ARTICLE

pubs.acs.org/EF

Effect of Additives on Decomposition of Sodium Carbonate:

Precombustion CO2 Capture Sorbent Regeneration
Ranjani V. Siriwardane,*,† James A. Poston, Jr.,† Clark Robinson,†,‡ and Thomas Simonyi†,‡
†
National Energy Technology Laboratory, United States Department of Energy, 3610 Collins Ferry Road, Post Office Box 880,
Morgantown, West Virginia 26507-0880, United States
‡
Parsons Infrastructure and Technology Group, Post Office Box 618, Pittsburgh, Pennsylvania 15129, United States

ABSTRACT: The effect of various additives on the decomposition of sodium carbonate (Na2CO3) was evaluated using
temperature-programmed desorption, thermogravimetric analysis, X-ray diffraction, and X-ray photoelectron spectroscopy.
Incorporation of additives, CaO and Ca(OH)2, had a significant effect on lowering the decomposition temperature of Na2CO3,
while CaCO3, SiO2, and Al2O3 had no effect. The amount of additive, sweep gas flow rate, and heating rate were also found to play a
significant role in altering the decomposition temperature of Na2CO3. The formation of a carbonate-type intermediate in the
presence of CaO and Ca(OH)2 may have promoted the decomposition of Na2CO3.

’ INTRODUCTION focused on understanding the effect of additives on the decom-

The world’s energy supply is presently dominated by fossil position of sodium carbonate; the results of which are described
fuels, with coal, petroleum, and natural gas supplying roughly in this paper.
80% of the global primary energy demand.1 One of the many Several studies on the additive effect on carbonate decom-
environmental detriments resulting from the extraction, produc- position have been reported. Doping effects of altervalent cations
tion, and use of fossils fuels that has recently come to the on the thermal decomposition and electrical conduction proper-
forefront is the generation of greenhouse gases, with the com- ties of manganese(II) carbonate (MnCO3) have been studied by
bustion of fossil fuels being a major source of carbon dioxide Hassan et al.8 They observed that doping MnCO3II carbonate
(CO2).2 Until the advent of advanced sustainable energy sources, with Liþ and Al3þ ions enhanced the decomposition of the
carbonate. The authors explained that this enhanced effect is
technologies are needed that will allow use of fossil fuels while
due to the generation of hole defects, which are concentrated at
reducing greenhouse gas emissions. Existing commercial CO2
the reaction interface. The doping effect of Liþ is more pro-
capture technology is very expensive and energy-intensive.
nounced than that of the Al3þ ions.
Improved technologies for CO2 capture are necessary to achieve Wigmans et al.11 conducted a temperature-programmed de-
low energy penalties and reduce costs. Pressure swing adsorption sorption (TPD) study on carbonate decomposition in the
(PSA)/sorption (PSS) and temperature swing adsorption presence of carbon. They observed enhancement of sodium
(TSA)/sorption (TSS) are some of the potential techniques carbonate (Na2CO3) decomposition by carbon. They indicated
that could be applicable for removal of CO2 from gas streams.3,4 that the mechanism includes decomposition to Na2O and
According to a system analysis conducted in the Netherlands,5 subsequent reduction to Na metal. Addoun et al.12 studied the
the PSA/TSA systems would be even more energy-efficient for effect of alkali carbonate decomposition on the porosity of
integrated gasification combined cycle (IGCC) systems if the carbons. They observed that the nature of the cation affected
sorbents were operational at warm gas temperatures (250- the degree of microporosity in the following order: Li < Na < K <
350
C; 0
C = 273.15 K). Various alkali oxides and salts are Rb < Cs. The highest effects observed with Cs, Rb, and K are due
potential candidates for the CO2 capture process at these to the decomposition of Cs, Rb, and K carbonates on the carbon
temperatures.6-8 National Energy Technology Laboratory surface at a rather low temperature (T ∼ 227
C or 500 K)
(NETL) researchers have developed NaOH- and Mg(OH)2- because of the formation of alkaline oxide clusters and a high
based sorbents to remove CO2 at warm gas temperatures.9,10 degree of dispersion of the alkaline compound.
Removal of CO2 with alkali oxides and salts involves a chemical Kim et al. studied the effect of SiO2 on the decomposition of
reaction that results in the formation of alkali carbonates. Na2CO3.13 They observed that the addition of SiO2 greatly
Regeneration of these materials requires high temperatures enhanced the decomposition of Na2CO3. The initial decomposi-
because it involves the decomposition of the alkali carbonates tion product was identified as Na2SiO3, irrespective of the mixing
formed during absorption. The high regeneration energies of ratio of Na2CO3/SiO2. Said et al. studied the influence of iron
these sorbents would result in significant loss in system efficiency. additions on the thermal decomposition of zinc carbonate.14
Therefore, it is necessary to find a low energy regeneration
procedure to be able to use CO2 removal sorbents effectively. Received: November 3, 2010
Incorporation of various additives may result in the decomposi- Revised: January 18, 2011
tion of alkali carbonate at lower temperatures. This research Published: February 18, 2011

This article not subject to U.S. Copyright.

Published 2011 by the American Chemical Society 1284 dx.doi.org/10.1021/ef101486m | Energy Fuels 2011, 25, 1284–1293
Energy & Fuels ARTICLE

Figure 1. Effect of Na2CO3/CaO system composition on the thermal decomposition of Na2CO3.

They observed that the presence of iron retarded the decom- carried out with a CaO sample from ambient temperature (21
C) to
position. The retardation effect increased with an increasing ion 400
C and with a Na2CO3 sample from ambient temperature (21
C)
concentration up to 50 atomic percent. The authors indicated to 650
C.
that the presence of foreign ions causes lattice deformation, as X-ray photoelectron spectroscopy (XPS) was carried out using a
well as either a change in the valence of ions or the appearance Physical Electronics (PHI) model 3057 X-ray photoelectron spectros-
of additional cationic or anionic vacancies in the lattice, which, copy subsystem with a spherical capacitance analyzer and a model 04-
in turn, strongly affects the thermal decomposition rate. The 548 X-ray source. The system consists of separately pumped preparation
authors also indicated that the introduction of iron ions into and detection chambers that are routinely operated within the pressure
ZnO lattice positions during thermoanalysis decreased the hole range of 10-8-10-10 Torr (from 1.3  10-6 to 1.3  10-8 Pa). The
concentration either on the surface or in the bulk. However, system was calibrated in accordance with PHI procedures,15 with
studies on the effect of sodium carbonate decomposition photoemission lines Eb of Cu 2p3/2 = 932.7 eV and Eb of Au 4f7/2 =
with additives used in present work are not reported in the 84 eV, using an aluminum anode. All reported intensities are experi-
mentally determined peak areas divided by the instrumental sensitivity
literature.
factors. The binding energies were corrected for sample charging using
the reference binding energy (BE) of adventitious C (1s) as 284.6 eV.15
’ EXPERIMENTAL SECTION The spectrometer sample holder is resistively heated and capable of
TPD studies were conducted in a lab-scale fixed-bed flow reactor withstanding temperatures up to 600
C. The sample preparation
(Micromeritics Autochem 2910 atmospheric flow reactor) at 14.7 psi chamber is separated from the main analysis chamber by a gate valve.
(∼1.01  105 Pa). The outlet gas stream was analyzed using a Pfeiffer The samples were placed in the preparation chamber, evacuated, and
Vacuum Thermostar mass spectrometer. In the TPD studies, N2 was then heated. After the preparation chamber reached the desired pres-
introduced to the samples (1.7  10-3 L) at a flow rate of 10 cm3/min sure, the sample was transferred to the analyzing chamber for data
and the temperature was increased at a rate of 10
C/min while acquisition. The sample was maintained at the desired temperature
measuring the CO2 concentration with the mass spectrometer. For throughout the experiment, including data acquisition. The data were
comparison, in some TPD studies, a nitrogen flow rate of 50 cm3/min analyzed with analysis software developed by RBD Instruments.
and heating rates of 5 and 25
C/min were used. The chemicals used in these series of experiments were obtained from
Thermogravimetric analysis (TGA) experiments were conducted Alfa Asear (Johnson Matthey). The sodium carbonate used in the TPD,
on a TA Instruments model 2050 thermogravimetric analyzer. Samples TGA, and XRD studies was monohydrated (Na2CO3 3 H2O). The
(90 mg) were heated to 850
C under N2, and weight gain was recorded calcium oxide and calcium carbonate used were anhydrous. The sodium
isothermally as a function of time. N2 was introduced at a flow rate of carbonate used in the XPS studies was also anhydrous.
90 cm3/min, and the temperature was increased at a rate of 10
C/min
up to 850
C and held isothermally for 120 min.
High-temperature in situ X-ray diffraction (XRD) studies were carried ’ RESULTS
out using an Anton Paar HTK 1200N high-temperature stage connected
to an Anton Paar TCU-1000N temperature control unit. The high- TPD. The measured CO2 concentration and corresponding
temperature stage was interfaced to a Panalytical PW 3040 X-Pert Pro temperature data as a function of time during TPD measure-
XRD system equipped with a 50 kV PW 3373/00 Cu LFF high-power ments of Na2CO3 3 H2O/CaO mixtures with varying composi-
ceramic tube with a Cu anode and a PW 3011/20 detector. The X-ray tions are shown in Figure 1. For pure Na2CO3 3 H2O, the increase
wavelength used was Cu KR1 at 1.540 56 Å. System calibration was in the CO2 concentration initiates around 780
C and the CO2
carried out using a polysilicon-pressed disk with Si Æ111æ referenced to maximum occurs around 850
C, with a narrow CO2 peak. When
28.443
(2θ). Sample data were acquired at 45 kV and 40 mA in a line Na2CO3 3 H2O was mixed with CaO in the ratio of 1:3 Na2
focus mode. The in situ high-temperature studies were conducted in CO3 3 H2O/CaO, the increase in the CO2 concentration initiated
ambient air, with data analysis carried out using X’Pert Highscore Plus at around 550
C, a temperature that is significantly lower than
software supplied by Panalytical. For comparative purposes, in addition that of pure Na2CO3 3 H2O. The CO2 curve was also broader
to the Na2CO3-CaO sample, in situ temperature studies were also than that of pure Na2CO3 3 H2O, indicating not only that CO2

1285 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293

Energy & Fuels ARTICLE

Figure 2. Effect of the flow rate on Na2CO3 3 H2O/CaO (1:4) thermal decomposition.

Figure 3. Effect of the heating rate on Na2CO3 3 H2O/CaO (1:4) thermal decomposition.

decomposition initiates around 550
C but also that it continues 850
C remained essentially unchanged. The CO2 concentration
at a wide range of temperatures. When the Na2CO3 3 H2O of the secondary peak was found to increase. A small temperature
concentration was further decreased to 1:4 Na2CO3 3 H2O, difference between the low-temperature peaks was also observed.
CO2 was initially observed around 520
C because of the This observation suggests that higher flow rates may facilitate a
decomposition of carbonate, with more CO2 being observed in lower carbonate decomposition temperature through the in-
the 520
C temperature range than that observed with the creased removal of CO2 from the gas stream. The two peaks
Na2CO3 3 H2O/CaO (1:3) sample. The CO2 curve for observed at each flow rate represent two types of absorption sites
the Na2CO3 3 H2O/CaO (1:6) sample was similar to that of from which CO2 production is thermodynamically favorable. As
the 1:4 samples. For these three Na2CO3/CaO systems, an observed from Figure 2, the flow rate has little effect on the
increase in the CaO concentration in the Na2CO3 3 H2O/CaO desorption peak at 850
C but does have a significant effect on
mixture results in a decrease in the decomposition temperature of the lower temperature desorption peak.
Na2CO3, with an optimum CaO concentration beyond which The effect of the heating rate on the decomposition of
the decomposition temperature of Na2CO3 is not affected. Na2CO3 3 H2O/CaO (1:4) is shown in Figure 3. The initial
The effect of the N2 flow rate on the decomposition of decomposition temperature appears to be similar for all heating
Na2CO3 3 H2O/CaO (1:4) is shown in Figure 2. At 10 cm3/ rates; however, the shape of the CO2 concentration curve was
min flow rate, the CO2 decomposition initiates around 550
C, affected by the heating rate. A sharp curve was observed at a
with a secondary peak observed at 718
C and a maximum peak heating rate of 25
C/min, with a peak maximum at 850
C.
observed around 850
C. At a flow rate of 50 cm3/min N2, two When the heating rate was decreased to 10
C/min, a secondary
distinct desorption peaks were observed. The decomposition peak appeared around 725
C. When the heating rate was further
temperature and intensity of the peak observed initially at decreased to 5
C/min, the secondary peak appeared at a
1286 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293
Energy & Fuels ARTICLE

Figure 4. Effect of Na2CO3 3 H2O/Ca(OH)2 composition on Na2CO3 thermal decomposition.

Figure 5. Effect of CaCO3 addition on Na2CO3 thermal decomposition.

temperature of 625
C. Subsequently, a lower heating rate more effective than the addition of CaO in lowering the decom-
appears to facilitate Na2CO3 decomposition at a lower tempe- position temperature of Na2CO3.
rature. When CaCO3 was added to Na2CO3 3 H2O (Figure 5), the
Because the addition of CaO decreased the decomposition temperature of decomposition did not initiate at 550
C, as
temperature of Na2CO3, the effect of the addition of Ca(OH)2 was observed with CaO and Ca(OH)2. Similarly, when Na2
was also investigated. The CO2 decomposition curves for various CO3 3 H2O was mixed with Al2O3 and SiO2, lower decomposi-
Na2CO3 3 H2O/Ca(OH)2 compositions are shown in Figure 4. tion temperatures were not observed, as shown in Figure 6. Thus,
After the addition of Ca(OH)2, the initial decomposition tem- CaO and Ca(OH)2 have a unique ability to affect the decom-
perature of Na2CO3 decreased from 785 to 550
C, similar to the position temperature of Na2CO3.
observations with CaO. However, the shapes of the CO2 curves TGA. Thermogravimetric data for the decomposition of pure
with Ca(OH)2 during decomposition of the carbonate were Na2CO3 3 1H2O, pure CaO, and a mixture of Na2CO3 3 H2O/
significantly different from those of Na2CO3 and Na2CO3/CaO. CaO (1:3) are summarized in Table 1. When pure CaO was
Unlike the TPD curve for pure Na2CO3 3 H2O, which had a sharp heated to 800
C, about 4.6 wt % was lost between 288 and
maximum at 850
C, the Na2CO3 3 H2O/Ca(OH)2 (ratio 1:1) 405
C and 1.2 wt % was lost between 405 and 657
C, believed
CO2 peak maximum still appeared at 850
C, but the peak was to be due to the loss of H2O. When pure Na2CO3 3 H2O was
broad. With a Na2CO3 3 H2O/Ca(OH)2 ratio of 1:3, a very broad heated to 800
C, about 14.8 wt % was lost between 62 and
maximum between 725 and 850
C was observed. When the 163
C, which is close to the calculated amount of water from the
Na2CO3 3 H2O/Ca(OH)2 ratio was increased to 1:4, the max- formula (14.5 wt %). No significant weight was lost after 162
C.
imum appeared at 750
C, with the peak being narrower than that This observation indicates that carbonate decomposition takes
for the 1:3 ratio. Thus, the addition of Ca(OH)2 appears to be place above 800
C.

1287 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293

Energy & Fuels ARTICLE

Figure 6. Effect of (a) SiO2 and (b) Al2O3 addition on Na2CO3 decomposition.

Table 1. Weight Loss as a Function of the Temperature (
C) of CaO, Na2(CO3) 3 H2O, and Selected Na2(CO3) 3 H2O/CaO
Systems
compound temperature (
C) Δwt (%) temperature (
C) Δwt (%) temperature (
C) Δwt (%) temperature (
C) Δwt (%)

CaO 288-405 4.6 405-657 1.2 >657 0

Na2CO3 3 H2O 062-163 14.8 162-800 0.3
Na2CO3 3 H2O/CaO 033-337 2.5 337-439 15.3 439-652 5.1 652-800 2.6
ratio 1:3
Na2CO3 3 H2O/CaO 045-344 2.6 344-436 15.1 436-700 4.5 700 1.5
ratio 1:3

For the Na2CO3 3 H2O/CaO (1:3) system, the initial weight The total weight loss was 25.5% up to 800
C and 23.7% up to
loss was 2.5% at temperatures between 33 and 337
C. A 700
C during repeated tests. The theoretical weight loss
repeated test produced results of 2.6% initial weight loss at combined with both H2O loss and carbonate decomposition
temperatures between 45 and 344
C. This weight loss is believed calculated from the formula was about 26.7%, a value close to that
to be due to the loss of water (H2O). The percent of H2O of the experimentally determined value. This indicates that, for
calculated from the formula is about 6 wt %, indicating that all Na2CO3 3 H2O/CaO, the carbonate decomposition takes place
H2O was not removed fully at 337/344
C. For the Na2 below 800
C, which is different from that of pure Na2-
CO3 3 H2O/CaO (1:3) system, the major weight loss was ob- CO3 3 H2O. These data are consistent with the TPD and XRD
served to be between the temperatures 337 and 439
C during data that will be discussed later.
repeated tests, with an approximate weight loss of 15%, indicat- XRD. Analysis of the CaO in situ high-temperature XRD data
ing that additional H2O loss and carbonate decomposition may showed that the sample is composed of the CaO phase from 21 to
have taken place in this temperature range. It appears that 400
C, with the presence of Ca(OH)2 appearing at 150
C and
carbonate decomposition initiates around 337-439
C with disappearing at 400
C. Analysis of the Na2CO3 standard at
the Na2CO3 3 H2O/CaO (1:3) system. It is difficult to determine 21
C revealed the sample to be Na2CO3 3 H2O. The peak shape,
the exact starting temperature of carbonate decomposition, full width at half-maximum (fwhm), and poor reference fit
because some H2O loss may have occurred in that temperature indicated that the material was not entirely crystalline. The
range. The weight loss as a result of carbonate decomposition presence of Na2CO3 was observed upon heating the sample to
continued from 439 to 652
C and from 436 to 700
C, 100
C and continued to be observed through 300
C. From 350
respectively, followed by slow decomposition after 652
C up to 650
C, it was difficult to observe the presence of Na2CO3
to 800
C for the test. During the repeated test, the time at because of the poor reference match for the main diffraction
800
C was extended, which resulted in additional weight loss. patterns. The peak shape, fwhm, and poor reference fit indicated
1288 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293
Energy & Fuels ARTICLE

Figure 7. In situ high-temperature XRD data from the Na2CO3 3 H2O/CaO (1:3) system.

Figure 8. O 1s region of Na2CO3 during heating.

that the sample was disordered and not entirely crystalline. The observed was CaO, although the reference fit was poor, indicat-
presence of an oxide was not observed. ing that the sample was disordered. At no temperature (21-
XRD in situ temperature studies on the Na2CO3 3 H2O/CaO 800
C) was the formation of any crystalline Na-Ca-type
(1:3 ratio) sample were carried out from ambient temperature compound observed.
(21
C) to 800
C (Figure 7). Analysis of the Na2CO3 3 H2O/ XPS Analysis. In situ high-temperature XPS analysis was
CaO (1:3) sample at ambient temperature showed the sample to conducted on Na2CO3 and a mixture (1:3) of Na2CO3/CaO.
be composed primarily of CaO and Ca(OH)2 with a small The sample was heated in the sample preparation chamber to the
concentration of Na2CO3 3 H2O. The definitive presence of desired temperature, with a standing pressure of 10-6-10-7
Na2CO3 was not observed at any temperature ranging from Torr, and was transferred to the analysis chamber for data
100 to 800
C, indicating that the Na2(CO3) material was acquisition while maintaining temperature.
disordered and not entirely crystalline. No presence of any other The O 1s spectral region for Na2CO3 (Figure 8) contained
Na-based compound could be identified from 300 to 800
C. one distinct peak centered at 530.5 eV. The BE value of the O 1s
The formation and presence of CaCO3 was observed from 300 peak centered at 530.5 eV is considerably lower than that
to 600
C. At 700 and 800
C, the only Ca-based compound reported in the literature for Na2CO3,16-18 suggesting that the

1289 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293

Energy & Fuels ARTICLE

Figure 9. C 1s region of the Na2CO3/CaO (1:3) system during heating.

surface composition of the Na2(CO3) standard may have been The BE value for the corresponding Na 1s peak (figure not
non-stoichiometric or in a hydroxide (OH-) form. The O 1s BE shown) at ambient temperature was 1070 eV, a value less than
value increased with an increased temperature approaching a that reported in the literature for Na2(CO3).16,17 This BE value is
value of 531.8 ( 0.1 eV, which corresponds to the oxygen in consistent with observations of the C 1s and O 1s spectral regions
Na2(CO3), between the temperature range of 398 and and further indicates that the surface composition is non-
511
C.16-18 At 533
C, a decrease in the O 1s BE value to stoichiometric. The BE value of the Na 1s peak continued to
530.3 eV was observed. The value of 530.3 eV is slightly higher increase with an increasing temperature up to 398
C, where the
than that reported in the literature (529.7 eV) for Na2O.18,19 The Na 1s BE value approached that reported in the literature for
presence of a small peak corresponding to oxygen of carbonate Na2(CO3).16,17,20 This temperature is higher than the tempera-
indicated that complete decomposition had not taken place. This ture observed in the C 1s and O 1s spectral regions (326
C),
was also confirmed by the continued presence of the C 1s where the corresponding BE values were in the range reported in
carbonate peak (not shown). the literature for Na2(CO3).16-18 This temperature difference is
The C 1s spectral region of Na2(CO3) at ambient tempera- not fully understood but may be related to surface rearrange-
ture, consisted of two distinct peaks (not shown) centered at ment. At 446
C, the Na 1s BE value increased to 1071.9 eV, with
289 and 284.6 eV. The C 1s BE value of 289 eV is lower than that the BE value remaining at 1071.9 ( 0.1 eV throughout the
reported in the literature for Na2(CO3).16,17 A similar tempera- temperature at 535
C, the maximum temperature.
ture correlation was observed in the C 1s region as that observed Analysis of the Na2CO3/CaO mixture (1:3) at ambient
in the O 1s region, with the C 1s carbonate peak approaching BE temperature revealed two distinct peaks in the C 1s region
values consistent with those reported in the literature for (Figure 9) centered at BE values of 284.6 and 289.5 eV; values
carbonate at 326
C during Na2(CO3) decomposition.16,17 At correlating to ubiquitous carbon and carbonate carbon.15-17 The
temperatures below 326
C, the carbonate peak was the carbonate carbon peak was the dominant spectral peak. When
dominant peak with a ubiquitous carbon/carbonate carbon ratio the Na2(CO3)/CaO mixture was heated, no significant changes
(Cu/Cc) of approximately 0.5 and 0.4 for Na2(CO3) at ambient were observed until 200
C, where a slight decrease in the
and 204
C, respectively. At 326
C, the Cu/Cc ratio was carbonate C 1s spectral peak BE value to 289.2 eV was observed.
approximately 0.8, indicating that the peak intensities were The carbonate carbon peak in the C 1s region remained the
approaching equality. At 398
C, the Cu/Cc ratio was approxi- dominant spectral peak. At 300
C, the initial intensities of the
mately 1.4, indicating that the carbonate peak was no longer the ubiquitous carbon and carbonate carbon peaks were approxi-
dominant C 1s peak and that decomposition was underway. An mately equal, but with increased time, the intensity of the
increase in the carbonate BE value to 289.9 eV, a value slightly carbonate peak continued to decrease, making the ubiquitous
higher than that reported for Na2(CO3), was observed.16,17 The carbon peak the dominant peak in the C 1s region, as shown in
carbonate peak intensity continued to decrease with increased Figure 9. At 350
C, the maximum temperature, the C 1s
temperature, with corresponding BE values of approximately carbonate peak further decreased in intensity, indicating further
290.1 ( 0.1 eV in the temperature range of 446-511
C. At decomposition, as well as a shift in BE to approximately 290.2 eV,
511
C, the presence of the C 1s ubiquitous peak was not indicating a non-stoichiometric surface.
observed. At 535
C, the maximum temperature, the C 1s In the O 1s region at ambient temperature, a single slightly
carbonate BE value increased to 290.6 eV, with the peak broad and asymmetrical peak with a BE value of 531 eV was
intensity being very low. observed, as shown in Figure 10. Spectral deconvolution

1290 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293

Energy & Fuels ARTICLE

Figure 10. O 1s region of the Na2CO3/CaO (1:3) system during heating.

Figure 11. Ca 2p region of the Na2CO3/CaO (1:3) system during heating.

indicates that a secondary peak may be present at approximately Na2O. The broadness and asymmetry of the O 1s peak continued
531.8 eV. A BE value of 531 eV is close to the value reported in to increase at 350
C.
the literature for CaO, and a BE value of 531.8 eV is close to the Observation of the Ca 2p spectral region revealed two distinct
value of either CaCO3 or Na2(CO3).16-18 The asymmetry and peaks (Figure 11) in the Ca 2p region with BE values of 346.8 and
broadening of the O 1s peak, as shown in Figure 10, continued 350.1 eV (Δ 3.4 eV), values close to that reported for
with both increasing time and temperature. At 300
C, decon- CaO.15,16,20-22 When the Na2(CO3)/CaO mixture was heated,
volution indicates the formation of an additional peak at 529.6 changes observed in the Ca 2p region with temperature were
eV, a value that correlates to values reported in the literature for different from those observed in the C 1s and O 1s regions, with
Na2O.18,19 Increased time resulted in the increased formation of significant changes not being observed until 300
C, as shown in

1291 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293

Energy & Fuels ARTICLE

Figure 11. At 300
C, both a peak shift toward higher BE and an at a lower temperature, as opposed to heating rates of 10 and
asymmetry on the lower BE side of the Ca 2p3/2 peak were 25
C/min. The lower heating rate results in increased resonance
observed. This shift toward higher BE continued with time, with time at a given temperature range, which facilitates the Na2-
a final BE value of 347.3 eV, a BE value that is slightly higher than (CO3) thermal decomposition process by driving reactions 1 and
that reported in the literature for Ca(CO3).15,23,24 At 350
C, the 2 and 3 and 4 farther to the right than would have been achieved
BE value of the Ca 2p peak further increased to 347.4 eV, a BE under a higher heating rate. A higher sweep gas flow rate was also
value higher than that reported for Ca(CO3). found to facilitate thermal decomposition of Na2(CO3/CaO).
The Na 1s spectral peak (not shown) was of low intensity and The peaks around 700 and 850
C were observed at both flow
broad, with a BE value of 1071.3 eV, which correlates to rates; however, the peaks were more distinct, and the intensity of
Na2(CO3).16,17,20 Aside from a slight increase in intensity at the 700
C peak was greater at the higher flow rate. The two
200
C, the peak intensity decreased with both time and peaks observed at each flow rate represent two types of absorp-
temperature. At 350
C, the Na 1s peak was no longer observed. tion sites where CO2 desorption is favorable. We speculate that
Given the low intensity of the Na 1s peak and the high intensity, the low-temperature peak (∼700
C) represents NaCO3 3 3
asymmetry, and broadness of the O 1s peak, the dominant 3 CaO decomposition, while that of the high-temperature peak
oxygen-containing species on the sample surface were not Na- (850
C) represents NaCO3 decomposition. Faster CO2 re-
based. moval from the product stream at a higher flow rate may have
On the basis of the XPS spectroscopic data of the C 1s and O facilitated the Na2CO3 3 3 3 CaO decomposition process shown
1s region, the decomposition of pure Na2(CO3) initiated around in reactions 1 and 2, which in turn contributed to the higher CO2
398
C, while for the Na2(CO3)/CaO mixture (1:3), the concentration at 700
C.
decomposition initiated at 300
C. This observation indicates
that the presence of CaO lowers the decomposition temperature
of Na2(CO3), which is consistent with the TGA and TPD data. ’ CONCLUSIONS
These decomposition temperatures are significantly lower than The presence of CaO or Ca(OH)2 in Na2CO3 3 H2O lowered
those observed during decomposition at atmospheric pressure, the decomposition temperature of Na2CO3, while the presence
because XPS analysis was conducted in vacuum as opposed to of SiO2, Al2O3, and CaCO3 did not show any effect. Ca(OH)2
atmospheric pressure. Similar observations regarding lower had a more pronounced effect than that of CaO, with the
decomposition temperatures in vacuum have been previously concentration, flow rate of sweep gas, and heating rate all
observed.25,26 affecting the decomposition temperature of Na2CO3. The for-
mation of crystalline phases between CaO and Na2CO3 were not
’ DISCUSSION observed during heating in XRD, but the formation of CaCO3-
type species were observed. XPS studies also confirmed that the
The presence of both CaO and Ca(OH)2 decreased the
presence of CaO lowered the decomposition temperature of
decomposition temperature of Na2(CO3). XRD studies did
Na2CO3, with the formation of Ca-CO3-type species observed
not indicate the presence of any crystalline compounds formed
during heating. The formation of carbonate-type species with
between Na and Ca. However, the formation of CaCO3-type
CaO and Na2CO3 may have facilitated the decomposition of
species was observed from both XPS and XRD at intermediate
Na2CO3.
temperatures prior to carbonate decomposition. Thus, the
decomposition of Na2(CO3) may be enhanced by CaO through
’ AUTHOR INFORMATION
the formation of an intermediate carbonate complex, as shown in
reactions 1 and 2. Corresponding Author
Na2 ðCO3 Þ þ CaO T Na2 3 3 3 CO3 - Ca 3 3 3 O ð1Þ *Telephone: 304-285-4513. Fax: 304-285-4403. E-mail: ranjani.
siriwardane@netl.doe.gov.
Na2 3 3 3 CO3 - Ca 3 3 3 O T Na2 O þ CaO þ CO2 ð2Þ
Ca(OH)2 had an even more pronounced effect on the decom- ’ ACKNOWLEDGMENT
position temperature of Na2(CO3) than that of CaO. The The authors thank Dr. Robert Romanosky for his continued
formation of NaOH may have facilitated the carbonate decom- support throughout the years.
position, as shown in reactions 3 and 4.
Na2 ðCO3 Þ þ CaðOHÞ2 f 2NaOH 3 3 3 CO3 3 3 3 Ca ð3Þ ’ REFERENCES
(1) Goldeemberg, J.; Olende, S. A.; El-Ashry, M.; Davis, G.; Johans-
2NaOH 3 3 3 CO3 3 3 3 Ca f H2 O þ CaO þ CO2 ð4Þ son, T.; Keith, D.; Jinghai, L.; Nakicenovic, N.; Pachauri, R.; Shafie-Pour,
M.; Shpilrain, E.; Socolow, R.; Yamaji, K.; Luguang, Y. Lighting the Way:
When Ca(CO3) was added to Na2(CO3), the decomposition Toward a Sustainable Energy Future, Report; InterAcademy Council
temperature did not change. The presence of carbonate in (IAC): Amsterdam, the Netherlands, Oct 2007.
Ca(CO3) possibly hindered the formation of a Na 3 3 3 (2) Energy Information Administration (EIA). Greenhouse Gases,
carbonate 3 3 3 Ca intermediate. This would indicate that the Climate Change, and Energy; U.S. Department of Energy: Washington,
D.C., 2008; Report DOE/EIA-X012.
Ca 3 3 3 carbonate intermediate is important in facilitating the
(3) Skarstrom, C. W. Method and apparatus for fractionating
decomposition of Na2(CO3). gaseous mixtures by adsorption. U.S. Patent 2,944,627, 1960.
It was also observed that the heating (temperature) rate and (4) Guerrin de Montgareuil, P.; Domine, D. Process for separating a
sweep gas flow rate had an effect on the thermal decomposition binary gaseous mixture by adsorption. U.S. Patent 3,155,468, 1964.
temperature of Na2(CO3)/CaO. A lower heating rate of 5
C/ (5) Van Kessel, L. B. M.; Saeijs, J. C. P. L.; Lalbahadoersing, V.;
min facilitated the faster decomposition of Na2(CO3) initiating Arendsen, A. R. J.; Stavenga, M.; Heesink, A. B. M.; Temmink, H. M. G.

1292 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293

Energy & Fuels ARTICLE

IGCC Power Plant: CO2 Removal with High Temperature Adsorbents, (26) Poston, J. A., Jr.; Siriwardane, R. V.; Fisher, E. P.; Miltz, A. L.
Final Report; TNO: Delft, the Netherlands, April 1998; TNO Rapport Thermal decomposition of the rare earth sulfates of cerium(III) cerium-
R98/135. (IV), lanthanum(III) and samarium(III). Appl. Surf. Sci. 2003, 214
(6) Golden, M. C.; Shivaji, S. Regenerative carbon dioxide (CO2) (104), 83–102.
removal system. U.S. Patent 6,322,612, 2001.
(7) Siriwardane, R. V. Regenerable sorbents for CO2 capture from
moderate and high temperature applications. U.S. Patent 7,314,847 B1,
Jan 2008.
(8) Hassan, E. A.; Katib, S. M. A.; El-Qurashi, M. A. M.; El-Salaam,
K. M. A. The doping effect of altervalent cations on the thermal
decomposition and electrical conduction properties of manganese(II)
carbonate. J. Therm. Anal. Calorim. 1994, 41 (2-3), 337–346.
(9) Siriwardane, R. V.; Robinson, C.; Shen, M.; Simonyi, T. Novel
regenerable sodium-based sorbents for CO2 capture at warm gas
temperatures. Energy Fuels 2007, 21 (4), 2088–2097.
(10) Siriwardane, R. V.; Stevens, R. W., Jr. Novel regenerable
magnesium hydroxide based sorbents for CO2 capture at warm gas
temperatures. Ind. Eng. Chem. Res. 2009, 48 (4), 2135–2141.
(11) Wigmans, T.; Doorn, J. V.; Moulijn, J. A. Temperature-
programmed desorption study of Na2CO3-containing activated carbon.
Fuel 1983, 62, 190–195.
(12) Addoun, A.; Dentzer, J.; Ehrburger, P. Porosity of carbons
obtained by chemical activation: Effect of the nature of the alkaline
carbonates. Carbon 2002, 40 (7), 1140–1143.
(13) Kim, J. W.; Lee, Y. D.; Lee, H. G. Decomposition of Na2CO3 by
interaction with SiO2 in mold flux of steel continuous casting. ISIJ Int.
2001, 41, 116–123.
(14) Said, A. A.; Hassan, E. A.; El-Salaam, K. M. A.; Mohamed, M. M.
Influence of ion additions on the thermal decomposition of basic zinc
carbonate. J. Therm. Anal. Calorim. 1990, 36 (4), 1331–1345.
(15) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D.
Handbook of X-ray Photoelectron Spectroscopy; Physical Electronics, Inc.:
Eden Prairie, MN, 1992.
(16) Gelius, U.; Heden, P. F.; Hedman, J.; Lindberg, B. J.; Manne, R.;
Norberg, R.; Nordling, C.; Siegbahn, K. Molecular spectroscopy by
means of ESCA III. Carbon compounds. Phys. Scr. 1970, 2, 70.
(17) Hammond, J. S.; Holubka, J. W.; Duckie; Devries, J. E. The
application of X-ray photo-electron spectroscopy to a study of interfacial
composition in corrosion-induced paint de-adhesion. Corros. Sci. 1981,
21 (3), 239–253.
(18) Wagner, C. D.; Zatko, D. A.; Raymond, R. H. Use of the oxygen
KLL lines in identification of surface chemical states by electron
spectroscopy for chemical analysis use of the oxygen KLL lines in
identification of surface chemical states by electron spectroscopy for
chemical analysis. Anal. Chem. 1980, 52, 1445.
(19) Barrie, A.; Street, F. J. An auger and X-ray photoelectron
spectroscopic study of sodium metal and sodium oxide. J. Electron
Spectrosc. Relat. Phenom. 1977, 7, 1.
(20) Siriwardane, R. V.; Cook, J. M. Interactions of NO and
SO2 with iron deposited on silica. J. Colloid Interface Sci. 1985, 104,
250.
(21) Van Doveren, H.; Verhoeven, J. A. Th. XPS spectra of Ca, Sr, Ba
and their oxides. J. Electron Spectrosc. Relat. Phenom. 1980, 21 (3),
265.
(22) Franzen, H. F.; Merrick, J.; Umana, M.; Khan, A. S.; Peterson,
D. T.; McCreary, J. R.; Thorn, R. J. XPS spectra and crystalline potentials
in alkaline-earth chalco-genides and hydrides. J. Electron Spectrosc. Relat.
Phenom. 1977, 11, 439.
(23) Landis, W. J.; Martin, J. R. X-ray photoelectron spectroscopy
applied to gold-decorated mineral standards of biological interest. J. Vac.
Sci. Technol., A 1984, 2, 1108.
(24) Christie, A. B.; Lee, J.; Sutherland, I.; Walls, J. M. An XPS study
of ion-induced compositional changes with group II and group IV
compounds. Appl. Surf. Sci. 1983, 15, 224.
(25) Siriwardane, R. V.; Poston, J. A., Jr.; Fisher, E. P.; Shen, M.-S.;
Miltz, A. L. Decomposition of the sulfates of copper, iron(II), iron(III),
nickel, and zinc: XPS, SEM, DRIFTs, XRD, and TGA study. Appl. Surf.
Sci. 1999, 152 (3-4), 219–236.

1293 dx.doi.org/10.1021/ef101486m |Energy Fuels 2011, 25, 1284–1293