You are on page 1of 18

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/319796192

Recent Progress in Photocatalytic CO2 Reduction Over Perovskite Oxides

Article  in  Solar RRL · September 2017


DOI: 10.1002/solr.201700126

CITATIONS READS

116 2,223

3 authors:

Run Shi Geoffrey I.N. Waterhouse


Technical Institute of Physics and Chemistry University of Auckland
94 PUBLICATIONS   6,795 CITATIONS    350 PUBLICATIONS   15,786 CITATIONS   

SEE PROFILE SEE PROFILE

Tierui Zhang
Technical Institute of Physics and Chemistry
269 PUBLICATIONS   21,228 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Magnetic removal of céium ions using gamma-poly(glutamic aid)-coated magnetite particles with the enhanced effect of zoelite supplementation View project

Photonic Band Gap Materials View project

All content following this page was uploaded by Tierui Zhang on 19 September 2017.

The user has requested enhancement of the downloaded file.


REVIEW
Photocatalytic CO2 Reduction www.solar-rrl.com

Recent Progress in Photocatalytic CO2 Reduction Over


Perovskite Oxides
Run Shi, Geoffrey I.N. Waterhouse, and Tierui Zhang*

So why not simply bypass these prob-


Photocatalytic CO2 reduction has attracted widespread attention in recent lems by switching from fossil fuels to
years, due its potential to reduce anthropogenic CO2 emissions from fossil renewables? Solar power generation shows
fuel combustion whilst also yielding fuels and valuable chemical feedstocks. an annual growth of about 33%, which
encompasses three distinct approaches:
Among the various semiconductor photocatalysts capable of driving CO2
photothermal, photovoltaic, and photoca-
reduction under direct sunlight, perovskite oxides are arguably the most talysis.[3] The first two of these approaches
promising due to their high catalytic activity, good stability, long charge are mature technologies with wide utiliza-
diffusion lengths, and compositional flexibility that allows precise bandgap tion. The final approach, semiconductor
and band edge tuning. This review aims to showcase recent advances in the photocatalysis, offers great potential for
transforming earth-abundant materials
design of perovskite oxides and their derivatives for photocatalytic CO2
(H2O, CO2, and N2) into high energy
reduction, placing particular emphasis on structure modulation, defect density chemicals (e.g., H2, CH4, CH3OH,
engineering, and interface construction as rational approaches for enhancing and NH3, among others) via complex
solar-driven CO2 conversion to CH4, CO, and other valuable oxygenates photoreactions that harness solar photons
(CxHyOz). as the energy source.[4] In the context of 1)
conservation of existing fossil fuel reserves
and 2) curbing anthropogenic CO2 emis-
sions to slow climate change caused by
global warming, solar-driven photocatalytic
1. Introduction CO2 reduction to fuels represents an
exciting opportunity.[5] However, CO2 is
1.1. Brief Overview of Photocatalytic CO2 Reduction one of the most thermodynamically stable
In spite of dwindling fossil fuel reserves and environmental carbon compounds due to the much higher bond energy of C5 5O
concerns caused by their combustion, modern societies remain (750 kJ mol1) compared with C─H bonds (411 kJ mol1) and
high dependent on fossil fuels for electricity generation, C─C bonds (336 kJ mol1). For this reason, CO2 activation by
transportation, and as a feedstock to the chemical industry.[1] conventional thermal catalysis requires high temperatures and
Annual anthropogenic CO2 emissions from fossil fuel combus- thus a high energy input, creating secondary pollution.[6]
tion have steadied at 9 Gt (1 Gt ¼ 109 t) over last decade, with Semiconductor photocatalysis using sunlight to drive CO2
current level of CO2 in the atmosphere 400 ppm, approxi- activation at near ambient temperatures is therefore a far more
mately 43% higher than the level recorded in pre-industrial attractive approach from a sustainability perspective. Nature
times.[2] CO2 accumulation in the atmosphere, coupled with achieves this same transformation with ease through the process
rapid deforestation across the planet, has seriously damaged the of photosynthesis, thereby providing valuable clues for photo-
balance of the earth’s carbon cycle and triggered global warming. catalyst development for solar-driven CO2 reduction.[7]
Large-scale CO2 capture or recycling could potentially halt the Photocatalytic CO2 reduction is complex, leading to a diverse
rate of progression of global warming, motivating widespread range of products via multi-step reaction pathways.[2,8] Equations
fundamental, and applied research in both these areas. 1–6 (Table 1) list half equations to the various products
commonly formed in electrochemical CO2 reduction in water,
along with the theoretical reductions potentials (E0) for each half
R. Shi, Prof. T. Zhang equation referenced against the normal hydrogen electrode
Key Laboratory of Photochemical Conversion and Optoelectronic (NHE) at pH 7.
Materials, Technical Institute of Physics and Chemistry, Chinese
Academy of Sciences, Beijing 100190, P. R. China From Table 1, it is evident that thermodynamically stable
E-mail: tierui@mail.ipc.ac.cn products will not be kinetically favored, since more electrons are
R. Shi needed to drive the reaction to those thermodynamically stable
College of Materials Science and Opto-Electronic Technology, products. Equations 2–6 should be considered as the culmina-
University of Chinese Academy of Sciences, Beijing 10049, P. R. China tion of a series of photon-assisted single electron reactions that
Prof. G. I.N. Waterhouse lead to the same products as those described in Table 1. It is
School of Chemical Sciences, The University of Auckland, Auckland generally accepted that CO2 reduction starts with the chemi-
1142, New Zealand
sorption of CO2 on the surface of the photocatalyst. Various
DOI: 10.1002/solr.201700126 adsorbed species such as carbonate-like species can be formed

Sol. RRL 2017, 1700126 1700126 (1 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

Table 1. Electrochemical reactions involved in CO2 reduction with


Run Shi studied materials science
water and their corresponding reduction potential E0 (V vs. NHE
and engineering at the Tianjin
at pH 7).
Polytechnic University where he
received his BSc. in 2012. He is
Entry Equation Product E0 (V)
currently pursuing his Ph.D. in
1 CO2 þ e ! CO
2 Carbonate anion radical 1.90 materials science at the Technical
þ  0.61
2 CO2 þ 2H þ 2e ! HCOOH Formic acid Institute of Physics and Chemistry,
3 CO2 þ 2Hþ þ 2e ! CO þ H2 O Carbon monoxide 0.53 Chinese Academy of Sciences, in the
4 CO2 þ 4Hþ þ 4e ! HCHO þ H2 O Formaldehyde 0.48
lab of Prof. Tierui Zhang. His
þ 
current research interest focuses on
5 CO2 þ 6H þ 6e ! CH3 OH þ H2 O Methanol 0.38
controlled synthesis of noble-metal-free photocatalysts for
6 CO2 þ 8Hþ þ 8e ! CH4 þ 2H2 O Methane 0.24 hydrogen evolution and CO2 reduction.

Prof. Geoffrey I.N. Waterhouse


on metal oxide surfaces through bidentate or monodentate completed a Ph.D. in Chemistry at
coordination.[9] A single electron reduction of CO2 then occurs, the University of Auckland in 2003.
leading to the formation of (Equation 1).[3] The strongly negative He is a principal investigator in the
redox potential of 1.90 V is due to the huge energy barrier that MacDiarmid Institute (a New
needs to be overcome to transform linear CO2 into bent on Zealand CoRE) and a Chair
accepting one electron from the photocatalyst surface. This first Professor in both the School of
single electron transfer process is therefore considered as the Materials Science and Engineering
rate limiting step in many photocatalytic CO2 reduction at the South China University of
reactions. The product selectivity of CO2 reduction depends Technology and the College of
on many factors, such as the specific photocatalyst active sites Chemistry and Material Science at Shandong Agricultural
and reaction conditions. For example, copper is the active site for University. His research interests include photocatalysis,
the CO2 reduction to hydrocarbons, whereas silver and rhodium photonic crystals, and biosensors.
afford high selectivities to CO.[10] HCOOH and CH3OH are
usually formed in liquid phase reaction systems, whilst they are Prof. Tierui Zhang is a full professor
seldom formed in the photoreduction of gas phase CO2.[11] at the Technical Institute of Physics
CO2 photoreduction efficiency is generally discussed in terms and Chemistry, Chinese Academy of
of CO2 conversion rate, the formation rate of products, the Sciences. He obtained his Ph.D.
turnover frequency (TOF) and the quantum efficiency of the degree in Chemistry in 2003 at Jilin
process.[3] In recent years, emphasis has been placed on University, China. He then worked
improving the reliability of the CO2 reduction data. since the as a postdoctoral researcher in the
opposite half-reaction involving water oxidation and oxygen labs of Prof. Markus Antonietti, Prof.
evolution was seldom reported or discussed in pioneer works.[9] Charl F. J. Faul, Prof. Hicham
Indeed, product formation rates in these early works often Fenniri, Prof. Z. Ryan Tian, Prof.
increased in the initial stage and then decreased sharply at longer Yadong Yin, and Prof. Yushan Yan. His current scientific
reaction times. Since CO2 reduction efficiencies are generally interests focus on catalyst nanomaterials for efficient and
low, many of the carbon products detected may not have clean production, and utilization of hydrogen.
originated from CO2 photoreduction, but rather from some
residual carbon species in the reaction system such as surface
adsorbed organic species or solvent, thereby detrimentally
impacting the final results. generated oxygen exceeds the amount expected based on the
To avoid such problems and serious misinterpretations of reduced products, then this can point to an undetected carbon-
experimental data, it is extremely important to use rigorous containing species, such as an alcohol or formic acid, with low
testing protocols and advanced scientific techniques when saturation vapor pressures at liquid-gas interface or high
studying CO2 photoreduction reactions. Stoichiometric method adsorption energies at the solid-gas interface.[13] Compared to
is useful for establishing the source of the CO2 reduction water as the electron donor and hydrogen source, hydrous
products, and product selectivities. Accordingly, each carbon- hydrazine with a lower oxidation potential was found to be more
containing product of CO2 photoreduction in water will produce effective in suppressing the oxidation of bimetallic AuCu
a stoichiometric amount of O2. If O2 is not detected, or detected nanoparticles loaded on SrTiO3/TiO2 during photocatalytic
but in an amount less than the theoretical value, then three reactions.[14] Ye et al. used hydrogen as an additional electron
possible explanations can be forwarded: 1) water oxidation did donor to illustrate the key role of peroxide intermediates from
not occur; 2) the photocatalyst had a strong adsorption energy for water oxidation during CO2 photoreduction to CH4 over Ru
O2 preventing it from desorbing from the photocatalyst surface; loaded NaTaO3.[15] Wu et al. evaluated the performance of
or 3) the O2 was consumed through unwanted oxidation SrTiO3-based photocatalysts in different reactant gas mixtures,
reactions on the photocatalyst surface.[12] Conversely, if the and confirmed the synergistic effect of a CO2/CO gas mixture in

Sol. RRL 2017, 1700126 1700126 (2 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

enhancing CH3OH selectivity.[16] For CO2 reduction in an attractive advantages of perovskite oxides in photocatalysis is their
aqueous dispersion containing a NaTaO3-based photocatalyst, flexibility of composition and structure. Both alkali metals (Li, Na,
the presence of HCO3 or CO32 in solution was found to K, etc.) and alkaline-earth metals (Mg, Ca, Ba, etc.) can occupy the
suppress H2 evolution through a pH effect and/or additional A-sites, whereas the B-site can host a wide range of transition
CO2 release, thus resulting in drastically increased CO selectivity metals (Ti, Nb, Fe, etc.). Furthermore, both A and B sites can be
up to 90%.[17] These studies highlight the sensitivity of CO2 doped with other metal cations to form structures like AA0 BO3 or
photoreduction to experimental testing conditions. ABB0 O3. Similar to other oxide semiconductors, oxygen can be also
Isotope labeling experiments allow unambiguous assign- partially replaced by other non-metal anions (N, S, etc.) to allow
ment of reaction products as those of CO2 photoreduction bandgap narrowing for increased visible light absorption.[30]
rather than some other source (e.g., residual solvent or an Figure 1 shows the various elements used to date in the
adsorbed organic species from the photocatalyst synthesis). construction of perovskite oxide-based photocatalysts for CO2
Typically, 13CO2, D2O, or H218O are introduced as reactants reduction (including co-catalysts). By varying the A-site and B-
and the isotopic signals ducts monitored to confirm the site cation composition, properties such as the band structure,
source of generated CO or hydrocarbons.[18] This approach is charge transfer and adsorption of reactant molecules can be
especially important for photocatalysts synthesized from manipulated and optimized for a given photoreaction. Figure 2
carbon containing precursors.[19] In situ FTIR is a powerful illustrates how the bandgap and band edges in perovskite oxides
tool for monitoring reactant adsorption and intermediate (and selected oxynitrides) can be adjusted by changing the A-site
formation for many photocatalytic reactions including CO2 and B-site cations. For perovskite oxides, the conduction band
reduction,[20] CO hydrogenation,[21] and ethanol oxidation.[22] generally consists outer shell d orbitals of the B-site transitions
Detailed analysis about the effects of surface oxygen vacancies metals (e.g., Ti 3d, Nb 4d, Ta 5d), whereas the top of the valence
and metal active sites of perovskite oxides to the photo- band is composed of O 2p states. Generally, for B-site cations, as
catalytic CO2 reduction will be discussed in the following you go down a group the conduction band will shift to more
section. negative potentials and the bandgap will get larger (c.f. NaNbO3
Eg ¼ 3.3 eV, NaTaO3 Eg ¼ 4.0 eV).[31] For A-site cations, the
conduction band will shift to more positive potentials and the
1.2. Perovskite Oxides as Semiconductor Photocatalysts bandgap narrow on going down a group (c.f. LiTaO3 Eg ¼ 4.8 eV,
NaTaO3 Eg ¼ 4.0 eV, KTaO3 Eg ¼ 3.6 eV).[11a,32]Perovskite oxy-
Transition metal oxides and group IIIA metal oxides are the most nitrides, commonly derived from perovskite oxides though
widely studied photocatalytic CO2 reduction materials due to their nitrogen substitution for oxygen, possess a higher valence band
high activity, stability and low cost.[23] Adjustable bulk and surface and a narrower bandgap due to the formation of continuous N2 p
components of metal oxides provide good opportunity to modify energy levels above pristine O 2p orbitals.[33] Further, 2-D layered
the photoexcitation and CO2 activation process, thus contribute to structured perovskite oxides exhibit anisotropic excited state
enhanced conversion rate and selectivity.[24] Among them, charge transfer properties due to their unique crystal structures,
perovskite oxides of general formula ABO3 demonstrate great which allows enhancement of exciton transfer dynamic
potential in the development of solar cells,[25] solid oxide fuel processes in photocatalytic reactions.[34]
cells,[26] photo(electro)catalysts[27] and ferroelectrics.[28] Typical Linear configurations of B─O─B bonds (close to 180 ) in the
perovskite oxides have a cubic crystal structure, with the larger A- corner-shared BO6 octahedra of perovskite oxides and strain-
site cations occupying the corners of the cube and the smaller B- induced incipient ferroelectricity present in many perovskite
site cations located at the center of the cube and octahedrally oxides provide good electron mobility and promote the
coordinated by the face-centered oxygen atoms.[29] One of the most delocalization and separation of photoexcited electron-hole

Figure 1. Element composition of reported perovskite oxide-based photocatalysts for CO2 reduction.

Sol. RRL 2017, 1700126 1700126 (3 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

Figure 2. Band structures for a selection of perovskite oxides and oxynitrides and the corresponding redox potentials involved in photocatalytic CO2
reduction.

pairs, thereby enhancing the photocatalytic activity.[35] Figure 2 structure of perovskite oxides influences each of these three
illustrates that perovskite oxides are promising materials for processes. Accordingly, different perovskite oxides exhibit
CO2 photoreduction. In the following sections, we provide a distinct activities and product selectivies for CO2 reduction.
brief overview of CO2 reduction research conducted over the In order to achieve high CO2 reduction selectivity toward a
last decade based on perovskite oxide and derivative photo- particular product, deep understanding of the relationship
catalysts, placing emphasis on structure-activity relationships. between the component elements and pervoskite structure is
We will start by discussing perovskite composition and required. Furthermore, structure-activity relationships need to
structure, followed by defect engineering (including hetero- be established, thereby providing a solid platform for the
atom doping). Interfacial charge transfer between perovskite rational design of improved perovskite oxide-based photo-
oxides and metal co-catalysts, semiconductor heterostructures catalysts for CO2 reduction.
and organic complex will then be examined. Finally, we
summarize the performance of various perovskite oxide-based
photocatalyst systems for CO2 reduction, offering a short 2.1. ABO3 Perovskite Oxides
perspective on the challenges and possible future directions of
research in this area. 2.1.1. Tantalates and Niobates

Both tantalum and niobium are group VB elements. Thus,


2. Composition and Structure perovskite tantalates and niobates share many similar

As mentioned above, perovskite oxides offer outstanding


compositional and structural flexibility. Figure 3 shows the
composition and structural distribution of known perovskite
oxides for CO2 photoreduction. The pie chart can be broadly
divided into two categories: compounds with the ABO3 structure
(75%) and compounds with a layered structure (25%). The ABO3
perovskite oxides mainly consist of titanates, tantalates and
niobates. The layered perovskite oxides can be sub-divided into
four types: Aurivillius, Ruddlesden Popper (RP), (111), and (110)
layered structures. Bi2WO6 with a Aurivillius layered structure is
the most widely studies layered perovskite oxide for CO2
reduction. The photocatalytic CO2 reduction activity of RP phase
and (111)/(110) layered perovskite oxides have received far less
attention.
Photocatalytic CO2 reduction on perovskite oxides involves
three key steps: 1) light absorption by the semiconductor,
creating electron-hole pairs; 2) separation and migration of
photoexcited electron-hole pairs from the bulk to semiconduc- Figure 3. Composition and structure distribution of perovskite oxides for
tor the surface; and 3) CO2 reduction and water oxidation photocatalytic CO2 reduction. The percentages have been calculated
reactions at surface sites. The chemical composition and based on ca. 50 articles from 2007 to 2017.

Sol. RRL 2017, 1700126 1700126 (4 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

properties, including valence band energy levels and charge 2.1.2. Titanates and Others
transfer properties.[31–32,36] Ta 5d orbitals have a more negative
potential than Nb 4d orbitals. As a result, tantalate perovskites Titanate perovskite oxides are easy to prepare and show very
possess more negative conduction band energy levels and wider good photo-stability compared with tantalates and niobates. The
bandgaps than the corresponding niobates. For example, the photocatalytic CO2 reduction performance of MTiO3 photo-
conduction band of NaTaO3 is more negative by 0.7 eV than that catalysts (M ¼ Ca, Sr, and Pb) have been studied exhaustively.
of NaNbO3, while their O 2p orbital-valence bands are almost CaTiO3 prepared using flux-assisted methods exhibits a variety
identical due to the similar ionic radius of oxygen anions of morphologies, specific surface areas and porosities depending
coordinated to Ta5þ or Nb5þ. Recently, O’Shea et al. prepared on the flux precursors. Photocatalytic activities for CO2 reduction
NaNbO3 and NaTaO3 with similar surface area and crystal size and product selectivities were determined to be highly
via high temperature solid-state reactions.[31] The crystal dependent on CaTiO3 morphology.[12]
structures of both samples were predominantly orthorhombic SrTiO3 is the archetypal perovskite oxide and has received
phase, with the NaTaO3 also containing approximately 18% considerable attention in photocatalytic and photoelectrochem-
monoclinic phase. The photocatalytic CO2 reduction activity of ical applications due to its excellent charge transport properties
the NaNbO3 and NaTaO3 perovskites were superior to that of and bandgap of 3.2 eV which is identical to that of anatase
TiO2, with NaTaO3 showing higher CH4 selectivity than TiO2.[41] The [001] direction of ABO3 perovskites can be
NaNbO3. Further studies have shown that niobate perovskites considered as alternating stacks of AO and BO2 sheets. Thus,
with same elemental composition can exhibit vastly different SrTiO3 can be viewed as a sequence of alternating charge neutral
catalytic CO2 reduction performance due to variations in their SrO and TiO2 sheets. Pristine SrTiO3 generally shows poor CO2
crystal structure. For example, cubic phased NaNbO3 prepared at adsorption ability, though this can be improved enormously if
relatively lower temperatures showed a higher CH4 production surface oxygen vacancies and coordinatively unsaturated metal
rate than orthorhombic phase NaNbO3 synthesized through centers can be introduced into the alternating SrO and TiO2
phase transformation at higher temperatures.[37] DFT calcula- sheets.[42] Strategies for improving SrTiO3 performance includ-
tions revealed the migration of conduction band electrons was ing doping with active metal cations, introducing surface defects,
faster for cubic phase NaNbO3. The cubic phase has a smaller adding metal co-catalysts or forming heterojunctions with other
electron effective mass due to the more dispersed conduction semiconductors. For example, Ti3þ self-doped SrTiO3 possessed
band, thus facilitating the transfer of photoexcited electrons and a higher surface oxygen vacancy concentration than pristine
enhancing CO2 reduction. The transformation from cubic SrTiO3, and thus demonstrates improved CO2 adsorption
NaNbO3 to orthorhombic NaNbO3 occurs with heating in the properties.[43] Wu et al. achieved highly selective reduction of
range 400–600  C.[38] A mixed cubic-orthorhombic NaNbO3 CO2 to CH3OH by locating Pt/SrTiO3:Rh and Pt/CuAlGaO4 on
photocatalyst demonstrated a higher CO2 reduction activity than opposite sides of a Nafion membrane-separated twin reactor. The
pure cubic phase NaNbO3, which was rationalized in terms Pt/SrTiO3:Rh and Pt/CuAlGaO4 photocatalysts were responsible
superior charge separation at p-n junctions formed at the cubic- for CO2 reduction and water oxidation, respectively.[16] Hierar-
orthorhombic phase boundaries. chical structures offer enhanced charge separation through
Experimental studies have shown that Ta-based perovskite multi-dimensional electron transfer channels, and have been
oxides prepared by solvothermal methods generally exhibit successfully applied in many photocatalytic CO2 reduction
higher catalytic activity than those synthesized by the traditional systems, including leaf-templated 3-D SrTiO3 with abundant
solid-state reaction method, which can be attributed to the anisotropic porous channels (Figure 4) and basalt fiber micro-
smaller particle size, higher specific surface area and abundant rod-supported PbTiO3 core-shell composites.[44,45] Detailed
surface defects in solvothermally synthesized samples.[39] Using discussions about the modification of SrTiO3-based perovskites
small alkali or alkaline-earth metals such as Li or Ca as the A-site for CO2 photoreduction is provided in the following sections.
cation in tantalates and niobates widens the bandgap (>4 eV) but In addition to Ti, Ta, and Nb-based ABO3 structures, other B-
enhances CO2 adsorption, thus generally leading to higher site cations show promise for photocatalyst development for CO2
catalytic activity under UV irradiation.[11a,36,40]Addition of a co- reduction. Zou et al. reported BaCeO3 and BaZrO3 showed
catalyst can enhance charge separation and modify product promise for photocatalytic CO2 reduction.[46] The conduction
selectivities. Silver loaded NaTaO3 showed 90% CO selectivity for band of BaCeO3 mainly consists of Ce 4f orbitals, which have
CO2 reduction in aqueous solution.[17] In the presence of H2, Ru reduced orbital splitting compared to the d-orbitals that form the
loaded NaTaO3 exhibited a high activity for CO2 reduction to CH4 conduction band of Ti, Ta, and Nb-perovskite oxides, facilitating
(51.8 mmol h1 g1).[15] In general, the wide bandgaps of the transfer of photoexcited electrons. LaFeO3 is the only ABO3
tantalates limit their application for solar-driven photocatalytic type perovskite with strong visible light absorption properties,
CO2 reduction. Thermal ammonolysis of tantalum-based due to its narrow bandgap of 2.0 eV. LaFeO3 is thus a
perovskite oxides to yield oxynitrides (such as CaTaO2N, very promising candidate for CO2 reduction under visible
SrTaO2N, and BaTaO2N) can increase the valence band energy irradiation.[47]
via introduction of N 2p states, significantly narrowing the
bandgap and allowing visible absorption.[11a,33]However, the
oxynitrides are highly susceptible to photocorrosion due to their 2.2. Layered Perovskite Oxides
slow water oxidation rates, thereby requiring sacrificial agents or
hybridization with other semiconductor materials to overcome In addition to the typical ABO3 perovskite oxides, there are many
this limitation. derivative layered perovskite oxides whose structure consists of

Sol. RRL 2017, 1700126 1700126 (5 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

Figure 5. Crystal structures of typical layered perovskite oxides. (a)


Aurivillius, (b) Ruddlesden Popper, (c) (111) layered, and (d) (110) layered
perovskite oxides (red spheres: oxygen; blue spheres: A-site element; gray
Figure 4. 3-D hierarchical SrTiO3 network derived from a sacrificial leaf- octahedron: B-site element).
template using a sol–gel infiltration method. (a–c) Optical images shows
the leaf-like structure of the as-prepared SrTiO3. (d) SEM image of the
cross section of the venation architecture showing anisotropic macro- nanoplates that assemble into complex flower-like hierarchical
porosity. (e) Magnified SEM image of the rough surface of the walls. (f) structures. Such hierarchical structures are highly beneficial for
HRTEM image of the mesoporous structure of SrTiO3. Reproduced with charge separation and visible light-driven CO2 reduction.[51] In
permission.[44] Copyright 2013, Nature Publishing Group.
Ruddlesden-Popper layered perovskite structures, the (ABO3)n
perovskite layer is sandwiched between (A’O) spacing layer,
alternating layered stacks of discontinuous perovskite slabs.[27a] where A and A0 are either alkali, alkaline-earth or rare-earth
These possess infinite extended perovskite-like structures with cations. The A0 cations located at the perovskite boundary have
corner-sharing BO6 units arranged along the ab planes, similar to an asymmetrical coordination with eight neighboring oxygen
the general ABO3 type but with much more severe structural anions on one side and one oxygen atom on the opposite side.
distortions.[48] Layered perovskites find applications in a wide range Sr3Ti2O7 and Li2SrTa2O7 are examples of Ruddlesden-Popper
of research fields, due to their interesting properties that include layered perovskites and exhibit good CO2 adsorption and
ferroelectricity, magnetoresistance and superconductivity.[49] photoreduction performance.[52] The (110) and (111) layered
Along the c-axis of layered perovskites, spacing layers creating perovskites are members of a homologous series of layered
insulating properties with poor electron mobility.[48] This layered structures built from (110)/(111) perovskite slabs formed by A-
crystal structure and anisotropic electron transfer give layered oxide site metal cations and corner-shared BO6 octahedra.[53] Among
perovskites special optoelectronic and exciton transfer properties these structures, ALa4Ti4O15 (A ¼ Ca, Sr, and Ba) is the most
including quantum well effects and edge-state exciton dissocia- active (111) layered perovskite showing impressive photo-
tion.[50] Figure 5 lists the common types of layered perovskite oxide catalytic CO2 reduction and water splitting activity under UV
structures and their general formulae, in which n represents the light mainly due to the high reducing potentials of photo-
stack cycles of periodic BO6 units that comprise one perovskite-like generated electrons.[54]
layer. For Aurivillius, the structure consists of alternating layers of Layered perovskite oxides are typically prepared by solid-state
(Bi2O2) and (An1BnO3nþ1) perovskite-like layers, where low reaction or hydrothermal methods. The as-synthesized 2-D
valence A-site cations are 12-fold coordinated by oxygen and the platelet structures can be easily exfoliated to yield ultrathin
octahedral B-site is occupied by a high valence metal cations. nanosheets using tetra-n-butylammonium hydroxide as an
Typical Aurivillius phased perovskite oxides such as Bi2WO6 and intercalation agent.[55] The exfoliated nanosheets possess high
Bi2MoO6 have narrow bandgaps (2.6–2.7 eV) and tend to form specific surface areas and abundant exposed ab planes.[56] The

Sol. RRL 2017, 1700126 1700126 (6 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

Table 2. Performance comparison of different perovskite oxides for photocatalytic CO2 reduction.

Synthesis Co-catalyst (method, Incident Isotope Major Sel.e)


Perovskite method wt.%) light Reaction conditions labeling product Yielda) (%) References
LiTaO3 SS – UV–Vis GP (CO2, H2) – CO 0.0175 – [36]

NaTaO3 SS – UV–Vis GP (CO2, H2) – CO 0.0042 –


KTaO3 SS – UV–Vis GP (CO2, H2) – CO 0.0025 –
KTaO3 HT – UV–Vis LP (CO2, H2O) 13
CO2 CO 61.98b) 4.5 [39]

13 b)
Ag (PD, 0.01) UV–Vis LP (CO2, H2O) CO2 CO 152.62 12
13
NaTaO3 PC Ru (PD, 0.5) UV–Vis GP (CO2, H2O gas, H2) CO2 CH4 51.8 96f) [15]

13 f)
Pt (PD, 0.5) UV–Vis GP (CO2, H2O gas, H2) CO2 CO 139.1 99
NaTaO3 HT CuO (IMP, 2) UV–Vis LP (CO2, isopropanol) – acetone 337 71f) [59]

NaTaO3 SS CuxO (IMP, 2) <410 nm LP (CO2, H2O, 0.5 M KHCO3) 13


CO2, CO 137.5 c)
80 [18]

H218O
13 [17]
Ca-NaTaO3 SS Ag (CR, 2) UV–Vis LP (CO2, H2O, 0.1M NaHCO3) CO2 CO 296 91
13
Sr-NaTaO3 SS Ag (CR, 2) UV–Vis LP (CO2, H2O, 0.1M NaHCO3) CO2 CO 352 86
13
Ba-NaTaO3 SS Ag (CR, 3) UV–Vis LP (CO2, H2O, 0.1M NaHCO3) CO2 CO 250 84
NaTaO3 Templated-SG Au (IMP, 1) UV–Vis GP (CO2, H2O gas) – CO 0.173 17.3 [40]

NaTaO3 SS – UV–Vis GP (CO2, H2O, 2 bar, 50  C) – CO 50.7 78 [31]


NaNbO3 SS – UV–Vis GP(CO2, H2O, 2 bar, 50 C) – CO 50.6 74
KNbO3 SS Pt (PD, 0.5) UV–Vis GP (CO2, H2O gas) – CH4 70b) – [32]

NaNbO3 SS Pt (PD, 0.5) UV–Vis GP (CO2, H2O gas) – CH4 23 b)



Cub. NaNbO3 PC Pt (PD, 0.5) >300 nm GP (CO2, H2O gas) 13
CO2, D2O CH4 15 – [38]

Ort. NaNbO3 PC Pt (PD, 0.5) >300 nm GP (CO2, H2O gas) 13


CO2, D2O CH4 4 –
Cub./Ort. PC Pt (PD, 0.5) >300 nm GP (CO2, H2O gas) 13
CO2, D2O CH4 18 –
NaNbO3
NaNbO3/g-C3N4 HT Pt (PD, 0.5) >420 nm GP (CO2, H2O gas) – CH4 6.4 – [60]

KNbO3/g-C3N4 HT Pt (PD, 0.5) >420 nm GP (CO2, H2O gas) – CH4 2.5 – [61]

CaTaO2N/Ru(II) PC Ag (IMP, 1) >400 nm LP (DMA, 20% TEOA) 13


CO2 HCOOH 5.33 100 [11a]

NaTaON/N- Chemical – >400 nm GP (CO2, H2) 13


CO2 CO 43 75f) [33]

GQDs etching
Rh-SrTiO3 SS Pt (PD, 0.8) AM 1.5G LP (CO/CO2 ¼ 1:10, H2O, – CH3OH 1.0 91 [16]

2 mM Fe2þ)
Pt (PD, 0.8) AM 1.5G LP (CO2, H2O, 2 mM Fe2þ) – CH3OH 0.5 83

Ti -Vo-SrTiO3 Combustion Pt(0.3) Vis GP (CO2, H2O gas) – CH4 0.034c) – [43]

SrTiO3 Templated-SG Au (PD, 1) UV–Vis GP (CO2, H2O gas) – CO 0.35 44.1 [44]

SrTiO3 Commercial Au (PD, 0.5) þ Rh >400 nm GP (CO2, H2O gas) – CO 66.8 56 [62]

(IMP, 0.5)
SrTiO3 Commercial Au (PD, 0.5) þ Rh (PD, >400 nm GP (CO2, H2O gas) – CO 369.2 84
0.5)
Fe-SrTiO3 HT Pt (PD, 0.5) >420 nm GP (CO2, H2O gas, 1atm) – CH4 421b) – [63]

Co-SrTiO3 HT Pt (PD, 0.5) >420 nm GP (CO2, H2O gas, 1atm) – CH4 636b) 86
SS Pt (PD, 0.5) >420 nm GP (CO2, H2O gas, 1atm) – CH4 159 b)

Cr-SrTiO3 SC – >420 nm GP (CO2, H2O gas) – CH4 0.88 – [64]

SrTiO3/TiO2 LS – UV–Vis LP (CO2, H2O, 1 M KHCO3) – CH4 3.67 d) [65]

Pt (PD) UV–Vis LP (CO2, H2O, 1 M KHCO3) – CH4 11.37d)


Pd (PD) UV–Vis LP (CO2, H2O, 1 M KHCO3) – CH4 20.83d)
SrTiO3/TiO2 HT Au3Cu (ST) UV–Vis LP (CO2, N2H4H2O) 13
CO2 CO 3370 84f) [14]

SrTiO3/ZnTe SC – >420 nm GP (CO2, H2O gas) – CH4 2.38 – [66]

CaTiO3 SS (flux) Ag (PD, 0.5) UV–Vis LP (CO2, H2O, 1.1M NaHCO3) – CO 2.25 44 [12]

(Continued)

Sol. RRL 2017, 1700126 1700126 (7 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

Table 2. (Continued)

Synthesis Co-catalyst (method, Incident Isotope Major Sel.e)


a)
Perovskite method wt.%) light Reaction conditions labeling product Yield (%) References
CaTiO3/basalt HT – UV–Vis LP (CO2, H2O) – CO 9.14 80f) [67]

PbTiO3/basalt HT – UV–Vis GP (CO2, H2O, 1 atm, 30  C) – CH4 48.3 – [45]

BaCeO3 Pechini Ag (CR, 0.5) UV–Vis GP (CO2, H2O gas) – CH4 0.4 – [46a]

BaZrO3 Pechini Ag (CR, 0.3) UV–Vis GP (CO2, H2O gas) – CH4 0.54 – [46a]

Pt (CR, 0.5) UV–Vis GP (CO2, H2O gas) – CH4 0.30 –


Au (CR, 0.5) UV–Vis GP (CO2, H2O gas) – CH4 0.24 –
N-LaFeO3/TiO2 SG – >400 nm LP (CO2, H2O) – CO 18.8 58f) [47]


Bi2WO6 hollow Anion – >420 nm LP (CO2, H2O, 0 C) – CH3OH 16.3 – [68]

sphere exchange
Bi2WO6 NPs HT – >420 nm GP (CO2, H2O gas) – CH4 1.09 – [51a]

Vo-Bi2WO6 HT – >400 nm GP (CO2, H2O gas) – CH4 0.535 – [69]

– >700 nm GP (CO2, H2O gas) – CH4 0.049 –


Bi2WO6 HT Pt (PD, 0.5) 365 nm GP (CO2, H2O gas, 140  C) – CH4 2.7 – [70]

Bi2WO6/TiO2 HT Pt (PD, 0.5) 365 nm GP (CO2, H2O gas, 140  C) – CH4 6.7 –

– 365 nm GP (CO2, H2O gas, 140 C) – CH4 1.5 –
Bi2WO6 HT – >420 nm GP (CO2, H2O gas) – CO 0.81 – [58a]

Bi2WO6/g-C3N4 HT – >420 nm GP (CO2, H2O gas) – CO 5.19 –


Bi2WO6 HT – >420 nm LP (CO2, H2O, 4  C) – CH3OH 5.12 54f) [58a]


Bi2WO6/MoS2 HT – >420 nm LP (CO2, H2O, 4 C) – CH3OH 9.17 50f)

Bi2WO6/CQDs HT – >400 nm GP (CO2, H2O gas) – CH4 0.899 – [71]

– >700 nm GP (CO2, H2O gas) – CH4 0.051 –


Bi2WO6/PPy HT – >420 nm LP (CO2, H2O, 4  C) – CH3OH 7.6 69f) [72]


Bi2WO6/PANI HT – >420 nm LP (CO2, H2O, 4 C) – CH3OH 11.07 74f)

Bi2WO6/PTh HT – >420 nm LP (CO2, H2O, 4  C) -– CH3OH 14.12 73f)


Bi2MoO6 HT – >420 nm LP (CO2, H2O, 4  C) – CH3OH 6.2 – [51b]

13 [13]
La2Ti2O7 PC Ag (CR, 1) UV–Vis LP (CO2, H2O) CO2 CO 5.2 51
13
Ag (IMP, 1) UV–Vis LP (CO2, H2O) CO2 CO 0.7 17
Ag (PD, 1) UV–Vis LP (CO2, H2O) 13
CO2 – – –
BaLa4Ti4O15 PC Ag (CR, 2) UV–Vis LP (CO2, H2O) – CO 73.3 67 [54a]

Ag (IMP, 1) UV–Vis LP (CO2, H2O) – CO 29.7 60


Ag (PD, 1) UV–Vis LP (CO2, H2O) – CO 14.3 29

SS, solid-state; HT, hydrothermal; PC, polymerized complex; SG, sol–gel; SC, sonochemical; LS, liquid-state.
a)
In mmol g1 h1; b)In ppm g1 h1; c)In mmol m2 h1; d)In ppm cm2 h1; e)Selectivity ¼ N(major products)/N(all reduction products)  100%; f)Selectivity ¼ N(major
products)/N(CO2 reduction products)  100%, of which H2 production is not included.

3. Defect Engineering of Perovskite Oxides


intrinsic wide bandgap of most layered perovskite oxides can be
narrowed through nitrogen doping, redshifting the band-edge Atom scale defects inside or on the surface of crystals have a
position due to uniform bulk N-doping.[57] Forming 2-D strong impact on their optical and electronic properties, which is
heterojunction composites with other layered materials such especially significant for nanomaterials of finite dimensions.
as MoS2 and g-C3N4 facilitates charge separation because Defect engineering is therefore an excellent approach for
electron transport across anisotropic 2D interfaces is much more optimizing the performance of nano-sized materials for
efficient than point contact between 0D nanoparticles.[58] Table 2 applications in photovoltaics, catalysis and bio-related applica-
summarizes the performance of various perovskite oxide tions.[73] Heteroatom dopants and oxygen vacancies are of great
photocatalysts for photocatalytic CO2 reduction. The table importance for modifying the photocatalytic performance of
provides representative examples of the different structures perovskite oxides, improving light absorption, charge transfer
and modifications, as well as other relevant information such as and reaction dynamics via crystal symmetry transformations
synthesis method, photocatalytic testing conditions, catalytic and subtle variations in metal cation coordination.[52a,74] The
activity and selectivity to the major product. following sections illustrate the various approaches used to

Sol. RRL 2017, 1700126 1700126 (8 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

introduce defects into perovskite oxides including metal cation


and nitrogen doping, oxygen vacancy and the corresponding
effect toward CO2 photoreduction.

3.1. Metal Cation Doping

Metal cation doping in perovskite oxides involves substitution on


the A or B sites, which can significantly both alter the band
structure and create new active site for photocatalytic CO2
reduction.[34,75] The band structure variation comprises two
aspects: varied bandgap composition and the introduction of
discrete doping states. The d orbitals of incorporated metal
cations will overlap and hybridize with existing A/B site metals to
change the composition of both conduction and valence bands
and their energy potentials. Some heteroatoms, especially
transition metals, can form discrete doping states in the
bandgap of perovskite oxides with d-d transition sub-peak
absorption in the low photon energy region. For example, He
et al. reported that the bandgap of SrTiO3 could be reduced by
about 0.1 eV on A-site doping with Cr3þ, due to the formation of a
narrow sub-bandgap and a sub-peak located at around
600 nm.[64]
Although perovskite oxides doped with metallic cations have
been extensively studied in relation to photocatalytic hydrogen
evolution and pollutant degradation, it does not mean that
these same doping strategies are also effective for promoting
CO2 photoreduction. For the latter, product selectivity (adsorp-
tion/desorption, activation energy, etc.) also needs to be
considered. Recently, Kudo et al. systematically investigated
the effect of metal doping on the CO2 reduction performance of
NaTaO3-based photocatalysts.[17] Among alkali earth metals (Mg,
Ca, Sr, and Ba) and La, Ba doping afforded the highest CO Figure 6. Temperature programmed FTIR spectra for the CO2 adsorption
selectivity (95%). The selectivity to CO was much lower for the products on (a) SrTiO3 and (b) Rh-doped SrTiO3 at temperatures ranging
other cation dopants with only hydrogen being detected in the from room temperature to 500  C. The new peaks at 1515 cm1 are due to
reduction products in the absence of doping. Rh-doped SrTiO3 monodentate carbonates, whilst the 1460 and 1400 cm1 peaks indicate
has been studied intensively over the past decade, showing great the formation of polydentate carbonates. Reproduced with permis-
sion.[41a] Copyright 2016, Wiley-VCH.
potential in photocatalytic solar energy conversion due to the
multi-oxidation states of the Rh dopant which introduces new
electron donor/acceptor energy levels.[76] High oxidation state
Rh cations act as electron acceptors and can be photoreduced to formed and result in much improved CO2 adsorption and a high
lower oxidation state cations by accepting electrons from SrTiO3. CH4 production rate (Figure 7).[63] Al is special as a dopant in
Under visible excitation, the low oxidation state Rh cations can perovskite oxides because it acts as a structural mediator rather
be photoexcited and supply electrons into the conduction band of than band regulator or active site. Michalsky et al. found that
SrTiO3 through d-d transitions at much lower photon energies B-site Al doping in La1xSrxMnO3 enriched the amount of Mn at
compared to the band edge excitation. As mentioned above, for the surface of the perovskite oxide from 16% (for the undoped
photocatalytic CO2 reduction the adsorption of CO2 molecule on La1xSrxMnO3) to nearly 94% without introducing additional
the photocatalyst surface is an important factor that oxygen vacancies.[77] Surface enrichment with the active element
strongly influences activity. Recent findings demonstrate has previously been shown to enhance many catalytic
that Rh3þ-Vo centers formed on the surface of Rh-doped reactions.[78] Controlling the surface composition of perovskite
SrTiO3 provide strong CO2 adsorption sites, leading to the oxides by doping metal elements can theoretically provide an
formation of adsorbed monodentate carbonate species, con- abundance of surface active centers, offering a promising
firming the potential of Rh-doped SrTiO3 for CO2 reduction strategy for further enhancing the photocatalytic CO2 reduction
(Figure 6).[41a] Due to its 3d0 electron configuration, Ti4þ performance.
possesses a high electronegativity and cannot provide electrons The aforementioned results demonstrate that the local
for CO2 to form chemical adsorbates, explaining the poor CO2 environment of doping atoms is important to achieving
reduction activity of pristine SrTiO3. However, after doping with enhancements in photocatalyst performance. Doping can cause
transition metal cations (e.g., Co2þ, Fe2þ, etc.) which have 3d significant modification of the structural and electronic
electrons, strong electron donor/acceptor interactions will be properties of perovskite oxides, as well as the types and

Sol. RRL 2017, 1700126 1700126 (9 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

or B-site substitution (Figure 8).[64] Extended X-ray


absorption spectroscopic (EXAFS) is another power-
ful technique for probing the local environment of
dopant atoms (coordination number, bond param-
eters, etc.) in perovskite oxides, and like SAED can
settle speculation about possible doping sites in
the crystal.[79] In addition, doping modifies the
valence states in the A/B sites of perovskite oxides,
generally resulting in the co-introduction of lattice
oxygen defects in varying concentrations, as will be
discussed in the subsequent section.

3.2. Nitrogen Doping and Oxygen Vacancy


Figure 7. (a) Adsorption isotherms at 298 K for CO2 on (I) pure SrTiO3; (II) N-doped
TiO2; (III) SrTi0.98Fe0.02O3; (IV) SrTi0.98Ni0.02O3; and (V) SrTi0.98Co0.02O3. (b) Substitution of lattice oxygen in perovskite oxides
Corresponding CH4 yields during CO2 photoreduction over the same photocatalysts
with nitrogen is a commonly used approach to
loaded with Pt under visible light irradiation (λ > 420 nm). Reproduced with
permission.[63] Copyright 2015, Springer International Publishing AG.
narrow the bandgap and thus allow increased visible
light absorption during photocatalytic processes.
Due to a higher potential of the N 2p orbital relative
availability of CO2 adsorption sites. Accordingly, special to the O 2p orbital, the incorporation of nitrogen into oxides
attention must be paid to the effect of dopants on the physical shifts their bandgaps from UV and into the visible regime
and optical properties of perovskite oxides.[75b] Interplanar through shifting the valence band position to less positive
spacings typically change on doping due to the dopant having a potentials. Jing et al. reported that the bandgap of LaFeO3 was
different ionic radius to the cation it is replacing. Selected area reduced from 2.0 to 1.8 eV after high temperature ammonia
electron diffraction patterns (SAED) are invaluable for probing A treatment.[47] A further increase in N-doping concentration
resulted in the formation of a perovskite oxynitride
showing continuous N 2p energy levels.[33] Bandg-
aps measured for MTaO2N (M ¼ Ca, Sr, Ba) were 2.5,
2.1, and 1.8 eV, respectively, with these oxynitrides
exhibiting good photosensitization properties when
used in heterostructured systems for photocatalytic
CO2 reduction.[11a] In addition, the introduction of
nitrogen anions in the perovskite lattice generates
considerable oxygen vacancies, since nitrogen has
one less electron than oxygen. Sun et al. demon-
strated that the introduction of oxygen vacancies in
NaTaON by an in situ reduction strategy could
further narrow the bandgap of NaTaON and enhance
CO production activity (Figure 9a).[33] By investigat-
ing the ionic state and concentration of N in
LaTiO2N formed by thermal ammonolysis of
La2Ta2O7, it was demonstrated that dispersed
substitutional N atoms at O sites can effectively
upshift the valence band, while other types of
substitutional or interstitial N will lead to impurity
states at the higher energy region of the bandgap,
which may act as trap sites and decrease the
photocatalytic activity.[74b] Layered perovskite oxides
exhibit many advantages when preparing N-doped
materials.[80] Due to their unique 2D nanostructure,
N atoms can more easily substitute oxygen atoms
to achieve uniform bulk doping and thus a
controlled redshift of the band edge absorption
(Figure 9b).[57,80]
Figure 8. HRTEM images of (a) SrTiO3 and (b) Cr-doped SrTiO3. The corresponding Introducing anion vacancies into semiconductors
SAED patterns are shown in (c) SrTiO3 and (d) Cr-doped SrTiO3. The reduced lattice is an effective approach for modifying their photo-
spacing after doping indicates A-site substitution (relative ionic radius size catalytic performance.[81] Localized variations in
Ti4þ<Cr3þ<Sr2þ). Reproduced with permission.[64] Copyright 2015, Elsevier. metal coordination geometry, valence state and

Sol. RRL 2017, 1700126 1700126 (10 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

formation can also be triggered by strain effects at heterostruc-


tured interfaces or dislocation boundaries.[83] Further reduction
treatments (such as heating in H2) or heteroatom doping can
also introduce oxygen defects at varying concentrations. A recent
study compared the photocatalytic activity of pristine Bi2WO6
and reduced Bi2WO6. The latter showed enhanced visible and
near infrared absorption, making it a very efficient photocatalyst
for CO2 even under near infrared excitation.[69] The strong
absorption is mainly due to the high oxygen vacancy concentra-
tion in the layered structure. Ye et al. studied the effect of oxygen
vacancies on photocatalytic CO2 reduction on self-doped
SrTiO3.[43] Ti3þ centers formed in the vicinity of oxygen
vacancies enhanced visible light absorption through newly
formed in-gap states, whilst the surface oxygen vacancies
enhanced CO2 adsorption. However, the high initial activity for
CO2 photoreduction decreased with reaction time due to
continuous consumption of Ti3þ and oxygen vacancies, which
could be recovered by further annealing in an Ar atmosphere.
Subsequent research has confirmed the role of surface oxygen
vacancies on SrTiO3 in promoting CO2 chemical adsorption.[42]
The adsorption of CO2 on SrTiO3 (100) surfaces in the absence of
oxygen vacancies is weak, whilst argon-ion sputtered surfaces
containing oxygen vacancies show much improved CO2
adsorption accompanied by the formation of carbonates, as
observed using metastable induced electron spectroscopy
(MIES).
Since the oxygen vacancy concentration can strongly impact
the photocatalytic CO2 reduction activity of perovskite oxides, the
detection and quantification of oxygen vacancies has become
a hot research area. Commonly used methods include
electron paramagnetic resonance (ESR), Raman spectroscopy
and X-ray photoelectron spectroscopy (XPS), along with
density-functional theory (DFT) calculations.[43] The CO2
adsorption ability is also reported as a characterization
method to probe the surface oxygen vacancies on perovskite
oxides.[41a,43,84] With recent advancements in electron
Figure 9. (a) UV–Vis diffuse reflectance spectra for NaTaO3 and
microscopy, imaging atom-scale defect has become possi-
Vo-NaTaON perovskites. Reproduced with permission.[33] Copyright
ble.[84] For example, high-angle annular dark-field scanning
2016, Elsevier. (b) UV–Vis diffuse reflectance spectra of pristine ALaTa2O7
(A ¼ Li, Na, K, Rb, or Cs) layered perovskites (solid lines), and transmission electron microscopy (HAADF STEM) and
corresponding samples after treatment in a NH3 flow at 1073 K for 6 h electron energy loss spectra (EELS) were used successfully
(dotted lines). Reproduced with permission.[80] Copyright 2016, Royal by Choi et al. to study oxygen vacancies formed at strain-
Society of Chemistry. imposed boundaries of SrTiO3.[83] Regions with intense strain
were confirmed to lower the formation energy of oxygen
vacancies, thus explaining the preferential formation of
bonding configurations can create new defect energy levels and oxygen vacancies at heterostructured interfaces and grain
unsaturated coordination, creating new electron transport boundaries. Recently, the quantum wall effect of graphene
pathways and sites for reactant adsorption and reaction. Oxygen was applied to monitor the in situ formation of surface oxygen
vacancies in metal oxides (TiO2, BiOCl, etc.) have received a lot of vacancies at the interface with SrTiO3 during electrochemical
attention in recent years.[73] The oxygen vacancy-induced redox cycling (Figure 10).[82b]
magnetic, dielectric and catalytic properties of perovskite oxides
have been reported.[42,82] In particular, oxygen vacancies are very
important for photocatalytic CO2 reduction over perovskite
4. Heterostructures
oxides, where the extended light absorption and strong CO2
adsorption are essential to subsequent photoreduction reactions. Although defect engineering can extend the light absorption of
There are many ways to create oxygen vacancies. Nanoscale perovskite oxides far into the visible region, defect-induced
perovskite oxides with stepped surfaces and abundant step edges energy trap sites must be carefully excluded to avoid unwanted
have compositions that differ from the ideal ABO3 stoichiomet- charge recombination. Further, perovskite oxides can naturally
ric ratio, and in such cases oxygen vacancies are actually drive oxygen evolution reactions because of their deep
introduced during the synthesis process. Oxygen vacancy valence band composed of O 2p orbitals, but generally lack

Sol. RRL 2017, 1700126 1700126 (11 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

energy transfer (PRET) mechanisms. Metal nanoparticles


functionalized perovskite oxides generally show superior
photocatalytic CO2 reduction performance compared with the
pristine perovskite oxides.[12,39] The influence of metal co-
catalyst on CO2 reduction activity and product selectivity is
complicated, and highly dependent on the metal co-catalyst, the
co-catalyst loading, particle size, speciation and deposition
site. Fortunately, the catalytic activity and selectivity of different
metals for CO2 reduction has been studied in detail in the field of
electrocatalysis, which provides valuable insights regarding the
selection of suitable co-catalysts for photocatalytic CO2
reduction.[10b] For example, metals such as Au and Ag show
excellent CO selectivity, whereas CH4 is the main reduction
product over Cu. Hydrogen evolution competes with CO2
reduction over Ni and Pt. Such differences can be rationalized in
several ways. Firstly, the adsorption energy and bonding
geometry of CO2 are distinct for each metal, thus leading to
different photocatalytic activity and selectivity when the metal
nanoparticles are supported on perovskite oxide photocata-
lysts.[10a] Adsorbed CO generated from CO2 over Cu can be
further hydrogenated to hydrocarbons (e.g., CH4) because of the
Figure 10. Surface oxygen vacancy formation in a SrTiO3 thin film, moderate adsorption energies of the oxygen-containing inter-
monitored via induced hysteretic conductance behavior at a high quality mediates. On the contrary, metals that show a positive CO
graphene-STO interface. (a) The source–drain current (ISD) under gate adsorption energy will have high selectivities for CO2 reduction
voltage (VG) swept systematically from 1, þ0.5 V (purple closed curve) to CO (it will desorb immediately on formation), whereas metals
to 4.5 V (red closed curve) at 2 K. The arrows and letters from A to F that bind CO too strongly will be poisoned. Furthermore, the
indicate the direction and sequence of different sweeping stages. b) metal/semiconductor interface is another important factor that
Schematic diagrams of the proposed oxygen vacancies creation and influences the product selectivity.[3] For Schottky contacts
annihilation process during the redox cycling. Reproduced with between a metal (Au, Ag, Cu, etc.) and a semiconductor
permission.[82b] Copyright 2017, Wiley-VCH.
support, the work function difference between metal and
semiconductor creates a potential offset in the space charge
the active sites needed for CO2 reduction.[17] To suppress severe region, the existence of which will contribute to the directional
charge recombination common to most photocatalysts, perov- transfer of photogenerated electrons across the interface.[87] The
skite oxide-based composites are often constructed to separate photogenerated electrons will accumulate on the metals until
charges and achieve high performance for photocatalytic CO2 equilibration of the Fermi levels is achieved. CO2 reduction is a
reduction. The following section discuss the effects of adding multiple electron process. Accordingly, electron accumulation
transition metals, semiconductors and organic compounds on metal co-catalysts through Schottky contacts affords higher
perovskite oxides and its derivatives as a means of enhancing CO2 reduction activities compared with metals (Pt, Pd, Ni, etc.)
their photocatalytic properties. that form an Ohmic contact with the semiconductor support.[87b]
Among metal co-catalysts commonly used to promote CO2
photoreduction on perovskite oxide surfaces, Ag nanoparticles
4.1. Transition Metal Deposition are most widely used because of their high CO selectivity.[46a]
Interestingly, different Ag deposition methods can result in wide
Transition metals are irreplaceable in a number of important ranging CO2 photoreduction performance.[17] Photocatalysts
catalytic reactions, such as nitrogen fixation, oxygen reduction prepared by depositing Ag nanoparticles using chemical reaction
reaction in fuel cells and Fischer-Tropsch synthesis.[85] In methods typically show better reduction activity and CO
photocatalytic reactions, transition metal co-catalysts frequently selectivity than photocatalysts prepared using photodeposition
play critical roles to improve performance, which can be or impregnation methods. Tanaka et al. demonstrated using Ag
attributed to one or more of the following: (1) suppression of K-edge EXAFS spectra (Figure 11) that all three deposition
charge recombination in the semiconductor by accepting methods afforded metallic Ag nanoparticles on La2Ti2O7.[13]
electrons from the semiconductor conduction band; (2) serving However, the intensity of first Ag-Ag shell for supported Ag
as cathodic reduction sites; (3) improving the photostability of nanoparticles prepared by the chemical reduction method was
the semiconductor photocatalyst by preventing the accumulation much lower than that of the corresponding signal for Ag
of photoexcited electrons; and (4) extending visible light nanoparticles deposited by the other two methods, indicating a
absorption through localized surface plasmon resonance (LSPR) smaller Ag nanoparticle size, consistent with supporting TEM
of metal nanostructures.[86] Decoration of semiconductors with observations. In addition to the Ag nanoparticle size effect on
Ag or Au nanoparticles which show a peak at visible wavelengths performance, Kudo et al. established that Ag nanoparticles
also opens the possibility of visible light-driven plasmonic deposited on the edge of layered ALa4Ti4O15 (A ¼ Ca, Sr, and Ba)
photocatalysis via hot electron injection or plasmon resonance sheets contribute significantly to the high CO selectivity by

Sol. RRL 2017, 1700126 1700126 (12 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

nontrivial local atomic and electronic structures,


owing to the presence of dangling bonds and
unsaturated coordination.[91] Such interfacial crystal
distortions in perovskite oxides give rise to intrigu-
ing physical properties, that could potentially be
manipulated to enhance photoreactions. For exam-
ple, the La2O3-terminated and TiO2-terminated
interface in LaAlO3/SrTiO3 shows 2-D electron
gas properties with high electron mobility, attracting
wide interest as ferromagnets and in quantum
devices that utilize the quantum Hall effect.[92] It is
believed that the same unique interfacial charge
transfer properties could guide the design of novel
photocatalyst systems based on perovskite oxide
Figure 11. (a) Ag K-edge XANES and (b) Fourier transformed EXAFS spectra for Ag heterostructures.
nanoparticles loaded La2Ti2O7 prepared by chemical reduction (CR), photo-deposition The topotactic epitaxial growth of nanostructured
(PD), and impregnation (IMP). Ag foil and Ag2O were used as reference materials. SrTiO3/TiO2 creates abundant interfaces that
Reproduced with permission.[13] Copyright 2015, Elsevier.

creating distinct redox active sites on different planes, thereby


suppressing the back oxidation of CO to CO2.[54a]
In addition to their role as co-catalysts, Au nanoparticles
show a strong plasmonic absorption in visible region which
enables Au NPs to function as photosensitizers for visible light-
driven photocatalytic reactions.[88] However, it is demonstrated
that only trace amounts of CH4 are produced under Au
plasmonic stimulation, while CO and H2 are generated through
interband-transitions.[62,89] The low activity in the plasmonic
regime may be due to the short lifetime of hot electrons, with
their energy not sufficient to drive reactions with more negative
potentials than the CO2/CH4 redox couple.[90] On the contrary,
the interband transition electrons of Au can be excited into
empty sp states with sufficient energy to satisfy the potentials
required for further CO2 reduction reactions. A 0.5 wt.% Rh/
SrTiO3 photocatalyst co-loaded with Au nanoparticles showed
good activity for visible light-driven photocatalytic syngas
production in the presence of CO2 and water, with an adjustable
CO/H2 product ratio depending on the Au/Rh loading
(Figure 12).[62] A fraction of the electrons excited in Au
nanoparticles under visible light irradiation were transferred to
the Rh nanoparticles (a known effective catalyst for CH4
reforming), thus suppressing CH4 production and creating an
active site for syngas production.

4.2. Interfacial Heterojunctions

The wide bandgap of most perovskite oxides allows the


photogeneration of charge carriers capable of driving a wide
range of redox reactions. However, limited charge carrier
mobility and low light absorption at solar wavelengths greatly
hinder the photocatalytic performance. Interfacial heterojunc-
tions with other semiconductors or organic sensitzers allow
many of the inherent limitations of perovskite oxides to be
Figure 12. (a) Product evolution rate over 0.5 wt.% Rh-SrTiO3 co-loaded
overcome, thus affording highly efficient CO2 photoreduction with different amounts of Au (λ > 400 nm, 5 h). (b) Product evolution rate
through fast interfacial charge transfer processes and an over 0.5 wt.% Au-SrTiO3 co-loaded with different amounts of Rh
extended visible light response.[59,65,67] Polarity discontinuities (λ > 400 nm, 5 h). Reproduced with permission.[62] Copyright 2016, Royal
at the interfaces of metal oxide heterostructures can lead to Society of Chemistry.

Sol. RRL 2017, 1700126 1700126 (13 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

directed electron transmission and also a ferroelec-


tric effect caused by lattice distortion at the
interface.[94] The same distortion-induced ferroelec-
tric effect exists in many perovskite oxides.[28a] The
intrinsic electric field promotes the directional
migration of photoexcited electrons and holes,
thereby improving the overall charge separation
efficiency of the catalyst. The concept has been
successfully applied to photocatalytic CO2 reduction
based on SrTiO3/TiO2 coaxial nanotube arrays,
which provided efficient and stable photoexcited
charge separation.[14] Similar charge separation
improvements were also found in LaFeO3/TiO2
heterostructures using surface photovoltage spec-
troscopy.[47] The surface phase junction formed
between cubic/orthorhombic NaNbO3 yielded en-
hanced CH4 production during CO2 photoreduction
compared with either cubic or orthorhombic phase
NaNbO3.[38,95]
Loading SrTiO3 with amorphous CuxO (x ¼ 1 or
Figure 13. (a) Schematic of the experimental setup for single-particle spectroelec- 2) increased the rate of CO2 photoreduction to CO by
trochemistry. PL images of SrTiO3 mesocrystals with an increased aging time from (b) a factor of 4.[18] The formation of Cuþ during UV
24 h to c) 48 h. d) Typical PL spectrum, and (e) applied potential dependence of the PL
light irradiation confirmed the electron acceptor role
intensity obtained for 48 h aged SrTiO3 mesocrystals. The test region is labeled with a
[93] of CuxO, which acts as metal oxide co-catalyst for
dashed square in (c). Reproduced with permission. Copyright 2017, Wiley-VCH.
photocatalytic CO2 reduction. Similar electron
acceptor properties were found for carbon quantum
promote charge separation and transfer. Recently, Majima et al.
dots on Bi2WO6.[71] The formation of heterostructures between
used temporal and spatial spectroscopy to investigate the photo-
perovskite oxides and other non-oxide semiconductors such as g-
charge dynamics of an epitaxially grown SrTiO3 nanocube
C3N4[58b,61] and ZnTe[66] generates a type-II band alignment for
mesocrystal superstructure (Figure 13). The nanocubes were
[93]
charge separation, whilst the light absorption of the hetero-
in intimate contact within the mesocrystal superstructure,
structure is extended into the visible due to the narrower
ensuring effective electron transfer channel between nanocubes.
bandgap of the second semiconductor. Perovskite oxide/organic
Experimental evidence revealed a spatial concentration effect of
hybrid nanostructures represent a further avenue for CO2
photogenerated electrons due to the size disorder of the SrTiO3
photoreduction. For example, conducting polymers such as
nanocubes. Such interface directed charge transfer promotes
polythiophene, polyaniline and polypyrrole coupled with
charge separation, and is expected to be highly beneficial for
Bi2WO6 showed much higher photocurrents and CH3OH
multi-electron processes such as CO2 reduction. A 87.7% charge
production activity under visible light compared with pristine
separation efficiency was achieved for TiO2@SrTiO3 core-shell
Bi2WO6, indicating that the excellent conductivity of the
nanowires, which is due to the 1-D nanostructure-induced
conducting polymer was highly beneficial in aiding separation
of photoexcited electron/hole pairs in Bi2WO6.[72] Furthermore,
binuclear Ru(II) complexes coupled to perovskite semiconduc-
tors exhibit impressive visible light-driven CO2 reduction activity
and HCOOH selectivity (Figure 14).[11a] Both CaTaO2N and the
Ru(II) tris-bipyridyl complex show a strong light absorption up
to 500 nm, with the photoexcited electrons then being
transferred to the Ru(II) complex responsible for CO2 reduction
via intramolecular electron transfer. However, the valence band
energy level of CaTaO2N is not positive enough to drive the water
oxidation half-reaction, thus requiring the use of a sacrificial
reagent to prevent CaTaO2N photocorrosion.

5. Conclusions and Future Perspectives


In this review we have examined the current state of the art in
Figure 14. Schematic illustration of Z-Scheme CO2 reduction under
photocatalytic CO2 reduction based on perovskite oxides. We
visible light based on a hybrid system consists of a semiconductor and a highlighted the most important recent advances in the
binuclear Ru complex. Reproduced with permission.[11a] Copyright 2015, understanding of the structure and structural modification of
American Chemical Society. perovskite oxides, placing special emphasis on defect

Sol. RRL 2017, 1700126 1700126 (14 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

engineering and interfacial modulation through heteroatom Strategic Priority Research Program of the Chinese Academy of Sciences
doping, oxygen vacancy and heterostructures formed with metal (XDB17030300), and the Youth Innovation Promotion Association of the
co-catalysts, other semiconductors and metal-organic com- Chinese Academy of Sciences is gratefully acknowledged.
plexes. At present, photocatalytic CO2 reduction rates on
perovskite oxides typically fall in the mmol h1 range. This
means that current catalytic efficiencies are still too low to Conflict of Interest
warrant serious consideration for practical solar-to-fuel The authors declare no conflict of interest.
applications.
To improve the efficiency of perovskite oxide photocatalysts
for CO2 reduction, a number of technical hurdles need to be Keywords
carefully addressed and overcome. Metal heteroatom doping in
perovskite oxides shows promise, promoting the adsorption of CO2 photoreduction, defect engineering, perovskite oxides,
CO2, decreasing the activation energy for CO2 conversion and photocatalysis, solar fuel
enhancing the amount of photogenerated electrons on the
surface to accelerate the reduction reaction. However, the wide Received: August 1, 2017
choice of possible dopants, doping sites and synergistic effects of Revised: August 22, 2017
multi-dopants makes performance optimization a very compli- Published online:
cated task. Oxygen vacancies are frequently generated during
heteroatom doping through the associated crystal distortions
and charge compensation, which complicates the exact role of [1] B. Dudley, BP Statistical Review of World Energy, 2016. http://www.
the metal dopant in promoting CO2 photoreduction. To date, bp.com/statisticalreview
most information about the effect of metal co-catalysts on CO2 [2] M. McGee, Annual Global Carbon Emissions and Atmospheric
Accumulation, 2017. http://www.co2.earth
reduction has been inferred from mechanistic studies of
[3] S. N. Habisreutinger, L. Schmidt-Mende, J. K. Stolarczyk, Angew.
electrocatalytic CO2 reduction using active metal as working Chem. Int. Ed. 2013, 52, 7372.
electrode. Whilst there are obvious similarities between these [4] G. A. Olah, G. K. Prakash, A. Goeppert, J. Am. Chem. Soc. 2011, 133,
systems, the influence of the perovskite oxide support needs to 12881.
be considered when explaining the diverse CO2 photoreduction 
[5] A. Alvarez, A. Bansode, A. Urakawa, A. V. Bavykina, T. A. Wezendonk,
activities and product selectivities of metal-loaded perovskite M. Makkee, J. Gascon, F. Kapteijn, Chem. Rev. 2017, 117, 9804.
oxides. Other factors, such as the influence of the metal/ [6] K. Li, B. Peng, T. Peng, ACS Catal. 2016, 6, 7485.
perovskite interface on exciton transfer processes need deeper [7] a)Q. Yi, W. Li, J. Feng, K. Xie, Chem. Soc. Rev. 2015, 44, 5409;
investigation. Perovskite oxides typically possess a wide b)Y. F. Xu, M. Z. Yang, B. X. Chen, X. D. Wang, H. Y. Chen,
bandgap, which means that they are not particularly effective D. B. Kuang, C. Y. Su, J. Am. Chem. Soc. 2017, 139, 5660.
in utilizing the solar spectrum to drive photoreactions. Most [8] C. Huang, Z. Li, Z. Zou, MRS Commun. 2016, 6, 216.
[9] H. Yoshida, Energy Efficiency and Renewable Energy Through
reported perovskite oxides are only active under UV light (4%
Nanotechnology. Ed.: L. Zang Springer London, London 2011,
of total solar energy). Doping with non-metallic elements or p. 531.
forming solid solutions can increase the visible-light respon- [10] a)A. A. Peterson, J. K. Nørskov, J. Phys. Chem. Lett. 2012, 3, 251;
se.[57b,96] b)H. Yoshio, K. Katsuhei, S. Shin, Chem. Lett. 1985, 14, 1695.
Widespread research continues toward increasing the activity, [11] a)F. Yoshitomi, K. Sekizawa, K. Maeda, O. Ishitani, ACS Appl. Mater.
selectivity and stability of perovskite oxide-based photocatalysts Interfaces 2015, 7, 13092. b)V. Jeyalakshmi, R. Mahalakshmy,
for CO2 photoreduction. Heterostructured photocatalysts which K. R. Krishnamurthy, B. Viswanathan, Catal. Today 2016, 266, 160.
couple perovskite oxides with other narrower bandgap semi- [12] H. Yoshida, L. Zhang, M. Sato, T. Morikawa, T. Kajino, T. Sekito,
conductors, appear the most practical path forward for the S. Matsumoto, H. Hirata, Catal. Today 2015, 251, 132.
development of more efficient CO2 reduction photocatalysts, [13] Z. Wang, K. Teramura, S. Hosokawa, T. Tanaka, Appl. Catal. B-
benefitting from enhanced charge separation and increased Environ. 2015, 163, 241.
solar spectrum utilization. Dramatic improvements in photoex- [14] Q. Kang, T. Wang, P. Li, L. Liu, K. Chang, M. Li, J. Ye, Angew. Chem.
cited charge separation and photocatalyst performance may be Int. Ed. 2015, 54, 841.
achieved through manipulation and exploitation of the inherent [15] M. Li, P. Li, K. Chang, T. Wang, L. Liu, Q. Kang, S. Ouyang, J. Ye,
ferroelectric properties shown by many perovskite oxides. Chem. Commun. 2015, 51, 7645.
Considerable work remains before the efficiencies of perovskite [16] Y.-H. Cheng, V.-H. Nguyen, H.-Y. Chan, J. C. S. Wu, W.-H. Wang,
Appl. Energy 2015, 147, 318.
oxide-based photocatalyst can compete with conventional metal-
[17] H. Nakanishi, K. Iizuka, T. Takayama, A. Iwase, A. Kudo,
based electrocatalysts for CO2 reduction.
ChemSusChem 2017, 10, 112.
[18] S. Shoji, G. Yin, M. Nishikawa, D. Atarashi, E. Sakai, M. Miyauchi,
Chem. Phys. Lett. 2016, 658, 309.
Acknowledgements [19] C.-C. Yang, Y.-H. Yu, B. van der Linden, J. C. S. Wu, G. Mul, J. Am.
Support from the Ministry of Science and Technology of China Chem. Soc. 2010, 132, 8398.
(2014CB239402, 2013CB834505), the National Key Projects for [20] G. Liu, X. Meng, H. Zhang, G. Zhao, H. Pang, T. Wang, P. Li, T. Kako,
Fundamental Research and Development of China (2016YFB0600901, J. Ye, Angew. Chem. Int. Ed. 2017, 56, 5570.
2017YFA0206904, 2017YFA0206900), the National Natural Science [21] Z. Li, L. Zhong, F. Yu, Y. An, Y. Dai, Y. Yang, T. Lin, S. Li, H. Wang,
Foundation of China (U1662118, 51572270, 51772305, 21401207), the P. Gao, Y. Sun, M. He, ACS Catal. 2017, 7, 3622.

Sol. RRL 2017, 1700126 1700126 (15 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

[22] N. Erini, V. Beermann, M. Gocyla, M. Gliech, M. Heggen, R. E. Dunin- [48] A. M. Srivastava, J. F. Ackerman, W. W. Beers, J. Solid State Chem.
Borkowski, P. Strasser, Angew. Chem. Int. Ed. 2017, 56, 6533. 1997, 134, 187.
[23] a)Y. Ma, X. Wang, Y. Jia, X. Chen, H. Han, C. Li, Chem. Rev. 2014, 114, [49] a)H. N. Lee, D. Hesse, N. Zakharov, U. Gösele, Science 2002, 296;
9987; b)W. Wang, M. O. Tade, Z. Shao, Chem. Soc. Rev. 2015, 44, b)Y. Moritomo, A. Asamitsu, H. Kuwahara, Y. Tokura, Nature 1996,
5371; c)P. Liao, E. A. Carter, Chem. Soc. Rev. 2013, 42, 2401. 380, 141; c)Y. Maeno, H. Hashimoto, K. Yoshida, S. Nishizaki,
[24] a)Y. X. Pan, Y. You, S. Xin, Y. Li, G. Fu, Z. Cui, Y. L. Men, F. F. Cao, T. Fujita, J. G. Bednorz, F. Lichtenberg, Nature 1994, 372, 532.
S. H. Yu, J. B. Goodenough, J. Am. Chem. Soc. 2017, 139, 4123; [50] J.-C. Blancon, H. Tsai, W. Nie, C. C. Stoumpos, L. Pedesseau,
b)Y.-X. Pan, Z.-Q. Sun, H.-P. Cong, Y.-L. Men, S. Xin, J. Song, C. Katan, M. Kepenekian, C. M. M. Soe, K. Appavoo, M. Y. Sfeir,
S.-H. Yu, Nano Res. 2016, 9, 1689. S. Tretiak, P. M. Ajayan, M. G. Kanatzidis, J. Even, J. J. Crochet,
[25] a)L. Zhu, Z. Shao, J. Ye, X. Zhang, X. Pan, S. Dai, Chem. Commun. A. D. Mohite, Science 2017, 355, 1288.
2016, 52, 970; b)A. Bera, K. Wu, A. Sheikh, E. Alarousu, [51] a)Y. Zhou, Z. Tian, Z. Zhao, Q. Liu, J. Kou, X. Chen, J. Gao, S. Yan,
O. F. Mohammed, T. Wu, J. Phys. Chem. C 2014, 118, 28494; Z. Zou, ACS Appl. Mater. Interfaces 2011, 3, 3594; b)W. Dai, J. Yu,
c)I. Grinberg, D. V. West, M. Torres, G. Gou, D. M. Stein, L. Wu, H. Xu, X. Hu, X. Luo, L. Yang, X. Tu, CrystEngComm 2016, 18, 3472.
G. Chen, E. M. Gallo, A. R. Akbashev, P. K. Davies, J. E. Spanier, [52] a)V. Jeyalakshmi, R. Mahalakshmy, K. Ramesh, P. V. C. Rao, N. V. Choudary,
A. M. Rappe, Nature 2013, 503, 509; d)K. Singh, J. Nowotny, G. S. Ganesh, K. Thirunavukkarasu, K. R. Krishnamurthy, B. Viswanathan,
V. Thangadurai, Chem. Soc. Rev. 2013, 42. RSC Adv. 2015, 5, 5958; b)C. Galven, J. L. Fourquet, E. Suard, M. P. Crosnier-
[26] a)Y. Zhu, W. Zhou, Z. Shao, Small 2017, 13, 1603793; b)M. Li, Lopez, F. Le Berre, Dalton Trans. 2010, 39, 4191.
M. J. Pietrowski, R. A. De Souza, H. Zhang, I. M. Reaney, S. N. Cook, [53] F. Lichtenberg, A. Herrnberger, K. Wiedenmann, J. Mannhart, Prog.
J. A. Kilner, D. C. Sinclair, Nat. Mater. 2014, 13, 31; c)J. Suntivich, Solid State Chem. 2001, 29, 1.
H. A. Gasteiger, N. Yabuuchi, H. Nakanishi, J. B. Goodenough, [54] a)K. Iizuka, T. Wato, Y. Miseki, K. Saito, A. Kudo, J. Am. Chem. Soc.
Y. Shao-Horn, Nat. Chem. 2011, 3, 546; d)G. M. Rupp, A. K. Opitz, 2011, 133, 20863; b)Y. Miseki, H. Kato, A. Kudo, Energy Environ. Sci.
A. Nenning, A. Limbeck, J. Fleig, Nat. Mater. 2017, 16, 640. 2009, 2, 306.
[27] a)G. Zhang, G. Liu, L. Wang, J. T. Irvine, Chem. Soc. Rev. 2016, 45, [55] Y. Ebina, N. Sakai, T. Sasaki, J. Phys. Chem. B 2005, 109, 17212.
5951; b)E. Grabowska, Appl. Catal. B- Environ. 2016, 186, 97. [56] Y. Ebina, T. Sasaki, M. Harada, M. Watanabe, Chem. Mater. 2002,
[28] a)L. Liang, X. Kang, Y. Sang, H. Liu, Adv. Sci. 2016, 3, 1500358; 14, 4390.
b)C. M. Fernandez-Posada, A. Castro, J. M. Kiat, F. Porcher, O. Pena, [57] a)F. Meng, Z. Hong, J. Arndt, M. Li, M. Zhi, F. Yang, N. Wu, Nano
M. Alguero, H. Amorin, Nat. Commun. 2016, 7, 12772; c)J. F. Scott, Res. 2012, 5, 213; b)A. Mukherji, C. Sun, S. C. Smith, G. Q. Lu,
M. Dawber, Appl. Phys. Lett. 2000, 76, 3801. L. Wang, J. Phys. Chem. C 2011, 115, 15674.
[29] P. Gao, M. Grätzel, M. K. Nazeeruddin, Energy Environ. Sci. 2014, 7, [58] a)W. Dai, J. Yu, Y. Deng, X. Hu, T. Wang, X. Luo, Appl. Surf. Sci. 2017,
2448. 403, 230; b)M. Li, L. Zhang, X. Fan, Y. Zhou, M. Wu, J. Shi, J. Mater.
[30] a)S. Acharya, S. Martha, P. C. Sahoo, K. Parida, Inorg. Chem. Front. Chem. A 2015, 3, 5189.
2015, 2, 807; b)P. Kanhere, Z. Chen, Molecules 2014, 19, 19995. [59] T. Xiang, F. Xin, J. Chen, Y. Wang, X. Yin, X. Shao, Beilstein J.
[31] F. Fresno, P. Jana, P. Renones, J. M. Coronado, D. P. Serrano, V. A. de Nanotechnol. 2016, 7, 776.
la Pena O’Shea, Photochem. Photobiol. Sci. 2017, 16, 17. [60] H. Shi, G. Chen, C. Zhang, Z. Zou, ACS Catal. 2014, 4, 3637.
[32] H. Shi, Z. Zou, J. Phys. Chem. Solids 2012, 73, 788. [61] H. Shi, C. Zhang, C. Zhou, G. Chen, RSC Adv. 2015, 5, 93615.
[33] J. Hou, S. Cao, Y. Wu, F. Liang, L. Ye, Z. Lin, L. Sun, Nano Energy [62] D. Li, S. Ouyang, H. Xu, D. Lu, M. Zhao, X. Zhang, J. Ye, Chem.
2016, 30, 59. Commun. 2016, 52, 5989.
[34] P. V. Balachandran, J. Young, T. Lookman, J. M. Rondinelli, Nat. [63] J. Kou, J. Gao, Z. Li, H. Yu, Y. Zhou, Z. Zou, Catal. Lett. 2015, 145, 640.
Commun. 2017, 8, 14282. [64] Y. Bi, M. F. Ehsan, Y. Huang, J. Jin, T. He, J. CO2 Utilization 2015,
[35] a)J. H. Lee, A. Selloni, Phys. Rev. Lett. 2014, 112, 196102; b)A. Kudo, 12, 43.
H. Kato, S. Nakagawa, J. Phys. Chem. B 2000, 104, 571. [65] Y. Bi, L. Zong, C. Li, Q. Li, J. Yang, Nanoscale Res. Lett. 2015, 10, 345.
[36] K. Teramura, S.-i. Okuoka, H. Tsuneoka, T. Shishido, T. Tanaka, Appl. [66] M. F. Ehsan, M. N. Ashiq, F. Bi, Y. Bi, S. Palanisamy, T. He, RSC Adv.
Catal. B- Environ. 2010, 96, 565. 2014, 4, 48411.
[37] P. Li, S. Ouyang, G. Xi, T. Kako, J. Ye, J. Phys. Chem. C 2012, 116, 7621. [67] Y. Im, S.-M. Park, M. Kang, Bull. Korean Chem. Soc. 2017, 38, 397.
[38] P. Li, H. Xu, L. Liu, T. Kako, N. Umezawa, H. Abe, J. Ye, J. Mater. [68] H. Cheng, B. Huang, Y. Liu, Z. Wang, X. Qin, X. Zhang, Y. Dai, Chem.
Chem. A 2014, 2, 5606. Commun. 2012, 48, 9729.
[39] K. Li, A. D. Handoko, M. Khraisheh, J. Tang, Nanoscale 2014, 6, 9767. [69] X. Y. Kong, Y. Y. Choo, S. P. Chai, A. K. Soh, A. R. Mohamed, Chem.
[40] H. Zhou, P. Li, J. Guo, R. Yan, T. Fan, D. Zhang, J. Ye, Nanoscale 2015, Commun. 2016, 52, 14242.
7, 113. [70] S. Murcia-Lopez, V. Vaiano, M. C. Hidalgo, J. A. Navio, D. Sannino,
[41] a)Q. Wu, J. Cen, K. R. Goodman, M. G. White, G. Ramakrishnan, Photochem. Photobiol. Sci. 2015, 14, 678.
A. Orlov, ChemSusChem 2016, 9, 1889; b)T. Arai, S. Sato, T. Kajino, [71] X. Y. Kong, W. L. Tan, B.-J. Ng, S.-P. Chai, A. R. Mohamed, Nano Res.
T. Morikawa, Energy Environ. Sci. 2013, 6, 1274. 2017, 10, 1720.
[42] F. Voigts, C. Argirusis, W. Maus-Friedrichs, Surf. Interface Anal 2012, [72] W. Dai, H. Xu, J. Yu, X. Hu, X. Luo, X. Tu, L. Yang, Appl. Surf. Sci. 2015,
44, 301. 356, 173.
[43] K. Xie, N. Umezawa, N. Zhang, P. Reunchan, Y. Zhang, J. Ye, Energy [73] H. Li, J. Li, Z. Ai, F. Jia, L. Zhang, Angew. Chem. Int. Ed. 2017, https://
Environ. Sci. 2011, 4, 4211. doi.org/10.1002/ange.201705628
[44] H. Zhou, J. Guo, P. Li, T. Fan, D. Zhang, J. Ye, Sci. Rep. 2013, 3, 1667. [74] a)V. Jeyalakshmi, R. Mahalakshmy, K. R. Krishnamurthy,
[45] J. Y. Do, Y. Im, B. S. Kwak, S.-M. Park, M. Kang, Ceram. Int. 2016, B. Viswanathan, Catal. Today 2017, https://doi.org/10.1016/j.
42, 5942. cattod.2017.02.050; b)Z. Ma, K. Wu, R. Sa, Q. Li, C. He, Z. Yi,
[46] a)J. Wang, C. Huang, X. Chen, H. Zhang, Z. Li, Z. Zou, Appl. Surf. Sci. Int. J. Hydrogen Energy 2015, 40, 980; c)E. Enriquez, A. Chen,
2015, 358, 463; b)X. Chen, J. Wang, C. Huang, S. Zhang, H. Zhang, Z. Harrell, P. Dowden, N. Koskelo, J. Roback, M. Janoschek, C. Chen,
Z. Li, Z. Zou, Catal. Sci. Technol. 2015, 5, 1758. Q. Jia, Sci. Rep. 2017, 7, 46184.
[47] M. Humayun, Y. Qu, F. Raziq, R. Yan, Z. Li, X. Zhang, L. Jing, Environ. [75] a)Y. Sakai, J. Yang, R. Yu, H. Hojo, I. Yamada, P. Miao, S. Lee,
Sci. Technol. 2016, 50, 13600. S. Torii, T. Kamiyama, M. Lezaic, G. Bihlmayer, M. Mizumaki,

Sol. RRL 2017, 1700126 1700126 (16 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.solar-rrl.com

J. Komiyama, T. Mizokawa, H. Yamamoto, T. Nishikubo, Y. Hattori, [85] a)M. Ali, F. Zhou, K. Chen, C. Kotzur, C. Xiao, L. Bourgeois, X. Zhang,
K. Oka, Y. Yin, J. Dai, W. Li, S. Ueda, A. Aimi, D. Mori, Y. Inaguma, D. R. MacFarlane, Nat. Commun. 2016, 7, 11335; b)Y. Bing, H. Liu,
Z. Hu, T. Uozumi, C. Jin, Y. Long, M. Azuma, J. Am. Chem. Soc. L. Zhang, D. Ghosh, J. Zhang, Chem. Soc. Rev. 2010, 39, 2184;
2017, 139, 4574; b)D. Wang, J. Ye, T. Kako, T. Kimura, J. Phys. Chem. c)A. Y. Khodakov, W. Chu, P. Fongarland, Chem. Rev. 2007, 107, 1692.
B 2006, 110, 15824. [86] a)P. V. Kamat, J. Phys. Chem. Lett. 2012, 3, 663; b)W. Hou,
[76] a)R. Konta, T. Ishii, H. Kato, A. Kudo, J. Phys. Chem. B 2004, 108, S. B. Cronin, Adv. Funct. Mater. 2013, 23, 1612.
8992; b)Q. Wang, T. Hisatomi, Q. Jia, H. Tokudome, M. Zhong, [87] a)V. Subramanian, E. E. Wolf, P. V. Kamat, J. Am. Chem. Soc. 2004,
C. Wang, Z. Pan, T. Takata, M. Nakabayashi, N. Shibata, Y. Li, 126, 4943; b)A. Wood, M. Giersig, P. Mulvaney, J. Phys. Chem. B 2001,
I. D. Sharp, A. Kudo, T. Yamada, K. Domen, Nat. Mater. 2016, 15, 611. 105, 8810.
[77] M. Ezbiri, V. Becattini, M. Hoes, R. Michalsky, A. Steinfeld, [88] a)K. Wu, J. Chen, J. R. McBride, T. Lian, Science 2015, 349, 632;
ChemSusChem 2017, 10, 1517. b)C. Gomes, Silva, R. Juarez, T. Marino, R. Molinari, H. García, J. Am.
[78] S. T. Hunt, M. Milina, A. C. Alba-Rubio, C. H. Hendon, J. A. Dumesic, Chem. Soc. 2011, 133, 595.
Y. Roman-Leshkov, Science 2016, 352, 974. [89] a)M. Liu, W. Zhou, T. Wang, D. Wang, L. Liu, J. Ye, Chem. Commun.
[79] I. Mikulska, M. Valant, I. Arcon, D. Lisjak, A. Belik, J. Am. Ceram. Soc. 2016, 52, 4694; b)Y. Sugano, Y. Shiraishi, D. Tsukamoto, S. Ichikawa,
2015, 98, 1156. S. Tanaka, T. Hirai, Angew. Chem. Int. Ed. 2013, 52, 5295; c)S. Neatu,
[80] H. Suzuki, O. Tomita, M. Higashi, R. Abe, J. Mater. Chem. A 2016, 4, J. A. Macia-Agullo, P. Concepcion, H. Garcia, J. Am. Chem. Soc. 2014,
14444. 136, 15969.
[81] X. Jiao, Z. Chen, X. Li, Y. Sun, S. Gao, W. Yan, C. Wang, Q. Zhang, [90] a)W. Hou, W. H. Hung, P. Pavaskar, A. Goeppert, M. Aykol,
Y. Lin, Y. Luo, Y. Xie, J. Am. Chem. Soc. 2017, 139, 7586. S. B. Cronin, ACS Catal. 2011, 1, 929; b)C. Voisin, N. Del Fatti,
[82] a)A. Karaphun, S. Hunpratub, S. Phokha, T. Putjuso, E. Swatsitang, D. Christofilos, F. Vallee, J. Phys. Chem. B 2001, 105, 2264.
J. Mater. Sci.: Mater. Electron. 2017, 28, 8294; b)K. T. Kang, H. Kang, [91] A. Ohtomo, H. Y. Hwang, Nature 2004, 427, 423.
J. Park, D. Suh, W. S. Choi, Adv. Mater. 2017, 29, 1700071; c)J. Zhang, [92] R. Ohshima, Y. Ando, K. Matsuzaki, T. Susaki, M. Weiler, S. Klingler,
K. Xie, H. Wei, Q. Qin, W. Qi, L. Yang, C. Ruan, Y. Wu, Sci. Rep. 2014, H. Huebl, E. Shikoh, T. Shinjo, S. T. B. Goennenwein, M. Shiraishi,
4, 7082. Nat. Mater. 2017, 16, 609.
[83] S. Y. Choi, S. D. Kim, M. Choi, H. S. Lee, J. Ryu, N. Shibata, [93] P. Zhang, T. Ochi, M. Fujitsuka, Y. Kobori, T. Majima, T. Tachikawa,
T. Mizoguchi, E. Tochigi, T. Yamamoto, S. J. Kang, Y. Ikuhara, Nano Angew. Chem. Int. Ed. 2017, 56, 5299.
Lett. 2015, 15, 4129. [94] F. Wu, Y. Yu, H. Yang, L. N. German, Z. Li, J. Chen, W. Yang,
[84] X. Lin, J. C. Lu, Y. Shao, Y. Y. Zhang, X. Wu, J. B. Pan, L. Gao, L. Huang, W. Shi, L. Wang, X. Wang, Adv. Mater. 2017, 1701432.
S. Y. Zhu, K. Qian, Y. F. Zhang, D. L. Bao, L. F. Li, Y. Q. Wang, [95] P. Li, S. Ouyang, Y. Zhang, T. Kako, J. Ye, J. Mater. Chem. A 2013,
Z. L. Liu, J. T. Sun, T. Lei, C. Liu, J. O. Wang, K. Ibrahim, 1, 1185.
D. N. Leonard, W. Zhou, H. M. Guo, Y. L. Wang, S. X. Du, [96] G. Zhang, S. Sun, W. Jiang, X. Miao, Z. Zhao, X. Zhang, D. Qu,
S. T. Pantelides, H. J. Gao, Nat. Mater. 2017, 16, 717. D. Zhang, D. Li, Z. Sun, Adv. Energy Mater. 2017, 7, 1600932.

Sol. RRL 2017, 1700126 1700126 (17 of 17) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

View publication stats

You might also like