Available online at www.sciencedirect.

com

ScienceDirect
Solar Energy 103 (2014) 425437
www.elsevier.com/locate/solener

Thermochemical CO2 splitting reaction with supported

LaxA1xFeyB1yO3 (A = Sr, Ce, B = Co, Mn; 0 6 x, y 6 1)
perovskite oxides
Qingqing Jiang a,b, Jinhui Tong a,b, Guilin Zhou a,b, Zongxuan Jiang a, Zhen Li a,b, Can Li a,
a

State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian National Laboratory for Clean Energy,
Dalian 116023, China
b
Graduate University of Chinese Academy of Sciences, Beijing 100049, China
Received 13 December 2013; received in revised form 19 February 2014; accepted 21 February 2014
Available online 15 March 2014
Communicated by: Associate Editor Michael Epstein

Abstract
An ecient redox material for two-step thermochemical CO2 splitting reaction requires high chemical yield at relatively low reduction
temperature. Herein, the oxides with perovskite structure of formula LaxA1xFeyB1yO3 (A = Sr, Ce, B = Co, Mn; 0 6 x, y 6 1) start to
release O2 at 800 C and the largest O2 production is 11.8 ml/gperovskite at 1300 C. However, for these unsupported LaxA1xFeyB1yO3
materials, the CO production is low in spite of high reduction yield. ZrO2, Al2O3 and SiO2 are thus considered as supports to disperse
LaxA1xFeyB1yO3 materials and dierent supports induce great dierences in the reaction activity. By A-site or B-site substitution of
LaFeO3, the O2 releasing temperature has fallen from 1230 C to 800-950 C and the CO production is enhanced 23 times. LaFe0.7Co0.3O3 (25 wt%)/SiO2 shows the highest reaction activity among these investigated materials with the O2 production of 4.0 ml/gmaterial
(16.0 ml/gperovskite) and CO production of 7.6 ml/gmaterial (30.4 ml/gperovskite) when it is reduced at 1300 C and re-oxidized at 1100 C,
and the activity is relatively stable even after 10 cycles of the reaction. By contrast, the CO production is 4.5 ml/g for CeO2 when it
is reduced at 1400 C. The estimated activation energy for the reduction step of LaFe0.7Co0.3O3 (25 wt%)/SiO2 is around 89149 KJ/
mol according to dierent models. The CO generation step of LaFe0.7Co0.3O3 (25 wt%)/SiO2 is mainly controlled by bulk diusion
(D1) at 1000 C and then it turns to rst order surface reaction (F1) at 1100 C.
2014 Elsevier Ltd. All rights reserved.

Keywords: Solar energy; CO2 reduction; Perovskite oxide; Kinetics

1. Introduction
Ecient utilization of solar energy is an important subject. The conversion of concentrated solar energy into
chemical fuels via two-step thermochemical H2O and CO2
splitting reaction which uses the entire solar spectrum is
Corresponding author. Tel.: +86 411 84379070; fax: +86 411

84694447.
E-mail address: canli@dicp.ac.cn (C. Li).
http://dx.doi.org/10.1016/j.solener.2014.02.033
0038-092X/ 2014 Elsevier Ltd. All rights reserved.

an attractive option for solar energy utilization (Chueh

et al., 2010; Steinfeld, 2005; Furler et al., 2012; Muhich
et al., 2013; Jiang et al., 2014). The two-step thermochemical splitting reaction with metal oxide as redox material
can be represented as follows:
MOox solar thermal energy MOred O2

1:1

MOred H2 O=CO2 MOox H2 =CO

1:2

Total reaction:

426

Q. Jiang et al. / Solar Energy 103 (2014) 425437

H2 O=CO2 solar energy H2 =CO O2

1:3

Various metal oxides have been investigated in the two

step thermochemical splitting reaction, among which
Fe3O4, ZnO and CeO2 are the most popular redox materials. In order to obtain high chemical yield at relatively low
reduction temperature, the addition of dopants into Fe3O4
and CeO2 were extensively studied with a certain degree of
success (Gal et al., 2011; Gal and Abanades, 2012; Meng
et al., 2011; Schee and Steinfeld, 2012; Fresno et al.,
2009; Kodama et al., 2008). However, the reduction to
nonstoichiometric MOred below 1400 C shows limited
chemical yield for most of these investigated redox oxides.
The perovskite oxides, ABO3, have been applied in the
elds of catalysis, solid oxide fuel cells, and oxygen permeable membranes due to their unique structural features
(Merino et al., 2006; Wang et al., 2004; Yang et al., 2003;
He et al., 2011; Lu et al., 2012). Their basic structural
framework keeps unchanged with wide variety of constituent elements and composition, therefore the valence of cations and the oxygen vacancy concentration can be changed
widely (Tanaka and Misono, 2001). Furthermore, the thermal stability of perovskite oxides is better than their corresponding transition metal oxides.

La1xSrxMO3 (M = Mn, Fe) have been investigated for

the two-step thermochemical steam reforming of methane
(Evdou et al., 2008a,b). It is desired to investigate the
two step thermochemical H2O and CO2 splitting reaction
with perovskite oxides because it is a more attractive way
for energy utilization from a long term perspective. The
latest results from McDaniel et al. (2013) showed that
Sr- and Mn-doped LaAlO3d exhibited excellent reaction
performance in two step thermochemical CO2 and H2O
splitting reaction. The H2 yields were 9  greater, and the
CO yields 6  greater, than those produced by ceria, when
it was reduced at 1350 C and re-oxidized at 1000 C.
In the present work, the perovskite oxides of formula
LaxA1xFeyB1yO3 (A = Sr, Ce, B = Co, Mn; 0 6 x,
y 6 1) based on Fe3+/Fe2+ cycle were utilized in two-step
thermochemical CO2 splitting reaction. It is found that
unsupported LaxA1xFeyB1yO3 materials present negligible CO production in spite of high reduction yield. What
seems surprising is that mixing SiO2 with the perovskite
oxides apparent enhances the redox activity. These
A- and B-site substituted LaFeO3 show an enhanced reducibility at low temperature and in some cases, the CO generation activity is also largely increased. The CO production

Fig. 1. (a) The O2 evolution rate-time proles from 800 C to 1300 C and (b) the CO generation rate-time proles at 1000 C of LaFexB1xO3 (B = Mn
and Co; x = 1.0, 0.7 and 0).

Q. Jiang et al. / Solar Energy 103 (2014) 425437

for LaFe0.7Co0.3O3 (25 wt%)/SiO2 is 7.6 ml/gmaterial

(30.4 ml/gperovskite) when it is reduced at 1300 C and reoxidized at 1100 C. By contrast, the CO production is
4.5 ml/g for CeO2 when it is reduced at 1400 C. Cycle performance shows that the redox activity is stable even after
10 cycles of high temperature reaction. The calculated activation energies of the reduction step and the main limitation of the CO generation reaction was given by kinetic
analysis.
2. Experimental section
2.1. Synthesis method
The citrate method was employed to produce nanoparticles of these LaxA1xFeyB1yO3 (A = Sr, Ce; B = Mn,
Co, Ni) perovskite oxides (Merino et al., 2006). Aqueous
solutions of starting nitrates were mixed with an aqueous
solution of citric acid. The molar ratio between the total
amount of nitrates and citric acid was 1:2. The resulting
solution was evaporated at 80 C and then completely
dried overnight at 120 C. The spongy material thus
obtained was crushed and calcined at 700 C for 4 h. The
perovskite oxides were supported on commercial supports
(SiO2, ZrO2 and Al2O3) before reaction testing. The perov-

427

skite oxides were well-mixed with supports in which the

weight ratio was 1:3. The perovskite oxides and commercial supports were mechanically mixed in a mortar for half
an hour. After that, the mixture was heated at 900 C in a
mue burner for 10 h.
2.2. Characterization of samples
The synthesized samples were characterized by X-ray
powder diraction (XRD) on a Rigaku D/Max-2500/PC
powder diractometer. Each sample powder was scanned
using Cu Ka radiation with an operating voltage of
40 kV and an operating current of 200 mA. The scan rate
of 5 min1 was applied to record the XRD patterns in
the range of 2080 at a step size of 0.02. The specic surface areas of the samples were taken with a Micromeritics
ASAP 2000 adsorption analyzer.
2.3. Reaction activity testing
The two-step thermochemical splitting reaction was carried out at a laboratory scale in a xed bed conguration.
The vertical alumina tubular reactor was placed inside an
electric furnace. The argon ow (purity 99.9996%) rstly
passed through a deoxidation tube to get rid of the residual

Fig. 2. SEM images of (a) LaFeO3, (b) LaFeO3 after high temperature treatment and (c) LaFe0.7Co0.3O3 (25 wt%)/SiO2 after 10 cycles.

428

Q. Jiang et al. / Solar Energy 103 (2014) 425437

O2 before it passed into the reactor. The O2 concentration

in the experimental system was about 20 ppm.
The reduction reaction was performed with an argon
ow rate of 100 ml/min. For unsupported LaxA1xFeyB1yO3
(A = Sr, Ce; B = Mn, Co) samples, 0.5 g of the materials
were heated to 800 C with a 20 C/min heating rate and
then to 1300 C with a 10 C/min heating rate. To investigate the eect of dierent supports on the reaction activity,
0.4 g of the mixed materials were heated to 800 C with a
20 C/min heating rate and then to 1300 C with a
10 C/min heating rate. To investigate the eect of A-site
or B-site substitution on the reaction activity, 0.4 g of the
mixed materials were heated to 1200 C with a 20 C/min
heating rate and then to 1300 C with a 5 C/min heating
rate. The temperature plateau at 1300 C was maintained
for a while until O2 releasing reaction completed. The produced gas mixture was analyzed with a gas chromatograph
(Aligent 6890) equipped with a 5A molecular sieve column
and a TCD detector, which took gas sample at the reactor
outlet every ca. 2 min.
The reduced state of the oxide was maintained under the
protection of Ar with a ow rate of 500 ml/min before CO2
was exposed to the reactor. After the electric furnace was
fell to the certain temperature (9001100 C), and then
CO2 (purity 99.999%) with a ow rate of 500 ml/min was
injected to react with the oxygen-decient materials. The

CO gas product was analyzed by a gas chromatograph

(Aligent 6890) equipped with a GDX-01 column and a
FID detector, which took gas sample at the reactor outlet
every ca. 1 or 2 min. Numerical integration of the molar
ow rate-time curves gave the production of O2 and CO.
In this study, the results were expressed in terms of O2
and CO production (ml) per gram of active perovskite oxides (ml/gperovskite) and per gram of total mixed materials
(ml/gmaterial).
3. Results and discussion
3.1. LaFexB1xO3 as redox materials
LaFexB1xO3 (B = Mn and Co; x = 1.0, 0.7 and 0) were
heated from 800 C to 1300 C with a heating rate of
10 C/min under an argon (purity 99.9996%) ow of
100 ml/min. As shown in Fig. 1(a), LaFeO3 starts to release
O2 at 1210 C and the nal amount of oxygen released is
1.3 ml/gperovskite. By contrast, LaMnO3 and LaCoO3
start to release O2 at 800 C and the O2 production is
10.6 ml/gperovskite and 11.8 ml/gperovskite, respectively. The
results agree well with the thermodynamic analysis which
indicates that the reduction of Mn3O4 to MnO and of
Co3O4 to CoO is much easier than that of Fe3O4 to FeO
(Kodama and Gokon, 2007). As a result, doping Mn and

Fig. 3. (a) The O2 evolution rate-time proles from 800 C to 1300 C and (b) the CO generation rate-time proles at 1000 C of LaFe0.7Co0.3O3 dispersed
on dierent supports for the rst cycle.

Q. Jiang et al. / Solar Energy 103 (2014) 425437

429

Fig. 4. XRD patterns of (a) LaFe0.7Co0.3O3, Al2O3 and LaFe0.7Co0.3O3 (25 wt%)/Al2O3 after reaction, (b) LaFe0.7Co0.3O3, ZrO2 and LaFe0.7Co0.3O3
(25 wt%)/ZrO2 after reaction, and (c) LaFe0.7Co0.3O3 (25 wt%)/SiO2 after 10 cycles of CO2 splitting reaction.

Fig. 5. The CO generation activity of (a) LaAl11xFexO19 (b) CoxFe3xO4 at 1000 C. The value of x is according to the content of iron in LaFe0.7Co0.3O3
(25 wt%)/Al2O3 and LaFe0.7Co0.3O3 (25 wt%)/ZrO2.

Co into LaFeO3 makes the thermal reduction step thermodynamically more favorable in comparison to pure LaFeO3.
The corresponding O2 production is increased from
1.3 ml/gperovskite up to 8.5 ml/gperovskite for LaFe0.7Mn0.3O3
and 6.4 ml/gperovskite for LaFe0.7Co0.3O3, respectively.
The generation of CO for these reduced LaFexB1xO3
materials shown in Fig. 1(b) demonstrates that CO2 splitting reaction activity is poor at 1000 C. These samples

present low CO generation rate and negligible CO production at 1000 C in spite of high reduction yield. SEM
images show that the samples are aggregated into large
particles after high temperature treatment and the poor
CO generation activity may be due to sintering (Fig. 2(a
and b)). Therefore, the perovskite oxides are dispersed
on supports to enhance the high temperature thermal
stability.

430

Q. Jiang et al. / Solar Energy 103 (2014) 425437

Fig. 6. XRD patterns of (a) SrFeO3, La0.7Sr0.3FeO3, La0.7Ce0.3FeO3 and LaFeO3 and (b) LaMnO3, LaCoO3, LaFe0.7Co0.3O3, LaFe0.7Mn0.3O3 and
LaFeO3.

Table 1
BET surface areas and crystallite sizes of these perovskite oxides.

3.2. The eect of dierent supports on the reaction activity

Samples

Crystallite size
)
(A

Surface area
(m2 g1)

LaFeO3
La0.7Sr0.3FeO3
La0.7Ce0.3FeO3
LaFe0.7Co0.3O3
LaFe0.7Mn0.3O3
SiO2
LaFe0.7Co0.3O3/SiO2 after10 cycles
The crystallite size is calculated
from XRD

233
316
465
249
207

10.7
9.7
9.6
9.2
18.5
2.9
0.6

ZrO2, Al2O3 and SiO2 were thus used as the supports.

The loading of LaFe0.7Co0.3O3 was set to 25% on a weight
basis. These supported materials were reduced at temperature up to 1300 C (with a heating rate 10 C/min) under an
argon ow of 100 ml/min.
Indeed, The O2 evolution activity is largely enhanced by
dispersing LaFe0.7Co0.3O3 on supports of ZrO2, Al2O3 and
SiO2, as shown in Fig. 3(a). Dierent supports (ZrO2,
Al2O3 and SiO2) induce great dierences in the O2 produce activity. The maximum O2 evolution rate for

Fig. 7. (a) The O2 evolution rate-time proles, (b) CO generation rate-time proles at 1100 Cof LaxA1xFeO3 (A = Sr and Ce) (25 wt%)/SiO2 for the rst
cycle.

Q. Jiang et al. / Solar Energy 103 (2014) 425437

LaFe0.7Co0.3O3 (25 wt%)/SiO2 is 13  greater, for LaFe0.7Co0.3O3 (25 wt%)/Al2O3 is 5.6  greater, and for LaFe0.7Co0.3O3 (25 wt%)/ZrO2 is 2.9  greater, than that
produced by unsupported LaFe0.7Co0.3O3. LaFe0.7Co0.3O3
(25 wt%)/SiO2 is considered to be the best candidate which
could be reduced to a greater extent at same reduction temperature. The O2 evolution curves are quite dierent from
each other mainly due to dierent high temperature solid
state reactions occur between LaFe0.7Co0.3O3 and
supports.
As shown in Fig. 3(b), the CO generation activity
follows the order of intensities LaFe0.7Co0.3O3 (25 wt%)/
SiO2 > LaFe0.7Co0.3O3 (25 wt%)/ZrO2 > LaFe0.7Co0.3O3
(25 wt%)/Al2O3 > LaFe0.7Co0.3O3. The dierent productivities of LaFe0.7Co0.3O3 dispersed on dierent supports are
thus attributed to the dierent structural changes after high
temperature treatment.
The XRD patterns shown in Fig. 4(a) indicates that Fesubstituted hexaaluminate (LaAl11xFexO19) is formed for
LaFe0.7Co0.3O3 (25 wt%)/Al2O3 after high temperature
treatment. The CO production of LaAl11xFexO19 is negligible (Fig. 5(a)) and thus Al2O3 supported LaFe0.7Co0.3O3
exhibits low CO generation activity. For LaFe0.7Co0.3O3
(25 wt%)/ZrO2 after high temperature treatment, phases
of La2Zr2O7, CoxFe3xO4, and ZrO2 are observed. The
reaction activity of LaFe0.7Co0.3O3 (25 wt%)/ZrO2 is simi-

431

lar to that of CoxFe3xO4 (Fig. 5(b)) which proves the

active species is CoxFe3xO4. For LaFe0.7Co0.3O3
(25 wt%)/SiO2, silica is transformed to stable tridymite
phase after high temperature treatment. The high temperature stable tridymite phase is the main factor for high reaction activity. High temperature solid state reaction occurs
between silica and perovskite oxides after high temperature
treatment. The redox active agents are not perovskite phase
but the mixed materials, including La2SiO5, FexSiyOz,
tridymite and others. The active redox element is Fe3+,
however, the reaction activity of LaxA1xFeyB1yO3/SiO2
is quite dierent from their corresponding ferrites. The
mixed structure may be the other reason for the enhanced
reaction activity.
3.3. The eect of A-site substitution of LaFeO3 on the
reaction activity
Ce4+- and Sr2+-substituted LaFeO3 were synthesized
and tested in the two step thermochemical CO2 and H2O
splitting reaction. XRD patterns of SrFeO3, La0.7Sr0.3FeO3,
La0.7Ce0.3FeO3 and LaFeO3 are exhibited in Fig. 6(a). For
La0.7Sr0.3FeO3, the pure orthorhombic perovskite-type
structure is formed, and single component oxide such as
La2O3, and Fe2O3, has not been detected, indicating that
the perovskite-type structure is preserved with the partial

Fig. 8. (a) The O2 evolution rate-time proles and (b) the CO generation rate-time proles at 1100 C of LaFexCo1xO3 (x = 1, 0.9, 0.7, 0.5) (25 wt%)/
SiO2 for the rst cycle.

432

Q. Jiang et al. / Solar Energy 103 (2014) 425437

substitution of A-site elements. For La0.7Ce0.3FeO3, tiny

CeO2 phase is observed. The crystallite sizes were estimated
to be <50 nm using Scherrers equation, as represented in
Table 1. The samples synthesized by citrate method exhibit
low specic surface area, about 10 m2 g1.
From Fig. 7(a) it can be seen that the O2 releasing temperature has fallen from 1230 C to 970 C by partial substitution of La3+ with Sr2+ and the maximum O2 evolution

rate is increased 2 times after the introduction of Ce4+ into

LaFeO3. The O2 production for these SiO2 supported
LaxA1xFeyB1yO3 samples is shown in Fig. 10(a) and it
indicates that the O2 production is enhanced 23 times by
A-site substitution.
However, the high reduction yield of La0.7Sr0.3FeO3
(25 wt%)/SiO2 does not result in high CO generation rate
(Fig. 7b). The CO generation rate of La0.7Sr0.3FeO3

Fig. 9. The (a) O2 evolution rate-time proles and (b) CO generation rate-time proles at 1100 C of LaFexB1xO3 (B = Mn and Co) (25 wt%)/SiO2 for
the rst cycle.

Fig. 10. (a) O2 and (b) CO production of LaxA1xFeyB1yO3 (A = Sr, Ce; B = Co, Mn) (25 wt%)/SiO2 for the rst cycle.

Q. Jiang et al. / Solar Energy 103 (2014) 425437

(25 wt%)/SiO2 is only half of LaFeO3 (25 wt%)/SiO2. The

CO production during the rst 20 min of La0.7Sr0.3FeO3
(25 wt%)/SiO2 is also half of LaFeO3 (25 wt%)/SiO2.
Nevertheless, the total CO production is increased from
2.0 ml/gmaterial
(8.1 ml/gperovskite)
to
2.8 ml/gmaterial
(11.2 ml/gperovskite) by substituted La3+ with Sr2+.
For La0.7Ce0.3FeO3 (25 wt%)/SiO2, although the CO
generation rate does not show much improvement, the
enhanced reducibility leads to higher CO and H2 production as compared to LaFeO3 (25 wt%)/SiO2. The total
CO production is increased from 2.0 ml/gmaterial
(8.1 ml/gperovskite) to 3.1 ml/gmaterial (12.4 ml/gperovskite) by
substituted La3+ with Ce4+.
The defect chemistry of LaxA1xFeO3 has been widely
investigated. The partial substitution of Sr2+ for La3+ leads
to higher oxidation states of Fe ion and the simultaneous
formation of structural defects (Evdou et al., 2008a,b; Ferri
and Forni, 1998; Royer et al., 2005). The enhanced reducibility could be attributed to the modied reducibility of Fe
ion. However, the vacancies for compensating the Sr2+
doping acceptors are considered stable and could not
utilize during the CO2 and H2O splitting reaction. Therefore, the high reduction yield and low fuel productivity
for La0.7Sr0.3FeO3 (25 wt%)/SiO2 is reasonable. According
to the previous TPR results, the addition of Ce4+ into the
perovskite lattice could enhance the reducibility of Fe3+
species due to the interaction between Fe3+Fe2+ and
Ce4+Ce3+ redox couples (Royer et al., 2005). Therefore,
the addition of Ce4+ into LaFeO3 favours the reduction
process and the corresponding CO production.
3.4. The eect of B-site substitution of LaFeO3 on the
reaction activity
Combination of two dierent ions at the B-site brings
about synergistic eects and is a powerful tool for enhancement of the reducibility. Co- and Mn- are thus considered
for the substitution of Fe in LaFeO3. The studied materials
are LaFe1xCoxO3 (x = 0.9, 0.7, 0.5) and LaFe0.7Mn0.3O3.
The XRD patterns are presented in Fig. 6(b) conrming
the formation of perovskite-type structure.
The O2 evolution activity for LaFexCo1xO3 as a function of Co content is shown in Fig. 8(a). It can be seen that

433

Fig. 12. Cycle performance of LaFe0.7Co0.3O3 (25 wt%)/SiO2 for 10 times

with reduction temperature at 1300 C and CO generation reaction at
1100 C.

the O2 evolution activity increases as the Co content

increasing. The CO generation rate reaches the maximum
when x is 0.3 as the x value varies from 0 to 0.5
(Fig. 8(b)). The initial CO generation rate of LaFe0.7Co0.3O3
(25 wt%)/SiO2 is 4 times faster than that of LaFeO3
(25 wt%)/SiO2 and the CO production for the rst 20 min
is increased from 1.2 ml/gmaterial (4.6 ml/gperovskite) up to
4.3 ml/gmaterial (17.2 ml/gperovskite). The total CO production
for LaFexCo1xO3 (x = 0.7 and 0.5) (25 wt%)/SiO2 is almost
three times larger than that of LaFeO3 (25 wt%)/SiO2
(Fig. 10(b)). The substitution of Fe3+ with Co3+ favors the
reduction of Fe3+ to Fe2+, therefore, the reaction activity
enhances with the molar ratio of Co to Fe increasing, however, Fe3+/Fe2+ is the eective redox pair, if too much
Fe3+ is substituted, the CO generation activity could be
decreased.
LaFe0.7Mn0.3O3 (25 wt%)/SiO2 exhibits the highest O2
evolution activity among these A-site and B-site substituted
LaFeO3 (Fig. 9(a)). The maximum O2 evolution rate is
0.37 ml min1 g1
material for LaFe0.7Mn0.3O3 (25 wt%)/SiO2,
by contrast, 0.13 ml min1 g1
material for LaFe0.7Co0.3O3
(25 wt%)/SiO2, 0.13 ml min1 g1
material for La0.7Ce0.3FeO3
(25 wt%)/SiO2 and 0.1 ml min1 g1
for LaFeO3
material

Fig. 11. (a) The O2 rate-time proles at 1200 C, 1300 C, and 1400 C, (b) the corresponding CO generation activity at 1100 C of LaFe0.7Co0.3O3/SiO2.

434

Q. Jiang et al. / Solar Energy 103 (2014) 425437

(25 wt%)/SiO2. The initial CO generation rate of LaFe0.7Mn0.3O3 (25 wt%)/SiO2 is similar to that of LaFeO3
(25 wt%)/SiO2, and it is much lower than that of
LaFe0.7Co0.3O3 (25 wt%)/SiO2 (Fig. 9(b)), nevertheless, the
CO productivity is comparable to that of LaFe0.7Co0.3O3
(25 wt%)/SiO2, as shown in Fig. 10(b).
Fig. 11 exhibits the CO2 splitting activity for LaFe0.7Co0.3O3 (25 wt%)/SiO2 as the reduction temperature varies
from 1200 C to 1400 C. The O2 production is largely
increased as reduction temperature increases from
1200 C to 1400 C, from 1.6 ml/gmaterial up to 6.3 ml/gmaterial.
The corresponding CO production is only 0.4 ml/gmaterial
when the reduction temperature is 1200 C which indicates

that O2 production below 1200 C is not due to the reduction

of Fe3+ to Fe2+, but the reduction of Co3+ to Co2+. The corresponding CO generation activity of that reduced at 1400 C
is much lower than that reduced at 1300 C which proves that
the reaction performance of LaFe0.7Co0.3O3 (25 wt%)/SiO2 is
dierent from ferrites.
Cycle performance for LaFe0.7Co0.3O3 (25 wt%)/SiO2
was performed with the reduction temperature at 1300 C
and CO generation step at 1100 C. As shown in Fig. 12,
the O2 and CO production for LaFe0.7Co0.3O3
(25 wt%)/SiO2 becomes even higher as the increasing of cycle
number. The ratio of CO production to O2 production is
more and more close to 2 as the increasing of cycle number.

Fig. 13. (a) O2 evolution rate of LaFe0.7Co0.3O3 (25 wt%)/SiO2 under heating rate of 10 C/min, (be) Arrhenius plot for the non-isothermal reduction of
LaFe0.7Co0.3O3 (25 wt%)/SiO2 according to dierent models. F1 means rst order reaction; F2 means second order reaction; R2 and R3 means
geometrical-contracting models.

Q. Jiang et al. / Solar Energy 103 (2014) 425437

The CO production is 7.6 ml/gmaterial (30.4 ml/gperovskite) for

the 10th cycle, by contrast, the CO production is 4.5 ml/g for
CeO2 when it is reduced at 1400 C. It indicates that SiO2
supported LaxA1xFeyB1yO3 (A = Sr, Ce; B = Co, Mn)
pervoskite oxides are promising redox materials for fuel production via two-step solar thermal splitting reaction at relatively low reduction temperature.
3.5. Kinetic analysis
The kinetic studies for the O2 evolution reaction were
performed under non-isothermal experimental condition
with a constant heating rate of 10 C/min. According to
the literatures, the model can be expressed as follows
(Levenspiel, 1972; Galwey and Brown, 1999; Francis
et al. (2010); Schunk and Steinfeld, (2009)


da=dT b
Ea
ln
3:1
ln A 
f a
RT
f a 1  a

3:2

435

And a is dened as:

a

Total O2 t
Total O2

The curve of ln

3:3
h

da=dT b
f a

versus 1/T should be linear if the

hypothetical model is reasonable, and then activation

energy Ea can be calculated from slope of the line.
The O2 evolution rate-time proles of LaFe0.7Co0.3O3
(25 wt%)/SiO2 from 800 C to 1300 C with a heating rate
of 10 C/min is shown in Fig. 13(a). The reaction order (F1
and F2) and geometrical-contracting (R2 and R3) models
have been attempted for f (a). The experimental data are
processed according to (3.1). The linear regression for the
experimental data is shown in Fig. 13(be). Both surface
reaction order (F1 and F2) and geometrical-contracting
(R2 and R3) models show excellent linear t. It is dicult
to precisely determine the actual reaction model for the
reduction reaction of LaFe0.7Co0.3O3 (25 wt%)/SiO2. The
estimated activation energy according to dierent models

Fig. 14. (a) the CO generation rate-time proles for LaFe0.7Co0.3O3 (25 wt%)/SiO2 from 900 C to 1100 C (b and c) normalized rate data compared to
solid state reaction model of the CO generation reaction for LaFe0.7Co0.3O3 (25 wt%)/SiO2 at 1000 C and 1100 C. F1 means rst order reaction; D1
means rst order diusion reaction model.

436

Q. Jiang et al. / Solar Energy 103 (2014) 425437

is 108 KJ/mol for F1, 149 KJ/mol for F2, 89 KJ/mol for
R2, and 97 KJ/mol for R3, respectively.
The CO generation rate can be expressed as follows:

thermal splitting reaction at relatively low reduction

temperature.

da
kf a
dt

Acknowledgements

3:4

For isothermal experiment, if the normalized rate data

agree well with a given kinetic model, (3.4) can be
expressed as (Gotor et al., 2000; Perez-Maqueda et al.,
2002; Yang et al., 2009):
da=dt
f a

da=dta0:5 f aa0:5

3:5

Fig. 14(a) shows the CO generation rate-time proles for

LaFe0.7Co0.3O3 (25 wt%)/SiO2 from 900 C to 1100 C. It
can be seen that the CO generation rate increases with the
increasing of temperature. The normalized rate data for
LaFe0.7Co0.3O3 (25 wt%)/SiO2 compared to various solid
state reaction models are shown in Fig. 14(b and c). The
CO generation reaction of LaFe0.7Co0.3O3 (25 wt%)/SiO2
agrees well with a bulk diusion model (D1) at 1000 C
(0.2 < a < 0.8). When the CO generation reaction is performed at 1100 C, the reaction turns to the rst order surface reaction (0.1 < a < 1.0).
4. Conclusions
The oxides with perovskite structure of formula
LaxA1xFeyB1yO3 (A = Sr, Ce; B = Co, Mn) were utilized in two-step thermochemical CO2 splitting reaction.
For unsupported pervoskite oxides, the CO generation
activity was negligible in spite of promising O2 evolution
production at low reduction temperature. Thus, commercial ZrO2, Al2O3 and SiO2 were thus considered as the
supports. Dierent supports have great inuence on the
reaction activity and SiO2 was found to be the best candidate among these investigated supports. By A-site or B-site
substitution of LaFeO3, the O2 evolution temperature was
reduced by 200400 C and the fuel production was
enhanced 23 times. LaFe0.7Co0.3O3 (25 wt%)/SiO2 exhibited the highest reaction activity among these investigated
perovskite oxides. The maximum CO generation rate for
LaFe0.7Co0.3O3 (25 wt%)/SiO2 was 4 times greater than
that produced by LaFeO3 (25 wt%)/SiO2. The CO production for LaFe0.7Co0.3O3 (25 wt%)/SiO2 was 7.6 ml/gmaterial
(30.4 ml/gperovskite) when the sample was reduced at
1300 C, by contrast, the CO production was 4.5 ml/g for
CeO2 when it was reduced at 1400 C. The estimated activation energy for the reduction step of LaFe0.7Co0.3O3
(25 wt%)/SiO2 was around 89149 KJ/mol according to
dierent models. The kinetic analysis of the CO generation
step of LaFe0.7Co0.3O3 (25 wt%)/SiO2 indicated that the
reaction was mainly controlled by bulk diusion (D1) at
1000 C and then it turned to rst order surface reaction
(F1) at 1100 C. Overall, SiO2 supported LaxA1xFeyB1yO3
(A = Sr, Ce; B = Co, Mn) pervoskite oxides are promising
redox materials for fuel production via two-step solar

This work was nancially supported by National

Program on Key Basic Research Project (Grant
2009CB220010), National Natural Science Foundation of
China (Grant 21061140361), and Solar Energy Action
Plan of Chinese Academy of Sciences (Grant KGCX2YW-393-1).
References
Chueh, W.C., Falter, C., Abbott, M., Scipio, D., Furler, P., Haile, S.M.,
Steinfeld, A., 2010. High-ux solar-driven thermochemical dissociation
of CO2 and H2O using nonstoichiometric ceria. Science 330, 1797
1800.
Evdou, A., Nalbandian, L., Zasplis, V.T., 2008a. Perovskite membrane
reactor for continuous and isothermal redox hydrogen production
from the dissociation of water. J. Membr. Sci. 325, 704711.
Evdou, A., Zaspalis, V., Nalbandian, L., 2008b. La(1x)SrxMnO3d
perovskites as redox materials for the production of high purity
hydrogen. Int. J. Hydrogen Energy 33, 55545562.
Ferri, D., Forni, L., 1998. Methane combustion on some perovskite-like
mixed oxides. Appl. Catal. B: Environ. 16, 119126.
Francis, T.M., Lichty, P.R., Weimer, A.W., 2010. Manganese
oxide dissociation kinetics for the Mn2O3 thermochemical
water-splitting cycle. Part 1: Experimental. Chem. Eng. Sci.
65, 37093717.
Fresno, F., Fernandez-Saavedra, R., Gomez-Mancebo, M.B., Vidal, A.,
Miguel Sanchez, V., Rucandio, M.I., Quejido, J.A., Romero, M., 2009.
Solar hydrogen production by two-step thermochemical cycles: evaluation of the activity of commercial ferrites. Int. J. Hydrogen Energy
34, 29182924.
Furler, P., Furler, P., Schee, J.R., Steinfeld, A., 2012. Syngas
production by simultaneous splitting of H2O and CO2 via ceria
redox reactions in a high-temperature solar reactor. Energy
Environ. Sci. 5, 60986103.
Gal, A.L., Abanades, S., 2012. Dopant incorporation in ceria for
enhanced water-splitting activity during solar thermochemical hydrogen generation. J. Phys. Chem. C 116, 1351613523.
Gal, A.L., Abanades, S., Flament, G., 2011. CO2 and H2O splitting for
thermochemical production of solar fuels using nonstoichiometric
ceria and ceria/zirconia solid solutions. Energy Fuels 25, 4836.
Galwey, A.K., Brown, M.E., 1999. Thermal Decomposition of Ionic
Solids, rst ed. Elsevier, New York.
Gotor, F.J., Criado, J.M., Malek, J., Koga, N., 2000. Kinetic analysis of
solid-state reactions: the universality of master plots for analyzing
isothermal and nonisothermal experiments. J. Phys. Chem. A 104,
1077710782.
He, Y.F., Zhu, X.F., Yang, W.S., 2011. The role of a-site ion nonstoichiometry in the oxygen absorption properties of Sr1+xCo0.8Fe0.2O3
oxides. AIChE J. 57, 8795.
Jiang, Q.Q., Zhou, G.L., Jiang, Z.X., Li, C., 2014. Thermochemical CO2
splitting reaction with CexM1xO2d (M = Ti4+, Sn4+, Hf4+, Zr4+,
La3+, Y3+ and Sm3+) solid solutions. Solar Energy 99, 5566.
Kodama, Tatsuya, Gokon, Nobuyuki, 2007. Thermochemical cycles for
high-temperature solar hydrogen production. Chem. Rev. 107, 4048
4077.
Kodama, T., Gokon, N., Yamamoto, R., 2008. Thermochemical two-step
water splitting by ZrO2-supported NixFe3xO4 for solar hydrogen
production. Solar Energy 82, 7379.
Levenspiel, O., 1972. Chemical Reaction Engineering. John Wiley & Sons,
New York.

Q. Jiang et al. / Solar Energy 103 (2014) 425437

Lu, H., Zhu, L.L., Kim, J.P., Son, S.H., Park, J.H., 2012. Perovskite
La0.6Sr0.4B0.2Fe0.8O3-d (B = Ti, Cr, Co) oxides: structural, reduction
tolerant, sintering, and electrical properties. Solid State Ionics 209
210, 2429.
McDaniel, A.H., Miller, E.C., Arin, D., Ambrosini, A., Coker, E.N.,
OHayre, R., Chueh, W.C., Tong, J.H., 2013. Sr- and Mn-doped
LaAlO3dfor solar thermochemical H2 and CO production. Energy
Environ. Sci. 6, 24242428.
Meng, Q.L., Lee, C.I., Shigeta, S., Kaneko, H., Tamaura, Y., 2011.
Reactivity of CeO2-based ceramics for solar hydrogen production via a
two-step water-splitting cycle with concentrated solar energy. Int. J.
Hydrogen Energy 36, 1343513441.
Merino, N.A., Barbero, B.P., Ruiz, P., Cadus, L.E., 2006. Synthesis,
characterisation, catalytic activity and structural stability of LaCo1yFeyO3k perovskite catalysts for combustion of ethanol and propane.
J. Catal. 240, 245257.
Muhich, C.L., Evanko, B.W., Weston, K.C., Lichty, P., Liang, X.H.,
Martinek, J., Musgrave, C.B., Weimer, A.W., 2013. Ecient generation of H2 by splitting water with an isothermal redox cycle. Science
341, 540542.
Perez-Maqueda, L.A., Criado, J.M., Gotor, F.J., Malek, J., 2002.
Advantages of combined kinetic analysis of experimental data
obtained under any heating prole. J. Phys. Chem. A 106, 28262868.

437

Royer, S., Alamdari, H., Duprez, D., Kaliaguine, S., 2005. Oxygen storage
capacity of La1xAxBO3 perovskites (with A = Sr, Ce; B = Co, Mn)
relation with catalytic activity in the CH4 oxidation reaction. Appl.
Catal. B: Environ. 58, 273288.
Schee, J.R., Steinfeld, A., 2012. Thermodynamic analysis of ceriumbased oxides for solar thermochemical fuel production. Energy Fuels
26, 19281936.
Schunk, L.O., Steinfeld, A., 2009. Kinetics of the thermal dissociation of
ZnO exposed to concentrated solar irradiation using a solar-driven
thermogravimeter in the 18002100 K range. AIChE J. 55, 14971503.
Steinfeld, A., 2005. Solar thermochemical production of hydrogena
review. Solar Energy 78, 603615.
Tanaka, H., Misono, M., 2001. Advances in designing perovskite
catalysts. Curr. Opin. Solid State Mater. Sci. 5, 381387.
Wang, C.H., Chen, C.L., Weng, H.S., 2004. Surface properties and
catalytic performance of La1xSrxFeO3 perovskite-type oxides for
methane combustion. Chemosphere 57, 11311138.
Yang, Tan, Liang, Gu, X.H., Jin, W.Q., Zhang, L.X., Xu, N.P., 2003.
New series of Sr(Co, Fe, Zr)O3-dperovskite-type membrane materials
for oxygen permeation. Ind. Eng. Chem. Res. 42, 22992305.
Yang, H.C., Eun, H.C., Cho, Y.Z., Lee, H.S., Kim, I.T., 2009. Kinetic
analysis of dechlorination and oxidation of PrOCl by using a nonisothermal TG method. Thermochim. Acta 484, 7781.