Thermochemical capture of carbon dioxide on lithium aluminates (LiAlO 2 and Li5AlO4): A new option for the CO 2 absorption

Article abstract of DOI:10.1021/jp902501v

Lithium aluminates (LiAlO₂ and Li₅AlO₄) were synthesized, characterized, and tested as possible CO₂ captors. LiAlO₂ did not seem to have good qualities for the CO₂ absorption. On the contra

Full text of DOI:10.1021/jp902501v

J. Phys. Chem. A 2009, 113, 6919–6923
6919
Thermochemical Capture of Carbon Dioxide on Lithium Aluminates (LiAlO2 and Li5AlO4):
A New Option for the CO2 Absorption
,‡
´
Tatiana Avalos-Rendo´n, Julio Casa-Madrid, and Heriberto Pfeiffer*
Instituto de InVestigaciones en Materiales, UniVersidad Nacional Auto´noma de Me´xico, Circuito exterior s/n,
Cd. UniVersitaria, Del. Coyoaca´n, CP 04510, Me´xico DF, Mexico
ReceiVed: March 19, 2009; ReVised Manuscript ReceiVed: May 1, 2009
Lithium aluminates (LiAlO2 and Li5AlO4) were synthesized, characterized, and tested as possible CO2 captors.
LiAlO2 did not seem to have good qualities for the CO2 absorption. On the contrary, Li5AlO4 showed excellent
behavior as a possible CO2 captor. Li5AlO4 was thermally analyzed under a CO2 flux dynamically and
isothermically at different temperatures. These results clearly showed that Li5AlO4 is able to absorb CO2 in
a wide temperature range (200-700 °C). Nevertheless, an important sintering effect was observed during the
thermal treatment of the samples, which produced an atypical behavior during the CO2 absorption at low
temperatures. However, at high temperatures, once the lithium diffusion is activated, the sintering effect did
not interfere with the CO2 absorption. Eyring’s model was used to determine the activation enthalpies of the
CO2 absorption (15.6 kJ/mol) and lithium diffusion (52.1 kJ/mol); the last one is the limiting process.
1. Introduction
2LiAlO2 + CO2 f Li2CO3 + Al2O3
2Li5AlO4 + 5CO2 f 5Li2CO3 + Al2O3
(1)
(2)
In the last 10 years different lithium and sodium ceramics
have been tested as possible carbon dioxide (CO2) captors. These
ceramics present a chemisorption reaction with CO2, producing
the respective alkaline carbonate and a residual metal oxide.
For example, lithium metazirconate (Li2ZrO3) produces Li2CO3
and ZrO2 because of the CO2 absorption.1-3 Into this field, the
lithium ceramics most studied until to now are: lithium
zirconates (Li2ZrO3 and Li6Zr2O7), lithium silicates (Li4SiO4 and
Li2SiO3), and, more recently, lithium cuprate (Li2CuO2), lithium
titanate (Li4TiO4), and lithium ferrite (LiFeO2).1-13
The reactions proposed above are presented assuming that
CO2 is absorbed through the same mechanism observed for other
lithium ceramics24-26 in which lithium aluminates react to
produce lithium carbonate in addition to a residual oxide, Al2O3.
The theoretical CO2 absorption capacities for LiAlO2 and
Li5AlO4 are 0.33 and 0.87 gCO /gceramic, respectively. Li5AlO4
2
has a very high theoretical CO2 absorption capacity in com-
parison with that of the other lithium ceramics (Figure 1). In
fact, only lithium oxide (Li2O) possesses a higher theoretical
capacity (1.46 gCO /gLi O) with the disadvantages of its high
Among these materials, two of the most promising lithium
ceramics are Li4SiO4 and Li4TiO4. These ceramics present good
kinetic behaviors, and their theoretical CO2 absorption capacities
2
2
7,10
are the best, 0.73 (Li4SiO4) and 0.63 (Li4TiO4) gCO /gceram
.
2
reactivity and corrosive characteristics.
As can be deduced from previous data, CO2 absorption capacity
depends on the molecular weight of each ceramic. Consequently,
the lighter the ceramic is, the higher the CO2 absorption capacity
would be. On the basis of this idea, lithium aluminates would
present advantages over other lithium ceramics, even silicates,
which were the lighter ceramics tested until now.
Therefore, the aim of this work was to study and demonstrate
whether LiAlO2 and Li5AlO4 are able to capture CO2 by a
mechanism similar to that reported previously for other lithium
ceramics. Special attention was given to Li5AlO4 because of its
high theoretical CO2 absorption capacity, which is mainly caused
by its high Li/Al molar ratio equal to 5, and the fact that
aluminum is a lighter atom than any other element tested, such
as zirconium, copper, or even silicon.
Lithium aluminates (LiAlO2 and Li5AlO4) have been used
for different applications. LiAlO2 has been proposed as a breeder
ceramic into the fusion reactors, a solid electrolyte for lithium
batteries, and ceramic carrier material in the fabrication process
of electrolyte tiles for molten carbonate fuel cells, among other
applications.14-18 However, there is almost no information about
Li5AlO4. There are only a few papers in which Li5AlO4 has
been synthesized and tested for some electrical applications.19-21
Li5AlO4 ceramic presents two different crystal polymorphs
R-Li5AlO4 and ꢀ-Li5AlO4, where both polymorphs have orthor-
hombic crystal structures.22,23
2. Experimental Section
LiAlO2 and Li5AlO4 were synthesized by solid state reaction
using, in both cases, lithium oxide (Li2O, Aldrich) and aluminum
oxide (Al2O3, Aldrich). Initially, reagents were mixed mechani-
cally and then heat-treated under different conditions. Whereas
LiAlO2 powders were thermally treated at 900 °C for 6 h,
Li5AlO4 powders were heated at 900 °C for 24 h with two
intermediate milling processes. To improve the synthesis of both
lithium aluminates, 10 wt % of lithium excess was used because
of the high tendency of lithium to sublimate.24
If lithium aluminates were able to absorb CO2, then following
reactions may occur
X-ray diffraction (XRD) patterns were obtained from a
diffractometer (Bruker AXS, D8 Advance) coupled to a copper
anode X-ray tube. Compounds were identified conventionally
* Corresponding author. Tel: +52 (55) 5622 4627. Fax: +52 (55) 5616
1371. E-mail: pfeiffer@iim.unam.mx.
‡ Member of the American Chemistry Society.
10.1021/jp902501v CCC: $40.75  2009 American Chemical Society
Published on Web 06/02/2009

´
6920 J. Phys. Chem. A, Vol. 113, No. 25, 2009
Avalos-Rendo´n et al.
Figure 1. Comparison of the maximum theoretical CO2 absorption
capacities of different lithium ceramics. Li4SiO4** and Li4SiO4 represent
the CO2 absorption capacities, assuming a total and half lithium
conversion to Li2CO3, respectively.
Figure 3. Scanning electron micrographs of the (A) Li5AlO4 and (B)
LiAlO2 samples. The rectangle inset shows a general view of the sample
powders.
0714 for LiAlO2. Therefore, both ceramics could be considered
to be pure materials, at least at the XRD detection level. Then,
particle size and morphology of the samples were obtained by
SEM. Figure 3 shows the micrographs of Li5AlO4 and LiAlO2.
Li5AlO4 presented a particle size average of 60 µm, where the
particle size was determined by standard procedures (inset of
Figure 3A). These particles seemed to be highly sintered dense
agglomerates, where their surface did not present any kind of
texture (Figure 3A). This morphology can be merely explained
by the high and long thermal treatment, in addition to the high
lithium mobility, which tends to propitiate high sintering levels.
A similar effect could be established for the LiAlO2 sample in
which the average particle size was equal to 50 µm. Perhaps
the only difference observed on LiAlO2 is that these particles
seemed to be slightly less sintered (Figure 3B). Again, it can
be explained by the shorter thermal-treatment time and lower
lithium quantities present in the sample.
Once Li5AlO4 and LiAlO2 were characterized, these materials
were thermally treated under a CO2 flux to analyze whether
they were able to act as CO2 captors. Figure 4 presents the
dynamic thermograms of LiAlO2 and Li5AlO4 into a CO2 flux.
It is more than evident that both ceramics presented very
different behaviors. LiAlO2 practically did not absorb CO2. This
ceramic was able to increase its weight percentage by only 0.13
wt % between 540 and 830 °C. This may be explained by its
high thermal stability and its high dense structure.
Figure 2. XRD patterns of the LiAlO2 and Li5AlO4 powder samples.
by the corresponding Joint Committee Powder Diffraction
Standards (JCPDS) files. We determined the particle size and
morphology by scanning electron microscopy using a Stereoscan
440, Leica-Cambridge microscope. The samples were covered
with gold to avoid the lack of electrical conductivity, and the
particle size was determined using standard procedures. Finally,
different thermal analyses were performed in a Q500HR
equipment from TA Instruments. Initially, a set of samples was
heat-dynamically treated with a heating rate of 5 °C/min from
room temperature to 850 °C in a CO2 flux (Praxair, grade 3.0).
Then, Li5AlO4 sample was tested isothermically at different
temperatures in the same CO2 flux.
However, Li5AlO4 presented a very high CO2 absorption. It
is clear that two different sorption processes took place; the
first one between 200 and 380 °C and a second one between
590 and 750 °C. Although this kind of thermal trend has already
been observed for other lithium ceramic, for example Li2CuO2,9,27
it is not the most common behavior among the lithium ceramics,
3. Results and Discussion
Figure 2 shows the XRD patterns of both lithium aluminates,
Li5AlO4 and LiAlO2. In both cases, the diffraction patterns fitted
to their respective JCPDS files: 70-0432 for Li5AlO4 and 18-

Thermochemical Capture of CO2 on Lithium Aluminates
J. Phys. Chem. A, Vol. 113, No. 25, 2009 6921
Figure 4. Dynamic thermogravimetric curves of Li5AlO4 and LiAlO2
in a CO2 flux. The rectangular inset shows an enlargement of the LiAlO2
thermogram, 530-880 °C.
where the absorption is produced in just one step.4,11 In other
words, the superficial and bulk absorption processes are not
distinguishable on lithium ceramics. Nevertheless, in the
Li2CuO2 case, the whole absorption process was divided into
two thermal steps: Initially, at low temperatures, a superficial
reaction is produced. At this moment, an external lithium
carbonate and residual oxide shell are formed over the surface
of the ceramic particles, inhibiting the CO2 absorption process.
Then, when the temperature is increased and the lithium
diffusion is activated, the reaction continues through the bulk
of the ceramic, completing the CO2 chemisorption.9,27 For
Li5AlO4, a similar behavior could be described. The superficial
reaction takes place between 200 and 380 °C, where the sample
increased its weight by 4.5 wt %. Then, an increment of weight,
equal to 3.8 wt % (8.3 wt % in total), was observed between
380 and 590 °C. In this temperature range, CO2 absorption is
highly controlled by the diffusion process, considering that the
Li2CO3 external shell has already been formed during the first
process. Finally, between 580 and 750 °C, the lithium diffusion
is activated, and the process is completed through the bulk of
the particles. In this case, the weight increased by 47.2 wt %
(55.5 wt % in total). In addition, it can be observed in Figure
4 that at temperatures higher than 780 °C, the CO2 desorption
process began.
To analyze the CO2 absorption process on Li5AlO4 further,
different isothermal experiments were performed (Figure 5). At
the lowest temperature (250 °C), the isothermal showed an
exponential behavior, which had not reached the plateau after
4 h, and it absorbed 3.35 wt %. Then, samples treated at 300
and 350 °C presented the same exponential behavior, increasing
their weights by 3.7 and 4.22 wt %. Then, the sample treated
at 400 °C presented an atypical behavior. Although this sample
showed a fast CO2 absorption at short times (in comparison
with the previous isotherms), the final absorption was smaller
than that observed for the samples heat treated at 300 and 350
°C. The same effect, but more dramatically, was observed for
the sample heat treated at 450 °C, where the sample increased
its weight by only 0.65 wt %. This atypical behavior has been
reported for the CO2 absorption on other alkaline ceramics such
as Na2ZrO3 and Li2CuO2.27,28 This behavior has been associated
with a sintering process produced during the heating of the
samples, which produces an important decrement of the surface
area. This phenomenon is usually observed at low temperatures
because once the diffusion process is activated, sintering and
Figure 5. Isotherms of CO2 absorption on Li5AlO4 at different
temperatures into a flux of CO2. (A) Isotherms obtained between 250
and 450 °C. (B) Isotherms obtained between 450 and 700 °C.
surface area are not preponderant factors on the CO2 absorption.
Therefore, in the isothermal analyses at 500, 550, 600, and 650
°C, the weight gained increased again, absorbing more CO2 as
a function of the temperature (Figure 5B). This means that
although the sintering effect must be produced in those samples,
the lithium diffusion was activated. Finally, the behavior
presented at 700 °C should be pointed out, where the quantity
and rate of the CO2 absorption were very high and fast,
presenting a totally different isothermal shape. In the first 3 min,
the sample absorbed 35.7 wt %. Then, the absorption rate,
between 3 and 30 min, decreased, getting a CO2 maximum total
absorption of 47.8 wt %. After that time, the isotherm reached
the plateau. This means that in the first few minutes, the CO2
absorption efficiency reached 53.7%, and after 30 min, the
maximum efficiency under these thermal conditions, 68.5%, was
obtained.
These results clearly show that the sintering process affects
CO2 absorption on Li5AlO4. Therefore, to corroborate and
eliminate the presence of this effect, a second set of isotherms
was performed. In this case, all of the Li5AlO4 samples were
initially heated to 675 °C at 100 °C/min with a subsequent
isothermal treatment of 60 min. Then, each sample was cooled
to its respective isothermal temperature for the CO2 absorption.
All of these procedures were carried out under an inert
atmosphere (N2). Once the sample reached the corresponding
temperature, the flux gas was switched from N2 to CO2 and the

´
6922 J. Phys. Chem. A, Vol. 113, No. 25, 2009
Avalos-Rendo´n et al.
Figure 6. CO2 absorption isotherm of Li5AlO4 at 350 °C after a
sintering process at 675 °C.
Figure 7. Isotherms of CO2 absorption on Li5AlO4 at different
temperatures (400 to 675 °C) in a flux of CO2 after a sintering process
at 675 °C.
isothermal experiments were performed. As could be expected,
all of the isothermals followed the typical behavior. In other
words, the CO2 absorption on Li5AlO4 increased as a function
of the temperature.
TABLE 1: Kinetic Parameters Obtained from Li5AlO4
Isotherms Fitted to Exponential Models
temp (°C)
k1 (1/s)
k2 (1/s)
The isotherms performed at the lowest temperatures (300,
350, and 400 °C) clearly showed the presence of two different
processes (Figure 6). Initially, during the first 3000 s, CO2 is
absorbed only over the surface. Then, at larger times, the
absorption occurred on the bulk, although the total quantity of
CO2 absorbed over time is very poor. In fact, the only difference
among these three samples is that the total CO2 absorbed slightly
increased progressively as a function of the temperature (0.15,
0.16, and 0.57 wt % for 300, 350, and 400 °C, respectively).
The behavior and the quantity of CO2 absorbed are very different
than those observed when the samples were not previously
sintered. In these cases, as Li5AlO4 powders were sintered, their
surface area must have dramatically decreased, reducing the
possibility of CO2 reacting with the surface, and because the
lithium diffusion is very slow at those temperatures, the final
CO2 absorbed decreased.
However, all other isotherms performed at higher tempera-
tures (450, 500, 550, 600, 650, and 675 °C) presented an
exponential typical behavior. In these cases, samples increased
their weights from 0.88 (450 °C) to 34.2 wt % (675 °C) after
4 h. These results are presented in Figure 7. Additionally, it is
clearly evident that CO2 absorption at short times was dramati-
cally increased as a function of the temperature. However, the
isothermal slopes became almost the same in long time periods.
It can be explained as follows: in short time periods, the increase
in weight depends on the CO2 absorption reaction, whereas in
long time periods, it depends on the lithium diffusion.
To prove these qualitative observations, we fitted all isotherms
to simple (300-400 °C) or double (450-675 °C) exponential
models because there are two different processes taking place:
the CO2 absorption and the lithium diffusion. In the first three
samples, the isotherms were divided and adjusted to a simple
exponential model, whereas in the other cases, a double
exponential model was used because the reaction and diffusion
processes were undistinguished. From these results, it was
possible to obtain two different constant values, k1 and k2, which
represent the kinetic constant values of the CO2 absorption and
lithium diffusion, respectively. Table 1 shows the different
parameters obtained from the isotherms. From these data, it can
be seen that CO2 absorption constants (k1) are at least one order
of magnitude higher than lithium diffusion constants (k2). In
300a
350a
400a
450
500
550
600
650
675
0.00073
0.00097
0.00092
0.00116
0.00175
0.00153
0.00276
0.00257
0.00736
6.7802 × 10-6
5.2696 × 10-6
1 × 10-5
2 × 10-5
3 × 10-5
8 × 10-5
2.7 × 10-4
2.7 × 10-4
6.8 × 10-4
a Isotherms fragmented and fitted to a simple exponential model.
All other isotherms were fitted to a double exponential model.
other words, the limiting step of the total process is the lithium
diffusion. Additionally, CO2 absorption and diffusion constant
values were enhanced as a function of the temperature, as was
already qualitatively described in Figures 6 and 7. In both cases,
the processes were improved by one (k1) or two (k2) orders of
magnitude, increasing the temperature from 300 to 675 °C.
If both processes, CO2 absorption (k1) and lithium diffusion
(k2), follow a linear trend as a function of temperature, then the
gradients of these best fit lines should fit the Eyring’s model.
These results are illustrated in Figure 8. It is clear that plots of
ln(k/T) versus 1/T describe linear trends, fitting Eyring’s model.
Therefore, the activation enthalpies (∆H) of the two different
processes could be determined. The ∆H values obtained were
15.6 kJ/mol for CO2 absorption and 52.1 kJ/mol for lithium
diffusion. Therefore, this result clearly shows that lithium
diffusion on Li5AlO4 is more dependent on temperature in
comparison to the CO2 absorption. It is very clear that lithium
diffusion is the limiting step in the whole temperature range,
although it is highly activated as a function of the temperature
in comparison with the CO2 absorption process.
4. Conclusions
Lithium aluminates (LiAlO2 and Li5AlO4) were synthesized
by solid state reaction and then characterized by XRD and SEM.
In both cases, pure materials were obtained, and they presented
a considerably large particle size. Then, during the dynamic
thermal analyses, both ceramics presented very different be-
haviors for the CO2 absorption. Whereas LiAlO2 practically did

Thermochemical Capture of CO2 on Lithium Aluminates
J. Phys. Chem. A, Vol. 113, No. 25, 2009 6923
supported by the projects 23418-CONACYT-SEMARNAT,
99102-CONACYT, and IN100609-PAPIIT-UNAM. We thank
L. Ban˜os and E. Fregoso for technical help.
References and Notes
(1) Ochoa-Ferna´ndez, E.; Rønning, M.; Yu, X.; Grande, T.; Chen, D.
Ind. Eng. Chem. Res. 2008, 47, 434–442.
(2) Hwang, K. S.; Lee, Y. H.; Hwangbo, S. Mater. Sci. Poland 2007,
25, 969–975.
(3) Yi, K. B.; Eriksen, D. Ø. Sep. Sci. Technol. 2006, 41, 283–296.
(4) Nakagawa, K.; Ohashi, T. J. Electrochem. Soc. 1998, 145, 1344–
1346.
(5) Rodriguez, M. T.; Pfeiffer, H. Thermochim. Acta 2008, 473, 92–97.
(6) Pfeiffer, H.; Bosch, P. Chem. Mater. 2005, 17, 1704–1710.
(7) Togashi, N.; Okumura, T.; Oh-ishi, K. J. Ceram. Soc. Jpn. 2007,
115, 324–328.
(8) Kato, M.; Essaki, K.; Nakagawa, K.; Suyama, Y.; Terasaka, K.
J. Ceram. Soc. Jpn. 2005, 113, 684–686.
(9) Palacios-Romero, L. M.; Pfeifer, H. Chem. Lett. 2008, 37, 862–
863.
(10) Escobedo-Bretado, M.; Guzma´n-Velderrain, V.; Lardizabal-Guti-
errez, D.; Collins-Mart´ınez, V.; Lopez-Ortiz, A. Catal. Today 2005, 107-
108, 863–867.
(11) Essaki, K.; Kato, M.; Uemoto, H. J. Mater. Sci. 2005, 18, 5017–
5019.
(12) Kalinkin, A. M.; Kalinkina, E. V.; Zalkind, O. A.; Makarova, T. I.
Colloid J. 2008, 70, 33–41.
(13) Yamaguchi, T.; Niitsuma, T.; Nair, B. N.; Nakagawa, K. J. Membr.
Sci. 2007, 294, 16–21.
(14) Johnson, C. E.; Hollenberg, G. W. J. Nucl. Mater. 1981, 103, 547–
556.
Figure 8. Eyring’s plots for the rate constants of CO2 absorption (k1)
and lithium diffusion (k2).
not absorb CO2 at all, Li5AlO4 presented a high CO2 absorption.
Li5AlO4 absorbed CO2 in a wide temperature range from 200
to 700 °C, but it sintered as a function of temperature. At low
temperatures, the CO2 absorption decreased because of the
sintering effect, which implies a reduction of the surface area.
However at high temperatures, this phenomenon was not
observed because lithium diffusion was activated.
To avoid the sintering effect and to perform a kinetic analysis,
a different set of samples was homogeneously sintered before
the isothermal experiments were carried out. All of these
isotherms were fitted to exponential models, which described
two different processes: the CO2 absorption reaction and the
lithium diffusion from the core of the particles to the surface.
It could be determined that the diffusion process is the limiting
step of the whole process. Additionally, the ∆H values obtained
for the CO2 absorption and lithium diffusion processes were
15.6 and 52.1 kJ/mol, respectively. This means that lithium
diffusion is more dependent on the temperature.
Last but not least, it has to be mentioned that aluminum is a
lighter element in comparison with the other elements used as
structural ceramics for this application, for example silicates
(Li4SiO4 and Li2SiO3), cuprates (Li2CuO2), zirconates (Li2ZrO3
and Li6Zr2O7), or titanates (Li4TiO4). Therefore, because Li5AlO4
has the best theoretical CO2 chemisorption capacity per gram
of ceramic among the lithium ceramics and because of the
results obtained, Li5AlO4 may become an important case of
study as a CO2 captor.
(15) Oda, T.; Oya, Y.; Tanaka, S.; Weber, W. J. J. Nucl. Mater. 2007,
367, 263–268.
(16) Sundaram, N. T. K.; Subramania, A. Electrochim. Acta 2007, 52,
4987–4993.
(17) Li, T. K.; Ng, D. H. L. Mater. Sci. Eng., A 2007, 445-446, 652–
656.
(18) Ribeiro, R. A.; Silva, G. G.; Mohallem, N. D. S. J. Phys. Chem.
Solids 2001, 62, 857–864.
(19) Andreev, O. L.; Zelyutin, G. V.; Martemyanova, Z. S.; Batalov,
N. N. Inorg. Mater. 2001, 37, 177–179.
(20) Follstaedt, D. M.; Biefeld, R. M. Phys. ReV. B 1978, 18, 5928–
5937.
(21) La-Ginestra, A.; Lo-Jacono, M.; Porta, P. J. Therm. Anal. 1972, 4,
5–17.
(22) Stewner, F.; Hoppe, R. Z. Anorg. Allg. Chem. 1971, 380, 241–
243.
(23) Hoppe, R.; Ko¨nig, H. Z. Anorg. Allg. Chem. 1977, 430, 211–217.
(24) Pfeiffer, H.; Knowles, K. M. J. Eur. Ceram. Soc. 2004, 24, 2433–
2443.
(25) Mosqueda, H. A.; Vazquez, C.; Bosch, P.; Pfeiffer, H. Chem. Mater.
2006, 18, 2307–2310.
(26) Xiong, R.; Ida, J.; Lin, Y. S. Chem. Eng. Sci. 2003, 58, 4377–
4385.
(27) Palacios-Romero, L. M.; Lima, E.; Pfeiffer, H. J. Phys. Chem. A
2009, 113, 193–198.
(28) Alce´rreca-Corte, I.; Fregoso-Israel, E.; Pfeiffer, H. J. Phys. Chem.
C 2008, 112, 6520–6525.
Acknowledgment. This work has been performed into the
PUNTA IMPULSA-UNAM framework, and it was financially
JP902501V