A Novel CeO2–xSnO2/Ce2Sn2O7 Pyrochlore Cycle for Enhanced

Solar Thermochemical Water Splitting

Chongyan Ruan and Yuan Tan
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences,
Dalian 116023, China
University of Chinese Academy of Sciences, Beijing 100049, China
Lin Li, Junhu Wang, Xiaoyan Liu, and Xiaodong Wang
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences,
Dalian 116023, China

DOI 10.1002/aic.15701
Published online March 17, 2017 in Wiley Online Library (wileyonlinelibrary.com)

A novel CeO2–xSnO2/Ce2Sn2O7 pyrochlore stoichiometric redox cycle with superior H2 production capacities is identi-
fied and corroborated for two-step solar thermochemical water splitting (STWS). During the first thermal reduction step
(14008C), a reaction between CeO2 and SnO2 occurred for all the CeO2–xSnO2 (x 5 0.05–0.20) solid compounds, form-
ing thermodynamically stable Ce2Sn2O7 pyrochlore rather than metastable CeO2-d. Consequently, substantially higher
reduction extents were achieved owing to the reduction of CeIV to CeIII. Moreover, in the subsequent reoxidation with
H2O (8008C), H2 production capacities increased by a factor of 3.8 as compared to the current benchmark material
ceria when x 5 0.15, with the regeneration of CeO2 and SnO2 and the concomitant reoxidation of CeIII to CeIV. The
H2O-splitting performance for CeO2–0.15SnO2 was reproducible over seven consecutive redox cycles, indicating the
C 2017 American Institute of Chemical Engineers AIChE J, 63: 3450–3462, 2017
material was also robust. V
Keywords: solar fuel, pyrochlore, stoichiometric, water splitting, H2 production

Introduction High-temperature reduction

Solar resource, with its unmatched magnitude and availabil-
ity, is generally considered as a permanent solution to ensuring
energy security and combating climate change.1–3 Currently,
several plausible pathways to convert sunlight to fuel inter-
mediates, including photochemical,4,5 photoelectrochemi-
cal,6,7 thermochemical,8–10 and their combinations11,12 are
under consideration. In particular, solar thermochemical water
splitting (STWS), which utilizes the entire solar spectrum and
inherently operates at high temperatures, offers the potential
to convert abundant yet intermittent solar energy to renewable
hydrogen with high solar to hydrogen production rates and
efficiencies.8–10,13–16
Two-step thermochemical cycles, which eliminate the need
for in situ gaseous separation of O2 and H2,17–21 and lower the
required temperature compared to direct water thermoly-
sis,13,22 are most commonly operated in STWS. A typical two-
step STWS cycle involving metal oxide as reactive intermedi-
ate can be exemplified as follows:
Additional Supporting Information may be found in the online version of this
article.
Correspondence concerning this article should be addressed to X. Wang at
xdwang@dicp.ac.cn.
C 2017 American Institute of Chemical Engineers
V with pure Fe3O4 has been experimentally demonstrated, but
1
MOox ! MOred 1 O2 ðgÞ (1)
2
Low-temperature oxidation with H2O
MOred 1H2 OðgÞ ! MOox 1H2 ðgÞ (2)
where MOox and MOred represent the oxidized and the corre-
sponding reduced forms of the metal oxide.
During the first high-temperature reduction step, MOox is
typically reduced at low oxygen partial pressure driven by
concentrated solar energy. During the following oxidation
step, H2O is introduced to reoxidize the reduced metal oxide
back to the original oxidized state, producing hydrogen and
thus establishing a cyclic process.
Over the past decades, numerous efforts have been made to
identify viable materials for efficient two-step
STWS.14–16,23–25 However, substantial challenges still remain
in discovering feasible metal-oxide based redox cycles with
moderate operating temperatures and large gravimetric fuel
production capacities for cost-effective solar hydrogen pro-
duction. In early studies, systems based on stoichiometric
chemistries (e.g., Fe3O4 ! 3FeO 1 1=2O226,27; ZnO !
Zn 1 1=2O228,29; SnO2 ! SnO 1 1=2O230,31) were extensive-
ly studied on account of the great oxygen exchange capacities
and thus fuel productions per mass of oxide. Cycle feasibility

3450 August 2017 Vol. 63, No. 8 AIChE Journal

the kinetics are slow and magnetite must be combined with an
oxygen conducting refractory material such as yttrium-
stabilized zirconia (YSZ) or ZrO2 to suppress problems associ-
ated with melting and sintering, which limit their overall
solar-to-fuel efficiency. Conversely, for the ZnO/Zn and
SnO2/SnO processes, the volatile products require rapid
quenching to avoid recombination and to achieve products
separation, yielding great challenges for innovative reactor
design even though the global solar-to-fuel efficiency is
promising.24
Recently, two-step STWS cycles based on nonstoichiomet-
ric chemistries has attracted considerable interest for the sim-
plicity of implementation. Three classes of nonstoichiometric
oxides have been proposed as candidate materials for commer-
cial two-step STWS: doped-hercynite (CoxFe1–xAl2O4), perov-
skite (SrxLa1–xMnyAl1–yO3), and ceria (CeO2). Muhich et al.19
validated the doped-hercynite cycle governed by an O-
vacancy mechanism using DFT calculations in conjunction
with experimental measurements. Near-isothermal H2O split-
ting was achieved by cycling CoxFe1–xAl2O4 materials
between 1500 and 13508C. Slow reaction kinetics is currently
the most significant limitation for the doped-hercynite active
materials. Recent studies on perovskites (ABO3) highlight this
class of oxides as impressive candidates for solar-to-fuel con-
version. McDaniel et al.20 evaluated the nonstoichiometric
SrxLa1–xMnyAl1–yO3–d perovskites and demonstrated encour-
aging fuel productivity at a reduction temperature of 13508C.
Based on the highly amenability to doping, perovskites with
higher H2 production capacity and more rapid reaction kinetics
still remain to be identified for practical solar fuel applica-
tions. Undoped ceria has emerged as a highly attractive candi-
date with inherent fast reaction kinetics and superior
crystallographic stability, making it the current benchmark
material for comparing the performance of new redox cycles
for two-step STWS.8,21,32,33 Chueh et al.8,33 proposed a non-
stoichiometric ceria cycle via an O-vacancy mechanism as
illustrated below:
Higher temperature
d ples conditioned under various redox treatment (initial sam-
CeO2 ! CeO22d 1 O2 (3)
2
Lower temperature
CeO22d 1dH2 O ! CeO2 1dH2 (4)
The viability of the nonstoichiometric ceria cycle was success-
fully demonstrated with stable solar fuels production in a
cavity-based solar reactor under realistic solar concentrating
conditions.8,21,34 Nevertheless, the reduction to nonstoichio-
metric CeO22d shows restricted thermochemical fuel yield
despite the fast kinetics and outstanding thermal stability.
Rather high temperatures, near 20008C, were necessary to
decompose CeIVO2 to CeIII 2 O3 for efficient fuel production,
35
which impeded the practicality of undoped ceria in STWS
applications.
Doping ceria with divalent (Ca, Mg, Sr, Cu),36,37 trivalent
(Y, La, Sc, Sm, Gd, Pr, Fe, Ni, Rh),32,38–40 and tetravalent (Zr,
Hf)41–45 cations to form CeO2-based mixed oxides or solid sol-
utions has proven to be an effective way to alter its thermody-
namic and kinetic properties. The divalent and trivalent
dopants introduce stable intrinsic oxygen vacancies within the
ceria lattice and thus result in higher oxygen diffusion rates
when codoped with tetravalent dopants. The tetravalent dop-
ants induce lattice strain in bulk ceria, which in turn decreases
AIChE Journal August 2017 Vol. 63, No. 8 Published on behalf of the AIChE
the oxygen-vacancy formation energy and in consequence
enhances the reduction extents. Among them Hf-doped ceria
exhibits the best redox performance in term of the fuel produc-
tion for CO and H2, with 35% more CO and 41% more H2
evolving for Ce0.9Hf0.1O2 than undoped ceria.36,42 Moreover,
in a recent study, rhodium-doped ceria was employed in solar
thermochemical cycles with stable solar H2 and even hydro-
carbon fuels productivity by incorporating a catalytic pro-
cess.40 Despite the extensive effort, the above doping metals
based on nonstoichiometric chemistries have achieved limited
success in increasing H2 production capacities due to the par-
tial reduction of CeIV to CeIII. In previous studies, Sn doped
ceria solid solutions exhibited numerous applications as oxy-
gen gas sensors and oxygen ion conductors by virtue of their
amazing oxygen exchange properties, which relied on the
redox cycle between CeIV and CeIII within the Ce2Sn2O7-
Ce2Sn2O8 complex oxides known as pyrochlores.46 The pyro-
chlore oxides associated with the general formula A2B2O7 can
be described as cubic fluorite consisting of ordered cations and
vacancies with 1/8 of the anions missing,47 which have been
the subject of great interest motivated by the wide variety of
the chemical substitutions at the A, B, O sites, significant
oxide ion conductivity and radiation tolerance under extreme
conditions.48–54 Therefore, it would be highly anticipated that
the CeO2–xSnO2 solid compounds can potentially enhance the
reduction extents and thus improve two-step STWS perfor-
mance. However, the lack of studies related to CeO2–xSnO2
solid compounds in two-step STWS, in particular, is surprising
given the superior oxygen exchange capacities.
In light of this evidence, we synthesized a series of ceria-tin
oxide solid compounds (CeO2–xSnO2) for assessment of their
two-step STWS capabilities as compared to benchmark mate-
rial ceria. The productivity and overall reaction kinetics of the
prepared materials for two-step STWS cycled between 1400
and 8008C were systematically investigated in identical exper-
imental conditions. In parallel, crystalline phase evolution,
element-specific chemical state, microstructural morphology
change, and elemental composition distributions of the sam-
ples, samples subjected to thermal reduction, postcycling
samples), were characterized and critically analyzed via pow-
der X-ray diffraction (XRD), Raman spectroscopy, X-ray
absorption near edge structure (XANES) spectroscopy,
M€ossbauer spectroscopy, and electron microscopy (SEM,
HRTEM, STEM-EDX). Finally, insights into reaction mecha-
nism governing the two-step redox cycle were deduced based
on the experimental findings and characterization evidence.
Essentially, a new CeO2–xSnO2/Ce2Sn2O7 pyrochlore stoi-
chiometric redox cycle taking advantage of reversible change
of the oxidation state of Ce, that is, CeIV $ CeIII was identi-
fied to facilitate two-step STWS.
Experimental Section
Preparation of redox materials
All the chemicals used in the coprecipitation synthesis of
CeO2–xSnO2 (x 5 0.05, 0.10, 0.15, and 0.20, i.e., molar ratio
of Ce:Sn 5 0.95:0.05, 0.90:0.10, 0.85:0.15, and 0.80:0.20) sol-
id compounds were of analytical grade without further treat-
ment. Typically, stoichiometric amounts of Ce(NO3)36H2O
and SnCl22H2O were mixed and dissolved in deionized water
under continuous stirring. To dissolve SnCl22H2O precursor,
dilute hydrochloric acid was added slowly to the magnetically
DOI 10.1002/aic 3451
 Table 1. Nominal and Measured Material Compositions of
CeO2–xSnO2 Solid Compounds with Associated Sample
Names
Nominal Ce:Sn Measured Ce:Sn
Sample Name Molar Ratio Molar Ratio
CeO2-0.05SnO2 0.95:0.05 0.96:0.04
CeO2-0.10SnO2 0.90:0.10 0.92:0.08
CeO2-0.15SnO2 0.85:0.15 0.85:0.15
CeO2-0.20SnO2 0.80:0.20 0.79:0.21
stirred solution until pH 5 1 to obtain a transparent solution.
Subsequently, ammonium hydroxide was added dropwise to
the transparent solution until pH 5 9, under vigorously stirring
for 2 h at room temperature. The resultant slurry was aged for
another 1 h, filtered, washed thoroughly with deionized water,
and then dried at 1208C overnight. Finally, the dried solid
products were ground into a fine powder with an agate mortar
and pestle followed by thermal treatment in the flowing air
according to the program: heating from room temperature to
10008C at 58C min21, and finally heating to 14008C at 28C
min21 and holding for 4 h. This resulted in sintered, light yel-
low dense powders. For comparison, pure CeO2 sample was
prepared following the same procedure as described above
without adding SnCl22H2O precursor. It must be mentioned
that the sintering process at the high temperature relevant to
thermal cycling is crucially important to stabilize the structure
properties of the samples as prepared. Fluctuations of the sam-
ples’ performance during two-step STWS can thus be
minimized.
A summary of desired molar ratio of Ce:Sn, actual molar
ratio of Ce:Sn as determined by inductively coupled plasma-
atomic emission spectrometry (ICP-AES) analysis, and the
associated sample identifications are presented in Table 1.
Inductively coupled plasma2mass spectrometry (ICP-MS)
analysis detected no residual chlorine from the as synthesized
CeO2–0.15SnO2 sample sintered at 14008C for 4 h (Supporting
Information Table S1).
Figure 1. Experimental setup for two-step thermochemical water splitting.
[Color figure can be viewed at wileyonlinelibrary.com]
3452 DOI 10.1002/aic Published on behalf of the AIChE
Reactivity tests for thermochemical cycling
The experimental setup for two-step water splitting is
shown schematically in Figure 1. Powder Sample (0.5 g) was
loosely packed in a stagnation flow reactor as reported else-
where,9,18,55 allowing gases to access all the redox material.
The main features of the reactor are briefly summarized here.
A stainless gas-handing manifold, two concentric Al2O3 tubes
make up the reactor. Inlet gases flow along the open ended
inner tube (7 mm i.d.) downward toward the round-bottom,
close ended outer tube (14 mm i.d.) to mix well with the mate-
rial tested, and then turn by 1808 to exit via an annular region
between the two concentric tubes. The concentric alumina
tubes reactor was placed horizontally in the infrared image
furnace (ULVAC-RIKO, VTH-E44) capable of rapid heating
and cooling. Irradiation from the infrared image lamps was
regulated by a temperature program controller (ULVAC-
RIKO, TPC-5000), and temperature was monitored using an
R-type thermocouple (63.88C) enclosed in an alumina sheath
in direct contact with the sample.
Gas flows were regulated with digital mass flow controllers.
Prior to entering the reactor, traces of O2 in purge gas Ar
(99.9996% purity) was removed by passing through the deoxi-
dizer. During a typical thermochemical cycle, thermal reduc-
tion was first accomplished by increasing the sample
temperature from 8008C to 14008C at 1008C min21 in an Ar
flow (500 mL min21), and holding at 14008C for 10 min. Fol-
lowing thermal reduction, the temperature was ramped down
to 8008C at 2008C min21 in the flowing Ar. Subsequently,
H2O splitting was initiated by delivering Ar gas saturated with
purified water (500 mL min21, c(H2O) 5 4.2 mol %) to the
thermally reduced sample via a water bubbler inside a water
bath at 308C for 8 min. The passage for the steam to the reac-
tor was heated by a heater band to prevent condensation. Both
the high-temperature reduction and low-temperature H2O
splitting steps were carried out at atmospheric pressure.
Product gases (O2, H2) exiting the flow reactor were con-
stantly monitored and recorded using a quadrupole mass
August 2017 Vol. 63, No. 8 AIChE Journal
spectrometer (MS; InProcess Instruments GAM200) after con-
densing out of excess steam by a cryogenic trap. Calibration
of the mass spectrometry was performed daily prior to each
experiment with six known concentrations of O2 and H2
(1000–5000 ppm O2, 1000–5000 ppm H2, balance Ar) to an
accuracy of 0.001 mol %.
The transient O2 and H2 evolution rates were calculated as
xi Vtotal
Vi 5 (5)
mred
where Vi denotes the volumetric rates of O2 or H2 produced
per unit of mass of CeO2–xSnO2 sample; xi denotes the mole
fraction of O2 or H2 in the effluent gas, monitored by the mass
spectrometer; Vtotal is the total volumetric flow rate regulated
by the digital mass flow controller; and mred is the mass of
CeO2–xSnO2 redox material.
For quantitative determination of effluent gas, O2 and H2
yields per cycle were obtained from numerical integration of
gas evolution rate-time curves.
Characterization techniques
At the end of each processing step, the redox material was
cooled down under Ar atmosphere. Afterward, the powder
was transferred into an air-closed container under N2-atmo-
sphere for ex situ structure characterization. The crystal phase
transformation of CeO2–xSnO2 (x 5 0.05–0.20) redox samples
(samples as-synthesized, samples subjected to thermal reduc-
tion, samples after H2O splitting reaction), were identified by
X-ray powder diffraction (XRD) using a PANalytical diffrac-
tometer (Cu Ka, 40 kV, 40 mA). A continuous mode was used
for collecting data in the 2h scan range of 15–808 with an
angular step size of 0.028 and a counting time of 30 s per step.
Raman spectroscopy was conducted using a LabRam
HR800 confocal Raman microprobe (HORIBA Jobin Yvon,
France) equipped with a 532-nm wavelength He-Ne laser.
Phonon modes in the powders were investigated to fingerprint
the chemical structure of CeO2–xSnO2 (x 5 0.05–0.20) sam-
ples after various processing steps.
X-ray absorption near edge structure (XANES) spectra of
the samples (diluted with approximated 90% by mass of BN
and pressed into self-supporting wafers) at the Ce LIII-edge
were measured at beamline BL14W1 of Shanghai Synchrotron
Radiation Facility (SSRF, Shanghai, China). The electron stor-
age ring was operated at 3.5 GeV with the electron current of
140 2 210 mA. Spectra were collected in transmission mode
at room temperature using a solid-state detector. The output
beam was monochromatized by a Si(111) double-crystal
monochromator. CeO2 and Ce(NO3)36H2O wafers were used
as standard references with known oxidation states of CeIV
and CeIII, respectively. The XANES data were recorded from
5520 to 6120 eV and analyzed using a standard curve-fitting
procedure with ATHENA software program.
119
Sn Mo€ssbauer spectroscopic measurements were carried
out in transmission mode at room temperature using a constant
acceleration spectrometer (Topologic Systems. Inc.) with a
single-line c-ray source of Ca119mSnO3. The absorbers con-
taining approximately 10 mg cm22 of 119Sn were prepared
under argon atmosphere inside the glove box, and the sample
holder was then sealed to avoid contact with air. Velocity
scale calibration was performed using the magnetic sextet of a
high-purity a-Fe foil absorber as a standard, and 57Co (Rh
matrix) was used as the source. Experimental spectra were fit-
ted to appropriate combinations of Lorentzien profiles by
AIChE Journal August 2017 Vol. 63, No. 8 Published on behalf of the AIChE
least-squares method using the MossWinn 4.0 program. The
119
Sn isomer shifts were given relative to the center of the
CaSnO3 spectrum recorded at room temperature.
Surface microstructures were analyzed via scanning elec-
tron microscopy (JSM-7800F Field Emission SEM) operated
at 4 kV. Samples were placed on conductive tapes and then
sputtered with carbon. The morphology and crystallinity of the
materials were also characterized using a JEOL JEM-2000
EM field emission high-resolution transmission electron
microscope (HRTEM) operating at 200 kV. Samples after var-
ious processing steps were crushed into fine powders, followed
by ultrasonic dispersion in ethanol for 15 min. The resulting
suspension was drop-casted on copper TEM grids coated with
carbon and allowed to dry for analysis. Elemental composition
and distribution were determined via energy dispersive X-ray
spectroscopy (EDX).
Nitrogen sorption isotherm at 77 K was performed on a
Micromeritics ASAP 2460 apparatus. The samples were out-
gassed at 3008C and 1-m Torr for 4 h prior to the analysis.
Brunauer-Emmett-Teller (BET) were evaluated from the
adsorption isotherm within the pressure range 0.05 < P/P0<
0.3.
Results and Discussion
Thermal chemical H2O splitting
For thermochemical cycling applications, pure CeO2 acts as
a good baseline for comparison of new materials to evaluate
the degree of enhancement in solar fuels productivity that can
be achieved. We therefore compared the transient O2 and H2
evolution rates during a complete H2O-splitting redox cycle
obtained with CeO2–0.15SnO2 and pure CeO2 under identical
experimental conditions. The results are presented in Figures
2a, b. During the thermal reduction step (Figure 2a), O2 evolu-
tion starts at above 9008C for both CeO2–0.15SnO2 and pure
CeO2. Intriguingly, O2 evolution from CeO2–0.15SnO2
increases more rapidly initially in comparison with pure CeO2,
attaining a peak rate of 2.83 mL min21 g21 at 11008C and
0.30 mL min21 g21 at 14008C for CeO2–0.15SnO2 and pure
CeO2, respectively. Notably, fully 4.73 more O2 is released
for CeO2–0.15SnO2 (4.87 mL g21) than pure CeO2 (1.03 mL
g21). During the following oxidation with H2O at 8008C (Fig-
ure 2b), H2 is immediately produced on introduction of the
steam in both cases. Pure CeO2 exhibits an initial rapid
increase to a peak rate of 5.08 mL min21 g21 followed by
exponential decline. Whereas for CeO2–0.15SnO2, the peak
H2 evolution rate drops to 2.96 mL min21 g21 followed by a
comparably slower decrease rate. Conversely, compared to
pure CeO2, CeO2–0.15SnO2 did not achieve its thermodynam-
ic equilibrium, as indicates by long tail of H2 evolution after
reoxidation at 8008C for 8 min. And it was due to the compa-
rably short reoxidation period (see Supporting Information
Figure S1 for further information). More importantly, the
enhanced reduction extent for CeO2–0.15SnO2 leads to a
higher H2 yield during the subsequent oxidation with H2O, for
which H2 production capacity (CeO2–0.15SnO2, 7.18 mL g21;
pure CeO2, 1.87 mL g21) increases by a factor of 3.8 despite
the initial H2 evolution rate being lower on introduction of the
steam.
To get further insight into the influence of SnO2 content on
the thermal reduction capabilities and the subsequent H2O
splitting kinetics, CeO2–xSnO2 (x 5 0.05–0.20) together with
pure CeO2 were assessed systematically. The thermochemical
DOI 10.1002/aic 3453
 splitting performances are also sufficiently favored owing to
the enhanced reduction extents.
To assess the stability of the water-splitting performance,
CeO2–0.15SnO2 was subjected to seven consecutive redox
cycles at 14008C reduction/8008C reoxidation. As seen in Fig-
ure 3b, the O2 and H2 productivity are roughly constant with-
out obvious deactivation from cycle 1 to cycle 7. The average
yields of O2 and H2 for CeO2–0.15SnO2 are 4.64 and 7.05 mL
g21, respectively, with very slightly decreases compared with
the values reported in Supporting Information Table S2, indi-
cating rather stable fuel productivity. ICP-AES analysis shows
the molar ratio of Ce:Sn for CeO2–0.15SnO2 sample after
cycling is consistent with that of initial sample sintered at
14008C for 4 h (Supporting Information Table S3).
To probe into the underlying reason for the enhanced reduc-
tion extents with sufficiently favored H2O splitting performan-
ces of ceria-tin oxide solid compounds when compared with

Figure 2. Comparison of transient (a) O2 evolution rates

during thermal reduction at 14008C in Ar, (b) H2
evolution rates during the following reoxidation
at 8008C in H2O/Ar for CeO2–0.15SnO2 and pure
CeO2.
Experimental conditions were as follows: Ar sweep gas
flow rate 5 1000 mL min21 g21 during thermal reduc-
tion; cH2O 5 4.2 mol %, total gas flow rate 5 1000 mL
min21 g21 during the following reoxidation. Inset shows
the integrated O2 and H2 productivity, respectively.
[Color figure can be viewed at wileyonlinelibrary.com]

behaviors are depicted in Figure 3a (see Supporting Informa-

tion Figure S2 and Table S2 for details). We note that for all
the CeO2–xSnO2 (x 5 0.05–0.20) solid compounds, total
amounts of oxygen evolved are improved significantly when
compared with pure CeO2, and increase linearly with increas-
ing SnO2 content. Given that all the samples were evaluated
under the same conditions, we thus presume the enhanced O2
yields mainly stem from more favorable reduction thermody-
namics. H2 yields, in excellent correspondence with the
expectations, also increase linearly with the improved reduc-
tion extent for x 5 0.05–0.15. It is noteworthy that H2 yields Figure 3. (a) O2 and H2 yield of CeO2–xSnO2 (x 5 0.05–
increase by a factor of 1.3 and even 3.8 for x 5 0.05 and 0.20) solid compounds as compared to pure
x 5 0.15, respectively (Supporting Information Figure S2 and CeO2. (b) O2 and H2 yield of CeO2–0.15SnO2
Table S2). However, for higher SnO2 content of x 5 0.20, H2 over seven consecutive H2O splitting cycles.
yield decreases remarkably albeit the highest O2 yield is The same experimental conditions were employed and
observed, which may result from the severe sintering as will illustrated as follows: Ar sweep gas at a flow rate of
be shown in the subsequent discussion. In summary, the reduc- 1000 mL21 min21 g21 during the thermal reduction at
14008C, and H2O/Ar at cH2O 5 4.2 mol %, for which
tion extents for CeO2–xSnO2 (x 5 0.05–0.20) are substantially total gas flow rate 5 1000 mL21 min21 g21 during the
improved in comparison with pure CeO2, and increase linearly reoxidation at 8008C. [Color figure can be viewed at
with increasing SnO2 content. Moreover, the subsequent H2O wileyonlinelibrary.com]

3454 DOI 10.1002/aic Published on behalf of the AIChE August 2017 Vol. 63, No. 8 AIChE Journal
pure CeO2, crystalline phase, element-specific chemical state,
microstructural morphology, and elemental composition distribu-
tions of the samples (initial samples, samples subjected to thermal
reduction, samples after the subsequent H2O splitting cycle) of
various processing steps, were characterized via XRD together
with Raman spectroscopy, XANES spectroscopy, M€ossbauer
spectroscopy, as well as electron microscopy (SEM, HRTEM,
STEM-EDX).
Sample characterization
Figure 4a displays the X-ray powder diffraction patterns of
initial CeO2–xSnO2 (x 5 0.05–0.20) solid compounds thermal-
ly calcined at 14008C for 4 h. Pure CeO2 is also shown as a
reference. For all the as synthesized samples, reflection peaks
indexable to the cubic fluorite CeO2 (JCPDS 00-043-1002)
and rutile type tetragonal SnO2 (JCPDS 01-077-0447) are
observed, indicating the samples prior to use are in the form of
mixed metal oxides. Additionally, the intensity of SnO2 dif-
fraction peaks, as expected, monotonically increases with
increasing SnO2 amount. The X-ray powder diffraction pat-
terns of CeO2–xSnO2 (x 5 0.05–0.20) solid compounds after
thermal reduction at 14008C in flowing Ar for 10 min are
shown in Figure 4b. Interestingly, for all the samples thermally
reduced, reflection peaks assignable to the SnO2 completely
disappear (with traces of SnO2 only detectable for x 5 0.20)
along with the emergence of a new superstructure reflection
attributable to the Fd3m (no. 227) symmetry of a typical
Ce2Sn2O7 pyrochlore phase. The newly emerged reflection
peaks (highlight by the gray bars in Figures 4b, c) can be
indexed as (222), (400), (440), and (622) planes of Ce2Sn2O7
pyrochlore structure.56 Furthermore, the increase of the corre-
sponding Ce2Sn2O7 pyrochlore reflections is coupled with the
decrease of the cubic fluorite CeO2 phase. This points out that
a reaction between CeO2 and SnO2 has occurred, resulting in
the reduction or even disappearance of the reflection peaks of
CeO2 and SnO2. No extra peaks originating from the other
phases are observed for pure CeO2 sample thermally reduced
(Figure 4b). The above XRD observations confirm that during
thermal reduction in flowing Ar, a stoichiometric reaction
occurs between CeO2 and SnO2, forming thermodynamically
favorable Ce2Sn2O7 pyrochlore. These observations are also in
agreement with the enhanced O2 production capacities, which
increase linearly with increasing SnO2 content as shown in
Figure 3a. Figure 4. X-ray powder diffraction patterns of CeO2–xSnO2
Notably, after the subsequent H2O splitting step with the solid compounds: (a) as synthesized samples
thermally reduced CeO2–xSnO2 (x 5 0.05–0.20) solid com- calcined at 14008C, (b) samples thermally
pounds, the intensities of the Ce2Sn2O7 pyrochlore reflection reduced in flowing Ar, (c) samples after the sub-
peaks decrease remarkably, with only traces of Ce2Sn2O7 sequent H2O splitting step.
pyrochlore phase present when x 5 0.05–0.15 (Figure 4c). Crystalline phases: (•) CeO2; (*) SnO2; (䉬) Ce2Sn2O7.
This reveals that the reduced Ce2Sn2O7 pyrochlore is almost The gray bars highlight reflection peaks assignable to
completely oxidized. From these results, it is concluded that Ce2Sn2O7 pyrochlore. Pure CeO2 after various process-
during reoxidation with H2O, Ce2Sn2O7 pyrochlore is oxidized ing steps are also shown as references. [Color figure can
be viewed at wileyonlinelibrary.com]
back to CeO2 and SnO2. Additionally, no diffraction peak
characteristic of SnO2 is resolved, indicating a high dispersion
of SnO2 in ceria-tin oxide solid compounds after the reoxida-
tion step. On the contrary, in the case of CeO2–xSnO2 (x 5 0.05–0.20) solid compounds. Raman spectra of samples
(x 5 0.20), Ce2Sn2O7 pyrochlore is still visible, thus implying after various processing steps are compared in Figure 5a (see
a fraction of the reduced Ce2Sn2O7 pyrochlore has not been Supporting Information Figures S3 and S4 for more details).
oxidized back by H2O, resulting in the remarkable decrease in The spectra are plotted on the same relative intensity scale.
H2 yield presented in Figure 3a. All the as synthesized samples calcined at 14008C exhibit
As an effective tool to analyze the M-O bonds as well as lat- spectral features indicative of ideal crystalline CeO2, with an
tice defects, Raman spectroscopy was employed to further intense band centered at ca. 465 cm21 (Figure 5a) related to
characterize the crystalline phase evolution of CeO2–xSnO2 the F2g Raman active mode of the fluorite-type structure

AIChE Journal August 2017 Vol. 63, No. 8 Published on behalf of the AIChE DOI 10.1002/aic 3455

Figure 5. (a) Raman spectra of CeO2–xSnO2 solid com-
pounds after various processing steps (as
synthesized samples calcined at 14008C,
samples thermally reduced in flowing Ar,
samples after the subsequent H2O splitting
cycle) with different SnO2 content. (b) Plot of
DF2g Raman shift of samples thermally
reduced as compared to the as synthesized
samples (blue shift) and samples after the
subsequent H2O splitting cycle as compared
to samples thermally reduced (red shift) as a
function of SnO2 molar ratio.
[Color figure can be viewed at wileyonlinelibrary.com]
(Supporting Information Figure S5a).57 The F2g peak origi-
nates from the symmetrical stretching vibration of the oxygen
atoms located around eight-fold coordination CeIV ions (Sup-
porting Information Figure S5d).58–60 Furthermore, weaker
Raman bands at ca. 479 cm21, which is the characteristic Eg
vibration mode of cassiterite SnO2,61 are observed with
x 5 0.15 and x 5 0.20 (Supporting Information Figure S3a).
However, this low-intensity band at ca. 479 cm21 is absent
with x 5 0.05 and x 5 0.10 (Supporting Information Figure
S3a) attributable to the comparably lower SnO2 content. The
above observations clearly confirm that SnO2 has not been
incorporated into the CeO2 framework of the initial samples,
as evidenced by the X-ray diffraction patterns in Figure 4a.
Importantly, all the samples thermally reduced in following
Ar generally decrease remarkably in the intensity of the F2g
peaks, accompanied by a shift to lower wavenumbers with
broader modes (as displayed in Figure 5a). By closer inspec-
tion, a continuous blue shift of 25 to 210 cm21 is measured
3456 DOI 10.1002/aic Published on behalf of the AIChE
for all the samples thermally reduced as compared to the as
synthesized samples, for which DF2g Raman shift is almost
linear with SnO2 molar ratio, Figure 5b. These phenomena are
qualitatively consistent with other Ce12xRExO22y systems,
where RE 5 La, Pr, Nd, Eu, Gd, and Sm.58–60,62–64 For exam-
ple, Banerji et al.59 and Giannici et al.63 reported that as the
M2O3 (M 5 Gd, Sm, Yb) concentration in CeO2 increases, the
Raman spectrum changes considerably with increases in width
and a decrease of the intense mode of CeO2 at 465cm21. In a
recent study, Li et al.65 also demonstrated that the insertion of
Sn ions into CeO2 lattice to form a solid solution results in a
systematic shift of the band at 465cm21 to lower frequencies
with increasing Sn content. Considering that the blue-shift
(red-shift) of Raman F2g peak is associated with lattice expan-
sion (contraction) and weakening (strengthening) of local
cation-anion bonds,58,60 the decreased intensity and the blue
shift of F2g peaks resulting from lattice expansion explicitly
indicate the incorporation of SnO2 into CeO2 fluorite structure
after thermal reduction, with the reduction of CeIV (0.97 Å at
octahedral coordination) ions to bigger CeIII (1.143 Å at octa-
hedral coordination) as supported by the XANES results
below. This is in good accordance with conclusion from afore-
mentioned XRD observations that a continuous reaction
between CeO2 and SnO2 occurs with increasing SnO2 content.
It is noteworthy that after the subsequent H2O splitting step,
the intensity of the F2g peaks increase with a red shift ranging
between 13 to 15 cm21 for cycled samples when x 5 0.05–
0.15 (Figures 5a, b), implying the regeneration of CeO2
fluorite-type structure and lattice contraction with the oxida-
tion of CeIII to smaller CeIV ions. However, for the cycled
sample with x 5 0.20, the F2g peak is neither enhanced in
intensity nor shifted back to higher wavenumbers, indicating
CeO2 and SnO2 are not regenerated completely during reoxi-
dation with H2O. These results are in line with the XRD inves-
tigations shown in Figure 4c. Meanwhile, cassiterite SnO2,
which falls below XRD detection limits (Figure 4c), is identi-
fied with characteristic weak Raman bands at ca. 479 cm21
(Supporting Information Figure S3c). This suggests the migra-
tion of SnO2 out of CeO2 fluorite structure which alleviates
CeO2 lattice distortions after reoxidation with H2O.
To provide element-specific electronic configurations of the
metals studied and to give better insight into redox chemistry,
valence states of Ce and Sn in CeO2–0.15SnO2 solid com-
pound featuring the highest H2 yield were determined by Ce
LIII-edge XANES spectroscopy and 119Sn M€ossbauer spectros-
copy. Figure 6 shows the normalized XANES spectra for
CeO2–0.15SnO2 after various processing steps together with
the spectra of standard reference compounds at the Ce LIII
edge. XANES spectrum of CeIII(NO3)36H2O (red) is dominat-
ed by one intense absorption band at around 5728 eV. The
spectrum of CeIVO2 (black) exhibits a typical double white
line 5 eV apart, which stems from the two cerium ground-state
electronic configurations of 4f0 and 4f1.66,67 For the initial
CeO2–0.15SnO2 (green), the spectrum shows exclusively fea-
tures characteristic of CeIV. Negligible shifts in the edge posi-
tion of the Ce LIII absorption edge is observed as compared to
the reference spectra of CeO2. The inset highlights the shift of
the absorption edge position of CeO2–0.15SnO2 solid com-
pound at three different stages of thermochemical H2O split-
ting cycling. We note that for CeO2–0.15SnO2 after thermal
reduction in following Ar (blue), a shift in the edge position to
lower photon energy is evident concomitant with appearance
of a new peak corresponding to CeIII (Figure 6) when
August 2017 Vol. 63, No. 8 AIChE Journal
 s21). This finding implies an evolution from crystalline SnO2
to Ce2Sn2O7 pyrochlore occurs after thermal reduction, corre-
lating well with the XRD results in Figure 4. The relatively
higher IS value suggests an increase in Sn 5s electron density
due to the electron transfer from neighboring Ce to Sn in
Ce2Sn2O7 pyrochlore, while the larger QS value indicating a
more asymmetric environment originating from the surround-
ing Ce atoms as the original Sn neighbors in SnO2 are replaced
with Ce in Ce2Sn2O7 pyrochlore (Supporting Information Fig-
ures S5c, f). Note that no detectable spectroscopic indications

Figure 6. Normalized XANES spectra for CeO2–

0.15SnO2 solid compounds after various
processing steps at the Ce LIII edge togeth-
er with the spectra of standard reference
compounds. Inset represents the shift of
the absorption edge position of CeO2–
0.15SnO2 solid compound as synthesized
(green), reduced (blue), and cycled (brown).
[Color figure can be viewed at wileyonlinelibrary.com]

compared with the as synthesized sample (green), indicating a

reduction from CeIV to CeIII happens. For CeO2–0.15SnO2
samples after the subsequent H2O splitting cycle (brown),
there is a shift in the edge position back to higher photon ener-
gy, which confirms that CeIII reverts back to CeIV during reox-
idation with H2O. Additionally, a slight shift to higher photon
energy compared to the initial spectrum is observed, revealing
the incomplete reoxidation of cerium, which is likely due to
the comparably short reoxidation period (8 min).
119
Sn M€ossbauer spectroscopy is traditionally considered as
the most sensitive tool for probing the oxidation state and the
local structural symmetry of Sn on the atomic scale.68,69 Two
characteristic hyperfine parameters can be extracted from the
measured spectra: isomer shift (IS) and quadrupole splitting
(QS). IS reflects the changes to the s-electrons density, which
is induced by electron-donating or electron-withdrawing
groups and, to a greater extent, by the change of the valence
state. QS indicates the local structural symmetry of the probed
Sn nucleus, with asymmetric environments corresponding to
large splitting of the nuclear energy levels.
The experimental and fitted 119Sn M€ossbauer spectra of
CeO2–0.15SnO2 solid compounds (as synthesized, reduced,
and cycled) are illustrated in Figure 7 and the resulting hyper-
fine parameters of the corresponding fits are included in Sup-
porting Information Table S4. The spectrum of the initial
CeO2–0.15SnO2 (Figure 7a) shows a symmetric Lorentzian Figure 7. Room temperature 119Sn Mo € ssbauer spectra
doublet that can be unambiguously ascribed to typical crystal- of CeO2–0.15SnO2 solid compounds (a) as
line SnO2. The isomer shift close to 0 mm s21 (IS 5 0.088 mm synthesized, (b) thermally reduced in flowing
s21) is characteristic of the SnIV oxidation state and the quad- Ar, and (c) after the subsequent H2O splitting
rupole splitting of QS 5 0.528 mm s21 arises from the intrin- cycle. Structures inset corresponding to the
sic asymmetry of the SnO6 octahedron in the well-crystallized local coordination of (a) SnO2 moiety, (b)
SnO2 (Supporting Information Figures S5b, e). For CeO2– Ce2Sn2O7 moiety, (c) SnO2 and Ce2Sn2O7
0.15SnO2 sample thermally reduced (Figure 7b), the moieties as assigned in literature.
M€ossbauer experimental data can be fitted to a symmetric Lor- The red sphere represents an oxygen atom, whereas
entzian doublet with a higher value of an isomer shift and a green and brown spheres identify tin and cerium atoms,
respectively. [Color figure can be viewed at wileyonline-
quadrupole splitting which can be assigned to SnIV in library.com]
Ce2Sn2O7 pyrochlore (IS 5 0.163 mm s21, QS 5 0.654 mm

AIChE Journal August 2017 Vol. 63, No. 8 Published on behalf of the AIChE DOI 10.1002/aic 3457
Figure 8. Representative SEM images of initial CeO2–xSnO2 samples thermally calcined at 14008C with different
SnO2 content: (a) x 5 0.05, (b) x 5 0.10, (c) x 5 0.15, (d) x 5 0.20, and (e) pure CeO2.

for the reduced Sn species, such as SnII and Sn0, are observed,
which clearly demonstrates no change in SnIV oxidation state
occurs for CeO2–0.15SnO2 sample thermally reduced. For
CeO2–0.15SnO2 samples after the subsequent H2O splitting
cycle (Figure 7c), the spectrum consists of two symmetric Lor-
entzian doublets corresponding to SnO2 and Ce2Sn2O7 with an
area fraction of, respectively, 73% and 27%. This result is
well in accordance with the O2 and H2 productivity shown in
Supporting Information Table S2 (O2 yield: 4.87 mL g21, H2
yield: 7.18 mL g21), where 74% of the reduced Ce2Sn2O7 was
reoxidized to produce H2. To sum up, the XANES and
M€ossbauer spectroscopy results provide unambiguous evi-
dence that the thermochemical H2O splitting cycling with
CeO2–0.15SnO2 rely on the facile and reversible change of the
oxidation state of Ce (CeIVO2 $ CeIII 2 Sn2O7); meanwhile,
instead of an innocent spectator, the presence of Sn facilitates
the reversibility of the reduction/oxidation between CeIV/CeIII
for forming the thermodynamically stable Ce2Sn2O7 pyro-
chlore intermediate.
Structures and morphologies of the as-synthesized and ther-
mochemical cycled samples were studied by scanning electron
microscopy (SEM). Figure 8 shows the typical SEM micro-
graphs of initial CeO2–xSnO2 (x 5 0.05–0.20) samples prior to
thermochemical cycling. For comparison, a SEM image of
pure CeO2 is given in Figure 8e. Pure CeO2 mainly consists of
small particles with primary diameters of around 0.5 lm.
CeO2–xSnO2 (x 5 0.05, 0.10) exhibits a relatively homoge-
neous microstructure with smooth surfaces and grain sizes of
approximately 1 lm (Figures 8a, b). However, lower homoge-
neity is observed with x 5 0.15 and x 5 0.20 (Figures 8c, d).

3458 DOI 10.1002/aic Published on behalf of the AIChE August 2017 Vol. 63, No. 8 AIChE Journal
Figure 9. (a) High-resolution TEM image of initial CeO2–0.15SnO2 samples thermally calcined at 14008C, along with
the respective FFT patterns characteristics of cubic fluorite CeO2. Representative STEM image and EDX
elemental map of initial CeO2–0.15SnO2 sample thermally calcined at 14008C (b–e), sample thermally
reduced in flowing Ar (f–i), and sample after thermochemical H2O splitting cycling (j–m).
[Color figure can be viewed at wileyonlinelibrary.com]

For x 5 0.15, small particles and large agglomerates appear
mixed with sizes ranging between 1 and 3 lm (Figure 8c).
Further increasing SnO2 content to x 5 0.20, small grains are
almost vanished and only large agglomerates more than 5 lm
are evident (Figure 8d). The BET surface area of the as-
synthesized CeO2–xSnO2 (x 5 0.05–0.20) and pure CeO2 are
summarized in Supporting Information Table S5. The lowest
surface area of 0.21 m2 g21 was measured for CeO2–
0.20SnO2. After thermochemical H2O splitting cycling, parti-
cle growth is observed over all the samples (Supporting Infor-
mation Figure S6). The grain boundaries of all the cycled
samples appear to be completely fused. Remarkably, CeO2–
0.20SnO2 undergoes dramatic sintering with grains significant-
ly larger than CeO2–xSnO2 (x 5 0.05–0.15) and pure CeO2.
CeO2–0.15SnO2 solid compound was further analyzed by
TEM and STEM-EDX. For the initial CeO2–0.15SnO2 sample
thermally calcined at 14008C, the clear lattice fringes observed
in high-resolution TEM and the corresponding fast Fourier
transform (FFT) diffraction pattern (Figure 9a) indexed to
cubic fluorite CeO2 are revealed. However, cassiterite SnO2 is
not identified, although some variability in contrast across the
investigated region is evident (refer to Supporting Information
Figure S7 for further details). Conversely, STEM-EDX analy-
sis does reveal the presence of CeO2 and SnO2 with Ce and Sn
emissions (Figures 9b–e). Moreover, EDX mapping of Ce and
Sn displays an inhomogeneous distribution through the as-
synthesized sample prior to thermochemical cycling, as evi-
dent by the Ce/Sn segregation phases (Figures 9b-e).

AIChE Journal August 2017 Vol. 63, No. 8 Published on behalf of the AIChE DOI 10.1002/aic 3459
Following thermal reduction in the following Ar, CeO2–
0.15SnO2 sample in the reduced state exhibits significant mix-
ing of the CeO2 and SnO2 phases with a homogeneous disper-
sion of Ce and Sn elements (Figures 9f–i), further confirming
that a reaction between CeO2 and SnO2 occurs after thermal
reduction. Additionally, for sample after thermochemical H2O
splitting cycling, Ce and Sn elements are also well mixed (Fig-
ures 9j–m). See Supporting Information Figure S8 for further
details. In accordance, we can conclude that of the initial
CeO2–xSnO2 (x 5 0.05–0.20) solid compounds, SnO2 was dis-
tributed unevenly on CeO2, forming mixed metal oxides. After
thermal reduction, a reaction between CeO2 and SnO2 occurs,
resulting in a homogeneous dispersion of Ce and Sn elements.
Reaction mechanism for the novel CeO2–xSnO2/
Ce2Sn2O7 pyrochlore cycle
The experimental findings of two-step thermochemical H2O
splitting performance and the characterization evidence pre-
sented above provide a basis for refining the mechanism for
the novel CeO2–xSnO2/Ce2Sn2O7 pyrochlore cycle. We con-
clude that instead of proceeding via the nonstoichiometric O-
vacancy mechanism for pure CeO2 as embodied by Eqs. 3 and
4, a new CeO2–xSnO2/Ce2Sn2O7 pyrochlore stoichiometric
redox cycle is involved for CeO2–xSnO2 in two-step thermo-
chemical H2O splitting. This new redox cycle is illustrated as
follows:
High-temperature reduction
1 1
CeIV O2 1SnO2 ! CeIII 2 Sn2 O7 1 O2 ðgÞ (6)
2 4
Low-temperature reoxidation with H2O
1 III 1 1
Ce 2 Sn2 O7 1 H2 OðgÞ ! CeIV O2 1 SnO2 1 H2 ðgÞ (7)
2 2 2 5. Xiang Q, Cheng B, Yu J. Graphene-based photocatalysts for solar-
To summarize, during the first high-temperature reduction
step (14008C), a stoichiometric reaction between inhomoge-
neous distributed CeO2 and SnO2 occurred, forming thermo-
dynamically stable Ce2Sn2O7 pyrochlore instead of metastable
CeO2-d (Eq. 6). As a result, enhanced reduction extent was
accomplished on account of the reduction of CeIV to CeIII.
During the subsequent low-temperature reoxidation with H2O
(8008C), the reduced Ce2Sn2O7 pyrochlore was reoxidized
with CeIII reverted back to CeIV. Meanwhile CeO2 and SnO2
were regenerated with a homogeneous dispersion, completing
the redox cycle between CeIV and CeIII at modest temperatures
(Eq. 7).
Conclusions
In this contribution, two-step STWS operating via a stoi-
chiometric reaction mechanism was successfully demonstrated
with the successive insertion/extraction of SnO2 within the
intriguing CeO2–xSnO2/Ce2Sn2O7 pyrochlore cycle. The nov-
el pyrochlore cycle was corroborated using X-ray diffraction,
Raman spectroscopy, XANES spectroscopy, M€ossbauer spec-
troscopy, in combination with electron microscopy (SEM,
HRTEM, STEM-EDX). The first thermal reduction proceeded
via a stoichiometric reaction between CeO2 and SnO2, forming
thermodynamically stable Ce2Sn2O7 pyrochlore rather than
metastable CeO2-d. Consequently, enhanced reduction extents
were achieved for all the CeO2–xSnO2 (x 5 0.05–0.20) solid
compounds owing to the reduction of CeIV to CeIII. O2 produc-
tion capacities were improved significantly when compared
with pure CeO2, which increased linearly with increasing
3460 DOI 10.1002/aic Published on behalf of the AIChE
SnO2 content. During the subsequent low-temperature reoxi-
dation with H2O, CeO2, and SnO2 were regenerated, complet-
ing the redox cycle between CeIV and CeIII at 14008C
reduction/8008C reoxidation. Encouragingly, H2 yield for
CeO2–0.15SnO2 was improved by a factor of 3.8 due to the
enhanced reduction extent when compared with the state-of-
the-art CeO2, with stable O2 and H2 productivity over seven
cycles.
It may be anticipated that this novel compound-forming
stoichiometric redox chemistry can be readily implemented
and exploited with other pyrochlore oxides exhibiting wide
varieties of chemical substitution, thus opening up new oppor-
tunities for identifying even more promising materials and
accelerating realization of solar fuels production based on
thermochemical cycles.
Acknowledgments
This work was supported by National Natural Science
Foundation of China (21476232, 21573232, 21522608,
21676269) and the “Strategic Priority Research Program” of
the Chinese Academy of Sciences (XDB17020100).
Literature Cited
1. Lewis NS, Nocera DG. Powering the planet: chemical challenges in
solar energy utilization. Proc Natl Acad Sci USA. 2006;103:15729–
15735.
2. Trainham JA, Newman J, Bonino CA, Hoertz PG, Akunuri N.
Whither solar fuels? Curr Opin Chem Eng. 2012;1:204–210.
3. Centi G, Perathoner S. Towards solar fuels from water and CO2.
ChemSusChem. 2010;3:195–208.
4. Liu J, Liu Y, Liu N, Han Y, Zhang X, Huang H, Lifshitz Y, Lee S-
T, Zhong J, Kang Z. Metal-free efficient photocatalyst for stable vis-
ible water splitting via a two-electron pathway. Science. 2015;347:
970–974.
fuel generation. Angew Chem Int Ed. 2015;54:11350–11366.
6. Luo J, Im J-H, Mayer MT, Schreier M, Nazeeruddin MK, Park N-G,
Tilley SD, Fan HJ, Graetzel M. Water photolysis at 12.3% efficiency
via perovskite photovoltaics and earth-abundant catalysts. Science.
2014;345:1593–1596.
7. Cox CR, Lee JZ, Nocera DG, Buonassisi T. Ten-percent solar-to-
fuel conversion with nonprecious materials. Proc Natl Acad Sci
USA. 2014;111:14057–14061.
8. Chueh WC, Falter C, Abbott M, Scipio D, Furler P, Haile SM,
Steinfeld A. High-flux solar-driven thermochemical dissociation of
CO2 and H2O using nonstoichiometric ceria. Science. 2010;330:
1797–1801.
9. Muhich CL, Evanko BW, Weston KC, Lichty P, Liang X, Martinek
J, Musgrave CB, Weimer AW. Efficient generation of H2 by split-
ting water with an isothermal redox cycle. Science. 2013;341:540–
542.
10. Roeb M, Sattler C. Isothermal water splitting. Science. 2013;341:
470–471.
11. Meng X, Wang T, Liu L, Ouyang S, Li P, Hu H, Kako T, Iwai H,
Tanaka A, Ye J. Photothermal conversion of CO2 into CH4 with H2
over group VIII nanocatalysts: an alternative approach for solar fuel
production. Angew Chem Int Ed. 2014;53:11478–11482.
12. Chanmanee W, Islam MF, Dennis BH, MacDonnell FM. Solar pho-
tothermochemical alkane reverse combustion. Proc Natl Acad Sci
USA. 2016;113:2579–2584.
13. Fletcher EA, Moen RL. Hydrogen and oxygen from water. Science.
1977;197:1050–1056.
14. Funk JE. Thermochemical hydrogen production: past and present.
Int J Hydrogen Energy. 2001;26:185–190.
15. Kodama T, Gokon N. Thermochemical cycles for high-temperature
solar hydrogen production. Chem Rev. 2007;107:4048–4077.
16. Perkins C, Weimer AW. Solar-thermal production of renewable
hydrogen. AIChE J. 2009;55:286–293.
17. Arifin D, Aston VJ, Liang X, McDaniel AH, Weimer AW. CoFe2O4
on a porous Al2O3 nanostructure for solar thermochemical CO2 split-
ting. Energy Environ Sci. 2012;5:9438.
August 2017 Vol. 63, No. 8 AIChE Journal
18. Scheffe JR, McDaniel AH, Allendorf MD, Weimer AW. Kinetics and
mechanism of solar-thermochemical H2 production by oxidation of a
cobalt ferrite–zirconia composite. Energy Environ Sci. 2013;6:963.
19. Muhich CL, Ehrhart BD, Witte VA, Miller SL, Coker EN,
Musgrave CB, Weimer AW. Predicting the solar thermochemical
water splitting ability and reaction mechanism of metal oxides: a
case study of the hercynite family of water splitting cycles. Energy
Environ Sci. 2015;8:3687–3699.
20. McDaniel AH, Miller EC, Arifin D, Ambrosini A, Coker EN, O’hayre
R, Chueh WC, Tong J. Sr- and Mn-doped LaAlO32d for solar thermo-
chemical H2 and CO Production. Energy Environ Sci. 2013;6:2424.
21. Furler P, Scheffe JR, Steinfeld A. Syngas production by simulta-
neous splitting of H2O and CO2 via ceria redox reactions in a high-
temperature solar reactor. Energy Environ Sci. 2012;5:6098–6103.
22. Ohya H, Yatabe M, Aihara M, Negishi Y, Takeuchi T. Feasibility of
hydrogen production above 2500 K by direct thermal decomposition
reaction in membrane reactor using solar energy. Int J Hydrogen
Energy. 2002;27:369–376.
23. Loutzenhiser PG, Meier A, Steinfeld A. Review of the two-step
H2O/CO2-splitting solar thermochemical cycle based on Zn/ZnO
redox reactions. Materials. 2010;3:4922–4938.
24. Scheffe JR, Steinfeld A. Oxygen exchange materials for solar ther-
mochemical splitting of H2O and CO2: a review. Mater Today.
2014;17:341–348.
25. Muhich CL, Ehrhart BD, Al-Shankiti I, Ward BJ, Musgrave CB,
Weimer AW. A review and perspective of efficient hydrogen genera-
tion via solar thermal water splitting. Wiley Interdiscip Rev Energy
Environ. 2016;5:261–287.
26. Gokon N, Hasegawa T, Takahashi S, Kodama T. Thermochemical
two-step water-splitting for hydrogen production using Fe-YSZ par-
ticles and a ceramic foam device. Energy. 2008;33:1407–1416.
27. Gokon N, Murayama H, Umeda J, Hatamachi T, Kodama T. Mono-
clinic zirconia-supported Fe3O4 for the two-step water-splitting ther-
mochemical cycle at high thermal reduction temperatures of 1400–
16008C. Int J Hydrogen Energy. 2009;34:1208–1217.
28. Stamatiou A, Loutzenhiser PG, Steinfeld A. Solar syngas production
via H2O/CO2-splitting thermochemical cycles with Zn/ZnO and FeO/
Fe3O4 redox reactions. Chem Mater. 2010;22:851–859.
29. Stamatiou A, Loutzenhiser PG, Steinfeld A. Syngas production from
H2O and CO2 over Zn particles in a packed-bed reactor. AIChE J.
2012;58:625–631.
30. Charvin P, Abanades S, Lemont F, Flamant G. Experimental study
of SnO2/SnO/Sn thermochemical systems for solar production of
hydrogen. AIChE J. 2008;54:2759–2767.
31. Lev^eque G, Abanades S, Jumas J-C, Olivier-Fourcade J. Characteri-
zation of two-step tin-based redox system for thermochemical fuel
production from solar-driven CO2 and H2O splitting cycle. Ind Eng
Chem Res. 2014;53:5668–5677.
32. Chueh WC, Haile SM. Ceria as a thermochemical reaction medium
for selectively generating syngas or methane from H2O and CO2.
ChemSusChem. 2009;2:735–739.
33. Chueh WC, Haile SM. A Thermochemical study of ceria: exploiting
an old material for new modes of energy conversion and CO2 miti-
gation. Philos Trans R Soc A. 2010;368:3269–3294.
34. Marxer D, Furler P, Scheffe J, Geerlings H, Falter C, Batteiger V,
Sizmann A, Steinfeld A. Demonstration of the entire production
chain to renewable kerosene via solar thermochemical splitting of
H2O and CO2. Energy Fuels. 2015;29:3241–3250.
35. Abanades S, Flamant G. Thermochemical hydrogen production from
a two-step solar-driven water-splitting cycle based on cerium oxides.
Sol Energy. 2006;80:1611–1623.
36. Meng Q-L, Lee C-I, Ishihara T, Kaneko H, Tamaura Y. Reactivity
of CeO2-based ceramics for solar hydrogen production via a two-
step water-splitting cycle with concentrated solar energy. Int J
Hydrogen Energy. 2011;36:13435–13441.
37. Scheffe JR, Steinfeld A. Thermodynamic analysis of cerium-based
oxides for solar thermochemical fuel production. Energy Fuels.
2012;26:1928–1936.
38. Kaneko H, Ishihara H, Taku S, Naganuma Y, Hasegawa N, Tamaura
Y. Cerium ion redox system in CeO2 – xFe2O3 solid solution at high
temperatures (1,273–1,673 k) in the two-step water-splitting reaction
for solar H2 generation. J Mater Sci. 2008;43:3153–3161.
39. Call F, Roeb M, Schm€ ucker M, Sattler C, Pitz-Paal R. Ceria doped
with zirconium and lanthanide oxides to enhance solar thermochemi-
cal production of fuels. J Phys Chem C. 2015;119:6929–6938.
40. Lin FJ, Rothensteiner M, Alxneit I, van Bokhoven JA, Wokaun A.
First demonstration of direct hydrocarbon fuel production from water
AIChE Journal August 2017 Vol. 63, No. 8 Published on behalf of the AIChE
and carbon dioxide by solar-driven thermochemical cycles using rho-
dium-ceria. Energy Environ Sci. 2016;9:2400–2409.
41. Abanades S, Legal A, Cordier A, Peraudeau G, Flamant G, Julbe A.
Investigation of reactive cerium-based oxides for H2 production by ther-
mochemical two-step water-splitting. J Mater Sci. 2010;45:4163–4173.
42. Scheffe JR, Jacot R, Patzke GR, Steinfeld A. Synthesis, characterization,
and thermochemical redox performance of Hf41, Zr41, and Sc31 doped
ceria for splitting CO2. J Phys Chem C. 2013;117:24104–24114.
43. Hao Y, Yang C-K, Haile SM. Ceria–zirconia solid solutions (Ce1–
xZrxO22d, x 0.2) for solar thermochemical water splitting: a ther-
modynamic study. Chem Mater. 2014;26:6073–6082.
44. Rothensteiner M, Bonk A, Vogt UF, Emerich H, van Bokhoven JA.
Structural changes in Ce0.5Zr0.5O22d under temperature-swing and
isothermal solar thermochemical looping conditions determined by
in situ Ce K and Zr K edge X-ray absorption spectroscopy. J Phys
Chem C. 2016;120:13931–13941.
45. Petkovich ND, Rudisill SG, Venstrom LJ, Boman DB, Davidson JH,
Stein A. Control of heterogeneity in nanostructured Ce1–xZrxO2 bina-
ry oxides for enhanced thermal stability and water splitting activity.
J Phys Chem C. 2011;115:21022–21033.
46. Tolla B, Demourgues A, Isnard O, Menetrier M, Pouchard M,
Rabardel L, Seguelong T. Structural investigation of oxygen inser-
tion within the Ce2Sn2O7-Ce2Sn2O8 pyrochlore solid solution by
means of in situ neutron diffraction experiments. J Mater Chem.
1999;9:3131–3136.
47. Subramanian MA, Aravamudan G, Rao GVS. Oxide pyrochlores2a
review. Prog Solid State Chem. 1983;15:55–143.
48. Thomson JB, Armstrong AR, Bruce PG. A new class of pyrochlore
solid solution formed by chemical intercalation of oxygen. J Am
Chem Soc. 1996;118:11129–11133.
49. Sickafus KE, Mineevini L, Grimes RE, Valde JA, Ishimare M, Li F,
McClellan KJ, Hartmann T. Radiation tolerance of complex oxides.
Science. 2000;289:748–751.
50. Wright CS, Fisher J, Thompsett D, Walton RI. Hydrothermal synthe-
sis of a cerium(IV) pyrochlore with low-temperature redox proper-
ties. Angew Chem Int Ed. 2006;45:2442–2446.
51. Yamamoto T, Suzuki A, Nagai Y, Tanabe T, Dong F, Inada Y,
Nomura M, Tada M, Iwasawa Y. Origin and dynamics of oxygen
storage/release in a Pt/ordered CeO2–ZrO2 catalyst studied by time-
resolved XAFS analysis. Angew Chem Int Ed. 2007;119:9413–9416.
52. Sickafus KE, Grimes RW, Valdez JA, Cleave A, Tang M, Ishimaru
M, Corish SM, Stanek CR, Uberuaga BP. Radiation-induced
amorphization resistance and radiation tolerance in structurally relat-
ed oxides. Nat Mater. 2007;6:217–223.
53. Oh SH, Black R, Pomerantseva E, Lee JH, Nazar LF. Synthesis of a
metallic mesoporous pyrochlore as a catalyst for lithium-O2 batter-
ies. Nat Chem. 2012;4:1004–1010.
54. Shamblin J, Feygenson M, Neuefeind J, Tracy CL, Zhang F,
Finkeldei S, Bosbach D, Zhou H, Ewing RC, Lang M. Probing dis-
order in isometric pyrochlore and related complex oxides. Nat
Mater. 2016;15:507–511.
55. Scheffe JR, Allendorf MD, Coker EN, Jacobs BW, McDaniel AH,
Weimer AW. Hydrogen production via chemical looping redox
cycles using atomic layer deposition-synthesized iron oxide and
cobalt ferrites. Chem Mater. 2011;23:2030–2038.
56. Tolla B, Demourgues A, Pouchard M, Rabardel L, Fournes L,
Wattiaux A. Oxygen exchange properties in the new pyrochlore solid
solution Ce2Sn2O7-Ce2Sn2O8. Solid State Chem Cryst Chem. 1999;2:
139–146.
57. Guzman J, Carrettin S, Corma A. Spectroscopic evidence for the supply
of reactive oxygen during CO oxidation catalyzed by gold supported on
nanocrystalline CeO2. J Am Chem Soc. 2005;127:3286–3287.
58. McBride JR, Hass KC, Poindexter BD, Weber WH. Raman and X-
ray studies of Ce12xRexO22y, where Re5La, Pr, Nd, Eu, Gd, and
Tb. J Appl Phys. 1994;76:2435.
59. Banerji A, Grover V, Sathe V, Deb SK, Tyagi AK. CeO2-Gd2O3
system: unraveling of microscopic features by Raman spectroscopy.
Solid State Commun. 2009;149:1689–1692.
60. Rupp JLM, Fabbri E, Marrocchelli D, Han J-W, Chen D, Traversa
E, Tuller HL, Yildiz B. Scalable oxygen-ion transport kinetics in
metal-oxide films: impact of thermally induced lattice compaction in
acceptor doped ceria films. Adv Funct Mater. 2014;24:1562–1574.
61. Johari A, Bhatnagar MC, Rana V. Effect of substrates on structural
and optical properties of tin oxide (SnO2) nanostructures. J Nanosci
Nanotechnol. 2012;12:7903–7908.
62. Yang N, Shi Y, Schweiger S, Strelcov E, Belianinov A, Foglietti V,
Orgiani P, Balestrino G, Kalinin SV, Rupp JLM, Aruta C. Role of
DOI 10.1002/aic 3461
 associated defects in oxygen ion conduction and surface exchange
reaction for epitaxial Samaria-doped ceria thin films as catalytic
coatings. ACS Appl Mater Interfaces. 2016;8:14613–14621.
63. Giannici F, Gregori G, Aliotta C, Longo A, Maier J, Martorana A. Structure
and oxide ion conductivity: local order, defect interactions and grain bound-
ary effects in acceptor-doped ceria. Chem Mater. 2014;26:5994–6006.
64. Shi Y, Bork AH, Schweiger S, Rupp JLM. The effect of mechanical
twisting on oxygen ionic transport in solid-state energy conversion
membranes. Nat Mater. 2015;14:721–727.
65. Li LL, Xu J, Yuan Q, Li ZX, Song WG, Yan CH. Facile synthesis
of macrocellular mesoporous foamlike Ce-Sn mixed oxides with a
nanocrystalline framework by using triblock copolymer as the single
template. Small. 2009;5:2730–2737.
66. Elfallah J, Boujana S, Dexpert H, Kiennemann A, Majerus J, Touret
O, Villain F, Lenormand F. Redox processes on pure ceria and on
Rh/CeO2 catalyst monitored by X-ray absorption (Fast acquisition
mode). J Phys Chem. 1994;98:5522–5533.
3462 DOI 10.1002/aic Published on behalf of the AIChE
67. Zhang F, Wang P, Koberstein J, Khalid S, Chan S-W. Cerium oxida-
tion state in ceria nanoparticles studied with X-ray photoelectron
spectroscopy and absorption near edge spectroscopy. Surf Sci. 2004;
563:74–82.
68. de Kergommeaux A, Faure-Vincent J, Pron A, de Bettignies R,
Malaman B, Reiss P. Surface oxidation of tin chalcogenide nano-
crystals revealed by 119Sn-M€ossbauer spectroscopy. J Am Chem Soc.
2012;134:11659–11666.
69. Protesescu L, Rossini AJ, Kriegner D, Valla M, de Kergommeaux
A, Walter M, Kravchyk KV, Nachtegaal M, Stangl J, Malaman B,
Reiss P, Lesage A, Emsley L, Coperet C, Kovalenko MV. Unrav-
eling the core-shell structure of ligand-capped Sn/SnOx nanopar-
ticles by surface-enhanced nuclear magnetic resonance,
M€ ossbauer, and X-ray absorption spectroscopies. Acs Nano. 2014;
8:2639–2648.
Manuscript received Nov. 16, 2016, and revision received Feb. 13, 2017.
August 2017 Vol. 63, No. 8 AIChE Journal