + 
a
 a

 a

(
*
K s,q,m
 M m L q   HL 
= 
mb
(
−q)
 H +  a
mb
−q)
a
(II.19)
Example:
−
U(HPO 4 ) 2 ⋅ 4H 2 O(cr) + H + U UHPO 2+
4 + H 2 PO 4 + 4 H 2 O(l)
*
−
 UHPO 2+


4   H 2 PO 4 
K s ,1,1 = * K s ,1 = 
.
+
 H 
II.1.6.4 Equilibria involving the addition of a gaseous ligand
A special notation is used for constants describing equilibria that involve the addition of
a gaseous ligand, as outlined in Eq.(II.20).
ML q −1 + L(g) U ML q
K p, q =
MLq 
ML  p
q −1 

 L


(II.20)
The subscript “p” can be combined with any other notations given above.
Example:
CO 2 (g) U CO 2 (aq)
Kp =
[CO 2 (aq)]
pCO2
22
II. Standards, Conventions, and Contents of the Tables
6−
+
3 UO 2+
2 + 6 CO 2 (g) + 6 H 2 O(l) U (UO 2 )3 (CO3 ) 6 + 12 H
*
b p,6,3
 (UO 2 )3 (CO3 )66 −   H + 
= 
3
 UO 2+
 6
2  pCO 2
12
UO 2 CO3 (cr) + CO 2 (g) + H 2 O(l) U UO 2 (CO3 ) 22 − + 2 H +
*
 UO 2 (CO3 ) 22 −   H + 
K p, s, 2 = 
pCO2
2
In cases where the subscripts become complicated, it is recommended that K or
β be used with or without subscripts, but always followed by the equation number of
the equilibrium to which it refers.
II.1.6.5 Redox equilibria
Redox reactions are usually quantified in terms of their electrode (half cell) potential, E,
which is identical to the electromotive force (emf) of a galvanic cell in which the electrode on the left is the standard hydrogen electrode, SHE1, in accordance with the “1953
Stockholm Convention” [93MIL/CVI]. Therefore, electrode potentials are given as reduction potentials relative to the standard hydrogen electrode, which acts as an electron
donor. In the standard hydrogen electrode, H 2 (g) is at unit fugacity (an ideal gas at unit
pressure, 0.1 MPa), and H + is at unit activity. The sign of the electrode potential, E, is
that of the observed sign of its polarity when coupled with the standard hydrogen electrode. The standard electrode potential, E ο , i.e., the potential of a standard galvanic cell
relative to the standard hydrogen electrode (all components in their standard state, cf.
Section II.3.1, and with no liquid junction potential) is related to the standard Gibbs
energy change ∆ r Gmο and the standard (or thermodynamic) equilibrium constant K ο as
outlined in Eq.(II.21).
Eο = −
1
RT
∆ r Gmο =
ln K ο
nF
nF
(II.21)
and the potential, E, is related to E ο by:
E = E ο − (R T / n F)∑ ν i ln ai
1
The definitions of SHE and NHE are given in Section II.1.1.
(II.22).
II.1 Symbols, terminology and nomenclature
23
For example, for the hypothetical galvanic cell:
Pt
H2(g, p = 1 bar)
HCl(aq, aH + = 1,
f H2
= 1)
Fe(ClO4)2 (aq,
aFe2+ = 1)
Fe(ClO4)3 (aq,
aFe3+ = 1)
Pt
(II.23)
where denotes a liquid junction and a phase boundary, the reaction is:
Fe3+ +
1
H 2 (g) U Fe 2+ + H +
2
(II.24)
Formally Reaction (II.24) can be represented by two half cell reactions, each
involving an equal number of electrons, (designated “ e − ”), as shown in the following
equations:
Fe3+ + e − U Fe 2+
(II.25)
1 H (g) U H + + e − .
2 2
(II.26)
The terminology is useful, although it must be emphasised that “ e − ” here does
not represent the hydrated electron.
The equilibrium constants of the two half cell reactions may be written:
K ο (II.25) =
K ο (II.26) =
aFe2+
(II.27)
aFe3+ ⋅ ae−
aH+ ⋅ ae−
f H2
=1
(II.28)
The “Stockholm Convention” implies that the constant K ο (II.26) = 1, which
also defines the numerical value of ae− to 1 for the standard hydrogen electrode. In addition, ∆ r Gmο (II.26) = 0, ∆ r H mο (II.26) = 0, ∆ r Smο (II.26) = 0 by definition, at all temperatures, and therefore ∆ r Gmο (II.25) = ∆ r Gmο (II.24).
The equilibrium constant of the more general cell reaction:
Ox +
1
n H2(g) U Red + n H+
2
(II.29)
and of the half cell reaction
Ox + n e − U Red
(II.30)
together with K ο (II.26) yield
K ο (II.29) = K ο (II.30) × K ο (II.26)n =
aRed
a
1
1
= Red
aOx (a − /(a − (SHE)=1)) n aOx a n−
e
e
e
(II.31)
24
II. Standards, Conventions, and Contents of the Tables
The activity scale is thus fixed by the above convention. Further
a 
RT ln(10)
log10  Red 
(II.32)
nF
 aOx 
Combination of Eqs.((II.21), (II.31) and (II.32)) yields the relationship between ae− and
the electrode potential vs. SHE:
E(II.29) = Eo(II.29) −
– log10 ae− =
F
E (II.29)
RT ln(10)
(II.33)
The splitting of redox reactions into two half cell reactions by introducing the
symbol “ e − ” is highly useful even though the e − notation does not in any way refer to
solvated electrons. When calculating the equilibrium composition of a chemical system,
both “ e − ”, and H + can be chosen as components and they can be treated numerically in
a similar way: equilibrium constants, mass balances, etc. may be defined for both. However, while H + represents the hydrated proton in aqueous solution, the above equations
use only the activity of “ e − ”, and never the concentration of “ e − ”. Concentration to
activity conversions (or activity coefficients) are never needed for the electron (cf. Appendix B, Example B.3).
In the literature on geochemical modelling of natural waters, it is customary to
represent the “electron activity” of an aqueous solution with the symbol “pe” or
“pε”( = − log10 ae− ) by analogy with pH ( = − log10 aH+ ), and the redox potential of an
aqueous solution relative to the standard hydrogen electrode is usually denoted by either
“Eh” or “ EH ” (see for example [82DRE], [84HOS], [86NOR/MUN],
[96STU/MOR2],). The two representations are interrelated via Eq.(II.33) yielding the
expression
pe =
F
EH .
RT ln(10)
The symbol E ′ο is used to denote the so-called “formal potential” [74PAR].
The formal (or “conditional”) potential can be regarded as a standard potential for a
particular medium in which the activity coefficients are independent (or approximately
so) of the reactant concentrations [85BAR/PAR] (the definition of E ′ο parallels that of
“concentration quotients” for equilibria). Therefore, from
RT
(II.34)
∑ νi ln ci
nF
E ′ο is the potential E for a cell when the ratio of the concentrations (not the activities)
on the right-hand side and the left-hand side of the cell reaction is equal to unity, and
E = E ′ο −
∆G
RT
(II.35)
∑ νi lnŠ γi = − r m
nF
nF
mi
where the γ i are the molality activity coefficients and Š is ( ci ) , the ratio of molality
to molarity (cf. Section II.2). The medium must be specified.
E ′ο = E ο −
II.1 Symbols, terminology and nomenclature
25
II.1.7 pH
Because of the importance that potentiometric methods have in the determination of
equilibrium constants in aqueous solutions, a short discussion on the definition of “pH”
and a simplified description of the experimental techniques used to measure pH will be
given here. For a comprehensive account see [2002BUC/RON].
The acidity of aqueous solutions is often expressed in a logarithmic scale of the
hydrogen ion activity. The definition of pH as:
pH = − log10 aH+ = − log10 (mH+ γ H+ )
can only be strictly used in the limiting range of the Debye-Hückel equation (that is, in
extremely dilute solutions). In practice the use of pH values requires extra assumptions
on the values for single ion activities. In this review values of pH are used to describe
qualitatively ranges of acidity of experimental studies, and the assumptions described in
Appendix B are used to calculate single ion activity coefficients.
The determination of pH is often performed by emf measurements of galvanic
cells involving liquid junctions [69ROS], [73BAT]. A common setup is a cell made up
of a reference half cell (e.g. Ag(s)|AgCl(s) in a solution of constant chloride concentration), a salt bridge, the test solution, and a glass electrode (which encloses a solution of
constant acidity and an internal reference half cell):
Pt(s)
Ag(s)
AgCl(s)
KCl(aq)
salt
test
bridge
solution
a
b
KCl(aq)
AgCl(s)
Ag(s)
Pt(s)
(II.36)
where stands for a glass membrane.
The emf of such a cell (assuming Nernstian behaviour of the glass electrode) is
given by:
RT
E = E *+
ln aH+ + E j
F
where E * is a constant, and E j is the liquid junction potential. The purpose of the salt
bridge is to minimise the junction potential in junction “b”, while keeping constant the
junction potential for junction “a”. Two methods are most often used to reduce and control the value of E j . An electrolyte solution of high concentration (the “salt bridge”) is a
requirement of both methods. In the first method, the salt bridge is a saturated (or nearly
saturated) solution of potassium chloride. A problem with a bridge of high potassium
concentration, is that potassium perchlorate might precipitate1 inside the liquid junction
when the test solution contains a high concentration of perchlorate ions.
1
KClO4(cr) has a solubility of ≈ 0.15 M in pure water at 25°C
26
II. Standards, Conventions, and Contents of the Tables
In the other method the salt bridge contains the same high concentration of the
same inert electrolyte as the test solution (for example, 3 M NaClO4). However, if the
concentration of the background electrolyte in the salt bridge and test solutions is reduced, the values of E j are dramatically increased. For example, if both the bridge and
the test solution have [ClO −4 ] = 0.1 M as background electrolyte, the dependence of the
liquid junction at “b” on acidity is E j ≈ – 440 × [H+] mV·dm3·mol–1 at 25°C [69ROS]
(p.110), which corresponds to a correction at pH = 2 of ≥ 0.07 pH units.
Because of the problems in eliminating the liquid junction potentials and in
defining individual ionic activity coefficients, an “operational” definition of pH is given
by IUPAC [93MIL/CVI]. This definition involves the measurement of pH differences
between the test solution and standard solutions of known pH and similar ionic strength
(in this way similar values of γ H+ and E j cancel each other when emf values are subtracted).
In order to deduce the stoichiometry and equilibrium constants of complex
formation reactions and other equilibria, it is necessary to vary the concentrations of
reactants and products over fairly large concentration ranges under conditions where the
activity coefficients of the species are either known, or constant. Only in this manner is
it possible to use the mass balance equations for the various components together with
the measurement of one or more free concentrations to obtain the information desired
[61ROS/ROS], [90BEC/NAG], [97ALL/BAN], p. 326 – 327. For equilibria involving
hydrogen ions, it is necessary to use concentration units, rather than hydrogen ion activity. For experiments in an ionic medium, where the concentration of an “inert” electrolyte is much larger than the concentration of reactants and products we can ensure that,
as a first approximation, their trace activity coefficients remain constant even for moderate variations of the corresponding total concentrations. Under these conditions of
fixed ionic strength the free proton concentration may be measured directly, thereby
defining it in terms of – log10[H+] rather than on the activity scale as pH, and the value
of – log10[H+] and pH will differ by a constant term, i.e., log10 γ H+ . Equilibrium constants deduced from measurements in such ionic media are therefore conditional constants, because they refer to the given medium, not to the standard state. In order to
compare the magnitude of equilibrium constants obtained in different ionic media it is
necessary to have a method for estimating activity coefficients of ionic species in mixed
electrolyte systems to a common standard state. Such procedures are discussed in Appendix B.
Note that the precision of the measurement of – log10[H+] and pH is virtually
the same, in very good experiments, ± 0.001. However, the accuracy is generally considerably poorer, depending in the case of glass electrodes largely on the response of the
electrode (linearity, age, pH range, etc.), and to a lesser extent on the calibration method
employed.
II.1 Symbols, terminology and nomenclature
27
II.1.8 Order of formulae
To be consistent with CODATA, the data tables are given in “Standard Order of Arrangement” [82WAG/EVA]. This scheme is presented in Figure II-1 below, and shows
the sequence of the ranks of the elements in this convention. The order follows the ranks
of the elements.
For example, for uranium, this means that, after elemental uranium and its
monoatomic ions (e.g., U 4+ ), the uranium compounds and complexes with oxygen
would be listed, then those with hydrogen, then those with oxygen and hydrogen, and so
on, with decreasing rank of the element and combinations of the elements. Within a
class, increasing coefficients of the lower rank elements go before increasing coefficients of the higher rank elements. For example, in the U–O–F class of compounds and
complexes, a typical sequence would be UOF2 (cr) , UOF4 (cr) , UOF4 (g) , UO 2 F(aq) ,
UO 2 F+ , UO 2 F2 (aq) , UO 2 F2 (cr) , UO 2 F2 (g) , UO 2 F3− , UO 2 F42 − , U 2 O3 F6 (cr) , etc.
[92GRE/FUG]. Formulae with identical stoichiometry are in alphabetical order of their
designators. An exception is made for ammonium salts, which are included in group 1
before the alkali metals.
Figure II-1: Standard order of arrangement of the elements and compounds based on
the periodic classification of the elements (from [82WAG/EVA]).
%
$
$
! " # # & ' () ' & % $
&
& , ! ' & $
* + &
28
II. Standards, Conventions, and Contents of the Tables
II.1.9 Reference codes
The references cited in the review are ordered chronologically and alphabetically by the
first two authors within each year, as described by CODATA [87GAR/PAR]. A reference code is made up of the final two digits of the year of appearance (if the publication
is not from the 20th century, the year will be put in full). The year is followed by the
first three letters of the surnames of the first two authors, separated by a slash.
If there are multiple reference codes, a “2” will be added to the second one, a
“3” to the third one, and so forth. Reference codes are always enclosed in square brackets.
II.2 Units and conversion factors
Thermodynamic data are given according to the Système International d'unités (SI
units). The unit of energy is the joule. Some basic conversion factors, also for nonthermodynamic units, are given in Table II-4.
Table II-4: Unit conversion factors
To convert from
to
multiply by
(non-SI unit symbol)
(SI unit symbol)
ångström (Å)
metre (m)
1×10–10 (exactly)
standard atmosphere (atm)
pascal (Pa)
1.01325×105 (exactly)
bar (bar)
pascal (Pa)
1×105 (exactly)
thermochemical calorie (cal)
−1
entropy unit e.u. cal ⋅ K ⋅ mol
joule (J)
−1
−1
J ⋅ K ⋅ mol
4.184 (exactly)
−1
4.184 (exactly)
Since a large part of the NEA-TDB project deals with the thermodynamics of
aqueous solutions, the units describing the amount of dissolved substance are used very
frequently. For convenience, this review uses “M” as an abbreviation of “ mol ⋅ dm −3 ”
for molarity, c, and, in Appendices B and C, “m” as an abbreviation of “ mol ⋅ kg −1 ” for
molality, m. It is often necessary to convert concentration data from molarity to molality
and vice versa. This conversion is used for the correction and extrapolation of equilibrium data to zero ionic strength by the specific ion interaction theory, which works in
molality units (cf. Appendix B). This conversion is made in the following way. Molality
is defined as mB moles of substance B dissolved in 1 kilogram of pure water. Molarity
is defined as cB moles of substance B dissolved in (ρ − cB M ) kilogram of pure water,
where ρ is the density of the solution in kg·dm–3 and MB the molar weight of the solute
in kg·mol–1.
From this it follows that:
mB =
cB
ρ − cB M
.
II.2 Units and conversion factors
29
When the ionic strength is kept high and constant by an inert electrolyte, I, the
ratio mB /cB can be approximated by:
mB
≈
cB
1
ρ − cI M I
where cI is the concentration of the inert electrolyte in mol·dm–3 and MI its molar mass
in kg·mol–1.
Baes and Mesmer [76BAE/MES], (p.439) give a table with conversion factors
(from molarity to molality) for nine electrolytes and various ionic strengths. Conversion
factors at 298.15 K for twenty one electrolytes, calculated using the density equations
reported by Söhnel and Novotný [85SOH/NOV], are reported in Table II-5.
Example:
1.00 M NaClO 4 1.05 m NaClO4
1.00 M NaCl
1.02 m NaCl
4.00 M NaClO 4 4.95 m NaClO 4
6.00 M NaNO3
7.55 m NaNO3
It should be noted that equilibrium constants need also to be converted if the
concentration scale is changed from molarity to molality or vice versa. For a general
equilibrium reaction, 0 = ∑ B ν B B , the equilibrium constants can be expressed either in
molarity or molality units, K c or K m , respectively:
log10 K c = ∑ ν B log10 cB
B
log10 K m = ∑ ν B log10 mB
B
With (mB / cB ) = Š, or (log10 mB − log10 cB ) = log10Š, the relationship between
K c and K m becomes very simple, as shown in Eq.(II.37).
log10 K m = log10 K c + ∑ ν B log10Š
(II.37)
B
∑ B ν B is the sum of the stoichiometric coefficients of the reaction, cf. Eq.
(II.53) and the values of Š are the factors for the conversion of molarity to molality as
tabulated in Table II-5 for several electrolyte media at 298.15 K. In the case of very
dilute solutions, these factors are approximately equal to the reciprocal of the density of
the pure solvent. Then, if the solvent is water, molarity and molality may be used interchangeably, and K c ≈ K m . The differences between the values in Table II-5 and the
values listed in the uranium NEA-TDB review [92GRE/FUG] (p.23) are found at the
highest concentrations, and are no larger than ± 0.003 dm3·kg–1, reflecting the accuracy
expected in this type of conversion. The uncertainty introduced by the use of Eq.(II.37)
in the values of log10 K m will be no larger than ± 0.001 ∑ B ν B .
30
II. Standards, Conventions and Contents of the Tables
Table II-5: Factors Š for the conversion of molarity, cB, to molality, mB, of a substance
B, in various media at 298.15 K (calculated from densities in [85SOH/NOV])
Š = mB / cB (dm3 of solution per kg of H2O)
c (M)
HClO4
NaClO4
LiClO4
NH4ClO4
Ba(ClO4)2
HCl
NaCl
LiCl
0.10
1.0077
1.0075
1.0074
1.0091
1.0108
1.0048
1.0046
1.0049
0.25
1.0147
1.0145
1.0141
1.0186
1.0231
1.0076
1.0072
1.0078
0.50
1.0266
1.0265
1.0256
1.0351
1.0450
1.0123
1.0118
1.0127
0.75
1.0386
1.0388
1.0374
1.0523
1.0685
1.0172
1.0165
1.0177
1.00
1.0508
1.0515
1.0496
1.0703
1.0936
1.0222
1.0215
1.0228
1.50
1.0759
1.0780
1.0750
1.1086
1.1491
1.0324
1.0319
1.0333
2.00
1.1019
1.1062
1.1019
1.2125
1.0430
1.0429
1.0441
3.00
1.1571
1.1678
1.1605
1.3689
1.0654
1.0668
1.0666
4.00
1.2171
1.2374
1.2264
1.0893
1.0930
1.0904
5.00
1.2826
1.3167
1.1147
1.1218
1.1156
6.00
1.3547
1.4077
1.1418
c (M)
KCl
NH4Cl
MgCl2
CaCl2
NaBr
HNO3
NaNO3
LiNO3
0.10
1.0057
1.0066
1.0049
1.0044
1.0054
1.0056
1.0058
1.0059
0.25
1.0099
1.0123
1.0080
1.0069
1.0090
1.0097
1.0102
1.0103
0.50
1.0172
1.0219
1.0135
1.0119
1.0154
1.0169
1.0177
1.0178
0.75
1.0248
1.0318
1.0195
1.0176
1.0220
1.0242
1.0256
1.0256
1.00
1.0326
1.0420
1.0258
1.0239
1.0287
1.0319
1.0338
1.0335
1.50
1.0489
1.0632
1.0393
1.0382
1.0428
1.0478
1.0510
1.0497
2.00
1.0662
1.0855
1.0540
1.0546
1.0576
1.0647
1.0692
1.0667
3.00
1.1037
1.1339
1.0867
1.0934
1.0893
1.1012
1.1090
1.1028
4.00
1.1453
1.1877
1.1241
1.1406
1.1240
1.1417
1.1534
1.1420
1.1974
1.1619
1.1865
1.2030
1.1846
1.2033
1.2361
1.2585
1.2309
5.00
1.2477
6.00
1.1423
c (M)
NH4NO3
H2SO4
Na2SO4
(NH4)2SO4
H3PO4
Na2CO3
K2CO3
NaSCN
0.10
1.0077
1.0064
1.0044
1.0082
1.0074
1.0027
1.0042
1.0069
0.25
1.0151
1.0116
1.0071
1.0166
1.0143
1.0030
1.0068
1.0130
0.50
1.0276
1.0209
1.0127
1.0319
1.0261
1.0043
1.0121
1.0234
0.75
1.0405
1.0305
1.0194
1.0486
1.0383
1.0065
1.0185
1.0342
1.00
1.0539
1.0406
1.0268
1.0665
1.0509
1.0094
1.0259
1.0453
1.50
1.0818
1.0619
1.0441
1.1062
1.0773
1.0170
1.0430
1.0686
2.00
1.1116
1.0848
1.1514
1.1055
1.0268
1.0632
1.0934
3.00
1.1769
1.1355
1.2610
1.1675
1.1130
1.1474
4.00
1.2512
1.1935
1.4037
1.2383
1.1764
1.2083
5.00
1.3365
1.2600
1.3194
1.2560
1.2773
6.00
1.4351
1.3365
1.4131
1.3557
II.3 Standard and reference conditions
31
II.3 Standard and reference conditions
II.3.1 Standard state
A precise definition of the term “standard state” has been given by IUPAC [82LAF].
The fact that only changes in thermodynamic parameters, but not their absolute values,
can be determined experimentally, makes it important to have a well-defined standard
state that forms a base line to which the effect of variations can be referred. The IUPAC
[82LAF] definition of the standard state has been adopted in the NEA-TDB project. The
standard state pressure, p ο = 0.1 MPa (1 bar), has therefore also been adopted, cf. Section II.3.2. The application of the standard state principle to pure substances and mixtures is summarised below. It should be noted that the standard state is always linked to
a reference temperature, cf. Section II.3.3.
• The standard state for a gaseous substance, whether pure or in a gaseous mixture, is
the pure substance at the standard state pressure and in a (hypothetical) state in
which it exhibits ideal gas behaviour.
• The standard state for a pure liquid substance is (ordinarily) the pure liquid at the
standard state pressure.
• The standard state for a pure solid substance is (ordinarily) the pure solid at the
standard state pressure.
• The standard state for a solute B in a solution is a hypothetical liquid solution, at
the standard state pressure, in which mB = mο = 1 mol·kg–1, and in which the activity coefficient γ B is unity.
It should be emphasised that the use of superscript, ο , e.g., in ∆ f H mο , implies
that the compound in question is in the standard state and that the elements are in their
reference states. The reference states of the elements at the reference temperature (cf.
Section II.3.3) are listed in Table II-6.
Table II-6: Reference states for some elements at the reference temperature of 298.15 K
and standard pressure of 0.1 MPa [82WAG/EVA], [89COX/WAG], [91DIN].
O2
gaseous
Al
crystalline, cubic
H2
gaseous
Zn
crystalline, hexagonal
He
gaseous
Cd
crystalline, hexagonal
Ne
gaseous
Hg
liquid
Ar
gaseous
Cu
crystalline, cubic
Kr
gaseous
Ag
crystalline, cubic
Xe
gaseous
Fe
crystalline, cubic, bcc
F2
gaseous
Tc
crystalline, hexagonal
Cl2
gaseous
V
crystalline, cubic
(Continued on next page)
32
II. Standards, Conventions and Contents of the Tables
Table II-6: (continued)
Br2
liquid
Ti
crystalline, hexagonal
I2
crystalline, orthorhombic
Am
crystalline, dhcp
S
crystalline, orthorhombic
Pu
crystalline, monoclinic
Se
crystalline, hexagonal (“black”)
Np
crystalline, orthorhombic
Te
crystalline, hexagonal
U
crystalline, orthorhombic
N2
gaseous
Th
crystalline, cubic
P
crystalline, cubic (“white”)
Be
crystalline, hexagonal
As
crystalline, rhombohedral (“grey”)
Mg
crystalline, hexagonal
Sb
crystalline, rhombohedral
Ca
crystalline, cubic, fcc
Bi
crystalline, rhombohedral
Sr
crystalline, cubic, fcc
C
crystalline, hexagonal (graphite)
Ba
crystalline, cubic
Si
crystalline, cubic
Li
crystalline, cubic
Ge
crystalline, cubic
Na
crystalline, cubic
Sn
crystalline, tetragonal (“white”)
K
crystalline, cubic
Pb
crystalline, cubic
Rb
crystalline, cubic
B
β, crystalline, rhombohedral
Cs
crystalline, cubic
II.3.2 Standard state pressure
The standard state pressure chosen for all selected data is 0.1 MPa (1 bar) as recommended by the International Union of Pure and Applied Chemistry IUPAC [82LAF].
However, the majority of the thermodynamic data published in the scientific
literature and used for the evaluations in this review, refer to the old standard state pressure of 1 “standard atmosphere” (= 0.101325 MPa). The difference between the thermodynamic data for the two standard state pressures is not large and lies in most cases
within the uncertainty limits. It is nevertheless essential to make the corrections for the
change in the standard state pressure in order to avoid inconsistencies and propagation
of errors. In practice the parameters affected by the change between these two standard
state pressures are the Gibbs energy and entropy changes of all processes that involve
gaseous species. Consequently, changes occur also in the Gibbs energies of formation
of species that consist of elements whose reference state is gaseous (H, O, F, Cl, N, and
the noble gases). No other thermodynamic quantities are affected significantly. A large
part of the following discussion has been taken from the NBS tables of chemical thermodynamic properties [82WAG/EVA], see also Freeman [84FRE].
The following expressions define the effect of pressure on the properties of all
substances:
 ∂H 
 ∂V 

 = V −T 
 = V (1 − αT )
 ∂T  p
 ∂p T
(II.38)
II.3 Standard and reference conditions
33
 ∂C p 
 ∂ 2V 

 = −T  2 
 ∂T  p
 ∂p T
(II.39)
 ∂S 
 ∂V 
  = −V α = − 

∂
p
 ∂T  p
 T
(II.40)
 ∂G 

 = V
 ∂p T
(II.41)
where α ≡
1  ∂V 


V  ∂T  p
(II.42)
For ideal gases, V = RT / p and α = R / pV = 1 /T. The conversion equations
listed below (Eqs. (II.43) to (II.50)) apply to the small pressure change from 1 atm to 1
bar (0.1 MPa). The quantities that refer to the old standard state pressure of 1 atm are
assigned the superscript ( atm ) , while those that refer to the new standard state pressure of
1 bar carry the superscript (bar ) .
For all substances the changes in the enthalpy of formation and heat capacity
are much smaller than the experimental accuracy and can be disregarded. This is exactly
true for ideal gases.
∆ f H (bar) (T ) − ∆ f H (atm) (T ) = 0
C
(bar )
p
(T ) − C
(atm)
p
(T ) = 0
(II.43)
(II.44)
For gaseous substances, the entropy difference is:
 p (atm) 
S (bar) (T ) − S (atm) (T ) = R ln  (bar)  = R ln 1.01325
p

= 0.1094 J ⋅ K −1 ⋅ mol−1
(II.45)
This is exactly true for ideal gases, as follows from Eq.(II.40) with α =
R / pV.The entropy change of a reaction or process is thus dependent on the number of
moles of gases involved:
 p (atm) 
∆ r S (bar) − ∆ r S (atm) = δ ⋅ R ln  (bar ) 
 p

= δ × 0.1094 J ⋅ K −1 ⋅ mol −1
where δ is the net increase in moles of gas in the process.
(II.46)
34
II. Standards, Conventions and Contents of the Tables
Similarly, the change in the Gibbs energy of a process between the two standard state pressures is:
 p (atm) 
∆ r G (bar) − ∆ r G (atm) = − δ ⋅ RT ln  (bar ) 
 p

= − δ ⋅ 0.03263 kJ ⋅ mol −1 at 298.15 K.
(II.47)
Eq.(II.47) applies also to ∆ f G (bar) − ∆ f G (atm) , since the Gibbs energy of formation describes the formation process of a compound or complex from the reference
states of the elements involved:
.
∆ f G (bar) − ∆ f G (atm) = − δ × 0.03263 kJ ⋅ mol −1 at 298.15 K.
(II.48).
The changes in the equilibrium constants and cell potentials with the change in the
standard state pressure follow from the expression for Gibbs energy changes, Eq.(II.47)
∆ r G (bar) − ∆ r G (atm)
RT ln 10
 p (atm) 
ln  (bar) 
(atm)
p

 = δ ⋅ log  p
= δ⋅ 
10 
(bar) 
ln10
p


log10 K (bar) − log10 K (atm) = −
= δ ⋅ 0.005717
(II.49)
∆ r G (bar) − ∆ r G (atm)
nF
 p (atm) 
RT ln  (bar) 
 p

= δ⋅
nF
E (bar) − E (atm) = −
0.0003382
V at 298.15K
(II.50).
n
It should be noted that the standard potential of the hydrogen electrode is equal
to 0.00 V exactly, by definition.
H + + e − U 1 H 2 (g)
E ο def
(II.51).
= 0.00V
2
This definition will not be changed, although a gaseous substance, H 2 (g) , is
involved in the process. The change in the potential with pressure for an electrode potential conventionally written as:
= δ⋅
Ag + + e − U Ag (cr)
II.4 Fundamental physical constants
35
should thus be calculated from the balanced reaction that includes the hydrogen electrode,
1
Ag + + H 2 (g) U Ag (cr) + H +
2
Here δ = – 0.5 Hence, the contribution to δ from an electron in a half cell reaction is the
same as the contribution of a gas molecule with the stoichiometric coefficient of 0.5.
This leads to the same value of δ as the combination with the hydrogen half cell.
Example:
+
Fe(cr) + 2 H U Fe
2+
δ=1
+ H 2 (g)
δ = −1
CO 2 (g) U CO 2 (aq)
NH 3 (g) +
1
2
5
4
O 2 U NO(g) +
Cl 2 (g) + 2 O 2 (g) + e
−
3
2
H 2 O(g) δ = 0.25
−
U ClO 4
δ = −3
E
(bar)
− E
log10 K
∆rG
(bar)
∆f G
(bar)
(bar)
(atm)
= 0.00017 V
− log10 K
− ∆rG
(atm)
− ∆f G
(atm)
(bar)
= − 0.0057
= − 0.008 kJ ⋅ mol
= 0.098 kJ ⋅ mol
−1
−1
II.3.3 Reference temperature
The definitions of standard states given in Section II.3 make no reference to fixed temperature. Hence, it is theoretically possible to have an infinite number of standard states
of a substance as the temperature varies. It is, however, convenient to complete the
definition of the standard state in a particular context by choosing a reference temperature. As recommended by IUPAC [82LAF], the reference temperature chosen in the
NEA-TDB project is T = 298.15 K or t = 25.00°C. Where necessary for the discussion,
values of experimentally measured temperatures are reported after conversion to the
IPTS-68 [69COM]. The relation between the absolute temperature T (K, kelvin) and the
Celsius temperature t (°C) is defined by t = (T − To ) where To = 273.15 K.
II.4 Fundamental physical constants
To ensure the consistency with other NEA TDB Reviews, the fundamental physical
constants are taken from a publication by CODATA [86COD]. Those relevant to this
review are listed in Table II-7. Note that updated values of the fundamental constants
can be obtained from CODATA, notably through its Internet site. In most cases, recalculation of the NEA TDB database entries with the updated values of the fundamental
constants will not introduce significant (with respect to their quoted uncertainties) excursions from the current NEA TDB selections.
36
II. Standards, Conventions and Contents of the Tables
Table II-7: Fundamental physical constants. These values have been taken from
CODATA [86COD]. The digits in parentheses are the one-standard-deviation uncertainty in the last digits of the given value.
Quantity
Value
Units
speed of light in vacuum
Symbol
c
299 792 458
m·s–1
permeability of vacuum
µ0
4π×10–7 = 12.566 370 614…
10–7 N·A–2
єο
1/µ0 c = 8.854 187 817…
10–12 C2·J–1·m–1
permittivity of vacuum
2
Planck constant
h
6.626 0755(40)
10–34 J·s
elementary charge
e
1.602 177 33(49)
10–19C
Avogadro constant
NA
6.022 1367(36)
1023 mol–1
Faraday constant
F
96 485.309(29)
C·mol–1
molar gas constant
R
8.314 510(70)
J·K–1·mol–1
Boltzmann constant, R/NA
k
1.380 658(12)
10–23 J·K–1
1.602 177 33(49)
10–19 J
1.660 5402(10)
10–27 kg
Non-SI units used with SI:
electron volt, (e/C) J
atomic mass unit,
1u = mu =
eV
u
1 m (12 C )
12
II.5 Uncertainty estimates
One of the principal objectives of the NEA-TDB development effort is to provide an
idea of the uncertainties associated with the data selected in the reviews. In general the
uncertainties should define the range within which the corresponding data can be reproduced with a probability of 95%. In many cases, a full statistical treatment is limited or
impossible due to the availability of only one or few data points. Appendix C describes
in detail the procedures used for the assignment and treatment of uncertainties, as well
as the propagation of errors and the standard rules for rounding.
II.6 The NEA-TDB system
A data base system has been developed at the NEA Data Bank that allows the storage of
thermodynamic parameters for individual species as well as for reactions. The structure
of the data base system allows consistent derivation of thermodynamic data for individual species from reaction data at standard conditions, as well as internal recalculations
of data at standard conditions. If a selected value is changed, all the dependent values
will be recalculated consistently. The maintenance of consistency of all the selected
data, including their uncertainties (cf. Appendix C), is ensured by the software developed for this purpose at the NEA Data Bank. The literature sources of the data are also
stored in the data base.
The following thermodynamic parameters, valid at the reference temperature
of 298.15 K and at the standard pressure of 1 bar, are stored in the data base:
II.6 The NEA-TDB system
37
∆ f Gmο
the standard molar Gibbs energy of formation from the elements in
their reference states (kJ·mol–1)
∆ f H mο
the standard molar enthalpy of formation from the elements in their
reference states (kJ·mol–1)
S mο
the standard molar entropy (J·K–1·mol–1)
C pο,m
the standard molar heat capacity (J·K–1·mol–1).
For aqueous neutral species and ions, the values of ∆ f Gmο , ∆ f H mο , S mο and
C
correspond to the standard partial molar quantities, and for individual aqueous
ions they are relative quantities, defined with respect to the aqueous hydrogen ion, according to the convention [89COX/WAG] that ∆ f H mο ( H + , T) = 0 and that S mο ( H + , T)
= 0. Furthermore, for an ionised solute B containing any number of different cations and
anions:
ο
p ,m
∆ f H mο (B± , aq) = ∑ ν + ∆ f H mο (cation, aq) + ∑ ν − ∆ f H mο (anion, aq)
−
+
ο
m
ο
m
ο
m
S (B± , aq) = ∑ ν + S (cation, aq) + ∑ ν − S (anion, aq)
+
−
As the thermodynamic parameters vary as a function of temperature, provision
is made for including the compilation of the coefficients of empirical temperature functions for these data, as well as the temperature ranges over which they are valid. In
many cases the thermodynamic data measured or calculated at several temperatures
were published for a particular species, rather than the deduced temperature functions.
In these cases, a non-linear regression method is used in this review to obtain the most
significant coefficients of the following empirical function for a thermodynamic parameter, X:
X (T ) = a X + bX ⋅ T + c X ⋅ T 2 + d X ⋅ T −1 + eX ⋅ T −2 + f X ⋅ ln T + g X ⋅ T ln T
+ hX ⋅ T +
iX
+ j X ⋅ T 3 + k X ⋅ T −3 .
(II.52)
T
Most temperature variations can be described with three or four parameters, a,
b and e being the ones most frequently used. In the present review, only C pο,m (T ) , i.e.,
the thermal functions of the heat capacities of individual species, are considered and
stored in the data base. They refer to the relation:
C pο,m (T ) = a + b × T + c × T 2 + d × T −1 + e × T −2 + i × T −1/ 2
(where the subindices for the coefficients have been dropped) and are listed in Table
III–3.
The pressure dependence of thermodynamic data has not been the subject of
critical analysis in the present compilation. The reader interested in higher temperatures
and pressures, or the pressure dependency of thermodynamic functions for geochemical
38
II. Standards, Conventions and Contents of the Tables
applications, is referred to the specialised literature in this area, e.g., [82HAM],
[84MAR/MES], [88SHO/HEL], [88TAN/HEL], [89SHO/HEL], [89SHO/HEL2],
[90MON], [91AND/CAS].
Selected standard thermodynamic data referring to chemical reactions are also
compiled in the data base. A chemical reaction “r”, involving reactants and products
‘B”, can be abbreviated as:
0 = ∑ ν rB B
(II.53)
B
where the stoichiometric coefficients ν are positive for products, and negative for reactants. The reaction parameters considered in the NEA-TDB system include:
r
B
log10 K rο
the equilibrium constant of the reaction, logarithmic
∆ r Gmο
the molar Gibbs energy of reaction (kJ·mol–1)
∆ r H mο
the molar enthalpy of reaction (kJ·mol–1)
∆ r Smο
the molar entropy of reaction (J·K–1·mol–1)
∆ r C pο,m
the molar heat capacity of reaction (J·K–1·mol–1)
The temperature functions of these data, if available, are stored according to Eq.(II.52).
The equilibrium constant, K rο , is related to ∆ r Gmο according to the following
relation:
log10 K rο = −
∆ r Gmο
RT ln(10)
and can be calculated from the individual values of ∆ f Gmο (B) (for example, those given
in Table III–1 and Table IV–1), according to:
log10 K rο = −
1
∑ ν Br ∆f Gmο (B)
RT ln(10) B
(II.54).
II.7 Presentation of the selected data
The selected data are presented in Chapters III and IV. Unless otherwise indicated, they
refer to standard conditions (cf. Section II.3) and 298.15K (25°C) and are provided with
an uncertainty which should correspond to the 95% confidence level (see Appendix C).
Chapters III contains a table of selected thermodynamic data for individual
compounds and complexes of nickel (Table III–1), a table of selected reaction data
(Table III–2) for reactions concerning nickel species and a table containing selected
thermal functions of the heat capacities of individual species of nickel (Table III–3).
The selection of these data is discussed in Chapter V.
II.7 Presentation of the selected data
39
Chapter IV contains, for auxiliary compounds and complexes that do not contain nickel, a table of the thermodynamic data for individual species (Table IV–1) and a
table of reaction data (Table IV–2). Most of these values are the CODATA Key Values
[89COX/WAG]. The selection of the remaining auxiliary data is discussed in
[92GRE/FUG], [99RAR/RAN] and [2001LEM/FUG].
All the selected data presented in Table III–1, Table III–2, Table IV–1 and
Table IV–2 are internally consistent. This consistency is maintained by the internal consistency verification and recalculation software developed at the NEA Data Bank in
conjunction with the NEA-TDB data base system, cf. Section II.6. Therefore, when using the selected data for nickel species, the auxiliary data of Chapter IV must be used
together with the data in Chapter III to ensure internal consistency of the data set.
It is important to note that Table III–2 and Table IV–2 include only those species for which the primary selected data are reaction data. The formation data derived
there from and listed in Table III–1 are obtained using auxiliary data, and their uncertainties are propagated accordingly. In order to maintain the uncertainties originally
assigned to the selected data in this review, the user is advised to make direct use of the
reaction data presented in Table III–2 and Table IV–2, rather than taking the derived
values in Table III–1 and Table IV–1 to calculate the reaction data with Eq.(II.54). The
later approach would imply a twofold propagation of the uncertainties and result in reaction data whose uncertainties would be considerably larger than those originally assigned.
The thermodynamic data in the selected set refer to a temperature of 298.15 K
(25.00°C), but they can be recalculated to other temperatures if the corresponding data
(enthalpies, entropies, heat capacities) are available [97PUI/RAR]. For example, the
temperature dependence of the standard reaction Gibbs energy as a function of the standard reaction entropy at the reference temperature ( T0 = 298.15 K), and of the heat capacity function is:
∆ r Gmο (T ) = ∆ r H mο (T0 ) +
ο


T ∆ r C p ,m (T)
ο
ο
C
(T)
d
T
T
S
(
T
)
dT  ,
∆
−
∆
+
 r m 0 ∫T
∫T0 r p,m

0
T


T
and the temperature dependence of the standard equilibrium constant as a function of
the standard reaction enthalpy and heat capacity is:
log10 K ο (T ) = log10 K ο (T0 ) −
−
1
RT ln(10)
∫
T
T0
∆ r H mο (T0 )  1 1 
 − 
R ln(10)  T T0 
∆ r C pο,m (T ) dT +
1
R ln(10)
T
∆ r C pο,m (T )
T0
T
∫
dT ,
40
II. Standards, Conventions and Contents of the Tables
where R is the gas constant (cf. Table II-7).
In the case of aqueous species, for which enthalpies of reaction are selected or
can be calculated from the selected enthalpies of formation, but for which there are no
selected heat capacities, it is in most cases possible to recalculate equilibrium constants
to temperatures up to 100 to 150°C, with an additional uncertainty of perhaps about 1 to
2 logarithmic units, due to neglecting the heat capacity contributions to the temperature
correction. However, it is important to observe that “new” aqueous species, i.e., species
not present in significant amounts at 25°C and therefore not detected, may be significant
at higher temperatures, see for example the work by Ciavatta et al. [87CIA/IUL]. Additional high-temperature experiments may therefore be needed in order to ascertain that
proper chemical models are used in the modelling of hydrothermal systems. For many
species, experimental thermodynamic data are not available to allow a selection of parameters describing the temperature dependence of equilibrium constants and Gibbs
energies of formation. The user may find information on various procedures to estimate
the temperature dependence of these thermodynamic parameters in [97PUI/RAR]. The
thermodynamic data in the selected set refer to infinite dilution for soluble species. Extrapolation of an equilibrium constant K, usually measured at high ionic strength, to K ο
at I = 0 using activity coefficients γ, is explained in Appendix B. The corresponding
Gibbs energy of dilution is:
∆ dil Gm = ∆ r Gmο − ∆ r Gm
(II.55)
= − R T ∆ r ln γ ±
(II.56)
Similarly ∆ dil Sm can be calculated from ln γ ± and its variations with T, while:
(II.57)
∆ dil H m = R T 2 ∂ (∆ r ln γ ± ) p
∂T
depends only on the variation of γ with T, which is neglected in this review, when no
data on the temperature dependence of γ’s are available. In this case the Gibbs energy of
dilution ∆ dil Gm is entirely assigned to the entropy difference. This entropy of reaction is
calculated using the ∆ r Gmο = ∆ r H mο − T ∆ r Smο equation, the above assumption ∆ dil H m =
0, and ∆ dil Gm .
Part II
Tables of selected data
41
Chapter III
III Selected nickel data
This chapter presents the chemical thermodynamic data set for nickel species that has
been selected in this review. Table III–1 contains the recommended thermodynamic
data of the nickel compounds and species, Table III–2 the recommended thermodynamic data of chemical equilibrium reactions by which the nickel compounds and complexes are formed, and Table III–3 the temperature coefficients of the heat capacity data
of Table III–1 where available.
The species and reactions in the tables appear in standard order of arrangement. Table III–2 contains information only on those reactions for which primary data
selections are made in Chapter V of this review. These selected reaction data are used,
together with data for key nickel species and auxiliary data selected in this review, to
derive the corresponding formation data in Table III–1. The uncertainties associated
with values for key nickel species and the auxiliary data are in some cases substantial,
leading to comparatively large uncertainties in the formation quantities derived in this
manner.
The values of ∆ r Gmο for many reactions are known more accurately than would
be calculated directly from the uncertainties of the ∆ f Gmο values in Table III–1 and auxiliary data. The inclusion of a table for reaction data (Table III–2) in this report allows
the use of equilibrium constants with total uncertainties that are based directly on the
experimental accuracies. This is the main reason for including both Table III–1 and
Table III–2.
The selected thermal functions of the heat capacities, listed in Table III–3, refer
to the relation
C pο,m (T) = a + b × T + c × T 2 + d × T –1 + i × T -1/2
A detailed discussion of the selection procedure is presented in Chapter V. It
may be noted that this chapter contains data on more species or compounds than are
present in the tables of Chapter III. The main reasons for this situation are the lack of
information for a proper extrapolation of the primary data to standard conditions in
some systems and lack of solid primary data in others.
A warning: The addition of any aqueous species and their data to this internally
consistent data base can result in a modified data set, which is no longer rigorous and
can lead to erroneous results. The situation is similar when gases or solids are added.
43
III. Selected nickel data
44
Table III–1: Selected thermodynamic data for nickel compounds and complexes. All
ionic species listed in this table are aqueous species. Unless noted otherwise, all data
refer to the reference temperature of 298.15 K and to the standard state, i.e., a pressure
of 0.1 MPa and, for aqueous species, infinite dilution (I = 0). The uncertainties listed
below each value represent total uncertainties and correspond in principle to the statistically defined 95% confidence interval. Values obtained from internal calculation, cf.
footnotes (a) and (b), are rounded at the third digit after the decimal point and may
therefore not be exactly identical to those given in Part V. Systematically, all the values
are presented with three digits after the decimal point, regardless of the significance of
these digits. The data presented in this table are available on computer media from the
OECD Nuclear Energy Agency.
C οp ,m
∆ f Gmο
∆ f H mο
S mο
(kJ · mol–1)
0.000
±0.000
(kJ · mol–1)
0.000
±0.000
(J · K–1 · mol–1)
29.870
±0.200
(J · K–1 · mol–1)
26.070
±0.100
23.360
±0.100
Compound
Ni(cr)
Ni(g)
384.686
±8.400
(a)
430.100
±8.400
182.190
±0.080
Ni(l)
14.035
±0.272
(a)
17.500
±0.250
41.490
±0.300
Ni2+
– 45.773
±0.771
(b)
– 55.012
±0.878
NiO(cr)
– 211.660
±0.422
(a)
– 239.700
±0.400
NiOH+
– 228.458
±1.111
(b)
– 287.042
±1.914
β-Ni(OH)2
– 457.100
±1.400
(a)
– 131.800
±1.400
(d)
38.400
±0.400
(b)
– 542.300
±1.500
– 64.046
±6.454
(c)
(b)
80.000
±0.800
82.010
±0.300
–
– 590.519
±9.735
(b)
Ni2OH3+
– 268.181
±5.913
(b)
– 349.955
±6.252
(b)
– 242.636
±27.917
(b)
Ni4(OH)4
– 974.567
±3.203
(b)
– 1173.370
±10.601
(b)
– 137.004
±34.126
(b)
NiF+
– 335.458
±1.132
(b)
– 380.862
±3.193
(b)
– 86.360
±10.305
(b)
NiF2(cr)
– 609.852
±8.001
(a)
– 657.300
±8.000
NiCl+
– 177.447
±3.512
(b)
73.520
±0.400
– 46.100
±7.500
44.400
±0.100
Ni(OH)3
4+
(c)
63.210
±2.000
(c)
(Continued on next page)
III Selected nickel data
45
Table III–1 (continued)
∆ f Gmο
∆ f H mο
S mο
C οp ,m
Compound
NiCl2(cr)
(kJ · mol–1)
– 258.743
±0.154
(a)
(kJ · mol–1)
– 304.900
±0.110
(J · K–1 · mol–1)
98.140
±0.300
(J · K–1 · mol–1)
71.670
±0.300
230.000
±15.000
(c)
NiCl2. 2 H2O(cr)
– 754.382
±1.256
(b)
– 913.374
±1.653
(a)
186.200
±3.600
NiCl2. 4 H2O(cr)
– 1234.946
±1.040
(b)
– 1514.048
±9.179
(b)
249.861
±30.980
(b)
NiCl2. 6 H2O(cr)
– 1713.669
±0.846
(b)
– 2104.700
±1.800
340.964
±6.673
(a)
– 195.408
±2.421
(a)
– 213.500
±2.400
121.400
±1.000
75.400
±1.000
(c)
– 94.360
±0.898
(a)
– 96.420
±0.840
139.100
±1.000
77.280
±1.000
(c)
β-Ni(IO3)2
– 323.735
±1.743
(b)
– 487.112
±4.216
(b)
213.496
±14.075
(b)
Ni(IO3)2. 2 H2O(cr)
– 802.068
±1.832
(b)
– 1087.672
±5.175
(b)
270.058
±17.403
(b)
α-NiS
– 87.792
±1.007
(a)
– 88.100
±1.000
60.890
±0.340
49.760
±0.300
(c)
β-NiS
– 91.330
±1.007
(a)
– 94.000
±1.000
52.970
±0.330
47.080
±0.300
(c)
NiS2(cr)
– 123.832
±7.396
(a)
– 128.000
±7.000
80.000
±8.000
Ni3S2(cr)
– 211.172
±1.624
(a)
– 217.200
±1.600
133.500
±0.700
118.200
±0.800
(c)
Ni9S8(cr)
– 746.803
±9.255
(a)
– 760.000
±9.000
481.000
±7.000
407.000
±5.000
(c)
– 63.098
±2.524
(b)
– 762.688
±1.572
(a)
– 873.280
±1.570
101.300
±0.200
97.600
±0.100
(c)
α-NiSO4. 6 H2O
– 2225.467
±1.059
(a)
– 2683.817
±1.014
(b)
334.450
±1.000
327.900
±1.000
β-NiSO4. 6 H2O
– 2224.915
±1.815
(a)
– 2677.287
±1.034
(b)
354.500
±5.000
NiSO4. 7 H2O(cr)
– 2462.699
±0.929
(b)
– 2977.327
±0.978
(b)
378.950
±1.000
NiBr2(cr)
NiI2(cr)
NiHS+
NiSO4(cr)
(b)
364.600
±4.000
(Continued on next page)
III. Selected nickel data
46
Table III–1 (continued)
∆ f Gmο
∆ f H mο
S mο
C οp ,m
Compound
NiSO4(aq)
(kJ · mol–1)
– 803.191
±0.893
(b)
(kJ · mol–1)
– 958.692
±1.260
(b)
Ni(NO3)2. 2 H2O(cr)
– 1012.732
±1.555
(b)
Ni(NO3)2. 4 H2O(cr)
– 1620.392
±1.441
(b)
Ni(NO3)2. 6 H2O(cr)
– 2214.452
±1.571
(b)
– 159.421
±5.775
(b)
– 1159.167
±1.820
(b)
– 2031.107
±4.843
(b)
HNiP2O7
– 2064.270
±4.766
(b)
NiAs(cr)
– 66.583
±4.606
(a)
+
NiNO3
NiHPO4(aq)
2–
NiP2O7
–
NiAs2(cr)
– 70.820
±4.240
(J · K–1 · mol–1)
– 49.326
±3.135
(J · K–1 · mol–1)
(b)
463.000
±2.000
50.760
±6.000
50.800
±4.000
(c)
490.000
±40.000
(c)
(c)
– 90.100
±8.000
Ni5As2(cr)
– 237.638
±21.914
(a)
– 244.600
±20.000
196.200
±30.000
Ni11As8(cr)
– 715.758
±39.338
(a)
– 743.000
±35.000
518.000
±60.000
Ni3(AsO3)2(cr, hyd)
– 1253.627
±9.251
(b)
Ni3(AsO4)2. 8 H2O(cr)
– 3491.557
±8.823
(b)
– 4179.000
±20.000
540.771
±73.331
NiHAsO4(aq)
– 776.918
±4.426
(b)
NiCO3(cr)
– 636.416
±1.292
(b)
– 713.320
±1.424
(a)
85.400
±2.000
90.300
±4.100
– 1920.881
±1.004
(b)
– 2312.992
±3.147
(a)
311.100
±10.000
405.400
±50.000
– 597.646
±2.441
(b)
NiCO3. 5.5 H2O(cr)
NiCO3(aq)
(a)
(Continued on next page)
III Selected nickel data
47
Table III–1 (continued)
∆ f Gmο
∆ f H mο
S mο
C οp ,m
Compound
(kJ · mol–1)
449.601
±10.129
2–
4
Ni(CN)
(b)
(kJ · mol–1)
353.688
±14.745
(b)
(J · K–1 · mol–1)
245.032
±36.559
(J · K–1 · mol–1)
(b)
Ni(CN)5
626.244
±12.938
(b)
490.638
±19.449
(b)
278.786
±51.082
(b)
NiSCN+
36.596
±4.080
(b)
9.588
±6.463
(b)
7.543
±25.373
(b)
124.273
±8.047
(b)
76.788
±11.348
(b)
137.802
±46.516
(b)
215.089
±12.069
(b)
145.188
±15.645
(b)
261.556
±66.172
(b)
– 1288.441
±5.002
(a)
– 1396.000
±5.000
3–
Ni(SCN)2(aq)
–
Ni(SCN)3
Ni2SiO4(oliv)
128.100
±0.200
ο
(a) Value calculated internally using equation ∆ f Gmο = ∆ f H mο − T ∑ S m,i
.
i
(b) Value calculated internally from reaction data (see Table III–2).
(c) Temperature coefficients of this function are listed in Table III–3.
(d) Value calculated internally from reaction data for a different key species.
123.200
±0.500
(c)
III. Selected nickel data
48
Table III–2: Selected thermodynamic data for reactions involving nickel compounds
and complexes. All ionic species listed in this table are aqueous species. Unless noted
otherwise, all data refer to the reference temperature of 298.15 K and to the standard
state, i.e., a pressure of 0.1 MPa and, for aqueous species, infinite dilution (I = 0). The
uncertainties listed below each value represent total uncertainties and correspond in
principal to the statistically defined 95% confidence interval. Values obtained from internal calculation, cf. footnote (a), are rounded at the third digit after the decimal point
and may therefore not be exactly identical to those given in Part V. Systematically, all
the values are presented with three digits after the decimal point, regardless of the significance of these digits. The data presented in this table are available on computer media from the OECD Nuclear Energy Agency.
Species
Reaction
log 10Ko
o
Ni(cr) U Ni2+ + 2 e
Ni2+
8.019
±0.135
o
∆r G m
(kJ . mol–1)
(b)
∆r H m
(kJ . mol–1)
o
∆r S m
(J . K–1 . mol–1)
–
– 45.773
±0.771
H2O(l) + Ni2+ U H+ + NiOH+
NiOH+
– 9.540
±0.140
–
Ni(OH)3
4+
(a)
157.085
±0.856
190.000
±10.000
110.396
±33.663
(a)
– 8.163
±0.457
9.500
±3.000
59.240
±10.178
(a)
4+
–
F + Ni2+ U NiF+
–
Cl + Ni2+ U NiCl+
0.080
±0.600
NiCl2. 2H2O(cr)
– 48.986
±27.776
60.505
±5.708
4H2O(l) + 4Ni2+ U 4H+ + Ni4(OH)4
1.430
±0.080
NiCl+
45.900
±6.000
166.675
±9.704
H2O(l) + 2Ni2+ U H+ + Ni2OH3+
– 27.520
±0.150
NiF+
(a)
–
– 10.600
±1.000
Ni4(OH)4
– 2.196
±6.300
3H2O(l) + Ni2+ U 3H+ + Ni(OH)3
– 29.200
±1.700
Ni2OH3+
53.800
±1.700
54.455
±0.799
– 0.457
±3.425
NiCl2. 4H2O(cr) U 2H2O(g) + NiCl22H2O(cr)
– 4.099
±0.123
23.400
±0.700
(Continued on next page)
III Selected nickel data
49
Table III–2 (continued)
Species
NiCl2. 4H2O(cr)
Reaction
log 10Ko
o
∆r G m
(kJ . mol–1)
(c)
21.560
±0.600
(c)
(a)
25.958
±16.879
(a)
– 12.167
±0.226
2.593
±0.843
(a)
5.660
±0.810
63.974
±2.777
(a)
–
2IO3 + Ni2+ U β– Ni(IO3)2
– 25.287
±0.114
7.300
±4.000
2H2O(l) + 2IO3 + Ni2+ U Ni(IO3)2. 2H2O(cr)
–
– 29.339
±0.571
– 21.600
±5.000
–
HS + Ni2+ U NiHS+
5.180
±0.200
α– NiSO4. 6H2O
109.296
±13.422
17.380
±0.080
5.140
±0.100
NiHS+
(a)
–
4.430
±0.020
Ni(IO3)2. 2H2O(cr)
286.567
±30.253
107.000
±9.000
2Cl + 6H2O(l) + Ni2+ U NiCl2. 6H2O(cr)
– 3.045
±0.014
β– Ni(IO3)2
o
∆r S m
(J . K–1 . mol–1)
NiCl2. 6H2O(cr) U 2H2O(g) + NiCl24H2O(cr)
– 3.777
±0.105
NiCl2. 6H2O(cr)
o
∆r H m
(kJ . mol–1)
– 29.568
±1.142
6H2O(l) + Ni2+ + SO4 U α– NiSO4. 6H2O
2–
– 4.485
±0.200
β– NiSO4. 6H2O
α– NiSO4. 6H2O U β– NiSO46H2O
6.530
±0.200
NiSO4. 7H2O(cr)
7H2O(l) + Ni2+ + SO4 U NiSO4. 7H2O(cr)
2–
2.267
±0.019
NiSO4(aq)
– 12.940
±0.110
2–
Ni2+ + SO4 U NiSO4(aq)
2.350
±0.030
Ni(NO3)2. 2H2O(cr)
– 13.414
±0.171
2H2O(l) + 2NO3 + Ni2+ U Ni(NO3)2. 2H2O(cr)
–
27.640
±1.000
Ni(NO3)2. 4H2O(cr)
4H2O(l) + 2NO3 + Ni2+ U Ni(NO3)2. 4H2O(cr)
–
– 8.360
±0.800
(Continued on next page)
III. Selected nickel data
50
Table III–2 (continued)
Species
Reaction
log 10Ko
Ni(NO3)2. 6H2O(cr)
o
∆r G m
(kJ . mol–1)
o
∆r H m
(kJ . mol–1)
o
∆r S m
(J . K–1 . mol–1)
6H2O(l) + 2NO3 + Ni2+ U Ni(NO3)2. 6H2O(cr)
–
– 30.760
±1.000
–
+
NO3 + Ni2+ U NiNO3
+
NiNO3
0.500
±1.000
NiHPO4(aq)
– 2.854
±5.708
2–
HPO4 + Ni2+ U NiHPO4(aq)
3.050
±0.090
4–
– 17.410
±0.514
2–
Ni2+ + P2O7 U NiP2O7
2–
NiP2O7
8.730
±0.250
–
HNiP2O7
3–
259.062
±50.537
(a)
163.821
±3.996
– 160.396
±2.854
2–
HAsO4 + Ni2+ U NiHAsO4(aq)
– 16.553
±1.712
CO2(g) + H2O(l) + Ni2+ U 2H+ + NiCO3(cr)
40.870
±1.027
CO2(g) + 6.5H2O(l) + Ni2+ U 2H+ + NiCO3. 5.5H2O(cr)
– 10.630
±0.100
NiCO3(aq)
47.900
±15.000
3–
– 7.160
±0.180
NiCO3. 5.5H2O(cr)
– 29.339
±1.427
2AsO4 + 8H2O(l) + 3Ni2+ U Ni3(AsO4)2. 8H2O(cr)
2.900
±0.300
NiCO3(cr)
(a)
2H2O(l) + 2HAsO2(aq) + 3Ni2+ U 6H+ + Ni3(AsO3)2(cr, hyd)
28.100
±0.500
NiHAsO4(aq)
269.768
±33.880
–
– 28.700
±0.700
Ni3(AsO4)2. 8H2O(cr)
30.600
±10.000
HP2O7 + Ni2+ U HNiP2O7
5.140
±0.250
Ni3(AsO3)2(cr, hyd)
– 49.831
±1.427
60.676
±0.571
2–
CO3 + Ni2+ U NiCO3(aq)
4.200
±0.400
– 23.974
±2.283
(Continued on next page)
III Selected nickel data
51
Table III–2 (continued)
Species
Reaction
log 10Ko
o
∆r G m
(kJ . mol–1)
–
o
∆r H m
(kJ . mol–1)
o
∆r S m
(J . K–1 . mol–1)
2–
4CN + Ni2+ U Ni(CN)4
2–
Ni(CN)4
30.200
±0.120
– 172.383
±0.685
–
– 180.700
±4.000
– 27.896
±13.611
(a)
– 191.100
±8.000
– 95.324
±28.489
(a)
– 11.800
±5.000
– 4.925
±16.788
(a)
– 21.000
±8.000
– 18.935
±26.866
(a)
– 29.000
±10.000
– 39.449
±33.717
(a)
3–
5CN + Ni2+ U Ni(CN)5
3–
Ni(CN)5
28.500
±0.500
– 162.679
±2.854
–
Ni2+ + SCN U NiSCN+
NiSCN+
1.810
±0.040
Ni(SCN)2(aq)
–
Ni2+ + 2SCN U Ni(SCN)2(aq)
2.690
±0.070
–
Ni(SCN)3
– 10.332
±0.228
–
– 15.355
±0.400
–
Ni2+ + 3SCN U Ni(SCN)3
3.020
±0.180
– 17.238
±1.027
(a) Value calculated internally using ∆ r Gmο = ∆ r H mο − T ∆ r S mο .
(b) Value calculated from a selected standard potential.
(c) Value of log10 K ο calculated internally from ∆ r Gmο .
III. Selected nickel data
52
Table III–3: Selected temperature coefficients for heat capacities marked with (c) in
Table III–1, according to the form C pο,m (T ) = a + bT + cT 2 + dT −1 + eT −2 + iT −1/ 2 . The
functions are valid between the temperatures Tmin and Tmax (in K). The notation E±nm
indicates the power of 10. Units for C pο,m are J · K–1 · mol–1.
Compound
Ni(cr)
Ni(cr)
a
b
3.36472E+01 – 2.47530E– 02
– 4.93235E+01
c
d / i*
4.35458E– 05
0
– 3.61955E+05
3.76214E+06
450 600
0
– 6.52809E+09
600 631
1.676480E–01 – 7.48518E–05
Tmin Tmax
298 450
Ni(cr)
1.05023E+05 – 2.29221E+02
Ni(cr)
5.29406E+04 – 1.65828E+02
1.29936E–01
0
0
631 640
Ni(cr)
1.82638E+02 – 4.20407E–01
2.90521E– 04
0
0
640 690
Ni(cr)
1.23677E+01
4.45300E– 06
0
2.46766E+06
690 1728
NiO(cr)
4.11064E+03 – 5.30241E+00
2.43067E+07
298 519
NiO(cr)
– 8.776E+00
2.22800E–02
4.2232E–02
1.40772E–01
e
3.52061E– 03 5.30393E+04(i)
–7.5267E– 06 7.87250E+02(i)
3.6067E+06
519 1800
NiF2(cr)
6.55050E+01
1.53770E– 02
0
0
– 6.11603E+05
250 1450
NiCl2(cr)
7.37949E+01
1.22269E– 02
0
0
– 5.12775E+05
298 1281
NiBr2(cr)
7.23163E+01
1.46270E– 02
0
0
– 1.13805E+05
298 1200
NiI2(cr)
7.35775E+01
1.64250E– 02
0
0
– 1.05775E+05
200 550
α-NiS
4.66760E+01
1.99807E– 02
0
0
– 2.55499E+05
298 660
α-NiS
5.00783E+01
1.49007E– 02
0
0
– 3.36916E+05
660 1000
β-NiS
5.40080E+01 – 1.43700E– 02
3.22820E– 05
0
– 4.90363E+05
298 660
298 835
Ni3S2(cr)
1.65794E+02 – 1.03160E– 01
1.20000E– 04
0
– 2.44462E+06
Ni9S8(cr)
1.07925E+03 – 2.05994E+00
1.98000E– 03
0
– 2.08396E+07 298 640
NiSO4(cr)
1.18934E+02
3.67610E– 02
0
0
– 2.86810E+06 298 1000
NiAs(cr)
3.65400E+01
4.18700E– 02
0
0
1.58000E+05
298 800
Ni11As8(cr)
3.80120E+02
3.67270E– 01
0
0
1.80000E+03
298 800
NiCO3(cr)
8.87010E+01
3.89100E– 02
0
0
– 1.23400E+06
298 700
0
– 3.34214E+06
270 1570
Ni2SiO4(oliv) 1.47616E+02
4.86860E– 02 – 1.50143E– 05
* : One single column is used for the two coefficients d and i. The coefficient concerned is indicated in parentheses behind each value, when any of the two values is
different from zero.
Chapter IV
IV Selected auxiliary data
This chapter presents the chemical thermodynamic data for auxiliary compounds and
complexes which are used within the NEA’s TDB project. Most of these auxiliary species are used in the evaluation of the recommended nickel data in Table III–1, Table
III–2, and Table III–3. It is therefore essential to always use these auxiliary data in conjunction with the selected data for nickel. The use of other auxiliary data can lead to
inconsistencies and erroneous results.
The values in the tables of this chapter are either CODATA Key Values, taken
from [89COX/WAG], or were evaluated within the NEA’s TDB project, as described in
the corresponding chapters of the uranium review [92GRE/FUG], the technetium review [99RAR/RAN], the Update review [2003GUI/FAN], the selenium review
[2005OLI/NOL] and the present Review.
Table IV–1 contains the selected thermodynamic data of the auxiliary species
and Table IV–2 the selected thermodynamic data of chemical reactions involving auxiliary species. The reason for listing both reaction data and entropies, enthalpies and
Gibbs energies of formation is, as described in Chapter III, that uncertainties in reaction
data are often smaller than the derived S mο , ∆ f H mο and ∆ f Gmο , due to uncertainty accumulation during the calculations.
All data in Table IV–1 and Table IV–2 refer to a temperature of 298.15 K, the
standard state pressure of 0.1 MPa and, for aqueous species and reactions, to the infinite
dilution standard state (I = 0).
The uncertainties listed below each reaction value in Table IV–2 are total uncertainties, and correspond mainly to the statistically defined 95% confidence interval.
The uncertainties listed below each value in Table IV–1 have the following significance:
53
54
IV Selected auxiliary data
• for CODATA values from [89COX/WAG], the ± terms have the meaning: “it
is probable, but not at all certain, that the true values of the thermodynamic
quantities differ from the recommended values given in this report by no
more than twice the ± terms attached to the recommended values”.
• for values from [92GRE/FUG], [99RAR/RAN], [2003GUI/FAN], the on-going
NEA TDB review on Selenium (to be published) and the present Review, the
± terms are derived from total uncertainties in the corresponding equilibrium
constant of reaction (cf. Table IV–2), and from the ± terms listed for the necessary CODATA key values.
CODATA [89COX/WAG] values are available for CO2(g), HCO3− , CO32 − ,
H 2 PO 4− and HPO 24 − . From the values given for ∆ f H mο and S mο the values of ∆ f Gmο
and, consequently, all the relevant equilibrium constants and enthalpy changes can be
calculated. The propagation of errors during this procedure, however, leads to uncertainties in the resulting equilibrium constants that are significantly higher than those
obtained from experimental determination of the constants. Therefore, reaction data for
CO2(g), HCO3− , CO32 − , which were absent form the corresponding Table IV–2 in
[92GRE/FUG], are included in this volume to provide the user of selected data for species of nickel (cf. Chapter III) with the data needed to obtain the lowest possible uncertainties on reaction properties.
Note that the values in Table IV–1 and Table IV–2 may contain more digits
than those listed in either [89COX/WAG] or in the discussions on auxiliary data selection of NEA TDB reviews, because the data in the present chapter are retrieved directly
from the computerised data base and rounded to three digits after the decimal point
throughout.
IV Selected auxiliary data
55
Table IV–1: Selected thermodynamic data for auxiliary compounds and complexes
adopted in the NEA TDB project,
including the CODATA Key Values
[89COX/WAG]. All ionic species listed in this table are aqueous species. Unless noted
otherwise, all data refer to 298.15 K and a pressure of 0.1 MPa and, for aqueous species, a standard state of infinite dilution (I = 0). The uncertainties listed below each
value represent total uncertainties and correspond in principle to the statistically defined
95% confidence interval. Values in bold typeface are CODATA Key Values and are
taken directly from Ref. [89COX/WAG] without further evaluation. Values obtained
from internal calculation, cf. footnotes (a) and (b), are rounded at the third digit after the
decimal point. Systematically, all the values are presented with three digits after the
decimal point, regardless of the significance of these digits. The reference listed for
each entry in this table indicates the NEA TDB Review where the corresponding data
have been adopted as NEA TDB Auxiliary data. The data presented in this table are
available on computer media from the OECD Nuclear Energy Agency.
Compound and
NEA TDB Review
∆ f Gmο
∆ f H mο
S mο
(kJ · mol )
249.180
±0.100
O2(g)
[92GRE/FUG]
0.000
0.000
205.152
±0.005
29.378
±0.003
H(g)
[92GRE/FUG]
203.276
±0.006
217.998
±0.006
114.717
±0.002
20.786
±0.001
H+
[92GRE/FUG]
0.000
0.000
0.000
0.000
H2(g)
[92GRE/FUG]
0.000
0.000
130.680
±0.003
28.836
±0.002
O(g)
[92GRE/FUG]
(kJ · mol )
231.743 (a)
±0.100
(a)
–1
–1
C οp ,m
–1
–1
(J · K · mol )
161.059
±0.003
(J · K–1 · mol–1)
21.912
±0.001
OH
[92GRE/FUG]
– 157.220
±0.072
(a)
– 230.015
±0.040
– 10.900
±0.200
H2O(g)
[92GRE/FUG]
– 228.582
±0.040
(a)
– 241.826
±0.040
188.835
±0.010
33.609
±0.030
H2O(l)
[92GRE/FUG]
– 237.140
±0.041
(a)
– 285.830
±0.040
69.950
±0.030
75.351
±0.080
–
H2O2(aq)
[92GRE/FUG]
– 191.170
±0.100
He(g)
[92GRE/FUG]
0.000
0.000
126.153
±0.002
20.786
±0.001
Ne(g)
[92GRE/FUG]
0.000
0.000
146.328
±0.003
20.786
±0.001
(Continued on next page)
IV Selected auxiliary data
56
Table IV–1: (continued)
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
0.000
(kJ · mol–1)
0.000
(J · K–1 · mol–1)
154.846
±0.003
(J · K–1 · mol–1)
20.786
±0.001
Kr(g)
[92GRE/FUG]
0.000
0.000
164.085
±0.003
20.786
±0.001
Xe(g)
[92GRE/FUG]
0.000
0.000
169.685
±0.003
20.786
±0.001
F(g)
[92GRE/FUG]
62.280
±0.300
(a)
79.380
±0.300
158.751
±0.004
22.746
±0.002
F
[92GRE/FUG]
– 281.523
±0.692
(a)
– 335.350
±0.650
– 13.800
±0.800
F2(g)
[92GRE/FUG]
0.000
0.000
202.791
±0.005
HF(aq)
[92GRE/FUG]
– 299.675
±0.702
– 323.150
±0.716
88.000
±3.362
HF(g)
[92GRE/FUG]
– 275.400
±0.700
– 273.300
±0.700
173.779
±0.003
– 655.500
±2.221
92.683
±8.469
Compound and
NEA TDB Review
Ar(g)
[92GRE/FUG]
–
–
(a)
HF2
[92GRE/FUG]
– 583.709
±1.200
Cl(g)
[92GRE/FUG]
105.305
±0.008
(a)
121.301
±0.008
165.190
±0.004
Cl
[92GRE/FUG]
– 131.217
±0.117
(a)
– 167.080
±0.100
56.600
±0.200
Cl2(g)
[92GRE/FUG]
0.000
0.000
223.081
±0.010
–
–
– 37.669
±0.962
–
10.250
±4.044
–
– 7.903
±1.342
(a)
– 104.000
±1.000
162.300
±3.000
ClO4
[92GRE/FUG]
–
– 7.890
±0.600
(a)
– 128.100
±0.400
184.000
±1.500
HCl(g)
[92GRE/FUG]
– 95.298
±0.100
(a)
– 92.310
±0.100
186.902
±0.005
HClO(aq)
[92GRE/FUG]
– 80.023
±0.613
HClO2(aq)
[92GRE/FUG]
– 0.938
±4.043
ClO
[92GRE/FUG]
ClO2
[92GRE/FUG]
ClO3
[92GRE/FUG]
31.304
±0.002
(a)
29.137
±0.002
(a)
21.838
±0.001
33.949
±0.002
29.136
±0.002
(Continued on next page)
IV Selected auxiliary data
57
Table IV–1: (continued)
Compound and
NEA TDB Review
Br(g)
[92GRE/FUG]
–
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
82.379 (a)
±0.128
(kJ · mol–1)
111.870
±0.120
(J · K–1 · mol–1)
175.018
±0.004
(J · K–1 · mol–1)
20.786
±0.001
(a)
– 121.410
±0.150
82.550
±0.200
(a)
30.910
±0.110
245.468
±0.005
0.000
±0.000
152.210
±0.300
Br
[92GRE/FUG]
– 103.850
±0.167
Br2(aq)
[92GRE/FUG]
4.900
±1.000
Br2(g)
[92GRE/FUG]
3.105
±0.142
Br2(l)
[92GRE/FUG]
0.000
±0.000
36.057
±0.002
–
– 32.095
±1.537
BrO3
[92GRE/FUG]
–
19.070
±0.634
(a)
– 66.700
±0.500
161.500
±1.300
HBr(g)
[92GRE/FUG]
– 53.361
±0.166
(a)
– 36.290
±0.160
198.700
±0.004
29.141
±0.003
HBrO(aq)
[92GRE/FUG]
– 81.356
±1.527
(b)
I(g)
[92GRE/FUG]
70.172
±0.060
(a)
106.760
±0.040
180.787
±0.004
20.786
±0.001
I
[92GRE/FUG]
– 51.724
±0.112
(a)
– 56.780
±0.050
106.450
±0.300
I2(cr)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
116.140
±0.300
I2(g)
[92GRE/FUG]
19.323
±0.120
(a)
62.420
±0.080
260.687
±0.005
IO3
[92GRE/FUG]
– 126.338
±0.779
(a)
– 219.700
±0.500
118.000
±2.000
HI(g)
[92GRE/FUG]
1.700
±0.110
(a)
26.500
±0.100
206.590
±0.004
29.157
±0.003
HIO3(aq)
[92GRE/FUG]
– 130.836
±0.797
0.000
±0.000
32.054
±0.050
22.750
±0.050
277.170
±0.150
167.829
±0.006
23.674
±0.001
BrO
[92GRE/FUG]
–
–
S(cr)(orthorhombic)
[92GRE/FUG]
S(g)
[92GRE/FUG]
2–
S
[92GRE/FUG]
0.000
±0.000
236.689
±0.151
(a)
36.888
±0.002
120.695
±11.610
(Continued on next page)
IV Selected auxiliary data
58
Table IV–1: (continued)
Compound and
NEA TDB Review
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(J · K–1 · mol–1)
228.167
±0.010
(J · K–1 · mol–1)
32.505
±0.010
S2(g)
[92GRE/FUG]
(kJ · mol–1)
79.686 (a)
±0.301
(kJ · mol–1)
128.600
±0.300
SO2(g)
[92GRE/FUG]
– 300.095
±0.201
(a)
– 296.810
±0.200
248.223
±0.050
2–
SO3
[92GRE/FUG]
2–
S2O3
[92GRE/FUG]
– 487.472
±4.020
– 519.291
±11.345
2–
– 744.004
±0.418
(a)
– 909.340
±0.400
18.500
±0.400
HS
[92GRE/FUG]
–
12.243
±2.115
(a)
– 16.300
±1.500
67.000
±5.000
H2S(aq)
[92GRE/FUG]
– 27.648
±2.115
(a)
– 38.600
±1.500
126.000
±5.000
H2S(g)
[92GRE/FUG]
– 33.443
±0.500
(a)
– 20.600
±0.500
205.810
±0.050
(a)
– 886.900
±1.000
131.700
±3.000
SO4
[92GRE/FUG]
–
HSO3
[92GRE/FUG]
–
39.842
±0.020
34.248
±0.010
– 528.684
±4.046
HS2O3
[92GRE/FUG]
– 528.366
±11.377
H2SO3(aq)
[92GRE/FUG]
– 539.187
±4.072
–
– 755.315
±1.342
–
– 755.315
±1.342
– 886.900
±1.000
131.700
±3.000
0.000
±0.000
0.000
±0.000
49.221
±0.050
25.550
±0.100
HSO4
[92GRE/FUG]
HSO4
[2001LEM/FUG]
Te(cr)
[92GRE/FUG]
– 265.996
±2.500
(a)
– 321.000
±2.500
69.890
±0.150
60.670
±0.150
N(g)
[92GRE/FUG]
455.537
±0.400
(a)
472.680
±0.400
153.301
±0.003
20.786
±0.001
N2(g)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
191.609
±0.004
29.124
±0.001
348.200
±2.000
275.140
±1.000
107.710
±7.500
– 206.850
±0.400
146.700
±0.400
TeO2(cr)
[2003GUI/FAN]
–
N3
[92GRE/FUG]
–
NO3
[92GRE/FUG]
– 110.794
±0.417
(a)
(c)
(a)
(Continued on next page)
IV Selected auxiliary data
59
Table IV–1: (continued)
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
321.372
±2.051
(kJ · mol–1)
260.140
±10.050
(J · K–1 · mol–1)
147.381
±34.403
(J · K–1 · mol–1)
HN3(aq)
[92GRE/FUG]
NH3(aq)
[92GRE/FUG]
– 26.673
±0.305
– 81.170
±0.326
109.040
±0.913
NH3(g)
[92GRE/FUG]
– 16.407
±0.350
(a)
– 45.940
±0.350
192.770
±0.050
– 79.398
±0.278
(a)
– 133.260
±0.250
111.170
±0.400
Compound and
NEA TDB Review
+
NH4
[92GRE/FUG]
– 7.500
±2.000
P(am)(red)
[92GRE/FUG]
P(cr)(white, cubic)
[92GRE/FUG]
35.630
±0.005
0.000
0.000
41.090
±0.250
23.824
±0.200
P(g)
[92GRE/FUG]
280.093
±1.003
(a)
316.500
±1.000
163.199
±0.003
20.786
±0.001
P2(g)
[92GRE/FUG]
103.469
±2.006
(a)
144.000
±2.000
218.123
±0.004
32.032
±0.002
P4(g)
[92GRE/FUG]
24.419
±0.448
(a)
58.900
±0.300
280.010
±0.500
67.081
±1.500
– 1284.400
±4.085
– 220.970
±12.846
3–
PO4
[92GRE/FUG]
4–
P2O7
[92GRE/FUG]
– 1025.491
±1.576
– 1935.503
±4.563
2–
– 1095.985
±1.567
(a)
– 1299.000
±1.500
– 33.500
±1.500
H2PO4
[92GRE/FUG]
–
– 1137.152
±1.567
(a)
– 1302.600
±1.500
92.500
±1.500
H3PO4(aq)
[92GRE/FUG]
– 1149.367
±1.576
– 1294.120
±1.616
161.912
±2.575
HPO4
[92GRE/FUG]
3–
HP2O7
[92GRE/FUG]
– 1989.158
±4.482
H2P2O7
[92GRE/FUG]
2–
– 2027.117
±4.445
H3P2O7
[92GRE/FUG]
–
– 2039.960
±4.362
H4P2O7(aq)
[92GRE/FUG]
– 2045.668
±3.299
– 2280.210
±3.383
274.919
±6.954
As(cr)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
35.100
±0.600
24.640
±0.500
(Continued on next page)
IV Selected auxiliary data
60
Table IV–1: (continued)
Compound and
NEA TDB Review
–
2
AsO
[92GRE/FUG]
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
– 350.022 (a)
±4.008
(kJ · mol–1)
– 429.030
±4.000
(J · K–1 · mol–1)
40.600
±0.600
(J · K–1 · mol–1)
AsO4
[92GRE/FUG]
– 648.360
±4.008
(a)
– 888.140
±4.000
– 162.800
±0.600
As2O5(cr)
[92GRE/FUG]
– 782.449
±8.016
(a)
– 924.870
±8.000
105.400
±1.200
116.520
±0.800
As4O6(cubic)
[92GRE/FUG]
– 1152.445
±16.032
(a)
– 1313.940
±16.000
214.200
±2.400
191.290
±0.800
As4O6(monoclinic)
[92GRE/FUG]
– 1154.009
±16.041
(a)
– 1309.600
±16.000
234.000
±3.000
As4O6(g)
This Review
– 1092.716
±16.116
(a)
– 1196.250
±16.000
408.600
±6.000
– 402.925
±4.008
(a)
– 456.500
±4.000
125.900
±0.600
H2AsO3
[92GRE/FUG]
– 587.078
±4.008
(a)
– 714.790
±4.000
110.500
±0.600
H3AsO3(aq)
[92GRE/FUG]
– 639.681
±4.015
(a)
– 742.200
±4.000
195.000
±1.000
2–
– 714.592
±4.008
(a)
– 906.340
±4.000
– 1.700
±0.600
H2AsO4
[92GRE/FUG]
–
– 753.203
±4.015
(a)
– 909.560
±4.000
117.000
±1.000
H3AsO4(aq)
[92GRE/FUG]
– 766.119
±4.015
(a)
– 902.500
±4.000
184.000
±1.000
3–
HAsO2(aq)
[92GRE/FUG]
–
HAsO4
[92GRE/FUG]
– 4248.400
±24.000
(As2O5)3. 5 H2O(cr)
[92GRE/FUG]
Sb(cr)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
45.520
±0.210
25.260
±0.200
Bi(cr)
[2001LEM/FUG]
0.000
±0.000
0.000
±0.000
56.740
±0.420
25.410
±0.200
C(cr)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
5.740
±0.100
8.517
±0.080
C(g)
[92GRE/FUG]
671.254
±0.451
(a)
716.680
±0.450
158.100
±0.003
20.839
±0.001
CO(g)
[92GRE/FUG]
– 137.168
±0.173
(a)
– 110.530
±0.170
197.660
±0.004
29.141
±0.002
CO2(aq)
[92GRE/FUG]
– 385.970
±0.270
(a)
– 413.260
±0.200
119.360
±0.600
(Continued on next page)
IV Selected auxiliary data
61
Table IV–1: (continued)
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
– 394.373 (a)
±0.133
(kJ · mol–1)
– 393.510
±0.130
(J · K–1 · mol–1)
213.785
±0.010
(J · K–1 · mol–1)
37.135
±0.002
– 527.900
±0.390
(a)
– 675.230
±0.250
– 50.000
±1.000
– 586.845
±0.251
(a)
– 689.930
±0.200
98.400
±0.500
CN
[2005OLI/NOL]
166.939
±2.519
(b)
147.350
±3.541
(b)
101.182
±8.475
(b)
HCN(aq)
[2005OLI/NOL]
114.368
±2.517
(b)
103.750
±3.536
(b)
131.271
±8.440
(b)
HCN(g)
[2005OLI/NOL]
119.517
±2.500
(a)
129.900
±2.500
201.710
±0.100
Compound and
NEA TDB Review
CO2(g)
[92GRE/FUG]
2–
CO3
[92GRE/FUG]
–
HCO3
[92GRE/FUG]
–
–
∆ f Gmο
(a)
SCN
[92GRE/FUG]
92.700
±4.000
76.400
±4.000
144.268
±18.974
Si(cr)
[92GRE/FUG]
0.000
0.000
18.810
±0.080
19.789
±0.030
Si(g)
[92GRE/FUG]
405.525
±8.000
(a)
450.000
±8.000
167.981
±0.004
22.251
±0.001
– 856.287
±1.002
(a)
– 910.700
±1.000
41.460
±0.200
44.602
±0.300
SiO2(α– quartz)
[92GRE/FUG]
2–
– 1175.651
±1.265
– 1381.960
±15.330
– 1.488
±51.592
SiO(OH)3
[92GRE/FUG]
–
– 1251.740
±1.162
– 1431.360
±3.743
88.024
±13.144
Si(OH)4(aq)
[92GRE/FUG]
– 1307.735
±1.156
SiO2(OH)2
[92GRE/FUG]
2–
– 2269.878
±2.878
–
– 2332.096
±2.878
3–
– 3048.536
±3.870
3–
– 3291.955
±3.869
4–
– 4075.179
±5.437
3–
– 4136.826
±4.934
Si2O3(OH)4
[92GRE/FUG]
Si2O2(OH)5
[92GRE/FUG]
Si3O6(OH)3
[92GRE/FUG]
Si3O5(OH)5
[92GRE/FUG]
Si4O8(OH)4
[92GRE/FUG]
Si4O7(OH)5
[92GRE/FUG]
(b)
– 1456.960
±3.163
(b)
189.973
±11.296
(b)
(Continued on next page)
IV Selected auxiliary data
62
Table IV–1: (continued)
Compound and
NEA TDB Review
SiF4(g)
[92GRE/FUG]
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
– 1572.773 (a)
±0.814
(kJ · mol–1)
– 1615.000
±0.800
(J · K–1 · mol–1)
282.760
±0.500
(J · K–1 · mol–1)
73.622
±0.500
Ge(cr)
[92GRE/FUG]
0.000
Ge(g)
[92GRE/FUG]
331.209
±3.000
0.000
31.090
±0.150
23.222
±0.100
(a)
372.000
±3.000
167.904
±0.005
30.733
±0.001
– 521.404
±1.002
(a)
– 580.000
±1.000
39.710
±0.150
50.166
±0.300
GeF4(g)
[92GRE/FUG]
– 1150.018
±0.584
(a)
– 1190.200
±0.500
301.900
±1.000
81.602
±1.000
Sn(cr)
[92GRE/FUG]
0.000
0.000
51.180
±0.080
27.112
±0.030
Sn(g)
[92GRE/FUG]
266.223
±1.500
(a)
301.200
±1.500
168.492
±0.004
21.259
±0.001
Sn2+
[92GRE/FUG]
– 27.624
±1.557
(a)
– 8.900
±1.000
– 16.700
±4.000
SnO(tetragonal)
[92GRE/FUG]
– 251.913
±0.220
(a)
– 280.710
±0.200
57.170
±0.300
47.783
±0.300
SnO2(cassiterite,
tetragonal)
[92GRE/FUG]
Pb(cr)
[92GRE/FUG]
– 515.826
±0.204
(a)
– 577.630
±0.200
49.040
±0.100
53.219
±0.200
0.000
64.800
±0.300
26.650
±0.100
20.786
±0.001
GeO2(tetragonal)
[92GRE/FUG]
0.000
Pb(g)
[92GRE/FUG]
162.232
±0.805
(a)
195.200
±0.800
175.375
±0.005
Pb2+
[92GRE/FUG]
– 24.238
±0.399
(a)
0.920
±0.250
18.500
±1.000
PbSO4(cr)
[92GRE/FUG]
– 813.036
±0.447
(a)
– 919.970
±0.400
148.500
±0.600
B(cr)
[92GRE/FUG]
0.000
0.000
5.900
±0.080
11.087
±0.100
B(g)
[92GRE/FUG]
521.012
±5.000
(a)
565.000
±5.000
153.436
±0.015
20.796
±0.005
B2O3(cr)
[92GRE/FUG]
– 1194.324
±1.404
(a)
– 1273.500
±1.400
53.970
±0.300
62.761
±0.300
B(OH)3(aq)
[92GRE/FUG]
– 969.268
±0.820
(a)
– 1072.800
±0.800
162.400
±0.600
B(OH)3(cr)
[92GRE/FUG]
– 969.667
±0.820
(a)
– 1094.800
±0.800
89.950
±0.600
86.060
±0.400
(Continued on next page)
IV Selected auxiliary data
63
Table IV–1: (continued)
Compound and
NEA TDB Review
BF3(g)
[92GRE/FUG]
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
– 1119.403 (a)
±0.803
(kJ · mol–1)
– 1136.000
±0.800
(J · K–1 · mol–1)
254.420
±0.200
(J · K–1 · mol–1)
50.463
±0.100
Al(cr)
[92GRE/FUG]
0.000
Al(g)
[92GRE/FUG]
289.376
±4.000
Al3+
[92GRE/FUG]
0.000
28.300
±0.100
24.200
±0.070
(a)
330.000
±4.000
164.554
±0.004
21.391
±0.001
– 491.507
±3.338
(a)
– 538.400
±1.500
– 325.000
±10.000
Al2O3(corundum)
[92GRE/FUG]
– 1582.257
±1.302
(a)
– 1675.700
±1.300
50.920
±0.100
79.033
±0.200
AlF3(cr)
[92GRE/FUG]
– 1431.096
±1.309
(a)
– 1510.400
±1.300
66.500
±0.500
75.122
±0.400
Tl+
[99RAR/RAN]
– 32.400
±0.300
Zn(cr)
[92GRE/FUG]
0.000
0.000
41.630
±0.150
25.390
±0.040
Zn(g)
[92GRE/FUG]
94.813
±0.402
(a)
130.400
±0.400
160.990
±0.004
20.786
±0.001
Zn2+
[92GRE/FUG]
– 147.203
±0.254
(a)
– 153.390
±0.200
– 109.800
±0.500
ZnO(cr)
[92GRE/FUG]
– 320.479
±0.299
(a)
– 350.460
±0.270
43.650
±0.400
Cd(cr)
[92GRE/FUG]
0.000
0.000
51.800
±0.150
26.020
±0.040
Cd(g)
[92GRE/FUG]
77.230
±0.205
(a)
111.800
±0.200
167.749
±0.004
20.786
±0.001
Cd2+
[92GRE/FUG]
– 77.733
±0.750
(a)
– 75.920
±0.600
– 72.800
±1.500
CdO(cr)
[92GRE/FUG]
– 228.661
±0.602
(a)
– 258.350
±0.400
54.800
±1.500
– 1464.959
±0.810
(a)
– 1729.300
±0.800
229.650
±0.400
Hg(g)
[92GRE/FUG]
31.842
±0.054
(a)
61.380
±0.040
174.971
±0.005
Hg(l)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
75.900
±0.120
Hg2+
[92GRE/FUG]
164.667
±0.313
170.210
±0.200
– 36.190
±0.800
CdSO4. 2.667 H2O(cr)
[92GRE/FUG]
(a)
20.786
±0.001
(Continued on next page)
IV Selected auxiliary data
64
Table IV–1: (continued)
Compound and
NEA TDB Review
2+
2
Hg
[92GRE/FUG]
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
153.567 (a)
±0.559
(kJ · mol–1)
166.870
±0.500
(J · K–1 · mol–1)
65.740
±0.800
(J · K–1 · mol–1)
HgO(montroydite,
red)
[92GRE/FUG]
Hg2Cl2(cr)
[92GRE/FUG]
– 58.523
±0.154
(a)
– 90.790
±0.120
70.250
±0.300
– 210.725
±0.471
(a)
– 265.370
±0.400
191.600
±0.800
Hg2SO4(cr)
[92GRE/FUG]
– 625.780
±0.411
(a)
– 743.090
±0.400
200.700
±0.200
Cu(cr)
[92GRE/FUG]
0.000
0.000
33.150
±0.080
24.440
±0.050
Cu(g)
[92GRE/FUG]
297.672
±1.200
(a)
337.400
±1.200
166.398
±0.004
20.786
±0.001
Cu2+
[92GRE/FUG]
65.040
±1.557
(a)
64.900
±1.000
– 98.000
±4.000
CuCl(g)
[2003GUI/FAN]
76.800
±10.000
CuSO4(cr)
[92GRE/FUG]
– 662.185
±1.206
Ag(cr)
[92GRE/FUG]
0.000
Ag(g)
[92GRE/FUG]
246.007
±0.802
Ag+
[92GRE/FUG]
(a)
– 771.400
±1.200
109.200
±0.400
0.000
42.550
±0.200
25.350
±0.100
(a)
284.900
±0.800
172.997
±0.004
20.786
±0.001
77.096
±0.156
(a)
105.790
±0.080
73.450
±0.400
AgCl(cr)
[92GRE/FUG]
– 109.765
±0.098
(a)
– 127.010
±0.050
96.250
±0.200
Ti(cr)
[92GRE/FUG]
0.000
0.000
30.720
±0.100
25.060
±0.080
Ti(g)
[92GRE/FUG]
428.403
±3.000
(a)
473.000
±3.000
180.298
±0.010
24.430
±0.030
TiO2(rutile)
[92GRE/FUG]
– 888.767
±0.806
(a)
– 944.000
±0.800
50.620
±0.300
55.080
±0.300
TiCl4(g)
[92GRE/FUG]
– 726.324
±3.229
(a)
– 763.200
±3.000
353.200
±4.000
95.408
±1.000
Th(cr)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
51.800
±0.500
26.230
±0.050
Th(g)
[92GRE/FUG]
560.745
±6.002
602.000
±6.000
190.170
±0.050
20.789
±0.100
(a)
(Continued on next page)
IV Selected auxiliary data
65
Table IV–1: (continued)
Compound and
NEA TDB Review
ThO2(cr)
[92GRE/FUG]
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
– 1169.238 (a)
±3.504
(kJ · mol–1)
– 1226.400
±3.500
(J · K–1 · mol–1)
65.230
±0.200
(J · K–1 · mol–1)
Be(cr)
[92GRE/FUG]
0.000
Be(g)
[92GRE/FUG]
286.202
±5.000
– 580.090
±2.500
BeO(bromellite)
[92GRE/FUG]
0.000
9.500
±0.080
16.443
±0.060
(a)
324.000
±5.000
136.275
±0.003
20.786
±0.001
(a)
– 609.400
±2.500
13.770
±0.040
25.565
±0.100
0.000
32.670
±0.100
24.869
±0.020
20.786
±0.001
Mg(cr)
[92GRE/FUG]
0.000
Mg(g)
[92GRE/FUG]
112.521
±0.801
(a)
147.100
±0.800
148.648
±0.003
Mg2+
[92GRE/FUG]
– 455.375
±1.335
(a)
– 467.000
±0.600
– 137.000
±4.000
MgO(cr)
[92GRE/FUG]
– 569.312
±0.305
(a)
– 601.600
±0.300
26.950
±0.150
37.237
±0.200
MgF2(cr)
[92GRE/FUG]
– 1071.051
±1.210
(a)
– 1124.200
±1.200
57.200
±0.500
61.512
±0.300
Ca(cr)
[92GRE/FUG]
0.000
0.000
41.590
±0.400
25.929
±0.300
Ca(g)
[92GRE/FUG]
144.021
±0.809
(a)
177.800
±0.800
154.887
±0.004
20.786
±0.001
Ca2+
[92GRE/FUG]
– 552.806
±1.050
(a)
– 543.000
±1.000
– 56.200
±1.000
CaO(cr)
[92GRE/FUG]
– 603.296
±0.916
(a)
– 634.920
±0.900
38.100
±0.400
42.049
±0.400
CaF(g)
[2003GUI/FAN]
– 302.118
±5.104
– 276.404
±5.100
229.244
±0.500
33.671
±0.500
CaCl(g)
[2003GUI/FAN]
– 129.787
±5.001
– 103.400
±5.000
241.634
±0.300
35.687
±0.010
Sr(cr)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
55.700
±0.210
Sr2+
[92GRE/FUG]
– 563.864
±0.781
(a)
– 550.900
±0.500
– 31.500
±2.000
SrO(cr)
[92GRE/FUG]
– 559.939
±0.914
(a)
– 590.600
±0.900
55.440
±0.500
SrCl2(cr)
[92GRE/FUG]
– 784.974
±0.714
(a)
– 833.850
±0.700
114.850
±0.420
(Continued on next page)
IV Selected auxiliary data
66
Table IV–1: (continued)
Compound and
NEA TDB Review
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
– 982.360
±0.800
(J · K–1 · mol–1)
194.600
±2.100
(J · K–1 · mol–1)
Sr(NO3)2(cr)
[92GRE/FUG]
(kJ · mol–1)
– 783.146 (a)
±1.018
Ba(cr)
[92GRE/FUG]
0.000
±0.000
0.000
±0.000
62.420
±0.840
152.852
±5.006
185.000
±5.000
170.245
±0.010
Ba(g)
[2003GUI/FAN]
20.786
±0.001
Ba2+
[92GRE/FUG]
– 557.656
±2.582
(a)
– 534.800
±2.500
8.400
±2.000
BaO(cr)
[92GRE/FUG]
– 520.394
±2.515
(a)
– 548.100
±2.500
72.070
±0.380
BaF(g)
[2003GUI/FAN]
– 349.569
±6.705
– 324.992
±6.700
246.219
±0.210
BaCl2(cr)
[92GRE/FUG]
– 806.953
±2.514
– 855.200
±2.500
123.680
±0.250
Li(cr)
[92GRE/FUG]
0.000
0.000
29.120
±0.200
24.860
±0.200
Li(g)
[92GRE/FUG]
126.604
±1.002
(a)
159.300
±1.000
138.782
±0.010
20.786
±0.001
Li+
[92GRE/FUG]
– 292.918
±0.109
(a)
– 278.470
±0.080
12.240
±0.150
Na(cr)
[92GRE/FUG]
0.000
0.000
51.300
±0.200
28.230
±0.200
Na(g)
[92GRE/FUG]
76.964
±0.703
(a)
107.500
±0.700
153.718
±0.003
20.786
±0.001
Na+
[92GRE/FUG]
– 261.953
±0.096
(a)
– 240.340
±0.060
58.450
±0.150
NaF(cr)
[2001LEM/FUG]
– 546.327
±0.704
(a)
– 576.600
±0.700
51.160
±0.150
NaCl(cr)
[2001LEM/FUG]
– 384.221
±0.147
– 411.260
±0.120
72.150
±0.200
50.500
0.000
64.680
±0.200
29.600
±0.100
20.786
±0.001
(a)
NaNO3(cr)
[2003GUI/FAN]
34.747
±0.300
– 467.580
±0.410
K(cr)
[92GRE/FUG]
0.000
K(g)
[92GRE/FUG]
60.479
±0.802
(a)
89.000
±0.800
160.341
±0.003
K+
[92GRE/FUG]
– 282.510
±0.116
(a)
– 252.140
±0.080
101.200
±0.200
(Continued on next page)
IV Selected auxiliary data
67
Table IV–1: (continued)
∆ f Gmο
∆ f H mο
S mο
C οp ,m
(kJ · mol–1)
(kJ · mol–1)
– 436.461
±0.129
(J · K–1 · mol–1)
(J · K–1 · mol–1)
KCl(cr)
This Review
KBr(cr)
This Review
– 393.330
±0.188
KI(cr)
This Review
– 329.150
±0.137
Compound and
NEA TDB Review
Rb(cr)
[92GRE/FUG]
0.000
Rb(g)
[92GRE/FUG]
53.078
±0.805
Rb+
[92GRE/FUG]
– 284.009
±0.153
Cs(cr)
[92GRE/FUG]
0.000
Cs(g)
[92GRE/FUG]
49.556
±1.007
Cs+
[92GRE/FUG]
0.000
76.780
±0.300
31.060
±0.100
(a)
80.900
±0.800
170.094
±0.003
20.786
±0.001
(a)
– 251.120
±0.100
121.750
±0.250
0.000
85.230
±0.400
32.210
±0.200
(a)
76.500
±1.000
175.601
±0.003
20.786
±0.001
– 291.456
±0.535
(a)
– 258.000
±0.500
132.100
±0.500
CsCl(cr)
[2001LEM/FUG]
– 413.807
±0.208
(a)
– 442.310
±0.160
101.170
±0.200
52.470
CsBr(cr)
[2001LEM/FUG]
– 391.171
±0.305
– 405.600
±0.250
112.940
±0.400
52.930
(a)
(b)
ο
Value calculated internally using ∆ f Gmο = ∆ f H mο − T ∑ S m,
i .
i
Value calculated internally from reaction data (see Table IV–2).
IV Selected auxiliary data
68
Table IV–2: Selected thermodynamic data for reactions involving auxiliary compounds
and complexes used in the evaluation of thermodynamic data for the NEA TDB Project
data. All ionic species listed in this table are aqueous species. Unless noted otherwise,
all data refer to 298.15 K and a pressure of 0.1 MPa and, for aqueous species, a standard
state of infinite dilution (I = 0). The uncertainties listed below each value represent total
uncertainties and correspond in principle to the statistically defined 95% confidence
interval. Systematically, all the values are presented with three digits after the decimal
point, regardless of the significance of these digits. The reference listed for each entry in
this table indicates the NEA TDB Review where the corresponding data have been
adopted as NEA TDB Auxiliary data.The data presented in this table are available on
computer media from the OECD Nuclear Energy Agency.
Species and
NEA TDB Review
Reaction
log10 K ο
∆ r Gmο
(kJ · mol–1)
HF(aq)
[92GRE/FUG]
–
HF2
[92GRE/FUG]
–
ClO
[92GRE/FUG]
∆ r H mο
(kJ · mol–1)
–
3.180
±0.020
–
– 18.152
±0.114
12.200
±0.300
101.800
±1.077
(a)
– 2.511
±0.685
3.000
±2.000
18.486
±7.090
(a)
42.354
±0.742
19.000
±9.000
– 78.329
±30.289
(a)
– 64.600
±10.078
(a)
–
F + HF(aq) U HF2
0.440
±0.120
–
HClO(aq) U ClO + H+
– 7.420
±0.130
–
HClO2(aq) U ClO2 + H+
HClO(aq)
[92GRE/FUG]
Cl2(g) + H2O(l) U Cl + H+ + HClO(aq)
HClO2(aq)
[92GRE/FUG]
H2O(l) + HClO(aq) U 2H+ + HClO2(aq) + 2 e
– 1.960
±0.020
11.188
±0.114
–
– 4.537
±0.105
– 55.400
±0.700
25.900
±0.600
(b)
–
316.230
±3.996
–
BrO
[92GRE/FUG]
HBrO(aq) U BrO + H+
HBrO(aq)
[92GRE/FUG]
Br2(aq) + H2O(l) U Br + H+ + HBrO(aq)
–
(J · K–1 · mol–1)
F + H+ U HF(aq)
ClO2
[92GRE/FUG]
–
∆ r S mο
– 8.630
±0.030
49.260
±0.171
30.000
±3.000
–
– 8.240
±0.200
47.034
±1.142
(Continued on next page)
IV Selected auxiliary data
69
Table IV–2 (continued)
Species and
NEA TDB Review
Reaction
log10 K ο
∆ r Gmο
∆ r H mο
(kJ · mol–1)
HIO3(aq)
[92GRE/FUG]
2–
S
[92GRE/FUG]
2–
SO3
[92GRE/FUG]
(kJ · mol–1)
–
0.788
±0.029
–
– 4.498
±0.166
2–
HS U H+ + S
– 19.000
±2.000
108.450
±11.416
–
2–
–
2–
H2O(l) + SO4 + 2 e U 2OH + SO3
– 31.400
±0.700
(b)
179.230
±3.996
–
2–
–
2–
3H2O(l) + 2SO3 + 4 e U 6OH + S2O3
H2S(aq)
[92GRE/FUG]
H2S(aq) U H+ + HS
–
HSO3
[92GRE/FUG]
– 39.200
±1.400
(b)
223.760
±7.991
–
– 6.990
±0.170
2–
–
H+ + SO3 U HSO3
7.220
±0.080
66.000
±30.000
359.590
±100.630
(a)
16.000
±5.000
88.891
±16.840
(a)
– 26.828
±0.457
– 15.000
±10.000
39.671
±33.575
(a)
52.725
±0.126
52.090
±0.210
– 2.130
±0.821
(a)
– 41.212
±0.457
2–
–
H+ + S2O3 U HS2O3
H2SO3(aq)
[92GRE/FUG]
H+ + HSO3 U H2SO3(aq)
1.590
±0.150
– 9.076
±0.856
–
1.840
±0.080
2–
– 10.503
±0.457
–
HSO4
[92GRE/FUG]
H+ + SO4 U HSO4
HN3(aq)
[92GRE/FUG]
H+ + N3 U HN3(aq)
NH3(aq)
[92GRE/FUG]
NH4 U H+ + NH3(aq)
–
(a)
39.899
±0.970
HS2O3
[92GRE/FUG]
–
(J · K–1 · mol–1)
H+ + IO3 U HIO3(aq)
S2O3
[92GRE/FUG]
2–
∆ r S mο
1.980
±0.050
– 11.302
±0.285
–
4.700
±0.080
+
– 9.237
±0.022
(Continued on next page)
IV Selected auxiliary data
70
Table IV–2 (continued)
Species and
NEA TDB Review
Reaction
log10 K ο
∆ r Gmο
(kJ · mol–1)
HNO2(aq)
[92GRE/FUG]
3–
PO4
[92GRE/FUG]
4–
P2O7
[92GRE/FUG]
3.210
±0.160
2–
– 18.323
±0.913
– 11.400
±3.000
23.219
±10.518
(a)
70.494
±0.171
14.600
±3.800
– 187.470
±12.758
(a)
– 3.600
±1.000
126.000
±3.363
(a)
8.480
±0.600
69.412
±2.093
(a)
22.200
±1.000
21.045
±4.673
(a)
3–
HPO4 U H+ + PO4
– 12.350
±0.030
3–
4–
HP2O7 U H+ + P2O7
– 9.400
±0.150
2–
53.656
±0.856
–
H3PO4(aq)
[92GRE/FUG]
H+ + H2PO4 U H3PO4(aq)
7.212
±0.013
– 41.166
±0.074
–
2.140
±0.030
2–
– 12.215
±0.171
3–
HP2O7
[92GRE/FUG]
H2P2O7 U H+ + HP2O7
2–
H3P2O7 U H+ + H2P2O7
H2P2O7
[92GRE/FUG]
– 6.650
±0.100
–
37.958
±0.571
2–
– 2.250
±0.150
12.843
±0.856
–
H3P2O7
[92GRE/FUG]
H4P2O7(aq) U H+ + H3P2O7
H4P2O7(aq)
[92GRE/FUG]
2H3PO4(aq) U H2O(l) + H4P2O7(aq)
CO2(aq)
[92GRE/FUG]
H+ + HCO3 U CO2(aq) + H2O(l)
CO2(g)
[92GRE/FUG]
CO2(aq) U CO2(g)
–
(J · K–1 · mol–1)
–
H+ + HPO4 U H2PO4
3–
(kJ · mol–1)
∆ r S mο
H+ + NO2 U HNO2(aq)
H2PO4
[92GRE/FUG]
–
∆ r H mο
– 1.000
±0.500
– 2.790
±0.170
5.708
±2.854
15.925
±0.970
–
6.354
±0.020
1.472
±0.020
– 36.269
±0.114
(a)
– 8.402
±0.114
(a)
(Continued on next page)
IV Selected auxiliary data
71
Table IV–2 (continued)
Species and
NEA TDB Review
Reaction
log10 K ο
∆ r Gmο
∆ r H mο
(kJ · mol–1)
–
HCO3
[92GRE/FUG]
2–
–
10.329
±0.020
–
HCN(aq)
[2005OLI/NOL]
HCN(g) U HCN(aq)
2–
– 9.210
±0.020
0.902
±0.050
52.571
±0.114
43.600
±0.200
– 30.089
±0.772
(a)
– 5.149
±0.285
– 26.150
±2.500
– 70.439
±8.440
(a)
75.000
±15.000
– 191.460
±50.340
(a)
25.600
±2.000
– 101.950
±6.719
(a)
25.400
±3.000
8.613
±10.243
(a)
2–
Si(OH)4(aq) U 2H+ + SiO2(OH)2
– 23.140
±0.090
132.080
±0.514
–
SiO(OH)3
[92GRE/FUG]
Si(OH)4(aq) U H+ + SiO(OH)3
Si(OH)4(aq)
[92GRE/FUG]$
2H2O(l) + SiO2(quar) U Si(OH)4(aq)
–
2–
Si2O3(OH)4
[92GRE/FUG]
–
Si2O2(OH)5
[92GRE/FUG]
3–
Si3O6(OH)3
[92GRE/FUG]
3–
Si3O5(OH)5
[92GRE/FUG]
(a)
– 58.958
±0.114
HCN(aq) U CN + H+
SiO2(OH)2
[92GRE/FUG]
(J · K–1 · mol–1)
CO3 + H+ U HCO3
CN
[2005OLI/NOL]
–
(kJ · mol–1)
∆ r S mο
– 9.810
±0.020
– 4.000
±0.100
55.996
±0.114
22.832
±0.571
2–
2Si(OH)4(aq) U 2H+ + H2O(l) + Si2O3(OH)4
– 19.000
±0.300
108.450
±1.712
–
2Si(OH)4(aq) U H+ + H2O(l) + Si2O2(OH)5
– 8.100
±0.300
46.235
±1.712
3–
3Si(OH)4(aq) U 3H+ + 3H2O(l) + Si3O6(OH)3
– 28.600
±0.300
163.250
±1.712
3–
3Si(OH)4(aq) U 3H+ + 2H2O(l) + Si3O5(OH)5
– 27.500
±0.300
156.970
±1.712
(Continued on next page)
IV Selected auxiliary data
72
Table IV–2 (continued)
Species and
NEA TDB Review
Reaction
log10 K ο
∆ r Gmο
(kJ · mol–1)
4–
Si4O8(OH)4
[92GRE/FUG]
3–
Si4O7(OH)5
[92GRE/FUG]
∆ r H mο
(kJ · mol–1)
4–
4Si(OH)4(aq) U 4H+ + 4H2O(l) + Si4O8(OH)4
– 36.300
±0.500
207.200
±2.854
3–
4Si(OH)4(aq) U 3H+ + 4H2O(l) + Si4O7(OH)5
– 25.500
±0.300
145.560
±1.712
(a) Value calculated internally using ∆ r Gmο = ∆ r H mο − T ∆ r S mο .
(b) Value calculated from a selected standard potential.
(c) Value of log10 K ο calculated internally from ∆ r Gmο .
∆ r S mο
(J · K–1 · mol–1)
Part III
Discussion of data selection
73
Chapter VEquation Section 22
Discussion of data selection for
nickel
V.1 Elemental nickel
V.1.1 Nickel gas
In this review, it is assumed that Ni forms an ideal gas phase. Our recommendations are
based on the critically evaluated data given in [73HUL/DES], [76MAH/PAN] and
[98CHA]. Gaseous ions e.g., Ni+(g), Ni2+(g) and Ni–(g) also evaluated in [98CHA],
[76MAH/PAN] and [73HUL/DES] are not comprehensively treated by this review.
An interest for the interpretation of spectra of highly-ionised atoms has arisen
from the investigation of hot plasmas generated to achieve nuclear fusion. These activities have produced a substantial increase in spectroscopic information and the preparation of new compilations of energy levels. The energy levels of nickel atoms and its
ions, as derived from analyses of atomic spectra, have been published by Corliss and
Sugar [81COR/SUG]. Based on this publication, the thermodynamic data for Ni+(g) and
Ni–(g) have been calculated by [98CHA]. The source of data for the calculation of the
enthalpy of formation, the heat capacity and the entropy for Ni2+(g) ions were the energy levels deduced from NiII and NiIII spectra. Such calculations were carried out by
Mah and Pankratz [76MAH/PAN].
The enthalpy of formation for Ni(g) was calculated by [98CHA] by analysis of
the vapour pressure data over liquid nickel obtained by [57MOR/ZEL]. This calculations gave ∆ vap H mο (Ni, 298.15 K) = (412.29 ± 0.50) kJ·mol–1 (3rd law) and
(415.57 ± 5.15) kJ·mol–1 (2nd law).
Using the 3rd law ∆ vap H mο (Ni, 298.15 K) and ∆ f H mο (Ni, l, 298.15 K) the enthalpy of formation ∆ f H mο (Ni, g, 298.15 K) was calculated [98CHA]. The value
∆ f H mο (Ni, g, 298.15 K) = (430.1 ± 8.4) kJ·mol–1
obtained by [98CHA] is the same as the one chosen by [76MAH/PAN] and
[73HUL/DES]. This value is also selected in the present review.
75
V.1 Elemental nickel
76
Rutner and Haury [74RUT/HAU] determined the “best value” of the heat of
vaporisation and sublimation of nickel by a statistical treatment of data available in the
literature and their own results obtained by the Langmuir technique. However, the difference of the entropy of sublimation determined by the 2nd and 3rd law analysis
− (10.5 ± 16.3) J·K–1·mol–1, being a criterion of the precision of measurements, was
rather high. Therefore, Chase [98CHA] has not used results of [74RUT/HAU] for the
calculation of the enthalpy of formation for Ni(g). Other vapour pressure studies were
discussed by Hultgren et al. [73HUL/DES].
The source of data for the calculation of the heat capacity and entropy of Ni(g),
tabulated in [98CHA], [76MAH/PAN] and [73HUL/DES], were the calculations of
electronic energy levels and quantum weights done by [70ROT] and [68MOO/MER].
All tabulated entropy values agree within 0.004 J·K–1·mol–1, therefore, the present reviewers select the values:
S mο (Ni, g, 298.15 K) = (182.19 ± 0.08) J·K–1·mol–1
C pο,m (Ni, g, 298.15 K) = (23.36 ± 0.10) J·K–1·mol–1
which are the same as those recommended by [98CHA]. With the selections of ∆ f H mο
and S mο the Gibbs energy can be calculated to be:
∆ f Gmο (Ni, g, 298.15 K) = (384.7 ± 8.4) kJ·mol–1.
V.1.2 Nickel crystal
Ni metal has a fcc (A1) structure existing up to the melting point of the solid at
(1726 ± 4) K [73HUL/DES]. Ferromagnetic Ni is frequently designated Ni(α) and paramagnetic Ni, Ni(β) [60KEL].
The melting point selected by Hultgren et al. [73HUL/DES] and also retained
by this review, Tfus = (1726 ± 4) K, is based on the following determinations:
[66VOL/KOH]: 1725 K
[63GEO/FER]: 1726 K
[54ORI/JON]:
(1725 ± 4) K.
The value of the Curie point, Tc = 631 K, is taken from the study of Connelly et
al. [71CON/LOO] using a calorimetric method which permits continuous observation of
C p ,m versus T, and from the study of Vollmer et al. [66VOL/KOH] using a high temperature adiabatic calorimeter.
Pure Ni metal is defined as the nickel reference phase. As such, its Gibbs energy of formation and enthalpy of formation are zero by definition at 298.15 K and
0.1 MPa.
The source of data for the entropy determination of nickel has been the temperature dependence of the heat capacities. The measurements of the molar heat of
V.1 Elemental nickel
77
nickel in the temperature range 1.1 to 19.0 K were performed by Keesom and Clark
[35KEE/CLA]. The standard molar entropy of Ni determined by Busey and Giauque
[52BUS/GIA2], using measurements of the heat capacity from 15 to 300 K, was equal
to 29.86 J·K–1·mol–1.
In the present review, the selected value of the standard molar entropy of crystalline nickel, Ni(cr), is based on the above determination:
S mο (Ni, cr, 298.15 K) = (29.87 ± 0.20) J·K–1·mol–1.
This is essentially the same as the values given in the NIST-JANAF Thermochemical Tables [98CHA] (29.87 ± 0.21) J·K–1·mol–1, [76MAH/PAN] (29.87 ± 0.08)
J·K–1·mol–1, [73HUL/DES] (29.87 ± 0.08) J·K–1·mol–1 and in [77BAR/KNA] 29.87
J·K−1·mol–1.
In the literature, numerous heat capacity studies of Ni metal, especially in the
vicinity of the Curie point, are reported. The studies upon which our adopted values are
based are as follows: [36BRO/WIL], [38SYK/WIL], [55KRA/WAR], [65PAW/STA] ,
[66VOL/KOH] and [71CON/LOO]. The experimental results of these studies and a
polynomial fitting curve derived for this review are shown in Figure V-1.
Figure V-1: The standard molar heat capacity of nickel as a function of temperature.
The data are taken from (+)[36BRO/WIL], ( )[38SYK/WIL], ( )[55KRA/WAR],
( )[65PAW/STA], ( )[66VOL/KOH] and ( )[71CON/LOO]. The dotted line corresponds to the thermal heat capacity function given by [77BAR/KNA]. The fitted curve
(solid line) refers to the thermal heat capacity function recommended by this review.
40
38
C°p ,m / J·K–1·mol–1
36
34
32
30
28
26
24
22
20
200
400
600
800
1000
T/K
1200
1400
1600
1800
V.1 Elemental nickel
78
The heat capacity calculated using the temperature coefficients given by Barin
and Knacke [77BAR/KNA] and the experimental results disagree conspicuously in the
temperature range 640 < T/K < 1728 (Figure V-1). Based on the fitted curve, this review
selects the heat capacity of Ni at 298.15 K:
C pο,m (Ni, cr, 298.15 K) = (26.07 ± 0.10) J·K–1·mol–1
and the temperature coefficients of the heat capacity function: C pο,m (T) = a + bT + cT2 +
eT–2, which are listed in Table V-1. The recommended value for the heat capacity at
298.15 K is consistent with the values selected by the US Bureau of Mines
[76MAH/PAN], 25.98 J·K–1·mol–1, the US Geological Survey [95ROB/HEM], 25.99
J·K–1·mol–1, the value selected by Hultgren et al. [73HUL/DES], 26.07 J·K–1·mol–1, the
NIST-JANAF Thermochemical Tables [98CHA], 25.987 J·K–1·mol–1, and also with the
[77BAR/KNA] compilation, 26.09 J·K–1·mol–1. The uncertainty assigned to the selected
value of C pο,m (Ni, cr, 298.15 K) was calculated using the experimental data of
[36BRO/WIL] and [52BUS/GIA] obtained from measurements in the vicinity of 298 K.
The temperature coefficients of the heat capacity function for Ni derived for
the present review also allow calculation of heat capacities which coincide very well
with the experimental data in the temperature range 640 < T/K < 1728, as shown in
Figure V-1.
Table V-1: The temperature coefficients of the heat capacity function for Ni(cr):
C pο,m (T) = a + bT + cT 2 + eT –2. The notation E±nn indicates the power of 10, units for
C pο,m are J·K–1·mol–1. The sources of experimental data used for the determination of
these coefficients are given in Figure V-1.
Tmin/K
Tmax/K
3.36472 E+01
a
– 2.47530E–02
b
4.35458E–05
c
– 3.61955E+05
e
298
450
– 4.93235E+01
1.676480E–01
– 7.48518E–05
3.76214E+06
450
600
1.05023E+05
– 2.29221E+02
1.40772E–01
– 6.52809E+09
5.29406E+04
– 1.65828E+02
1.29936E–01
1.82638E+02
– 4.20407E–01
1.23677E+01
2.22800E–02
600
631
0
631
640
2.90521E– 04
0
640
690
4.45300E– 06
2.46766E+06
690
1728
V.1.3 Nickel liquid
The enthalpy of formation of Ni(l) can be calculated from that of Ni(cr) by adding the
enthalpy of fusion and the difference in enthalpy, ( H m (1728 K) – H mο (298.15 K)) between the solid and liquid phase. The value
∆ f H mο (Ni, l, 298.15 K) = (17.50 ± 0.25) kJ·mol–1
given by [98CHA] is essentially the same as that recommended by the present review.
V.1 Elemental nickel
79
The enthalpy of fusion ∆ fus H mο = (17.15 ± 0.42) kJ·mol–1, used for the calculation of the enthalpy of formation, is based on the calorimetric study of Vollmer et al.
[66VOL/KOH] ∆ fus H mο = (16.90 ± 0.25) kJ·mol–1, and Geoffray et al. [63GEO/FER]
∆ fus H mο = (17.47 ± 0.23) kJ·mol–1. An another reported value, (18.43 ± 0.27) kJ·mol–1
obtained by the electrical explosion study i.e., by rapid heating using a current of high
density [71LEB/SAV] was significantly different and, therefore, disregarded. Another
study performed using a similar experimental technique i.e., fast pulse heating gave
∆ fus H mο = (18.02 ± 0.54) kJ·mol–1 [93OBE/KAS]. Although this value is higher than the
calorimetric one, the respective error limits overlap.
The enthalpy of fusion retained by this review is identical with that given in the
NIST-JANAF Tables [98CHA]. The compilations [73HUL/DES] and [77BAR/KNA]
adopted the value determined by [63GEO/FER].
The heat capacity of nickel in the liquid region (1728 – 1822 K) was measured
by Vollmer et al. [66VOL/KOH] using an adiabatic high temperature calorimeter with
an inert gas atmosphere, C pο,m (Ni, l) = 38.99 J·K–1·mol–1 was obtained. Another study
yielding the liquid phase heat capacity is that of Geoffray et al. [63GEO/FER],
C pο,m (Ni, l) = 43.10 J·K–1·mol–1. In this review, the constant value independent of the
temperature, C pο,m (Ni, l) = (41.1 ± 2.1) J·K–1·mol–1 is based on these two studies. The
uncertainty was calculated assuming that the uncertainty in both studies was 2%, i.e.,
the same as reported by [66VOL/KOH]. The thermochemical tables [98CHA] and
[76MAH/PAN] adopted the heat capacity of nickel in the liquid phase, 38.911
J·K−1·mol–1, and 38.91 J·K–1·mol–1, respectively which is close to the one measured by
[66VOL/KOH].
The entropy at 298.15 K can be calculated in a manner analogous to the one
used for the enthalpy of formation of nickel in the liquid region. The value selected by
this review:
S mο (Ni, l, 298.15 K) = (41.49 ± 0.30) J·K–1·mol–1
is almost identical with the one chosen by NIST-JANAF Thermochemical Tables
[98CHA].
The selections above yield:
∆ f Gmο (Ni, l, 298.15 K) = (14.04 ± 0.27) kJ·mol–1.
V.2 Simple nickel aqua ions
V.2.1 Ni2+
In every review of thermodynamic data care should be taken that the same and simultaneously the most reliable auxiliary quantities are employed. Consequently CODATA
values (see NEA-TDB auxiliary data) were taken whenever they had been available.
The standard enthalpy of formation and the partial molar entropy of the bivalent nickel
80
V.2 Simple nickel aqua ions
ion have been re-evaluated in the present review. The recommendations of recent reviews by Plyasunova et al. [98PLY/ZHA] as well as Archer [99ARC] (see Sections
V.2.1.2 and V.2.1.3) have been taken into account.
V.2.1.1 Gibbs energy of formation of Ni2+
In principle the standard Gibbs energy of formation of Ni2+ can be obtained directly
from potentiometric data. Plyasunova et al. [98PLY/ZHA] maintain, however, that no
accurate value for the standard electrode potential of Ni2+ | Ni according to:
Ni2+ + H2(g) U Ni(cr) + 2H+
(V.1)
can be produced by emf measurements, because the truly reversible equilibrium appears
to be unattainable. Thus, in Plyasunova et al.’s review the results given by Murata
[28MUR], Haring and Vanden Bosche [29HAR/BOS], Colombier [34COL2], as well as
Carr and Bonilla [52CAR/BON] were simply averaged and taken as the result of potentiometric data on the Ni2+ | Ni couple. The present reviewers, however, re-evaluated
these and other data [62LOP], [62VAG/UVA], see Appendix A.
From the point of view of thermodynamics the standard electrode potential of
Ni2+ | Ni can be obtained most accurately from a cell without a liquid junction, such as
that used by Haring and Vanden Bosche [29HAR/BOS] and discussed thoroughly by
Archer [99ARC]. The former authors measured potentials of cells of the type:
Ni | NiSO4(m) | Hg2SO4 | Hg
at 25°C for 0.05 < m < 0.16 mol·kg–1. The recalculated mean value of E° (SIT model
[97GRE/PLY2]) of the 26 cells measured was (0.8500 ± 0.0025) V, see Figure V-2.
Combined with the standard potential of the Hg2SO4 | Hg electrode
((0.6128 ± 0.0031) V, [89COX/WAG]) a value of E°(Ni2+ | Ni, 298.15 K) =
− (0.2372 ± 0.0040) V has been obtained which agrees almost perfectly with the NBS
tables (– 0.236 V) [82WAG/EVA], [89BRA].
Attempts to measure the standard electrode potential of nickel using the similar
cell:
Ni | NiCl2, solvent | Hg2Cl2 | Hg
failed when the solvent was water [29HAR/BOS], see also [61TAN2].
For the selected standard Gibbs energy of formation of Ni2+ an uncertainty of
± 0.77 kJ·mol–1 has to be taken into account,
∆ f Gmο (Ni2+, 298.15 K) = – (45.77 ± 0.77) kJ·mol–1.
V.2 Simple nickel aqua ions
81
Figure V-2: Standard electrode potential of Ni | NiSO4 | Hg2SO4 | Hg at 25°C. (Thick
solid curve: calculated according to [29HAR/BOS]; thin solid curves: corresponding
error limits; ○ experimental data [29HAR/BOS]; thick dash curve: calculated using previously selected thermodynamic quantities [82WAG/EVA], [89COX/WAG]).
0.972
0.968
E/V
0.964
0.960
0.956
▬▬ :
0.952
0.04
: E/V= =0.8500
0.858–−(RT/F)
(RT/F)lnlnaaNiSO
E/V
NiSO4
4
0.06
0.08
0.10
0.12
0.14
−1
m (NiSO4) / mol·kg
0.16
0.18
82
V.2.1.1.1
V.2 Simple nickel aqua ions
Temperature coefficient of the standard electrode potential Ni2+ | Ni
The formation reaction of Ni2+ is the reverse of the reaction of Equation (V.1):
Ni(cr) + 2H+ U Ni2+ + H2(g),
(V.2)
and the corresponding standard entropy may be calculated by:
∆ f Smο (Ni2+, 298.15 K) = – n F (d E°(Ni2+ | Ni)/d T)298.
Bratsch [89BRA] recommended the value (d E°/d T)298 = (0.146 ± 0.010)
mV·K–1 which leads to ∆ f Smο (Ni2+, 298.15 K) = – (28 ± 2) J·K–1·mol–1. According to
Archer [99ARC] this value can be traced to the Ph.D. thesis of Muldrow [58MUL], who
measured the enthalpy of solution of NiCl2(cr) in water and obtained ∆ f H mο (Ni2+,
298.15 K) = − (54.040 ± 0.800) kJ·mol–1. Combined with ∆ f Gmο (Ni2+, 298.15 K) from
the NBS Tables [82WAG/EVA], (d E°(Ni2+ | Ni)/d T)298 = 0.1465 mV·K–1 follows
[58MUL]. When Ko and Hepler [63KO/HEP] derived ∆ f H mο (Ni2+, 298.15 K) = – 53.14
kJ·mol–1 by combination of the enthalpies of formation and solution in water of
NiCl2(cr) with the enthalpy of formation of Cl– and ∆ f Gmο (Ni2+, 298.15 K) from Carr
and Bonilla [52CAR/BON], almost the same value of (d E°(Ni2+ | Ni)/d T)298 was obtained. From Reaction (V.2) with NEA-TDB auxiliary values and the data for Ni(cr),
optimised in this review, the partial molar entropy of the nickel ion can then be calculated to be: S mο (Ni2+, 298.15 K) = – (128.9 ± 3.0) J·K–1·mol–1. This result necessarily
agrees with the NBS value [82WAG/EVA]. It must be emphasised that
(d E°(Ni2+ | Ni)/d T)298 [89BRA] was obtained from previously determined entropies
and not independently by investigating the temperature dependence of E°(Ni2+ | Ni), as
might be concluded from the review of Plyasunova et al. [98PLY/ZHA]. Consequently
the selection of S mο (Ni2+, 298.15 K) has been deferred to Section V.2.1.3.
V.2.1.2 Enthalpy of formation of Ni2+
A straightforward calculation of ∆ f H mο (Ni2+, 298.15 K) can be based on Reactions
(V.3), (V.4), (V.5), see Equations (V.6), (V.7), (V.8):
NiCl2(cr) U Ni2+ + 2Cl–
2+
NiBr2(cr) U Ni + 2Br
NiI2(cr) U Ni2+ + 2I–.
–
(V.3)
(V.4)
(V.5)
Thomsen [1883THO] measured the enthalpy of dissolution of NiCl2(cr) in water as early as the 19th century, however, the data reported in [58MUL], [90EFI/FUR]
and [95MAN/KOR] have been evaluated, see Table V-2. Other data were considered to
be of lower quality [70EFI/KUD], [75BAR/MAS], [75WEE/KOE]. As NiCl2(cr) is very
hygroscopic, a systematic error may be introduced when moisture is not excluded effectively. It was not possible to decide which of the three results chosen is the most reliable. Consequently the unweighted average was calculated, and the uncertainty (2σ)
V.2 Simple nickel aqua ions
83
assigned covers the whole range of expectancy, thus ∆ sol H mο (V.3) = − (83.9 ± 1.7)
kJ·mol–1 at 298.15 K. For the calculation of ∆ f H mο (Ni2+) according to Equation (V.6),
∆ f H mο (NiCl2, cr, 298.15 K) = – (304.90 ± 0.11) kJ·mol–1 has been selected in this review and ∆ f H mο (Cl–) = – (167.08 ± 0.10) kJ·mol–1 has been taken from CODATA
[89COX/WAG] as adopted by the NEA-TDB reviews:
∆ f H mο (Ni2+) = ∆ sol H mο (V.3) – 2 ∆ f H mο (Cl–) + ∆ f H mο (NiCl2, cr).
(V.6)
This results in ∆ f H mο (Ni2+) = – (54.64 ± 1.71) kJ·mol–1.
Table V-2: Enthalpy of Reaction (V.3)
− ∆ sol H mο / kJ ⋅ mol−1
References
(83.41 ± 1.40)
[58MUL]
(84.89 ± 0.27)
[90EFI/FUR]
(83.48 ± 0.57)
[95MAN/KOR]
Efimov and Furkalyuk [90EFI/FUR] determined ∆ f H mο (Ni2+) =
− (55.18 ± 0.58), − (55.27 ± 0.50), – (56.22 ± 0.42) kJ·mol–1 with Equations (V.6), (V.7)
and (V.8), using the enthalpies of nickel halide formation determined by [88EFI/EVD],
[89EVD/EFI] and [90EFI/EVD]. In the present review the selected enthalpy of formation values for ∆ f H mο (NiBr2, cr) and ∆ f H mο (NiI2, cr), – (213.5 ± 2.4) kJ·mol-1 and
− (96.42 ± 0.84) kJ·mol–1 are used with the enthalpies of formation for the halide ions
∆ f H mο (Br–) = – (121.41 ± 0.15) kJ·mol–1 and ∆ f H mο (I–) = − (56.78 ± 0.05) kJ·mol–1
taken from CODATA [89COX/WAG] as adopted by the NEA-TDB reviews:
∆ f H mο (Ni2+) = ∆ sol H mο (V.4) – 2 ∆ f H mο (Br–) + ∆ f H mο (NiBr2, cr)
ο
m
ο
m
2+
ο
m
ο
m
–
∆ f H (Ni ) = ∆ sol H (V.5) – 2 ∆ f H (I ) + ∆ f H (NiI2, cr).
ο
m
2+
(V.7)
(V.8)
–1
This results in ∆ f H (Ni ) = – (57.4 ± 2.4) kJ·mol from Equation (V.7) and
∆ f H (Ni2+) = – (56.2 ± 0.9) kJ·mol–1 from Equation (V.8). Together with ∆ f H mο (Ni2+)
= – 52.10, – 51.85 kJ·mol–1 determined by Vasil’ev et al. [84VAS/VAS],
[86VAS/DMI], there are five independent values which, giving equal weight to each
one, result in a mean value of ∆ f H mο (Ni2+, 298.15 K) = – (54.4 ± 4.9) kJ·mol–1.
ο
m
This value agrees with the one selected by Plyasunova et al. (– (54.1 ± 2.5)
kJ·mol–1) from essentially the same sources.
Archer [99ARC] cited values taken, partly from the same, partly from different
sources, resulting in – (54.2 ± 2.1) kJ·mol–1 which again agrees within the error limits
with the earlier assessments [84VAS/VAS], [86VAS/DMI], [90EFI/FUR],
[98PLY/ZHA]. As the uncertainty of the calorimetric determinations of ∆ f H mο (Ni2+,
298.15 K) is rather high (± 4.9 kJ·mol–1), the enthalpy of Reaction (V.2) has been calculated from the Gibbs energy using the definition in Equation (V.9). For the calculation
V.2 Simple nickel aqua ions
84
of ∆ f Smο (Ni2+, 298.15 K) = ∆ r Smο (V.2) the selected values of S mο (Ni, cr, 298.15 K),
S mο (Ni2+, 298.15 K), see Sections V.1.2 and V.2.1.3, as well as S mο (H2, g, 298.15 K)
from the selected auxiliary data of Table IV-1 were used:
H=G+TS
ο
m
ο
m
2+
(V.9)
2+
ο
m
∆ f H (Ni , 298.15 K) = ∆ f G (Ni , 298.15 K) + 298.15· ∆ r S (V.2)
∆ f H mο (Ni2+, 298.15 K) = (– (45.77 ± 0.77) + 0.29815·(– (30.99 ± 1.41))) kJ·mol–1
∆ f H mο (Ni2+, 298.15 K) = – (55.01 ± 0.88) kJ·mol–1.
This value is similar to the result obtained by the ∆ sol H mο measurements, but
the uncertainty is much less.
V.2.1.3 Partial molar entropy of Ni2+
V.2.1.3.1
NiSO4·7H2O(cr) cycle
Dissolution:
NiSO4·7H2O(cr) U Ni2+ + SO 24 − + 7H2O(l)
(V.10)
The solubility of NiSO4·7H2O(cr) at 25°C has been taken from the authors recommended by Linke and Seidell on p. 1219 [65LIN/SEI]. The mean value and error
limits have been estimated by averaging the six different solubilities listed at 25°C
[23VIL/BEN], [24TAN], [39ROH]:
40.6 g (NiSO4)/100g (H2O)
which yields:
m (sat) = (2.62 ± 0.05) mol·kg–1 (NiSO4·7H2O(cr)).
The value of ln γ ± (sat) = – (3.326 ± 0.007) has been derived from the data of
Robinson and Stokes [59ROB/STO]. The water activity can be calculated according to:
ln aw = – [M (H2O) · m(NiSO4) · 2 ·φ ] / 1000
(V.11)
where φ is the osmotic coefficient also listed by Robinson and Stokes [59ROB/STO].
The value of ln aw (sat) = – (0.0707 ± 0.0030) has been derived from those data. The
error limits were estimated by inserting the higher and lower values of both m and φ into
Equation (V.11).
The standard Gibbs energy of Reaction (V.10) can be calculated according to
Equation (V.12), as was shown for example by Larson et al. [68LAR/CER]:
∆ sol Gmο (V.10) = – RT{2·[ln m (sat) + ln γ ± (sat)] + 7·ln aw (sat)}
(V.12)
which leads to the selection:
∆ sol Gmο (V.10) = (12.940 ± 0.110) kJ·mol–1.
Stout et al. [66STO/ARC] determined the standard enthalpy of Reaction (V.13):
NiSO4·7H2O(cr) U NiSO4·6H2O(α) + H2O(l)
(V.13)
V.2 Simple nickel aqua ions
85
to be ∆ r H mο (V.13) = (7.682 ± 0.105) kJ·mol–1 ((1836 ± 25) cal·mol–1).
This value can be combined with the recalculated standard enthalpy of Reaction (V.14) determined by Goldberg et al. [66GOL/RID]:
NiSO4·6H2O(α) U Ni2+ + SO 24 − + 6H2O(l)
(V.14)
∆ r H mο (V.14) = (4.485 ± 0.200) kJ·mol–1
to result in:
∆ sol H mο (V.10) = (12.167 ± 0.226) kJ·mol–1.
The entropy of Reaction (V.10) has been calculated by rearranging Equation
(V.9):
∆ sol S mο (V.10) = { ∆ sol H mο (V.10) – ∆ sol Gmο (V.10)}/298.15
∆ sol S mο (V.10) = – (2.59 ± 0.84) J·K–1·mol–1.
The selected value for partial molar entropy of Ni2+ follows from Equation
(V.15):
S mο (Ni2+) = S mο (NiSO4·7H2O, cr) – S mο ( SO 24 − ) – 7 S mο (H2O, l) + ∆ sol S mο (V.10) (V.15)
S mο (Ni2+, 298.15 K) = – (131.8 ± 1.4) J·K–1·mol–1
where S mο (NiSO4·7H2O, cr) = (378.95 ± 1.00) J·K–1·mol–1 as assessed in the present
review (see Section V.5.1.2.2.1) and the values for S mο ( SO 24 − ) and S mο (H2O, l) are
auxiliary values from Table IV-1.
V.2.1.3.2
NiCl2·6H2O(cr) cycle
Another possible cycle for determining the partial molar entropy of Ni2+ is based on the
dissolution reaction:
NiCl2·6H2O(cr) U Ni2+ + 2Cl– + 6H2O(l)
(V.16)
Rard [92RAR2] calculated a value of ∆ sol Gmο (V.16) = – (17.38 ± 0.04) kJ·mol–1
from his precise isopiestic vapour-pressure and solubility experiments. Thomsen
[06THO] measured ∆ r H m (V.17) = – (85.06 ± 1.80) kJ·mol–1 (– 20.33 kcal·mol–1) for
the reaction:
NiCl2(cr) + 6H2O(l) U NiCl2·6H2O(cr)
(V.17)
at 291 – 293 K. The uncertainty assigned in the present work to ∆ r H m (V.17) accounts
for the deviation of the actual measurements and correction to the standard temperature
(298.15 K) [06THO], [98PLY/ZHA]. As discussed in Section V.2.1.2, based on experimental heat of solution measurements, ∆ sol H mο (V.3) = − (83.9 ± 1.7) kJ·mol–1 at 298.15
K. From Reaction (V.3) and (V.17) ∆ sol H mο (V.16) = (1.2 ± 2.5) kJ·mol–1 follows. The
entropy of Reaction (V.16) can be calculated in a manner quite analogous to that used in
Section V.2.1.3.1:
V.2 Simple nickel aqua ions
86
∆ sol S mο (V.16) = { ∆ sol H mο (V.16) – ∆ sol Gmο (V.16)}/298.15
∆ sol S mο (V.16) = (62.3 ± 8.3) J·K–1·mol–1.
If a reliable independent value were available for S mο (NiCl2·6H2O, cr,
298.15 K), the partial molar entropy of Ni2+ would follow from Equation (V.18):
S mο (Ni2+) = S mο (NiCl2·6H2O, cr) – 2 S mο (Cl–) – 6 S mο (H2O, l) + ∆ sol S mο (V.16) (V.18)
For example, if S mο (NiCl2·6H2O, cr) = 344.3 J·K–1·mol–1 from [63KO/HEP] is
used with auxiliary data for S mο (Cl–) and S mο (H2O, l) from Table IV-1,
S mο (Ni2+) = – (126.3 ± 8.3) J·K–1·mol–1.
However, in [63KO/HEP], the value of S mο (NiCl2·6H2O, cr) was taken from
unpublished heat capacity data, and the value accepted in the present review is not independent of the value selected for ∆ f H mο (Ni2+) (cf. Section V.4.1.3.3).
Although the entropy values of the NiSO4·7H2O(cr) and the NiCl2·6H2O(cr)
cycles agree within the error limits, the former has been preferred, because it is more
precise and has a more reliable experimental basis.
V.2.1.3.3
Entropy of Ni2+ from its ∆ f Gmο and ∆ f H mο
When the independently determined values for ∆ f Gmο (Ni2+, 298.15 K) =
− (45.77 ± 0.77) kJ·mol–1 (see Section V.2.1.1) and ∆ f H mο (Ni2+, 298.15 K) =
− (54.4 ± 4.9) kJ·mol–1 (see Section V.2.1.2) are employed, ∆ f Smο (Ni2+) can be calculated by Equation (V.19):
∆ f Smο (Ni2+) = [ ∆ f H mο (Ni2+) – ∆ f Gmο (Ni2+)]/298.15.
(V.19)
The partial molar entropy of Ni2+ follows from Equation (V.20)
S mο (Ni2+) = ∆ f Smο (Ni2+) – S mο (H2, g) + S mο (Ni, cr)
(V.20)
S mο (Ni2+) = – (131.4 ± 16.6) J·K–1·mol–1.
A comparison of Equations (V.15), (V.18), (V.20) shows that three completely
independent paths have led to S mο (Ni2+) values that agree with each other within the
experimental uncertainty. The result of the Section V.2.1.3.1, however, is the most precise one. Plyasunova et al. [98PLY/ZHA] selected S mο (Ni2+, 298.15 K) = − (130 ± 3)
J·K−1·mol–1, which is very close to the finally selected value given in Equation (V.15):
S mο (Ni2+, 298.15 K) = – (131.8 ± 1.4) J·K–1·mol–1.
V.2.1.4 Heat capacity of Ni2+
Spitzer et al. [78SPI/SIN], Perron et al. [81PER/ROU], and Smith-Magowan et al.
[82SMI/WOO] reported apparent molar heat capacity values for NiCl2(aq); values for
Ni(ClO4)2(aq) were reported by Spitzer et al. [78SPI/SIN2] and Pan and Campbell
[97PAN/CAM]. There are also literature values for Ni(NO3)2(aq) [79SPI/OLO],
V.2 Simple nickel aqua ions
87
NiBr2(aq) [79VOR/VAS] and NiSO4(aq) [85DRA/MAD]. Values of C pο,m (Ni2+, T), relative to the standard state C pο,m (H+, T) = 0 J·K–1·mol–1 can be determined if the apparent
molar heat capacities for the salts NiX2(aq, T) are extrapolated to I = 0, and values for
the partial molar heat capacities of the corresponding acids, C pο,m (HX, aq, T), are
known. The measurements of Smith-Magowan et al. [82SMI/WOO] (50 to 300°C) were
limited at each temperature, were carried out at a relatively high pressure (17.7 MPa),
and are not useful in the selection of a value for C pο,m (Ni2+, 298.15 K). Similarly, the
measurements of Drakin et al. [85DRA/MAD] cannot be used because they were done
at higher temperatures (50 to 90°C) and concentrations.
The procedures for an extrapolation of the electrolyte apparent molar heat capacities to I = 0 differ considerably in the different references. For example, the apparent molar heat capacities for Ni(ClO4)2(aq) found by Spitzer et al. [78SPI/SIN2] are
approximately 3 J·K–1·mol–1 less negative than those found by Pan and Campbell [97PAN/CAM]. Yet, the reported partial molar heat capacities for 25°C differ by 6
J·K–1·mol–1. Apparent molar heat capacities for low molalities (< 0.07 m) are needed for
a reasonable extrapolation to I = 0. Unfortunately, uncertainties in the apparent molar
heat capacities increase with decreasing electrolyte concentration (small uncertainties in
the solute concentrations become significant). This is especially evident in the lowconcentration results of Perron et al. [81PER/ROU], which are so badly scattered that
the values as extrapolated to I = 0 must be considered unreliable.
The same problems in extrapolation are found to a somewhat lesser degree in
calculations of values of the partial molar heat capacities for the (1:1) aqueous acids.
Reanalysis and comparison of the apparent molar heat capacity data for inorganic acids
is not within the scope of the present review. Here, the partial molar heat capacity values C pο,m (HNO3, aq, 298.15 K) = – 71.0 J·K–1·mol–1 from Patterson and Woolley
[2002PAT/WOO], C pο,m (HCl, aq, 298.15 K) = – 123.2 J·K–1·mol–1 [2002PAT/WOO],
and C pο,m (HClO4, aq, 298.15 K) = – 26.3 J·K–1·mol–1 from Oakes and
Rai [2001OAK/RAI] (based on the experimental work of Hovey and Hepler
[89HOV/HEP] and Lemire and Campbell [96LEM/CAM]) have been retained. No correction has been applied for values obtained at different pressures to 0.6 MPa, as these
corrections are markedly less than other uncertainties in the measurements.
The unweighted average of the values for C pο,m (Ni2+, 298.15 K) in Table V-3
(neglecting the discrepant value from Perron et al. [81PER/ROU]) is – 46.1 J·K–1·mol–1
(relative to C pο,m (H+) = 0 J·K–1·mol–1), which is very close to the value – 47.0
J·K−1·mol−1 from the most extensive set of experimental data [97PAN/CAM]. The selected value is:
C pο,m (Ni2+, 298.15 K) = – (46.1 ± 7.5) J·K–1·mol–1.
The unweighted average has been used, and the uncertainty of 7.5 J·K–1·mol–1
has been selected, because most of the uncertainty in the value of the partial molar heat
V.2 Simple nickel aqua ions
88
capacity arises from systematic experimental errors and the method of extrapolation to
infinite dilution.
Over the temperature range of the measurements, the value of C pο,m (Ni2+)
shows only a very shallow maximum [97PAN/CAM]. Within the uncertainties of the
measurements, the value for 298.15 K can be used over the temperature range between
298.15 and 358.15 K. The values of the apparent molar heat capacities for nickel salts
depend strongly on the electrolyte concentrations [97PAN/CAM]1.
Table V-3: Partial molar heat capacity values for Ni(II) salts in aqueous solution at
298.15 K.
C οp ,m (Ni-Salt, aq, 298.15 K)/
J·K–1·mol–1
salt
Ni(ClO4)2(aq)
C οp ,m (Ni2+, 298.15 K)/
J·K–1·mol–1
reference
– 99.6
– 47.0
[97PAN/CAM]
– 93.1
– 40.5
[78SPI/SIN]
NiCl2(aq)
– 294.0
– 306
– 47.6
– 59.6(a)
[78SPI/SIN]
Ni(NO3)2(aq)
– 186.5
– 44.5
[79SPI/OLO]
[81PER/ROU]
NiBr2
– 310
– 51(b)
[79VOR/VAS]
(a) The apparent molar heat capacities of Perron et al. [81PER/ROU] for low concentrations are too badly
scattered to give a reliable value for C οp ,m (Ni2+, 298.15 K). A second value for C οp ,m (NiCl2, aq, 298.15 K)
from the same paper (– 295.3 J·K–1·mol–1) is based primarily on higher concentration values.
(b) Calculated using a value for C οp ,m (Br–, 298.15 K) based on the value for HCl(aq) [2002PAT/WOO] and
C οp ,m (Br–, 298.15 K) – C οp ,m (Cl–, 298.15 K) = 6 J·K–1·mol–1 [96HEP/HOV], [96CRI/MIL].
V.2.2 Other oxidation states
Nickel(III) and nickel(IV) complexes are well documented, but have not been treated in
this review. Whereas a ‘transient’ Ni3+ species presumably exists, there is no aqua ion
representative of Ni(IV).
V.2.2.1 Ni3+
Although the trivalent oxidation state of Ni can be stabilised by certain arrangements of
donor ligands, no stable aqua ion of Ni(III) appears to exist in dilute aqueous solutions
at ambient conditions [97RIC]. Vogel [68VOG] found strong evidence that a redox
couple composed of tri- and divalent nickel can exist in potassium hydroxide melts containing a mass fraction of water w(H2O) = 0.15 at 498 and 523 K. The standard potential
measured, E°(Ni3+ | Ni2+) = 1.168 and 1.174 V at 498 and 523 K, respectively, cannot be
compared with the value estimated for 298 K in predominantly aqueous solutions.
1
After this review was completed, results from an extensive experimental study (from 278.15 to 393.15 K
at 0.35 MPa) were reported [2004BRO/MER] for aqueous Ni(NO3)2 solutions. The new apparent molar
heat capacity values are markedly more negative than those reported by Spitzer et al. [79SPI/OLO].
V.2 Simple nickel aqua ions
89
Bhattacharya et al. [86BHA/MUK] synthesised the tris(2,2'-bipyridine 1,1'dioxide) Ni(III) complex, Ni(bpyO 2 )33+ , and measured the potential of Reaction (V.21)
in acetonitrile against the standard calomel electrode (SCE).
Ni(bpyO 2 )33+ + e– U Ni(bpyO 2 )32 +
(V.21)
These authors observed a linear correlation between the M3+ | M2+ and
M(bpyO 2 )33+ | M(bpyO 2 )32 + potentials (M being Cr, Fe, Mn, Co) versus the SCE in
aqueous and acetonitrile solutions, respectively. A correlation of similar quality is obtained when the standard electrode potentials of M3+ | M2+ [89BRA] are linearly fitted
against the potentials measured by [86BHA/MUK], see Figure V-3. As the potential of
Ni(bpyO 2 )33+ | Ni(bpyO 2 )32 + can be measured, the experimentally inaccessible standard
electrode potential of Ni3+ | Ni2+ can be estimated as E°(Ni3+ | Ni2+, 298.15 K) =
(2.56 ± 0.27) V. Twice the standard deviation of the regression analysis has been taken
as an uncertainty limit. With this result and ∆ f Gmο (Ni2+, 298.15 K) = − (45.77 ± 0.77)
kJ·mol–1 (see Section V.2.1.1) the Gibbs energy of formation of Ni3+ has been roughly
estimated to be ∆ f Gmο (Ni3+, 298.15 K) = (202 ± 26) kJ·mol–1. Although this result is of
general interest, it is not sufficiently well defined to be a selected value in this review.
Figure V-3: Estimation of the Ni3+ | Ni2+ standard electrode potential. Solid straight line:
least square fit of SHE data; dash straight line: least square fit of SCE data; ▼ SHE poο
tentials; ▲ SCE potentials ; V extrapolated value of E298
(Ni3+ | Ni2+, SHE); U extrapoο
3+
2+
lated value of E298 (Ni | Ni , SCE).
3.0
2.5
E°298 / V, M3+ ‌ M2+
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.2
-0.8
-0.4
0.0
0.4
0.8
E°298 / V(SCE), M(bpyO 2 )33+ ‌ M(bpyO 2 )32 +
1.2
1.6
90
V.3 Oxygen and hydrogen compounds and complexes
V.3 Oxygen and hydrogen compounds and complexes
V.3.1 Aqueous nickel(II) hydroxo complexes
V.3.1.1 Hydroxo complexes in acidic or neutral solutions
The Ni(II) ion is a rather weak acid in aqueous solution. The hydrolysis reactions of
Ni(II) may be defined by the following generalised equilibrium process:
mNi2+ + nH2O(l) U Ni m (OH) n2 m − n + nH+;
*
b n,m
(V.22)
The formation of five water soluble hydroxo complexes is generally recognised: NiOH+, Ni(OH)2(aq), Ni(OH)3− , Ni2OH3+, Ni 4 (OH) 44 + .
Several experimental difficulties may arise during the study of the hydrolysis
of Ni(II): (i) nickel forms a relatively insoluble hydroxide, i.e., the hydroxide or basic
salt precipitates even in solutions with a low degree of hydrolysis, and therefore the
solutions have a low buffer capacity; (ii) because of the formation of oligomer hydroxo
complexes, the mole fractions of different hydroxo species depend strongly on the experimental solution concentrations; (iii) the kinetics of formation of oligomer species is
relatively slow. The first mentioned problem can be avoided by using concentrated solutions, but this introduces complications (ii) and (iii). Consequently, the choice of experimental methods and conditions is a critical factor in determining the accuracy of the
hydrolysis constant values. The published data often refer to high ionic strengths and
concentrations of Ni(II), while parallel measurements at different temperatures are usually missing. For these reasons, a limited number of sets of serviceable experimental
data are available to evaluate the thermodynamic properties of Ni(II) hydroxo complexes.
Only a few compilations and reviews have been published on the thermodynamic data of Ni(II) hydrolysis. Baes and Mesmer [76BAE/MES] have critically reviewed the available experimental data, and provided recommended values for I = 0,
0.1, 1.0 and 3.0 mol·kg–1. A slightly different set of stability constants has been recommended by Smith and Martell [76SMI/MAR] in their compilation. In a recent review of
Plyasunova et al. [98PLY/ZHA], critical evaluations, and in some cases re-evaluations,
of literature data have been carried out, using the SIT approach. The values selected by
these authors are given in Appendix A.
The hydrolysis of Ni(II) has mainly been studied by potentiometric titrations,
although some solubility, kinetic and calorimetric studies have also been published.
V.3 Oxygen and hydrogen compounds and complexes
91
In all cases when sufficient experimental data were reported, a re-evaluation of
data was performed for the purposes of this review (see discussions in Appendix A).
Such re-assessments were required for several reasons:
(i)
Several background electrolytes ((Na)ClO4, (Li)ClO4, (Na)Cl, K(Cl) and
(K)NO3) have been used to keep the anion concentration constant in experimental solutions, and thereby to limit changes in activity coefficients. Generally, the
cations of these background electrolytes were singly-charged, and in most studies varying and relatively high Ni(II) concentrations were used (up to 1.5 M). In
such cases, multiply-charged Ni(II) species were a substantial fraction of the total cationic species in the solutions. In a few cases polynuclear Ni hydroxo complexes even predominated. Consequently, the constancy of the activity coefficients is questionable due to both the medium effect and the changing ionic
strength. Moreover, at high chloride or nitrate concentrations, the complexation
of these anions with Ni(II) cannot be ignored. Therefore, a re-evaluation of data
was performed to minimise the influence of the medium effects (often using only
a portion of the reported data) and to take into account the presence of NiX+ (X–
= Cl– or NO3− ) complexes.
(ii)
In publications that appeared before 1965, the experimental data were interpreted
in terms of formation only of the mononuclear NiOH+ complex. However, after
the work of Sillén and his co-workers [65BUR/LIL], the formation of
Ni 4 (OH) 4+
as the predominant hydrolytic species at [Ni2+]tot > 1 mM, became
4
widely accepted. Consequently, as pointed out in [98PLY/ZHA], almost all previously reported values should be corrected. In such cases, re-evaluation was
performed including oligonuclear species, especially Ni 4 (OH) 4+
4 .
(iii)
In addition, although the experimental methods (e.g., pH-measurements) were
already well established in the 1950s, the evaluation of experimental data has
developed considerably since that time (in a great part of the cited works, graphical evaluation was used). Therefore, re-assessment of data was performed to
check (and if necessary complete) the proposed equilibrium model.
The formation of two main species is generally recognised in this pH region:
NiOH , and Ni 4 (OH)44 + . Due to the presence of oligonuclear species, the onset of the
hydrolysis reaction of Ni(II) is strongly concentration dependent. Under natural conditions above pH = 6.9, and at low Ni(II) concentrations (< 0.0001 M), only the dominant
aqua complex and the species NiOH+ may be significant. In more concentrated solutions ([Ni2+]tot > 0.005 M), however, the tetramer Ni 4 (OH) 44 + is the major hydroxo
complex before precipitation occurs. The formation of a minor species Ni2OH3+ has
been suggested by several authors [65BUR/LIL], [65BUR/LIL2], [66BUR/IVA],
[71BUR/ZIN], [71OHT/BIE], [78BUR/KAM], because by adding this complex to the
model a better fit of the data could be achieved. Some other species were also postulated
by different authors ( Ni 2 (OH) 22 + [69KEN2], Ni 2 (OH)62 − [71BHA/SUB], Ni3 (OH)33+
+
92
V.3 Oxygen and hydrogen compounds and complexes
[66BUR/IVA]). However, the formation of these species has never been independently
confirmed. Actually the re-evaluation of data reported in [66BUR/IVA] indicated the
presence of Ni 4 (OH) 44 + rather than Ni3 (OH)33+ (see discussion in Appendix A).
The following equations describe the formation of the first mentioned three
species:
Ni2+ + H2O(l) U NiOH+ + H+
*
(V.23)
2Ni2+ + H2O(l) U Ni2OH3+ + H+
*
(V.24)
4Ni2+ + 4H2O(l) U Ni 4 (OH) 44 + + 4H+
*
(V.25)
b1,1
b1,2
b 4,4
The reported and re-calculated hydrolysis constants are listed in Table V-4.
As mentioned above, several early studies reported the formation of NiOH+ as
the only hydrolytic species, but the experiments were conducted at relatively high concentrations of Ni(II), where a considerable amount of Ni 4 (OH) 44 + species can be expected. To avoid its presence, the [Ni2+]tot should not exceed 0.001 M, and the pH
should not exceed 8.6 to preclude precipitation. The studies of Perrin [64PER] and of
Funahashi and Tanaka [69FUN/TAN] were the only ones in which conditions appropriate to the determination of a reliable formation constant for NiOH+ were used. The reevaluation of the experimental data reported in [63BOL/JAU], [65BUR/LIL],
[65BUR/LIL2], [71BUR/ZIN] resulted in acceptable log10 *b1,1 values. In a few cases, a
simple correction was applied to take into account the formation of the NiCl+ complex
[71OHT/BIE], [80MIL/BUG]. However, some of the reported values cannot be corrected, as insufficient experimental details or data have been communicated
[52CHA/COU], [52GAY/WOO], [53SCH/POL], [56CUT/KSA], [59ACH],
[63SHA/DES]. Tremaine and LeBlanc have reported a sharp increase of the first hydrolysis constant of Ni(II) between 150 and 300°C [80TRE/LEB2]. These data indicate
that the formation of NiOH+ is more pronounced at higher temperatures. The mole fraction of NiOH+ never exceeds 15% at 25°C, but reaches 70% at 300°C. Although the
reported data seem to be reliable, the temperature and pressure range used is well above
of the scope of this review. The authors’ extrapolation to the standard state resulted in
considerably different values from those determined at 25°C.
V.3 Oxygen and hydrogen compounds and complexes
93
Table V-4: Experimental equilibrium data (logarithmic values) for the hydrolysis of
Ni(II) in acidic or neutral solutions.
Method
Medium
Im
t /°C
log10 *b n ,m
reported(a)
log10 *b n ,m
retained(b)
Reference
Ni2+ + H2O(l) U NiOH+ + H+
gl
gl
gl
gl
gl
gl
KCl
NiCl2
NiCl2, NiSO4
Ni(ClO4)2
KClO4, Ni(NO3)2
(Na)ClO4
gl
(K)NO3
gl
gl
kin
kin
gl
gl
pH (sp)
sol
(Na)ClO4
(Na)Cl
NaClO4
?
(Na)Cl
(Na)Cl
(Na)ClO4
HCl, NaOH
gl
(K)Cl
sol
NaClO4
0.1
30
25
~ 0.08 25
→0
25
→0
25
25
0.25
25
1.05
30
1.05
40
1.05
50
1.05
→0
15
20
25
30
36
42
3.50
25
3.20
25
0.10
25
→0
?
3.20
25
3.20
60
0.82
25
→0
150
200
250
300
0.51
25
1.57
0.01
25
→0
– 9.4
– (10.64 ± 0.24)
–9
– 8.94
– (10.92 ± 0.12)
– (9.76 ± 0.05)
–
– (9.65 ± 0.04)
– (9.51 ± 0.04)
– (9.32 ± 0.06)
– (10.22 ± 0.04)
– (10.05 ± 0.04)
– (9.86 ± 0.03)
– (9.75 ± 0.07)
– (9.58 ± 0.06)
– (9.43 ± 0.03)
–
–
– (9.5 ± 0.1)
– (11.0 ± 0.2)
≤ – 10.5
–
– (8.93 ± 0.1)
– (8.05 ± 0.48)
– (6.95 ± 0.22)
– (6.03 ± 0.20)
– (5.23 ± 0.36)
– (9.87 ± 0.40)
– (10.06 ± 0.40)
– (8.35 ± 0.10)
–
–
–
–
–
– (9.84 ± 0.30)(c)
– (9.71 ± 0.30)(c)
– (9.72 ± 0.30)(c)
– (9.62 ± 0.30)(c)
– (9.66 ± 0.30)(c)
– (10.22 ± 0.20)(d)
– (10.07 ± 0.20)(d)
– (9.80 ± 0.20)(d)
– (9.74 ± 0.20)(d)
– (9.59 ± 0.20)(d)
– (9.38 ± 0.20)(d)
– (9.79 ± 0.12)(c)
– (9.53 ± 0.20)(c)
– (9.5 ± 0.2)
–
– (10.23 ± 0.20)(c)
– (8.45 ± 0.30)(c)
–
–
–
–
–
– (9.81 ± 0.50)(c)
– (9.93 ± 0.50)(c)
–
[52CHA/COU]
[52GAY/WOO]
[53SCH/POL]
[56CUT/KSA]
[59ACH]
[63BOL/JAU]
[64PER]
[65BUR/LIL]
[65BUR/LIL2]
[69FUN/TAN]
[71HOH/GEI]
[71OHT/BIE]
[71BUR/ZIN]
[75CLA/KEP]
[80TRE/LEB2]
[80MIL/BUG]
[97MAT/RAI]
(Continued on next page)
V.3 Oxygen and hydrogen compounds and complexes
94
Table V-4 (continued)
Method
Medium
Im
log10 *b n ,m
reported(a)
t /°C
log10 *b n ,m
retained(b)
Reference
2Ni2+ + H2O(l) U Ni2OH3+ + H+
gl
(Na)ClO4
3.50
25
– (10.0 ± 0.5)
– (9.8 ± 0.5)(c)
[65BUR/LIL]
gl
gl
gl
gl
gl
(Na)Cl
(Na)NO3
(Na)Cl
(Na)Cl
(Na)Br
3.20
3.30
3.20
3.20
3.30
25
25
25
60
25
– (9.3 ± 0.2)
– (9.6 ± 0.2)
– (10.5 ± 0.5)
– (8.5 ± 0.2)
– (9.5 ± 0.1)
– (8.95 ± 0.40)(c)
–
– (10.0 ± 0.6)(c)
– (8.58 ± 0.60)(c)
– (8.82 ± 0.40)(c)
[65BUR/LIL2]
[66BUR/IVA]
[71OHT/BIE]
[71BUR/ZIN]
[78BUR/KAM]
[63BOL/JAU]
4Ni2+ + 4H2O(l) U Ni 4 (OH) 44+ + 4H+
gl
(Na)ClO4
0.25
25
–
– (27.34 ± 0.30)(c)
1.05
25
–
– (27.66 ± 0.40)(c)
1.05
30
–
– (27.10 ± 0.30)(c)
1.05
40
–
– (26.39 ± 0.30)(c)
1.05
50
–
– (25.24 ± 0.30)(c)
gl
(Na)ClO4
3.50
25
– (27.37 ± 0.02)
– (27.36 ± 0.05)(c)
[65BUR/LIL]
gl
(Na)Cl
3.20
25
– (28.42 ± 0.05)
– (27.52 ± 0.20)(c)
[65BUR/LIL2]
gl
(Na)ClO4
1.61
25
– (27.03 ± 0.06)
– (27.00 ± 0.12)
[69KOL/KIL]
gl
(Na)Cl
3.20
25
– (28.55 ± 0.02)
– (27.57 ± 0.30)(c)
[71OHT/BIE]
gl
(Na)Cl
3.20
60
– (25.33 ± 0.02)
– (24.06 ± 0.30)(c)
[71BUR/ZIN]
gl
(Li)ClO4
3.48
25
– (27.32 ± 0.08)
– (27.25 ± 0.16)
[73KAW/OTS]
sp
0.8 M (Na)ClO4
0.82
25
– (27.00 ± 0.25)
– (26.99 ± 0.40)(c)
[75CLA/KEP]
gl
(Na)Br
3.3
25
– (28.18 ± 0.05)
– (27.15 ± 0.30)(c)
[78BUR/KAM]
gl
(K)Cl
0.51
25
– (28.04 ± 0.10)
–
[80MIL/BUG]
1.03
– (28.14 ± 0.06)
–
1.57
– (28.28 ± 0.04)
–
2.13
– (28.24 ± 0.04)
–
3.31
– (28.58 ± 0.06)
–
– (27.0 ± 0.1)
– (27.0 ± 0.3)
gl
NiCl2
→0
25
[91SPI/TRI]
(a) Reported values.
(b) Accepted values corrected to molal scale. The accepted values reported in Appendix A are expressed on
the molar or molal scales, depending on which units were used originally by the authors.
(c) Re-evaluated value, see Appendix A.
(d) Corrected to I = 0 by means of the SIT.
A single least squares SIT analysis was done using all the accepted (appropriately weighted) constants for 25°C in perchlorate and chloride media (Figure V-4 and
Figure V-5). The available data for chloride media are somewhat less consistent than
those for perchlorate media. This analysis results in the selected value of:
log10 *b1,1ο ((V.23), 298.15 K) = – (9.54 ± 0.14).
V.3 Oxygen and hydrogen compounds and complexes
95
In this analysis, the value recalculated from Perrin [64PER] for I = 0 and 25ºC
(weighted according to its estimated uncertainty) was also included (see Appendix A).
From the slopes in Figure V-4 and Figure V-5, ∆ε( ClO −4 ) = – (0.088 ± 0.061) kg·mol–1
and ∆ε(Cl–) = – (0.063 ± 0.072) kg·mol–1 are calculated. Using the NEA-TDB values of
0.37, 0.17, 0.14 and 0.12 kg·mol–1 for ε(Ni2+, ClO −4 ), ε(Ni2+,Cl–), ε(H, ClO −4 ) and
ε(H+,Cl–), respectively, leads to the selected values:
ε(NiOH+, ClO −4 ) = (0.14 ± 0.07) kg·mol–1
and
ε(NiOH+, Cl–) = – (0.01 ± 0.07) kg·mol–1 .
From the selected log10 *b1,1ο value, ∆ r Gmο ((V.23), 298.15 K) = (54.5 ± 0.8)
kJ·mol–1 can be derived. The Gibbs energy of formation of NiOH+ is calculated using
the selected values for Ni2+ and H2O(l).
∆ f Gmο (NiOH+, 298.15 K) = – (228.5 ± 1.1) kJ·mol–1.
When the discussion of Perrin [64PER] in Appendix A is also considered, the
ε(NiOH , X–) values (X = Cl–, NO3− , ClO −4 ) show a larger than expected spread – 0.01,
– 0.27, 0.14 kg·mol–1, though the value for ε(NiOH+, NO3− ) is not a selected value in the
present review.
+
The log10 *b1,1ο ((V.23), 298.15 K) value reported above is somewhat higher
than recommended by Baes and Mesmer (– (9.86 ± 0.03), [76BAE/MES]), but agrees
well with the one obtained by [98PLY/ZHA] viz. – (9.50 ± 0.36).
From the temperature dependence of the first hydrolysis constant determined
by Perrin [64PER], ∆ r H mο = (51.7 ± 3.4) kJ·mol–1 can be obtained. The re-evaluation of
the data reported in [68ARN] yielded ∆ r H m = (54.3 ± 2.0) kJ·mol–1 for 3.2 m NaCl medium. Combining the accepted hydrolysis constants of [65BUR/LIL2] and
[71BUR/ZIN], ∆ r H m = (59 ± 10) kJ·mol–1 can be calculated for the formation of the
species NiOH+ in 3.2 m NaCl medium. From the log10 *b1,1 values reported for 150 –
300°C in [80TRE/LEB2], ∆ r H m = (87 ± 15) kJ·mol–1 can be derived, but this value was
not considered in this review, as it is probably valid only in the temperature range covered by the measurements. The temperature variation of log10 *b1,1 values accepted in
[63BOL/JAU] is small compared to the uncertainties in the values. Consequently, no
enthalpy value has been derived from these constants. As the relatively high uncertainties of the remaining three data cannot be used to extract a reliable ionic strength dependence, it was assumed to be negligible. Assigning statistical weights to the data , the
value of:
∆ r H mο ((V.23), 298.15 K) = (53.8 ± 1.7) kJ·mol–1
is selected. The standard enthalpy of formation of NiOH+ is:
∆ f H mο (NiOH+, 298.15 K) = – (287.0 ± 1.9) kJ·mol–1.
V.3 Oxygen and hydrogen compounds and complexes
96
Figure V-4: Extrapolation to Im = 0 of the experimental data for Reaction (V.23) in
NaClO4 media.
-8.5
Ni2+ + H2O(l) U NiOH+ + H+
log10β1,1 + 2D - log10aw
-9.0
-9.5
-10.0
-10.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
−1
I / mol·kg
Figure V-5: SIT analysis of the experimental data for Reaction (V.23) in chloride media, including log10 *b1,1ο ((V.23), 298.15 K) obtained in NaClO4.
-8.5
Ni2+ + H2O(l) U NiOH+ + H+
log10β1,1 + 2D − log10aw
-9.0
-9.5
-10.0
-10.5
0
1
2
3
−1
I / mol · kg
V.3 Oxygen and hydrogen compounds and complexes
97
The dinuclear Ni2OH3+ species is always a minor component, and its formation
is considered to account for small deviations between the observed and calculated titration curves. Therefore, the rather wide scatter of the reported values is not surprising.
As stated in [76BAE/MES], the reported constants for this minor species should be considered as the upper limits for its stability, if indeed such a species is present. In some
cases, the accuracy of the reported experimental data was insufficient to take into account the formation of Ni2OH3+ during the re-evaluation [63BOL/JAU], [75CLA/KEP],
thus all accepted constants refer to high ionic strength. Therefore, the extrapolation to
Im = 0 was made assuming ε(Ni2OH3+, ClO −4 ) = (0.50 ± 0.15) kg·mol–1, based on the
estimated value for ε(Be2OH3+, ClO −4 ), see Table B-4, (ε(Ni2OH3+, ClO −4 ) =
(0.59 ± 0.16) kg·mol–1 is reported in [98PLY/ZHA], based on the corrected data from
[59ACH], [63BOL/JAU], [75CLA/KEP]). In this way, ∆ε((V.24), ClO −4 ) is estimated
to be − (0.1 ± 0.2) kg·mol–1, which leads to the selection:
ο
log10 *b1,2
((V.24), 298.15K) = – (10.6 ± 1.0).
From this value, ∆ r Gmο ((V.24), 298.15 K) = (60.5 ± 5.7) kJ·mol–1 and
∆ f Gmο (Ni2OH3+, 298.15 K) = – (268.2 ± 5.9) kJ·mol–1
can be derived.
Two reliable values are available concerning the enthalpy of reaction to form
Ni2OH3+. From the re-evaluation of the calorimetric data reported in [68ARN] ∆ r H m =
(45.9 ± 1.6) kJ·mol–1 can be calculated. Combining the accepted hydrolysis constants of
[65BUR/LIL2] and [71BUR/ZIN], ∆ r H m = (20 ± 20) kJ·mol–1 can be obtained. Both
values refer to 3.2 m NaCl medium. The more precise calorimetric value is accepted,
but with an increased uncertainty, and negligible dependence on the ionic strength is
assumed.
∆ r H mο ((V.24), 298.15 K) = (45.9 ± 6.0) kJ·mol–1
is selected in this review, which leads to:
∆ f H mο (Ni2OH3+, 298.15 K) = – (350.0 ± 6.3) kJ·mol–1.
At Ni(II) concentrations higher than 0.005 M, the Ni 4 (OH) 4+
complex is
4
dominant in the acidic pH region. Most of the accepted values concerning the formation
of the tetrameric hydroxo species have been re-evaluated [63BOL/JAU], [65BUR/LIL],
[65BUR/LIL2], [71BUR/ZIN], [75CLA/KEP], [78BUR/KAM] and corrected considering complexation with the background anion [65BUR/LIL2], [71BUR/ZIN],
[71OHT/BIE], [78BUR/KAM].
The SIT analysis of the accepted constants for perchlorate medium (Figure V6) resulted in the selected value of:
ο
log10 *b 4,4
((V.25), 298.15 K) = – (27.52 ± 0.15).
V.3 Oxygen and hydrogen compounds and complexes
98
From the slope in Figure V-6, ∆ε((V.25), ClO −4 ) = (0.16 ± 0.05) kg·mol–1 can
be calculated. Using the selected values for ε(Ni2+, ClO −4 ) and ε(H+, ClO −4 ), ∆ε((V.25)
ClO −4 ) leads to a value of ε( Ni 4 (OH) 44 + , ClO −4 ) = (1.08 ± 0.08) kg·mol–1.
In sodium chloride medium, only two values are accepted at 25°C
[65BUR/LIL2], [71OHT/BIE]; both are valid for Im = 3.2 mol·kg–1. Combining them
ο
with the above selected log10 *b 4,4
((V.25), 298.15 K), ∆ε((V.25), Cl–) = (0.23 ± 0.07)
–1
kg·mol can be derived, which leads to a value of ε( Ni 4 (OH) 44 + ,Cl–) = (0.43 ± 0.08)
kg·mol–1. A similar treatment, using the value accepted for NaBr medium, results in
∆ε((V.25), Br–) = (0.11 ± 0.07) kg·mol–1. Since no value is currently available for ε(H+,
Br–), only an estimate can be given for ε( Ni 4 (OH) 44 + ,Br–). Assuming ε(H+,Br–) =
ε(H+,Cl–), ε( Ni 4 (OH) 44 + ,Br–) = (0.71 ± 0.12) kg·mol–1 can be estimated.
Figure V-6: Extrapolation to Im = 0 of the experimental data for Reaction (V.25) in perchlorate media.
-26.0
4Ni2+ + 4H2O(l) U Ni 4 (OH) 44 + + 4H+
log10 β4,4 − 4D − 4 log 10 aw
-26.5
-27.0
-27.5
-28.0
-28.5
-29.0
-29.5
0.0
0.5
1.0
1.5
2.0
−1
I / mol·kg
2.5
3.0
3.5
V.3 Oxygen and hydrogen compounds and complexes
99
ο
From the selected log10 *b 4,4
value, ∆ r Gmο ((V.25), 298.15 K) = (157.1 ± 0.9)
kJ·mol can be obtained. The Gibbs energy of formation of Ni 4 (OH) 44 + is
–1
∆ f Gmο ( Ni 4 (OH) 44 + , 298.15 K) = – (974.6 ± 3.2) kJ·mol–1.
From the temperature dependence of log10 *b 4,4 values recalculated from the
data in [63BOL/JAU], ∆ r H m = (174 ± 15) kJ·mol–1 can be derived for 1.0 M NaClO4
medium (see Appendix A). The re-evaluation of the calorimetric data reported in
[68ARN] yielded ∆ r H m = (192.9 ± 1.0) kJ·mol–1 for 3.2 m NaCl medium (see Appendix
A). Combining the accepted hydrolysis constants of [65BUR/LIL2] and [71BUR/ZIN],
∆ r H m = (186 ± 6) kJ·mol–1 can be obtained for the formation of Ni 4 (OH) 44 + species in
3.2 m NaCl medium.
Based on these three values in two different media, and the probability that the
most accurate result is the value determined calorimetrically in 3.2 m NaCl(aq), the
value:
∆ r H mο ((V.25), 298.15 K) = (190 ± 10) kJ·mol–1
is selected in this review, which leads to
∆ f H mο ( Ni 4 (OH) 44 + , 298.15 K) = – (1173.4 ± 10.6) kJ·mol–1.
Electrostatic models such as the one presented in [94PLY/GRE] have been
used to estimate the temperature dependence of formation constants and entropies of
hydrolysis reactions of polynuclear hydroxo complexes, such as Ni 4 (OH) 44 + . Such
semi-theoretical treatments are beyond the scope of this review.
V.3.1.2 Hydroxo complexes in alkaline solutions
Three further complexes have been reported to form in Ni(II) solutions above ca. pH =
9: Ni(OH)2(aq), Ni(OH)3− and Ni(OH) 24 − , based on the increasing solubility of
Ni(OH)2(cr) in alkaline solutions. According to a re-evaluation, carried out in this review, only the hydrolysis constant of the second species (V.27) can be determined with
reasonable certainty:
*
(V.26)
Ni + 3H2O U Ni(OH) + 3H ;
*
(V.27)
Ni2+ + 4H2O U Ni(OH) 24 − + 4H+;
*
(V.28)
Ni2+ + 2H2O U Ni(OH)2(aq) + 2H+;
2+
−
3
+
b 2,1
b 3,1
b 4,1 .
The values of these hydrolysis constants selected by [76BAE/MES] and
[98PLY/ZHA], (see Appendix A) are based on a few experimental points of a single
paper [49GAY/GAR]. These data can, however, be equally well described when only
Ni(OH)3− is assumed to be present.
V.3 Oxygen and hydrogen compounds and complexes
100
Three publications [49GAY/GAR], [80TRE/LEB2], [89ZIE/JON] report thermodynamic properties for the Ni(OH)2(aq) and Ni(OH)3− species. The hydrolysis constants for reactions (V.26) and (V.27) can be derived from solubility equilibria:
Ni(OH)2(cr) U Ni(OH)2(aq)
K sο,2,1
(V.29)
Ni(OH)2(cr) + OH–U Ni(OH)3−
K sο,3,1
(V.30)
Ni(OH)2(cr) + 2H+ U Ni2+ + 2H2O
*
(V.31)
with
2+
and assuming that hydrolysis of Ni
mental solutions used.
K sο,0
+
to NiOH was essentially complete in the experi-
Several hydrolysis models can be considered in interpreting nickel hydrolysis
data obtained from experiments in basic solutions (see Figure V-7).
Figure V-7: Solubility of Ni(OH)2(s) in alkaline solutions at 298.15 K. • data from
[49GAY/GAR], estimated error: ± 0.5 log10 units, dash-dotted line: log10[Ni(II)]tot =
log10 K sο,3,1 + log10[OH–]ini, dotted line: log10 [Ni(II)]tot = log10 K sο,3,1 – ∆ε·[OH–]ini +
−
log10[OH–]ini; solid line: log10[Ni(II)]tot = log10 ( K sο,2,1 + K sο,3,1 [OH − ]ini 10−∆ε·[OH ]ini )
-3
0
0
K s,2,1, K s,3,1, ∆ε
0
K s,3,1, ∆ε
0
K s,3,1
exp. data [49GAY/GAR]
−1
log10([Ni(II)]tot/mol·kg )
-4
-5
-6
-7
-8
-9
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
−
-0.5
0.0
−1
log10([OH ]/mol·kg )
0.5
1.0
1.5
V.3 Oxygen and hydrogen compounds and complexes
101
Gayer and Garrett [49GAY/GAR] reported log10 *K sο,0 = (10.8 ± 0.15),
log10 K
≈ – 7 and log10 K sο,3,1 = – (4.2 ± 0.6). By combining these values, and the
ο
ο
ionisation constant of water (pKw = 14.0), log10 *b 2,1
= – (17.8 ± 1.0) and log10 *b 3,1
=
− (29.0 ± 0.8) can be derived for reactions (V.26) and (V.27), respectively. Reevaluation of the solubility data reported by these authors in alkaline solutions, using
the SIT approach, assuming formation of both Ni(OH)3− and Ni(OH)2(aq) (but excluding the solubility values obtained in 8 M, 10 M and 15 M NaOH(aq)), results in
log10 K sο,2,1 ≈ – (7.20 ± 0.22), log10 K sο,3,1 = − (4.58 ± 0.11), and ∆ε = (0.84 ± 0.24)
ο
kg·mol–1. These equilibrium constants correspond to log10 *b 2,1
= − (18.00 ± 0.30) and
* ο
1
log10 b 3,1 = − (29.38 ± 0.30) .
ο
s ,2,1
A re-evaluation of the solubility data, using the SIT approach, and assuming
Ni(OH)3− (but not Ni(OH)2(aq)) is formed in the dissolution process, results in
log10 K sο,3,1 = − (4.40 ± 0.12), ∆ε = (0.66 ± 0.17) kg·mol–1. The equilibrium constant
ο
corresponds to log10 *b 3,1
= − (29.20 ± 0.48). If activity coefficient effects are neο
ο
glected, log10 K s ,3,1 = − (4.63 ± 0.22) and log10 *b 3,1
= − (29.43 ± 0.48).
ο
ο
Tremaine and LeBlanc obtained, log10 *b 2,1
= − (20.26 ± 1.77) and log10 *b 3,1
= – (29.78 ± 1.97) at 298.15 K, by extrapolation, based on the solubility of NiO in
NaOH solutions measured between 150 and 300°C [80TRE/LEB2]. A re-evaluation
revealed that their experimental data can also be explained with Ni(OH)3− as the only
solubility controlling species in alkaline solutions (see Appendix A). The revised exο
trapolation to T = 298.15 K resulted in log10 *b 3,1
= – (30.9 ± 1.3), in reasonable agreement with [80TRE/LEB2], although Ni(OH)2 was neglected. Ziemniak et al.
[89ZIE/JON] also measured the solubility of NiO between 18 and 289°C in alkaline
solutions containing HPO 24 − ion. A re-evaluation of their data by [95LEM] resulted in
ο
ο
log10 *b 2,1
= − (19.77 ± 1.00) and log10 *b 3,1
= – (30.9 ± 1.3). As the numerical value of
* ο
log10 b 3,1 remains within its error limits whether Ni(OH)2 is taken into account or not,
no thermodynamic quantities for the latter species are selected in this review. The solid
line of Figure V-7 shows that a value of log10 K sο,2,1 = – (7.2 ± 0.2) is also consistent with
the data. No thermodynamic quantities for the latter species are selected in this review,
ο
though – 7 can be tentatively assigned as its upper limit. The value for log10 *b 3,1
obtained by the re-evaluation of data reported in [49GAY/GAR] has been selected, but
with an increased uncertainty to reflect the lower values obtained from use of the higher
temperature data of [80TRE/LEB2] and [89ZIE/JON]:
ο
log10 *b 3,1
((V.27), 298.15 K) = – (29.2 ± 1.7).
1
Gayer and Garrett [49GAY/GAR] did not report on the characterisation of their solid phase, which may,
or may not, have been identical to β-Ni(OH)2, the solid discussed in Section V.3.2.2.1.1. Therefore, the
value log10 *K sο,0 = (10.8 ± 0.15) was used for these recalculations, rather than the value for β-Ni(OH)2,
log10 *K sο,0 = (11.02 ± 0.20), selected in the present review (also see the discussion of [49GAY/GAR] in
Appendix A).
102
V.3 Oxygen and hydrogen compounds and complexes
From this value, ∆ r Gmο ((V.27), 298.15 K) = (166.7 ± 9.7) kJ·mol–1 can be obtained, while the Gibbs energy of formation is
∆ f Gmο ( Ni(OH)3− , 298.15 K) = – (590.5 ± 9.7) kJ·mol–1.
ο
ο
From the temperature dependence of log10 *b 2,1
and log10 *b 3,1
measured be–1
tween 150 and 300°C, ∆ r H m (V.26) = (109 ± 15) kJ·mol and ∆ r H m (V.27) =
(119 ± 15) kJ·mol–1 were obtained by [80TRE/LEB2]. From the re-evaluation of Ziemniak's data [89ZIE/JON] reported in [95LEM], ∆ r H mο (V.26) = (94 ± 10) kJ·mol–1 and
∆ r H mο (V.27) = (126 ± 15) kJ·mol–1 can be estimated. The re-evaluation of the data of
Tremaine and LeBlanc results in ∆ r H mο ((V.27), 298.15 K) = (121.2 ± 6.5) kJ·mol–1
which agrees satisfactorily with the values given by [80TRE/LEB2] and [95LEM]. Although the quantities are not sufficiently accurate to allow selection of selected values
for the standard reaction enthalpies in question (see above), they can be used as rough
estimates.
We have not found any convincing evidence for formation of Ni(OH) 24 − .
V.3.2 Solid nickel oxides and hydroxides
V.3.2.1 Ni(II) oxide
Bunsenite (NiO) is extremely rare, and a member of the periclase group of minerals. It
had been discovered as early as 1868 in Johanngeorgenstadt, Erzgebirge, Saxony, Germany in a hydrothermal Ni-U vein. It is translucent to opaque, and occurs as wellformed fine sized crystals (up to 3 mm). Its hardness is 5.5, the colour pistachio-green,
the lustre adamantine, and the streak brownish black.
The lattice of NiO is face centred cubic, space group: Fm3m , with unit cell
dimensions a0 = 4.177 Å, Z = 4, ρ(calc.) = 6.806 g·cm–3 (according to JCPDS-ICDD
card No.4-835, [53SWA/TAT]), ρ(obs.) = 6.4 – 6.8 g·cm–3.
A very accurate technique for the determination of the standard entropy
S mο (298.15 K) of a solid crystalline compound is the integration of low-temperature
heat capacity data, C pο,m , between 0 and 298.15 K:
S mο (298.15 K) =
∫
298.15
0
C pο,m
T
dT .
(V.32)
In the case of nickel oxide several publications dealing with heat capacity
measurements in the temperature range 3.2 – 477.8 K are available [40SEL/DEW],
[57KIN], [74WHI], [79DUB/NAU], [93WAT]. Seltz et al. [40SEL/DEW] investigated
the heat capacity from 68 to 298 K, King [57KIN] performed low-temperature measurements between 54 and 296 K, and DuBose and Naugle [79DUB/NAU] determined
the specific heat of NiO from 14 to 280 K. Whereas bulk-like NiO powder samples
were employed in all these investigations, White [74WHI] used single crystals for the
determination of C pο,m from 3.2 to 18.75 K. Furthermore, Watanabe [93WAT] measured
V.3 Oxygen and hydrogen compounds and complexes
103
the heat capacity of single crystals of NiO over the temperature range from 135.7 to
477.8 K. The experimental data of all studies mentioned above coincide remarkably, as
can be seen in Figure V-8. The heat capacity values between 0 and 40 K are perfectly
described by a Debye – T 3 law. A fifth order polynomial fit to the experimental data in
the range 40 – 477.8 K results in a C pο,m function which can be easily integrated. The
smoothed C pο,m data as well as calculated entropy values are listed in Table V-5. The
standard entropy of nickel oxide at 298.15 K, S mο (NiO), is equal to 38.4 J·K–1·mol–1.
This result is somewhat higher than the value (38.0 ± 0.2) J·K–1·mol–1 obtained by King
[57KIN] which was adopted by Kelley and King [61KEL/KIN], Kellogg [69KEL], the
NBS tables [82WAG/EVA], Knacke et al. [91KNA/KUB] and Robie and Hemingway
[95ROB/HEM]. As our value for S mο (NiO) is based on a simultaneous evaluation of
five independent experimental studies of comparable accuracy, the selected standard
entropy of NiO amounts to:
S mο (NiO, cr, 298.15 K) = (38.4 ± 0.4) J·K–1·mol–1.
The uncertainty has been chosen, such that both values (the result of the present evaluation and the value of King [57KIN]) fall within the error limits.
Table V-5: Smoothed C pο,m data and calculated values for the entropy of NiO.
T
C pο,m
S mο (NiO)
T
C pο,m
S mο (NiO)
(K)
(J·K–1·mol–1)
(J·K–1·mol–1)
(K)
(J·K–1·mol–1)
(J·K–1·mol–1)
10
0.02
0.006
160
27.09
15.865
20
0.13
0.047
170
28.93
17.563
30
0.45
0.159
180
30.68
19.266
40
0.96
0.376
190
32.32
20.969
50
2.98
0.800
200
33.85
22.666
60
5.13
1.536
210
35.29
24.353
70
7.37
2.494
220
36.63
26.026
80
9.66
3.627
230
37.87
27.682
90
11.97
4.898
240
39.03
29.318
100
14.28
6.279
250
40.10
30.933
110
16.57
7.748
260
41.11
32.526
120
18.82
9.286
270
42.04
34.095
130
21.00
10.880
280
42.93
35.640
140
23.12
12.514
290
43.76
37.161
150
25.15
14.179
298.15
44.41
38.383
300
44.55
38.658
V.3 Oxygen and hydrogen compounds and complexes
104
Figure V-8: Experimental results of heat capacity measurements on nickel oxide from
3.2 to 1809.7 K. ▲ [40SEL/DEW], ■ [74WHI], ○ [57KIN], × [79DUB/NAU], V
[93WAT], + [90HEM], □ [58KIN/CHR]; solid line: polynomial fit (present assessment).
90
80
o
−1 –1
/ J·K−–11··mol
CC°
/ J·mol
K
p ,m
p,m
70
60
50
40
30
20
10
0
0
200
400
600
800
1000
1200
1400
1600
1800
T/K
The heat content of nickel oxide was investigated at temperatures above
298.15 K by Tomlinson et al. [55TOM/DOM] as well as King and Christensen
[58KIN/CHR]. The experimental data of these contributions are in close agreement.
King and Christensen interpreted their results in terms of two phase transitions occurring at 525 K and 565 K, respectively. Hemingway [90HEM] reinvestigated the hightemperature heat capacity of bunsenite (NiO). Both the experimental data of
[58KIN/CHR], [90HEM] and the low-temperature heat capacities are illustrated in
Figure V-8. The data of Hemingway clearly demonstrate that in the high-temperature
region only the antiferromagnetic-paramagnetic phase transition can be observed. The
corresponding Néel temperature (the antiferromagnetic-paramagnetic transition temperature) was found to be equal to 519 K and the following polynomial fits were given
[90HEM]:
–3
[C pο,m ]519
(T/K)2 –
298.15 (NiO, cr) = (4110.720 – 5.302412 (T/K) + 3.52061×10
53039.297 (T/K)–0.5 + 2.43067×107 (T/K)–2) J·K–1·mol–1
(V.33)
V.3 Oxygen and hydrogen compounds and complexes
105
valid over the range from 298.15 to 519 K, and
–2
[C pο,m ]1800
(T/K) – 7.5267×10–6 (T/K)2 +
519 (NiO, cr) = (– 8.776 + 4.2232×10
–0.5
787.25 (T/K) + 3.6067×106 (T/K)–2) J·K–1·mol–1
(V.34)
valid from 519 to 1800 K. Equation (V.33) has to be modified somewhat in order to
obtain a heat capacity function which is consistent with the value at 298.15 K listed in
Table V-5:
–3
[C pο,m ]519
(T/K)2 –
298.15 (NiO, cr) = (4110.64 – 5.302412 (T/K) + 3.52061×10
–0.5
7
–2
53039.297 (T/K) + 2.43067×10 (T/K) ) J·K–1·mol–1
(V.35)
valid over the range from 298.15 to 519 K. Equations (V.34) and (V.35) describe the
temperature dependence of the heat capacity above 298.15 K satisfactorily. Moreover,
Equation (V.34) is consistent with the data of King and Christensen [58KIN/CHR]
above 846 K, see the solid line in Figure V-8. Therefore, the Reviewers have selected
the heat capacity function Equations (V.35) and (V.34), where C pο,m at 298.15 K
amounts to:
C pο,m (NiO, cr, 298.15 K) = (44.4 ± 0.1) J·K–1·mol–1.
The uncertainty corresponds to the deviation between Equations (V.33) and
(V.35) at 298.15 K.
Boyle et al. [54BOY/KIN] determined the enthalpy of formation of nickel oxide directly by means of combustion calorimetry, ∆ f H mο (NiO) = – (239.7 ± 0.4)
kJ·mol−1. The application of a solid-state electrochemical cell to determine the Gibbs
energy of formation of nickel oxide at high temperatures was originally introduced by
Kiukkola and Wagner [57KIU/WAG]. By employing electrochemical cells of the type:
Reference electrode | YSZ or CSZ | NiO, Ni, Pt
(V.36)
the temperature dependence of the Gibbs energy function of NiO was studied by a number of authors, where yttria- or calcia-stabilised zirconia served as the solid electrolyte
and the reference electrode consisted of Pt, O2(air) or Fe, FeOx (wuestite) or Cu, Cu2O
[63RAP], [65TRE/SCH], [65STE/ALC], [67RIZ/BID], [68CHA/FLE], [69MOR/SAT],
[70HUE/SAT],
[75MOS/FIT2],
[76BER2],
[78IWA/FUJ],
[79KEM/KAT],
[80TRA/BRU], [84COM/PRA], [84RAY/PET], [86HOL/NEI], [93NEI/POW],
[94KAL/FRA]. Some of these references are discussed in Appendix A. Basically, the
results of all these experimental studies agree within the limits of uncertainty. Especially, Charette and Flengas [68CHA/FLE], Berglund [76BER2], Comert and Pratt
[84COM/PRA], Holmes et al. [86HOL/NEI] and O'Neill and Pownceby [93NEI/POW]
performed highly reliable and precise electrochemical measurements of the Gibbs energy of formation as a function of temperature, as shown in Figure V-9.
Equation (V.37) allows the calculation of the Gibbs energy of formation of
NiO as a function of temperature:
V.3 Oxygen and hydrogen compounds and complexes
106
∆ f Gm (T ) = ∆ f H mο (298.15 K) − T ∆ f Smο (298.15 K) +
(V.37)
∆ f C p ,m
T
∆
C
T
−
T
T
d
d
p
f
,m
∫298.15
∫298.15 T
The auxiliary data for O2(g) at 298.15 K can be found in the NEA TDB auxiliary data set, heat capacities at higher temperatures were taken from CODATA
[89COX/WAG] and from Robie and Hemingway [95ROB/HEM] in tabular and analytical form, respectively. When the selected data for Ni(cr), the calorimetric value for
the standard enthalpy of formation of NiO, according to [54BOY/KIN], the heat capacity function and the standard entropy of NiO (selected above) are used, the predicted
temperature dependence of the Gibbs energy of formation agrees remarkably well with
experimental data obtained from various high-temperature electrochemical measurements, see Figure V-9. Thus, we select for the standard enthalpy of formation of nickel
oxide at 298.15 K:
T
∆ f H mο (NiO, cr, 298.15 K) = – (239.7 ± 0.4) kJ·mol–1
and the corresponding derived value for the Gibbs energy of formation at 298.15 K is:
∆ f Gmο (NiO, cr, 298.15 K) = – (211.66 ± 0.42) kJ·mol–1.
Figure V-9: Comparison between the temperature dependence of the Gibbs energy of
formation for NiO obtained from electrochemical as well as chemical reduction/oxidation equilibrium measurements and the prediction based on the present selection of thermodynamic properties for NiO.
-80
-90
−1
∆∆fG°
(NiO)//kJ·mol
kJ·mol–1
fG°
m (NiO)
m
-100
-110
-120
[26PEA/COO]
[33WAT]
[38BOG]
[42FRI/WEI]
[73RAU/GUE]
[76BER2]
[86HOL/ONE]
[84COM/PRA]
[68CHA/FLE]
present review
-130
-140
-150
-160
-170
-180
-190
600
800
1000
1200
T/K
1400
1600
1800
V.3 Oxygen and hydrogen compounds and complexes
107
In addition, equilibrium constants for the reduction of NiO with H2 and CO are
reported in [26PEA/COO], [73RAU/GUE] and [33WAT], [38BOG], [42FRI/WEI],
[64ALC/BEL], [67ANT/WAR], respectively. The equilibrium constants for the reactions:
NiO(cr) + H2(g) U Ni(cr) + H2O(g)
NiO(cr) + CO(g) U Ni(cr) + CO2(g)
can be transformed into the Gibbs energy of formation of NiO by using the pertinent
auxiliary quantities for H2(g), H2O(g), CO(g) and CO2(g) listed in the CODATA tables
[89COX/WAG] and Robie and Hemingway [95ROB/HEM]. The equilibrium data for
the reduction/oxidation reactions given by [26PEA/COO], [33WAT], [42FRI/WEI],
[64ALC/BEL], [67ANT/WAR], [73RAU/GUE] are in perfect agreement with calculated values for the Gibbs energy of formation using the thermodynamic data selected in
the present compilation, as can be deduced from Figure V-9 (the results of
[64ALC/BEL], [67ANT/WAR] are discussed in Appendix A). However, the data of
Bogatzki [38BOG] deviate somewhat from the calculated line at lower temperatures,
probably because thermodynamic equilibrium was not attained below 1000 K.
The solid – gas equilibria:
NiO(cr) + H2(g) U Ni(cr) + H2O(g)
NiO(cr) + CO(g) U Ni(cr) + CO2(g)
were likewise studied by [21WOH/BAL], [32SKA/DAB] and [36KAP/SIL], respectively, who obtained equilibrium constants deviating by approximately one order of
magnitude from those predicted according to the thermodynamic model of the present
assessment. The presumably erroneous results of these early works may be caused by (i)
insufficient equilibration of the solid material with the gas phase, (ii) deposition of carbon by the Boudouard reaction, or (iii) inaccurate chemical analysis of the composition
of the gas phase.
Also, the relative mass change of NiO was monitored as a function of the oxygen partial pressure at 1473 K by means of thermogravimetry [89KIT]. Hence, the decomposition of NiO into O2(g) and Ni was found to occur at log10(p(O2)/bar) = – 7.74
which concurs well with log10(p(O2)/bar) = – 7.84 calculated from the thermodynamic
data of the present compilation.
V.3.2.1.1
Solubility measurements
Due to the kinetically inert nature of nickel oxide with respect to its dissolution in aqueous media the solubility of NiO has been studied only at elevated temperatures so far
[80TRE/LEB2], [89ZIE/JON]. These studies are not suitable for the calculation of any
thermodynamic properties of NiO because of the high uncertainty of the measured solubilities compared to the high-temperature emf data and the low-temperature heat capac-
108
V.3 Oxygen and hydrogen compounds and complexes
ity data discussed above. Moreover, the evaluation of the solubility experiments performed at hydrothermal conditions may cause an additional uncertainty for the solubility constant of NiO owing to the lack of heat capacity functions for the ionic species
including the hydroxo complexes. Thus, the calculated value for the solubility constant
of NiO at 298.15 K, according to the reaction:
NiO(cr) + 2H+ U Ni2+ + H2O(l),
derived from the thermodynamic data accepted in the present assessment, is :
log10 *K sο,0 (NiO, cr, 298.15 K) = (12.48 ± 0.15).
V.3.2.2 Ni(II) hydroxides, Ni(OH)2(cr)
Recently the thermodynamic property values of Ni(II) hydroxide have been reviewed by
Archer [99ARC] and critically evaluated by Plyasunova et al. [98PLY/ZHA]. The data
of Ni(OH)2(cr) reported are rather uncertain and, as this solid phase is of considerable
importance for the deposition and remobilisation of nickel, they have been re-evaluated
on the basis of recent solubility measurements [2002GAM/WAL].
V.3.2.2.1
V.3.2.2.1.1
Crystallography and mineralogy of nickel hydroxide
β-Ni(OH)2
In a comprehensive review on bivalent metal hydroxides it was pointed out that the unit
cell dimensions of synthetic β-Ni(OH)2 were determined by Lotmar and Feitknecht for
the first time [36LOT/FEI], [77OSW/ASP]. Like most M(OH)2 hydroxides, the crystal
structure is of the CdI2- or brucite- type, trigonal, space group: P 3 m1, with unit cell
dimensions a = 3.126 Å, c = 4.605 Å, Z = 1 (according to JCPDS-ICDD card
No. 14−117).
The natural occurrence of Ni(II) hydroxide in the Lord Brassey mine, Heazlewood, Tasmania, was first reported by Williams [60WIL2]. Almost pure Ni(OH)2 was
found in the Vermion region, northern Greece, whereas Mg-bearing (Ni, Mg)(OH)2 occurs in Unst, Shetland, Scotland [81MAR/ECO], [82LIV/BIS]. The X-ray powder diffraction patterns of the former mineral match closely with synthetic β-Ni(OH)2. It was
named theophrastite after Theophrastos, the first Greek mineralogist, 373/372–288/287
B.C. [81MAR/ECO] (approved by IMA 1980). Theophrastite from the Vermion region
is emerald green, translucent with vitreous luster and has a pale green streak. The Mohs
hardness is 3.5. The specific gravity is ρ(obs.) = 4.0 and ρ(calc.) = 3.95 g·cm–3. It is a
gangue mineral in ore consisting of magnetite (Fe3O4), chromite (FeCr2O4) and Nisulphides as minor component. Theophrastite is formed from Ni-bearing solutions between 80 ≤ t(°C) ≤ 115 in alkaline moderately oxidising media [83ECO/MAR].
V.3 Oxygen and hydrogen compounds and complexes
V.3.2.2.1.2
109
α-Ni(OH)2
Other varieties of crystallised divalent nickel hydroxide, α-Ni(OH)2 and α∗-Ni(OH)2,
differ from the thermodynamically stable β-form by the presence of a layer of water
molecules in the van der Waals gap. Bode et al. [66BOD/DEH] proposed an
α−3Ni(OH)2·2H2O unit cell (a = 3.08 Å, c = 8.09 Å), whereas Braconnier et al.
[84BRA/DEL]1 synthesised a compound with a stoichiometric composition of
α∗−Ni(OH)2·0.75H2O and the following unit cell dimensions: a = 3.08 Å, c = 23.41 Å.
In any case the intercalated crystal water of α-Ni(OH)2 causes an extension of the structure along the crystallographic c-axis, whereas the other dimension remains similar to
that of theophrastite. The α-, α∗- and β-forms of nickel hydroxide have the same trigonal structure. The β-form consists of well-ordered brucite type layers with a separation
of 0.46 nm, whereas the hydrated α-form shows larger layer distances and turbostratic
character as a result of intercalated water and distorted stacking, respectively. The
α−form may also intercalate different anions so that the interlayer separation depends
on the type as well as the extent of these ions and varies between 0.8 and 0.9 nm. By
electron microscopy the very small crystallites appear as formless aggregates
[84BRA/DEL]. Although it plays an important role in the charge/discharge cycle of
nickel batteries no thermodynamic data can be definitively assigned to it (see Sections
V.2.2 and V.3.2.3 about Ni(III, IV) ions and hydroxides). The natural occurrence of
α−Ni(OH)2 has never been reported as it is probably too unstable to persist under ambient conditions.
V.3.2.2.2
Heat capacity and entropy of Ni(OH)2(cr)
Sorai et al. [69SOR/KOS] measured the heat capacity of Ni(OH)2(cr) at low temperatures. These data have been used to determine the standard entropy, S mο (298.15 K),
which has been accepted for the selected thermodynamic data of nickel compounds
[76MAH/PAN]. Sorai et al. [69SOR/KOS] investigated three samples, (iii), (ii), and (i),
of Ni(OH)2(cr) with different particle sizes, and reported C pο,m values up to 104 K and
300 K for the coarsest (iii) and the finer crystals (ii, i), respectively. As shown in Figure
V-10 the heat capacity and consequently the entropy turned out to depend on the particle size and thus on the specific surface area s.
When the excess surface heat capacity estimated by Sorai et al. between 80 and
130 K was assumed to remain constant up to 300 K values of C pο,m (Ni(OH)2, cr,
298.15 K) = 81.7 J·K–1·mol–1 and S mο (Ni(OH)2, cr, 298.15 K) = 79.4 J·K–1·mol–1 have
been extrapolated for the coarsest material. When, however, the entropy increases linearly with s, extrapolation to s = 6 m2·g–1 (according to Sorai et al. the specific surface
area of the coarsest sample) results in S mο (Ni(OH)2, cr, 298.15 K) = 80.0 J·K–1·mol–1, in
1
The latter phase has been called α∗ to emphasise that the composition and the intersheet distance are close
to those of the turbostratic α-Ni(OH)2, whereas the texture and structure are different.
V.3 Oxygen and hydrogen compounds and complexes
110
perfect agreement with the value selected by Mah and Pankratz [76MAH/PAN]. This
value has been retained, but the uncertainty was increased from ± 0.4 to ± 0.8:
S mο (Ni(OH)2, β, 298.15 K) = (80.0 ± 0.8) J·K–1·mol–1.
(V.38)
These increased error limits probably account also for the uncertainty introduced by Enoki and Tsujikawa’s slightly differing results for bulky Ni(OH)2 between
4.2 to 35 K [78ENO/TSU], see Figure V-10. In addition the interpolated C pο,m value for
sample (ii) at 298.15 K has been preferred to the extrapolated one of sample (iii). An
uncertainty has been selected, so that both values fall within the error limits:
C pο,m (Ni(OH)2, β, 298.15 K) = (82.01 ± 0.30) J·K–1·mol–1.
Figure V-10: The heat capacity of Ni(OH)2(cr) as a function of temperature (▲
s(Ni(OH)2(iii)) = 6 m2·g–1 [69SOR/KOS]; ☼ s(Ni(OH)2(ii)) = 64 to 75 m2·g–1 estimated
by the reviewer; × s(Ni(OH)2(i)) = 320 m2·g–1 [69SOR/KOS]; thick line: bulky Ni(OH)2
[78ENO/TSU]).
24
22
20
18
p,m
1
−1–1
C°
J·K–1−·mol
Cp ,m //J·mol
·K
16
14
12
10
8
6
4
2
0
0.0
0.5
1.0
1.5
2.0
2.5
ln (T / K)
3.0
3.5
4.0
4.5
5.0
V.3 Oxygen and hydrogen compounds and complexes
V.3.2.2.3
111
Thermodynamic analysis of solubility data of Ni(OH)2(s)
For a methodological review of the experimental determination of solubilities of sparingly soluble ionic solids in aqueous media see [2003GAM/KON].
All auxiliary variables discussed in Section V.2.1 were used in the thermodynamic analysis of solubility data.
The determination of the solubility product of Ni(OH)2(cr) is in principle a
straightforward method to evaluate the standard Gibbs energy of formation,
∆ f Gmο (Ni(OH)2, cr). It presumes, however, that the solubility product (defined by Equation (V.39)) or any appropriate solubility constant (as for example defined by Equation
(V.40)) of the respective phase has indeed been determined.
β-Ni(OH)2 U Ni2+ + 2OH–
Solubility product:
+
2+
β-Ni(OH)2 + 2H U Ni + 2H2O(l)
Dissolution:
K s ,0
(V.39)
*
(V.40)
K s ,0
Standard thermodynamic quantities from solubility measurements on hydroxides can be obtained using Equations (V.41) – (V.45).
∆ sol Gmο = ∆ sol H mο – T ∆ sol S mο = – R T ln *K sο,0
(V.41)
∆ sol H mο = R T 2 (∂ln *K sο,0 /∂T)
(V.42)
∆ f Gmο (Ni(OH)2, β) = ∆ f Gmο (Ni2+) + 2 ∆ f Gmο (H2O, l) – ∆ sol Gmο
ο
m
ο
m
2+
ο
m
∆ f H (Ni(OH)2, β) = ∆ f H (Ni ) + 2 ∆ f H (H2O, l) – ∆ sol H
S mο (Ni2+) = S mο (Ni(OH)2, β) – 2 S mο (H2O, l) + ∆ sol S mο
(V.43)
ο
m
(V.44)
(V.45)
As discussed in Section V.3.2.2.2, the entropy of nickel hydroxide was calorimetrically measured with sufficient accuracy [69SOR/KOS], [78ENO/TSU] that, in
fact, the value of the entropy of dissolution contributes to the knowledge of the partial
molar entropy S mο (Ni2+), see Equation (V.45).
Actually most solubility data of Ni(OH)2(cr), reported so far and listed in Table
V-6, suffer from an uncertainty in the physical state of the solid investigated [18ALM],
[25BRI], [25WIJ], [38OKA], [43NAS], [49GAY/GAR], [50AKS/FIA], [53SCH/POL],
[54DOB], [54FEI/HAR], [56CUT/KSA], [56JEN/PRA], [67BLA], [69MAK/SPI],
[73KAW/OTS], [73NOV/COS], [80CHI/SAB], [96POU/DRE]. This is particularly surprising for Feitknecht and Hartmann [54FEI/HAR], because Lotmar and Feitknecht
[36LOT/FEI] were the first to determine the structure of β-Ni(OH)2, and the concept
that only well defined solid phases result in reliable solubility constants of metal oxides
and hydroxides was developed by Schindler in the same laboratory [63SCH],
[63FEI/SCH]. Moreover, the comprehensive review on Ni(II) hydroxide mentioned in
Section V.3.2.2.1, is also based to a considerable extent on the work performed by
Feitknecht and his co-workers [77OSW/ASP]. There it has been clearly pointed out
that, apart from the well defined β-Ni(OH)2, a number of basic salts of changing com-
112
V.3 Oxygen and hydrogen compounds and complexes
position exist. When nickel hydroxide is precipitated from aqueous NiCl2, Ni(NO3)2, or
NiSO4 with NaOH or KOH solutions it is always contaminated with basic salts. The
solubility of the latter varies depending on the anion and the molar ratio OH/Ni. This
means that solubility studies on poorly defined nickel hydroxide or the respective basic
salts are useless as an experimental basis to derive accurate thermodynamic functions of
nickel hydroxide. They may, however, serve to find out relevant information concerning
removal of Ni2+ from radioactive effluents.
Sometimes solubility products have to be rejected, because they were calculated on the basis of quite inadequate experimental results, see Appendix A [18ALM],
[53SCH/POL], [67BLA], [80CHI/SAB].
One of the most careful solubility studies reported so far was carried out by
Mattigod et al. [97MAT/RAI], but according to their preparative method the respective
results refer to a microcrystalline β-Ni(OH)2 that probably contained chloride. It is
noteworthy that Poulson and Drever’s [96POU/DRE] study of a commercially available
nickel hydroxide (Johnson Matthey Chemical Co.) resulted in a solubility constant
which after being extrapolated to the same temperature ( log10 *K sο,0 = 12.08) agrees
within 0.18 log10-units with that of Mattigod et al. [97MAT/RAI].
A re-evaluation of ∆ f Gmο and ∆ f H mο of β-Ni(OH)2 has been based on the results of [2002WAL/GAT] and [2002GAM/WAL].
In Figure V-11, the solubility constant log10 *K s ,0 , is plotted as a function of
temperature. It is quite obvious that after having rejected the most unreliable results
[18ALM], [25WIJ], [67BLA], [80CHI/SAB], the maximum error limits according to
Plyasunova et al. [98PLY/ZHA] turned out to be too pessimistic—albeit the usual experimental error is exceeded by an order of magnitude. The results of several authors
scatter closely around the curve adjusted to log10 K sο,0 of Mattigod et al. which seems to
describe the solubility of finely divided β-Ni(OH)2, [25BRI], [54DOB], [56JEN/PRA],
[96POU/DRE], [97MAT/RAI], whereas still higher solubilities presumably refer to
amorphous Ni(OH)2 or even basic salts. Measurements carried out with carefully prepared coarse theophrastite led to lower solubilities, closer error limits, and a comparatively reliable prediction of thermodynamic quantities [2002GAM/WAL].
V.3 Oxygen and hydrogen compounds and complexes
113
Table V-6: Literature data(a) for log10 *K sο,0 (Ni(OH)2).
Medium
I (mol·kg–1)
NiCl2, NaCl 0.030 – 0.065
Ni(NO3)2,
0.3 – 0.4
T (K)
log10 *K sο,0
291.15
298.15
(a)
Method
Remarks
Reference
12.20
H, titr.
Ni(OH)2–xClx(s)?
[25BRI]
13.36
[OH ]→
incipient ppt.
[25WIJ]
NH4NO3
–
[NH3]/[ NH +4 ]
Ni(NO3)2
dilute
298.15
13.50
pH, titr.
incipient ppt.
[38OKA]
NiCl2, KCl
0.002 – 2.05
298.15
12.92
H, titr.
Ni(OH)2–xClx(s)?
[43NAS]
NiCl2
0.038 – 0.15
298.15
10.78
pH, [Ni(II)]
saturated?
[49GAY/GAR]
NiCl2
0.038 – 0.15
308.15
10.78
pH, [Ni(II)]
sat. at 35°C ?
[49GAY/GAR]
(b)
(c)
NiSO4
0.006 – 5.37
291.15
13.40
H, QH , Sb
Ni(OH)2 pptd.
[50AKS/FIA]
NiSO4
0.030 – 0.065
348.15
9.48
pH, titr.
Figs. 3 and 5.1
[54DOB]
NiCl2
0 corr
298.15
13.30
pH, ?
Ni(OH)2, fresh
[54FEI/HAR]
NiCl2
0 corr
298.15
10.80
pH, ?
Ni(OH)2, aged
[54FEI/HAR]
NaClO4
1.05
293.15
12.56
pH, [Ni(II)]
incipient ppt.
[56CUT/KSA]
NiCl2
0.028 – 0.28
298.15
12.12
pH, [Ni(II)]
Ni(OH)2 pptd.
[56JEN/PRA]
302.15
11.75
pH, [Ni(II)]
Ni(OH)2 pptd.
[56JEN/PRA]
298.15
12.46
pH, [Ni(II)]
incipient ppt.
[69MAK/SPI]
CH3COONa 0.10 – 0.50
NiCl2
0.001 – 4.7 M
(d)
max
LiClO4
3.48
298.15
12.99
pH, Z
incipient ppt.
[73KAW/OTS]
NaCl
0.56
298.15
12.49
pH, [Ni(II)]
slow ppt.
[73NOV/COS]
NaCl
0.56
298.15
10.89
pH, [Ni(II)]
fast ppt.
[73NOV/COS]
NaClO4
1.05
298.15
10.38
pH, [Ni(II)]
fast ppt.
[73NOV/COS]
NaNO3
0.02 – 0.2
295.15
12.24
pH, [Ni(II)]
Johnson Matthey
[96POU/DRE]
NaClO4
0.01
298.15
11.90
pH, [Ni(II)]
X-ray,β- Ni(OH)2
[97MAT/RAI]
NaClO4
0.5 – 3.0, SIT
298.15
11.02
pH, [Ni(II)]
X-ray, β- Ni(OH)2
[2002GAM/WAL]
(a) Whenever possible the entries of column 4 have been recalculated from the experimental data.
(b) quinhydrone electrode.
(c) antimony electrode.
(d) maximum number of H+ set free per Ni.
V.3 Oxygen and hydrogen compounds and complexes
114
Figure V-11: Solubility constant of Ni(OH)2 as a function of temperature. Thick solid
curve: optimised from experimental data of Gamsjäger et al. [2002GAM/WAL]; thin
solid curves: corresponding error limits; thick dash-dotted curve: calculated according
to [98PLY/ZHA]; thin dash-dotted curves: corresponding maximum error limits; thick
dash curve: calculated with ∆ f H mο (Ni(OH)2, cr) adjusted to coincide with the result of
Mattigod et al. [97MAT/RAI]; ▲ [25BRI], ⊗ [38OKA], ∆ [43NAS], ∇ [49GAY/GAR],
■ [50AKS/FIA], ● [54DOB], × [54FEI/HAR] (aged), ♦ [54FEI/HAR] (freshly prepared),
[56CUT/KSA], ∗ [56JEN/PRA], [56JEN/PRA], ○ [69MAK/SPI], ▼
[73KAW/OTS], + [73NOV/COS] (high value: slow, lower values: fast precipitation), ⊕
[96POU/DRE], □ [97MAT/RAI], [2002GAM/WAL].
14
Ni(OH)2(s) + 2 H+ U Ni2+ + 2 H2O(l)
13
* o
log1010*KK°ss,0
log
,0
12
11
10
9
8
7
6
270
280
290
300
310
320
330
340
350
360
T/K
In Table V-7 the thermodynamic properties of Ni(OH)2 obtained in
[2002GAM/WAL] and selected by Plyasunova et al. are compared. The latter authors
overlooked a direct calorimetric determination of S mο of Ni(OH)2(cr) [69SOR/KOS],
[76MAH/PAN] which has been discussed in Section V.3.2.2.2 and Appendix A
([69SOR/KOS], [78ENO/TSU]). For the optimisation algorithm applied, the calorimetrically measured S mο (β-Ni(OH)2) = 80.0 J·K–1·mol–1 (Equation (V.38)) and S mο (Ni2+) =
– 131.8 J·K–1·mol–1 (Equation (V.15)) were kept fixed. Based on solubility data, the
V.3 Oxygen and hydrogen compounds and complexes
115
precision and reliability of ∆ f Gmο and ∆ f H mο (β-Ni(OH)2) can be improved considerably, leading to the selections:
∆ f Gmο (Ni(OH)2, β, 298.15 K) = – (457.1 ± 1.4) kJ·mol–1
∆ f H mο (Ni(OH)2, β, 298.15 K) = – (542.3 ± 1.5) kJ·mol–1.
Table V-7: Thermodynamic properties of Ni(OH)2(cr).
Compound
log10 *K sο,0 (V.40)
∆ sol H mο
S mο
–1
–1
Reference
–1
(kJ·mol )
(J·K ·mol )
Ni(OH)2(cr)(a)
(10.5 ± 1.3)
– (78.7 ± 10.2)
(73 ± 10)
β-Ni(OH)2(b)
(11.02 ± 0.20)#
– (84.36 ± 1.24)#
β-Ni(OH)2(b)
[98PLY/ZHA]
[2002GAM/WAL]
(80.0 ± 0.8)
[69SOR/KOS], [76MAH/PAN]
(a) Uncertainties are taken from Plyasunova et al. [98PLY/ZHA].
(b) Uncertainties based on ∆ log10 *K s ,0 = ± 0.2 (2σ), estimated from the experiments.
#
From selections in this review.
According to the present thermodynamic model the equilibrium temperature
for:
β-Ni(OH)2 U NiO(cr) + H2O(l)
amounts to T/K= (503 ± 31). The uncertainty has been obtained from error propagation
of the estimated uncertainties for the corresponding enthalpy and entropy of reaction (T
= 503 K): ∆ r H mο = (26.40 ± 1.55) kJ·mol–1; ∆ r Smο = (52.50 ± 0.89) J·K–1·mol–1. The
lower value for the decomposition temperature of Ni(OH)2 found by [89ZIE/JON],
468 K, may be caused by a scarcely crystallised and, therefore, less stable Ni(OH)2.
V.3.2.3 Ni(III, IV) hydroxides
Nickel hydroxides of oxidation state two and higher have been used as the active material in the positive electrodes of several alkaline batteries for more than a hundred years
[99MCB]. Bode et al. [66BOD/DEH] established the scheme depicted below.
charge →
← discharge
β-Ni(OH)2 U β-NiOOH
dehydratation ↑
↓ overcharge
α-Ni(OH)2 U γ-NiOOH
The charge and discharge cycles of nickel batteries involve two different pairs
of solid phases. Oxidation of β-Ni(OH)2 produces β-NiOOH, oxidation of α-Ni(OH)2
produces γ-NiOOH. The end-products of these cycles are interconnected by dehydration
and overcharge. For the crystallographic properties of Ni(II) hydroxides and Ni(III, IV)
hydroxides see Section V.3.2.2.1 and Sections V.3.2.3.1, V.3.2.3.2, respectively.
V.3 Oxygen and hydrogen compounds and complexes
116
V.3.2.3.1
β-NiOOH
Glemser and Einerhand [50GLE/EIN] synthesised β-NiOOH and determined the dimensions of its hexagonal unit cell to be a0 = 2.81 Å, c0 = 4.84 Å. This Ni(III) oxide
hydroxide is the primary oxidation product of electrodes containing β-Ni(OH)2.
Conway and Gileadi [62CON/GIL] measured the reversible potential of the
β−NiOOH | β−Ni(OH)2 electrode as a function of the degree of oxidation in 1 molal
KOH. The mercury | mercury oxide system, Hg | HgO, was used as reference electrode.
Similar measurements were carried out by Barnard et al. [81BAR/RAN]. In this review
the experimental data according to both teams of authors have been based on the standard hydrogen electrode (SHE) and extrapolated to I = 0 with the SIT model
[97GRE/PLY2]. In Figure V-12 the reversible potential of Reaction (V.46) has been
plotted versus the oxidation state of nickel which is numerically equal to x (NiIII).
β-NiOOH + ½ H2(g) U β-Ni(OH)2
(V.46)
Figure V-12: β-Ni(OH)2 | β-NiOOH system. Variation of reversible potential with the
oxidation state of Ni at 298.15 K and I = 0. Solid curve: regular model; experimental
data: ■ [62CON/GIL]; ○ [81BAR/RAN].
1.44
1.42
1.40
E/V
1.38
1.36
1.34
1.32
1.30
1.28
0.0
0.2
0.4
0.6
0.8
1.0
x (Ni(III))
When nickel (II, III) hydroxides form regular solid solutions in the whole
range of compositions between β-Ni(OH)2 and β-NiOOH the experimental data can be
V.3 Oxygen and hydrogen compounds and complexes
117
evaluated with Equation (V.47), where n = 1, A ( = ∆ G ex /[x·(1 – x)]) is the excess interaction parameter and x the mole fraction of Ni(III):
RT
(1 − x) A
E = E° –
(V.47)
⋅ ln
+
⋅ (1 − 2 x)
nF
x
nF
A non-linear least squares analysis of these experimental data led to A =
(3.4 ± 1.0) kJ·mol–1, E° = (1.361 ± 0.006) V and ∆ r Gmο (298.15 K) = – (131.3 ± 0.6)
kJ·mol–1. Now the standard Gibbs energy of formation can be derived:
∆ f Gmο (NiOOH, β, 298.15 K) = – (325.4 ± 6.0) kJ·mol–1
(V.48)
It must be emphasised, however, that potentiometric measurements are thermodynamically comparable only when exactly the same phases are involved in the respective redox reactions. As shown in Figure V-11 the solubility constant and consequently the Gibbs energy of formation of β-Ni(OH)2 presumably depends on the particle
size (compare thick solid and thick dash curves which differ by ≈ 0.8 log10K-units).
Thus the total uncertainty in Equation (V.48) is composed of an experimental random
error 2σ ≈ 1.5 kJ·mol–1 and a systematic error due to the solid phase used of ≈ 4.5
kJ·mol–1.
Jain et al. [98JAI/ELM] determined the standard potential of:
NiOOH(s) + H2O(l) + e– U Ni(OH)2(s) + OH–
(V.49)
against the silver | silver chloride electrode. When their result is recalculated versus
SHE the following standard potential is obtained, E°(298.15 K) = 1.340 V. The Gibbs
energy of NiOOH formation derived from this result differs by only 2 kJ·mol–1 from the
Equation (V.48) value. It might be that not quite the same pair of phases have determined the potentials measured by Barnard et al. [81BAR/RAN] and Jain et al.
[98JAI/ELM] respectively.
Jain et al. measured the following value (∂ ∆ r E°(V.50)/∂T)p = – 0.84 mV·K–1
for the temperature coefficient of the galvanic cell (V.50).
β-NiOOH + H2O(l) + Ag(cr) + Cl– U β-Ni(OH)2 + AgCl(cr) + OH–
(V.50)
This value is regarded to be in reasonable agreement with an earlier value of
− 0.5 mV·K–1. On the basis of these data S mο (NiOOH, β, 298.15 K) = 77.3 or 44.5
J·K−1·mol–1 is obtained, when the other standard entropies were taken from CODATA
[89COX/WAG] as adopted by the NEA TDB Project and Equation (V.38). With the socalled “Latimer” entropy contributions listed in Tables XIII, XIV and XVI of
[93KUB/ALC] S mο (NiOOH, β) can be estimated. It is either 55.8 J·K–1·mol–1, when the
OH– contribution for a divalent cation is accepted, or 62.6 J·K–1·mol–1, when NiO+ is
considered to be a monovalent cation. To sum up the situation, the mean of these observed and estimated values (60 ± 27) J·K−1·mol–1, which seems too rough an estimate
to be recommended.
V.3 Oxygen and hydrogen compounds and complexes
118
V.3.2.3.2
γ-NiOOH
γ-NiOOH is produced either by oxidation of α-Ni(OH)2 or on overcharge of β-Ni(OH)2.
This nickel oxide hydroxide was first prepared by Glemser and Einerhand
[50GLE/EIN]. It crystallises in a rhombohedral system with cell dimensions of a0 =
2.8 Å, c0 = 20.65 Å and has a layer structure with a spacing of 7.2 Å between layers.
γ-NiOOH always contains small quantities of water and alkali metal ions between the
layers, whereas β-NiOOH does not. When γ-NiOOH is prepared by the method of
Glemser and Einerhand its chemical composition is NiOOH·0.51H2O.
The standard potential measurements of Reaction (V.51) are difficult to interpret thermodynamically.
γ-NiOOH + ½ H2(g) U α-Ni(OH)2
(V.51)
Clearly the γ-NiOOH | α-Ni(OH)2 electrode plays an important role in the operation of alkaline nickel batteries and has been investigated extensively, but thermodynamic data of α-Ni(OH)2 still remain unknown. Consequently, emf data of the above
electrode measured against a suitable reference electrode contain contributions of both
the γ- as well as the α-phases and thus do not suffice to determine the standard Gibbs
energy of γ-NiOOH formation. Moreover, the γ-phase can be charged up to a formal
oxidation number of z(Ni) ≤ 3.6. Balej and Divisek re-evaluated four series of experimental data [70MAC], [70TYS/KSE], [80BAR/RAN], [81BAR/RAN] implying the
following model [93BAL/DIV], [93BAL/DIV2], [97BAL/DIV].
Figure V-13: α-Ni(OH)2 | γ-NiOOH system. Variation of reversible potential with the
oxidation state of Ni at 298.15 K and I = 0. Solid curve: regular model; dotted curve
[93BAL/DIV2]; experimental data: ■ [70MAC], ○ [70TYS/KSE], ▼ [80BAR/RAN], ●
[81BAR/RAN].
1.40
E/V
1.35
1.30
1.25
1.20
0.0
0.2
0.4
x (Ni(IV))
0.6
0.8
V.3 Oxygen and hydrogen compounds and complexes
119
The so-called “α-Ni(OH)2 | γ-NiOOH” system forms regular solid solutions of
Ni(II) and Ni(IV) components in the whole range of compositions between α-Ni(OH)2
and NiO2·xH2O(s), without any participation of Ni(III) oxide hydroxides. Thus, Reaction (V.51) has to be modified.
NiO2·xH2O(s) + H2(g) U α-Ni(OH)2 + xH2O(l)
(V.52)
On the basis of Reaction (V.52), an equation analogous to (V.47) can be fitted
to the experimental data, where z = 2, A ( = ∆G ex /[x·(1 – x)]) is the regular excess interaction parameter and x the mole fraction of Ni(IV). A non-linear least squares analysis
of these experimental data led to A = (2.31 ± 1.10) kJ·mol–1, E° = (1.322 ± 0.004) V and
∆ r Gmο (298.15 K) = – (255.1 ± 0.8) kJ·mol–1. For the calculation of ∆ f Gmο (NiO2·xH2O, s,
298.15 K) = (– 201.6 – 237.14 x) kJ·mol–1 the arbitrary assumption ∆ f Gmο (Ni(OH)2, α)
= ∆ f Gmο (Ni(OH)2, β) has to be made.
To summarise the situation, it can be said that unless the equilibrium potential
measurements of the system NiO2·xH2O | α-Ni(OH)2 are confirmed independently from
the structural and stoichiometric point of view, the respective results do not qualify for
thermodynamic recommendations. They can be used, however, for the calculation of
E-pH diagrams [97BAL/DIV].
V.4 Group 17 (halogen) compounds and complexes
V.4.1 Nickel halide compounds
V.4.1.1 Solid nickel fluoride, NiF2(cr)
Catalano and Stout [54STO/CAT], [55CAT/STO], [55STO/CAT] reported lowtemperature (adiabatic calorimetry) heat capacity measurements (12 K to 300 K) for this
solid, and calculated S mο (NiF2, cr, 298.15 K) = 73.60 J·K–1·mol–1. A λ-transition at
(73.22 ± 0.05) K was reported, and a similar temperature was associated with anomalous behaviour in the magnetic anisotropy of a single crystal of NiF2 [54MAT/STO]. No
heat capacity measurements for temperatures below 12 K have been found in the literature. Binford and Hebert [70BIN/HEB] reported drop calorimetry results for temperatures between 381.9 K and 1461.7 K, and calculated a “best fit” curve to match their
results to the earlier heat capacity measurements. The C pο,m results of Catalano and
Stout appear to be slightly higher than would merge smoothly with the drop calorimetry
results.
In the present review, a common function is fit to the combined results of the
two studies above 250 K. The heat capacity results are weighted using an uncertainty of
0.2 J·K–1·mol–1 for measurements between 250 and 270 K, and 0.3 J·K–1·mol–1 at higher
temperatures. For weighting purposes, the enthalpy differences are assigned an uncertainty of 0.2 kJ·mol–1. The results are constrained so that H m (T ) − H mο (298.15 K) is
V.4 Group 17 (halogen) compounds and complexes
120
zero at 298.15 K (a similar procedure appears to have been used by Binford and Herbert, but without the constraint at 298.15 K), and the resulting equation is:
[ H m (T /K) − H mο (298.15 K)] / J·mol–1 = 65.505(T /K) + 0.0076887(T /K)2 +
611603 (T /K)–1 –22265.1
(V.53)
and therefore,
–1
–1
[C pο,m ]1450K
= 65.505 + 0.015377(T /K) – 611603(T/K)–2 (V.54)
250K (NiF2, cr)/ J·K ·mol
Based on Equation (V.54), the heat capacity of NiF2(cr) at 298.15 K is 63.21
J·K–1·mol–1, whereas the value reported by Catalano and Stout [55CAT/STO] was 64.06
J·K–1·mol–1. Equation (V.54) is accepted in the present review, and as a consequence the
reported entropy value at 298.15 K [55CAT/STO] should also be adjusted slightly (by
0.08 J·K–1·mol–1).
Therefore, the selected values are:
S mο (NiF2, cr, 298.15 K) = (73.52 ± 0.40) J·K–1·mol–1
and
C pο,m (NiF2, cr, 298.15 K) = (63.21 ± 2.00) J·K–1·mol–1,
where the uncertainties are estimated values.
V.4.1.1.1
Enthalpy of formation of NiF2(cr)
Several groups have determined values for ∆ f H mο (NiF2, cr) from enthalpies of reaction
at temperatures between 500 K and 1000 K. Jellinek and Rudat [28JEL/RUD] investigated the reaction:
H2(g) + NiF2(cr) U Ni(cr) + 2HF(g)
by a transpiration method (573 K to 773 K). Domange [37DOM] studied:
H2O(g) + NiF2(cr) U NiO(cr) + 2HF(g)
from 773 K to 973 K, and Hood and Woyski [51HOO/WOY] studied:
NiCl2(cr) + 2HF(g) U NiF2(cr) + 2HCl(g)
from 477 K to 830 K. Extrapolations based on these higher temperature experiments led
to values of ∆ f H mο (NiF2, cr, 298.15 K) between – 650 kJ·mol–1 and − 670 kJ·mol−1
[67RUD/DEV].
The enthalpy of formation of NiF2(cr), was measured directly by bomb calorimetry at 298.15 K, and reported as – (657.72 ± 1.67) kJ·mol–1 by Rudzitis et al.
[67RUD/DEV]. This complicated experimental work appears to have been carried out
with great care.
In addition, Chattopadhyay and coworkers [75CHA/KAR] determined Gibbs
energies of reaction from measurements of galvanic potentials for cells of the type
V.4 Group 17 (halogen) compounds and complexes
121
Pt|M’,M’F2|CaF2|NiF2,Ni|Pt between 850 and 1000 K (M’ = Fe, Co). However, use of
these data to determine a value for ∆ f H mο (NiF2, cr, 298.15 K) would require assessment
of an enthalpy of formation value for CoF2(cr) or FeF2(cr), and this falls outside the
scope of the present review. Lofgren and McIver [66LOF/MCI] carried out measurements for the cell Mg,MgF2|CaF2|NiF2|Ni, and ∆ f H mο (NiF2, cr, 298.15 K) =
− (657.0 ± 1.4) kJ·mol–1 has been calculated from a third-law analysis of their results
(Appendix A).
As discussed in Appendix A, the early work of Jellinek and Rudat cannot be
used, but the measurements in the two later high-temperature studies [37DOM],
[51HOO/WOY] lead to inconsistent values of ∆ f H mο (NiF2, cr, 298.15 K),
− (666.13 ± 0.64) kJ·mol–1 and – (652.76 ± 0.93) kJ·mol–1, respectively. Neither of these
is in particularly good agreement with the calorimetric value of Rudzitis et al.
[67RUD/DEV].
Brunetti and Piacente [96BRU/PIA] reported an extensive set of torsion balance and Knudsen cell measurements for the determination of the enthalpy of sublimation of NiF2(cr) between 950 and 1250 K. The measurements were in reasonable agreement with, but somewhat lower than the sparse earlier sets of Farber et al.
[58FAR/MEY] and Ehlert et al. [64EHL/KEN]. Pankratz [84PAN], and Pankratz et al.
[84PAN/STU] combined the results from the earlier studies with vapour pressure results
for higher temperatures [65CAN], and the enthalpy of formation value for NiF2(cr) at
298.15 K from Rudzitis et al. [67RUD/DEV], to determine the enthalpy of formation of
NiF2(g). They also used spectroscopic data to estimate the thermal functions for the gas.
Brunetti and Piacente [96BRU/PIA] indicated that the thermodynamic data from Pankratz et al. are inconsistent with their measurements, and attributed the problem to errors
in either the enthalpy of formation (at 298.15 K) or the thermal functions of the solid
(and would show better agreement if the value of ∆ f H mο (NiF2, cr, 298.15 K) obtained
from the work of Hood and Woyski [51HOO/WOY] were used). The discrepancy reported by Brunetti and Piacente [96BRU/PIA] remains unresolved. The present reviewers could find no apparent substantive error in the experiments used to derive the different values for NiF2(cr), though confirmation of the function for C pο,m (T) at temperatures
above 300 K would be useful. Vogt [2001VOG] has reported further spectroscopic information for NiF2(g), but recalculation of the thermal functions is beyond the scope of
this review.
Some of the discrepancies may suggest problems with auxiliary data (e.g.,
∆ f H mο (NiCl2, cr, 298.15 K), ∆ f H mο (NiO, cr, 298.15 K)). Also, it is possible that there is
some undetected systematic error in the work of Domange [37DOM], as his value is
markedly more negative than any of the others. In the present review, the value:
∆ f H mο (NiF2, cr, 298.15 K) = – (657.3 ± 8.0) kJ·mol–1
is selected. It is the average of the values based on the potentiometric work of Lofgren
and McIver [66LOF/MCI] and the calorimetric work of Rudzitis et al. [67RUD/DEV],
V.4 Group 17 (halogen) compounds and complexes
122
but the uncertainty has been greatly increased to reflect the inconsistency with other
values.
Together with the selected value for S mο (NiF2, cr, 298.15 K), this selection
yields:
∆ f Gmο (NiF2, cr, 298.15 K) = – (609.9 ± 8.0) kJ·mol–1.
V.4.1.2 Solid nickel chloride, NiCl2(cr)
Anhydrous nickel chloride, NiCl2, has a layered CdCl2 structure (rhombohedral space
group R3m [63FER/BRA]) of cubic close packed chloride ions with 1/2 octahedral
holes filled by nickel ions. At T > 52 K a transition of the antiferromagnetic to the
paramagnetic state occurs [52BUS/GIA2].
V.4.1.2.1
Low-temperature heat capacity and entropy of NiCl2(cr)
Trapeznikova et al. measured the heat capacity of anhydrous Ni(II) chloride in the range
of 14 to 130 K [36TRA/SCH]. Their results were not confirmed by Busey and Giauque
who took great care to prepare high purity NiCl2 and measured its heat capacity between
15 and 300 K [52BUS/GIA2]. Kostryukova assessed the singularities of the magnetic
energy spectrum of NiCl2 by heat capacity measurements at liquid helium temperatures
[68KOS], [69KOS], [75KOS]. In Figure V-14 the smoothed values of C p ,m / T are plotted versus T for temperatures up to 30 K. Figure V-14 clearly shows that the usual extrapolation to T = 0 with Debye's T 3 law is not applicable up to 15 K.
Thus, in the present review the standard entropy of NiCl2 was re-evaluated by
integration of the original and smoothed C p ,m values listed in [52BUS/GIA2] and
[75KOS] respectively. Two different linear spline functions, valid for the temperature
range below and above the Néel temperature (TN = 52.32 K) were fitted to these data,
shown in Figure V-15. The sum of these integrals resulted in the selected value:
S mο (NiCl2, cr, 298.15 K) = (98.14 ± 0.30) J·mol–1·K–1.
This value is slightly higher than the one given in [52BUS/GIA2], which may
be due to the more realistic extrapolation to T = 0 on the basis of Kostryukova's data
[75KOS] as well as to the influence of different fitting functions employed by
[52BUS/GIA2] (unknown) and in the present review. The uncertainty selected is greater
than the difference arising from using the more complete set of fitting functions.
The standard heat capacity derived by interpolation of the fitted C p ,m –
functions is essentially equal to the value given by Busey and Giauque (17.13
cal·mol−1·K–1) who stated that near 300 K heat leaks may have introduced an error of
several tenths of a percent into their C p ,m –values. Thus the following value as well as
its uncertainty are selected:
C pο,m (NiCl2, cr, 298.15K) = (71.67 ± 0.30) J·mol–1·K–1.
(V.55)
V.4 Group 17 (halogen) compounds and complexes
123
Figure V-14: Molar heat capacity of anhydrous NiCl2 between 2 and 30 K.
–– –– : smoothed values of [75KOS], –– –– : smoothed values of [52BUS/GIA2]
Dotted line: T 3–law up to 2 K, dash line T 3–law up to 15 K, solid line: C p ,m / T =
0.0011715·T 2 + 0.0091211·T up to 3.2 K [75KOS].
0.25
0.15
0.10
p, m
–1−1
–2 −1 −2–1
(C°
p ,m ·T
(C°
· T ))// J·K
J·mol·mol
·K
0.20
0.05
0.00
0
5
10
15
20
25
30
35
T/K
Figure V-15: Heat capacity of anhydrous NiCl2(cr). : experimental data from
[52BUS/GIA2]. : smoothed data of [75KOS]. Dash line: linear spline T > TN, dotted
line: linear spline T < TN
0.5
0.3
–1
–2
−1
1
−2
(C°
J·K −·mol
(Cp ,m ··TT )) // J·mol
·K
–1
0.4
p,m
0.2
0.1
0.0
0
50
100
150
T/K
200
250
300
350
V.4 Group 17 (halogen) compounds and complexes
124
V.4.1.2.2
The high-temperature heat capacity of NiCl2(cr)
Coughlin conducted high temperature heat content measurements of nickel chloride
over the temperature range from 298.15 to 1336 K [51COU]. The experimentally determined heat contents listed in [51COU] were fitted to the following function up to the
highest temperature before any premelting was observed:
[ H m (T ) − H mο (298.15 K)] / J·mol−1 = a·((T / K) 2 − 298.152 ) +
b·((K / T ) − 298.15−1 ) +
c·((T / K) − 298.15)
(V.56)
With the standard heat capacity of NiCl2 according to Equation (V.55) as a
constraint, c can be eliminated from Equation (V.56).
 ∂ ( H m (T ) − H mο (298.15)) 
−2
C pο,m (T ) = 
 = 2·a·T − b·T + c
T
∂

p
c = C pο,m (298.15) − 2·a·(298.15) − b·(298.15) −2
(V.57)
From a least squares analysis of the ( H m (T ) − H mο (298.15)) data listed in
[51COU] a C pο,m (T ) –function was obtained which is numerically slightly different from
the one given by Coughlin, but the calculated C pο,m –values differ only marginally. For
the sake of consistency with the selected value of C pο,m (NiCl2, cr, 298.15 K) the following C pο,m (T ) –function is selected:
–2
[C pο,m ]1281
(T / K)
298 (NiCl2, cr) = 73.7949 + 1.22269×10
5
–2
– 5.12775×10 · (T / K) /J·mol–1·K–1
V.4.1.2.3
(V.58)
Enthalpy of formation of NiCl2(cr)
The enthalpy of formation of NiCl2 can be determined by a third-law analysis of the
equilibrium studies on the reaction:
NiCl2(cr) + H2(g) U Ni(cr) + 2 HCl(g)
(V.59)
carried out by Berger and Crut, [21BER/CRU], Crut, [24CRU], Jellinek and Uloth,
[26JEL/ULO], Sano, [37SAN2], [53SAN], Busey and Giauque, [53BUS/GIA], as well
as Shchukarev et al. [54SHC/TOL]. An inspection of Figure V-16 reveals that the data
of [53BUS/GIA], [53SAN], and [54SHC/TOL] appear to be the most reliable and best
suited for this analysis. This was confirmed by linear fits through the data pairs T,
− RT·ln(Kp /bar) which resulted in lower standard errors for [53SAN], [53BUS/GIA]
and [54SHC/TOL] than for [24CRU] or [26JEL/ULO].
An alternative method consists in subjecting high-purity nickel metal to chlorination in a calorimetric bomb according to reaction:
Ni(cr) + Cl2(g) U NiCl2 (cr)
(V.60)
The advantage of the latter method is that the standard enthalpy of formation
V.4 Group 17 (halogen) compounds and complexes
125
∆ r H mο (NiCl2, cr, 298.15 K) can be determined directly at the reference temperature,
[84LAV/TIM].
A third independent method has been developed by Efimov et al. and is described in Appendix A [88EFI/EVD].
Figure V-16: ∆ r Gmο –function of Reaction (V.59). ∆ r Gmο = – RT·lnK°, linear fit of
experimental T; experimental data from:
[53BUS/GIA],
[54SHC/TOL],
[26JEL/ULO],
[21BER/CRU] [24CRU]. ∆ r Gmο data: solid
[37SAN2] [53SAN],
line [53BUS/GIA], dotted line [53SAN], dashed line [54SHC/TOL]
30
NiCl2(cr) + H2(g) U Ni(cr) + 2 HCl(g)
G° / kJ·mol–1
r m m/ kJ·mol
∆∆rG°
−1
20
10
0
-10
-20
550
600
650
700
750
800
850
T/K
The third-law analyses of the equilibrium constants obtained by Busey and Giauque [53BUS/GIA], Sano [53SAN] and Shchukarev et al. [54SHC/TOL] for reaction
(V.59) are summarised in Figure V-17. A combination of the standard enthalpy of
HCl(g) formation recommended by CODATA [89COX/WAG] with the reaction enthalpies of Figure V-17 resulted in ∆ f H mο (NiCl2, cr, 298.15 K) / kJ·mol–1 =
− (305.01 ± 0.17) [53BUS/GIA]; − (305.5 ± 2.4) [54SHC/TOL], – (307.07 ± 0.56)
[53SAN]. Busey and Giauque's data show the least uncertainty. The results of Shchukarev et al. led to a similar value with a significantly enhanced uncertainty. Moreover,
Busey and Giauque's as well as Sano's values scatter around the mean, whereas Shchukarev et al.'s results increase with increasing temperature.
V.4 Group 17 (halogen) compounds and complexes
126
In principle equilibrium constants of other reactions such as:
NiO(cr) + 2 HCl(g) U NiCl2(cr) + H2O(g)
can also be subjected to the third-law method, see Appendix A [77KAS/SAK].
Figure V-17: Third-law analysis of data for Reaction (V.59). Experimental data from
[53BUS/GIA],
[54SHC/TOL], and
[53SAN]. Solid line: ∆ r H mο / kJ·mol–1 =
(120.41 ± 0.10) [53BUS/GIA], dotted line: ∆ r H mο = (122.45 ± 0.56) kJ·mol–1, [53SAN],
dash line: ∆ r H mο = (120.99 ± 2.4) kJ·mol–1 [54SHC/TOL], dash-dotted line: linear
regression of [54SHC/TOL] data.
123
121
r
m
–1−1
∆∆rH°
m / kJ·mol
H°
/ kJ·mol
122
120
NiCl2(cr) + H2(g) U Ni(cr) + 2 HCl(g)
119
550
600
650
700
750
800
850
T/K
Lavut et al. [84LAV/TIM] determined the standard enthalpy of NiCl2 formation very accurately ∆ f H mο (NiCl2, cr, 298.15 K) = – (304.78 ± 0.16) kJ·mol–1. This result overlaps with the value obtained from the third-law analysis of Busey and Giauque's data. Consequently the mean value of these investigations was selected in this
review:
∆ f H mο (NiCl2, cr, 298.15 K) = – (304.90 ± 0.11) kJ·mol–1.
(V.61)
V.4 Group 17 (halogen) compounds and complexes
V.4.1.2.4
127
Gibbs energy of formation of NiCl2(cr)
The Gibbs energy of formation of NiCl2(cr) can be calculated as a function of temperature using the selected values of ∆ f H mο , S mο , C pο,m for NiCl2(cr) and for Ni(cr) as well
as the respective CODATA values for Cl2(g) [89COX/WAG], adopted (Chapter VI) as
NEA TDB auxiliary data (see Figure V-18).
Gee and Shelton measured the emf of solid electrolyte cells and thus obtained,
for example, ∆ r Gmο of the displacement reaction (V.62) directly [76GEE/SHE].
Fe(cr) + NiCl2(cr) U FeCl2(cr) + Ni(cr)
(V.62)
The Gibbs energy of formation of FeCl2 appeared to be well established and
thus the Fe, FeCl2 electrode was taken as the reference electrode. Figure V-18 shows
that values measured in this manner and calculated ∆ f Gmο (NiCl2, cr) values agree with
each other satisfactorily. This is considered to be a strong argument in favour of the
selected thermodynamic quantities for NiCl2(cr) and Ni(cr).
Figure V-18: Gibbs energy of formation of NiCl2(cr) versus temperature. Values
(straight line) and uncertainty (dotted lines) derived from [76GEE/SHE]. Dashed line
calculated from the selected values of ∆ f H mο , S mο , C pο,m .
-160
-170
Ni(cr) + Cl2(g) U NiCl2(cr)
–1
∆mfG°
(NiCl
, cr)
/kJ·mol
∆fG°
(NiCl
/ kJ·mol
2, cr)
m
2
−1
-180
-190
-200
-210
-220
-230
-240
-250
500
600
700
T/K
800
900
V.4 Group 17 (halogen) compounds and complexes
128
V.4.1.3 Hydrated NiCl2 solids
Thermodynamic data have been reported for the nickel chloride hexahydrate, tetrahydrate and dihydrate, and the existence of a heptahydrate has been reported at low temperatures (below – 30°C) [33BOY]. At 0.1 MPa, in contact with saturated aqueous solutions, the hexahydrate was reported [16DER/YNG] to be the stable form to approximately 36.25°C, and dehydration of the tetrahydrate to the dihydrate occurs above 60°C
[33BOY]. The lower hydrates can also be prepared by equilibration with aqueous HCl
solutions at room temperature [23FOO].
V.4.1.3.1
Gibbs energy of formation of NiCl2·6H2O
Many solubility measurements for NiCl2·6H2O(cr) in water have been reported
[33BOY], [37PEA/ECK], [79OIK], [79OJK/MAK], [79PET/SHE], [79PET/SHE2],
[80PET/SHE], [83KIM/IMA], [85FIL/CHA], [86FIL/CHA], [87RAR3], [89LIL/TEP].
NiCl2·6H2O(cr) U Ni2+ + 2Cl–+ 6H2O(l)
(V.63)
Rard [92RAR2] has done an excellent assessment of the solubility of
NiCl2·6H2O and the associated activity and osmotic coefficients of the saturated solution at 298.15 K. Based primarily on results from [37PEA/ECK], [87RAR3] and
[92RAR2], Rard calculated ((4.9155 ± 0.0022) mol·kg–1),
K ο (V.63) = (1108 ± 37)
and
∆ r Gmο (V.63) = – (17.38 ± 0.08) kJ·mol–1.
Rard’s value is selected in the present review and, with the auxiliary data for
Cl– (Chapter IV) and the selected value for ∆ f Gmο (Ni2+, 298.15 K) (Table III.1), the
value:
∆ f Gmο (NiCl2·6H2O, cr) = – (1713.7 ± 0.8) kJ·mol–1
is calculated.
V.4.1.3.2
Enthalpy of formation of NiCl2·6H2O
Surprisingly, the only reported direct measurement of the heat of solution of the hexahydrate in water appears to be that of Thomsen [1883THO] (4.85 kJ·mol–1 at 18.5°C
[20BUR]). Archer [99ARC] has reported that an attempt to correct this value to 25°C
and infinite dilution (cf. Appendix A) resulted in the value 0.14 kJ·mol–1. Ashcroft
[70ASH] reported the heat of solution of NiCl2·6H2O(cr) in 1 M HCl(aq) as
(6.10 ± 0.22) kJ·mol–1, and Gurrieri et al., [73GUR/CAL] reported (7.00 ± 0.13)
kJ·mol−1. These two values do not agree within the authors’ (2σ) uncertainties, even
though the final solution concentrations of the nickel salt were apparently quite similar
in the experiments. Kakolowicz and Giera [75KAK/GIE] reported the heat of solution
of NiCl2·6H2O(cr) in 4.36 M HCl(aq) as (21.32 ± 0.13) kJ·mol–1.
V.4 Group 17 (halogen) compounds and complexes
129
The final nickel concentrations in Thomsen’s experiments were quite high (approximately 0.14 m), thus applying a proper heat of dilution correction to determine
∆ r H mο (V.63) is problematic [99ARC]. The final nickel concentrations in the experiments of Ashcroft [70ASH] and Gurrieri et al. [73GUR/CAL] were about a factor of
five lower, but to use these values a correction for transfer of the aqueous species from
the dissolution medium to water is needed.
The enthalpy of solution measurements of Manin and Korolev [95MAN/KOR]
for dissolution of anhydrous nickel chloride in HCl(aq) indicate that if the final solutions have nickel chloride concentrations of less than 0.01 M, the enthalpy of solution of
anhydrous nickel chloride is approximately 4 kJ·mol–1 more positive than in water. This
is in semi-quantitative agreement with results of the other measurements [1883THO],
[70ASH], [73GUR/CAL]. Manin and Korolev [95MAN/KOR] also reported heats of
solution of NiCl2(cr) in water as a function of the final solution concentration (for
0.0042 to 0.030 m). The variation in these values could be applied equally well to the
heats of solution of the hydrated salt, but the experimental concentration range is not
adequate to determine heats of dilution to I = 0 to better than ± 1 kJ·mol–1. As discussed
in Appendix A, using a reasonable value for the enthalpy of dilution, the heat of solution of NiCl2·6H2O(cr) to infinite dilution in water at 298.15 K can be estimated as
(0.64 ± 1.80) kJ·mol–1. From this, ∆ f H mο (NiCl2·6H2O, cr) = – (2104.41 ± 2.01) kJ·mol–1
can be calculated.
Another approach is to compare the heats of solution of the anhydrous salt and
the hydrate in the same medium (with appropriate minor corrections for changes in the
heat of formation of water in the medium). However, except for Thomsen [1883THO],
none of these researchers reported enthalpy of solution values for the anhydrous solid,
NiCl2(cr) and the hexahydrate in the same medium. As discussed in Appendix A, the
heat of hydration of NiCl2(cr) to NiCl2·6H2O(cr) at 298.15 K can be estimated as
− (85.06 ± 1.80) kJ·mol–1, and ∆ f H mο (NiCl2·6H2O, cr) = – (2104.9 ± 1.8) kJ·mol–1.
In the present review, the weighted average of the two values based on the
measurements of Thomsen [1883THO] is selected:
∆ f H mο (NiCl2·6H2O, cr, 298.15 K) = – (2104.7 ± 1.8) kJ·mol–1,
but with an uncertainty somewhat larger than the statistical uncertainty because the values are based, in part, on the same measurement.
V.4.1.3.3
The entropy and heat capacity of NiCl2·6H2O(cr)
Based on the selected values of ∆ f Gmο (NiCl2·6H2O,
∆ f H mο (NiCl2·6H2O, cr, 298.15 K), the selected entropy is:
cr,
298.15 K)
and
S mο (NiCl2·6H2O, cr, 298.15 K) = (341.0 ± 6.7) J·K–1·mol–1.
Robinson and Friedberg [60ROB/FRI] reported measurements of the heat capacity of NiCl2·6H2O(cr) between 1.6 K and 20.0 K, and found that the Néel tempera-
130
V.4 Group 17 (halogen) compounds and complexes
ture is at 5.34 K [73HAM/FRI]. Donaldson and Edmonds [65DON/EDM] extended the
measurements to lower temperatures, with measurements between 0.7 K and 4 K. Similar measurements were reported by Okaji and Watanabe [76OKA/WAT]. Although Ko
and Hepler [63KO/HEP] reported S mο (NiCl2·6H2O, cr, 298.15 K) = 344.3 kJ·mol–1,
based on unpublished C p ,m data of Friedberg, the primary C p ,m data apparently were
never published. In the present review, no heat capacity value at 298.15 K is selected for
NiCl2·6H2O(cr).
V.4.1.3.4
Gibbs energy and enthalpy of formation of nickel chloride
tetrahydrate
Measurements of water vapour pressures of mixtures of nickel chloride hydrates
[16DER/YNG] and [40BEL], and over pure solids [93UVA/TIM] have been reported.
Water vapour pressure values reported for the equilibrium:
NiCl2·6H2O(cr) U NiCl2·4H2O(cr) + 2H2O(g)
(V.64)
by Derby and Yngve [16DER/YNG] and the equation of Bell [40BEL] between 20 and
36°C are in good agreement, and may even be based on the same experimental results
(cf. Appendix A), whereas Uvaliev et al. [93UVA/TIM] reported slightly lower values.
The latter (isopiestic) work was focussed on determining transition pressures at constant
temperature (25 and 35°C) rather than on the temperature dependence of the water vapour pressure. The equilibrium water vapour pressure at 298.15 K is (1.30 ± 0.15) kPa,
based on the average of the values reported by [16DER/YNG] (1.40 kPa) and Uvaliev et
al. [93UVA/TIM] (1.19 kPa). The selected value is:
∆ r Gmο (V.64) = (21.56 ± 0.60) kJ·mol–1.
The uncertainty has been assigned to span the uncertainties in these inconsistent results.
The enthalpy of reaction, determined only from the results of Derby and Yngve
[16DER/YNG], is (102.4 ± 1.8) kJ·mol–1; and (107.8 ± 5.8) kJ·mol–1 if the values from
[93UVA/TIM] are included. If the water vapour pressure is fixed at the selected value at
298.15 K, the fitted slope of log10 pH2 O against (T/K)–1 leads to a calculated enthalpy of
reaction of (111.2 ± 5.6) kJ·mol–1. In the present review a median value spanning all
these values is selected:
∆ r H mο (V.64) = (107 ± 9) kJ·mol–1.
Using the values selected above for the hexahydrate, and auxiliary data from Chapter
IV, yields:
∆ f Gmο (NiCl2·4H2O, cr, 298.15 K) = – (1234.9 ± 1.0) kJ·mol–1
and
∆ f H mο (NiCl2·4H2O, cr, 298.15 K) = – (1514.0 ± 9.2) kJ·mol–1.
V.4 Group 17 (halogen) compounds and complexes
131
From these and auxiliary values from Chapter IV,
S mο (NiCl2·4H2O, cr, 298.15 K) = (250 ± 31) J·K–1·mol–1
is calculated.
Bell [40BEL] reported on measurements of the vapour pressure of D2O over
mixtures of the tetradeuterate and hexadeuterate, and it was found that the ratio of equilibrium values of pH2 O / pD2 O was greater than 1.0 near 25°C.
V.4.1.3.5
The entropy and heat capacity of nickel chloride dihydrate
Polgar et al. [72POL/HER] measured the specific heat of NiCl2·2H2O(cr) for temperatures between 1.2 and 24.5 K. Peaks were observed at 6.31 and 7.26 K, and at lower
temperatures the heat capacity values are not a simple function of T 3. The authors estimated that the contribution to the molar entropy between 0 and 1.2 K is 0.03 R. Using
an adiabatic calorimeter, Juraitis et al. [90JUR/DOM] measured the heat capacity of the
solid between 80 and 281 K, with emphasis on the effects of the structural transition
near 220 K. It is not clear from these results (only reported graphically) whether the
results from 250 to 281 K can be extrapolated smoothly to 298.15 K.
A series of polynomials were fitted to the values of C p ,m / T over the entire
temperature range from 1.2 K to 281 K. The uncertainty in the extrapolation below 1.5
to 0 K is of the order of ± 0.2 J·K–1·mol–1. A single simple polynomial was fitted to the
combined results from 20 to 25 K [72POL/HER] and 80 to 100 K [90JUR/DOM]. In the
present review, it is assumed that there are no further anomalies in this temperature
range, and the uncertainty contribution to S mο (NiCl2·2H2O, cr) over this range is estimated as ± 2.0 J·K–1·mol–1. The uncertainties in the set of C pο,m values above 240 K (see
Appendix A) independently introduce an uncertainty of ± 3.0 J·K−1·mol–1 in the calculated value of S mο . In the present review,
S mο (NiCl2·2H2O, cr, 298.15 K) = (186.2 ± 3.6) J·K–1·mol–1
is selected, where the uncertainty is an estimate. From the work of Juraitis et al.
[90JUR/DOM] (see Appendix A),
C pο,m (NiCl2·2H2O, cr, 298.15 K) = (230 ± 15) J·K–1·mol–1
is selected in the present review. Again, the uncertainty is estimated.
V.4.1.3.6
Gibbs energy and enthalpy of formation of nickel chloride dihydrate
Derby and Yngve [16DER/YNG] and Uvaliev et al. [93UVA/TIM] both reported water
vapour pressure measurements for the dehydration of nickel chloride tetrahydrate to the
dihydrate.
NiCl2·4H2O(cr) U NiCl2·2H2O(cr) + 2 H2O(l)
(V.65)
132
V.4 Group 17 (halogen) compounds and complexes
The values of log10 pH2 O from the earlier study do vary monotonically with
1/T, though they are in fair agreement with the results from Uvaliev et al.
[93UVA/TIM] at 25 and 35°C. In the present review, the average of the discordant values from the two studies at 298.15 K is used to calculate a value for ∆ r Gmο (V.65), but
no value for ∆ r H mο (V.65) is selected,
∆ r Gmο (V.65) = (23.4 ± 0.7) kJ·mol–1.
Using the values selected above for the tetrahydrate, and auxiliary data from
Chapter IV,
∆ f Gmο (NiCl2·2H2O, cr, 298.15 K) = – (754.4 ± 1.3) kJ·mol–1.
Based on this, auxiliary data from Chapter IV, and S mο (NiCl2·2H2O, cr,
298.15 K) from Section V.4.1.3.5,
∆ f H mο (NiCl2·2H2O, cr, 298.15 K) = – (913.4 ± 1.7) kJ·mol–1.
V.4.1.4 Solid nickel bromide, NiBr2(cr)
There are two different crystallographic forms of paramagnetic NiBr2(cr) [34KET].
Samples prepared by sublimation, or by melting and slow cooling, have a CdCl2 structure, rhombohedral space group R 3m [33BIJ/NIE], [34KET], [80ADA/BIL]. Samples
obtained by low-temperature dehydration of the hydrated salts or by recrystallisation of
these solids from organic solvents have a “Wechselstruktur” [34KET]. Day et al.
[76DAY/DIN] describe the structure as a disordered intergrowth of the CdCl2 and CdI2
lattices [33BIJ/NIE]. Both forms have been used in experiments to determine the
chemical thermodynamic properties of anhydrous nickel bromide, often without a recognition that two forms can exist. NiBr2(cr) orders to an antiferromagnetic form below
52 K, and to a helimagnetic structure below 23 K [83KAT/SUG].
V.4.1.4.1
Low-temperature heat capacity and entropy of NiBr2(cr)
There are two sets of low-temperature (8 to 300 K) heat capacity (adiabatic calorimetry)
measurements [78STU/FER], [82WHI/STA] that are in good agreement between 30 and
250 K (as discussed in Appendix A). Both sets of results show a small excess heat capacity between 40 and 55 K, near the Néel temperature. White and Staveley
[82WHI/STA] also found a small, sharp peak near 20 K, and apparently this is associated with the transition from an antiferromagnetic structure to an incommensurate phase
with helimagnetic ordering [76DAY/DIN], [81ADA/BIL]. No corresponding peak was
found in the earlier study [78STU/FER], and the reported heat capacities below 15 K
decrease much more slowly with decreasing temperature than suggested by the results
of [78STU/FER]. Unlike the results of White and Staveley, the values reported by Stuve
et al. below 18 K do not fit well to a T 3 relationship (see Figure V-19). Above 250 K,
the heat capacity values reported by White and Staveley increase rapidly with increasing
temperature, and the authors suggested [82WHI/STA] that the sample may have been
V.4 Group 17 (halogen) compounds and complexes
133
contaminated by moisture (at 298.15 K, their value of C pο,m (NiBr2, cr) is approximately
83 J·K–1·mol–1, as opposed to 75.4 J·K–1·mol–1 from Stuve et al.).
Stuve et al. [78STU/FER] integrated their heat capacity values to obtain
S mο (NiBr2, cr, 298.15 K) = (122.42 ± 0.25) J·K–1·mol–1. In the present review, the experimental C p ,m (T) values reported in both papers were integrated. The results of White
and Staveley [82WHI/STA] are accepted to 250 K (an entropy contribution of 108.23
J·K−1·mol–1). The contribution to the entropy from 0 to 8 K (0.09 J·K–1·mol–1) was estimated by assuming C pο,m /T varies as T 2 (Figure V-19), though there are no measurements to support this assumption. The results of Stuve et al. [78STU/FER] are accepted
between 250 K and 298.15 K (an entropy contribution of 13.14 kJ·mol–1). The sum of
these contributions provides the selected value:
S mο (NiBr2, cr, 298.15 K) = (121.4 ± 1.0) J·K–1·mol–1 .
The uncertainty is an estimate based on the spread of the entropy contributions
from the two studies in different temperature ranges, and considering that there may be
an unknown contribution from the extrapolation from 8 to 0 K.
The reported C pο,m (NiBr2, cr, 298.15 K) value of Stuve et al. [78STU/FER]
C pο,m (NiBr2, cr, 298.15 K) = (75.40 ± 1.00) J·K–1·mol–1
is selected in the present review with a markedly increased uncertainty.
Figure V-19: Values from the heat capacity studies of Stuve et al. [78STU/FER] and
White and Staveley [82WHI/STA]. The curves (from bottom to top) are based on the
equation C pο,m (T) = bT 2 for b = 0.000500, 0.000750, 0.001000, 0.001350 and 0.002000.
0.7
[78STU/FER]
[82WHI/STA]
0.5
0.4
0.3
p
-2–2·mol
-1–1
(C°p ,mC·T/T–1)/ /J·K
J·K ·mol
0.6
0.2
0.1
0.0
0
2
4
6
8
10
T/T
K/ K
12
14
16
18
20
V.4 Group 17 (halogen) compounds and complexes
134
V.4.1.4.2
The high-temperature heat capacity of NiBr2(cr)
The fitted heat capacity equation reported by the same authors [78STU/FER] reproduces all their high-temperature experimental H m (T ) − H mο (298.15) values to better
than 0.5% and is fixed to the 298.15 K value from the low-temperature measurements.
The values from the equation deviate slightly from the experimental values below
298.15 K. If this heat capacity equation is used with the high-temperature equilibrium
measurements of Shchukarev et al. [54SHC/TOL], the derived values of ∆ f H mο (NiBr2,
cr, 298.15 K) vary with the measurement temperature (see the Appendix A discussion
of [54SHC/TOL]). In the present review, this drift has been attributed to experimental
difficulties in the work of Shchukarev et al., but it might also indicate problems with the
drop calorimetry work of Stuve et al. It is also not clear whether the more stable (or the
same) crystalline form of NiBr2 was used for all the measurements [34KET],
[54SHC/TOL]. Nevertheless, in the present review, the high-temperature heat capacity
equation of Stuve et al. is selected:
–1
–1
[C pο,m ]1200K
= 72.3163 + 0.014627T/K – 113805 (T/K)–2.
298.15K (NiBr2, cr)/J·K ·mol
V.4.1.4.3
Enthalpy of formation of NiBr2(cr)
Stuve et al. [78STU/FER] measured the heats of heat of solution of samples of
NiBr2(cr), (possibly the “Wechselstruktur” form), NiSO4(cr) and H2SO4·6H2O(l) in 4.36
m HCl(aq) and, from these measurements, as described in Appendix A, ∆ f H mο (NiBr2,
cr, 298.15 K) = − (211.214 ± 1.752) kJ·mol–1. Evdokimova and Efimov [89EVD/EFI]
measured the heats of solution of Br(l), nickel and NiBr2(cr), (probably the “Wechselstruktur” form) in an aqueous KBr, Br2, HBr mixture. From these measurements,
∆ f H mο (NiBr2, cr) = − (211.93 ± 0.90) kJ·mol–1, 2σ uncertainties (Appendix A).
Several groups, Crut [24CRU], Jellinek and Uloth [26JEL/ULO2] and Shchukarev et al. [54SHC/TOL] have measured the equilibrium constant for the reaction:
NiBr2(cr) + H2(g) U Ni(cr) + 2HBr(g).
at temperatures between 623 K and 928 K. The most extensive, and least scattered set of
values is that of [54SHC/TOL]. As discussed in Appendix A, these measurements can
be used to calculate ∆ f H mο (NiBr2, cr) = – (215.02 ± 0.91) kJ·mol–1. Values based on the
results of [24CRU] and [54SHC/TOL] are in rough agreement (– 214 to – 217
J·K−1·mol–1). A value of ∆ f H mο (NiBr2, cr) based on the sparse enthalpy of solution
measurements of Paoletti [65PAO] would depend on use of the selected value of
∆ f H mο (Ni2+, 298.15 K) (also see the discussion of [90EFI/FUR] in Section V.2.1.2 and
Appendix A).
It is unclear whether the enthalpy of formation values based on the measurements of Stuve et al., Shchukarev et al. and Evdokimova and Efimov are for the identical solid. The solid of Evdokimova and Efimov was almost certainly the “Wechselstruktur” form, and that of Shchukarev et al. was possibly the rhombohedral form, at least for
V.4 Group 17 (halogen) compounds and complexes
135
some of their measurements. This seems to be reflected in the less negative value of
∆ f H mο (NiBr2, cr) found by Evdokimova and Efimov. In the absence of other information, the average of the more precise values of Evdokimova and Efimov and of Shchukarev et al. is selected, and the uncertainty is selected to span the uncertainties of the
these measurements:
∆ f H mο (NiBr2, cr, 298.15 K) = – (213.5 ± 2.4) kJ·mol–1.
This value results in the following derived Gibbs energy of formation:
∆ f Gmο (NiBr2, cr, 298.15 K) = – (195.4 ± 2.4) kJ·mol–1.
Based on this enthalpy of formation value and the selected value for
∆ f H mο (Ni2+, 298.15 K), the enthalpy of solution of NiBr2(cr) in water (at infinite dilution) is – (83.9 ± 2.6) kJ·mol–1. This value is less negative than, but in marginal agreement with the – (86.18 ± 0.40) kJ·mol–1 reported by [90EFI/FUR], and is also in agreement, within the uncertainty limits, with the rather poorly documented result of Paoletti
[65PAO] (– 82.1 kJ·mol–1).
V.4.1.5 Hydrated solid, NiBr2·xH2O
Although several hydrated compounds have been reported, information on wellcharacterised samples is sparse. The isolation of the hexahydrate (over a desiccant at
room temperature) has been reported [66GME], [77BHA/CAR]. Conversion of the
hexahydrate to the trihydrate may occur near 30°C.
The low-temperature heat capacity of NiBr2·6H2O was reported by Spence et
al. [59SPE/FOR] (between 4.0 and 13 K), and by Bhatia et al. [77BHA/CAR] (between
1.5 and 30 K). There is a λ–transition in the heat capacity measurements. Spence et al.
indicate that the transition is at 6.5 K, whereas Bhatia et al. report that the transition is
at (8.30 ± 0.02) K, the same temperature as the paramagnetic-antiferromagnetic transition found in their magnetic susceptibility measurements. The two groups report similar
values for the entropy contribution of the transition (9.62 J·K−1·mol–1 [59SPE/FOR],
(9.41 ± 0.21) J·K–1·mol–1 [77BHA/CAR]). Because no values could be found for the
heat capacity of NiBr2·6H2O for temperatures between 30 and 298.15 K, no value for
S mο (NiBr2·6H2O, cr) is selected.
A value for ∆ f H mο (NiBr2·3H2O) has been reported in several standard sets of
tables [36BIC/ROS], [52ROS/WAG], [82WAG/EVA]. This value has been traced to
enthalpy of solution measurements of Crut [24CRU], [24CRU2] (from – 79.1 kJ·mol–1
for the enthalpy of hydration of the anhydrous salt [24CRU]). However, the stoichiometry of the hydrated salt is not specified, and there are ambiguities in the reported values
(see the discussion in Appendix A). In the present review, no chemical thermodynamic
values are selected for the nickel bromide hydrates for 298.15 K.
V.4 Group 17 (halogen) compounds and complexes
136
V.4.1.6 Solid nickel iodide, NiI2(cr)
V.4.1.6.1
Entropy of NiI2(cr)
Worswick et al. [74WOR/COW] reported low-temperature (adiabatic calorimetry) heat
capacity measurements (10 K to 300 K) for this solid, and calculated S mο (NiI2, cr) =
138.7 J·K–1·mol–1. A λ-transition at 58.9 K was reported. There are only limited confirmatory heat capacity measurements in the literature. Based on magnetic susceptibility
measurements, Van Uitert et al. [65UIT/WIL] reported a λ-transition as occurring at
approximately 75 K. Billerey et al. [77BIL/TER] found heat-capacity transitions at
59.5 K and near 76 K, and these were later identified by Kuindersma et al.
[81KUI/SAN] as a structural transition and a magnetic phase transition, respectively.
The raw data from Billerey et al. are not available, but the contributions to the entropy
from the two transitions are estimated to be 0.40 and 0.37 J·K–1·mol–1 (see Appendix A).
Examination of the data from Worswick et al. indicates that their value for S mο (NiI2, cr)
included a contribution of 0.34 J·K–1·mol–1 as an excess entropy in the region 50 K to
85 K. There are no reported heat capacity values at temperatures below 10 K, though
the magnetic behaviour of the solid is very complex at low temperatures [81KUI/SAN],
[90PAS/TAY], [92FRE/FAL]. The entropy, adjusted for the contribution of the transition at 76 K, is 139.1 J·K–1·mol–1, and this value is accepted in the present review. A
rather large uncertainty of ± 1.0 J·K–1·mol–1 is assigned to the selected value:
S mο (NiI2, cr, 298.15 K) = (139.1 ± 1.0) J·K–1·mol–1 .
V.4.1.6.2
Heat capacity of NiI2(cr)
Worswick et al. [74WOR/COW] reported low-temperature (adiabatic calorimetry) heat
capacity measurements (10 to 300 K) and differential scanning calorimetry (DSC) results from 300 to 550 K. The equation (V.66) (200 to 550 K) is not in the standard form
generally used for solids (e.g., [93KUB/ALC], p. 166, Equation (116)), however, the
original DSC data are unavailable.
–1
–1
[C pο,m ]550K
= 65.5 + 0.0515 T/K – 0.0000397 (T/K)2
200K (NiI2, cr) / J·K ·mol
(V.66)
Equation (V.67), in a more standard form, was calculated to fit the adiabatic
calorimetry results from 201.55 to 297.39 K and values from equation (V.66) at higher
temperatures (as discussed in [74WOR/COW]). Equation (V.67) closely reproduces the
values from the authors’ equation (within 0.5 J·K–1·mol–1 between 300 and 550 K, and
within the experimental uncertainties at lower temperatures). Equation (V.67) is accepted in the present review for the specified temperature range, but this equation
should not be used at higher temperatures. NiI2(cr) is known to decompose at temperatures slightly above 700 K.
–1
–1
[C pο,m ]550K
= 73.5775 + 0.016425T – 1.057753×105 (T/K)–2 (V.67)
200K (NiI2, cr) / J·K ·mol
The value of C pο,m (NiI2, cr, 298.15 K) based on equation (V.67),
V.4 Group 17 (halogen) compounds and complexes
137
C pο,m (NiI2, cr, 298.15 K) = (77.28 ± 1.00) J·K–1·mol–1
is selected in the present review with a markedly increased uncertainty (see Appendix A).
V.4.1.6.3
Enthalpy of formation of NiI2(cr)
Values for the enthalpy of formation of NiI2(cr) have been reported (or can be calculated) based on enthalpy of solution measurements [65PAO/SAB], [90EFI/EVD] and
from high temperature equilibrium measurements [80OPP/TOS]. The work of Oppermann and Toschew [80OPP/TOS] considers many of the difficulties in determining
values from the rather complex Ni-I system at higher temperatures, and the problems in
doing a proper extrapolation to determine thermodynamic values at 298.15 K. However,
as discussed in Appendix A, this process can only be carried out within very large uncertainty bounds, by using the selected C pο,m (T ) equation for NiI2(cr), and drawing tentative conclusions related to the thermodynamic functions for NiI2(g) and Ni2I4(g). The
earlier measurements of Jellinek and Uloth [26JEL/ULO2] suffer from many of the
same problems, but are also too sparse to be properly re-interpreted. The value reported
by Bartovská et al. [74BAR/CER] is based, in part, on estimated values.
A value of ∆ f H mο (NiI2, cr) based on the enthalpy of solution measurements of
Paoletti et al. [65PAO/SAB] or Efimov and Furkalyuk [90EFI/FUR] would depend on
use of the selected value of ∆ f H mο (Ni2+, 298.15 K). However, experimental results reported by Efimov et al. [88EFI/EVD], Efimov and Evdokimova [90EFI/EVD] provide a
value of ∆ f H mο (NiI2, cr, 298.15 K) that is independent of the value for ∆ f H mο (Ni2+,
298.15 K). From their work (see Appendix A), the value,
∆ f H mο (NiI2, cr, 298.15 K) = – (96.42 ± 0.84) kJ·mol–1
is selected in the present review. This value is in agreement with – (91 ± 8) kJ·mol–1
based on Oppermann and Toschew [80OPP/TOS] (see Appendix A). This value results
in the following derived Gibbs energy of formation:
∆ f Gmο (NiI2, cr, 298.15 K) = – (94.4 ± 0.9) kJ·mol–1.
Based on the selected values of ∆ f H mο (NiI2, cr, 298.15 K) and ∆ f H mο (Ni2+,
298.15 K), the enthalpy of solution (at infinite dilution) of NiI2(cr) in water is
− (71.8 ± 1.2) kJ·mol–1, in reasonable agreement with the rather poorly documented
result of Paoletti [65PAO] (– 70.5 kJ·mol–1) for slightly higher ionic strength. The value
of – (73.38 ± 0.12) kJ·mol–1 reported by Efimov and Furkalyuk [90EFI/FUR] is somewhat more negative.
V.4.1.7 Solid nickel iodate compounds
Pracht et al. [97PRA/LAN] have summarised and examined phase relationships in the
Ni(IO3)2-H2O system. They concluded that, as reported in earlier papers [73COR],
[73NAS/SHI], two tetrahydrates, a dihydrate and a higher hydrate, perhaps a decahy-
138
V.4 Group 17 (halogen) compounds and complexes
drate, may have regions of metastability or stability near 300 K. Meusser [01MEU]
measured the solubilities of several different hydrated nickel iodates, and the most stable form near room temperature was the “β-dihydrate”, which is the Ni(IO3)2·2H2O of
later workers. The solubility results of Meusser [01MEU] and Cordfunke [73COR] for
the dihydrate are quite similar, and extrapolation [73COR] and interpolation [01MEU]
to 298.15 K provide values of – (5.17 ± 0.06) and – (5.17 ± 0.04) for log10 K sο,0 ((V.68),
298.15 K) (see Appendix A). This agreement is rather surprising as neither worker reported a measurement at 298.15 K, and there are doubts concerning whether equilibrium
was attained, especially in the work of Cordfunke [73COR].
Ni(IO3)2·2H2O(cr) U Ni2+ + 2 IO3− + 2H2O(l)
(V.68)
Fedorov et al. [73FED/SHM] measured the solubility of a solid with the reported composition of a trihydrate, Ni(IO3)2·3H2O, in aqueous lithium perchlorate/nitrate mixtures at 298.15 K and ionic strengths from 0.5 to 4.0 M. Their solubility
results near 25°C, on extrapolation to I =0, are similar to those from the other two studies. Based on the values obtained from experiments using solutions without nitrate, a
value of the solubility product of Ni(IO3)2·3H2O, log10 K sο,0 = – (5.09 ± 0.16) is calculated (see Appendix A).
It is clear from all the studies that the kinetics of the interconversion of the hydrates is slow at room temperature. The evidence for a true trihydrate is slim, yet might
be the result of catalysed precipitation of a more stable form that is difficult to synthesize in other media. It might also reflect insufficient periods of equilibration, or that the
solid was not properly handled prior to analysis. There does not seem to be any thermodynamic information (vapour pressure measurements) available on the interconversion
of the hydrates.
In the present review the solubilities of Fedorov et al. [73FED/SHM] are assumed to relate to the dihydrate rather than the “trihydrate”. The three values for the
dihydrate at 298.15 K are then in marginal agreement within the statistical (2σ) uncertainties. The average from the three studies is selected:
log10 K sο,0 (V.68) = − (5.14 ± 0.10),
where the uncertainty of 0.10 in log10 K sο,0 is assigned to reflect uncertainty in the crystalline form.
From the solubility product,
∆ r Gmο ((V.68), 298.15 K) = (29.34 ± 0.57) kJ·mol–1.
Using this selected value and the auxiliary data in Table IV-1,
∆ f Gmο (Ni(IO3)2·2H2O, cr, 298.15 K) = – (802.1 ± 1.8) kJ·mol–1.
The effective enthalpy of the dissolution reaction can be calculated from the
temperature dependence of the solubility products, provided that the heat capacity of
V.4 Group 17 (halogen) compounds and complexes
139
reaction is small over the experimental temperature range. For a dissolution reaction of
an ionic salt, it is unlikely that the latter condition is met; nevertheless, as shown in Appendix A, the values based on the data of Meusser [01MEU] and Cordfunke [73COR]
do not appear to change significantly if the temperature range used for calculation of
∆ r H m (V.68) is changed. On this basis, the average of the calculated temperatureaveraged values for the enthalpy of reaction (V.68), 21.4 kJ·mol–1 (281 to 323 K)
[01MEU] and 21.7 kJ·mol–1 (302 to 323 K) [73COR], is selected for 298.15 K, but with
a slightly increased estimated uncertainty:
∆ r H mο ((V.68), 298.15 K) = (21.6 ± 5.0) kJ·mol–1.
Using this value and the auxiliary data in Table IV-1,
∆ f H mο (Ni(IO3)2·2H2O, cr, 298.15 K) = – (1087.7 ± 5.2) kJ·mol–1
S mο (Ni(IO3)2·2H2O, cr, 298.15 K) = (270.1 ± 17.4) J·K–1·mol–1
are calculated.
The heat capacity of Ni(IO3)2·2H2O(cr) has been measured [64CHA/BOE],
[72KLA/DOK] between 1 and 30 K, but no higher temperature measurements appear to
have been reported.
There are two forms of anhydrous nickel iodate, and there is some evidence
[73NAS/SHI], [97PRA/LAN] that the yellow (hexagonal) β form is more stable than
the green α form. The measurements of Meusser for the solubility of an anhydrous
nickel iodate, probably the β-Ni(IO3)2 [73NAS/SHI], at four temperatures leads to
log10 K sο,0 ((V.69), Ni(IO3)2, β, 298.15 K) = – (4.43 ± 0.02), and an average value of
∆ r H m (V.69) = – (7.3 ± 0.7) kJ·mol–1 (303.15 K to 363.15 K). It is clear that at 298.15 K
the dihydrate is less soluble than the anhydrous salt. In the absence of other data, the
value of the solubility product is selected, but with an increased uncertainty:
β-Ni(IO3)2 U Ni2+ + 2 IO3−
(V.69)
∆ r Gmο ((V.69), 298.15 K) = (25.3 ± 0.1) kJ·mol–1.
Using this value and the auxiliary data in Table IV-1,
∆ f Gmο (Ni(IO3)2, β, 298.15 K) = – (323.7 ± 1.7) kJ·mol–1.
The enthalpy of reaction value is somewhat suspect because it is an average
value from measurements over a rather large temperature range (not including
298.15 K), and no attempt has been made to include heat capacity values in the data
analysis. Nevertheless, the value is selected here with an increased estimated uncertainty.
∆ r H mο ((V.69), 298.15 K) = – (7.3 ± 4.0) kJ·mol–1.
Using this value and the auxiliary data in Table IV-1,
140
V.4 Group 17 (halogen) compounds and complexes
∆ f H mο (Ni(IO3)2, β, 298.15 K) = – (487.1 ± 4.2) kJ·mol–1
S mο (Ni(IO3)2, β, 298.15 K) = (213.5 ± 14.1) J·K–1 mol–1
are selected.
V.4.2 Aqueous nickel halide complexes
V.4.2.1 Introduction
Halide anions, with the exception of fluoride, form rather unstable complexes with
Ni(II) in aqueous solution. This is mostly due to the strong hydration of Ni(II). Thus,
water can efficiently compete with the essentially electrostatic Ni(II)-halide interaction.
Consequently, high and varying excesses of ligand anions over Ni(II) have been used to
assess the stability of the complexes formed. Under such conditions a large part of the
medium ions are replaced by the complex-forming electrolyte, causing substantial
change in the activity coefficients, and thus, the constant medium principle is violated in
these measurements. According to Harned's rule, the logarithms of the activity coefficients in a mixture of two electrolytes vary linearly with their mole fractions. This variation, which depends mainly on the hydration of the ions involved, introduces an inherent and electrolyte dependent uncertainty into the determined stability constants of weak
complexes [89BJE]. The hypothesis generally made in the earlier volumes of this series,
was that in all studies of weak complexation, to a first approximation, the activity coefficients do not change significantly as the weak complex-forming anion substitutes for
perchlorate, provided that the total ionic strength is maintained constant. In the present
review the same procedure is accepted, emphasising the approximative validity of this
hypothesis. Therefore, if considerable medium changes occurred during the experiments, we assign higher uncertainty to the accepted formation constants than was published in the original report, regardless of the quality of the experimental work. Furthermore, as it is almost impossible to distinguish between a medium effect and the
formation of higher complexes, for lack of solid evidence, only NiX+ species are accepted in this review, if a substantial medium effect can be assumed.
In some cases mixed or ‘semi-thermodynamic’ formation constants are reported [36JOB], [71PAA/HUM], [72RET/HUM], [88BJE], [90BJE]. In these sets of
measurements, the formation of weak complexes was studied at low concentrations of
the metal ion and up to the highest possible concentrations of the highly soluble salts
(NX) of the complex-forming anions (X–). The ‘semi-thermodynamic’ formation constants can be written as:
[ NiX n ]
Kn =
NiX
[ n −1 ] ⋅ [ X ] ⋅ γ ±
where γ± is the molar activity coefficient of the complex forming electrolyte (NX) in
molar units [88BJE], [90BJE]. Others [71PAA/HUM], [72RET/HUM] used the so
called complete equilibrium constant ( K'n ), in which the water is also taken into consideration in the equilibrium process,
V.4 Group 17 (halogen) compounds and complexes
141
[NiXn–1(H2O)7–n]3–n + X– U [NiXn(H2O)6–n]2–n + H2O(l)
K n' =
 NiX n ( H 2 O )6 − n  ⋅ aw
 NiX n −1 ( H 2 O )7 − n  ⋅ [ X ] ⋅ γ ±
Although these constants probably provide a better description of the given
system in the case of a high and varying excess of complex-forming anion than do those
determined using the constant medium principle, their chemical meaning is not clear
[90BEC/NAG]. These ‘semi-thermodynamic’ formation constants are not equivalent to
the value extrapolated to zero ionic strength using the SIT, except for the case when the
ratio γ [NiXn ( H2 O ) ] / γ [NiXn−1 ( H2 O ) ] = 1 over the whole range of the ionic strength studied.
6− n
7 −n
Besides the weakness of complex formation, the simultaneous presence of inner- and outer-sphere complexes, especially in the cases of chloride and bromide, further complicates the determination of the correct solution speciation. Using most of the
experimental methods the measured formation constant is the sum of the formation constants of two types of complexes, [NiL(H 2 O)5 ]inner and [Ni(H 2 O)6 L]outer :
β1exp =
[NiL(H 2 O)5 ]inner + [Ni(H 2 O)6 L]outer
= β1inner + β1outer
[Ni(H 2 O)6 ] ⋅ [L]
However, the d-d transitions of Ni(II) are certainly more influenced by innerthan outer-sphere complexation. In the case of weak complexes, inner-sphere interaction
can be expected to dominate only at markedly reduced water activity, i.e., in case of
concentrated solutions. Consequently, a visible spectrophotometric study, using high
concentrations of the complex-forming anion, mostly provides information on innersphere complexes, and it is therefore not strictly comparable with e.g., potentiometric
results obtained at considerably lower ligand concentrations.
V.4.2.2 Solution structural studies
A number of NMR [68LIN/APR], [79WEI/HER], [79WEI/MUL], [83GOW/DOD],
EXAFS [80LAG/FON], [83LIC/PAS], [99HOF/DAR], solution X-ray [80CAM/LIC],
[82WAK/ICH], [82MAG/PAS], [85MAG/MOR], [99WAI/KOU], neutron diffraction
[83NEI/END] and neutron scattering [2002CHI/SIM] studies are available concerning
the structure of concentrated NiCl2 and NiBr2 solutions. In presence of 2 M NiCl2 and
4 − 6 M LiCl/HCl, solution X-ray [82MAG/PAS] and EXAFS [83LIC/PAS] studies
clearly demonstrated the formation of inner-sphere chloro complexes, although the average number of Ni2+ − Cl– contacts are relatively low (nCl– = 0.7-1.4). In stoichiometric
solutions the situation is less clear. Some authors [80LAG/FON], [80CAM/LIC],
[83NEI/END], [99HOF/DAR] have detected the formation of ion pairs only, but the
majority of reports agree with the presence of inner-sphere complexes at higher concentrations. The first solid evidence in favour of the existence of inner-sphere
[Ni(H2O)5Cl]+ complex in stoichiometric NiCl2 solutions was the detection of slow
chloride exchange by 35Cl NMR at – 5 and + 1°C [83GOW/DOD]. This observation
V.4 Group 17 (halogen) compounds and complexes
142
demonstrates the presence of inner-sphere complexes, since outer-sphere interaction is
unlikely to produce slow ligand-exchange. Solution X-ray diffraction studies of concentrated NiBr2 solutions [82WAK/ICH], [85MAG/MOR] confirmed the above observations, since the very large backscattering amplitude of a Br scatterer results in better
identification of Ni2+ − Br– contacts than the Ni2+ − Cl– ones. Although in [99HOF/DAR]
the authors report the formation of ion pairs at 298 K, the temperature dependence of
XAFS spectra clearly indicate the formation of tetrahedral [NiBr4]2– complex above
698 K. The average number of Ni2+ − X– contacts reported in most of the publications
mentioned above allowed the calculation of the stability constants of inner-sphere complexes. Although the applied ionic strengths are generally very high, some of these constants are reported in Sections V.4.2.4 and V.4.2.5.
V.4.2.3 Aqueous Ni(II) - fluoro complexes
The reported stability constants for Reaction (V.70) are listed in Table V-8.
Ni2+ + F– U NiF+
(V.70)
Table V-8: Experimental stability constants (logarithmic values) of the species NiF+.
Method
ise-F
Medium
t (°C)
0.05 M (CH3)4NClO4
25
(a)
log10 K1
log10 K1
reported
accepted
Reference
(1.32 ± 0.05)
(1.32 ± 0.30)
[83SOL/BON]
ise-F
0.1 M NaClO4
25
(1.02 ± 0.12)
(1.02 ± 0.20)
[81KUL/BLO]
kin
0.1 M NaClO4
25
(1.1 ± 0.1)
(1.10 ± 0.20)
[69FUN/TAN]
ise-F
0.25 M NaClO4
25
(0.91 ± 0.10)
(0.91 ± 0.30)
Mironov et al.(b)
ise-F
0.5 M NaClO4
25
(0.76 ± 0.08)
(0.75 ± 0.20)
Mironov et al.(b)
ise-F
1.0 M NaClO4
25
(0.68 ± 0.09)
(0.66 ± 0.20)
Mironov et al.(b)
ise-F
1.0 M NaClO4
25
(0.34 ± 0.04)
(0.32 ± 0.30)
[72BON/HEF]
pol
1.0 M Na(ClO4, F)
25
(0.48 ± 0.05)
(0.46 ± 0.30)
[71BON]
qh
1.0 M NaClO4
20
(0.66 ± 0.05)
(0.64 ± 0.10)
[56AHR/ROS]
(0.66 ± 0.10)(c)
25
ise-F
2.0 M NaClO4
25
(0.65 ± 0.06)
(0.61 ± 0.20)
[81KUL/BLO]
ise-F
3.0 M NaClO4
25
(0.76 ± 0.06)
(0.69 ± 0.20)
Mironov et al.(b)
(a) Corrected to molal scale. The accepted values reported in Appendix A are expressed on the
molar or molal scales, depending on which units were used originally by the authors.
(b) Average value from the studies reported by Mironov’s group [83AVR/BLO],
[81KUL/BLO], [76KUL/BLO].
(c) Corrected to 298.15 K using the selected value of ∆ r H mο .
The majority of data in Table V-8 were obtained in NaClO4 solutions using a
fluoride selective electrode, but some pH-metric [56AHR/ROS], kinetic [69FUN/TAN]
and polarographic [71BON] results are included, too. The SIT analysis of the experimental stability constants is depicted in Figure V-20.
V.4 Group 17 (halogen) compounds and complexes
143
Figure V-20: Extrapolation to Im = 0 of the experimental data for Reaction (V.70) in
NaClO4 media. Experimental data from: : [56AHR/ROS], : [69FUN/TAN], :
[71BON], : [72BON/HEF], : [81KUL/BLO], : [83SOL/BON], : [76KUL/BLO],
[81KUL/BLO], [83AVR/BLO].
2.0
1.8
log10 β1 + 4D
1.6
1.4
1.2
Ni2+ + F– U NiF+
1.0
0.8
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
−1
I / mol·kg
The value given by Arhland and Rosengren [56AHR/ROS] determined at
293.15 K has been corrected to 298.15 K using the enthalpy of reaction selected below.
The constant determined in 0.05 M tetraethylammonium perchlorate [83SOL/BON] was
also considered (but with a somewhat higher uncertainty). Although there are notable
differences between the values published by different authors, the data in Figure V-20
show acceptable consistency. We should note, however, that most of these data were
reported by the Mironov’s group [83AVR/BLO], [81KUL/BLO], [76KUL/BLO]. Only
two of the five constants reported by other laboratories are consistent with the data of
Mironov et al. The weighted linear regression using 11 data points yielded the selected
value of:
log10 b1ο ((V.70), 298.15 K) = (1.43 ± 0.08).
This value agrees well with that published in [81KUL/BLO] calculated by
means of the Vasil’ev equation. The resulting ∆ε(V.70) value is – (0.049 ± 0.060)
kg·mol–1. Using the selected values for ε(Ni2+, ClO −4 ) and ε(Na+, F–), ∆ε(V.70) leads to
a value of ε(NiF+, ClO −4 ) = (0.34 ± 0.08) kg·mol–1.
V.4 Group 17 (halogen) compounds and complexes
144
From the above formation constant ∆ r Gmο ((V.70), 298.15 K) = – (8.2 ± 0.5)
kJ·mol can be calculated. The Gibbs energy of formation of NiF+ is derived using the
selected values for Ni2+ and F–. The selected value is:
–1
∆ f Gmο (NiF+, 298.15 K) = – (335.5 ± 1.1) kJ·mol–1.
The formation of higher complexes ( NiFn2 − n , n > 1) is not reported in the literature, not even in presence of more than a thousand-fold excess of fluoride over Ni(II)
[71BON].
Reaction enthalpies for the formation of the NiF+ complex are collected in
Table V-9. All data in Table V-9 except those of [74ARU] were determined in Mironov’s laboratory. In order to determine ∆ r H mο ((V.70), 298.15 K), the SIT equation
adapted to enthalpy format [97GRE/PLY2] has been used (cf. Figure V-21). The values
at I = 0.1 and 0.25 M [81KUL/BLO] were not considered, as the amount of NiF+
formed is probably too small to yield reliable enthalpies. The value reported by Aruga
was also neglected for the SIT, due to the different ionic medium used. The analysis of
the remaining four points resulted in the selected standard reaction enthalpy:
∆ r H mο (V.70) = (9.5 ± 3.0) kJ·mol–1
and ∆εL(V.70) = – (5 ± 35) × 10–5 kg·K–1·mol–1. The former value is fundamentally different from that reported in [81KUL/BLO] ( ∆ r H mο ((V.70), 298.15 K) = – (2.9 ± 2.5)
kJ·mol–1; the method of extrapolation to I = 0 is not reported). It agrees well, however,
with the general observation that the stabilities of most aqueous monofluoride complexes are governed predominantly by electrostatic interactions and thus are entropy
controlled [74HEF], [90GRZ].
Table V-9: Experimental enthalpy values for Reaction (V.70).
Method
Medium
cal
0.1 M NaClO4
t /°C
∆ r H m (kJ·mol–1)
reported
25
0
∆ r H m (kJ·mol–1) Reference
accepted
[81KUL/BLO]
cal
0.25 M NaClO4
25
(3.3 ± 1.6)
cal
0.5 M NaClO4
25
(6.7 ± 0.8)
(6.7 ± 2.0)
[81KUL/BLO]
cal
0.5 M self(a)
25
(8.1 ± 0.2)(b)
(4.4 ± 2.0)(c)
[74ARU]
cal
1.0 M NaClO4
25
(7.9 ± 1.2)
(7.9 ± 2.0)
[81KUL/BLO]
cal
2.0 M NaClO4
25
(7.9 ± 1.6)
(7.9 ± 2.0)
[81KUL/BLO]
cal
3.0 M NaClO4
25
(6.5 ± 2.0)
(6.5 ± 2.0)
Mironov et. al.(d)
(a)
(b)
(c)
(d)
[81KUL/BLO]
The solution of 0.17 M Ni(NO3)2 and 0.5 M NaF were mixed in different ratios.
Calculated from log10 b1 = 0.31, which was derived from the datum in [72BON/HEF].
Recalculated data using the selected value for log10 b1ο ((V.70), 298.15 K) and ∆ε (V.70).
Average value from the studies reported by Mironov’s group [76KUL/BLO] and [81KUL/BLO].
V.4 Group 17 (halogen) compounds and complexes
145
From the above ∆ r H mο value, the following selected standard enthalpy of formation can be derived for NiF+:
∆ f H mο (NiF+, 298.15 K) = – (380.9 ± 3.2) kJ·mol–1.
The above selections result in:
S mο (NiF+, 298.15 K) = – (86.4 ± 10.3) J·K–1·mol–1.
Figure V-21: Extrapolation to Im = 0 of the experimental enthalpies for Reaction (V.70)
in NaClO4 media from [76KUL/BLO] and [81KUL/BLO] (squares) and [74ARU] (triangle). The data marked with open square were not considered in the SIT treatment,
since the amount of NiF+ formed is too small. According to [97GRE/PLY2],
Ψ(Im) = −
A I
3∆ ( z 2 )
× L m
4
1 + 1.5 I m
where ∆(z2) is the stoichiometric sum of the charge numbers squared of the reactants (in
the present case – 4), AL is the Debye-Hückel parameter for the enthalpy, at 298.15 K
and 1 bar AL = 1.986 kJ·kg1/2·mol–3/2.
16
Ni2+ + F– U NiF+
∆rH m−Ψ(I m) / kJ·mol−1
14
12
10
8
6
4
2
0
0
0.5
1
1.5
2
I / mol·kg−1
2.5
3
3.5
4
146
V.4 Group 17 (halogen) compounds and complexes
V.4.2.4 Aqueous Ni(II) - chloro complexes
The formation of Ni(II)-chloro complexes has been the subject of numerous investigations using several experimental methods (cf. Table V-10). Since in most of these studies the composition of the ionic medium was varied over a wide range, and no attention
was paid to the medium effect, the literature reports on the complex formation are rather
divergent concerning both the reported formation constants and stoichiometry of the
formed complexes. Beside the formation of NiCl+ and NiCl2(aq) complexes, the presence of other species is also proposed in aqueous solution at room temperature: e.g.,
Ni2Cl3+ [60LIS/ROS], NiCl3− [64GRI/LIB] or NiCl24 − [68AND/KHA], [71KLY/GLE],
[85SAP/CHI]. This reflects the fact that conventional methods are not always applicable
to determine the stability of weak metal complexes if a high concentration of the ligand
is needed to achieve a sufficient degree of complex formation. The majority of the experiments have shown that no anionic complexes are formed even at very high chloride
concentrations. On the other hand, the formation of Ni2Cl3+ reported in [60LIS/ROS]
was reconsidered and finally rejected in a subsequent paper of the authors
[66KEN/LIS]. Therefore in this review only the formation constants for NiCl+ and
NiCl2(aq) complexes are considered. The formation constants reported for Reaction
(V.71),
Ni2+ + Cl– U NiCl+
(V.71)
are listed in Table V-10. For reasons mentioned in Appendix A we do not consider the
data reported in [57KIV/LUO], [58TRE], [62TRI/CAL], [65MOR/REE],
[74GRA/WIL] and [76MUR/KUR]. Most of the accepted data are, however, also subject to substantial experimental errors, due to the medium effect, and in such cases we
assigned significantly higher uncertainty to the selected constants than reported in the
original publications. The data in [75LIB/TIA] are free of significant medium effects,
and this is the only data set where the systematic errors can be assumed identical for
each point. Therefore, these data were used to determine the ion interaction coefficient
between NiCl+ and ClO −4 , in spite of the fact that the applied ionic strength (Im = 3 – 9
m) is well above of the recommended range for the SIT analysis.
The plot ( log10 b1 + 4D) versus Im is depicted in Figure V-22. The weighted
linear regression of data in Figure V-22 results in the values of log10 b1ο (V.71) =
− (0.37 ± 0.27) and ∆ε(V.71) = – (0.073 ± 0.040) kg·mol–1. Using the selected values for
ε(Ni2+, ClO −4 ), ε(Ni2+,Cl–), (in the absence of data to allow calculation of ε(Ni2+,Cl−)
while also considering association; see, for example, Section V.6.1.2), ∆ε(V.71) leads to
a value of:
ε(NiCl+, ClO −4 ) = (0.47 ± 0.06) kg·mol–1.
There is some uncertainty concerning this calculation. In [75LIB/TIA]
Ni(ClO4)2 was used as a constant ionic medium (with a chloride concentration of ~ 0.01
m), therefore, ∆ε can be calculated as ∆ε = ε(NiCl+, ClO −4 ) – ε(Ni2+,Cl–) – ε(Ni2+, ClO −4 ).
V.4 Group 17 (halogen) compounds and complexes
147
In Appendix B.2 of previous TDB reviews it is mentioned that since the formation of
chloride complexes is taken into account, the value of ε(Ni2+, Cl–) as listed, for example,
in Appendix B of [92GRE/FUG], [2001LEM/FUG] should be replaced by
ε(NiCl+, ClO −4 ) in all calculations when chloride is part of the ionic medium. This listed
value of ε(Ni2+,Cl–) was obtained [80CIA] by neglecting association of nickel and chloride. Following this process, and using ε(Ni2+, ClO −4 ) = (0.37 ± 0.03) kg·mol–1 (see Section V.4.3), from the above determined ∆ε = – (0.073 ± 0.04), ε(NiCl+, ClO −4 ) = (0.667
± 0.06) kg·mol–1 can be calculated. This value of ε(NiCl+, ClO −4 ) is too high, taking into
account the relatively accurate value for ε(NiF+, ClO −4 ).
In this review, ε(NiCl+, ClO −4 ) = (0.47 ± 0.06) kg mol–1 is used to determine ∆ε
in other media to calculate log10 b1ο values from the rest of the data in Table V-10.
According to Appendix B.2, the same treatment should be applied, e.g., for
NO3− , i.e., ε(Mx+, NO3− ) should be replaced by ε(Mx+, ClO −4 ) if nitrate is part of the medium, but this recommendation is often disregarded.
Using the remaining experimental values for NaClO4 media (from Table V-10)
and ε(Ni2+, ClO −4 ) = 0.37 to determine ε(NiCl+, ClO −4 ), from the equation ∆ε =
ε(NiCl+, ClO −4 ) – ε(Na+, Cl–) – ε(Ni2+, ClO −4 ), ε(NiCl+, ClO −4 ) = 0.51 kg·mol–1 can be
obtained. This may support the value calculated using ε(Ni2+, Cl–) from the data in
[75LIB/TIA], however, this is not particularly convincing because most of the remaining data in the other papers have relatively low accuracy, due to substantial medium
effects.
Table V-10: Experimental formation constants (logarithmic values) for the NiCl+ species.
Method Ionic medium
t (°C)
log10 b1
(a)
log10 b1
(b)
log10 b1ο (c)
Reference
pol
2 M Na(ClO4, Cl)
25
– (0.25 ± 0.15)
[57KIV/LUO]
cix
?
– (0.66 ± 0.01)
[58TRE]
cry
1.5 M (Ca, Ni)
(NO3, Cl)
0.048 m KClO4
– 0.15
0.62
cry
0.258 m KClO3
– 0.8
0.23
ise-Cl 2 M (Na, Ni)ClO4
12
25
40
–2
– 0.17
– 0.15
(0.38 ± 0.08)
cry
1.3 M K(NO3/Cl)
sp
1.5 M Na(ClO4, Cl)
?
– (0.85 ± 0.25)
[59KEN]
(0.23 ± 0.30)
(0.79 ± 0.40)
[59KEN]
– (0.25 ± 0.20)(d)
– (0.21 ± 0.20)(d)
– (0.18 ± 0.20)(d)
(0.34 ± 0.30)(d)
(0.84 ± 0.30)
(0.88 ± 0.30)
(0.91 ± 0.30)
(1.14 ± 0.40)
[60LIS/ROS]
[62FAU/CRE]
[62TRI/CAL]
sp
5.7 M H(Cl, ClO4)
25
– (0.5 ± 0.2)
– (0.62 ± 0.40)
(0.16 ± 0.50)
[63NET/DRO]
cix
3 M H(ClO4, Cl)
25
– (0.64 ± 0.06)
– (0.51 ± 0.50)(d)
(0.35 ± 0.60)
[64GRI/LIB]
cix
0.69 H(ClO4, Cl)
20
(0.23 ± 0.03)
[65MOR/REE]
(Continued on next page)
V.4 Group 17 (halogen) compounds and complexes
148
Table V-10 (continued)
Method Ionic medium
t (°C)
log10 b1
(a)
– (1.02 ± 0.05)
sp
10 M Li(NO3, Cl)
30
sol
4 M Na(ClO4, Cl)
25
sp
3 M Li(ClO4, Cl)
25
pol
1 M Na(ClO4, Cl)
25
sp
6 M (Na, Li)(ClO4, Cl)
0 corr(g)
25
– (0.5 ± 0.1)
– (1.30 ± 0.09)
kin
1 M Na(ClO4, Cl)
25
(0.07 ± 0.08)
ise-Cl 1 M NaClO4 + 0.1 M
HClO4
ise-Cl 3 M Na(ClO4,Cl)
25
(0.00 ± 0.06)
25
(0.69 ± 0.04)
ise-Cl 1 m Ni(ClO4)2
1.5 m Ni(ClO4)2
2 m Ni(ClO4)2
2.5 m Ni(ClO4)2
3 m Ni(ClO4)2
cix
1 M Na(ClO4, Cl)
25
sp/gl 3 M Na(ClO4, Cl)
25
nmr
4 m KCl
25
25
log10 b1ο (c)
(b)
– (1.17 ± 0.50)
Reference
[66FLO]
– (0.71 ± 0.30)(d)
(0.68 ± 0.40) [70HAL/VAN]
(0.18 ± 0.50)(f)
– (0.57 ± 0.03)
– (0.63 ± 0.40)
(0.36 ± 0.50)
[70MIR/MAK]
– (0.22 ± 0.08)
– (0.24 ± 0.40)
(0.66 ± 0.50)
[71BON]
– 0.2(e)
– 1.1
– 1.1
– 1.0
– 0.92
– 0.80
0.0
– (0.50 ± 0.13)
(h)
sp
0 corr
25
– (1.90 ± 0.10)
sol
6 m Na(ClO4, Cl)
0 corr(j)
25
– (0.69 ± 0.08)
– 0.83
(i)
log10 b1
[71PAA/HUM]
(0.00 ± 0.40)(d)
– (0.02 ± 0.30)
(0.90 ± 0.50)
[73HUT/HIG]
(0.90 ± 0.40)
[74BIX/LAR]
[74GRA/WIL]
– (1.10 ± 0.20)
– (1.10 ± 0.20)
– (1.00 ± 0.20)
– (0.92 ± 0.20)
– (0.80 ± 0.20)
– (0.37 ± 0.27)
– (0.57 ± 0.40)
(0.68 ± 0.50)
[78FOR]
– (1.26 ± 1.00)
(0.18 ± 1.00)
[79WEI/HER]
– (0.69 ± 0.20)
(0.65 ± 0.30)
[75LIB/TIA]
[76MUR/KUR]
(d)
[88BJE]
[89IUL/POR]
(a) Reported values.
(b) Accepted values corrected to molal scale. The accepted values reported in Appendix A are expressed on
the molar or molal scales, depending on which units were used originally by the authors.
(c) Corrected to I = 0 by means of the SIT, using the selected ion interaction coefficients.
(d) Re-evealuated value, see Appendix A.
(e) The authors suggested solely the formation of NiCl2(aq) complex, but they also calculated the stability constant of NiCl+ assuming its unique formation.
(f) Re-evaluated value taking into account of the effect of medium changes, see discussion on
[70HAL/VAN].
(g) ‘Semi-thermodynamic’ constant, in which the activity of chloride (using γ± of the complex forming
electrolyte) and water were taken into account.
(h) The average number of bound chloride per Ni(II) is reported.
(i) ‘Semi-thermodynamic’ constant, in which the activity of chloride (using γ± of the complex forming
electrolyte) was taken into account.
(j) Corrected to I = 0 by means of the SIT, using estimated ion interaction coefficients.
V.4 Group 17 (halogen) compounds and complexes
149
Figure V-22: Extrapolation to I = 0 of the log10 b1 values reported in [75LIB/TIA]
for Reaction (V.71) in Ni(II) perchlorate solution. The solid line is calculated with
log10 b1ο = – (0.37 ± 0.27), and ∆ε = – (0.073 ± 0.040).
1
0.8
Ni2+ + Cl– U NiCl+
log10 β1 + 4D
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
0
1
2
3
4
5
6
7
8
9
10
−1
I / mol·kg
In [89IUL/POR] the authors proposed ε(NiCl+, ClO −4 ) = 0.202 kg·mol–1, estimated from ε(Ni2+, ClO −4 ) = 0.375 kg·mol–1 and ε(Na+, Cl–) using Equation (B.22) (see
Appendix B). Although Iuliano’s value is considerably lower, we recommend the value
derived in the present review for ε(NiCl+, ClO −4 ), since it is based on experimental evidence.
All accepted constants for NaClO4 media are within the ionic strength range Im
= 0 – 6 m. The SIT representation of these values is depicted in Figure V-23. Using the
ion interaction coefficient ε(Ni2+, ClO −4 ) together with the above determined value of
ε(NiCl+, ClO −4 ), ∆ε = (0.07 ± 0.10) kg·mol–1 can be calculated for Reaction (V.71) in
NaClO4 medium. From the plot of experimental data in Figure V-23 a very similar
value, ∆ε = (0.11 ± 0.06) kg·mol–1 can be obtained. However, this value must be treated
with caution, due to the already mentioned medium change, which affects most of the
data points in Figure V-23. Therefore, ∆ε((V.71), NaClO4) = (0.07 ± 0.10) kg·mol–1 was
used to calculate the log10 b1ο values for NaClO4 media given in Table V-10. The SIT
extrapolation resulted in somewhat lower log10 b1ο values in LiClO4 (∆ε((V.71),
V.4 Group 17 (halogen) compounds and complexes
150
LiClO4) = (0.0 ± 0.1) kg·mol–1) and HClO4 (∆ε((V.71), HClO4) = – (0.04 ± 0.10)
kg·mol−1) media.
Figure V-23: Extrapolation to I = 0 of the accepted experimental data for Reaction
(V.71) in NaClO4 media (see text).
1.5
Ni2+ + Cl– U NiCl+
log10 β1 + 4D
1.0
0.5
0.0
0
1
2
3
I / mol·kg
4
5
6
7
−1
These extrapolated values are considerably higher than the value determined
from the data in [75LIB/TIA]. This is probably a result of the disregarded medium
changes. Obviously, the attribution of the global effect (complex formation and changes
in activity coefficients) to complex formation alone results in higher constants.
The well documented experimental results in [70HAL/VAN] offered the possibility to use the SIT to assess the medium effect caused by the replacement of the
original background electrolyte (see Appendix A). In this work the background medium
was changed gradually from 4 M NaClO4 to 2 M NaClO4/2 M NaCl. As explained in
the discussion on [70HAL/VAN], the modified SIT treatment, in which the changes in
ionic medium were taken into account, resulted in a value approximately half a unit
lower in the log10 b1ο value (0.20 ± 0.40) than obtained by neglecting this effect.
The medium changes, caused by the replacement of ClO −4 by Cl– ions, also
were accounted for by SIT in [89IUL/POR]. Due to the different ε(NiCl+, ClO −4 ) value
V.4 Group 17 (halogen) compounds and complexes
151
used (see Appendix A), however, the authors reported a considerably lower log10 b1ο
than was recalculated in this review from their data.
Bjerrum [88BJE] reported a ‘semi-thermodynamic’ formation constant
( log10 b1ο ' ), which is much lower than any other log10 b1ο value in Table V-10. Due to
the unknown ratio of γ Ni2+ / γ NiCl+ (see Appendix A), this constant is not compatible
with the assumptions of the SIT model. Since Bjerrum applied a spectrophotometric
method using the d-d transitions of Ni(II) and very high chloride concentrations, his
constant probably refers to the formation of the inner-sphere NiCl(H 2 O)5+ complex
( log10 b1ο,inner ). Libus and Tialowska estimated a similar value for log10 b1ο,inner (– 2.0,
[75LIB/TIA]). Bjerrum’s value is also within the uncertainty range of the ‘semithermodynamic’ formation constant derived from the data reported in [79WEI/HER]
( log10 b1ο ' = – (1.03 ± 1.00), see Appendix A). However, the NMR relaxation measurements in [79WEI/HER] also demonstrate the incompatibility of log10 b1ο ' as defined by
Bjerrum (see [88BJE] and [89BJE]) and the log10 b ο values extracted by means of the
SIT analysis. Using the recommended ion interaction coefficients to calculate ∆ε(V.71)
for KCl medium, log10 b1ο ((V.71), 298.15 K) = (0.16 ± 1.00) is derived from the experimental data in [79WEI/HER], which is considerably higher than Bjerrum’s
log10 b1ο ' value.
Due to the medium effect, most of the constants listed in Table V-10 represent
only the upper limits of the true values. Therefore, log10 b1ο ((V.71), 298.15 K) is obtained as the average of the following two data: – (0.37 ± 0.27) (determined from the
data in [75LIB/TIA]) and (0.52 ± 0.38) (average of the remaining log10 b1ο values reported in Table V-10 for 298 K in (H/Li/Na)ClO4 and KCl media). This treatment resulted in the selected value:
log10 b1ο ((V.71), 298.15 K) = (0.08 ± 0.60).
From this value, ∆ r Gmο ((V.71), 298.15 K) = – (0.5 ± 3.4) kJ·mol–1 can be derived. The Gibbs energy of formation of NiCl+ is calculated using the selected values
for Ni2+ and Cl–.
∆ f Gmο (NiCl+, 298.15 K) = – (177.4 ± 3.5) kJ·mol–1.
Several authors reported equilibrium constants for Reaction (V.72), which are
collected in Table V-11:
NiCl+ + Cl– U NiCl2(aq).
(V.72)
For reasons mentioned in Appendix A, and since in cases in which there are
considerable changes in the ionic medium it is impossible to distinguish between a medium effect and the formation of higher complexes, we do not consider the data reported
in [57KIV/LUO], [65MOR/REE] and [70HAL/VAN]. ‘Semi-thermodynamic’ constants
are reported for Reaction (V.72) in [71PAA/HUM] and [88BJE] (see Appendix A),
based on a spectrophotometric method using the d-d transitions of Ni(II) and very high
V.4 Group 17 (halogen) compounds and complexes
152
chloride concentrations ([Cl–] ≥ 9 M). Therefore, these constants most probably refer to
the formation of the inner-sphere complex NiCl2(H2O)4(aq). Iuliano et al. used solubility measurements to determine the equilibrium constant for Reaction (V.72). They reported a higher value for K2 than for K1, which is rather unlikely considering the low
affinity of Ni(II) for chloride. Therefore, this review does not find it justified to select a
recommended value for K2 (V.72).
Table V-11: Experimental equilibrium constants (logarithmic values) for
Reaction (V.72).
Method
Ionic medium
pol
2 M Na(ClO4, Cl)
t /°C
25
log10 K 2 (a)
(0.20 ± 0.15)
Reference
[57KIV/LUO]
cix
0.69 H(ClO4, Cl)
20
– (0.27 ± 0.05)
[65MOR/REE]
sol(b)
4 M Na(ClO4, Cl)
25
– (0.89 ± 0.01)
[70HAL/VAN]
sp
0 corr(c)
25
– (2.8 ± 0.3)
[71PAA/HUM]
sp
0 corr(d)
25
– (2.3 ± 0.6)
[88BJE]
sol
6 M Na(ClO4, Cl)
25
– (0.5 ± 0.2)
[89IUL/POR]
0 corr(e)
– 0.3
(a) Reported values.
(b) Corrected to I = 0 by means of the SIT, using the selected ion interaction coefficients.
(c) ‘Semi-thermodynamic’ constant, in which the activity of chloride (using γ± of the complex forming electrolyte) and water were taken into account.
(d) ‘Semi-thermodynamic’ constant, in which the activity of chloride (using γ± of the complex forming electrolyte) was taken into account.
(e) Corrected to I = 0 by means of the SIT, using estimated ion interaction coefficients.
Higher chloro complexes NiCl2n − n with n = 3 – 4 readily form in several nonaqueous solutions [59GIL/NYH]. As was mentioned previously, such complexes are
unlikely to be present in detectable concentrations in aqueous solution at room temperature. The log10 b 4 in aqueous solution was estimated to be – 14 in [76BIA/CAR] (see
Appendix A). Accepting this value, the calculated concentration of NiCl24 − is negligible, even in highly concentrated chloride solutions. In aqueous solutions such complexes can be only expected at elevated temperature in melts [66ANG/GRU] or under
hydrothermal conditions [96UCH/GOR]. The only exception is reported in
[69GRI/SCA]. Using a concentrated aqueous solution of (CH3)4NCl ( > 5 M) as background electrolyte, the formation of tetrahedral NiCl24 − complex was detected by UVVIS spectrophotometry at as low a temperature as 50°C. This exceptional behaviour
was attributed to the specific effect of tetramethylammonium cation, which reduces the
hydration of chloride anion enhancing its coordination ability [69GRI/SCA].
V.4 Group 17 (halogen) compounds and complexes
153
Only a few studies have reported reaction enthalpies for the formation of NiCl+
species (Table V-12), and such data are not available for NiCl2(aq). Only the experimental results reported in [66KEN/LIS] for 298 K could be re-evaluated using
log10 b1ο ((V.71), 298.15 K) and ∆ε((V.71), NaClO4), therefore no value is selected in
the present review.
Table V-12: Reaction enthalpies reported for Reaction (V.71).
Method Medium
cal
t (°C)
2 M (Na, Ni)ClO4
∆ r H m (kJ·mol–1)(a)
25
(2.1 ± 0.2)
(c)
∆ r H m (kJ·mol–1)(b) Reference
(8.2 ± 2.5)(d)
[66KEN/LIS]
cal
2 M (Na, Ni)ClO4
40
(1.8 ± 0.2)
[66KEN/LIS]
cal
3 M Li(ClO4/Cl)
25
(9.6 ± 0.4)
[74BLO/RAZ]
(e)
(a) Reported value.
(b) Accepted value.
(c) Calculated using log10 b1 = – 0.17 (25°C) and log10 b1 = – 0.15 (40°C) from [60LIS/ROS].
(d) Re-evaluated value, see Appendix A.
(e) Calculated using log10 b1 = – 0.57 from [70MIR/MAK].
V.4.2.5 Aqueous Ni(II) – bromo complexes
Only a few studies are available concerning the complex formation between Ni(II) and
bromide ions. Most of the studies reported the formation of a monobromo complex
(NiBr+), but at very high concentrations of the bromide ion the formation of bis[36JOB], [72RET/HUM], [90BJE] and tetrakis- bromo complexes [36JOB],
[68AND/KHA2] are also reported. The graphical data presented in [36JOB] were later
re-evaluated in [72RET/HUM], and were found to be consistent with the formation of
NiBr+ and NiBr2(aq) complexes (see comments on [36JOB]). The qualitative work in
[68AND/KHA2] using spectrophotometry may well suggest formation of 5 – 10% of
the tetrakis - bromo complex, but only in extremely concentrated (18 M) HBr solutions
(see Appendix A). Nevertheless, the assessment of values for formation of negatively
charged complexes in aqueous solutions is not justified.
The available quantitative information on the formation of NiBr+, according to
Reaction (V.73), is collected in Table V-13.
Ni2+ + Br– U NiBr+
(V.73)
There are large differences between the values listed in Table V-13. This is the
consequence of the inherent difficulties that occur when formation constants of weak
complexes are determined (see also Sections V.4.2.1 and V.4.2.4). The values reported
in [63NET/DRO], [70MIR/MAK] and [73HUT/HIG] are of low precision, due to a considerable medium effect. The ‘semi-thermodynamic’ formation constant reported in
[90BJE] probably gives an exact description of complex formation in 3.5 – 11 M LiBr,
V.4 Group 17 (halogen) compounds and complexes
154
but this constant is not compatible with the assumptions of the SIT model (see also Sections V.4.2.1 and V.4.2.4 and [89BJE]). The most reliable data set is given in
[78LIB/KOW]. The SIT representation of the data reported in [78LIB/KOW] is depicted in Figure V-24. Although the ionic strengths in the experiments of [78LIB/KOW]
are beyond the optimal range for SIT, the experiments are relatively free of interfering
medium effects. Therefore, the estimation of ε(NiBr+, ClO −4 ) has been based on this
data set. This probably leads to a more accurate value than reliance on the other experimental values in Table V-13, or on Equation (B.22). From the plot in Figure V-24,
∆ε(V.73) = – (0.046 ± 0.09) kg·mol–1 and log10 b1ο = − (0.5 ± 0.6) are obtained. Using
the recommended values for ε(Ni2+, Br–) (in the absence of data to allow calculation of
ε(Ni2+, Br–), while also considering association; see Section V.4.2.5.1) and
ε(Ni2+, ClO −4 ),
ε(NiBr+, ClO −4 ) = (0.59 ± 0.10) kg·mol–1
can be derived. Using this unexpectedly large value, it is possible to estimate the ∆ε
values for all background electrolytes listed in Table V-13 (∆ε((V.73), NaClO4) =
(0.19 ± 0.11) kg·mol–1, ∆ε((V.73), HClO4) = (0.08 ± 0.11) kg·mol–1, ∆ε((V.73), LiClO4)
= (0.07 ± 0.11) kg·mol–1, ∆ε((V.73), KCl) = (0.22 ± 0.11) kg·mol–1). These coefficients
were used to calculate the log10 b1ο values given in Table V-13.
Table V-13: Experimental formation constants (logarithmic values) of the NiBr+
complex.
Method Ionic medium
ise-Br
t /°C
log10 b1
(a)
log10 b1
(b)
log10 b1ο (c)
Reference
2 M (Na, Ni)ClO4
25(d) – (0.12 ± 0.02) – (0.16 ± 0.40)
(1.20 ± 0.50)
[61LIS/WIL]
sp
5.7 M H(Br, ClO4)
25
– (0.30 ± 0.22) – (0.42 ± 0.50)
(1.28 ± 0.60)
[63NET/DRO]
sp
3 M Li(ClO4, Br)
25
– (0.82 ± 0.03) – (0.89 ± 0.50)
(0.35 ± 0.60)
[70MIR/MAK]
kin
1 M Na(ClO4, Br)
25
– (0.09 ± 0.04) – (0.11 ± 0.50)
(0.89 ± 0.60)
[73HUT/HIG]
ise-Br
1.5 m Ni(ClO4)2
25
nmr
sp
– 1.3
– (1.30 ± 0.30) – (0.50 ± 0.60) [78LIB/KOW]
2 m Ni(ClO4)2
– 1.3
– (1.30 ± 0.30)
2.5 m Ni(ClO4)2
– 1.3
– (1.30 ± 0.30)
3 m Ni(ClO4)2
– 1.16
– (1.16 ± 0.30)
4 m KBr
(e)
0 corr
– (1.10 ± 0.60)
25
25
– (2.37 ± 0.08)
(0.80 ± 0.60)
[79WEI/HER]
[90BJE]
(a) Reported values.
(b) Accepted values corrected to the molal scale. The accepted values reported in Appendix A are expressed on the molar or molal scales, depending on which units were used originally by the authors.
(c) Corrected to I = 0 by means of the SIT, using the recommended ion interaction coefficients.
(d) Data for several temperatures, see Appendix A.
(e) ‘Semi-thermodynamic’ constant, in which the activity of bromide (using γ± of the complex
forming electrolyte) was taken into account.
V.4 Group 17 (halogen) compounds and complexes
155
As already mentioned in previous sections, experiments in which there was a
considerable medium effect lead to formation constants that represent only upper limits
for the true values. Therefore, the weighted average of the following two values is calculated to assess the most probable formation constant at infinite dilution: − 0.5
( log10 b1ο calculated from the data in [78LIB/KOW], weight = 2) and 0.90 (average of
the remaining log10 b1ο values in Table V-13, weight = 1). This treatment yielded
log10 b1ο ((V.73), 298.15 K) = – (0.03 ± 1.30). The uncertainty range was assigned to
cover all the reported constants in Table V-13. Taking into account this large uncertainty, the above value can not be recommended, but can be used as the most probable
value, until more precise data are published for Reaction (V.73).
Figure V-24: Extrapolation to I = 0 of the log10 b1 values reported in [78LIB/KOW] for
Reaction (V.73) in Ni(II) perchlorate media.
2
Ni2+ + Br– U NiBr+
1.5
log10 β1 + 4D
1
0.5
0
-0.5
-1
-1.5
-2
0
1
2
3
4
5
6
7
8
9
10
−1
I / mol·kg
The formation of the dibromo complex NiBr2(aq) is dominant only at very high
bromide concentrations (8 – 12 M LiBr or HBr). Therefore, only ‘semi-thermodynamic’
stability constants ( log10 K 2ο ' ) are available for this species. The reported values are as
V.4 Group 17 (halogen) compounds and complexes
156
follows: log10 K 2ο ' = – (2.97 ± 0.13) [90BJE], log10 K 2ο ' = – (4.25 ± 0.06) [72RET/HUM]
and log10 K 2ο ' = – 3.7 (calculated from the data in [36JOB] by [72RET/HUM]). These
data are not directly comparable since in the calculation of the last two values the activity of water was also taken into account. Thus these values refer to the following equilibrium:
NiBr(H 2 O)5+ + Br– U NiBr2(H2O)4(aq) + H2O(l).
The higher value reported in [90BJE] is reasonable when the very low water
activity in concentrated LiBr or HBr solutions is taken into account.
Two quantitative results are available for the enthalpy of Reaction (V.73)
(Table V-14). The lack of experimental details and some unidentified experimental errors (see Appendix A), however, do not allow selection of accepted values for
∆ r H m (V.73).
Table V-14: Reaction enthalpies reported for Reaction (V.73).
Method
ise-Br
cal
Medium
2 M (Na, Ni)ClO4
3 M Li(ClO4/Br)
t /°C
∆ r H m / kJ·mol–1 (a)
(8 ± 5)
(b)
25
25
(11.7 ± 0.4)
(c)
Reference
[61LIS/WIL]
[74BLO/RAZ]
(a) Reported value.
(b) Calculated using the reported temperature dependence of log10 b1 (V.73), see Appendix A.
(c) Calculated using log10 b1 = – 0.82 from [70MIR/MAK].
V.4.2.5.1
Determination of the Ni2+ – Br– ion interaction coefficient
Because the value of log10 b1ο ((V.73), 298.15 K) given above is low and rather uncertain, the description of the Ni(II)-bromo system in terms of ion-ion interactions may be
in many cases more straightforward. Therefore, the ion interaction coefficient
ε(Ni2+,Br−) has also been derived in this review from the osmotic and mean activity coefficients of NiBr2 solutions, using Equations (V.74) and (V.75), see Section V.4.3.
The only relevant experimental data are reported in [78LIB/MAC]. Goldberg
et al. reported recommended values for activity and osmotic coefficients of NiBr2 solutions [79GOL/NUT], based on the data in [78LIB/MAC]. In [79GOL/NUT] a corrected
data set for the experimentally determined osmotic coefficients was given. Reference
was made to a personal communication from Z. Libus that provided a corrected version
of the erroneous tables in [78LIB/MAC]. Therefore, only the activity coefficients were
taken from [78LIB/MAC].
V.4 Group 17 (halogen) compounds and complexes
157
Plots of log10 γ ± and φ as a function of the molality, m, of NiBr2 (m = Im/3)
are depicted in Figure V-25 and Figure V-26, respectively. The experimental data can
be reasonably described in the concentration range of 0 – 3 m (Im = 0 – 9 m) using
ε(Ni2+,Br–) = (0.268 ± 0.020) kg·mol–1 (see the solid lines in Figure V-25 and Figure V26).
V.4.2.6 Aqueous Ni(II) – iodine complexes
V.4.2.6.1
Aqueous Ni(II) – iodo complexes
No quantitative data are available for the formation of NiI 2q − q complexes. The formation
of NiI+ and NiI2(aq) was postulated in [74KUT/GAL], to explain the polarographic prewaves of Ni(II) reduction in 5 M Ca(NO3)2 solutions containing 0.2 – 1 M KI. The spectrophotometric data reported in [68AND/KHA2] indicate formation of a considerable
amount of NiI 24 − in 9 – 11 M HI solutions (see Appendix A).
Figure V-25: Plot of log10 γ ± versus molality of aqueous NiBr2 solutions at 298.15 K.
Solid line: SIT, ε(Ni2+, Br–) = 0.268; : Experimental data from [78LIB/MAC].
1.2
1
10
logg10 γγ±
±
0.8
0.6
0.4
0.2
0
-0.2
-0.4
0
0.5
1
1.5
2
2.5
3
[NiBr2] / mol·kg−1
3.5
4
4.5
V.4 Group 17 (halogen) compounds and complexes
158
Figure V-26: Plot of osmotic coefficient versus molality of aqueous NiBr2 solutions at
298.15 K. Solid line: SIT, ε(Ni2+, Br–) = 0.268; : Experimental data from
[78LIB/MAC], [79GOL/NUT].
3.2
2.8
φ
2.4
2
1.6
1.2
0.8
0
0.5
1
1.5
2
2.5
3
[NiBr2] / mol·kg
3.5
4
4.5
−1
V.4.3 Determination of the Ni2+ – ClO4− ion interaction coefficient
The ion interaction coefficients can be derived from the ionic strength dependence of
the osmotic and mean activity coefficients of the corresponding electrolyte solutions.
Two papers report relevant experimental data for Ni(ClO4)2 solutions. Libus and
Sadowska used an isopiestic method to determine the osmotic coefficients in 0.1 – 3.5
m Ni(ClO4)2 solutions [69LIB/SAD], and the mean activity coefficients were calculated
from the Gibbs-Duhem equation. Malatesta et al. determined the activity coefficients in
0.0001 – 0.9 m Ni(ClO4)2 solutions from the emf of liquid-membrane cells
[99MAL/CAR]. Goldberg et al. in [79GOL/NUT] have also reported recommended
values for activity and osmotic coefficients of Ni(ClO4)2 solutions, but their data are
based on [69LIB/SAD].
The ion interaction coefficient ε(Ni2+, ClO −4 ) was previously derived in
[89IUL/POR] (ε(Ni2+, ClO −4 ) = 0.375 kg·mol–1) from the experimental data reported in
[69LIB/SAD]. Since no uncertainty assignment is given in [89IUL/POR], and in the
meantime, another relevant data set has been published [99MAL/CAR], the value of
ε(Ni2+, ClO −4 ) has been re-determined in the present review.
V.4 Group 17 (halogen) compounds and complexes
159
According to the specific ion-interaction theory the mean activity coefficient
γ ± of Ni(ClO4)2 can be expressed as (V.74):
log10 γ ± = −
2 A Im
1 + 1.5 I m
+
4
ε(Ni 2 + , ClO −4 ) I m ,
9
(V.74)
while the osmotic coefficient φ reads:
φ = 1−
 2 ln(10)
2 A ln(10) 
1
− 2 ln 1 + 1.5 I m  +
ε (Ni 2 + , ClO −4 ) I m
1 + 1.5 I m −
3
9
1.5 I m 
1 + 1.5 I m

(V.75)
(
)
with A denoting the Debye-Hückel parameter.
In [69LIB/SAD] the osmotic coefficients have been determined by the isopiestic method.
The value of ε(Ni2+, ClO −4 ) was recalculated based on the smoothed data of Table II in [69LIB/SAD] and using Equation (V.75). The limited validity of Equation
(V.75) was taken into account by weighting schemes which reduce the influence of the
data measured at higher ionic strength. So each data point was either not weighted for
comparison or weighted by the inverse of its x value (I) and the inverse of its x2 value
(I 2), respectively. In Figure V-27 it is shown that the φ data of [69LIB/SAD] can be
reasonably reproduced by Equation (V.75) up to m(Ni2+, ClO −4 ) ≈ 2 mol·kg–1.
In [99MAL/CAR] the activity coefficients have been obtained by measuring
the emf of liquid-membrane cells. In Table III of [99MAL/CAR] the estimated uncertainties are given along with the activity coefficients of Ni(ClO4)2. The value of
ε(Ni2+, ClO −4 ) was recalculated using Equation (V.74). The statistically estimated uncertainty (± 0.004) seems to be rather low. In Figure V-28 experimental data log10 γ ± are
plotted versus m(Ni(ClO4)2. They fit nicely into the SIT approach when an uncertainty
of ± 0.03 for ε is provided for.
The mean value of the entirely independent investigations (emf measurements
[99MAL/CAR] and isopiestic measurements [69LIB/SAD]) is selected:
ε(Ni2+, ClO −4 ) = (0.37 ± 0.03) kg·mol–1.
The uncertainty is rather pessimistic but takes into account that the weighting
scheme influences the numerical value of ε(Ni2+, ClO −4 ) considerably.
V.4 Group 17 (halogen) compounds and complexes
160
Figure V-27: Osmotic coefficient plotted as a function of molality of aqueous Ni(ClO4)2
solutions at 298.15 K.
3.5
smoothed data of [69LIB/SAD]
3.0
2.5
φ
2.0
1.5
2+
−
weighting scheme ε (Ni , ClO4 )
none
0.393
x
0.375
2
x
0.360
1.0
0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
−1
[Ni(ClO4)2] / mol·kg
Figure V-28: Plot of log10 γ ± vs. molality of aqueous Ni(ClO4)2 solutions at 298.15 K.
0.05
data from [99MAL/CAR]
−1
ε = (0.36 ± 0.03) / kg·mol
0.00
-0.05
log10 γ±
-0.10
-0.15
-0.20
-0.25
-0.30
0.0
0.2
0.4
0.6
0.8
−1
[Ni(ClO4)2] / mol·kg
1.0
V.5 Group 16 (chalcogen) compounds and complexes
161
V.5 Group 16 (chalcogen) compounds and complexes
V.5.1 Sulphur compounds and complexes
V.5.1.1 Nickel sulphides
V.5.1.1.1
Solid nickel sulphides
V.5.1.1.1.1
V.5.1.1.1.1.1
Crystallography and mineralogy of nickel sulphides
Heazlewoodite, Ni3S2(cr)
Ni3S2(cr) can be synthesised by fusing nickel and sulphur (3 Ni : 2 S) in an evacuated
silica glass tube. The crystal structure of this synthetic product was determined by Peacock [47PEA]. It is trigonal, space group: R32, with unit cell dimensions a0 = 5.741 Å,
c0 = 7.139 Å, Z = 1 (for rhombohedral cell a0 = 4.080 Å, α = 89°26’), ρ(calc.) = 5.87
g·cm–3; ρ(exp.) = 5.82 g·cm–3 (according to JCPDS-ICDD card No.8-126). Fleet refined
the crystal structure of artificial α-Ni3S2 using X-ray powder intensity data, and essentially confirmed Peacock’s results a0 = 5.7465 Å, c0 = 7.1349 Å, Z = 1, ρ(calc.) = 5.86
g·cm–3 (according to JCPDS-ICDD card No.30-863) [77FLE].
Heazlewoodite is the natural form of the compound Ni3S2(cr) found intergrown
with magnetite, Fe3O4, in narrow bands in the serpentine at Heazlewood, Tasmania. It
occurs massive and granular, without crystal form or cleavage; it is characterised by the
following properties: fracture even; brittle, hardness 4, lustre metallic, colour pale yellowish bronze, streak light bronze, non-magnetic, opaque.
V.5.1.1.1.1.2
α-Ni7S6
This Ni-sulphide phase is orthorhombic with a0 = 3.274 Å, b0 = 16.157 Å, c0 =
11.359 Å, space group: Bmmb. Based on ρ(exp.) = 5.36 g·cm–3, the unit-cell content is
22.5 Ni and 19.3 S [72FLE]. There are three non-equivalent S sites and five nonequivalent Ni sites per unit cell. Four of the Ni sites are in square pyramidal coordination with S and one is in tetrahedral coordination. Each Ni site is related to at least one
neighbouring Ni site by an interatomic distance similar to that of metallic Ni (2.492 Å).
It is apparent that metallic (Ni-Ni) bonding has been significant in stabilising this structure.
V.5.1.1.1.1.3
Pentlandite, (Fe, Ni)9S8(cr)
Pentlandite (JPCRD card 8-90, [52ROB/BRO]) is an important ore of nickel. It is commonly associated with other sulphides such as pyrite (FeS2), chalcopyrite, (CuFeS2) and
pyrrhotite (Fe1–xS) in basic igneous rock intrusions. The largest deposit of this important
ore of nickel is at Sudbury, Ontario. Important deposits are mined in Manitoba, Russia,
and Western Australia. However, the pure end-member, Ni9S8, of pentlandite does not
seem to occur in nature and has not been synthesised [87FLE], thus no thermodynamic
data are proposed for this phase in the present review.
162
V.5.1.1.1.1.4
V.5 Group 16 (chalcogen) compounds and complexes
Godlevskite, Ni9S8(cr)
Godlevskite forms light yellow microscopic crystals with metallic lustre and gray
streak. The hardness is 4 – 5. Fleet showed that godlevskite has a crystal structure based
on a distorted cubic close-packed array of 32 S atoms per unit cell, with 20 Ni atoms in
tetrahedral coordination and 16 in square-pyramidal coordination [87FLE]. The ideal
stoichiometry was established to be Ni9S8(cr). The empirical formula of the specimen
was (Ni8.7Fe0.3)S8, the unit cell is orthorhombic, characterised by a0 = 9.3359 Å, b0 =
11.2185 Å, c0 = 9.4300 Å, Z = 4, space group: C222, ρ(calc.) = 5.273 g·cm–3.
Originally the X-ray single-crystal and powder diffraction patterns of
godlevskite were reported to be consistent with the space groups C222, Cmm2, Amm2,
and Cmmm, with a0 = 9.18 Å, b0 = 11.29 Å, c0 = 9.47 Å, and the stoichiometric formula
was assumed to agree with β-Ni7S6 (according to JCPDS-ICDD card No. 8-107)
[69KUL/ERS].
V.5.1.1.1.1.5
Millerite, β-NiS
Millerite is a low temperature hydrothermal mineral found in cavities in carbonate rocks
and as an alteration product of other nickel minerals. The high temperature form α-NiS
has not been found in nature. The β-NiS phase is rhombohedral with hexagonal axes a0
= 9.620 Å; c0 = 3.149 Å, space group: R3m, Z = 9, ρ(calc.) = 5.374 g·cm–3 (according to
JCPDS-ICDD card No.12-41). The crystal structure of β-NiS has been refined by Grice
and Ferguson [74GRI/FER] as well as Rajamani and Prewitt [74RAJ/PRE]. The colour
of the mineral is brassy yellow, its luster is metallic, crystals are opaque, fracture is uneven, hardness is 3 – 3.5, ρ(exp.) = 5.3 – 5.5 g·cm–3, the streak is dark green to almost
black. Associated minerals are calcite, CaCO3, quartz, SiO2, dolomite, CaMg(CO3)2,
bravoite (basically a nickel-rich pyrite), Fe0.7Ni0.2Co0.1S2, chalcopyrite, CuFeS2, grossular, Ca3Al2(SiO4)3, fluorite, CaF2, and pyrrhotite, Fe1–xS. Notable occurrences include
south central Indiana, Keokuk, Iowa, Gap Mine, Pennsylvania and Sterling Mine, New
York, USA; Freiberg, Germany; Glamorgan, Wales, England and Sherbrooke and
Planet Mines, Quebec.
V.5.1.1.1.1.6
Polydymite, Ni3S4(cr)
Gricaenko et al. [53GRI/SLU] investigated a sample from the Ural mountains, Russia,
with respect to its X-ray powder patterns and synthesised polydymite, Ni3S4, from
aqueous solution. The d values of the natural mineral and the artificial product agreed
with each other. Polydymite has a spinel structure, the dimension of the cubic unit cell
is a0 = 9.48 Å, Z = 8, space group: Fd3m, ρ(calc.) = 4.83 (4.75) g·cm–3 (the numerical
value in parenthesis agrees with Ni3S4), ρ(exp.) = 4.81 g·cm–3 (according to JCPDSICDD card No. 8-107).
The colour of the mineral is violet grey, copper red, light grey, or steel grey,
lustre is metallic, its hardness is 4.5 – 5.5, its cleavage [001] is indistinct, the streak is
black grey, the crystals are opaque. Polydymite occurs in Heazlewood, Tasmania, Aus-
V.5 Group 16 (chalcogen) compounds and complexes
163
tralia, Port Radium, Great Bear Lake, Canada, and in various sites in Germany and
USA. Examples for the latter localities are the Grüne Au Mine, Siegen District, Rhineland-Palatinate and the Gold Hill district, Boulder, Colorado.
V.5.1.1.1.1.7
Vaesite, NiS2(cr)
The synthesis and X-ray diffraction pattern of NiS2(cr) has been reported by de Jong
and Willems [27JON/WIL]. This nickel sulphide has a pyrite, FeS2, type structure. Material approaching NiS2(cr) in composition and having a pyrite like lattice has been
found in the Kasompi mine of Zaire (formerly the Belgian Congo) by Johannes Vaes of
the Union Minière du Haut Katanga, whom it was named after. Vaes was an important
contributor to the mineral development of Zaire.
Two specimens close to the composition NiS2(cr), but containing minor
amounts of Fe or Fe and Co, were examined microscopically, by an X-ray study, and
analysed chemically [45KER]. The crystal structure is cubic, space group PA3, with
lattice constant a0 = 5.670 Å, Z = 4, ρ(calc.) = 4.45 g·cm–3 (according to JCPDS-ICDD
card No. 11-99). The density listed in this card has been calculated for the natural mineral. For pure NiS2(cr) a slightly different value, ρ(calc.) = 4.476 g·cm–3, would be obtained with a0 = 5.670 Å. The colour of the mineral is grey, its lustre is metallic, the
crystals are opaque, its cleavage [001] is perfect. Vaesite occurs with pyrite randomly
disseminated in dolomite. The type locality is the Kasompi mine, Katanga, Zaire.
V.5.1.1.1.2
Ni – S phase diagram
The phase diagram of the binary system Ni – S, given by Elliott [65ELL], is based on a
systematic investigation of the phase relationships in this system performed by Kullerud
and Yund [62KUL/YUN] as well as Sokolova [56SOK]. Craig and Scott [76CRA/SCO]
presented a phase diagram which is in accordance with experimental studies of Kullerud
and Yund [62KUL/YUN] as well as Arnold and Malik [75ARN/MAL]. Apart from the
elements the following mixture phases and stoichiometric compounds were reported:
liquid phase (miscibility gap at high mole fractions of sulphur), Ni3±xS2 (nonstoichiometric high-temperature modification), Ni1–xS(α) (hexagonal, nonstoichiometric high-temperature form), Ni7±xS6 (non-stoichiometric high-temperature
modification), Ni7S6 (stoichiometric low-temperature modification), Ni3S2 (heazlewoodite), NiS(β) (millerite), Ni3S4 (polydymite), and NiS2 (vaesite). Whereas a
monotectic reaction between NiS2 and the two liquid phases of the miscibility gap was
proposed in [62KUL/YUN], Arnold and Malik [75ARN/MAL] re-investigated the
phase relations for the sulphur-rich part of the phase diagram and found a syntectic
equilibrium between NiS2 and two liquid phases, NiS2 U L1 + L2. In 1980 Sharma and
Chang [80SHA/CHA] optimised the Gibbs energy functions of all phases in the system
Ni – S in order to calculate the phase diagram for the first time. Based on experimental
work of Rau [76RAU] as well as Lin et al. [78LIN/HU] the high-temperature modification Ni3±xS2 was shown to consist of two non-stoichiometric phases, viz. β1-Ni3S2 and
β2-Ni4S3, where both phases exhibit a broad homogeneity range.
V.5 Group 16 (chalcogen) compounds and complexes
164
In contrast to earlier investigations [54ROS], [62KUL/YUN] the composition
of the stoichiometric low-temperature phase, coexisting with Ni3S2 and NiS(β), has
been found to be Ni9S8 [94STO/FJE], [96SEI/FJE]. Moreover, the thermodynamic
properties of Ni7±xS6, Ni9S8, NiS (high- and low-temperature forms), and Ni3S2 were reinvestigated experimentally by Stølen et al. [91STO/GRO], [94STO/FJE],
[95GRO/STO]. In particular, in the case of NiS the enthalpy of transition estimated by
Biltz et al. [36BIL/VOI] was revised and a new value was obtained by [95GRO/STO]
based on careful heat capacity measurements employing an adiabatic-shield calorimeter.
Thus, a new optimisation of the Gibbs energy functions of some phases in the system
Ni – S has been necessary in order to recalculate the phase diagram and to recommend
reliable thermodynamic data for the solid nickel sulphides.
The phase diagram of the binary system Ni – S is depicted in Figure V-29 and
Figure V-30. The calculation of the phase diagram is based on Gibbs energy equations
listed in Table V-15. It is worth mentioning that the small homogeneity range of Ni7S6
has been neglected, i.e., this phase is treated as a stoichiometric compound.
Figure V-29: Phase diagram for the binary system Ni – S with T plotted versus mole
fraction of sulphur. ○ [10BOR], ■ [54ROS], ▲ [70NAG/ING], V [75MEY/WAR], ×
[75RAU2].
1800
1600
T/K
1400
1200
1000
800
600
0.0
0.1
0.2
0.3
0.4
0.5
xS
0.6
0.7
0.8
0.9
1.0
V.5 Group 16 (chalcogen) compounds and complexes
165
Figure V-30: Section of the calculated phase diagram for the composition range
0.30 < xS < 0.55. Dashed lines have been extrapolated, indicating regions where phase
equilibria cannot be calculated by using the Gibbs energy functions of Table V-15.
α-NiS
1100
1000
liquid
β1-Ni3S2
β2-Ni4S3
800
600
500
0.30
0.35
0.40
0.45
β-NiS
Ni9S8
Ni3S2
700
Ni7S6
T/K
900
0.50
0.55
xS
Table V-15: Gibbs energy functions for the binary system Ni – S relative to metallic
nickel, Ni(s), and gaseous sulphur, S2(g); ∆ f Gmο = a + bT + cT ln(T), Li = a + bT +
cT ln(T).
Range of validity
a (J·mol–1)
b (J·K–1·mol–1) c (J·K–1·mol–1)
Liquid phase (three-suffix Margules model)
Ni
900 < T < 1900
17472.000
NiS
900 < T < 1900
– 123776.00 152.91400
– 13.513000
S
900 < T < 1900
– 65357.000 165.39600
– 13.513000
LNi–NiS
900 < T < 1900;
– 13691.322 2.8470168
0.00000000
– 104089.64 18.798027
0.00000000
28684.680
– 29.748923
0.00000000
35296.291
4.2627929
0.00000000
– 10.120000
0.00000000
0 < xS < 0.45
LNiS–Ni
900 < T < 1900;
0 < xS < 0.45
LNiS–S
900 < T < 1900;
0 < xS < 0.45
LS–NiS
900 < T < 1900;
0 < xS < 0.45
(continued on next page)
V.5 Group 16 (chalcogen) compounds and complexes
166
Table V-15: (continued)
Range of validity
a (J·mol–1)
b (J·K–1·mol–1) c (J·K–1·mol–1)
β1-Ni3S2 (subregular model)
Ni
845 < T < 1040
– 102153.21 216.88700
0.00000000
S
845 < T < 1040
0.00000000
0.00000000
0.00000000
L1
845 < T < 1040;
– 93628.460 – 256.37000
0.00000000
595976.19
0.00000000
0.35 < xS < 0.50
L2
845 < T < 1040;
– 908.72000
0.35 < xS < 0.50
β2-Ni4S3 (subregular model)
Ni
845 < T < 1040
307930.46
– 177.62880
0.00000000
S
845 < T < 1040
0.00000000
0.00000000
0.00000000
L1
845 < T < 1040;
– 896420.15 515.25100
0.00000000
– 500648.48 150.70800
0.00000000
0.35 < xS < 0.50
L2
845 < T < 1040;
0.35 < xS < 0.50
NiS(α) (sublattice model)
Ni:S
500 < T < 1250
– 161700.00 216.85000
– 18.707460
Va:S
780 < T < 1250
– 108112.14 106.48130
0.00000000
LVa-Va
780 < T < 1250
– 296408.36 3313.7040
– 419.61114
Ni3S2
500 < T < 850
– 331689.70 163.98400
0.00000000
Ni7S6
675 < T < 850
– 9005.7621 – 7709.455
1064.58400
NiS(β)
500 < T < 700
– 155944.00 86.675500
0.00000000
Ni9S8
500 < T < 700
– 1259000.0 667.67880
0.00000000
Ni3S4
500 < T < 630
– 560503.30 338.15692
0.00000000
NiS2
600 < T < 1250
– 244721.70 160.65761
0.00000000
Stoichiometric phases
V.5.1.1.1.2.1
Ni – S melt
Sharma and Chang [80SHA/CHA] proposed a three-suffix Margules model for the excess Gibbs energy function of the Ni – S melt. In the present review this concept is
likewise adopted. The molar excess Gibbs energy is given by the sum of binary interactions between the three species (Ni, S and NiS) of the liquid phase
Gmex = ∑ xi x j ( Li xi + L j x j ) ,
(V.76)
i, j
where the temperature dependence of the binary interaction parameters is defined as
Li = a + bT + cT ln(T).
(V.77)
The liquidus line in the Ni-rich part of the phase diagram agrees excellently
with various experimental studies, as can be seen in Figure V-29. Hence, the Gibbs en-
V.5 Group 16 (chalcogen) compounds and complexes
167
ergy function of the liquid phase, listed in Table V-15, is consistent with [08BOR],
[10BOR], [54ROS], [70NAG/ING], [73VEN/GEI], [75MEY/WAR].
V.5.1.1.1.2.2
β1-Ni3S2 and β2-Ni4S3
According to Sharma and Chang [80SHA/CHA] the excess Gibbs energy function of
the mixture phases β1-Ni3S2 and β2-Ni4S3, containing the components Ni and S, are
modelled by application of a subregular model
Gmex = xS (1 − xS )[ L1 + L2 (1 − 2 xS )]
(V.78)
with the temperature dependence of the parameters L1 and L2 being likewise given by
Equation (V.77).
The predicted homogeneity range of the high-temperature phases β1-Ni3S2 and
β2-Ni4S3 is in accordance with phase equilibrium studies performed by Bornemann
[10BOR] as well as sulphur partial pressure measurements reported by Rau [76RAU]
and Lin et al. [78LIN/HU].
V.5.1.1.1.2.3
Ni3S2(cr) (heazlewoodite)
The standard enthalpy of formation of heazlewoodite, Ni3S2(cr), was determined directly by drop calorimetry [86CEM/KLE], resulting in,
∆ f H mο (Ni3S2, cr, 298.15 K) = – (217.2 ± 1.6) kJ·mol–1,
which is selected in the present evaluation. In another calorimetry investigation the
reaction between elemental nickel and sulphur did not go to completion [92VID/KOR].
The result disagreed with [86CEM/KLE], and was not considered further in this review.
Heat capacity measurements at low temperatures were performed by Weller
and Kelley [64WEL/KEL] and Stølen et al. [91STO/GRO]. From these data the standard entropy at 298.15 K can be deduced, leading to S mο (Ni3S2, cr, 298.15 K) =
(133.89 ± 0.84) J·K–1·mol–1 [64WEL/KEL] and S mο (Ni3S2, cr, 298.15 K) = 133.2
J·K−1·mol–1 [91STO/GRO], respectively. Since these values agree well with each other,
S mο (Ni3S2, cr, 298.15 K) = (133.5 ± 0.7) J·K–1·mol–1
is selected for the standard entropy of Ni3S2(cr). High temperature heat capacity measurements were carried out by Ferrante and Gokcen [82FER/GOK] and Stølen et al.
[91STO/GRO]. From these two data sets a heat capacity function valid from 298.15 to
835 K has been obtained in the present assessment, see Appendix A for details. At
298.15 K the heat capacity is reported to be 118.2 J·K–1·mol–1 [91STO/GRO], 117.7
J·K–1·mol–1 [82FER/GOK], and 117.65 J·K–1·mol–1 [64WEL/KEL]. The heat capacity at
298.15 K, selected in the present assessment, corresponds to:
C pο,m (Ni3S2, cr, 298.15 K) = (118.2 ± 0.8) J·K–1·mol–1.
C
ο
p ,m
The error limits (± 0.8 J·K–1·mol–1) are chosen such that the selected value for
is consistent with all experimental data. The above selections lead to:
V.5 Group 16 (chalcogen) compounds and complexes
168
∆ f Gmο (Ni3S2, cr, 298.15 K) = – (211.2 ± 1.6) kJ·mol–1.
Figure V-31 shows a plot of the Gibbs energy of reaction for:
3Ni(cr) + 2H2S(g) U 2H2(g) + Ni3S2(cr)
(V.79)
as a function of temperature. The experimental data reported by Rosenqvist [54ROS]
and Liné and Lafitte [63LIN/LAF] are in close agreement with calculated values using
the present selection of thermodynamic properties of Ni3S2(cr).
Figure V-31: Gibbs energy of reaction for (V.79) plotted versus temperature.
● [54ROS], solid line: calculation using the thermodynamic quantities selected in this
review.
-90
3 Ni + 2 H2S(g) U Ni3S2 + 2 H2(g)
∆rG°m / kJ·mol–1
-95
-100
-105
-110
660
680
700
720
740
760
780
800
820
T/K
V.5.1.1.1.2.4
Ni7S6(cr) and Ni9S8(cr)
The standard entropy of Ni9S8(cr) was determined from low temperature heat capacity
measurements by Stølen et al. [94STO/FJE]. Unfortunately, the thermodynamic properties were measured on a mixture of (⅓Ni3S2 + ⅔Ni9S8) rather than on pure Ni9S8(cr).
Taking into account the low-temperature heat capacity function of Ni3S2(cr)
[91STO/GRO] the standard entropy at 298.15 K is calculated to be:
S mο (Ni9S8, cr, 298.15 K) = (481 ± 7) J·K–1·mol–1.
V.5 Group 16 (chalcogen) compounds and complexes
169
This quantity is selected with fairly large error bounds because the entropy was
not determined directly on pure Ni9S8(cr). The heat capacity function valid from 298.15
to 640 K is likewise obtained from measurements on the mixture of (⅓Ni3S2 + ⅔Ni9S8)
carried out by the same authors [94STO/FJE], see Appendix A for details. No additional
investigations on Ni9S8(cr) are reported in literature so far. The selected heat capacity of
Ni9S8 at 298.15 K is:
C pο,m (Ni9S8, cr, 298.15 K) = (407 ± 5) J·K–1·mol–1.
According to Cemic and Kleppa [86CEM/KLE] the standard enthalpy of formation of the high-temperature form of Ni7S6(cr) at 298.15 K can be determined by
drop calorimetry, where the reaction temperature was 833 K, yielding ∆ f H mο =
− (582.8 ± 5.7) kJ·mol–1. However, Seim et al. [96SEI/FJE] state that the hightemperature form of Ni7S6(cr) cannot be obtained by quenching samples from temperatures above 675 K to room temperature. The formation of several metastable phases of
different structures was reported. If, in turn, the quenched samples of Cemic and Kleppa
consisted of a mixture of ⅓Ni3S2 + ⅔Ni9S8 (which should form from Ni7S6(cr) below
675 K according to the phase diagram), the standard enthalpy of formation of this mixture at 298.15 K would amount to – 579 kJ·mol–1 which coincides remarkably well with
the experimental result of Cemic and Kleppa. The value – 579 kJ·mol–1 is based on a
combination of the standard enthalpy of formation for Ni3S2(cr) with that for Ni9S8(cr)
( ∆ f H mο = – 760 kJ·mol–1) which is consistent with the Gibbs energy equation (Table V15) used for the calculation of the phase diagram (Figure V-29). Thus, for the standard
enthalpy of formation of Ni9S8(cr) at 298.15 K,
∆ f H mο (Ni9S8, cr, 298.15 K) = – (760 ± 9) kJ·mol–1
can be selected, introducing large error bounds because no direct experimental determination of this quantity has been performed so far. The above selections lead to:
∆ f Gmο (Ni9S8, cr, 298.15 K) = – (746.8 ± 9.3) kJ·mol–1.
In Figure V-32 the temperature dependence of the partial pressure of sulphur is
depicted for phase equilibria involving Ni9S8(cr), Ni7S6(cr), Ni3S2(cr), and β2-Ni4S3.
Usually, the partial pressure of sulphur is obtained from measurements of the ratio
p(H2S)/p(H2) after a hydrogen containing gas phase has been equilibrated with the solid
phases. Figure V-32 reveals that calculations of phase equilibria, using the present
thermodynamic model, coincide well with experimental data taken from various sources
[39SCH/FOR], [52SUD], [54ROS], [78LIN/HU]. The present model disagrees, however, with the experimental data of Jellinek and Zakowski [25JEL/ZAK], probably because they investigated inadequately defined nickel phases.
V.5 Group 16 (chalcogen) compounds and complexes
170
Figure V-32: Partial pressure of sulphur for various phase equilibria in the binary system Ni – S plotted against temperature. V [39SCH/FOR], ▲ [52SUD], ● [54ROS], □
[78LIN/HU], solid lines: calculations using the thermodynamic data of the present
assessment.
-6
-7
log10 [p(S2) / bar]
-8
-9
Ni7S6
-10
β2-Ni4S3
-11
Ni3S2
Ni9S8
-12
-13
-14
600
650
700
750
800
850
T/K
V.5.1.1.1.2.5
α−NiS and β−NiS
The high-temperature modification of NiS crystallises in a hexagonal NiAs-type structure and exhibits a wide range of homogeneity which was studied extensively as a function of temperature by Rau [75RAU2]. Based on the experimental data of Rau a statistical thermodynamic model [69LIG/LIB], [77LIN/IPS] was introduced by Sharma and
Chang [80SHA/CHA], describing the defect interactions in this non-stoichiometric
phase. It can be shown that the statistical thermodynamic model applied in
[80SHA/CHA] is equivalent to a sublattice model which is used in the present assessment. α-NiS is envisaged to be composed of two sublattices, one of which contains only
sulphur atoms and the other which consists of both nickel atoms and vacancies. This
takes into account that deviations from the ideal stoichiometry (NiS) lead exclusively to
nickel deficient compositions. The Gibbs energy function for this phase can be written
as:
ο
ο
Gm = yNi y S GNi:S
+ yVa y S GVa:S
+ RT ( yNi ln yNi + yVa ln yVa + yS ln yS ) + Gmex
(V.80)
V.5 Group 16 (chalcogen) compounds and complexes
171
and the non-ideal interaction between vacancies on the nickel sublattice can be expressed as :
2
Gmex = yVa
LVa − Va
(V.81)
The symbols yNi and yVa denote the site fractions of nickel atoms and vacancies,
respectively, on the nickel sublattice and the site fraction of sulphur atoms on the sulphur sublattice is always equal to yS = 1. The temperature dependence of the defect interaction parameter LVa–Va is given by Equation (V.77). The Gibbs energies of the referο
ο
ence states GNi:S
and GVa:S
refer to the nickel sublattice completely occupied by Ni atoms (stoichiometric composition NiS) and vacancies (empty nickel sublattice), respectively.
The homogeneity range of α-NiS, predicted by the thermodynamic model presented in Table V-15, agrees well with experimental data reported by Rau [75RAU2],
see Figure V-29. The standard enthalpy of formation of NiAs-type α-NiS (stoichiometric composition) was determined by Cemic and Kleppa [86CEM/KLE] employing drop
calorimetry,
∆ f H mο (NiS, α, 298.15 K) = – (88.1 ± 1.0) kJ·mol–1
and this value is selected in the present data assessment. The enthalpy of α-NiS formation determined by [71ARI/MOR] agrees with the selected value within the limits of
uncertainty, but is less well documented. As the high-temperature modification of Ni(II)
sulphide can be quenched easily to room temperature, the heat capacity function of
stoichiometric α-NiS can be derived from calorimetric measurements [95GRO/STO]
over the range from 298.15 to 900 K, see Appendix A. Furthermore, the heat capacity
function of millerite, β-NiS, was determined from 298.15 to 660 K [95GRO/STO], see
Appendix A for details. Several investigations dealt with the enthalpy of transition between millerite and NiAs-type nickel monosulphide which are summarised in Table V16. The most reliable one seems to be the study of Grønvold and Stølen [95GRO/STO].
Hence, combining this value for the enthalpy of transition with the enthalpy of formation of α−NiS and the pertinent heat capacity functions leads to the selected value for
the standard enthalpy of formation of β-NiS:
∆ f H mο (NiS, β, 298.15 K) = – (94.0 ± 1.0) kJ·mol–1.
The standard entropy of millerite was determined at 298.15 K by application of
low temperature heat capacity measurements [64WEL/KEL],
S mο (NiS, β, 298.15 K) = (52.97 ± 0.33) J·K–1·mol–1
which is likewise selected in the present evaluation. Combining this value with the entropy of transition, ( ∆ trs H mο /Ttrs) = (9.99 ± 0.08) J·K–1·mol–1 [95GRO/STO], and the
respective heat capacity functions results in the standard entropy for stoichiometric
α−NiS,
S mο (NiS, α, 298.15 K) = (60.89 ± 0.34) J·K–1·mol–1.
172
V.5 Group 16 (chalcogen) compounds and complexes
According to Weller and Kelley [64WEL/KEL] the heat capacity of β-NiS at
298.15 K amounts to 47.11 J·K–1·mol–1 which is in excellent agreement with the result
of Grønvold and Stølen [95GRO/STO],
C pο,m (NiS, β, 298.15 K) = (47.08 ± 0.30) J·K–1·mol–1,
selected in the present review. In addition, it should be mentioned that the heat capacity
of millerite was likewise measured in the past by Tilden [02TIL], [02TIL2], C pο,m =
47.35 J·K–1·mol–1, and Regnault [1841REG], C pο,m = 48.61 J·K–1·mol–1, respectively. The
error limits of the selected value (± 0.30 J·K–1·mol–1) are chosen such that the selected
value for C pο,m is consistent with the experimental data of [02TIL], [02TIL2],
[64WEL/KEL] and [95GRO/STO]. The selected value for the heat capacity of
stoichiometric α-NiS at 298.15 K is:
C pο,m (NiS, α, 298.15 K) = (49.76 ± 0.30) J·K–1·mol–1.
Table V-16: Temperatures and enthalpies of transition for β-NiS U α-NiS.
Ttrs / K
∆ trs H mο ( kJ·mol–1)
669
2.6
[36BIL/VOI]
652
6.4
[77CON/SRI]
Reference
667
5.9
[80DUB/CLA]
660
6.6
[95GRO/STO]
The temperature dependence of the Gibbs energy function for stoichiometric
α−NiS, i.e., the nickel sublattice is completely occupied, is shown in Figure V-33. Taking the enthalpy of formation, determined by Cemic and Kleppa [86CEM/KLE], and the
C p ,m function as well as the standard entropy of α-NiS, measured by Grønvold and
Stølen [95GRO/STO], the ∆ f Gm function of stoichiometric α-NiS does not match the
values proposed by Sharma and Chang [80SHA/CHA], as can be deduced from Figure
V-33. The Gibbs energy function for stoichiometric α-NiS, retained in the present review (cf. Table V-15), is consistent with both the high temperature data of Sharma and
Chang and the experimental results obtained at temperatures below 850 K according to
Cemic and Kleppa as well as Grønvold and Stølen. The Gibbs energy functions for the
species S (empty nickel sublattice) and the parameter for the non-ideal interaction between vacancies on the nickel sublattice are adopted from [80SHA/CHA].
The above selections lead to:
∆ f Gmο (NiS, α, 298.15 K) = – (87.8 ± 1.0) kJ·mol–1
and
∆ f Gmο (NiS, β, 298.15 K) = – (91.3 ± 1.0) kJ·mol–1.
V.5 Group 16 (chalcogen) compounds and complexes
173
Figure V-33: Gibbs energy function for the stoichiometric composition of NiAs-type
Ni(II) sulphide, α-NiS, plotted versus temperature.
-50
Ni + 1/2 S2(g) U NiS
-60
–1
∆fG°
/ kJ·mol
∆mfG°
/ kJ·mol
m
−1
-70
-80
-90
-100
[86CEM/KLE],[95GRO/STO]
[80SHA/CHA]
optimized line
-110
-120
-130
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
T/K
V.5.1.1.1.2.6
NiS2(cr), vaesite
Cemic and Kleppa [86CEM/KLE] measured the standard enthalpy of formation of vaesite, NiS2(cr), at 298.15 K by drop calorimetry and obtained ∆ f H mο (NiS2, cr, 298.15 K)
= – (124.9 ± 1.0) kJ·mol–1. The optimised Gibbs energy function of NiS2(cr) listed in
Table V-15 is based on decomposition pressure measurements [64LEE/ROS] and phase
equilibrium studies [75ARN/MAL]. An experimental determination of the heat capacity
function at high temperatures has not yet been reported in literature. Applying the estimated heat capacity function for NiS2(cr) published in the JANAF-tables [98CHA],
C pο,m (NiS2, cr) = (64.4378 + 0.020753 (T/K)) J·K–1·mol–1
(V.82)
the Gibbs energy function given in Table V-15 yields for the standard enthalpy of formation and the standard entropy at 298.15 K: ∆ f H mο (NiS2, cr, 298.15 K) = – 121.1
kJ·mol–1 and S mο (NiS2, cr, 298.15 K) = 88 J·K–1·mol–1, respectively. This value for
∆ f H mο agrees reasonably well with that given by [86CEM/KLE].
Ogawa [77OGA] has published low-temperature measurements of the heat capacity of NiS2(cr) from 17 to 347 K. Integration of these results between 0 and 298.15 K
leads to the standard entropy of NiS2(cr), S mο (NiS2, cr, 298.15 K) = (79.6 ± 2.4)
V.5 Group 16 (chalcogen) compounds and complexes
174
J·K−1·mol–1, see Appendix A for details. Taking this value for S mο and the heat capacity
function estimated in the JANAF tables (Equation (V.82)), the re-evaluation of decomposition pressure measurements, reported by Leegaard and Rosenqvist [64LEE/ROS],
results in ∆ f H mο (NiS2, cr, 298.15 K) = – 128.3 kJ·mol–1. This value, however, is not
consistent with the phase diagram data [75ARN/MAL] and the Gibbs energy function
listed in Table V-15. As the quantity S mο (NiS2, cr, 298.15 K) = 79.6 J·K–1·mol–1 is based
on low-temperature heat capacities,
S mο (NiS2, cr, 298.15 K) = (80 ± 8) J·K–1·mol–1
and
∆ f H mο (NiS2, cr, 298.15 K) = – (128 ± 7) kJ·mol–1
are selected. The discrepancy between the Gibbs energy function, necessary to predict
the high-temperature phase equilibria between NiS2(cr), NiS(α) and the liquid phases,
and the selected thermodynamic properties of vaesite is taken into account by the fairly
large error bounds.
From the above selections, the selected Gibbs energy of formation is:
∆ f Gmο (NiS2, cr, 298.15 K) = – (123.8 ± 7.4) kJ·mol–1.
V.5.1.1.1.2.7
Ni3S4(cr), polydymite
Experimental investigations on the thermodynamic properties of polydymite, Ni3S4(cr),
have not yet been published. Hence, thermodynamic data for this compound cannot be
recommended in the present assessment.
V.5.1.1.1.2.8
Ni4.5Fe4.5S8(cr), pentlandite
Cemic and Kleppa [87CEM/KLE] determined the enthalpy of formation of synthetic
pentlandite, Ni4.5Fe4.5S8(cr), by drop calorimetry and obtained ∆ f H mο (Ni4.5Fe4.5S8, cr,
298.15 K) = – (837 ± 15) kJ·mol–1.
V.5.1.1.1.3
Discussion of selected thermodynamic properties of nickel sulphides
A comparison of predicted and experimentally determined temperatures for invariant
three-phase equilibria is given in Table V-17. Except for the peritectoid decomposition
of Ni9S8(cr) into NiS(α) and Ni7S6(cr), the predicted temperatures, employing the thermodynamic quantities of the present review (cf. Table V-15), agree well with experimental data [62KUL/YUN], [75ARN/MAL], [94STO/FJE]. The peritectoid decomposition of Ni9S8(cr) was determined by means of DTA [94STO/FJE]. The discrepancy between the experimental temperature and the predicted value may be caused by (i) an
inconsistency of the present thermodynamic model or (ii) the equilibrium temperature
was overestimated in the DTA experiment, as onset temperatures usually depend on the
heating rate.
V.5 Group 16 (chalcogen) compounds and complexes
175
Table V-17: Temperatures for invariant three-phase equilibria in the binary
system Ni − S.
T / K (calc.)
T / K (exp.)
β2-Ni4S3 + Ni7S6(cr) + Ni3S2(cr)
797.1
797
[62KUL/YUN]
Ni7S6(cr) + β2-Ni4S3 + NiS(α)
850.1
850.5
[94STO/FJE]
Reference
Ni7S6(cr) + Ni3S2(cr) + Ni9S8(cr)
675.3
675.2
[94STO/FJE]
Ni9S8(cr) + NiS(α) + Ni7S6(cr)
677.7
709
[94STO/FJE]
Ni3S4(cr) + NiS(α) + NiS2(cr)
628.8
629
[62KUL/YUN]
liquid + NiS2(cr) + NiS(α)
1259.8
NiS2(cr) + L1 + L2
1305.7
1266
[75ARN/MAL]
1258
[62KUL/YUN]
1295
[75ARN/MAL]
The partial pressure of sulphur for phase equilibria, involving NiS(α), β2-Ni4S3
and NiS2(cr), is plotted versus temperature in Figure V-34. Experimental data from
various sources [36BIL/VOI], [39SCH/FOR], [54ROS], [59LAF], [64LEE/ROS],
[78LIN/HU], [90OSA/ROS] compare well with thermodynamic calculations of the present review (Gibbs energy functions of Table V-15, solid lines) as well as with
[80SHA/CHA] (dashed lines). In the case of NiS2 the selected thermodynamic data
(dotted line) coincides even better with experimental decomposition pressures than that
proposed by [80SHA/CHA].
Finally, thermodynamic data can be selected for the following phases of the
binary system Ni – S: Ni3S2(cr), Ni9S8(cr), NiS(β), NiS(α) (stoichiometric composition),
and NiS2(cr). The heat capacity functions for these compounds are summarised in Table
V-18. The standard enthalpies of formation, the standard entropies and the heat capacities at 298.15 K are listed in Table V-19 to Table V-21. The selected quantities of this
assessment compare reasonably well with those of other data compilations [74MIL],
[95ROB/HEM], [98CHA]. The thermodynamic functions reported by [78SCH/MIL] are
inconsistent with any of these data sets, and thus were disregarded.
V.5 Group 16 (chalcogen) compounds and complexes
176
Figure V-34: Plot of equilibrium partial pressure of sulphur versus temperature. ▼
[36BIL/VOI], V [39SCH/FOR], ▲ [54ROS], ■ [59LAF], ● [64LEE/ROS], ○
[90OSA/ROS], solid lines: calculations using Gibbs energy functions
[78LIN/HU],
of Table V-15, dashed lines: calculations using the thermodynamic model of
[80SHA/CHA], dotted line: calculation using the selected thermodynamic data for NiS2.
0
-1
log10 [p(S2) / bar]
-2
NiS2
-3
-4
α-NiS
-5
-6
β2-Ni4S3
-7
-8
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1000 K / T
Table V-18: Heat capacity functions of nickel sulphides; C pο,m = a + bT + cT 2 + eT –2.
(These are selected values except for NiS2(cr), vaesite).
Compound
Range of validity
a
–1
b
–1
(J·K ·mol )
Ni3S2(cr), heazlewoodite 298.15 < T/K < 835
165.79352
–2
c
–1
–3
(J·K ·mol )
(J·K ·mol )
– 0.10316
0.00012
Ni9S8(cr)
298.15 < T/K < 640 1079.2452
– 2.0599400 0.00198
NiS(β), millerite
298.15 < T/K < 660
– 0.01437
NiS(α)
298.15 < T/K < 660
NiS2(cr), vaesite
54.008
e
–1
3
(J·K ·mol–1)
– 2444624
– 20839639
0.000032282 – 490363
46.676
0.0199807 0.00000000
– 255499.4
660 < T/K < 1000
50.07833
0.0149
0.00000000
– 336916.92
estimated
64.4378
0.020753
0.00000000
0.00000000
V.5 Group 16 (chalcogen) compounds and complexes
177
Table V-19: Standard enthalpy of formation of nickel sulphides at 298.15 K.
∆ f H mο (kJ·mol–1)
Compound
[74MIL]
[95ROB/HEM]
Ni3S2(cr), heazlewoodite – (215.9 ± 12.6) – (216.3 ± 3.0)
Ni9S8(cr)
– (742 ± 38)
NiS(β), millerite
– (94.1 ± 4.2)
[98CHA]
this review
– (216.3 ± 5.0) – (217.2 ± 1.6)
– (760 ± 9)
– (91.0 ± 3.0)
– (87.9 ± 6.3)
NiS(α)
– (94.0 ± 1.0)
– (88.1 ± 1.0)
NiS2(cr), vaesite
– (133.9 ± 8.4)
– (131 ± 17)
– (128 ± 7)
Table V-20: Standard entropy of nickel sulphides at 298.15 K.
S mο (J·K–1·mol–1)
Compound
[74MIL]
[95ROB/HEM]
Ni3S2(cr), heazlewoodite (133.9 ± 2.5) (133.2 ± 0.3)
Ni9S8(cr)
(440 ± 38)
NiS(β), millerite
(53.0 ± 0.8)
[98CHA]
this review
(133.9 ± 0.4) (133.5 ± 0.7)
(481 ± 7)
(53.0 ± 0.4)
(53.0 ± 0.4)
NiS(α)
(52.97 ± 0.33)
(60.89 ± 0.34)
NiS2(cr), vaesite
(67.8 ± 8.4)
(72.0 ± 8.4)
(80 ± 8)
Table V-21: Heat capacity of nickel sulphides at 298.15 K.
C pο,m (J·K–1·mol–1)
Compound
[74MIL]
Ni3S2(cr), heazlewoodite
117.6
[95ROB/HEM]
118.16
[98CHA]
117.7
(407 ± 5)
Ni9S8(cr)
NiS(β), millerite
NiS(α)
V.5.1.1.2
this review
(118.2 ± 0.8)
47.11
47.11
47.11
(47.08 ± 0.30)
(49.76 ± 0.30)
Solubilty of NiS(s) and aqueous nickel hydrogen sulphido species
Apart from theoretical estimations of the stability constants of aqueous nickel sulphide
complexes by Dyrssen [85DYR], [88DYR], [90DYR/KRE], only two experimental
studies on the thermodynamic properties of nickel sulphide complexes are available so
far [94ZHA/MIL], [96LUT/RIC]. These experimental data have been recalculated by
applying the SIT model [97GRE/PLY2], see Table V-22.
V.5 Group 16 (chalcogen) compounds and complexes
178
Table V-22: Recalculated stability constants and SIT parameters for nickel sulphide
complexes at T = 298.15 K and I = 0.
log10 b iο
Reaction
2+
–
(V.83): Ni + HS U NiHS
+
(V.84): 2Ni2+ + HS– U Ni2HS3+
2+
–
(V.85): 3Ni + HS U Ni3HS
5+
∆ε
Reference
(5.18 ± 0.20)
– (0.97 ± 0.39) [94ZHA/MIL], [96LUT/RIC]
(9.92 ± 0.10)
– (0.05 ± 0.22) [96LUT/RIC]
(14.008 ± 0.099)
(0.59 ± 0.22) [96LUT/RIC]
Based on these stability constants the distribution of sulphide-bearing aqueous
Ni(II) species has been calculated as a function of total Ni(II) molality in aqueous solutions saturated with H2S at T = 298.15 K and a constant ionic strength of 1.00 mol·kg–1
NaCl, see Figure V-35. At fairly low pH values high Ni(II) molalities are stable in solutions saturated with H2S. At these conditions ([Ni(II)]tot > 10–3.5 mol·kg–1) the uncommon complexes Ni2HS3+ and Ni3HS5+ become the most dominant species in aqueous
solution. As this situation seems to be unrealistic and no studies on the structure of these
complexes are reported in literature, we select thermodynamic data only for the aqueous
species NiHS+:
log10 b1,1ο ((V.83), 298.15 K) = (5.18 ± 0.20).
This selection, with NEA-TDB selected auxiliary data, yields:
∆ f Gmο (NiHS+, 298.15 K) = – (63.1 ± 2.5) kJ·mol–1.
Owing to the high uncertainty of the second dissociation constant of H2S the
following dissolution reaction of nickel sulphide:
NiS(cr) + 2H+ U Ni2+ + H2S(aq)
*
K sο,0
(V.86)
K sο,0 .
(V.87)
is thermodynamically better defined than
NiS(cr) U Ni2+ + S2–
To the best of our knowledge no reliable studies on the solubility of nickel sulphides in aqueous media have been published yet. Some investigations refer to amorphous solid phases, resulting in too high values for the solubility constants
[14THI/GES], [70CAR/LAI]. In many cases reaction times were far too short to attain
solid-solute equilibrium, yielding values for the solubility constants that are too low
[14THI/GES], [24MOS/BEH], [47DON]. Clearly thermodynamic quantities based essentially on experimental data of [14THI/GES] are unreliable [93MAK/VOE]. Recalculated values for both the solubility constant, log10 *K sο,0 , corresponding to the dissolution reaction (V.86) and the solubility product, log10 K sο,0 , using the currently accepted
value for the second dissociation constant of hydrogen sulphide (H2S(aq) U 2H+ + S2–
log10 K 2ο = – 25.99 [92GRE/FUG]), are listed in Table V-23.
V.5 Group 16 (chalcogen) compounds and complexes
179
Figure V-35: Distribution of nickel sulphide complexes as a function of total molality of
nickel (II) in aqueous solutions saturated with H2S, T = 298.15 K, I = 1.00 mol·kg–1
NaCl. Solid line: Ni2+; dashed line: NiHS+; dotted line: Ni2HS3+; dash-dotted line:
Ni3HS5+.
-2
10
-4
[NixHS2x−1] / mol·kg−1
10
+
NiHS
-6
10
2+
Ni
3+
Ni2HS
-8
10
p(H2S) = 1 bar
-10
10
−1
I = 1.00 mol·kg NaCl
5+
Ni3HS
-12
10
-6
10
-5
10
-4
10
-3
10
−1
[Ni(II)]tot / mol·kg
-2
10
-1
10
180
V.5 Group 16 (chalcogen) compounds and complexes
Table V-23: Solubility constants for Ni(II) sulphide at T = 298.15 K and I = 0.
log10 *K sο,0
log10 K sο,0
Reference
– 0.5
– 27.5
[1883THO], [09BRU/ZAW]
– 3.98
– 30.98
[24MOS/BEH], [31KOL]
– 17.8
[70CAR/LAI]
9.20
– (1.53 ± 0.30)(a)
– (2.10 ± 0.30)
(b)
– 28.5(a)
this review
(b)
this review
– 29.1
(a) NiS(α).
(b) NiS(β), millerite.
The solubility data scatter remarkably and none of these is consistent with the
solubility constants selected in this data assessment. The solubility constant given by
Bruner and Zawadzki [09BRU/ZAW] is based on a calorimetric value for the enthalpy
of reaction for the precipitation of NiS from aqueous solution:
Ni2+ + S2– U NiS,
reported by Thomson [1883THO]. Thiel and Gessner [14THI/GES] carried out a comprehensive study of the dissolution behaviour of NiS prepared by precipitation from
aqueous solutions at various conditions. They found three fractions, viz. I-NiS, II-NiS
and III-NiS, with different solubilities in hydrochloric acid. The variations of the solubility of these fractions, observed by the authors, can be explained in terms of different
dissolution kinetics rather than in terms of different thermodynamic properties, since
reaction times were certainly too short to attain equilibrium between the solid phase and
the aqueous solution. Licht [88LIC2] has recalculated the results of [14THI/GES] and
obtained Gibbs energies of formation for I-NiS, II-NiS and III-NiS, respectively. The
solubilities of Ni(II) sulphides in solutions of hydrochloric acid according to [88LIC2],
[14THI/GES] as well as the present assessment are depicted in Figure V-36. Whereas
the solubility of I-NiS, which seems to correspond to an amorphous material, is too
high, the solubilities of II-NiS and III-NiS are by far too low because of too short reaction times. The fractions II-NiS and III-NiS are apparently crystalline products. This
assumption is likewise confirmed by Dönges [47DON] who has pointed out that, besides amorphous products, the metastable high-temperature modification, NiS(α), is
precipitated from aqueous solutions. Furthermore, it should be mentioned that millerite,
NiS(β), is the only nickel sulphide identified so far in natural low-temperature anoxicsulphidic environments according to Thoenen [99THO].
V.5 Group 16 (chalcogen) compounds and complexes
181
Figure V-36: Solubility of Ni(II) sulphides in aqueous solutions saturated with H2S as a
function of initial concentration of hydrochloric acid at 298.15 K. Dashed lines refer to
I-NiS, II-NiS and III-NiS as experimentally investigated by Thiel and Gessner
[14THI/GES] and thermodynamically interpreted by Licht [88LIC2]. Solid lines correspond to NiS(α) and NiS(β) using the thermodynamic data selected in the present compilation. Dotted lines are related to calculations taking into account the uncommon
aqueous species Ni2HS3+ and Ni3HS5+, respectively.
0
I-NiS
−1
log10([Ni(II)]tot / mol·kg )
-2
-4
-6
-8
-10
-12
α-NiS
β-NiS
II-NiS
III-NiS
-2.0
-1.5
-1.0
-0.5
0.0
0.5
−1
log10([HCl]initial / mol·kg )
V.5.1.2 Nickel sulphates
V.5.1.2.1
Aqueous nickel (II) sulphato complexes
The complexation reactions of Ni2+ with SO 24 − have been the subject of a large number
of investigations (Table V-24). These include studies by cryoscopy [55BRO/PRU],
[56KEN], [58KEN], [71ISO], conductivity [1895FRA], [56FIA/SHE], [73KAT],
[76SHI/TSU],
[79FIS/FOX],
[93SHE],
[2000TSI/MOL],
[74MAK/MAS],
potentiometry
and
other
electrochemical
techniques
[59NAI/NAN],
[2001MOL/TSI],
[61TAN/OGI], [63TAN/SAI], [65POT], [68PRA], [70TUR/RUV], solubility
[70MIR/MAK],
spectroscopy
[64LAR],
[70MIR/MAK],
[67GES/NEU],
[78BEC/LIU],
[80LIB/SAD]
calorimetry
[69IZA/EAT],
[77ASH/HAN],
[74BLO/RAZ], [85DRA/MAD] and isopiestic measurements [80LIB/SAD],
[83HOL/MES]. There is a broad consensus, primarily based on the spectroscopic evidence, that, for temperatures below 100°C, association of these ions in dilute solutions
182
V.5 Group 16 (chalcogen) compounds and complexes
is predominantly an outer-sphere interaction, with sulphate gradually displacing water
molecules in the inner sphere around Ni2+ at higher temperatures and concentrations
[77ASH/HAN], [78BEC/LIU]. Indeed, for 25°C, it is not clearly defined what does and
what does not constitute a nickel sulphate complex. The nature of a specific type of experiment and the interpretation of experimental results affect any proposed value for the
formation constant of NiSO4(aq) according to reaction (V.88):
Ni2+ + SO 24 − U NiSO4(aq).
(V.88)
In particular, the reported value of the formation constant extrapolated to zero
ionic strength is usually dependent on the particular method used to assign values to the
ionic activity coefficients. In the present review, the objective is to assign a value that is
consistent with the use of a particular formulation of the specific ion interaction equations (the SIT approach, cf. Appendix B). Among the assumptions inherent in the SIT
treatment is that although the association constant depends on the total ionic strength of
the solution, on the ionic interactions between Ni2+ and anions of the supporting electrolyte, and on the ionic interactions between SO 24 − and cations of the supporting electrolyte, all specific interactions between Ni2+ and SO 24 − are accounted for by the NiSO4
formation constant. Nevertheless, Holmes and Mesmer [83HOL/MES] have shown that
a Pitzer-type ion-interaction model can be used to provide a satisfactory interpretation
of isopiestic data without assuming formation of a discrete NiSO4 complex, even at
140°C.
Several studies on ionic strength effects have used nickel sulphate as a model
2:2 electrolyte [55BRO/PRU], [73KAT], [75TAM]. In such cases, for the sake of consistency, it has been necessary to abandon the more sophisticated treatments, and to reevaluate experimental data for complexation using the SIT. Perhaps because of the interest in nickel sulphate as a “typical” 2:2 electrolyte in aqueous solution, few of the
association studies have been done as a function of concentration of a supporting electrolyte. Exceptions include the cryoscopic study of Kenttämaa [56KEN], [58KEN] and
the polarographic studies of Tanaka and Ogino, [61TAN/OGI], and Tanaka, Saito and
Ogino [63TAN/SAI]. Some spectroscopic studies have also been done using supporting
electrolytes, often at very high concentrations [70TUR/RUV], [77ASH/HAN].
In the temperature range (0 to 50°C) in which an outer-sphere complex predominates over inner sphere complexes, several studies [59NAI/NAN], [73KAT],
[76SHI/TSU] have been used to determine the enthalpy of the complexation reaction
(V.88). In addition, the enthalpy of reaction has been determined calorimetrically by
Izatt et al. [69IZA/EAT], though a useful value can be obtained from their enthalpy
titrations only if an independent method is used to determine the association constant [73POW] (cf. Appendix A). The comparison of derived values for the enthalpy of
association also provides a tool for evaluating the association constant data.
V.5 Group 16 (chalcogen) compounds and complexes
183
Table V-24: Experimental values for the association constants and other thermodynamic
quantities for nickel sulphate.
Recalculated value(a)
Method
t/°C
Medium (aq)
Reported value
cond
25
0 (extrap.)
K1 = 250
(220.7 ±13.4)
[32MON/DAV],
recalc.of
[1895FRA]
(161.3 ± 20.0)
[55BRO/PRU]
Reference
fpt
0
0.005–0.09 self
K1 = 125–244
fpt
– 0.2
0.0484 m KClO4
K1 = 51
[56KEN]
fpt
– 0.8
0.258 m KClO3
K1 = 19
[58KEN]
– 2.8
1.15 m KNO3
K1 = 4.90
0
< 0.05 m HCl
K1 = 121
(98.0 ± 23.0)
10
< 0.05 m HCl
K1 = 151
(126.8 ±16.0)
15
< 0.05 m HCl
K1 = 174
(166.2 ± 21.8)
25
< 0.05 m HCl
K1 = 211
(172.8 ± 31.9)
35
< 0.05 m HCl
K1 = 247
(210.2 ± 31.2)
45
< 0.05 m HCl
K1 = 289
pot
∆ r H mο = 13.85 kJ·mol–1
[59NAI/NAN]
(236.9 ± 30.0)
see text
∆ r S mο = 90.8 J·K–1·mol–1
pol
25
0.2 M KNO3
K1 = 11.6
[59ROS/ROS],
recalc. of
[58KEN]
[61TAN/OGI]
pol
25
1.0 M NaClO4
K1 = 3.7
[63TAN/SAI]
fpt
– 3.06
1.15 m KNO3
K1 = 5.01
K2 = 33.6
K2 = 26
ir
RT
1.0 M self
β inner sphere / β outer sphere = 0.1
[64LAR]
pot
35
0 (extrap.)
K1 = 111
[68PRA]
cal
25
0 (extrap.)
K1 = 646
∆ r H mο = 1.71 kJ·mol–1
[69IZA/EAT]
see text
∆ r S mο = 59.0 J·K–1·mol–1
spec/sol
25.0
3.0 M LiClO4
K1 = 1.8
pol
25.0
5.0 M NaClO4
K1 = 15.6
fpt
0
0.005–0.03 self
K1 = 118–250
(153.8 ± 20.0)
[71ISO]
0.0
0
K1 = 156
(213.8 ± 4.4)
[73KAT]
5.0
0
K1 = 162
(217.9 ± 4.9)
cond
[70MIR/MAK]
[70TUR/RUV]
10.0
0
K1 = 168
(223.9 ± 4.9)
15.0
0
K1 = 176
(230.3 ± 4.7)
20.0
0
K1 = 183
(238.3 ± 5.3)
25.0
0
K1 = 187
(247.3 ± 5.8)
30.0
0
K1 = 191
(257.1 ± 7.1)
35.0
0
K1 = 201
(267.8 ± 7.8)
40.0
0
K1 = 209
(279.9 ± 8.8)
45.0
0
K1 = 217
25.0
0
∆ r H mο = 5.31 kJ·mol–1
(295.6 ± 9.4)
(5.3 ± 2.0)
(Continued on next page)
V.5 Group 16 (chalcogen) compounds and complexes
184
Table V-24: (continued)
Method
t/°C
cal
25.0
cal
25.0
Medium (aq)
Recalculated value(a)
Reported value
0
∆ r S mο = 61.5 J·K–1·mol–1
∆ r H mο =5.72–5.93 kJ·mol–1
1.5 M LiClO4/
∆ r H mο = 2.59 kJ·mol–1
Reference
[73POW],
recalc of
[69IZA/EAT]
[74BLO/RAZ]
0.75M Li2SO4
var
25.
0.018 self +?
K1 = 280
0.048 self +?
K1 = 170
[74MAK/MAS]
0.069 self +?
K1 = 110
cond
0.0
0.
K1 = 270
cond
15.0
0.
K1=457 (0.1 MPa)
[75TAM],
recalc. of
[73KAT]
(187.7 ± 2.1)
[76SHI/TSU]
K1=227 (160 MPa)
25.0
0.
(193.8 ± 1.9)
K1=490 (0.1 MPa)
K1=269 (160 MPa)
40.0
0.
(226.7 ± 2.8)
K1=575 (0.1 MPa)
K1=313 (160 MPa)
uv
25.0
0
25.0
0
25.0
0
∆ r S mο = 109 J·K–1·mol–1
(0.101 MPa)
∆1V° = 8.6 cm3·mol–1
47
K1=6.8
60
K1=5.5
25.0
15.0
[77ASH/HAN]
K1=5.4
0
∆ r H mο = 5.31 kJ·mol–1
∆ r S mο
cond
(6.0 ± 2.5) kJ·mol–1
(0.101 MPa)
25.0(b) 5 M Na2SO4/NaClO4 K1= 5.89
70
cond
∆ r H mο = 6.90 kJ·mol–1
0.
–1
[77KAT]
recalc.of
[73KAT]
–1
= 61.5 J·K ·mol
K1=200 (0.1 MPa)
(183.3 ± 30.4)
[79FIS/FOX]
K1=105 (203 MPa)
25.0
0.
K1=196 (0.1 MPa)
(199.4 ± 28.5)
K1=93.5 (203 MPa)
25.0
cond
0
∆1V° = 11.6 cm3·mol–1
K1=169
25.0
0.
25.0
0
cond
25.0
< 0.06 M
K1=230.1
cond.
20.0
0
K1=139
∆ r H mο = 5.24 kJ·mol–1
[81YOK/YAM]
recalc. of
[73KAT]
∆ r S mο = 60.2 J·K–1·mol–1
[93SHE]
[2000TSI/MOL],
[2001MOL/TSI]
(a) Uncertainties are the statistical uncertainties from the re-analysis of the data. The overall estimated uncertainties are discussed in the text.
(b) This value was reported by the authors based on an extrapolation to 25°C.
V.5 Group 16 (chalcogen) compounds and complexes
185
Table V-25 provides a summary of the results from re-analysis of the conductance data for NiSO4(aq) using the series form [80PET/TAB]1 of the Lee-Wheaton
equation [79LEE/WHE], modified so that the distance parameter is forced to the appropriate value for the SIT model used in the TDB. No value could be recalculated from
the work of Shehata [93SHE] because values of the concentration-dependent conductances were not reported. Results from the other conductance studies at 25°C are compared in Figure V-37. The molar conductances agree well at low concentrations at 25°C.
At higher concentrations (≥ 10–3 M), values are more divergent, especially those from
the early study of Frank [1895FRA]. The values of Fisher and Fox [79FIS/FOX] and
Shimizu, Tsuchihashi and Furumi [76SHI/TSU] are in very good agreement. As can be
seen from Table V-25, the agreement is slightly poorer for 15°C. Values of Λo from
analysis of the molar conductances from these two studies are substantially smaller than
the values calculated from the molar conductances reported by Katayama [73KAT].
Nevertheless, all four studies lead to values of log10 K ο ((V.88), 298.15 K) between 2.29
and 2.39.
Table V-25: Results of reanalysis of NiSO4(aq) conductance data using a form of the
Lee-Wheaton equation that is compatible with the SIT procedure.
t/°C
0
KA(a)
± σKA
Assessed
(statistical)
(2σ)
log10KA
Λo
Reference
213.8
± 4.4
2.33
68.82
[73KAT]
5
217.9
± 4.9
2.34
80.48
[73KAT]
10
223.9
± 4.9
2.35
92.96
[73KAT]
15
230.3
± 4.7
2.36
106.01
[73KAT]
183.3
± 30.4
2.26
105.28
[79FIS/FOX]
187.7
± 2.1
2.27
103.63
[76SHI/TSU]
238.3
± 5.3
2.38
119.82
[73KAT]
118.9
± 7.2
2.08
123.34
[2000TSI/MOL]
247.3
± 5.8
2.39
134.35
[73KAT]
220.7
± 13.4
60
2.34
136.57
[1895FRA]
193.8
± 1.9
30
2.29
132.99
[76SHI/TSU]
199.4
± 28.5
60
2.30
132.73
[79FIS/FOX]
30
257.1
± 7.1
2.41
149.50
[73KAT]
35
267.8
± 7.8
2.43
165.16
[73KAT]
40
279.9
± 8.8
2.45
181.47
[73KAT]
226.7
± 2.8
2.36
179.12
[76SHI/TSU]
295.6
± 9.4
2.47
198.58
[73KAT]
20
25
45
20
(a) The constant KA from the Lee-Wheaton treatment is assumed to be equivalent to K ο (V.88).
1
In 1979, Dr. A.D. Pethybridge kindly provided one of the reviewers (R.J. L.) with a version of the computer program that has been used here (with modifications) for the analysis of conductance data.
V.5 Group 16 (chalcogen) compounds and complexes
186
As discussed in Appendix A, it appears that the reported molar conductance
values from at least two of the studies [73KAT], [76SHI/TSU] are for “rounded” concentration values. If the values reported are “smoothed” molar conductances, the true
experimental scatter may have been greater. The “statistical” uncertainties listed in
Table V-25 are the values from the data re-analysis [80PET/TAB]. Fitting of the
(unavailable) primary data from those studies probably would have led to larger
calculated statistical uncertainties in KA (assumed equivalent to K ο (V.88)). Estimated
(2σ) uncertainties for all four values of K ο ((V.88), 298.15 K) are listed in Table V-25.
Based solely on the conductance data at 25°C, at I = 0 the value of K ο ((V.88),
298.15 K) is (228.1 ± 15.5), or log10 K ο ((V.88), 298.15 K) = (2.36 ± 0.03).
Figure V-37: The variation of molar conductances of aqueous nickel sulphate solutions
:
with concentration at 25°C. : [1895FRA], : [73KAT], : [76SHI/TSU],
[79FIS/FOX]
260
250
molar conductance
Λ / S·cm2·mol−1
240
230
220
210
200
190
180
0
5
10
15
4
20
25
30
35
40
−3
10 ·[NiSO4(aq)] / mol·dm
Earlier reviews [82WAG/EVA], [99ARC] gave substantial weight to the emf
study of Nair and Nancollas [59NAI/NAN]. As shown in Appendix A, recalculation of
values from that study to be consistent with the SIT procedure described in Appendix B,
requires values for the dissociation constant of HSO −4 , and several interaction coefficients, ε(Ni2+,Cl–), ε(H+,Cl–), ε(H+, HSO −4 ) and ε(Ni2+, HSO −4 ). The values calculated
for the formation constant of NiSO4(aq) are extremely sensitive to the activity coefficient model, to the values used for the interaction coefficients and to the values assumed
V.5 Group 16 (chalcogen) compounds and complexes
187
for the first protonation constant for the sulphate ion. As discussed in Appendix A, this
work can be used to determine a value at I = 0 of K ο ((V.88), 298.15 K) = (173 ± 80),
or log10 K ο ((V.88), 298.15 K) = (2.24 ± 0.20).
If the estimated uncertainty bounds are considered, the value from Nair and
Nancollas overlaps with those from the conductance studies. Using the weighted average of the conductance and emf data ((228.1 ± 15.5) and (173 ± 80), respectively,
weighted according to their uncertainties), the selected value of K ο ((V.88), 298.15 K)
is:
K ο ((V.88), 298.15 K) = (226.1 ± 15.2)
log10 K ο ((V.88), 298.15 K) = (2.35 ± 0.03).
or
None of the studies at higher ionic strength at 25°C is a systematic study in
which the concentration of a single supporting electrolyte was varied. The results are
summarised in Table V-26.
Table V-26: Experimental values for the first association constant of nickel sulphate
determined in the presence of supporting electrolytes (see Appendix A). Values of
log10 K1ο were determined using the SIT (Appendix B).
Medium
t / °C
log10 K1
(a)
log10 K1ο
Reference
0.0484 m KClO4
– 0.2
1.70
2.34
[56KEN]
0.258 m KClO3
– 0.8
(1.27 ± 0.30)
(2.35 ± 0.30)
[58KEN]
1.15 m KNO3
– 2.8
(0.69 ± 0.20)
(2.16 ± 0.20)
25
(1.06 ± 0.20)
(2.13 ± 0.20)
[61TAN/OGI]
(1.95 ± 0.20)
[63TAN/SAI]
0.2 M KNO3
1.0 M NaClO4
25
(0.57 ± 0.20)
3.0 M LiClO4
25
0.26
1.07(b)
[70MIR/MAK]
5.0 M NaClO4
25
1.19
1.70(c)
[70TUR/RUV]
(a)
The added nickel sulphate has a substantial effect on the ionic strength and the derived value of log10K1.
(b)
The value is anomalously low compared to all other reported values.
(c)
The ionic strength is beyond the range for use of the SIT (Appendix B).
Except for the spectroscopic value of Mironov et al. [70MIR/MAK], the results
are qualitatively in agreement with the results of the conductivity and emf studies at low
ionic strengths. The association constant values from the polarographic studies
[61TAN/OGI], [63TAN/SAI], [70TUR/RUV] do tend to be somewhat smaller than the
values selected above. The estimated uncertainties in the values from [61TAN/OGI] and
[63TAN/SAI] are large, and the raw data are unavailable for re-analysis. Therefore,
those results are not incorporated in the weighted-average selected value. The cryoscopic results from experiments in different saturated solutions [56KEN], [58KEN] are
in excellent agreement with the value (2.33 ± 0.06) for 0°C from the conductance data
of Katayama and the average value of (2.20 ± 0.06) from the two cryoscopic studies in
188
V.5 Group 16 (chalcogen) compounds and complexes
water [71ISO], [55BRO/PRU]. They are also in marginal agreement with the value
(1.99 ± 0.10) recalculated from the emf data of Nair and Nancollas [59NAI/NAN].
At high sulphate concentrations, there is some evidence for formation of
Ni(SO 4 ) 22 − in several studies (e.g., [58KEN], [59ROS/ROS], [63TAN/SAI]). However,
the evidence is reasonably ambiguous, and may only reflect systematic errors in the
experiments [58KEN]. No value is selected for K 2ο in the present review. There is considerable spectroscopic evidence that an inner-sphere sulphato complex exists at high
sulphate concentrations and may become more important as the temperature is increased
[77ASH/HAN], [78BEC/LIU]. Near 25°C, inner-sphere association is only about 10%
of the outer-sphere association. No values are selected in this review for the formation
constants of the inner sphere sulphato complexes [78BEC/LIU]. The polymeric species
Ni 2 (SO 4 ) 44 − proposed by Brintzinger and Osswald [34BRI/OSS] also is not considered
credible [41KIS/ACS].
V.5.1.2.1.1
Enthalpy of formation of NiSO4(aq)
There have been several studies [59NAI/NAN], [73KAT], [76SHI/TSU] of the temperature dependence of the formation constant of the outer-sphere complex NiSO4(aq), and,
as discussed above, several other measurements at temperatures between – 2.8 and
45°C. The recalculated values of log10 K1ο are plotted in Figure V-38.
The sets of values from most of the conductance and cryoscopic measurements
are in good agreement. The temperature-dependencies of the values from these studies
follow similar parallel curves. The higher temperature results of Nair and Nancollas
[59NAI/NAN] follow the same type of curve for temperatures ≥ 15°C (288.15 K), but
the values for the two lowest temperatures are inconsistent with the cryoscopic and conductivity results.
The recalculated values of log10 K1ο from Katayama [73KAT] are used to calculate (5.3 ± 2.0) kJ·mol–1 for the average value of ∆ r H mο (V.88) between 0 and 45°C.
Similarly, ∆ r H mο = (6.0 ± 2.5) kJ·mol–1 (average for 15 to 40°C) from the sparse data of
Shimizu, Tsuchihashi and Furumi [76SHI/TSU] and ∆ r H mο = (12.6 ± 3.0) kJ·mol–1 (average for 0 to 45°C, 2σ uncertainty) from the work Nair and Nancollas [59NAI/NAN].
Izatt et al. [69IZA/EAT] determined the enthalpy of reaction at 25°C by a titration calorimetry method. The intent was to determine both the association constant and
the enthalpy of association in a single experiment. Unfortunately, the enthalpy and association constant values, as derived iteratively [69IZA/EAT] or by the method of least
squares [73POW], are highly correlated. Powell showed that useful enthalpy values
could be obtained only if a value was assumed for K1ο . Starting from the 25°C value
from Nair and Nancollas [59NAI/NAN], Powell used the data of Izatt et al.
[69IZA/EAT] to obtain ∆ r H mο = (5.8 ± 0.2) kJ·mol–1. The calculation requires values for
the enthalpy of protonation of sulphate and the enthalpy of dilution of the tetramethylammonium sulphate titrant. Izatt et al. [69IZA/EAT] and Powell [73POW] also incor-
V.5 Group 16 (chalcogen) compounds and complexes
189
porated values for an association constant and enthalpy of association for the
Me4NClO4(aq) ion pair. Determination of the values for Me4NClO4(aq) required using
an explicit value for the association constant between Na+ and SO 24 − . As discussed in
Appendix A, a reasonable value from this work is ∆ r H mο = (5.8 ± 2.0) kJ·mol–1, based on
Powell’s recalculation and an increased estimated uncertainty. In the present review, we
exclude the value from Nair and Nancollas [59NAI/NAN], and accept the weighted
average of the results from the other three studies [69IZA/EAT], [73KAT],
[76SHI/TSU]:
∆ r H mο ((V.88), 298.15 K) = (5.66 ± 0.81) kJ·mol–1.
From the selected values of log10 K1ο and ∆ r H mο , the following quantities are
calculated:
∆ f Gmο (NiSO4, aq, 298.15 K) = – (803.2 ± 0.9) kJ·mol–1
∆ f H mο (NiSO4, aq, 298.15 K) = – (958.7 ± 1.3) kJ·mol–1
S mο (NiSO4, aq, 298.15 K) = – (49.3 ± 3.1) J·K–1·mol–1.
and
Figure V-38: Values of log10 K1ο for nickel sulphate association at different
temperatures. : [1895FRA], : [55BRO/PRU], : [58KEN], : [59NAI/NAN], :
[61TAN/OGI], : [63TAN/SAI], : [71ISO], : [73KAT], : [76SHI/TSU], :
[79FIS/FOX], : [2000TSI/MOL].
2.5
2.4
log10K1
o
2.3
2.2
2.1
2.0
270
280
290
300
T/K
310
320
V.5 Group 16 (chalcogen) compounds and complexes
190
V.5.1.2.1.2
Heat capacity values
Drakin, Madanat and Karlina [85DRA/MAD] reported measurements of the apparent
molar heat capacities for nickel sulphate solutions at 50°C, 70°C and 90°C. The data are
too sparse to be used to estimate the partial molar heat capacity of the ion pair.
V.5.1.2.2
Solid nickel sulphates
Several hydrated nickel sulphate solids from NiSO4·7H2O to NiSO4·H2O, including two
forms of NiSO4·6H2O, a tetrahydrate, and a dihydrate have been reported [04STE/JOH],
[39SIM/KNA], [39ROH], [64KOH/ZAS], though the domains of stability for and kinetics of interconversion between some of the intermediate hydrates remain contentious. In
saturated solutions at 1 bar, the heptahydrate converts to the α-hexahydrate just above
300 K. Therefore, samples of the heptahydrate often contain amounts of the hexahydrate, and samples of the α-hexahydrate can have the heptahydrate as a contaminant.
This has complicated attempts to measure the physical properties for the individual solids. Dehydration of the monohydrate to anhydrous NiSO4(s) occurs in a stream of water-saturated N2 only above 630 K [78SIR/SEM]. Decomposition of NiSO4(s) to NiO(s)
and sulphur oxides becomes important at temperatures above 900 K [83GAU/BAL].
V.5.1.2.2.1
NiSO4·7H2O(cr)1
The heptahydrate is the stable nickel sulphate hydrate at 298.15 K. Stout et al.
[66STO/ARC] reported low-temperature heat capacity measurements (from 16.1 K to
temperatures above 279 K) for samples that were slightly hypostoichiometric and hyperstoichiometric in water (NiSO4·6.867H2O(cr) and NiSO4·7.301H2O(cr)) with respect
to the heptahydrate, and combined the results with those from an earlier study
[41STO/GIA] to obtain S mο (NiSO4·7.00H2O, cr, 298.15 K). There do not appear to be
any other comparable measurements for this solid, and the entropy and heat capacity
values for NiSO4·7.00H2O(cr) at 298.15 K from Stout et al. are selected in the present
review. Though some of the auxiliary data used by Stout et al. differ slightly from those
used in the present review, such differences are much less than the rather small estimated experimental uncertainties (cf. Appendix A for [66STO/ARC]). The selected
values are:
S mο (NiSO4·7.00H2O, cr, 298.15 K) = (378.95 ± 1.00) J·K–1·mol–1
C pο,m (NiSO4·7.00H2O, cr, 298.15 K) = (364.6 ± 4.0) J·K–1·mol–1.
The value of the difference between the enthalpy of solution of
α-NiSO4·6.00H2O(cr) and NiSO4·7.00H2O(cr) in water ((7.68 ± 0.10) kJ·mol–1) has
been reported by Stout et al. [66STO/ARC]. In the present review,
∆ sol H mο (NiSO4·6.00H2O, α, 298.15 K) = (4.485 ± 0.200) kJ·mol–1 in water. Using this
and the enthalpy of hydration, the value of ∆ sol H mο (NiSO4·7.00H2O, cr, 298.15 K) =
1
The dissolution of this compound is discussed in Section V.2.1.3.1.
V.5 Group 16 (chalcogen) compounds and complexes
191
(12.17 ± 0.23) kJ·mol–1 in water is obtained. The value 15.644 kJ·mol–1 was reported by
Przepiera et al. [93PRZ/WIS] as the heat of solution of the heptahydrate in water to
produce a solution with a 5 × 10–4 mass ratio of salt to water. If the value is assumed to
be for the enthalpy of solution to this stated dilution (approximately 0.002 m), the application of a heat of dilution correction to I = 0 gives a value (12.5 kJ·mol–1) in good
agreement with that based on Goldberg et al. and Stout et al.
Using the selected values of ∆ f Gmο (Ni2+) = – (45.77 ± 0.77) kJ·mol–1 (Section
V.2.1.1), and ∆ sol Gmο = (12.94 ± 0.11) kJ·mol–1 (Section V.2.1.3) with standard auxiliary
data,
∆ f Gmο (NiSO4·7.00H2O, cr, 298.15 K) = – (2462.7 ± 0.9) kJ·mol–1.
With the selected entropy value,
S mο (NiSO4·7.00H2O, cr, 298.15 K) = (378.9 ± 1.0) J·K–1·mol–1
is calculated and
∆ f H mο (NiSO4·7.00H2O, cr, 298.15 K) = – (2977.3 ± 1.0) kJ·mol–1.
Bell [40BEL] reported vapour pressure measurements that showed that at
298.15 K removal of water from NiSO4·7H2O(cr) is less favoured than removal of D2O
from NiSO4·7D2O (see Appendix A). In this review, no values are selected for the thermodynamic quantities of NiSO4·7D2O.
V.5.1.2.2.1.1
α-NiSO4·6H2O
In contact with saturated solutions, the α-hexahydrate becomes the stable nickel sulphate hydrate near 302 K, and is transformed to the β-hexahydrate at approximately
327 K. Stout et al. [66STO/ARC] reported low-temperature heat capacity measurements
(from 13.7 to 321.0 K) for samples that were slightly hyperstoichiometric
(NiSO4·6.010H2O) in water with respect to the hexahydrate, and combined the results
with those for 1.1 to 21.8 K from an earlier study [64STO/HAD] and with results for the
heptahydrate to obtain S mο (NiSO4·6.00H2O, 298.15 K). There do not appear to be any
other comparable measurements for this solid, and the entropy and heat capacity values
for NiSO4·6.00H2O at 298.15 K from Stout et al. are selected in the present review.
Though some of the auxiliary data used by Stout et al. differ slightly from those used in
the present review, such differences are much less than the rather small estimated experimental uncertainties (cf. Appendix A for [66STO/ARC]). The Reviewers select:
S mο (NiSO4·6.00H2O, α, 298.15 K) = (334.45 ± 1.00) J·K–1·mol–1
C pο,m (NiSO4·6.00H2O, α, 298.15 K) = (327.9 ± 1.0) J·K–1·mol–1.
Values of the enthalpy of solution of α-NiSO4·6.00H2O in water have been reported by Stout et al. [66STO/ARC] and Goldberg et al. [66GOL/RID] and, allowing
for the enthalpies of dilution, the values from the two sources are in good agreement.
192
V.5 Group 16 (chalcogen) compounds and complexes
The more extensive set of measurements by Goldberg et al. resulted in more dilute final
solutions of nickel sulphate, and are preferred here. Goldberg et al. used enthalpy of
dilution values of Lange [59LAN] and Lange and Miederer [56LAN/MIE] to extrapolate the enthalpy of solution to I = 0, and obtained ∆ sol H mο = 4.81 kJ·mol–1 (as is normal
for 2:2 electrolytes, the experimental heat of dilution for NiSO4 at low ionic strength
changes much more strongly with concentration than would be predicted from simple
Debye-Hückel theory, cf. Appendix A for [66GOL/RID]). The largest uncertainty in this
heat of solution value comes from the extrapolation to I = 0. In the present review the
selected value is:
∆ sol H mο ((V.16), NiSO4·6.00H2O, α, 298.15 K) = (4.485 ± 0.200) kJ·mol–1.
This is one of the key quantities in the derivation of thermodynamic quantities
of Ni2+ (cf. Section V.2.1.3).
Based on ∆ f H mο (NiSO4·7.00H2O, 298.15 K) = – (2976.94 ± 0.95) kJ·mol–1,
and the difference between the enthalpies of solution of α-NiSO4·6.00H2O and
NiSO4·7.00H2O in water ((7.68 ± 0.10) kJ·mol–1) reported by Stout et al. [66STO/ARC],
∆ f H mο (NiSO4·6.00H2O, α, 298.15 K) = – (2683.8 ± 1.0) kJ·mol–1.
The value of ∆ f Gmο (NiSO4·6.00H2O, α, 298.15 K) is calculated from the selected values of ∆ f H mο (NiSO4·6.00H2O, 298.15 K) and S mο (NiSO4·6.00H2O, α,
298.15 K) and the TDB auxiliary data in Table IV-1:
∆ f Gmο (NiSO4·6.00H2O, α, 298.15 K) = – (2225.5 ± 1.1).
V.5.1.2.2.1.2
Other hydrated nickel sulphate solids
Solubility data [34BEN/THI], [39ROH] indicate that the β-hexahydrate is the stable
solid in contact with saturated solutions of nickel sulphate in water for temperatures
between 325 and 358 K, where lower hydrates begin to predominate. As might be expected, addition of sulphuric acid results in the appearance of the lower hydrates at
lower temperatures [39ROH].
Based on their differential scanning calorimetry results, Rabbering et al.
[75RAB/WAN] reported ∆ r H m = (6.53 ± 0.17) kJ·mol–1 for the reaction:
α-NiSO4·6.00H2O U β-NiSO4·6.00H2O.
(V.89)
Even a difference in the heat capacities of the two solids of 3 J·K–1·mol–1
would change this value by only 0.1 kJ·mol–1 (approximately) between 298.15 K and
the temperature of the experiment. Therefore, a value of the transition enthalpy
∆ r H mο ((V.89), α → β, 298.15 K) = (6.53 ± 0.20) kJ·mol–1 is assumed, and
∆ f H mο (NiSO4·6.00H2O, β, 298.15 K) = – (2677.3 ± 1.0) kJ·mol–1
V.5 Group 16 (chalcogen) compounds and complexes
193
is retained for selection. If the transition temperature is 325 K, and the heat capacities of
the two solids are assumed to be temperature independent and equal between 298.15
and 325 K, the selected value for molar entropy is:
S mο (NiSO4·6.00H2O, β, 298.15 K) = (354.5 ± 5.0) J·K–1·mol–1
where the uncertainty has been increased to allow for the approximations in the calculation and the uncertainty in the transition temperature. These selections yield:
∆ f Gmο (NiSO4·6.00H2O, β, 298.15 K) = – (2224.9 ± 1.8) kJ·mol–1.
Rabbering et al. [75RAB/WAN] and Siracusa et al. [78SIR/SEM] carried out
thermal decomposition studies by a variety of techniques, and found evidence for decomposition of the β-hexahydrate to a tetrahydrate at temperatures near 400 K and to
the monohydrate near 440 K. Nandi et al. [79NAN/DES] found that the dehydration
pattern for thermogravimetric measurements carried out at 5 K·min–1 were different for
well-crystallised and powdered samples of the heptahydrate and hexahydrate. However,
the intermediate hydrates were not well characterised.
Several studies [62CAI/POB], [64KOH/ZAS], [75RAB/WAN], [78SIR/SEM]
have shown that the monohydrate does not readily lose water below 500 K. Goldberg et
al. [66GOL/RID] determined the heat of solution of a nickel sulphate hydrate having the
composition NiSO4·5.059H2O, and a value for the enthalpy of hydration of
NiSO4·H2O(s) to the hexahydrate was calculated by assuming that the mixed solid was a
mixture of NiSO4·6.000H2O(s) and NiSO4·H2O(s). It seems more likely that the solid
was a mixture of NiSO4·6.000H2O(s) and NiSO4·4H2O(s) [65JAM/BRO]. Kohler and
Zäske [64KOH/ZAS] reported equations for water vapour pressures as a function of
temperature for mixtures of the hexahydrate and tetrahydrate, the tetrahydrate and a
dihydrate, and the dihydrate and a monohydrate. Simon and Knauer [39SIM/KNA] also
identified the tetrahydrate as a probable intermediate product in the isobaric decomposition of the heptahydrate. The results from these studies are in rough agreement, but in
the present review, no thermodynamic values are selected for these lower hydrates.
V.5.1.2.2.1.3
NiSO4(cr)
Although Goldberg et al. [66GOL/RID] indicated that samples of NiSO4(cr) dissolved
too slowly in water to allow determination of its heat of solution, Adami and King
[65ADA/KIN] reported the heat of solution in 4.360 m HCl(aq) at 303.15 K. The thermodynamic cycle involved comparison with the heat of dissolution of metallic nickel in
a similar medium. The results of [65ADA/KIN] have been recalculated in the present
review (Appendix A) to obtain the selected value:
∆ f H mο (NiSO4, cr, 298.15 K) = – (873.28 ± 1.57) kJ·mol–1.
Weller [65WEL] measured the heat capacity of anhydrous nickel sulphate between 52.0 and 296.3 K. Stuve et al. [78STU/FER] also measured the heat capacity by
adiabatic calorimetry, but primarily at lower temperatures, between 9.7 and 69.6 K (a
V.5 Group 16 (chalcogen) compounds and complexes
194
sharp thermal transition was observed at 35.5 K), and used their results with those of
Weller [65WEL] to determine:
C pο,m (NiSO4, cr, 298.15 K) = (97.6 ± 0.1) J·K–1·mol–1
S mο (NiSO4, cr, 298.15 K) = (101.3 ± 0.2) J·K–1·mol–1.
In the absence of other values, these values of S mο (NiSO4, 298.15 K),
C (NiSO4, 298.15 K) are selected in the present review (with an estimated uncertainty
of ± 0.2 J·K−1·mol–1 for S mο ). These selections yield:
ο
p ,m
∆ f Gmο (NiSO4, cr, 298.15 K) = – (762.7 ± 1.6) kJ·mol–1.
Stuve et al. [78STU/FER] also reported the results of a limited set of drop
calorimetry experiments (402.9 to 1001.5 K). The authors fitted an equation to the experimental enthalpy differences such that the heat capacity values meshed smoothly
with the value obtained from adiabatic calorimetry for 298.15 K. The authors indicated
that their equation1 was applicable, within 0.6 per cent for temperatures between 298
and 1200 K. As discussed below, NiSO4 decomposes towards the upper end of this temperature range, and extrapolation of the heat capacities to 1200 K does not seem justified. Conversion of the equation from calorie to joule units leads to:
–2
[C pο,m ]1000K
(T/K) – 2.8681 × 106 (T/K)–2)
298.15K (NiSO4, cr) = (118.934 + 3.6761 × 10
J·K–1·mol–1
and this equation for C pο,m (T) from Stuve et al. [78STU/FER] is accepted in the present
review.
There have been several studies of the equilibrium conversion of NiSO4 to NiO
from 880 K to 1200 K according to:
NiSO4 U NiO + SO3
with possible concurrent decomposition of SO3:
SO3 U SO2 + ½O2.
Both pressure measurements [25MAR], [78KAR/MAL] and electrochemical
measurements [66ING], [64ALC/SUD], [73SKE/ESP], [77SAD/KAW], [83GAU/BAL]
have been reported. The measurements lead to values of ∆ f Gmο that agree within 5 – 6
kJ·mol–1 near 900 K, and 7 – 8 kJ·mol–1 at 1150 K. The values from Alcock et al.
[64ALC/SUD] show better agreement with the vapour-pressure study of Karwan et al.
[78KAR/MAL] than do the other electrochemical studies. In the present review, no attempt has been made to calculate heat-capacity values from the second derivatives of
equations fitted to these measurements.
1
The rather similar equation for H m (T)
–
H mο (298 K), probably based on prepublication results from
[78STU/FER], is incorrect as published in the review of Mah and Pankratz [76MAH/PAN].
V.5 Group 16 (chalcogen) compounds and complexes
195
V.5.2 Selenium compounds and complexes
Nickel-selenium compounds have recently been reviewed by the NEA TDB review
team on Se, (see [2005OLI/NOL] for the corresponding NEA TDB critical review).
V.5.3 Tellurium compounds and complexes
V.5.3.1 Nickel tellurides
The binary phase diagram nickel/tellurium as summarised by Lee and Nash
[90LEE/NAS] shows a complex behaviour with six intermetallic solid phases. This diagram is essentially based on a seminal paper authored by Klepp and Komarek
[72KLE/KOM]. The phases β1, β2, β1′ and NiTe2–x(cr) have variable stoichiometry.
In an early stage of this review it was decided that non-stoichiometric alloy
systems will not be covered. Thus only the two γ-phase line components, NiTe0.775(γ1)
and NiTe0.85(γ2) among the six solid phases, need to be discussed here. The thermodynamic data of the Ni – Te system have been critically assessed by Ball et al.
[92BAL/DIC].
V.5.3.1.1
NiTe0.775(γ1)
NiTe0.775(cr) is orthorhombic and has the following unit cell parameters: a = 3.916 Å, b
= 6.860 Å, c = 12.32 Å [72KLE/KOM]. The γ1-phase appears to be stable up to
1015.5 K, when it decomposes peritectoidally into γ2 and β2. There have not been any
heat capacity measurements reported for γ1. Ball et al. [92BAL/DIC] obtained estimates
of C pο,m (NiTe0.775, cr, 298.15 K) = 45.8 (48.7) J·K–1·mol–1 and S mο (NiTe0.775, cr, 298.15
K) = 71.0 (67.6) J·K–1·mol–1 by taking a weighted average of heat capacities of the β
and NiTe2–x(cr) phases. The values in brackets have been calculated with the cationic
and anionic contributions to the heat capacity at 298 K and the revised ‘Latimer’ entropy contributions [93KUB/ALC]. If reasonable uncertainties of ± 5.0 J·K−1·mol–1 are
chosen for ∆C pο,m and ∆S mο , both methods of estimation lead to results that are consistent within these limits.
The enthalpy of formation of NiTe0.775(cr) has been estimated from emf measurements in the γ + NiTe2–x(cr) two-phase field [92BAL/DIC]. Neither of these standard
entropy or standard enthalpy estimations ( ∆ f H mο (NiTe0.775, cr, 298.15 K) =
− (41.7 ± 4.4) kJ·mol–1) is based directly on measurements of the NiTe0.775(cr) phase, so
no recommendation can be given in this review. This is not of any particular importance
here as the γ1 phase is of peripheral interest only.
V.5.3.1.2
NiTe0.85(γ2)
Based on thermoanalytical studies Klepp and Komarek [72KLE/KOM] assigned somewhat arbitrarily a composition of NiTe0.85(cr) to the high temperature γ2 phase. An attempt to characterise this phase by its X-ray powder diffraction pattern failed, because
below 963 K it decomposes quickly into NiTe0.775(cr) + NiTe2–x(cr) and the phase equi-
196
V.5 Group 16 (chalcogen) compounds and complexes
librium could not be frozen in by quenching [72KLE/KOM]. Therefore, estimated
thermodynamic quantities for the γ2 phase are even less reliable than for the γ1 phase.
According to Ball et al. [92BAL/DIC], the same values of the standard heat capacity
and entropy were estimated for both γ phases (based on x(Ni) + x (Te) = 1), resulting in
C pο,m (NiTe0.85, cr, 298.15 K) = 47.7 (50.7) J·K–1·mol–1 and S mο (NiTe0.85, cr, 298.15 K) =
74.0 (70.7) J·K–1·mol–1. The values in brackets have been calculated with the cationic
and anionic contributions to the heat capacity at 298 K and the revised ‘Latimer’ entropy contributions [93KUB/ALC]. The latter estimation method clearly leads also to
identical heat capacity and entropy values for both γ phases based on x(Ni) + x (Te) = 1.
The enthalpy of formation of NiTe0.85(cr) has again been estimated from emf
measurements in the γ + NiTe2–x(cr) two-phase field [92BAL/DIC]. Without confirmation by experimental data the estimated standard entropy and standard enthalpy
( ∆ f H mο (NiTe0.85, cr, 298.15 K) = – (50.1 ± 5.6) kJ·mol–1) of this phase cannot be recommended in this review.
V.6 Group 15 compounds and complexes
V.6.1 Nitrogen compounds and complexes
V.6.1.1 Solid nickel nitrates
The hydrated nickel nitrate solids have been the subject of sporadic thermodynamic
studies over the last 150 years, but the basic thermodynamic quantities for these materials are not well defined. Though careful synthesis yields solids with various
Ni(NO3)2:H2O ratios, the regions of stability (if any) for the lower hydrates are still not
well established [1899FUN], [34SIE/SCH], [98ELM/GAB], [2001MIK/MIG]. When
hydrated nickel nitrate is heated at a pressure of approximately 0.1 MPa, the last water
of hydration is lost between 450 and 500 K. Release of nitrogen oxides occurs at temperatures below 600 K, and continued heating leads to formation of basic nickel nitrates
and then NiO [98ELM/GAB], [2001MIK/MIG].
V.6.1.1.1
Ni(NO3)2·9H2O(cr)
This solid has been reported as a stable solid in the Ni(NO3)2 – H2O system for temperatures below 270 K [34SIE/SCH]. The temperature for the equilibrium conversion of the
nonahydrate to the hexahydrate in contact with saturated aqueous solution has not been
well established. No values for Ni(NO3)2·9H2O(cr) are selected in the present review.
V.6.1.1.2
Ni(NO3)2·6H2O(cr)
The stable hydrate in equilibrium with a solution saturated in nickel nitrate at 298.15 K
is Ni(NO3)2·6H2O(cr) [34SIE/SCH]. The water content of the commercially available
solid tends to vary slightly. The solid can easily lose water on exposure to dry air
[72AUF/CAR], [98ELM/GAB], but has also been reported to be slightly deliquescent in
V.6 Group 15 compounds and complexes
197
moist air [72AUF/CAR]. In a closed system, the hydrate begins to melt (or partially
dissolve in its water of hydration) at 328 K [2001MIK/MIG]. Auffredic, Carel and
Weigel [72AUF/CAR] measured the enthalpy of solution of the hexahydrate in water at
298.15 K. Calorimetric measurements were done for ratios of H2O:Ni(NO3)2·6H2O between 500 (0.11 m) and 10000 (0.0056 m), and a value of (31.23 ± 1.00) kJ·mol–1 for
∆ sol H mο was reported. For the more dilute solutions, there is considerable scatter in the
measurements as a function of molality, and the reported value of ∆ sol H mο was simply
an extrapolated value from almost random values for concentrations between 0.01 m
and 0.0056 m. As discussed below, and in the Appendix A discussion for
[72AUF/CAR], the lack of concentration dependence over this concentration range is
not unexpected. The average value of the enthalpies of solution for final concentrations
below 0.01 kg·mol–1 is (31.28 ± 0.87) kJ·mol–1. A reasonable estimate of the integral
heat of dilution of a nickel nitrate solution from a solution with a solvent:salt ratio of
8000:1 is – (0.52 ± 0.10) kJ·mol–1 [59LAN] (see the Appendix A discussion of
[72AUF/CAR]). In the present review, this estimated heat of dilution is applied to the
average value of the enthalpies of solution for final concentrations below 0.01 kJ·mol–1,
and for
Ni(NO3)2·6.00H2O(cr) U Ni2+ + 2 NO3− + 6H2O(l)
(V.90)
∆ r H mο ((V.90), 298.15 K) = (30.76 ± 1.00) kJ·mol–1
is selected. Values based on measurements of the vapour pressure of water over the
hexahydrate [72AUF/CAR] were not incorporated in the calculation of the assessed
value because, as discussed below, the stoichiometry of the dehydrated salt in equilibrium with the hexahydrate and H2O(g) is not firmly established. This selection yields,
using NEA-TDB auxiliary data:
∆ f H mο (Ni(NO3)2·6.00H2O, cr, 298.15 K) = – (2214.5 ± 1.6) kJ·mol–1.
Goldberg et al. [79GOL/NUT], reviewed the available literature data, and provided tables of recommended values for the activity coefficients and osmotic coefficients for solutions of Ni(NO3)2 in water at 298.15 K. However, there were no studies
that allowed the tables to be extended to saturated solutions of Ni(NO3)2·6.00H2O(cr) at
298.15 K (5.47 m [34SIE/SCH]), and no major studies seem to have been published
since that review. No values of ∆ sol Gm at saturation are estimated here.
Heat capacity measurements for Ni(NO3)2·6H2O(cr) were reported by Vasileff
and Grayson-Smith [50VAS/GRA] for temperatures from 65 to 300 K. The authors estimated the entropy change between 65 and 273 K to be 364.4 J·K–1·mol–1 (87.1
cal·K−1·mol–1). From the set of smoothed specific heats, the selected value of the molar
heat capacity of the solid at 298.15 K is calculated to be
C pο,m (Ni(NO3)2·6H2O, cr, 298.15 K) = (463.0 ± 2.0) J·K–1·mol–1,
198
V.6 Group 15 compounds and complexes
where the uncertainty has been estimated in the present review. The temperature range
of the measurements is not adequate for the calculation of a value for
S mο (Ni(NO3)2·6H2O, cr, 298.15 K).
V.6.1.1.3
Ni(NO3)2·4H2O(cr), Ni(NO3)2·3H2O(cr), Ni(NO3)2·2H2O(cr)
Dehydration of the hexahydrate leads to several lower hydrates, but the hydrate formed
seems to depend markedly on the method used to carry out the dehydration. In few of
the experiments were the dehydrated solids thoroughly characterised, nor was it established that the solids were stable over long periods.
Chukurov et al. [73CHU/DRA] and Auffredic, Carel and Weigel
[72AUF/CAR] have reported measurements of the heat of solution of a compound with
the stoichiometry Ni(NO3)2·4H2O(cr) in water (8.49 kJ·mol–1 and 8.66 kJ·mol–1, respectively). The latter group prepared their solid by dehydration of the hexahydrate over
calcium chloride at room temperature. They also reported [72AUF/CAR2] a value of
the heat of solution (– 27.41 kJ·mol–1) of Ni(NO3)2·2H2O(cr) (prepared by dehydration
of the tetrahydrate over calcium chloride at 50°C). Recalculation of the heats of dilution
(cf. Appendix A for [72AUF/CAR]) gives heats of solution, corrected to “infinite dilution”, for the dihydrate of – (27.64 ± 1.00) kJ·mol–1 and for the tetrahydrate
(8.23 ± 1.00) kJ·mol–1 (based on [72AUF/CAR]) and (8.49 ± 1.00) kJ·mol–1 (based on
[73CHU/DRA]).
In contrast to Funk [1899FUN] who reported a stable trihydrate phase (55 to
95°C), Sieverts and Schreiner [34SIE/SCH] found that both the tetrahydrate (55 to
80°C) and dihydrate (85 to 120°C) were stable phases in the Ni(NO3)2 – H2O system,
and could also be formed at equilibrium at 25°C in the Ni(NO3)2 – H2O – HNO3 system
at high nitric acid mole fractions. Studies of the dehydration of the hexahydrate using
differential thermal analysis [74WEN], [98ELM/GAB], differential scanning calorimetry [99MIK/MIG] and thermogravimetric analysis [66KAL/PUR], [2001MIK/MIG]
suggest that although a tetrahydrate can be obtained, it is easily converted to the trihydrate at slightly higher temperatures. For example, Kalinichenko and Purtov
[66KAL/PUR] reported that dehydration of the tetrahydrate to the trihydrate occurred
between 164°C and 170°C.
Sano [37SAN] and Auffredic, Carel and Weigel [72AUF/CAR], reported
measurements of the vapour pressure of water over Ni(NO3)2·6H2O(cr) as a function of
temperature (Figure V-39). The measured vapour pressures differ systematically by a
factor of 1.5 at similar temperatures. Sano interpreted his results in terms of the reaction:
Ni(NO3)2·6.00H2O(cr) U Ni(NO3)2·3.00H2O(cr) + 3H2O(g)
(V.91)
whereas Auffredic, Carel and Weigel interpreted their results in terms of the reaction:
Ni(NO3)2·6.00H2O(cr) U Ni(NO3)2·4.00H2O(cr) + 2H2O(g).
(V.92)
V.6 Group 15 compounds and complexes
199
As can be seen from the figure, the slope of the plot of – log10 pH2 O against 1/T
is similar for both sets of results. This may indicate that whichever lower hydrate was
involved in the equilibrium, it was the same solid for both sets of experiments. If the
vaporisation corresponds to Reaction (V.92), the calculated enthalpy of vaporisation
(110 kJ·mol–1) corresponds well with 116 kJ·mol–1, derived from the enthalpies of solution of the hexahydrate [72AUF/CAR] and the tetrahydrate [72AUF/CAR],
[73CHU/DRA]. This suggests that the vaporisation follows Reaction (V.92), rather than
(V.91).
Similarly, the enthalpy for Reaction (V.93),
Ni(NO3)2·4.00H2O(cr) U Ni(NO3)2·2.00H2O(cr) + 2H2O(g)
(V.93)
∆ r H mο = 124 kJ·mol–1, as derived from the reported calorimetric data [72AUF/CAR],
[72AUF/CAR2], is similar to the value obtained from measurements of the vapour pressure of water over the tetrahydrate (122 kJ·mol–1) [72AUF/CAR2]. Nevertheless no
value is selected in the present review for the Gibbs energy of Reaction (V.92) or (V.93)
because there was no proper characterisation of the lower hydrate at equilibrium during
any of these measurements.
Figure V-39: Measured water vapour pressures over hydrated Ni(NO3)2·6H2O(cr)
[37SAN], [72AUF/CAR] and Ni(NO3)2·4H2O(cr) [72AUF/CAR2].
-1.4
-1.6
log10(p(H2O)/bar)
-1.8
-2.0
-2.2
-2.4
[37SAN]
[72AUF/CAR]
[72AUF/CAR2]
-2.6
-2.8
-3.0
2.9
3.0
3.1
3.2
3
10 ·K/T
3.3
3.4
3.5
200
V.6 Group 15 compounds and complexes
In the present review, the average calorimetry-based value is selected for the
enthalpy of dissolution of the tetrahydrate:
Ni(NO3)2·4.00H2O(cr) U Ni2+ + 2 NO3− + 4H2O(l)
(V.94)
∆ r H mο ((V.94), 298.15 K) = (8.36 ± 0.80) kJ·mol–1.
The selected uncertainty is slightly larger than the statistical uncertainty because the uncertainty in the heat of dilution is common to both results. The corresponding value for the dihydrate [72AUF/CAR2] has also been selected:
Ni(NO3)2·2.00H2O(cr) U Ni2+ +2 NO3− + 2H2O(l)
(V.95)
∆ r H mο ((V.95), 298.15 K) = – (27.64 ± 1.00) kJ·mol–1.
In neither case should the selection of a value for the enthalpy of solution be
taken to mean that it is certain that the solid is a stable phase in the Ni(NO3)2-H2O system. These selections yield, using NEA-TDB auxiliary data:
∆ f H mο (Ni(NO3)2·4.00H2O, cr, 298.15 K) = – (1620.4 ± 1.4) kJ·mol–1
and
∆ f H mο (Ni(NO3)2·2.00H2O, cr, 298.15 K) = – (1012.7 ± 1.6) kJ·mol–1.
V.6.1.1.4
Ni(NO3)2(anhydrous)
Except for salts containing singly charged cations, synthesis of anhydrous nitrate salts is
very difficult, as the final steps of dehydration tend to cause loss of nitrogen oxides or
nitric acid. Although some authors [82MU/PER], [2001MIK/MIG] claim anhydrous
nickel nitrate forms as an intermediate decomposition product during heating of the
hydrate, only Chukurov et al. [73CHU/DRA] claim to have carried out any chemical
thermodynamic measurements ( ∆ f H mο = – (402.1 ± 2.9) kJ·mol–1, from determination of
the heat of solution in dimethyl sulphoxide). The summary of their paper, all that was
available to the reviewers, contains no details on the preparation, handling or analysis of
the anhydrous salt. No values for the thermodynamic quantities for anhydrous Ni(NO3)2
are selected in the present review.
V.6.1.2 Aqueous Ni(II)-nitrato complexes
There are conflicting opinions concerning the interaction of Ni(II) with nitrate ion. Mironov et al. have failed to detect nitrato complexes of Ni(II) [70MIR/MAK]. On the
other hand, neutron diffraction of 2 and 4.3 m Ni(NO3)2 solutions revealed ion pair formation between Ni2+ and nitrate ions [97HOW/NEI]. Quantitative information on nitrate
complexation (ion pair formation) of Ni(II) has been published in [73FED/SHM] and
[73HUT/HIG]. The Russian authors reported formation of mono-, bis- and triscomplexes [73FED/SHM], while only NiNO3+ was reported in [73HUT/HIG]. In light
of the results collected for similar weak complexes of Ni(II) (e.g., for halogeno complexes), the formation of higher complexes ( Ni(NO3 ) 2n − n , n > 1) is unlikely, except in
V.6 Group 15 compounds and complexes
201
highly concentrated nitrate solutions. Therefore, the formation of such complexes is not
considered in this review. The available association constant values for Reaction (V.96):
Ni2+ + NO3− U NiNO3+
(V.96)
are reported in Table V-27.
The experimental data reported in [73FED/SHM] were re-evaluated taking into
account only formation of the NiNO3+ species (see Appendix A). To minimise the medium effect, only half of the experimental data, for which [ NO3− ] ≤ [ ClO −4 ], were taken
into account. The recalculated constants are also listed in Table V-27.
Table V-27: Formation constants (logarithmic values) of the NiNO3+ complex (ion
pair).
t (°C)
Method Medium
log10 b1
log10 b1
(a)
(0.04 ± 0.08)
(b)
(0.02 ± 0.40)
log10 b1ο (c)
(0.97 ± 0.60) [73HUT/HIG]
kin
1 M Na(ClO4, NO3)
25
45
(0.03 ± 0.12)
sol
4 M Li(ClO4, NO3)
25
– (0.30 ± 0.04)
– (0.01 ± 0.40)(d)
3 M Li(ClO4, NO3)
– (0.55 ± 0.09)
– (0.30 ± 0.40)(d)
2 M Li(ClO4, NO3)
– (0.44 ± 0.06)
– (0.21 ± 0.40)(d) (0.49 ± 0.45)
1 M Li(ClO4, NO3)
– (0.22 ± 0.03)
– (0.20 ± 0.40)(d)
0 corr
Reference
(0.01 ± 0.40)
[73FED/SHM]
(0.45 ± 0.08)
(e)
(a) Reported values.
(b) Accepted values corrected to molal scale.
(c) Corrected to I = 0 by means of the SIT.
(d) Recalculated value, taking into account the solely formation of NiNO3+ species (see Appendix A).
(e) Extrapolated value, using the Vasil’ev equation.
Extrapolation of log10 b1 values to Im = 0 in lithium perchlorate media are
shown in Figure V-40. The weighted linear regression resulted in the values of
log10 b1ο (V.96) = (0.49 ± 0.45) and ∆ε((V.96), LiClO4) = – (0.08 ± 0.14) kg·mol–1.
From the latter value ε( NiNO3+ , ClO −4 ) = (0.44 ± 0.14) kg·mol–1 can be derived. Calculation of this value also requires the value of ε(Ni2+, NO3− ) and, in the absence of data to
allow calculation of ε(Ni2+, NO3− ) while also considering association, the value derived
in Section V.6.1.2.1 is used. Based on this result, ∆ε((V.96), NaClO4) = (0.04 ± 0.15)
kg·mol–1 can be estimated, and this is used to determine log10 b1ο from the value of
log10 b1 reported in [73HUT/HIG]. Both values of log10 b1ο in Table V-27 are derived
from constants at higher ionic strength that have large uncertainties because of medium
effects. Therefore, in this review the lower value, calculated from [73FED/SHM], is
selected with an increased uncertainty:
log10 b1ο ((V.96), 298.15 K) = (0.5 ± 1.0).
V.6 Group 15 compounds and complexes
202
Only a single value is reported for the enthalpy of Reaction (V.96) [75ARU].
Recalculating the experimental data using the tentative values of log10 b1ο ((V.96),
298.15 K) and ∆ε((V.96), Ni(ClO4)2) = – (0.11 ± 0.15) kg·mol–1, ∆ r H m ((V.96),
298.15 K) = (6.0 ± 4.0) kJ·mol–1 can be estimated for I = 1 M (Ni(ClO4)2), from the data
reported in [75ARU] (see Appendix A). The selected value yields, using NEA-TDB
auxiliary data:
∆ f Gmο ( NiNO3+ , 298.15 K) = – (159.4 ± 5.8) kJ·mol–1.
Figure V-40: Extrapolation to I = 0 of the constants for Reaction (V.96) in lithium perchlorate media based on the data in [73FED/SHM]. The recalculated and original values
are marked with full and open squares, respectively (see Appendix A).
2
Ni2+ + NO3− U NiNO3+
log10 β1 + 4D
1.5
1
0.5
0
-0.5
0
1
2
I / mol·kg
3
−1
4
5
V.6 Group 15 compounds and complexes
V.6.1.2.1
203
Determination of the Ni2+ – NO −3 ion interaction coefficient
Due to the low and rather uncertain value of log10 b1ο ((V.96), 298.15 K), the description of the Ni(II)-nitrate system in terms of ion-ion interactions may be in many cases
more straightforward. Therefore, the ion interaction coefficient ε(Ni2+, NO3− ) is also
derived in this review from the osmotic and mean activity coefficients of Ni(NO3)2 solutions, using Equations (V.74) and (V.75).
The most reliable values for the mean activity and osmotic coefficient of aqueous Ni(NO3)2 solutions at 298.15 K are given in [82SAD/LIB] and [82SAR/COV]. Plots
of log10 γ ± and φ as a function of the molality, m, of Ni(NO3)2 (m = Im/3) are depicted
in Figure V-41 and Figure V-42, respectively. The two data sets of activity coefficients
values are somewhat different, but the disagreement is acceptable, taking into account
the similarity of osmotic coefficients. A fit of Equations (V.74) and (V.75) to the experimental data up to an ionic strength of 15 mol·kg–1 yields ε(Ni2+, NO3− ) =
(0.182 ± 0.010) kg·mol–1, see solid lines in Figure V-41 and Figure V-42.
Figure V-41: Plot of log10 γ ± versus molality of aqueous Ni(NO3)2 solutions at
298.15 K. Experimental data from: [82SAD/LIB] ( ); [82SAR/COV] ( ). Solid line:
SIT, ε(Ni2+, NO3− ) = 0.182.
0.8
0.6
log10 γ±
0.4
0.2
0
-0.2
-0.4
-0.2
0.4
1.0
1.6
2.2
2.8
3.4
[Ni(NO3)2] / mol·kg−1
4.0
4.6
5.2
V.6 Group 15 compounds and complexes
204
Figure V-42: Osmotic coefficient plotted as a function of molality of aqueous Ni(NO3)2
solutions at 298.15 K. Experimental data from: [82SAD/LIB] ( ); [82SAR/COV] ( ).
Solid line: SIT, ε(Ni2+, NO3− ) = 0.182.
2.6
2.4
2.2
2
φ
1.8
1.6
1.4
1.2
1
0.8
-0.2
0.4
1
1.6
2.2
2.8
3.4
[Ni(NO3)2] / mol·kg
4
4.6
5.2
−1
V.6.2 Phosphorus compounds and complexes
V.6.2.1 Solid nickel phosphorus compounds
V.6.2.1.1
Nickel phosphides
The thermodynamic measurements for nickel phosphides were recently summarised by
Schlesinger [2002SCH] (in that reference, red phosphorus (V) was used as the standard
state for phosphorus). Schlesinger concluded that the details of the Ni-P phase diagram
remain controversial, and that the available data are insufficient to allow calculation of
assessed thermodynamic values for any nickel phosphides. No values are selected in the
present review.
V.6.2.1.2
Nickel phosphates
A number of nickel phosphate solids have been reported, such as Ni3(PO4)2·8H2O
Ni3(PO4)2·7H2O
[66GME],
Ni3(PO4)2·1.25H2O
[86TOD/HAS],
[66GME],
NiHPO4·3H2O [87CUD/LEC], (NiHPO4)2·3H2O [87CUD/LEC], Ni(H2PO4)2·2H2O
V.6 Group 15 compounds and complexes
205
[90TRO] and Ni3(PO4)2 [94HAS/TOD]. Nevertheless, chemical thermodynamic data for
these solids are almost non-existent.
The most thoroughly studied solid is Ni3(PO4)2·8H2O, which is reported
[96DOJ/NOV] to undergo three phase transitions below 300 K (at 174, 247 and 289 K).
Though solubility measurements have been reported for Ni3(PO4)2·8H2O
[61CHU/ALY2], [82CHI/SAB], Ni3(PO4)2·7H2O [76PAN/PAT] and NiHPO4
[76PAN/PAT], none are of adequate quality to allow chemical thermodynamic quantities to be calculated for these solids (cf. Appendix A).
V.6.2.2 Aqueous nickel phosphorus species
V.6.2.2.1
Simple nickel phosphato complexes
The literature on complex formation between phosphate and Ni(II) ions in solution is
not extensive (see Table V-28). Most measurements have been based on pH titrations
[67SIG/BEC], [72FRE/STU], [76TAY/DIE], [96SAH/SAH], though there have also
been polarographic [91SWI/PAW] and solubility [89ZIE/JON] studies. Caminiti
[82CAM] showed that in acidic highly concentrated solutions oxygen atoms from phosphate groups could replace water molecules in the inner sphere of hydrated Ni2+ ions.
However, under most conditions the complexes are weak, and difficult to identify unambiguously because of protonation equilibria involving both the ligand and the complexes. Based on a curve-fitting analysis of potentiometric titration curves, Childs
[70CHI] suggested a wide range of species: MH 2 PO +4 , MHPO4(aq), MH 3 (PO 4 ) 2− ,
M2H2(PO4)2(aq) and MPO −4 . Few of these have been reported in the Ni(II) – phosphate
system.
Most of the studies [67SIG/BEC], [72FRE/STU], [76TAY/DIE],
[91SWI/PAW], [96SAH/SAH] were inspired by possible parallels between phosphate
complexation and biochemical interactions between phosphate esters and metal ions.
These studies have been carried out over a fairly limited pH range (usually between 4
and 6) at low ionic strength ( ≤ 0.2 M). In none of these publications was the primary
data published. Most papers did report experimental values for one or more protonation
constants of the phosphate ion itself. This provides some basis for qualitative evaluation
of the experiments. Except in the work of Taylor and Diebler [76TAY/DIE], the authors
interpreted their results in terms of a single complex, NiHPO4(aq),
Ni2+ + HPO 24 − U NiHPO4(aq),
(V.97)
and except for the polarographic result of Swiatek and Pawlowski [91SWI/PAW], the
reported values are consistent within the experimental uncertainties. For reasons discussed in Appendix A, the result from the polarographic study is not used here. The
other values have been corrected to I = 0, and the weighted average of the three results
at 25°C provides the selected values:
log10 K ο ((V.97), 298.15 K) = (3.05 ± 0.09)
V.6 Group 15 compounds and complexes
206
and
∆ r Gmο ((V.97), 298.15 K) = – (17.4 ± 0.5) kJ·mol–1.
The value log10 K ο ((V.97), 298.15 K) = (2.88 ± 0.30) obtained from the work
of Frey and Stuehr [72FRE/STU] is compatible with this selection, which yields, using
NEA-TDB auxiliary data:
∆ f Gmο (NiHPO4, aq, 298.15 K) = – (1159.2 ± 1.8) kJ·mol–1.
Ziemniak, Jones and Combs [89ZIE/JON] reported solubilities of Ni(II) oxide
in aqueous sodium phosphate solutions (pH298 > 10) for temperatures from 290 to
560 K. As discussed in Appendix A and elsewhere [95LEM], it is difficult to interpret
these solubility results quantitatively because of the limited range in hydroxide concentration and the wide variation in the ionic strength. The values for the formation constants of NiHPO4(aq) from the pH studies [67SIG/BEC], [96SAH/SAH] and for the
species, Ni(OH) 2 HPO 24 − , as proposed by Ziemniak, Jones and Combs [89ZIE/JON],
[95LEM] are not inconsistent for the pH values at which the experimental measurements were obtained.
Table V-28: Experimental values for the association constants and other thermodynamic
quantities for nickel phosphate complexes.
Method
t (°C)
Medium (aq)
Reported value
Recalculated value (a)
log10 K ο
Reference
ligand: HPO 24−
Ni2+ + HPO 24− U NiHPO4(aq)
pot
25.0
0.1 M NaClO4
log10 K1 = 2.08
(2.97 ± 0.30)
[67SIG/BEC]
pot
15.0
0.1 M KNO3
log10 K1 = 2.00
(2.88 ± 0.30)
[72FRE/STU]
pot
25.0
0.1 M NaClO4
log10 K1 = 2.11
(3.00 ± 0.20)
[76TAY/DIE]
pol
25.0
0.2 M NaClO4
log10 K1 = 3.26
pot
25.0
0.1 M NaNO3
log10 K1 = 2.20
[91SWI/PAW]
(3.07 ± 0.10)
[96SAH/SAH]
β-Ni(OH)2(cr) + HPO 24− U Ni(OH) 2 HPO 24−
sol
25.0
ligand: H 2 PO
0.1 M NaClO4
log10 K = – (5.0 ± 1.0)
[89ZIE/JON]
−
4
Ni2+ + H 2 PO 4− U H 2 NiPO +4
log10 K1 = 0.5
pot
25.0
0.1 M Li/NaClO4
(a) Uncertainties are the statistical uncertainties from the re-analysis of the data.
The overall estimated uncertainties are discussed in the text.
[76TAY/DIE]
V.6 Group 15 compounds and complexes
V.6.2.2.2
207
Aqueous nickel diphosphato complexes
As in the NEA-TDB uranium review [92GRE/FUG], the only polyphosphate(V) species
considered in the present review are the pyrophosphates (diphosphato complexes).
Other polyphosphoric acid species have negligible equilibrium concentrations at total
phosphate concentrations < 0.045 mol·dm–3 and at temperatures below 200°C
[74MES/BAE]. There have been potentiometric (pH measurement) [56YAT/VAS],
[64HAM/MOR], [73PER/SEC], calorimetric [56YAT/VAS2], spectrophotometric
[58VAI/RAM] and amperometric [78KAR/HUB] studies of the formation of pyrophosphate complexes of Ni(II) (Table V-29).
The complexes of the highly charged pyrophosphate ion with nickel are generally stronger than the phosphate complexes, but interpretation of the experiments is beset by the same difficulties as the interpretation of the phosphate studies with respect to
unambiguous identification of the species. In the literature, results of studies of association of Ni2+ with pyrophosphate have been discussed in terms of the following equilibria:
Ni2+ + P2 O74 − U NiP2 O72 −
(V.98)
Ni2+ + HP2 O37− U NiHP2 O7−
2+
Ni + 2 P2 O
4−
7
6−
7 2
U Ni(P2 O )
(V.99)
(V.100)
The results of Vaid and Rama Char [58VAI/RAM] were obtained under illdefined conditions, and are in poor agreement with results from the other studies. The
equilibrium constant of Yatsimirskii and Vasilev for Reaction (V.98) is consistent with
other values, but was obtained at moderately high and variable ionic strength. Values
from the measurements of Hammes and Morrell [64HAM/MOR] in 0.1 M Me4NCl(aq)
for Reactions (V.98) and (V.99), after correction to I = 0, are log10 K ο (V.98) =
(8.73 ± 0.25) and log10 K ο (V.99) = (5.14 ± 0.25), at 298.15 K.
Perlmutter-Hayman and Secco [73PER/SEC] did careful measurements of
nickel pyrophosphate complexation at four different ionic strengths (KNO3(aq)), though
for reasons discussed in Appendix A, only their values obtained in 0.1 M and 0.2 M
KNO3(aq) are considered to be consistent. As discussed in Appendix A, the authors
neglected association of pyrophosphate with K+, though the values they used for pyrophosphate protonation were determined in the corresponding aqueous KNO3 solutions.
The apparent 298.15 K values (corrected to I = 0) of log10 K (V.98)
((8.73 ± 0.25) [64HAM/MOR], (7.69 ± 0.25) and (7.78 ± 0.30) [73PER/SEC]) are in
poor agreement; the value of Frey and Stuehr [72FRE/STU] for 288.15 K (7.94 ± 0.25)
is in only marginal agreement with that of Perlmutter-Hayman and Secco (7.53 ± 0.25),
and the value at 303.15 K of Karwelk and Huber [78KAR/HUB] is lower than would be
expected from the results of Perlmutter-Hayman and Secco. The disagreement is primarily the result of neglect of complexation of the singly-charged cations, K+ and
NH +4 , with pyrophosphate. As indicated in Appendix A, use of reasonable potassium
V.6 Group 15 compounds and complexes
208
pyrophosphate complexation constant values brings the results of Perlmutter-Hayman
and Secco into agreement with those of Hammes and Morrell. The raw data of Perlmutter-Hayman and Secco [73PER/SEC] are unavailable for re-analysis, and in this review,
the selected value for log10 K ο (V.98) is the value based on the work of Hammes and
Morrell [64HAM/MOR] at 298.15 K:
log10 K ο ((V.98), 298.15 K) = (8.73 ± 0.25).
Hence,
∆ r Gmο ((V.98), 298.15 K) = – (49.83 ± 1.43) kJ·mol–1.
The apparent formation constants from Perlmutter-Hayman and Secco
[73PER/SEC] and Frey and Stuehr [72FRE/STU] for Reaction (V.99) at 288.15 K agree
well within the uncertainty bounds. After extrapolation to I = 0, the formation constants
from Perlmutter-Hayman and Secco ((5.02 ± 0.25) and (5.02 ± 0.30), based on their
measurements in 0.1 and 0.2 M KNO3(aq), respectively) and from Hammes and Morrell
(5.14 ± 0.25) also seem to be in good agreement. Unfortunately, the neglect of K+ association with pyrophosphate [73PER/SEC] (see Appendix A) and the unavailability of
the raw data of Perlmutter-Hayman and Secco [73PER/SEC] preclude a proper reanalysis of the data from that study. The value of log10 K (V.99) of Hammes and
Morrell [64HAM/MOR] is selected in the present review for 298.15 K:
log10 K ο ((V.99), 298.15 K) = (5.14 ± 0.25).
Hence,
∆ r Gmο ((V.99), 298.15 K) = – (29.34 ± 1.43) kJ·mol–1.
There appears to be only one reported value [56YAT/VAS] for the cumulative
second association constant (Reaction (V.100)), b 2 = (1.54 ± 0.50) × 107. As discussed
in Appendix A, this value was calculated from results obtained for solutions of ionic
strength that ranged from 0.2 to 1.0 M. This value suggests that significant quantities of
Ni(P2 O7 )62− should form only in solutions of moderate to high ionic strength where
formation of NiP2 O72 − is less strongly favoured. Furthermore, the effect of association
of Na+ with pyrophosphate [94STE/FOT] on the reported value has not been established. No value for the second association constant is selected in the present review.
The values for the enthalpies of Reactions (V.98) and (V.99) from the formation constant measurements for 0.1 M KNO3 solutions [73PER/SEC] ( ∆ r H mο ((V.98),
(278 – 308) K) = (30.6 ± 7.4) kJ·mol–1, ( ∆ r H mο ((V.99), (278 – 308) K) = (47.9 ± 10.0)
kJ·mol–1) are accepted to be the same as the values at I = 0 without correction for either
ionic strength or association of K+ with pyrophosphate. The uncertainties are estimates,
∆ r H mο ((V.98), 298.15 K) = (30.6 ± 10.0) kJ·mol–1
∆ r H mο ((V.99), 298.15 K) = (47.9 ± 15.0) kJ·mol–1.
V.6 Group 15 compounds and complexes
209
The value for Reaction (V.98) is not inconsistent with the rough calorimetric
value reported by Yatsimirskii and Vasilev (see discussion of [56YAT/VAS] in Appendix A). No value is selected for the enthalpy of Reaction (V.100).
It must be emphasised that the values selected in this review for formation of
NiHP2 O and NiP2 O72 − should not be used for solutions more than 0.01 M in alkali
metal ions unless explicit values are introduced for the pyrophosphate-alkali metal ion
association constants (the values of De Stefano et al. [94STE/FOT] may be useful, cf.
Appendix A for that reference).
−
7
The corresponding Gibbs energy of formation are:
∆ f Gmο ( NiP2 O72 − , 298.15 K) = – (2031.1 ± 4.8) kJ·mol–1
∆ f Gmο ( NiHP2 O7− , 298.15 K) = – (2064.3 ± 4.8) kJ·mol–1
Table V-29: Experimental values for the association constants and other thermodynamic
quantities for nickel pyrophosphate complexes.
Method
t /°C
Medium (aq)
Reported value
Recalculated value (a, b)
log10 K ο
Reference
ligand: P2 O74− (pyrophosphate)
Ni2+ + n P2 O74− U Ni(P2 O7 ) 2n−4 n
sol
25.0
var
log10 K1 = 5.82
[56YAT/VAS]
log10 b 2 = 7.19
log10 K s ,0 (Ni2P2O7) = – 12.77
cal
25.0
var
∆ r H1 = (17.61 ± 0.17) kJ·mol–1
[56YAT/VAS2]
log10 K 2 = 1.60
∆ r H 2 = – (9.25 ± 0.25) kJ·mol–1
sp
25.0
var
log10 K1 = 3.62
pot
25.0
0.1 M Me4NCl
log10 K1 = 6.98
(8.73 ± 0.25)
[64HAM/MOR]
pot
15.0
0.1 M KNO3
log10 K1 = 6.22
(7.94 ± 0.25)
[72FRE/STU]
pot
5.0
0.1 M KNO3
log10 K1 = 5.63
15.0
0.1 M KNO3
log10 K1 = 5.81
25.0
0.02 M KNO3
log10 K1 = 6.39
25.0
0.05 M KNO3
log10 K1 = 6.04
25.0
0.1 M KNO3
log10 K1 = 5.94
(7.69 ± 0.25)
25.0
0.2 M KNO3
log10 K1 = 5.60
(7.78 ± 0.30)
35.0
0.1 M KNO3
log10 K1 = 6.21
0.1 M KNO3
amp
30.0
0.1 M NH4NO3
[58VAI/RAM]
[73PER/SEC]
(7.53 ± 0.25)
∆ r H m ((278 − 308) K) = (30.6 ± 7.4) kJ·mol–1
log10 K1 = 5.35
(7.11 ± 0.30)
[78KAR/HUB]
(Continued on next page)
V.6 Group 15 compounds and complexes
210
Table V-29 (continued)
Method
t /°C
Medium (aq)
Reported value
Recalculated value (a, b)
log10 K ο
Reference
ligand: HP2 O37− (hydrogen pyrophosphate)
Ni2+ + HP2 O37− U HNiP2 O 7−
pot
25.0
0.1 M Me4NCl
log10 K1 = 3.83
(5.14 ± 0.25)
[64HAM/MOR]
pot
15.0
0.1 M KNO3
log10 K1 = 3.50
(4.79 ± 0.25)
[72FRE/STU]
5.0
0.1 M KNO3
log10 K1 = 3.15
15.0
0.1 M KNO3
log10 K1 = 3.55
25.0
0.02 M KNO3
log10 K1 = 4.01
25.0
0.05 M KNO3
log10 K1 = 3.78
pot
[73PER/SEC]
(4.84 ± 0.25)
25.0
0.1 M KNO3
log10 K1 = 3.71
(5.02 ± 0.25)
25.0
0.2 M KNO3
log10 K1 = 3.39
(5.02 ± 0.30)
35.0
0.1 M KNO3
log10 K1 = 4.07
0.1 M KNO3
∆ r H m ((278 − 308) K) = (47.9 ± 10.0) kJ·mol–1
(a) Uncertainties are the estimated uncertainties as discussed in the text.
(b) Values from most studies were re-calculated neglecting explicit association of K+ or NH +4 with pyrophosphate, and cannot be averaged directly with the values obtained in solutions of tetraalkylammonium
salts.
V.6.3 Arsenic compounds
V.6.3.1 Nickel arsenides
The nickel-arsenic phase diagram [61YUN], [87SIN/NAS] is moderately complex, with
the compounds Ni5As2(cr), Ni11As8(cr) (nickel-deficient Ni3As2(cr)), NiAs(cr) and
NiAs2(cr) (with a β to α conversion near 870 K).
V.6.3.1.1
NiAs(cr)
There are at least two sources available for the enthalpy of formation of NiAs(cr). Predel and Ruge [72PRE/RUG] reported – 36.0 kJ·g-at–1 (– 72.0 kJ·mol–1) for the formation
enthalpy based on the heat of dissolution in liquid tin at an unstated temperature. As
discussed in Appendix A, a third law analysis results in the value ∆ f H mο (NiAs, cr,
298.15 K) = – (73.60 ± 8.00) kJ·mol–1. Skeaff et al. [85SKE/MAI] used an electrochemical technique to determine the Gibbs energies of reaction of NiAs(cr) with oxygen
to form NiO (bunsenite) and As4O6(g) at temperatures from 711 to 833 K. As discussed
in Appendix A, a third law analysis with the heat capacity equation of Muldagalieva et
al. [95MUL/CHU], and auxiliary data consistent with those used in the present review,
leads to ∆ f H mο (NiAs, cr, 298.15 K) = – (69.73 ± 5.00) kJ·mol–1. The values are consistent within the lower uncertainties. The selected value of ∆ f H mο is the weighted average
of the two experimental values, and the selected value of S mο is the recalculated value
from Skeaff et al. [85SKE/MAI]:
V.6 Group 15 compounds and complexes
211
∆ f H mο (NiAs, cr, 298.15 K) = – (70.82 ± 4.24) kJ·mol–1
S mο (NiAs, cr, 298.15 K) = (50.76 ± 6.00) J·K–1·mol–1.
The heat capacity equation of Muldagalieva et al. [95MUL/CHU] is selected in
the present review,
–3
[C pο,m ]800K
(T/K) + 1.58×105 (T/K)–2) J·K–1·mol–1
298.15K (NiAs, cr) = (36.54 + 41.87 × 10
as is the value of
C pο,m (NiAs, cr, 298.15 K) = (50.8 ± 4.0) J·K–1·mol–1.
The heat capacity equation was claimed by the authors to be applicable over
the temperature range 298.15 to 675 K, and in the present review it is accepted over a
slightly expanded range to 800 K to accommodate the recalculation of the equilibrium
data of Skeaff et al. [85SKE/MAI]. Nozue et al. [97NOZ/KOB] measured the specific
heat of NiAs(cr) for temperatures between 1.7 K and 30 K, but data linking these values
to those near 300 K are lacking. The above selections yield:
∆ f Gmο (NiAs, cr, 298.15 K) = – (66.6 ± 4.6) kJ·mol–1.
V.6.3.1.2
Ni11As8(cr)
The heat capacity equation of Muldagalieva et al. [93MUL/ISA] is selected in the present review,
–3
[C pο,m ]800K
(T/K) + 0.018×105 (T/K)–2 J·K–1·mol–1
298.15K (Ni11As8, cr) = 380.12 + 367.27×10
as is the value of:
C pο,m (Ni11As8, cr, 298.15 K) = (490 ± 40) J·K–1·mol–1.
The heat capacity equation was claimed by the authors to be applicable over
the temperature range 298.15 to 700 K, and in the present review it is accepted over a
slightly expanded range to 800 K to accommodate the recalculation of the equilibrium
data of Skeaff et al. [85SKE/MAI].
Skeaff et al. [85SKE/MAI] used an electrochemical technique to determine the
Gibbs energies of reaction of Ni11As8(cr) with oxygen to form NiO (bunsenite) and
As4O6(g) at temperatures from 805 to 991 K. As discussed in Appendix A, a third law
analysis with the heat capacity equation of Muldagalieva et al. [93MUL/ISA] and auxiliary data consistent with those used in the present review leads to the selected values:
∆ f H mο (Ni11As8, cr, 298.15 K) = – (743.0 ± 35.0) kJ·mol–1
S mο (Ni11As8, cr, 298.15 K) = (518 ± 60) J·K–1·mol–1.
The above selections yield:
∆ f Gmο ( Ni11As8, cr, 298.15 K) = – (715.8 ± 39.3) kJ·mol–1.
V.6 Group 15 compounds and complexes
212
V.6.3.1.3
Ni5As2(cr)
Skeaff et al. [85SKE/MAI] also determined the Gibbs energies of reaction of Ni5As2(cr)
with oxygen to form NiO (bunsenite) and As4O6(g) at temperatures from 937 to 1139 K.
No experimental heat capacity values appear to have been reported for Ni5As2(cr).
Based on the approximation ∆ r C p ,m = 0 used by the original authors, and auxiliary data
from the present review, ∆ f H mο (Ni5As2, cr, 298.15 K) = – (244.6 ± 20.0) kJ·mol–1,
S mο (Ni5As2, cr, 298.15 K) = (196.2 ± 30.0) J·K–1·mol–1. Stolyrova [81STO] has reported
a much less negative enthalpy of formation, – (204.0 ± 5.8) kJ·mol–1, for the same solid,
and Skeaff et al. [85SKE/MAI] noted that her reported value is inconsistent with the
observed stability of Ni5As2(cr) solid with respect other phases [87SIN/NAS]. For that
reason, the values of Skeaff et al. are selected in the present review:
∆ f H mο (Ni5As2, cr, 298.15 K) = – (244.6 ± 20.0) kJ·mol–1
S mο (Ni5As2, cr, 298.15 K) = (196.2 ± 30.0) J·K–1·mol–1.
The above selections yield:
∆ f Gmο (Ni5As2, cr, 298.15 K) = – (237.6 ± 21.9) kJ·mol–1.
V.6.3.1.4
NiAs2(cr)
An experimental value for ∆ f H mο of – (90.06 ± 2.30) kJ·mol–1 (1σ uncertainty) has been
reported by Stolyrova [82STO] for this compound. Skeaff et al. [85SKE/MAI] were
unable to establish a proper equilibrium with this compound in their apparatus, although
some experiments suggested approximately the same value. In the present review, the
value is selected, but with an increased uncertainty:
∆ f H mο (NiAs2, cr, 298.15 K) = – (90.1 ± 8.0) kJ·mol–1.
V.6.3.2 Nickel arsenates
V.6.3.2.1
Solid nickel arsenates
Several anhydrous [58TAY/HEY] and hydrated [66GME] nickel arsenates have been
identified, but quantitative chemical thermodynamic data for these solids are sparse.
Chukhlantsev [56CHU3] published values for the solubility of Ni3(AsO4)2·8H2O
(annabergite) and its apparent solubility product [56CHU4] (K = (3.1 ± 2.4)×10–26, at
20°C for:
Ni3(AsO4)2·8H2O(cr) U 3Ni2+ + 2 AsO34− + 8 H2O(l).
(V.101)
Nishimura et al. [88NIS/ITO] measured the solubility at 25°C of what was
probably the same solid, based on measurements over a wider range of pH (4 to 8), and
reported log10 K (V.102) = 9.9 ( log10 K (V.101) = – 26.76) for:
Ni3(AsO4)2·8H2O(cr) + 4H+ U 3Ni2+ + 2 H 2 AsO 4− + 8H2O(l).
(V.102)
V.6 Group 15 compounds and complexes
213
Langmuir et al. [99LAN/MAH] re-analysed these solubility data using a model
that included formation of the complex NiHAsO4(aq), and reported log10 K (V.101) =
− 28.38. The experiments of Nishimura et al. [88NIS/ITO] which involved equilibration
for periods between one week and four months, indicate greater stability for the solid
than was determined by Chukhlantsev [56CHU4]. The value of log10 K ο (V.101) selected in the present review:
log10 K ο (V.101) = – (28.1 ± 0.5)
is based on the work of Nishimura et al. [88NIS/ITO] (see Appendix A), the auxiliary
data for the present review, and the effect of formation of NiHAsO4(aq) as estimated by
Langmuir et al. [99LAN/MAH].
From this, ∆ r Gm (V.101) can be calculated and
∆ f Gmο (Ni3(AsO4)2·8H2O, cr, 298.15 K) = – (3491.6 ± 8.8) kJ·mol–1.
Omarova and Sharipov measured the heats of solution of NiSO4·7H2O,
Na3AsO4, Na2SO4, H2O and Ni3(AsO4)2·8H2O in aqueous HCl, and reported
∆ f H mο (Ni3(AsO4)2·8H2O, s, 298.15 K) = – (4179.0 ± 9.2) kJ·mol–1 [80OMA/SHA]. Artamonova and Kasenov [89ART/KAS] reported ∆ f H mο (Ni3(AsO4)2·8H2O, s, 298.15 K)
= – (4065.4 ± 21.4) kJ·mol–1, based on the heat of reaction of Na3AsO4(cr) with an
aqueous nickel chloride solution to precipitate hydrated Ni3(AsO4)2. In the latter experiment, the less negative value found for the enthalpy of formation suggests that the
initially precipitated compound might not have been well crystallised, and that the value
of Omarova and Sharipov [80OMA/SHA] is to be preferred. In the present review, in
the absence of information concerning the auxiliary data used by the authors, the value
of Omarova and Sharipov [80OMA/SHA] is selected with an estimated uncertainty of
20 kJ·mol–1:
∆ f H mο (Ni3(AsO4)2·8H2O, 298.15 K) = – (4179.0 ± 20.0) kJ·mol–1,
which leads to:
S mο (Ni3(AsO4)2·8H2O, 298.15 K) = (540.8 ± 73.3) J·K–1·mol–1.
V.6.3.2.2
Aqueous nickel arsenato complexes
Langmuir et al. [99LAN/MAH] reported an estimated value of log10 K = 2.90 at
298.15 K for
Ni2+ + HAsO 24 − U NiHAsO4(aq)
(V.103)
based on work described in a thesis by Whiting [92WHI].
The value is similar to the value (3.05 ± 0.09) for the corresponding phosphate
complex (cf. Section V.6.2.2.1), and is an acceptable analogue value. In the present review, the value from Langmuir et al. [99LAN/MAH] is selected, but because of the
V.6 Group 15 compounds and complexes
214
unavailability of experimental values for comparison, an uncertainty of ± 0.3 is assigned,
log10 K ο ((V.103), 298.15 K) = (2.9 ± 0.3).
This selection implies:
∆ f Gmο ( NiHAsO4, aq, 298.15 K) = – (776.9 ± 4.4) kJ·mol–1.
V.6.3.3 Nickel arsenites
V.6.3.3.1
Solid nickel arsenites
Several nickel arsenites [66GME] have been identified, but there are very limited
chemical thermodynamic data for these solids. Chukhlantsev [57CHU] dissolved samples of nickel orthoarsenite in dilute nitric acid solutions at 20°C over 12 hours to form
solutions with aqueous nickel concentrations of 8.7 × 10–3 mol·dm–3 at a “pH” value of
6.75 and 3.1×10–3 at a “pH” value of 7.10 mol·dm–3. The author did not report a solubility product based on their measurements but, as discussed in Appendix A, if the dissolution of the solid is assumed to correspond to the reaction:
Ni3(AsO3)2·xH2O(cr, hydr.) + 6H+ U 3Ni2+ + 2HAsO2(aq) + (2 + x)H2O(l) (V.104)
the selected equilibrium constant can be calculated:
log10 K ο ((V.104), 298.15 K) = (28.7 ± 0.7).
From this, ∆ r Gm (V.104) can be calculated and, if the water of hydration is not
explicitly included in the chemical formula,
∆ f Gmο (Ni3(AsO3)2, cr, hydr, 298.15 K) = – (1253.6 ± 9.3) kJ·mol–1.
V.7 Group 14 compounds and complexes
V.7.1 Carbon compounds and complexes
V.7.1.1 Nickel carbonates
V.7.1.1.1
Solid nickel carbonates
V.7.1.1.1.1
NiCO3(cr)
V.7.1.1.1.1.1
V.7.1.1.1.1.1.1
Crystallography and mineralogy of nickel carbonate
NiCO3(cr), gaspéite
The crystal structure of NiCO3(cr) was definitively determined by Isaacs [63ISA]. It is
of the calcite type, hexagonal, space group: R 3 c, Z = 6, with unit cell dimensions a0 =
4.609 Å, c0 = 14.737 Å (according to JCPDS-ICDD card No. 12-771).
V.7 Group 14 compounds and complexes
215
The natural occurrence of a new mineral (Ni0.49Mg0.43Fe0.08)CO3 as a vein enclosed in siliceous dolomite in the Gaspé Peninsula, Quebec, was reported by Kohls and
Rodda [66KOH/ROD]. The name gaspéite was selected for the nickel carbonate end
member (approved by IMA 1966). The vein consists of essentially pure magnesian
gaspéite with minor amounts of annabergite (Ni3(AsO4)2·8H2O), magnesite (MgCO3)
and dolomite (CaMg(CO3)2). Enclosed in the magnesian gaspéite are a small amount of
serpentine ((Mg, Fe)3Si2O5(OH)4) and a (Cr, Al) spinel. Well-formed crystals of millerite (NiS), niccolite (NiAs), annabergite, gersdorffite (NiAsS) and magnesite are
found outside the vein in the buff siliceous dolomite.
The X-ray powder diffraction patterns of magnesian gaspéite differ slightly
from synthetic NiCO3(cr), see JCPDS-ICDD card No. 23-437. Gaspéite from the Gaspé
Peninsula is light green, its luster is vitreous to dull and it has a yellow-green streak.
The Mohs hardness is 4.5 to 5, the specific gravity is ρ(obs.) = (3.71 ± 0.01) and
ρ(calc.) = 3.748 g·cm–3.
V.7.1.1.1.1.1.2
Ni2(CO3)(OH)2, nullaginite
Nullaginite occurs as a secondary mineral in serpentinised peridotites in the Nullagine
district (Western Australia). Mineral and name were approved by the IMA Commission
in 1978. It is bright green with a luster varying from dull to silky. Composition and Xray powder diffraction pattern of nullaginite were investigated by Nickel and Berry
[81NIC/BER]. Its crystal system is monoclinic, space group: P21/m, cell dimension: a0
= 9.236 Å, b0 = 12.001 Å, c0 = 3.091 Å, Z = 4, β = 90.48°, Vcell = 342.60 Å3, ρ(obs.) =
3.56 and ρ(calc.) = 4.099 g·cm–3 (according to JCPDS-ICDD card No 35-501). The
large discrepancy between ρ(calc.) and ρ(obs.) is attributed to the presence of admixed
water in the specimen. The ideal composition of Nullaginite is Ni2(CO3)(OH)2, but it
occurs also together with Cr, Mg, Fe, and Cu and belongs to the rosasite group. This
group is composed of minerals with either a monoclinic or triclinic symmetry and a
general formula of M2(CO3)(OH)2. The M ion can be Co2+, Cu2+, Zn2+, Mg2+ and/or
Ni2+. Minerals of this group composed of colour stimulating metals such as cobalt, copper and nickel are extremely colourful, usually green to blue. Most of these minerals
range from rare to extremely rare. No synthesis of nullaginite has been reported and no
thermodynamic data have been determined so far.
V.7.1.1.1.1.1.3
Ni2(CO3)(OH)2·H2O, otwayite
The crystal structure of otwayite was investigated by Nickel et al. [77NIC/ROB]. Its
fibre-rotation and X-ray powder diffraction patterns can be indexed on an orthorhombic
unit cell with a = 10.18 Å, b = 27.4 Å, c = 3.22 Å, Vcell = 898.16 Å3, Z = 8, ρ(obs.) =
3.41 and ρ(calc.) = 3.346 g·cm–3 (according to JCPDS-ICDD card No. 29-868).
Otwayite occurs in Heazlewood (Tasmania), in the Nullagine region, but also
together with Widgiemoolthalite in Widgiemooltha (both in Western Australia). Mineral and formula were approved by the IMA in 1977. Otwayite was named after Charles
Albert Otway (1922 –), miner and prospector of Cosnells in Australia [77NIC/ROB].
V.7 Group 14 compounds and complexes
216
No synthesis of otwayite has been reported and no thermodynamic data have been determined so far.
V.7.1.1.1.1.1.4
Ni3(CO3)(OH)4·4H2O, zaratite
Isaacs studied the structure of zaratite [63ISA]. The cell dimensions are: a = 6.16 Å, Z =
1; Vcell = 233.74 Å3, ρ(calc.) = 2.31 g·cm–3, crystal system: isometric, space group: unknown (in part amorphous).
Zaratite occurs in Cape Ortegal, Galicia, (Spain); in Broken Hill, New South
Wales, (Australia); and in Pennsylvania, (USA). The mineral was named after Antonio
Gil y Zárate (1793 – 1861). Data for physical and chemical properties vary significantly
[63ISA]. The average chemical composition is close to Ni3(CO3)(OH)4·4H2O, however,
analytical results scatter in the following range: NiO 56.9 – 61.2%, CO2 13.5 – 15.7%,
H2O 23.2 – 27.1%. Zaratite is rare, forming emerald green secondary crusts on other
nickel minerals. Its hardness is 3.5. From X-ray investigation of artificially synthesised
as well as natural specimens it was concluded that zaratite is not a single mineral and
should not be called one. Consequently no thermodynamic data can be ascribed to this
poorly defined mixture of nickel hydroxide carbonates.
V.7.1.1.1.1.1.5
Ni5(CO3)4(OH)2·4–5H2O, widgiemoolthalite
The crystal structure of widgiemoolthalite was investigated by Nickel et al.
[93NIC/ROB]. It has an ideal composition of Ni5(CO3)4(OH)2·4–5H2O. The monoclinic
unit cell is similar to that of hydromagnesite, space group: P21/c. The cell dimensions
are: a = 10.06 Å, b = 8.75 Å, c = 8.32 Å, Z = 2; β = 114.3°, Vcell = 667.48 Å3, ρ(obs.) =
3.13 and ρ(calc.) = 2.97 g·cm–3 (according to JCPDS-ICDD card No. 46-1398).
Widgiemoolthalite is found together with a number of other secondary nickel
minerals in the weathered zone in the Widgie 132N mine in Widgiemooltha
[93NIC/ROB]. It is bluish green and occurs mainly as a fibre. The mineral was discovered in 1992, named after its location and approved by IMA 1993. In reality the composition of widgiemoolthalite is close to (Ni, Mg)5(CO3)4.15(OH)1.7·5.12 H2O. No
thermodynamic data of this mineral are known.
V.7.1.1.1.1.2
Heat capacity and entropy of NiCO3(cr)
A high temperature heat capacity function claimed to be valid at 298 ≤ (T/K) ≤ 700 for
NiCO3(cr) is given in thermodynamic data compilations:
C pο,m (NiCO3, cr) = (a + b (T/K) + e (T/K)–2) J·K–1·mol–1
–3
[C pο,m ]700K
(T/K) – 12.34×105(T/K)–2) J·K–1·mol–1
298K (NiCO3, cr) = (88.701 + 38.91×10
(V.105)
Originally the following coefficients were given: a = 92.05, b = 38.91×10–3
and e = – 12.34×105 [77BAR/KNA]. In Knacke et al. [91KNA/KUB] and
Kubaschewski et al. [93KUB/ALC] the coefficients b and e have been retained, but a
V.7 Group 14 compounds and complexes
217
has been slightly modified (a = 88.701). The only source of original C pο,m measurements
up to 300 K found so far is [64KOS/KAL]. The latter authors estimated the maximum
error of the standard entropy to be ± 0.4 J·K–1·mol–1, however, they presented their results only graphically:
S mο (NiCO3, cr, 298.15 K) = (85.4 ± 2.0) J·K–1·mol–1.
Robie and Hemingway assigned an uncertainty of ± 2.0 J·K–1·mol–1 to this
quantity, which appears reasonable because, as mentioned above, the C pο,m function can
only be evaluated graphically [95ROB/HEM]. Both the numerical value for this quantity and its error limits have been selected for this compilation. The so-called ‘Latimer’
contributions would result in S mο (NiCO3, cr, 298.15 K) = (81.7 ± 5.0) J·K–1·mol–1
whereby the actually selected value falls within these error limits [93KUB/ALC].
The following value can be read off directly from Fig. 1 of [64KOS/KAL]:
C pο,m (NiCO3, cr, 298.15 K) = (90.3 ± 4.1) J·K–1·mol–1.
The uncertainty is assigned in such a way that the C pο,m values calculated according to Equation (V.105) as well as estimated according to the cationic and anionic
contributions, given by Kubaschewski et al. [93KUB/ALC], fall within the error limits.
Although neither the experimental nor the semi-empirical basis for Equation (V.105)
could unambiguously be retraced, it has been retained in the current review.
V.7.1.1.1.1.3
Thermodynamic analysis of solubility data on gaspéite
Solubility measurements are especially valuable for the determination of an internally
consistent set of thermodynamic data for nickel carbonates. Other methods turned out to
be less convenient. Thus the calorimetric determination of the dissolution enthalpy according to Reaction (V.106) is bound to be inaccurate, because the result depends to a
high extent on the state of carbon dioxide. It is well known that keeping CO2 quantitatively either in solution or in the gas phase is quite difficult.
NiCO3(cr) + 2H+ U Ni2+ + CO2(g) + H2O(l)
V.7.1.1.1.1.3.1
(V.106)
Analysis of hydrothermal decomposition
Another method to determine thermodynamic data of nickel carbonate consists in the
investigation of the decomposition equilibrium (V.107). Tareen et al. [91TAR/FAZ]
determined the standard molar quantities of formation of gaspéite, NiCO3(cr) from
hydrothermal T, p decomposition data of this reaction, but probably overestimated the
precision achieved: ∆ f Gmο = – (628.36 ± 1.24) kJ·mol–1 (– (632.3 ± 6.0) kJ·mol–1),
∆ f H mο = – (703.38 ± 1.24) kJ·mol–1 (– (709.2 ± 6.0) kJ·mol–1).
NiCO3(s) U NiO(s) + CO2(g)
(V.107)
The numerical values in brackets have been found when the experimental results of [91TAR/FAZ] were re-evaluated according to the present state of the art, see
Appendix A.
V.7 Group 14 compounds and complexes
218
V.7.1.1.1.1.3.2
Analysis of solubility studies
It was shown recently that solubility data obtained from NiCO3·5.5H2O(cr), hellyerite,
were erroneously ascribed to NiCO3(cr), gaspéite [2001GAM/PRE]. Thus, in fact it was
Reiterer [80REI] who determined the solubility constant of gaspéite defined by Reaction (V.106) for the first time. He employed the pH variation method to investigate the
temperature dependence of log10 *K p, s ,0 (V.106) at constant ionic strength I = 1.0
mol·kg−1 (Na)ClO4. Kloger [82KLO] carried out additional experiments at I = 0.2
mol·kg–1 (Na)ClO4 and 363.15 K.
The solubility data of synthetic NiCO3 were thermodynamically analysed to
obtain the respective standard molar quantities of formation ∆ f Gmο and ∆ f H mο . An
analogous set of equations was used as described in Section V.3.2.2.3 (Equations (V.41)
– (V.44)). Solubility constants were extrapolated to infinite dilution using the Specific
Ion-interaction Theory [97GRE/PLY2]. All calculations were based on auxiliary quantities which were either recommended by CODATA (see NEA TDB auxiliary data set) or
recently critically evaluated (see Section V.2.1).
The following set of reliable thermodynamic quantities has been derived
[82GAM/REI], [98GAM/KON], [2002WAL/PRE]:
log10 *K p, s ,0 ((V.106), 298.15 K) = (7.16 ± 0.18)
∆ f Gmο (NiCO3, cr, 298.15 K) = – (636.4 ± 1.3) kJ·mol–1
∆ f H mο (NiCO3, cr, 298.15 K) = – (713.3 ± 1.4) kJ·mol–1.
These values obtained from solubility measurements of pure synthetic
NiCO3(cr) fall well within the error limits of those re-evaluated from decomposition
studies, but are clearly more precise.
V.7.1.1.1.2
V.7.1.1.1.2.1
NiCO3·5.5H2O(cr)
Crystallography and mineralogy of nickel carbonate hydrate
The preparation and X-ray powder diffraction pattern of nickel carbonate hexahydrate,
NiCO3·6H2O(cr), were first described by Rossetti-François [52ROS]. The crystal system
is monoclinic-prismatic, space group: C2/c, cell dimensions: a0 = 10.770 Å, b0 = 7.299
Å, c0 = 18.681 Å, Z = 8; β = 94.00°, Vcell = 1464.94 Å3, ρ(obs.) = 1.97 g·cm–3, ρ(calc.) =
1.975 g·cm–3 (according to JCPDS-ICDD cards 12-276 and 24-523, where the empirical
formulae are NiCO3·6H2O(cr) in the former and NiCO3·5.5H2O(cr) in the latter).
The natural occurrence of hellyerite, NiCO3·5.5H2O(cr), in the Lord Brassey
nickel mine near Heazlewood (Tasmania), was first reported by Williams et al.
[59WIL/THR]. The Mohs hardness is 2.5, its luster is vitreous (glassy) and its streak is
white. The X-ray powder diffraction pattern was found to be similar to that of synthetic
nickel carbonate hexahydrate, prepared by the method of Rossetti-François.
V.7 Group 14 compounds and complexes
219
The single crystal structure determination by Threadgold [63THR] of the natural product may be not perfect due to extensive twinning of the material. Hellyerite
forms predominantly twin crystals to enhance the low crystal symmetry. Therefore the
crystal structure needs to be refined to resolve its subtleties.
Hellyerite has a prominent layered structure parallel to (001) and is made up of
two distinct layers, a NiCO3(H2O)4 layer and an almost planar sheet of H2O molecules
[63THR]. This layer structure with its unknown arrangement of H2O molecules is presumably responsible for chemical and thermogravimetric analyses of hellyerite resulting
in a stoichiometric formula of NiCO3·5.5H2O(cr) i.e., NiCO3(H2O)4·1.5H2O(cr) instead
of NiCO3·6H2O(cr) [2001GAM/PRE]. Crystallographic investigations to solve the hellyerite structure definitively are in progress.
V.7.1.1.1.2.2
Heat capacity and entropy of NiCO3·5.5H2O(cr)
Apparently
no
experimental
low-temperature
heat
capacity data
of
NiCO3(H2O)4·1.5H2O(cr) have been reported so far. A crude estimate was obtained by
analogy to magnesium carbonate and its hydrates. When lansfordite (MgCO3·5H2O(cr)),
nesquehonite (MgCO3·3H2O(cr)) and magnesite (MgCO3(cr)) are compared the heat
capacity can roughly be approximated by Equation (V.108) [99KON/KON],
C pο,m (MgCO3·nH2O, cr, 298.15 K) = (76.09 + n × 58.0) J·K–1·mol–1.
(V.108)
With Equation (V.105) the following value can tentatively be used for the heat
capacity of NiCO3·5.5H2O(cr):
C pο,m (NiCO3·5.5H2O, cr, 298.15 K) = (405.4 ± 50.0) J·K–1·mol–1.
The entropy of NiCO3·5.5H2O(cr) was determined by solubility measurements,
S mο (NiCO3·5.5H2O, cr, 298.15 K) = (311.1 ± 10.0) J·K–1·mol–1,
see Section V.7.1.1.1.2.3, and is selected in this review.
The standard entropy obtained from the thermodynamic analysis of solubility
data agrees reasonably well with S mο (NiCO3·5.5H2O, cr, 298.15 K) = 307.9 J·K−1·mol–1
estimated according to Latimer [52LAT].
V.7.1.1.1.2.3
V.7.1.1.1.2.3.1
Thermodynamic analysis of solubility data on hellyerite
Solubility measurements of hellyerite, NiCO3·5.5H2O(cr)
As pointed out in Section V.7.1.1.1.1.3 solubility measurements of metal carbonates
provide an important method to determine the thermodynamic quantities. For hellyerite,
however, these data can be unambiguously assigned only when the stoichiometric composition is known. As formula and crystal structure are strongly interdependent, they
should be determined simultaneously. So far neither has been established definitively
and the thermodynamic quantities derived from solubilities must be used with caution.
V.7 Group 14 compounds and complexes
220
With hellyerite prepared by a new method, solubility measurements were carried out at different temperatures and at I = 1.0 m (Na)ClO4 [2001GAM/PRE]. In addition, the solubility was also determined at 25°C and different ionic strengths
[2002WAL/PRE]. In either case the pH variation method was used to study the dissolution reaction according to Reaction (V.109):
NiCO3·5.5H2O(cr) + 2H+ U Ni2+ + CO2(g) + 6.5H2O(l)
(V.109)
The thermodynamic calculations of this compilation are tentatively based on
the formula NiCO3·5.5H2O(cr) which agrees with Threadgold’s crystal structure analysis [63THR] and the composition of the synthetic product [2001GAM/PRE]. From the
weak temperature dependence of the solubility constant, log10 *K p, s ,0 , of hellyerite (see
Figure V-43) ∆ f H mο (NiCO3·5.5H2O, cr) and S mο (NiCO3(H2O)4·1.5H2O, cr) were calculated using a non-linear least squares optimising routine.
Figure V-43: Temperature dependence of hellyerite solubility. I = 1.0 mol·kg–1 NaClO4,
● 278.15, × 288.15, ■ 298.15, ▲ 308.15, 313.15 K, solid line: specific ion-interaction
formalism [97GRE/PLY2], dashed line: Pitzer model [91PIT].
11.6
NiCO3·5.5H2O(cr) +2 H+ U Ni2+ + CO2(g) + 6.5 H2O
11.4
*
log10 Kp,s,0
11.2
11.0
10.8
10.6
10.4
270
280
290
300
T/K
310
320
V.7 Group 14 compounds and complexes
221
Thermodynamic auxiliary data for H2O(l) and CO2(g) were taken from Table
IV-1 (T = 298.15 K) and from CODATA [89COX/WAG] (T ≠ 298.15 K). Data for Ni2+
were used as selected in this review (see Section V.2.1). Nickel hydroxo and carbonato
complexes are negligible [76BAE/MES], [2003BAE/BRA] in the pH ranges where the
experiments of [2001GAM/PRE] and [2002WAL/PRE] were carried out.
Figure V-44 shows the solubility constant of NiCO3·5.5H2O(cr) plotted as a
function of ionic strength at 298.15 K. Both the Pitzer equations [91PAP/PIT], [91PIT],
[99KON/KON] and the SIT, [97GRE/PLY2] were applied to model the ionic strength
dependence of log10 *K p, s ,0 . Whereas the Pitzer equations seem to predict the ionic
strength dependence of the solubility constant more accurately than the SIT approach,
the experimental data as a function of temperature coincide best with the latter, see
Figure V-43.
Figure V-44: Ionic strength dependence of hellyerite solubility. t = 25°C, Ionic medium:
I = 0.5 to 3.0 mol·kg–1 NaClO4, [2002WAL/PRE], solid line: specific ion-interaction
formalism [97GRE/PLY2], dashed line: Pitzer model [91PIT].
11.5
NiCO3·5.5H2O(cr) +2 H+ U Ni2+ + CO2(g) + 6.5 H2O
11.4
11.3
11.2
11.0
*
log10 Kp, s, 0
11.1
10.9
10.8
10.7
10.6
10.5
0.0
0.2
0.4
0.6
0.8
1.0
1.2
−1
(I / mol·kg )
1.4
1.6
1.8
2.0
1/2
A plot of log10 (aNi2+ ⋅ aH6.52 O ⋅ aH−2+ ) versus log10 pCO2 shown in Figure V-45 describes the thermodynamic analysis of the data presented in [11AGE/VAL],
[30MUL/LUB] and [2002WAL/PRE] in comparison with the optimised line (slope =
− 1.00). The activities of Ni2+ and H+ were calculated by using the SIT model. This result reveals a ratio of n[Ni]/n[CO2] close to unity in the solid phase, as is expected for
NiCO3·5.5H2O(cr). A re-evaluation of the experimental data of Müller and Luber
V.7 Group 14 compounds and complexes
222
[30MUL/LUB] as well as Ageno and Valla [11AGE/VAL] resulted in log10 *K p,ο s ,0 =
(10.56 ± 0.10), which compares favourably with
log10 *K p,ο s ,0 ((V.109), 298.15 K) = (10.63 ± 0.10)
calculated from experimental data of [2001GAM/PRE], [2002WAL/PRE]. Thus it can
be concluded that Ageno and Valla [11AGE/VAL] studied NiCO3·5.5H2O(cr) and not
NiCO3 ( log10 *K p,ο s ,0 = (7.13 ± 0.18)), see Appendix A. Obviously the solubility constants of hellyerite (Equation (V.109)) and gaspéite (Equation (V.106)) differ by 3.5
orders of magnitude (see Section V.7.1.1.1.1.3). This presents a further example that
solubility data measured at different temperatures and ionic strengths provide a useful
basis for the calculation of thermodynamic quantities when and only when the solid
samples investigated are analytically and structurally unambiguously characterised.
Moreover, a careful thermodynamic analysis showed that the experimental information
of rather old solubility studies either falls in line with the optimised set of thermodynamic properties obtained [11AGE/VAL], [30MUL/LUB] or reveals that presumably a
basic nickel carbonate was investigated [38SMU].
Figure V-45: Thermodynamic analysis of hellyerite solubility. Data of [11AGE/VAL]:
■, datum of [30MUL/LUB]: ▲, data of [2002WAL/PRE]: ◊ I = 0.5, ○ I = 1.0, + I = 2.0,
× I = 3.0 mol·kg–1 NaClO4. Solid straight line: calculated with log10 *K p,ο s ,0 = 10.63 selected in this review.
12.5
T = 298.15 K
12.0
−2
11.0
10.5
2
6.5
+
log10 {a Ni · aH O · aH }
11.5
2+
10.0
9.5
9.0
8.5
NiCO3·5.5H2O(cr) +2 H+ U Ni2+ + CO2(g) + 6.5 H2O
8.0
-1.5
-1.0
-0.5
0.0
0.5
log10 {p(CO2)/bar}
1.0
1.5
2.0
V.7 Group 14 compounds and complexes
223
Thus the selected standard Gibbs energy and standard enthalpy of formation
have been derived from solubility measurements at various temperatures and ionic
strengths.
∆ f Gmο (NiCO3·5.5H2O, cr, 298.15 K) = – (1920.9 ± 1.0) kJ·mol–1
∆ f H mο (NiCO3·5.5H2O, cr, 298.15 K) = – (2313.0 ± 3.1) kJ·mol–1.
A convenient overview of thermodynamic data on nickel carbonates and hydroxide is provided by a predominance diagram as depicted in Figure V-46.
Figure V-46: Three-dimensional predominance diagram for the system Ni2+–CO2–H2O.
aNi2+ = 10–4; black lines, solid phases: NiCO3, β-Ni(OH)2; gray lines, solid phases:
NiCO3·5.5H2O(cr), β-Ni(OH)2; aqueous phase: Ni2+.
.5H O
NiCO3·5 2
1
NiCO3
-1
2
0
log10 [pCO / p ]
0
H)2
β - Ni(O
-2
2+
Ni
-3
-4
290
310
T/
V.7.1.1.2
330
K
β - Ni(OH)2
350
370
4.0
5.0
6.0
7.0
8.0
9.0
10.0
pH
Aqueous nickel carbonato complexes
Recently the formation of carbonato complexes in the system Ni2+ – H2O – CO2 has
been critically discussed and re-evaluated in a seminal review of Hummel and Curti
[2003HUM/CUR]. So far only one paper has been published describing an attempt to
determine the equilibrium constant, β1, of Reaction (V.110) experimentally
[87EMA/FAR].
Ni2+ + HCO3− U NiHCO3+
(V.110)
224
V.7 Group 14 compounds and complexes
Emara et al. clearly misinterpreted their data and did not provide enough information to allow recalculation. Consequently, the stability constant of NiHCO3+ reported in [87EMA/FAR] cannot be included in this review.
Values of equilibrium constants for Reaction (V.110) estimated by various
procedures differ considerably 0.96 ≤ log10 b1ο ≤ 3.08 (T = 298.15 K) [76ZHO/BEZ],
[77MAT/SPO], [79MAT/SPO], [84FOU/CRI]. The stability constant of the carbonato
complex according to Reaction (V.111) has also been estimated leading to an even larger discrepancy: 2.56 ≤ log10 K1ο ≤ 6.87 (T = 298.15 K) [76ZHO/BEZ], [77MAT/SPO],
[79MAT/SPO], [81TUR/WHI], [84FOU/CRI].
Ni2+ + CO32 − U NiCO3(aq)
K1
(V.111)
One estimate exists for the equilibrium constant ( log10 K 2ο = 3.24) of Reaction
(V.112) [77MAT/SPO].
NiCO3(aq) + CO32 − U Ni(CO3 ) 22 −
K2
(V.112)
As the basis of the individual estimation procedures is rather dubious, variations of up to more than four log-units in these stability constants are to be expected
[2003HUM/CUR]. Again neither of these values appeared suitable to be included in this
review.
Hummel and Curti proposed estimating K1 for Equation (V.111) using either
the good correlation between the equilibrium constants of Ni(II) and Co(II) complexes
and the poor data available for K1 of CoCO3(aq) or the rather poor correlation between
Ni(II) and Zn(II) complexes and the excellent data for K1 for ZnCO3(aq)
[2003HUM/CUR]. Both methods result in similar lower and upper bounds 4
< log10 K1ο ((V.111), 298.15 K) < 5.5.
A comparison of the stabilities of transition metal hydrogen carbonato as well
as carbonato complexes led to 1 < log10 b1ο ((V.110), 298.15 K) < 2 and log10 K 2ο <
( log10 K1ο – 2) [2003HUM/CUR]. Even the careful and competent guesswork of
Hummel and Curti resulted in rough estimates only. Fortunately, in a recent paper,
Baeyens et al. [2003BAE/BRA] investigated Ni-carbonato and -oxalato complexes by
an ion exchange method. Ni-carbonato complexes were investigated at constant ionic
strength I = 0.5 mol·dm–3 NaClO4 / NaHCO3 and (22 ± 1)°C. The experimentally obtained complexation constant ( log10 K1 (295.15 K) = (2.9 ± 0.3)) was extrapolated to I =
0 with the SIT approach to give [97GRE/PLY2]:
log10 K1ο ((V.111), 298.15 K) = (4.2 ± 0.4).
This result was finally selected for the present review. The somewhat higher
uncertainty was assigned, because Baeyens et al. carried out their measurements at
22°C instead of 25°C, and used a relatively simple approximation to extrapolate
log10 K1ο to I = 0. This selection yields:
V.7 Group 14 compounds and complexes
225
∆ f Gmο (NiCO3, aq, 298.15 K) = – (597.6 ± 2.4) kJ·mol–1.
Only upper bounds can be given for the stabilities of NiHCO3+ and
Ni(CO3 ) 22 − : log10 b1ο ((V.110), 298.15 K) < 1.4 and log10 K 2ο ((V.112), 298.15 K) < 2
[2003BAE/BRA]. Both upper bounds compare well with the lower limits of the range
predicted by Hummel and Curti [2003HUM/CUR], but do not qualify for being included in the list of selected values in this review. Figure V-47 shows the nickel speciation when NiCO3(cr) dissolves in dilute sodium carbonate solutions under atmospheric
conditions. The formation of NiCO3(aq) becomes evident only at pH > 8.5. At pH > 9
NiCO3(aq) predominates.
Figure V-47: Calculated solubility of NiCO3(cr) in dilute sodium carbonate solutions
under atmospheric conditions.
-3.5
−1
25°C, I = 1.0 mol·kg NaClO4
log10(p (CO2) / bar) = − 3.5
2+
[Ni ]
+
[NiOH ]
[NiCO3]
[Ni(II)]tot
-4.5
−1
log10([Ni(II)] / mol·kg )
-4.0
-5.0
-5.5
-6.0
-6.5
-7.0
-7.5
-8.0
7.4
7.6
7.8
8.0
8.2
8.4
pH
8.6
8.8
9.0
9.2
9.4
226
V.7 Group 14 compounds and complexes
V.7.1.2 Nickel cyanides
V.7.1.2.1
Aqueous Ni(II) cyano complexes
The first quantitative data on the aqueous Ni(II) cyanide complexes were generated at
the end of XIXth century, when Varet reported the heat of formation of the Ni(CN) 24 −
complex [1896VAR]. Varet's ∆ r H m value is in good agreement with more recent determinations, but there has been an increase of nearly twenty orders of magnitude in the
reported log10 b 4 value of Ni(CN) 24 − , since it was first measured by Masaki in 1932
[32MAS]. Besides the numerical value of log10 b 4 , conflicting results also have been
obtained concerning the composition and protonation state of the species formed. Therefore, the discussion is divided into two parts.
V.7.1.2.1.1
Complexes in neutral and alkaline solution
Most authors have agreed that the formation of NiCN+, Ni(CN)2(aq) and Ni(CN)3− can
not be detected in the equilibrated solutions. Although, Shibata et al. claimed to detect
Ni(CN)2(aq) and possibly Ni(CN)3(H2O)– by electrophoresis of γ-ray irradiated
K2Ni(CN)4 solutions [66SHI/FUJ], the assignment of the measured peaks was rather
speculative. The absence of the above-mentioned species indicates that the stepwise
formation constants K1, K2 and K3 are much less than 4 b 4 , and thus K4 is probably
greater than 1014 [68KOL/MAR]. This very large value is likely related to the transformation of an octahedral to a square planar complex in the fourth stepwise complex formation process.
The first reported value for log10 b 4 [32MAS] was not accurate, since the electrode reactions involving Ni(II) are not reversible [50HUM/KOL].
Ni2+ + 4CN– U Ni(CN) 24 −
(V.113)
Reliable values were reported from spectrophotometric [59FRE/SCH], kinetic
[61MAR/BYD], [68KOL/MAR], [76PER], and pH-metric [63CHR/IZA],
[71IZA/JOH], [74PER] measurements. Freund and Schneider used an incorrect pK HCN
value, therefore the experimental data reported in [59FRE/SCH] were re-evaluated (see
discussion in Appendix A). The experimentally determined log10 b 4 values are collected in Table V-30. Although only a limited number of data are available in NaClO4
media, the precision of the constants is assumed to be good enough to perform an SIT
analysis (Figure V-48). The weighted linear regression using five data points yielded the
selected value of:
log10 b 4ο ((V.113), 298.15 K) = (30.20 ± 0.12).
From the slope in Figure V-48, ∆ε((V.113), NaClO4) = – (0.465 ± 0.045)
kg·mol–1 can be calculated. Using the selected value for ε(Ni2+, ClO −4 ), and ε(Na+, CN−)
= (0.07 ± 0.03) kg·mol–1 reported in [92BAN/BLI], ε(Na+, Ni(CN) 24 − ) = (0.185 ± 0.081)
kg·mol–1 can be derived. In KClO4 media, too narrow an ionic strength range is covered
to perform the SIT analysis [59FRE/SCH], but after correction for the Debye-Hückel
V.7 Group 14 compounds and complexes
227
contributions, the formation constants determined in these low ionic strength solutions
are almost constant, and are nearly identical to the selected log10 b 4ο ((V.113), 298.15 K)
value (see insert in Figure V-48).
At ratios [CN–]/[Ni2+] > 4, the pale yellow solution of diamagnetic Ni(CN) 24 −
becomes successively orange, red and deep red upon addition of cyanide ion. The reason for this change is evidently the formation of higher complexes, but the composition
and stability of these have been disputed for a long time. Several early reports
[23JOB/SAM], [43SAM], [59KIS/CUP], [59BLA/GOL], [59BLA/GOL2] interpreted
the colour change as evidence for the formation of hexacyanonickelate(II) ion,
Ni(CN)64 − . Some of them reported surprisingly high equilibrium constant values for the
reaction:
Ni(CN) 24 − + 2CN– U Ni(CN)64 −
[59KIS/CUP], [59BLA/GOL2]. Later, the formation of a weak pentacyano complex and
the non-existence of the earlier-suggested Ni(CN)64 − ion were definitely proven
[60MCC/JON], [62BEC/BJE], [64GEE/HUM], [65COL/PET], [71PIE/HUG]. The reported equilibrium constants for the reaction:
Ni(CN) 24 − + CN– U Ni(CN)35−
(V.114)
are collected in Table V-30. Since the pentacyano complex is rather unstable, high cyanide concentrations were used to achieve its formation, which resulted in considerable
replacement of the original background electrolyte by NaCN. Due to this medium effect, considerably higher uncertainties have been assigned to the equilibrium constants
than originally reported. The SIT analysis of the constants determined in NaClO4 media
(Figure V-49), resulted in log10 K 5ο ((V.114), 298.15 K) = – (1.70 ± 0.36), and
∆ε((V.114), NaClO4) = (0.00 ± 0.11) kg·mol–1. From the latter value, ε(Na+, Ni(CN)35− )
= (0.25 ± 0.14) kg·mol–1 can be derived. Using equilibrium constant from above, the
overall formation constant of the Ni(CN)35− species is
Ni2+ + 5 CN– U Ni(CN)35−
log10 b ((V.115), 298.15 K) = (28.5 ± 0.5).
ο
5
(V.115)
V.7 Group 14 compounds and complexes
228
Table V-30: Experimental equilibrium data for the Ni(II) cyanide system.
Method
t (°C)
Im
Medium
log10 b x
log10 b x
reported
retained(a)
Reference
Ni2+ + 4CN– U Ni(CN) 24−
pot
NaCN
0.78 M
25
(11.83 ± 0.17)
> 24
pol
KCl
0.10
?
sp
K(ClO4)
0.1
25
[32MAS]
[50HUM/KOL]
(29.37 ± 0.20)(b)
0.048
(29.45 ± 0.20)(b)
0.014
(29.79 ± 0.20)(b)
0.0028
(30.12 ± 0.20)(b)
[59FRE/SCH]
(31.0 ± 0.08)
0 corr
kin
NaCl
1.03
25
31.5
gl
(self)
0 corr
25
(30.1 ± 0.2)
[61MAR/BYD]
kin
NaClO4
0.10
25
gl
(self)
0 corr
10
25
(30.22 ± 0.05)
(30.22 ± 0.15)
40
(27.43 ± 0.09)
(27.43 ± 0.20)
(30.1 ± 0.2)
[63CHR/IZA]
(30.5 ± 0.3)
(30.48 ± 0.6)
[68KOL/MAR]
(32.20 ± 0.20)
(32.20 ± 0.30)
[71IZA/JOH]
gl
NaClO4
3.50
25
(31.12 ± 0.08)
(30.85 ± 0.15)
[74PER]
kin
NaClO4
3.50
25
(31.08 ± 0.05)
(30.81 ± 0.15)
[76PER]
[60MCC/JON]
Ni(CN)
sp
2−
4
–
+ CN U Ni(CN)
Na(ClO4)
3−
5
1.43
15.4
– (0.66 ± 0.02)
– (0.69 ± 0.30)
25.2
– (0.72 ± 0.02)
– (0.75 ± 0.30)
33.6
– (0.77 ± 0.02)
– (0.80 ± 0.30)
ir
Na(ClO4)
4.95
25
– (0.55 ± 0.02)
– (0.65 ± 0.30)
sp
Na(ClO4)
2.84
23
– (0.70 ± 0.02)
[63PEN/BAI]
[64GEE/HUM]
– (0.77 ± 0.30)
25
sp
KF-KCN
4.00
25
(0.03 ± 0.01)(c)
[65COL/PET]
sp
Na(ClO4)
2.21
20
– (0.77 ± 0.01)
[71PIE/HUG]
25
(a) Accepted values corrected to molal scale.
(b) Re-evaluated value, see Appendix A.
(c) In molal units.
(d) Corrected to 25°C, using the selected enthalpy values.
– (0.84 ± 0.30)(d)
V.7 Group 14 compounds and complexes
229
Figure V-48: Extrapolation to I = 0 of experimental formation constants of Ni(CN) 24 −
(V.113) determined in NaClO4 medium (insert shows the corresponding plot of data
reported for KClO4 medium).
30.5
Ni2+ + 4CN– U Ni(CN) 24 −
34
30.0
log10 β4 + 4D
33
29.5
32
0.00
0.02
0.04
0.06
0.08
0.10
0.12
31
30
0.0
0.5
1.0
1.5
2.0
−1
I / mol·kg
2.5
3.0
3.5
Figure V-49: Extrapolation to I = 0 of experimental equilibrium constants for Reaction
(V.114) determined in NaClO4 medium.
0.0
Ni(CN) 24 − + CN– U Ni(CN)35−
-0.5
log 10 K5 − 4D
-1.0
-1.5
-2.0
-2.5
-3.0
0
1
2
3
−1
I / mol·kg
4
5
V.7 Group 14 compounds and complexes
230
The selected thermodynamic formation constants correspond to:
∆ r Gmο ((V.113), 298.15 K) = – (172.4 ± 0.7) kJ·mol–1,
∆ r Gmο ((V.115), 298.15 K) = – (162.7 ± 2.9) kJ·mol–1
and thus the standard molar Gibbs energies of formation are:
∆ f Gmο ( Ni(CN) 24 − , 298.15 K) = (449.6 ± 10.1) kJ·mol–1,
∆ f Gmο ( Ni(CN)35− , 298.15 K) = (626.6 ± 12.9) kJ·mol–1.
The reaction enthalpy of formation of the tetracyano complex has been studied
calorimetrically by Varet [1896VAR] and by Izatt and his co-workers [63CHR/IZA],
[71IZA/JOH] (see Table V-31). The value determined in [63CHR/IZA] is selected in
this review:
∆ r H mο ((V.113), 298.15 K) = – (180.7 ± 4.0) kJ·mol–1.
Izatt et al. also determined the standard enthalpy of Reaction (V.113) at two
other temperatures (10 and 40°C), which allowed calculation of the standard molar heat
capacity of the reaction ∆ r C pο,m ((V.113), 298.15 K) = (150.6 ± 42.0) J·K–1·mol–1
[71IZA/JOH].
Neglecting the ionic strength dependence, ∆ r H mο ((V.114), 298.15 K) =
− (10.4 ± 4.0) kJ·mol–1 can be estimated from the temperature variation of log10 K 5
[60MCC/JON]. The combination of the above two enthalpy values yielded:
∆ r H mο ((V.115), 298.15 K) = – (191.1 ± 8.0) kJ·mol–1.
Table V-31: Experimental enthalpy values for Reactions (V.113) and (V.114).
Method
Medium
Im
t (°C)
∆ r H m (kJ·mol–1)
∆ r H m (kJ·mol–1)
reported
retained
Reference
Ni2+ + 4CN– U Ni(CN) 24−
cal
?
?
14
– (180 ± 4.2)
cal
(self)
0 corr
25
– (180.7 ± 2.0)
cal
(self)
0 corr
[1896VAR]
– (180.7 ± 4.0)
[63CHR/IZA]
[71IZA/JOH]
10
– (189.1 ± 1.2)
– (189.1 ± 2.0)
40
– (183.7 ± 0.8)
– (183.7 ± 2.0)
– 12.5
– (10.4 ± 4.0)(a)
Ni(CN) 24− + CN– U Ni(CN)35−
sp
Na(ClO4)
1.43
15 – 34
[60MCC/JON]
(a) Calculated from the temperature dependence of the corresponding formation constant, see Appendix A.
From these data
∆ f H mο ( Ni(CN) 24 − , 298.15 K) = (353.7 ± 14.7) kJ·mol–1,
and
V.7 Group 14 compounds and complexes
231
∆ f H mο ( Ni(CN)35− , 298.15 K) = (490.6 ± 19.4) kJ·mol–1
are selected in this review.
These above selections yield:
S mο ( Ni(CN) 24 − , 298.15 K) = (245.0 ± 36.6) J·K–1·mol–1,
S mο ( Ni(CN)35− , 298.15 K) = (278.8 ± 51.1) J·K–1·mol–1.
V.7.1.2.1.2
Protonated complexes
The cyano complexes of several metal ions are reported to undergo protonation in the
acidic pH-range [87BEC], but rather contradictory results are available for Ni(II). The
protonation constants of Ni(CN) 24 − ( log10 K1 = (5.4 ± 0.2), log10 K 2 = (4.5 ± 0.2),
log10 K 3 = (2.6 ± 0.4) at t = 25°C and I = 0.1 M NaClO4) were first reported by Kolski
and Margerum, based on a kinetic study of the acid dissociation of Ni(CN) 24 −
[68KOL/MAR]. Their equilibrium measurements between pH = 4 – 5, using a
spectrophotometric method, resulted in systematic changes in the calculated log10 b 4
values. This was taken as further proof for the presence of protonated species, although
the UV-VIS spectrum of Ni(CN) 24 − was found to be unaffected by the protonation. Two
successive protonations of Ni(CN) 24 − have been reported using a pH-metric method in
[74KAB], but the results are uncertain, and thus this work was not considered in the
present review (see Appendix A). Recently, Monlien et al. reported a kinetic study on
the cyanide exchange with the Ni(CN) 24 − complex [2002MON/HEL]. The kinetic data
confirmed the presence of protonated species, although the protonation did not affect
the 13C NMR spectrum of Ni(CN) 24 − . On the other hand, spectrophotometric
[59FRE/SCH] and careful potentiometric [63CHR/IZA], [71IZA/JOH], [74PER] studies
performed between pH = 4 – 7 did not indicate the existence of protonated species in
notable concentrations, although their kinetic role in the acid dissociation of Ni(CN) 24 −
was later corroborated [76PER].
To sum up, the kinetic role of the protonated species in some reactions of
Ni(CN) 24 − is well established, but the presence of such species in equilibrated solutions
is at least questionable. Additional experimental data would be needed for a definitive
conclusion.
V.7.1.3 Nickel thiocyanates
V.7.1.3.1
Aqueous nickel thiocyanato complexes
The thiocyanate ion is a widely studied ligand, mainly due to its ambidentate nature.
Aside from some general reviews [75NEW2], [79GOL/KOH], a critical survey of stability constants of thiocyanate complexes has been published [97BAH/PAR]. The latter
publication listed the majority of available thermodynamic data for the Ni(II) – thiocyanate system, but recommended values are given only for log10 b1 (V.116) at I = 1.0
and 1.5 M.
V.7 Group 14 compounds and complexes
232
Thiocyanic acid is a moderately strong acid. Only a few badly scattered data
are available for estimation of its protonation constant [97BAH/PAR]. Among them,
log10 K1 = – 0.87 at I = 2.0 M [82NEV/ANG] seems to be the most reliable. Thus, the
protonation of the thiocyanate anion is negligible even at the most acidic conditions
used [64TRI/CAL], [68MAL/TUR] for studying the Ni(II)-thiocyanate interaction.
The thiocyanate anion is N-coordinated in most of its Ni(II) compounds
[77CHE/BRO]. The aqueous Ni(II)-thiocyanate complexes have octahedral geometry,
although the [Ni(SCN)4(H2O)2]2– species is probably in equilibrium with a minimal
concentration of tetrahedral [Ni(SCN)4]2– [88BJE2]. Some early papers reported qualitative information on the Ni(II) – thiocyanate system [38CSO], [42MAJ] based on spectrophotometric studies. Later, several experimental techniques, namely ion exchange,
spectrophotometric, polarographic, distribution, IR, potentiometric and kinetic methods,
have been used to derive equilibrium data for the Ni(II)-thiocyanate complexes. Depending on the ligand-to-metal ratio, the formation of four species is generally recognised:
Ni2+ + qSCN– U Ni(SCN) 2q − q
(V.116)
with q = 1 – 4. The reported formation constants are listed in Table V-32. For reasons
discussed in Appendix A, the data reported in [61DAV/SMI], [61MOH/DAS],
[62FRO/LAR], [62TRI/CAL], [71DAS/DAS], [71LAN/CUM], [74NET/BAT] were not
considered in the present review. In some cases [53FRO2], [64TRI/CAL],
[74DIC/HOF], the formation constants reported for 20°C were corrected to 25°C, using
the enthalpy values selected below. A majority of the accepted experimental data refers
to the formation of the NiSCN+ species in perchlorate media. SIT analysis of these data
shows acceptable consistency (cf. Figure V-50). The weighted linear regression using
12 data points yielded the selected value of:
log10 b1ο ((V.116), q = 1, 298.15 K) = (1.81 ± 0.04).
Since no experimental data are available, in the above treatment ε(H+, SCN–)
and ε(Li , SCN–) ion interaction parameters were assumed to be equal to ε(Na+, SCN−).
From the slope in Figure V-50, ∆ε((V.116), q = 1, ClO −4 ) = − (0.109 ± 0.025) kg·mol–1
can be calculated. Using the selected values for ε(Ni2+, ClO −4 ) and ε(Na+, SCN–),
∆ε((V.116), q = 1, ClO −4 ) leads to a value of ε(NiSCN+, ClO −4 ) = (0.31 ± 0.04) kg·mol–1.
+
Less data are available for the formation of the Ni(SCN)2(aq) and Ni(SCN)3−
species. The SIT treatment of the accepted data (cf. Figure V-51 and Figure V-52) resulted in the following selected thermodynamic formation constants:
log10 b 2ο ((V.116), q = 2, 298.15 K) = (2.69 ± 0.07),
log10 b 3ο ((V.116), q = 3, 298.15 K) = (3.02 ± 0.18).
From the slopes, ∆ε((V.116), q = 2, ClO −4 ) = – (0.091 ± 0.043) kg·mol–1 and
∆ε((V.116), q = 3, ClO −4 ) = (0.14 ± 0.12) kg·mol–1 can be derived. These parameters
V.7 Group 14 compounds and complexes
233
lead to the values of ε(Ni(SCN)2(aq), Na+ + ClO −4 ) = (0.38 ± 0.06) kg·mol–1 (also see
[97ALL/BAN]) and ε(Na+, Ni(SCN)3− ) = (0.66 ± 0.13) kg·mol–1. The use of an interaction coefficient like ε(Ni(SCN)2(aq), Na+ + ClO −4 ) for a neutral species is not a standard
procedure within the context of the SIT as described in Appendix B, but appears to be
justified in the current case. Further experimental work could prove useful.
Several experimental values are published for log10 b 4 (or log10 K 4 )
[64TRI/CAL], [71LAN/CUM], (or [88BJE2]). In [64TRI/CAL] there was a rather small
excess of ligand ( cSCN− ≤ 0.5 M), and for both ionic strengths K3 < K4 is reported. Considering that no geometrical change of Ni(II) occurs during the stepwise complex
formation, the reported K4 values are probably overestimated. The method used in
[71LAN/CUM] and [90BJE2] is not compatible with the SIT. Although there is good
evidence for the existence of Ni(SCN) 24 − [88BJE2], the available data cannot be used to
derive a selected value.
The above listed selected log10 b q values correspond to:
∆ r Gmο ((V.116), q = 1, 298.15 K) = – (10.33 ± 0.22) kJ·mol–1,
∆ r Gmο ((V.116), q = 2, 298.15 K) = – (15.35 ± 0.40) kJ·mol–1,
∆ r Gmο ((V.116), q = 3, 298.15 K) = – (17.24 ± 1.03) kJ·mol–1,
and hence
∆ f Gmο (NiSCN+, 298.15 K) = (36.60 ± 4.08) kJ·mol–1,
∆ f Gmο (Ni(SCN)2, aq, 298.15 K) = (124.27 ± 8.04) kJ·mol–1,
∆ f Gmο ( Ni(SCN)3− , 298.15 K) = (215.09 ± 12.07) kJ·mol–1.
Table V-32: Experimental equilibrium data for the Ni(II) – thiocyanate system.
Method
cix
sp
sp
pot
sp
ir
pol
sp
Im
t (°C)
log10 b x
reported
log10 b x
retained(a)
Reference
Ni2+ + SCN– U NiSCN+
(Na)ClO4
1.05
0 corr(c)
0 corr
(1.18 ± 0.02)
–
(1.67 ± 0.03)
(1.50 ± 0.04)
(1.82 ± 0.02)
1.38
1.18
(1.17 ± 0.02)
(1.762 ± 0.005)
–
–
(1.12 ± 0.06)(b)
(1.69 ± 0.15)
[53FRO2]
Ni(NO3)2
Ni(ClO4)2 + KSCN
20
25
23
25
35
1.4
Medium
?
(NaClO4)
(Na, H)ClO4
(Na)ClO4
0.5 M
3.5
1.05(c)
0 corr
0.43
?
25
(1.77 ± 0.10)(d)
(1.20 ± 0.08)(d)
[58YAT/KOR]
[61MOH/DAS]
[61DAV/SMI]
[62FRO/LAR]
[62TRI/CAL]
[62WIL2]
(Continued on next page)
V.7 Group 14 compounds and complexes
234
Table V-32 (continued)
Method
dis
Medium
Ni2+ + SCN– U NiSCN+
(NaClO4)
Im
t (°C)
log10 b x
reported
log10 b x
accepted(a)
1.62
20
25
20
25
15
25
35
25
25
25
35
45
?
5
15
25
35
25
45
20
25
25
?
25
(1.14 ± 0.02)
–
(1.19 ± 0.02)
–
(1.34 ± 0.02)
(1.24 ± 0.02)
(1.17 ± 0.03)
(1.3 ± 0.2)
(1.34 ± 0.01)
(1.98 ± 0.02)
(1.85 ± 0.02)
(1.74 ± 0.02)
2.22
1.55
1.46
1.37
1.29
(1.09 ± 0.06)
(1.00 ± 0.03)
1.83
–
(1.14 ± 0.02)
0.3 – 1.44
1.1
–
(1.08 ± 0.10)(b)
–
(1.09 ± 0.10)(b)
(1.33 ± 0.10)
(1.23 ± 0.10)
(1.16 ± 0.10)
(1.29 ± 0.20)
(1.27 ± 0.10)
20
25
20
25
20
25
?
5
15
25
35
25
25
(1.64 ± 0.04)
–
(1.76 ± 0.06)
–
(1.68 ± 0.05)
–
2.43
2.23
2.04
1.86
1.69
(1.58 ± 0.05)
1.6
3.50
pol
(H)ClO4
0.67
dis
sp
pot
(Na)ClO4
(LiClO4)
KSCN
0.26
3.48
0 corr
aix
pol
KSCN
(KNO3)
kin
(NaClO4)
0 corr
0.2
0.2
0.2(c)
0.2
1.05
kin
cal
ir
dis
cix
dis
0 corr
(NaClO4)
(LiClO4)
(NaClO4)
1.05
3.48
1.05
Ni2+ + 2SCN– U Ni(SCN)2(aq)
(Na)ClO4
1.05
(NaClO4)
1.62
3.50
aix
pol
KSCN
(KNO3)
cal
dis
(NaClO4)
(NaClO4)
0 corr
0.2
0.2
0.2(c)
0.2
1.05
1.05
Reference
[64TRI/CAL]
[68MAL/TUR]
[69SUB/COR]
[70MIR/MAK]
[71DAS/DAS]
[71LAN/CUM]
[71TUR/MAL]
(1.07 ± 0.10)
(0.98 ± 0.10)
–
(1.79 ± 0.10)(b)
(1.12 ± 0.02)
(1.08 ± 0.10)
–
(1.54 ± 0.06)(b)
–
(1.63 ± 0.10)(b)
–
(1.48 ± 0.10)(b)
[73HUT/HIG]
[74DIC/HOF]
[74KUL3]
[74NET/BAT]
[76MUR/KUR]
[53FRO2]
[64TRI/CAL]
[71LAN/CUM]
[71TUR/MAL]
(1.54 ± 0.05)
(1.56 ± 0.20)
[74KUL3]
[76MUR/KUR]
(Continued on next page)
V.7 Group 14 compounds and complexes
235
Table V-32 (continued)
t (°C)
log10 b x
reported
log10 b x
accepted(a)
Reference
(1.81 ± 0.07)
–
(1.70 ± 0.10)
–
(1.3 ± 0.3)
–
2.62
(1.6 ± 0.2)
–
(1.66 ± 0.10)(b)
–
(1.52 ± 0.20)(b)
–
(1.01 ± 0.30)(b)
[53FRO2]
0 corr
1.05
20
25
20
25
20
25
?
25
Ni2+ + 4SCN– U Ni(SCN) 24−
(NaClO4)
1.62
3.50
KSCN
0 corr
NaSCN
0 corr
20
20
?
25
Method
cix
Im
Medium
Ni2+ + 3SCN– U Ni(SCN)3−
(Na)ClO4
1.05
dis
(NaClO4)
1.62
3.50
aix
cal
dis
aix
sp
KSCN
(NaClO4)
(1.54 ± 0.20)
(2.04 ± 0.20)
(1.54 ± 0.35)
2.15
log10 K 4 = – (0.92 ± 0.18)(e)
[64TRI/CAL]
[71LAN/CUM]
[74KUL3]
[64TRI/CAL]
[71LAN/CUM]
[88BJE2]
(a) Accepted values corrected to the molal scale. The accepted values reported in Appendix A are expressed
on the molar or molal scales, depending on which units were used originally by the authors.
(b) Corrected to 25°C, using the accepted enthalpy values.
(c) Data for several ionic strengths, see Appendix A.
(d) Re-evaluated value, see Appendix A.
(e) ‘Semi-thermodynamic’ constant, the activity of thiocyanate ion (using γ ± of NaSCN) was taken into
account.
Figure V-50: Extrapolation to Im = 0 of the accepted formation constants reported for
the formation of NiSCN+ in perchlorate media.
2.4
log10 β1 + 4D
2.2
2.0
1.8
1.6
0.0
0.5
1.0
1.5
2.0
2.5
−1
I / mol·kg
3.0
3.5
4.0
V.7 Group 14 compounds and complexes
236
Figure V-51: Extrapolation to Im = 0 of the accepted formation constants reported for
the formation of Ni(SCN)2(aq) in perchlorate media.
3.4
3.2
log10 β2 + 6D
3.0
2.8
2.6
2.4
2.2
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
−1
I / mol·kg
Figure V-52: Extrapolation to Im = 0 of the accepted formation constants reported for
the formation of Ni(SCN)3− in perchlorate media.
3.6
log10 β3 + 6D
3.2
2.8
2.4
2.0
1.6
0.0
0.5
1.0
1.5
2.0
2.5
−1
I / mol·kg
3.0
3.5
V.7 Group 14 compounds and complexes
237
Five reliable reports are available for the reaction enthalpies of the formation of
Ni(SCN) 2x − x (x = 1 – 3) complexes. The reported values are collected in Table V-33.
Table V-33: Experimental reaction enthalpy values for the formation of Ni(II)thiocyanate complexes.
Method
Im
Medium
t (°C)
∆ r H m (kJ·mol–1)
∆ r H m (kJ·mol–1)
reported
retained for selection
Reference
Ni2+ + SCN– U NiSCN–
< 0.06
25
– (10.46 ± 0.40)
(H)ClO4
0.67
15 – 35
– 14.43
(KNO3)
0.2
5 – 35
cal
NiClO4 + KSCN
pol
pol
– (9.25 ± 2.00)(a)
[67NAN/TOR]
– (14.5 ± 2.0)
[68MAL/TUR]
–
– (14.3 ± 2.0)(b)
[71TUR/MAL]
– (8.3 ± 2.0)
[73HUT/HIG]
(b)
kin
(NaClO4)
1.05
25 – 45
–
cal
(NaClO4)
1.05
25
– (12.02 ± 0.15)
(12.0 ± 0.5)
– (29.5 ± 2.0)(b)
[71TUR/MAL]
[74KUL3]
Ni2+ + 2SCN– U Ni(SCN)2(aq)
pol
(KNO3)
0.2
5 – 35
–
cal
(NaClO4)
1.05
25
– (20.92 ± 1.15)
– (20.9 ± 1.2)
[74KUL3]
25
– (29.12 ± 5.15)
– (29.1 ± 5.2)
[74KUL3]
Ni2+ + 3SCN– U Ni(SCN)3−
cal
(NaClO4)
1.05
(a) Re-evaluated value, see Appendix A.
(b) Calculated from the temperature dependence of the corresponding formation constant, see Appendix A.
These data do not allow correct evaluation of the ionic strength dependence,
therefore it was assumed that reaction enthalpies are independent of the ionic strength.
Assigning higher weight to the most reliable calorimetric data published in [74KUL3],
and considering the above assumption, the weighted averages of the retained values are
as follows:
∆ r H mο ((V.116), q = 1, 298.15 K) = – (11.8 ± 5.0) kJ·mol–1,
∆ r H mο ((V.116), q = 2, 298.15 K) = – (21 ± 8) kJ·mol–1,
∆ r H mο ((V.116), q = 3, 298.15 K) = – (29 ± 10) kJ·mol–1,
From the above selected ∆ r H mο values, the following standard enthalpies of
formation can be derived:
∆ f H mο (NiSCN+, 298.15 K) = (9.59 ± 6.46) kJ·mol–1,
∆ f H mο (Ni(SCN)2, aq, 298.15 K) = (76.79 ± 11.35) kJ·mol–1,
∆ f H mο ( Ni(SCN)3− , 298.15 K) = (145.19 ± 15.65) kJ·mol–1.
The above selections yield:
S mο (NiSCN+, 298.15 K) = (7.5 ± 25.4) J·K–1·mol–1,
S mο (Ni(SCN)2, aq, 298.15 K) = (137.8 ± 46.5) J·K–1·mol–1,
238
V.7 Group 14 compounds and complexes
S mο ( Ni(SCN)3− , 298.15 K) = (261.5 ± 66.2) J·K–1·mol–1.
V.7.2 Silicon compounds and complexes
V.7.2.1 Solid nickel silicates
V.7.2.1.1
V.7.2.1.1.1
Nickel orthosilicate Ni2SiO4(cr)
Crystal structure and phase transitions
The only stable silicate in the system NiO – SiO2 is the orthosilicate 2NiO·SiO2
[54EIT], [66GME], [63PHI/HUT], [69LEV/ROB], [91KNA/KUB]. The existence of
the pure metasilicate NiSiO3 has not been confirmed by any X-ray investigation
[66GME]. However, an estimation of the thermodynamic data and the stability for the
nickel metasilicate (pyroxene) was performed by Campbell and Roeder [68CAM/ROE]
and by Navrotsky [71NAV]. In the present review, the thermodynamic data for the
NiSiO3 are not evaluated because the pure nickel pyroxene is unstable relative to silica
and the nickel olivine at a total pressure of one atmosphere [68CAM/ROE].
The 2NiO·SiO2, orthorhombic crystals, have the olivine structure, but at higher
temperature and pressure a spinel transformation can occur. This transformation was the
subject of numerous studies [62RIN], [65AKI/FUJ], [68NAV/KLE], [73NAV],
[73NAV2], [76NAV/KAS], [74YAG/MAR], [74MA] because of great geophysical interest. The approximate equilibrium pressure for the olivine – spinel transition in
Ni2SiO4 at 650°C was found to be 18 kb [62RIN]. This value is about 40% lower than
that determined by [74MA] who found a linear pressure – temperature dependence of
the transformation, expressed by the equation: p/bar = 23000 + 11.8 t /°C. The following thermodynamic data for the olivine (ol.) – spinel (sp.) transition in Ni2SiO4 were
obtained by Navrotsky [73NAV] by measuring the heats of solution of both polymorphs
in a molten oxide at 986 K: the enthalpy of the transition Ni2SiO4(ol.) → Ni2SiO4(sp.),
∆ trs H mο (986 K) = (5.9 ± 2.9) kJ·mol–1, the heat content increments, ( H m (986 K) –
H mο (298 K)), olivine, (107.7 ± 1.8) kJ·mol–1, spinel, (106.2 ± 0.8) kJ·mol−1 and the entropy of the olivine – spinel transition ∆ trs S mο = 12.6 to 14.6 J·K–1·mol–1. Although the
spinel polymorph of Ni2SiO4 also exists in metastable form at atmospheric pressure, the
thermodynamic data for this high pressure compound are not comprehensively treated
in the present review.
Ma [74MA] found that Ni2SiO4 olivine melts incongruently at high pressures
to NiO plus melt and that it is a stable phase until melting occurs. Since such melting
behaviour persists at a pressure as low as 4.9 kbar, he expected that similar melting behaviour also occurs at atmospheric pressure. The melting curve extrapolated to atmospheric pressure gave a melting temperature of 1848 K [74MA]. The phase diagram
NiO-SiO2, drawn from data obtained by quenching and direct observational techniques
shows, however, that the decomposition temperature of the nickel olivine is
V.7 Group 14 compounds and complexes
239
(1818 ± 5) K [63PHI/HUT], [69LEV/ROB]. The dissociation reaction occurring prior to
melting is
Ni2SiO4(cr) U 2NiO(cr) + SiO2(cris.)
(V.117)
where (cris.) means cristobalite.
The uncertainty of the melting and decomposition temperature of Ni2SiO4 olivine was minimised by refining the thermodynamic data concerning this silicate
[87NEI]. The decomposition temperature of Ni2SiO4 may therefore be given as
(1820 ± 5) K [87NEI]. This is in perfect agreement with Phillips et al. [63PHI/HUT].
V.7.2.1.1.2
Crystallography and mineralogy of nickel orthosilicate
Liebenbergite (Ni2SiO4) is an end-member mineral of olivines being a group of minerals
of the general formula X2SiO4, where X is a divalent metal cation (Mg, Fe, Mn, Ni, Ca,
and Co) [82BRO]. The pure nickel olivine (liebenbergite) does not occur naturally
[73WAA/CAL], [2001HEN/RED]. The natural occurrence of liebenbergite
(Ni, Mg)2SiO4 of an approximate empirical formula Ni1.5Mg0.5SiO4 has been reported
from Scotia Talc mine, Bon Accord, Barbeton distr., Transvaal, South Africa
[73WAA/CAL]. Liebenbergite was named after W. R. Liebenberg, director of the National Institute of Metallurgy of South Africa. The crystal system of liebenbergite is
orthorombic (space group Pbnm; Z = 4) [87BOS]. Cell data for the end-member
Ni2SiO4 are a = 4.727 Å, b = 10.120 Å, c = 5.911 Å, V = 284.98 Å3 [2001HEN/RED].
The X-ray powder diffraction pattern is listed on JCPDS-ICDD card No. 15 –388.
V.7.2.1.1.3
Heat capacity and standard entropy
Heat capacity measurements of the nickel silicate olivine were reported by [82WAT],
[65EGO/SMI], and [84ROB/HEM]. However, the heat capacities reported by Watanabe
in the temperature range 350 to 700 K [82WAT] are 4.6% greater than the heat capacities obtained by [84ROB/HEM] with the more accurate low-temperature calorimeter.
Therefore, the results of Watanabe are not incorporated in this review. Egorov and
Smirnova [65EGO/SMI] have determined the specific heat capacities of the nickel orthosilicate in the temperature range between 555 to 1570 K using a copper block calorimeter. The well-described, careful measurements allowed a determination of C pο,m to
within ± 0.5%. These high-temperature data were used in this review to extend the heat
capacity equation up to 1570 K. Robie et al. [84ROB/HEM] have measured the heat
capacities of Ni2SiO4-olivine between 5 and 387 K by cryogenic adiabatic-shield calorimetry and between 360 and 1000 K by differential scanning calorimetry. The standard
molar heat capacity at 298.15 K, determined from these low-temperature measurements,
was reported to be (123.20 ± 0.20) J·K–1·mol–1 [84ROB/HEM]. This value, the only one
experimentally determined by accurate calorimetric measurements, is selected by the
present review:
C pο,m (Ni2SiO4, olivine, 298.15 K) = (123.20 ± 0.50) J·K–1·mol–1.
V.7 Group 14 compounds and complexes
240
An assumption of a higher uncertainty for the accepted value of C pο,m (Ni2SiO4,
olivine, 298.15 K) was necessary to achieve the coincidence of the standard heat capacity value at 298.15 K obtained from the low temperature measurements [84ROB/HEM]
and from the high temperature thermal heat capacity function selected by this review
(see below).
According to [84ROB/HEM] between 300 and 1300 K the heat capacities can
be represented by the equation:
C pο,m (T) = (289.73 – 0.024015 T + 131045 T –2 – 2779.0 T –1/2) J·K–1·mol–1
to within ± 0.5%.
It should be mentioned that these authors carried out measurements only in the
temperature range between 5 and 1000 K. Therefore, the extrapolation using the above
equation up to 1300 K seems to be less reliable. In this review, we have used the calorimetric heat capacity measurements of [65EGO/SMI] and [84ROB/HEM] and fitted the
thermal heat capacity function in the temperature range 270 < (T/K) < 1570. The selected experimental data and the fitted curve are shown in Figure V-53. The coincidence
of both data sets is satisfactory. Thus, we can select the following thermal function of
the heat capacity for Ni2SiO4-olivine between 270 and 1570 K:
C pο,m (T) = (147.61612 + 0.0486860 T – 1.50143·10–5- T 2 – 3.34214·106 T –2) J·K–1·mol–1
with an uncertainty of ± 0.5%.
Figure V-53: The standard molar heat capacity of Ni2SiO4-olivine as a function of temperature. The data are taken from [84ROB/HEM] ( ) and [65EGO/SMI] ( ). The fitted curve (solid line) corresponds to the thermal heat capacity function selected in this
review.
190
180
p, m
–1 −1 −–1
1
C°pC,m0 / J·K
/ J·mol·mol
·K
170
160
150
140
130
120
110
0
200
400
600
800
T/K
1000
1200
1400
1600
V.7 Group 14 compounds and complexes
241
The standard molar entropy at 298.15 K for Ni2SiO4-olivine as determined by
calorimetric measurements by Robie et al. [84ROB/HEM] is:
S mο (Ni2SiO4, olivine, 298.15 K) = (128.10 ± 0.20) J·K–1·mol–1.
This value was also selected by the present review. The entropy value calculated from the temperature dependence of the Gibbs energy of the reaction of nickel
with silica and oxygen e.g., as estimated by [68CAM/ROE], [76MAH/PAN], was not
accepted by this review.
V.7.2.1.1.4
Enthalpy of formation
The enthalpy of formation of Ni2SiO4 olivine from the component oxides was measured
by solution calorimetry in a molten oxide solvent at (965 ± 2) K by Navrotsky
[71NAV].
The value of ∆ r H m ((V.118), 965 K) for the reaction:
2NiO(cr) + SiO2(q) U Ni2SiO4(ol.)
(V.118)
was reported to be – (13.9 ± 1.9) kJ·mol–1. Using the heat capacities of the components,
the standard molar enthalpy of this reaction at 298.15 K was calculated: ∆ r H mο ((V.118),
298.15 K) = – (6.3 ± 6.7) kJ·mol–1 [78KOT/MUL]. The main interest, for this review, is
the determination of the standard enthalpy of formation of Ni2SiO4-olivine from the
elements in their standard state i.e., ∆ f H mο for the reaction:
2Ni(cr) + Si(cr) + 2O2(g) U Ni2SiO4(cr)
(V.119)
The enthalpy of Reaction (V.118), together with the enthalpies of the formation of SiO2(q), where (q) means quartz, and NiO(cr) and the thermal functions of heat
capacity of all components can of course be used for the calculation of
∆ f H mο (298.15 K). Generally, the enthalpy of the reaction based solely on calorimetric
data is the least reliable quantity, and contributes the major part of the uncertainty in the
determination of the Gibbs energy of reaction. On the other hand, the standard entropy
and the high-temperature heat capacities as well as the Gibbs energy of a reaction do
provide more accurate data [87NEI]. Thus, using the third-law method, i.e., the equation:
∆ r H mο (298.15 K) = ∆ r Gmο (T) –
∫
T
298.15
∆ r C pο,m dT +
T ( ∆ r Smο (298.15 K) +
∫
T
298.15
(∆ r C pο,m / T ) dT),
(V.120)
the standard molar enthalpy of a reaction at 298.15 K can be calculated. Based on his
emf-measurements and the heat capacity values, among others those for the Ni2SiO4olivine determined by [84ROB/HEM], O´Neill [87NEI] used this method to calculate
∆ r H mο (298.15 K) for the reaction:
2Ni(cr) + SiO2(cr) + O2(g) U Ni2SiO4(cr)
(V.121)
242
V.7 Group 14 compounds and complexes
to be – 482.42 kJ·mol–1. Using this value it was possible to calculate the Gibbs energy
of Reaction (V.118) and to achieve very good agreement between the calculated and
experimentally observed decomposition temperature of Ni2SiO4-olivine, i.e.,
(1820 ± 5) K [87NEI].
Robie and Hemingway [95ROB/HEM] used their accurate heat capacity measurements, combined with results from molten salt calorimetry, thermal decomposition
of the Ni2SiO4-olivine into its constituent oxides, and equilibrium studies, both by CO
reduction and solid state electrochemical cell measurement for Reaction (V.121)
[87NEI], and calculated the following standard molar enthalpy of formation of Ni2SiO4olivine (liebenbergite): ∆ f H mο (298.15 K) = – (1396.5 ± 3.0) kJ·mol–1.
In the present review, the standard molar enthalpy of formation of
Ni2SiO4−olivine at 298.15 K was also determined by the third-law method using the
thermal heat capacity function of Ni2SiO4-olivine, the equation representing the Gibbs
energy of Reaction (V.121) in the temperature range 970 – 1770 K, both derived by this
review as described in Sections V.7.2.1.1.3 and V.7.2.1.1.5, as well as thermodynamic
data of elemental nickel selected by this review. Elemental silicon, oxygen and silica
(quartz) data are taken from the NEA-TDB auxiliary dataset. The value obtained and
selected is:
∆ f H mο (Ni2SiO4, olivine, 298.15 K) = – (1396.0 ± 5.0) kJ·mol–1
and is in an excellent agreement with the value recommended by [95ROB/HEM]. A
higher uncertainty than calculated was ascribed to ∆ f H mο (298.15 K) to overcome the
problem of inconsistency arising during calculation of the decomposition temperature of
Ni2SiO4-olivine into NiO and SiO2-cristobalite. Using the selected value of ∆ f H mο it is
possible to calculate the decomposition temperature to be (1820 ± 5) K. It should be
noted that our calculations of ∆ f H mο were carried out using the thermodynamic data for
SiO2-quartz and not for SiO2-cristobalite. However, no thermodynamic data of
SiO2(cristobalite) are available in the CODATA compilation [89COX/WAG] or in the
TDB auxiliary data appendices used till now. The reason for it is probably the fact, that
the thermodynamic data of SiO2(cristobalite) are less reliable. This issue was also discussed briefly by [87NEI]. Using the thermodynamic data of SiO2-cristobalite given by
[95ROB/HEM] and the same procedure for the calculation of ∆ f H mο (298.15 K) of
Ni2SiO4-olivine as described above for SiO2-quartz, we have calculated:
∆ f H mο (298.15 K) = – (1401.15 ± 3.40) kJ·mol–1 and the proper decomposition temperature of Ni2SiO4-olivine (1820 ± 5)°C. Different sets of values for low and high temperature SiO2(cristobalite) have been proposed by [98CHA]. The other value of the standard
molar enthalpy of formation of Ni2SiO4-olivine proposed in the literature
∆ f H mο (298.15 K) = – 1378.2 kJ·mol–1 [61LEB/LEV] was not used by this review (see
Appendix A).
V.7 Group 14 compounds and complexes
V.7.2.1.1.5
243
Gibbs energy of formation
The Gibbs energy of formation of Ni2SiO4-olivine from metallic nickel, oxygen and
silica according to Reaction (V.121) has been determined by equilibration of this silicate or metallic nickel and silica with a CO/CO2 atmosphere [61LEB/LEV],
[63BUR/ABB], by determination of the oxygen fugacity over the mixture of nickel oxide-silica and the gas atmosphere of known CO2/H2-ratio [68CAM/ROE], or electrochemically by use of cells with a solid electrolyte [64TAY/SCH], [75LEV/GOL],
[84ROG/BOR], [98ROG/KOZ], [2000ROG/KOZ], [87NEI] and [86JAC/KAL]. The
results of the paper [86JAC/KAL] were not used for further evaluations, because they
were presented only in graphical form. Thus, additional uncertainties in recovering the
values from the original report were expected. A comparison of all results obtained by
the determination of the Gibbs energy of formation of Ni2SiO4-olivine from metallic
nickel, oxygen and silica, reported in the literature, is shown on Figure V-54. Although,
different reactions were investigated, comparison was possible by recalculation, using
the original data reported in the papers under consideration, to obtain the Gibbs energy
of Reaction (V.121). Reaction (V.121) is the overall reaction in the solid electrolyte
cells used for the determination of ∆ r Gmο . Thus, it has been directly investigated in papers [64TAY/SCH], [75LEV/GOL], [84ROG/BOR], [98ROG/KOZ], [2000ROG/KOZ],
[87NEI]. The study of Campbell and Roeder [68CAM/ROE] also directly refers to the
equilibrium of Reaction (V.121). On the other hand, the studies of nickel orthosilicate
reported by Lebedev and Levitskii [61LEB/LEV] and by Burdese et al. [63BUR/ABB]
are concerned with the equilibrium of the reaction:
2Ni(cr) + SiO2(cr) + 2CO2(g) U Ni2SiO4(cr) + 2CO(g).
(V.122)
The calculations of the Gibbs energy of Reaction (V.121) based on the Gibbs
energy determinations of Reaction (V.122) requires the ∆ r Gmο of Reaction (V.123)
2CO(g) + O2(g) U 2CO2(g).
(V.123)
Using the ∆ r Gmο values for Reactions (V.122) and (V.123) reported by
[61LEB/LEV] and [63BUR/ABB], the Gibbs energies of Reaction (V.121) were calculated and are also shown on the diagram ∆ r Gmο versus T, cf. Figure V-54, together with
results of the solid electrolyte cell measurements.
A comparison of the Gibbs energy of Reaction (V.121) as obtained by different
authors, Figure V-54, shows very good agreement between all solid electrolyte emfstudies and the oxygen fugacity study of Campbell and Roeder [68CAM/ROE]. The
results of Lebedev and Levitskii [61LEB/LEV] and of Burdese et al. [63BUR/ABB] are
significantly different. Therefore, they were disregarded by the present review and were
not used for the non-linear curve fit. It should be mentioned that no significant difference between the value of the Gibbs energy of Reaction (V.121) was found when SiO2
cristobalite or quartz was used in solid electrolyte emf-studies reported by O´Neill
V.7 Group 14 compounds and complexes
244
[87NEI]. The equation describing the fitted curve shown in Figure V-54 represents the
Gibbs energy of Reaction (V.121) in the temperature range 970 – 1770 K:
∆ r Gm (T) = (– (522.89819 ± 8.94994) + (0.39674 ± 0.05714) T –
(0.02629 ± 0.00700) T lnT) kJ·mol–1.
As already discussed in Section V.7.2.1.1.4, this equation was used in the present review for the calculation of the molar enthalpy of formation of Ni2SiO4-olivine at
298.15 K. The standard entropy change corresponding to Reaction (V.119)
∆ r Smο (298.15 K) was calculated using S mο (Ni2SiO4, olivine, 298.15 K) and the standard
entropy of elemental nickel selected in this review. The other data i.e., the standard entropies of elemental silicon, oxygen and silica (quartz) at 298.15 K were taken from the
NEA TDB auxiliary dataset. The Gibbs energy of formation of Ni2SiO4-olivine at
298.15 K was selected to be:
∆ f Gmο (Ni2SiO4, olivine, 298.15 K) = – (1288.44 ± 5.00) kJ·mol–1.
The selected value is in good agreement with ∆ f Gmο = – (1289.0 ± 3.1) kJ·mol–1
reported by [84ROB/HEM].
Figure V-54: Standard Gibbs energy of Reaction (V.121) as a function of temperature.
The data are taken from [68CAM/ROE] , [64TAY/SCH] , [75LEV/GOL] ,
[84ROG/BOR] , [98ROG/KOZ] , [2000ROG/KOZ] , [87NEI] +, [87NEI]
and [63BUR/ABB] . The solid line refers to the nonSiO2(cr) , [61LEB/LEV]
linear curve fit used by this review. The data of [61LEB/LEV] and [63BUR/ABB] were
disregarded.
-160
-180
2Ni(cr) + SiO2(cr) + O2(g) U Ni2SiO4(cr)
0
−1 –1
∆∆rrG
G°
kJ·mol
// kJ·mol
m
m
-200
-220
-240
-260
-280
-300
-320
1000
1100
1200
1300
1400
T/K
1500
1600
1700
1800
V.7 Group 14 compounds and complexes
245
V.7.3 Germanium compounds and complexes
V.7.3.1 Solid nickel germanates
V.7.3.1.1
V.7.3.1.1.1
Nickel orthogermanate Ni2GeO4(cr)
Crystal structure of nickel orthogermanate
Ni – orthogermanate has a spinel structure and can easily be synthesised from the component oxides [66GME]. Germanates have long been regarded as suitable analogues for
silicates, based on crystal chemistry systematics and, therefore, used for investigations
to predict the strength of silicate spinel in the transition zone of the Earth
[2001LAW/ZHA]. Even at high temperatures and pressures Ni2GeO4-spinel is a stable
phase [66GME]. The cubic face-centred cell of the nickel orthogermanate, space group:
Fd3m, has dimensions a = 8.221 Å, Z = 8 (according to JCPDS-ICDD card No. 10 –
266).
V.7.3.1.1.2
Enthalpy of formation from NiO(cr) and GeO2(cr)
Despite it being of geophysical interest, there are few thermodynamic data regarding
nickel orthogermanate reported in the literature so far. Moreover, the reported data are
incomplete. The direct determination of thermodynamic data for Ni2GeO4(cr) was performed by Navrotsky and Kleppa [68NAV/KLE] and by Navrotsky [71NAV]. These
authors used oxide-melt solution calorimetry and determined the enthalpy of formation
of Ni2GeO4-spinel from NiO(cr) and GeO2(quartz) at 965 K to be – (39.8 ± 1.7)
kJ·mol−1. The enthalpy of formation values of Ni2GeO4(cr) from the constituent oxides
given by other authors [78KOT/MUL] and [72KUB] were taken from the original work
of [68NAV/KLE] or [71NAV]. However, the value reported by [78KOT/MUL],
− (18.1 ± 1.7) kJ·mol–1, seems to be erroneously quoted. Lack of the thermal heat capacity function and the Gibbs energy of formation or entropy values for Ni2GeO4(cr) do not
allow accurate calculation of thermodynamic data for this germanate at 298.15 K. Thus,
the present review does not recommend any standard thermodynamic data for
Ni2GeO4(cr).
V.8 Group 13 compounds and complexes
V.8.1 Boron compounds and complexes
Intermetallics are not treated in this review, therefore, thermodynamic data regarding
borides are not included in this chapter.
V.8.1.1 Solid nickel borates
V.8.1.1.1
Crystal structure
Depending on the temperature, nickel borates can form crystals and glassy phases with
different degrees of polymerisation, which are very often difficult to identify from the
crystallographic point of view [66GME]. The numerous water-free and water-
246
V.8 Group 13 compounds and complexes
containing nickel borate phases were characterised only by their stoichiometric composition and not by X-ray diffraction methods [66GME]. The X-ray diffraction pattern is
known for Ni3B2O6 (JCPDS card No. 22-745, 26-1284 and 75-1809).
V.8.1.1.2
Thermodynamic data for NiO·2B2O3(s)
So far no direct determination of thermodynamic data for water-free solid nickel borates
has been reported in the literature. One of the reasons for this is probably the difficulty
in obtaining well-defined crystalline phases. The only available data, published in a
compilation of thermodynamic data for borate systems [74SLO/JON], are estimated
values of the standard enthalpy of formation and the entropy for NiO·2B2O3 at
298.15 K. The value of ∆ r H mο (NiO·2B2O3, s, 298.15 K) = – (62.8 ± 20.0) kJ·mol–1 reported by [74SLO/JON] was obtained by a linear relationship between the enthalpy of
reaction of the crystalline oxides and the ratio of cationic charge z to radius r. In view of
the uncertainties with respect to the ionic radii selected, for (i) the z·r–1 versus ∆ r H mο
correlation and for (ii) the Ni2+ ion, this value has not been accepted by this review.
According to Latimer´s method the standard entropies of predominantly ionic
compounds can be obtained additively from values determined for the cationic and anionic constituents [51LAT]. Using the revised contribution of Ni2+ [93KUB/ALC] and
the value for B4 O72 − derived by Richter and Vreuls [79RIC/VRE], the standard entropy
was estimated to be S mο (NiO·2B2O3, s, 298.15 K) = (129.9 ± 6.0) J·K–1·mol–1, which
only slightly differs from the value given by [74SLO/JON] (137.6 ± 6.0). Generally this
method is quite reliable, although the uncertainty would have to be increased to the 2σ
level (± 12.0). The problem, however, is that no well defined solid compound is known
to which this entropy could be assigned. Thus it was not accepted by this review.
V.8.1.1.3
Solubility of Ni(BO2)2·4H2O(s)
Shchigol [61SHC] studied the solubility of nickel orthoborate, Ni(BO2)2·4H2O(s). An
average value of the solubility product of Ni(BO2)2·4H2O K sο,0 = 2.16×10–9 mol3·dm–9 at
22°C was obtained. The recalculation (see Appendix A) resulted in
log10 K sο,0 (Ni(BO2)2·4H2O, s) = – (8.80 ± 0.19). The nickel orthoborate used in the study
[61SHC] was not structurally characterised by X-ray diffraction. It was synthesised by
reacting nickel hydroxide or carbonate with orthoboric acid and only analytically investigated with respect to its stoichiometric composition. Reliable thermodynamic quantities can be determined from solubility measurements only when the solid phase is unambiguously characterised with respect to its structure. The results of solubility measurements of the nickel orthoborate given by [61SHC] do not fulfil this criterion, therefore, no thermodynamic data for Ni(BO2)2·4H2O(s) were selected in this review.
No evidence of Ni(BO2)2·4H2O formation was found in a solubility study on
the system NiCl2-ZnCl2-H3BO3-H2O [89BAL].
V.8 Group 13 compounds and complexes
247
V.8.1.2 Aqueous nickel borato complexes
The existence of at least one soluble nickel borate complex was postulated by Shchigol
[61SHC]. Nickel orthoborate Ni(BO2)2·4H2O(s) dissolves better in orthoboric acid solutions than in pure water. The formation of the nickel borate complex in solution was
believed to occur according to the reaction:
Ni(BO2)2·4H2O(s) + H3BO3(aq) U HNi(BO2)3(aq) + 5 H2O(l)
(V.124)
The formula assigned to the proposed complex species was Ni(BO 2 )3− .
The dissociation constant of the aqueous nickel borate complex, according to
Reaction (V.125), was calculated by Shchigol [61SHC], b 3,1 (V.125) = 2.82×10–9
mol3·dm–9.
Ni(BO 2 )3− U Ni2+ + 3 BO −2
(V.125)
An inspection of the experimental data revealed that they were misinterpreted,
although Equation (V.124) may be correct, see Appendix A. Consequently the reported
value of the dissociation constant was not selected by the present reviewers.
248
V.8 Group 13 compounds and complexes
Chapter VI
VI Discussion of auxiliary data
selection
VI.1 Group 15 auxiliary species
VI.1.1 Additional consistent auxiliary data for As compounds
As was the case during preparation of an earlier TDB volume [92GRE/FUG], auxiliary
data for arsenic compounds are not being critically assessed at this time, and several self
consistent values from the U.S. National Bureau of Standards compilation have been
accepted as an interim measure. The values of ( H m (T) – H mο (298 K)) for As(cr) are
from Herrick and Feber [68HER/FEB], although it is recognised that the value of
S mο (As, cr, 298.15 K) from the same paper is likely a better value than the TDB value,
and differs from it by 0.6 J·K–1·mol–1. The high-temperature thermodynamic quantities
for As4O6(g) are based on the work of Behrens and Rosenblatt, but are adjusted to account for the differences in the values of S mο (As4O6, cubic, 298.15 K). The consistent
values: ∆ f H mο = (As4O6, g, 298.15 K) = – (1196.25 ± 16.00) kJ·mol–1 and S mο = (As4O6,
g, 298.15 K) = (408.6 ± 6.0) J·K–1·mol–1, and the enthalpy and entropy differences of
Behrens and Rosenblatt [72BEH/ROS] are used in the present review. The similar
measurements of Murray et al. [74MUR/POT] on the equilibrium between gas-phase
As4O6 and the monoclinic solid are less well documented, and have not been taken into
account here. The main contributions to uncertainties in the values for the values for
As4O6(g) are from the uncertainties in the corresponding values for As4O6(cubic).
VI.2 Other auxiliary species
VI.2.1 Auxiliary data for KCl(cr), KBr(cr) and KI(cr)
The enthalpy of formation values for these solids in Glushko et al. [82GLU/GUR] are
good, well-evaluated values that are very close to CODATA consistent. The value for
∆ f H mο (KCl, cr, 298.15 K) used in the first book in the CODATA series [87GAR/PAR]
(but not included in the CODATA selected data [89COX/WAG]) is only 0.02 kJ·mol–1
more positive than the value from [82GLU/GUR].
249
VI Discussion of auxiliary data selection
250
Detailed assessments of the aqueous enthalpy of solution values at 298.15 K
for KCl(cr), KBr(cr) and KI(cr) were given in [82GLU/GUR]. In the present review, to
obtain TDB-consistent values for the enthalpies of formation, these enthalpy of solution
values are combined with the TDB enthalpy of formation values for K+ and Cl–, Br– and
I– for 298.15 K.
Table VI-1: Enthalpy of formation values for KCl(cr), KBr(cr), and KI(cr) at 298.15 K
solid
∆ sln H mο / kJ·mol–1
[82GLU/GUR]
∆ f H mο (KX, cr) / kJ·mol–1
selected
(17.241 ± 0.018)
– (436.461 ± 0.129)
KBr(cr)
(19.780 ± 0.080)
– (393.330 ± 0.188)
KI(cr)
(20.230 ± 0.100)
– (329.150 ± 0.137)
KCl(cr)
The selected values of the enthalpies of formation differ by ≤ 0.15 kJ·mol–1
from those tabulated by Glushko et al. [82GLU/GUR]. A TDB reassessment of the enthalpies of solution is not attempted, as it would involve analysis of a very large body of
experimental data, and it is unlikely that the final result would be more accurate than the
values from [82GLU/GUR]. It is possible that the enthalpy of solution values selected
by Glushko et al. [82GLU/GUR] for these solids influenced the CODATA selections
for the enthalpies of formation of the aqueous ionic species – the discussion in the
CODATA volume [89COX/WAG] appears to lack the details that would be necessary
to check this point. However, any extra uncertainties and inconsistencies from this
source are expected to be small.
Part IV
Appendices
251
Appendix A Equation Section 1
This appendix comprises discussions relating to a number of key publications, which
contain experimental information cited in this review. These discussions are fundamental in explaining the accuracy of the data concerned and the interpretation of the experiments, but they are too lengthy or are related to too many different Sections to be
included in the main text. The notation used in this appendix is consistent with that used
in the present book, and not necessarily consistent with that used in the publication under discussion.
[1883THO]
Thomsen measured the heats of solution of NiCl2(cr) and NiCl2·6H2O(cr) in water. The
final solution had a water to salt ratio of 400:1 (0.139 m). The reported heat of solution
of the hexahydrate was 1.16 kcal·mol–1 (4.85 kJ·mol–1) at 291.65 K. In the experiment,
dilution was to 400 moles of H2O per mole of salt. Archer [99ARC] indicated that the
NBS tables [82WAG/EVA] corrected the heat of solution value to I = 0 and 298.15 K,
and obtained 0.138 kJ·mol–1 using – 188 J·K–1·mol–1 for the heat capacity of reaction,
and – 3.502 kJ·mol–1 for the integral heat of dilution. Based on the enthalpy of solution
data of Manin and Korolev [95MAN/KOR] and the enthalpy of dilution (see Section
V.4.1.3.2, Reference [59HAM] and the Appendix A discussion of [72AUF/CAR]), the
enthalpy of dilution to I = 0 of the 1:400 solution of nickel chloride solution is here estimated as – (3.0 ± 1.0) kJ·mol–1. The total uncertainty from the heat of solution measurements themselves and the heat capacity correction is estimated as 1.5 kJ·mol–1, and
therefore the heat of solution of NiCl2·6H2O to infinite dilution in water at 298.15 K can
be estimated as (0.64 ± 1.80) kJ·mol–1. This leads to ∆ f H mο (NiCl2·6H2O, cr) =
− (2103.33 ± 2.01) kJ·mol–1.
The reported heat of solution of the anhydrous salt was – 19.17 kcal·mol–1
(− 80.21 kJ·mol–1), again at 291.65 K and dilution to 400 moles H2O per mole salt. If it
is assumed that the heat of dilution is reasonably independent of the change in temperature, and that the heat capacity changes for the hydration reaction are not strong functions of temperature, the heat of hydration of NiCl2 to NiCl2·6H2O at 298.15 K can be
estimated as – 85.06 kJ·mol–1. Thomsen and other early experimentalists found that
NiCl2(cr) dissolved only with difficulty [66GME]. This might cause the reliability of the
direct enthalpy of solution measurements of NiCl2(cr) to be cast in some in doubt. However, the heat of solution of the anhydrous solid is not inconsistent with the results of
Manin and Korolev [95MAN/KOR].
253
254
A. Discussion of selected references
Based on the NEA TDB auxiliary value for ∆ f H mο (H2O, l, 298.15 K), and the
selected enthalpy of formation of NiCl2(cr) (– (304.90 ± 0.11) kJ·mol–1), Thomsen’s
values would lead to an enthalpy of hydration of – (2104.9 ± 1.8) kJ·mol–1, where the
uncertainty (estimated here) is slightly lower than if the two heat of solution measurements were independent, because it is probable there were compensating errors.
[1895FRA]
In this early paper, results are reported for the conductance of nickel sulphate solutions
for six concentrations between 9.8 × 10–4 M and 3.1 × 10–2 M. The work appears to
have been done carefully. Re-analysis is problematic because only two of the solutions
were < 0.0025 M in nickel sulphate [79LEE/WHE]. Reanalysis using the solutions at
the three lowest concentrations led to K1 = (221 ± 13). The statistical deviation is undoubtedly smaller than the true uncertainty because of the limited data and the difficulties in properly calibrating the apparatus as discussed by Robinson and Stokes
[59ROB/STO]. In the present review, we estimate the uncertainty in K1 at 25°C to be
± 60.
[1899FUN]
Solubility measurements are reported for Ni(NO3)2·6H2O(cr) (– 21°C to 41°C) and
Ni(NO3)2·3H2O(cr) (58°C to 95°C) in water. The equilibrium mixtures below 0°C may
also have contained the nonahydrate. The analyses on the solids indicated that good
crystals of a tetrahydrate salt could not be obtained.
[01MEU]
Meusser measured the solubility of several hydrated forms of Ni(IO3)2, and of anhydrous Ni(IO3)2, between 0 and 90°C. Near 25°C, the least soluble compound was identified by the author as the β dihydrate. Later authors [73NAS/SHI], [97PRA/LAN] have
concluded that the α-dihydrate of Meusser was actually an ill-defined amorphous solid,
and Meusser’s β form is what is now normally referred to as the dihydrate,
Ni(IO3)2·2H2O. The solubility of this solid was measured at 8, 18, 50, 75 and 80°C.
Equilibration times range from 24 h at the lowest temperature to 10 minutes at 80°C.
The reported experimental (percent weight as Ni(IO3)2) values were converted
to molalities, and the apparent solubility products were corrected to I = 0. These were
plotted against (T/K)–1 to obtain values for log10 K sο,0 at 298.15 K and an average value
of ∆ r H mο (A.1) over the temperature range of the measurements.
Ni(IO3)2·2H2O U Ni2+ + 2 IO3− + 2H2O(l)
ο
s ,0
(A.1)
If only the values to 50°C are included, log10 K ((A.1), Ni(IO3)2·2H2O,
298.15 K) = – (5.17 ± 0.11) and ∆ r H mο (A.1) = (21.4 ± 6.1) kJ·mol–1, whereas, if the
points at 75 and 80°C are included, log10 K sο,0 ((A.1), Ni(IO3)2·2H2O, 298.15 K) =
− (5.17 ± 0.08) and ∆ r H mο (A.1) = (22.5 ± 2.2) kJ·mol–1.
A. Discussion of selected references
255
A similar analysis based on the solubility measurements (30°C to 90°C) for a
yellow anhydrous nickel iodate, probably the β–Ni(IO3)2 [73NAS/SHI], leads to
log10 K sο,0 ((A.1), Ni(IO3)2·2H2O, 298.15 K) = – (4.427 ± 0.015) and ∆ r H mο (A.1) =
− (7.31 ± 0.69) kJ·mol–1. This appears to be the origin of the values listed for Ni(IO3)2 in
the US NBS tables [82WAG/EVA]. In the present review, greater uncertainties (± 0.1 in
log10 K sο,0 and 4.0 in ∆ r H m ) are estimated because of the temperature extrapolation and
the difficulties in establishing the purity of the solid at pseudo-equilibrium.
The tetrahydrates are difficult to purify, and convert easily to the dihydrate in
the presence of liquid water [73NAS/SHI]. This agrees with observation that the measured solubilities in water are greater than those of the dihydrate at all temperatures
[01MEU]. However, the solubilities reported for the tetrahydrate are not used further in
the present review.
[06THO]
See the entry for [1883THO].
[07WEI]
The solubility of crystalline (millerite) as well as precipitated NiS in pure water at
291.15 K was determined by means of conductivity measurements. The solubility of
millerite was found to be 1.629 × 10–5 mol·dm–3 whereas that of a precipitated product
amounts to 3.987 × 10–5 mol·dm–3. No experimental details are given.
[09BRU/ZAW]
Based on the enthalpy of formation of NiS(s) given by Thomsen [1883THO]
( ∆ f H mο (NiS, cr, 298.15 K) = – 72.76 kJ·mol–1) the solubility product was calculated,
leading to log10 K sο,0 = – 24.15 at T = 298.15 K. The entropy of formation for NiS(s) was
assumed to be zero, which is likewise confirmed by the thermodynamic data of the present assessment. The authors used values for the second dissociation constant of
H2S(aq),
H2S(aq) U 2H+ + S2–
ο
log10 K 2ο = – 23.26 and the standard potential of Ni2+, ENi/Ni
2+ = 0.219 V, which are no
longer considered reliable. Employing the currently accepted values, log10 K 2ο = – 25.99
ο
[92GRE/FUG] and ENi/Ni
2+ = 0.2372 V (this assessment), the recalculated solubility
product amounts to log10 K sο,0 = – 27.5.
[10BOR]
The phase relations in the system Ni – S in the composition range from 0 to 31 wt%
sulphur were investigated by thermal (cooling curves) and microscopic techniques. The
eutectic between metallic nickel and Ni3S2(cr) was found to lie at 918 K and 21.5
256
A. Discussion of selected references
wt% S. Moreover, it was reported that the high-temperature modification of Ni3S2(cr)
melts incongruently at a temperature around 1083 K.
[11AGE/VAL]
Seemingly the solubility of NiCO3 was studied by Ageno and Valla for the first time,
however, in their preliminary communication the authors described no attempt to characterise the chemical and physical state of the samples investigated. Regardless of this
uncertainty the result was ascribed to anhydrous NiCO3 and accepted by thermodynamic compilations for calculation of ∆ f Gmο (NiCO3, cr, 298.15 K) [35KEL/AND],
[52LAT].
Ageno and Valla equilibrated their solid phase with water under a carbon dioxide containing atmosphere in a closed vessel at 25°C. The Ni(II) concentration and the
partial pressure of CO2 were determined experimentally. Column 2 of their table on
carbonate solubilities is labelled incorrectly. Under the conditions prevailing in their
study an elementary stoichiometric consideration predicts:
2[Ni2+] ≈ [ HCO3− ].
Consequently it has been assumed that the numerical values listed in the solubility table under NiCO3/2 have to be ascribed to 2 [Ni(II)] instead.
[14THI/GES]
Nickel sulphide was prepared by adding Na2S or (NH4)2S to aqueous solutions of nickel
chloride (typically 0.04 mol·dm–3) at ambient (around 298 K) and elevated temperatures, respectively. In addition, NiS was precipitated by reaction of H2S with NiCl2 solutions containing sodium acetate (typically 0.1 mol·dm–3) at ambient as well as elevated
temperatures. In some cases the precipitated product was aged in the mother liquor for
10 weeks.
The solubility of NiS at room temperature was investigated by treating both
freshly prepared and aged solid phases with HCl solutions (2 mol·dm–3) saturated with
H2S. The amount of dissolved nickel(II) was determined by means of electrolysis. Three
different fractions of the solid sulphide could be discerned with respect to their solubility behaviour. Fraction I dissolved rapidly in hydrochloric acid at room temperature.
The solubility was estimated to be 0.1 mol·dm–3 at pH = 2. This high value indicates
that fraction I corresponds to an amorphous product. Fraction II showed a solubility in
H2S saturated 2 M HCl of approximately 1 mmol·dm–3 at room temperature. As the reaction time was in the range from 0.5 to 42 hours, equilibrium was certainly not attained. Finally, the solubility of fraction III in H2S saturated 2 M HCl was reported to be
approximately 0.02 mmol·dm–3 at room temperature. At temperatures around the boiling
point of the hydrochloric acid solution the solubility amounted to 0.22 mmol·dm–3 with
the reaction time ranging from 0.5 to 2 hours. Again, this time is not expected to be sufficient to attain equilibrium between the aqueous and the solid phase. According to
A. Discussion of selected references
257
Dönges [47DON] the hexagonal NiAs-type modification of nickel sulphide is obtained
when the precipitation is performed slowly by using H2S. Hence, it can be concluded
that fractions II and III correspond to crystalline NiS of different particle sizes depending on the conditions of precipitation and ageing. The different nickel(II) concentrations
found by dissolving these fractions may be caused by the variation of the dissolution
rate with the particle size. However, thermodynamic solubility constants cannot be derived from these experimental studies. As none of these nickel sulphide phases was
characterised by X-ray diffraction analysis the original notation I, II, and III-NiS has
been retained in this review.
[16DER/YNG]
Using a static method, water vapour pressures were determined as a function of temperature above mixtures of the tetrahydrate and hexahydrate (1.8°C to 36.25°C) and the
dihydrate and tetrahydrate (25.15°C to 79.06°C). The work established the existence of
the tetrahydrate as a separate phase, and the temperature of the hexahydrate/tetrahydrate
conversion at equilibrium with water vapour was reported as 36.25°C. The reported
water vapour pressure above the tetrahydrate/hexahydrate mixtures at 25°C was
(1.40 ± 0.05) kPa, where the uncertainty is estimated in the present review.
[18ALM]
Almqvist suspended freshly precipitated nickel hydroxide in water for 24h, filtered off
the residue, and determined the Ni(II) content of the clear solution. The value of the
Ni(OH)2 solubility thus obtained ([Ni(II)] = 1.37 × 10–4 mol·dm–3) implies that, owing
to the mass and charge balance of Equation (A.2), either the saturated solution is
strongly alkaline or the uncharged hydroxo complex Ni(OH)2 is rather stable. As neither
phenomenon has been observed, Almqvist’s solubility is too high, as clearly documented in Figure 5 of [97MAT/RAI]. Therefore, it has not been taken into account in
the calculation of a value of log10 K s ,0 . Probably the high solubility observed by Almqvist was due to colloidal Ni(OH)2 particles which passed through the filter.
[Ni(II)] = [Ni2+] + [NiOH+] + [Ni(OH)2]
[OH–] = 2 [Ni2+] + [NiOH+]
(A.2)
[21BER/CRU]
Berger and Crut investigated the equilibrium of the reaction,
NiCl2(cr)+H2(g) U Ni(cr) + 2 HCl(g)
(A.3)
using a static method. When the experimental data depicted in Figure V-16 and listed in
Table A-1 were compared with those of [53BUS/GIA], [53SAN], [54SHC/TOL] they
did not seem to be of adequate quality for use in a third law analysis.
A. Discussion of selected references
258
Table A-1: Equilibrium constants of Reaction (A.3) [21BER/CRU], [24CRU].
T/K
– RT ln(Kp / bar) / kJ·mol–1
log10(Kp / atm)
log10(Kp / bar)
583
– 2.214
– 2.2083
24.6475
613
– 1.627
– 1.6213
19.0269
668
– 0.784
– 0.7783
9.9532
718
– 0.0556
– 0.0499
0.6857
[23VIL/BEN]
The solubility of nickel sulphate was determined by the floating equilibrium method.
The following results were obtained at 25°C: 40.469 g NiSO4 per 100 g H2O and 40.720
g NiSO4 per 100 g H2O (floating equilibrium), 40.417 g NiSO4 per 100 g H2O (gravimetry).
[24CRU]
Crut described the method applied and the experiments carried out previously
[21BER/CRU] in more detail. The reported constants are the same as those given in
[21BER/CRU]. Only one chloride data set at 668 K is not identical. This was considered to be a misprint, as the other Kp values were exactly the same in [24CRU] as in
[21BER/CRU]. The reported measurements were primarily for reduction of the chloride. Only very limited measurements for the bromide system were carried out (at 718
and 848 K). Also, see the discussion of [24CRU2].
[24CRU2]
Crut [24CRU2] reported the heat of solution of anhydrous and hydrated NiBr2 as – 19.9
kcal·mol–1 (– 83.26 kJ·mol–1) and 0 kcal·mol–1, respectively. The paper also reports a
somewhat different derived value for the heat of hydration of the anhydrous salt as
− 19 kcal·mol–1 (– 79.50 kJ·mol–1). This value is consistent with the heats of solution of
anhydrous and hydrated NiBr2, – 18.9 kcal·mol–1 (– 79.08 kJ·mol–1) and 0 kcal·mol–1,
probably from the same experiments, reported in an earlier paper by the same author
[24CRU]. The extent of hydration of the hydrated salt is not reported, nor is the salt
purity, the final solution concentrations nor the measurement temperature. Because
nickel bromide prepared at high temperatures dissolved slowly in water at room temperature, a sample of anhydrous NiBr2(cr) was prepared in a manner to enhance its rate
of dissolution in water. The dissolution rate difference may reflect a structural difference that could have been affected the value determined in the heat of solution measurements.
Bichowsky and Rossini [36BIC/ROS], in the forerunner to the US NBS
Tables, estimated ∆ f H mο (NiBr2·3H2O, cr) using the value – 18.8 kcal·mol–1 (– 78.62
kJ·mol–1) for the enthalpy of hydration of NiBr2(cr) to NiBr2·3H2O(cr) from Crut’s 1923
A. Discussion of selected references
259
dissertation (unavailable to the current reviewers). This value corresponds approximately to the values reported in [24CRU]. The authors appear to have based their identification of the solid as NiBr2·3H2O on work reported in the thesis or elsewhere. There is
a similar, but smaller discrepancy in the values for the cobalt bromides from the same
two papers [24CRU], [24CRU2]. The analysis and value for ∆ f H mο (NiBr2·3H2O, cr)
from Bichowsky and Rossini [36BIC/ROS] is the origin of the value for the same solid
(at 298.15 K) in the 1952 version of the US NBS tables [52ROS/WAG] (and probably
of the slightly different value in the 1982 tables [82WAG/EVA]).
[24MOS/BEH]
The solubility of NiS at 293.15 K in H2S saturated 1 M H2SO4 was reported to be
1.04 × 10–4 mol·dm–3. After an equilibration time of 10 hours the dissolved amount of
nickel(II) was determined gravimetrically. The solid phase was precipitated at elevated
temperatures (333 – 343 K) by adding H2S to an aqueous solution of NiSO4·7H2O. No
characterisation of the solid product was performed. The reaction time of 10 hours
seems to be too short for equilibrium between the solid phase and the aqueous solution
to be attained.
[24TAN]
The solubilities of the crystalline nickel sulphate hydrates were determined to calculate
the respective temperature coefficients. For 25°C 40.55 and 40.26 g NiSO4 per 100 g
H2O were given.
[25BRI]
Britton titrated a nickel chloride solution with sodium hydroxide and employed a hydrogen electrode to follow the pH change. As only 1.66 moles of NaOH were needed to
complete the reaction it is quite clear that a basic chloride had at least been coprecipitated with the pure nickel hydroxide. Thus, it is doubtful whether Britton’s experimental data refer to Ni(OH)2 at all. In addition Jena and Prasad have previously reported that Britton miscalculated the solubility product [56JEN/PRA]. Re-evaluation of
the buffer region of Britton’s Figure 2 has shown that log10 K sο,0 = – (16.3 ± 0.3) is consistent with the experimental data, and has been used instead of the value originally
given (– 18.06) for the respective entry in Figure V-11 and Table V-6.
[25JEL/ZAK]
The partial pressure of sulphur in equilibrium with nickel sulphide was determined at
788.15 and 903 K. The ratio p(H2S)/p(H2) in the gas phase was measured by means of
chemical analysis, resulting in sulphur partial pressures by taking into account the wellknown equilibrium:
H2S(g) U H2(g) + ½ S2(g).
A. Discussion of selected references
260
The solid phase, NiS, was prepared by reaction of aqueous solutions of NiCl2
with (NH4)2S. The product was characterised by chemical analysis, but no X-ray
diffraction analysis was performed. As the results (log10[p(S2)/bar] = – 9.0 for the
proposed equilibrium between NiS and Ni7S6 at 788.15 K and log10[p(S2)/bar] = – 10.0
for the proposed equilibrium between Ni and β1−Ni3S2 at 903 K) do not coincide with
any other literature data, this investigation is not considered further in the present
assessment.
[25WIJ]
In the course of a study of nickel ammine complexes, in two experiments turbidity occurred, which was interpreted as an indication of formation of Ni(OH)2. When the solubility product ( log10 K s ,0 = − 13.8, 25°C) is extrapolated to I = 0 by means of the Davies
equation the following constants were obtained ( log10 K sο,0 = – 14.64, log10 *K sο,0 =
13.36) [62DAV]. These values are in line with others which refer to freshly precipitated
probably amorphous Ni(OH)2, see Table V-6 [38OKA], [54FEI/HAR].
[26JEL/ULO]
The reduction equilibrium of nickel(II) chloride was studied by a flow method. The
results were probably flawed by vaporisation of NiCl2, thus they were not subjected to a
third law analysis. The experimental data depicted in Figure V-16 are listed in
Table A-2.
Table A-2: Equilibrium constants of NiCl2(cr) + H2(g) U Ni(cr) + 2 HCl(g).
T/K
log10(Kp / atm)
log10(Kp / bar)
– RT ln(Kp / bar) / kJ·mol–1
573
– 2.574
– 2.5683
28.1739
613
– 1.398
– 1.3923
16.6060
723
– 0.136
– 0.1303
1.8033
[26JEL/ULO2]
The flow measurements for hydrogen reduction of NiBr2(cr) (683 K to 833 K) and
NiI2(cr) (783 K to 983 K) suffer from problems similar to those discussed for the measurements on NiCl2(cr) [26JEL/ULO]. Furthermore, above 700 K, NiI2 has been reported
to be subject to thermal decomposition [74WOR/COW]. The results of Jellinek and
Uloth are not used in the present assessment.
[27JON/WIL]
Powdered NiS was heated with S at 170°C for 30 to 40 h. The empirical formula of the
product was NiS2.18. The X-ray powder pattern was quite similar to the pattern for FeS2.
A. Discussion of selected references
261
A cubic lattice constant a0 = 5.74 Å was determined.
[28JEL/RUD]
The equilibrium of the reaction:
NiF2(cr) + H2(g) U Ni(cr) + 2HF(g)
(A.4)
was studied from 573 K to 773 K. The measured equilibrium constants varied with flow
rate. Therefore, although the derived value of ∆ f H mο (NiF2, cr, 298.15 K) is consistent
with those from other types of experiments [67RUD/DEV], the values of Jellinek and
Rudat are not used in the assessment of ∆ f H mο (NiF2, cr, 298.15 K) in the present review.
[28MUR]
Murata reviewed older measurements with – 0.597 V ≤ E°(NiSO4(aq) | Ni) ≤ – 0.183 V
and concluded that the state of the nickel electrode was responsible for this conspicuous
disagreement [1894NEU], [04EUL], [04MUT/FRA], [04SIE], [09SCH], [09SCH2],
[10SCH], [18SMI/LOB]. In his own work he used powdered nickel which was reduced
with hydrogen directly in the following cell
Ni | NiSO4 | KCl(sat.) | KCl(0.1 M), Hg2Cl2 | Hg.
Murata’s data have been recalculated with the standard potential of the
Hg2Cl2 | Hg electrode ((0.2681 ± 0.0026) V, [89COX/WAG]) and the chloride activity
of a 0.1 M KCl solution (aKCl = 0.077) [59ROB/STO]. The activity of Ni2+ in the solution was calculated using the SIT model. The corrected values are E°(Ni2+ | Ni) =
− (0.246 ± 0.003) V at 25°C and – (0.245 ± 0.005) V at 18°C. As the measurements
were carried out under hydrogen, which according to other authors interferes, and no
correction for the liquid junction was applied, Haring and Vanden Bosche’s
[29HAR/BOS] result has been preferred to Murata’s.
[29HAR/BOS]
Haring and Vanden Bosche reviewed the literature on the standard potential of
Ni2+ | Ni, discussed Murata’s work and included the following papers into the discussion [00KUS], [01PFA], [05COF/FOR], [08THO/SAG], [26GLA]. They pointed out
that in many of the numerous earlier studies, which led to results with E°(Ni2+ | Ni)
ranging from – 0.621 to – 0.178 V, appropriate corrections for activity coefficients and
liquid junction potentials were lacking. Moreover, the absence of interfering gases such
as oxygen or hydrogen and a constant temperature were not always ensured.
These authors carried out most careful measurements using the cell:
Ni | NiSO4(m) | Hg2SO4 | Hg
which involves no liquid junction. Finely divided, strain-free, electroplated nickel was
used and precautions were taken to exclude oxygen and hydrogen completely. The
A. Discussion of selected references
262
measured potentials were constant within 1-5 mV for 18 days. The cell behaved completely reversibly. The reversibility of the electrode was demonstrated by the Nernst
behaviour of the measured potentials with the activity of NiSO4, see also Figure V-2.
Moreover, after anodic and cathodic polarisation and short circuiting, which produced a
temporary disturbance, the equilibrium value was quickly regained.
The results are listed in Table A-3. It was necessary to recalculate the standard
potential (E°(Ni2+| Ni, 298.15 K) = − (0.231 ± 0.002) V [29HAR/BOS]), because the
SIT model leads to slightly different activity coefficients and the standard potential of
the Hg|Hg2SO4 electrode had to be changed to the value consistent with the NEA-TDB
auxiliary data. Using the function ln γ± = F (m (NiSO4)) given in the caption of Figure
A-5 (page 295 under the discussion of [59ROB/STO]) for the recalculation of E°(Ni2+|
Ni, 298.15 K) results in a value which agrees within 0.1 mV with the one derived using
the SIT. ∆ f Gmο (Ni2+) selected in this review has been based on the latter value.
An attempt by these authors to measure the standard potential of the nickel
electrode with the similar cell:
Ni | NiCl2(0.05 mol·kg–1) | Hg2Cl2 | Hg
was unsuccessful. Obviously nickel corroded in the chloride medium.
Table A-3: Electrode potential of the cell: Ni | NiSO4 (m) | Hg2SO4 | Hg [29HAR/BOS]
m (NiSO4)
E/ V
–1
(mol·kg )
E° / V
E° / V
E° / V
[29HAR/BOS]
SIT
[59ROB/STO]
0.0504
0.967
0.852
0.84848
0.84794
0.0504
0.967
0.851
0.84848
0.84794
0.0504
0.969
0.853
0.85048
0.84994
0.0506
0.968
0.852
0.84953
0.84900
0.0506
0.967
0.851
0.84853
0.84800
0.0507
0.971
0.855
0.85256
0.85204
0.0508
0.968
0.852
0.84959
0.84907
0.0509
0.967
0.851
0.84862
0.84811
0.0511
0.968
0.852
0.84968
0.84917
0.0511
0.967
0.851
0.84868
0.84817
0.0512
0.967
0.851
0.84871
0.84821
0.0515
0.970
0.854
0.85180
0.85131
(Continued on next page)
A. Discussion of selected references
263
Table A-3 (continued)
m (NiSO4)
E/ V
–1
(mol·kg )
E° / V
E° / V
E° / V
[29HAR/BOS]
SIT
[59ROB/STO]
0.0515
0.968
0.852
0.84980
0.84931
0.0516
0.968
0.852
0.84982
0.84934
0.0525
0.971
0.854
0.85308
0.85263
0.1017
0.958
0.852
0.84967
0.85007
0.1018
0.958
0.852
0.84969
0.85008
0.1020
0.958
0.852
0.84971
0.85011
0.1024
0.958
0.852
0.84977
0.85017
0.1052
0.958
0.852
0.85016
0.85056
0.1503
0.953
0.852
0.85023
0.85046
0.1504
0.954
0.853
0.85124
0.85147
0.1519
0.953
0.852
0.85038
0.85060
0.1543
0.953
0.852
0.85060
0.85080
0.1543
0.952
0.851
0.84960
0.84980
0.1617
0.954
0.852
0.85226
0.85241
[30MUL/LUB]
Müller and Luber determined a single value of the solubility of nickel carbonate hydrate, which they believed to be NiCO3·6H2O, at p(CO2) = 50 atm, without giving the
temperature of equilibration explicitly [30MUL/LUB]. When it is tentatively assumed
that the experiment was carried out at 25°C, Müller and Luber’s result agrees fairly well
with results from [11AGE/VAL] and [2001GAM/PRE], [2002WAL/PRE] indicating
that the same nickel carbonate phases were used in these studies.
[31KOL]
The solubility product of NiS was calculated by taking into account the solubility of
nickel(II) sulphide in 1 M H2SO4 (T = 293.15 K), determined by Moser and Behr
[24MOS/BEH], leading to log10 K sο,0 = – 26.96. The author used a value for the second
dissociation constant of H2S(aq),
H2S(aq) U 2H+ + S2–,
log10 K 2ο = – 21.97 which is no longer considered to be correct. Employing the currently
accepted value, log10 K 2ο = – 25.99 [92GRE/FUG], the solubility product can be recalculated, resulting in log10 K sο,0 = − 30.98.
264
A. Discussion of selected references
[32MAS]
Potentiometric measurements of the nickel(II) – cyanide system are described. The addition of Ni(CN)2, NiSO4 or Ni(NO3)2 to sodium cyanide solution resulted in the formation of a single nickel(II) containing species, Ni(CN) 24 − . Its formation constant at 25°C
and in 0.78 M NaCN solution was found to be log10 b 4 = (11.83 ± 0.17). The author
used nickel metal as the working electrode. Since the electrode reactions involving
nickel(II) are not reversible [50HUM/KOL], the reported formation constant is not reliable.
[32MON/DAV]
This paper provides a calculated value of 250 for K1 for nickel sulphate complexation
based on the earlier measurements of Franke [1895FRA] (at the two lowest molarities
for which results were reported, 9.768 × 104 and 19.53 × 104 M) at 25°C. The value of
Λ° was based on estimates of the molar ionic conductivities of sulphate (158
S·cm2·mol−1), and for Ni2+ (112.2 S·cm2·mol–1). There are no new experimental data.
[34BRI/OSS]
This paper reports the results of a dialysis study using aqueous solutions 1 M in ammonium sulphate and 0.05 M in various other metal sulphates. The sparse results for the
nickel system are interpreted in terms of formation of Ni 2 (SO 4 ) 44 − . There are insufficient details to derive useful complexation constants from the data. Also see
[41KIS/ACS].
[34COL2]
Colombier measured the potential of NiSO4(0.5 M) | Ni against a calomel reference
electrode at 20°C. When the electrolyte was appropriately degassed and kept air-free the
same equilibrium potentials were found for massive or powdered nickel, the latter having been electrolytically deposited or prepared by hydrogen reduction of NiO. As the
author presented no details about the experimental data, the reference potential selected,
how the liquid junction potential and the activity coefficients were accounted for, the
final result, E°(Ni2+ | Ni, 293.15 K) = – (0.227 ± 0.002) V, cannot be recalculated and
was not being considered any further.
[34SIE/SCH]
Sieverts and Schreiner [34SIE/SCH] carried out solubility measurements from – 27.8°C
to 119.8°C. They found Ni(NO3)2·9H2O below – 11°C and Ni(NO3)2·6H2O to 50°C.
They also reported that both the tetrahydrate (55 to 80°C) and dihydrate (85 to 120°C)
were stable phases in the Ni(NO3)2 – H2O system, and could be formed at equilibrium at
25°C in the Ni(NO3)2 – H2O – HNO3 system at high nitric acid mole fractions. The solution compositions at saturation do not differ markedly from those of Funk [1899FUN]
except at the lowest temperatures.
A. Discussion of selected references
265
[35KEE/CLA]
This paper reports the measurements of the specific heat capacity of nickel in the temperature range from 1.1 to 19.0 K. The hypothesis was advanced that the deviation from
Debye’s T 3 law is connected with the energy of electron conduction.
[36BIL/VOI]
The decomposition pressure of NiS2(cr) was measured as a function of temperature by
means of a tensimetric method. The solid phase, NiS2(cr), was prepared by solid state
reaction between NiS(cr) and sulphur. The composition of the product was checked by
chemical analysis (no X-ray diffraction analysis was performed).
Furthermore, the phase transformation of millerite into its high-temperature
modification was studied by an early DTA-like method. The transition temperature was
reported to be 669 K and the heat of transition was estimated to be in the range between
2.1 and 2.9 kJ·mol–1. Millerite was precipitated at room temperature by bubbling H2S
gas through an ammonia-containing aqueous solution of NiCl2. The solid compound
was characterised by chemical analysis.
[36BRO/WIL]
The mean heat capacities from – 80 to 120°C have been determined for nickel and a few
other metals. The errors were estimated to amount to ± 0.1%. A discussion of the theoretical basis for the equations describing the temperature dependence of the heat capacity is given together with some calculations of heat capacities using data which involve
no measurements of heat e.g., the expansion coefficient, compressibility etc.
[36JOB]
This is a spectrophotometric study on (among others) nickel(II) bromide complexes
formed in aqueous HBr solution. The author identified two bromo complexes, NiBr2(aq)
and NiBr42 − . The former species is reported to be dominant in 6 – 7 M, the latter above
10 M HBr. Since the ionic strength could not be maintained constant, the activity of
bromide, instead of its concentration, was taken into account in the mass action expressions. These “semi-thermodynamic” constants include the ratios g NiBrn / g NiBr( n−2) (where
n = 2 or 4), which are assumed to be nearly constant with increasing HBr concentrations. The reported constants are log10 b 2ο′ = – 3.24 and log10 b 4ο′ = – 8.12. The graphical data presented in [36JOB] were later re-evaluated in [72RET/HUM], and were
found consistent with the formation of NiBr+ and NiBr2(aq) complexes, if the activity of
water is also taken into account in the mass action expressions. In [72RET/HUM] the
recalculated value of log10 b 2ο′ is also reported ( log10 b 2ο′ = – 3.7). Since negatively
charged halogeno complexes of nickel(II) have not been identified in aqueous solutions
neither with fluoride nor with chloride, only the latter value is considered in this review.
266
A. Discussion of selected references
[36LOT/FEI]
Lotmar and Feitknecht investigated variations in ionic distances of bivalent basic metal
halogenides with layered structures of the CdI2 type. In this context they needed the unit
cell dimensions of nickel hydroxide as well as cobalt hydroxide.
[36TRA/SCH]
Trapeznikova et al. studied the heat capacity of NiCl2 at temperatures between 13 and
130 K. They used a glass calorimeter and a platinum thermometer heater. The discrepancy between their results and those of Busey and Giauque probably indicates that their
glass calorimeter and their thermometer did not attain thermal equilibrium
[52BUS/GIA].
[37BEL]
Measurements are reported for dissociation pressures of nickel chloride hexadeuterate
to tetradeuterate. The transition temperatures for dehydration of the hexahydrate and the
hexadeuterate at 1 atm are reported as 36.25 and 35.9°C, respectively. It was noted that
the solvent pressure ratio over the Ni–Cl solvates ( pD2 O :pH2 O ) rises with increasing temperature from 25 to 35°C. There are no experimental numbers reported for the hexahydrate/tetrahydrate equilibrium.
[37DOM]
The equilibrium of the reaction:
NiF2(cr) + H2O(g) U NiO(cr) + 2HF(g)
(A.5)
was studied from 773 K to 973 K. The measured equilibrium constants showed only a
small variation with flow (at higher flow rates). In the present review, values of
∆ r H mο ((A.5), 298.15 K) are calculated using the reported equilibrium constants, selected values of S mο and C p ,m (T) from Tables III-1 and III-3 for NiF2(cr) and NiO(cr),
and the tabulated values for H2O(g) and HF(g) in the CODATA compilation
[89COX/WAG]. Then, using ∆ f H mο values from Tables III-1 and IV-1 for NiO(cr),
H2O(g), and HF(g), an average value, ∆ f H mο (NiF2, cr, 298.15 K) = – (666.13 ± 0.64)
kJ·mol–1, is calculated. The drift in the calculated value depends only slightly on the
measurement temperature (< 1 kJ·mol–1 for the 200 K range). The value calculated in
the present review is 18 kJ·mol–1 more negative than the value recalculated by
Rudzitis et al. from the same data, though the values of ∆ r H m (298 K) are in good
agreement. Only part of the discrepancy can be explained as resulting from use of different auxiliary data (primarily the value for ∆ f H mο (HF, g, 298.15 K)).
[37SAN]
Sano reported the values for the vapour pressure of water over (initial) Ni(NO3)2·6H2O
from 300 K to 326 K using a Jackson spring manometer. The product of the dehydration
A. Discussion of selected references
267
reaction was assumed to be Ni(NO3)2·3H2O, and the plot of – log10 pH2 O against 1/T is
linear, but no estimate of the uncertainties is given. In this paper, the thermodynamic
parameters for the dehydration reaction were calculated incorrectly from the data (the
equilibrium constant depends on pH3 2 O , not pH2 O as assumed by the author). No evidence was presented that the dehydration product was the trihydrate rather than the tetrahydrate, the dihydrate or a non-equilibrium mixture of hydrated salts.
[37SAN2]
See [53SAN].
[38OKA]
Three titrations of pure as well as NaCl- or Na2SO4-containing nickel nitrate solutions
with sodium hydroxide were carried out at 25°C. From the pH measured at each point
taken on the titration curve (0.07 ≤ n(NaOH) / n(Ni(NO3)2) ≤ 1.62) the solubility product was calculated and extrapolated to zero ionic strength. A mean value of log10 K sο,0 =
– 14.50 was obtained and has been taken as a basis for the value listed in Figure V-11
and Table V-6, respectively.
[38SMU]
Smurov precipitated nickel carbonate by mixing aqueous sodium carbonate and nickel
chloride solutions. The solid phase obtained was washed free of chloride and used for
the solubility study, but no attempt was made to characterise its stoichiometry and/or its
crystal structure. The precipitate was equilibrated with water in a temperature range
between 278.15 and 353.15 K. For each temperature the partial pressure of carbon dioxide was varied from 0.0005 atm to pCO + pH O = 1 atm. Smurov’s data have been ex2
2
trapolated to zero ionic strength under the assumption that the solid phase had been hellyerite. As shown in Figure A-1 the solubility constants thus derived disagree not only
with the values of hellyerite but also with those of gaspéite. The temperature dependence is more pronounced than is usually observed with neutral carbonates. This probably indicates that a basic nickel carbonate was investigated instead. However, without
stoichiometric and structural evidence, no thermodynamic quantities can be assigned to
it.
A. Discussion of selected references
268
Figure A-1: Solubility of precipitated nickel carbonate. ■ Experimental data [38SMU];
solid curve: experimental data fitted to log10 *K p , s ,0 = A + B/T + C·lnT; dotted curve:
solubility constant of hellyerite, NiCO3·5.5H2O; dash curve: solubility constant of
gaspéite, NiCO3.
11.5
11.0
10.5
9.5
*
0
log10 K p,s,0
10.0
9.0
8.5
8.0
7.5
7.0
6.5
260
280
300
320
340
360
T/K
[38SYK/WIL]
The specific heat capacity – temperature dependence for four different types of nickel
have been determined and the effects of the method of manufacture, the heat treatment
and the chemical composition have been ascertained. The results, together with those
obtained by previous investigators were reviewed and an attempt was made to evaluate
the most probable C pο,m versus T curve for nickel in the temperature range 100 to
600°C.
[39ROH]
After an equilibration period of four days at 25°C the solubility of NiSO4·7H2O was
found to be 41.2 g NiSO4 per 100 g H2O.
The 6 values taken from [23VIL/BEN], [24TAN], [39ROH], and cited in the
respective Appendix A entries, were used to calculate mean and standard deviation (2σ)
of the NiSO4·7H2O solubility in water (see Section V.2.1.3.1).
A. Discussion of selected references
269
[39SCH/FOR]
The high-temperature form of nickel(II) sulphide, NiS, was treated with hydrogen gas at
three different temperatures, viz. 673, 773 and 873 K. The H2S(g) content in the gas
phase was measured at constant temperature as a function of the sulphur content of the
solid phases. Equilibrium compositions of the gas phase for univariant equilibria, involving NiS, Ni3S2 and Ni6S5 (according to Stølen et al. [94STO/FJE] the composition
of this phase is rather Ni7S6 (high-temperature modification) and Ni9S8 (lowtemperature modification), respectively), are given. No experimental details are reported.
[39SIM/KNA]
Results of study of the decomposition of nickel sulphate heptahydrate through nickel
sulphate hexahydrate and nickel sulphate tetrahydrate at 0.015 bar (11 torr) are presented. X-ray diffraction patterns are presented for some of the different solids. Detailed
measurement results were not tabulated, but the graphical presentation and the reported
thermodynamic quantities for the dehydration reactions are in reasonable agreement
with results reported in the later study of Kohler and Zäske [64KOH/ZAS].
[40BEL]
This paper contains measurements of the water vapour pressures above samples of
NiSO4·7H2O (at 20, 25 and 30°C) and NiSO4·7D2O (at 20 and 25°C). The values for the
NiSO4·7H2O are very similar to those obtained by Schumb [23SCH], Bonnell and Burridge [35BON/BUR] and [66STO/ARC]. The vapour pressures above the deuterate
were found to be about 80% of the vapour pressures above the ordinary hydrate. The
enthalpy of removal of D2O from NiSO4·7D2O was found to be approximately 11
kJ·mol–1 less favourable than dehydration of NiSO4·7H2O.
Equations for dissociation pressures of nickel chloride hexahydrate to tetrahydrate and the nickel chloride hexadeuterate to tetradeuterate are presented. The equation
for the deuterate reaction is based on the experimental vapour pressures reported in
[37BEL]. The equation for the hexahydrate/tetrahydrate equilibrium matches the experimental points from Derby and Yngve [16DER/YNG] so well that it raises the question of whether the reported equation is from a fit to the earlier data [16DER/YNG],
rather than based on new experiments. The transition temperature reported in [37BEL]
matches that in [16DER/YNG] exactly (within 0.01 K).
[41KIS/ACS]
This paper discusses problems with the dialysis method as used by Brintzinger and coworkers (cf. [34BRI/OSS]).
270
A. Discussion of selected references
[41STO/GIA]
Results were reported for careful heat capacity and magnetic susceptibility measurements carried out between 1 and 15 K on a sample of “NiSO4·7H2O”. The results indicated a maximum in the heat capacity for 1.8 K. In a later paper [66STO/ARC], the results were reinterpreted and combined with results for higher temperatures to determine
the third-law entropy of the solid at 298.15 K.
[42FRI/WEI]
Mixtures of metallic nickel and nickel oxide were treated with CO/CO2 gas mixtures at
temperatures between 1044 and 1289 K. The composition of the gas phase after equilibration with the solid phases was determined by chemical analysis, yielding equilibrium
constants for the reaction:
Ni(cr) + CO2(g) U NiO(cr) + CO(g)
as a function of temperature.
[43NAS]
Näsänen titrated an excess of potassium hydroxide drop-wise with NiCl2 solutions and
determined the equivalence point potentiometrically with a hydrogen electrode. The
solubility product was calculated from the minimum of the buffer capacity. A recalculation taking Näsänen’s experimental conditions into account has led to log10 K sο,0 =
− 15.09 and log10 *K sο,0 = 12.92, respectively. The latter value has been listed in Figure V-11 and Table V-6. It agrees reasonably well with log10 K sο,0 = – 15.21
( log10 *K sο,0 = 12.79) given in the original paper. These values are at best valid for amorphous Ni(OH)2 immediately after its precipitation, but in the course of titration some
Ni(OH)2–xClx may have formed as well and thus influenced the result.
[45KER]
Two sulphides have been identified as lying close to the Ni and Co end members of the
Co-, Ni-, Fe-disulphide system. In the light of this study the pyrite group forms an
isostructural series with pyrite, FeS2, vaesite NiS2 and cattierite CoS2 as end members,
while bravoite (Co, Ni, Fe)S2 is intermediate.
[47DON]
Amorphous NiS was prepared by reaction of nickel chloride solutions with Na2S and
(NH4)2S. If the nickel sulphide is precipitated by bubbling H2S(g) through a 1 M NiCl2
solution containing sodium acetate, crystalline NiS is obtained which shows the hexagonal NiAs-type structure according to X-ray diffraction analysis. After stirring the
solid phase in 1 M HCl for 10 minutes the nickel content of the aqueous phase was analysed spectrophotometrically. The solubility of NiS was reported to be in the range from
A. Discussion of selected references
271
0.006 to 0.142 mol·dm–3. As the reaction time of 10 minutes is by far too short to attain
equilibrium, no further evaluation of these data has been performed.
[47PEA]
Peacock investigated two specimens of heazlewoodite from Heazlewood, Tasmania, by
chemical and X-ray diffraction analyses. The empirical formula of one of the heazlewoodite samples was close to Ni3.04Fe0.02S2. The X-ray diffraction patterns of the natural
and the artificial Ni3S2 phases agreed very well with each other and also with those reported by Westgren [38WES]. The cell dimensions given by the latter author, however,
were incompatible with the respective d spacings. Consequently, the cell dimensions
given by Peacock and not those of Westgren were accepted for JCPDS-ICDD card
No.8-126.
[49GAY/GAR]
Nickel hydroxide was prepared by adding an equal volume of 0.02 M NaOH to a hot
solution of 0.0034 M Ni(NO3)2.The precipitate was washed free of alkali with water.
Two sets of samples were prepared for each concentration of HCl (0.0025 – 0.1 m) or
NaOH (0.0016 – 15 m). One was equilibrated for 5 to 7 days at 35°C and then for 7
days at 25°C, the other was equilibrated for 5 to 7 days at 25°C only. In this way the
authors intended to approach equilibrium from supersaturation as well as undersaturation. Inspection of Figure V-11 shows that in the acid range an aqueous suspension of
β–Ni(OH)2 saturated at higher temperatures is undersaturated at lower temperatures. As
a rule of thumb a temperature decrease of 10 K slows down chemical reactions, such as
the dissolution reaction according to Equation (V.40), by at least a factor of two. Long
ago Schindler pointed out that the method applied by Gayer and Garrett may not be effective with slowly dissolving metal hydroxides [63SCH]. Thus, it is not quite sure that
in their work the investigated solutions were indeed saturated. When it is assumed that
Gayer and Garrett's result is valid at 308.15 K it agrees very well with the experimental
data of Gamsjäger et al. [2002GAM/WAL].
Jena and Prasad [56JEN/PRA] as well as Mattigod et al. [97MAT/RAI] presume that there were systematic errors in pH measurements by Gayer and Garrett, but
this cannot be proven just by shifting the reported pH values until their own results are
confirmed. Moreover, Gayer and Garrett determined the concentration of nickel(II) employing the dimethylglyoxime method; in addition the data in alkaline solutions were
checked by radioactive tracer experiments. They calibrated the glass electrode for pH
measurements in a then conventional way. Thus, errors between 0.5 and 0.8 pH units as
suggested by Mattigod et al. are improbable. The solubility of β–Ni(OH)2 certainly depends on the particle size, but whereas it is quite certain that Mattigod et al. equilibrated
their solutions with the finest portion of their samples, no such information is available
of Jena and Prasad's or Gayer and Garrett's experiments.
272
A. Discussion of selected references
The re-evaluation of Gayer and Garrett's data resulted in log10 *K s ,0 = 10.78 (as
compared to 10.8 given) only when the second column of their Table II was considered
to contain the molality, mH+ , and not the activity, aH+ , as stated in the original paper.
The ionic strength was not controlled, instead activity coefficients of Harned
and Owen were used [58HAR/OWE]. The nickel(II) concentrations, over the pH range
studied, were sufficiently low to neglect the formation of Ni 4 (OH) 4+
4 . The authors reported the values, log10 *K s ,1,2 ≈ – 7 and log10 K s ,1,3 = – (4.2 ± 0.6), for the reactions:
Ni(OH)2 (cr) U Ni(OH)2 (aq)
(A.6)
Ni(OH)2(cr) + OH– U Ni(OH)3−
(A.7)
The authors mentioned, that the dissolution of nickel hydroxide in alkaline solutions up to 15 M NaOH was so slight that only estimates of its solubility could be
made. When the reported experimental data were re-evaluated it became evident that, at
best, only one constant, K s ,1,3 , can be determined. The solubility of Ni(OH)2(cr) increases with increasing NaOH molality, but the solubility curve flattens out at [OH–] >
1 m. This may be due to ion-ion interactions at the high ionic strengths, but there is no
evidence whatever for formation of Ni(OH) 24 − . Equation (A.7) can be fit reasonably to
the data at lower NaOH molalities. When the SIT model is applied:
log10 [Ni(II)]tot = log10 K s ,1,3 – ∆ε·[OH–]ini + log10[OH–]ini
The logarithmic solubility curve, as obtained by regression analysis, turns
sharply downwards at high hydroxide molalities. This demonstrates that the SIT approach is no longer valid, because the mass law requires an increasing solubility with an
increasing NaOH molality.
The regression analysis results in log10 K s ,1,3 = – (4.40 ± 0.12) and ∆ε =
(0.66 ± 0.17) kg·mol–1. Where ∆ε = ε(Na+, Ni(OH)3− ) – ε(Na+,OH–) and ε(Na+,OH–) =
0.04 kg·mol–1. The value of ε(Na+, Ni(OH)3− ) = (0.70 ± 0.12) kg·mol–1 is an unusually
high value that leads to the strange looking solubility curve of Figure V-7. This indicates that the present analysis of the sparse data is probably oversimplified, especially
for high base concentrations.
[50AKS/FIA]
Aksel’rud and Fialkov precipitated Ni(OH)2, which was not characterised any further,
and reacted it with nickel sulphate solutions at 18°C, determined the equilibrium pH
with hydrogen, quinhydrone and antimony electrodes, plotted y = log10[Ni2+] +
2log10KW + 2pH versus x = [Ni2+], and extrapolated graphically to zero concentration.
The authors used log10KW valid at 25°C instead of 18°C. In Figure V-11 and Table V-6
the similarly extrapolated but appropriately corrected value is given. It should be emphasised that by equilibrating Ni(OH)2 with NiSO4 solutions only minute amounts of
the solid phase dissolve and thus the solubility relates to its finest probably amorphous
fraction.
A. Discussion of selected references
273
[50GLE/EIN]
Glemser and Einerhand investigated β-NiOOH, γ-NiOOH and other nickel oxide hydroxides of higher oxidation state by X-ray powder diffractometry and electron micrography.
[50HUM/KOL]
Polarographic studies on the reduction of the Ni(CN) 24 − ion have been performed. The
results confirmed the irreversibility of the reduction, and indicated a minimum formation constant of about 1024. Polarograms of saturated solutions of Ni(CN)2 showed
waves due to the aquonickel(II) ion and the Ni(CN) 24 − complex, but no free cyanide
ions could be detected. This indicates that the solid is actually Ni[Ni(CN)4] (see also
[51LON]. The solubility product of Ni[Ni(CN)4] in 0.1 M NaCl solution was reported to
be log10 K s = – 8.77.
[50VAS/GRA]
Heat capacity measurements were reported for temperatures from 65 to 300 K for several hydrated transition-metal salts, including Ni(NO3)2·6H2O. Results were not listed
for each measurement, although the specific heat values are indicated as points on a
graph. “Smoothed” values are reported at 5 to 20 K intervals over the temperature
range. A rather broad second-order transition for Ni(NO3)2·6H2O was found between
140 K and 170 K, with a peak at 149 K. The authors estimated the entropy change between 65 K and 273 K to be 87.1 cal·K–1·mol–1 (364.4 J·K–1·mol–1). No magnetic transition was found in the temperature range of the transition. The authors stated that the
random error “should not exceed 0.001 cal·K–1·g–1” (1.22 J·K–1·mol–1). From the set of
smoothed specific heats, the value of the molar heat capacity of the solid at 298.15 K is
calculated to be (463.0 ± 2.0) J·K–1·mol–1, where the uncertainty has been estimated in
the present review. The temperature range of the measurements is not adequate for the
calculation of a value for S mο (Ni(NO3)2·6H2O, 298.15 K).
[51COU]
Coughlin measured the enthalpy of nickel chloride from 298.15 to 1336 K, see Figure
A-2. The NiCl2 samples for this investigation were provided by Busey and Giauque
[52BUS/GIA]. The adopted melting point was Tfus(NiCl2, cr) = 1303 K.
A. Discussion of selected references
274
Figure A-2: High temperature enthalpy of anhydrous NiCl2. Solid curve:
H m (T ) − H mο (298.15) = a · (T − 298.15) + b ·(T 2 − 298.152 ) + c·(T −1 − 298.15−1 ) , with
a = 7.37949·10–2, b = 6.11344·10–6, c = 512.775. : Experimental data from [51COU].
180
160
( Hm(T) − H°m(298.15) ) / kJ·mol
−1
140
120
100
80
60
40
20
0
-20
200
400
600
800
1000
1200
1400
T/K
[51HOO/WOY]
The equilibrium of the reaction:
NiCl2(cr) + 2 HF(g) U NiF2(cr) + 2 HCl(g)
(A.8)
was studied from 477 K to 830 K. The concentrations of HCl(g) and HF(g) were measured in a gas stream that had been brought to equilibrium over the solids. At the temperatures used, the measured equilibrium constants were independent of the flow rates
of the nitrogen carrier gas containing HCl(g) or HF(g). In the present review, values of
∆ r H mο ((A.8), 298.15 K) are calculated using the reported equilibrium constants, selected values of S mο and C pο,m (T) from Tables III-1 and III-3 for NiF2(cr) and NiCl2(cr),
and the tabulated values for HCl(g) and HF(g) in the CODATA compilation
[89COX/WAG]. Then, using ∆ f H mο values from Tables III-1 and IV-1 for NiCl2(cr),
HCl(g), and HF(g), an average value, ∆ f H mο (NiF2, cr, 298.15 K) = – (652.76 ± 0.93)
kJ·mol–1, is calculated. Calculated values of ∆ f H mο (NiF2, cr, 298.15 K) appear to be
systematically more negative (by 1–2 kJ·mol–1) when based on the higher temperature
measurements.
A. Discussion of selected references
275
[51LAT]
In this paper the original “Latimer” entropy contributions of metals and negative ions in
solid compounds are listed.
[51LON]
The exchange of cyanide ion between the Ni(CN) 24 − ion and the bulk solution was
complete within 30 seconds, which was too fast to be followed properly by the method
used. A study of the Ni2+ – Ni(CN) 24 − exchange was complicated by formation of the
precipitate (Ni(CN)2) in the mixture of nickel(II) and tetracyanonickelate ions. The results indicated two non-equivalent metal ions in the solid, which confirmed that its
structure is indeed Ni[Ni(CN)4]. The nickel(II) exchange between Ni(CN) 24 − and four
nickel(II) complexes indicated a direct bimolecular interaction.
[52BUS/GIA]
Busey and Giauque measured the heat capacity of nickel from 15 to 300 K. The entropy, heat content and free energy functions have been calculated. The authors used
99.98% nickel, in contrast to the numerous low temperature heat capacity studies
quoted in this paper where rather low purity Ni-metal was investigated. Calculations of
thermodynamic properties of nickel were extended to 800 K on the basis of available
data. However, the results of such extrapolation seem to be less reliable. The standard
molar entropy of Ni determined by Busey and Giauque was equal to 29.86 J·K–1·mol–1.
[52BUS/GIA2]
The heat capacity of anhydrous nickel chloride was studied between 15 to 300 K. The
anomalous region associated with the transition from the antiferromagnetic to the paramagnetic state was investigated in detail. A maximum of C pο,m (T) was found near 52 K.
This temperature corresponds with the temperature of the magnetic susceptibility
maximum within the accuracy of the magnetic measurements.
[52CAR/BON]
Carr and Bonilla reported an approximate value of the standard electrode potential of
nickel from the following cell:
Ni | NiSO4 | KCl(sat.) | KCl(sat.), Hg2Cl2 | Hg.
Their data have been recalculated with the standard potential of the
Hg2Cl2 | Hg electrode ((0.2681 ± 0.0026) V, [89COX/WAG]) and the chloride activity
of a saturated KCl solution a (KCl) = 2.829 [89LOB]. The activity of Ni2+ in the solution was calculated using the SIT model supported by ChemSage. The results are listed
in Table A-4.
A. Discussion of selected references
276
Table A-4: Standard electrode potential of Ni2+ | Ni at 25°C.
NiSO4/M
log10([NiSO4]/m)
log10 aNi2+
E (vs. std. CE)
E (vs. SHE)
E° (vs. SHE)
(mV)
(mV)
(mV)
(mol·dm–3)
0.0001
– 4.0014
– 4.10627
– 590.4
– 349.0
– 227.54
0.0010
– 2.999
– 3.15545
– 571.3
– 329.9
– 236.56
0.0100
– 1.9987
– 2.41766
– 548.6
– 307.2
– 235.69
0.1000
– 0.9989
– 1.81732
– 527.7
– 286.3
– 232.54
1.0000
0.0017
– 1.26378
– 499.2
– 257.8
– 220.42
The authors rejected the results in rows 1, 4, and 5 altogether. As Figure A-3
shows only the data in row 5 deviate considerably from the curve predicted with a value
of E°(Ni2+ | Ni, 298.15 K) = − (0.2331 ± 0.0082) V. This value agrees within the error
limits with the value in the NBS tables [82WAG/EVA] and the result of [29HAR/BOS]
as well as the compilation of [89BRA]. The reversibility of the cell was tested thoroughly and thus the value of [52CAR/BON] confirms the result of [29HAR/BOS] by an
independent method. When only the results in row 2 and 3 are accepted, as recommended by the authors, [29HAR/BOS] and [52CAR/BON] agree within 1 mV. The
uncertainty from the use of an uncorrected non-thermodynamic diffusion potential remains, and as a consequence the result of [29HAR/BOS] has been preferred.
Figure A-3: Plot of E (vs. SHE) versus log10([NiSO4]/m)
-260
Ni|NiSO4|KCl (sat.)|KCl (sat.), Hg2Cl2|Hg
E / mV (vs. SHE)
-280
-300
-320
o
-340
25 C
exp. data (vs. SHE) [52CAR/BON]
2+
E / mV = − 233.1 − (R T /2 F) ln (a (Ni ))
error limits: ± 8.2 mV
-360
-4
-3
-2
-1
−1
log10 (m (NiSO4) / mol·kg )
0
A. Discussion of selected references
277
[52CHA/COU]
Approximate hydrolysis constants for a number of cations, including Ni2+, were determined by potentiometric titrations at 30°C and 0.1 M (K)Cl ionic strength. The formation of NiOH+ as the sole hydrolytic species was assumed. However, the total metal
concentration used (about 0.01 M) is too high to neglect formation of the tetranuclear
hydroxo complex. The experimental data are not reported. Therefore, the results in this
paper have not been considered further in the present review.
[52GAY/WOO]
The pH values of solutions prepared from purified NiCl2 ([Ni2+]t = 0.0156 to 0.5 M)
were measured to obtain data on the magnitude of the hydrolysis reactions. Only the
formation constant, *b1,1 , of NiOH+:
Ni2+ + H2O U NiOH+ + H+,
was taken into account to explain the variation of the measured pH values. This paper
was a short communication, and thus, several experimental details (such as the method
of calculation and the source of the activity coefficients) were not reported. Furthermore, the [Ni2+]t used was far too high to neglect formation of polynuclear hydroxo
complexes. In addition, the effect of formation of chloro complexes was not addressed.
The experimental data are inadequate for proper re-evaluation; therefore the reported
constant was not considered in this review.
[52ROS]
An aqueous solution of nickel nitrate saturated with carbon dioxide was treated with
ammonium carbonate at 0°C. After one week at 5°C bluish green crystals were formed.
The analysis of this precipitate agreed within the error limits with the formula
NiCO3·6H2O. The X-ray powder diffraction pattern was taken and listed.
[52SUD]
The reduction equilibria of synthetic nickel sulphides with hydrogen gas were measured
with the help of a flow method. Ni7S6 was prepared by precipitation (reaction of H2S
with an ammonia containing nickel sulphate solution). The solid product was annealed
at 661 K in an evacuated silica tube for one hour. The composition was shown to be
close to Ni7S6 by gravimetric analysis; no X-ray diffraction analysis was carried out. As
the reduction experiments were mainly performed at temperatures below 675 K, the gas
phase, characterised by a certain ratio p(H2S)/p(H2), is in equilibrium with Ni3S2 and
Ni9S8 in accordance with [94STO/FJE]. Thus, the ratios p(H2S)/p(H2) given as a function of temperature should correspond to the equilibrium:
½Ni9S8 + H2(g) U
3
2
Ni3S2 + H2S(g).
In addition, the author prepared Ni3S2 by reduction of Ni7S6 with H2. Again,
the solid product was characterised by chemical analysis, but no X-ray diffraction
A. Discussion of selected references
278
analysis was performed. Ni3S2 was partly reduced with hydrogen gas at temperatures
ranging from 855 to 1000 K and the ratio p(H2S)/p(H2) in the gas phase was measured.
The results were erroneously referred to the equilibrium between heazlewoodite (lowtemperature form of Ni3S2) and metallic nickel. However, at temperatures above 834 K
the high-temperature form (β1-Ni3S2) is stable [80SHA/CHA], [91STO/GRO]. Moreover, the temperature for the eutectic between Ni, β1-Ni3S2 and the liquid phase is 905 K
[80SHA/CHA], so that above this temperature some of the material equilibrated with
the gas phase was molten. For these reasons the data of the experiments on Ni3S2 are
discarded.
[53BUS/GIA]
Busey and Giauque investigated the equilibrium reaction:
NiCl2(cr) + H2(g) U Ni(cr) + 2 HCl(g)
(A.9)
in order to provide evidence from the application of the third law of thermodynamics
that a ferromagnetic substance approaches zero entropy at 0 K. This plausible implication was experimentally confirmed for the first time by this study.
A third law analysis of the reported equilibrium constants was carried out employing the Gibbs energy functions of Ni(cr) and NiCl2(cr) selected in this review,
whereas the Gibbs energy functions of H2(g) and HCl(g) were taken from
[89COX/WAG]. The experimental data depicted in Figure V-18 are listed in Table A-5.
Table A-5: Equilibrium constants of Reaction (A.9)
T/K
Kp / atm
Kp / bar
661.67
676.65
693.00
645.33
676.18
707.11
738.02
707.02
646.19
692.14
692.12
707.42
722.73
737.89
630.18
0.2503
0.3949
0.6365
0.1488
0.3913
0.9568
2.135
0.954
0.1514
0.6223
0.6238
0.9565
1.423
2.095
0.0872
0.2536
0.4001
0.6449
0.1508
0.3965
0.9695
2.1633
0.9666
0.1534
0.6305
0.6321
0.9692
1.4419
2.1228
0.08836
– RT ln(Kp / bar) / kJ·mol–1
7.5476
5.1532
2.5272
10.1516
5.2011
0.1822
– 4.7349
0.1994
10.0720
2.6539
2.6400
0.1842
– 2.1989
– 4.6180
12.7133
A. Discussion of selected references
279
[53FRO2]
The formation constants of Ni(SCN) 2x − x (x = 1 – 3) complexes have been determined at
20°C, at pH = 3, in 1 M (Na)ClO4 solution, using Amberlite IR-105 cation exchanger.
The thiocyanate concentration ranged from 0 to 0.5 M, [Ni2+]tot varied between 1 and 5
mM. The equilibrium nickel(II) concentration was determined spectrophotometrically
using dithiooxalate, and that of the thiocyanate ion by a Volhard titration. The results
obtained were qualitatively confirmed by spectrophotometric measurements of the
aqueous nickel(II) – thiocyanate system. Considering the accepted enthalpy values for
the formation of Ni(SCN) 2x − x (x = 1 – 3) species, the reported formation constants were
corrected to 298.15 K: log10 b1 = (1.14 ± 0.06), log10 b 2 = (1.58 ± 0.06), log10 b 3 =
(1.72 ± 0.10).
[53GRI/SLU]
Ni3S4 was synthesised by precipitation of NiS from aqueous solution with H2S. The
suspended NiS was oxidised with oxygen to NiS2 which eventually disproportionated
into Ni3S4 and H2SO4.
[53SAN]
This paper seems to be a slightly updated English version of [37SAN2]. Sano studied
the reduction equilibria of cobalt(II) and nickel(II) chloride by hydrogen. He used the
method of observing the total pressure and the partial pressure of hydrogen by means of
a palladium membrane. Busey and Giauque tried to account for the difference of their
results and Sano's results by assuming that thermal equilibrium was not attained between Sano’s thermocouple and the reaction vessel. If Sano's temperatures were lower
by about 14 K, the equilibrium constants would agree quite well with those of
[53BUS/GIA]. In Figure V-18 the original experimental data have been depicted. They
are listed in Table A-6.
Table A-6: Equilibrium constants of Reaction (A.9) [53SAN]
T/K
log10(Kp / atm)
log10(Kp / bar)
– RT ln(Kp / bar) / kJ·mol–1
792
0.7345
0.7402
– 11.2236
760
0.4141
0.4198
– 6.1083
– 2.8853
738
0.1985
0.2042
709
– 0.1783
– 0.1726
2.3426
661
– 0.8016
– 0.7959
10.0716
[53SCH/POL]
Schwab et al. determined the solubility product of freshly precipitated Ni(OH)2 by
measuring the pH when one half of the Ni2+ originally present in NiCl2 solutions had
been precipitated by adding NaOH. The solid phase was not investigated any further.
280
A. Discussion of selected references
All experiments were carried out at “room temperature”. The authors reported
log10 K s ,0 = – 15.5. It is difficult to recalculate this value from the experimental data
given; in any case, two different values of the pH of half precipitation can be found in
[53SCH/POL]. If the first (pH = 7.75) is used, this results in log10 K s ,0 = – 15.5. Use of
the second value (pH = 8.4) would lead to log10 K s ,0 = – 14.2. Thus, it was concluded
that to extrapolate either of these results to I = 0 is not justified, and consequently neither was included in the final SIT calculations. In addition the hydrolysis of nickel (II)
was studied by potentiometric titrations of NiSO4, Ni(NO3)2 and NiCl2 solutions. The
initial concentration of nickel(II) was 2.44 mM for all titrations. The ionic strength was
not controlled. The experimental data were interpreted assuming that NiOH+ was the
only hydrolysis product formed. For all three solutions, irrespective of the anion present, log10 *b1,1 = – 9 was reported. This value was not considered further in this review.
[53SWA/TAT]
Swanson and Tatge recorded the pattern of a NiO sample obtained from Johnson, Matthey & Co., Ltd. The estimated purity of 99.99% was corroborated by spectroscopic
analysis, which showed only faint traces of Mg, Si, and Ca. After comparison with eight
other published patterns the cell dimensions given by [53SWA/TAT] were accepted for
JCPDS-ICDD card No.4-835.
[54DOB]
Nickel sulphate solutions were titrated with sodium hydroxide and vice versa at 75°C.
Each titration was carried out within 2 to 3 hours which appears to be a rather short period for equilibration even at 75°C. Ni(OH)2 and NiSO4·3Ni(OH)2 were precipitated in
dilute and concentrated solutions, respectively. Neither solid phase was investigated
with respect to its structure, e.g., by X-ray analyses. It is difficult to recalculate the results properly, because the ionic strengths are not given in Dobrokhotov’s Table 1. The
value for Ni(OH)2 listed by the author ( log10 K s ,0 = – 16.2, log10 *K s ,0 = 9.25 at 75°C)
agrees surprisingly well with our recent results. The recalculation of Dobrokhotov’s
digitalised data from Figures 3 and 5.1 with the SIT model resulted in log10 *K s ,0 = 9.48
at 75°C, whereby complex formation between Ni2+ and SO 24 − had been taken into account (see Figure V-11 and Table V-6).
[54FEI/HAR]
This is an abstract of a conference paper containing no experimental details at all. The
authors maintain that the solubility products of Ni(OH)2, freshly precipitated and aged
at 110°C, differ by 2.5 orders of magnitude. As mentioned in Section V.3.2.2.3, the
solid phases were not characterised with respect to their X-ray patterns.
[54ROS]
This important contribution deals with the investigation of phase relations in the system
Ni—S by measurement of the partial pressure of sulphur in equilibrium with various
A. Discussion of selected references
281
solid phases. The nickel sulphides were prepared by solid state reaction from the elements. The solid sulphides were equilibrated with hydrogen gas at temperatures between 673 and 1273 K. The partial pressure of sulphur is given by the ratio
p(H2)/p(H2S) which is recorded in situ by density measurements employing a buoyancy
balance. The compositions of the solid phases were determined by chemical analysis
before and after equilibration with hydrogen. From these equilibrium studies the stability ranges of the different sulphides as well as their Gibbs energies of formation were
derived. Gibbs energy functions are given for the following reactions:
Ni+H2S U ½Ni3S2+H2
–1
[∆ r Gm ]808K
673K = (– 75479.4 + 32.217 (T/K)) kJ·mol ,
2Ni3S2+H2S U Ni6S5+H2
–1
[∆ r Gm ]798K
673K = (– 12761.2 – 7.7404 (T/K)) kJ·mol ,
Ni6S5+H2S U 6Ni1–xS+H2
–1
[∆ r Gm ]833K
673K = (– 21840.5 + 14.477 (T/K)) kJ·mol ,
3
2
Ni+H2S U ½Ni3+xS2+H2
923K
[∆ r Gm ]808K
= (– 49998.8 – 13.514 (T/K)) kJ·mol–1,
Ni3+xS2+H2S U 3Ni1–xS+H2
–1
[∆ r Gm ]1083K
833K = (– 69454.4 + 70.291 (T/K)) kJ·mol ,
Ni1–xS+H2S U NiS2+H2
–1
[∆ r Gm ]1073K
673K = (– 19664.8 + 55.815 (T/K)) kJ·mol .
3
2
It is worth mentioning that according to Stølen et al. [94STO/FJE], the composition of the phase Ni6S5 is actually Ni7S6.
[54SHC/TOL]
Shchukarev et al. measured the equilibrium concentrations of HBr(g) and H2(g) at equilibrium over solid samples of nickel and nickel bromide for temperatures from 623 K to
928 K. The primary reaction was:
NiBr2(cr) +H2(g) U Ni(cr) + 2HBr(g).
The values of the logarithm (base 10) of the reported equilibrium constant values (converted to the 1 bar standard state) are listed in Table A-7, where:
2
 pHBr(g)
/ bar 
.
log10 K p = log10 
 pH (g) / bar 
 2

The authors indicated that equilibrium may not have been established successfully at the lowest temperatures (below 700 K), and that almost complete conversion of
the NiBr2 to Ni at 928 K made the value for the highest temperature somewhat suspect.
The reported equilibrium constant values were used here with the selected entropy (Section V.4.1.4.1) and equations and values for C pο,m (NiBr2, cr, T), Section V.4.1.4.2,
C pο,m (Ni, cr, T), Section V.1.2, and C pο,m (H2, g) and C pο,m (HBr, g) [89COX/WAG], to
calculate values for ∆ f H mο (NiBr2, cr). The authors suggested that there was evidence
for a transition for NiBr2(cr) just below 800 K (and an enthalpy of transition of 24
kJ·mol–1). Although two crystalline modifications of NiBr2 are known [34KET], the
evidence from these sparse measurements [54SHC/TOL] seems to be insufficient to
conclude that there is a transition with such a large transition enthalpy. The authors ap-
A. Discussion of selected references
282
parently did not confirm the crystalline form of the nickel bromide after measurements
at each temperature. The calculated values of ∆ f H mο for 298.15 K become progressively
more negative with increasing temperatures of the experiments on which the ∆ f H mο
values are based (and the results for nickel chloride, as obtained by the same authors,
also drift slightly with temperature). The average value of ∆ f H mο (NiBr2, cr) from the
measurements done between 700 and 900 K is – (215.02 ± 0.91) kJ·mol–1.
Table A-7: Enthalpy of formation values calculated from equilibrium gas compositions.
log10 K p
T/K
∆ f H mο (NiBr2, cr)
623
– 2.915
– 210.86
683
723
748
773
823
873
928
– 1.930
– 1.392
– 1.070
– 0.777
– 0.295
0.130
0.497
– 212.25
– 214.37
– 214.36
– 214.48
– 215.48
– 216.42
– 218.24
Shchukarev et al. investigated the decomposition reaction:
NiCl2 U Ni + Cl2
by studying the equilibrium:
NiCl2(cr) + H2(g) U Ni(cr) + 2 HCl(g)
(A.10)
and combining the latter with the well known formation equilibrium of hydrogen chloride:
H2 + Cl2 U 2 HCl.
A third law analysis of the reported equilibrium constants was carried out using
the same auxiliary Gibbs energy functions described under [53BUS/GIA]. The experimental data are depicted in Figure V-18 and listed in Table A-8.
Table A-8: Equilibrium constants of Reaction (A.10) [54SHC/TOL]
T/K
Kp / atm
Kp / bar
– RT ln(Kp / bar) / kJ·mol–1
573
0.013
0.01317
20.6272
623
693
723
743
773
823
0.078
0.59
1.23
2.00
4.00
9.8
0.07903
0.5978
1.2463
2.0265
4.0530
9.9299
13.1460
2.9643
– 1.3236
– 4.3633
– 8.9944
– 15.7079
A. Discussion of selected references
283
[55BRO/PRU]
This paper reports cryoscopic measurements for NiSO4 aqueous solutions (0.00 > t /°C
> – 0.21), and demonstrates the dependence of the calculated value of K1 on the assumptions in the complexation model. Based on the arguments of Brown and Prue, it is
clear that use of the SIT as is done in the TDB project imposes constraints on the data
analysis, and that these constraints should lead to a specific value for K1. However, recalculations show that the association constant is also strongly dependent on the
concentration at which the data set is truncated. For the purposes of the present review,
recalculations were done using the SIT, but only results for molalities ≤ 0.03 were used
in the selection of the value of log10 K1 . When data for higher concentrations were included, the calculated value for the association constant increased, and log10 K1 = 300
was obtained if the entire data set was used.
[55CAT/STO]
This adiabatic calorimetry study (12 K to 300 K) appears to be the only lowtemperature heat capacity study for NiF2(cr). From their large series of results the authors calculated S mο (NiF2, cr, 298.15 K) = (73.60 ± 0.17) J·K–1·mol–1. A λ transition was
found at (73.22 ± 0.05) K. The experimental values seem to have a non-systematic uncertainty of less 0.3 J·K–1·mol–1. In a later paper, the authors indicated that the heat capacity values above 270 K may have a slightly greater uncertainty than those for lower
temperatures [55STO/CAT]. Slightly lower values between 270 K and 300 K mesh better with the available high-temperature drop calorimetry results [70BIN/HEB].
[55KRA/WAR]
Krauss and Warncke used two different calorimetric methods to determine the specific
heat of high-purity nickel. The measurements in the temperature range between 180 and
600°C were carried out using a continuous calorimetric technique and between 500 and
1160°C by means of “reverse calorimetry”. The source of heat in the last method was a
Pt-rod of known temperature and heat capacity which was placed into the calorimeter
containing a nickel sample. The separation of the determined heat capacity into a latticevibration term, a magnetic term and a residual term was briefly discussed in this paper.
The reported data were used by this review for fitting of the thermal heat capacity function for nickel crystal.
[56AHR/ROS]
The stabilities of the fluoride complexes of nickel(II), copper(II) and zinc(II) have been
determined by a pH-metric method using a quinhydrone electrode at I = 1 M
Na(ClO4, F) and at 293.15 K. The pK value of HF2− was reported in an earlier paper (pK
= 2.93, [56AHR/LAR]). Under the conditions used ([F–]t ≤ 400 mM, [Ni2+]t ≤ 100 mM)
only the monocomplexes (MF+) were found to exist in appreciable concentration.
284
A. Discussion of selected references
[56CUT/KSA]
Potentiometric titrations of NiClO4 solutions with NaOH were performed at
(20 ± 0.1)°C. To determine the solubility constant, the pH of incipient precipitation was
measured for nickel concentrations of 0.0041 – 0.034 M at constant ionic strength, I =
1.0 M NaClO4. The value for log10 K sο,0 , obtained by extrapolating these experimental
data to I = 0 with the SIT model is listed in Table V-6. The hydrolysis constant was determined at nickel concentrations ranging from 0.0025 to 0.184 M. The ionic strength
was not controlled, and the activity coefficients were calculated using an extended Debye-Hückel expression. Such a general expression is normally assumed to be valid at
ionic strengths below 0.1 M. In the course of this study, the ionic strengths varied from
0.007 to 0.55 M, with most data pairs at I > 0.1 M. The formation of the single complex
NiOH+ was assumed, and a value of log10 *b1,1ο = – 8.94 was obtained from graphical
extrapolation to I = 0. The experimental data used in the calculations referred to very
low degrees of hydrolysis, thus, large unavoidable errors render the result dubious. For
this reason, the reported hydrolysis constant was not included in the detailed evaluation.
[56FIA/SHE]
This was a scoping study. Conductance measurements were carried out for dilute solutions of nickel sulphate. However, the data were incorrectly analysed in terms of a 1:2
nickel:sulphate complex, and the data are not available for re-analysis.
[56JEN/PRA]
Nickel hydroxide was prepared by adding an excess of sodium hydroxide solution to a
solution of nickel chloride. The precipitate was freed from alkali by washing with water
and was then dried in an air-oven at 98 – 100°C. This solid phase was equilibrated either with HCl (0.0189 – 0.1892 M) at 25°C or buffer solutions of (1:1) acetic acid (0.04
– 0.20 M) and sodium acetate at 29°C. The authors preferred the result of the buffer solutions, however a recalculation shows that the mean values of either series agree within
an experimental error of ± 0.1 log10K units and thus both have been recorded in Figure V-11 and Table V-6.
[56KEN]
This paper reports cryoscopic measurements for NiSO4 aqueous solutions in saturated
potassium perchlorate solutions (eutectic freezing point 272.99 K, freezing point depressions, ∆T, are between 0.0673 K and 0.181 K). Thus, the measurements were done
with less variation of ionic strength than for measurements done in water. The results of
this study are assessed in our discussion of a later paper [58KEN] by the same author.
[56LAN/MIE]
Lange and Miederer determined the relative apparent molar enthalpy, φL, of NiSO4 and
demonstrated that:
A. Discussion of selected references
285
 ∂ φL 
0.5
–1.5
lim 
 = 28.9 kJ·kg ·mol
m→0
m
∂

T , p
leads to the slope, which is over two times higher than the value of AL predicted by the
Debye-Hückel limiting law. Lange and Miederer’s experimental data were re-evaluated
using the chord method described in [58HAR/OWE]. A still higher limiting slope has
been obtained:
 ∂ φL 
0.5
–1.5
lim 
 = (37.5 ± 2.6) kJ·kg ·mol .
m→0
m
∂

T , p
The φL function thus derived was employed to extrapolate the enthalpies of
NiSO4·6H2O(cr) solution determined by Goldberg et al. to zero ionic strength
[66GOL/RID].
[56SOK]
The application of thermal, metallographic, X-ray, conductivity and specific gravity
measurements revealed the phase relations in the binary system Ni – S in the region
from 30.0 to 50.0 at% S. The phase diagram given by Elliott [65ELL] is based on results of this study as well as the work of Kullerud and Yund [62KUL/YUN].
[56YAT/VAS]
The solubility product at 25.0°C for hydrated Ni2P2O7, and 1:1 and 1:2 complexation
constants between Ni2+ and P2 O74 − were determined from a small set of solubility measurements in aqueous solutions of sodium pyrophosphate ranging in concentration from
0.032 to 0.123 M. The experiments seem to have been done carefully. Hydrolysis of
pyrophosphate was considered to be small, and was neglected in the data analysis. This
was probably not a bad approximation, as it can be calculated that < 5% of the unassociated pyrophosphate in the final solutions was protonated. The species NiHP2 O7−
[64HAM/MOR], [73PER/SEC] was even less important under the experimental conditions. Because of the presence of many highly charged species, the ionic strength of the
experimental solutions was high (0.5 to 1.2 M), and varied from solution to solution.
Estimation of interaction coefficients would be highly speculative for some of the species, and it does not seem useful to attempt to use the value in this paper for the first
association constant to determine thermodynamic values for I = 0. The value of 5.82 for
log10 K1 , is consistent with later studies in constant ionic strength media
[64HAM/MOR], [73PER/SEC].
The value determined for the cumulative second association constant (Reaction
(A.11)) was b 2 = 1.54 × 107
Ni2+ + 2 P2 O74 − U Ni(P2 O7 )62−
(A.11)
–8
(actually the cumulative dissociation constant, (6.5 ± 2.0) × 10 ) was reported). Reexamination of the experimental data suggests that the 1:2 complex does form, but evidence for its formation is only marginal except in those solutions with higher initial
A. Discussion of selected references
286
pyrophosphate concentrations (> 0.08 M), and log10 K 2 is probably < 1. The neglect of
pyrophosphate complexation with sodium ions (an inherent assumption required within
the context of the SIT) [94STE/FOT], introduces a further uncertainty. Therefore, no
value for K2 is selected in the present review.
In the absence of added sodium pyrophosphate, the solubility of the nickel pyrophosphate in 0.5 M to 1.0 M KNO3(aq) was reported as 3.4 × 10–4 M, which is not
inconsistent with the solubility product proposed by the authors (for higher ionic
strengths) if the lower hydroxide concentration, and consequent formation of HNiP2 O7− ,
is considered.
[56YAT/VAS2]
Measurements of the enthalpy of solution of aliquots of a 1.92 M solution of nickel nitrate into solutions of sodium pyrophosphate (0.025 M to 0.184 M) were used to determine the enthalpy of the complexation reactions of Ni2+ with P2 O74 − . The final nickel:
pyrophosphate ratios were also varied at the four different pyrophosphate concentrations. The values ∆ r H1 = (17.61 ± 0.17) kJ·mol–1 and ∆ r H 2 = – (9.25 ± 0.25) kJ·mol–1
at 25°C were reported. A value of K 2−1 = 2.5 × 10–2 derived from the calorimetric results
is also reported. It is not clear that the enthalpies of the two complexation reactions are
as easily separable as suggested by the authors, especially at the higher pyrophosphate
concentrations. The complexation constants cannot be accurately calculated for these
solutions (see the discussion above for [56YAT/VAS]), and therefore there are large
uncertainties in the enthalpies of the complexation reactions. Recalculations suggest that
the value for ∆ r H1 is more positive than the reported value, that ∆ r H 2 is more negative than the reported value, and that in none of the solutions does a single complex account for > 90% of the nickel in solution. For lower pyrophosphate concentrations, the
measured enthalpies also must include a contribution for deprotonation of small
amounts (< 5%) of monohydrogenpyrophosphate.
[57CHU]
The nickel orthoarsenite, not identified except by the Ni:As ratio, was found to dissolve
in dilute nitric acid solutions at 20°C over 12 hours to form solutions with aqueous
nickel concentrations of 8.7 × 10–3 mol·dm–3 at a “pH” value of 6.75 and 3.1 × 10–3
mol·dm–3 at a “pH” value of 7.10. The Ni:As ratio of the solid was found to be unchanged after the experiment. The author did not report a solubility product based on
their measurements. If the dissolution of the solid is assumed to correspond to the reaction:
Ni3(AsO3)2·xH2O + 6H+ U 3Ni2+ + 2HAsO2(aq) + (2 + x)H2O
(A.12)
and the concentrations are assumed to be sufficiently low that the ∆ε terms in the activity coefficient equations are negligible, the average value of log10 K (A.12) is
(28.69 ± 0.16). In the near-neutral solutions, ionisation of HAsO2 can be assumed small,
A. Discussion of selected references
287
as can the extent of any complex formation. If the initial “pH” values are compared to
those calculated based on the final concentrations of Ni2+, there are discrepancies of
19% and 31%. This suggests that the measured “pH” values were not particularly accurate. In the present review, the average value for 20°C is accepted as the value for 25°C,
but with an increased uncertainty, i.e., log10 K ((A.12), 298.15 K) is (28.7 ± 0.7).
[57KIU/WAG]
This pioneering work demonstrates the usefulness of emf measurements on galvanic
cells, involving solid electrolytes for the determination of the standard molar Gibbs energy of formation of oxides, sulphides, selenides, and tellurides at elevated temperatures. These measurements have been carried out by means of the “open cell stacked
pellet” technique. The reference electrode was a metal – metal oxide system of known
Gibbs energy of formation. The emf of the cell Fe, FexO (wüstite) (0.85 ZrO2 +
0.15 CaO) Ni, NiO and the standard Gibbs energy change of the reaction:
Ni(s) + ½O2(g) U NiO(s)
were determined in the temperature range from 1023.15 to 1413.15 K. The reported
reproducibility of emf was ± 2 mV or better, corresponding to an uncertainty in
∆ f Gm (T) of ± 0.418 kJ·mol–1. The standard molar Gibbs energy of formation of nickel
oxide was the following function of temperature: ∆ f Gm = (– 235.171 + 0.08567 (T/K))
kJ·mol–1.
[57KIV/LUO]
The formation constants of nickel(II) and zinc(II) chloro complexes were determined by
a polarographic method at 25°C using lead(II) as the indicator ion. The ionic strength
was held at 2 M using sodium perchlorate. The concentration of chloride ion was varied
between 0.4 – 1.4 M. The authors reported the formation of two nickel(II) chloro species: NiCl+ and NiCl2(aq) and their formation constants β1 = (0.56 ± 0.20) and β2 =
(0.9 ± 0.5). From these data, it follows that K2 (= β2/β1 = 1.6) is three times higher than
K1 (= β1). This is very unlikely taking into account the low complex forming ability of
chloride ion with nickel(II), and is probably the result of some unidentified experimental error. Consequently, these data could not be used to formulate conclusions concerning the nature and stability of chloro complexes.
[57MOR/ZEL]
This report describes a carrier-gas method for measuring vapour pressure of liquid metals. Vapour pressures of nickel were determined at temperatures 1813 to 1893 K. The
results can be represented by the following equation: log10 pmm = – 21030/T + 9.689,
where pmm is vapour pressure (in mm of Hg). The heat of vaporisation of nickel at 0 K
was determined to be 427.898 kJ·mol–1.
288
A. Discussion of selected references
[58KEN]
This paper reports cryoscopic measurements for NiSO4 aqueous solutions in saturated
potassium chlorate solution (freezing point 272.55 K) and saturated potassium nitrate
solution (freezing point 270.32 K). The eutectic freezing point depressions, ∆T, are between 0.0296 K and 0.178 K. Thus, together with the measurements reported in
[56KEN], the variation in the complexation constant, K1, with ionic strength was examined.
The experimental association constant values, K1, were extrapolated to the
saturation ionic strength of the supporting electrolyte at the melting point. Therefore,
the reported values at different limiting ionic strengths are also values for slightly different temperatures. For the cryoscopic measurements using potassium chlorate, the
ionic strength contribution of the supporting electrolyte is generally less than that from
the nickel sulphate. Only in the potassium nitrate solutions is the contribution of the
nickel sulphate to the total ionic strength reasonably minor. Therefore, the extrapolation
method used by the author is probably appropriate only for the potassium nitrate and
(perhaps) the potassium chlorate systems.
The author’s extrapolations to determine K1 in the saturated solution of supporting electrolytes are not strictly compatible with the TDB use of the SIT, but the reported K1 values are probably adequate if the uncertainties are increased. The values
log10K1 = (1.27 ± 0.30) (– 0.8°C, I = 0.26 mol·kg–1 KClO3(aq)) and log10K1 =
(0.69 ± 0.20 (– 2.8°C, I = 1.15 mol·kg–1 KNO3(aq)) are accepted for use in the present
assessment. In this review, the SIT is used to determine the values at I = 0 (with the
effective interaction coefficient for Ni2+ with NO3− (0.18 ± 0.01) kg·mol–1, neglecting
nitrate association (Section V.6.1.2.1), and the estimated interaction coefficient for Ni2+
with ClO3− (0.28 ± 0.10) kg·mol–1). The latter value was estimated as the average of the
interaction coefficients of Ni2+ with perchlorate and nitrate, with an increased uncertainty. The results are log10 K1 = (2.35 ± 0.30) (– 0.8°C) and log10 K1 = (2.16 ± 0.20)
(– 2.8°C).
The curve fits to Kenttämaa’s data could be used to determine values of the
second cumulative association constant, but these values of β2 are strongly dependent
on activity coefficient assumptions and on the values used for the heats of solution.
Kenttämaa did not propose β2 values (and considered that the value corresponding to β2
from the fitting process also incorporated systematic errors from the experiment and
data analysis). Rossotti and Rossotti [59ROS/ROS] did report β2 values based on the
measurements done by Kenttämaa using potassium nitrate mixtures. We concur with the
opinion of the original author, and do not accept these values of approximately log10 K 2
= 1 at I = 0.3 m and 1.2 m, although they are roughly comparable to a value reported by
Tanaka, Saito and Ogino [63TAN/SAI] ( log10 K 2 = 1.4 in 1.2 M NaClO4(aq)CH3COONa mixtures at 25°C).
A. Discussion of selected references
289
[58MUL]
Muldrow synthesised anhydrous NiCl2 and determined ∆ sol H m of reaction:
NiCl2(cr) U Ni2+ + 2Cl–
(A.13)
at 298.15 K. The mean value of ∆ sol H mο (A.13) has been obtained from column 4 of
Table A-9 by Equation (A.14)
∆ sol H mο (A.13) = ∆ sol H m (A.13)– φL,
(A.14)
see Section V.2.1.3.2 (NiCl2·6H2O(cr) cycle). For the calculation of φL the last term of
Equation (IX.44) in [97GRE/PLY2] had to be neglected, because the temperature derivative of ε(Ni2+, Cl–) is unknown. Systematic errors may have arisen from the fact that
the NiCl2 samples were not completely dry (n(H2O) / n(NiCl2) ≈ 0.04 – 0.08).
Table A-9: Enthalpy of solution of NiCl2(cr) in H2O(l) at 298.15 K [58MUL].
b (NiCl2)
∆ sol H m (A.13)
–1
–1
φ
∆ sol H mο (A.13)
L
–1
(mol·kg )
(kJ·mol )
(kJ·mol )
(kJ·mol–1)
0.00545
– 82.508
0.667
– 83.175
0.00971
– 83.596
0.854
– 84.451
0.00192
– 82.341
0.417
– 82.758
0.00200
– 82.718
0.424
– 83.142
0.00373
– 82.760
0.564
– 83.323
0.00555
– 83.011
0.672
– 83.683
0.00138
– 81.881
0.357
– 82.238
0.00279
– 82.508
0.494
– 83.003
0.00685
– 83.303
0.736
– 84.040(a)
0.01036
– 83.387
0.878
– 84.265
(a) Carried out in dilute acid.
[58TRE]
The method of frontal analysis on an ion-exchange column was used to study the formation of weak chloro complexes of nickel(II), cobalt(II) and copper(II). Although, the
reported dissociation constant for the complex NiCl+ (Kdiss = (4.6 ± 0.1), which can be
converted to log10 K1 = – (0.66 ± 0.01)) is close to other reported values, some experimental data, such as the temperature of the measurements, were not published. Therefore, the reported constant is rejected in this review.
[58VAI/RAM]
The only part of this work that was related quantitatively to pyrophosphate complexation of nickel was a sparse set of spectrophotometric measurements from which an “in-
A. Discussion of selected references
290
stability constant” of 2.4 × 10–4 was derived. The pH of the solutions is not indicated,
though apparently the sodium salt was used. The association quotient is smaller by several orders of magnitude than values from other studies, and there are insufficient experimental details to determine if the value might refer to a protonated complex. The
reported value is not used in the present review.
[58YAT/KOR]
The formation of MSCN+ complexes (M = Mn(II), Fe(II), Co(II) and Ni(II)) has been
studied by a spectrophotometric method, by detecting the decrease of the absorbance of
Fe(SCN)2+ complex at 465 and 495 nm with increasing concentrations of M(II). The
concentration of Fe(III) and SCN– was 5 and 0.5 mM, respectively, while the concentration of Ni(II) ranged from 0.02 to 0.5 M. Consequently, the ionic strength varied between 0.13 and 1.56 M. The Davies equation was used to calculate the activity coefficients. For the purposes of the present review, the reported data were re-evaluated using
the SIT (cf. Figure A-4). From the plot of this figure, log10 b1ο = (1.69 ± 0.15) and ∆ε =
– (0.075 ± 0.030) kg·mol–1 can be derived.
Figure A-4: Extrapolation to Im = 0 (using the SIT) of the formation constants for the
species NiSCN+ as reported in [58YAT/KOR].
2.2
log10 β1 + 4D
2.0
1.8
1.6
1.4
1.2
0.0
0.5
1.0
1.5
−1
I / mol·kg
[59ACH]
The pH of slightly alkaline 0.01 M K2SO4 and KClO4 solutions was measured before
and after addition of a Ni(NO3)2 solution under N2 atmosphere. The concentration of
A. Discussion of selected references
291
nickel(II) in the titration vessel varied between 0.003 and 0.024 M. The ionic strength
was varied during the measurements, and the Güntelberg equation was used to calculate
the activity coefficients. Since the inertness of sulphate media is highly questionable,
only the data in perchlorate solution will be discussed further. The solutions were
equilibrated for 30 minutes. The experimental data were interpreted only in terms of the
formation of NiOH+. Therefore, we re-evaluated the reported experimental data assuming the presence of both NiOH+ and Ni 4 (OH) 44 + . The following values were obtained:
log10 *b1,1 = − (10.85 ± 0.10) and log10 *b 4,4 = – (26.8 ± 1.0). This calculation revealed,
however, that the degree of hydrolysis is very low. The highest concentration of NiOH+
and Ni 4 (OH) 44 + was 5.4 × 10–7 and 4.4 × 10–9 M, respectively. Therefore, above data
were not considered in the present review.
[59BLA/GOL]
The spin-lattice relaxation times of water protons have been measured to study complex
formation between nickel(II) and cyanide ion in concentrated aqueous ammonia solution (ρ = 0.88 g·cm–3). The relaxation time increased up to the ratio [CN–]/[Ni2+] = 4/1,
where a sharp breakpoint was observed, according to the transformation of the paramagnetic Ni(NH 3 )62+ to the diamagnetic Ni(CN) 24 − . At higher cyanide concentration the
relaxation time decreased with increasing cyanide concentration, and after a second
breakpoint at the ratio [CN–]/[Ni2+] = 6/1 the relaxation time was unaffected by further
cyanide addition. From these observations the authors were led to the conclusion that
the addition of cyanide ion to Ni(CN) 24 − results in quantitative formation of Ni(CN)64 − .
This finding was disputed later by many authors [60MCC/JON], [62BEC/BJE],
[63PEN/BAI], [64GEE/HUM], [65COL/PET], [71PIE/HUG], and now the nonexistence of Ni(CN)64 − is well established.
[59BLA/GOL2]
In this work an effort has been made to determine the formation constants of the
Ni(CN)64 − species, using relaxation times derived from 1H NMR measurements. As
already mentioned in the discussion of [59BLA/GOL], the existence of this species can
be ruled out, even in the case of high excess of cyanide ion.
[59FRE/SCH]
Complex formation in the nickel(II) – cyanide system has been investigated by spectrophotometry (267.5 nm) at 24.92°C and at several ionic strengths (I = 0.0028 to 0.1 M)
using potassium perchlorate as the ionic medium, in the pH range from 5.3 to 7.7. The
nickel(II) concentration was ~ 4 × 10–5 M. The [Ni2+]/[CN–] ratio was varied from 0.05
to 0.8 to obtain data so that the Job's method could be applied, and otherwise kept at
around ¼. The time required for the equilibration ranged from a few days for solutions
of higher pH to several weeks for those of lower pH. Acetate and phosphate buffers
were used to maintain the pH, but correction was made only to account for the formation of the acetato complex. Only formation of the complex Ni(CN) 24 − was detected,
292
A. Discussion of selected references
with a formation constant of log10 b 4ο = (31.0 ± 0.1). The authors used pK HCN = 9.398,
which is the average value of some early determinations. Since the above value is not
correct, and log10 b 4 depends on pK HCN to the fourth power, Christensen et al.
[63CHR/IZA] recalculated the thermodynamic formation constant reported by Freund
ο
and Schneider, using the more reliable pK HCN
= 9.216, which resulted in log10 b 4ο =
(30.3 ± 0.1). The method of this recalculation is unknown; therefore, a re-evaluation
was carried out in the present review using the reported experimental data. Since the
concentration of the acetate and phosphate buffers were not reported, corrections taking
into account the acetato- and phosphato- complexes of nickel(II) was not possible. The
values of pK HCN at different ionic strengths were calculated using the data reported in
[92BAN/BLI]. The re-evaluation indicated that only Ni(CN) 24 − formed in the equilibrated solutions, with log10 b 4 = (29.39 ± 0.14) (I = 0.1 M), (29.46 ± 0.13) (I = 0.048
M), (29.79 ± 0.10) (I = 0.014 M), (30.12 ± 0.12) (I = 0.0028 M). For the purposes of
this review, slightly larger uncertainties, ± 0.20, were estimated for each of these constants. The ionic strength range is too narrow to perform an SIT analysis, but these values for log10 b 4ο determined at low ionic strengths are nearly identical to the value
of log10 b 4ο selected in the present review.
[59KEN]
This paper reports cryoscopic measurements of NiCl2 solutions in saturated potassium
chlorate (0.258 m, f.p. = 272.35 K) and potassium perchlorate (0.0485 m, f.p. = 272.99
K) solutions. The author determined the freezing point depression at different NiCl2
concentrations (in KClO3 c NiCl2 = 0.0193 – 0.121 m and ∆T = 0.10 – 0.61°C, in KClO4
c NiCl2 = 0.0234 – 0.1202 m and ∆T = 0.12 – 0.59°C), then extrapolated to c NiCl2 = 0 at
the saturating ionic strength of the supporting electrode. The ionic strength contribution
of NiCl2 is dominant in KClO4 solutions, therefore the extrapolation used by the author
can be accepted, with an increased uncertainty, only for the data measured in KClO3
solutions. The value of log10 b1 = (0.23 ± 0.30) (– 0.8°C, I = 0.258 m KClO3) is accepted in this review. From this value, log10 b1ο = (0.79 ± 0.40) (– 0.8°C) can be calculated using the SIT, assuming the interaction coefficients ε(Ni2+, ClO3− ) to be
(0.28 ± 0.10) kg·mol–1 (see also [58KEN]) and ε(NiCl+, ClO3− ) = 0.5 × {ε(Ni2+, ClO3− )
+ ε(K+, Cl–)} = (0.135 ± 0.100) kg·mol–1.
[59KIS/CUP]
Spectrophotometric measurements have been performed in the binary nickel(II)-cyanide
and copper(I)-cyanide, and ternary nickel(II)-copper(I)-cyanide systems. It was concluded that Ni(CN) 24 − and Ni(CN)64 − form in the binary nickel system. The equilibrium
constant for the reaction:
Ni(CN) 24 − + 2 CN– U Ni(CN)64 −
was determined in 5 M potassium acetate ( log10 K = (3.33 ± 0.15)) and 10 M potassium
nitrite ( log10 K = (3.42 ± 0.15)) solutions (the temperature of the measurements was not
A. Discussion of selected references
293
reported). Today, the non-existence of Ni(CN)64 − is well established [60MCC/JON],
[62BEC/BJE], [63PEN/BAI], [64GEE/HUM], [65COL/PET], [71PIE/HUG]. According
to Beck and Bjerrum [62BEC/BJE], the results reported in [59KIS/CUP] do not support
the formation of a stable hexacyano complex, but rather of an unstable pentacyano species.
[59LAF]
The phase relationships between nickel monosulphide and Ni3S2 or Ni7S6 were studied
by equilibrating NiS with hydrogen gas at temperatures ranging from 677.15 to
1006.15 K. The high-temperature form of NiS was in equilibrium with Ni7S6 from 677
to 823 K and with Ni3S2 (high-temperature modification) between 823 and 1006 K, respectively. The H2S(g) content of the gas, formed due to reduction of the monosulphide
with hydrogen, was measured iodometrically and the composition of the solid phases
was determined by X-ray diffraction analysis, magnetic susceptibility measurements,
and chemical analysis. From the ratio of the partial pressures between H2S(g) and H2(g)
the partial pressure of S2 can be easily calculated, leading to a temperature dependence
of p(S2) which is in close agreement with the data reported by Rosenqvist [54ROS].
[59NAI/NAN]
The paper reports on a series of potentiometric measurements (0 to 45°C). The measurements were carried out in acidic media (HCl(aq)), and a Davies-type equation was
used to calculate the activity coefficients (I < 0.05 m). The reported formation constant
(A.15) for NiSO4(aq) at 25°C is (211 ± 24),
Ni2++ SO 24 − U NiSO4(aq).
(A.15)
Sufficient data are provided to permit recalculation using the SIT and the TDB
selected values [92GRE/FUG] for the first protonation constant (and enthalpy of protonation) for the sulphate ion. As in the original analysis, the initial concentrations, measured potential of the cell:
H2(g), Pt| HCl(aq, m1), MSO4(aq, m2)|AgCl(s)|Ag(s),
the selected value for the equilibrium constant (I = 0) for:
H+ + SO 24 − U HSO −4 ,
(A.16)
and the selected activity coefficient equations (in this case the SIT equations) were
used. A model assuming formation of a single 1:1 nickel sulphate complex was used,
and calculations were carried out so that the ionic strength values in the activity coefficient equations were identical to those calculated from the final species concentrations.
The values log10 K ο ((A.16), 298.15 K) = (1.98 ± 0.05) and ∆ r H mο ((A.16),
298.15 K) = (22.4 ± 1.1) kJ·mol–1 were those selected in a previous TDB review
[92GRE/FUG]. From these, the association constants for Reaction (A.16) were calculated assuming constant values for C pο,m over the temperature range 0 to 45°C. As dis-
294
A. Discussion of selected references
cussed below, this is probably not adequate. The values ε(Ni2+, Cl–) = (0.17 ± 0.02) and
ε(H+, Cl–) = (0.12 ± 0.01) were from the same source (see also Section V.4.2.4). The
value ε(H+, HSO −4 ) = (0.01 ± 0.02) is the value used by Grenthe et al. [92GRE/FUG] in
their Appendix A discussion of [76PAT/RAM]. No specific value was adopted for
ε(Ni2+, HSO −4 ), as this interaction was assumed to be incorporated in the value of the
formation constant for the NiSO4(aq) complex. Preliminary calculations indicated that
the reported data for experiment 8 at 298.15 K [59NAI/NAN] could not be used in recalculations. The concentrations reported in Table 1 of the paper are not compatible
with the reported complexation constant value or the reported final ionic strength.
The calculations showed that the values calculated for the formation constant
of NiSO4(aq) are extremely sensitive to the activity coefficient model and to the values
assumed for the first protonation constant for the sulphate ion. For example, for
298.15 K, the formation constant was recalculated to be (173 ± 64). However, if the
value for log10 K ο (A.16) was changed from 1.98 to 1.96, well within the assessed uncertainty bound, the value became (140 ± 72). The uncertainties here are merely the 2σ
uncertainties in the set of seven values recalculated from [59NAI/NAN], and do not
include uncertainties in the auxiliary data. Even though the solute concentrations are
low (I < 0.05), and the overall changes to the activity coefficients from the SIT interaction terms are small, the calculated value of log10 K ο (A.15) increases from (173 ± 64)
to (201 ± 35) if all the ε values are set to zero (but the maximum change in g Cl− from
the interaction terms is 0.004).
In the present review, the value log10 K ο ((A.15), 298.15 K) = 173 is accepted,
and an uncertainty of ± 80 is estimated. It is not clear whether the values of K(A.16) at
temperatures other than 298.15 K are sufficiently well defined to generate usable values
of K(A.15). The calculated values of K(A.15) at the six different temperatures can be
used to calculate ∆ r H m = – (13.8 ± 1.0) kJ·mol–1 if the enthalpy of reaction is assumed
to be constant from 273 to 318 K. However, Figure V-38 suggests that the values of
K(A.15) at the two lowest temperatures diverge from those obtained from conductance
and cryoscopic studies.
[59ROB/STO]
The data m, γ± and φ of NiSO4 required in the plots of Figure A-5 and Figure A-6 were
taken from Table 16, Appendix 8.10 of this textbook. The activity coefficient γ±(sat) has
been calculated by a third order polynomial of ln γ± vs. m , as depicted in Figure A-5.
For the activity of water, aW(sat), φ was extrapolated to the saturated solution molality
with a fourth order polynomial in m, as depicted in Figure A-6.
A. Discussion of selected references
295
Figure A-5: Extrapolation of activity coefficients to saturated solution at 25°C. Solid
curve: calculated with a third order polynomial in m : ln γ± = A + B1·m0.5 + B2·m +
B3·m1.5 with A = – (0.90807 ± 0.02571), B1 = – (3.60463 ± 0.09626), B2 =
(1.42317 ± 0.10959) and B3 = – (0.07339 ± 0.3843), with R2 = 0.9998, and σ = 0.00699.
● data from [59ROB/STO], extrapolated value.
-1.4
-1.6
-1.8
-2.0
ln γ±
-2.2
-2.4
-2.6
-2.8
-3.0
-3.2
-3.4
-3.6
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
−1 0.5
(m / mol·kg ) (NiSO4)
Figure A-6: Extrapolation of osmotic coefficients to saturated solution at 25°C. Solid
curve: calculated with a fourth order polynomial in m: φ = A + B1·m + B2·m2 + B3·m3 +
B4·m4; with A = 0.61907, B1 = – 0.48613, B2 = 0.46844, B3 = – 0.16846, B4 = 0.02584,
and R2 = 0.99831 and σ = 0.00326. ● data from [59ROB/STO], extrapolated value,
with m(NiSO4) = 0.7488 ;φ = 0.7488.
0.80
0.75
0.70
φ
0.65
0.60
0.55
0.50
0.45
-0.5
0.0
0.5
1.0
1.5
2.0
−1
m (NiSO4) / mol·kg
2.5
3.0
296
A. Discussion of selected references
[59WIL/THR]
Hellyerite forms a colourful association with bright green, amorphous zaratite,
Ni3CO3(OH)4·4H2O, along shear planes in serpentine containing heazlewoodite (Ni3S2)
and other nickel sulphides. The crystals are of a cobalt or pale blue. The empirical formula was found to be Ni(CO3)1.18·5.67H2O. The mineral was named after Henry Hellyer, first Surveyor General of the Van Dieman’s Land Company, who was responsible
for exploring and surveying much of North Western Tasmania during 1826 – 1830 (approved by IMA 1959). Its rarity is probably due to the fact that hellyerite is relatively
unstable and, if not kept in a cool, air-tight environment (Heazlewood is typically rather
cold and damp), it decomposes to powdery zaratite-like phases, plus otwayite,
Ni2(CO3)(OH)2·H2O, and other minerals.
[60LIS/ROS]
See comments under [66KEN/LIS].
[60MCC/JON]
See the discussion for [65COL/PET].
[60WIL2]
Williams discovered a new mineral at Heazlewood, Tasmania, and identified it as nickel
hydroxide, Ni(OH)2, by its X-ray powder pattern.
[61CHU/ALY2]
A value of log10 K s ,0 for hydrated Ni3(PO4)2 at 20°C is reported as – 30.3. The value is
based on sparse measurements, and no account is taken of complex formation between
Ni2+ and HPO 24 − or other phosphate species. The results from this study were not used
further in the present review.
[61DAV/SMI]
Spectrophotometric measurements were performed at 242.5 – 252.5 nm to determine
the association constant for the NiSCN+ species. The concentration of the metal ion was
in five- to twenty-fold excess over that of the thiocyanate ion so that the kinetic data
could be interpreted. The ionic strength was 0.5 M, the background electrolyte is not
specified. The association constants were measured between 1.4 and 45°C, but only the
value for 1.4°C is reported in the paper (although two interpolated values for 5.2 and
9.3°C are also given). Due to the lack of experimental details, the reported constants
were not considered further in this review.
[61LEB/LEV]
Lebedev and Levitskii investigated the equilibrium of the reaction:
Ni2SiO4 + 2CO U 2Ni + SiO2 + 2CO2
A. Discussion of selected references
297
in the range 800 – 1100°C using a circulation technique with an automatic and a volumetric gas analyser. The equilibrium of this reaction was approached both by reduction
of nickel orthosilicate and by oxidation of nickel. Combining the experimentally determined Gibbs energy change for the above reaction with the Gibbs energy for the reaction:
2 CO(g) + O2(g) U 2 CO2(g),
the authors [61LEB/LEV] calculated the Gibbs energy for the reaction:
2Ni + SiO2 + O2 U Ni2SiO4
in the range 800 – 1100°C. These ∆ r Gmο values are significantly different from results
obtained by another direct equilibration technique [68CAM/ROE] and the emf-studies
shown in Figure V-54. The results of Lebedev and Levitskii were, therefore, disregarded by this review. A possible explanation of the observed discrepancy seems to be
difficulty in attaining equilibrium for the reduction of nickel orthosilicate by carbon
monoxide in the temperature range chosen by [61LEB/LEV]. The values of the standard
enthalpy of formation ∆ f H mο (Ni2SiO4, olivine, 298.15 K) = – 1378.21 kJ·mol–1 and the
Gibbs energy of formation ∆ f Gmο (Ni2SiO4, olivine, 298.15 K) = − 1265.24 kJ·mol–1
proposed by Lebedev and Levitskii were not used by this review and are significantly
different from the recommended values.
[61LIS/WIL]
The emf of the cell Ag,AgBr|NaBr+M(ClO4)2+NaClO4|NaBr+NaClO4|AgBr,Ag (where
M = Ni(II) or Co(II)) has been measured at I = 2 M (Na, M)ClO4 and between 273 –
323 K in order to determine the composition and stability of the nickel(II) and cobalt(II)
complexes formed with bromide ion. The concentration of nickel(II) ranged from 0.08
to 0.31 M, while the chloride concentrations were nearly constant ([Cl–] ~ 0.008 M).
The complex formation induced 1 to 6.5 mV differences between the electrodes. The
average value of log10 b1 at 298.15 K is – (0.12 ± 0.02). This value has been accepted
with an increased uncertainty (± 0.40). From the temperature dependence of log10 b1 ,
∆ r H m appears to fall from about 13 to 2.7 kJ·mol–1 over the range 0 to 50°C (see Figure
A-7). The authors could not explain this behaviour, which is probably due to some unidentified experimental error. Assuming constant ∆ r H m between T = 273 – 323 K, and
that the medium effect is additive in the logarithmic scale, from these data ∆ r H mο for the
reaction,
Ni2+ + Br– U NiBr+
can be obtained as a rough estimation, ∆ r H mο (A.17) = (8 ± 5) kJ·mol–1.
(A.17)
A. Discussion of selected references
298
Figure A-7: Plot of log10 b1 (A.17) reported in [61LIS/WIL] vs. 1/T.
0
-0.05
-0.1
log10 β1
-0.15
-0.2
-0.25
-0.3
-0.35
-0.4
0.003
0.0031
0.0032
0.0033
0.0034
0.0035
0.0036
0.0037
K/T
[61MAR/BYD]
The kinetics of the reaction:
Ni(edta)2– + 4CN– U Ni(CN) 24 − + edta4–
(A.18)
have been studied by spectrophotometry (285 nm) at 25°C in I = 0.11 and 1.01 M NaCl
solutions around pH 11. From the kinetic data, log10 b 4 (A.18) = 31.52 was derived at I
= 1.01 M, using pK HCN = 9.36. Although, the latter pK is rather far from the correct
value [92BAN/BLI], considering the pH used, the error in pK HCN has little effect on the
reported log10 b 4 (A.18) value.
[61MOH/DAS]
The thermodynamic stability constant of the NiSCN+ complex was determined by spectrophotometric (t = (25 ± 1)°C) and potentiometric (t = 35°C) methods. The Davies
equation was used to calculate the activity coefficients, and this is not compatible with
the SIT. Therefore, the potentiometric data were re-evaluated, using the SIT approach.
The resulting log10 b1 – Im plot is rather scattered and the ionic strength range was relatively narrow (see Figure A-8). Therefore the reported data were not considered further
in this review.
A. Discussion of selected references
299
Figure A-8: log10 b1 – Im plot obtained from the potentiometric data (t = 35°C) reported
for the Ni2+ – SCN– system in [61MOH/DAS].
log10 β1 + 4D
2.2
2.1
2
1.9
0
0.1
0.2
0.3
0.4
0.5
-1
I / mol˙kg
[61SHC]
Shchigol synthesised nickel orthoborate by reacting nickel hydroxide or carbonate with
orthoboric acid and determined its stoichiometric composition analytically. No attempt
was made to characterise this substance structurally. First, the solubility of nickel orthoborate, Ni(BO2)2·4H2O(s), was studied at 22°C by equilibrating the solid phase with
HCl solutions. Second, solubility measurements in KOH solutions were interpreted by
assuming simultaneous equilibrium between Ni(BO2)2·4H2O(s), Ni(OH)2(s) and the
aqueous medium. A third method consisted of precipitating nickel nitrate with borax
solutions. It should be mentioned that neither the solid nickel borate nor the precipitated
nickel hydroxide were structurally characterised. An average value of the solubility
product of Ni(BO2)2·4H2O(s) K s ,0 = 2.16×10–9 mol3·dm–9 was obtained by these three
methods. A recalculation of the solubility product was based on the experimental data
obtained from undersaturation (method 1) and oversaturation (method 3). The acid dissociation constant of boric acid was taken to be log10 K aο = – 9.623 at 22°C
[58HAR/OWE] and the Davies equation [62DAV] was used for extrapolation to I = 0.
A value of log10 K sο,0 = – (8.80 ± 0.19) was found in reasonable agreement with Shchigol’s result ( log10 K sο,0 = – 8.666), see Figure A-9.
Shchigol found that the solubility of Ni(BO2)2·4H2O in aqueous media increases with increasing concentration of orthoboric acid. In electrolysis experiments the
A. Discussion of selected references
300
solution of the anode compartment was enriched in nickel indicating its presence in the
complex borate anions.
The formation of the nickel and cobalt borate complexes in solution was believed to occur according to the reaction:
M(BO2)2(s) + nH3BO3(aq) U HnM(BO2)2+n(aq) + nH2O(l)
(A.19)
The value n calculated from the solubility measurements on Ni(BO2)2·4H2O(s)
at different concentrations of H3BO3 was calculated to be 0.7. Thus the co-ordination
number for Ni2+ relative to the metaborate ion was assumed to be 3 and the proposed
complex species was Ni(BO 2 )3− . While Figure A-10 shows that for nickel n = 1 and for
cobalt n = 2, Shchigol’s conclusion that the dissolved Ni2+ essentially consists of
Ni(BO 2 )3− must be refuted. Where do the balancing cations come from ? When it was
attempted to explain the data by the formation of a neutral species like NiH(BO2)3, the
solubility product did not remain constant. Additional experiments (careful pH measurements, reagents free of protolytic impurities) are necessary to establish the formula
and stability of nickel borate complexes. After standing 5 to 6 days, fine crystals of
nickel hexaborate NiB6O10 were precipitated from the complex nickel borate solution
containing an excess of orthoboric acid possibly according to the reaction:
HNi(BO2)3(aq) + 3H3BO3(aq) U NiB6O10(s) + 5H2O(l)
(A.20)
Figure A-9: Solubility product of Ni(BO2)2·4H2O(s) at 22°C from data reported in
[61SHC].
oversaturation,
exp. slope: – 1.39, undersaturation, exp. slope both series: –
1.06 ; theor. slope: – 2.00
-1.8
-2.0
2+
log a (Ni )
-2.2
-2.4
-2.6
-2.8
-3.5
-3.4
-3.3
-3.2
-3.1
-
log a (BO2 )
-3.0
-2.9
A. Discussion of selected references
301
Figure A-10: Nickel and cobalt borate complexes from data in [61SHC].
M2+ : Co2+,
Co(BO2)2·2H2O(s) + 2 H3BO3(aq) U CoH2(BO2)4(aq) + 4 H2O(l),
2+
2+
M : Ni ,
Ni(BO2)2·4H2O(s) + H3BO3(aq) U NiH(BO2)3(aq) + 6 H2O(l),
([Ni2+]tot – [Ni2+]) = f ([H3BO3].
0.035
t = 22°C
0.030
0.020
0.015
2+
[M ]tot / mol·dm
-3
0.025
0.010
0.005
0.000
-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
-3
[H3BO3] /mol·dm
[61TAN2]
The standard potentials of cobalt and nickel in methanol were determined using the following cell without liquid junction:
Ni | NiCl2, solvent | Hg2Cl2 | Hg.
Whereas for the cobalt cell measurements were carried out successfully in pure
water, methanol was the solvent for the nickel cell. Apparently the author experienced
the same difficulties with nickel in the aqueous medium which were already described
by [29HAR/BOS]. Thus, the results of this work provide no relevant information on
thermodynamic data of Ni2+ in aqueous solution.
[61TAN/OGI]
Polarographic measurements were carried out in an acetate buffer (the assumed “pH”
was 3.66), and concentrations of sulphate and acetate were varied in the presence of
1.00 × 10–3 M nickel nitrate and 1.09 × 10–3 M nitrilotriacetate at a constant ionic
strength of 0.2 M (maintained with potassium nitrate). A dropping mercury electrode
was used, and polyoxyethylene lauryl ether (1–2 × 10–6 M) was used as a maximum
suppressor. Although results are reported for 15, 25 and 35°C, the assumed “pH” value
for the buffer at 15 and 35°C is not stated. The acetate molarities were varied from
0.025 to 0.100 M at 0.03 M sulphate, and the potassium sulphate molarities were varied
from 0.01 to 0.05 M at 0.05 M acetate. The results indicated that for the experimental
302
A. Discussion of selected references
conditions only a monosulphato complex was formed for Ni(II), with no evidence for
higher complexes or a mixed sulphato-acetato complex. Values of (14.3 ± 0.4),
(11.6 ± 0.9) and (14.6 ± 0.7) dm3·mol–1 were reported for K1 at 15, 25 and 35°C, respectively. The reported statistical uncertainties are undoubtedly underestimates of the true
uncertainties. For the purposes of the present assessment, the value log10 K1 =
(1.06 ± 0.20) at 25°C is used; the measurements at 15 and 35°C are rejected because of
the lack of experimental details. At this low ionic strength, assuming that the ε value for
the Ni2+–acetate interaction is the same as the value for the Ni2+– NO3− interaction has
little effect on the value calculated for I = 0 using the SIT, and log10 K1ο = (2.13 ± 0.20)
is obtained.
[62BEC/BJE]
The effect of different salts on the visible spectra of tetracyanonickelate(II) has been
investigated, in order to elucidate the formation of further complexes, and to solve the
discrepancy between two earlier papers, [59BLA/GOL] and [60MCC/JON]. All salts
(KCN, KSCN, KI, KBr, KCl, KF, KNO3) added to Ni(CN) 24 − caused an increase in
absorbance around 375 – 500 nm, but the effects were rather different. The data qualitatively confirmed the formation of four very weak complexes: Ni(CN)35− , Ni(CN)4SCN3–
Ni(CN)4I3–, and to a much lesser extent Ni(CN)4Br3–. Numerical estimates of their stability were not provided.
The above-mentioned salts and some other compounds (NH3, pyridine, ethylenediamine, butylamine, ethanol, methanol, dioxane, dimethylformamide) were also
added to solutions containing 0.04 Ni(CN) 24 − and 0.98 M KCN. Relatively small concentrations of all these compounds result in a decrease of absorbance between 375 –
500 nm. Since ammonia has nearly the same effect as fluoride ion or dioxane, the authors rejected the possibility of further complex formation (such as Ni(CN)5F4–), and
explained the observation as a strong decrease of activity coefficients corresponding to
the mass action expression:
F = f Ni(CN)2− f CN− / f Ni(CN)5− .
4
3
[62CON/GIL]
Conway and Gileadi carried out electrochemical kinetic studies at the nickel oxide
(NiII–NiIII) electrode. A polarisation decay method was employed in which the reversible potential was approached both from the anodic and the cathodic directions.
[62FAU/CRE]
Dissociation of some unstable chloro complexes has been studied by cryoscopic titration in saturated KNO3 (I = 1.38 m, f.p. = 270.34 K). Using [Ni(NO3)2] = 0.02 M and
[KCl] = 0.05 – 0.25 M; the freezing-point depressions were found to be between 0.065
and 0.393 K. The recalculation of the reported data using non-linear regression resulted
in a value ( log10 b1 = (0.36 ± 0.06)) almost identical to that reported by the authors
A. Discussion of selected references
303
( log10 b1 = (0.38 ± 0.08)). The value of log10 b1 = (0.36 ± 0.30) is accepted for use in
the present assessment. Assuming ε(NiCl+, NO3− ) equal to 0.5((ε(Ni2+, NO3− ) +
ε(K+, Cl–)) = 0.09 kg·mol–1 the extrapolation using the SIT resulted log10 b1ο =
(1.17 ± 0.40) (– 2.0°C).
[62FRO/LAR]
The frequency and intensity of the C–N vibration have been measured for the SCN– ion
(at 2066 cm–1) and some MSCN+ complexes of 3d metal ions. In the presence of metal
ions, new peaks at ~ 2100 cm–1 appeared in all systems, and these peaks correspond to
formation of the metal complexes. The intensities of the free and bound SCN– ion were
used to determine stability constants of the monothiocyanato complexes which formed.
The NaSCN concentration was 0.36 M, [Ni2+]tot varied from 0.44 to 1.32 M, and (2.64 –
2[Ni2+]tot) M NaClO4 was used as a background electrolyte. Therefore, during the measurements, the ionic strength ranged from 3.44 to 4.32 M. Moreover, the temperature of
the measurements was not reported in the paper. Although, the reported stability constant for the NiSCN+ complex ( log10 b1 = 1.18) is close to the value in other papers,
taking into account the uncertain experimental conditions, the above value was not considered further in this review. Based on the comparison of different evaluation methods
of the experimental data, the authors suggested the formation of 80% inner-, and 20%
outer-sphere complexes.
[62KUL/YUN]
A systematic study of the phase relations in the binary system Ni – S was performed in
the temperature range between 473 and 1303 K. The phase equilibria were investigated
by means of quenching experiments, differential thermal analysis (DTA) and, to a lesser
extent, high temperature X-ray powder diffraction. The phases present in the quenched
samples were detected by application of X-ray diffraction analysis and reflecting microscopy. In the latter case the samples were polished and etched with concentrated
nitric acid. The following stoichiometric phases were found by the authors: Ni3S2
(heazlewoodite), Ni7S6 (low temperature modification), NiS (millerite, rhombohedral
structure), Ni3S4 (polydymite), and NiS2 (vaesite). The high temperature modifications
of Ni3S2, Ni7S6, and NiS (hexagonal NiAs-type structure) were reported to show a nonstoichiometric homogeneity range. Whereas the homogeneity range of Ni7S6 was fairly
narrow, the high temperature phases Ni1–xS and Ni3+xS2 showed large stability fields in
the respective phase diagram. In the sulphur-rich portion of the system a two-liquid
phase field over a wide range of composition was revealed. The authors proposed a
monotectic reaction between NiS2 and the two immiscible liquids at 1264 K.
[62LOP]
In this work a similar set-up to the one described previously by [52CAR/BON] was
used for the measurement of the Ni2+ | Ni standard electrode potential. To avoid the
formation of a Ni(OH)2 surface layer on the Ni electrode the NiSO4 solution was acidi-
A. Discussion of selected references
304
fied with H2SO4. According to other authors the nickel electrode corrodes in acid solution [52CAR/BON]. As a consequence the result, E°(Ni2+ | Ni, 295.15 K) =
− (0.248 ± 0.004) V, (recalculated with the CODATA value for the saturated calomel
electrode and corrected for an obvious misprint in the error limits) has been rejected.
[62RIN]
A direct search was made for the spinel polymorph of Ni2SiO4(cr) using high-pressure
techniques. An olivine – spinel transition was discovered at 650°C and (18000 ± 5000)
bars. This result is in excellent agreement with the theoretical prediction. This enables
the free energy of transition and the density of the closer packed polymorph to be found.
From these data the transition pressure can be calculated.
[62TRI/CAL]
The decrease of the kinetic wave heights of the Ti(IV)-SCN complex at a Hg dropping
electrode, in the presence of a second cation, allowed the stability constants of Ni(II),
Co(II), Cd(II) and Mn(II) with thiocyanate ions to be determined at different ionic
strengths (I = 0.7 – 5 M (Na,H)ClO4). Due to the weak complex formation, this method
could not be applied to the detection of formation of the NiCl+ complex. Therefore, the
authors also determined b1 (NiSCN + ) in a solution of 1.5 M (Na,H)ClO4 by means of
spectrophotometry, using the UV-band at 275 nm, attributed to the NiSCN+ species.
Then the decrease of the absorbance at 275 nm with increasing chloride concentration
allowed calculation of b1 (NiCl+ ) . The following values were reported for the NiSCN+
complex.
Table A-10: log10 b1 versus Im.
0.73
Im
log10 b1
1.05
1.62
2.21
2.84
6.58
(1.21 ± 0.02) (1.17 ± 0.02) (1.12 ± 0.02) (1.10 ± 0.04) (1.11 ± 0.02) (1.27 ± 0.05)
Several experimental details are not given in the publication: (i) the temperature of the measurement, (ii) the chloride concentration used to determine b1 (NiCl+ )
(Chemical Abstracts reported [Cl–] < 1.3 M). The composition of the background electrolyte ((Na, H)ClO4) is also dubious. Therefore, the reported data were not taken into
account in this review.
[62VAG/UVA]
Vagramian and Uvarov carried out emf measurements in the range of 18 ≤ t ≤ 250°C
with the cell:
Ni | NiSO4(0.501 mol·kg–1) | Hg2SO4 | Hg.
A. Discussion of selected references
305
Between 200 and 250°C the potential became a linear function of the temperature and was extrapolated to 25°C. The resultant E°(298.15 K) = 0.952 V combined
with the standard potential of the Hg2SO4 | Hg electrode ((0.6127 ± 0.0030) V, NEATDB auxiliary data) would lead to E°(Ni2+ | Ni, 298.15 K) = – 0.339 V. The value
originally given (– 0.270 V) is based on an obsolete standard potential value for the
Hg2SO4 | Hg electrode (0.682 V). Finally, this result has been rejected due to the questionable calculation of the NiSO4 activity at t > 200°C and the rather long extrapolation
to 25°C, although the slope (dE°/dT) of the solid straight line in Figure 1 of
[62VAG/UVA] can be predicted accurately from the quantities selected by NBS and
CODATA [82WAG/EVA] or using NEA-TDB auxiliary data.
[62WIL2]
The dissociation constants of CoSCN+ and NiSCN+ have been determined from UV
spectrophotometric measurements at 25°C. The first set of data was measured at constant ionic strength (I = 0.417 M (Na)ClO4) and was used to calculate the molar absorption coefficients of the MSCN+ complexes. A second set of data, measured at varying
ionic strength, yielded the thermodynamic dissociation constants of thiocyanate complexes. The necessary activity coefficients were calculated from the Davies equation.
ο
The reported K diss
for the NiSCN+ complex is (0.0173 ± 0.0002), which is equivalent to
ο
log10 b1 = (1.762 ± 0.005). Since the activity model used by the author is not equivalent
to the SIT treatment, the above value was not selected. Instead, the reported experimental data were re-evaluated for the purposes of this review. The recalculation of the first
data set (measured at I = 0.417 M (Na)ClO4) resulted in the value of log10 b1 =
(1.21 ± 0.08), and ε (NiSCN+, 273 nm) = (149 ± 5). Using latter value for ε (NiSCN+,
273 nm), log10 b1ο = (1.77 ± 0.10) can be calculated from SIT analysis of the second
data set.
[63BOL/JAU]
The hydrolysis of nickel(II) perchlorate ([Ni2+]t = 2 – 20 mM) in 0.25 M and 1.0 M NaClO4 media has been investigated by potentiometric titration between – log10[H+] = 7
and 7.7 and at temperatures ranging from 25 – 50°C ( ± 0.05°C). Only the formation of
NiOH+,
Ni2+ + H2O U NiOH+ + H+
*
b1,1 ,
was taken into account to explain the experimental data. The values of log10 *b1,1 =
− 9.76, ∆ r Gm (298.15 K) = (13.25 ± 0.10) kcal·mol–1 and ∆ r H m (298.15 K) =
(7.76 ± 1.00) kcal·mol–1 are reported, probably for both ionic strengths used, since no
“sensible” dependence of log10 *b1,1 with ionic strength is reported.
Under the conditions used, a substantial amount of Ni 4 (OH) 44 + would be expected to form; therefore, the reported experimental data have been re-evaluated for the
purpose of this review, taking into account the formation of the mononuclear and
A. Discussion of selected references
306
tetranuclear hydroxo species. Considering the formation constants reported by Baes and
Mesmer [76BAE/MES], the presence of the dinuclear Ni2OH3+ species can be neglected
under the conditions used. The recalculation resulted in the following constants (2σ
uncertainties).
Table A-11: Recalculated constants.
t (°C) / I (M)
log10 *b1,1
log10 *b 4,4
25 / 0.25
30 / 1.0
40 / 1.0
50 / 1.0
– (9.84 ± 0.02)
– (9.74 ± 0.02)
– (9.64 ± 0.06)
– (9.68 ± 0.20)
– (27.34 ± 0.08) – (27.12 ± 0.10) – (26.41 ± 0.20) – (25.26 ± 0.20)
Taking into account the limited experimental data available, the uncertainty
± 0.30 is accepted for all the above values. From the temperature dependence of the
formation constants (see Figure A-11 and Figure A-12) log10 *b1,1 = – (9.73 ± 0.30) and
log10 *b 4,4 = – (27.68 ± 0.40) can be extrapolated for t = 25°C and I = 1.0 M. The temperature dependence of log10 *b1,1 is too small, compared to the experimental error, to
derive a value for the enthalpy of reaction, but ∆ r H m (298.15 K, 1.0 M) = (174 ± 15)
kJ·mol–1 can be obtained for the formation of Ni 4 (OH) 44 + .
Figure A-11: Temperature
[63BOL/JAU].
dependence
of
log10 *b1,1
values
-8.5
log10 *β1,1
-9.0
-9.5
-10.0
-10.5
0.0031
0.0032
K/T
0.0033
derived
from
A. Discussion of selected references
307
Figure A-12: Temperature dependence of log10 *b 4,4 values derived from [63BOL/JAU].
-24
log10 *β4,4
-25
-26
-27
-28
0.0031
0.0032
0.0033
K/T
[63BUR/ABB]
This study is similar to [61LEB/LEV]. The equilibrium of the reaction:
Ni2SiO4 + 2CO U 2Ni + SiO2 + 2CO2
was investigated in the temperature range 700 – 1200°C. Results obtained by Burdese et
al. do not coincide with results of other studies discussed in Section V.7.2.1.1.5 and
shown in Figure V-54. Thus, they were not used for the evaluation of thermodynamic
quantities selected by this review.
[63CHR/IZA]
The thermodynamic formation constant and heat of formation of Ni(CN) 24 − in aqueous
solution have been determined at 25°C potentiometrically and calorimetrically, respectively. No background electrolyte was used, and the Debye-Hückel equation was applied to extrapolate the experimental thermodynamic data to zero ionic strength. Although, the activity model used is not compatible with the SIT, most of the experiments
referred to I < 0.007 M; therefore, the reported value ( ∆ r H mο = – 180.7 kJ·mol–1 at
25ºC) is accepted as valid at I = 0, but the uncertainty is estimated in this review as
± 4.0 kJ·mol–1.
A. Discussion of selected references
308
[63GEO/FER]
The authors measured the heat capacity from 1120 to 1919°C and the heat of nickel
fusion by a calorimetric method. The reported heat of fusion is ∆ fus H mο = (17.47 ± 0.23)
kJ·mol–1.
[63ISA]
The X-ray diffraction patterns of hydrothermally synthesised anhydrous nickel carbonate were taken, indexed and the constants of the unit cell ascertained. The T, p phase
diagram of NiCO3(cr) was determined semi-quantitatively. Solid solution formation
between NiCO3 and MgCO3 was established. In addition hellyerite, NiCO3·5.5H2O, and
zaratite were investigated. The latter turned out not to be a single mineral, but a composite of amorphous and fibrous components.
[63KO/HEP]
Ko and Hepler claimed that S. Friedberg (Department of Physics, Carnegie Inst. Technol., 1961) gave them S mο (NiCl2·6H2O, cr) = 344.3 J·K–1·mol–1 “from unpublished heat
capacity data”. No subsequent article containing these heat capacity data was found.
Hence, no documentation of this value can be provided for.
[63LIN/LAF]
The partial pressure ratio p(H2S)/p(H2) of the gas phase in equilibrium with the hightemperature modification of Ni3S2 was determined as a function of composition at
853.15 and 933.15 K, respectively. The same experimental procedure was performed as
described by Laffitte [59LAF]. The ratios p(H2S)/p(H2) for the coexistence of Ni3+xS2
with metallic nickel at 853.15 and 933.15 K are listed in Table A-12. The experimental
values coincide well with those predicted by the present compilation of thermodynamic
properties of nickel sulphides.
Table A-12: Comparison between experimental data [63LIN/LAF] and calculated values of log10[p(H2S)/p(H2)] for the coexistence of Ni3+xS2 with metallic nickel.
T/K
log10[p(H2S)/p(H2)], exp.
log10[p(H2S)/p(H2)], calc.
853.15
– 3.04
– 3.03
933.15
– 2.90
– 2.79
[63NET/DRO]
The stability constants of NiCl+ and NiBr+ complexes have been determined by a
spectrophotometric method at an ionic strength of 5.7 M (H(X, ClO4), where X = Cl– or
Br−). The concentration of the complex forming anion was varied from 0 to 5.44 M. The
calculation of the formation constants was based on the slight changes of extinction
A. Discussion of selected references
309
coefficients at 395 and 400 nm. The values of the formation constants accepted in this
review at I = 5.7 M for the NiCl+ and NiBr+ complexes are log10 b1 = – (0.5 ± 0.4) and
– (0.3 ± 0.5), respectively.
[63PEN/BAI]
See discussion for [65COL/PET].
[63PHI/HUT]
Phillips et al. investigated the phase equilibria in the system NiO – Al2O3 – SiO2 by
quenching and direct observational techniques. The crystalline phases were identified
by petrographic microscopy and X-ray diffraction methods. The equilibrium diagram of
the system NiO – SiO2 drawn from the experimental data shows that the only intermediate compound is Ni2SiO4, which has the olivine structure. The authors found, that unlike
other olivines which melt congruently, nickel olivine has an upper temperature of stability 1545°C and at temperatures 1545 and 1650°C, NiO and SiO2(cristobalite) coexist in
equilibrium.
[63SHA/DES]
Solutions of unspecified nickel(II) salts ([Ni]tot = 0.029 to 0.0855 M) were titrated with
NaOH solution at (28.0 ± 0.5)°C at an ionic strength of 1.0 M using NaClO4 as the ionic
medium. The pH values of the solutions were determined by unspecified “external electrodes”, which were calibrated using a single buffer solution (pH = 7). The added volume of NaOH solution was measured in “drops”, assuming 20.6 drops per cm3. The
uniform drop-size over the range of 0 – 25 cm3 is questionable. The results were analysed assuming the formation of NiOH+ as the sole hydrolytic species ( log10 *b1,1 =
− (10.03 ± 0.08)), although under the conditions used, a considerable amount of
Ni 4 (OH) 44 + would be expected. Therefore we re-evaluated the reported data, assuming
(i) only the presence of Ni 4 (OH) 44 + ( log10 *b 4,4 = – (27.85 ± 0.3)) and (ii) the formation
of both NiOH+ and Ni 4 (OH) 44 + ( log10 *b1,1 = – (10.03 ± 0.04) and log10 *b 4,4 =
− (30.2 ± 0.7)). Both calculations revealed substantial errors in the measured pH, probably due to the relatively low precision of the experimental work. Therefore, these constants were not considered further in this review. The correction applied by Plyasunova
et al. in [98PLY/ZHA] for the data in [63SHA/DES] resulted in log10 *b1,1 =
− (10.39 ± 0.17), log10 *b1,2 = − (10.4 ± 0.7) and log10 *b 4,4 = – (28.2 ± 1.0). Even these
constants do not provide a better description of the experimental data.
[63TAN/SAI]
This work is a continuation of an earlier study [61TAN/OGI]. Polarographic measurements were carried out at 25°C in an acetate buffer (the assumed “pH” was 3.72), and
concentrations of sulphate and acetate were varied. The sulphate complexation constant
was then obtained by accounting for acetate complex formation. At this higher ionic
strength, the reported nickel sulphate association constant was K1 = 3.7 dm3·mol–1. This
310
A. Discussion of selected references
was determined using an extrapolation procedure that should have minimised any effect
of acetate. The SIT is used with ε(Ni2+, ClO −4 ) equal to (0.37 ± 0.03) kg·mol–1 and
log10 K1ο = 1.95 is calculated, with an estimated uncertainty of ± 0.20.
At the higher acetate concentrations (to 0.5 M) and sulphate concentrations (to
0.20 M) there was evidence for formation of Ni(SO 4 ) 22 − (K = 26 dm6·mol–2),
Ni(OAc)(SO4)– (K = 1 dm6·mol–2) and Ni(OAc)(SO 4 )32− (K = 3 dm9·mol–3). There is
insufficient information to re-analyse the experimental data, and the values of the higher
complexation constants are not accepted in the present review.
[63THR]
In the course of structure refinement Threadgold points out that the water content of
natural hellyerite corresponds to the composition NiCO3(H2O)4·1.5H2O, whereas the
nickel content indicates NiCO3(H2O)4·2H2O [59WIL/THR]. From the packing model of
the structure it was concluded that it is not possible for hellyerite to contain two water
molecules per formula unit in the interlayer water sheet. Consequently, the agreement of
the diffraction pattern with Rossetti-François’ NiCO3(H2O)4·2H2O [52ROS] was considered a problem. It should be emphasised, however, that a new synthesis also resulted
in crystals with the empirical formula NiCO3(H2O)4·1.5H2O [2001GAM/PRE],
[2002WAL/PRE].
[64ALC/BEL]
This work deals with the experimental determination of the oxygen activity in lead
melts by performing emf measurements over the temperature range from 820 to 1120 K.
Galvanic cells of the type Pb(l) | YSZ | NiO, Ni were applied, where the Ni / NiO –
oxygen buffer served as the reference electrode. The Gibbs energy function for the reference electrode was obtained from measurements of the solid-gas equilibrium:
NiO(cr) + CO(g) U Ni(cr) + CO2(g)
provided by Tomlinson and Young (quoted as a private communication). Tomlinson
and Young determined the CO/CO2 ratios in equilibrium with Ni/NiO by chemical
analysis as a function of temperature. They obtained the following Gibbs energy function for the reaction mentioned above: ∆ r Gm (T) = (– (47.15 ± 0.71) + (0.293 ± 0.711) ×
10–3 (T/K)) kJ·mol–1. Based on this equation, Alcock and Belford derived for the Gibbs
energy function of formation for NiO: ∆ f Gm (T) = (– 235.27 + 0.08682 (T/K)) kJ·mol–1,
820 < T/K < 1120, with a standard deviation of ± 1.1 kJ·mol–1.
[64GEE/HUM]
The authors repeated (and extended) the measurements reported by Blackie and Gold in
[59BLA/GOL]. It was found that the solutions, for which the ratio of cyanide to
nickel(II) was between 4 and 8, have the same relaxation time as the solvent (both in
17 M ammonia solution and in water), in sharp contrast with the findings of Blackie and
A. Discussion of selected references
311
Gold [59BLA/GOL]. At a much higher excess of cyanide, a marked change in the UVVIS spectra, between 300 – 520 nm, was observed. This spectral change was used to
determine the stability constant for the reaction:
Ni(CN) 24 − + CN– U Ni(CN)35−
at 23°C and I = 2.5 M Na(ClO4). The concentration of the tetracyanonickelate(II) ion
ranged between 0.4 and 70 mM, while the concentration of cyanide ion varied from
0.05 to 2.5 M. The reported stability constant ( log10 K 5 = – (0.69 ± 0.02)) was not corrected to 25°C, as the difference is assumed to be negligible.
[64GRI/LIB]
Equilibria established in a solution containing a cation (Cu2+, Co2+, Ni2+ or Cd2+) and
chloride have been studied by means of chromatography on ion-exchange papers. Complex formation constants were determined for Ni2+ at 25°C and I = 3 M H(ClO4, Cl).
The concentration of nickel(II) was 0.4 M, while the chloride concentration varied between 0 – 3.0 M. The reported constants are β1 = (0.23 ± 0.03) M–1; β2 = (0.06 ± 0.03)
M–2; β3 = (0.001 ± 0.006) M–3. The authors stated that the results are less accurate then
ones obtained by means other standard methods. Due to the considerable change in the
ionic medium during the measurements, only the formation of the first complex is considered in this review. After recalculation of the experimental data using non-linear regression, a value of log10 b1 = – (0.45 ± 0.50) is accepted.
[64HAM/MOR]
Complexation concentration quotients for 25°C and a medium of 0.1 M (CH3)4NCl(aq)
were reported for the formation of NiP2 O72 − ( log10 K (A.21) = 6.98) and NiHP2 O7−
( log10 K (A.22) = 3.83).
Ni2+ + P2 O74 − U NiP2 O72 −
(A.21)
Ni2+ + HP2 O37− U NiHP2 O7−
(A.22)
The ligand solutions were prepared by titration of the free acid with
tetramethylammonium hydroxide (to a pH value of approximately 10). Aliquots
containing the ligand, metal (as Ni(NO3)2), and tetramethylammonium chloride were
titrated with 0.0685 M HCl(aq), with the primary measurements being carried out at pH
values between 4.5 and 6.5. The electrodes were probably calibrated against pH buffers
rather than concentration standards, and the hydrogen ion activity coefficient, g H+ was
assigned the value of 0.796; the SIT procedure described in Appendix B would have led
to essentially the same value, γ H+ = 0.80.
For Reaction (A.21) in 0.1 M (CH3)4NCl, log10 K (A.21) = 6.98 and
log10 K (A.21) = (8.73 + 0.1 ∆ε).
ο
For Reaction (A.22) in 0.1 M (CH3)4NCl, log10 K (A.22) = 3.83 and
log10 K (A.22) = (5.14 + 0.1 ∆ε).
ο
312
A. Discussion of selected references
As the ionic strength is low, the effect of the ∆ε term on the final calculated
values of log10 K ο is limited. Rather than use estimated ε values for interactions involving the pyrophosphate moieties (some of which are highly charged), it is simply assumed that ∆ε is zero. As a consequence, the uncertainty in the log10 K ο values is increased to ± 0.25 for the values from these measurements in 0.1 M Me4NCl solutions.
[64KOH/ZAS]
Thermal decomposition of hydrated nickel sulphates was carried out as a function of
temperature at 0.026 bar (“20 torr”). Samples were equilibrated at constant temperature
(± 0.1 K) and pressure for at least 24 hours. Equations were reported for water vapour
pressures as a function of temperature for mixtures of hydrates
(NiSO4·7H2O/NiSO4·6H2O; NiSO4·6H2O/NiSO4·4H2O; NiSO4·4H2O/NiSO4·2H2O;
NiSO4·2H2O/NiSO4·1H2O), and for a saturated solution (18 to 59°C). Rehydration of
the monohydrate at 60°C with water vapour pressure of 0.078 bar was reported to generate a mixture of the mono-, di- and tetrahydrates. Decomposition of the monohydrate
led to an amorphous solid, and vapour pressure measurements over mixtures of the anhydrous solid and the monohydrate were not reproducible.
The authors indicated that they did not feel it was useful to use their empirical
equations to calculate dissociation enthalpies and entropies, and values calculated from
the equations are not used in the present review.
[64KOS/KAL]
The temperature functions of the heat capacities of hydrothermally synthesised calcite
type carbonates MnCO3, FeCO3, CoCO3, and NiCO3 were investigated between 70 and
300 K. The respective standard entropies were calculated by the integration of C p ,m /T
versus T curves.
[64LAR]
The author used infrared spectroscopic results to conclude that only about 10% of the
nickel sulphate ion pairs are inner-sphere complexes in 0.3 M and 1 M NiSO4. The derived association constants are useful for comparisons for the different di- and trivalent
systems studied, but are not suitable for calculating a selected value in the present review.
[64LEE/ROS]
The sulphur pressures (sulphur activities) for the equilibrium between NiS2 and Ni1–xS
were studied at temperatures ranging from 673 to 873 K. Pure NiS2 was prepared by
solid state reaction between sulphur and NiS. The solid product was checked by X-ray
diffraction analysis before and after the decomposition experiments. The equilibrium
sulphur pressure for the decomposition of NiS2 into Ni1–xS and S2(g) was measured by
means of a Rodebush manometer [30ROD/HEN].
A. Discussion of selected references
313
[64PER]
The first hydrolysis constant ( *K1 ≡ *b1,1 ) of nickel(II) was determined at 20°C by potentiometric titrations using various ionic strength and nickel(II) nitrate concentrations:
0.0016 M ≤ I ≤ 1.503 M, 0.00050 M ≤ [Ni(II)]tot ≤ 0.01000 M. In another set of titrations, the temperature was varied from 15 to 42°C, applying three different ionic
strengths: 0.0016 ([Ni(II)]tot = 0.0005 M), 0.0130 and 0.0430 M ([Ni(II)]tot = 0.001 M).
In all titrations, KNO3 was used to adjust the ionic strength. The author assumed that the
only important hydrolysed species was NiOH+. For the experiments made at [Ni(II)]tot ≤
0.001 M, the formation of Ni 4 (OH) 44 + (and Ni2(OH)3+) can, indeed, be ignored, in the
pH range of the experiments. The formation of NiNO3+ was not considered by the author, although at 20°C, where a wide range of ionic strength (I = 0.0016 – 1.5) was
used, its concentration can not be neglected. Therefore, for the purposes of this review,
the reported constants were corrected for the formation of NiNO3+ , and the SIT approach was used to derive log10 *b1,1ο (see Figure A-11, page 306). From the plot in
Figure A-13, for the reaction:
Ni2+ + H2O U NiOH+ + H+
(A.23)
log10 *b1,1ο = – (10.07 ± 0.07) and ∆ε = – (0.38 ± 0.10) kg·mol–1 are obtained. From the
latter value and ε(Ni2+, NO3− ) = (0.182 ± 0.010) kg·mol–1, ε(NiOH+, NO3− ) =
− (0.27 ± 0.12) kg·mol–1 can be derived.
Figure A-13: Extrapolation to I = 0 of the equilibrium constants log10 *b1,1 (A.23) reported in [64PER] at 20°C, using the SIT.
-8.8
log10*β1,1 + 2D
-9.2
-9.6
-10
-10.4
-10.8
0
0.5
1
1.5
−1
I / mol·kg
2
A. Discussion of selected references
314
At the other temperatures, only a few percent of NiNO3+ can be expected (Imax
= 0.046 M), thus no correction was applied, but the SIT was used to derive log10 *b1,1ο .
The reported log10 *b1,1ο values and those recalculated using the SIT are collected in
Table A-13.
Table A-13: Reported and recalculated log10 *b1,1ο .
t (°C)
15
20
(reported)
– 10.22
– 10.05
– 9.86
– 9.75
– 9.58
– 9.43
log10 *b1,1ο (recalculated) – 10.22
– 10.07
– 9.80
– 9.74
– 9.59
– 9.38
*
log10 b
ο
1,1
25
30
36
42
From the plot of log10 *b1,1ο vs. 1/T (see Figure A-14) ∆ r H mο (A.23) =
(51.7 ± 3.4) kJ·mol–1 can be derived.
Figure A-14: Plot of log10 *b1,1ο (A.23) recalculated from [64PER] vs. 1/T.
-21
-21.5
ln *β
o
1,1
-22
-22.5
-23
-23.5
-24
0.00315
0.0032
0.00325
0.0033
0.00335
K/T
0.0034
0.00345
0.0035
A. Discussion of selected references
315
[64TAY/SCH]
This is one of the pioneering studies of thermodynamic data by emf-measurements with
a solid electrolyte galvanic cell. Zirconia, doped with CaO and MgO, was used as an
electrolyte to determine the free energy of the reaction:
Ni + 0.5 SiO2 + 0.5O2 U 0.5Ni2SiO4,
and the free energy of formation of numerous silicates and titanates at temperatures
from 750 to 1200°C. The data of Taylor and Schmalzried coincide well with the newest
emf-study of the above reaction and were used by this study to determine the Gibbs
energy of formation of Ni2SiO4.
[64TRI/CAL]
The formation constants of thiocyanate complexes of Ni(II), Cd(II) and Ni(II) were determined by an extraction method at 20°C. The aqueous media were 1.5 and 3.0 M
(Na)ClO4 solutions. The authors used rather acidic conditions (pH = 0.5 to 1.0), as only
the protonated form of the ligand (HSCN) can be extracted into the organic phase
(Ni(SCN)2(aq) is reported to be weakly extractable). The method used to determine the
concentration of the extracted HSCN was not reported. The concentration of the thiocyanate ion ranged from 0.01 to 0.55 M, while [Ni2+]tot was ~ 0.1 M. The experimental
data were described by the formation of four Ni(SCN) 2x − x (x = 1 – 4) complexes. For
both ionic strengths used, K3 < K4 is reported. Considering that no geometrical change
of nickel(II) occurs during the stepwise complex formation [90BJE2], the reported K4
values are probably overestimated. Using the accepted enthalpy values for the formation
of Ni(SCN) 2x − x (x = 1 – 3) species, the reported formation constants were corrected to
298.15 K and the uncertainties were increased: log10 b1 = (1.10 ± 0.10), log10 b 2 =
(1.69 ± 0.10), log10 b 3 = (1.62 ± 0.20) (I = 1.5 M), and log10 b1 = (1.15 ± 0.10),
log10 b 2 = (1.61 ± 0.10), log10 b 3 = (1.21 ± 0.30) (I = 3.0 M).
[64WEL/KEL]
Low-temperature heat capacities were determined for NiS (millerite) and Ni3S2
(heazlewoodite) between 50 and 298.15 K. The solid compounds were obtained by reaction of certain amounts of nickel oxide with sulphur. All products were characterised
by X-ray diffraction analysis. Standard entropy values at 298.15 K were derived by integration of the respective heat capacity functions. In the case of millerite, the standard
entropy is equal to S mο (NiS, β, 298.15 K) = (52.97 ± 0.33) J·K–1·mol–1 and the heat
capacity at 298.15 K is C pο,m (NiS, β, 298.15 K) = 47.11 J·K–1·mol–1, which agrees well
with the result published by Grønvold and Stølen [95GRO/STO], C pο,m (NiS, β,
298.15 K) = 47.08 J·K–1·mol–1. The standard entropy of Ni3S2 was found to be
S mο (Ni3S2, cr, 298.15 K) = (133.89 ± 0.84) J·K–1·mol–1 which is in close agreement with
the value given by Stølen et al. [91STO/GRO] ( S mο (Ni3S2, cr, 298.15 K) = 133.2
J·K−1·mol–1). Weller and Kelley [64WEL/KEL] obtained C pο,m (Ni3S2, cr, 298.15 K) =
A. Discussion of selected references
316
117.65 J·K−1·mol–1, which is in excellent agreement with the values recommended by
Ferrante and Gokcen [82FER/GOK], C pο,m (Ni3S2, cr, 298.15 K) = 117.7 J·K–1·mol–1, and
Stølen et al. [91STO/GRO], C pο,m (Ni3S2, cr, 298.15 K) = 118.2 J·K–1·mol–1.
[65ADA/KIN]
Adami and King reported heats of solution in 4.360 m HCl(aq) at 303.15 K. The thermodynamic cycle used heats for the following reactions:
H2SO4·7.068H2O(l) U 2H+(sln) + SO 24 − (sln) + 7.068H2O(sln)
Ni(cr) + 2H+(sln, H2(sat)) U Ni2+(sln, H2(sat)) + H2(g)
7.068H2O(l) U 7.068H2O(sln)
NiSO4(cr) U Ni2+(sln) + SO 24 − (sln)
to obtain a value of (3.724 ± 0.460) kJ·mol–1 for the heat of the reaction:
Ni(cr) + H2SO4·7.068H2O(l) U NiSO4(cr) + 7.068H2O(l) +H2(g)
(A.24)
at 303.15 K. The authors corrected the value for the heat of Reaction (A.24) to
(3.598 ± 0.460) kJ·mol–1, at 298.15 K.
The reported experimental uncertainties are 1σ uncertainties. Also, the authors
estimated the uncertainly in the reaction involving dissolution of nickel with evolution
of hydrogen as ± 0.3 kJ·mol–1, after applying a correction to account for heat absorbed
in vaporising the water and hydrogen chloride that accompanied the evolution of hydrogen. Solution calorimetry involving reactions resulting in gas evolution is notoriously
difficult, and in the present assessment the 1σ uncertainty has been doubled. Based on
this, a value of (3.724 ± 1.400) kJ·mol–1 is calculated for the heat of the Reaction (A.24)
at 303.15 K (2σ uncertainty).
The difference in the values of ∆ r C pο,m for 298.15 K and of ∆ r C pο,m for the average reaction temperature, 300.65 K, are small enough that they can be ignored relative
to the uncertainties in the values of C pο,m . Using the values at 298.15 K, C pο,m (NiSO4, cr,
298.15 K) = (97.63 ± 0.10) J·K–1·mol–1 (see Section V.5.1.2.2.1.3, and [78STU/FER]),
C pο,m (Ni, cr, 298.15 K) = (26.07 ± 0.10) J·K–1·mol–1 (see Section V.1.2), C pο,m (H2, g,
298.15 K) = (28.836 ± 0.002) J·K–1·mol–1 (cf. Table IV-1), C pο,m (H2O, l, 298.15 K) =
(75.351 ± 0.080) J·K–1·mol–1 (cf. Table IV-1) and C pο,m (H2SO4·7.068H2O, l, 298.15 K) =
(614 ± 10) J·K–1·mol–1 [56HOR/BRA], [82WAG/EVA], we obtain ∆ r C pο,m = (20 ± 10)
J·K–1·mol–1, ∆( ∆ r H mο ) = (0.10 ± 0.10) kJ·mol–1, and ∆ r H mο = (3.59 ± 1.40) kJ·mol–1.
Then, using ∆ f H mο (H2O, l, 298.15 K) = – 285.83 kJ·mol–1 and ∆ f H mο (H2SO4·7.068H2O,
l, 298.15 K) = – 2897.05 kJ·mol–1 [82WAG/EVA], ∆ f H mο (NiSO4, cr, 298.15 K) can be
calculated. The value for ∆ f H mο (H2SO4·7.068H2O, l, 298.15 K) is adjusted to
− (2897.12 ± 0.64) kJ·mol–1 to allow for the differences in the values for ∆ f H mο ( SO 24 − ,
298.15 K) from Table IV-1 and from the US NBS table [82WAG/EVA]; the uncertainty
A. Discussion of selected references
317
is estimated to incorporate the uncertainty for ∆ f H mο for the sulphate ion and the “hydration” to H2SO4·7.068H2O.
From these values, ∆ f H mο (NiSO4, cr, 298.15 K) is calculated to be
− (873.28 ± 1.57) kJ·mol–1. The difference between this number and the value originally
reported by Adami and King is related primarily to the different values used for
∆ f H mο (H2SO4, l, 298.15 K). Adami and King used values from the 1952 US NBS Circular 500 [52ROS/WAG].
[65AKI/FUJ]
In this paper the olivine – spinel transition in Fe2SiO4 and Ni2SiO4 have been investigated over the temperature range 700 to 1500°C in the pressure range 20 to 40 kilobars
for Ni2SiO4 using a tetrahedral-anvil type of high-pressure apparatus. For both silicates
the transition has proved to be reversible. The equation for the transition curve for
Ni2SiO4 was determined to be p = 19 + 0.012 t, where p is in kilobars and t is in degrees
centigrade. On the basis of this equation, the standard Gibbs energy of formation of
spinel from the olivine was calculated to be (1200 + 0.98 T) cal·mol–1 for Ni2SiO4,
where T is absolute temperature.
[65BUR/LIL]
The hydrolysis of nickel(II) was studied at (25.0 ± 0.1)°C in 3 M (Na)ClO4 medium by
potentiometric titrations using glass electrodes. The concentration of nickel(II) ranged
from 0.1 to 0.8 M. Since [ClO −4 ] was kept at 3 M during the measurements, the ionic
strength ranged from 3.1 to 3.8 M. This is a common problem of investigations using
rather concentrated nickel(II) solutions. Nevertheless, this is one of the highest quality
papers concerning the hydrolysis of nickel(II). Both graphical and computer evaluation
of the experimental data indicated that the major species is Ni 4 (OH) 44 + , under the experimental conditions used. The remaining small deviations between the observed and
calculated Z values (number of OH– bound per nickel(II)) were explained by small analytical errors and the presence of Ni2OH3+ as a minor complex. The optimum fit resulted
log10 *b 4,4 = – (27.37 ± 0.02 (3σ)) and log10 *b1,2 = – (10.0 ± 0.5). The authors listed
some additional minor species, remarking that the experimental data give no support for
their existence.
Due to the relatively important change of ionic strength (and thus activity coefficients) between the measurements at different nickel(II) concentrations, we reevaluated the reported experimental data referring to both [Ni2+]tot = 0.1 to 0.4 M and
[Ni2+]tot = 0.1 to 0.8 M. Using the whole dataset, and the model proposed by the authors,
log10 *b 4,4 = – (27.37 ± 0.04) (3σ) and log10 *b1,2 = – (9.42 ± 0.10) were obtained. The
agreement between the recalculated and original values is excellent for the main hydrolytic species, and reasonable for the minor complex. The difference is very likely due to
the more sophisticated data analysis software now available (for all calculations in this
chapter the PSEQUAD program was used [91ZEK/NAG]). Introducing the species
318
A. Discussion of selected references
NiOH+ into the equilibrium model, a substantial (40%) decrease of the fitting parameter
was obtained. From the data referring to [Ni2+]tot = 0.1 – 0.4 M, log10 *b 4,4 =
− (27.43 ± 0.04) (3σ), log10 *b1,1 = – (9.86 ± 0.12) and log10 *b1,2 = – (9.87 ± 0.50), using the whole dataset log10 *b 4,4 = – (27.42 ± 0.03), log10 *b1,1 = – (9.89 ± 0.06) and
log10 *b1,2 = − (9.78 ± 0.16) were calculated. It is important to note that the authors also
tried to include NiOH+ in the equilibrium model, but their calculations gave no real
support for its existence. Taking into account the above values, log10 *b 4,4 =
− (27.43 ± 0.05) (3σ), log10 *b1,1 = – (9.86 ± 0.12) and log10 *b1,2 = – (9.8 ± 0.5) are accepted for use in this review.
[65BUR/LIL2]
The hydrolysis of nickel(II) was studied at (25.0 ± 0.1)°C in 3 M (Na)Cl medium by a
method identical to the one described in [65BUR/LIL]. The total nickel(II) concentration was varied between 0.2 and 1.4 M, which resulted in a change of ionic strength
from 3.2 to 4.4 M. The analysis of experimental data indicated the presence of
Ni 4 (OH) 44 + as the dominant species ( log10 *b 4,4 = – (28.42 ± 0.05)), while the deviations at lower pH values suggested the presence of Ni2OH3+ ( log10 *b1,2 = – (9.3 ± 0.2)).
For log10 *b1,1 a value of – 10.0 was obtained from the calculations, but was rejected as
too uncertain. During these measurements, a large part of the Na+ ion concentration was
replaced by Ni2+, consequently the constancy of activity coefficients at different
nickel(II) concentrations is questionable. Furthermore, the formation of chloro complexes (both binary and ternary) was not considered.
For the purpose of this review, an attempt has been made to minimise the influence of the medium effect (using only the data reported for [Ni2+]tot = 0.2 and 0.4 M)
and to take into account the presence of NiCl+ complex. For the reaction:
Ni2+ + Cl– U NiCl+
(A.25)
log10 b1 = – 0.59 can be calculated at Im = 3.2 m in NaCl medium, using the recommended value for log10 b1ο (A.25) and ε(NiCl+, Cl–) = ½{ε(Ni2+, Cl–) +
ε(Na+, Cl–)} = (0.1 ± 0.03) kg·mol–1.
Considering the whole data set and the same model as proposed, log10 *b 4,4 =
− (28.32 ± 0.09) (3σ) and log10 *b1,2 = – (9.41 ± 0.13) can be calculated, which are in
reasonable agreement with the published values (see also the discussion of
[65BUR/LIL]). Using only the data sets reported for [Ni2+]tot = 0.2 and 0.4 M, and taking into account the formation of NiCl+, log10 *b 4,4 = – (27.52 ± 0.10 (3σ)) and
log10 *b1,2 = – (8.55 ± 0.06) can be derived. Introducing the species NiOH+ into the
equilibrium model, a 50% decrease in the fitting parameter was found, and log10 *b 4,4 =
– (27.55 ± 0.06) (3σ), log10 *b1,1 = – (9.56 ± 0.10) and log10 *b1,2 = − (8.95 ± 0.20) were
obtained. These values are accepted for use in this review, with increased uncertainties:
log10 *b 4,4 = – (27.55 ± 0.20), log10 *b1,1 = – (9.56 ± 0.20) and log10 *b1,2 =
− (8.95 ± 0.40).
A. Discussion of selected references
319
[65COL/PET]
Penneman and his co-workers published three consecutive papers concerning formation
of the Ni(CN) 24 − ion [60MCC/JON], [63PEN/BAI], [65COL/PET]. In the first paper
[60MCC/JON], the complexes formed in aqueous solutions of Ni(CN) 24 − and cyanide
ions were studied by IR and spectrophotometric methods. The results indicated the formation of only one new species ( Ni(CN)35− ), even at a very high excess of cyanide concentrations. The magnetic measurements indicated diamagnetic behaviour of the pentacyano species, which suggests square pyramid geometry around the metal ion. From the
spectrophotometric data, measured at 430, 450 and 470 nm, log10 K 5 = − (0.72 ± 0.02)
was determined at 25.2°C and I = 1.34 M (NaClO4). From the IR measurements,
log10 K 5 = – (0.52 ± 0.20) was obtained. Identical spectrophotometric studies at 15.4
and 33.6°C have been performed. From the temperature dependence of log10 K 5 (see
Figure A-15) ∆ r H m = – (10.4 ± 0.2) kJ·mol–1 can be obtained. The uncertainty in this
value does not include systematic uncertainties, nor the uncertainty in determining an
enthalpy of reaction from equilibrium data over a small temperature range. In the present review, we assign an uncertainty of ± 4.0 kJ·mol–1 to this value.
While [60MCC/JON] was in press, two papers by Blackie and Gold
[59BLA/GOL], [59BLA/GOL2] appeared, with substantially different conclusions,
therefore Penneman et al. reinvestigated the Ni(CN) 24 − – CN– system [63PEN/BAI],
using IR spectroscopy at 25°C and I = 4 M NaClO4. The authors suggested the formation of two new species Ni(CN)35− and Ni(CN)64 − . The latter complex was taken into
account to explain some deviations at very high cyanide concentrations ([CN–] > 2.5
M). The values of log10 K 5 = – (0.55 ± 0.02) and log10 K 6 = – (1.0 ± 1.0) have been reported. Due to the substantial medium effect, only formation of the pentacyano complex
is considered in this review, and its formation constant is accepted with an increased
uncertainty (± 0.30). In a further experiment, the authors found that the addition of NaCl
to the solution of 2 M NaCN and 0.075 M Ni(CN) 24 − (the main species is Ni(CN)35− )
resulted in a significant decrease in the peak height attributed to Ni(CN)35− . This observation was explained as formation of Ni(CN)5Cl4– ( log10 K = – 0.66).
Two years later, Coleman et al. in [65COL/PET] tried to solve the discrepancy
between their two earlier papers [60MCC/JON] and [63PEN/BAI]. In this third paper
the authors again used IR and spectrophotometric methods to study the Ni(CN) 24 − – CN–
system in a 4 m KF – KCN mixture at 25°C (30°C for IR). They showed that, under the
conditions used, the formation of a single species, Ni(CN)35− , is sufficient to account for
the experimental data. Based on the spectrophotometric data, a value of log10 K 5 =
(0.03 ± 0.01) (in molal units) has been reported. From the IR measurements at 30°C,
log10 K 5 = – (0.08 ± 0.06) was derived. In light of these findings, the authors explained
their earlier results reported in [63PEN/BAI] (the formation of Ni(CN)64 − and
Ni(CN)5Cl4–) by the neglected medium effect caused by the substantial change of
NaCN/NaClO4 (NaCl/NaClO4) ratio during the measurements.
A. Discussion of selected references
320
Figure A-15: Temperature dependence of the association constant for the reaction
Ni(CN) 24 − + CN– U Ni(CN)35− reported in [60MCC/JON].
-0.6
log10 K5
-0.7
-0.8
-0.9
0.0032
0.0033
0.0034
0.0035
K/T
[65EGO/SMI]
This is a very careful study of the specific heat capacity of Ni2SiO4 and Zn2SiO4, the
only one for the nickel orthosilicate for temperatures as high as 1570 K. Egorov and
Smirnova used a copper block calorimeter and 99 NiCr / Constantan thermocouples for
the measurement of the temperature change of calorimeter. The average, least square,
precision of measurements performed in the range between 555 – 293 K and 1570 –
293 K was ± 0.4. This high temperature data were used in this review to fit the heat
capacity equation up to 1570 K.
[65LIN/SEI]
According to Linke and Seidell’s critical evaluation of the nickel sulphate solubility in
water the stable phase from the eutectic temperature to 31°C is NiSO4·7H2O. The reported solubilities differ by 1 – 3%.
[65MOR/REE]
The formation of nickel(II) chloro complexes in aqueous solution has been studied using a cation-exchange method at 20°C, in a solution of 0.691 M H(ClO4,Cl). The distri-
A. Discussion of selected references
321
bution of nickel at equilibrium between the resin (Amberlite IR-120) and the solution
has been determined spectrophotometrically. The reported formation constants are β1 =
(1.7 ± 0.1) M–1; β2 = (0.92 ± 0.10) M–2. Due to the almost complete change of the ionic
medium during the measurements ([Cl–] = 0 to 0.659 M) the real formation of the biscomplex is doubtful. As the presented data do not allow an independent recalculation,
they were not considered in this review.
[65PAO]
The heat of solution of NiBr2 in water was reported as – (82.09 ± 0.17) kJ·mol–1 based
on limited measurements and analysis. The solid was obtained by dehydration over sulphuric acid in a vacuum desiccator, and the Ni and Br analyses were within 0.2% of
theoretical. The number of samples measured was not reported, but some values in the
same paper were based on only two measurements. Neither the dilution of the final
solution nor the sample size was given. The enthalpy value probably was not corrected
for dilution, as there is no description of a correction procedure. The uncertainty
reported was probably 1σ. The resulting value is approximately 4 kJ·mol–1 less
exothermic than that of Efimov and Furkalyuk [90EFI/FUR] (although the final
solutions probably were somewhat different).
[65PAO/SAB]
A value of – (70.46 ± 0.33) kJ·mol–1 was determined for the enthalpy of solution of
NiI2(cr) in water containing (n–Pr)4NI (0.0005 m) and a “small quantity” of HI. The
final solution concentrations were approximately 0.00025 m in nickel. The solid was
obtained by dehydration over sulphuric acid in a vacuum desiccator, and the Ni and I
analyses were within 0.2% of theoretical. The reported value was an average of “at least
two determinations”. The authors concluded from similar measurements on the dissolution of ZnI2(cr) that the effects of association were small; yet, their own measurements
suggest that the uncertainties from this source could be greater than 1 kJ·mol–1. The resulting value is less exothermic than that of Efimov and Furkalyuk [90EFI/FUR],
though the final solutions also are not identical.
[65PAW/STA]
The specific heat capacity of high-purity nickel and several nickel-rich alloys from
room temperature to 600°C is reported. The data, obtained with an adiabatic calorimeter, were reproducible to within ± 0.25 per cent and have an estimated absolute error of
less than ± 0.5 per cent. This high quality data were credited by this review and used for
the determination of the heat capacity function.
[65POT]
This reports a study of the complex dielectric constants of 1 M NiSO4 in aqueous solution as a function of temperature.
A. Discussion of selected references
322
[65WEL]
Heat capacity measurements, 52.0 to 296.3 K, were reported for anhydrous NiSO4(cr).
A maximum entropy for 298.15 K was estimated by extrapolation to 0 K and incorporation of an estimated magnetic contribution. These heat capacity results were incorporated with results at lower temperatures in the later analysis of Stuve et al.
[78STU/FER].
[66BOD/DEH]
Bode et al. investigated the reactions and reaction products on the nickel oxide electrode of the alkaline storage batteries. The following aspect of Bode et al.’s model is
well established by now: the charged (NiOOH) as well as the discharged (Ni(OH)2)
material of nickel oxide hydroxide electrodes can exist in two forms. In this paper the
synthesis and structure of α-Ni(OH)2 has been described and its formula
3Ni(OH)2·2H2O was determined. This nickel hydroxide phase consists of brucite type
layers of Ni(OH)2 with larger layer distances than those of β-Ni(OH)2. This is a result of
interlayers of water. The definitive characterisation of its structure was not possible,
because no single crystals could be grown.
[66BUR/IVA]
The hydrolysis of nickel(II) has been studied at (25.0 ± 0.1)°C in 3 M (Na)NO3 medium
by a method identical to the procedure described in [65BUR/LIL]. The nickel(II) concentration ranged from 0.2 to 1.0 M, which resulted in a change of ionic strength from
3.2 M to 4.0 M. The analysis of experimental data indicated the presence of Ni3 (OH)33+
( log10 *b 3,3 = – (21.58 ± 0.03)) as major and Ni2OH3+ ( log10 *b1,2 = − (9.6 ± 0.2)) as
minor complexes. Since no spectroscopic evidence was found up to 4 M nitrate concentration, for the formation of the species NiNO3+ , the formation of binary or ternary nitrato complexes was not considered by the authors. In fact, a substantial amount of nitrato complex is present at 3 M nitrate concentration, therefore the reported constants
are underestimated. In any other medium (NaClO4, NaCl, KCl, NaBr), the tetramer is
reported to dominate, and the formation of a trimer has never been confirmed by others;
therefore, we re-evaluated the reported experimental data. Considering the suggested
hydroxo complexes and neglecting the formation of NiNO3+ , log10 *b 3,3 =
− (21.61 ± 0.03) and log10 *b1,2 = – (9.32 ± 0.11) can be calculated, in good agreement
with the reported values. However, an equally good fit to the experimental data can be
achieved by assuming the presence of Ni 4 (OH) 44 + , NiOH+ and Ni2OH3+ ( log10 *b 4,4 =
− (28.00 ± 0.04), log10 *b1,1 = − (9.70 ± 0.11) and log10 *b1,2 = − (9.28 ± 0.2)). An attempt was made to consider the formation of NiNO3+ species. However, the tentatively
accepted equilibrium constant for the reaction,
Ni2+ + NO3− U NiNO3+ ,
(A.26)
log10 b ((A.26), 298.15 K) = (0.5 ± 1.0)) represents only the higher limit of the ‘true’
constant due to the neglected medium effect. Using the SIT approach, log10 b1 (A.26),
ο
1
A. Discussion of selected references
323
298.15 K) = − (0.3 ± 1.0) can be calculated for 3.31 m NaNO3 medium. Accepting this
value, log10 *b 4,4 = – (26.52 ± 0.05), log10 *b1,1 = – (9.14 ± 0.12) and log10 *b1,2 =
− (8.77 ± 0.25) were obtained. These constants are much higher than the values of
[65BUR/LIL] and [65BUR/LIL2], which is very likely the consequence of the overestimated formation constants of NiNO3+ . Fixing the value of log10 *b 4,4 at – 27.4,
log10 *b1,1 = – (9.49 ± 0.15), log10 *b1,2 = – (9.02 ± 0.15) and log10 b1 ((A.26), 298.15 K)
= – (0.84 ± 0.04) can be calculated. Considering the great uncertainty of
log10 b1ο ((A.26), 298.15 K), none of the above values are accepted for use in this review.
[66FLO]
The formation and stability of the NiCl(H 2 O)5+ complex have been studied by a spectrophotometric method, in 10 M Li(NO3, Cl) medium at 303 K, to explain the polarographic behaviour of nickel(II) in concentrated chloride media. The concentration of
LiCl was varied between zero and 10.0 M. This fundamental change in the ionic medium results in a large uncertainty in the reported formation constant (β1 =
(0.094 ± 0.009)), although the author noted that the absorption spectra of 0.03 M
nickel(II) in 10 M LiCl, 10 M HCl, 5 M CaCl2 and 5 M MgCl2 are almost identical.
This indicates a relatively minor impact of the medium effect. Therefore, log10 b1 =
− (1.0 ± 0.5) is accepted in this review. The ionic strength applied is too high to extrapolate the reported constant to infinite dilution using the SIT. In addition, LiNO3 can not
be regarded as an ‘innocent’ background electrolyte, since nickel(II) forms an ion pair
with the nitrate ion.
An attempt to determine the stability constants of the nickel bromo complexes
was complicated by the slow formation of bromine from oxidation of bromide by the
nitrate ion. However, the results indicated that the bromo and chloro complexes of
nickel(II) have similar stability.
[66GOL/RID]
In this paper, heats of solution in water at 298.15 K were reported for
α-NiSO4·6.000H2O(s) (final solution compositions 0.00009 m to 0.051 m) and for a
mixed hydrate with a nominal composition of NiSO4·5.059H2O (final solution compositions 0.02 m). The mixed solid was assumed to be a mixture of NiSO4·6.000H2O(s) and
NiSO4·H2O(s), and the enthalpy of hydration of NiSO4·H2O(s) to the hexahydrate was
calculated. The standard enthalpies of solution for the solids were calculated using heat
of dilution values reported by Lange [59LAN], and Lange and Miederer [56LAN/MIE].
The enthalpy of solution of NiSO4·6.000H2O(s) in water was reported as
∆ sol H mο ((A.27), 298.15 K) = 1.15 kcal·mol–1 (4.81 kJ·mol–1).
NiSO4·6H2O(cr) U Ni2+ + SO 24 − + 6H2O(l)
(A.27)
A. Discussion of selected references
324
The largest uncertainty in this value comes from the extrapolation to I = 0.
Measurements by Lange and co-workers [56LAN/MIE] on CdSO4, CuSO4, ZnSO4, and
NiSO4 lead to values for the integral heats of dilution that approach I = 0 with a slope
almost twice the expected Debye-Hückel limiting slope, AL. If an extended DebyeHückel treatment (such as the SIT [97ALL/BAN]) is used to describe the behaviour of a
strong 2:2 electrolyte, the electrolyte will be found to behave as if it is associated (as we
have noted for NiSO4). However, because higher order electrostatic terms have been
omitted, the extended Debye-Hückel theory is inadequate for extrapolation of enthalpies
of dilution and heat capacity values to I = 0 [97MAL/ZAM], [98ARC/RAR]. Thus, the
large slope found by Lange and Miederer [56LAN/MIE] is not unexpected.
In Figure A-16 the original data are plotted versus the square root of the ionic
strength. The experimental enthalpies of dilution for NiSO4 reported by Lange and
Miederer [56LAN/MIE] were subtracted from ∆ sol H m values of [66GOL/RID], resulting in values of ∆ sol H mο . The mean value obtained from Equation (A.28)
∆ sol H mο (A.27) = ∆ sol H m (A.27) – φL
ο
m
(A.28)
–1
is ∆ sol H (A.27) = (4.485 ± 0.200) kJ·mol , and this value is accepted in the present
review. For comparison φL was calculated with Equation (IX.44) of [97ALL/BAN],
where the last term had to be neglected because the temperature derivative of
ε(Ni2+, SO 24 − ) is unknown. Figure A-16 shows that relying on the Debye-Hückel limiting law leads to an error of ~ 1.7 kJ·mol–1 in the enthalpy of solution of NiSO4·6H2O(cr)
at infinite dilution.
It seems likely that the NiSO4·5.059H2O was a mixture of NiSO4·6H2O(s) and
NiSO4·4H2O(s) rather than a mixture of the hexahydrate and monohydrate. Jamieson et
al. [65JAM/BRO], using samples dehydrated at 100°C, noted changes in dissolution
kinetics and enthalpies for hydrated nickel sulphates with less than 29% water (the water content of the tetrahydrate would be 31%). They also showed that vacuum dehydration at 40°C gave products with a different heat of hydration per mole of water of
hydration. Goldberg et al. [66GOL/RID] reported similar difficulties in dissolving more
highly dehydrated samples, and presented no proof that dehydration of the hexahydrate
over P2O5 led to formation of the monohydrate.
A. Discussion of selected references
325
Figure A-16: Molar enthalpy of solution of NiSO4·6H2O(cr) at 25°C. Experimental data
from Goldberg et al. [66GOL/RID] ○; dotted curve: calculated according to Equation
(IX.44) in [97ALL/BAN]; dash dotted curve: calculated according to the chord method,
see Equations (8 – 2 – 19) and (8 – 2 – 20) of Harned and Owen [58HAR/OWE]; solid
curve: calculated with an adjustable Debye-Hückel parameter, AL, and a term linear in I,
∆ sol H mο ((A.27), 298.15 K) derived from dilution enthalpies [56LAN/MIE] ×; mean
value of ∆ sol H mο ((A.27), 298.15 K) ¹, dash line.
9
∆solHm / kJ·mol
−1
8
7
6
5
4
0.0
0.1
0.2
0.3
−1
(I / mol·kg )
0.4
0.5
0.5
[66KEN/LIS]
This paper reports calorimetric results for the heats of formation of (among others)
NiCl+ and NiBr+ complexes at an ionic strength of 2.0 at 298 and 313 K. In a typical
experiment 350 cm3 of the nickel(II) perchlorate solution (0.67 M) was mixed with 50
cm3 2.0 M sodium halide solution. In the first step, values of the stability constants of
the complexes formed were taken from two earlier papers from Lister's laboratory
[60LIS/ROS] and [61LIS/WIL]. In [60LIS/ROS] the authors measured (among others)
the emf of the cell Ag,AgCl | NaCl+M(ClO4)2+NaClO4 | NaCl+NaClO4 | AgCl,Ag
(where M = Ni(II) or Co(II), [M] = 0.12 – 0.31 M and [Cl–] = 0.008 M) at I = 2 M
(Na,M)ClO4 and at 285, 298 and 313 K, to determine the composition and stability of
the nickel(II) complexes formed with chloride ion. The complex formation induced 1-6
mV differences between the electrodes. The experimental data were interpreted by tak-
326
A. Discussion of selected references
ing into account the formation of two complexes, NiCl+ (K1) and Ni2Cl3+(K2). The latter
complex was considered to explain a small drift in the K1 value. The experimental data
reported in this earlier paper were reinterpreted in this paper by Kennedy and Lister
[66KEN/LIS], as the temperature dependence of K1 and K2 determined earlier was not
consistent with their calorimetric results. Taking into consideration only the formation
of NiCl+, a coherent model can be given. Therefore, the authors recalculated the earlier
reported K1 values for 298 and 313 K neglecting the formation of Ni2Cl3+. For the purpose of this review we re-evaluated the experimental data reported in [60LIS/ROS] for
all the three temperatures. This calculation resulted in only slightly different values for
298 and 313 K than were tabulated in [66KEN/LIS]. The following formation constants
are accepted in this review: log10 b1 = – (0.21 ± 0.20) (285 K), – (0.16 ± 0.20) (298 K)
and – (0.14 ± 0.20) (313 K). It is worth mentioning that the impact of the medium effect
on these constants, induced by the relatively high concentrations of Ni(ClO4)2
(0.12 − 0.31 M) used, is probably considerably lower than for most of the studies reported in Table V-12.
From the equilibrium constants at 285 and 298 K, a considerably higher reaction enthalpy ( ∆ r H m = (5.6 ± 1.4) kJ·mol–1) can be calculated using the integrated van’t
Hoff equation than from the data of 298 and 313 K ( ∆ r H m = (3.1 ± 1.2) kJ·mol–1). The
latter value is closer to the calorimetrically determined reaction enthalpies ( ∆ r H m =
(2.1 ± 0.3) kJ·mol–1 and (1.8 ± 0.2) kJ·mol–1 at 298 and 313 K, respectively). In the present review we re-evaluated the calorimetric data at 298 K of the equilibrium:
Ni2+ + Cl– U NiCl+
(A.29)
using the recommended value for log10 b1ο and ∆ε((A.29), NaClO4). At I = 2 M, β1 =
0.1 can be calculated. Since the authors used self-media for the calorimetric measurements, by mixing 0.632 M Ni(ClO4)2 and 2 M NaCl solutions, the β1 value calculated
above is only an approximation for the conditions used. Nevertheless, using this constant ∆ r H m ((A.29), 298 K) = (8.1 ± 2.5) kJ·mol–1 can be derived.
[66KOH/ROD]
The nickel mineral discovered at Lemieux Township, Gaspé Peninsula, Canada was
investigated by X-ray, wet chemical, spectrographic, infrared, differential thermal, and
petrographic analyses. It was found to be a previously undescribed nickel, magnesium,
iron carbonate. It was named after the locality of its discovery.
[66LOF/MCI]
Lofgren and McIver carried out measurements for the cells Mg,MgF2|CaF2|NiF2|Ni and
Mg,MgF2|CaF2,YF3,(3% by weight)|NiF2|Ni. The unweighted average of the values
from the three sets of measurements judged by the authors to be unaffected by moisture
is ∆ r Gm (A.30)= – (451.90 ± 0.70) kJ·mol–1.
NiF2(cr) + Mg(cr) U MgF2(cr) + Ni(cr)
(A.30)
A. Discussion of selected references
327
A third law analysis of the results using auxiliary values from the present review, and with heat capacity functions for Mg(cr) and MgF2(cr) consistent with those in
the CODATA review [89COX/WAG], leads to ∆ f H mο (NiF2, cr, 298.15 K) =
− (657.0 ± 1.4) kJ·mol–1.
The authors also determined potentials for the cells: Mg,MgF2|CaF2,YF3,(3%
by weight)|UF4,UF3|Pt and Pt,UF3,UF4|CaF2|NiF2|Ni and Pt,UF3,UF4|CaF2,YF3(3% by
weight)|NiF2|Ni.
The combined measurements from these two cells were consistent with, indeed
almost identical to, those from the simpler Mg,MgF2|CaF2|NiF2|Ni cell. Results reported
for the cell Fe,FeF2|CaF2|NiF2|Ni cannot be used without a proper re-evaluation of the
thermodynamic quantities for FeF2(cr), and such an evaluation is outside the scope of
the present review.
[66SHI/FUJ]
A solution of K2Ni(CN)4 was irradiated with γ-rays for 15 minutes, then the chemical
species in the irradiated sample (labelled with 57Ni by the reaction 58Ni(γ,n)57Ni)) was
separated by paper electrophoresis. The authors detected four main peaks. Beside the
Ni2+ and Ni(CN) 24 − ions, the presence of Ni(CN)2 and Ni(CN)3(H2O)– species was also
suggested, though the assignment of the peaks is rather speculative.
[66STO/ARC]
Results were reported for low-temperature calorimetric heat capacity measurements of
samples of hydrated nickel sulphate with compositions NiSO4·6.010H2O,
NiSO4·6.867H2O and NiSO4·7.301H2O. These were combined with results from earlier
heat-capacity measurements [41STO/GIA], [64STO/HAD] to determine smoothed sets
of entropy and heat capacity values from 1 to 300 K for α-NiSO4·6H2O and
NiSO4·7H2O. The authors corrected the measurements for NiSO4·7.301H2O for the heat
of fusion of ice (5.879 kJ·mol–1 at 269.75 K) on the basis of solubility data at the temperature of the NiSO4·7H2O-ice eutectic [39ROH]. The heat of dehydration,
NiSO4·7H2O U NiSO4·6H2O + H2O(NiSO4(aq, sat))
was reported as (7.98 ± 0.21) kJ·mol–1.
The authors also measured the enthalpies of solution of samples of
NiSO4·6.000H2O and NiSO4·6.940H2O in water such that the final solutions had essentially the same compositions (0.1 M in NiSO4). Thus, in that medium, the enthalpy of
the dehydration reaction was (7.68 ± 0.10) kJ·mol–1. The authors assumed that this enthalpy difference was unaffected by the final medium, and that it was equal to the difference at infinite dilution in water. This value is accepted in the present review, and
used in the calculation of the standard enthalpy of solution of the heptahydrate.
A. Discussion of selected references
328
The authors also carried out vapour pressure measurements to determine the
Gibbs energy of hydration of α-NiSO4·6H2O at 293.15, 298.15 and 303.15 K. Their
results were within the experimental uncertainties of values determined by others
[23SCH], [35BON/BUR]. From the enthalpy of solution measurements and the vapour
pressure measurements the difference between S mο (NiSO4·6H2O, α) and
S mο (NiSO4·7H2O) was calculated, and compared to the difference in the S mο values calculated from the low-temperature heat capacity measurements.
The differences were of the order of 0.7 J·K–1·mol–1. This is small considering
the many experimental quantities used in the calculation. Though they are dependent on
fewer assumptions, and are therefore the basis of the selected values in the present review, the third-law entropy values for the two hydrated solids may incorporate compensating errors. Therefore, an uncertainty of 1.0 J·K–1·mol–1 is estimated in the present
review for the values of S mο for the two solids at 298.15 K. For NiSO4·6H2O, the uncertainties in the heat capacity values near 298.15 K are estimated in this review to be ± 1.0
J·K–1·mol–1, and the same uncertainty could be assigned to values for
C pο,m (NiSO4·7H2O) for temperatures below, but close to the NiSO4·7H2O-ice eutectic.
However, because of the experimental problems encountered near room temperature, an
uncertainty of ± 4.0 J·K–1·mol–1 is assigned to C pο,m (NiSO4·7H2O, 298.15 K).
[66VOL/KOH]
Vollmer et al. determined the enthalpy of fusion ∆ fus H mο = (17.15 ± 0.42) kJ·mol–1 and
the heat capacity of nickel in the liquid region up to 1820 K. Based on the earlier work
of one of the authors, R. Kohlhaas, the heat capacity of nickel in the solid state in the
temperature range 300 to 1725 K is also reported in this paper. The reported values
were used to determine the thermal heat capacity function recommended by this review.
The heat capacity of nickel in the liquid region, C pο,m (Ni, l) = 39.0 J·K–1·mol–1, was
found to be independent of the temperature.
[67ANT/WAR]
A mixture of metallic nickel and nickel oxide was treated with a CO2/CO gas mixture at
temperatures ranging from 853 to 1289 K. The partial pressure of CO in the equilibrated
gas mixture was measured by means of a 14C tracer method after the carbon dioxide had
been removed by a liquid nitrogen trap. The authors found for the temperature dependence of the partial pressure ratio: log10 ( pCO / pCO ) = – 2381 (K/T) – 0.025. From this
2
relationship, the Gibbs energy of formation of NiO was derived:
∆ f Gm (T) = (– 236.81 + 0.087446 (T/K)) kJ·mol–1, 853 < T/K < 1289.
[67BLA]
Blaszkiewicz determined the solubility of Ni(OH)2 by a 63Ni tracer method and found
[Ni(II)] = 1.49 × 10–5 mol·dm–3. Although this value is an order of magnitude lower than
Almkvist’s, the mass and charge balance with Ni2+ and NiOH+ still leads to a much too
A. Discussion of selected references
329
high value of K s ,0 or else Ni(OH)2(aq) is unrealistically stable. Consequently this result
was also rejected. At room temperature directly measured aqueous solubilities of nickel
hydroxide are an insufficient basis for the determination of the respective solubility
constant [18ALM], [67BLA]. As the system is not adequately buffered, freshly precipitated Ni(OH)2 forms colloidal particles, and the concentration of dissolved Ni(II) at the
solubility minimum is at or below the instrumental detection limit [97MAT/RAI].
[67GES/NEU]
The solubility of NiSO4 hydrate is reported for temperatures from 160 to 240°C, and in
ammonium nitrate solutions to 0.768 m to 300°C. Activity coefficients over the concentration and temperature range were calculated.
[67NAN/TOR]
Differential calorimetry has been used to measure the enthalpies of formation of the
MSCN+ complexes (M = Mn(II), Co(II), Ni(II), Cu(II), Zn(II), Cd(II) and Pb(II)) in
aqueous solutions at 25°C. No background salt was used to attempt to maintain constant
ionic strength. The concentration of nickel(II) perchlorate varied between 5.04 and 8.9
mM, while [SCN–]tot ranged from 2.44 to 4.31 mM. Thus only the mono-thiocyanato
complex was formed in the calorimeter:
Ni2+ + SCN– U NiSCN+.
(A.31)
The concentrations of NiSCN+ formed were calculated by means of the thermodynamic stability constant reported in [62WIL2] and the Davies equation. For the
purposes of the present review, the experimental data given in [67NAN/TOR] have
been re-evaluated using the recommended values for log10 b1ο (A.31) and ∆ε((A.31),
ClO −4 ). The recalculated reaction enthalpy (– (9.25 ± 0.40) kJ·mol–1) is somewhat
higher than the original value (– (10.46 ± 0.40) kJ·mol–1). Although, these values refer
to NaClO4 medium, the ionic strength in [67NAN/TOR] was always less than 0.063 M.
Therefore, the recalculated reaction enthalpy is accepted with an increased uncertainty
(– (9.25 ± 2.00) kJ·mol–1).
[67RUD/DEV]
Rudzitis et al. used fluorine bomb calorimetry to determine the enthalpy of formation of
anhydrous nickel fluoride as – 657.7 kJ·mol–1 at 298.15 K. During the reaction, only
approximately 50% of the nickel underwent reaction, and the fraction of unreacted material was determined by mechanical and magnetic separation. Despite these difficulties,
the authors estimated that the overall 2σ uncertainty in their value of ∆ f H mο (NiF2, cr,
298.15 K) was 1.7 kJ·mol–1 (0.4 kcal·mol–1), slightly greater than the statistical uncertainties of the eight heat of combustion measurements. Because of the need to separate
the nickel fluoride from the residual nickel at the end of the experiments, a bias in the
final enthalpy of formation value cannot be ruled out. The paper provides an assessment
of earlier fluoride enthalpy of formation determinations.
330
A. Discussion of selected references
[67SIG/BEC]
The determination of the complexation constant between Ni2+ and HPO 24 − was almost
incidental to the main set of experiments, which were intended to compare the effects of
addition of bipyridyl and organophosphate ligands on nickel complexation. Nevertheless, this is one of the studies most often cited when an association constant is required
for nickel-phosphate complexation (and the values were cited and used by the same
group in later publications, e.g., [81BAN/KAD]). Potentiometric titrations were carried
out at 25°C using solutions 0.01 or 0.02 M in nickel and 2 × 10–4 M in HPO 24 − , with
0.02 M NaOH(aq) as titrant, in a 0.1 M perchlorate medium (NaClO4(aq)). To determine the complexation constant, the titration curves for solutions containing the metal
and ligand were compared with those with the metal or ligand omitted.
The primary data were not reported except in a figure showing the smoothed
data on a coarse scale, and, therefore, the data analysis cannot be properly evaluated.
Also, the stoichiometry of the complex was not clearly established. The value
log10 K (A.32) = 2.08 is given for the equilibrium constant for
Ni2+ + HPO 24 − U NiHPO4(aq)
(A.32)
but the estimated uncertainty is not reported. In the same paper, the association constant
value for the reaction:
H+ + HPO 24 − U H 2 PO −4
(A.33)
is reported as (6.70 ± 0.02) in 0.1 M NaClO4(aq), whereas the value calculated using the
SIT (cf. Appendix B) is (6.781 ± 0.015). Correction to zero ionic strength (Appendix B)
gives log10 K ο (A.32) = (2.97 ± 0.30), where the uncertainty has been estimated in the
present review. The small ∆ε correction has been based on the value for ε(Na+, HPO 24 − )
from Table B-5 (rather than using the equation in Table B-6), and ε(Ni2+, ClO −4 ) = 0.37
kg·mol–1.
[68AND/KHA2]
This is qualitative work on the complex formation of nickel(II) in concentrated aqueous
HBr (0.1 – 18 M) and HI (0.1 – 11 M) solutions, done by comparing the electronic absorption spectra of nickel(II) at varying HX concentrations with those of crystalline
Ni(II) complexes. The authors suggested the formation of NiBrx2 − x (x = 1 – 6) and
NiI 2x − x (x = 1 – 4) complexes. Considering the molar absorption coefficient of NiBr42 −
complex in nitromethane (εmax(698 nm) = 173 [59GIL/NYH]), only the formation of
approximately 5 –10 % NiBr42 − complex can be estimated in 18 M HBr solution from
the published electronic spectra, while the higher bromo complexes can not be justified.
On the other hand, the spectra presented suggest, indeed, the formation of a NiI 24 − complex in considerable amounts in 9 – 11 M HI solutions. The absorption coefficients reported for 11 M HI solution (εmax(~ 800 nm) = 400, εmax(~510 nm) = 2600) suggest the
formation of a tetrahedral complex.
A. Discussion of selected references
331
[68ARN]
Calorimetric studies by acid titration of hydrolysed nickel(II) solutions have been performed at 25°C in 3 M (Na)Cl ionic medium in order to obtain the enthalpy of the hydrolysis reactions. The experimental data have been interpreted using the hydrolysis
mechanisms proposed by Burkov and Lilich [65BUR/LIL2] and Biederman and Ohtaki
(cited as a personal communication to the author [68ARN]). The latter values
( log10 *b 4,4 =
− (28.57 ± 0.03),
log10 *b1,1 = – (10.5 ± 0.1) and
log10 *b1,2 =
− (10.3 ± 0.2)) are slightly different from those reported in [71OHT/BIE]. The concentration of nickel(II) ranged from 0.055 to 0.750 M (I = 3.055 to 3.750 M). A value of
(180 ± 2) kJ·mol–1 was obtained for the heat of formation of the main hydrolysis product
Ni 4 (OH) 44 + , irrespective of which assumptions were made about the minor species. On
the other hand, the rather uncertain enthalpy values for the formation of the possible
minor species (NiOH+, Ni2OH3+) strongly depended on the model (formation constants)
used. The best fit to the data was obtained using the hydrolysis model proposed by
Biederman and Ohtaki: ∆ r H m = (64.4 ± 6.0) and (16.0 ± 8.4) kJ·mol–1 for the formation
of NiOH+ and Ni2OH3+, respectively.
The formation constants used in the calculations were obtained without considering chloride complexation. Therefore, for the purpose of this review we re-evaluated
the reported experimental data, using the selected formation constants for 3 M chloride
medium and assuming a value of (8.2 ± 2.5) kJ·mol–1 for the formation of NiCl+ under
the conditions used (the latter value was calculated from the data measured at I = 2 M
[66KEN/LIS]). The recalculation resulted in the following enthalpy of reaction values,
valid for 298 K and I = 3M, ∆ r H m = (54.3 ± 2.0) kJ·mol–1, (45.9 ± 1.6) kJ·mol–1 and
(192.9 ± 1.0) kJ·mol–1 for the formation of NiOH+, Ni2OH3+ and Ni 4 (OH) 44 + , respectively.
[68CAM/ROE]
This paper describes a study of the stability of olivine and pyroxene in the
Ni − Mg − Si – O system at temperatures of 1300 – 1500°C at controlled oxygen fugacities using gas mixtures of known mixing ratio. The Gibbs energy of formation of nickel
oxide and the Gibbs energy change for the reaction:
2Ni + SiO2 + O2 U Ni2SiO4
determined in this temperature range coincide very well with results of all emf-studies
discussed by this review. Moreover, a close approach to ideal behaviour of the magnesium – nickel olivine solid solutions was reported.
[68CHA/FLE]
A “closed system” solid electrolyte electrochemical cell has been designed to investigate the thermodynamic properties of metal oxides. This technique gave a rapid cell
response and highly stable potentials through the elimination of mixed potentials. The
A. Discussion of selected references
332
standard Gibbs energy of formation of NiO was found to be the following function of
the temperature: ∆ f Gmο (T) = (– 233.651 + 0.085 (T/K)) kJ·mol–1. The accuracy calculated from the maximum deviation from the computed least-squares line is equal
to ± 0.209 kJ·mol–1. This corresponds to the maximum deviation in cell potential (emf)
of ± 1.0 mV.
[68KOL/MAR]
The rate of formation and dissociation of Ni(CN) 24 − has been studied at 25°C and I =
0.1 M NaClO4. The rate of formation between pH 5.5 to 7.5 showed a second-order
dependence on both [CN–] and [HCN], at higher pH the reaction order in [CN–] was
greater than 2, but the total cyanide dependence remained fourth order. The fact that the
formation rate depended on [HCN]2 suggested that the reaction product might be
Ni(HCN)2(CN)2(aq). Equilibrium measurements, using a spectrophotometric method
between pH 4 – 5, resulted in systematic change in the calculated log10 b 4 values. This
was taken as further proof for the presence of protonated species, although the UV-VIS
spectrum of Ni(CN) 24 − was found to be unaffected by the protonation. The evaluation of
combined kinetic and spectrophotometric data resulted in log10 b 4 = (30.5 ± 0.3) and
three protonation constants of the tetracyanonickelate(II) species: log10 K H,0 =
(5.4 ± 0.2), log10 K H,1 = (4.5 ± 0.2), log10 K H,2 = (2.6 ± 0.2), for the reactions:
Ni(HCN) x (CN) 4x −− 2x + H+ U Ni(HCN) x +1 (CN) 4x −− 2x −+11
(where x = 0, 1, 2). On the other hand, spectrophotometric [59FRE/SCH] and potentiometric [63CHR/IZA], [71IZA/JOH], [74PER] studies performed in the same pH-range
did not indicate the existence of protonated species in notable concentration, although
their kinetic role in the acid dissociation of Ni(CN) 24 − was later corroborated [76PER].
The value for log10 b 4 is accepted, but with an increased uncertainty (± 0.6).
[68KOS]
Kostryuka and Zarubina measured the heat capacity of NiCl2 between 1.8 and 16 K to
ascertain the singularities of the energy spectrum of layered antiferromagnets. The data
have been presented graphically only.
[68LAR/CER]
Larson et al. obtained for the reaction:
NiSO4·7H2O(cr) U Ni2+ + SO 24 − + 7H2O(l)
ο
m
–1
–1
(A.34)
∆ sol G (A.34) = 12.929 kJ·mol (3.09 kcal·mol ) which agrees very well with the value
recalculated in Section V.2.1.3.1. A misprint seems to have occured in this paper, as the
standard enthalpy of reaction (A.34) is given as ∆ sol H mο (A.34) = 2890 instead of 2986
cal·mol–1. The latter value follows from the data of [66STO/ARC] and the original value
given by [66GOL/RID].
A. Discussion of selected references
333
[68MAL/TUR]
The instability constants of NiSCN+ were determined at 15, 25 and 35°C by a polarographic method in aqueous solutions at I = (0.65 ± 0.03) M using (H)ClO4 as background electrolyte. The thiocyanate concentration ranged from 0.5 mM to 7 mM, with
[Ni2+]tot = 55.7 mM, thus the formation of higher complexes (Ni(SCN)x, x > 1) was neglected. The determination is based on the measurement of the decrease in catalytic
polarographic current of the Ti(IV) – SCN complexes in the absence and in the presence
of nickel(II). Although, the authors applied rather acidic conditions, and thus the thiocyanate ion was partly protonated, the reported values are accepted with increased uncertainties. From the temperature dependence of log10 b1 (see Figure A-17) ∆ r H m =
− (14.5 ± 1.2) kJ·mol–1 can be derived.
Figure A-17: Temperature dependence of log10 b1 values of the NiSCN+ complex reported in [68MAL/TUR].
1.6
log10 β1
1.4
1.2
1.0
0.8
0.0032
0.0033
0.0034
0.0035
K/T
[68NAV/KLE]
The enthalpies of formation from the oxides of thirty-two 2-3 and 2-4 spinels, among
others Ni2GeO4, have been determined by solution calorimetry in molten oxide solvents
at 970 K. Regularities in the thermodynamics of spinel formation and the olivine –
spinel transition for orthosilicates of Mg, Fe, Mn, Co, and Ni have been discussed.
[68PRA]
This is a description of a general method used by Prasad and co-workers to carry out
potentiometric measurements to determine complexation constants, and describes how
the measurements were extrapolated to I = 0. For NiSO4 only a single value is reported
( K1−1 = 9.0 × 10–3 at 35°C). This suggests weaker complexation than was found in other
334
A. Discussion of selected references
work (e.g., [59NAI/NAN]). For experimental details and results, reference was made to
unpublished work by S.C. Sircar, P.K. Jena and B. Prasad. However, no related paper
on nickel sulphate complexation by these authors seems to have been reported in
Chemical Abstracts, and the work may never have been published. The unavailability of
the primary data precludes recalculation of the results using the standard SIT procedures, and the reported value is not used in the present assessment. The data reported for
the nickel(II) – thiocyanate system in this paper, were later published in detail in
[71DAS/DAS].
[68VOG]
This paper presented evidence for the existence, in potassium hydroxide melts (0.85 w
KOH) between 498 and 523 K, of reversible homogeneous nickel redox couples in
which nickel exhibits valencies greater than two. The data can be interpreted by the following two simultaneous redox reactions:
Ni2+ U Ni3+ + e–
Ni3+ + 4OH– U NiO2(s) + 2H2O + e–.
The nature of the black NiO2 crystals (+4-valent nickel or peroxidic oxygen)
was left undecided.
[69FUN/TAN]
Kinetic and mechanistic studies of the reaction between 4-(2-pyridylazo)resorcinol
(PAR) and (among others) NiOH+ and NiF+ were performed at an ionic strength of 0.10
M (NaClO4) and at (298.15 ± 0.20) K. Nickel(II) perchlorate solutions were mixed with
ligand solution, then borate buffer solution (8 × 10–3 M) was added to adjust the pH (7 –
9.5). The concentration ranges were as follows: [Ni2+] ~ 5 – 50 × 10–7 M, [PAR] ~ 10–5
M, [F–] ~ 0.05 M. The influence of the buffer was examined and it was concluded that it
had no effect on the rate constants. The kinetic results also provided data for the equilibrium constant for the reaction:
Ni2+ + OH– U NiOH+
(A.35)
log10 K (A.35) = (4.3 ± 0.1)) and for
Ni2+ + F– U NiF+,
(A.36)
log10 K (A.36) = (1.1 ± 0.1)).
Using log10 K w = – 13.78, log10 K (A.35) can be can be converted to the corresponding first hydrolysis constant of nickel(II):
Ni2+ + H2O U NiOH+ + H+,
log10 *K1ο = – (9.5 ± 0.1). In this review, these values are accepted, but with the uncertainties increased to ± 0.2.
A. Discussion of selected references
335
[69IZA/EAT]
Titration calorimetry was used to determine values of log10 K , ∆ r H mο and ∆ f S mο , reported for the formation of NiSO4(aq) for 25°C and I = 0. The data analysis is suspect
because of a possible correlation between the value for the calculated equilibrium constant and the enthalpy of association. A good reassessment was reported by Powell
[73POW], who pointed out that if the data of Izatt et al. were used with the association
constant at 298.15 K set equal to the value from Nair and Nancollas [59NAI/NAN],
then an enthalpy of ion-pair formation of 5.7 to 5.9 kJ·mol–1 was obtained, not the reported 1.7 kJ·mol–1 [69IZA/EAT].
In the present review, several approaches were tried to determine the enthalpy
of association within the constraints of the SIT formulation in Appendix B. As in earlier
reviews [92GRE/FUG], [2001LEM/FUG], for low ionic strengths the assumption was
made that ∆ r H m is independent of ionic strength. Powell [73POW] suggested that the
effect of this approximation for sulphate complexation was < 1 J·K–1·mol–1. Izatt et
al. [69IZA/EAT] and Powell [73POW] incorporated values for an association constant
and enthalpy of association for the Me4NClO4(aq) ion pair in their calculations. A full
evaluation of these values has not been attempted here. Determination of the values for
Me4NClO4(aq) required using an explicit value for the association constant between Na+
and SO 24 − , whereas the SIT, as described in Appendix B, would treat these interactions
using an interaction coefficient rather than a specific association constant.
Initially, the effect of the enthalpy of protonation of sulphate has been ignored.
Using I = 0 values for the formation constant of NiSO4(aq), as recalculated from the
data of Nair and Nancolas [59NAI/NAN] and of Katayama [73KAT], the molar concentration of NiSO4(aq) formed at each step of the titration was calculated iteratively. The
average enthalpy of ion pair formation is 5.8 kJ·mol–1 and 4.7 kJ·mol–1, respectively.
However, it is apparent that the calculated enthalpy of association is strongly dependent
on the ionic strength of the solution. If the values at the measured ionic strengths are
extrapolated to I = 0, the limiting value (Figure A-18) is 1.2 kJ·mol–1 or 0.4 kJ·mol–1
depending on whether the association constant data of Nair and Nancolas
[59NAI/NAN] or from Katayama [73KAT] are used. This cyclic inconsistency may be
a result of the neglect of the heats of the secondary association reactions and heats of
dilution. It may also represent a failure of the SIT to adequately represent the system. At
present, the enthalpy of association as calculated by Powell is accepted, but with an
increased uncertainty ∆ r H mο = (5.8 ± 2.0) kJ·mol–1.
A. Discussion of selected references
336
Figure A-18: Heats of reaction ( ) determined from the calorimetric data of
[69IZA/EAT] by using the 298.15 K values of log10 K1 from [73KAT]. The line
represents a least-squares fit to these enthalpies as a function of ionic strength.
10000
∆r Hm / J·mol
−1
8000
6000
4000
2000
0
0.00
0.02
0.04
0.06
0.08
I / mol·kg
0.10
0.12
0.14
−1
[69KOL/KIL]
The reaction rate and mechanism of the acid catalysed decomposition of Ni 4 (OH) 44 +
have been studied by stopped-flow measurements. Potentiometric titrations of 0.46 M
nickel(II) perchlorate solutions at 25°C and 1.5 M ionic strength, maintained by NaClO4, were also performed to study the hydrolysis under the conditions used for the
kinetic measurements. The authors suggested the presence of the Ni 4 (OH) 44 + species,
only, and log10 *b 4,4 = – (27.03 ± 0.06) was obtained. This value compares favourably
with the corresponding value reported in [65BUR/LIL]. The reported data indicate a
slight tendency for a decrease in the calculated log10 *b 4,4 values with decreasing pH,
which may be due to the presence a minor hydroxo complex. Therefore, the reported
experimental data were re-evaluated for the purpose of this review, but no further species can be identified with any confidence.
[69KOS]
The measurements carried out in [68KOS] were discussed from the point of view of the
magnetic spectrum of layered antiferromagnets. The molar heat capacities of NiCl2 from
1.8 to 16 K have been listed.
A. Discussion of selected references
337
[69KUL/ERS]
On a godlevskite specimen from the Cu – Ni sulphide deposit Noril'sk, Siberia, Russia,
Kulagov et al. carried out chemical and mineralogical investigations as well as a crystal
structure analysis. Whereas the mean values of the chemical analyses agreed with
(Ni, Fe, Co)S the authors maintained that the upper limits of the metal and the lower
limits of the sulphur contents are consistent with Ni7S6.
[69LIB/SAD]
Osmotic coefficients have been determined by an isopiestic method for (among others)
the aqueous solution of Ni(ClO4)2 in the concentration range m = 0.1 – 3.5 mol·kg–1 at
25°C. The mean activity coefficients were calculated from the Gibbs-Duhem equation.
For the purposes of this review, these data, together with those listed in [99MAL/CAR],
have been used to derive the ion interaction coefficient ε(Ni2+, ClO −4 ), see also the discussion on [99MAL/CAR].
[69MAK/SPI]
Makovskaya and Spivakovskii added sodium hydroxide to 0.001 – 4.7 M NiCl2 solutions at 25°C until a slight precipitate formed. The pH and the chloride activity were
measured with glass and silver chloride electrodes, respectively. The concentration of
Ni(II) in the filtered solutions was determined by complexometric, photometric, and
polarographic methods. Ni(OH)2, Ni(OH)1.5Cl0.5, and Ni(OH)1.75Cl0.25 were detected by
thermodynamic analysis, but the latter phase disappeared during aging. None of these
phases were characterized by X-ray analyses. The solubility product of the freshly precipitated nickel hydroxide turned out to be approximately an order of magnitude higher
than after an undefined aging period. So, only the value of the aged phase ( log10 K sο,0 =
− 15.54) has been accepted for Table V-6 and Figure V-11.
[69MOR/SAT]
The standard Gibbs energy of the reaction:
2Ni + O2 U 2NiO
was investigated as a function of temperature in the temperature range from 700 to
about 1100°C. The Gibbs energy function reads: ∆ f Gmο (T) = – 476.976 +
0.175770 (T/K)) kJ·mol–1. The uncertainty of emf measurements, caused by different
kinds of solid electrolytes, was around ± 3 mV.
[69RAN/BRE]
The potential of nickel sulphide electrodes was measured against the saturated calomel
electrode at 298.15 K. The sulphide electrodes were prepared by covering platinum with
nickel sulphide which was obtained by precipitation in aqueous solutions of low acidity
(1 M Na2S solution was dropped into 0.5 M NiSO4 solution). The composition of the
338
A. Discussion of selected references
sulphide was altered by anodic as well as cathodic polarisation. Unfortunately, X-ray
diffraction analysis revealed that the nickel sulphides were amorphous. Hence, no reliable thermodynamic data can be evaluated from this study.
[69SOR/KOS]
For the calculation of S mο (Ni(OH)2, cr, 298.15 K) according to Equation (A.37) the
original C pο,m data have been plotted versus lnT:
S m = ∫ C p ,m d lnT .
(A.37)
Figure V-10 shows the experimental C p ,m values of three Ni(OH)2 samples (iii,
ii, i) up to the temperature where the data for each have been measured. At T ≤ 4 K series (iii) and (i) essentially coincide. In the range of the magnetic transition the heat capacities decrease in the order C p ,m (iii) > C p ,m (ii) > C p ,m (i), but at temperatures higher
than at the minimum of C p ,m = f (lnT), this order is reversed.
Using the CurveExpert routine the functions C p ,m (ii) = f (lnT) and C p ,m (i) =
f (lnT) were fitted to linear splines and integrated according to Equation (A.37)
[97HYA]. Finally, the Debye T 3 extrapolation down to T = 0 was applied. As no data
above T = 104 K were measured for sample (iii), 0.4 and 1.2 J·K–1·mol–1 were subtracted from C p ,m (ii) and C p ,m (i) above 104 K, respectively, and the values thus obtained were added to series (iii) [69SOR/KOS]. Again the CurveExpert routine was
used to integrate the true as well as the dummy part of this C p ,m (iii) function.
[69SUB/COR]
The rate of extraction of nickel(II)- and zinc(II)-dithizonates into CHCl3 has been studied in presence of several auxiliary ligands (L), e.g., thiocyanate, at 25°C and at ionic
strength of 0.25 M using NaClO4. The concentrations used were not reported, but were
probably sufficiently low compared to the total ionic strength, since the concentration of
the Ni(ClO4)2 stock solution was 1 mM. The variation of the apparent rate constant of
the extraction as a function of the concentration of the auxiliary ligand, allowed calculation the stability constants of the MLx complexes. The authors reported log10 b1 =
(1.3 ± 0.2) for the stability of the species NiSCN+.
[70ASH]
The heat of solution of nickel chloride hexahydrate into 1.0M HCl(aq) was measured,
and (6.10 ± 0.22) kJ·mol–1 was reported (2σ uncertainties). The final solution was 0.02
M in Ni. A double correction to 0.0 M HCl(aq) and to I = 0 [95MAN/KOR] would be
required to use the reported value to calculate the enthalpy of solution of the salt at infinite dilution in water. The corrections themselves have large uncertainties, and this heat
of solution value is not used further in the present review.
A. Discussion of selected references
339
[70BIN/HEB]
The authors reported drop calorimetry results obtained from samples initially at temperatures between 381.9 K and 1461.7 K. The scatter in the enthalpy difference results,
based on measurements done at similar temperatures, seems to be of the order of 200
kJ·mol–1. Direct fitting of the drop calorimetry results without incorporating some low
temperature heat capacity constraint leads to a poor extrapolation for temperatures below 400 K. The authors fitted an equation to their values of ( H m (T ) − H mο (298.15 K) ),
“weighting the point at 298.15 K for the best fit of all results”. This seems to have entailed the use of some of the heat capacity results of Catalano and Stout [55CAT/STO].
The reported value of C pο,m (298.15 K) (62.93 J·K–1·mol–1) differs by 1.1 J·K–1·mol–1
from the value of [55CAT/STO], though the S mο value from that source is used. Also,
the expression reported by Binford and Hebert
[ H m (T ) − H mο (298.15 K) ] /J·mol–1 = 4.184·(15.677(T/K) + 1.7344 × 10–3 (T/K)2
+ 1.4872 × 105(T/K)–1 –5161.4)
is non-zero (693 J·mol–1) for 298.15 K. In the present review, the drop calorimetry results are used together with the heat capacity results of Catalano and
Stout [55CAT/STO] to derive a properly constrained expression for
( H m (T ) − H mο (298.15 K) ) and C pο,m (T). Above 1300 K, the experimental values of
( H m (T ) − H mο (298.15 K) ) appear to be slightly larger (by 0.5 kJ·mol–1 to 1.3
J·K−1·mol−1) than might be expected by extrapolation of values from lower temperatures. Because there are only two measurements in that temperature range, these values
have not been treated differently from the other measurements in the fitting process.
[70CAR/LAI]
The solubility product of NiS was determined by linear voltage sweep voltammetry
using a mercuric sulphide coated electrode (hanging mercury drop). The peak potential
for the exchange reaction between the mercuric sulphide coated mercury electrode and
Ni2+ ions of a nickel perchlorate solution,
HgS + 2e– + Ni2+ U Hg + NiS
(A.38)
yields the ratio between the solubility constants of HgS and NiS. At 298.15 K the authors obtained log10 K sο,0 = – 17.8 for the solubility product of NiS. Unfortunately, neither any further experimental details nor the second dissociation constant for H2S(aq)
are indicated in this contribution. Thus, the data cannot be recalculated and compared
with results of different studies.
[70EFI/KUD]
The enthalpies of solution in water of nickel chloride, cobalt chloride and the coordination compounds RbNiCl3, RbCoCl3, Rb2CoCl4 and Rb3CoCl5 were determined in a calorimeter with an “isothermal jacket”. The authors reported for the reaction,
340
A. Discussion of selected references
NiCl2(cr) U Ni2+ + 2Cl–
(A.39)
an enthalpy of solution of − (83.26 ± 0.42) kJ·mol for 10000 ≦ n(H2O) / n(NiCl2, cr)
≦ 40000); this leads to a value of ∆ sol H mο (A.39), between – 83.62 kJ·mol–1 and – 83.93
kJ·mol–1, which is approximately 1 kJ·mol–1 less negative than the value reported in
[90EFI/FUR]. The latter (– (84.89 ± 0.27) kJ·mol–1) has been better documented and
thus is accepted in this review.
–1
[70HAL/VAN]
The formation of nickel(II) chloro complexes has been studied at 298.15 K in 4 M
Na(ClO4,Cl). The solution containing a known total concentration of nickel (0.1, 0.2 or
0.5 M) and chloride (0 – 2.0 M) was saturated with AgCl labelled with radioactive
110
Ag. The measured radioactivity was correlated with the free chloride concentration
taking into account the solubility of AgCl and the different chloro complexes of silver(I). The experimental data have been interpreted in terms of the unique formation of
the bis-complex NiCl2 (i.e., β1 = (0.000 ± 0.015) M–1; β2 = (0.129 ± 0.002) M–2). This
interpretation solely in terms of the formation of the NiCl2 complex is questionable, and
probably was adopted because of effects from the marked change of the ionic medium
during the measurement. The medium effects, however, are less than in most of the
other work on chloro complexation of Ni(II). Therefore, we re-evaluated the published
experimental data. Taking into account only the bis-complex, a better fit can indeed be
achieved ( log10 b 2 = – (0.89 ± 0.02)), compared to the result obtained using the assumption of the unique formation of NiCl+ ( log10 b1 = – (0.62 ± 0.04)). Nevertheless, in this
review we prefer the latter value with a higher uncertainty ( log10 b1 = − (0.62 ± 0.30)).
This choice is also confirmed by several equilibrium studies performed in non-aqueous
solvents, which have higher precision since the complexes are more stable than in aqueous solution. For example, in methanol, ethanol and propanol, K1 is much higher than
K2 [95KAH/CRO]. Using the recommended values for ε(Na+, Cl–), ε(Ni2+, ClO −4 ) and
the ε(NiCl+, ClO −4 ) derived from the data published in [75LIB/TIA], the extrapolation
to I = 0 resulted in log10 b1ο = (0.68 ± 0.40).
The experimental work is well documented in this paper, which offers the possibility of using the SIT to assess the medium effect caused by the replacement of the
original background electrolyte. The activity of a cation M of charge zM in a mixture of
two electrolytes (NX and NY, with a molality mX and mY) at constant ionic strength can
be expressed using the SIT as follows:
log10 γM = – zM2 D + ε(M, X) mX + ε(M, Y) mY
where D is the Debye-Hückel term. The same equation for an anion Y is:
log10 γY = – zY2 D + ε(N, Y) (mX + mY).
Substituting the log10 γi values in Equations (B.3) (Appendix B) with these expressions, applying it to the equilibrium:
A. Discussion of selected references
341
Ni2+ + Cl– U NiCl+
and rearranging leads to:
log10 b1 + 4D – {ε(Na+, Cl–) + ε(Ni2+, ClO −4 )}Im – ε(Ni2+, ClO −4 ) mY – ε(Ni2+, Cl–) mY
= log10 b1ο – ε(NiCl+, ClO −4 )Im + ε(NiCl+, ClO −4 ) mY – ε(NiCl+, Cl–) mY
provided that N = Na+, M = Ni2+, X = ClO −4 , Y = Cl– and Im = mX + mY. Since all quantities on the left-hand side of this equation are known, it can be written:
Q = log10 b1ο – ε(NiCl+, ClO −4 )Im + ε(NiCl+, ClO −4 ) mY – ε(NiCl+, Cl–) mY
The three unknown quantities on the right-hand side are strongly correlated
with each other; thus, to solve this equation we should accept the ε(NiCl+, ClO −4 ) value
derived from the data published in [75LIB/TIA]. Then the equation is transformed into
Q’ = log10 b1ο – ε(NiCl+, Cl–) mY
(A.40)
Plotting the Q’ values vs. mCl− gives a straight line, with the intercept and
slope equal to log10 b1ο and – ε(NiCl+, Cl–), respectively.
Before applying this equation to the data published in [70HAL/VAN], two further simplifications should be used. With increasing chloride concentration the ionic
strength of the solution expressed in molarity is constant, but in the molal scale it
changes. Since no density data are available for such an electrolyte mixture, Im was calculated using the equation:
M − × a + (4 − M Cl− ) × b
I m = Cl
4
where MCl– is the concentration of chloride in molar units, and a and b are the molality
of 4 M NaCl and NaClO4 solution, respectively. The density of solutions with different
compositions was calculated assuming that 4 M NaCl and 4 M NaClO4 solutions mix
ideally. With these assumptions, the following dataset can be calculated.
The plot of Q’ versus mCl− is depicted in Figure A-19. From this plot
log10 b = (0.20 ± 0.03) and ε(NiCl+, Cl–) = – (0.16 ± 0.02) are calculated. This value of
log10 b is 0.5 log units lower than that calculated above from the accepted log10 b1
value at I = 4 M Na(ClO4, Cl) using the conventional SIT extrapolation. This difference
is, in great part, due to the medium effect. But it is also the consequence of the sensitivity of function Q’ to the values of different interaction coefficients used (especially on
whether the value of ε(NiCl+, ClO −4 ) is correct), and on the validity of the assumptions
made concerning Im and mCl–. In this review we accept the value obtained from the plot
in Figure A-19, but with a considerably increased uncertainty:
ο
1
ο
1
log10 b1ο = (0.20 ± 0.50).
A. Discussion of selected references
342
Table A-14: Dataset.
[Cl–]tot
[Cl–]free
[Ni2+]tot
β1
log10 b1 (molal)
mCl−
Im
D
Q’
0.25
0.2375
0.50
0.1078
– 1.0600
0.3058
4.9139
0.2610
0.2453
0.30
0.2835
0.50
0.1201
– 1.0129
0.3665
4.9067
0.2609
0.2754
0.38
0.3525
0.50
0.1335
– 0.9672
0.4572
4.8958
0.2609
0.2955
0.40
0.3900
0.20
0.1345
– 0.9639
0.4873
4.8922
0.2608
0.2904
0.45
0.4440
0.10
0.1433
– 0.9363
0.5474
4.8850
0.2608
0.3011
0.50
0.4651
0.50
0.1616
– 0.8841
0.6074
4.8778
0.2608
0.3364
0.65
0.6299
0.20
0.1770
– 0.8447
0.7863
4.8561
0.2606
0.3254
0.70
0.6880
0.10
0.1979
– 0.7962
0.8456
4.8489
0.2606
0.3572
0.75
0.7000
0.50
0.1587
– 0.8919
0.9047
4.8416
0.2605
0.2449
0.90
0.8640
0.20
0.2543
– 0.6872
1.0809
4.8200
0.2604
0.4000
1.00
0.9049
0.50
0.2596
– 0.6783
1.1975
4.8055
0.2603
0.3760
1.15
1.0960
0.20
0.3377
– 0.5640
1.3710
4.7838
0.2602
0.4413
1.20
1.1700
0.10
0.3656
– 0.5295
1.4285
4.7766
0.2601
0.4596
1.25
1.1241
0.50
0.2994
– 0.6162
1.4858
4.7694
0.2601
0.3567
1.40
1.3292
0.20
0.4120
– 0.4777
1.6567
4.7477
0.2599
0.4471
1.45
1.4099
0.10
0.4741
– 0.4167
1.7133
4.7405
0.2599
0.4921
1.50
1.3262
0.50
0.4019
– 0.4885
1.7697
4.7333
0.2598
0.4044
1.65
1.5570
0.20
0.5578
– 0.3461
1.9378
4.7116
0.2597
0.4993
1.75
1.5325
0.50
0.5024
– 0.3915
2.0489
4.6971
0.2596
0.4225
1.90
1.7840
0.20
0.7737
– 0.2040
2.2143
4.6755
0.2595
0.5634
1.95
1.8863
0.10
0.9322
– 0.1230
2.2690
4.6682
0.2594
0.6288
2.00
1.7362
0.50
0.6433
– 0.2842
2.3236
4.6610
0.2594
0.4523
2.15
1.9989
0.20
1.5441
0.0961
2.4861
4.6393
0.2592
0.7866
0.20
0.2000
0.10
0.0007
– 3.2532
0.2450
4.9211
0.2610
– 1.9308
0.95
0.8890
0.10
1.7597
0.1529
1.1393
4.8127
0.2603
1.2236
1.70
1.6106
0.10
5.2113
0.6244
1.9934
4.7044
0.2597
1.4541
0.40
0.3940
0.20
0.0785
– 1.1980
0.4873
4.8922
0.2608
0.0563
A. Discussion of selected references
343
Figure A-19: Extrapolation to I = 0 of the formation constant of NiCl+ using Equation
(A.40) and the experimental data published in [70HAL/VAN]. The outlying points
(open diamond) were not taken into account in the calculation.
2
1.5
1
Q'
0.5
0
-0.5
-1
-1.5
-2
-2.5
0
0.5
1
1.5
−
m (Cl ) / mol·kg
2
2.5
−1
[70MAC]
Potential sweep experiments were performed on film nickel hydroxide electrodes in an
effort to determine the redox mechanism. The over-all reaction for α-nickel hydroxide
was found to be:
2[3Ni(OH) 2 ·2H 2 O] + 6OH − + 3 KOH U 2[3NiOOH· 7 H 2 O· 3 KOH] + 3H 2 O + 6e − .
2
2
4
The rate-controlling step was assumed to be proton diffusion.
[70MIR/MAK]
This paper reports association constants for the inner- and outer-sphere complexes (β1
and ω1, respectively) of Co(II), Ni(II), and Cu(II) hexa-aqua ions with chloride, bromide, thiocyanate, sulphate, and nitrate ions, based on spectrophotometric and solubility
measurements in 3 M Li(ClO4, X) media at 298.15 K. The authors used the UV-region
of the electronic spectra to determine the sum of β1 and ω1, while the value of β1 was
determined from the VIS-region. This is obviously only a rough approximation. Only
the thiocyanate ion is proposed to form an inner-sphere complex (approximately 10% in
the case of nickel(II)). The sulphate value reported is an average of values determined
by UV spectroscopy and from the effects of nickel ions on the solubility of calcium sulphate. Apparently the two techniques gave reasonable agreement, as the average value
of log10 b1ο was reported as (1.8 ± 0.2). This value is smaller than most, and there are
insufficient details in the paper to allow the reviewer to properly assess or recalculate
the results. In attempting to assess the values for the present review, the SIT was used to
344
A. Discussion of selected references
calculate log10 K1ο = 1.07, again indicating that the value obtained by Mironov et al. is
anomalously low.
The authors provided insufficient experimental details to assess the reliability
of their data, although the sum of the β1 and ω1 values are in agreement with other reports.
[70NAG/ING]
The partial pressure of sulphur over Ni – S melts was derived from the equilibrium
weight of the melt in gas streams of known H2S/H2 compositions. The mass of the melt
was monitored continuously, while the composition of the gas stream was maintained
constant by appropriate flow meters. Experimental results for melts saturated with
nickel as well as for homogeneous melts, with compositions ranging from 33 to 42
at% S, were obtained at temperatures between 973 and 1373 K. The sulphur partial
pressure for a given gas composition was calculated from the well-known equilibrium
constants for the reaction:
H2S(g) U H2(g) + ½S2(g).
[70ROT]
Experimental levels of the configuration 3d94p, 3d84s4p and 3d95p of (NiI) were compared with corresponding calculated values. The electrostatic interactions between the
configurations were considered explicitly.
[70TUR/RUV]
This paper describes polarographic measurements done at (25.0 ± 0.1)°C, I = 5.0 M
(NaClO4), 1 × 10–3 M Ni(NO3)2, 0.0 – 0.2M Na2SO4 and with enough HClO4 to adjust
the pH to 3. A value of K1 = 15.6 is reported for nickel sulphate association. The ionic
strength is too high (6.6 m) to properly apply the SIT. If the SIT is used with
ε(Ni2+, ClO −4 ) equal to (0.37 ± 0.03) kg·mol–1, log10 K1ο = 1.70 is calculated.
[71ARI/MOR]
The enthalpies of formation were determined for several nickel sulphides, including
Ni3S2, NiS (various compositions of the high-temperature form) and NiS2. The results
are summarised in Table A-15. No experimental details are given. The mean value for
the standard enthalpy of formation of the NiAs-type NiS is calculated to be ∆ f H mο =
− (85.8 ± 2.4) kJ·mol–1 which agrees within the limits of uncertainty with the value recommended in the present compilation ( ∆ f H mο = – (88.1 ± 1.0) kJ·mol–1).
A. Discussion of selected references
345
Table A-15: Experimental results for the enthalpy of formation for various nickel sulphides determined by calorimetric measurements.
Composition
Phases
∆ f H mο (kJ·mol–1)
NiS0.67
Ni3S2
– (202.1 ± 3.9)
NiS0.98
Ni3S2 + NiS(α)
– (85.0 ± 2.6)
NiS1.00
NiS(α)
– (84.9 ± 2.6)
NiS1.01
NiS(α)
– (87.4 ± 2.5)
NiS1.02
NiS(α)
– (84.9 ± 2.5)
NiS1.04
NiS(α)
– (86.5 ± 2.4)
NiS1.053
NiS(α)
– (85.4 ± 2.4)
NiS1.06
NiS(α) + NiS2
– (85.7 ± 2.4)
[71BHA/SUB]
The hydrolysis of nickel(II) was studied in 0.5 M NaClO4 medium at 30°C. The concentration of the metal ion ranged from 0.01 to 0.04 M, while the pH varied between 6.8
and 7.1. The graphical analysis of the data, assuming only one hydroxo species, indicated the presence of Ni 2 (OH)62 − with a log10 *b 6,2 = – (42.94 ± 0.04). The authors used
too narrow a range of metal ion concentrations (and too narrow a range of pH for each
[Ni2+]tot) to allow a satisfactory evaluation of the composition of the oligonuclear species. The reliability of the results is questionable; therefore, the results from this publication were not used further in this review.
[71BON]
The author earlier suggested a simple calculation method for interpreting metal-ion
complexation behaviour from polarographic studies of irreversible electrode processes
[70BON]. The present paper was devoted to a test of the validity of the proposed
method. One of the major conditions is that the heterogeneous charge-transfer rate constant should not change markedly with addition of the ligand to the aquo complex, i.e.,
the removal of coordinated water also should remain the rate determining step of the
reduction in the presence of ligand. This condition was found to be fulfilled in case of
simple 1:1 complexes of nickel(II), due to the slow water exchange in both the aquo and
[Ni(H2O)5X]+ complexes. Therefore, the proposed simple method, combined with the
De Ford-Hume equation, was used to determine the stability of the monofluoro,
monochloro and mononitrato complexes of nickel(II). During the measurements, to
achieve a measurable effect of complex formation, the background electrolyte (NaClO4)
was entirely replaced by the sodium salt of the corresponding anion. The author mentioned that, taking into account the limits of the method, the strong medium effect, and
the weak complexation (especially in the cases of chloride (β1 = 0.6) and nitrate (β1 =
0.4)), the reported values can be considered only as order of magnitude estimates.
346
A. Discussion of selected references
[71BUR/ZIN]
The hydrolysis of nickel(II) in 3 M (Na)Cl was studied at 60°C by potentiometric titrations. The nickel(II) concentration was [NiCl2]tot = 0.2 to 1.5 M and pH range was from
4.9 to 6.3. The reported constants are log10 *b 4,4 = − (25.33 ± 0.02) and log10 *b1,2 =
− (8.5 ± 0.2). The medium effect and thus the change of ionic strength was very important in these measurements (I varied from 3.2 to 4.5 M, i.e., pure NiCl2 solution was
used for the highest nickel(II) concentration). Therefore, the reported experimental data
were re-evaluated for the purposes of this review. For the reaction:
Ni2+ + Cl– U NiCl+
(A.41)
at Im = 3.2 m, and T = 333.15 K, log10 b1 (A.41) = – 0.44 can be estimated for the NiCl+
complex in NaCl medium, accepting the values of log10 b1 ((A.41), 298.15 K) = – 0.59
(see discussion on [65BUR/LIL2]) and ∆ r H m = (8.2 ± 2.5) kJ·mol–1 [66KEN/LIS]. Considering the whole dataset and the same model proposed by the authors, log10 *b 4,4 =
− (25.29 ± 0.05) (3σ) and log10 *b1,2 = – (8.50 ± 0.09) can be calculated, in good agreement with the published values. If only the data sets reported for [Ni2+]tot = 0.2, 0.4, 0.6
and 0.8 M are used, taking into account the formation of NiCl+, log10 *b 4,4 =
− (24.04 ± 0.06) (3σ) and log10 *b1,2 = – (7.92 ± 0.10) can be derived. If the species
NiOH+ is added to the equilibrium model, a substantial (30%) decrease of the fitting
parameter was achieved, and the values log10 *b 4,4 = – (24.09 ± 0.03) (3σ), log10 *b1,1 =
– (8.48 ± 0.10) and log10 *b1,2 = – (8.58 ± 0.40) were obtained. The latter values, with
increased uncertainties, are accepted for use in this review: log10 *b 4,4 =
− (24.09 ± 0.30), log10 *b1,1 = – (8.48 ± 0.30) and log10 *b1,2 = – (8.58 ± 0.60). Combining these constants with the values from [65BUR/LIL2], ∆ r H m = (59 ± 10) kJ·mol–1,
(20 ± 20) kJ·mol–1 and (186 ± 6) kJ·mol–1 are calculated for the formation of NiOH+,
Ni2OH3+ and Ni 4 (OH) 44 + species, respectively (in 3.2 m aqueous NaCl).
[71CON/LOO]
Connelly et al. measured the specific heat of pure single-crystalline nickel near the Curie temperature using a calorimetric technique which permits direct observation of
C pο,m (T) as a continuous function of T, with a temperature resolution of about 10–2 K.
Special attention has been devoted to the determination of the analytical form of the
magnetic contribution to C pο,m (T).
[71DAS/DAS]
Potentiometric measurements have been performed at 25, 35 and 45°C to determine
thermodynamic parameters of the mono-thiocyanato complexes of five 3d metal ions,
among them nickel(II). The activity coefficients of the different species were calculated
by the Davies equation. Since the Davies equation is not compatible with the SIT, and
no experimental data are provided, the reported thermodynamic functions were not considered in this review.
A. Discussion of selected references
347
[71HOH/GEI]
A stopped-flow method was used to study the hydrolysis of nickel(II). The ionic
strength of the solutions was not controlled ([Ni2+]tot = 5×10–4 to 2.5×10–3 M, [OH]tot ~
3×10–4 M); the measurement temperature was not specified in the paper. The results
indicated, that using different time scales, the deprotonation of Ni(H 2 O)62 + ion and the
polynuclear hydroxo species can be separated. The deprotonation reaction for the
mononuclear species is estimated to occur within 10–7 s, while the polynuclear complex
formation needs several seconds (20 min. is reported to reach the equilibrium in
[65BUR/LIL] and [69KOL/KIL]). The pH of the solutions, measured between 2 – 1000
ms by spectrophotometry using phenolphthalein as the indicator (the value used for the
pK of the indicator was not reported), was extrapolated to t = 0 s. From these data
log10 *b1,1 = (11.0 ± 0.2) has been derived. This value has not been considered further in
this review, because of the lack of experimental details.
[71ISO]
Freezing point measurements were carried out in a manner similar to that of Brown and
Prue [55BRO/PRU], and similar results were obtained. Analysis of the saturated solutions was done by comparison of the conductivity with the conductivity of standard solutions. An extended Debye-Hückel treatment was used to analyse the data, and values
of K1 =118 to 250 dm3·mol–1 were obtained for values of the Debye-Hückel “a” parameter of 0.4 to 1.4 nm. The author reported that the calculated association constant also
depends on the choice of the maximum solute concentration for which the data analysis
can be considered reliable. In the present review the data were reanalysed using the SIT,
and only data for total nickel sulphate molalities ≤ 0.03 were used.
[71IZA/JOH]
The thermodynamic parameters for the formation of several metal-cyanide complexes,
among others those of Ni(CN) 24 − , have been determined using pH-metric and calorimetric methods at 10, 25 and 40°C. In case of nickel(II), the thermodynamic data were determined by titration of Ni(ClO4)2 solutions with NaCN solutions. The ionic strength of
the solutions were I < 0.02 M in all cases. The Debye-Hückel equation, related to the
SIT model, was used to correct the formation constants to thermodynamic constants
valid at I = 0. Since previous experiments indicated that the dependence of ∆ r H mο in the
ionic strength in dilute aqueous solutions is small compared to the experimental error,
the measured heats of reaction ( ∆ r H m = – 189.1 kJ·mol–1 at 10ºC; ∆ r H m = – 183.7
kJ·mol–1 at 40ºC) were taken to be valid at I = 0, but the uncertainties were estimated in
this review as ± 2.0 kJ·mol–1. From the values of ∆ r H m as a function of temperature,
average ∆ r C pο,m values were calculated.
348
A. Discussion of selected references
[71LAN/CUM]
The distribution of 63Ni(II) between anion exchange resin and MSCN (M = Na+, K+,
NH +4 and Li+) solutions was determined using liquid scintillation techniques. The temperature of the measurements was not given in the paper. The thiocyanate concentration
ranged between 0.05 and 5.0 M, while [63Ni(II)]tot was varied from 4.26×10–7 to
2.92×10–6 M. No background electrolyte was used, thus the ionic strength varied over a
wide range. The authors reported “semi-thermodynamic” formation constants (see
[89BJE]) for the four thiocyanato complexes detected ( Ni(SCN) 2x − x with x = 1 – 4),
using literature data for the main activity coefficients of KSCN and NaSCN solutions,
and estimated values for NH4SCN. Since the temperature of the measurements was not
specified, and the “semi-thermodynamic” formation constants are not compatible with
values extrapolated to zero ionic strength using SIT, the reported data were not considered further in this review.
[71LEB/SAV]
This paper describes the measurements of the heat of fusion of Ni and other refractory
metals by an electrical explosion method i.e., the rapid heating by a current of high density. The metal wires 0.05 – 0.1 mm in diameter and about 1 cm in length were heated
by single square current pulses of density 1.5 – 4.0×1010 A·cm–2. Under these conditions, one can neglect the self-induction, the skin effect, and the energy losses, so the
wire resistance and energy supplied can be calculated if the voltage and the time of the
pulse are known. The maximum possible systematic error in energy determination was
8%. The authors assume that in the case of Ni this error is 1.5%. Unfortunately,
∆ fus H mο (Ni) determined by the electrical explosion method ∆ fus H mο = (18.43 ± 0.27)
kJ·mol–1 is significantly different from the values obtained by calorimetric studies of
Vollmer et al. [66VOL/KOH] ∆ fus H mο = (16.90 ± 0.25) kJ·mol–1, and Geoffray et al.
[63GEO/FER] ∆ fus H mο = (17.47 ± 0.23) kJ·mol–1 upon which our selected value is
based. One possible reason for the observed differences could be the low-purity of
nickel, 99.5%, used in the study of Lebedev and Savvatimskii.
[71NAV]
Navrotsky has measured the enthalpies of formation of Mg2SiO4, Co2SiO4, Ni2SiO4,
Zn2SiO4, MgGeO3, Co2GeO4, CoGeO3, Ni2GeO4 and Zn2GeO4 from the component
oxides by solution calorimetry in a molten oxide solvent at (965 ± 2) K. Regularities in
the thermodynamics of formation of the mentioned silicates and germanates were discussed. In the case of Ni2GeO4, 9PbO·3CdO·4B2O3 was used as a solvent and the enthalpy of formation from oxides was determined to be – 39.75 kJ·mol–1. This value is
identical with that previously given by Navrotsky and Kleppa [68NAV/KLE]. Using the
calorimetrically determined enthalpy of formation from oxides for Ni2SiO4, an attempt
was made to estimate the temperature of decomposition of Ni2SiO4 into NiO and SiO2.
The obtained value, 2065 K, is significantly higher then the value accepted by this review (1820 ± 5) K.
A. Discussion of selected references
349
[71OHT/BIE]
The hydrolysis of nickel(II) ion was studied by potentiometric titrations in 3 M (Na)Cl
medium at 25°C. The concentration of nickel(II) ranged between 0.0145 and 1.0 M,
thus the ionic strength varied from 3.0145 to 4 M. The evaluation of emf data, performed with both graphical and numerical methods, indicated the dominant formation
of Ni 4 (OH) 44 + , together with some minor species NiOH+ and Ni2OH3+. The best fit to
data was obtained by the following formation constants: log10 *b 4,4 = – (28.55 ± 0.02),
log10 *b1,1 = – (10.5 ± 0.1) and log10 *b1,2 = – (10.5 ± 0.5). The authors attempted to consider the impact of formation of the NiCl+ complex, but they considerably underestimated this effect.
Accepting for the reaction,
Ni2+ + Cl– U NiCl+
(A.42)
log10 b1 (A.42) = – 0.59 in 3.2 m NaCl medium (see the discussion on [65BUR/LIL2]),
ca. 38 – 42 % of nickel(II) is present as the chloro complex. Considering this in the
mass action equations, the corrected formation constants are as follows: log10 *b 4,4 =
− (27.60 ± 0.02), log10 *b1,1 = − (10.26 ± 0.10) and log10 *b1,2 = – (10.0 ± 0.5).
Taking into account the medium effect caused by replacement of NaCl by
NiCl2, the changing ionic strength and the uncertainties of the latter correction
log10 *b 4,4 = – (27.6 ± 0.3), log10 *b1,1 = – (10.26 ± 0.20) and log10 *b1,2 = – (10.0 ± 0.6)
are accepted in this review.
[71PAA/HUM]
This is a spectrophotometric study on nickel(II) chloride complexes formed in aqueous
NaClO4/LiCl solution. Complex formation was studied first at I = 6 M. The lithium
chloride concentration was varied between 0 and 6 M; the ionic strength was maintained
by sodium perchlorate. Under these conditions the results indicated the unique formation of the NiCl+ complex and its conditional stability constant is reported to be β1 =
(0.32 ± 0.05). During this study, both the anion and the cation of the original background electrolyte were entirely replaced, which induced additional uncertainty in the
value determined for the formation constant. The impact of the ClO −4 → Cl– change is
difficult to quantify, but the discussion on [70HAL/VAN] may give an approximation.
The impact of the Na+ → Li+ change is easier to assess. As discussed in [97GRE/PLY2]
the following expression can be applied for log10 b1 using the SIT at constant ionic
strength and total perchlorate concentration:
∆ log10 b1 = log10 b1,NaClO4 – log10 b1,(Na,Li)ClO 4 = [ε(Na+, Cl–) – ε(Li+, Cl–)] mLi+
At I = 6 M, and in case of a complete Na+ → Li+ exchange, ∆ log10 b1 = – 0.48
can be calculated. It means that the replacement of both the cation and the anion of the
background electrolyte each induces (as a rough approximation) an increase of 0.5 in
the determined value of log10 b1 compared with the “true” value in NaClO4 (it is very
350
A. Discussion of selected references
difficult to say if these effects are additive or not). Therefore, the reported constant can
not be accepted.
At LiCl concentrations above 6 M, the formation of a further complex
(NiCl2(aq)) was detected. Since the ionic strength between [LiCl] = 6–13.68 M could
not be maintained constant, the activity of chloride ion and water was taken into account, while the ratios γ Ni ( H O )2+ / γ NiCl( H O )+ and γ NiCl( H O )+ / γ NiCl ( H O ) (aq) were assumed
2
2
2
2
2
5
4
6
5
to be nearly constant with increasing LiCl concentrations and were included in the reο'
ported “semi-thermodynamic” formation constants ( b x = [ NiCl x (H 2 O)62−− xx ]×aw /
[ NiCl x −1 (H 2 O)37−− xx ]×aCl–, where x = 1, 2). The reported values are b1ο ' = (4.9 ± 0.8)×10–2,
and b 2ο ' = (7.8 ± 4.1)×10–5. Although the use of such constants generally provide higher
correctness in case of weak complexes (see the discussion on [89BJE]), b1ο ' was obtained from the rather uncertain value of β1 valid for I = 6 M, thus only K 2ο ' = b 2ο ' / b1ο '
can be accepted in this review ( log10 K 2ο ' = – (2.8 ± 0.5)).
[71PIE/HUG]
The equilibrium constant for the reaction,
Ni(CN) 24 − + CN– U Ni(CN)35−
K5
was determined by spectrophotometry (400 – 500 nm) at 20°C, I = 2 M (NaClO4(aq)
containing 0.01 M NaOH (pH ~ 12)). The formation of the hexacyano complex was not
observed even at a 100-fold excess of CN– over Ni(CN) 24 − . The reported stability constant ( log10 K 5 = – 0.77) was corrected to 25°C ( log10 K 5 = – 0.80), using the selected
reaction enthalpy.
[71TUR/MAL]
The effects of temperature (t = 5 to 35°C) and ionic strength (I = 0.075 to 0.2 M, KNO3
+ KSCN) on the stability and kinetics of formation of nickel(II)-thiocyanate complexes
were investigated at pH = 6.0 – 6.2, by detecting the catalytic polarographic currents in
the nickel(II) – thiocyanate system. The metal ion concentration varied from 0.2 to 9.5
mM, while the thiocyanate concentration ranged between 4 and 40 mM. Under the conditions used, the formation of Ni(SCN)+ and Ni(SCN)2(aq) was detected. The reported
formation constants are accepted, but uncertainties of ± 0.1 and ± 0.2 are assigned to
log10 b1 and log10 b 2 , respectively. From the temperature dependence of the formation
constants (see Figure A-20) ∆ r H m = – (14.3 ± 0.5) kJ·mol–1 and ∆ r H m = − (29.5 ± 0.5)
kJ·mol–1 can be derived for the formation of Ni(SCN)+ and Ni(SCN)2(aq), respectively.
These values are selected in this review with both uncertainty values increased to ± 2.0
kJ·mol–1.
A. Discussion of selected references
351
Figure A-20: Temperature dependence of log10 b1 (filled squares) and log10 b 2 (open
squares) values for the Ni(II) – (SCN)– system (I = 0.2 M) as reported in
[71TUR/MAL].
2.5
log10 βx
2.0
1.5
1.0
0.0032
0.0033
0.0034
0.0035
0.0036
K /T
[72AUF/CAR]
The heats of solution in water of the hexahydrate (actually Ni(NO3)2·6.055H2O) and the
tetrahydrate were measured as a function of final solution concentration. Extrapolated
“infinite dilution” values of 31.23 kJ·mol–1 and 8.66 kJ·mol–1 were reported. The authors did separate extrapolations for the two sets of measurements (and also for the
measurements using the dihydrate [72AUF/CAR2]). The enthalpy of dilution correction
to obtain a value for the enthalpy of solution at I = 0 should be identical (within the experimental uncertainties) for the hexahydrate, tetrahydrate and dihydrate
[72AUF/CAR2], and the correction should be consistent with whatever correction is
applied to the heat of solution results of Chukurov et al. [73CHU/DRA] (for the tetrahydrate). The experimental scatter is fairly large (as much as 0.4 kJ·mol–1) for final
nickel concentrations < 0.015 m. It would appear that the enthalpy of dilution curve is
similar to those for potassium sulphate and barium nitrate [59LAN].
For each of the three salts [72AUF/CAR], [72AUF/CAR2], the values for the
five lowest concentrations were averaged to estimate a value at the third-lowest concentration with a solvent:salt ratio of 8000:1. The difference between each experimental
heat of solution measurement and the measurement (for the same hydrate) resulting in a
solution with a solvent:salt ratio of 8000:1 (0.00694 m) was calculated. Only values
from the measurements with final solution concentrations below 0.056 m were used.
Different values from 650 J·mol–1 to 200 J·mol–1 were assumed for the heat of dilution
to zero molality of the solution with a solvent:salt ratio of 8000:1 (0.00694 m). Chukurov et al. [73CHU/DRA] provided an equation for heat of solution of the tetrahydrate in
the concentration range 0.007 m to 0.03 m, and difference values based on that equation
A. Discussion of selected references
352
and the same assumed heat of dilution were also calculated. The derived heats of dilution were plotted as a function of m½ (Figure A-21), as was a line with the limiting slope
for 1:2 and 2:1 electrolytes [59LAN].
If the enthalpy of dilution from a solution with a solvent:salt ratio of 8000:1 to
infinite dilution is assumed to be of the order of (520 ± 100) J·mol–1, all the experimental results enthalpy of solution are in reasonable agreement with each other. However,
the values including those of Chukurov et al. [73CHU/DRA] at the lower concentrations) are compatible with the limiting-law slopes found for other 1:2 and 2:1 electrolytes [59LAN].
Figure A-21: Heats of dilution of hydrated nickel nitrate salts based on the experimental
heats of solution and an assumed heat of dilution to infinite dilution of 520 J·mol–1 for a
solution with a solvent:salt ratio of 8000:1. Heat of solution values for [73CHU/DRA]
were based on the equation of the concentration dependence as reported in that paper
(see text and Appendix A entry for [73CHU/DRA]).
heat of dilution / J·mol
−1
2000
1500
theoretical limiting slope
[73CHU/DRA] tetrahydrate
[72AUF/CAR] tetrahydrate
[72AUF/CAR] hexahydrate
[72AUF/CAR2] dihydrate
1000
500
0
0.00
0.05
0.10
0.15
0.20
0.25
−1 0.5
(m / mol·kg )
Using the estimated heat of dilution and average values of the heats of solution
at concentrations at or below 0.01 m, the heats of solution corrected to “infinite dilution” for the hexahydrate, tetrahydrate and dihydrate are (30.76 ± 1.00) kJ·mol–1,
(8.23 ± 1.00) kJ·mol–1 and – (27.64 ± 1.00) kJ·mol–1. The uncertainties have been estimated in the present review based on the values shown in the authors’ Figure 1 and systematic uncertainties generally associated with similar measurements.
A. Discussion of selected references
353
The equilibrium water vapour pressure at which the hexahydrate begins to lose
water to form a lower hydrate was determined from 15°C to 42°C. The vapour pressure
was found to follow the equation:
log10( pH2 O /atm) = ((7.7226 – 3039.2·T–1 ) ± 0.078)
or
pH2 O /bar = 1.01325 × 10((7.7226 − 3039.2⋅T
= 10((7.7283− 3039.2⋅T
−1
−1
) ± 0.078)
) ± 0.078)
and at 298.15 K, if the reaction is assumed to be:
Ni(NO3)2·6.00H2O(cr) U Ni(NO3)2·4.00H2O(cr) + 2H2O(g)
(A.43)
∆ r Gmο ((A.43), 298.15 K) = (28.14 ± 0.89) kJ·mol–1 and ∆ r H mο ((A.43), 298.15 K) =
(116.4 ± 10.9) kJ·mol–1. If the heat of vaporisation is calculated from the enthalpy of
solution measurements, and the accepted values for the enthalpies of formation of
H2O(l, g), ∆ r H mο ((A.43), 298.15 K) = (110.4 ± 1.3) kJ·mol–1. The enthalpy of reaction
value from the vapour pressure measurements agrees with the value from calorimetry
within the combined uncertainties, though the uncertainty is much lower for the value
from calorimetry. Unfortunately, the method used to measure the water vapour pressure
does not allow unambiguous identification of the lower hydrate.
[72AUF/CAR2]
The heat of solution in water of the dihydrate (Ni(NO3)2·2.00H2O) was measured as a
function of final solution concentration. An extrapolated “infinite dilution” value of
− (27.41 ± 1.00) kJ·mol–1 was reported, and a recalculation has been done in the present
review (see the Appendix A for [72AUF/CAR]).
The water vapour pressure over mixtures of the tetrahydrate and dihydrate
were also measured from 32°C to 73°C. The vapour pressure was found to follow the
equation
log10( pH2 O /atm)= ((7.5527 – 3179.4 ·T–1 ) ± 0.094)
or
pH2 O /bar = 1.01325 × 10((7.5527 − 3179.4⋅T
= 10((7.5584 − 3179.4⋅T
−1
−1
) ± 0.094)
) ± 0.094)
and at 298.15 K, if the reaction is assumed to be:
Ni(NO3)2·4.00H2O(cr) U Ni(NO3)2·2.00H2O(cr) + 2H2O(g)
ο
m
–1
ο
m
(A.44)
∆ r G ((A.44), 298.15 K) = (35.45 ± 1.09) kJ·mol and ∆ r H ((A.44), 298.15 K) =
(121.7 ± 9.3) kJ·mol–1. If the heat of vaporisation is calculated from the enthalpy of solution measurements, and the accepted values for the enthalpies of formation of
H2O(l, g), ∆ r H mο ((A.44), 298.15 K) = (124.0 ± 1.3) kJ·mol–1. The enthalpy of reaction
value from the vapour pressure measurements agrees with the value from calorimetry
354
A. Discussion of selected references
within the combined uncertainties, though the uncertainty is much lower for the value
from calorimetry. Unfortunately, the method used to measure the water vapour pressure
does not allow unambiguous identification of the lower hydrate, and no value for
∆ r Gmο ((A.44), 298.15 K) can be accepted based on this work.
[72BON/HEF]
The fluoride complexes of the divalent ions from Mn(II) to Zn(II) have been studied by
potentiometric titration, using a fluoride selective electrode at 298.15 K and I = 1.0 M
NaClO4. Under the experimental conditions employed (25 cm3 0.05 M metal(II) perchlorate solution was titrated by 10 cm3 0.004 M NaF solution), only monofluoride
complexes are formed but in rather low concentrations (at the end of titrations [NiF+] ~
2.2×10–6 M can be calculated). This yielded only a moderate effect (1 – 2 mV) in the
measured electrode potential, therefore we estimate higher uncertainty in K1 ((2.2 ± 1.0)
for NiF+) in the present review, than suggested by the authors. The authors reported an
anomalous stability sequence in the Mn(II)-Zn(II) series which does not fit an IrvingWilliams type trend. This behaviour was later confirmed by several authors (see e.g.,
[83SOL/BON]).
[72FLE]
The high temperature phase α-Ni7S6 was synthesised by heating Ni sponge and S crystals in an evacuated silica glass tube at 504°C for 9 days; the Ni sponge was reduced by
hydrogen at 900°C before use.
[72FRE/STU]
Complexation concentration quotients for 15°C and a medium of 0.1 M KNO3(aq) were
reported for the formation of NiHPO4 (K(A.47) = 100), NiP2 O72 − ( log10 K (A.45) =
6.22) and NiHP2 O7− ( log10 K (A.46)= 3.50).
Ni2+ + P2 O74 − U NiP2 O72 −
(A.45)
Ni2+ + HP2 O37− U NiHP2 O7−
(A.46)
Ni2+ + HPO 24 − U NiHPO4(aq)
(A.47)
The values for the pyrophosphate species were cited and used in a later paper
from this group [78FRE/STU]. The potentiometric measurements were carried out by
titration of the free acid with KOH(aq) solution in the presence and absence of Ni(NO3)2
in the solutions. The metal:ligand concentrations were varied from 1:2 to 2:1. Ligand
concentrations were 2×10–3 to 5×10–3 M. The authors estimated 5% to 10% uncertainties in the experimentally determined stability constant values. The electrodes were
calibrated against pH buffers rather than concentration standards, and the hydrogen ion
activity coefficient, γ H+ was assigned the value of 0.83; the SIT procedure described in
Appendix B would have led to γ H+ = 0.79. If the concentration quotients are adjusted
A. Discussion of selected references
355
for this, the value for formation of NiHPO4 is K = 105) and for NiP2 O72 − ,
log10 K (A.45) = 6.24 (the value for log10 K (A.46) is unchanged).
Correction to zero ionic strength (Appendix B) gives log10 K ο (A.47) =
(2.88 ± 0.30), where the uncertainty has been assigned in the present review. The small
∆ε correction has been based on the value for ε(Na+, HPO 24 − ) from Table B-5 (rather
than using the equation in Table B-6), and ε(Ni2+, NO3− ) = 0.18 kg·mol–1.
For Reaction (A.45) in 0.1 M KNO3, log10 K (A.45) at 15°C = (7.94 + 0.1∆ε).
For Reaction (A.46) in 0.1 M KNO3, log10 K (A.46) at 15°C = (4.79 + 0.1∆ε).
As the ionic strength is low, the effect of the ∆ε term on the final calculated
values of log10 K ο is limited. Rather than use estimated ε values for interactions involving the pyrophosphate moieties (some of which are highly charged), it is simply assumed that ∆ε is zero. As a consequence, the uncertainty in the log10 K ο values is increased to ± 0.25 for the values from measurements in 0.1 M KNO3 solutions. As discussed in Appendix A for [73PER/SEC], the fact that the calculations were done without directly considering the effect of association of K+ with pyrophosphate means that
these are “apparent” values (useful for comparisons, but not adequate as selected values). The raw data are unavailable for re-analysis.
[72KLE/KOM]
Based on thermal and X-ray measurements the complete Ni–Te phase diagram was constructed.
[72POL/HER]
Polgar et al. carried out calorimetric measurements of the specific heat of
NiCl2·2H2O(cr) for temperatures between 1.2 and 25.5 K. Peaks were observed at 6.31
and 7.26 K, and at lower temperatures the heat capacity values are not a simple function
of T 3. The reasons for these magnetic transitions are not clear. The authors estimated
that the contribution to the molar entropy between 0 and 1.2 K is 0.03 R. In the present
review, the uncertainty in the extrapolation below 1.5 K is estimated to be of the order
of ± 0.2 J·K–1·mol–1.
[72PRE/RUG]
Predel and Ruge reported a value of – 8600 cal·g-atom–1 for the heat of formation of
NiAs(cr) based on an unreported value for the heat of solution of NiAs in liquid tin over
an unspecified temperature range. Based on another paper from the same group
[72PRE/RUG2], it would appear that the reaction was carried out between 600 and
900 K. The value is not claimed to be a value corrected to 298.15 K (though Mah and
Pankratz [76MAH/PAN], Barin et al. [77BAR/KNA], and Skeaff et al. [85SKE/MAI]
appear to have accepted it as the value for that temperature). If the value is assumed to
be an average value obtained at 800 K, and the extrapolated heat capacity function of
A. Discussion of selected references
356
Muldagalieva et al. [95MUL/CHU] is applied (with values for H m (800 K) –
H mο (298 K) for Ni(cr) from the present review and for As(cr) from Herrick and Feber
[68HER/FEB]), the value:
∆ f H mο (NiAs, cr, 298.15 K) = – 71.965 – 30.21 + 15.42 + 13.15 kJ·mol–1,
i.e., ∆ f H mο (NiAs, cr, 298.15 K) = – 73.60 kJ·mol–1 can be calculated. The authors
claimed that their results were accurate within 5%. In the present review, this is considered to be a 1σ uncertainty, and an overall uncertainty of ± 8.00 kJ·mol–1 is accepted.
[72RET/HUM]
The “semi-thermodynamic” equilibrium constant ( K 2ο ' ) for the reaction:
NiBr(H 2 O)5+ + Br– U NiBr2(H2O)4(aq) + H2O(l)
was estimated from spectrophotometric measurements in 8.5 – 11.7 M LiBr solutions.
In the evaluation of K 2ο ' , several simplifying assumptions were made: (i) the activity
coefficient of bromide ion can be approximated by the mean activity coefficients of
LiBr, (ii) the ratio of the activity coefficients of the two Ni bromide complexes is constant with increasing LiBr concentration, (iii) in 7 – 8 M LiBr solution the only
nickel(II)-containing species is the NiBr(H 2 O)5+ complex.
[73CHU/DRA]
The heat of solution of Ni(NO3)2·4H2O in water was found to follow the equation:
∆ sol H mο = (8.494 + 7.448 m0.5) kJ·mol–1
for final solution concentrations of nickel from 0.007 m to 0.03 m. Although the value
as extrapolated to I = 0 (8.49 kJ·mol–1) is very similar to the one obtained by Auffredic
et al. [72AUF/CAR] (8.66 kJ·mol–1), the concentration dependence (and hence, it is
assumed, the actual values for the measurements at finite concentrations) is less similar—with differences of 0.2 kJ·mol–1 to 0.4 kJ·mol–1 in the concentration range of the
measurements (see the Appendix A discussion for [72AUF/CAR] for further details).
Using the estimated heat of dilution from 0.007 m, and the value of the heat of solution
as calculated using the authors’ equation for that concentration, the heat of solution for
the tetrahydrate, corrected to “infinite dilution”, is (8.59 ± 1.00) kJ·mol–1. The uncertainty is estimated in the present review.
The authors also reported the heats of dissolution of “anhydrous” Ni(NO3)2
(− (117.2 ± 1.67) kJ·mol–1), water (– (5.31 ± 0.08) kJ·mol–1) and Ni(NO3)2·4H2O
(− (63.18 ± 0.84) kJ·mol–1) in dimethyl sulphoxide. This would indicate an enthalpy of
(75.23 ± 1.90) kJ·mol–1 for the dehydration reaction:
Ni(NO3)2·4H2O U Ni(NO3)2 + 4H2O.
Except for salts containing singly charged cations, synthesis of anhydrous nitrate salts is very difficult, as the final steps of dehydration tend to cause loss of nitrogen
A. Discussion of selected references
357
oxides or nitric acid. Only the summary of the paper was available to the reviewers, not
the full deposited document, and, in the summary, there were no details on the preparation of the anhydrous salt. The experimental approach used to determine the enthalpy of
formation is reasonable, and may lead to a useful value if the deposited paper contains
proper analysis results for the solid and solutions. No value for anhydrous Ni(NO3)2 is
selected from these data.
[73COR]
Cordfunke measured the solubility of Ni(IO3)2·2H2O between 29°C and 61°C. It is not
clear from the text whether a pure sample of the dihydrate was used for the measurements or whether the samples were contaminated with the tetrahydrate. Also, the equilibration times are not reported.
The results were shown graphically, and it was necessary to retrieve values by
digitisation of the points plotted in the author’s Figure 2. These values (percent weight
as Ni(IO3)2) were converted to molalities, and the apparent solubility products were
corrected to I = 0. These were plotted against (T/K)–1 to obtain values for log10 K sο,0 at
298.15 K and an average value of ∆ r H mο (A.48) over the temperature range of the measurements.
Ni(IO3)2·2H2O U Ni2+ + 2 IO3− + 3H2O(l)
(A.48)
If only the values to 50°C are included, log10 K sο,0 ((A.48), Ni(IO3)2·2H2O,
298.15 K) = – (5.18 ± 0.04) and ∆ r H mο (A.48) = (21.7 ± 4.7) kJ·mol–1, whereas, if the
points for temperatures above 50°C also are included, log10 K sο,0 ((A.48), Ni(IO3)2·2H2O,
298.15 K) = − (5.17 ± 0.04) and ∆ r H mο (A.48) = (20.5 ± 2.4) kJ·mol–1.
[73FED/SHM]
The solubility of Ni(IO3)2 in nitrate containing LiClO4 solutions (I = 0.5, 1.0, 2.0, 3.0,
4.0 M) has been measured to determine the stability and composition of the nitrato
complexes formed with nickel(II). The authors reported the formation of Ni(NO3 ) 2x − x
complexes (x = 1 – 3). Since, during the measurements, the original background electrolyte (LiClO4) was entirely replaced by LiNO3, the formation of bis- and tris-complexes
can not be verified. Therefore, the experimental data published in [73FED/SHM] were
re-evaluated taking into account solely the formation of NiNO3+ species. To minimise
the medium effect, only the experimental data [LiNO3]/[LiClO4] ≤ 1 (see Figure A-22)
were taken into account. The non-linear regression of these data resulted in the following values: log10 b1 = − (0.20 ± 0.40) (I = 1.0 M), log10 b1 = − (0.21 ± 0.40) (I = 2.0 M),
log10 b1 = − (0.30 ± 0.40) (I = 3.0 M), log10 b1 = − (0.01 ± 0.40) (I = 4.0 M). The uncertainties were assigned by the reviewer, taking into account the still substantial medium
effect (also see Section V.6.1.2).
A. Discussion of selected references
358
Figure A-22: Experimental [73FED/SHM] and calculated solubility of nickel(II) in the
aqueous phase with increasing nitrate concentrations.
0.035
s(Ni2+) / M
0.03
0.025
I=1M
I=2M
0.02
I=3M
I=4M
0.015
0
0.5
1
1.5
2
2.5
3
3.5
4
[NO3-] / M
The values for the solubility of Ni(IO3)2·3H2O in aqueous 0.5 M to 4.0 M
LiClO4 solutions (i.e., those solutions with no added nitrate) were used to calculate the
value of the solubility product of Ni(IO3)2·3H2O (Figure A-23) from a standard SIT
plot. This gives the value log10 K sο,0 = – (5.09 ± 0.16).
Figure A-23: Experimental [73FED/SHM] and calculated values of the solubility of
Ni(IO3)2·3H2O in aqueous LiClO4 solutions.
-5.00
log10 Ks - 6D
-5.40
-5.80
-6.20
-6.60
0.00
1.00
2.00
3.00
−1
I / mol·kg
4.00
5.00
A. Discussion of selected references
359
[73FLE]
The data concerning NiO are essentially the same as in [68CHA/FLE].
[73GUR/CAL]
The heat of solution of nickel chloride hexahydrate into 1.0M HCl(aq) was measured,
and (6.996 ± 0.133) kJ·mol–1 was reported (2σ uncertainties). The final solution was
0.02 to 0.03 M in Ni. A double correction to 0.0 M HCl(aq) and to I =0 [95MAN/KOR]
would be required to use the reported value to calculate the enthalpy of solution of the
salt at infinite dilution in water. The corrections themselves have large uncertainties,
and this heat of solution value is not used further in the present review.
[73HUT/HIG]
A kinetic method, based on the metal ion catalysed reaction of:
[Co(EDTA)Cl]2– U [Co(EDTA)]– + Cl–,
has been used to determine the stability constants of several [MA]– complexes, among
others those of the chloro-, bromo-, nitrato- and thiocyanato complexes of nickel(II)
(pH = 4.5 – 5.5, T = 298 K, I = 1 M Na(ClO4, X)). The concentration of nickel(II) was
varied up to 0.051 M. The highest concentrations of the complex-forming anions (X–)
for nickel(II) in a given series were [Br–] = 0.49 M, [Cl–] = 0.278, 0.466 and 0.844 M,
[ NO3− ] = 0.422 M and [SCN–] = 0.15, 0.318 M. The formation constant determined at
[Cl–] = 0.844 M was not considered in this review, in order to minimise the need to assess the medium effect. The authors acknowledged several disadvantages of the chosen
reaction. The most important is that the monocationic complex [NiX]+, not just Ni2+,
may catalyse the reaction. This effect was taken into account by a simple correction
factor. Therefore, we assigned higher uncertainties to the reported values. The formation constants of chloro-, nitrato- and thiocyanato- complexes also were determined at
318.15 K. The stability of the nitrato complex does not change at higher temperatures,
but those of the chloro and thiocyanato complexes decrease moderately; this results in
negative ∆ r H m values. This observation is the opposite to what was observed in
[60LIS/ROS] and [66KEN/LIS] for the NiCl+ complexes, but is in reasonable agreement with other reports [67NAN/TOR], [74KUL3] in the case of NiSCN+. Based on the
entropies of association, the formation of mostly outer-sphere complexes (ion pairs) was
suggested in the case of nickel(II), except for thiocyanate. From the temperature dependence of the formation constant for the NiSCN+ species ∆ r H m = – (8.3 ± 1.0)
kJ·mol−1 can be derived. This value is accepted with an increased uncertainty ( ∆ r H m =
− (8.3 ± 2.0) kJ·mol–1).
[73KAT]
This is a report on a careful conductometric study. Sets of data were obtained for 5°C
intervals for temperatures between 0 and 45°C. The sets of equivalent conductivities are
reported for specific concentrations from 0.0002 to 0.003 M, and have the appearance
360
A. Discussion of selected references
of being smoothed values from unreported primary data. In the data analysis, the ionsize parameter in the Debye-Hückel equation was optimised for each temperature. It is
not clear that this is consistent with the use of the Fuoss and Accascina conductance
equation [59FUO/ACC]. The result is a slightly greater temperature dependence of the
association constant than if a constant value were used for the ion-size parameter. Nevertheless, the temperature dependence is calculated to be less than found in the potentiometric study of Nair and Nancollas [59NAI/NAN].
For the purposes of the present review, the data have been re-analysed using a
version of the Lee–Wheaton equation modified so that the activity-coefficient equation
matches the SIT equation usually used in the TDB project (Appendix B). At comparable
temperatures, the calculated nickel sulphate association constants are slightly larger than
those recalculated from the results of other conductivity studies [1895FRA],
[76SHI/TSU], [79FIS/FOX]. In the present review, we estimate the uncertainty in K1 at
25°C to be ± 20. The same uncertainty is estimated for 20 and 30°C, but for temperatures above 30°C and below 20°C, an uncertainty of ± 30 is estimated.
The average enthalpy of reaction for 0 to 45°C was recalculated, and found to
be 5.3 kJ·mol–1. This value is accepted for 25°C, and an uncertainty of ± 2.0 kJ·mol–1 is
assigned, in part by considering separately the average enthalpies of reaction calculated
from the association constants for the four highest temperatures (7.5 kJ·mol–1) and for
the four lowest temperatures (3.3 kJ·mol–1).
[73KAW/OTS]
The hydrolytic reactions of nickel(II) ion were studied at 25°C in water and dioxanewater mixtures containing 3 M (Li)ClO4 as an ionic medium. The total nickel(II) ion
concentration, B, varied between 0.025 – 0.8 M (I = 3.025 – 3.8 M) in water and the
average number of hydrogen ions set free per nickel atom, Z, was measured as a function of pH. In the aqueous solutions a range of 5.7 ≤ pH ≤ 7.3 was studied. Approximate
values of the solubility constants of Ni(OH)2(s) ( log10 *K s ,0 = (13.5 ± 0.5)) and
Ni(OH)ClO4(s) ( log10 *K s ,0 = (6.9 ± 0.5)) were estimated from maximum values of Z.
The former constant, converted into the molality scale with density data taken from
Lobo and Quaresma [89LOB/QUA] (Part B), and extrapolated to zero ionic strength,
has been given in Table V-6 and Figure V-11. At lower Z values, the function Z = Z(pH)
can be interpreted by the formation of Ni 4 (OH) 44 + as the predominant hydrolysis product. None of the minor species NiOH+ and Ni2OH3+ could be established from the data
in this study. In aqueous solution log10 *b 4,4 = – (27.32 ± 0.08) has been obtained,
which is in good agreement with the corresponding value reported in [65BUR/LIL].
This value has been accepted for use in the present review, but with the uncertainty increased to ± 0.16.
A. Discussion of selected references
361
[73NAV]
The enthalpy of the olivine – spinel transition of Ni2SiO4 was obtained by measuring the
heat of solution of both polymorphs in a molten oxide solvent, 2PbO·B2O3, at 713°C.
The following values were found: enthalpy of the transition Ni2SiO4(oliv) →
Ni2SiO4(spin), ∆ trs H m (986 K) = (5.9 ± 2.9) kJ·mol–1, the heat content increments,
H m (986 K) – H mο (298 K), olivine, (107.7 ± 1.8) kJ·mol–1, spinel, (106.2 ± 0.8)
kJ·mol−1 and the entropy of olivine – spinel transition ∆ trs S mο = 12.6 to 14.6 J·K–1·mol–1.
[73NAV2]
Navrotsky reported an experimental approach that was used to obtain values of Gibbs
energy differences between olivine, spinel, and phenacite forms of silicates and germanates from the thermodynamics of terminal solid solutions in ternary systems. This is
applied to the systems NiO − MgO – GeO2 and CoO – MgO – GeO2 at 1200°C in air
and to the system NiO − MgO – GeO2 at 800°C and 0.57 kbar water pressure. The
Gibbs energy of transformation from the olivine to spinel structure for Ni2GeO4 at
1200°C was estimated to be – 34.31 kJ·mol–1.
[73NOV/COS]
The authors point out that Ni(OH)2 phases of different solubilities were obtained depending on the speed of precipitation. The solubility products given at 25°C for I = 0.55
M NaCl and I = 1.0 M NaClO4 seem to have been calculated from Equation (A.49):
log10 K s ,0 = log10([Ni2+]·[H+]–2) + 2 log10 K W .
(A.49)
The authors used Ni2+ and presumably H+ concentrations in the first term on
the right hand side of Equation (A.49) but the ionic activity product of H2O in the second term. Thus, adequately corrected numerical values are given in Table V-6 and Figure V-11.
[73PER/SEC]
The complexation of Ni2+ with partially protonated pyrophosphate ion was studied as a
function of ionic strength (KNO3, 0.02 M to 0.20 M), temperature (5 to 35°C) for 0.1 M
KNO3 in solutions with pH values from 4.5 to 7.0. Values are reported for the formation
constants of NiP2 O72 − and NiHP2 O7− .
Ni2+ + P2 O74 − U NiP2 O72 −
(A.50)
Ni2+ + HP2 O37− U NiHP2 O7−
(A.51)
Values for the first and second protonation constants for P2 O74 − also were determined for the same ionic media and temperatures. The ionic strengths are sufficiently
low that it is unlikely that meaningful ∆ε values can be established for any of these reactions based on the data in this paper alone. Furthermore, the pH electrode appears to
have been calibrated against standard buffer solutions, with no allowance being made
362
A. Discussion of selected references
for different junction potentials in the solutions of different ionic strength. The values of
the reported 25°C constants after correction for the extended Debye-Hückel terms (nD,
cf. Appendix B), but without consideration of complex formation between K+ and pyrophosphate ions, are plotted in Figure A-24.
The corrected complexation constants of pyrophosphate with nickel show a
marked ionic strength dependence that would correspond to ∆ε values of (2.5 ± 0.7)
kg·mol–1 and (1.7 ± 0.7) kg·mol–1. The former, at least, is inconsistent with any reasonable selection of ε values (though the same is not true for the ∆ε values for the first protonation constant for pyrophosphate, which also involves highly charged ions). Based
on these calculations, only the results from the higher ionic strength solutions (I = 0.1 M
and 0.2 M) are used in the present review.
For Reaction (A.50) in 0.1 M KNO3:
log10 K (A.50) = 5.94 and log10 K ο (A.50) = (7.69 + 0.1∆ε) at 25°C, log10 K (A.50) =
5.81 and log10 K ο (A.50) = (7.53 + 0.1∆ε) at 15°C, in 0.2 M KNO3, log10 K (A.50) =
5.60 and log10 K ο (A.50) = (7.78 + 0.2∆ε) at 25°C.
For Reaction (A.51) in 0.1 M KNO3:
log10 K (A.51) = 3.71 and log10 K ο (A.51) = (5.02 + 0.1∆ε) at 25°C, log10 K (A.51) =
3.55 and log10 K ο (A.51) = (4.84 + 0.1∆ε) at 15°C, in 0.2 M KNO3, log10 K (A.51) =
3.39 and log10 K ο (A.51) = (5.02 + 0.2∆ε) at 25°C.
As the ionic strength is low, the effect of the ∆ε term on the final calculated
values of log10 K ο is limited. Rather than use estimated ε values for interactions involving the pyrophosphate moieties (some of which are highly charged), it is simply assumed that ∆ε is zero. As a consequence, the uncertainty in the log10 K ο values is increased to ± 0.25 for the values from measurements in 0.1 M KNO3 solutions, and to
± 0.3 for the values from measurements in 0.2 M KNO3 solutions.
There is a further complication. Within the context of a standard (or extended)
Debye-Hückel treatment of activity coefficients, the interaction of the highly-charged
pyrophosphate ion with alkali metal ions can be considered most easily by assuming
weak complex formation [49MON], [57LAM/WAT]. The same is true if the Specific
Ion Interaction approach described in Appendix B is used. De Stefano et al.
[94STE/FOT] measured the protonation constants for pyrophosphate, as a function of
temperature between 5 and 45°C, in aqueous solutions of (CH3)4NCl, NaCl and KCl
(0.00 to 0.75 M). The values for a 0.1 M KCl(aq) medium agreed well (within 0.05 in
log10 K ) with the values reported for a 0.1 M KNO3(aq) medium in the studies of
Perlmutter-Hayman and Secco [73PER/SEC]. From their apparent protonation constants, De Stefano et al. [94STE/FOT] calculated values of the formation constants of
KP2 O37− and KHP2 O72 − (and of the formation constants for the corresponding sodiumion species).
A. Discussion of selected references
363
Selection of values for the association constants for pyrophosphate ion with alkali metal ions is beyond the scope of the present review. If the formation constants
values for KP2 O37− and KHP2 O72 − from [94STE/FOT] are used with the values of the
pyrophosphate protonation constants from the same source, and if the nickel pyrophosphate complexation constants from Hammes and Morrell [64HAM/MOR] are used, the
calculated concentrations of Ni2+ and NiP2 O72 − are within 15% of those calculated by
Perlmutter-Hayman and Secco (by neglecting the association of K+). This provides
strong support for the value of K(A.50) of Hammes and Morrell [64HAM/MOR]. The
species NiHP2 O7− is calculated to be less important (generally < 20% of the total nickel
in solution), and its calculated concentration is approximately 50% higher when
association is neglected. This would suggest that the formation constant for NiHP2 O7−
reported by Perlmutter-Hayman and Secco is greater than would be the case if
association of pyrophosphate with K+ were considered explicitly. The raw data are
unavailable for re-analysis.
From the temperature dependencies of the equilibrium constants, the enthalpies
of reaction in 0.1 M KNO3(aq) can be calculated as, ∆ r H m (A.50) = (47.93 ± 10.16)
kJ·mol–1, and ∆ r H m (A.51) = (30.56 ± 7.42) kJ·mol–1.
Figure A-24: Variation of formation constants from Perlmutter-Hayman and Secco
[73PER/SEC] with ionic strength. For the first and second protonation constants of pyrophosphate, n = – 8 and – 6; for Reactions (A.50) and (A.51), n = – 16 and – 12.
12
4-
11
10
protonation constant for P2O7
3-
protonation constant for HP2O7
-
formation constant for HNiP2O7
2-
formation constant for NiP2O7
log10K - nD
9
8
7
6
5
0.00
0.05
0.10
0.15
−3
I / mol·dm
0.20
364
A. Discussion of selected references
[73POW]
This paper presents a reassessment of the data of Izatt et al. [69IZA/EAT]. The author
used the K1 value for the nickel sulphate association constant that was determined by
Nair and Nancollas [59NAI/NAN], and reinterpreted the results from the later titration
calorimetry experiments. The variation of the enthalpy of ion-pair formation with ionic
strength was also reported. The recalculated value agrees well with those from some
conductance studies, but is still low with respect to the value from [59NAI/NAN].
[73RAU/GUE]
A mixture of metallic nickel and nickel oxide (particle size around 1 µm) was treated
with mixtures of water vapour and hydrogen gas at temperatures ranging from 711 to
860 K. The ratio of equilibrium partial pressures pH O / pH was determined by a tensimetric method using a palladium-membrane apparatus. The authors applied a third
law analysis to their data and arrived at ∆ f H mο = – (239.5 ± 1.0) kJ·mol–1 for the standard enthalpy of formation of NiO at 298.15 K.
2
2
[73VEN/GEI]
The activity of sulphur in dilute solutions of sulphur in pure nickel melts was investigated at three different temperatures, viz. 1773, 1823 and 1848 K, for concentrations up
to 0.7 wt% S. A gas mixture of H2(g) and H2S(g) was equilibrated with liquid nickel at a
constant gas stream through the furnace. The compositions of both the gas mixture and
the equilibrated liquid phase were determined by chemical analysis. When S2(g) is the
reference state for sulphur, the activities of sulphur in the nickel melt can be calculated
from the well-known gas phase equilibrium:
H2S(g) U H2(g) + ½S2(g).
Since this contribution is confined to highly diluted nickel melts only, the experimental results are not re-evaluated in the present assessment.
[73WAA/CAL]
Liebenbergite, a natural nickel mineral is described mineralogically in this work.
[74ARU]
The heat of association, in aqueous solution, of fluoride ion with metal ions of the
Mn(II)-Zn(II) series has been measured by a direct calorimetric method at I = 0.5 M and
298.15 K. The measurements were performed in a self medium (to 83 cm3 of a 0.1665
M Ni(NO3)2 solution were added three amounts of 5 cm3 0.5 M NaF solution). The molar enthalpies of the reaction were calculated using the stability constants reported in
[72BON/HEF]. For the purpose of the present review, the reported experimental data
for the reaction,
Ni2+ + F– U NiF+
(A.52)
A. Discussion of selected references
365
have been recalculated using the recommended values for log10 b1ο ((A.52), 298.15 K)
and ∆ε (A.52). At I = 0.5 m log10 b1 ((A.52), 298.15 K) = (0.76 ± 0.08) can be calculated. Taking into account the different ionic media used by Aruga, a greater uncertainty
in log10 b1 ( ± 0.25) was assigned. The recalculated formation enthalpy of the NiF+ species is (4.4 ± 2.0) kJ·mol–1.
[74BIX/LAR]
Chloride selective electrode measurements were used to determine the stability constants of nickel(II) chloride complexes at 25°C in 1.00 M NaClO4 and 0.10 M HClO4.
The total chloride and metal ion concentrations ranged from 0.0003 – 0.01 M and 0.02 –
0.04 M, respectively. The authors noted that the response of the chloride selective electrode was not linear below a chloride concentration of 0.2 mM. The effect of complex
formation on the measured potential was rather modest (0.4 – 1.1 mV), therefore we
assigned a higher uncertainty to the formation constants than was reported by the authors.
[74BLO/RAZ]
The enthalpy of formation of M(H2O)6X(2–y)+ complexes (where M = Ni(II), Co(II), or
Cu(II) and X–y = Cl–, Br– or SO 24 − ) was determined, using the stability constants reported in [70MIR/MAK] and experimental results on heat of mixing of 0.05 M
M(ClO4)2 solution (I = 3.0 M maintained with LiClO4) with the solution of the corresponding lithium salt (I = 3 M). The authors did not reporte sufficient experimental details to allow the reliability of their data to be assessed.
[74DIC/HOF]
The kinetics of formation and dissociation of Ni(SCN)2(aq) have been studied in water
and in several organic solvents, using the pressure-jump and shock wave relaxation
technique at 20°C. The concentration of Ni(SCN)2(aq) ranged between 0.001 and 0.1
M. In water, only the formation of the monothiocyanato complex was observed. No
background electrolyte was used, and the activity coefficients were calculated by an
extended Debye-Hückel expression. Although, this activity model is not compatible
with the SIT, the ionic strengths were low. Therefore, the reported result was corrected
to 25°C and the resulting value was accepted with an increased uncertainty ( log10 b1ο =
(1.79 ± 0.10)).
[74GRA/WIL]
The ternary and quaternary complexes of nickel(II) and palladium(II) with asparaginate
(asn), chloride and hydroxide ions were studied using glass and silver-silver chloride
ion selective electrodes (I = 3 M NaClO4, T = 298.15 K). The formation of
Ni(asn) x (OH)(2y − x − y ) complexes was not reported, only the ternary (quaternary) complexes with chloride ions, Ni(asn) x Cl(OH)(1y − x − y ) , in addition to the binary NiCl+ complex, were found to exist. Taking into account the low affinity of chloride for nickel(II)
366
A. Discussion of selected references
this is unlikely. Therefore, the reported log10 K (NiCl+ ) = (0.687 ± 0.040), which is
much higher than other reported values, was not taken into account in this review.
[74GRI/FER]
The structure of a millerite crystal from Marbridge Mine, Malartic, Quebec, with the
empirical formula Ni0.981Fe0.016Co0.004S has been refined. The hexagonal axes were
found to be a0 = 9.607(1) Å and c0 = 3.143(1) Å. Within the lattice each Ni is coordinated by five S atoms and two Ni atoms. The observed Ni-S bond lengths are comparable to the expected value for a covalent bond. Molecular orbital theory was invoked to
show that the millerite structure with five-fold coordination around Ni is more stable
than the nickeline structure (α-NiS) with six-fold coordination about each Ni. Thereby it
is rationalised that the low temperature phase β-NiS occurs in nature and the high temperature phase α-NiS does not.
[74HEF]
This is a review paper on fluoride complexes of some 50 metal ions. The author found
that the stability constants of these complexes correlate very well with simple electrostatic parameters such as ion size and charge. Only a few metal ions, all found in the
lower right hand side of the Periodic Table, seem to be exceptions from this general
rule. The enthalpy of formation is positive even for the strongest complexes. Only Be2+
and Ag+ appear to have negative ∆ r H m values. Consequently, the formation of the great
majority of fluoro complexes is entropy controlled.
[74KAB]
The protonation of tetracyanonickelate(II) ion was investigated by a pH-metric method
at temperatures from 25 – 50°C and at ionic strengths of 0.01, 0.02, 0.05, 0.1, 0.2 and
0.5 M, adjusted using NaCl. The author reported two successive protonations of the
Ni(CN) 24 − ion ( log10 K1 and log10 K 2 , respectively). Titrations of aqueous Na2Ni(CN)4
solutions by 0.1 M HCl were performed while nitrogen gas was bubbled through the
solution; therefore HCN volatilisation could have been occurring. Several authors have
reported slow equilibration in acidic solutions of Na2Ni(CN)4, sometimes lasting several
days [63CHR/IZA], [68KOL/MAR], [74PER]. During the titration performed in
[74KAB], equilibrium was probably not attained. The dependence of the determined
protonation constants on the temperature and ionic strength shows notable deviation
from what would be expected (see e.g., Figure A-25 and Figure A-26). Therefore, the
reported data were not considered further in this review.
A. Discussion of selected references
367
Figure A-25: Temperature dependence of log10 K 2 for the reaction Ni(CN) 24 − + H+ U
Ni(HCN)(CN)3− at different ionic strengths using NaCl.
5
4.8
log10 K 2
I = 0.01 M
4.6
I = 0.02 M
I = 0.05 M
I = 0.1 M
4.4
I = 0.2 M
I = 0.5 M
4.2
4
0.00305
0.00315
0.00325
0.00335
K/T
Figure A-26: Extrapolation to I = 0 of experimental equilibrium constants for the reaction Ni(HCN)(CN)3− + H+ U Ni(HCN) 2 (CN) 2 (aq) ( log10 K 2 ) at different temperatures.
5
log10 K 2 + 2D
4.9
T = 25 °C
4.8
T = 30 °C
T = 35 °C
T = 40 °C
4.7
T = 45 °C
T = 50 °C
4.6
4.5
0
0.1
0.2
0.3
.
I / mol kg
0.4
-1
0.5
0.6
368
A. Discussion of selected references
[74KUL3]
The stability constants and the enthalpy changes for the formation of nickel(II) (and
copper(II)) thiocyanate complexes were determined by calorimetric titrations at 25°C in
1 M (NaClO4) aqueous solution. The concentrations of nickel(II) and thiocyanate
ranged from ~ 0.01 – 0.1 M and 0 – 0.8 M, respectively. The formation of NiSCN+,
Ni(SCN)2(aq) and Ni(SCN)3− was assumed according to [53FRO2]. Both graphical and
computer evaluations of the experimental data were performed. The formation constants
obtained by the computer programme "Kalori" are as follows: log10 b1 = (1.12 ± 0.04)
(3σ), log10 b 2 = (1.57 ± 0.14), log10 b 3 = (1.28 ± 1.00) (two years later the author reported [76KAR/KUL] somewhat different constants: log10 b1 = (1.12 ± 0.02) (3σ),
log10 b 2 = (1.57 ± 0.04), log10 b 3 = (1.25 ± 0.20)). The reported "best" values are
slightly different from both above mentioned sets, since Fronaeus' data [53FRO2] were
also taken into account: log10 b1 = (1.14 ± 0.02) (3σ), log10 b 2 = (1.58 ± 0.05),
log10 b 3 = (1.6 ± 0.2). The latter constants have been used to calculate the corresponding stepwise enthalpy values: ∆ r H m = – (12.02 ± 0.15), – (8.9 ± 1.0) and − (8.2 ± 4.0)
kJ·mol–1 for the three successively formed complexes, respectively. The enthalpy of
reaction values corresponding to the three cumulative formation constants are ∆ r H m =
− (12.0 ± 0.5), – (20.9 ± 1.2) and – (29.1 ± 5.2) kJ·mol–1, respectively, where the uncertainties have been estimated.
[74MA]
In this work the olivine – spinel transformation in Ni2SiO4 at high pressures was investigated over a p – t range of 20 – 40 kbar and 650 – 1200°C, using a piston-cylinder
apparatus. The transformation curve is essentially a straight line and can be expressed
by an equation: p(bar) = 23300 + 11.8·t (°C). The melting behaviour of Ni2SiO4 olivine
was investigated in a p – t range of 5 – 13 kbar and 1600 – 1700°C. The results show
that Ni2SiO4, olivine melts incongruently at high pressures to NiO plus melt. The reported dissociation of Ni2SiO4, olivine, to NiO plus cristobalite prior to melting
[63PHI/HUT] was not observed. The melting curve has a p – t slope of approximately
105 bar·K–1. It can be extrapolated to atmospheric pressure to give a melting temperature of approximately 1575°C. This value is 75°C lower than that given by
[63PHI/HUT].
[74MAK/MAS]
Values for the association of nickel with sulphate for temperatures from 25 to 95°C and
ionic strengths to 2.6 M were measured using a variety of pH titration, conductivity and
transport number measurement techniques. Although the reported association constants
are in rough agreement with those from other studies, there are insufficient experimental
details (or results) to use this work in deriving a recommended value for K1.
A. Discussion of selected references
369
[74NET/BAT]
Complex formation of Ni(II), Pr(III) and Ho(III) with thiocyanate was studied in aqueous solutions (I = 3 M (LiClO4)) by IR spectroscopy. The concentration of thiocyanate
was 0.36 M, the concentration of metal ions ranged from 0.36 to 0.86 M. The formation
constants of the mono-complexes (MSCN+) were calculated by three different methods,
which resulted in considerably different values. The temperature of the measurements is
not given in the paper. Therefore, the reported data were not considered further in this
review.
[74PER]
The formation of complexes between nickel(II) and cyanide ion was studied by spectrophotometric and potentiometric measurements with a glass electrode at 25°C, at an ionic
strength of 3 M using aqueous sodium perchlorate as the ionic medium. The
potentiometric data are best described by the formation of two complexes NiCN+ and
Ni(CN) 24 − ( log10 b1 = (7.03 ± 0.2) and log10 b 4 = (31.06 ± 0.03)). On the other hand,
the spectrophotometric measurements gave no positive evidence for the formation of
NiCN+; therefore the log10 b1 value given above should be regarded as a maximum
value. If only formation of Ni(CN) 24 − was postulated, log10 b 4 = (31.12 ± 0.08) was
obtained, with a somewhat higher uncertainty in the fitted parameter. This value is accepted in the present review, but the uncertainty has been estimated here as ± 0.15. The
potentiometric investigation was performed at 4 < pH < 5.5, but protonated tetracyano
complexes, as reported in [68KOL/MAR], were not detected.
[74RAJ/PRE]
The structure of a single crystal of millerite from Quebec, Canada, with hexagonal axes
a0 = 9.6190(5) Å, c0 = 3.1499(3) Å and composition Ni1.03Fe0.016Co0.004S has been refined. The Ni and S atoms are in five-fold coordination (tetragonal pyramidal) with each
other. The average Ni-S distance, 2.310 Å, is consistent with a divalent Ni in five-fold
coordination. Rajmani and Prewitt suggested that the metal-metal bonding and the formation of the trinuclear cluster stabilise the millerite structure.
[74RUT/HAU]
This paper describes vapour pressure measurements over Ni in the range 1277 – 1658 K
as obtained using the Langmuir technique. The results yield the enthalpy of sublimation
(447.7 ± 24.3) kJ·mol–1 and (431.8 ± 16.7) kJ·mol–1 according to 2nd and 3rd law analysis, respectively. In addition Rutner and Haury [74RUT/HAU] determined the “best
values” of the parameters in the vapour – pressure equations of solid and liquid nickel,
and the heat of vaporisation and sublimation of nickel by a statistical treatment of data
available in the literature and their own results. The results of Rutner and Haury
[74RUT/HAU] were not used by [98CHA] for the determination of the enthalpy of formation for Ni(g) because of an inconsistency with the melting data.
A. Discussion of selected references
370
[74SLO/JON]
In their Figure 2 Slough and Jones plotted z·r–1 versus ∆ r H mο of reactions,
MO(cr) + 2B2O3(cr) U MB4O7(s)
(A.53)
M2O(cr) + 2B2O3(cr) U M2B4O7(s)
(A.54)
and
ο
m
where r are presumably the Pauling radii [40PAU] and ∆ r H were determined experimentally by reactions of the crystalline oxides, see Figure A-27. The value for NiB4O7,
given by [74SLO/JON], can only be estimated when the Goldschmidt radius of Ni2+ (r =
0.078 nm) [26GOL] and not the Pauling radius (r = 0.070 nm) is employed. The effective ionic radii of Shannon (for coordination number 6 [76SHA]) lead to an inferior
correlation and a completely different ∆ r H mο for the respective nickel reaction, even on
the plot based on the Pauling radii. This example demonstrates that methods of thermodynamic data estimation also must be reviewed critically.
Figure A-27: ∆ r H mο (298 K) of reaction of B2O3(cr) with MO or M2O.
2 B 2O 3 (cr) + MO (cr) U MB 4 O 7 (s)
2 B 2O 3 (cr) + M 2 O (cr) U M 2B 4O 7 (s)
30
25
−1
z·r
2+
20
Li
/ nm
−1
Ca
+
2+
Ni [26GOL]
2+
Ni [76SHA]
15
Na
10
K
Pauling ionic radii [40PAU]
Effective ionic radii [76SHA]
5
0
-100
-200
+
-300
+
-400
–1 −1
∆mrH°
/ kJ·mol
∆rH°
/ kJ·mol
m
[74WOR/COW]
Heat capacities from 10 K to 300 K were determined by adiabatic calorimetry, and by
differential scanning calorimetry (DSC) from 300 K to 550 K. In the paper, only
rounded values are supplied for the adiabatic calorimetry results for NiI2(cr) except for
values near the (structural phase [81KUI/SAN]) transition at approximately 59 K. The
authors also supply an equation (A.55)
A. Discussion of selected references
–1
–1
[C pο,m ]550K
= 65.5 + 0.0515 T/K – 0.0000397 (T/K)2
200K (NiI2, cr) / J·K ·mol
371
(A.55)
that provides a fit to the combined heat capacity results from adiabatic calorimetry and
differential scanning calorimetry between 200 K and 550 K, but do not separately provide measured values above 300 K. This equation is not in a form normally used for the
temperature dependence of heat capacities of solids near room temperature
[93KUB/ALC].
The adiabatic calorimetry heat capacity data were available as supplementary
material from the British Library Lending Division (SUP 21075), and were used in the
preparation of the current review. The authors integrated the low-temperature heat capacity measurements and reported 138.7 J·K–1·mol–1 for S mο (NiI2, cr). No contribution
was added for the magnetic phase transition at 75 K to 76 K reported by Billerey et al.
[77BIL/TER] and Kuindersma et al. [81KUI/SAN], and this may have led to an underestimation of between 0.3 J·K–1·mol–1 and 0.5 J·K–1·mol–1 in the value of S mο above
80 K. Indeed, the two most relevant values from the authors “Run I” (for 77.69 K and
79.36 K) show no evidence whatsoever of an anomaly. However, the authors’ “Run I”
values also do not provide a well-defined peak for the anomaly near 59 K. The “Run II”
experiments covered only the temperature range of the 59 K anomaly, and appear to
have been done with much more care and with mush smaller temperature increments.
The values of entropies of reaction as derived from adiabatic calorimetry and from
measured dissociation pressures and heats of solution of NiI2, Ni(NH3)2I2 and
Ni(NH3)6I2 were found to be in good agreement. This indicates that it is likely that assuming a smooth extrapolation of C p ,m between 10 K and 0 K does not introduce any
substantial error in the calculated S mο value. The adiabatic calorimetry results between
200 K and 300 K showed a random scatter of approximately ± 0.4 J·K–1·mol–1, unusually large for this type of measurement.
The experimental scatter for the differential scanning calorimetry values was
reported to be within 4%, but it is not reported how well the authors’ equation corresponded to the measured values, or how well measurements obtained from the two
techniques agree near 300 K. Unfortunately, the DSC data were not part of the supplementary material.
To obtain an equation for C p ,m (T) in a more standard form, values of the heat
capacity were calculated at 25 K intervals between 300 K and 550 K using the authors’
equation (A.55). An equation of the form (A + B(T/K) + C(T/K)–2) was fitted to these
values and to the experimental low-temperature adiabatic calorimetry values from
200 K to 300 K. Values from the adiabatic calorimetry and from the equation based on
the DSC measurements were weighted in a ratio of 0.4:2.0. Because the number of DSC
measurements used in the original fitting by Worswick et al. is unknown, these weightings are arbitrary by definition. The obviously erroneous datum at 229.92 K was not
used. The resulting equation (A.56),
–1
–1
[C pο,m ]550K
= 73.5775 + 0.016425T – 1.057753×105 (T/K)–2 (A.56)
200K (NiI2, cr) / J·K ·mol
372
A. Discussion of selected references
fit the low temperature data as well as the authors’ equation (and well within the experimental scatter). As would be expected, agreement between values from the two
equations was poorer for 300 K to 550 K, but the differences were always less than 0.5
J·K–1·mol–1 —again much less than the experimental uncertainties.
The values from equations (A.55) and (A.56) for 298.15 K agree well with the
reported smoothed value from the adiabatic calorimetry results (78.4 J·K–1·mol–1) and
with each other (within 0.05 J·K–1·mol–1). Because the actual DSC results are unavailable, and there is considerable scatter in the adiabatic calorimetry results above 200 K,
an uncertainty of ± 1.0 J·K–1·mol–1 is assigned to the heat capacity value for nickel iodide at 298.15 K.
[74YAG/MAR]
Yagi et al. have investigated structures of the spinel polymorphs of Fe2SiO4 and
Ni2SiO4 in detail using single crystals synthesised at high pressures and temperatures.
Ni2SiO4 spinel with a = 8.044(1) Å had a normal spinel structure.
[75ARN/MAL]
The sulphur-rich part of the system Ni – S above 1253 K was re-investigated by metallographic methods. Vaesite (NiS2) was prepared from NiS and sulphur by solid state
reaction. Mixtures of NiS2 and S were heated in evacuated silica glass tubes. After
quenching, polished samples were examined microscopically. In contrast to Kullerud
and Yund [62KUL/YUN], who reported a congruent melting point of vaesite at 1280 K
and a monotectic reaction at 1264 K, the heating experiments indicated that NiS2 melts
syntectically, forming two immiscible liquids at (1295 ± 3) K, i.e., NiS2 U L1 + L2. The
liquid L2 was estimated to contain 0.5 wt% Ni, while the composition of the liquid L1
amounted to (52.0 ± 1.8) wt% Ni.
[75ARU]
Enthalpies of association of nitrate and chlorate with Mn(II), Co(II), Ni(II), Cu(II) and
Zn(II) have been determined by direct calorimetry in an aqueous medium at 298 K and
an ionic strength I = 1 M. In a typical experiment, 2.5 and 5.0 cm3 of 1 M NaNO3 solution were added to 93 cm3 of 0.34 M Ni(ClO4)2 solution, and this was repeated twice
under identical conditions. The association between Na+ and nitrate ion was considered
during the calculations (Kass = 0.25 M–1). The molar enthalpies of the reaction:
Ni2+ + NO3− U NiNO3+
(A.57)
were calculated using the stability constants reported in [73HUT/HIG]. For the purpose
of the present review, the reported experimental data have been recalculated using the
tentative values for log10 b1ο ((A.57), 298.15 K) and ∆ε((A.57), (NiClO4)2) =
− (0.11 ± 0.15) kg·mol–1. At I = 1.05 m log10 b1 ((A.57), 298.15 K) = – (0.21 ± 1.00)
can be calculated. The recalculated formation enthalpy of the NiNO3+ species is (6 ± 4)
A. Discussion of selected references
373
kJ·mol–1. The relatively high uncertainty is assigned due to the large error in
log10 b1ο ((A.57), 298.15 K).
[75BAR/MAS]
Barvinok et al. measured the enthalpy of solution of NiCl2(cr) in 2.0 M HCl at
298.15 K. A value for the molar enthalpy of solution, ∆ sol H m = – (74.43 ± 0.17)
kJ·mol−1, was reported. As other experimental details are lacking, this single value was
disregarded in this review.
[75CLA/KEP]
The kinetics of hydrolysis of nickel(II) were studied by mixing nickel(II) sulphate or
nickel(II) perchlorate solutions with NaOH solutions in a stopped-flow apparatus at 25
and 41.7°C. The ionic strength was maintained at 0.8 M by additions of sodium sulphate
or perchlorate. The hydrolysis occurred in two stages, as already mentioned in
[71HOH/GEI]: (i) the first stage (probably the formation of NiOH+) was complete
within the mixing time (2 ms), (ii) the second stage resulted in the formation of
Ni 4 (OH) 44 + within a few seconds (after a longer time interval, slight precipitation of
nickel hydroxide was also noted). In perchlorate medium some runs were made to follow the changes of pH, after addition of base, using Neutral Red as an indicator (the
value used for the pK of the indicator was not reported) at [Ni2+]tot = 0.203 M. The pH
was measured just after mixing and at equilibrium. From the pH values measured at “t =
0 s”, log10 *b1,1 = – (8.93 ± 0.10) was obtained for the reaction :
Ni2+ + H2O U NiOH+ + H+.
(A.58)
Since this value is considerably lower than other literature values, the authors
concluded that some oligomerisation occurred even during mixing. From the pH values
at equilibrium, the authors calculated the formation constants of Ni 4 (OH) 44 + assuming it
as the only hydrolysis species that was formed. Taking into account the relatively low
pH values at equilibrium (pH = 6.50 – 6.75), the formation of NiOH+ should also be
considered. Since only 6 experimental data points were reported, the value of log10 *b1,1
was fixed at – 9.87 (calculated from the selected log10 *b1,1 ((A.58), 298.15 K) for Im =
0.82), and only the log10 *b 4,4 value was refined during our re-assessment. With this
constraint, log10 *b 4,4 = − (26.98 ± 0.20) was obtained, a value close to the reported
value. Taking into account the limited number of experimental data, log10 *b 4,4 =
− (27.0 ± 0.4) is accepted for use in this review.
[75KAK/GIE]
The heat of solution of nickel chloride hexahydrate into 4.36 M HCl(aq) was measured
((21.32 ± 0.13) kJ·mol–1). Few experimental details are provided, and a double correction (to 0.0 M HCl(aq) and to I = 0) with large uncertainties would be required to use
the reported value to calculate the enthalpy of solution of the salt at infinite dilution in
water. This heat of solution value is not used further in the present review.
374
A. Discussion of selected references
[75KOS]
The molar heat capacities of NiCl2 from 2 to 30 K have been investigated. Smoothed
values in this region were listed. The results were compared with the predictions of the
spin-wave theory.
[75LEV/GOL]
The authors investigated the equilibrium potentials of the following galvanic cell with a
solid electrolyte: Pt | (Ni + SiO2) Ni2SiO4 | O2– | Ni, NiO | Pt. Based on the Gibbs energy
of the cell reaction determined in the temperature range of 1228 to 1355 K, the standard
Gibbs energy of formation and the standard molar enthalpy of formation of nickel orthosilicate at 298.15 K were calculated. We used the raw data of this work to represent
the Gibbs energy of the reaction:
2Ni + SiO2 + O2 U Ni2SiO4
as a function of temperature.
[75LIB/TIA]
Emf and spectrophotometric methods have been used to determine the stability constants of MCl+ complexes in M(ClO4)2 medium (M = Mn2+, Co2+, Ni2+ and Zn2+) at different concentrations.
The following cell was used for the potentiometic determination of formation
constants: Ag | AgCl,HCl(m1),Mg(ClO4)2(m) || M(ClO4)2(m),HCl(m1), AgCl | Ag, where
m1 was fixed at approximately 0.01 m, while the molalities of the metal perchlorate (m)
were 1.0, 1.5, 2, 2.5 and 3 m. The emf of the cells varied between 1.0 and 26.4 mV. The
calculation was based on the following plausible assumptions: (i) the activity coefficient
of the chloride is identical in the equimolar solution of Mg(ClO4)2 and M(ClO4)2, (ii)
the liquid-junction potential of the above cell is negligible, (iii) Mg(II) does not form
complexes or ion pairs with chloride. In the spectrophotometric study, the strong absorption of the CuCl+ complex at 272 nm was used to assess the formation constants of
the MCl+ complexes. At this wavelength the molar absorptivities of both the M(H2O)2+
and the MCl+ complexes were assumed to be negligible. The free chloride concentration
and, thus, the formation constants of the MCl+ complexes were calculated using the
formation constant and the ελ=272nm value of the CuCl+ complex determined earlier
[73LIB]. The two methods resulted in nearly coincident values.
The main advantage of the above experimental procedure is that in Ni(ClO4)2
media the medium effect can be eliminated. In the case of nickel, the results are interpreted in terms of the existence of [Ni(H2O)5]Cl+ and [NiCl(H2O)5]+ complexes. The
overall stability constant of the NiCl+ species (β1 = 0.37 at I = 0) is due mainly to the
stability of the outer-sphere ion pair. The β1 value for the inner-sphere complex was
reported to be around 0.01 M–1.
A. Discussion of selected references
375
[75MEY/WAR]
The equilibrium partial pressure of sulphur over Ni – S melts was determined by bubbling H2/H2S gas mixtures through the melts at temperatures between 1373 and 1873 K.
The compositions of the melt varied from nickel saturation to 27 wt% sulphur. The
composition of the gas phase was monitored with the help of an optical interferometer,
and the composition of the melt was determined by chemical analysis. The sulphur partial pressure for a given gas composition was calculated from the well-known equilibrium constants for the reaction:
H2S(g) U H2(g) + ½S2(g).
In addition, the liquidus line of the binary phase diagram Ni – S was obtained
in the region from 0 to 20 wt% sulphur.
[75MOS/FIT]
The data concerning NiO are essentially the same as in [75MOS/FIT2].
[75MOS/FIT2]
The least-squares line representing the temperature dependence of ∆ f Gmο (NiO) was
determined to be ∆ f Gm (T) = (– 244.019 + 0.092 (T/K)) kJ·mol–1. However, no detailed
information concerning uncertainties of the measured values was available.
[75RAU2]
The partial pressure of sulphur in equilibrium with the non-stoichiometric hightemperature modification of Ni1–xS was measured as a function of composition in the
temperature range from 781 to 1250 K. The sulphur pressure was determined directly
by a tensimetric method. In addition, the solid sample was reduced with hydrogen gas.
The amount of H2S(g) formed during the reduction as well as the ratio between H2S(g)
and H2(g) in the gas phase were analysed by tensimetric measurements. From these data
both the composition of the solid phase and the temperature dependence of the sulphur
activity in Ni1–xS, with S2(g) being the reference state for sulphur, could be calculated
from the well-known equilibrium:
H2S(g) U H2(g) + ½S2(g).
Moreover, phase boundaries of the homogeneity range of Ni1–xS are likewise
obtained from this equilibrium study.
[75TAM]
This is a reanalysis of some of the data of Katayama [73KAT] using different activity
coefficient assumptions.
A. Discussion of selected references
376
[75WEE/KOE]
Weenk et al. determined the enthalpies of solution of a series of alkalitrichloronickelates ANiCl3 (A = Cs, Rb, K, Tl, NH4) and their constituting chlorides (NiCl2 and ACl)
in a medium of 0.1 M HCl at 298.15 K with n(H2O) / n(NiCl2) = 1200. A solution calorimeter of the isoperibol type was used. The molar enthalpy of solution of NiCl2 was
found to be ∆ sol H mο = – (76.82 ± 0.42) kJ·mol–1 for the reaction:
NiCl2 U Ni2+ + 2 Cl–.
This value refers to an ionic strength I ≈ 0.24 mol·kg–1, which is too high for
the usual extrapolation to I = 0 to be applicable. Consequently this single value was
disregarded in this review.
[76BAE/MES]
This is a well-known standard work concerning metal ion hydrolysis in water. Selected
experimental values for 298.15 K were converted to zero ionic strength using the equations from Pitzer and Brewer [61LEW/RAN]. For the water soluble nickel(II) hydroxo
complexes, Ni x OH 2y x − y the following log10 *b yο, x values are reported: log10 *b1,1ο =
ο
ο
ο
− (9.86 ± 0.03), log10 *b 2,1
= – (19 ± 1) log10 *b 3,1
= – (30.0 ± 0.5), log10 *b 4,1
< – 44,
* ο
* ο
log10 b1,2 = – (10.7 ± 0.5) and log10 b 4,4 = – (27.74 ± 0.02). No thermodynamic data
have been taken directly from this reference.
[76BER2]
A high quality data set was obtained by accurate measurements of the Gibbs energy of
formation of NiO by using a solid state emf method with a gaseous hydrogen/water
buffer atmosphere. The result of the weighed linear regression for the temperature range
from 825 to 1675 K was given by the equation:
∆ f Gmο (NiO) = – (235.071 ± 0.141) + (0.08620 ± 0.00010)·(T/K) kJ·mol–1.
[76BIA/CAR]
A method using variation of solvent composition to study the formation equilibria of
complexes and its application to the nickel(II)-chloride system is reported. In pure acetone the formation of the complex NiCl24 − was detected, and this species readily decomposed with increasing water content. The extrapolation of experimental results, determined between 0 and 0.05 mole fraction of water, to pure water, resulted in a value of
log10 b (NiCl42 − ) ~ – 14 as a rough approximation.
[76GEE/SHE]
The emf measured for solid electrolyte cells of the type:
Pt | M’, M’Cl2 | electrolyte | M, MCl2 | Pt
provided Gibbs energies of formation for CrCl2, MnCl2, CoCl2 and NiCl2 at
A. Discussion of selected references
377
470 ≤ T / K ≤ 910. Depending on the temperature range either PbCl2 or BaCl2 were
used as the solid electrolyte. The Fe, FeCl2 electrode served as reference electrode.
From results two-parameter Gibbs energy of formation equations have been derived,
which obviously contain the uncertainties of the thermodynamic parameters ascribed to
the substances of the reference half-cell. The Gibbs energy equation related to NiCl2 has
been plotted in Figure V-18.
[76KAR/KUL]
This is a description of the "Kalori" computer programme (see the discussion of
[74KUL3]). As one of the applications, the nickel(II) – thiocyanate system was mentioned. Using the identical dataset, the authors reported stability constants and enthalpy
values that differ slightly from those reported for the same complexes in the original
paper [74KUL3].
[76KUL/BLO]
See comments under [81KUL/BLO].
[76MUR/KUR]
A solvent extraction method was used to determine, the stability of NiCl+ and NiSCN+
complexes, among others. The aqueous phase was 1 M NaClO4 solution, while the organic phase was CCl4 containing 0.05 M thenoyltrifluoroacetone and 0.03 M trioctylphosphine. The concentration of nickel(II) was ~ 1.0 × 10–4 M, while that of the complex forming anion varied up to 0.4 M for thiocyanate and ca 0.85 M in case of chloride. As the amount of metal extracted was very low in the case of chloride, even at high
chloride concentrations, only the formation constant reported for the thiocyanate complex is considered in this review.
[76NAV/KAS]
The enthalpy of formation from the oxides of Mg2SnO4 and Co2SnO4 were found by
oxide solution calorimetry. Using thermochemical data for the formation of olivine, for
olivine – spinel transitions and for the transformation of quartz to stishovite, pressures
for the disproportionation of silicate spinels were calculated to be in the range 150 – 200
kbar.
[76PAN/PAT]
Pant and Pathak measured the solubilities of Ni3(PO4)2·7H2O and NiHPO4 in water from
30 to 60°C, and reported the metal concentration in solution (0.012 M to 0.027 M), the
metal to phosphate ratio in solution (near 2.0 in all cases) and the “pH” (between 7.1
and 7.7, without allowance for the change in Kw with temperature). In the absence of a
good set of values for the complexation constants for Ni2+ with phosphate ligands as a
function of temperature, and good analyses of the final solids, no useful thermodynamic
quantities can be extracted from the information in this paper.
378
A. Discussion of selected references
[76PER]
The kinetics of formation of Ni(CN) 24 − have been studied spectrophotometrically at
367.5 nm in the pH-range 4.7 – 7.0, at I = 3 M NaClO4(aq) and t = 25°C. The author
found that, at the equilibrium, only Ni2+ and Ni(CN) 24 − exist in substantial concentrations, although the complexes NiCN+, Ni(CN)2(aq) and Ni(CN)3− are kinetically significant. In contrast to the acidic dissociation of Ni(CN) 24 − [76PER/EKS], its association
from nickel(II) can be described, within the experimental errors, without the protonated
complexes. This is probably due to the higher pH range used in this study. From the
kinetic data presented in this paper, the author determined the formation constant of
Ni(CN) 24 − ( log10 b 4 = (31.08 ± 0.05), at I = 3 M NaClO4 and t = 25°C, using pK HCN =
9.484 [71PER]). This value of log10 b 4 is retained in the present review, but with the
uncertainty increased to ± 0.15.
[76PER/EKS]
The acidic dissociation of Ni(CN) 24 − was studied spectrophotometrically in the pH
range 0 < pH < 3, at I = 3 M NaClO4 and t = 25°C, in order to elucidate the presence of
protonated H x Ni(CN) 4x − 2 species. Persson [74PER] reported Ni(CN) 24 − as the only
complex existing in the nickel(II) – cyanide system. On the other hand, Kolski and
Margerum, in [68KOL/MAR], reported extensive formation of protonated complexes.
Persson and Ekström [76PER/EKS] acknowledged the formation of protonated species
and their role in the acidic decomposition, but their data indicated that the protonated
species have steady state concentrations during the decomposition. Therefore, according
to the authors, the pK values of the protonated complexes should be much lower than
suggested in [68KOL/MAR].
[76RAU]
The activity of sulphur in the high temperature modification of Ni3S2 was determined as
a function of composition at temperatures between 815 and 1044 K. The solid sample
was reduced with hydrogen gas in a stepwise manner. The amount of H2S(g) formed
during the reduction as well as the ratio between H2S(g) and H2(g) in the gas phase were
obtained from tensimetric measurements. From these data both the composition of the
solid phase and the temperature dependence of the sulphur activity in Ni3S2, with S2(g)
being the reference state for sulphur, could be calculated from the well-known equilibrium H2S(g) U H2(g) + ½S2(g). Neither the initial compound nor the solid products
formed owing to reduction with hydrogen gas were checked by X-ray diffraction analysis.
[76SHI/TSU]
This was a conductance study done at pressures from 0.1 to 160 MPa and at 15, 25 and
40°C. The sets of equivalent conductivities are reported for a very limited range of concentrations from 0.0001 to 0.0003 M, and have the appearance of being smoothed val-
A. Discussion of selected references
379
ues from unreported primary data. The data analysis used the simple Onsager equation
to extrapolate the NiSO4(aq) dissociation constant to I = 0, and values for ∆ r H m and ∆V
are reported. The reported association constants are greater than those found in many
other studies. Re-analysis of the “1 atm.” results (assumed to be identical at 0.1 MPa)
using a version of the Lee-Wheaton equation that is consistent with the SIT equation
used in the present review (Appendix B) leads to results that are consistent with the
other studies. In the present review, we estimate the uncertainty in K1 at 25°C (0.1 MPa)
to be ± 30. The high-pressure results are consistent with values found in a later study of
Fisher and Fox [79FIS/FOX].
The average enthalpy of reaction, calculated from the equally weighted values
for the temperature range 15 to 40°C, is 6 kJ·mol–1, in excellent agreement with the
value obtained from the data of Katayama [73KAT]. An uncertainty of ± 2.5 kJ·mol–1 is
estimated in the present review.
[76SMI/MAR]
This is a well-known compilation of stability constants including a number of metal
ions and a variety of inorganic ligands. For the water soluble nickel(II) hydroxo complexes the following log10 b y , x for the reaction:
xNi2+ + yOH– U Ni x (OH) 2y x − y
ο
ο
are reported: log10 b1,1ο = 4.1, log10 b 2,1
= 8, log10 b 3,1
= 11, log10 b1,2 = 4.2 (I = 3 M) and
* ο
log10 b 4,4 = 28.3 (I = 0), (29.37 ± 0.03) (I = 3 M). No thermodynamic data have been
taken directly from this reference.
[76TAY/DIE]
This paper reports a pH study of complexation between HPO 24 − and Ni2+ at 25°C in
0.1 M NaClO4:
Ni2+ + HPO 24 − U NiHPO4(aq)
(A.59)
Nickel(II) concentrations were between 7 × 10–3 and 3.3 × 10–2 M, and the initial ligand concentration was approximately 2 × 10–2 M. The association constant between Ni2+ and HPO 24 − was reported as (130 ± 6) dm3·mol–1, and the association constant between Ni2+ and H 2 PO 4− to form NiH 2 PO 4+ as (3.5 ± 1) dm3·mol–1. There is insufficient information to recalculate the results. For the protonation constants for
HPO 24 − , the reported log10 K values are within ± 0.2 of values from other studies, and
in the present review a similar uncertainty is assigned to the reported NiHPO4(aq) formation constant ( log10 K = (2.11 ± 0.20)). Correction to zero ionic strength (Appendix B) gives log10 K ο (A.59) = (3.00 ± 0.20), where the uncertainty has been estimated
in the present review. The small ∆ε correction has been based on the value for
ε(Na+, HPO 24 − ) from Table B-5 (rather than using the equation in Table B-6), and
ε(Ni2+, ClO −4 ) = 0.37 kg·mol–1. The evidence for H 2 NiPO 4+ is not considered to be sufficient to assign a formation constant for this species.
380
A. Discussion of selected references
[77ASH/HAN]
UV spectroscopic measurements (I = 5 mol·dm–3, NaClO4/Na2SO4) were compared for
Na2SO4 concentrations of 0.1 mol·dm–3 and 1.0 mol·dm–3. Measurements for 47, 60 and
70°C were reported. The authors extrapolated their measurements to 25°C, assuming the
effects of temperature were small ( ∆ r H m was assumed to be 0 kJ·mol–1), and reported
log10 K1 (298.15 K, I = 5 M NaClO4) = 0.77. The assumption concerning ∆ r H m is reasonable in a general sense, but probably introduces an uncertainty of the order of ± 0.2
in the value of log10 K1 (298.15 K). The ionic strength (6.6 m) is too high to properly
apply the SIT. If the SIT is used with ε(Ni2+, ClO −4 ) equal to (0.37 ± 0.03), log10 K1ο =
1.3 is calculated.
[77BAR/KNA]
This volume contains the Planck function, the enthalpy and the Gibbs energy tabulated
at 100 K intervals for pure substances. It continued the work started by [73BAR/KNA].
[77BIL/TER]
In this paper, the heat capacity and magnetisation of NiI2 were reported from measurements done between 20 K and 150 K. The heat capacity measurements showed peaks at
59.5 K and 76 K (later ascribed by Kuindersma et al. [81KUI/SAN] to a structural
phase transition and a magnetic phase transition, respectively), and a shoulder and a
peak were found at the corresponding temperatures in the magnetic measurements. The
measurements are reported only in graphical form. In the present review, the values of
C p ,m (T) from the authors’ Figure 2(a) were digitized and used to calculate values of
C p ,m /T. Equations were fit to the values over short ranges of temperature and to a set of
values including points for temperatures above and below the anomalies. If the contributions from the two transitions are neglected, the contribution to the entropy between
50 K and 85 K is calculated to be 27.2 J·K–1·mol–1. The additional contributions to the
entropy from the transitions near 59.5 K and 76 K are estimated to be 0.40 J·K−1·mol–1
and 0.37 J·K–1·mol–1, respectively.
[77CHE/BRO]
The 14N nuclear quadripole resonance spectra were measured for several Pd(II) and
Ni(II) thiocyanate complexes. The thiocyanate is N-bonded in all nickel(II) compounds.
[77FLE]
Fleet investigated synthetic α-Ni3S2 and demonstrated unambiguously that there are four
similar Ni–Ni ‘bonds’ per Ni atom. The short Ni–Ni distances in heazlewoodite are only
slightly greater than those for metallic Ni thus indicating metallic bonding.
A. Discussion of selected references
381
[77KAS/SAK]
The chloride-formation reaction of nickel oxide and other metal oxides with HCl gas
were studied at temperatures from 25 to 500°C and at 1.0 atm. The experiments were
carried out by using gas analysis, a “thermobalance” and X-ray analysis to determine
compositions for the reaction:
NiO (cr) + 2 HCl (g) U NiCl2 (cr) + H2O (g).
A third-law analysis of the reported equilibrium constants at 298.15 and
773.15 K resulted in ∆ f H mο (NiCl2, cr, 298.15 K) = – (316.2 ± 6.4) kJ·mol–1. This value
has a comparatively large uncertainty, does not overlap with the values obtained by the
definitive studies of [53BUS/GIA] and [84LAV/TIM], and is therefore rejected.
[77KAT]
This paper (in Japanese) uses the same nickel sulphate conductiometric data reported
previously by the same author [73KAT], and compares the results with those for other
electrolytes.
[77NIC/ROB]
The otwayite was investigated with respect to its chemical composition, its optical and
physical properties and powder diffraction patterns. No synthesis of otwayite has been
attempted so far.
[77OGA]
The heat capacities of FeS2, CoS2, NiS2 and binary solid solutions of these sulphides
were measured over the temperature range from 15 to 350 K in order to investigate the
magnetic contribution to the specific heat at low temperatures. The solid samples were
not characterised by X-ray diffraction analysis.
The temperature dependence of the heat capacity data for NiS2 is depicted in
Figure A-28. Between 0 and 20.5 K, a Debye – T 3 function is fitted to the data, and at
higher temperatures a spline fit of polynomials of the general form (a + bT + cT2 + dT –2
+ eT −½) is applied, leading to the solid line in Figure A-28. The integration
298.15 K
ο
± 2.4) J·K–1·mol–1. The integration of the C p ,m
∫0 (C p ,m / T )dT resultsο in Sm = (79.6
–1
function of FeS2 yields S m = 54.4 J·K ·mol–1 which deviates from the value listed in the
JANAF tables [98CHA], S mο = 52.9 J·K–1·mol–1, by 3%. Thus, the same relative error is
assumed for the standard entropy of NiS2, leading to the uncertainty of ± 2.4
J·K−1·mol−1.
A. Discussion of selected references
382
Figure A-28: Heat capacity of NiS2 plotted versus temperature. ○ [77OGA], solid line:
polynomial fit to the experimental data.
80
70
1 ·mol
−1–1
C°
J·K−–1
,m // J·K
Cpp,m
·mol
60
50
40
30
20
10
0
0
50
100
150
200
250
300
350
T/K
[77OSW/ASP]
Oswald and Asper reviewed the preparation and properties of bivalent metal hydroxides.
[78BEC/LIU]
In this study, the effect of temperature (0 to 180°C) on formation of inner-sphere and
outer-sphere nickel sulphate complexes was investigated by 17O nmr spectroscopy. The
solutions used were 0.097 m NiSO4 in 1.07 m and 2.34 m Li2SO4(aq). Results indicated
that inner-sphere complexation increased with increasing temperature, and that at least
two different inner-sphere complexes were formed, involving displacement of one and
two inner-sphere water molecules, respectively. It was not clearly established whether
the second complex was formed by bidentate complexation of a single inner-sphere
sulphate ion, or by unidentate complexation of two sulphate ions. Near 25°C, the estimated value of the formation constant for the first inner-sphere complex was approximately an order of magnitude less than the value of the formation constant as measured
by methods [59NAI/NAN], [73KAT] that did not distinguish between inner- and outersphere complexation. The complex formed by displacement of two inner-sphere water
molecules became predominant at temperatures above 160°C.
A. Discussion of selected references
383
[78BUR/KAM]
A potentiometric study of the hydrolysis of nickel(II), similar to the studies of
[65BUR/LIL], [65BUR/LIL2], [66BUR/IVA] and [71BUR/ZIN], at 25°C, and in 3 M
(Na)Br medium is described. The metal ion concentration ranged from 0.2 to 1.4 M (I =
3.2 – 4.4 M). The best fit to the experimental data was obtained assuming the presence
of Ni 4 (OH) 44 + and Ni2OH3+ species, with formation constants log10 *b 4,4 =
− (28.18 ± 0.05) and log10 *b1,2 = – (9.5 ± 0.1). During these measurements, a large part
of the Na+ ion is replaced by Ni2+; consequently the constancy of activity coefficients at
different nickel(II) concentrations is questionable. Furthermore, the formation of bromo
complexes (both binary and ternary) was not considered.
Therefore, the reported experimental data were re-evaluated for the purposes of
this review. At I = 3.0 M, equilibrium constant for the reaction:
Ni2+ + Br– U NiBr+
(A.60)
log10 b1 (A.60) = – 0.49 can be calculated for the NiBr+ complex in NaBr medium, using the recommended value log10 b1ο (A.60) and ε(NiBr+, Br–) = ½{ε(Ni2+, Br–) +
ε(Na+, Br−)} = (0.16 ± 0.03) kg·mol–1. Considering the whole dataset and the same
model as proposed, log10 *b 4,4 = − (28.22 ± 0.04) (3σ) and log10 *b1,2 = – (9.34 ± 0.08)
can be calculated, in good agreement with the published values. Using only the data sets
reported for [Ni2+]tot = 0.2, 0.4 and 0.8 M (to reduce the influence of medium effects),
and taking into account the formation of NiBr+, log10 *b 4,4 = – (27.19 ± 0.06) (3σ) and
log10 *b1,2 = − (8.82 ± 0.10) can be derived. If the formation of NiOH+ is also considered, the fit to the data does not improve. Thus, the values for the other two constants
are selected in this review, with increased uncertainties: log10 *b 4,4 = − (27.19 ± 0.30)
and log10 *b1,2 = – (8.82 ± 0.40).
[78ENO/TSU]
Enoki and Tsujikawa measured the heat capacity of hydrothermally aged, bulky
Ni(OH)2 crystals from 4.2 to 35 K. Whereas the magnetic entropy was found to be
higher by 0.37 J·K–1·mol–1 than that of Sorai et al. [69SOR/KOS], C p ,m values above
the transition temperature were lower. However, as the results were presented graphically in the rather narrow temperature range mentioned above, they were taken into account only by increasing the error limits.
[78FOR]
The equilibria in the nickel(II)-imidazole-chloride ternary system have been investigated by means of pH-metric and spectrophotometric titrations in 3 M Na(ClO4, Cl)
solution. The latter method was used to determine the stability constant of the binary
NiCl+ complex in the nickel(II)-chloride system. The concentration range studied was
[Ni2+] = 0.096 to 0.300 M and [Cl–] = 1.25 to 3.00 M. The log10 b1 value obtained
(− (0.47 ± 0.10)) is in good agreement with that determined from potentiometric data in
384
A. Discussion of selected references
an indirect way from the ternary system (– (0.53 ± 0.10)). In this review, the mean value
of these constants is accepted with an increased uncertainty ( log10 b1 = – (0.50 ± 0.40)).
[78KAR/HUB]
The stability constant of a complex of pyrophosphate with Ni2+ (presumably NiP2 O72 −
though this is not specified by the authors) was measured by an amperometric titration
procedure. Small amounts of the nickel nitrate salt were titrated with the pyrophosphate
in an ammonium nitrate buffer solution (0.1 M), adjusted to a pH value of 8 with aqueous ammonia. A PbO2 indicating electrode was used with a platinum foil counter electrode and a saturated calomel reference electrode. The procedure was found to be very
sensitive to pyrophosphate even in the presence of phosphate.
At 30°C, the measured value of log10 K was 5.35, and the authors used estimated activity coefficients to obtain log10 K = 6.35 at I = 0. The original data are not
reported and cannot be re-analysed. For reaction:
Ni2+ + P2 O74 − U NiP2 O72 −
in 0.1 M KNO3, log10 K at 30°C = (7.11 + 0.1∆ε)
As the ionic strength is low, the effect of the ∆ε term on the final calculated
value of log10 K ο is limited. Rather than use estimated ε values for interactions involving the pyrophosphate moieties (some of which are highly charged), it is simply assumed that ∆ε is zero. As a consequence, the uncertainty in the log10 K ο value is increased to ± 0.30 for the values from measurements in 0.1 M NH4NO3 solutions.
[78LIB/KOW]
The experimental procedure is very similar to that used in [75LIB/TIA]. In the 1978
work, a potentiometric method was used to determine the stability constants of MBr+
complexes in M(ClO4)2 medium (M = Mn2+, Co2+, Ni2+ and Zn2+) at different concentrations.
The following cell was used for the potentiometric determination of formation
constants: Ag|AgCl,HBr(m1),Mg(ClO4)2(m) || M(ClO4)2(m),HBr(m1),AgCl|Ag, where
m1 was fixed at around 0.01 m, while the molalities of the metal perchlorate (m) were
1.5, 2, 2.5 and 3 m. The emf of the cells varied between 1.6 and 12 mV. Among the
metal ions studied, nickel(II) forms the least stable bromo complex. The variation of the
stability constants within the series is dictated by the formation of inner-sphere complexes to different degrees.
[78LIN/HU]
The activity of sulphur in the high-temperature modification of Ni3S2 was determined as
a function of composition of the solid solution at various temperatures ranging from 823
to 1023 K. The sulphide was equilibrated with gas mixtures of H2(g) and H2S(g) and the
A. Discussion of selected references
385
composition of the gas phase was obtained from the flow rates of the gas streams. When
S2(g) is the standard state for sulphur the activity of sulphur can be calculated from the
well-known thermodynamic data for the gas phase equilibrium H2S(g) U H2(g) +
1/2 S2(g). The composition dependence of the sulphur activity at constant temperature
revealed that the high-temperature modification of Ni3S2 is composed of two nonstoichiometric solid solutions, viz. β1-Ni3S2 and β2-Ni4S3. The homogeneity range as
well as the phase relationships of these phases were described by means of a sub-regular
model. Based on the thermodynamic properties of these mixture phases (β1-Ni3S2 and
β2-Ni4S3) a new phase diagram for the binary system Ni – S was calculated, opposing
the nickel-rich portion of the phase diagram proposed by Kullerud and Yund
[62KUL/YUN] as well as Rau [76RAU].
[78SCH/MIL]
The phase relations in the binary system Ni – S were investigated in the composition
range from xNi = 0.33 to xNi = 0.83 at temperatures between 509 and 1287 K by employing isothermal sublimation in Knudsen effusion cells. The authors derived Gibbs energies of formation for Ni3S2, Ni7S6, NiS, Ni3S4, and NiS2 as a function of temperature as
well as enthalpies of formation for T = 298.15 K. No original experimental data are reported. The results given by the authors (see Table A-16) are not consistent with any
other thermodynamic study on nickel sulphides, published so far. Thus, the data given
in this paper are discarded.
Table A-16: Gibbs energy functions for various nickel sulphides.
Reaction
3Ni + S2 U Ni3S2
7
3
Ni + S2 U ⅓Ni7S6
2Ni +S2 U 2NiS
3
2
Ni + S2 U ½Ni3S4
Ni + S2 U NiS2
∆ r Gmο
(kJ·mol–1)
Temperature range
(K)
∆ r H mο (298.15 K)
(kJ·mol–1)
152.297 – 0.065689 T
925 – 1070
– 65.27
167.36 – 0.09037 T
640 – 840
– 90.79
184.93 – 0.11004 T
670 – 1200
– 72.38
212.97 – 0.08703 T
600 – 625
– 79.08
180.33 – 0.12761 T
509 – 730
– 87.86
[78STU/FER]
For anhydrous NiBr2, the authors report low temperature C p ,m measurements between
7.9 K and 303 K, drop calorimetry measurements to 1150 K, and heat of solution measurements at 298 K. Samples of NiBr2(cr), NiSO4(cr) and H2SO4·6H2O(l) were dissolved
in 4.36 m HCl(aq). Prior to use, the NiBr2 was heated for several weeks at 200°C, and
the X–ray diffraction pattern was claimed to be in good agreement with the ASTM pattern. Nevertheless, it is not certain which crystalline form of the solid was used
[34KET]. The values were combined with the results of earlier measurements to determine ∆ r H m (A.61) = (12.320 ± 0.319) kJ·mol–1.
A. Discussion of selected references
386
2[HCl·12.731H2O](l) +2NaBr(cr) +NiSO4(cr) U
19.462H2O(l) +2NaCl(cr) +H2SO4·6H2O(l) +NiBr2(cr)
(A.61)
The auxiliary data used by the authors were updated using the values in Table A-17.
Table A-17: Auxiliary data used in the recalculation of NaBr heat of formation values
with the data from [78STU/FER].
Reaction #
∆ r H mο
Equation
Reference
(A.62)
0.5H2(g) + 0.5Cl2(g) + 12.731H2O(l) U HCl·12.731H2O
– (162.344 ± 0.167)
a
(A.63)
Na(cr) + 0.5Br2(l) U NaBr(cr)
– (361.160 ± 0.200)
b
(A.64)
Ni(cr) + S(rh) + 2O2(g) U NiSO4(cr)
– (873.280 ± 1.570)
c
(A.65)
Na(cr) + 0.5Cl2(g) U NaCl(cr)
– (411.260 ± 0.120)
d
(A.66)
H2(g) + 2O2(g) + 6H2O + S(rh) U H2SO4·6H2O(l)
– (874.295 ± 0.418)
a
(a) [82WAG/EVA] adjusted to be consistent with values in Table IV.1
(b) [82GLU/GUR] (selected auxiliary datum)
(c) This review. Section V.5.1.2.2.1.3.
(d) Table IV.1 ([82GLU/GUR], selected in [2001LEM/FUG])
Thus, for
Ni(cr) +Br2(l) U NiBr2(cr)
(A.67)
∆ r H mο (A.67) = ∆ r H mο (A.61) + 2 ∆ r H mο (A.62) + 2 ∆ r H mο (A.63) + ∆ r H mο (A.64) –
2 ∆ r H mο (A.65) – ∆ r H mο (A.66),
and
∆ f H mο (NiBr2, cr) = – (211.214 ± 1.752) kJ·mol–1.
The method used is a reasonable way to determine ∆ f H mο (NiBr2, cr), but any
consequent determination of the value of the enthalpy of formation of Ni2+ with this
value for ∆ f H mο (NiBr2, cr) is linked to the value from the sulphate cycle. The heat of
solution of Ni(c) in the sulphuric acid solution is included in part of the cycle to determine the required ∆ f H mο (NiSO4, cr) auxiliary value.
The authors’ integration of their low-temperature heat capacity measurements
(adiabatic calorimetry) led to S mο (NiBr2, cr, 298.15 K) = (122.42 ± 0.25) J·K–1·mol–1.
The reported value of C pο,m (NiBr2, cr, 298.15 K) was (75.40 ± 0.02) J·K–1·mol–1. The
heat capacity results showed no sharp transition above 7.8 K, though there was a small
excess heat capacity between 40 and 55 K, near the Néel temperature. There was no
indication of the transition near 20 K [81ADA/BIL], [82WHI/STA], and the reported
heat capacity values for temperatures between 7.8 K and 15 K are markedly higher than
the corresponding values measured by White and Staveley [82WHI/STA]. Also, as
shown in Figure V-19, for the purposes of extrapolation to 0 K, an equation in T 3 does
not provide a satisfactory representation of the experimental C p ,m values between
7.89 K and 18 K. It appears that the values reported for temperatures below 20 K are
A. Discussion of selected references
387
unreliable. In the present review, the contribution to the entropy from integration of
C p ,m /T between 22 K and 45.4 K was calculated to be only 0.4 J·K–1·mol–1 greater than
the contribution as calculated from the results of White and Staveley; and the contribution to the entropy between 46.5 K and 250.0 K was calculated to be 94.17 J·K−1·mol–1,
in good agreement with 94.31 J·K–1·mol–1 from the later work [82WHI/STA].
The authors’ fitted heat capacity equation reproduces all the high-temperature
experimental H m (T ) − H mο (298.15) values to better than 0.5% and is fixed to the
298.15 K value from the low-temperature measurements. The values from the equation
deviate slightly from the experimental values below 298.15 K.
In the same report, the authors reported low-temperature calorimetry heat capacity measurements (9 to 70 K) for anhydrous NiSO4(cr). Limited sets of dropcalorimetry measurements were carried out for the same solid (403 to 1001.5 K), and
these were used to generate equations for thermodynamic functions for NiSO4 (cr) from
298 to 1200 K.
[79FIS/FOX]
This paper presents a thorough discussion of the pressure dependence of ion association
parameters for several 2:2 electrolytes, including NiSO4(aq). New conductance data are
reported for 15 and 25°C to 2000 atm (analysed using the Fernandez-Prini modification
of the Fuoss-Hsai equation). The trends in the association constants for Ni2+ with sulphate are in agreement with earlier work [76SHI/TSU]. Reanalysis of the 1 atm results
(assumed to be identical at 0.1 MPa) using a version of the Lee-Wheaton equation that
is consistent with the SIT equation used in the present review (Appendix B) leads to
results that are in good agreement with the other studies. The calculated association
constants have a rather large (10%) statistical uncertainty, but it does not appear that the
data are unusually badly scattered. In the present review, we estimate the uncertainty in
K1 at 25°C (0.1 MPa) to be ± 60. The high-pressure results are consistent with values
found in the study of Shimizu et al. [76SHI/TSU].
[79GOL/NUT]
Goldberg et al. reviewed the available literature data, and provided recommended values for the activity coefficients and osmotic coefficients for solutions of Ni(NO3)2 in
water at 298.15 K. The authors commented on the “unusually large amount of scatter”
between (and within) the osmotic coefficient measurements. The tables extend only to
solutions 4.623 m in Ni(NO3)2 and, based on the study of Sieverts and Schreiner
[34SIE/SCH], the saturation solubility is 5.47 m at 298.15 K. Therefore, in the present
review, no thermodynamic quantities for Ni(NO3)2 solids are based on this careful assessment.
388
A. Discussion of selected references
[79KEM/KAT]
Emf measurements on an electrochemical cell involving the Ni – NiO system were carried out in the temperature range from 1191 to 1823 K. No values of ∆ f Gmο (NiO) have
been reported hitherto at temperatures above 1726 K (the melting point of nickel). For
this review, however, the results in the lower temperature range are of primary interest.
The results were expressed by the authors as ∆ f Gmο (NiO) = ((− (232.463 ± 0.251) +
(0.083596 ± 0.000167) (T/K)) kJ·mol–1. The standard deviation of the ∆ f Gmο values
amounted to ± 0.209 kJ·mol–1.
[79RIC/VRE]
The standard molar entropy of solid NiB4O7 was estimated by the linear dependence of
entropies on the radius of the compound’s cation constituent. The extrapolation to
r(cation) → 0 led to the borate contribution of the standard molar entropy depending on
the cationic charge.
[79WEI/HER]
Proton and deuteron relaxation rates in aqueous solutions of NiX2 (X = Cl, Br) in the
concentration range 0.2 – 4 m are reported. The deuteron relaxation rates were found to
depend substantially linearly upon the NiX2 concentration. Comparison of the concentration dependences of the deuteron relaxation rate in these solutions with those in diamagnetic MgX2 solutions yielded information about the average number (nX) of halide
ions bound in the first coordination sphere of Ni2+. For NiCl2 solutions nCl =
(0.12 ± 0.02)mNiCl2, for NiBr2 nBr = (0.085 ± 0.02)mNiBr2 was found. Some additional
measurements with solutions containing 0.02 m NiX2 and 1 – 4 m KX were also performed, and nCl = (0.045 ± 0.04)mKCl, and nBr = (0.06 ± 0.04)mKBr were found. From the
latter data the equilibrium constant for the reaction:
Ni2+ + Cl– U NiCl+
(A.68)
log10 b1 ((A.68), 298.15 K) = – (1.26 ± 1.00) and for:
Ni2+ + Br– U NiBr+,
(A.69)
log10 b1 ((A.69), 298.15 K) = – (1.11 ± 0.60) can be derived in 4.0 m KX solutions.
Using the recommended ion interaction coefficients to calculate ∆ε (A.68) and ∆ε
(A.69) for KX background electrolytes, log10 b1ο ((A.68), 298.15 K) = (0.18 ± 1.00) and
log10 b1ο ((A.69), 298.15 K) = (0.8 ± 0.6) can be derived. For the sake of comparison,
we also calculated the “semi-thermodynamic” formation constants ( log10 b1ο ' ) as defined in [88BJE], using γ± (4 m KCl) = 0.577 and γ± (4 m KBr) = 0.608 [49ROB/STO]:
log10 b1ο ' ((A.68), 298.15 K) = – (1.03 ± 1.00) and log10 b1ο ' ((A.69), 298.15 K) =
− (0.89 ± 0.60).
A. Discussion of selected references
389
[80BAR/RAN]
Reversible potentials of partially charged α- and β-Ni(OH)2 electrodes have been measured up to an oxidation state of ≈ 2.5 over a range of KOH concentrations from 0.01 –
10.0 M. Couples derived from the parent α- and β-Ni(OH)2 systems can be distinguished by the relative change in KOH level on oxidation and reduction as well as the
difference in the formal potentials with respect to Hg | HgO | KOH.
[80CHI/SAB]
The solubility of Ni(OH)2 in pure water was measured from 298.15 to 313.15 K. Only
the solubility products calculated by the assumption 2[Ni(II)] = [OH–] have been given
(at 25°C log10 K sο,0 = – 11.82). Thus, the conclusion has been confirmed that by direct
reaction of Ni(OH)2(cr) with water no reliable solubility data can be obtained [18ALM],
[67BLA]. Thus, these results were not included in either Table V-6 or Figure V-11.
[80DUB/CLA]
The phase transition between millerite, NiS(β), and the NiAs-type high-temperature
modification, NiS(α), was studied by calorimetric and emf measurements. NiS(α) was
prepared by solid state reaction from the elements. Millerite was obtained by annealing
NiS(α) at 627 K for 3 – 4 days. Both solid compounds were checked by X-ray
diffraction analysis. The calorimetric measurements of enthalpy increments during
heating or cooling the sample yielded the temperature and the enthalpy of
transition (β → α), respectively, Ttrs = (667 ± 2) K, ∆ trs H mο = (5.86 ± 0.36) kJ·mol–1.
In addition, the measurement of the emf of the electrochemical cell
C (graphite) | Ni | NiCl2 | NiS | C (graphite) as a function of temperature allowed the
determination of an independent value for the entropy of transition, ∆ trs S mο =
(9.62 ± 0.84) J·K–1·mol–1.
[80LIB/SAD]
Isopiestic results (relative to KCl(aq)) are reported for NiSO4 at 25°C for nickel sulphate molalities between 0.1271 and 1.8526 mol·kg–1, and changes in the UV and visible spectra as a function of the concentration of nickel sulphate are discussed. The authors only present semi-quantitative estimates for the degree of association, and conclude that except for concentrations above 1 M, complexation between Ni2+ and SO 24 − is
primarily outer-sphere complexation.
[80MIL/BUG]
The hydrolysis of nickel(II) in (K)Cl medium was studied by emf titrations at 25°C. In
the first series of titrations, the chloride concentration was held constant at 2.0 M, and
[Ni2+]t was varied from 0.05 – 0.50 M (I = 2.05 – 2.5 M). In the second series, the
[Ni2+]t was held constant at 0.10 M while the chloride concentration was varied from 0.5
to 3.0 M (I = 0.6 – 3.1 M). Taking into account the smaller medium effect, the results of
A. Discussion of selected references
390
the second series are more reliable. The experimental data were interpreted by assuming
that Ni 4 (OH) 44 + was the main hydrolysis product formed. At I = 0.5 and 1.5 M the
value of log10 *b1,1 was also estimated, although the authors noted that the formation of
NiOH+ is not completely certain. The formation of the NiCl+ complex was not considered in the calculation; therefore a correction was applied for the purposes of this review, using the recommended log10 b1ο (NiCl+) and ∆ε(298.15 K, KCl) =
− (0.085 ± 0.05) kg·mol–1 (see the discussion of [71OHT/BIE]). The corrected formation constants are as follows:
Table A-18: Corrected formation constants.
I (M)
0.5
1.0
1.5
2.0
3.0
–
–
*
– (9.87 ± 0.40)
*
– (28.04 ± 0.10) – (28.14 ± 0.06) – (28.28 ± 0.06) – (28.24 ± 0.04) – (28.58 ± 0.06)
log10 b1,1 reported
log10 b 4,4 reported
log10 b1 (NiCl+)
– 0.58
–
– 0.65
log10 *b1,1 corrected – (9.82 ± 0.50)
–
– (10.06 ± 0.40)
– 0.67
– 0.67
– (9.95 ± 0.50)
–
– 0.63
–
*
log10 b 4,4 corrected – (27.83 ± 0.30) – (27.80 ± 0.30) – (27.80 ± 0.30) – (27.63 ± 0.30) – (27.66 ± 0.30)
The SIT-plot of the recalculated data, together with the accepted values for
ο
for the reaction:
[65BUR/LIL2] and [71OHT/BIE], as well as the selected log10 *b 4,4
4Ni2+ + 4H2O U Ni 4 (OH) 44 + + 4H+,
(A.70)
is presented in Figure A-29. As can be seen, the analysis of the data from [80MIL/BUG]
ο
results in the determination of a considerably different log10 *b 4,4
value than the value
selected in this review. The different medium cation can not be the reason of this deviation, since both the aquo ion and hydroxo complex of nickel(II) are positively charged.
Therefore, the above data were not considered further in this review.
A. Discussion of selected references
391
Figure A-29: SIT-plot of recalculated data from [80MIL/BUG] (open diamond),
ο
((A.70),
[65BUR/LIL2] and [71OHT/BIE] (open squares), and the selected log10 *b 4,4
298.15 K) (filled square).
-27.5
-28.0
*
log10 β4,4 − 4D − 4log10 aw
-27.0
-28.5
-29.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
−1
I / mol·kg
[80OPP/TOS]
Oppermann and Toschew studied the thermal decomposition and the sublimation of
crystallised nickel(II) iodide using a membrane manometer and transport ampoules. The
temperature functions of the total pressure (membrane manometer), as well as of the
sublimation pressure (transport ampoules), in a range from 800 to 1000 K, were thermodynamically analysed, and the following quantities were obtained: ∆ f H mο (NiI2, cr,
298.15 K) = – (83.7 ± 8.4) kJ·mol–1, ∆ f H mο (NiI2, g, 298.15 K) = (130.5 ± 20.9)
kJ·mol−1, S mο (NiI2, cr, 298.15 K) = (146 ± 8) J·K–1·mol–1·, S mο (NiI2, g, 298.15 K) =
(335 ± 8) J·K–1·mol–1·, and S mο (Ni2I4, g, 298.15 K) = (536 ± 13) J·K–1·mol–1·. In this
review the uncertainties given by the authors were accepted.
The recalculation of these quantities suffered from the total pressure function
having been presented only graphically. For the sake of comparison with independent
measurements, an attempt was made to recalculate ∆ f H mο (NiI2, cr, 298.15 K) and
S mο (NiI2, cr, 298.15 K), regardless of the uncertainty implicit in reading the relevant
coordinates from a Figure. Based on the auxiliary data used throughout this review the
second-law and approximate third-law methods employed by Oppermann and Toschew
resulted in ∆ f H mο (NiI2, cr, 298.15 K) = – (91.45 ± 4.18) kJ·mol–1 and ∆ f H mο (NiI2, cr,
298.15 K) = – (91.00 ± 8.37) kJ·mol–1, respectively. Within the uncertainty limits, both
of these values overlap with those reported by the original authors. The recalculated
enthalpies of formation also agree with the value selected in this review
(− (96.42 ± 0.84) kJ·mol–1), but their uncertainty is 5 to 10 times greater.
392
A. Discussion of selected references
As a reliable value for S mο (NiI2, cr, 298.15 K) is a prerequisite for the third-law
analysis, this quantity only can be determined by the second-law method. The standard
entropy of nickel(II) iodide is recalculated to be S mο (NiI2, cr, 298.15 K) = (138.0 ± 8.4)
J·K–1·mol–1, which, within the large uncertainty limits, agrees very well with the value
selected in this review ( S mο (NiI2, cr, 298.15 K) = (139.1 ± 1.0) J·K–1·mol–1).
[80REI]
In this work neutral anhydrous transition metal carbonates MCO3 (M = Mn, Fe, Co, Ni,
Zn) were synthesised by a hydrothermal autoclave method. The solubility of the well
crystallised pure carbonates (size of rhombohedral crystals 10 – 100 µm) in
HClO4/NaClO4 solutions of constant ionic strength (1.0 mol·kg–1) and at various partial
pressures of CO2 was studied by means of the pH variation method. At T = 323 K equilibrium was usually attained within a few days. Only NiCO3 was so inert that it had to
be equilibrated at 363 K ≥ T ≥ 348 K. When the Gibbs energy of formation was calculated from the solubility constant obtained it was noted that it is remarkably more negative than the then-accepted value [77BAR/KNA].
[80SHA/CHA]
The thermodynamic properties of all phases in the binary system Ni – S were optimised
for the first time, allowing the calculation of the phase diagram for this system. The
liquid phase was modelled by applying the three suffix Margules equation, the hightemperature phases β1-Ni3S2 and β2-Ni4S3 were described using sub-regular models, and
the non-stoichiometric hexagonal modification of nickel monosulphide was modelled
by application of a statistical model [69LIG/LIB] adopted previously for pyrrhotite, Fe1–
xS, [77LIN/IPS], [79SHA/CHA]. The phase Ni7S6, which shows a narrow homogeneity
range [94STO/FJE], was assumed to be a stoichiometric compound, neglecting its transformation into Ni9S8 at 675 K [94STO/FJE]. The phases Ni3S2, NiS (low-temperature
form), Ni3S4, NiS2 were likewise assumed to be stoichiometric compounds. The calculated phase diagram enables excellent predictions of available experimental phase diagram data [62KUL/YUN], [75ARN/MAL].
However, the phase relations of Ni7S6 and Ni9S8, respectively, were not considered properly. Furthermore, it should be mentioned that the thermodynamic data for
NiS2, the low- and high-temperature forms of NiS are not in perfect coincidence with
decomposition pressure measurements [54ROS], [64LEE/ROS] as well as heat capacity
measurements published recently [95GRO/STO].
[80TRE/LEB2]
The solubility of NiO, characterised by X-ray diffraction analysis, was measured in a
flow apparatus from 423 to 573 K. The solid phase was treated with acidic solutions,
containing HCl with molalities ranging from 10–5 to 5 × 10–4 mol·kg–1, or basic solutions, containing NaOH with molalities ranging from 10–6 to 4 × 10–2 mol·kg–1. An ex-
A. Discussion of selected references
393
tended Debye-Hückel expression was used by the authors in order to calculate the activity coefficients. The pH of the leach solutions was measured with a glass electrode, and
varied between 3 – 12. The nickel content of solutions equilibrated with NiO was determined continuously by means of an atomic-absorption spectrometer. As dilute solutions were investigated, only the formation of mononuclear hydroxo complexes
(NiOH+, Ni(OH)2 and Ni(OH)3− ) was considered. The values for the hydrolysis constants at 25°C were taken from [64PER] and [49GAY/GAR]. From the temperature
dependence of the solubility data the authors derived the first, second and third hydrolysis constants, see Table A-19.
Table A-19: Stability constants for the complexes NiOH+, Ni(OH)2 and Ni(OH)3−
a:
t /°C
log10 *b1,1 (A.71)
log10 *b 2,1 (A.72)
log10 *b 3,1 (A.73)
298.15
– (9.86 ± 0.03)(a)
– (12.01 ± 1.64)(c)
< 19(b)
– (20.26 ± 1.77)(c)
– (29.03 ± 0.50)(b)
– (29.78 ± 1.97)(c)
373
– (9.37 ± 0.85)(c)
– (16.86 ± 0.95)(c)
– (26.03 ± 1.12)(c)
423
– (8.05 ± 0.48)
– (15.18 ± 0.58)
– (24.19 ± 0.76)
473
– (6.95 ± 0.22)
– (13.81 ± 0.35)
– (22.68 ± 0.55)
523
– (6.03 ± 0.20)
– (12.65 ± 0.32)
– (21.42 ± 0.50)
573
– (5.23 ± 0.36)
– (11.67 ± 0.43)
– (20.35 ± 0.57)
Taken from [64PER].
b: Taken from [49GAY/GAR].
c:
Extrapolated from 150 – 300°C.
From these data the authors obtained:
Ni2+ + H2O U Ni(OH)+ + H+
∆ r H mο (A.71) = (87 ± 15) kJ mol–1,
(A.71)
Ni2+ + 2 H2O U Ni(OH)2(aq) + 2 H+
∆ r H mο (A.72) = (109 ± 15) kJ mol–1
(A.72)
Ni2+ + 3 H2O U Ni(OH)3− + 3 H+
∆ r H mο (A.73) = (119 ± 15) kJ mol–1.
(A.73)
The experimental data from this paper have been re-evaluated in the present
assessment. New values for the third hydrolysis constant, valid from 423 to 573 K, were
determined by non-linear curve fitting. The equilibrium values of the time-independent
solubility for each run are plotted versus the molality of HCl or NaOH of the starting
solutions in Figure A-30 and Figure A-31. The solid lines in Figure A-30 correspond to
the prediction of the solubility of NiO in acidic solutions according to the thermodynamic model of the present compilation including the data for NiOH+. In the range
[HCl]initial > 10–4.9 mol·kg–1 the predominant species is Ni2+. The calculated lines are
biased to somewhat higher nickel concentrations which may be caused (i) by insuffi-
A. Discussion of selected references
394
cient equilibration of the solid phase (NiO) with the aqueous media or (ii) by uncertainties of the solubility calculations for hydrothermal conditions owing to the lack of heat
capacity functions for the aqueous species. However, the deviation of the predicted lines
from the experimental data is less than 0.15 logarithmic units.
Figure A-30: Solubility of NiO in acidic solutions as a function of initial molality of
HCl. Symbols refer to experimental data [80TRE/LEB2], the solid lines correspond to
the thermodynamic model of the present assessment.
-3
−1
log10([Ni(II)]tot / mol·kg )
-4
-5
573 K
523 K
473 K
423 K
-6
-7
-8
-5.0
-4.8
-4.6
-4.4
-4.2
-4.0
-3.8
-3.6
-3.4
-3.2
−1
log10([HCl]initial / mol·kg )
The results in basic solutions indicate the occurrence of the hydrolysis species
Ni(OH) owing to the increase of the solubility at higher NaOH concentrations, see
Figure A-31. Due to the scatter of solubility data, the stability constants for Ni(OH)2(aq)
were not determined. For [NaOH]initial < 10–4 mol·kg–1, the predominant species is
NiOH+. The stability constants of the complex Ni(OH)3− are listed in Table A-20 for the
temperature range 423 – 573 K, neglecting the formation of Ni(OH)2(aq).
−
3
A. Discussion of selected references
395
Figure A-31: Solubility of NiO in basic solutions as a function of initial molality of
NaOH. Symbols refer to experimental data [80TRE/LEB2], the lines have been obtained from non-linear least squares fits to the experimental data in order to optimise the
stability constants for Ni(OH)3− ;
, : 423 K; , : 473 K; , : 523 K; , : 573 K.
-4
−1
log10([Ni(II)]tot / mol·kg )
-5
-6
-7
-8
-9
-10
-6.5
-6.0
-5.5
-5.0
-4.5
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
−1
log10([NaOH]initial / mol·kg )
ο
(A.73) for Ni(OH)3− , I = 0
Table A-20: Stability constants log10 *b 3,1
T/K
ο
log10 *b 3,1
ο
log10 *b 3,1
this assessment
[80TRE/LEB2]
423
– 24.61
– 24.2
473
– 22.97
– 22.7
523
– 21.83
– 21.4
573
– 20.64
– 20.3
From the temperature dependence of the optimised data in Table A-20 the linear extrapolation of the stability constant to 298.15 K ( log10 *b 3,1 plotted versus 1/T)
results in:
ο
log10 *b 3,1
= – (30.9 ± 1.3).
A. Discussion of selected references
396
The corresponding enthalpy value for 298.15 K is found to be:
∆ r H mο ((A.73), 298.15K) = (121.2 ± 6.5) kJ·mol–1.
[81BAR/RAN]
Reversible potentials of partially charged α- and β-Ni(OH)2 electrodes have been measured. Experimental emf-oxidation state measurements for both the β | β- and the α | γphase couples agree well with theoretically derived expressions. From considerations of
the homogeneous emf-composition regions it is deduced that the oxidised species are
dissociated in both the β | β- and the α | γ-phase systems.
[81KUL/BLO]
The authors have reported the formation and thermodynamic parameters (K, ∆ r H m ,
∆ r Sm ) of fluoride complexes of the Mn(II) – Zn(II) series in three papers
[76KUL/BLO], [81KUL/BLO], [83AVR/BLO]. In the first paper [76KUL/BLO] they
reported the stability constants and thermodynamic functions of complex formation
determined at 298.15 K and at I = 3 M NaClO4, by means of a fluoride-selective electrode and a differential double calorimeter. The experiments were performed in the
presence of an excess of metal ions (0.1 M Me(ClO4)2) to assure the formation of only
monofluoro complexes.
In the second paper [81KUL/BLO] the authors reported the ionic strength
dependence of thermodynamic quantities using the same experimental techniques. The
initial concentration of the metal ions was varied with varying ionic strength maintained
with NaClO4 ([M] = 0.02 M for I = 0.1 M, [M] = 0.05 M for I = 0.25 M and [M] = 0.1
M for I = 0.5, 1.0, 2.0 and 3.0 M). The data revealed that the stability of MF+ complexes
varies in the following order: Mn > Co ~ Ni < Cu > Zn, which is not consistent with the
Irving-Williams series. The authors also reported thermodynamic properties extrapolated to I = 0, but a reassessment of these data using the SIT was done in this review.
The authors classified NiF+ as an outer-sphere complex, since the addition of fluoride
ions to a solution of nickel perchlorate is not accompanied by any changes in the electronic spectrum, and the thermodynamic quantities are very similar to those of the complex [Co(NH3)5NO2]F+, which is clearly an outer-sphere complex.
In the third paper [83AVR/BLO], the stability constant values for NiF+ have
been reported for 0 – 60 vol. % H2O/EtOH mixture (I = 0.25, 0.5 and 1.0 M NaClO4),
using a fluoride-selective electrode. In contrast with the earlier conclusions, the authors
suggest inner-sphere association for NiF+. This assumption was based on the determination of the so called complete equilibrium constant, in which the water is also taken into
consideration in the equilibrium process:
Ni(H 2 O)62 + + F– U NiF(H 2 O)5+ + H2O(l).
A. Discussion of selected references
397
The reported thermodynamic quantities seem to have been re-determined in
each study, since the three papers report somewhat different values for the parameters
referring to identical conditions. To avoid the over-representation of the data from this
laboratory during our SIT analysis, we used an average value for each ionic strength
studied.
[81MAR/ECO]
Marcopoulos and Economou found almost pure nickel hydroxide in the Vermion region
of northern Greece, described it comprehensively, named it theophrastite
[81MAR/ECO], and studied its genesis [83ECO/MAR]. The new mineral and its name
were approved of by the Commission on New Minerals and Mineral Names, the International Mineralogy Association, IMA.
[81NIC/BER]
Crystallography and composition of nullaginite, found as nodular grains and cross-fibre
vein lets in a nickel-rich assemblage from the Otway deposit, Nullagine district, Western Australia were investigated. The close relationship between nullaginite,
Ni2(OH)2CO3, and rosasite, (Cu, Zn)2(OH)2CO3, as well as glaukosphaerite,
(Cu, Ni)2(OH)2CO3, was pointed out.
[81SCH]
The partial pressure of oxygen for the equilibrium between β2-Ni4S3 and NiO at pSO2 =
1 atm was measured as a function of temperature between 968.6 and 1053.7 K. The emf
of the electrochemical cell Pt, β2-Ni4S3, NiO, SO2 (1 atm) | ZrO2 | O2 (0.0122 atm), Pt
can be easily transformed into the equilibrium partial pressure of oxygen. The solid
phases were checked by X-ray diffraction analysis. During equilibration sulphur was
incorporated in excess in the non-stoichiometric high-temperature phase β2-Ni4S3 due to
reaction with SO2 of the gas phase. The sulphur content of quenched samples was determined by standard analytical procedures. A composition of Ni2.515S2 was calculated
from these chemical analyses. However, it is not expected that the sulphur content of
β2-Ni4S3 remains constant when the temperature is varied. As the variation of the composition of β2-Ni4S3 in equilibrium with NiO and SO2 was not investigated as a function
of temperature, the results of this work are not suitable for derivation of the Gibbs energy function of this phase.
[81YOK/YAM]
The data of Katayama [73KAT] were re-interpreted and compared to a calculated theoretical value. No new data are presented.
[82CHI/SAB]
Chickerur et al. [82CHI/SAB] reported solubility product values for Ni3(PO4)2·8H2O of
4.07 × 10–25 and 1.95 × 10–21 at 30 and 35°C, respectively. The paper provides few de-
398
A. Discussion of selected references
tails, though there is a suggestion that the measurements were done in solutions with
“pH” values < 5.75. At higher temperatures, the authors indicated that Ni3(PO4)2·8H2O
converted to “NiHPO4·H2O”. No allowance was made for complexation, no raw data
were presented, and no values can be recalculated from the sparse information in this
paper.
[82FER/GOK]
Enthalpy increments from 298.15 to 1197 K were measured on Ni3S2 by means of drop
calorimetry, leading to heat capacity values in this temperature region. The samples
were prepared by solid state reaction from the elements and characterised properly
(X-ray diffraction). The transition temperature for the phase transformation between the
high- and low-temperature form was found to be 834 K. The heat capacity at 298.15 K
amounts to 117.7 J·K–1·mol–1, which is in fair agreement with that obtained by Stølen et
al. [91STO/GRO], C pο,m (298.15 K) = 118.2 J·K–1·mol–1.
[82GAM/REI]
The solubility constants measured at 363 K ≥ T ≥ 348 K were extrapolated to T =
298.15 K implying a linear dependence of log10 *K p , s ,0 for the reaction:
NiCO3(s) + 2H+ U Ni2+ + CO2(g) + H2O(l)
on 1/T [80REI]. Data obtained at I = 1.0 and 0.2 mol·kg–1 were extrapolated to I = 0
with the Davies approximation [62DAV]. On this basis ∆ f Gmο (NiCO3, cr, 298.15 K)
was calculated.
[82KLO]
The solubility of NiCO3(cr) synthesised by the hydrothermal autoclave method was
determined at T = 363.15 K and I = 0.2 mol·kg–1 NaClO4.
[82LIV/BIS]
Livingstone and Bish reported X-ray powder diffraction, infrared, thermal, chemical,
optical, and physical data for a Mg-bearing theophrastite from Hagdale Quarry, Unst,
held in the Scottish Mineral Collection. These authors endeavoured to establish a mineral name for the compound (Ni, Mg)(OH)2 by a submission to the IMA in late 1979.
Two months later IMA received another submission for pure Ni(OH)2 [81MAR/ECO].
In the vote the Unst material and name were narrowly defeated in favour of the Greek
Ni(OH)2.
The values of c determined by Livingstone and Bish are intermediate between
those for pure Ni(OH)2 and the isomorphous Mg(OH)2, brucite, indicating that a solidsolution series exists between these end-members.
A. Discussion of selected references
399
[82WAT]
Watanabe carried out measurements of the molar heat capacities of high pressure compounds relevant to the earth’s mantle by differential scanning calorimetry between 350
and 700 K. The investigated compounds were, among others, α-Ni2SiO4 and γ-Ni2SiO4.
In general, the heat capacity for the high pressure phase was found to be smaller than
that for the isochemical low-pressure phase i.e., C p ,m decreases in the phase sequence
olivine (α), modified spinel (β), and spinel (γ). The calculated standard entropies at 298
K for high-pressure phases are presented and entropy changes for various transformations of silicate compounds were discussed. The heat capacities of Ni2SiO4, olivine,
reported by Watanabe are 4.6% larger than the heat capacities obtained by
[84ROB/HEM] with the more accurate low-temperature calorimeter. Therefore, the
results of Watanabe are not used by this review.
[82WHI/STA]
The authors report low-temperature heat capacity measurements (adiabatic calorimetry)
from 8.03 K to 299.83 K. A sample of NiBr2 that had been resublimed at 600°C was
used. When plotted against temperature, the results show a shoulder at 46 K, and a spike
at 19.5 K (Figure A-32). The values between 20 K and 250 K are in reasonable agreement with the earlier work of Stuve et al. [78STU/FER] (cf. discussion in Appendix A).
At lower temperatures, the values are lower, and the value of C p ,m (NiBr2, cr, 8.0 K) is
only 30% of the value reported by Stuve et al. [78STU/FER]. As shown in Figure V-19,
for the purposes of extrapolation to 0 K, an equation in T 3 provides a satisfactory representation of the experimental C p ,m values between 8.03 K and 18 K. In the present review, polynomials were fitted to C p ,m /T, and integration resulted in S m (NiBr2, cr,
250 K) = 108.23 kJ·mol–1. The authors reported difficulties with their measurements
between 250 K and 300 K, and the values of C p ,m increase more rapidly than expected.
Integration of C p ,m /T between 250 K and 300 K gives a contribution to S m (NiBr2,cr)
that is 0.61 J·K–1·mol–1 greater than the contribution obtained by integration of values
from Stuve et al. over the same temperature range. The results from White and Staveley
for the temperature range 250 K to 300 K are not accepted in the present review.
[83AVR/BLO]
See comments under [81KUL/BLO].
[83ECO/MAR]
Marcopoulos and Economou found almost pure nickel hydroxide in the Vermion region
of northern Greece, described it comprehensively, named it theophrastite
[81MAR/ECO], and studied its genesis [83ECO/MAR]. The new mineral and its name
were approved of by the Commission on New Minerals and Mineral Names, the International Mineralogy Association, IMA.
A. Discussion of selected references
400
Figure A-32: Values of C p ,m (NiBr2, cr) as a function of temperature as reported by
White and Staveley [82WHI/STA].
100
-1
–1
C°
·mol-1–1
Cpp,m/ / J·K ·mol
80
60
10
8
40
6
4
2
20
0
5
10
15
20
25
0
0
50
100
150
200
250
300
T / K
T/K
[83GOW/DOD]
A 35Cl NMR study of NiCl2 solutions has shown that at – 5 and + 1°C the system is in
the slow-exchange region. This allowed the concentration of the bound chloride ion to
be calculated, and thus the stability constant for the inner-sphere Ni(H 2 O)5 Cl+ complex
at different concentrations of the salt. The average number of inner-sphere-bound chlorides varied between nCl− = 0.07 and 0.27, depending on the concentration of NiCl2. The
determined stability constants do not show notable temperature or ionic strength dependence, and log10 b1 = (0.034 ± 0.020) was determined from the whole dataset. The
applied ionic strength is out of the validity range of the SIT. Nevertheless, if it is used,
log10 b1ο for the reaction:
Ni2+ + Cl– U NiCl+
can be calculated to be – (0.95 ± 0.50). The method applied has substantial inherent
inaccuracy, but since outer-sphere interaction is unlikely to produce slow ligandexchange, the results verify the existence of an inner-sphere complex.
A. Discussion of selected references
401
[83HOL/MES]
The authors carried out isopiestic measurements on aqueous solutions of nickel sulphate
at 383.14 K and 413.22 K for nickel sulphate molalities between 1.52 and 5.66. The
results were interpreted using a modified Pitzer equation without invoking formation of
NiSO4(aq) as discrete species. The results at these high ionic strengths were not used in
the present assessment.
[83SOL/BON]
The thermodynamics (K, ∆ r H m , ∆ r Sm ) for the formation of monofluoride complexes
of the Mn(II) – Zn(II) series have been studied in methanol at I = 0.05 M (CH3)4NClO4
and 298.15 K with use of fluoride selective electrode potentiometry. Additionally, the K
values have been determined in aqueous solution (I = 0.05 M (CH3)4NClO4 and
298.15 K). In a typical measurement, a solution containing 2 – 40×10–5 M fluoride was
titrated with the solution of the given metal ion, to reach a final metal ion concentration
of 2 – 80×10–4 M. This work was devoted to gaining further insight into the anomalous
stability order of the monohalide complexes of 3d transition metal ions first reported in
[72BON/HEF]. The results indicated that the monofluoride complexes are a clear and
distinct exception to the Irving-Williams sequence; the stability of MF+ complexes was
found to be Mn ~ Fe > Co > Ni < Cu > Zn in both water and methanol. This stability
order does not follow even the trend predicted by ionic radii. Therefore the authors suggested that the monofluoride complexes have some characteristics of ion pairs rather
than those of inner-sphere complexes. The thermodynamic data indicated that the complex (ion pair) formation reactions are all distinctly entropy controlled.
[84BRA/DEL]
Sodium nickelate, NaNiO2, was prepared by direct reaction of sodium oxide and nickel
oxide at 600 – 800°C under a stream of oxygen. This compound was hydrolysed and
reduced by so-called soft chemical reactions in aqueous media at room temperature.
[84COM/PRA]
The standard molar Gibbs energy of formation of NiO was determined at temperatures
as low as 760 K up to 1275 K with an excellent accuracy of ± 0.039 kJ·mol–1 (which is
± 0.2 mV for the standard deviation of the emf). The obtained results were represented
by the equation: ∆ f Gmο (T) = – 232.450 + 0.083435 (T/K) kJ·mol–1.
[84LAV/TIM]
High-purity nickel (wimpurities < 6 ×10–5) was subjected to chlorination in a calorimetric
bomb with a microfurnace for sample heating. The microfurnace developed a temperature of about 1070 K over approximately 10 min at a power input of 75 W. Under these
conditions nickel chlorination proceeded practically to completion.
402
A. Discussion of selected references
[84ROB/HEM]
Robie et al. have carried out excellent measurements of the heat capacity of Ni2SiO4
olivine between 5 and 387 K by cryogenic adiabatic-shield calorimetry and between
360 and 1000 K by differential scanning calorimetry. At 298.15 K the molar heat capacity and entropy of Ni2SiO4 olivine were determined. This molar heat capacity and entropy were accepted by this review. The thermal heat capacity function proposed by
Robie et al. for the temperature range between 300 and 1300 K was, however, not
adopted by this review, because it is based on measurements which were made only at
temperatures up to 1000 K. Combining the heat capacity measurements with results of
molten salt calorimetry, thermal decomposition of Ni2SiO4 olivine into its constituent
oxides, and equilibrium studies, both by CO reduction and solid state electrochemical
cell measurements for the reaction:
2Ni + SiO2 + O2 U Ni2SiO4,
Robie et al. calculated for Ni2SiO4 olivine ∆ f Gmο (298.15 K) = – (1289.0 ± 3.1) kJ·mol–1
and ∆ f H mο (298.15 K) = – (1396.5 ± 3.0) kJ·mol–1.
[84ROG/BOR]
The authors carried out emf-measurements using a galvanic cell with a Ni2+ exchanged
β-alumina solid electrolyte to determine the Gibbs energy, enthalpy and entropy of formation of nickel orthosilicate from the constituent oxides. The Gibbs energy of the cell
reaction as a function of temperature, given by Róg and Borchard for the temperature
range 970 to 1370 K, coincides very well with results of other emf-studies of Ni2SiO4,
as shown in this review. The estimated temperature of decomposition of Ni2SiO4 to NiO
and SiO2, 1790 K, is significantly lower than the value recommended by this review
(1820 ± 5) K.
[84VAS/VAS]
Vasil’ev et al. determined ∆ r H m of Reaction (A.74) calorimetrically at 298.15 K. Metallic nickel was dissolved in 2.0 to 6.0 M HCl or HClO4 solutions containing 1.0 to
1.5% H2O2.
Ni(cr) + 2H+ + H2O2(aq) U Ni2+ + 2H2O(l)
(A.74)
When the authors plotted the measured enthalpies minus the Debye-Hückel
term versus the ionic strength (which appears to be dominated by the acid concentration) the data fell on straight lines with slopes depending on the H2O2 content, but the
intercepts coincided within the error limits. Thus, the results were extrapolated semiempirically to I = 0, resulting in ∆ r H mο (A.74) = – (432.59 ± 0.43) kJ·mol–1. As the SIT
model cannot predict the H2O2 dependence of the slopes this evaluation has been accepted as is. With the standard enthalpies of ∆ f H mο (H2O, l) = – (285.830 ± 0.040)
kJ·mol–1 and ∆ f H mο (H2O2, aq) = – (191.170 ± 0.100) kJ·mol–1 taken from CODATA
A. Discussion of selected references
403
and the NBS Tables, respectively, ∆ f H mο (Ni2+) = – (52.10 ± 0.45) kJ·mol–1 has been
obtained [82WAG/EVA], [89COX/WAG].
[85DRA/MAD]
Apparent molar heat capacities were determined for nickel sulphate solutions at 50, 70
and 90°C. The lowest molalities used were 0.3 mol·kg–1, and the extrapolation to I = 0 is
problematic, especially if a model that invokes association between nickel and sulphate
ions is to be used. The values from this study are not used in the present assessment.
[85MEH/TAR]
The partial pressure of oxygen for the solid phase equilibrium β2-Ni4S3 / NiO at pSO =
2
1 atm was determined in the temperature range 973 – 1173 K by performing emf measurements on the electrochemical cell Pt, O2 (air) | YSZ | NiO, β2-Ni4S3, SO2 (1 atm), Pt
or Au. Yttria-stabilised zirconia served as the solid electrolyte. The solid phases were
prepared by solid state reactions from the elements. Phase identification was carried out
by means of X-ray diffraction analysis and the composition of the solid nickel sulphide
was checked by chemical analysis. The authors found a composition of Ni2.666S2. The
variation of this composition with temperature was reported to be not appreciable, but
was not checked carefully by the authors. Hence, the results of this work are apparently
not suitable for the derivation of the Gibbs energy function for this phase.
[85SKE/MAI]
Skeaff et al. [85SKE/MAI] used an electrochemical technique to determine the Gibbs
energies of reaction of NiAs(cr) with oxygen to form NiO (bunsenite) and As4O6(g)
from 711 K to 833 K. Similar measurements were carried out with Ni11As8(cr) (805 K
to 991 K) and Ni5As2(cr) (937 K to 1139 K). Experiments with NiO-NiAs2 and
NiAs-NiAs2 mixtures were unsuccessful in that there was no indication of any change in
potential with changes in partial pressure of As4O6(g).
Based on the approximation that ∆ r C p ,m = 0 for
y
(2 x + 3 y )
xNiO(cr) + As4O6(g) U NixAsy(cr) +
O2(g)
4
4
the authors’ estimated heat capacity equations (in J·K–1·mol–1) were:
C pο,m (NiAs, cr)
= 63.7532 + 6.4616 ×10–3 (T/K) – 8.6016 ×105 (T/K)–2
C pο,m (Ni11As8, cr) = 605.483 + 70.7846 ×10–3 (T/K) – 66.3005 ×105 (T/K)–2
C pο,m (Ni5As2, cr) = 222.965 + 32.0067 ×10–3 (T/K) – 14.6911 ×105 (T/K)–2
and the following values were reported: ∆ f H mο (NiAs, cr, 298.15 K) = – 73.29 kJ·mol–1,
S mο (NiAs, cr, 298.15 K) = 45.4 J·K–1·mol–1, ∆ f H mο (Ni11As8, cr, 298.15 K) = – 777.04
kJ·mol–1, S mο (Ni11As8, cr, 298.15 K) = 468.9 J·K-1·mol-1, ∆ f H mο (Ni5As2, cr, 298.15 K) =
– 251.10 kJ·mol–1, and S mο (Ni5As2, cr, 298.15 K) = 190.6 J·K–1·mol–1.
A. Discussion of selected references
404
The authors used values for the heat capacity and entropy of As4O6(g) from the
analysis of Behrens and Rosenblatt [72BEH/ROS], and tied the value for ∆ f H mο (As4O6)
to the vapour pressure measurements of [72BEH/ROS] and Murray et al.
[74MUR/POT]. Behrens and Rosenblatt, in turn, relied on values calculated from the
thorough sets of low-temperature heat capacity measurements of Chang and Bestul
[71CHA/BES] for the As4O6 solids. As was discussed by Grenthe et al. [92GRE/FUG],
auxiliary data for arsenic compounds have not been critically reassessed in the TDB
Project, and at 298.15 K several of the self-consistent values from the U.S. National
Bureau of Standards compilation [82WAG/EVA] are accepted as an interim measure.
The values ∆ f H mο (As4O6, g, 298.15 K) = – (1196.2 ± 16.0) kJ·mol–1 and S mο (As4O6, g,
298.15 K) = (408.6 ± 6.0) J·K–1·mol–1 were obtained by adjustment of the corresponding
values reported by Behrens and Rosenblatt [72BEH/ROS], and have been used in the
present review.
The reported reaction enthalpies and entropies for Reactions (A.75), (A.76),
(A.77) are listed in Table A-21
NiO(cr) + 1 As4O6(g) U NiAs(cr) + 5 O2(g)
4
4
11 NiO(cr) + 2 As4O6(g) U Ni11As8(cr) + 23 O2(g)
2
5 NiO(cr) + 1 As4O6(g) U Ni5As2(cr) + 4 O2(g)
2
(A.75)
(A.76)
(A.77)
Table A-21: Enthalpies and entropies of reaction from Skeaff et al. [85SKE/MAI].
arsenide
∆ r H mο
(kJ·mol–1)
∆ r S mο
(J·K–1·mol–1)
Tr *
(K)
NiAs(cr)
465.458
159.057
772
Ni11As8(cr)
4254.395
1564.75
800
Ni5As2(cr)
1544.845
604.452
1038
* approximate mid-temperature for the measurements
Equations for C pο,m (NiAs) and C pο,m (Ni11As8) from work of Muldagalieva et al.
[93MUL/ISA], [95MUL/CHU] are used with auxiliary data for As4O6(g) (see Chapter
IV), O2(g) [98CHA], and values for NiO(cr) from the present review to obtain
∆ f H mο (NiAs, cr, 298.15 K) = – (69.73 ± 5.00) kJ·mol–1, S mο (NiAs, cr, 298.15 K) =
(50.76 ± 6.00) J·K–1·mol–1, ∆ f H mο (Ni11As8, cr, 298.15 K) = – (743.0 ± 35.0) kJ·mol–1
and S mο (Ni11As8, cr, 298.15 K) = (518 ± 60) J·K–1·mol–1. In the absence of the original
data except as points on a graph, but with the reported uncertainties in the values for
∆ r Gm from the equations, the uncertainties here are estimated by the reviewer. For both
solids, the heat capacity equations are used slightly beyond the range suggested by the
original authors, but the C p ,m (T) curves appear to be monotonic, and there are no better
A. Discussion of selected references
405
data. For Ni5As2(cr), no heat capacity equation was available, and based on the approximation ∆ r C p ,m = 0 used by the original authors, and auxiliary data from the present
review, ∆ f H mο (Ni5As2, cr, 298.15 K) = – (244.6 ± 20.0) kJ·mol–1, and S mο (Ni5As2, cr,
298.15 K) = (196.2 ± 30.0) J·K–1·mol–1 is obtained. The uncertainties have been increased to account for the lack of proper heat capacity data.
[86BHA/MUK]
Bhattacharya et al. report the electrochemical generation of the first discrete NiIIIO6
species, Ni(bpyO 2 )33+ . The cation is low spin and rhombic. This paper is interesting in
the present context, because the redox potential of Ni(bpyO 2 )33+ | Ni(bpyO 2 )32 + can be
measured and correlated to the redox potential of Ni3+ | Ni2+.
[86CEM/KLE]
The standard enthalpies of formation of various phases in the system Ni – S were determined by means by drop calorimetry. Synthetic NiS was prepared by solid state reaction from the elements. X-ray diffraction and microprobe analysis revealed that the
high-temperature modification of NiS was obtained. The minerals vaesite (NiS2) and
heazlewoodite (Ni3S2) were synthesised by reaction of NiS with sulphur and metallic
nickel, respectively. All samples were proved to be well-crystallised material. The enthalpy of formation for the stoichiometric composition of the hexagonal NiAs-type
nickel monosulphide, NiS(α), at 298.15 K was determined by dropping evacuated silica
ampoules containing a stoichiometric mixture of powdered metallic nickel and sulphur
from room temperature (298.15 K) into the hot calorimeter (1021 K) where the formation of NiS occurred. In a second experiment the silica ampoules containing NiS formed
during the first experiment were dropped from room temperature into the hot calorimeter. Obviously, the difference of the enthalpy effects corresponding to the dropping experiments directly yields the enthalpy of formation of the high-temperature form of NiS
at 298.15 K, ∆ f H mο = – (88.1 ± 1.0) kJ·mol–1.The enthalpy of formation of the lowtemperature modification, NiS(β), cannot be obtained directly because in the high temperature calorimeter NiS(α) always is formed, and this solid does not transform into
NiS(β) upon quenching.
The enthalpy of formation of heazlewoodite was obtained by dropping evacuated silica capsules containing a mixture of powdered metallic nickel and NiS from
room temperature into the calorimeter at 873 K. Again, in a second experiment the silica capsules containing the product of the first reaction (N3S2) were dropped into the
calorimeter. Taking into account the enthalpy of formation of NiS determined earlier,
the difference of the enthalpy changes associated with these experiments leads to the
enthalpy of formation of heazlewoodite at 298.15 K, ∆ f H mο = – (217.24 ± 1.60) kJ·mol–1.
The enthalpy of formation of vaesite (NiS2) at 298.15 K was derived from the
reaction:
A. Discussion of selected references
406
NiS2 + Ni U 2NiS(α)
at 873 K in the calorimeter. In a second experiment the heat content of the silica ampoule and NiS formed during the reaction was measured. The difference between the
enthalpy changes associated with these experiments results in the enthalpy of formation
of vaesite at 298.15 K, ∆ f H mο = – (124.9 ± 1.0) kJ·mol–1.
The enthalpy of formation for Ni7S6 was likewise determined by reaction of
NiS with Ni at 833 K in the calorimeter. At these conditions the non-stoichiometric
high-temperature modification of Ni7S6 is formed. However, when this product is
quenched, metastable modifications are obtained depending on the cooling rate
[96SEI/FJE]. Therefore, the enthalpy of formation reported by the authors instead
should be assigned to formation of metastable phases of Ni7S6 or to a mixture of Ni3S2
and Ni9S8.
All experimental results of this interesting paper are listed in Table A-22.
Table A-22: Enthalpies of formation for solid nickel sulphides at 298.15 K.
Solid phase
∆ f H mο (kJ·mol–1)
NiS(α), high-temperature form
– (88.1 ± 1.0)
Ni3S2
– (217.2 ± 1.6)
NiS2
– (124.9 ± 1.0)
Ni7S6, maybe ⅓Ni3S2 + ⅔Ni9S8
– (582.8 ± 5.7)
[86HOL/NEI]
The standard Gibbs energy of formation of NiO has been experimentally determined
over the temperature range from 900 to 1400 K using a galvanic cell with the solid electrolyte made of 15% calcia-stabilised zirconia. The measured value of ∆ f Gmο (NiO) at
1300 K was – 123.555 kJ·mol–1 with a precision of ± 0.057 kJ·mol–1 and an estimated
accuracy not worse than 0.200 kJ·mol–1. This precision is equivalent to an error of only
0.2 – 0.3 mV in the cell potential (emf). In comparison, most previous studies have reported a precision of 1 – 2 mV. Using the third law analysis, the authors obtained for the
enthalpy of formation of NiO, ∆ f H mο (NiO, 298.15 K) = – 240.110 kJ·mol–1.
[86HSI/CHA]
Equilibria between two solid phases in the ternary system Ni – S – O at pSO = 1 atm
2
were studied by using the electrochemical cell Pt, O2(air) | YSZ or CSZ | mixture of
phases, SO2 (1 atm), Au, where yttria- or calcia-stabilised zirconia served as the solid
electrolyte. The following phase equilibria were studied: NiS(α) / β2-Ni4S3; β2-Ni4S3 /
NiO; NiS(α) / NiO; and NiO / NiSO4. While the solid sulphides and NiO were prepared
by solid state reaction from the elements, NiSO4 was synthesised by dehydrating
NiSO4·6H2O. All solid phases were identified by X-ray diffraction analysis. The emf of
A. Discussion of selected references
407
the electrochemical cell, recorded as a function of temperature, corresponds directly to
the partial pressure of oxygen at pSO = 1 atm for the phase equilibria mentioned above.
2
Figure A-33 shows the temperature dependence of the oxygen partial pressure for
NiS(α) / β2-Ni4S3 as well as β2-Ni4S3 / NiO, where the experimental results of [81SCH],
[85MEH/TAR] are likewise included. It can be deduced from Figure A-33 that the
thermodynamic models of both [80SHA/CHA] and the present data assessment are in
close agreement with the experimental results for the phase equilibrium NiS(α) /
β2-Ni4S3. The experimental data from various literature sources [81SCH],
[85MEH/TAR], [86HSI/CHA] for the equilibrium β2-Ni4S3 / NiO coincide remarkably
well with each other despite the fact that Mehrotra et al. [85MEH/TAR] and Schaefer
[81SCH] gave constant but different values for the composition of the nonstoichiometric phase β2-Ni4S3. However, it is to be expected that the composition of β2Ni4S3, which is fixed by the phase equilibrium with NiO at pSO = 1 atm, should change
2
with temperature. Moreover, it is worth mentioning that the temperature dependence of
the sulphur content in β2-Ni4S3 in equilibrium with NiO has not yet been investigated
experimentally.
In addition, the variation of the oxygen partial pressure with temperature for
the equilibrium NiS(α) / NiO at pSO = 1 atm is depicted in Figure A-34. The experi2
mental data agree reasonably with equilibrium calculations using the present thermodynamic model.
Figure A-33: Plot of logarithm of the oxygen partial pressure versus temperature for the
phase equilibria α-NiS / β2-Ni4S3 and β2-Ni4S3 / NiO, respectively, at pSO = 1 atm. ●, ■
2
[86HSI/CHA]; U [81SCH]; dotted line: [85MEH/TAR]. Dashed line: thermodynamic
model according to [80SHA/CHA]; solid line: thermodynamic model of the present
assessment.
-11.0
NiO
-11.5
log10 [p(O2) / bar]
-12.0
β 2-Ni4S3
-12.5
NiS(α)
-13.0
-13.5
-14.0
920
940
960
980
1000
1020
T/K
1040
1060
1080
1100
A. Discussion of selected references
408
Figure A-34: Logarithm of the oxygen partial pressure plotted as a function of temperature for the equilibrium α-NiS / NiO at p(SO2) = 1 atm. ● [86HSI/CHA]. Solid line:
thermodynamic model of the present assessment.
-13.8
NiO
log10 [p(O2) / bar]
-14.0
-14.2
-14.4
-14.6
NiS(α)
-14.8
-15.0
870
880
890
900
910
920
930
940
T/K
[86JAC/KAL]
The emf-study of Ni2SiO4 done by Jacob et al. was not used for the evaluation of thermodynamic data by this review, because the results were presented only in graphical
form, making accurate recovery of experimental values difficult.
[86VAS/DMI]
Vasil’ev et al. reported another value of ∆ f H mο (Ni2+) = – (51.83 ± 0.88) kJ·mol–1 determined calorimetrically by i) dissolving Ni (cr) in solutions containing 0.3 M Br2, 0.05 M
HClO4, 0.95 to 3.95 M NaBr, ii) accounting for the enthalpy of mixing NiBr2 as well as
iii) Br2 to make up the final solutions, and iv) extrapolating to I = 0. The re-calculation
with the CODATA value for ∆ f H mο (Br–) = – (121.41 ± 0.15) kJ·mol–1 resulted in a
minute correction only, but the error limits appear too optimistic in view of four different contributions to the final result, and thus were set to 2σ = ± 2.0 kJ·mol–1. Thus,
∆ f H mο (Ni2+) = – (51.85 ± 2.00) kJ·mol–1 has been assigned to this experimental information.
A. Discussion of selected references
409
[87BEC]
The formation constants of the proton and metal ion complexes of cyanide ion have
been critically surveyed. The recommended value for pK οHCN is correct, but the tentative
values for higher ionic strengths seem to be erroneous [92BAN/BLI]. Several literature
values for the formation constants of Ni(CN) 24 − and Ni(CN) 24 − as well as reaction enthalpies for the formation of Ni(CN) 24 − were reported, but no recommended values
were given. In addition, the formation constants of several NiX(CN)x complexes (X =
multidentate organic ligands, x = 1 or 2) were listed.
[87BJE]
These “semi-thermodynamic” formation constants are not equivalent to the value determined by extrapolation to zero ionic strength using the SIT, except for the case when
the ratio:
γ[NiXn ( H2O )
6− n
]
/ γ [NiXn−1 ( H2O )
7−n
]
=1
in the whole range of the ionic strength studied.
[87CEM/KLE]
The standard enthalpy of formation of synthetic pentlandite, Fe4.5Ni4.5S8, and natural
violarite, (Fe0.2941Ni0.7059)3S4, were measured by solution drop calorimetry. Synthetic
pentlandite was prepared from the elements and checked by X-ray diffraction analysis.
First a stoichiometric mixture of FeS, NiS, Ni, and Fe was dropped from 298.15 K into
a melt of Ni0.6S0.4 at 1100 K. In a second experiment synthetic pentlandite was dropped
into the Ni0.6S0.4 melt. The enthalpy of formation of pentlandite from its elementary and
binary compounds is obtained from the difference of the enthalpy effects associated
with these two experiments, ∆ r H mο = – (78.41 ± 14.45) kJ·mol–1 for:
4 NiS + 4 FeS + 0.5 Ni + 0.5 Fe U Ni4.5Fe4.5S8.
Taking ∆ f H mο = – (101.67 ± 0.20) kJ·mol–1 for the standard enthalpy of formation of FeS from the JANAF tables [98CHA] and ∆ f H mο = − (88.1 ± 1.0) kJ·mol–1 for
the high-temperature modification of NiS [86CEM/KLE], the standard enthalpy of formation of pentlandite at 298.15 K is calculated to be ∆ f H mο = − (837 ± 15) kJ·mol–1. By
using a similar procedure, the authors arrive at ∆ f H mο = − (378 ± 12) kJ·mol–1 in the
case of violarite with the composition (Fe0.2941Ni0.7059)3S4.
[87EMA/FAR]
Emara et al. gave an example of how they made up their solutions. For a solution intended to contain 2.4×10–4 M HCO3− they started with 6.18×10–4 M NaHCO3, but in
reality this leads to 6.05×10–4 M HCO3− . Thus the results of this paper are entirely unreliable.
410
A. Discussion of selected references
[87FLE]
On a specimen which also originated from the Cu-Ni sulphide deposit Noril'sk, Siberia,
Russia, Fleet carried out a definitive crystal structure analysis of godlevskite. Discrepancies between the unit-cell parameters of [69KUL/ERS] and [87FLE] are attributable
to incorrectly indexed powder data.
Discrepancies between the chemical composition used by [69KUL/ERS] and
[87FLE] are attributable to advances in electron microprobe technology and calibration
procedures applied in the latter study. Although there are similarities in the unit-cells,
powder patterns, and S-atom arrays of godlevskite and pentlandite, godlevskite does not
have a pentlandite derivative crystal structure.
[87GAR/PAR]
The CODATA Task Group on Chemical Thermodynamic Tables presented recommended values for thermodynamic properties of selected compounds of calcium and
their mixtures in this prototype set of tables.
[87NEI]
This is an excellent study of NiO, CoO, Ni2SiO4 and Co2SiO4 using an electrochemical
technique. Cells were used with calcia-stabilised zirconia as an oxygen specific electrolyte. The fact that not only the overall cell reaction:
2 Ni + SiO2 + O2 U Ni2SiO4
but also the equilibria of Ni + NiO and Fe + FeO at the reference electrode were investigated in detail, make the obtained results very reliable. Moreover, some information is
also given for the two equilibria involving SiO2 both quartz and cristobalite. The Gibbs
energy of the above cell reaction measured between 972 and 1326 K was compared by
O’Neill with results of previous work and fairly good agreement has been found between all emf results. However, the discrepancy between the emf and calorimetric data
of [84ROB/HEM] remains unresolved. In order to approach this problem, O’Neill made
additional experiments to confirm the decomposition temperature of Ni2SiO4 to the oxides and found that Ni2SiO4 breaks down to NiO and SiO2-cristobalite at (1820 ± 5) K.
Taking the Gibbs energy of the cell reaction and the proper decomposition temperature
of Ni2SiO4, a consistent value of the standard molar enthalpy of the reaction under consideration has been ascertained.
[88BJE]
The formation constant of the NiCl+ complex has been determined by a
spectrophotometric study at 298.15 K in concentrated HCl (4.4 – 10.67 M) and LiCl
(2.94 – 12.46 M) solutions, and the formation constant of the NiCl2 complex was
tentatively estimated. The method used by the author results in a “semithermodynamic” constant, in which the activity of chloride ion is considered, but the
ratios γ Ni2+ / γ NiCl+ and γ NiCl+ / γ NiCl2 were assumed to be nearly constant with
A. Discussion of selected references
411
γ NiCl+ / γ NiCl2 were assumed to be nearly constant with increasing MCl concentration,
and were included in the reported formation constants (see discussion on [89BJE]). The
method used by the author is more appropriate to the study of weak complex formation
than the constant ionic medium principle if very high and varying concentrations of a
complex-forming anion is used. (see discussion on [89BJE]). Nevertheless, the formation constants determined in this way are not compatible with the log10 b ο values extracted from the SIT analysis
[88BJE2]
The “semi-thermodynamic” stability constant (see [89BJE]) for the reaction:
Ni(SCN)3− + SCN– U Ni(SCN) 24 −
was determined by a spectrophotometric method at 298.15 K in concentrated aqueous
NaSCN (2.14 – 10.5 M) solutions, using the main activity coefficients of NaSCN solutions reported in [56MIL/SHE]. The reported log10 K 4 value is − (0.92 ± 0.18). Although, the method used by the author is probably more appropriate to the study of
weak complex formation than the constant ionic medium principle, if high and varying
complex-forming anion concentrations are needed (see the discussion of [89BJE]), the
constants determined in this way are not compatible with the log10 b ο values extracted
from the SIT analysis. Therefore, the reported constant was not considered further in
this review.
[88EFI/EVD]
Efimov et al. devised an ingenious thermodynamic cycle to measure the enthalpy of
Reaction (A.78) at 298.15 K by solution calorimetry:
Br2(l) + Ni(cr) + 2 KCl(cr) U NiCl2(cr) + 2 KBr(cr).
(A.78)
Five calorimetric experiments were designed in such a way that the aqueous
species of the solutions used cancelled when the individual reactions were combined
according to Equation (A.78) and ∆ r H mο (A.78) = – (218.38 ± 0.36) kJ·mol–1. Then,
∆ f H mο (NiCl2, cr) = ∆ r H mο (A.78) – 2 ( ∆ f H mο (KBr, cr) – ∆ f H mο (KCl, cr) )
= – (303.71 ± 1.10)
can be recalculated using the selected standard enthalpies of formation for crystalline
KCl and KBr (cf. Table IV-1 and Chapter VI). The assigned uncertainty limits correspond to twice the standard deviation given by the authors. The value is consistent with,
but less precise than the value determined by Lavut et al., [84LAV/TIM] and the value
derived in this review from Busey and Giauque's equilibrium studies [53BUS/GIA].
[88LIC2]
Based on the Gibbs energy of formation for I, II, and III-NiS, derived from the experimental investigation of the solubility of nickel sulphide performed by Thiel and Gessner
A. Discussion of selected references
412
[14THI/GES], and a new value for the second dissociation constant of H2S, the solubility products for I, II, and III-NiS were recalculated. The author arrived at the following
values for the pertinent solubility products:
log10 K sο,0 (I-NiS) = – 24.3, log10 K sο,0 (II-NiS) = – 29.5, log10 K sο,0 (III-NiS) = – 31.3.
[88NIS/ITO]
Nishimura et al. measured the solubility for Ni3(AsO4)2·8H2O and, based on measurements at 25°C over a range of pH from 4 to 8, reported log10 K = 9.9 for
Ni3(AsO4)2·8H2O + 4H+ U 3Ni2+ +2 H 2 AsO −4 + 8H2O.
Langmuir et al. [99LAN/MAH], using this equilibrium constant, but different
auxiliary data, calculated log10 K s ,0 = – 27.02 for
Ni3(AsO4)2·8H2O U 3 Ni2+ + 2 AsO34− + 8 H2O
(A.79)
and log10 K s ,0 (A.79) = – 28.38 when corrected for associative complex formation. In
the present review, the correction to log10 K s ,0 for complexation is accepted. Then, if
the auxiliary data in Table IV-2 are used, log10 K s ,0 (A.79) = – 28.12 is calculated. The
authors did not propose uncertainty limits but, based on the figures in the original paper,
an uncertainty of ± 0.50 in the equilibrium constant seems reasonable and is accepted
here.
The authors also reported a value of log10 K = 1.8 for the solubility product for
NiHAsO4·2H2O from measurements done from pH 2 to 3.
NiHAsO4·2H2O + H+ U Ni2+ + H 2 AsO −4 + 2 H2O.
This value is accepted in the present review, again with an uncertainty of ± 0.5.
[89BJE]
The limitations inherent in the constant ionic medium principle have been discussed
with respect to the determination of weak stability constants for metal-anion complexes.
The author suggests that the use of the constant medium principle to study weak complexes, for which the background electrolyte should be almost entirely replaced by the
complex-forming electrolyte to enhance the complex formation, introduces an inherent
and electrolyte-dependent uncertainty, due to the change in activity coefficients. According to Harned's rule, the logarithms of the activity coefficients in a mixture of two
electrolytes vary linearly with their mole fraction. This variation, which depends mainly
on the hydration of the ions involved, results in an uncertainty that can be very large
relative to the stability constants of weak complexes. For this reason, the author proposed a different method to study weak complex formation. In a series of papers (see
e.g., [88BJE], [90BJE]) he published spectrophotometric studies of such systems, using
relatively small concentrations of the metal ion in solution with up to the highest possi-
A. Discussion of selected references
413
ble concentrations of highly soluble (pseudo)halogenides such as HCl, LiCl, LiBr or
NaSCN. His data are based on the following three approximations:
(i)
the logarithm of the main activity coefficient of the complex-forming electrolyte
(NX) increases linearly with the concentration above 3 M ( log10 γ ±NX = A + B[X]),
(ii) the ratio of the activity coefficients of MXn and MXn–1 does not change with increasing salt concentration (MX) and thus the following “semi-thermodynamic”
mass action expression can be used b nο ' = [MXn]/[MXn–1] × aX (where aX = [X]
× γ ±NX )
(iii) the absorption spectrum of a given complex does not change with respect to
changes in medium.
The latest isopiestic studies confirm the validity of assumption (i), while assumption (iii) is certainly more valid for d-d transitions than for charge-transfer bands.
In this respect, it is instructive to follow the scientific debate between J. Bjerrum and R.W. Ramette concerning the correctness of stability constants of copper(II)chloride complexes [83RAM/FAN], [86RAM], [86BJE/SKI], [87BJE]. Bjerrum stated
that maintaining a high constant ionic strength, but changing the background electrolyte,
results in unacceptable uncertainty for K3 and K4. On the other hand, Ramette claimed
that spectrophotometric data determined by Bjerrum et al. are also subject to a medium
effect.
These “semi-thermodynamic” formation constants (at I = 0) are not equivalent
to the values obtained by extrapolation to zero ionic strength using the SIT, except for
the case when the ratio:
γ [NiX
γ [NiX
n (H 2 O)6− n ]
n−1 (H 2 O)7− n
=1
]
in the whole range of the ionic strength studied.
[89BRA]
Standard electrode potentials of the elements and their temperature coefficients in water
at 298.15 K were listed for nearly 1700 half-reactions at pH = 0.000 and pH =13.996.
General and specific methods of estimation of thermodynamic quantities were described, but the original papers, from which the individual entries were derived, have
not been cited.
[89EVD/EFI]
The enthalpy of formation of NiBr2(cr) was determined from the heats of solution of
Ni(cr) and NiBr2(cr) in KBr-HBr(aq)-Br2(aq) mixtures (1.00KBr, 0.40Br2, 0.12HBr,
50.87H2O - probably molar ratios). The sample preparation appears to have been done
carefully, but the resultant solid probably had the “Wechselstruktur” [34KET].
414
A. Discussion of selected references
With solution 1 defined as (1.00KBr, 0.40Br2, 0.12HBr,50.87H2O), the reactions were:
Br2(l) + soln 1 U Br2(soln 2)
(A.80)
Ni(cr) + Br2(soln 2) U NiBr2 (soln 3)
(A.81)
NiBr2(soln 1) U NiBr2 (soln 3)
(A.82)
Table 1 in the translation contains some obvious typographical errors. For reaction (A.81), the last four reported heats are a factor of 10 too low. For reaction (A.82),
the reported heats are too low by a factor of 103. The reported uncertainties seem to be
average uncertainties, close to 1σ uncertainties. The reported values have been reaveraged, and the values are – (8.239 ± 0.216) kJ·mol–1, – (289.442 ± 0.712) kJ·mol–1,
and – (85.285 ± 0.311) kJ·mol–1 for the authors’ reactions 1, 2 and 3 respectively. The
correction of (0.462 ± 0.001) J·K–1·mol–1 applied by the authors to allow for the presence of some bromine in the gaseous state in the measurements of the heat of reaction
(A.80) is accepted. From these values, ∆ f H mο (NiBr2, cr) = − (211.93 ± 0.90) kJ·mol–1,
2σ uncertainties.
[89IUL/POR]
Chloride complexation with nickel(II) and copper(II) ions has been studied at 298.15 K
by solubility determinations of M(IO3)2 in aqueous solutions of 6 M Na(ClO4,Cl). The
chloride concentration varied from 0 to 6 M. Medium changes, caused by the replacement of ClO −4 by Cl– ions, were accounted for by the SIT. The solubility product of
Ni(IO3)2 is reported to be (5.1 ± 0.2) × 10–5, under the conditions used. The formation of
two complexes, NiCl+ and NiCl2, was detected. The authors also reported equilibrium
constants for infinite dilution by means of the SIT by using estimated interaction coefficients, ε(NiCl+, ClO −4 ) = 0.202 and ε(NiCl2(aq), NaClO4) = – 0.10. In the present review a considerably different value for ε(NiCl+, ClO −4 ) has been derived ((0.47 ± 0.06)
kg·mol−1); thus we calculated a markedly higher value of log10 b1ο from the data
[89IUL/POR] than the authors themselves. The reported value of K2 is higher than K1,
which is rather unlikely considering the low affinity of nickel(II) for chloride. Therefore, only the value of K1 is considered in this review.
[89ZIE/JON]
Ziemniak, Jones and Combs reported solubility measurements of NiO(cr) in aqueous
sodium phosphate solutions (pH values between 10.2 and 12.2) for temperatures from
290 K to 560 K. The experiments appear to have been done carefully, and it might be
expected that the measured solubilities should be suitable for derivation of chemical
thermodynamic parameters ( ∆ r Gmο , ∆ r Smο and ∆ r C pο,m ) for the dissolution reactions.
The authors analysed the data in terms of formation of nickel(II) hydroxo and mixed
phosphato-hydroxo complexes. Only mononuclear species of nickel were considered
due to the low nickel(II) concentrations.
A. Discussion of selected references
415
The authors concluded that a layer of the more stable hydrous nickel oxide
(Ni(OH)2) existed on the NiO surface below 468 K, and only at higher temperatures did
the NiO phase control the solubility behaviour of nickel(II). The data analyses provided
by the authors contained a number of inconsistencies [95LEM]. In the present evaluation, because the solid phase was poorly defined in the study of Ziemniak et al., any
ο
=
value used must be assigned a rather large uncertainty. The values log10 *b 2,1
* ο
− (19.77 ± 1.00) and log10 b 3,1 = – (30.86 ± 1.00) can be derived based on these data
and the analyses in [95LEM]. However, these values are not accepted in the present
review.
Ziemniak et al. proposed two aqueous nickel phosphate species,
Ni(OH)3 H 2 PO 42 − and Ni(OH) 2 HPO 24 − . These differ only by a (dissociated) water molecule in the inner sphere. The concentration of Ni(OH)3 H 2 PO 42 − was reported to be
small compared to the concentration of Ni(OH) 2 HPO 24 − for all solutions studied at
298 K, and the reverse was reported for solutions at 560 K. If the Gibbs energies of reaction proposed by the authors are used, the concentrations of these nickel-phosphate
species are approximately equal at 450 K. Even if both species exist, quantitative spectroscopic data for the relative concentrations of Ni(OH)3 H 2 PO 42 − and Ni(OH) 2 HPO 24 −
in a solution would be required to determine formation constants for the two species.
Lacking such information, the solubility data cannot be used to calculate independent
concentrations for the two species. From the point of view of the available variables
(total nickel concentration, phosphate concentration, sodium to phosphate ratio and pH)
the species Ni(OH)3 H 2 PO 42 − and Ni(OH) 2 HPO 24 − are mathematically redundant.
A re-analysis of the data was done by Lemire [95LEM]. He concluded that the
total equilibrium solution concentrations of nickel species over NiO(cr) (or β-Ni(OH)2
at low temperature) are higher in solutions containing substantial quantities of phosphate than they are in dilute aqueous solutions, and that this can be accounted for by
assuming formation of complexes of the form Ni(OH) 2 HPO 24 − , as proposed by Ziemniak et al.. However, in their experiments, the ionic strength increased (from 0.002 m to
0.6 m) in proportion to the added phosphate concentration and, from the available data,
it is not possible to separate the effects of the ionic medium and complexation. Ionic
strength effects seem to be less important than experimental uncertainties and model
errors. Less hydrolysed phosphato complexes (e.g., NiHPO4(aq)) do not appear to be
important under the conditions of the experiments. The re-analysis suggested that for
models that assume the stable solid at 298.15 K is β-Ni(OH)2, and that the solution species include only hydrolysis species and one nickel phosphate complex,
Ni(OH) 2 HPO 24 − , formed according to
β-Ni(OH)2(cr) + HPO 24 − U Ni(OH) 2 HPO 24 − ,
(A.83)
–1
the Gibbs energy of complex formation is approximately (27 ± 4) kJ·mol , and
log10 K (A.83) = – (5 ± 1).
A. Discussion of selected references
416
[90BJE]
This is very similar experimental work to that of [88BJE]. In this paper the formation of
bromo complexes of nickel(II) was studied in 3.5 – 11.1 M LiBr solution. The “semithermodynamic” formation constant of NiBr+ was determined, and that of the complex
NiBr2(aq) was estimated (see Table V-13).
[90BJE2]
The author demonstrated that from an aqueous solution that is dilute in nickel(II) ions,
but concentrated in NaSCN, the tetrahedral Ni(SCN) 24 − complex can be extracted with
hexanone. This shows that the octahedral [Ni(SCN)4(H2O)]2– complex, which forms in
presence of thiocyanate ion in very high excess, must be in equilibrium with small concentrations of the tetrahedral species.
[90EFI/EVD]
Efimov and Evdokimova determined the enthalpy of formation of NiI2(cr) at 298.15 K
by two related thermodynamic cycles involving dissolution of a series of compounds in
a mixture with the composition 1.00 KBr·0.40 Br2·0.12HBr·50.87H2O. In one cycle the
enthalpy of Reaction (A.84) was calculated in a manner similar to the enthalpy from the
cycle devised in [88EFI/EVD]:
Br2(l) + Ni(cr) + 2 KI(cr) U NiI2(cr) + 2 KBr(cr).
(A.84)
The authors made a correction of (0.462 ± 0.001) kJ·mol–1 to allow for the
amount of Br2 in the calorimeter sample bulb that was Br2(g) rather than Br2(l).
The other cycle was more direct:
I2(cr) + Ni(cr) U NiI2(cr).
(A.85)
Five calorimetric experiments were necessary to give ∆ r H mο (A.84) =
− (224.76 ± 0.38) kJ·mol–1, but only three values were required to determine
∆ r H mο (A.85) = – (96.42 ± 0.84) kJ·mol–1 (2σ uncertainties). The value given as the enthalpy of solution of KBr(cr) in the solvent mixture was (18.123 ± 0.039) kJ·mol–1,
whereas, in the earlier paper [88EFI/EVD] the value is reported as (18.145 ± 0.027)
kJ·mol–1. The value in [90EFI/EVD] is presumed here to be a deliberate (but minor)
correction, and ∆ r H mο (A.84) = – (224.76 ± 0.44) kJ·mol–1 is calculated,
∆ f H mο (NiI2, cr) = ∆ r H mο (A.85)
= – (96.42 ± 0.84) kJ·mol–1
(A.86)
∆ f H mο (NiI2, cr) = ∆ r H mο (A.84) – 2 ( ∆ f H mο (KBr, cr) – ∆ f H mο (KI, cr))
= – (96.40 ± 1.00) kJ·mol–1.
(A.87)
The two values are virtually identical. The calculations from the two cycles
both use exactly the same solution enthalpies for Ni(cr) and NiI2(cr), and the bulk of the
A. Discussion of selected references
417
uncertainty in either cycle can be attributed to the measurements of those quantities.
Differences in the ∆ f H mο (NiI2, cr) values result only from the values for the enthalpies
of formation of KBr(cr) and KI(cr) and the measured heats of solution of the elemental
halides and the potassium salts. Therefore, in this review, only the more precise of the
two values is accepted for further use:
∆ f H mο (NiI2, cr, 298.15 K) = – (96.42 ± 0.84) kJ·mol–1
(A.88)
This is virtually identical to the value given by the authors [90EFI/EVD] (if it
is assumed that their error limits correspond to one standard deviation).
[90EFI/FUR]
Efimov and Furkalyuk determined the enthalpy of Reactions:
NiCl2(cr) U Ni2+ + 2Cl–,
2+
(A.89)
–
NiBr2(cr) U Ni + 2Br ,
(A.90)
NiI2(cr) U Ni2+ + 2I–
(A.91)
calorimetrically by dissolving crystallised NiCl2, NiBr2 and NiI2 in 0.001 M HClO4 at
298.15 K. For the experiments with NiCl2 they used samples prepared as described in
[88EFI/EVD]. According to the analytical composition given there (w(Ni) =
(0.4529 ± 0.14) , w(Cl) = (0.5465 ± 0.10)), these samples contained less moisture than
those used in [58MUL] and [95MAN/KOR]. When Efimov and Furkalyuk’s experimental data were extrapolated to I = 0, like those listed in [58MUL] and [95MAN/KOR]
shown in Table A-23 and Table A-25, their results had to be corrected slightly. They
used perchloric acid but neglected its contribution to the ionic strength. The following
values were obtained: ∆ sol H mο (A.89) = – (84.89 ± 0.27) kJ·mol–1, which is about 1.5
kJ·mol–1 more negative than those found in [58MUL] and [95MAN/KOR],
∆ sol H mο (A.90) = – (86.18 ± 0.38) kJ·mol–1 and ∆ sol H mο (A.91) = – (73.38 ± 0.12)
kJ·mol−1.
Table A-23: Enthalpy of solution of NiCl2 in H2O at 298.15 K [90EFI/FUR].
m (NiCl2)
–1
∆ sol H mο (A.89)
–1
Φ
∆ sol H mο (A.89)
L
–1
(kg·mol )
(kJ·mol )
(kJ·mol )
(kJ·mol–1)
0.005495
– 84.169
0.6864
– 84.8554
0.005213
– 84.366
0.6718
– 85.0378
0.004584
– 84.233
0.6372
– 84.8702
0.004422
– 84.057
0.6278
– 84.6848
0.004410
– 84.354
0.6271
– 84.9811
A. Discussion of selected references
418
[90GRZ]
The stability of the monohalide complexes of the Mn2+–Zn2+ series was surveyed. The
order of constants both in water and organic solvents is distinctly different from the
stability order suggested by the Irving-Williams series. This behaviour seems to be
characteristic of complexes for which the stabilities are entropy controlled. The order of
stability constants has been correlated with the water-exchange rate and electrostriction
partial molar volumes of the hydrated cations.
[90JUR/DOM]
Using an adiabatic calorimeter, Juraitis et al. [90JUR/DOM] measured the heat capacity
of the solid between 80 K and 281 K, with emphasis on the effects of the structural transition near 220 K. The heat capacity values were recovered from the authors’ graph, and
polynomials were fitted to the C p ,m / T values to determine the entropy increments over
the temperature ranges 100 K to 200 K, 200 K to 220 K, 220 K to 240 K, and 240 K to
281 K. The results between 100 K and 200 K show considerable scatter, and it is not
clear from these results (only reported graphically) whether the values from 250 K to
280 K can be extrapolated smoothly to 298.15 K The experimental values from 250 K
to 273 K increase monotonically from 174 kJ·mol–1 to 209 J·K–1·mol–1, but the reported
value for 281 K is 207 J·K–1·mol–1. The highest-temperature calorimetry measurement at
281 K, closest to room temperature, is likely to have the greatest uncertainty. From the
simple quadratic equation fitted to the unweighted C p ,m / T values for temperatures
from 240 K to 281 K, C pο,m (NiCl2·2H2O, cr, 298.15 K) = 222.3 J·K–1·mol–1. In the present review, it is assumed that the C p ,m (T ) function has no anomalous behaviour above
240 K, and that differences from a monotonic curve are the result of experimental scatter. Based on that premise, C pο,m (NiCl2·2H2O, 298.15 K) = (230 ± 15) J·K–1·mol–1 is
selected in the present review, where the uncertainty is an estimate. These uncertainties
in the C pο,m values above 240 K introduce an uncertainty of ± 3 J·K–1·mol–1 in the contribution to the calculated value of S mο (298.15 K).
[90LEE/NAS]
The assessed Ni–Te phase diagram of the Handbook “Binary Alloy Phase Diagrams” is
essentially the same phase diagram as that published by Klepp and Komarek
[72KLE/KOM].
[90OSA/ROS]
The phase equilibria between β1-Ni3S2, β2-Ni4S3, and NiS(α) were studied by measuring
the equilibrium partial pressure of oxygen as a function of temperature.
The following galvanic cells were employed:
Pt, Au, NiS(α), β2-Ni4S3, NiO | CSZ | Fe, FeO, Pt
Pt, Au, β2-Ni4S3, β1-Ni3S2, NiO | CSZ | Fe, FeO, Pt
A. Discussion of selected references
419
Pt, Au, SO2 (0.010 atm), NiS(α), β2-Ni4S3 | CSZ | O2 (0.21 atm), Pt
where calcia stabilised zirconia (CSZ) served as the solid electrolyte. Figure A-35
shows the temperature dependence of the partial pressure of oxygen derived from these
measurements. The calculated lines, using the present thermodynamic model, agree
remarkably well with the experimental data which provides further evidence for the
reliability of the thermodynamic data recommended in this assessment. Furthermore, it
should be mentioned that the solid sulphides were prepared by solid state reaction from
the elements and NiO was obtained by thermal decomposition of NiCO3. All solid
phases were identified by X-ray powder diffraction analysis.
Figure A-35: Plot of logarithm of the oxygen partial pressure versus temperature for the
phase equilibria NiS(α) / β2-Ni4S3 and β2-Ni4S3 / NiO, respectively, at pSO = 1 atm. ▲
2
equilibrium between O2(g), NiO, NiS(α) and β2-Ni4S3; ● equilibrium between O2(g),
NiO, β2-Ni4S3 and β1-Ni3S2. Dashed line: thermodynamic model according to
[80SHA/CHA]; solid lines: thermodynamic model of the present assessment.
-11
-12
log10 [p(O2) / bar]
-13
-14
-15
-16
-17
-18
-19
-20
800
850
900
950
1000
1050
T/K
[91KNA/KUB]
This is the second updated and computerised edition of [73BAR/KNA], [77BAR/KNA].
A. Discussion of selected references
420
[91SPI/TRI]
The hydrolysis of aqueous nickel(II) chloride solutions ([NiCl2]tot = 0.03 to 1.5 M) has
been studied by potentiometric titrations at 25°C. The ionic strength was not controlled;
instead, the activity coefficients, based on literature values, were taken into account.
The formation of a single hydroxo complex was assumed. The evaluation of experimental data indicated the presence of the Ni 4 (OH) 44 + tetrameric species. A value of
ο
log10 *b 4,4
= − (27.0 ± 0.1) was reported for its thermodynamic hydrolysis constant. The
formation of chloro complexes was not taken into account, although these may only
have made a small contribution to the total solution concentration of nickel at [NiCl2]tot
≤ 0.1 M. Therefore, the reported value is accepted for use in this review, but with an
ο
increased uncertainty: log10 *b 4,4
= – (27.0 ± 0.3).
[91STO/GRO]
The heat capacity function of Ni3S2 was measured by adiabatic-shield calorimetry in the
temperature range from 5 to 350 K. High temperature measurements were carried out
up to a temperature of 1000 K. The solid phase heazlewoodite, Ni3S2, was synthesised
directly from the elements and characterised by X-ray diffraction analysis. The integration of the low temperature results yields the standard entropy of Ni3S2 at 298.15 K,
S mο = 133.2 J·K–1·mol–1. The high-temperature data are depicted in Figure A-36, where
the results of Ferrante and Gokcen [82FER/GOK] are likewise included. The heat capacity function:
2
[C pο,m ]835K
298.15K (Ni3S2, cr) = (165.79352 – 0.10316 (T/K) + 0.00012(T/K)
–2
–1
–1
– 2444624(T/K) ) J·K ·mol
(A.92)
valid from 298.15 to 835 K, was fitted to both sets of data mentioned above.
The steep increase of the C p ,m values of [91STO/GRO] at approximately
800 K can be attributed to the onset of the phase transition into the high-temperature
modification of Ni3S2 which was observed to occur at 834 K.
A. Discussion of selected references
421
Figure A-36: Heat capacity data for Ni3S2 as a function of temperature. ○
[91STO/GRO], ▲ [82FER/GOK]. Solid line: optimised heat capacity function (A.92).
200
C°p ,m / J·K–1·mol–1
180
160
140
120
100
200
300
400
500
600
700
800
900
T/K
[91SWI/PAW]
This polarographic study reports log10 K = 3.26, 25°C, 0.2 M NaClO4 for an ill-defined
Ni2+ complex with “phosphate” at pH 6.0. The concentration of nickel ion in solution
was 5 × 10–5 M and the ligand concentration was varied between 5 × 10–5 and 2.5 × 10–3
M. Figure 3 of the paper shows the effect of changes in “ PO34− ” on the polarograms (in
0.1 M NaClO4(aq)), but the paper does not state the chemical form of the added phosphate nor provide information as to the stoichiometry of the complex. The information
provided is not sufficient to recalculate a meaningful complexation constant, and the
results of the study are not used in the present review.
[91TAR/FAZ]
Tareen et al. determined T, p data for the decomposition equilibrium of NiCO3 to NiO,
NiCO3(s) U NiO(s) + CO2(g),
(A.93)
in the pressure range 300 – 2000 bar and at temperatures up to 700°C. The results of
this investigation relevant for the calculation of ∆ f H mο (NiCO3, 298.15 K) are listed in
Table A-24. The re-evaluation was based on the following set of equations:
A. Discussion of selected references
422
∆ f H mο (NiCO3, T) = ∆ f H mο (NiO, T) + ∆ f H mο (CO2, T) – T× ∆ r Smο ((A.93), T)
+ (Vm(NiO, 298.15 K) – Vm(NiCO3, 298.15 K))×(p – 1)
+ RT ln f(CO2) + RT ln y(CO2)
(A.94)
RT ln f(CO2) = RT ln p + B(CO2)p
 298.15 
 298.15 
B(CO2) (cm3·mol–1) = – 127 – 288× 
− 1 – 118× 
− 1
T


 T

∆ f H mο (NiCO3, 298.15 K) = ∆ f H mο (NiCO3, T) –
∫
T
298.15
2
(A.95)
(A.96)
∆ f C p ,m (NiCO3) dT (A.97)
At first Equations (A.94), (A.95) and (A.96) have been used to calculate
∆ f H mο (NiCO3, T), where Equation (A.96) was taken from [96KEH]. Then
∆ f H mο (NiCO3, T) was extrapolated to 298.15 K with Equation (A.97), see Table A-24.
As already noted by Tareen et al. the ∆ f H mο values obtained were slightly more negative at higher pressures than at lower ones [91TAR/FAZ]. This may either be due to
experimental uncertainties or to incorrect C p ,m and ln f(CO2) functions or both. When
the temperature and pressure dependence of the solid phase volumes, as in Equation
(A.94), is ignored, this certainly plays a minor role. By averaging the individual results
the error limit of the standard enthalpy has been estimated to be 2σ = ± 6.0 kJ·mol–1,
see Section V.7.1.1.1.1.3.1 and Table A-24 respectively.
Table A-24: ∆ f H mο (NiCO3, 298.15 K) from decomposition data [91TAR/FAZ].
isobars
T low
isobars
T high
T
p
(K)
(bar)
593.15
1020
623.15
1400
643.15
668.15
673.15
y(CO2)
∆ f H mο (NiCO3, T) ∆ f H mο (NiCO3, 298.15 K)
(kJ·mol–1)
(kJ·mol–1)
0.43
– 675.51
– 704.73
0.50
– 673.92
– 706.42
1550
0.46
– 673.84
– 708.54
1900
0.49
– 672.60
– 710.08
2070
0.45
– 672.81
– 710.85
608.15
1020
0.50
– 674.75
– 705.60
648.15
1400
0.43
– 674.33
– 709.59
653.15
1550
0.46
– 673.66
– 709.47
688.15
1900
0.42
– 673.01
– 712.74
708.15
2070
0.45
– 672.01
– 714.01
– (709.2 ± 6.0)
A. Discussion of selected references
423
[92BAL/DIC]
In this report1 the crystal structure, the phase transitions, the heat capacity function, the
entropy and the standard enthalpy of formation of the solid phases in the Ni–Te system
are comprehensively discussed and critically assessed. Thermodynamic functions in the
range of 298.15 to 1100 K were listed in intervals of 100 K for the following Ni–Te
phases with selected stoichiometries Ni0.333Te0.667(cr), Ni0.4Te0.6(cr), Ni0.476Te0.524(cr),
Ni0.54Te0.46(cr), Ni0.563Te0.437(cr), Ni0.6Te0.4(cr), NiTe2(g).
[92RAR2]
Isopiestic vapour-pressure experiments were performed for aqueous solutions of highpurity NiCl2 at 25°C. Values of the water activity, aW, the osmotic coefficient, φ, and
the mean activity coefficient, γ±, were recommended for this system. The solubility of
NiSO4·6H2O(cr) was determined definitively. Older solubility mesurements
[37PEA/ECK], [79OIK], [79OJK/MAK], [79PET/SHE], [83KIM/IMA], [85FIL/CHA],
[86FIL/CHA], [87RAR3] were reviewed and the reason for the discrepancies were discussed. Standard Gibbs energy of reaction,
NiCl2·6H2O(cr) U Ni2+ + 2Cl– + 6H2O(l),
(A.98)
∆ sol Gmο (A.98) = – (17.38 ± 0.04) kJ·mol–1 was based on the statistically weighted average of Rard’s as well as Pearce and Eckstrom’s [37PEA/ECK] solubility of
NiSO4·6H2O(cr).
[92VID/KOR]
The standard enthalpy of formation of heazlewoodite, Ni3S2, was measured directly by
reaction of elemental nickel with sulphur in a calorimeter. The reaction was initiated by
the energy of a light pulse. The heat of reaction at 298.15 K was determined to be
∆ f H mο = – (226.0 ± 2.2) kJ·mol–1. This value deviates by about 10 kJ·mol–1 from results
obtained by equilibrium experiments [54ROS] as well as drop calorimetry
[86CEM/KLE]. The products formed during the calorimetric process were shown to be
Ni3S2 and metallic nickel by X-ray diffraction analysis. Thus, a certain amount of sulphur evaporated during the reaction. It may be difficult to properly correct for the enthalpy associated with this process, and thus the results may be erroneous.
[93BAL/DIV]
Balej and Divisek re-evaluated qualitatively equilibrium or rest potentials of solid nickel
oxide hydroxides with 2 < z(Ni) ≤ 3.6 measured at ambient temperatures. The
experimental data were converted to standard conditions i.e., 25°C, aKOH = aH2 O = 1.0, E
versus SHE. From the functions E = f (z(Ni)), thus obtained, the authors concluded that
the so-called “β-NiOOH | β-Ni(OH)2” systems with z(Ni) < 3 are thermodynamically
1
In this report, Fig. 16 is erroneously titled ‘Phase diagram of the Ni–Te system’, it should read ‘⋅⋅⋅ the
Fe–Te system’ instead.
424
A. Discussion of selected references
unstable or metastable. In the whole range investigated (z(Ni) > 3.5) the so-called
“γ−NiOOH | α−Ni(OH)2” systems exhibit very flat S-shaped functions E = f (z(Ni))
with the lowest E values at the given z(Ni) and were regarded as thermodynamically
stable solid solutions of Ni(II) + Ni(IV) components only.
[93BAL/DIV2]
Balej and Divisek based a quantitative re-evaluation of critically selected data on the
equilibrium or rest potentials of the so-called “γ-NiOOH | α-Ni(OH)2” systems on the
assumption that these systems consist of homogeneous solid solutions with the endmembers Ni(OH)2 and NiO2·xH2O. Whereas the Ni(II) component is considered to be
completely undissociated, the Ni(IV) component fully deprotonates to give H+ and
NiO2·(x–1)H2O·OH–. While these authors convincingly describe the redox behaviour of
Ni(II, IV) oxide hydroxides, the stoichiometric and structural characterisation of the
hypothetical end-members is still lacking. So the thermodynamic quantities derived
from this model were not accepted for the present review.
[93KUB/ALC]
In Chapter 3 of this textbook the estimation of thermodynamic data is described for
compounds for which no experimental data are available. For the heat capacity of predominantly ionic, solid compounds at 298 K cationic and anionic group contributions
can be added. Moreover a set of equations is given to estimate the constants in the equa–3
tion ( [C pο,m ]700K
(T/K) – 12.34×105 (T/K)–2)
298K (NiCO3, cr) = (88.701 + 38.91×10
−1
−1
J·K ·mol ). The variables are the sum of the cationic and anionic contributions, the
melting temperature of the compound, and the number of atoms in the molecule.
[93MAK/VOE]
Based on thermodynamic quantities, derived from the experimental investigation of the
solubility of nickel sulphide performed by Thiel and Gessner [14THI/GES], solubility
constants as well as Pourbaix diagrams (redox potential plotted vs. pH) have been calculated.
[93MUL/ISA]
Muldagalieva et al. [93MUL/ISA] reported the equation:
–3
[C pο,m ]700K
(T/K) + 0.018×105 (T/K)–2 J·K–1·mol–1
298.15K (Ni11As8, cr) = 380.12 + 367.27×10
for 298.15 K to 700 K based on dynamic calorimetric measurements ( C pο,m (Ni11As8, cr,
298.15 K) = 490 J·K–1·mol–1). The experimental data were not provided, nor the goodness of fit of the equation, though the authors claimed that the experimental uncertainties were “no greater than 8%”, and this uncertainty is accepted in the present review to
be a 2σ value. In the temperature range of the equation, it would seem that the T –2 term
in the equation makes a negligible contribution to the value, suggesting that the fitting
was not done properly. The authors also reported an estimated standard entropy at
A. Discussion of selected references
425
298.15 K of 580.0 kJ·mol–1 for the same compound (that estimate, by Kumok, was
made in a reference unavailable to the present reviewers). The heat capacity values from
this study near 700 K are in rough agreement with those estimated by Skeaff et al.
[85SKE/MAI], though the variation with temperature is markedly different. For
298.15 K, the values from the two equations for C pο,m (Ni11As8) differ by 62 J·K–1·mol–1,
which is slightly greater than the experimental uncertainties.
[93NEI/POW]
Thermodynamic data of high accuracy and reliability for the Ni – NiO system at high
temperatures were obtained from electrochemical cells with yttria- or calcia-stabilised
zirconia electrolytes. The experimental results are in excellent agreement with available
calorimetric data as well as emf measurements of previous publications. The recommended Gibbs energy equation for the reaction:
2 Ni + O2 U 2 NiO
is ∆ f Gm (NiO) = (– 478.967 + 0.248514 (T/K) – 0.0097961 (T/K) ln(T/K)) kJ·mol–1,
valid from 700 to 1700 K. The uncertainty in ∆ f Gm was reported to be ± 0.200
kJ·mol−1.
[93NIC/ROB]
Widgiemoolthalite was chemically and crystallographically investigated. The diffuse Xray diffraction pattern suggests a high degree of structural disorder, and the pseudoorthorhombic cell indicates the possibility of twinning. Thus a definitive crystallographic characterisation was prevented, but the close similarity of its X-ray diffraction
pattern to that of hydromagnesite indicates that it is the nickel analogue of the latter.
[93OBE/KAS]
Obendrauf et al. have measured the thermophysical properties of nickel by a fast pulse
heating technique using a newly developed highly sensitive pyrometer. The values obtained for ∆ fus H mο (Ni) are slightly greater than, and the values for C pο,m (Ni, l) are significantly greater than the calorimetrically determined values on which the quantities in
this review are based.
[93PRZ/WIS]
The value 15.644 kJ·mol–1 was reported by Przepiera et al. [93PRZ/WIS] as the heat of
solution of nickel sulphate heptahydrate in water. Although Archer [99ARC] assumed
that this was an infinite dilution value, that information is not stated explicitly in the
original paper, and it might also be concluded that the value represents the heat of solution to produce a solution with a 5×10–4 mass ratio of salt to water. If the value is assumed to be for the enthalpy of solution to this stated dilution (approximately 0.002 m),
the application of a heat of dilution correction [66GOL/RID] to I = 0 gives a value (12.5
kJ·mol–1) that is in good agreement with the value based on results of Goldberg et al.
426
A. Discussion of selected references
[66GOL/RID] and Stout et al. [66STO/ARC] (also see [98PLY/ZHA]). As this ambiguity cannot be resolved, the value from this paper [93PRZ/WIS] is not used in determining thermodynamic quantities selected in the present review.
[93SHE]
The paper provides the results of a conductance study. The results were analysed using
the equation of d’Aprano [71APR]. No values are reported for the nickel sulphate system other than the limiting “equivalent conductance at infinite dilution” and the association constant. These values are not inconsistent with values from other conductance
studies, but reanalysis of the data to ensure consistency with the SIT model is not possible.
[93UVA/TIM]
The compositions of solid nickel chloride hydrates were determined at different water
vapour pressures. The water vapour pressures were set using an isopiestic arrangement
with aqueous H2SO4 solutions. At 25 and 35°C, the transition from the dihydrate to the
tetrahydrate was found to occur at 0.998 kPa and 1.70 kPa, respectively. At 25 and
35°C, the transition from the tetrahydrate to the hexahydrate was found to occur at 1.19
kPa and 2.61 kPa, respectively. The uncertainty in the reported transition water vapour
pressures is estimated in the present review as ± 0.05 kPa.
[94STE/FOT]
De Stefano et al. [94STE/FOT] measured the protonation constants for pyrophosphate,
as a function of temperature between 5 and 45°C, in aqueous solutions of (CH3)4NCl,
NaCl and KCl (0.00 to 0.75 M). These constants agree well (within 0.11 in log10 K )
with the values reported in the studies of Perlmutter-Hayman and Secco [73PER/SEC]
(KNO3(aq)) and Hammes and Morrell [64HAM/MOR] ((CH3)4NCl(aq)) in their studies
of nickel pyrophosphate complexation, and of Edwards et al. [73EDW/FAR]
((CH3)4NCl(aq)). From the apparent protonation constants, De Stefano et al.
[94STE/FOT] calculated values of the formation constants of KP2 O37− and KHP2 O72 −
(and of the formation constants for the corresponding sodium-ion species). The pyrophosphate protonation constant values at I = 0, differ by less than 0.2 in log10 K from
those selected by Grenthe et al. [92GRE/FUG].
[94STO/FJE]
The phase relations of the system Ni – S in the compositional range from 45 to 48 at% S
were studied by application of high temperature X-ray diffraction analysis, differential
thermal analysis (DTA), and adiabatic calorimetry. The solid compounds were prepared
by solid state synthesis from the elements. The structure of the orthorhombic high temperature modification of Ni7S6 was resolved by high temperature X-ray diffraction
analysis. This disordered high temperature phase transforms into metastable phases on
quenching [96SEI/FJE]. In contrast to the results of Kullerud and Yund [62KUL/YUN]
A. Discussion of selected references
427
the composition of the orthorhombic low temperature modification was found to be
Ni9S8 instead of Ni7S6. The stable non-stoichiometric high temperature phase is formed
eutectoidally from Ni3S2 and Ni9S8 at (675 ± 3) K and the decomposition into Ni3+xS2
and Ni1–xS occurs at (850 ± 2) K. Based on high temperature X-ray diffraction and
DTA, the decomposition of Ni9S8 into Ni7S6 and Ni1–xS was observed at (709 ± 5) K,
contradicting the findings of [62KUL/YUN] who obtained 673 K for the peritectoidal
decomposition of the low temperature phase of Ni7S6.
The heat capacity of a sample with a nominal composition of Ni7S6 was
measured by means of adiabatic calorimetry from 5 to 970 K. Below 675 K and above
850 K the specimen consisted of a mixture of two phases, viz. Ni9S8 + Ni3S2 and Ni3+xS2
+ Ni1–xS, respectively. From the heat capacity function for Ni3S2, determined earlier
[91STO/GRO], the heat capacity of Ni9S8 can be calculated for the temperature range
from 298.15 to 640 K:
2
[C pο,m ]640K
298.15K (Ni9S8, cr) = (1079.2452 – 2.05994·(T/K) + 0.00198·(T/K)
(A.99)
– 20839639·(T/K)–2 ) J·K–1·mol–1.
The exothermic and endothermic enthalpy effects occurring at 415 and 515 K,
respectively, are attributed to metastable phases of Ni7S6 which are formed during the
preparation of the samples. These enthalpy effects disappeared after annealing the samples at 810 K for two days.
Figure A-37: Heat capacity of “Ni7S6” consisting of a mixture of Ni3S2 and Ni9S8 as a
function of temperature. The solid line is derived from the heat capacity functions for
Ni3S2 and Ni9S8 recommended in this compilation. The various symbols refer to samples subjected to different heat treatment.
420
2/3 Ni9S8 + 1/3 Ni3S2
C°p ,m / J·K–1·mol–1
400
380
360
340
320
300
280
250
300
350
400
450
500
T/K
550
600
650
700
A. Discussion of selected references
428
[94ZHA/MIL]
The stability constant of the nickel sulphide complex NiHS+ in sea water (I = 0.7
mol·kg–1 NaCl) is determined at 298.15 K by applying cathodic stripping square wave
voltammetry. Typically, a nickel(II) standard solution was added to sea water, leading
to molalities of total nickel(II) ranging between 500 and 1500 nmol·kg–1. These solutions were titrated with Na2S solutions, where successive standard additions of sulphide
were in the range from 250 to 500 nmol·kg–1. The measurement of the stripping current
of free sulphide, deposited on the hanging mercury drop electrode, yielded the molality
of HS–. A proper evaluation of the experimental titration curves resulted in log10 b1 =
(5.3 ± 0.1) for the formation constant of NiHS+ in sea water at 298.15 K.
[95GRO/STO]
The heat capacity of synthetic NiS was measured by adiabatic-shield calorimetry from
260 to 1000 K. The solid phase was prepared from the elements by solid state reaction
and checked by X-ray diffraction analysis. The rhombohedral low temperature modification (millerite) was found to transform to the hexagonal non-stoichiometric NiAs
polytype at 660 K where the heat of transition amounts to ∆ trs H mο = (6591 ± 50) J·mol–1.
The high temperature form can easily be quenched to room temperature. Thus, the heat
capacity function for the stoichiometric composition of this phase can be derived from
298.15 to 900 K:
[C pο,m ]660K
298.15K (NiS, α) = (46.676 + 0.0199807 (T/K)
– 255499.4 (T/K)–2) J·K–1·mol–1
(A.100)
[C pο,m ]900K
660K (NiS, α) = (50.07833 + 0.0149 (T/K)
– 336916.92 (T/K)–2) J·K–1·mol–1.
(A.101)
In the case of millerite the heat capacity function is obtained for temperatures
up to 660 K:
–5
[C pο,m ]660K
(T/K)2
298K (NiS, β) = (54.008 – 0.01437 (T/K) + 3.2282 × 10
–2
–1
–1
(A.102)
– 490363 (T/K) ) J·K ·mol .
Figure A-38 shows the temperature dependence of the heat capacity of NiS,
where the dashed line corresponds to the metastable NiAs-type phase below 660 K. The
steep increase of C p ,m at temperatures above 900 K can be attributed to the formation of
small traces of Ni3+xS2.
A. Discussion of selected references
429
Figure A-38: Heat capacity of NiS as a function of temperature. ○, solid lines: stable
NiS phases. U, dashed line: metastable NiAs-type modification below 660 K.
80
C°p ,m / J·K–1·mol–1
70
60
50
40
200
300
400
500
600
700
800
900
1000
T/K
[95MAN/KOR]
Manin and Korolev reported having measured the enthalpy of solution of NiCl2(cr) into
water and into aqueous solutions of other electrolytes. The measurements were performed with an isoperibol calorimeter with a claimed accuracy of ± 0.6%. Their results
for the enthalpy of solution into H2O are listed in Table A-25. The mean value of
∆ sol H mο for the reaction:
NiCl2(cr) U Ni2+ + 2 Cl–
(A.103)
agrees nicely with the one derived from Muldrow’s data. According to the Cl content
(54.0%) determined by argentometric titration, the sample of NiCl2 used was not completely anhydrous (n(H2O) / n(NiCl2) ≈ 0.09). This may lead to systematic errors similar
to those discussed in the Appendix A entry for [58MUL].
A. Discussion of selected references
430
Table A-25: Enthalpy of solution of NiCl2 in H2O at 298.15 K [95MAN/KOR].
m (NiCl2)
–1
∆ sol H m (A.103)
–1
Φ
∆ sol H mο (A.103)
L
–1
(kg·mol )
(kJ·mol )
(kJ·mol )
(kJ·mol–1)
0.004230
– 82.67
0.5961
– 83.2661
0.004570
– 82.46
0.6169
– 83.0769
0.006320
– 82.92
0.7111
– 83.6311
0.009250
– 83.08
0.8371
– 83.9171
0.009260
– 82.83
0.8375
– 83.6675
0.009630
– 82.87
0.8515
– 83.7215
0.012410
– 82.10
0.9465
– 83.0465
0.014380
– 82.63
1.0054
– 83.6354
0.027340
– 82.10
1.2957
– 83.3957
0.030460
– 82.14
1.3499
– 83.4899
The variation of the enthalpy of solution of the anhydrous solid should parallel
the enthalpy of solution of the hydrated solid as a function of salt molality. The experimental values for dissolution in water drift to slightly more negative values at lower
NiCl2(aq) concentrations, but the measurements themselves have fairly large uncertainties of approximately ± 0.5 kJ·mol–1. The solution concentrations are not sufficiently
low to provide a good extrapolation to I = 0, but values based on a theoretical extrapolation [97GRE/PLY2] are provided in Table A-25. The authors’ Figure 1 compares the
values for the heats of solution in water to those taken from two compilations. However,
there is no evidence as to how the values from the compilations were based on experimental heats of dilution. (There is also a typographical error in the key to the points in
the authors’ Figure 1. The points designated “3” appear to be the data from
[95MAN/KOR]).
The heats of dissolution of anhydrous nickel chloride in aqueous solutions containing several different concentrations of HCl(aq) also were determined. For measurements with final solutions having nickel chloride molalities of ca. 0.007 m, the enthalpy
of solution of anhydrous nickel chloride for each mol% of HCl is about 3 kJ·mol–1 more
positive than in water.
[95MUL/CHU]
This paper is similar to the earlier paper on Ni11As8(cr) [93MUL/ISA], but contains
more details. The heat capacity equation, from dynamic calorimetric measurements, is
reported as:
–3
5
–2
–1
–1
[C pο,m ]675K
298K (NiAs, cr) = (36.54 + 41.87×10 (T/K) + 1.58×10 (T/K) ) J·K ·mol .
A. Discussion of selected references
431
[96BRU/PIA]
The authors reported an extensive set of torsion balance and Knudsen cell measurements for the determination of the enthalpy of sublimation of NiF2 between 950 and
1250 K. Third law and second law values are in much better agreement if the values of
(Gm (T ) − H mο (298.15 K)) / T for the sublimation reaction for CoF2(cr) are substituted
for those in the literature for NiF2(cr). The work appears to have been carefully carried
out, and if the interpretation were correct, this would indicate that (a) S mο (Ni, cr,
298.15 K) is 8 – 11 J·K–1·mol–1 less than based on [55CAT/STO], or (b) C p ,m (T) from
[70BIN/HEB] is incorrect (although the values used for CoF2 are from work by the
same group [67BIN/STR]) or (c) the value of ∆ f H mο (NiF2, cr, 298.15 K) based on
Rudzitis et al. [67RUD/DEV] is too negative by about 8 kJ·mol–1.
[96KEH]
The table presented gives second virial coefficients of about 110 inorganic and organic
gases as a function of temperature. Selected data from the literature have been fitted by
least squares to the equation:
B (cm3·mol–1) =
n
∑ a(i) [(T
i =1
0
/ T ) − 1]
i −1
where T0 = 298.15 K. For the present purpose, use has been made of the a(i) values of
CO2.
[96LUT/RIC]
The stability constants for the aqueous species NiHS+, Ni2HS3+ and Ni3HS5+ at 298.15 K
in seawater as well as sodium chloride solutions (artificial seawater) have been determined by application of square wave voltammetry. Solutions of sodium sulphide (1 – 10
µmol·kg–1) in diluted seawater of constant ionic strength ranging from 0.07 to 0.7
mol·kg–1 NaCl were titrated with nickel (II) solutions, where care was taken to remove
traces of oxygen from all solutions by purging with high purity argon. The molality of
free sulphide was determined from both the peak potential and the peak current at constant pH for the electrochemical reaction:
HS– + Hg U HgS + H+ + 2e–
occurring at the hanging mercury drop electrode. The titration curves were evaluated by
means of a standard method [51DEF/HUM], [77HEA/HEF], leading to stability constants for NiHS+, Ni2HS3+ and Ni3HS5+, respectively. Plots of the peak potential versus
pH indicate the existence of bisulfide (HS–) complexes at pH values above 7.
The ionic strength dependence of the nickel bisulphide stability constants is
depicted in Figure A-39. By applying the SIT concept [97GRE/PLY2] the respective
stability constants are extrapolated to zero ionic strength. As well as the stability constants, the optimisation procedure also yields the SIT interaction parameters ∆ε in chloride media, T = 298.15 K:
A. Discussion of selected references
432
Ni2+ + HS– U NiHS+
log10 b = (5.18 ± 0.20); ∆ε(A.104) = – (0.97 ± 0.39) kg·mol–1.
(A.104)
2Ni2+ + HS– U Ni2HS3+
log10 b = (9.92 ± 0.10); ∆ε(A.105) = – (0.05 ± 0.22) kg·mol–1.
(A.105)
ο
1
ο
2
3Ni2+ + HS– U Ni3HS5+
(A.106)
log10 b = (14.008 ± 0.099); ∆ε(A.106) = (0.59 ± 0.22) kg·mol–1.
ο
3
Figure A-39: Logarithm of stability constants of nickel bisulphide complexes in seawater (NaCl solutions) plus the Debye-Hückel term for ionic strength correction plotted
as a function of ionic strength, T = 298.15 K. The solid lines correspond to least squares
fits of the SIT model to the experimental data, ● [96LUT/RIC], ▲ [94ZHA/MIL].
(a) log10 b1ο = (5.18 ± 0.20);
∆ε(A.104) = – (0.97 ± 0.39) kg·mol–1.
(b) log10 b 2ο = (9.92 ± 0.10);
∆ε(A.105) = – (0.05 ± 0.22) kg·mol–1.
(c) log10 b 3ο = (14.008 ± 0.099); ∆ε(A.106) = (0.59 ± 0.22) kg·mol–.
6.4
6.2
10.20
(a)
10.15
(b)
10.10
5.8
10.05
log β2I + DH
5.6
5.4
5.2
10.00
9.95
5.0
9.90
4.8
9.85
4.6
9.80
0.0
0.2
0.4
0.6
0.8
0.0
1.0
0.2
0.4
-1
I / mol·kg
-1
I / mol·kg
14.4
14.2
(c)
14.0
log β3I + DH
log β1I + DH
6.0
13.8
13.6
13.4
13.2
13.0
0.0
0.2
0.4
0.6
-1
I / mol·kg
0.8
1.0
0.6
0.8
A. Discussion of selected references
433
[96POU/DRE]
Nickel hydroxide purchased from Johnson Matthey Chemical Co. was equilibrated with
0.002 to 0.200 mol·kg–1 NaNO3 over a pH range of approximately 5.5 – 6.5 at 22°C.
The experimental data were evaluated using the Davies equation in MINTEQA2 at
these ionic strengths [93ALL/BRO]. The solubilities obtained at I = 0.2 mol·kg–1 were
apparently lower than those at I = 0.002 – 0.02 mol·kg–1. The authors preferred the
higher values, because their model might become incorrect at I > 0.02 mol·kg–1. When
the activity data were recalculated assuming that Equation (A.107) is valid a mean value
of log10 K sο,0 = – (15.96 ± 0.10) was obtained at 295.15 K. log10 *K sο,0 = 12.24 recorded in
Table V-6 and Figure V-11 was based on this evaluation.
log10 aNi2+ = log10 K sο,0 − 2 log10 aOH−
(A.107)
[96SAH/SAH]
This paper reports log10 K1 = (2.20 ± 0.02) for the association constant of Ni2+ with
HPO 24 − at 25°C in 0.1 M NaNO3. The study appears to have been done carefully,
though several of the minor difficulties found for an earlier paper [67SIG/BEC] from
the same group still occur (see Appendix A for [67SIG/BEC]). Correction to zero ionic
strength (Appendix B) gives log10 K1ο = (3.07 ± 0.10), where the uncertainty has been
assigned in the present review. The small ∆ε correction has been based on the value for
ε(Na+, HPO 24 − ) from Table B-5 (rather than using the equation in Table B-6), and
ε(Ni2+, NO3− ) = 0.18 kg·mol–1.
[96SEI/FJE]
The relative stability and structural descriptions of four metastable phases with composition close to Ni7S6 were obtained from X-ray diffraction, transmission electron microscopy, calorimetry, and thermal analysis. The structures of these phases are closely
related to that of the disordered high-temperature phase Ni7S6 which is formed eutectoidally from Ni3S2 and Ni9S8 at 675 K [94STO/FJE]. When Ni7S6 is cooled to room
temperature, this phase decomposes into the stable compounds Ni3S2 and Ni9S8 or a
transformation of Ni7S6 into metastable modifications of similar composition occurs.
[97BAH/PAR]
This is a recent compilation of thermodynamic data of metal thiocyanate complexes.
The paper listed the majority of available thermodynamic data for the nickel(II) – thiocyanate system, but recommended values are given only for log10 b1 at I = 1.0
( log10 b1 = (1.14 ± 0.09)) and 1.5 M ( log10 b1 = (1.14 ± 0.02)).
[97BAL/DIV]
Revised E–pH and E–m(KOH) diagrams for nickel at 25°C are calculated and graphically represented using the thermodynamic data derived for hypothetical NiO2·xH2O
A. Discussion of selected references
434
[93BAL/DIV], [93BAL/DIV2]. This E–pH diagram is meant to supersede previous versions [63DEL/ZOU], [81SIL].
[97MAT/RAI]
Nickel hydroxide was precipitated by adding NaOH to a 0.2 M NiCl2 stock solution.
The final pH of the Ni(OH)2 suspension was approximately seven. The precipitate was
washed with water until it was free of chloride. The X-ray diffraction data of the freeze
dried precipitate matched with JCPDS data file 14-117 for the Ni(OH)2 mineral,
theophrastite. Approximately 300 mg of Ni(OH)2 were suspended in 30 cm3 of 0.01 M
NaClO4. The pH was adjusted between 7 and 12 by HClO4 and NaOH, respectively.
Thus, only a minute Ni(OH)2 portion, doubtlessly having consisted of the finest
particles, was actually dissolved. When equilibrium was approached from
oversaturation again finely dispersed Ni(OH)2 will have been precipitated. As
solubilities of metal oxides and hydroxides depend on the particle size [65SCH/ALT],
the numerical value of the solubility constant will probably be overestimated by this
method, see Figure V-11.
[97PAN/CAM]
This is probably the most comprehensive set of heat capacity results available for any
nickel salt in aqueous solution. Apparent molar heat capacities of aqueous Ni(ClO4)2
were measured calorimetrically from 25 to 85°C over a molality range of 0.02 to 0.80
mol·kg–1. Standard molar heat capacities of Ni2+ for the same temperature range were
obtained by using the additivity rule and data for HClO4(aq), given in literature. The
results for C pο,m (Ni2+) can be fitted with a conventional heat capacity model valid from
298.15 to 358.15 K using C pο,m (HClO4, aq, T) from Lemire et al. [96LEM/CAM]:
2+
7
2
–1
–1
[C pο,m ]358.15K
298.15K (Ni ) = (842.427 – 1.8166 (T/K) – 3.06653 × 10 /(T/K) ) J·K ·mol .
The values change only slowly with temperature in this range, with an apparent
shallow maximum value near 50°C. The authors extrapolated their results to 300°C using the Helgeson-Kirkham-Flowers equation [74HEL/KIR2], [81HEL/KIR],
[97PAN/CAM]. The mean value for the standard partial molar heat capacity of Ni2+ at
25°C in the authors’ Table III is – 44.0 kJ·mol–1 whereas the result in the authors’ Table
IV is C pο,m (Ni2+, 298.15 K) = – 49.6 J·K–1·mol–1. The difference reflects the sensitivity
of the C pο,m value to the use of two different, but reasonable, auxiliary values for the
partial molar heat capacity for HClO4(aq). As noted in Section V.2.1.4, results for other
nickel salt solutions lead to similar uncertainties in the value of C pο,m (Ni2+, 298.15 K).
[97RIC]
This monograph describes the synthesis, structure and reactivity of aqua ions. Only a
short paragraph deals with Ni3+, because this ion occurs only as a ‘transient’ species.
A. Discussion of selected references
435
[98GAM/KON]
A plot of log10 {[Ni 2+ ] × pCO } versus pH showed that the corresponding solubility con2
stant is almost independent of temperature at least in the range of 348.15 to 363.15 K.
Apparently other neutral transition metal carbonates behave similarly.
[98JAI/ELM]
The redox reaction of nickel hydroxide and nickel oxide hydroxide, the electrochemically active compounds at the positive electrode of a nickel battery, was investigated.
The thermodynamics of non-ideal solid solutions were applied to the reversible potential as a function of the state-of-discharge. In a temperature range 5 to 55°C two parameter activity coefficient models perform significantly better than one parameter
models.
[98PLY/ZHA]
Plyasunova et al. critically evaluated the standard thermodynamic quantities of Ni2+, its
hydrolysis reactions and hydroxo-complex formation on the basis of published experimental studies and the specific interaction theory (SIT) for activity coefficient modelling [97GRE/PLY2]. Recommended thermodynamic functions and interaction coefficients relevant for the present review and its compounds were presented, see Table A26.
Table A-26: Recommended formation constants ( *b nο, m ) and reaction enthalpies for the
hydroxo complexes formed in the acidic region, at 298.15 K, mNi2+ + nH2O(l) U
Ni m (OH) n2 m − n + nH+
Species
log10 *b nο,m
NiOH+
∆ r H mο (kJ·mol–1)
– (9.50 ± 0.36)
(50 ± 21)
Ni2OH3+
– (9.8 ± 1.2)
(43.3 ± 3.0)
Ni 4 (OH) 44+
– (27.9 ± 1.0)
(180 ± 2)
Table A-27: Recommended equilibrium constants ( *K sο, n, m ) and reaction enthalpies for
the hydroxo complexes formed in the alkaline region, at 298.15 K, mNi(OH)2(cr) +
nOH– U Ni m (OH) −2 mn + n
Species
log10 *K sο,n ,m
∆ r H mο (kJ·mol–1)
– (7.52 ± 0.80)
–
Ni(OH)3−
– (3.7 ± 1.8)
(7.1 ± 3.0)
Ni(OH) 24−
(6.43 ± 0.23)
– (14 ± 8)
Ni(OH)2(aq)
A. Discussion of selected references
436
Table A-28: Recommended ion interaction coefficients ε(j, k).
k
j
NiOH+
3+
Ni2OH
4+
4
Cl–
ClO −4
Na+
≥ 0.31
(0.30 ± 0.18)
–
(0.41 ± 0.56)
(0.59 ± 0.16)
–
(0.68 ± 0.38)
(0.90 ± 0.32)
–
Ni(OH)2(aq)
–
–
– (0.07 ± 0.03)
Ni(OH)3−
–
–
(0.18 ± 0.05)
–
–
(0.27 ± 0.06)
Ni 4 (OH)
Ni(OH)
2−
4
Table A-29: Standard thermodynamic functions of the aqueous species.
Species
Ni2+
+
NiOH
∆ f Gmο (kJ·mol–1)
∆ f H mο (kJ·mol–1)
S mο (J·K–1·mol–1)
– (45.5 ± 3.4)
– (54.1 ± 2.5)
– (130 ± 3)
– (228.4 ± 5.7)
– (290 ± 24)
– (74 ± 81)
Ni2OH
– (272 ± 14)
– (359 ± 20)
– (260 ± 86)
Ni4(OH) 44 +
– (971 ± 19)
– (1190 ± 27)
– (203 ± 90)
Ni(OH)2 (aq)
– (417 ± 11)
– (540 ± 13)
– (48 ± 32)
– (587 ± 15)
– (791 ± 18)
– (85 ± 47)
– (734.0 ± 8.6)
–
–
3+
Ni(OH)
−
3
Ni(OH)
2−
4
The analysis was useful in considering some of the references, but the values
from [98PLY/ZHA] are not used in the present review.
[98ROG/KOZ]
Exchange of sodium ions in β’’-alumina, performed in molten salts, made possible a
preparation of a series of divalent β’’-aluminas. In this way the solid electrolytes conducting zinc, cobalt, nickel, calcium, manganese and copper ions were prepared. Their
application to the solid-state galvanic cells have been presented. The Gibbs energy of
formation of nickel orthosilicate, from oxides, determined by [98ROG/KOZ] in the
range 973 – 1323 K was used by this review to fit best experimental values in the temperature range 970 – 1770 K.
[99ARC]
Archer detailed the sources of previous thermodynamic property values for nickel and
some of its compounds being of importance to environmental processes. This review is
particularly relevant to the origin of values contained in the NBS Thermodynamic Tables [82WAG/EVA].
A. Discussion of selected references
437
[99KON/KON]
Königsberger et al. presented a low-temperature thermodynamic model for the
Na2CO3–MgCO3–CaCO3–H2O system. The model was based on calorimetrically determined ∆ f H mο (298 K) values, S mο (298 K) values and C pο,m (T) functions taken from the
literature as well as on µο (298 K) values of solids derived from solubility measurements.
[99MAL/CAR]
The activity coefficients of (among others) aqueous Ni(ClO4)2 solutions ( mNi(ClO4 )2 =
7.3×10–5 – 0.89 mol·kg–1) have been determined from the emf of liquid-membrane cells.
For the purposes of this review, these data, together with the activity and osmotic coefficients listed in [69LIB/SAD], have been used to derive the ion interaction coefficient
ε(Ni2+, ClO −4 ). The selected value is ε(Ni2+, ClO −4 ) = (0.37 ± 0.03) kg·mol–1 (see Section
V.4.3). Ciavatta [80CIA] proposed to use ionic strength dependent ion interaction coefficients if the uncertainty is ± 0.03 or greater (Appendix B, Section B.3). The above
uncertainty is at the limit. Accepting Ciavatta’s proposal ε(Ni2+, ClO −4 ) = ε1 + ε2 log10 Im,
where ε1 = (0.287 ± 0.02) kg·mol–1 , ε2 = (0.121 ± 0.03) kg·mol–1 can be calculated.
However, these values provide a better fit to the experimental data only for m > 2.5
mol·kg–1 (I > 7.5 mol·kg–1), and lead to negative ion interaction coefficients for m <
0.0015 mol·kg–1. Therefore, the use of an ionic strength dependent ε(Ni2+, ClO −4 ) was
rejected in this review.
[99MCB]
McBreen comprehensively reviewed nickel hydroxide battery electrodes, the solid state
chemistry of nickel hydroxides, and the electrochemical reactions of the
Ni(OH)2 | NiOOH couple. Any critical discussion of the thermodynamic data of nickel
oxide hydroxides with higher oxidation states has to refer to this splendidly written account of nickel solid state electrochemistry.
[2000ROG/KOZ]
This paper of Róg et al. is a description of emf-measurements of Ni2SiO4 and Co2SiO4
with a solid electrolyte. A calcium fluoride-based composite containing SiO2 was applied as a solid electrolyte. The results obtained in the temperature range 823 to 1273 K
coincide perfectly with results of other emf-studies of Ni2SiO4 carried out by Róg et al.
[84ROG/BOR], [98ROG/KOZ] and also by other authors quoted in this review.
[2000TSI/MOL]
The main object of this conductance study was to obtain information on changes in association at a function of solvent composition. However, the value for the nickel sulphate association constant in water (calculated from the data using the Lee-Wheaton
equation) at 20°C is reported. For the purposes of the present review, the data were re-
438
A. Discussion of selected references
analysed with the distance parameter fixed to the value consistent with the SIT procedure described in Appendix B. The recalculated value of K1 is (118.6 ± 7.2) compared
to (139 ± 1) reported in the original paper. This value is markedly lower than the results
from other studies [73KAT], [76SHI/TSU], [79FIS/FOX].
[2001GAM/PRE]
In order to synthesise hellyerite, Rossetti-François’ method was applied and an aqueous
solution of nickel chloride was reacted with ammonium carbonate. Fragmented crystals
(d ≤ 2µm) and amorphous particles were precipitated [52ROS]. Therefore a new method
was developed. The initially precipitated amorphous particles were dissolved and larger
crystals (d = 0.05 mm) obtained when Ni2+ was kept in excess and the solution was
slowly acidified by CO2. The X-ray data of hellyerite prepared by this new method
agreed satisfactorily with the data given in the JCPDS-ICDD card 12- 276.
The aqueous solubility of hellyerite, then believed to have the stoichiometric
formula NiCO3·6H2O, was studied at varying temperatures and constant ionic strength
(I = 1.0 mol·kg–1). The crystals used were X-rayed before and after the dissolution experiment, but each time only the characteristic peaks of hellyerite were found. Data of
log10 {[Ni 2+ ] × pCO } plotted versus pH fell on straight lines with slopes of approxi2
mately – 2.0 for all temperatures.
From the experimental data obtained a preliminary set of the thermodynamic
quantities ∆ f Gmο , ∆ f H mο and S mο for hellyerite was derived using the ChemSage optimiser routine [95KON/ERI].
[2001HEN/RED]
This is a study of the non-convergent ordering of Mg and Ni in synthetic olivines by
means of the neutron powder diffraction and X-ray absorption spectroscopy (EXAFS).
The implications of the intracristalline ordering to olivine crystal chemistry and their
application to the consideration of thermodynamic relations, geothermometry, geospeedometry, and minor or trace element distribution control is discussed in detail.
[2001MOL/TSI]
The same data for nickel sulphate are reported as in [2000TSI/MOL]. However, values
for the specific conductance are also reported.
[2002GAM/WAL]
The re-evaluation of ∆ f Gmο and ∆ f H mο of β-Ni(OH)2 discussed in Section V.3.2.2.3. has
been based on the results of this paper.
Theophrastite, β-Ni(OH)2, was synthesised by an improved method based on
the hydrolysis of Na2[Ni(OH)4]. Pure single crystals were obtained with sizes up to
0.25 mm [2002WAL/GAT]. The solubility of β-Ni(OH)2 was measured at different
A. Discussion of selected references
439
temperatures and ionic strengths, and the data observed were thermodynamically analysed to obtain the respective standard molar quantities of formation ∆ f Gmο and ∆ f H mο .
For parameter optimisation non-linear least squares routines were used.
The tendency of precipitated Ni(OH)2 to form larger, well-shaped crystallites is
extraordinarily small (see Figure 13 of [77OSW/ASP]). Consequently it seemed doubtful that solubility equilibria according to the following equations
β-Ni(OH)2(cr) U Ni2+ + 2 OH–
β-Ni(OH)2(cr) + 2 H+ U Ni2+ + 2 H2O(l)
can be approached both from over- and undersaturation. Thus the method of pHvariation (i.e., the initial acidity was varied from 0.005 to 0.100 mol·kg–1 HClO4) was
employed [63SCH], but met in the case of β-Ni(OH)2 with two difficulties:
•
The equilibrium pH in the neutral range is poorly buffered by solute species.
•
At 25°C Ni(OH)2 is notoriously inert, thus the solid-solute buffer system
doesn’t work either.
However, when the experiments were carried out in a temperature range between 35 and 80°C constant pH values indicating equilibrium were attained within periods from 3 weeks to as little as 48 hours.
In a first series of solubility experiments on theophrastite the temperatures
were varied, and the ionic strength was kept constant at 1.0 mol·kg–1 NaClO4. Figure A40 shows results typical for the pH-variation method. Data of log10[Ni2+]tot plotted versus pH fall on straight lines with the theoretical slope of – 2.0.
At 50°C a second series of solubility measurements was carried out at different
ionic strengths varying from 0.5 to 3.0 mol·kg–1 NaClO4. Solubility constants were extrapolated to infinite dilution using the SIT [97GRE/PLY2]. Figure A-41 shows that the
experimental data of β-Ni(OH)2 coincide reasonably well with the optimised curve obtained when the SIT equation is employed.
A. Discussion of selected references
440
Figure A-40: Solubility of theophrastite as determined by the pH variation method;
t(°C): 80 ▲, 70 □, 60 ▼, 50 ○, 35 ■; solid straight lines; calculated with mean values of
log10 *K s ,0 and theoretical slope = – 2.0; I = 1.0 mol·kg–1 (Na)ClO4.
-1.0
-1.2
2+
−1
log10 ([Ni ]tot/mol·kg )
-1.4
-1.6
-1.8
-2.0
-2.2
-2.4
-2.6
-2.8
5.0
5.2
5.4
5.6
5.8
6.0
6.2
6.4
6.6
6.8
pH
Figure A-41: Ionic strength dependence of theophrastite solubility. Inert electrolyte:
NaClO4. I (mol·kg–1): 0.5 ♦, 1.0 ●, 2.0 ■, 3.0 ▲, + mean value and error bar
(∆ log10 *K s = ± 0.2). Solid curve: calculated according to the SIT model
[97GRE/PLY2] with the interaction parameter ε(Ni2+, ClO −4 ) = 0.37 kg·mol–1 selected
in this review.
10.8
T/K = 323.15
10.6
log10 Ks
10.4
10.2
10.0
9.8
0.0
0.2
0.4
0.6
0.8
1.0
−1 0.5
(I / mol·kg )
1.2
1.4
1.6
1.8
A. Discussion of selected references
441
[2002MON/HEL]
A kinetic study of cyanide exchange in [M(CN)4]2– square-planar complexes (M = Ni,
Pd and Pt) was performed as a function of pH using 13C NMR. The kinetic data confirmed the presence of the pentacyano complex at a higher excess of cyanide, and the
protonated species below pH 7. Surprisingly, the protonation does not affect the chemical shift of the 13C NMR complexed-cyanide signal (Kolski and Margerum detected
identical UV-VIS spectra for Ni(CN) 24 − and its protonated derivatives [68KOL/MAR]).
The observed kinetic behaviour was described using log10 b 4 = 31.0 for the formation
constant of Ni(CN) 24 − [59FRE/SCH], log10 K 5 = – 0.77 for the reaction:
Ni(CN) 24 − + CN– U Ni(CN)35−
[71PIE/HUG], and log10 K1,H = 5.4 and log10 K 2,H = 4.5 for the two successive protonations of Ni(CN) 24 − [68KOL/MAR]. Under the conditions used (25°C, Im = 0.6 m),
log10 b 4 = (29.75 ± 0.20) seems to be more reliable, but the impact of the erroneous β4
on the conclusions (e.g., the presence of protonated species) is rather difficult to judge.
[2002WAL/GAT]
A new method to synthesize β-Ni(OH)2 has been described. It is based on the hydrolysis
of Na2[Ni(OH)4] and leads to pure single crystalline nickel hydroxide with crystal sizes
up to 0.25 mm. The influence of temperature and concentration on the crystal size was
studied.
[2002WAL/PRE]
Solubility measurements of hellyerite were carried out at different ionic strengths of
NaClO4 (I = 0.5, 1.0, 2.0 and 3.0 mol·kg–1). These and all other solubility data on gaspéite [80REI], [82KLO] and hellyerite [11AGE/VAL], [30MUL/LUB], [2001GAM/PRE]
available were thermodynamically analysed to obtain the respective standard thermodynamic quantities ∆ f Gmο and ∆ f H mο . For hellyerite, S mο was also determined. Note that in
this paper all calculations on hellyerite were based on the stoichiometric formula
NiCO3·6H2O.
[2003BAE/BRA]
The stability constants of nickel carbonato complexes were investigated by measuring
the solid/liquid distribution ratio of Ni in the absence and presence of varying carbonate
concentrations. For this purpose a 63Ni tracer method was employed. A considerable
part of the uncertainty ( log10 K1ο , for the reaction Ni2+ + CO32 − U NiCO3(aq), is calculated to be (4.2 ± 0.4)) inherent in this method arises from the poorly known formation
constants of neutral and anionic hydroxo complexes of nickel.
442
A. Discussion of selected references
[2003HUM/CUR]
Hummel and Curti comprehensively reviewed complex formation between nickel and
carbonato and hydrogen carbonato ligands. Not only was the basis for experimental
measurements discussed, but also methods used for estimation of the formation constants [76ZHO/BEZ], [77MAT/SPO], [79MAT/SPO], [81TUR/WHI], [84FOU/CRI],
[87EMA/FAR]. Therefore a short summary of their arguments suffices.
As suggested by Garrels and Christ linear relationships between – log10 K1ο or
– log10 b1ο and the electronegativity of the metals were used by Zhorov et al. to determine the respective constants for Ni(II) [65GAR/CHR], [76ZHO/BEZ]. The former
case was calibrated with CuCO3(aq), MgCO3(aq) and CaCO3(aq), the latter with
MgHCO3+ and CaHCO3+ only, although Garrels and Christ warned against an overestimation of the then available data on alkaline earth hydrogen carbonato complexes.
Mattigod and Sposito correlated the complex formation constants with the sum of the
radii of the metal ion and the ligand forming the ion-pair. Their estimation formula for
the nickel carbonate system was calibrated to a single experimental data point
(CuCO3(aq)) [77MAT/SPO]. Two years later, Mattigod and Sposito correlated these
stability constants with a multiple parameter involving ionic charge, electronegativity,
integral weighting factors, and a contribution calculated from atomic shielding constants
[79MAT/SPO]. The numerical values obtained by the two estimation methods differ
considerably, but their empirical bases remained equally doubtful. Turner et al. used the
following correlation between the stability constants of carbonato and oxalato complexes:
log10 b MCO3 = – 0.011 + 1.042 log10 b MC2 O4
(A.108)
Hummel and Curti succeeded in revealing the rationale of Equation (A.108)
and pointed out that it leads to log10 K1ο (NiCO3, aq) = (5.4 ± 1.5) rather than log10 K1ο =
5.36 [81TUR/WHI]. Fouillac and Criaud [84FOU/CRI] applied different versions of
electrostatic models, but finally resorted to the numbers of Zhorov et al. [76ZHO/BEZ]
which they mistook for experimental data and therefore corrected to I = 0. This vicious
circle was probably one of the reasons which led to the subtitle “A thermodynamic elegy” for Hummel and Curti’s review [2003HUM/CUR].
B Appendix B
Ionic strength corrections1
Thermodynamic data always refer to a selected standard state. The definition given by
IUPAC [82LAF] is adopted in this review as outlined in Section II.3.1. According to
this definition, the standard state for a solute B in a solution is a hypothetical solution,
at the standard state pressure, in which mB = m = 1 mol·kg−1, and in which the activity
coefficient γB is unity. However, for many reactions, measurements cannot be made
accurately (or at all) in dilute solutions from which the necessary extrapolation to the
standard state would be simple. This is invariably the case for reactions involving ions
of high charge. Precise thermodynamic information for these systems can only be obtained in the presence of an inert electrolyte of sufficiently high concentration that ensures activity factors are reasonably constant throughout the measurements. This appendix describes and illustrates the method used in this review for the extrapolation of
experimental equilibrium data to zero ionic strength.
The activity factors of all the species participating in reactions in high ionic
strength media must be estimated in order to reduce the thermodynamic data obtained
from the experiment to the state I = 0. Two alternative methods can be used to describe
the ionic medium dependence of equilibrium constants:
• One method takes into account the individual characteristics of the ionic media
by using a medium dependent expression for the activity coefficients of the
species involved in the equilibrium reactions. The medium dependence is described by virial or ion interaction coefficients as used in the Pitzer equations
[73PIT] and in the specific ion interaction theory.
1
This Appendix essentially contains the text of the TDB-2 Guideline written by Grenthe and Wanner
[2000GRE/WAN], earlier versions of which have been printed in the previous NEA TDB reviews
[92GRE/FUG], [95SIL/BID], [99RAR/RAN], [2001LEM/FUG] and [2003GUI/FAN]. The equations presented here are an essential part of the review procedure and are required to use the selected thermodynamic values. The contents of Tables B.4 and B.5 have been revised.
443
444
B Ionic strength corrections
• The other method uses an extended Debye-Hückel expression in which the activity coefficients of reactants and products depend only on the ionic charge
and the ionic strength, but it accounts for the medium specific properties by introducing ion pairing between the medium ions and the species involved in the
equilibrium reactions. Earlier, this approach has been used extensively in marine chemistry, cf. Refs. [79JOH/PYT], [79MIL], [79PYT], [79WHI2].
The activity factor estimates are thus based on the use of Debye-Hückel type
equations. The “extended” Debye-Hückel equations are either in the form of specific
ion interaction methods or the Davies equation [62DAV]. However, the Davies equation should in general not be used at ionic strengths larger than 0.1 mol · kg−1. The
method preferred in the NEA Thermochemical Data Base review is a mediumdependent expression for the activity coefficients, which is the specific ion interaction
theory in the form of the Brønsted-Guggenheim-Scatchard approach. Other forms of
specific ion interaction methods (the Pitzer and Brewer “B-method” [61LEW/RAN]
and the Pitzer virial coefficient method [79PIT]) are described in the NEA Guidelines
for the extrapolation to zero ionic strength [2000GRE/WAN].
The specific ion interaction methods are reliable for intercomparison of experimental data in a given concentration range. In many cases this includes data at
rather low ionic strengths, I = 0.01 to 0.1 M, cf. Figure B-1, while in other cases, notably for cations of high charge ( ≥ + 4 and ≤ − 4), the lowest available ionic strength is
often 0.2 M or higher, see for example Figures V.12 and V.13 in [92GRE/FUG]. It is
reasonable to assume that the extrapolated equilibrium constants at I = 0 are more precise in the former than in the latter cases. The extrapolation error is composed of two
parts, one due to experimental errors, and the other due to model errors. The model
errors seem to be rather small for many systems, less than 0.1 units in log10 K ο . For
reactions involving ions of high charge, which may be extensively hydrolysed, one
cannot perform experiments at low ionic strengths. Hence, it is impossible to estimate
the extrapolation error. This is true for all methods used to estimate activity corrections.
Systematic model errors of this type are not included in the uncertainties assigned to
the selected data in this review.
It should be emphasised that the specific ion interaction model is approximate.
Modifying it, for example by introducing the equations suggested by Ciavatta
([90CIA], Eqs. (8−10), cf. Section B.1.4), would result in slightly different ion interaction coefficients and equilibrium constants. Both methods provide an internally consistent set of values. However, their absolute values may differ somewhat. Grenthe et al.
[92GRE/FUG] estimate that these differences in general are less than 0.2 units in
log10 K ο , i.e., approximately 1 kJ·mol−1 in derived ∆ f Gmο values.
B.1 The specific ion interaction equations
445
B.1 The specific ion interaction equations
B.1.1 Background
In the following discussion of the SIT model strict quantity calculus was sacrified in
order to improve readability (cf. [91STO]). The Debye-Hückel term, which is the dominant term in the expression for the activity coefficients in dilute solution, accounts for
electrostatic, non-specific long-range interactions. At higher concentrations, short
range, non-electrostatic interactions have to be taken into account. This is usually done
by adding ionic strength dependent terms to the Debye-Hückel expression. This method
was first outlined by Brønsted [22BRO], [22BRO2] and elaborated by Scatchard
[36SCA] and Guggenheim [66GUG]. Biedermann [75BIE] highlighted its practical
value, especially for the estimation of ionic medium effects on equilibrium constants.
The two basic assumptions in the specific ion interaction theory are described below.
• Assumption 1: The activity coefficient g j of an ion j of charge zj in the solution of ionic strength Im may be described by Eq. (B.1):
log10 g j = − z 2j D + ∑ ε( j , k , I m ) mk
(B.1)
k
D is the Debye-Hückel term:
D=
A Im
1 + B a j Im
(B.2)
where Im is the molal ionic strength:
I m = 1 ∑ mi zi2
2 i
A and B are constants which are temperature and pressure dependent, and aj is an ion
size parameter (“distance of closest approach”) for the hydrated ion j. The Debye1
1
Hückel limiting slope, A, has a value of (0.509 ± 0.001) kg 2 ⋅ mol− 2 at 25°C and 1 bar,
(cf. Section B.1.2). The term Baj in the denominator of the Debye-Hückel term has
1
1
been assigned a value of Baj = 1.5 kg 2 ⋅ mol− 2 at 25°C and 1 bar, as proposed by
Scatchard [76SCA] and accepted by Ciavatta [80CIA]. This value has been found to
minimise, for several species, the ionic strength dependence of ε( j , k , I m ) between Im =
0.5 m and Im = 3.5 m. It should be mentioned that some authors have proposed different
values for Baj ranging from Baj = 1.0 [35GUG] to Baj = 1.6 [62VAS]. However, the
parameter Baj is empirical and as such is correlated to the value of ε( j , k , I m ) . Hence,
this variety of values for Baj does not represent an uncertainty range, but rather indicates that several different sets of Baj and ε( j , k , I m ) may describe equally well the
experimental mean activity coefficients of a given electrolyte. The ion interaction coefficients at 25°C listed in Table B-4, Table B-5 and Table B-6 have thus to be used with
1
1
Baj = 1.5 kg 2 ⋅ mol− 2 .
446
B Ionic strength corrections
The summation in Eq. (B.1) extends over all ions k present in solution. Their
molality is denoted by mk, and the specific ion interaction parameters, ε( j , k , I m ) , in
general depend only slightly on the ionic strength. The concentrations of the ions of the
ionic medium are often very much larger than those of the reacting species. Hence, the
ionic medium ions will make the main contribution to the value of log10γj for the reacting ions. This fact often makes it possible to simplify the summation ∑ ε( j , k , I m )mk ,
k
so that only ion interaction coefficients between the participating ionic species and the
ionic medium ions are included, as shown in Eqs. (B.4) to (B.8).
• Assumption 2: The ion interaction coefficients, ε( j , k , I m ) are zero for ions of
the same charge sign and for uncharged species. The rationale behind this is that ε,
which describes specific short-range interactions, must be small for ions of the same
charge since they are usually far from one another due to electrostatic repulsion. This
holds to a lesser extent also for uncharged species.
Eq. (B.1) will allow fairly accurate estimates of the activity coefficients in
mixtures of electrolytes if the ion interaction coefficients are known. Ion interaction
coefficients for simple ions can be obtained from tabulated data of mean activity coefficients of strong electrolytes or from the corresponding osmotic coefficients. Ion interaction coefficients for complexes can either be estimated from the charge and size of
the ion or determined experimentally from the variation of the equilibrium constant
with the ionic strength.
Ion interaction coefficients are not strictly constant but may vary slightly with
the ionic strength. The extent of this variation depends on the charge type and is small
for 1:1, 1:2 and 2:1 electrolytes for molalities less than 3.5 m. The concentration dependence of the ion interaction coefficients can thus often be neglected. This point was
emphasised by Guggenheim [66GUG], who has presented a considerable amount of
experimental material supporting this approach. The concentration dependence is larger
for electrolytes of higher charge. In order to reproduce accurately their activity coefficient data, concentration dependent ion interaction coefficients have to be used, cf.
Lewis et al. [61LEW/RAN], Baes and Mesmer [76BAE/MES], or Ciavatta [80CIA].
By using a more elaborate virial expansion, Pitzer and co-workers [73PIT],
[73PIT/MAY], [74PIT/KIM], [74PIT/MAY], [75PIT], [76PIT/SIL], [78PIT/PET],
[79PIT] have managed to describe measured activity coefficients of a large number of
electrolytes with high precision over a large concentration range. Pitzer’s model generally contains three parameters as compared to one in the specific ion interaction theory.
The use of the theory requires knowledge of all these parameters. The derivation of
Pitzer coefficients for many complexes, such as those of the actinides would require a
very large amount of additional experimental work, since few data of this type are currently available.
B.1 The specific ion interaction equations
447
The way in which the activity coefficient corrections are performed in this review according to the specific ion interaction theory is illustrated below for a general
case of a complex formation reaction. Charges are omitted for brevity.
m M + q L + n H 2 O(l) U M m L q (OH) n + n H +
The formation constant of M m L q (OH) n , *b q , n , m , determined in an ionic medium (1:1 salt NX) of the ionic strength Im , is related to the corresponding value at zero
ionic strength, *b qο, n, m by Eq.(B.3).
log10 *b q , n , m = log10 *b qο, n , m + m log10 g M + q log10 g L + n log10 aH2 O
− log10 g q , n, m − n log10 g H+
(B.3)
The subscript (q,n,m) denotes the complex ion, M m L q (OH) n . If the concentrations of N and X are much greater than the concentrations of M, L, M m L q (OH) n and
H+, only the molalities mN and mX have to be taken into account for the calculation of
the term, ∑ ε( j , k , I m )mk in Eq. (B.1). For example, for the activity coefficient of the
k
metal cation M, γM, Eq. (B.4) is obtained at 25°C and 1 bar.
log10 g M =
− zM2 0.509 I m
1 + 1.5 I m
+ ε(M ,X ,I m )mX
(B.4)
Under these conditions, Im ≈ mX = mN Substituting the log10γ j values in Eq.
(B.3) with the corresponding forms of Eq. (B.4) and rearranging leads to:
log10 *b q , n , m − ∆ z 2 D − n log10 aH2 O = log10 *b qο, n , m − ∆ ε I m
(B.5)
where, at 25°C and 1 bar:
∆z 2 = (m zM − q zL − n) 2 + n − mzM2 − q zL2
D=
0.509 I m
1 + 1.5 I m
∆ε = ε(q, n, m, N or X) + n ε(H, X) − q ε(N, L) − m ε(M, X)
(B.6)
(B.7)
(B.8)
Here (m zM − q zL − n) , zM and zL are the charges of the complex,
M m L q (OH) n , the metal ion M and the ligand L, respectively.
Equilibria involving H2O(l) as a reactant or product require a correction for
the activity of water, aH2 O . The activity of water in an electrolyte mixture can be calculated as:
− φ ∑ mk
k
log10 aH2 O =
(B.9)
ln(10) ⋅ 1
M H2 O
where φ is the osmotic coefficient of the mixture and the summation extends over all
solute species k with molality mk present in the solution. In the presence of an ionic
B Ionic strength corrections
448
medium NX as the dominant species, Eq. (B.9) can be simplified by neglecting the
contributions of all minor species, i.e., the reacting ions. Hence, for a 1:1 electrolyte of
ionic strength Im ≈ mNX, Eq. (B.9) becomes:
log10 aH2 O =
− 2 mNX φ
ln(10) ⋅ 1
M H2 O
(B.10)
Alternatively, water activities can be taken from Table B-1. These have been
calculated for the most common ionic media at various concentrations applying Pitzer’s
ion interaction model and the interaction parameters given in [91PIT]. Data in italics
have been calculated for concentrations beyond the validity of the parameter set applied.
These data are therefore extrapolations and should be used with care.
Table B-1: Water activities aH2 O at 298.15 K for the most common ionic media at various concentrations applying Pitzer’s ion interaction approach and the interaction parameters given in [91PIT]. Data in italics have been calculated for concentrations beyond the validity of the parameter set applied. These data are therefore extrapolations
and should be used with care.
Water activities aH2 O at 298.15 K
c (M)
HClO4
NaClO4
LiClO4
NH4ClO4
Ba(ClO4)2
HCl
NaCl
LiCl
0.10
0.9966
0.9966
0.9966
0.9967
0.9953
0.9966
0.9966
0.9966
0.25
0.9914
0.9917
0.9912
0.9920
0.9879
0.9914
0.9917
0.9915
0.50
0.9821
0.9833
0.9817
0.9844
0.9740
0.9823
0.9833
0.9826
0.75
0.9720
0.9747
0.9713
0.9769
0.9576
0.9726
0.9748
0.9731
1.00
0.9609
0.9660
0.9602
0.9694
0.9387
0.9620
0.9661
0.9631
1.50
0.9357
0.9476
0.9341
0.9542
0.8929
0.9386
0.9479
0.9412
2.00
0.9056
0.9279
0.9037
0.8383
0.9115
0.9284
0.9167
3.00
0.8285
0.8840
0.8280
0.7226
0.8459
0.8850
0.8589
4.00
0.7260
0.8331
0.7309
0.7643
0.8352
0.7991
5.00
0.5982
0.7744
0.6677
0.7782
0.7079
6.00
0.4513
0.7075
0.5592
0.6169
(Continued on next page)
B.1 The specific ion interaction equations
449
Table B-1: (continued)
c (M)
KCl
NH4Cl
MgCl2
CaCl2
NaBr
HNO3
NaNO3
LiNO3
0.10
0.9966
0.9966
0.9953
0.9954
0.9966
0.9966
0.9967
0.9966
0.25
0.9918
0.9918
0.9880
0.9882
0.9916
0.9915
0.9919
0.9915
0.50
0.9836
0.9836
0.9744
0.9753
0.9830
0.9827
0.9841
0.9827
0.75
0.9754
0.9753
0.9585
0.9605
0.9742
0.9736
0.9764
0.9733
1.00
0.9671
0.9669
0.9399
0.9436
0.9650
0.9641
0.9688
0.9635
1.50
0.9500
0.9494
0.8939
0.9024
0.9455
0.9439
0.9536
0.9422
2.00
0.9320
0.9311
0.8358
0.8507
0.9241
0.9221
0.9385
0.9188
3.00
0.8933
0.8918
0.6866
0.7168
0.8753
0.8737
0.9079
0.8657
4.00
0.8503
0.8491
0.5083
0.5511
0.8174
0.8196
0.8766
0.8052
0.3738
0.7499
0.7612
0.8446
0.7390
0.6728
0.7006
0.8120
0.6696
5.00
0.8037
6.00
c (M)
NH4NO3 Na2SO4
(NH4)2SO4 Na2CO3
K2CO3
NaSCN
0.10
0.9967
0.9957
0.9958
0.9956
0.9955
0.9966
0.25
0.9920
0.9900
0.9902
0.9896
0.9892
0.9915
0.50
0.9843
0.9813
0.9814
0.9805
0.9789
0.9828
0.75
0.9768
0.9732
0.9728
0.9720
0.9683
0.9736
1.00
0.9694
0.9653
0.9640
0.9637
0.9570
0.9641
1.50
0.9548
0.9491
0.9455
0.9467
0.9316
0.9438
2.00
0.9403
0.9247
0.9283
0.9014
0.9215
3.00
0.9115
0.8735
0.8235
0.8708
4.00
0.8829
0.8050
0.7195
0.8115
5.00
0.8545
0.5887
0.7436
6.00
0.8266
0.6685
B Ionic strength corrections
450
Values of osmotic coefficients for single electrolytes have been compiled by
various authors, e.g., Robinson and Stokes [59ROB/STO]. The activity of water can
also be calculated from the known activity coefficients of the dissolved species. In the
presence of an ionic medium, N ν+ X ν − , of a concentration much larger than those of
the reacting ions, the osmotic coefficient can be calculated according to Eq. (B.11) (cf.
Eqs. (23−39), (23−40) and (A4−2) in [61LEW/RAN]).
1− f =

A ln(10) z+ z− 
1
1 + B a j I m − 2 ln(1 + B a j I m ) −

3
I m ( B a j ) 
1 + B a j I m 
 ν ν 
− ln(10) ε(N,X) mNX  + − 
 ν+ + ν− 
(B.11)
where ν + and ν − are the number of cations and anions in the salt formula
( ν + z+ = ν − z− ) and in this case:
Im =
1
z+ z− mNX (ν + + ν − )
2
The activity of water is obtained by inserting Eq. (B.11) into Eq. (B.10). It
should be mentioned that in mixed electrolytes with several components at high concentrations, it might be necessary to use Pitzer’s equation to calculate the activity of
water. On the other hand, aH2 O is nearly constant in most experimental studies of equilibria in dilute aqueous solutions, where an ionic medium is used in large excess with
respect to the reactants. The medium electrolyte thus determines the osmotic coefficient
of the solvent.
In natural waters the situation is similar; the ionic strength of most surface waters is so low that the activity of H2O(l) can be set equal to unity. A correction may be
necessary in the case of seawater, where a sufficiently good approximation for the osmotic coefficient may be obtained by considering NaCl as the dominant electrolyte.
In more complex solutions of high ionic strengths with more than one electrolyte at significant concentrations, e.g., (Na+, Mg2+, Ca2+) (Cl−, SO 24 − ), Pitzer’s equation
(cf. [2000GRE/WAN]) may be used to estimate the osmotic coefficient; the necessary
interaction coefficients are known for most systems of geochemical interest.
Note that in all ion interaction approaches, the equation for the mean activity
coefficients can be split up to give equations for conventional single ion activity coefficients in mixtures, e.g., Eq. (B.1). The latter are strictly valid only when used in combinations which yield electroneutrality. Thus, while estimating medium effects on standard potentials, a combination of redox equilibria with, H + + e − U 1 H 2 (g) , is nec2
essary (cf. Example B.3).
B.1 The specific ion interaction equations
451
B.1.2 Ionic strength corrections at temperatures other than
298.15 K
Values of the Debye-Hückel parameters A and B in Eqs. (B.2) and (B.11) are listed in
Table B-2 for a few temperatures at a pressure of 1 bar below 100°C and at the steam
saturated pressure for t ≥ 100°C. The values in Table B-2 may be calculated from the
static dielectric constant and the density of water as a function of temperature and pressure, and are also found for example in Refs. [74HEL/KIR], [79BRA/PIT],
[81HEL/KIR], [84ANA/ATK], [90ARC/WAN].
The term, Baj, in the denominator of the Debye-Hückel term, D, cf. Eq. (B.2),
1
1
has been assigned in this review a value of 1.5 kg 2 ⋅ mol− 2 at 25°C and 1 bar, cf. Section B.1.1. At temperatures and pressures other than the reference and standard state,
the following possibilities exist:
•
The value of Baj is calculated at each temperature assuming that ion sizes are
independent of temperature and using the values of B listed in Table B-2.
•
The value Baj is kept constant at 1.5 kg 2 ⋅ mol− 2 . Due the variation of B with
temperature, cf. Table B-2, this implies a temperature dependence for ion size
parameters. Assuming for the ion size is in reality constant, then it is seen that
this simplification introduces an error in D, which increases with temperature
and ionic strength (this error is less than ± 0.01 at t ≤ 100°C and I < 6 m, and
less than ± 0.006 at t ≤ 50°C and I ≤ 4 m).
•
The value of Baj is calculated at each temperature assuming a given temperature
variation for aj and using the values of B listed in Table B-2. For example, in the
aqueous ionic model of Helgeson and co-workers ([88TAN/HEL],
[88SHO/HEL], [89SHO/HEL], [89SHO/HEL2]) ionic sizes follow the relation:
a j (T ) = a j (298.15 K, 1 bar) + z j g (T , p ) [90OEL/HEL], where g(T, p) is a
temperature and pressure function which is tabulated in [88TAN/HEL],
[92SHO/OEL], and is approximately zero at temperatures below 175°C.
1
1
The values of ε( j , k , I m ) , obtained with the methods described in Section
B.1.3 at temperatures other than 25°C, will depend on the value adopted for Baj.. As
long as a consistent approach is followed, values of ε( j , k , I m ) absorb the choice of
Baj, and for moderate temperature intervals (between 0 and 200°C) the choice Baj =
1
1
1.5 kg 2 ⋅ mol− 2 .is the simplest one and is recommended by this review.
The variation of ε( j , k , I m ) with temperature is discussed by Lewis et al.
[61LEW/RAN], Millero [79MIL], Helgeson et al. [81HEL/KIR], [90OEL/HEL], Giffaut et al. [93GIF/VIT2] and Grenthe and Plyasunov [97GRE/PLY]. The absolute values for the reported ion interaction parameters differ in these studies due to the fact that
the Debye-Hückel term used by these authors is not exactly the same. Nevertheless,
common to all these studies is the fact that values of (∂ ε / ∂ T ) p are usually ≤ 0.005
kg·mol−1·K−1 for temperatures below 200°C. Therefore, if values of ε( j , k , I m ) obtained
B Ionic strength corrections
452
at 25°C are used in the temperature range 0 to 50°C to perform ionic strength corrections, the error in log10 g j / I m will be ≤ 0.13. It is clear that in order to reduce the uncertainties in solubility calculations at t ≠ 25°C, studies on the variation of ε( j , k , I m )
values with temperature should be undertaken.
Table B-2: Debye-Hückel constants as a function of temperature at a pressure of 1 bar
below 100°C and at the steam saturated pressure for t ≥ 100°C. The uncertainty in the
A parameter is estimated by this review to be ± 0.001 at 25°C, and ± 0.006 at 300°C,
while for the B parameter the estimated uncertainty ranges from ± 0.0003 at 25°C
to ± 0.001 at 300°C.
t(°C)
p(bar)
A ( kg 2 ⋅ mol− 2 )
B × 10−10 ( kg 2 ⋅ mol− 2 ⋅ mol−1 )
0
1.00
0.491
0.3246
5
1.00
0.494
0.3254
10
1.00
0.498
0.3261
15
1.00
0.501
0.3268
20
1.00
0.505
0.3277
25
1.00
0.509
0.3284
30
1.00
0.513
0.3292
35
1.00
0.518
0.3300
40
1.00
0.525
0.3312
50
1.00
0.534
0.3326
1
1
1
75
1.00
0.564
0.3371
100
1.013
0.600
0.3422
125
2.32
0.642
0.3476
150
4.76
0.690
0.3533
8.92
175
0.746
0.3593
200
15.5
0.810
0.365
250
29.7
0.980
0.379
300
85.8
1.252
0.396
1
B.1 The specific ion interaction equations
453
B.1.3 Estimation of ion interaction coefficients
B.1.3.1
Estimation from mean activity coefficient data
Example B.1:
The ion interaction coefficient ε(H + , Cl− ) can be obtained from published values of
g ±, HCl versus m