THE AMERIC.\N
\:OI, 5I, JULY_AUGUST, 19(19
MI\ERALOGIST,
KINETICS AND RELATIONS IN THE CALCITE-HYDROGEN
REACTION AND RELATIONS IN THE DOLOMITE-HYDROGEN AND SIDERITE-HYDROGEN SYSTENTS
The Unir:ers'ity
A. A. Gr-q.norNr
ANDC. A. S.q.rorrr,Departmentof Geol,ogy,
of Georgia,Athens,Georgia,30601.
Al]srnacr
Reactions of calcite, dolomite, and siderite rvith h1'drogen at elevated temperatures
and pressuresresult in the formation of inorganic hydrocarbons up to and including butane. The carbonate-hydrogen experiments were run in a "cold-seal" t1'pe vessel' The
analysesof the reaction products'n'eredone by mass spectrometry for the gases,and by'lvet
chemical, atomic absorption spectroscopy,and X ray methods for the solids.
Hydrocarbons appear to form directly from a carbonate surface gas reaction rather than
from a reaction betn-een generated gas and hydrogen. An inverse relation holds between
the complexity of the hydrocarbon formed and the temperature of formation
An evaluation of the kinetic data for calcite-hydrogen shou's the reaction to be pseudofirst-order. lior the experimental system the Arrhenius apparent activation energf is
18.000 callmole.
INtnoructtoN
Recent evidence (Ikorskii, 1964; Ikorskii and Romanikhim, 1961
1964)
Petersilie,1958; Petersilie,1963;Petersilie,1964;Zakrshevskaver,
indicates that reducing gaseswere present during the cr1-stallizittionof
s o m ei g n e o u sr o c k s .T h e r n r o d ln a m i c c a l c u l a t i o n s( F r e n c h ,1 Q 6 6 s) t r g g e s t
that in graphite-bearingrocks the equilibrium gasesshouldbe reducingin
composition.The predominant emphasisin experimental petrology to
date has been the studl'of systemsin air; under their own vapor pressures,carbonate and sulfi.desystemsfor example; or under an imposed
pressure,usually H2O or COz.The writers believe that the studl: sf na1ural and s1-ntheticsvstemsunder reducing atmosphereswill 1-ieldinformation that is pertinent to a number of petrologic problems. This
manuscript reports some of our fi.ndingsin the svstems calcite-h1'drogen,
dolomite-hydrogen,and siderite-hydrogen.
Calcite, dolomite, and sideriteunder the appropriate conditionsreact
with hydrogen to produce hydrocarbons. These reactions have been
studied under various P-T conditions.
ExprnmrNral
Pnocnounrs
The follor-ing materials were investigated in the hope that they rvouid have a catalytic
efiect upon the reactions: metallic nickel, platinum, copper, titanium, magnesium, and
iron; commercial mixtures of 0.5 percent palladium, platinum, and rhodium on alumina;
and dried silica gel, activated alumina, hematite, magnetite, chromic oxide, chromium
trioxide, and various "Kieselguhr" mixtures. The possible autocatalytic eflects o{ product
h1'drous and anhydrous oxides and p1'rolitic carbon also were investigated. None of the
1151
t152
A A. GIARDINI AND C A SALOTTI
100000 PSI
20000 Psl
i n t en sii e
fr
furnace
rld-seal"
vessel
gascylinders
Frc. 1. Schematic drawing of the apparatus used for studies in the calcite-hydrogen,
dolomite-h1'drogen, and siderite hydrogen systems.
above mentioned materials had any discernible efiect upon the reaction rate. Several of the
metals, particularly iron, increased the rate of pyrolitic dissociation of the generated
methane.
The experimental system is shown schematically in Figure 1. The bulk of the experiments u'ere carried out in a "cold seal" vessel made from Kelsay-Hayes Co. Udimet 500
stainless stee'|.A 20,000 psi Heise gauge measured the gas pressures.A platinum-rvound
muffle furnace rvas used to heat the lower section of the reaction vessel. A stainless steel
sheathed Cr-Al thermocouple was inserted into the reaction vessel through the cold seal
b)'means of a pressure fitting to measure the internal temperature External bottom i,vell
thermocouples also rvere used. A Leeds and Northrup model 60 control unit with a Fincor
saturable reactor and later an API digital setpoint SCR proportioning controller maintained the temperature automatically. Equilibrium, although approached, was not established in these experiments.
The procedure used for most of the experiments was as follows:
1 The carbonate mineral was weighed and loosely housed in platinum foil;
2. The sample was placed in the vessel, connected to the pressure system, flushed 2
limes with pure helium. at about 1000 psi, then charged with helium to about 1000 psi;
3 Upon reaching the operating temperature, the helium was vented and the s1'stem
given 3 hydrogen flushes and then pressurized to the desired hydrogen pressure;
4 After the experimenlal interval the excesshydrogen plus reaction gases were vented
(in some experiments a fraction rvas collected for mass spectrographic analysis) and the
vessel flushed and pressurized with helium;
5. The furnance was drawn from the vessel and the latter air-quenched under the internal helium pressure;
6 Alter cooiing, most of the samples were opened in air and placed in argon filled
containers. Some erperiments were opened in an argon-fi1leddry glove bor (Na clessicant)
and samples sealed in glass capillary tubes.
Most of the experiments rvere conducted at pressures belorv 10,000 psi. However, one
KI N ]JTI CS I N CARBONAT E-II VDROGE,A/.'YS?TMS
1153
E
()
E
aj4rlfisje?4,r,
O
I
E
b-2
Lr_
-3
Tt \ , "r.
Frc. 2, Standard-state free energiesof formation of methane, ethane, and propane by
reactions of the t1'pe: (2nll)HztnCO:C*Hz,+zlnHzO;
and the free-energy variation
o1 the water-gas reaction (data after Rossini el al.,1947).
experiment was run to investigate the effect of high pressures. A calcite cleavage rhomb
showed no discernible alteration after being held at 228.C under 80,000 psi hydrogen
pressurefor 23 hours.
The mass spectrographic analyses were done by Drs Richard and patricia crawford,
Chemistrl' Department, University of California, Lawrence Iladiation Laboratory, on a
Consolidated Electrodynamics Corporation model 21-103C anal),tical mass spectrometer.
The analytical sensitivity is 0.01c1 by volume. Samples rvith simple compositions were
compared rvith knorvn patterns using a step-wise regressionprogram, and more complex
compositions were interpreted using the computer step-wise regressionprogram of D. D.
Tunnicliff and P. A. Wadsworth at Shell Develooment Co
Al1 of the starting carbonates were minerals. The calcite was optical grade obtained from
\\iard's. The dolomite was also from Ward's. The siderite was donated by Dr. Dennis
ILadclifie. Al1 starting minerals were examined optically, hand picked for impurities,
crushed, and heated in 30 percent h1'drogen peroxide
A. A, GIARDINI AND C, A. SALOTTI
1154
CarcrrB-HvDRoGEN
In the calcite-hydrogen system using calcite fragments of 40 to 60
mesh,between535 870oC,and 200-8000psi of initial hydrogen, the following compoundswere observed:solid CaO; Ca (OH)t; graphite; and
carbon "soot." GaseousCHa; C2H6;CO; COz; and HzO also formed'
The experimental results are summarized in Table 1 and plotted in
Figures3, 4, and 5. The weight percentCOzin the remainingsolid and the
mole percent CHa formed are plotted against the duration of the run in
hours at 2000 psi (Pn,).
Below its dissociationtemperatureCa(OH)zis the stablesolid reaction
phase.In runs allowed to cool to room temperature under the reaction
gases,Ca(OH)2is alwayspresent.In experimentsrun abovethe dissociation temperatureof Ca(OH)2,CaO is presentif the reaclion gasesare replaced with helium at the operating temperature. Analyses of the reaction gasesindicate that water is the oxygenatedproduct. Simplifiedequations for the reactionsare the following:
(A)
(B)
CaCOr* 4H, :
C a C O a* 4 H : :
CaO t
CH* + 2HrO
C a ( O H ) z* C H ' * H r O
The changein free energy for reaction B is negative below approximatell. 325"C;whereasfor reactionA it is positive.However, the free enCalcite-Hydrogen
60soc, 20o0 psi(H2)
L o s( C ) r
Residualwt 96
cq
TIME in HOURS
Frc. 3. Linear plot of log C againsttime for calcite-hydrogendata from experiments71,
nature of
72,73, and 74; run at 605.C, 2000psi (H) illustratingthe pseudo-first-order
the reaction.
KIN ET I CS IN CARBONAT E.HY DROGEN SI/SZETI4S
z
a-l
C)
FI
I
h
F
e
tr
o
C)
I
E
z
!
4
9
=
EH
Fi,
O
Yq
6V
E
B5
^o
2X
<i
F
*!
A9
E
z
OH
q0
r'l <
z
E
4z
z>
O
*l
el
9
!
€
O
cn
,n
;-
z3
d
' ce
F
.;
F
z6
r
C)
:LQ
E:
E
a
(-)
.9 -u.9 g
U
Ee
<6aZ
1155
1156
A A. GIARDINI AND C. A. SALOTTI
roLE * cH.
H
o
U
R
S
Frc 4 p,otormo,e16.",:1":,:"'i,:i:.'i:::,::::
cozreandresidua,
svstem
runat 605"C;2000psi(H) ; for 2,4,8,and16hours'
in experiments
ascalcite
maining
ergli changefor reaction A is lesssensitiveto increasingtemperatureand
at about 540oc reactionA is thermodynamicallyfavored over reactionB.
Graphite is present mainly along calcium hydroxide surfaces,with
Iesseramounts occurringwithin the calcium hydroxide.At runs at higher
temperatures()700'C) a "soot-like" material formed in the bomb' It
was readily wiped clean from the sidesof the bomb. Onl1 rarely was this
,,soot-like" material ever associatedwith the solids within the platinum
cassette.Examination of the analytical data (Table I) reveals that at
higher temperatures(experiment81) there is a much larger weight lossof
carbon as COzin the solid than can be accountedfor as carbon in the generated CHa. This discrepancyis beyond analytical error. The "soot-like"
material appearsto be amorphous carbon formed through the thermal
dissociationof methane, and higher h1'drocarbonsif thel. formed' Any
higher hydrocarbonsformed by pyrolysis of the methane would be unstable and none were found in the gas analysis.Carbon and hydrogen are
the end products of methane pyrolysis, but equilibrium is difficult to attain. Catall'sis can hasten equilibrium. The dissociationof methane is
catalytically promoted b1'platinum, iron, nickel, all of which are present
in theseeroeriments.
K I N I'7' I CS I N CA RBONAT ]],-II YDIIOG I:'N S YS T EM S
WT
r157
o / oC 0 2 R E S I D U A LS O L I D
CALCITE-HYDROGEN
2,OOO psi (H2)
2 hour rung
T
'c
M O Lf
o/o CHa
Frc. 5 In the calcite-hydrogen system, plot of mole 7o CHn generated and residual
CO2 remaining as calcite in experiments run for 2 hours; 2000 psi (Hr);at 535, 605, 735,
790, and 870oC.
Analysis of the reaction gasesshows that the reaction compoundsare
restricted in number even though a large population of compoundsare
possiblein the C-H-O s\rstem.The gasespresentare the following: CHa;
C2H6;H2O; CO; and CO2.The latter two appearedin only one experiment (no. 75). This experimentwas unique in that the original hydrogen
pressurewas only 200 psi.l The appearanceof CO and COz at low pressuresmay be explained,at least in part, through thermodynamic calcuiations for a simplified C-H-O gaseoussystem (French, 1966).Thesecalculationsshow that decreasingpressurefavors the formation of COzand
water relative to methane and are in agreementwith our experimental
findings.
It appearsthat methane and, if within the stability field, homologrres
of methane, form directly rather than from reactionsbetween hydrogen
I Fugacities vary with pressure,consequentll'the
proportion oi each substancepresent
in the equiiibrium will also vary with pressure.
1158
A. A. GIAITDINI AND C. A SALOTTI
and CO or COz.Except at low pressuresand high temperatures,the experimental evidenceis that COzis not present.
Thermod-vnamicdata indicate that at moderate pressuresparallinic
hydrocarbonscan form below 350"c and olefinsbelow 250"c by reactions
betweencarbon dioxide and hydrogen; although the free energy changes
of such r.eactionsare much lessnegativethan carbon monoxide-hydrogen
also is pertireactions. The water-gas reaction (Hr+COr:COfHrO)
nent. Equilibrium constantsfor the water-gasreaction show that at temperaturesbelow 800oC,equilibrium very much favors the formation of
carbon dioxide and hydrogen.lThe point is that COz is thermodynamically stablewith regard to both the water-gasreaction and, at the temperatures of the experiments,in regard to hydrogenation, but it is not
found in the gas analysesexcept under the stated restricted conditions of
low pressureand high temperature.For the experimentalconditions investigated,therefore,a direct methanation of the calcite appearsto be
the plausible reaction.
The position that CO is not generallyan intermediate product in the
dissociationof calciLeis not as defensibleas the argument against CO2.
The reaction of CO with Hz forms the basis of the Fischer-TropschhyIn this synthesisthe best yields of hydrocarbonsocdrocarbons1'nthesis.
cur if the CO and I{z are introducedin the ratio of 1:2 by volume at about
with in100 psi total pressure'The hyd'rocarbonf ield rapidly decreases
2 shows
Figure
creasingh1-drogencontent and increasingtotal pressure.
of the
formation
that under most of our experimental conditions, the
is
therH2
and
Iighter hydrocarbons(Cr to Cn) by reaction between CO
CO
Hz*CaCO:r:
modynamically favorable. The initial reaction:
increase
substantial
*Ca(OH)z under standard conditions results in a
(+22.6+ kcal/mole) in free energy' The subsequentreaction: CO*3Hr
CH4+HrO f ields a large decrease(-34'01 kcal/mole) in free energy'The
sum of these two reactions, of course, equals the overall free energv
which
change for the reaction: CaCOaf4Hz:CHE*Ca(oH)z*Hzo
yields methane directly.
Again, however, the best evidencethat co is not a generalreaction
prod.uctis found in the gas analyses.Under conditionsof some of our experiments CO does occur in the presenceof excesshydrogen. This is
shown by the gas analvsesof experiments75 and 42 (Tables 1 and 3)'
Ilowever, CO is absentfrom all other gas analysesand it is reasonableto
assumethat it never was presentin detectablequantities.TheseobservaI The peristance of H:O and CO in some of the reaction gases below 800oC is another
indication of disequilibrium, and is due to the slowness o1 the subsequent water-gas reaction.
LLNGTII OF BONDS IN FRAMDWORK SILICAS
1159
tions, therefore,also appear to support a reaction involving direct methanation rather than hydrogenationof CO or COz.
The kineticsof the thermal dissociationof carbonates,especiallycalcite
and dolomite, has receivedconsiderableattention over the 1'ears.It has
been proposedthat the diffusion of releasedCOz is the rate controlling
step (Zadwadski and Bretsznajer, 1933a,1933b, 1038a,1938b; Maskill
and Turner, 1932;Britton, et al.,1952a,1952b,1952c).
The conductionof
heat to and within the samplehas also been held to be the determining
step. More recently a rate expressioninvolving actual and equilibrium
CO2pressuresand a temperature constant has been proposed(Hyatt et
o1..1958).
Aithough the calcite-hydrogenreactionis bimolecularand reactionbegins at temperatureswell below the "in air" calcinationof caicite,the reaction kinetics describingthe calcite-hvdrogens)-stempossess
similarities
to the kinetics for the thermal dissociationof calcite in vacuum. Hvdrogen appearsto simulate a vacuum environment.
The determination of anv reaction rate inevitable reducesto the determination of concentrationas a function of time. The methaneconcentration doesnot accuratelyreflectthe reactionbecauseof the subsequent
reaction: CHa: C+2H2. The weight percentcarbon dioxide remainingin
the solid is an appropriate concentrationmeasureof the reaction rate.
Becauseequal amounts of calcite, sieved to a common grain size,were
used for each run, the weight percent COr remaining in the solid can be
convertedinto conventionalunits, i.e. moles/cc,bv introducing the necessaryconstant. It is sufficienthere to recognizethat under the experimental conditionsweight percent CO: is a measureof the calcite concentration in the solid.
An unusual form of the concentrationunit is no inconvenienceif the
reaction is first order becausethe value of the rate constant for a first order reactionis independentof the concentrationunit. The hydrogen concentration is initialiy very large relative to calcite, and the percentage
changein the hydrogen concentrationis small even in runs of 16 hours.
During the experiments,therefore,the hydrogen concentrationfactor is
nearly constant, and because of this the reaction can be considered
pseudo-firstorderl and successfullytreated as such.
A p l o t o f t h e r a t e c o n s t a n t sc a l c u l a t e df r o m e x p e r i m e n t s7 1 , 7 2 , 7 3 , 7 4 ,
I A second-order reaction in terms of the variable o representing the decrease in con(b r) where a and b represent the
centration of a reactant in a given time is: dx/d.t:h(o-r)
initial concentrations. If b represents the initial hydrogen concentration in the calcitehydrogen reaction at 605oC and 2000 psi 2(Hr) then the value of (b r) for the interval
between 2 and. 16 hours is 0.998 to 0.985. The term (D-r) can be neglected leaving dr/dt
:k (a-r), the relation of a first order reaction
A. A. GIARDINI AND C. A. SALOTT]
1160
Calcite-Hvdrooen
799og.2eqplsi
RESIDUAL
wT%
c02
MOLE
qb
cHr
TIME iN HOURS
Frc. 6. PIot of calcite-hydrogen system showing molc a/o CHr generated and residual
COz remaining as calcite in experiments run at 700oC; 2000 psi (Hz) exp. rro.77,79 and
8000 psi (H) exp. no. 78; for 8, 4, and 2 hours respectively.
run at 605oC and 2000 psi P(H) for 2, 4, 8, 16 hours respectivelyis
shown in Figure 3. The deviation from linearity is not excessive,and the
agreementof the calculated rate constantsfor the same experimentsis
within the experimental error (Table 2). Both evaluationssupport the
assumptionof a pseudo-first-orderreaction.
T.Lrln 2. C,tr.crrt-HvonocEN Ar 605'C, 2000 rsr P(Hz)
I (hours)
2
4
8
16
wt % co'
remaining in solid
42.0
4 1. 0
3 7- 8
?A, '
fr (hourr)
0.019
0.016
0 018
0.015
mean: 0.017
o Initial COs Wt
/6 is 43.7
RI N ETI CS I N CARBONAT E-H YDROGEJ\ISITSTTMS
t16L
The Arrhenius equation:
dln k
It,,
dT
RT2
can be usedwith the data from experiments71, 91 and 80 to evaluatethe
variation of the rate constants with temperature, and to calculate the
Arrhenius apparent energy of activation (E,).
Figure 7 is a plot of the Arrhenius equation using data from experim e n t s n o s . 7 0 , 7 L , 9 1 , 8 0 , a n d 8 1 . T h e p r e s s u r e( 2 0 0 0p s i ) a n d t h e t i m e
(2 hours) were common to all experiments.The temperaturesare 535oC,
605oC,735oC,790oC,and 870oCrespectivel_v.
The apparent activation
energy is calculatedfrom the linear portion of the plot.
T h e A r r h e n i u se q u a t i o n :d l n k / d T : L " f R T 2 , b e c a u s e , Ei*s c o n s r a n t ,
C a l c i t eH
- ydrogen
psi
2 H o u r s2, O O O
logk= --EL . .-1Ea=-|8.OOO
cdtlmole
(Jl0
o
o
J
-l*
o
o
J
l/ T" .lor
l-rc. 7. Plot of 1og(1/l log ColC) or log A against the reciprocal of the absolute temperat u r e f o r e x p e r i m e n t s n u m b e r e d 7 0 , 7 7 , 9 2 , 8 0 , a n d 8 1T h e s e e x p e r i m e n t s s , c r e r u n a t 2 0 0 0
psi (Hz); 2 hours; at 535, 605, 735, 790, and 870o respectively. The Arrhenius activation
energy calculatecl{rom the slope of experiments 77,91, and 80 is 23,000cal/mole.
A. A. GIARDINI AND C A. SALOTTI
1162
- E"/ RT * C. This latcan be integrated with respectto T to give In k:
ter form was used with the data from experiments71,91, and 80 to calculate the Arrhenius apparent activation energy (1,). This activation
energy is 18,000cal/mole.
3. Dere Usao rN rnn Evalu.tTroN orr :|nr ActrvarroN
Talr-r
i::
llt
7r
91
80
81
'c
5JJ
605
735
790
870
| /r'
1 . 2 3 X 1 30
114X10-3
o 99xlo t
'94xto t
. 8 7X 1 0 - 3
c
co/C
los Co/C
430
430
378
34.5
163
!.o2
1 04
I t6
1.27
2.68
0.0086
0170
.0645
.1038
.4281
co
ENBnCv rN FrC. 7
l b c c o / *c ( + * ; )
0 004
008
.032
052
2r1
-2
-2
-1
-I
-0
40
05
50
28
61
The relation between the rate constant and the initial and final concentration of a reactant in a first-order reaction is as follows:
L -
_
I
t
l-
Co
-
.
c'
and the activation energv can be evaluatedfrom Figure 7.
The slope was determinedby using only the information from experiments 71, 91, and 80. Experiment 70 was excludedbecauseof the analytical uncertainty in determining so small a change in the amount of reacted.calcite. Experiment 81 was excluded becauseof the likelihood of a
changein the behavior of the system at higher temperatures.
The digressionfrom linearity of the sample at 870oc may result from
the interaction of a competing reaction. The competing reaction may be
either or both the thermal dissociation of calcite and the pyrolysis of
methane.
Usually a homogeneousreaction has a higher activation energy and so
is favored at high temperatures, whereasthe heterogeneousreaction pre,dominatesat lower temperatures. The pyrolysis methane decomposition
may also effect the calcite-hydrogen reaction rate. Britton, Gregg and
Winsor (1952a)and Britton Gregg,Winsor and Willing (1952)on the basis of a study of the thermal decomposition of calcite in vacuum, concluded that the rate of dissociation was proportional to the area of the
interface. Our studies support the conclusion that the calcite-hydrogen
reaction is surface dependent'
In earlier work where a single cleavagerhombohedron was used for the
calcite charge, it was common to recover a core of unreacted calcite enveloped by a cohesiverim of calcium hydroxide. Many of these samples
KI N L,TI CS I N CAIIBONATE-H YDROGI::N SYST].MS
!,9
r____j
Zfftm
1163
f.
Frc. 8. Unequal depth of reaction envelope oi Ca(OH)z developed on cLeavagerhombohedron of optical calcite. Experimental conditions: 66.5oC;10,000 psi (H);4 hours. The
unreacted calcite core was cleanly plucked from the Ca(OH): rim during grinding and for
illustrative purposes the calcite void was filled with piastic.
were immediately imbeddedin quick-settingplastic, and flats were carefully ground normal to the side cleavage surfaces.If the samples were
not potted, spontaneousspalling of the reaction film and shattering of
the core generallyoccurredon standing.In addition to normal cleavage,
conchoidal fracture was sometimesobserved on the shattered calcite.
The depth of alteration was measuredoptically and found to be very
anisotropicand crystallographicallysimilar for all crystals.The average
ratio of maximum to minimum depth of the anisotropic reaction rim (for
29 samples)is 2.1.
The crystallographic orientation of several specimens was determined by using thin sections,a petrographicmicroscope,and a universal
stage. The reaction rims and crystallographicorientation are shown in
Figure 5. In a sense,the reaction rims ma1' be comparedto etch figures.
Under normal conditions, the calcite I cr1stal system and class are
given as: trigonal, 32f m.Honess and Jones (1937)in their investigation
of etch figuresin carbonateminerals by opticallv active solvents,noted
1164
A. A. GIAIIDINI AND C. A. SALOTTI
that the six sidesof a cleavagerhombohedronof calcite always showed
the same figures and solubility rates, whereasdolomite showed unlike
solubility of the rhombohedral cleavagefaces.The crystallographicorientation of the unequal reaction rims, as determinedoptically on the remaining calcitecore of our specimens,showsthat the c axis doesnot pass
through the intersectionof the three similarly reactedfaces.The crystallographic anisotropy of the reaction implies more than a loss of the
symmetry center.
DorourrB aNr HvlnocpN
Dolomite and hydrogen were reacted under the following conditions:
(1) fragment size 40-60 mesh; (2) 520oCto 835"C; (3) 2000 to 5000 psi
initial hydrogen pressure;and (4) experimentslasting between 2 and 12
hours. Solid reactants observed include: CaC03(calcite); Ca(OH)z;
CaO; non-crystalline Mg(OH)r; elemental carbon or graphite, and
"soot". Gas analysesshow: CHa; CzHo;CO; and COz.Water is the oxygenated product from aII of the experiments.
The experimentalresults are summarizedin Table 4 and plotted in
Figures9 and 10. The weight percent COzremaining in the solid and the
mole percent CH+ formed are again plotted against time at 5000 psi for
620"C, and against temperature at 5000 psi for 4-hour runs.
In the thermal dissociationof doiomite under the vapor pressureof
the system, the reaction occurs in two stages.The first dissociation,
stoichiometricallysimplified,is given as follows:
caMg(coe): + A :
MgO *
C a C O rf
COz
With increasingtemperature, the remaining calcite decomposesto lime
and carbon dioxide with the exact dissociationtemperature determined
by the carbon dioxide pressure.If the P6e, r€ver exceeds100 mm,
dolomite decomposesin a singlestage dissociationas follows:
C a M g ( C O * ) r* A : M g O *
6aQ f
2COz
In the dolomite-hydrogenreaclion COz is not a reaction product under most of the experimental condilions and appears only as a minor
component of the reaction gasesat lower hydrogen pressures(Table 4).
It likely is incidental to the main reaction of direct methanation. The
basisfor this statement is found in the dolomite-hydrogengas analyses
and the reasoning previousll'used in discussingthe calcite-hydrogen
system.
The dolomite-hydrogenreaction occursin two steps.The first stageis
representedby the following equation:
KI N ETI CS I N CA RBONAT E-II YDROGIiIT STSZEMS
I 165
MoLE % CHI
Dolomite-Hydrogen
5OOOpsi,4Hours
T
t
RESTDUAL
wr % Cq
lrc 9. Plot ol mole /6 CHa generated and residual wt /6 COz in the solid for the dolomite-hydrogen s)'stem at 5000 psi (H);4 hour experimentsl at temperatures of 525, 550,
620, and 735'C.
4H, f
C a M g ( C O a ) z:
CaCOr* Mg(OH), + CHr * HrO.
At 5000 psi (Hr) reaction initiates as low as 520"C (Table 4). Calcite
formed from dolomite persistsat higher temperaturesthan does calcite
alone in the calcite-hydrogensystem. Similarities exist between the
thermal dissociationof carbonatesand the reaction of carbonateswith
hydrogen. In the dolomite-hydrogensystem CHa ma1.influencethe reaction in a manner similar to COz in the thermal decompositionof calcite. This possibility will be exploredin future studies.
Mg(OH), or NIgO f ormed in the first stageis non-crystallineto X-rays.
Chemicalanalvsesby atomic absorptionindicates17/6 MgO in the solid
reaction products.
A very minor amount of black lustrous material is present in the reaction products.It is similar in all appearancesto the graphite formed in
the calcite experiments."Soot-like" material also was present in the
bomb, and again was particularly noticeable at higher temperatures.
The reaction gaseswere those also presentin the calcite experiments.
( T a b l e 4 ) . C a r b o n d i o x i d ea p p e a r si n t w o e x p e r i m e n t s( n o s . 8 3 , 8 4 ) b u t
A. A. GIARDINI AND C. A, SALOTTI
1166
Dolomite-Hydrogen
620cc,5116psI
RESIDIJAL
MOLE
%
wr%
c02
cHc
TIME
in I1OURS
CFIz in the solid for the
Frc. 10. Plot of mole /6 CHa generated and the residual wtls
dolomite-hydrogen system at 620"C;5000 psi (H:); for 2, 4, 8, and 12 hour experiments.
only in very small amounts. Both of these experimentswere run at a
lower pressure of 2000 psi. The discussion on the reaction gasesfor the
calcite-hydrogen experiments applies also to the dolomite-hydrogen
system.
The kinetics of the dolomite-hydrogen system are considerably more
complex than in the calcite-hydrogensystem. An evaluation of the rate
constant for each successiveconcentration-reaction time pair at 620"C
assuming the reaction is first, second or third order, shows wide scatter,
and no trend for the rate constant.The wide divergencefrom linearity is
iilustrated by a plot of the Arhennius equation for a plot of the six temperature-concentrationpairs. An interpretation of kinetic data into physical terms for this system is not realistic with the limited data available'
SrprnrrB AND H2
The reactionbetweenone-halfgram of 40 to 60 meshsideritefragments
and hydrogen is more complex than the precedingcalcite-hydrogenand
dolomite hydrogen reactions.It is also the least studied with only four experiments. These four runs were between 400 to 605oC and 2000 to
5000psi (H2), all for 4 hours. A "thermal soak" under helium was usedin
KINETICS IN CARBONATE-IIYDROGEN SYSTEMS
bo
600
eelol6
:
o
I
oiolalo
2l "]:
-
lo
- lo
?lq
1
o lo lo
z
t^
t*
|]
n
4lo
I
I
FlSl{
lrlz
t,
ls
el
^l:
i$
L
I
o]-
lle
q
l^;
L
O
I
z
!
4>
F
oti
2'6
t-l >
YZ
(-)
z
>c
^a
=v
ltr
ol^
?
lSLo
F
E
@ F l
<o
v !
z::
ir>
<o
'rJ
47
z
9<
;z
<c-t 4
d
olo
o l*
loT
a-l
<-
z
B
3
O
.2q
a
3.>
a
t;
<r
rl
F
O
z
ts
O
Fr
F
(J
s
fl
O
6a
tr67
1168
A. A. GIARDINI
AND
C. A. SALOTTI
Tlrlr
5. Mess SprcrnocRApgrc Auar,ysr,s or RracrroN Gases" b c ANDWET Cnnurcer,
Amelrses or.Rrsroual COz rN UNrolctoo Sronnrrn Srornrrp HronocrN Svsrelr
87
Experiment No
605
2,000
Temperature ('C)
Pressure (psi)
Duration of run (hours)
Mole 076CHa in gas
Mole o/s CzHe in gas
Mole 0l CrHs in gas
Mole a/o CqHro in gas
Mole L/6 COz in gas
Mole a/6 H2 in gas
Mole 07i He in gas
400
2,000
0 .1 5
45.5
2,000
4
1.34
0.42
o.23
0.05
99.80
0.04
9+.90
0.08
95.10
0.09
n
525
5,000
AA
1. 4 5
0.12
0.03
4.45
0 28
0.10
003
004
95.30
3.07
" Analysis calculated on water-free basis.
b Starting hydrogen impurities given as follows: Iess than 5 ppm N2, less than 1 ppm
Oz, less than f ppm COz, iess than I ppm CO, Dew point-100oF.
" D e t e c t i o n l i m i t 0 . 0 1 ' .i .
each case to bring
the bomb
up to temperature
before
venting
and re-
placingh1'drogenfor the reactiongas.Becauseof the low temperaturefor
thermal dissociationof siderite,in those experimentsat higher temperatures somesideriteundoubtedly decomposedand the decompositiongas,
presumably CO2,was expelledalong with the helium. For this reasonthe
mole percent of gaseoushydrocarbonsdeterminedin the reaction gas for
these erperiments are minimum values (Table 5).
T h e s o l i d r e a c t i o np r o d u c t sf o r e x p e r i m e n t s4 0 ( 5 2 5 " C , 5 0 0 0p s i ) , 8 6
( 6 0 5 " C , 2 0 0 0p s i ) , a n d 8 7 ( 4 5 5 ' C , 2 0 0 0 p s i ) a r e i r o n ( F e ) a n d w i i s t i t e
(FeO). The solid reactionproducedfrom experiment89 (400"C, 2000psi)
is magnetite. In all of the experimentsthe reaction gaseswere sampled
and replacedwith helium at the reaction temperature. Minute and rare
flecks of what optically appearedto be graphite were observed,but no
certain identification was made.
The limited data availablesuggeststhat wiistite is the primar)r alteration solid. The ultimate IJ2O/H2 ratio and the temperature dictate
whether wiistite will be reducedto native iron or oxidized to n'iagnetite.
The simplified stoichiometricrelations would be the following:
FeCO3+ 5H2 :
FeO + CH4 + 2I{rO
if reducing:
H:r*FeO:Fe*HzO
KIN]JTICS IN CARBONATE-IIYDROGEN
SYSTEMS
1169
if oridizing:r
HrOf3FeO:FeeOrfHr
We have to date no experimentalevidencethat magnetite does not
form directllr under the appropriate conditions.'fhe lower the temperature, the drier the hydrogen needsto be to effect reduction. In Esperiment 87 run at 455"C at 2000 psi for 4 hours the solidswere wiistite and
iron. In experiment89 run at 400"C at 2000psi for 4 hours the solidswere
wiistite and magnetite.An extrapolationof Eastman and Evans' (1924)
"best values" for the equilibrium constant from 700oC to 450'C and
400oCfor the reactions,
FeO * H: :
Fe J- HrO,
and
Fe3O+*Hr:3FeO*HrO,
yield valuesof the ratio H2O/H2 significantlyhigher than valuesobtained
using the }J2O/H, ratios from experiment89.
The validity of extrapolatingthe Eastman and Evans data to 400'C is
questionable,but the magnitude of the differencebetween the extrapolated equilibrium ratio and our experimentalHzO/H, ratio ailows for a
considerablemargin of uncertaintl'. The extrapolatedratio indicatesthat
under the conditions of experiment 89, not only should magnetite not
form, but wiistite should readily reduce to iron. Although our experiments with sideritenever reachedequiiibrium, this discrepancvdoessuggest that relations in the Fe-C-H-O system differ significantlv from the
Fe-H-O system.
The reaction gasesinclude appreciablemethane and lesseramounts of
ethane,propane,and butane. Water was presentin all of the erperiments.
The appearanceof propaneand butane can best be understoodb1'considering the thermal stability of these hvdrocarbons.The temperature
necessarvfor sideriteto react at an appreciablerate is low enoughso that
ethane,propane and butane are thermally-stable.There is no direct evidenceto indicate how thesemethanehomologuesform. Indirect evidence
suggeststhat they do not form from methane. Someof the higher-temperature calcite and dolomite hydrogen runs were allowed to cool slowly
under the reaction gases(methane,water, hydrogen) and the higher hydrocarbonsdid not form.
Suulr:lnv AND CoNCLUSToNS
Inorganic hydrocarbonsup to and including butane form directll- in
carbonatemineral-hvdrogenreactionsat temperaturesas low as 400"C.
1 The fugacity of oxygen is proportional to PllrsfPsr.
tl70
A. A. GIARDINI AND C. A. SALOTTI
The experimental evidence indicates that hydrocarbons form directly
from a mineral surface-hydrogenreaction rather than from a subsequent
reaction betweengeneratedgasesand hydrogen.In general,the reaction
rate is rapid at temperatures well below the "in air" calcination temperatures of the carbonate minerals.
An inverse relation exists between the complexity of the hydrocarbon
formed and the temperature of the experiment.The lower the initial reaction temperature of the carbonate mineral-hydrogenpair, the more
complex the hydrocarbon.
Equilibrium was not generally established in these experiments' although it must have been attained locally at higher temperaturesat the
reaction surfaces.An evaluation of the kinetic data for calcite-hydrogen
shows the reaction to be pseudo-first order. For our experimental s1'stem
the Arrhenius apparent activation energy is 18,000cal/mole. The reactions between dolomite-hydrogen and siderite-hydrogen are more complex and an interpretation of the kinetics from our present data is not
possible.
This study is pertinent to a number of geologic problems. It defines a
simple processto produce inorganic hydrocarbons under conditions that
could reasonablyexist within the crust. These data indicate that the reactions are temperaturerather than pressuresensitiveand that the partial pressure of hydrogen is of greater importance than is the absolute
pressureof hydrogen in initiating the reaction.
The processof limestoneand dolomite assimilationby a magma is also
germane. If the gasesassociatedwith a magma are reducing (hydrogen
need only be a part of such a reducing gas) then COz need not develop.
The water generated in the reaction: carbonate mineralfhydrogen
:methane*water*metal
oxide or hydroxide, would inhibit the reacnot.
This water can' of course,be effectively
would
tion; carbon dioxide
incorporated in the rock-forming minfrom
the
system
by
being
removed
may not leave the system. There is
may
or
The
hydrocarbons
erals.
Iimited field evidenceto support such a hypothesis.Petersil'ye(1962)estimates that the rocks of the Khibina massif contain 7 X 108m3 of inorganically derived hydrocarbon gases.
AcrNowr-rncurx:rs
Part of this work was performed under the aegis of the Atomic Energy Commission.
The mass spectrographic analyses were done by Richard and Patricia Crawford; the COz
analyses were made by william Sunderland; Harry Mcsween and carl Myers ran a number of X-ray difiraction analyses. We thank John Lakner for cooperation and assistance
in the experimental aspects of the study, and our colleagues, RoIf Bargman and Francis
Tohnston. for assistance in the interpretation of kinetic data'
KIND,TICS
IN
CARBONATE-IIYDROGEN
SI/STE,US
1171
RonenrNcos
Bntrros,H T.S.,S.J.Gnecc,exoE.G
J . W r r - r - r N c ( 1 9 5 2H) e a t t r e a t m e n t o f d o l o m i t e :
III Precipitation of magnesia from sea water by calcined dolomite J Appl. Chem.
(London'), 2, 701-703.
aNo G W. WrNson (1952) Heat treatment ol dolomite: I Activitv of lime
and of magnesia in relation to their temperature of preparation. J. AppL Chem.
(London), 2, 693-697 .
AND E. G. J. Wrlr,rNc (1952) Heat treatment of dolomite: II
Activity of dolime in relation to its temperature of preparation from dolomite. J . A ppl.
Clrcn. (Lond.on),2, 698 700.
eNo C. W. Wrxson (1952) The calcination of dolomite: I'I'he kinetics of
the thermai decomposition of calcite and of magnesite II Thermal decomposition of
dolomite. Trans Farod.Soc ,48r 63-69.
Fruxcn, B M (1956) Some geologicalimplications of equilibrium betrveengraphite and a
C H O gas at high temperatures and pressures. Re',s.
Geophys.,4,223 253.
Gtanorsr, A A., C. A. Slrorrr, ANDJ. F. La<Nrn. (1968) S1'nthesisof graphite and hydrocarbons by reaction betrveencalcite and h1'drogen.Science,159,317-319
Gonsrxl, V. N, I A. Pnrrnsrr.'vB, AND V. A. Pnp,q.crrxrN (1965) Combustible gasesin
rocks of the contact zone of the Khibiny Alkalic pluton Dokl,.Akod,.lfazft SSSR, 162,
179,181.
Gnar, D. L., amo J. E Lelrln, (1955) Properties of calcium and magnesium carbonates.
E c o n .G e o l . ,5 0 A n n i t . V o l . , 6 3 9 7 1 3 .
AND GoLDSMTTH,
J. R. (1955) Dolomite-magnesian calcite relations at elevated
temperatures and CO: pressures.Geochim.Cosntoch,imAcla,7, 109-128
Horrss, A. P., axo J. R. Joxns (1937) Etch figure investigation with optically active solvents. Geol,.Soc.Amer. 8u11..48.667-722.
Hvlrr, E. P., I. B. Cutr-nn, ANDM. E. W,r.nswonrn (1958) Calcium carbonate decomposition in carbon dioxide atmosphere: .I. Amer. Cerom,.Soc.,4lr 70-74.
konsru,
S. V. (1964) Hydrocarbon gases and bitumens in rock-forming minerals of the
Khibina Stock. DoAl. Aka.d..Nauk SSSR. 157. 90-93.
AND A I,f RorteNrrnrN (1964) Forms of hydrocarbon gases in nepheline from
rocks of the Khibina Alkalic massif Geokhindya 1964 (3), 27G281, lChern.Abstr.60,
1s621.1
Mesrrll, W. eNl W. E. S. Tunrr:n (1932) Further experiments on rate of decomposition
of calcium carbonate. J. Soc.Glass Technol,.16,80-93d
Pntnnsrl'vr, I. A. (1958) Hydrocarbon gasesin the intrusive massifs of the central part of
the Kola Peninsula. DokI. Ahod.. nr'ozrftSSSR, 122, 1086-1089.
--(1962) Origin of hydrocarbon gases and dispersed bitumens of the Khibina Alkalic
Massif. Geochem,istry,l, 15-29.
--(1963) Organic matter in eruptive and metamorphic rocks of the Kola Peninsula.
In, A.P. Vinogradov (ed) TheChemistry oJ lhe l\arlh's Crust, Publ. Acad. Sci. U S.S.R.,
Vo]. 1,48-62.
(1964) The gas constitutent and trace bitumens of igneous and metamorphic rocks
of the Kola Peninsula (Abstr). Int. GeoI.Cong.22nd Sess.,Rep. Vol. Abstr., p. 9-10.
Rossrxr, F. D., and others (1947) Selected Values of Properties of Hydrocarbons. Iy'at
Bw. Stand..lU. S.l Ci.rc.461,
Slrorrr, C. A. axo A. A. Gr,q.norNr(1968) Reactions betv'een calcite, doiomite, siderite,
and hydrogen. Trons. Amer. Geophys.Union,49r 342.
t172
A. A. GIARDINI AND C. A. SALOTTI
Zlnowsxr, J. eNo S. Bnrrszmelonn (1933a) Heterogeneousreactions of the type A solid
*B gas:C solid: I. Deviations from constancy of the equilibrium pressure; apparent
equilibrium and its significance. Z. Physik Chem.,35, ffi-?8
(1933b) Heterogeneousreactions of the type A solid*B gas:C solid:
and II Kineticsof carbonateformationanddecomposition Z.Physih.Chem.,35r79-96.
(1938a) Heterogeneousreactions of the type A solid*B gas:C solid:
, and -III Experimental determination of the elementar-v reactions in the combined process.
Z. Physik. Chem.,4O, 158-182.
- (1938b) Mechanism of reaction of the type solid:solid*gas:
'---,
Trans.
and F arod.aySoc., 34, 951-959.
N. G. (1964) Oriein of the gases in rocks of the Khibina apatite deZlxnsnrvsrava,
posits. Dokl Akad.. Nauk,S.SSR, 154, 118-120. lChem. Abstr.60,9039 [1964]; Engl.
fiansl. Dokl. Acail. Sci.. USSR, Earth Sci. Sec. 154, 128-1311.
Manwscri'blreeeittel.Jul'y 27,1968; accefteillor pwbli'cotion,May 1,1969.