Central European
Science Journals
w
w
w
.
c
e
s
j
.
c
o
Central European Journal
of Chemistry
m
CEJC 2(2) 2004 347 362
X P S Study of Group IA Carbonates
A.V. S h c h u k a r e v 1 *, D.V. Korolkov 1
Chemical Faculty,
St.Petersburg State University,
Universitetsky pr. 2, 198 504 Petergof, St.Petersburg, Russia
Received 18 September 2003; accepted 22 December 2003
A b s t r a c t : The results of systematic XPS measurements of all alkali metal carbonates (Li,
Na, K, Rb and Cs) are presented. The first set of experiments was performed with "as
received" commercial carbonate powders under liquid nitrogen conditions using a precooling
procedure. A second set of experiments was performed under similar experimental conditions
after a preliminary grinding (mechanical activation) of the carbonates. In addition, Na2COa
"1H20, NaHCOa and KHCOa powders were studied. It was found that sample pre-cooling
allows distinction between hydrocarbonates and carbonate hydrates. Storage in air leads to
formation of hydrocarbonates at the surface of Li2COa and Na2COa. This phenomenon being
more pronounced in the former. In contrast, K2COa forms a hydrate with one H20 molecule.
Rb2COa and Cs2COa have hydrocarbonates as well as hydrates at the surface and this is more
pronounced for Cs2COa. Grinding of the carbonates results in the formation of hydrocarbonates
at the surface, the tendency to do so was found to increase down the group IA, namely, K <
Rb << Cs. For the most part, the hydrocarbonates formed were unstable in vacuum even under
liquid nitrogen conditions. Chemical trends in C ls and O ls binding energies in carbonates
and hydrocarbonates of the Group IA are discussed and related to the nature of the anion and
alkali cation.
@ Central European Science Journals. All rights reserved.
Keywords: Carbonates, hydrocarbonates, XPS
1
Introduction
G r o u p IA c a r b o n a t e s are widely used in industry, medicine, a n d play an i m p o r t a n t role in
nature. Their chemical properties are well k n o w n a n d t h e y differ in t h e r m a l stability from
o t h e r carbonates. T h e G r o u p IA c a r b o n a t e s exhibit t h e effect of increasing size a n d mass
on chemical a n d physical properties. Due to this effect t h e lattice energy of c a r b o n a t e s
* E-mail: [email protected]
Unauthenticated
Download Date | 6/15/17 6:58 PM
348
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
and the ease of their thermal decomposition decrease down the group [l]. From a chemical
point of view, such behaviour is connected to the decrease in first ionisation potential (I1)
of alkali metal atoms and determines the ionicity of the metal-carbonate bonds.
Alkaline metal carbonates are easily formed at oxide/hydroxide surfaces. Owing to
their high chemical stability, they can be used as reference compounds for analytical
purposes. The availability of XPS re%fence data on alkali metal carbonates is however
limited to carbonates of Li and Na [2, 3].The objective of this work was to obtain the binding energies re%fence data for all Group IA carbonates and to understand the chemical
changes which can be introduced by sample preparation techniques.
2
Experimental
All XPS spectra were recorded with a Kratos Axis Ultra electron spectrometer using mono
A1 Ks source operated at 225 W and a low-energy electron gun for charge compensation.
Survey scan spectra were recorded in 5 rain using pass energy of 160 eV and a 1 eV step.
No elements other than those of bulk composition were detected, except in the case of
lithium carbonate containing 1 at. % of Na and 0.3 at. % of sulphate. Narrow scans
of the C ls, O ls and the main metal line were acquired using a pass energy of 20 eV
and a 0.1 eV step. The hybrid lens mode was used for all measurements. The samples
were hand-pressed in a molybdenum holder with a 5 mm diameter through of variable
volume. To avoid loss of water and to minimize possible decomposition of the surface,
a liquid nitrogen precooling procedure was carried out before the measurements. This
procedure includes precooling the end of the sample trans%r rod (20 rain at 170~
and then waiting 30 s after sample loading before pumping of the introducing chamber.
After pumping to 4-5 x 10 -5 Pa, the sample was trans%rred to the precooled (-160~
manipulator where it was kept until a base vacuum of 4-6 x 10-7 Pa in the analysis
chmnber was reached. No frost layer was detected at the surface, despite not using a
glove box for sample introduction. After the measurements under liquid nitrogen, the
samples were kept overnight at the analysis position allowing warming in vacuum to
room T, and the XPS spectra were recorded the following day. Processing of the spectra
was accomplished with Kratos software. The binding energy (BE) scale was referenced
to the C ls line of aliphatic carbon contamination, set at 285.0 eV.
Li, Na and K carbonates used in this study were powders from Merck and had pro
analysi purity. Cs2COa powder (eztra pure) was also from Merck. Rb2COa powder
from Aldrich had 99.8 % purity. All carbonates were investigated ~%s received" and
measurements were repeated with samples ground in an agate mortar for 3 rain.
3
R e s u l t s and discussion
C ls and O ls spectra of all carbonates are presented in Fig.1 - 7. BE and FWHM values
Unauthenticated
are given in Table 1.
Download Date | 6/15/17 6:58 PM
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 34~362
3.1
349
Li2CO3
The Li + ion has a n exceptionally high charge-radius ratio comparable with Mg 2+. Li2CO3
is thermally much less stable than other alkali metal carbonates owing to poor packing
of relatively large carbonate anions and small lithium ions [1]. The Ia of 5.39 eV is the
highest among the Group [4].
The C ls spectrum (Fig. 1) contains a main narrow component at 290.1 eV corresponding to carbon in the carbonate ion. The O ls spectrum consists of two lines, the
main at 531.8 eV, and the minor at 533.2 eV. The intensity of the minor component
decreases after sample warming. This line can be assigned to an adsorbed water molecule
or an H-O-C bond in the hydrocarbonate ion. The Li/carbonate atomic ratio is 1.6 and
is less than a theoretical value of 2 calculated from the chemical %rmula and indicates
the presence of Li hydrocarbonate at the surface.
3.2
Na2CO3
Anhydrous sodium carbonate has BE values of both C ls and O ls lines 0.8 eV less than
those %r Li. Presence of hydrocarbonate traces at the surface can be detected %r the
sample measured under liquid nitrogen conditions (Fig. 2). Sodium hydrogen carbonate
(NaHCO3) has binding energies of main C ls and O ls lines 0.6 eV higher than the
carbonate (Table 1). In addition, a second component in the O ls spectrum appears
at 532.9 eV (H-O-C bond) (Fig. 3). This component slightly decreases in intensity
during the sample warming. The atomic ratio of two main oxygen lines is approx. 2:1
reflecting the structure of the HCO~ ion. Na2COa.IH20 (a hydrate) shows an additional
component in the O ls spectrum at 532.3 eV, which disappears during warming the
sample in vacuum. Again, the Na/carbonate atomic ratio is less than 2 (1.75, Table 1),
and seems to be due to some hydrolysis of the surface in air. In the case of NaHCO3 this
ratio is 1 in accordance with the chemical %rmula.
3.3
K2COa
Comparing with Li and Na, the surface chemistry of K2COa differs. The C ls and O ls
binding energies of the carbonate ion continue to decrease. Potassium carbonate has a
main C ls line at 289.0 eV and two O ls components at 530.7 eV and 532.0 eV. The
last one nearly disappears after warming in vacuum, whereas the first one and the C ls
component decrease in BE (Table 1, Fig. 4). In parallel, the K/carbonate atomic ratio
(1.9) does not change, indicating that the surface of potassium carbonate is significantly
hydrated in air. KHCO3 has much higher binding energies than both K2CO3 samples and
in contrast with hydrated carbonate does not lose the second O ls component responsible
%r the H-O-C bond in hydrocarbonate (Table 1, Fig. 5). It is interesting to note that
an increase in BE %r the K 2p3/2 line %r KHCO3 seems to be connected with increased
Unauthenticated
polarisibility of the K + ion.
Download Date | 6/15/17 6:58 PM
350
3.4
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Rb2CO3
Rb carbonate is very hygroscopic. The interaction with water is clearly seen in corresponding O ls spectra (Fig. 6), where part of the third component at 532.3 eV corresponds to a H20 molecule in the carbonate structure. Moreover, this interaction results
in additional formation of RbHCO3 appearing as the high BE component in the C ls
spectrum at 289.2 eV and the second O ls component at 530.7 eV (Fig. 6). This hydrocarbonate phase is rather stable in vacuum and the corresponding XPS lines can be found
in the spectra even after three days of keeping the sample in vacuum at room T, whereas
water molecules are nearly completely removed during warming. A slow decomposition
of the R b H C Q is confirmed by the increase in Rb/carbonate atomic ratio from 1.75 for
a pre-cooled sample to 2.0 after 3 days in vacuum. Both C ls and O ls BE values of
Rb2CO3 are comparable to potassium carbonate (Table 1) which may be a consequence
of the close ion crystal radii of K (1.44 A) and Rb (1.58 A) [1] and corresponding lattice
energies.
3.5
Cs2CO3
In the case of Cs carbonate, the formation of the hydrocarbonate phase due to storage
in air becomes more pronounced (Fig. 7). Within the Group, Cs has the highest crystal
radius and the lowest I1 (3.893 eV) which leads to the highest ionicity for the C s - O ( C Q )
bonds. This, in turn, results in the lowest BE values for C ls and O ls lines of the
carbonate ion (Table 1).
3.6
Influence of grinding on the surface composition
In XPS practice, grinding is often used to produce a fresh surface more or less free
of possible ea ~itu "artefacts" like hydrocarbon contamination, oxidation products, etc.
On the other hand, the surface of oxygen and moisture sensitive samples can change in
chemical composition during such a treatment.
As an example, C ls and O ls spectra of Cs carbonate subjected to grinding are
given in Fig. 8. A remarkable increase in the surface hydrocarbonate phase was detected
also for Rb2CO3. Most part of the hydrocarbonate %rmed is unstable in spectrometer
vacuum even under liquid nitrogen conditions, as was in the case of Cs carbonate (Fig.
8), and the surface composition aspires to that of unground sample. In contrast with
the initial powders, some hydrocarbonate formed by sample grinding is stable enough to
be detected at the surface even after briefly heating the sample up to 100~ in vacuum.
Na2CO3 and K2CO3 powders are much less sensitive to the grinding and give hydrates
Unauthenticated
as the main product.
Download Date | 6/15/17 6:58 PM
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 34~362
3.7
351
Chemical bonding and trends in C Is and O Is binding energies
The ls and 0 ls binding energy dependence on the first ionization potential of free alkali
metals for carbonates and hydrocarbonates is given in Fig. 9.
All O ls values consistently decrease with the decreasing 11. This effect is caused
by increasing delocalisation of valence electron density from the atom M to the oxygen with decreasing I1. Due to the mutual screening, core O ls levels are subjected to
destabilisation. This decreases their BE in accordance with an increase in their energy.
BE of O ls in solid CO2 (536.3 eV) [5] displays tile same trend. In tile case of
carbon dioxide, an effective negative charge at oxygen atoms is determined by valence
electron density displacement from carbon only, while in the carbonates this is associated
with an additional delocalisation along the M--+O bonds, increasing the negative charge
significantly.
The %rrnation of ~external" metal-carbonate bonds with considerable ionic character
influences the C--+O electron density displacement within the anion. The lower I~ of alkali
metal, the more negative the charge of oxygen, which in turn shifts the electron density
back to carbon decreasing its C ls BE value. This is clearly illustrated in Fig. 9.
The effect is diminished %r the corresponding hydrocarbonates where one metal atom
is replaced by the exceptionally small hydrogen atom. This H-O-C bond introduces an
additional ~constant" delocalisation of valence electron density. In this case, the hydrocarbonate anion undergoes alkali metal attack through one oxygen atom only.
Considering tile first ionisation potentials of H (13.595 eV), O (13.614 eV) and C
(11.256 eV) [4], we can suppose that formation of carbon dioxide will decrease tile I~ of
oxygen making it less than one %r hydrogen. In the case of hypothetical carbonic acid
%rrnation, electron density would be delocalised along the O--+H bonds causing a small
negative charge at the hydrogen atom. Such a structure cannot exist due to the huge
difference in corresponding electron affinities. Moreover, the formation of H2CO3 was not
confirmed by XPS even in the case of solid CO2/H20 stoichiometric mixture under liquid
nitrogen temperature [6].
4
Conclusions
The reference XPS data obtained in this study may be useful for identification of Group
IA carbonates forming in different industrial and natural processes.
Using the sample-precooling technique developed here, the surface chemistry of the
carbonate powders was elucidated. Two types of surface chemical reactions with atmospheric water occur, resulting in hydrocarbonate and/or carbonate hydrate phases, and
a specific behaviour of K2COa was noted.
The influence of grinding (mechanical activation) of the carbonates on the %rrnation
of surface hydrocarbonates phase was %und. This effect increases down the Group IA
K<Rb<<Cs and should be taken into account in sample preparation and treatment.
Unauthenticated
XPS spectra of C ls and O ls core levels and their chemical
shifts
solid 6:58
M2CO3
Download
Datein| 6/15/17
PM and
352
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
MHCO3 are in agreement with basic concepts of valence electron density distribution in
chemical bonds for the investigated compounds.
References
[1]
F.A. Cotton and G. Wilkinson: Advanced Inorganic Chemistry, Fifth Edition, John
Wiley &; Sons, 1988.
[2] NIST X-Ray Photoelectron Spectroscopy Database, NIST SRD 20, Version 3.2
(http://srdata.nist.gov/xps/).
[3] J. Chastain and R.G. King, Jr. (Ed.):
Handbook on X-Ray Photoelectron
Spectroscopy, Physical Electronics, Inc., 1995.
[4] Handbook of Chemistry and Physics, 59th Ed., C R C Press, Inc., 1979.
[5] T.R. Dillingham, D.M. Cornelison, K. Galle et al.: "A Study of Solid CO2 by XPS",
Surface Science Spectra, Vol. 4, No. 2, (1997), pp. 15~160.
[6] T.R. Dillingham and D.M Cornelison: "A Study of Solid CO2/H20 and CO2/CHaOH
by XPS", Surface Science Spectra, Vol. 6, N 2 (1999), pp. 146-152.
Unauthenticated
Download Date | 6/15/17 6:58 PM
A . V . S h c h u k a r e v , D . V . K o r o l k o v / C e n t r a l E u r o p e a n J o u r n a l of C h e m i s t r y 2(2) 2004
8
o
347 362
O
8
o
o].
o
8
o
,d
d
,..@
S
9
8
O
,..@
44
2
,z=
8
o
9
c~
~
~
c~
L)
o].
2
c5
9
0
a.
c~
o
~
G"
LO
o
C~
LO
LO
G"
G"
o
z
~
z
9
8
Unauthenticated
Download Date | 6/15/17 6:58 PM
353
354
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Fig. 1 C ls and O ls spectra of Li2COa 9 upper
warming.
taken under liquid nitrogen, bottom
after
Unauthenticated
Download Date | 6/15/17 6:58 PM
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Fig. 2 C ls and O ls spectra of Na2COa 9 upper
warming.
taken under liquid nitrogen, bottom
355
after
Unauthenticated
Download Date | 6/15/17 6:58 PM
356
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Fig. 3 C ls and O ls spectra of NaHCOa 9 upper
warming.
taken under liquid nitrogen, bottom
after
Unauthenticated
Download Date | 6/15/17 6:58 PM
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Fig. 4 C ls and O ls spectra of K2COa 9 upper
warming.
taken under liquid nitrogen, bottom
357
after
Unauthenticated
Download Date | 6/15/17 6:58 PM
358
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Fig. 5 C ls and O ls spectra of KHCOa 9 upper
warming.
taken under liquid nitrogen, bottom
after
Unauthenticated
Download Date | 6/15/17 6:58 PM
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Fig. 6 C ls and O ls spectra of Rb2COa 9 upper
warming.
taken under liquid nitrogen, bottom
359
after
Unauthenticated
Download Date | 6/15/17 6:58 PM
360
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Fig. 7 C ls and O ls spectra of Cs2COa 9 upper
warming.
taken under liquid nitrogen, bottom
after
Unauthenticated
Download Date | 6/15/17 6:58 PM
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
Fig. 8 C ls and 0 ls spectra of Cs2C03 after grinding: upper
bottom after warming.
taken under liquid nitrogen,
Unauthenticated
Download Date | 6/15/17 6:58 PM
361
362
A.V. Shchukarev, D.V. Korolkov / Central European Journal of Chemistry 2(2) 2004 347 362
290.5
Li
A
290.0
289.5
m 289.0
288.5
R'~~C
s
288.0
287.5
5.50
I
I
I
5.00
4.50
4.00
3.50
I1, eV
532.0
Li
B
531.5
531.0
"~ 530.5
K
530.0
529.5
5.50
,
,
,
5.00
4.50
4.00
3.50
ll, e V
Fig. 9 C ls (A) and O ls (B) BE dependence on first ionisation potential of alkali metals: o
carbonates, 9 hydrocarbonates.
Unauthenticated
Download Date | 6/15/17 6:58 PM