Solar fuels photoanode materials discovery by
integrating high-throughput theory and experiment
Qimin Yan
a,b,1,2
, Jie Yu
c,d,e,2
, Santosh K. Suram
c
, Lan Zhou
c
, Aniketa Shinde
c
, Paul F. Newhouse
c
, Wei Chen
d,3
, Guo Li
a,b,e
,
Kristin A. Persson
d,f
, John M. Gregoire
c,1
, and Jeffrey B. Neaton
a,b,e,g,1
a
Molecular Foundry, Lawrence Berkeley National Laboratory, Berkeley, CA 94720;
b
Department of Physics, University of California, Berkeley, CA 94720;
c
Joint Center for Artificial Photosynthesis, California Institute of Technology, Pasadena, CA 91125;
d
Environmental Energy Technologies Division, Lawrence
Berkeley National Laboratory, Berkeley, CA 94720;
e
Joint Center for Artificial Photosynthesis, Lawrence Berkeley National Laboratory, Berkeley, CA 94720;
f
Department of Materials Science and Engineering, University of California, Berkeley, CA 94720; and
g
Kavli Energy NanoSciences Institute at Berkeley,
Berkeley, CA 94720
Edited by Thomas E. Mallouk, The Pennsylvania State University, University Park, PA, and approved February 6, 2017 (received for review December 4, 2
016)
The limited number of known low-band-gap photoelectrocatalytic
materials poses a significant challenge for the generation of chem-
ical fuels from sunlight. Using high-throughput ab initio theory with
experiments in an integrated workflow, we find eight ternary van-
adate oxide photoanodes in the target band-gap range (1.2
–
2.8 eV).
Detailed analysis of these vanadate compounds reveals the key
role of VO
4
structural motifs and electronic band-edge character in
efficient photoanodes, initiating a genome for such materials and
paving the way for a broadly applicable high-throughput-discovery
and materials-by-design feedback loop. Considerably expanding
the number of known photoelectrocatalysts for water oxidation,
our study establishes ternary metal vanadates as a prolific class of
photoanode materials for generation of chemical fuels from sunlight
and demonstrates our high-throughput theory
–
experiment pipeline
as a prolific approach to materials discovery.
solar fuels materials
|
density-functional theory
|
high-throughput
experiments
|
complex oxides
|
photocatalysis
T
he use of predictive simulation in combination with experi-
ments for the accelerated discovery and rational design of
functional materials is a challenge of significant contemporary
interest. High-throughput computing and materials databases
(1
–
3), largely based on density-functional theory (DFT), have
recently enabled rapid screening of solid-state compounds with
simulation for multiple properties and functionalities (4
–
10).
Since their advent just a few years ago, these DFT-based data-
bases and analytics tools have already been used to identify more
than 20 new functional materials that were later confirmed by
experiments across a number of applications (8), motivating
concerted efforts to validate theory predictions with experiments
(11). However, in photoelectrochemistry for the renewable syn-
thesis of solar fuels, efficient metal-oxide photoanode materials
––
photoelectrocatalysts for the oxygen evolution reaction (OER)
––
remain critically missing (12). Forty years of experimental re-
search has yielded just 16 metal-oxide photoanode compounds
with band-gap energy in the desirable 1.2
–
2.8-eV range that
strongly overlaps with the solar spectrum. Prior high-throughput
computational screening studies have yet to expand this list (6, 7,
13), in part due to quantitative limitations in predictability of the
electronic structure
––
especially band-gap energy, E
g
, and the
valence band maximum (VBM) energy, E
VBM
––
from the chemical
composition and crystal structu
re. Our integration of ab initio
theory with high-throughput experiments has yielded a most prolific
materials discovery effort, as dem
onstrated by the identification of
12 water oxidation photoelectrocatalysts in the target band-gap
range, including our recently reported 4 copper vanadates (14) and
8 additional metal vanadates reported here.
Monoclinic BiVO
4
(15) has received substantial attention as a
solar fuels photoanode material due to its promising OER
photoactivity. It has a desirable 2.4-eV band gap derived from a
conduction band minimum (CBM) consisting of V 3d states, and
a VBM of mixed O 2p and Bi 6s character (16). We recently
identified another ternary vanadate,
β
-Mn
2
V
2
O
7
, that, although
not photoactive for the OER, exhibits a 1.8-eV band gap and
valence band alignment to the OER equilibrium energy, or OER
potential, a result of hybridization of Mn 3d with O 2p states.
β
-Mn
2
V
2
O
7
shares a common VO
4
structural motif with BiVO
4
:
Both compounds possess 3d
0
V cations tetrahedrally coordinated
by oxygen. Orbital hybridization resulting from this VO
4
motif
engenders significant baseline O 2p and V 3d character at the
VBM and CBM, creating an electronic structure
“
scaffold
”
that
enables the formation of a desirable E
g
and E
VBM
upon intro-
duction of an additional metal cation (17). For example, in BiVO
4
and
β
-Mn
2
V
2
O
7
, additional hybridization of O 2p with Mn 3d and
Bi 6s states, respectively, leads to an increase in the E
VBM
toward
the OER potential. Here, we hypothesize that this VO
4
-scaffold
phenomenon applies broadly to ternary vanadates. Indeed, we
observe that three additional previously known OER photoanodes
with band-gap energies between 1.2 and 2.8 eV [
α
-Ag
3
VO
4
(18),
FeVO
4
(19, 20), and
β
-Cu
3
V
2
O
8
(21)] are also ternary vanadates
with a VO
4
structural motif in the 3d
0
electronic configuration. For
these reasons, VO
4
-based ternary compounds are fertile ground
for both discovering metal-oxide photoanodes and seeding the
photoanode materials genome.
Current computational and experimental approaches differ in
the material properties they can most efficiently and effectively
characterize, and by leveraging their complementarity we can
both accelerate the materials discovery process and confirm the
validity of each approach. After several generations of integrated
computational and experimental materials screening workflows,
we arrived at the tiered screening pipeline of Fig. 1 in which we
(
i
) selectively mine a materials database with a well-defined
hypothesis to arrive at a subset of promising materials; (
ii
) screen
Significance
Combining high-throughput computation and experiment
accelerates the discovery of photoelectrocatalysts for water
oxidation and explains the origin of their functionality, estab-
lishing ternary metal vanadates as a prolific class of photo-
anode materials for generation of chemical fuels from sunlight.
Author contributions: Q.Y., K.A.P., J.M.G., and J.B.N. designed research; Q.Y., J.Y., S.K.S.,
L.Z., A.S., P.F.N., W.C., and G.L. performed calculations and experiments; Q.Y., S.K.S., K.A.P.,
J.M.G., and J.B.N. analyzed data; and Q.Y., J.M.G., and J.B.N. wrote the paper with
contributions from all authors.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1
To whom correspondence may be addressed. Email: qiminyan@temple.edu, gregoire@
caltech.edu, or jbneaton@berkeley.edu.
2
Present address: Department of Physics, Temple University, Philadelphia, PA 19122.
3
Present address: Department of Mechanical, Materials, and Aerospace Engineering, Illinois
Institute of Technology, Chicago, IL 60616.
This article contains supporting information online at
www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1619940114/-/DCSupplemental
.
3040
–
3043
|
PNAS
|
March 21, 2017
|
vol. 114
|
no. 12
www.pnas.org/cgi/doi/10.1073/pnas.1619940114
this materials subset for select properties with high-throughput
computation at appropriate levels of theory; and (
iii
) use com-
panion combinatorial experiments on the same subset to both
validate the calculations and characterize material performance
under device-relevant conditions. In the present work we dem-
onstrate the efficacy of this approach for the discovery of ternary
vanadate photoanodes and note that the strategies outlined herein
are broadly applicable for discovery of functional materials.
To evaluate the VO
4
-scaffold hypothesis, the 7 tiers of
screening commence with a query of the Materials Project (MP)
database (1) to identify 174 known VO
4
-based ternary vanadates.
Out of these 174 ternary vanadate compounds, 147 compounds
are indexed in the Inorganic Crystal Structure Database as previ-
ously synthesized materials, and the remaining 27 compounds are
the results of previous structure prediction calculations within the
MP. For each compound, the database also provides the DFT
formation energy above the convex hull in the composition phase
diagram (
Δ
H) and a coarse estimate of E
g
. These quantities are
computed using DFT with the generalized gradient approximation
of Perdew, Burke, and Ernzerhof (PBE) and Hubbard U (22)
corrections for metal cation d states with the Vienna Ab initio
Simulation Package (VASP) code (23). Our DFT-PBE
+
Ucalcu-
lations provide a high-throughput tier 2 screen that thresholds
Δ
H
and E
g
in an effort to avoid nonsynthesizable materials (6) and
known wide-gap insulators, respectively. To more accurately screen
and identify semiconductors with E
g
in the target range, we note
that a higher level of theory would be in principle required, such as
ab initio many-body perturbation theory (24), at significantly
greater computational expense. As a compromise that balances
computational efficiency with adequate accuracy, we proceed
in practice with DFT but use generalized Kohn
–
Sham states
obtained from the hybrid functional of Heyd, Scuseria, and Ern-
zerhof (HSE) with a modified mixing parameter,
α
,of0.17(
SI
Appendix
) that results in more tolerant acceptance criteria for tier
3. The HSE functional features improved treatment of short-range
exchange and correlation effects relative to PBE
+
U that treat
delocalizedspvalentandlocalizeddstatesonequalfooting(25)
and can lead to more predictive band gaps (26). Finally, in tier 4,
we evaluate E
VBM
using a combination of bulk HSE (
α
=
0.17) and
surface slab PBE
+
U calculations (27, 28) to determine if the VBM
meets the OER thermodynamic requirement (
SI Appendix
). Surface
dipoles that arise upon interfacing the metal oxides with water are
expected to slightly raise E
VBM
(28); evaluating E
VBM
relative to
vacuum, as we do here, is therefore a
particularly lenient screening
criterion for allowing the ensuing experiments to demonstrate op-
erational VBM alignment via photocurrent measurements.
To validate these screening criteria and their propensity for
identifying photoanode materials, we turn to combinatorial ex-
periments. The most common bottleneck in computation-guided
experimental work is synthesis of the target materials in a device-
relevant format, prompting our development of combinatorial
sputtering and annealing methods in which thin-film synthesis is
attempted for each target phase (and its off-stoichiometry vari-
ants) using a variety of reactive sputtering and annealing con-
ditions. The materials exhibiting target-phase purity in excess of
80% pass to tiers 6 and 7, where high-throughput optical spec-
troscopy (UV-vis) and photoelectrochemistry characterize E
g
(direct and indirect transitions) and the photocurrent density at
the Nernstian potential (J
O2/H2O
), respectively. It is important to
note that for these final tiers of screening, the optical and pho-
toelectrochemical properties are measured not only for the target
phase but also neighboring phases in the composition space, en-
abling identification of E
g
and J
O2/H2O
values that are representative
of the target phase. The high-throughput J
O2/H2O
measurements use
sufficiently energetic photons (
∼
3.2 eV) to excite semiconductors
across the entire E
g
range.
The list of tier 4 compounds is provided in
SI Appendix
and
includes several examples of multiple polytypes with a common
formula unit, and often a subset of these polytypes are synthe-
sized in high purity using the combinatorial synthesis techniques.
Other phases may not be accessible with these synthesis tech-
niques due to kinetic limitations or the presence of stable phases
that are absent from the MP database. In total, 17 high-purity
compounds were obtained, comprising tier 5 of the screening
Fig. 1.
Tiered screening pipeline for accelerated discovery of solar fuels
photoanodes. The number of compounds (bold) and screening criteria used
in this study for the seven-tier pipeline that integrates database mining (gray),
high-throughput computational screening (blue), and high-throughput ex-
perimental screening (red).
Fig. 2.
Landscape of photoactive structures identified by the pipeline. The experimentally measured band-gap energy and OER photocurrent density are
shown for 15 ternary vanadate photoanodes (
A
–
O
; see Table 1) organized by cation electronic configuration. The respective crystal structures share a
common VO
4
motif (gray tetrahedra with gray V and red O) and span a broad range of structures (colored by cation element).
Yan et al.
PNAS
|
March 21, 2017
|
vol. 114
|
no. 12
|
3041
APPLIED PHYSICAL
SCIENCES
pipeline. Band-gap estimation through automated Tauc analysis
experimentally verified the band-gap energy range for 16 of these
17 phases, demonstrating the fidelity of our DFT-HSE (
α
=
0.17)
band-gap calculations. The most striking acceptance rate in the
screening pipeline is the final screening criterion, where 15 of 16
phases exhibited photoactivity when biased at the OER poten-
tial, an exceptionally high hit rate of 94%.
The 15 metal-oxide photoanodes identified by the pipeline
are illustrated in Fig. 2, demonstrating that desirable E
g
and
photoactivity are obtained for ternary vanadates with a range
of cation elements and respective electronic configurations, a
marked departure from the d
0
configuration of photoanodes such
as TiO
2
and WO
3
. The band-gap and J
O2/H2O
values are also
listed in Table 1, which includes three photoanodes reported by
other investigators, validating the pipeline
’
s ability to replicate
known results. Several other phases in Table 1 have been ex-
plored as photoactive materials for different reactions or condi-
tions [
γ
-Cu
3
V
2
O
8
(29), Cr
2
V
4
O
13
(30), and
β
-Ag
3
VO
4
(14) for
photodegradation of organics and Ni
3
V
2
O
8
as a powder photo-
catalyst for water oxidation (31)] but have not been studied or
demonstrated as photoanodes. The materials identified by the
screening pipeline greatly expand the compendium of known
photoanodes with band-gap energy in the 1.2
–
2.8-eV range and
demonstrate the efficacy of integrating high-throughput compu-
tational and experimental screening.
We can further evaluate the VO
4
-scaffold hypothesis by
quantifying the chemical character of the valence and conduction
bands (with respect to the O 2p
–
V 3d scaffold) for the 91
ternary vanadates from tier 2 that possess a V 3d
0
electronic
configuration. The results, shown in Fig. 3, confirm our hy-
pothesis, demonstrating that E
g
can be tuned over a broad 3-eV
range through tailoring the VO
4
band-edge character via orbitals
from an additional cation element and crystal structure. Band-
gap tuning into the desirable range can be attained via moderate
hybridization at either or both of the conduction and valence
bands edges, and photoelectrochemical activity is obtained over
a diverse set of band-character parameters and band-gap ener-
gies. We note that hybridization that enables the tuning of E
VBM
is particularly critical for optimizing photoanode efficiency.
The results of this discovery pipeline, particularly when com-
bined with existing literature, identify ternary vanadates as a
remarkable class of photoanode materials. Out of the 28 known
photoanodes from present and previous work, 22 contain V with
16 of those exhibiting the pure VO
4
motif. Fig. 3 identifies the
band-character parameters as important descriptors for under-
standing and designing improved materials in this class. Addi-
tional characterization of the Cu-based phases discovered by this
pipeline has been reported by us (32) and others (33), and we
note that the collection of discoveries motivates additional studies,
including (
i
) the exploitation of band-edge tuning to optimize
energetics, (
ii
) the characterization of photoactivity over the full
solar spectrum, and (
iii
) the engineering of electrochemical stability
through choice of operating pH an
d protective or self-passivating
layers, all of which are enabled by the range of vanadate chemistries
identified by the pipeline. The cha
racterization and optimization of
stability is paramount to establishing a deployable photoanode, an
area requiring concerted future
theory and experimental effort,
particularly due to the recently demonstrated propensity for ternary
oxides to self-passivate under operational conditions (34).
For solar hydrogen generation with a tandem light absorber,
device models (35) indicate that lowering the photoanode band
gap from the BiVO
4
gap of 2.4 eV to the 1.8-eV value observed
with several tier-7 compounds can elicit a 2- to 3-fold enhancement
in device efficiency, highlighting the opportunities provided by our
discovery of low-band-gap photoanodes. As with well-established
photoanodes such as BiVO
4
and Fe
2
O
3
, each photoanode
reported here will require extensive research and development to
optimize its solar energy conversion efficiency. This successful
Table 1. The 15 phases identified by the screening pipeline are
listed with the corresponding direct (DA) and indirect (IA) band
gap values from tier 3 and tier 6 of the pipeline
Fig. 2
label
Phase
DFT-HSE band
gap, eV
UV-vis band
gap, eV
J
O2/H2O
,
mA cm
−
2
Refs.
DA
IA
DA
IA
ACr
2
V
4
O
13
2.56 2.55 2.52 2.30 0.139
B
orth-CrVO
4
2.38 2.21 2.59 2.38 0.20
C
mon-CrVO
4
2.48 2.48 2.27
<
2.27 0.036
D
tri-FeVO
4
2.14 2.10 2.51
<
2.3
1.3
19 20
E
α
-CoV
2
O
6
2.17 2.16 2.25
<
2.25 0.015
FCo
3
V
2
O
8
2.06 2.03 2.34 2.22 0.006
GNi
2
V
2
O
7
2.75 2.72 2.73
<
2.5
0.003
HNi
3
V
2
O
8
2.55 2.54 2.66
<
2.5
0.003
I
α
-Cu
2
V
2
O
7
1.98 1.84 2.43 2.06 1.6
32
J
β
-Cu
2
V
2
O
7
1.84 1.84 2.42 2.03 2.0
32
K
γ
-Cu
3
V
2
O
8
1.89 1.73 2.40 1.80 1.8
32
LCu
11
V
6
O
26
1.41 1.38 2.49 1.87 0.95
32
M
α
-Ag
3
VO
4
2.07 1.70 2.38 2.14 0.062
18
N
β
-Ag
3
VO
4
1.91 1.61 2.51
<
2.51 0.41
O
mon-BiVO
4
2.83 2.72 2.46 2.35 0.396
15
The photocurrent density from tier 7 is also provided along with references
for the copper vanadates identified by the pipeline of Fig. 1 (32) and the three
previously reported, experimentally discovered OER photoanodes.
Fig. 3.
Tuning band gap and band alignment with hybridization. The or-
bital character at the band edges [
Top
; band-character scale is 0
–
1 and de-
termined from HSE (
α
=
0.17) calculations] for 91 VO
4
-based ternary
vanadates is shown with false-color representation of the HSE (
α
=
0.17)
band-gap energy. Experimentally verified photoanodes are circled according
to the photocurrent density measured after 44 s of illumination toggling
(385 nm) in pH 9 electrolyte (labels indicate the cation element). The efficacy
of valence band hybridization in improving VBM alignment to the OER
equilibrium energy is demonstrated for 30 of these vanadates (
Bottom
).
3042
|
www.pnas.org/cgi/doi/10.1073/pnas.1619940114
Yan et al.
implementation of high-throughput materials discovery, where
complementary computational and experimental techniques are
carefully integrated, introduces a paradigm of materials research
in which hypothesis-driven studies can be conducted with signifi-
cantly improved efficiency.
ACKNOWLEDGMENTS.
The authors thank Anubhav Jain and Joel Haber for
helpful discussions. Computational work was supported by the Materials
Project Predictive Modeling Center through the US Department of Energy
(DOE), Office of Basic Energy Sciences, Materials Sciences and Engineering
Division, under Contract DE-AC02
–
05CH11231. Experimental work was per-
formed by the Joint Center for Artificial Photosynthesis, a DOE Energy In-
novation Hub, supported through the Office of Science of the US DOE
(Award DE-SC0004993). Work at the Molecular Foundry was supported by
the Office of Science, Office of Basic Energy Sciences, of the US DOE under
Contract DE-AC02
–
05CH11231. Computational resources were also provided
by the DOE through the National Energy Supercomputing Center, a DOE
Office of Science User Facility supported by the Office of Science of the US
DOE under Contract DE-AC02-05CH11231.
1. Jain A, et al. (2013) Commentary: The Materials Project: A materials genome approach
to accelerating materials innovation.
APL Mater
1(1):011002.
2. Saal JE, Kirklin S, Aykol M, Meredig B, Wolverton C (2013) Materials design and dis-
covery with high-throughput density functional theory: The Open Quantum Materials
Database (OQMD).
JOM
65(11):1501
–
1509.
3. Curtarolo S, et al. (2012) AFLOWLIB.ORG: A distributed materials properties re-
pository from high-throughput ab initio calculations.
Comput Mater Sci
58:227
–
235.
4. Curtarolo S, et al. (2013) The high-throughput highway to computational materials
design.
Nat Mater
12(3):191
–
201.
5. Hautier G, Miglio A, Ceder G, Rignanese GM, Gonze X (2013) Identification and design
principles of low hole effective mass p-type transparent conducting oxides.
Nat
Commun
4:2292.
6. Wu YB, Lazic P, Hautier G, Persson K, Ceder G (2013) First principles high throughput
screening of oxynitrides for water-splitting photocatalysts.
Energy Environ Sci
6(1):
157
–
168.
7. Castelli IE, et al. (2012) Computational screening of perovskite metal oxides for op-
timal solar light capture.
Energy Environ Sci
5(2):5814
–
5819.
8. Jain A, Shin Y, Persson KA (2016) Computational predictions of energy materials using
density functional theory.
Nat Rev Mater
1:15004.
9. Sendek AD, et al. (2017) Holistic computational structure screening of more than 12
000 candidates for solid lithium-ion conductor materials.
Energy Environ Sci
10(1):
306
–
320.
10. Gautier R, et al. (2015) Prediction and accelerated laboratory discovery of previously
unknown 18-electron ABX compounds.
Nat Chem
7:308
–
316.
11. Perim E, et al. (2016) Spectral descriptors for bulk metallic glasses based on the
thermodynamics of competing crystalline phases.
Nat Commun
7:12315.
12. Woodhouse M, Parkinson BA (2008) Combinatorial discovery and optimization of a
complex oxide with water photoelectrolysis activity.
Chem Mater
20(7):2495
–
2502.
13. Castelli IE, et al. (2012) New cubic perovskites for one- and two-photon water splitting
using the computational materials repository.
Energy Environ Sci
5(10):9034
–
9043.
14. Xu J, et al. (2012) Synthesis and visible light photocatalytic activity of
β
-AgVO3
nanowires.
Solid State Sci
14(4):535
–
539.
15. Sayama K, et al. (2003) Photoelectrochemical decomposition of water on nano-
crystalline BiVO4 film electrodes under visible light.
Chem Commun (Camb)
23(23):
2908
–
2909.
16. Walsh A, Yan Y, Huda MN, Al-Jassim MM, Wei SH (2009) Band edge electronic
structure of BiVO4: Elucidating the role of the Bi s and V d orbitals.
Chem Mater
21:
547
–
551.
17. Yan Q, et al. (2015) Mn2V2O7: An earth abundant light absorber for solar water
splitting.
Adv Energy Mater
5(8):1401840.
18. Chemelewski WD, Mabayoje O, Mullins CB (2015) SILAR growth of Ag3VO4 and
characterization for photoelectrochemical water oxidation.
J Phys Chem C
119(48):
26803
–
26808.
19. Mandal H, et al. (2016) Development of ternary iron vanadium oxide semiconductors
for applications in photoelectrochemical water oxidation.
RSC Advances
6(6):
4992
–
4999.
20. Morton CD, Slipper IJ, Thomas MJK, Alexander BD (2010) Synthesis and character-
isation of Fe
–
V
–
O thin film photoanodes.
J Photochem Photobiol Chem
216(2
–
3):
209
–
214.
21. Seabold JA, Neale NR (2015) All 1st row transition metal oxide photoanode for water
splitting based on Cu
3
V
2
O
8
.
Chem Mater
27(3):1005
–
1013.
22. Anisimov V, Aryasetiawan F, Lichtenstein AI (1997) First-principles calculations of the
electronic structure and spectra of strongly correlated systems: The LDA
+
U method.
J Phys Condens Matter
9(4):767
–
808.
23. Kresse G, Furthmuller J (1996) Efficiency of ab-initio total energy calculations for
metals and semiconductors using a plane-wave basis set.
Comput Mater Sci
6(1):
15
–
50.
24. Onida G, Reining L, Rubio A (2002) Electronic excitations: Density-functional versus
many-body Green
’
s-function approaches.
Rev Mod Phys
74(2):601
–
659.
25. Heyd J, Scuseria GE, Ernzerhof M (2006) Hybrid functionals based on a screened
Coulomb potential.
J Chem Phys
124(21):219906.
26. Heyd J, Peralta JE, Scuseria GE, Martin RL (2005) Energy band gaps and lattice pa-
rameters evaluated with the Heyd-Scuseria-Ernzerhof screened hybrid functional.
J Chem Phys
123(17):174101.
27. Moses PG, Miao M, Yan Q, Van de Walle CG (2011) Hybrid functional investigations of
band gaps and band alignments for AlN, GaN, InN, and InGaN.
J Chem Phys
134(8):
084703.
28. Stevanovi
c V, Lany S, Ginley DS, Tumas W, Zunger A (2014) Assessing capability of
semiconductors to split water using ionization potentials and electron affinities only.
Phys Chem Chem Phys
16(8):3706
–
3714.
29. Min W, Liu Q (2011) Synthesis and photocatalytic property of Cu
3
V
2
O
8
prepared by
liquid phase precipitation.
Adv Mat Res
236-238:1675
–
1678.
30. Kalal S, Pandey A, Ameta R, Punjabi PB (2016) Heterogeneous photo-Fenton-like
catalysts Cu2V2O7 and Cr2V4O13 for an efficient removal of azo dye in water.
Cogent
Chem
2(1):1143344.
31. Wang D, Tang J, Zou Z, Ye J (2005) Photophysical and photocatalytic properties of a
new series of visible-light-driven photocatalysts M3V2O8 (M
=
Mg, Ni, Zn).
Chem
Mater
17(20):5177
–
5182.
32. Zhou L, et al. (2015) High throughput discovery of solar fuels photoanodes in the
CuO-V2O5 system.
Adv Energy Mater
5(22):1500968.
33. Guo WL, et al. (2015) Synthesis and characterization of CuV2O6 and Cu2V2O7: Two
photoanode candidates for photoelectrochemical water oxidation.
J Phys Chem C
119(49):27220
–
27227.
34. Zhou L, et al. (2016) Stability and self-passivation of copper vanadate photoanodes
under chemical, electrochemical, and photoelectrochemical operation.
Phys Chem
Chem Phys
18(14):9349
–
9352.
35. Fountaine KT, Lewerenz HJ, Atwater HA (2016
) Efficiency limits for photoelectrochemical
water-splitting.
Nat Commun
7:13706.
Yan et al.
PNAS
|
March 21, 2017
|
vol. 114
|
no. 12
|
3043
APPLIED PHYSICAL
SCIENCES