NatRevMethodPrimers - 2023 3 42

nature reviews methods primers https://doi.org/10.
1038/s43586-023-00226-x
Primer Check for updates
Photocatalytic water splitting

Shunta Nishioka 1
, Frank E. Osterloh 2
, Xinchen Wang3, Thomas E. Mallouk4,5 & Kazuhiko Maeda 1,6
Abstract Sections
With the goal of achieving large-scale H2 production from renewable Introduction
resources, water splitting into H2 and O2 using semiconductor Experimentation

photocatalysts (sometimes called artificial photosynthesis) has Results
been studied for five decades. Unfortunately, the lack of rigour and
Applications
reproducibility in the data collection and analysis of experimental
Reproducibility and data
results has hindered progress in the field. This Primer provides a
deposition
comprehensive overview of proper characterization and evaluation
Limitations and optimizations
of photocatalysts for overall water splitting. In particular, the Primer
covers various pitfalls in photocatalysis research, best practices Outlook
for reproducibility and reliable methods for conducting rigorous

experiments. The recommendations are intended to reduce false
positives in the literature and to promote progress towards a practical
technology for producing H2 from water by using sunlight.
Department of Chemistry, School of Science, Tokyo Institute of Technology, Tokyo, Japan. 2Department of
1
Chemistry, University of California, Davis, Davis, CA, USA. 3State Key Laboratory of Photocatalysis on Energy and
Environment, College of Chemistry, Fuzhou University, Fuzhou, P. R. China. 4Department of Chemistry, University
of Pennsylvania, Philadelphia, PA, USA. 5International Center for Materials Nanoarchitectonics (WPI-MANA),
National Institute for Materials Science (NIMS), Tsukuba, Ibaraki, Japan. 6Living Systems Materialogy (LiSM)
Research Group, International Research Frontiers Initiative (IRFI), Tokyo Institute of Technology, Yokohama,
Kanagawa, Japan. e-mail: maedak@chem.titech.ac.jp
Nature Reviews Methods Primers | (2023) 3:42 1

0123456789();:
Primer
Introduction maximum of metal oxide photocatalysts is, in general, formed by the

H2 gas is a carbon-free clean energy carrier that releases a large amount oxygen 2p orbital and is usually around +3 V (versus SHE at pH 0)10. To
of energy (ΔH° = –284 kJ mol–1) when burned, with H2O as the only pro reduce protons, the CB minimum must be more negative than 0 V (ver-
duct. H2 is an important feedstock for industrially important chemicals sus SHE at pH 0), resulting in a bandgap equal to or greater than 3 eV.
such as hydrogen peroxide, methanol and ammonia. At present, com- Because the main component of the solar spectrum consists of visible
mercially available H2 is derived mainly from fossil fuels such as coal and near-infrared light, it is not possible for these wide-gap materials
(black H2) and natural gas (grey H2). During production of H2 from fossil to efficiently convert solar energy; UV/visible/near-infrared ratios of
fuels, CO2 emissions are inevitable. photon flux are calculated to be 3/56/41 from the data in ref. 11. This 3 eV
Photocatalytic water splitting, defined as the reaction that splits bandgap requirement can be circumvented in photoelectrochemical
H2O into H2 and O2 in a 2:1 ratio (Eq. 1) on a semiconductor photocatalyst, water splitting devices, which use additional electric energy input to
is a means of enabling light-to-chemical energy conversion through the drive both water oxidation and reduction reactions. In photoelectro-
production of clean, energy-dense H2 gas. The goal of photocatalytic chemical devices, separation of H2 and O2 products is easier because
water splitting research is to efficiently convert solar energy into they are produced at separate electrodes. However, photoelectro-
carbon-neutral (green) H2 on a large scale. As the solar energy that hits chemical water splitting requires a higher cost for implementation,
the Earth’s surface is four orders of magnitude larger than the energy as predicted by a techno-economic analysis by the US Department of
consumed by global human activities, the use of solar energy as the Energy in 2009 (ref. 12). According to that report, a particle suspension
energy input for photocatalytic water splitting is ideal. baggie-type reactor can produce H2 at a cost of US$1.60 per kg, which
Water splitting into H 2 and O 2 is an uphill reaction that is much lower than the cost of photoelectrochemical approaches.
involves a large positive change in the standard Gibbs free energy
(∆G° = 237 kJ mol–1): Types of water splitting photocatalysts
The simplest form of a water splitting photocatalyst is a single semi-
1 conductor particle that absorbs light and drives both the H2 and O2
H 2O → H 2 + O (1)
2 2 evolution reactions, in most cases, with the aid of an appropriate nano-
particle co-catalyst for H2 evolution. Although many photocatalysts
Its half-cell reactions are: are capable of evolving H2 or O2 in the presence of electron donors or
acceptors (sacrificial reagents), the number of visible-light photocata-
2H+ + 2e− → H2 (2) lysts that can split water into H2 and O2 without sacrificial reagents is
limited, owing to energy losses7,8.
+ 1 These energy losses are more tolerable in Z-scheme (or tandem)
H2O + 2h → O + 2H+ (3)
2 2 water splitting systems, which consist of two different photocatalysts
for H2 and O2 evolution13,14. In a Z-scheme (so named by its analogy to
The standard electrochemical potentials for water splitting, H+/H2 green plant photosynthetic systems), two separate photocatalysts
and O2/H2O, are 0 and +1.23 V (versus standard hydrogen electrode (SHE) for water reduction and oxidation are connected in series by using
at pH 0), respectively. For photocatalytic water splitting by a single light a reversible donor/acceptor mediator such as I3–/I– or Fe3+/Fe2+, or by
absorber, the conduction band (CB) minimum and the valence band (VB) direct electrical contact (Fig. 1b). In addition to suppressing the H2–O2
maximum should be positioned to straddle these redox potentials to recombination reaction, tandem photocatalysts must have good
perform water splitting, as shown in Fig. 1a. The minimum theoretical redox selectivity for the redox shuttle for water splitting to proceed.
bandgap required for water splitting by a single absorber is therefore For example, the I3– ion is more easily reduced than H2O (or H+), and
1.23 eV, which corresponds to a wavelength of approximately 1,000 nm. catalysis of that reaction by the H2 evolution photocatalyst must be sup-
In practice, however, the minimum bandgap is closer to 2.00 eV, to com- pressed. More details of Z-scheme water splitting, including historical
pensate for kinetic overpotentials, and entropy and charge separation aspects and materials development, can be found in review papers14,15.
losses1. Of course, the photocatalyst must have sufficient (photo)chemi- Reviews of other types of photocatalysts such as hetero-junctioned
cal stability under the operating conditions, and the impact of any back materials are also available16–18.
reactions such as H2–O2 recombination to form H2O should be minimized In this Primer, the details of performing photocatalytic water
during the photocatalytic water splitting process. For water splitting splitting experiments, focusing on the reaction set-up, and operating
using visible and/or near-infrared light (>400 nm) — the main compo- conditions are first introduced (Experimentation). The efficiency
nents of the solar spectrum — the photocatalysts in one-step or two-step metrics, data analysis and materials characterization are discussed
water splitting systems (Fig. 1) must have bandgaps smaller than 3 eV. in the context of the results of previous studies (Results). This Primer
Photocatalytic water splitting has been a popular research field for also deals with the possibility of performing photocatalytic water
five decades, and was triggered by the photo-assisted electrolysis of splitting on a large scale (Applications) and the factors that affect the
water using a single-crystal rutile TiO2 anode and a platinum cathode in reproducibility of photocatalytic water splitting (Reproducibility and
the presence of a chemical bias between the two electrodes2. UV light- data deposition). Finally, the main challenges and prospects for the
driven water splitting with TiO2 and SrTiO3 powdered photocatalysts field are discussed (Limitations and optimizations, Outlook).
was first reported in 1980 (refs. 3–6), and since then many metal oxide
photocatalysts that split pure water into H2 and O2 have been discov- Experimentation
ered7,8 (Fig. 2). Very recently, the quantum yield for water splitting has Materials
reached unity under UV light9. The development of visible light-responsive water splitting photocata-
Unfortunately, most metal oxide photocatalysts still require UV lysts is a momentous challenge. From the standpoint of stability, oxide-
light (shorter than 400 nm) to drive overall water splitting7,8. The VB based materials are among the most promising candidates for visible

0123456789();:
Primer
a One-step photoexcitation system b Two-step photoexcitation system
Potential Potential
(V vs SHE at pH 0) (V vs SHE at pH 0)
CB
(H+/H2) CB (H+/H2) H2
H2
0 0 CB e–
e–
H+ H+
e– D Eg
Eg (A/D)
Eg A
(O2/H2O) (O2/H2O) e–
H2O H2O
1.23 1.23 VB
O2 e– O2 e–
VB VB
Photocatalyst O2 evolution photocatalyst H2 evolution photocatalyst
Fig. 1 | Schematic energy diagrams of water-splitting photocatalysts. CB, conduction band; D, electron-donating species; Eg, bandgap energy
a,b, Photocatalytic water splitting with single-absorber (panel a) and of a semiconductor; SHE, standard hydrogen electrode; VB, valence band.
dual-absorber (panel b) photocatalysts. A, electron-accepting species;
light-responsive water splitting photocatalysts. Established strategies Single-absorber water splitting has been accomplished with some
for developing oxide-based photocatalysts for visible-light water split- mixed-anion photocatalysts, which include GaN:ZnO solid solutions32,
ting are presented, with representative examples (Fig. 3). Because the ZrO2/TaON (ref. 33) and Y2Ti2S2O5 (ref. 34).
materials aspects have been reviewed previously7,8,19–21, only the essen-
tial points are introduced in this Primer, such as the adjustment of the Spectral sensitization. Spectral sensitization is an approach in which
bandgap of metal oxides by doping with heteroatoms, the incorpora- a wide-gap semiconductor is modified with a narrow-gap substance
tion of anions and the sensitization with dyes, metal nanoparticles or (for example, an organic dye or metal) to enable the absorption of
non-oxide semiconductors. low-energy photons, followed by electron transfer to the wide-gap
For single-absorber systems (see Fig. 1), photocatalytic water split- semiconductor. HCa2Nb3O10 nanosheets sensitized by a ruthenium(II)
ting is accomplished when electrons and holes are used for water reduc- tris(diimine)-type complex show high activity for H2 evolution, and
tion and oxidation, respectively. Therefore, it is necessary to devise can be incorporated into Z-schemes for overall water splitting35. TiO2
ways to increase the efficiency of electron and hole utilization in materi- becomes an active photocatalyst for water splitting to H2 and O2 when
als, while minimizing electron–hole recombination. For example, this modified with gold nanoparticles36. In this case, the gold nanoparticles
can be done by improving the crystallinity and increasing the surface absorb visible light because of a localized surface plasmon resonance
area of a photocatalyst, although there is a trade-off between these two effect, followed by hot electron injection into the CB of TiO2. The injected
modifications. Besides, light absorbers exhibiting high photocatalytic electrons drive water reduction to form H2, and the positive holes left
activity should have high optical absorption coefficients, long charge behind on the gold nanoparticles are consumed by water oxidation.
carrier lifetimes and high charge carrier mobilities.
Other photocatalyst materials. An organic polymer semiconductor
Heteroatom doping of oxides. Doping a foreign element into a wide- photocatalyst, graphitic carbon nitride (g-C3N4)37, has been shown to
gap metal oxide can induce new visible-light active energy states in catalyse visible-light water splitting into H2 and O2 with the aid of water
the bandgap of the oxide. Dopants used for this purpose therefore reduction and oxidation co-catalysts38. Construction of a Z-scheme
provide donor or acceptor states in the bandgap of the metal oxide water splitting system with g-C3N4 is also possible39–41.
host, which must possess sufficient stability during the photochemical
process. Of course, the dopant ion must be close in size to the target Reaction set-up
ion for substitution in the host material, and, in principle, the charge Even though efficient photocatalysts for overall water splitting must be
balance of ions must be maintained in the substitution. Rhodium- able to evolve H2 /O2 in air under atmospheric pressure42, photocatalytic
doped SrTiO3 (SrTiO3:Rh)22 and rhodium and antimony co-doped water splitting tests are most often conducted under air-free condi-
SrTiO3 (ref. 23) hydrogen evolution photocatalysts are good examples tions (for example, in an argon gas atmosphere or under degassed
of this approach. The use of rutile TiO2 co-doped with tantalum and conditions). This is to enable quantification of the evolved H2 and
nitrogen24,25 or nitrogen and fluorine26 as O2-evolving photocatalysts O2 gases and to confirm the reaction stoichiometry (H2 /O2 = 2). Two
in a Z-scheme are other examples. reactor types have been established so far; one is a closed system and
the other is a flow system, in which different types of reaction cells can
Mixed anionization. Oxide-based mixed-anion compounds (includ- be used. In this section, the two kinds of systems are described, includ-
ing doped materials) such as oxynitrides, oxysulfides and oxyhalides ing reliable methods of product analysis and operating conditions. The
(such as combination of O2– and less electronegative anions) with experimental set-ups as well as the related reactors for photocatalytic
d0 or d10 electronic configurations also have strong absorption in the water splitting are illustrated in Fig. 4.
visible region, and most of these compounds are functional H2 and/or
O2-evolving photocatalysts in Z-scheme systems27–31. The visible-light Closed system. There are two types of reaction cells used in closed
response of these mixed-anion photocatalysts originates from the systems. One is a batch-type cell such as a test tube or cuvette, which is
p orbitals of the less electronegative anions, which form the states near typically sealed with a rubber septum. The other type is a custom-made
the VB maximum at potentials that are more negative than oxygen 2p. reaction cell, which is directly connected to a closed gas circulation

0123456789();:
Primer
1972: Photo-assisted electrolysis of water Fig. 2 | Key technological milestones in overall

water splitting by heterogeneous photocatalysts.
Photo-assisted water electrolysis, water splitting with
Bias H+
UV light UV light, Z-scheme water splitting under UV light,
e– Z-scheme water splitting under visible light, one-step
e– H2
water splitting under visible light, energy conversion
O2 h+
using photocatalytic sheets and water splitting
Platinum CE almost reaching unity. CE, counter electrode.
H2O
1980: Water splitting under UV light TiO2 photoelecrode
H2O
e– H2
h+
O2
H2O 1997: Z-Scheme water splitting under UV light

TiO2 or SrTiO3
λ < 280 nm
H+
Fe2+ Fe2+* H2
e– Fe3+
h +
O2
2001: Z-Scheme water splitting under visible light
H2O WO3
Visible light
I– h+ H2O
e–
e– IO –
3 H2
h+
O2 SrTiO3:Cr, Ta
WO3 2005: One-step water splitting under visible light
H2O
H+
H2
e–
h+
O2
2016: 1% of solar energy conversion by H2O

GaN:ZnO
photocatalyst sheets
H2O
e–
Gold h+
H2
e–
h+ e–
H2O
h+
SrTiO3:La,Rh 2020: Water splitting with a quantum efficiency
O2 of almost unity
BiVO4:Mo
O2 H2O
H2O H2
h+ e–
SrTiO3:Al

0123456789();:
Primer
system equipped with a vacuum pump. The reaction cells should be with a thermal conductivity detector (GC-TCD) via a gas-tight syringe.
optically transparent to the incident light and are typically made of An online GC-TCD system could also be used, to which the reaction
glass (quartz or Pyrex). system is directly connected. In any reaction system, the dead volume
In a batch-type system, a test tube or an optical cuvette is typi- of the sampling space should be as small as possible, because larger
cally used as the reaction cell. A transparent plastic thin-film baggie dead volumes will lead to larger experimental errors in the gas evolu-
(for example, a polypropylene bag) may be used as a reactor com- tion rate. For quantification of H2 and O2 gases that are simultaneously
ponent12,43,44. The reaction system (both the gas and liquid phases) is produced during photocatalytic water splitting, a molecular sieve 5A
purged with argon gas before irradiation. The use of N2 as a purging gas column of appropriate length, in which argon gas flows at a controlled
should be avoided because it becomes difficult to distinguish between rate, should be used.
the photocatalytically produced O2 and air leakage. The reaction cells Gas syringe injection is inexpensive and easy to set up, but air leak-
are irradiated by using an external light source or an all-in-one-type age is inevitable. As a result, reliable oxygen quantification is nearly
illuminator (for example, a merry-go-round system). Gas samples are impossible, especially when the amount of evolved O2 during photo-
taken by a gas-tight syringe for analysis, but O2 cannot be accurately catalytic water splitting is small. Online GC-TCD systems can minimize
determined by this way because of contamination from air during the negative impact of air leakage during gas sampling and injection,
the sampling. enabling precise oxygen quantification and multiple sampling, which
In a closed gas circulation system, the gas produced in the water is important for recording time course data for gas evolution. However,
splitting reaction is circulated through the system and sampled at it makes the reaction system complicated and expensive. In any case,
regular time intervals by a gas chromatograph directly connected a proper blank experiment is often necessary, irrespective of reaction
to the system. This closed gas circulation system allows for accurate systems.
measurement of both H2 and O2. Different types of reaction cells such Isotopically labelled water such as D2O and H218O is sometimes
as inner, side and top irradiation-type cells can be connected to the used for mechanistic investigations of photocatalytic water split-
closed gas circulation system. Water splitting experiments using this ting49–51. For such reactions, a mass spectrometer is required because
system are conducted under degassed conditions, in which saturated the detection of isotopically labelled gases is not possible with a
water vapour coexists with a small amount of a carrier gas (for exam- standard GC-TCD.
ple, argon). The low-pressure conditions enable easy detection of the Precise data acquisition also relies on calibration curves, which
evolved gases during the reaction, but they also provide a pressure bias should be taken using standard gases. Although photocatalytic water
that promotes the forward reaction42,45. splitting is usually conducted at room temperature by using a constant-
temperature bath, the temperature of the reactor under operating
Flow system. Flow systems, where the liquid phase is continuously conditions and the gas analyser will be different. In particular, the
bubbled with a carrier gas, have a great advantage for efficient water temperature of the reactor may rise during irradiation, and the tem-
splitting, owing to the suppression of the backward reaction (H2–O2 perature difference may cause an overestimation of the quantity of
recombination to water). However, flow systems can also dilute the evolved gases if the calibration curve is taken at room temperature,
products, requiring additional energy input downstream to isolate irrespective of the type of photocatalyst material.
pure H2 and O2 from the carrier gas.
In both closed and flow systems, photocatalyst particles should Operating conditions
be kept suspended by a magnetic stirrer, to optimize light absorption There are many reaction conditions, other than the photocatalyst
and to facilitate the diffusion of reactants and products for greater itself, that can influence the water splitting activity and need to be
reproducibility. However, long-term stirring (several months) will considered to ensure optimal activity. This section explains how dif-
damage the photocatalyst and lower its performance46. Photocatalyst ferent operating conditions can influence reactions and the appar-
particles may also be fixed onto a given substrate, such as an inert glass ent photocatalyst activity and describes how to safely conduct water
surface, to make a photocatalyst panel47,48, which may be advantageous splitting experiments.
in moving towards large-scale applications.
Hydrogen safety. Hydrogen/oxygen gas mixtures containing more
Product analysis than 4.0 vol% H2 or 4.8 vol% O2 are potentially explosive and pose a risk
As the products of a water splitting reaction are always in the gas phase of personal injury or even death52. To minimize this risk, photocata-
(H2 and O2), gas chromatography is the most convenient technique for lytic water splitting experiments, especially those using batch-type or
product analysis. A gas sample is injected into a gas chromatograph closed systems, should be carried out under controlled conditions and
a Heteroatom doping b Mixed anionization c Spectral sensitization Fig. 3 | Strategies for developing oxide-based
photocatalysts for visible-light water splitting.
e– a, Heteroatom doping. b, Mixed anionization.
CB CB CB
c, Spectral sensitization. CB, conduction band;
Acceptor level
VB, valence band.
Dye or nanoparticle
Donor level
VB
VB (anions’ p orbitals) VB

0123456789();:
Primer
A
Batch type system Closed gas circulation system Flow system
Gas chromatograph Sampling

Tube (using a syringe) Vacuum Sampling (connected to gas
pump (connected to gas Argon chromatograph)
Reaction chromatograph)
cell
Gas circulation
Visible pump
light
Reaction cell
Bubbling Irradiation Sampling

Reaction cell
Ba Suspension
Test tube Cuvette Top irradiation Side irradiation Inner irradiation Baggie
Lamp
Lamp
Optical filter
Bb Fixed bed
Panel
Photocatalyst sheet
Water layer
Gas (H2 + O2)

outlet
Water inlet
Fig. 4 | Reaction set-ups for photocatalytic water splitting. A, Batch type, closed gas circulation and flow reaction systems. B, Typical reaction cells for suspension
(part Ba) and fixed-bed (part Bb) reactions.
in a designated area, for example in a functioning certified chemical an evacuation/argon flush cycle; or conducting the experiment in a
fume hood with the sash closed. The use of a blast shield is recom- flow system. H2 and O2 concentrations should be monitored through
mended. Further, a laboratory worker should not work alone in the periodic measurements with the attached gas chromatograph. If the
laboratory when performing water splitting experiments. oxygen concentration reaches 4% by volume, the light source should
The risk of an explosion during the water splitting experiments be turned off and all gases should be removed from the flask using the
can be minimized by keeping the H2 and O2 concentrations in the sam- vacuum pump or by opening the flask to air.
ple gas space below 4 vol%. There are numerous ways to control the
concentration of these gases, including filling the reaction cell with Sacrificial reagents. There are many studies in the literature on photo
1.0 atm of argon gas at the beginning of the experiment; reducing the catalytic H2 or O2 evolution with sacrificial reagents53,54. Reagents
photon flux by moving the lamp away from the cell or by using a less such as methanol or triethanolamine donate electrons at favourable
intense lamp; reducing the amount of photocatalyst in the cell; reduc- electrochemical potentials, whereas sodium metaperiodate, sodium
ing the irradiation time; replacing the gas in the cell regularly by using persulfate, ferric nitrate or silver nitrate accept electrons. As a result,

0123456789();:
Primer
the photocatalytic H2 or O2 evolution process is promoted, respec- making it easy to generate more product during photocatalytic water
tively. The use of sacrificial reagents enables testing of the H2 and O2 splitting. Therefore, this cell geometry is sometimes used for the
evolution activity of a photocatalyst, but comes at the expense of the demonstration of highly efficient photocatalytic water splitting32,59.
energy conversion efficiency of the overall reaction45. Depending on However, it is often difficult to control the wavelength of light in the
the size of this electrochemical bias, these photocatalytic processes inner-irradiation cells because of the nested illumination geometry,
may actually be thermodynamically downhill and, therefore, do not although it is possible to use aqueous solutions as optical filters (for
convert any solar energy into fuels. For example, hydrogen evolution example, NaNO2 aq. for >400 nm; K2CrO4 aq. for >480 nm) by using a
(0 V versus SHE at pH 0) from a Pt/SrTiO3:Rh/[Fe(CN)6]3–/4– system double-layer inner-irradiation cell. For determining the AQY, a chemi-
(+0.36 V versus SHE at pH 0) generates only 0.36 eV of free energy per cal actinometry technique is required to measure the number of
converted photon55, whereas water oxidation with a BiVO4/sodium incident photons61.
metaperiodate (+1.63 V versus SHE at pH 0) suspension56 leads to a loss
of 0.40 eV per converted photon. The chemical bias from sacrificial Reaction pH. Photocatalytic water splitting is most often conducted in
reagents needs to be considered when assessing the energy conversion pure water. However, it may be necessary to adjust the pH of the reac-
efficiency of the photocatalytic system. Claims of energy conver- tion solution to utilize the full potential of a photocatalyst. The reaction
sion efficiency in sacrificial systems are thus rarely meaningful, and it pH will affect the chemical stability of a photocatalyst (both co-catalyst
should be noted that they also tend to involve the conversion of more and semiconductor components) as well as the efficiency of water
expensive molecules (for example, ethylenediaminetetraacetic acid reduction/oxidation on a co-catalyst62.
(EDTA) or triethanolamine) into less expensive ones (H2). However,
mechanistic studies of sacrificial systems can be instructive, and their Redox mediator. In most cases of Z-scheme water splitting, an appro-
apparent quantum yields (AQYs) can provide useful input into the priate redox mediator that consists of a reversible donor/acceptor pair
construction of non-sacrificial photosystems. In addition, confirming (such as Fe3+/Fe2+ or I3–/I–) must be dissolved in the reaction solution
the H2 or O2 evolution capability is an important step for new materials to increase the water splitting rate. The redox potentials of mediators
towards photocatalytic water splitting. and their adsorption properties on semiconductor surfaces have a
significant impact on the efficiency of Z-scheme water splitting63. More
Light wavelength and intensity. The wavelength of light is one of the details can be found in some excellent reviews14,15.
most important parameters that determine the efficiency of photo
catalytic water splitting. For photocatalyst screening purposes, it is Pressure. The pressure in a reaction set-up influences the rate of
desirable to use a light source capable of providing continuous illumi- photocatalytic water splitting64. In particular, the water splitting rate
nation at relatively high intensity. Except for measurements of the AQY is decreased if any backward reaction (for example, H2–O2 recombi-
or solar-to-hydrogen (STH) energy conversion efficiency, the incident nation to water, photoreduction of O2) occurs; H2–O2 recombination
light wavelength should be sufficiently shorter than the absorption occurs very quickly on the surface of noble metal particles (for exam-
edge of the photocatalyst. ple, platinum and rhodium). Nevertheless, they are intrinsically good
In principle, higher intensity of light will produce higher product co-catalysts for H2 evolution. The photoreduction of O2 may occur on
evolution rates, but the relationship is non-linear if catalysis or charge the surface of both the semiconductor and the co-catalyst. Therefore,
recombination is rate-limiting. High light intensity conditions may a degassed condition in the batch system or a flow system enables the
thus underestimate the efficiency of photocatalytic water splitting if evaluation of a photocatalytic water splitting reaction while minimizing
accumulated charge carriers in a semiconductor photocatalyst are not the back reaction.
fully used for the surface redox reaction, as exemplified by K4Nb6O17
and Rh–Cr mixed oxide/GaN:ZnO solid solutions49,57. Temperature. In a typical experiment, the temperature of the reaction
Furthermore, mechano-catalytic water splitting may be caused by solution is maintained near room temperature by using a temperature-
the conversion of mechanical energy supplied by rubbing the photo controllable water bath. The photovoltages of semiconductors typi-
catalyst powder against the bottom of the reaction cell by a stir bar cally decrease with increasing temperature. However, increasing the
without illumination58. NiO and Co3O4, usually used as co-catalysts reaction temperature can increase the rate of photocatalytic water
for photocatalytic water splitting, are typical mechano-catalysts for splitting, because there is an activation barrier to the surface redox
water splitting. Therefore, investigation of the operable wavelength reactions that involve charge transfer between the photocatalyst and
of light for a given photocatalyst must be conducted to preclude the water molecules49,64,65. It has also been predicted that above 353 K the
possibility of mechano-catalytic water splitting. photocatalytic activity decreases, owing to the rate-limiting adsorp-
tion of the reaction substrate, whereas below 273 K the activation
Light irradiation type. There are several choices of reaction cells and energy is ~40 kJ mol–1 (desorption of H2 is then rate-determining)
suitable light sources for each cell type59,60. An external light source and at 293–353 K is a few kilojoules per mole66. The predicted acti-
(for example, halogen lamp or LED) is common for a laboratory-scale vation energies of a few kilojoules per mole are consistent with the
water splitting experiment and can provide irradiation for any type of experimental values49.
reaction cell. The use of an external light source makes it easy to con-
trol the wavelength of the incident light and to calculate the number Results
of incident photons using a calibrated photodiode, which is essential The results of photocatalytic water splitting reactions tell us about the
in AQY calculations. efficiency of the photocatalyst and provide the basis for understanding
An inner irradiation-type reaction cell is sometimes used, the photocatalytic reaction processes. Data analysis combined with effi-
in combination with a high-pressure mercury lamp. The light intensity ciency metrics and characterization can give kinetic and mechanistic
of the mercury lamp with an inner-irradiation cell is generally high, insight, suggesting strategies for improving the activity.

0123456789();:
Primer
Efficiency metrics where A, R, Iab and Iir represent the reaction coefficient (2 for H2 and 4
Because photocatalytic activities are strongly dependent on the light for O2, see Eqs. 2 and 3), the H2 or O2 evolution rate, the rate of absorbed
wavelength and intensity, absolute efficiency metrics are very impor- photons and the rate of irradiated photons, respectively. In some cases,
tant. Two efficiency metrics are usually used. One is the quantum yield electron donors, especially organic molecules such as EDTA, supply
or AQY, and the other is STH conversion efficiency. two electrons per photon to the half reaction, because of the current
doubling effect67. In that case, the reaction coefficient, A, should be
Quantum yields. The quantum yield, especially the internal (or abso- set to half that of a one photon–one electron reaction for calculation
lute) quantum yield (IQY), expresses the ratio between the number of of the efficiency. Quantum yields are also useful for evaluating the
product molecules and the number of photons absorbed by the photo relationship between the photocatalytic activity and light absorption
catalyst (Eq. 4). Quantum yields need to be measured for a specific by the photocatalyst. This relationship can be studied by action spectra
wavelength of light at which the photon rate is uniquely determined that plot the AQY versus excitation wavelength. Theoretically, action
by the irradiance (watts per square metre) of the light source. This spectra should have the same shape as the absorption spectrum of
requires monochromatic light sources such as LEDs or band-pass filters. the photocatalyst.
Although quantum yield and IQY values are often used for homogeneous
systems, heterogeneous photocatalysts scatter light and make count- Solar to hydrogen energy conversion efficiency. STH conversion effi-
ing the number of absorbed photons difficult. Therefore, the number ciency expresses the ratio of the total power output to the input power
of photons irradiated is usually used for the calculation of quantum from sunlight (Eq. 6). The output power is calculated by the rate of gas
yields in powder suspension reaction systems (here AQY, Eq. 5): evolution in the water splitting reaction and the standard Gibbs free
energy of liquid water. The input power is given by the global standard
A×R spectrum, referred to as AM1.5G, with a total irradiance of 100 mW cm–2:
IQY = × 100 (4)
Iab
RH × ∆G ∘ 2R × ∆ G ∘
STH = × 100 = O × 100 (6)
A×R P×S P×S
AQY = × 100 (5)
Iir
where RH, Ro, ∆G°, P and S are the rates of H2 and O2 evolution
(moles per second), the standard Gibbs free energy of liquid water
Table 1 | Characterization techniques (237 × 103 J mol−1), the irradiance of simulated sunlight (100 mW cm–2)
and the irradiation area (square centimetres), respectively. For AQY
Techniques Characteristics investigated
measurements, there is no global standard spectrum, resulting in
Physical characterization certain differences in measurement conditions (for example, the full
X-ray diffraction Crystal phase, lattice constant, atomic width at half-maximum of the incident light wavelength). Therefore,
coordinate, isotropic displacement STH conversion efficiency is a more reliable metric for the evaluation
parameter, occupancy, electron of the water splitting efficiency. For systems that involve sacrificial
density
reagents, STH conversion efficiency is not a meaningful term as ∆G°
Scanning electron microscopy and Crystal phase, microstructure, space no longer refers to overall water splitting. But even in a photocatalytic
transmission electron microscopy group, morphology
system that relies on sacrificial reagents, it may be useful to perform the
Nitrogen adsorption experiment Adsorption/desorption properties, reaction under AM1.5G irradiation for more meaningful comparisons
specific surface area, pore size
of reaction rates.
distribution, type of pores
Spectroscopic characterization Characterization of reaction systems
X-ray absorption and extended X-ray Atomic and electronic local structure, Characterization of photocatalytic systems (for example, the photo-
fine-structure spectroscopy valence, bond length, inter-atomic catalyst and reaction solution) before and after the reaction can provide
distances
useful information about the reaction components. Physical, spec-
X-ray photoelectron spectroscopy Surface electronic state, valence, troscopic and (photo)electrochemical characterization techniques
and UV photoelectron spectroscopy composition, valence band position
are used to identify the initial state of the photocatalyst and possible
Infrared absorption spectrometry Local oscillation mode, chemical changes that occur during the photocatalytic reaction (Table 1).
and Raman spectroscopy component, crystal phase, surface
adsorbent
Physical characterization. The crystal structure of the photocatalyst
Inductively coupled plasma optical Chemical composition and should be determined before and after reaction by X-ray diffraction
emission spectroscopy component
measurements. However, X-ray diffraction cannot detect the forma-
Energy dispersive X-ray Chemical composition and tion of amorphous degradation products, and therefore by itself is
spectroscopy component
not a sufficient tool to rule out photocatalyst decomposition. Some
Near-infrared–UV–visible absorption/ Electron transition, bandgap, photocatalysts may decompose during a photocatalytic reaction by
diffuse-reflectance spectroscopy chemical environment self-oxidation; for example, ZnO is known to decompose by itself dur-
Emission spectroscopy Electron transition, defect structure ing the water oxidation reaction68. The morphology of a photocatalyst
(trap state) can be studied by scanning electron microscopy and/or transmis-
Surface photovoltage spectroscopy Effective bandgap, the majority carrier sion electron microscopy. Scanning electron microscopy is usually
type, depletion layer formation, defect used for observing the morphological changes in a representative
states, quasi-Fermi levels
sample of photocatalyst particles50, whereas transmission electron

0123456789();:
Primer
microscopy is more suitable for studying nano-sized particles such as splitting. A photocatalyst powder loaded onto a conductive substrate
co-catalysts69. Any beam-induced damage in the photocatalyst sam- (ideally a single crystal of the photocatalyst) can be used in (photo)
ple during the electron microscopy observation should be avoided. electrochemical measurements to reveal these properties. Owing to
The specific surface area of a photocatalyst can be measured by a the rectifying nature of the semiconductor–liquid interface, linear
nitrogen adsorption experiment at 77 K. If H2O is used as an adsorb- sweep voltammetry under (chopped) illumination enables one to
ate, the interaction of the photocatalyst with H2O molecules can be determine the direction of charge carrier flow as well as the majority
investigated70. carrier quasi-Fermi level (similar to the flat-band potential)77. Photo-
electrochemical scans78 and open circuit voltage measurements can
Spectroscopic characterization. The structural, electronic and opti- also determine the reactivity of the redox mediators with the semi-
cal properties of photocatalysts are investigated by spectroscopic conductor surface. These techniques provide insight into the charge
techniques. Information on the local structure of the photocatalyst separation and recombination properties of photocatalysts. It is noted
and its electronic state is acquired by X-ray absorption, extended X-ray that electrochemical properties of the conductive substrate should be
fine-structure spectroscopy, X-ray photoelectron spectroscopy (XPS), evaluated by a blank test because the substrate can contribute to the
infrared absorption spectrometry, Raman spectroscopy and other electrochemical behaviour79.
techniques. The combination of different spectroscopies compensates Open circuit potential measurements were recently applied to bet-
for the incomplete picture provided by any single technique and is ter understand photochemical charge separation in a tandem photo
powerful for elucidation of the structure of nano-sized co-catalysts catalyst for overall water splitting80. For example, the direct contact
and structure–activity relationships71. For example, X-ray absorption tandem (Z-scheme) photocatalyst was obtained by thermal fusion
reveals the oxidation state of the constituent elements in the bulk of of Ru/SrTiO3:Rh H2 evolution and BiVO4 O2 evolution photocatalyst
the photocatalyst particles, whereas XPS is surface-sensitive. X-ray particles73,80–82 (Fig. 5a). When an aqueous suspension of the catalyst
powder diffraction identifies the crystalline phases present, whereas is illuminated by a 400 nm LED, stoichiometric H2 and O2 evolution
extended X-ray fine-structure spectroscopy gives information about takes place. Interestingly, the AQY for overall water splitting varies
local nearest-neighbour inter-atomic distances. strongly with illumination intensity, as shown in Fig. 5b. At low intensi-
To reveal the chemical composition of a photocatalyst, inductively ties (8–14 mW cm–2), no water splitting takes place at all and the tandem
coupled plasma optical emission spectroscopy (or its equivalent) and appears non-functional. H2 evolution is observed at a light intensity
energy dispersive X-ray spectroscopy, usually associated with scanning of 17.7 mW cm–2 or higher. At first, the AQY values increase sharply
electron microscopy/transmission electron microscopy equipment, with light intensity, but above 171 mW cm–2 the AQY stays nearly con-
are employed. XPS can show not only the surface electronic states but stant. This shows that there are three activity regimes for the tandem
also surface atomic ratios, although the effective sampling depth of XPS photocatalyst. Open circuit potential measurements performed on
analysis is element-specific and in the nanometre regime. FTO-deposited films of the separate Ru/SrTiO3:Rh and BiVO4 compo-
Optical properties can be studied by UV–visible–near-infrared nents can explain this trend. These measurements reveal a similar light
absorption/diffuse-reflectance spectroscopy and by emission spec- intensity dependence of photocathodic behaviour of Ru/SrTiO3:Rh
troscopy. The change in steady-state emission can be used to compare and photoanodic behaviour of BiVO4 (ref. 80). The measured intensity-
the differences in activity of samples. The emission is derived from the dependent quasi-Fermi levels make it possible to derive the model
radiative recombination process of photogenerated charge carriers, shown in Fig. 5a for the tandem photocatalyst. Upon initial contact,
but whether the carriers that cause the emission are really involved in the two light absorbers form a weak p/n junction (Fig. 5a, left; see
the redox reaction on the photocatalyst surface must be ascertained by band bending at the SrTiO3:Rh/BiVO4 interface). Under mild excita-
other means. Time-correlated measurement of optical spectra is one tion, this junction promotes recombination of the minority carriers
of the most powerful techniques for observing charge carrier dynam- at the solid–solid interface, which leads to energy loss. This process is
ics in a photocatalytic reaction. In these measurements, the time- the reason for the threshold light intensity for water splitting. At higher
resolved changes in spectral features are recorded after excitation by light intensity (>17.7 mW cm−2), charge separation in each absorber is
a pulsed laser, enabling a kinetic analysis of charge transfer processes increasingly controlled by the junctions at the solid–liquid interfaces.
and catalytic redox reactions. In surface photovoltage spectroscopy, These junctions attract the minority charge carriers and allow the
a vibrating Kelvin probe electrode measures the contact potential dif- majority charge carriers to move to the solid–solid interface, where
ference change of an illuminated photocatalyst film sample to provide they can recombine in Z-scheme fashion. This equilibrates the majority
information about the effective bandgap, the majority carrier type, carrier Fermi levels across the tandem and, in turn, increases the band
depletion layer formation, defect states and quasi-Fermi levels72–74. bending at the two solid–liquid junctions (Fig. 5a, right). The photo
When correlating the results of these spectroscopic techniques with voltages of both light absorbers now add to produce a sufficient poten-
photocatalytic activity, it is always important to take into account the tial for overall water splitting. These processes are the reason for the
differences between the experimental environment for spectroscopic strongly increasing AQY in the 17.7–171 mW cm−2 intensity regime.
measurements and the photocatalytic reaction conditions. In situ or At higher light intensities, when the AQY is constant, the solid–liquid
operando spectroscopic techniques can visualize dynamic changes in junctions are no longer limiting the overall water splitting reaction.
the physicochemical properties of photocatalysts in their the working Instead, the activity of the tandem is limited by the electronic quality
state75,76. of the absorbers and by the slow kinetics of the water oxidation reac-
tion at BiVO4. Additionally, a current mismatch between the weakly
(Photo)Electrochemical characterization. The electrochemical light-absorbing Ru/SrTiO3:Rh nanocrystals and the much larger BiVO4
properties of photocatalysts, including their Fermi level, charge car- crystals may be to blame. Higher efficiency is possible, theoretically,
rier density, open circuit potential, flat-band potential and electri- by optimizing the Ru/SrTiO3:Rh/BiVO4 ratio and by adding a water
cal resistance, directly affect the efficiency of photocatalytic water oxidation co-catalyst.

0123456789();:
Primer
a b 1.2
SrTiO3:Rh SrTiO3:Rh
BiVO4
1.0
e– e–
Ruthenium BiVO4
Ruthenium 0.6
H2O/H2 0.8
EF,n
e– H2O/H2
(0.0 V)
AQY (%)
0.4
(0.0 V)
AQY (%)
EF e–
O2/H2O EF,p 0.6
O2/H2O 0.2
(1.23 V) EF,n
h+ (1.23 V) h +
EF,p
8 mW cm–2
0.4
0.0
0 5 10 15 20 25 30 35 40
h+
0.2 Light power (mW cm–2)
h+ 0 500 1,000 1,500 2,000
Light power (mW cm–2)
Fig. 5 | Mechanistic study of the light intensity dependence of a SrTiO3:Rh/BiVO4 b, Apparent quantum yield (AQY) at 400 nm as a function of irradiance. EF,n, quasi-
tandem photocatalyst. a, Charge transfer pathways for the SrTiO3:Rh/BiVO4 Fermi level of electron; EF,p, quasi-Fermi level of hole. Adapted with permission
tandem photocatalyst under low (left) and high (right) illumination intensity. from ref. 80. Copyright 2020 American Chemical Society.
Applications seawater splitting by a rhodium/chromium mixed oxide/GaN:ZnO

Large-scale photocatalytic water splitting photocatalyst, the low activity has been attributed to a mismatch
Photocatalyst panels/sheets in which photocatalyst powders are between the optimal operational pH and the oxidation of Cl– ions that
fixed on a solid substrate have been developed48,83. Laboratory-scale compete with water oxidation and induce undesirable side reactions.
experiments on photocatalytic water splitting systems are typically Although complete suppression of the undesirable Cl– oxidation may
done using a photocatalyst powder suspended in water to enable mass be very difficult, the production of value-added oxidation products
transfer of ions and soluble reactants and products. However, suspen- such as ClO– and Cl2 from Cl– oxidation during photocatalytic seawater
sion systems have several complications including dispersion (that splitting has been proposed87.
requires additional mechanical energy for stirring) and separation/
collection of the used (and sometimes deactivated) photocatalyst. Other potential applications
Approaches using photocatalyst panels/sheets may circumvent these One of the general characteristics of oxynitride-type photocatalysts
issues, making practical operation possible. More effective tracking is their low activity for water reduction to H2, even in the presence of
of sunlight is also possible by using photocatalyst panels/sheets in sacrificial electron donors89. This appears to be a critical bottleneck
all-day operation. A techno-economic analysis has suggested that a for achieving efficient water splitting with oxynitrides. Nevertheless,
sun-tracking concentrator array is more cost-effective than a fixed-bed the low activity for water reduction may be useful for applications
panel array12,43. A photocatalyst sheet consisting of Ru/SrTiO3:La,Rh, in CO2 photoreduction, in combination with metal complex (photo)
BiVO4:Mo and a conductive layer (such as gold or carbon) is capable catalysts90,91. In fact, CO2 to HCOOH conversion with ~80% selectiv-
of carrying out overall water splitting by a Z-scheme mechanism, with ity is possible using GaN:ZnO photocatalysts further modified with
a maximum STH conversion efficiency of 1.1% (at 331 K) at reduced gold nanoparticles and a binuclear ruthenium(II) supramolecular
pressures84. A prototype of a large-scale overall water splitting system photocatalyst under visible light92.
with a 100 m2 area has been developed using SrTiO3:Al modified with a
H2-evolving rhodium/chromium co-catalyst and an O2-evolving cobalt Reproducibility and data deposition
co-catalyst83. In this system, safe operation of overall water splitting The reproducibility of photocatalytic water splitting results depends
with ~0.76% STH conversion efficiency for 1,600 h and product gas col- strongly on photocatalyst preparation. At a minimum, this involves
lection/separation have been demonstrated. Using a similarly modified the synthesis of a semiconductor powder and the loading of a nano-
SrTiO3:Al photocatalyst, an STH value of 0.11% has been obtained with particulate co-catalyst. For semiconductor materials synthesized
a single-bed particle suspension ‘baggie’ reactor44. through multi-step procedures (for example, mixed-anion compounds
or composites of two different materials), care should be taken during
Seawater splitting the preparation. Even well-known metal oxide photocatalysts such as
Seawater, which covers approximately 70% of the Earth’s surface area, SrTiO3 may face a reproducibility issue, primarily because the activity
is the largest water resource on the planet. This makes it an attractive depends strongly on various factors, some of which are, unfortunately,
source of H2O for photocatalytic water splitting. There have been some difficult to analyse quantitatively. For example, standard characteri-
reports of seawater splitting by photocatalysts that are active for overall zation techniques such as X-ray diffraction, UV–visible absorption
water splitting into H2 and O2 (refs. 85–88), although the performance spectroscopy, XPS and surface area measurements often cannot
in seawater splitting was lower than that in pure water. For artificial distinguish between high and low-activity materials. This is because

0123456789();:
Primer
structural defects and/or vacancies existing in a photocatalyst have The photocatalyst’s stability should be tested to exclude the pos-
strong impacts on their activity. Non-stoichiometric TiO2 (refs. 93,94) sibility of non-catalytic production of H2/O2. This should be done by
and SrTiO3 (refs. 95,96) whose oxygen defect densities were carefully performing the photocatalytic measurement for a time sufficient to
controlled have been studied to understand the impact of defect den- achieve a turnover number greater than unity; the number of formed
sity on the photocatalytic reaction. The density of defects and their moles of H2 and O2 must exceed the number of moles of the catalyst.
levels in TiO2 have been evaluated experimentally97. However, quanti- If the turnover number is less than unity, a stoichiometric reaction can-
fication of defects in a photocatalyst and elucidation of their roles in not be ruled out. Also, a decay of H2/O2 rates over time usually indicates
photocatalytic water splitting remain a challenge. photocatalyst degradation. Some of the most stable photocatalysts
To minimize the reproducibility issue related to photocatalyst (RhCrOx/GaN:ZnO (ref. 46), CoOx/Pt/g-C3N4 (ref. 38) and CoOOH/
materials, the purity and origin of raw materials/reagents and the pro- RhCrOx/SrTiO3:Al (ref. 100)) can be operated for >500 h without decay
cedure should be stated clearly. Structural characterization of the syn- in the rates of H2/O2 evolution. The time-dependent degradation of
thesized photocatalyst should be used to elucidate the phase purity, the reaction rate is influenced by the light intensity; a higher density
chemical composition (both bulk and surface), particle morphology, of incident photon flux will result in more rapid degradation of the
specific surface area, light absorption properties and other parameters. water splitting rate so stability tests should be conducted in shorter
In heterogeneous (photo)catalysis, the reaction rate can be time periods. For unstable photocatalysts, the reason for deactivation
expressed in units of product formed per unit time (for example, moles should be identified by performing post-characterization analyses of
per hour). But to allow comparison with catalysts prepared in different the used photocatalyst and the reaction solution.
laboratories, absolute efficiency measures such as the AQY and/or STH
energy conversion efficiency should be given. This considers the fact Data deposition and reporting
that the rate of any photochemical reaction is governed by the number The details of general information and indispensable data of photo
of photons absorbed by the photocatalyst. catalytic reactions are presented in Table 2. A description of the
Great care is needed in reporting a reproducible AQY because the physicochemical properties of a photocatalyst and the experimen-
AQY can have a strong dependence on the light intensity used during tal details of activity testing are both necessary for reproducing the
the measurement57,80. As such, AQYs may contain some fluctuations. photocatalytic activity.
In this regard, reporting STH values in addition to the AQY is recom-
mended. Nevertheless, photocatalytic activity may differ with respect
to the irradiation geometry, even with the use of the same photocatalyst Glossary
material59,60.
Some studies have reported photocatalytic activity in units of
moles per hour per gram. However, this is not appropriate, because Back reactions Merry-go-round system
the rate of gas generation is not proportional to the mass of the photo Reactions that can lower the efficiency Samples placed equidistant from the
catalyst but, instead, depends only on the rate at which photons are of photocatalytic water splitting light source are rotated around the light
absorbed. The incident irradiance (milliwatts per square centimetre) (for example, H2 + ½O2 → H2O). source, and each sample is illuminated
and the illuminated cross-sectional area should also be given. with light of the same intensity.
For highest accuracy, measurements should be repeated three Black H2
or more times on different batches of a photocatalyst sample and the H2 that is produced from bituminous Sacrificial reagents
average activities should be reported together with the standard devia- coal via steam reforming, which Fast electron donors or acceptors that
tion. Reporting numerical values beyond the precision of experiments produces high quantities of CO2. may be irreversibly decomposed after
should be avoided. reaction and that modify the energetics
In photochemical reaction studies, it is useful to investigate the Conduction band (CB) and rate of the photocatalytic reaction.
dependence of the reaction rate on the wavelength of incident light. To minimum
do this, reporting AQYs as a function of the incident light wavelength in The energy region in a semiconductor Standard hydrogen electrode
the form of a photoaction spectrum is the best way (see also ‘Efficiency that corresponds to empty antibonding (SHE). The hydrogen electrode when
metrics’). This can be done using a monochromator or using LEDs. The orbitals and that conducts electrons the activities of hydrogen gas and
use of a cut-off (long-pass) filter may also be useful to determine when present from excitation. proton are both 1.
the longest wavelength that still drives the water splitting reaction.
Accurate detection and quantification of all products (including O2) Foreign element Valence band (VB) maximum
during water splitting is needed. Because both H2 and O2 are supposed An impurity added to semiconductors The energy region in a semiconductor
to be generated at a 2:1 ratio during the full water splitting reaction, the to control their characteristics that corresponds to filled bonding
confirmation that a stoichiometric amount of O2 is produced is very (for example, a metal cation that orbitals and that conducts holes when
important. Photocatalysts that contain a (photo)chemically unstable can contribute d electrons, or an present from excitation.
component should be evaluated with special care. For example, metal anion that is less electronegative
sulfides have inherent instability towards oxidation reactions98. Photo- than oxygen). Z-Scheme
oxidative dissolution of ZnO has also been reported68. So, performing A two-step photochemical reaction
an 18O-enriched H2O splitting experiment combined with mass spec- Grey H2 taking place in plants to utilize lower
troscopic analysis of the gaseous products is recommended for the H2 that is produced from natural gas energy photons and generate sufficient
evaluation of activity of such light-sensitive materials (for example, or methane via steam reforming, potential to drive photosynthetic
sulfide-related materials) for water splitting51 and for half-cell water which produces lower quantities reactions, The electron transfer pathway
oxidation in the presence of a sacrificial acceptor50,99. of CO2. resembles the letter Z.

0123456789();:
Primer
Limitations and optimizations efficiencies in suspension systems at room temperature (1.1% in the
Overall water splitting with an AQY of >95% at wavelengths between 350 optimized photocatalyst panel configuration at 331 K (ref. 84)). Higher
and 360 nm has recently been demonstrated using a CoOOH/RhCrOx/ efficiencies should be possible with more active proton reduction
SrTiO3:Al photocatalyst9. It has been also shown that scalable operation photocatalysts.
of photocatalytic water splitting with a 100 m2 area and safe separa-
tion of the produced H2/O2 mixture is feasible83. However, STH energy Undesirable electron transfer processes
conversion efficiencies achieved by these studies remain near 0.76%, To realize an efficient Z-scheme, any back electron transfer reaction
which is an order of magnitude lower than that required for practical needs to be suppressed. Possible back electron transfer reactions are
applications (10%). Higher theoretical STH efficiency may be possible those involved with an electron mediator. In a two-absorber system,
by using photocatalysts with smaller bandgaps and better sunlight each photocatalyst must oxidize or reduce the shuttle redox species
absorption. It has been recently demonstrated that water splitting for water splitting to proceed. In some cases, selective adsorption of
into H2 and O2 is possible using Y2Ti2S2O5 (refs. 34,101) and BaTaO2N redox species on a solid surface enables high selectivity of such forward
(ref. 102) as single absorbers, whose absorption edges are at ~650 nm. reactions. In an IO3–/I– redox system, for example, rutile TiO2 adsorbs
In the two-absorber systems (such as the Z-scheme), the current bot- IO3– anions on the surface with high selectivity and this enables efficient
tlenecks for water splitting are the lack of an efficient water reduction Z-scheme water splitting with Pt/anatase TiO2 (refs. 63,105,106).
photocatalyst and the undesirable electron transfer process. The best Electron (and hole) transfer to trap sites originating typically from
known H2-evolving catalysts SrTiO3:Rh (ref. 103), ZrO2/TaON (ref. 89) defects in semiconductors is another undesirable electron transfer
and dye-sensitized oxide nanosheets104, with proper modifications path in most of the semiconductor photocatalysts, not only for two-
aimed to control the electron transfer direction, give 0.1–0.2% STH absorber systems but also for single-absorber systems. Electron trans-
fer to defects becomes dominant, especially under low light intensity,
yielding lower efficiency80. This limitation may be addressed by using
Table 2 | Reporting standards for photocatalytic water a dye-sensitized photocatalyst because the system does not rely on
splitting the photoexcitation of the semiconductor component. Of course, an
undesirable back electron transfer reaction at the dye/semiconductor
Data type Details required interface occurs, and is characteristic of dye-sensitized photocatalyst
General information systems. This back reaction can, however, be suppressed by insertion
of an insulating blocking layer between them. For example, Al2O3 works
Specification Purity (including the detail of impurities) and
of materials commercial sources of available materials/ as a blocking layer for back electron transfer from the semiconductor
reagents used for photocatalyst preparation to the dye, improving the H2 evolution in a half-cell reaction107,108 and
and activity testing in Z-scheme water splitting35. In addition to this, active control of the
Details of photocatalyst Structural or chemical information to identify the selectivity of the redox species has been accomplished by surface
characterization phase purity, specific surface area, particle size modification of a dye-sensitized photocatalyst in an I3–/I– redox system.
and light absorption capability of a photocatalyst The back electron transfer to the oxidized redox mediator is a problem
Reporting of water splitting reaction conditions even in a system that allows for selective adsorption of a reactant. In the
Reaction set-up Type of reactors, light sources (intensity, case of dye-sensitized niobate photocatalysts, the use of the sacrificial
wavelength), cross-sectional area of the electron donor EDTA increased the H2 evolution rate by a factor of ~100,
illuminated reactor relative to the I– system in a half-cell reaction108. In other words, the
Operating conditions Temperature, pressure, carrier gas back reaction decreases the H2 evolution rate to one hundredth in this
system. The back electron transfer to I3– can be suppressed by modi-
Contents for activity test Masses (grams) of photocatalyst (and co-catalyst),
of additives (for example, redox mediator) and fication of the dye-sensitized photocatalyst with an anionic polymer
their concentrations, solution pH (polystyrene sulfonate (PSS)). Because of electrostatic repulsion of
Reporting of photocatalyst performance the I3– anion, this modification leads to an increased H2 evolution rate
in the half-cell reaction109.
Reaction rate Amount (moles) of H2 and O2 recorded by a gas
analyser (for example, gas chromatograph) in a
Dual modification of the dye-sensitized niobate photocatalyst
given time interval with Al2O3 and PSS achieved a 100-fold improvement in the STH of
the dye-sensitized Z-scheme water splitting system under simulated
AQY AQY at a given wavelength (nanometres) based on
the rate of formed products (H2 and O2) compared sunlight, relative to the unmodified photocatalyst104. The details of
with the flux of incident photons the reaction kinetics were studied by transient absorption spectros-
STH conversion STH free energy conversion efficiency measured copy measurements with a diffuse-reflectance apparatus104,109. The
efficiency under standard AM1.5G irradiation (100 mW cm–2) back reaction (I3– + 2e– → 3I–, which presumably proceeds through a
Stability Activity (product rate) versus time plot and
singly reduced, short-lived I3• intermediate) can be monitored by the
turnover number greater than unity to confirm attenuation of the absorbance change of I3–. Transient absorption
a photocatalytic process spectroscopy measurement in an aqueous NaI solution revealed that
Wavelength-dependent Reaction rate (or AQY) as a function of the incident Al2O3 modification promotes the I– oxidation reaction and PSS modifi-
activity/photoaction light wavelength cation suppresses back electron transfer to I3–. Further investigations
spectrum by laser flash photolysis revealed that the back reaction of electron
Isotope tracer Amount of formed labelled product transfer to I3– can be suppressed only at low concentrations of I3–. For
experiment further improvements of this system, the limitations of suppression
AQY, apparent quantum yield; STH, solar to hydrogen. of the back reaction to low concentrations of I3– must be addressed.

0123456789();:
Primer
Outlook In summary, although further research is needed to progress

Photocatalytic water splitting, producing H2 and O2, by sunlight would towards highly efficient and scalable production of H2 through photo-
be a solution to mitigate the energy and environmental problems on the catalytic water splitting, recent achievements made in the past ~10 years
Earth. Although there are several competing H2 production technolo- have reinvigorated those working in the field. The significant progress
gies already established in industry (such as steam reforming), large- made in recent years can be attributed, in part, to an interplay between
scale H2 production from renewable resources (water and sunlight) is scientists of different disciplines. Such interdisciplinary collaboration
very attractive. The tentative goal of research on photocatalytic water can reinforce the photocatalytic water splitting community, and pave
splitting is to develop a photocatalyst that exhibits a STH efficiency of the way for practical applications of photocatalytic water splitting.
5–10% with a lifetime of a few years.
A photocatalyst panel made from CrOx/Ru/SrTiO3:La,Rh and Published online: xx xx xxxx
BiVO4:Mo, working according to the Z-scheme principle, has recently
References
been reported to show STH efficiencies greater than 1%. It has also been 1. Bolton, J. R., Strickler, S. J. & Connolly, J. S. Limiting and realizable efficiencies of solar
reported that modified SrTiO3 achieved an AQY of >95% at wavelengths photolysis of water. Nature 316, 495–500 (1985).
To our knowledge, this is the first paper to quantify the theoretical efficiency limits
between 350 and 360 nm in overall water splitting. Since the first report
of solar water splitting.
of photocatalytic water splitting using SrTiO3 in 1980, it has taken 2. Fujishima, A. & Honda, K. Electrochemical photolysis of water at a semiconductor
about 40 years to achieve photocatalytic water splitting with a ~100% electrode. Nature 238, 37–38 (1972).
This pioneering work discusses photo-assisted electrolysis of water.
quantum yield. However, the STH efficiency of the SrTiO3 photocatalyst
3. Schrauzer, G. N. & Guth, T. D. Photolysis of water and photoreduction of nitrogen
is limited to <0.7% despite its near-unity quantum yield because of its on titanium dioxide. J. Am. Chem. Soc. 99, 7189–7193 (1977).
bandgap that only absorbs photons in the UV light region (<380 nm). 4. Domen, K., Naito, S., Soma, M., Onishi, T. & Tamaru, K. Photocatalytic decomposition
of water vapour on an NiO–SrTiO3 catalyst. J. Chem. Soc. Chem. Commun. https://
For both the one-step and the two-step water splitting systems,
doi.org/10.1039/C39800000543 (1980).
the development of a photocatalyst capable of absorbing longer wave- 5. Lehn, J. M., Sauvage, J. P. & Ziessel, R. Photochemical water splitting continuous
length photons and providing a high quantum yield is essential for generation of hydrogen and oxygen by irradiation of aqueous suspensions of metal
loaded strontium titanate. Nouv. J. Chim. 4, 623–627 (1980).
further improvement of the STH efficiency. Although some mixed-
6. Sato, S. & White, J. M. Photodecomposition of water over Pt/TiO2 catalysts. Chem. Phys.
anion photocatalysts possessing ~650 nm absorption edges such as Lett. 72, 83–86 (1980).
Y2Ti2S2O5 and BaTaO2N:Mg have recently achieved pure water splitting 7. Kudo, A. & Miseki, Y. Heterogeneous photocatalyst materials for water splitting.
Chem. Soc. Rev. 38, 253–278 (2009).
with proper co-catalyst modification, the AQYs of these photocatalysts 8. Wang, Q. & Domen, K. Particulate photocatalysts for light-driven water splitting:
still remain <1%, far below that required to satisfy 5–10% STH efficien- mechanisms, challenges, and design strategies. Chem. Rev. 120, 919–985 (2020).
cies. Considering the result of highly efficient water splitting by the 9. Takata, T. et al. Photocatalytic water splitting with a quantum efficiency of almost unity.
Nature 581, 411–414 (2020).
modified SrTiO3 photocatalyst, controlling the density of defects This contribution achieves a milestone in photocatalytic water splitting with a
and the morphology of the photocatalyst to maximize the photogen- quantum yield of almost unity.
erated charge carrier transport through the material appears to be 10. Scaife, D. E. Oxide semiconductors in photoelectrochemical conversion of solar energy.
Sol. Energy 25, 41–54 (1980).
a straightforward means of improving the photocatalytic activity. 11. Gueymard, C. SMARTS2, A Simple Model of the Atmospheric Radiative Transfer of
To do this, searching for a new preparation/modification method of Sunshine: Algorithms and Performance Assessment (Florida Solar Energy Center/Univ.
photocatalysts is important. Unlike metal oxide photocatalysts, for of Central Florida, 1995).
12. James, B. D., Baum, G. N., Perez, J. & Baum, K. N. Technoeconomic Analysis of
which synthetic methods and methods to control carrier and particle Photoelectrochemical (PEC) Hydrogen Production (Directed Technologies, 2009).
morphology are well established, mixed-anion photocatalysts that 13. Abe, R. Development of a new system for photocatalytic water splitting into H2 and O2
can absorb long-wavelength visible light still have many challenges in under visible light irradiation. Bull. Chem. Soc. Jpn. 84, 1000–1030 (2011).
14. Maeda, K. Z-Scheme water splitting using two different semiconductor photocatalysts.
terms of synthesis. Of course, exploring a new material and its photo ACS Catal. 3, 1486–1503 (2013).
catalytic functionality is important. These will be priorities for the next 15. Wang, Y. et al. Mimicking natural photosynthesis: solar to renewable H2 fuel synthesis
5–10 years of research development in this field. by Z-scheme water splitting systems. Chem. Rev. 118, 5201–5241 (2018).
16. Wang, H. et al. Semiconductor heterojunction photocatalysts: design, construction,
From the standpoint of reproducibility and optimization, there are and photocatalytic performances. Chem. Soc. Rev. 43, 5234–5244 (2014).
several ongoing debates in the water splitting research community. One 17. Moniz, S. J. A., Shevlin, S. A., Martin, D. J., Guo, Z.-X. & Tang, J. Visible-light driven
of the biggest problems in reproducing and optimizing photocatalytic heterojunction photocatalysts for water splitting—a critical review. Energy Environ. Sci.
8, 731–759 (2015).
performance is the photocatalyst material itself. This is because there 18. Xu, Q., Zhang, L., Cheng, B., Fan, J. & Yu, J. S-Scheme heterojunction photocatalyst.
is not sufficient technology to control and quantify the lattice defects Chem 6, 1543–1559 (2020).
19. Maeda, K. & Domen, K. Photocatalytic water splitting: recent progress and future
in a photocatalyst, which greatly affect photocatalytic performance.
challenges. J. Phys. Chem. Lett. 1, 2655–2661 (2010).
It is also desirable to standardize the method for measuring the photo 20. Maeda, K. Photocatalytic water splitting using semiconductor particles: history and
catalytic water splitting rate, as the efficiency may vary with respect to recent developments. J. Photochem. Photobiol. C. 12, 237–268 (2011).
21. Hisatomi, T., Kubota, J. & Domen, K. Recent advances in semiconductors for
the irradiation geometry.
photocatalytic and photoelectrochemical water splitting. Chem. Soc. Rev. 43,
Development of in situ or operando spectroscopic techniques 7520–7535 (2014).
is important to obtain information about the physicochemical state 22. Konta, R., Ishii, T., Kato, H. & Kudo, A. Photocatalytic activities of noble metal ion doped
SrTiO3 under visible light irradiation. J. Phys. Chem. B 108, 8992–8995 (2004).
of photocatalysts during photocatalytic water splitting. In powder
23. Asai, R. et al. A visible light responsive rhodium and antimony-codoped SrTiO3 powdered
photocatalysts, the oxidation and reduction reactions occur on the photocatalyst loaded with an IrO2 cocatalyst for solar water splitting. Chem. Commun.
same particle, so the information obtained from such measurements 50, 2543–2546 (2014).
24. Nakada, A. et al. Solar-driven Z-scheme water splitting using tantalum/nitrogen
includes information on both oxidation and reduction reactions. In
co-doped rutile titania nanorod as an oxygen evolution photocatalyst. J. Mater. Chem. A
this regard, it would be reasonable to construct a model photocatalyst 5, 11710–11719 (2017).
(typically, in the form of thin films) by focusing on either the reduction 25. Nishioka, S. et al. Enhanced water splitting through two-step photoexcitation by sunlight
using tantalum/nitrogen-codoped rutile titania as a water oxidation photocatalyst.
or oxidation reaction, and to measure it under conditions identical to Sustain. Energy Fuels 3, 2337–2346 (2019).
those of the photocatalytic water splitting55,76,80,104.

0123456789();:
Primer
26. Miyoshi, A. et al. Nitrogen/fluorine-codoped rutile titania as a stable oxygen-evolution 55. Wang, J., Zhao, J. & Osterloh, F. E. Photochemical charge transfer observed in nanoscale
photocatalyst for solar-driven Z-scheme water splitting. Sustain. Energy Fuels 2, hydrogen evolving photocatalysts using surface photovoltage spectroscopy. Energy
2025–2035 (2018). Environ. Sci. 8, 2970–2976 (2015).
27. Maeda, K. & Domen, K. New non-oxide photocatalysts designed for overall water splitting 56. Wang, J. & Osterloh, F. E. Limiting factors for photochemical charge separation in BiVO4/
under visible light. J. Phys. Chem. C 111, 7851–7861 (2007). Co3O4, a highly active photocatalyst for water oxidation in sunlight. J. Mater. Chem. A 2,
28. Kageyama, H. et al. Expanding frontiers in materials chemistry and physics with multiple 9405–9411 (2014).
anions. Nat. Commun. 9, 772 (2018). 57. Tabata, S., Ohnishi, H., Yagasaki, M., Ippommatsu, M. & Domen, K. Light-intensity
29. Maeda, K. et al. Recent progress on mixed-anion materials for energy applications. dependence in photocatalytic decomposition of water over K4Nb6O17 catalyst.
Bull. Chem. Soc. Jpn 95, 26–37 (2022). Catal. Lett. 28, 417–422 (1994).
30. Miyoshi, A. & Maeda, K. Recent progress in mixed‐anion materials for solar fuel 58. Ikeda, S. et al. Mechano-catalysis — a novel method for overall water splitting. Phys. Chem.
production. Sol. RRL 5, 521 (2020). Chem. Phys. 1, 4485–4491 (1999).
31. Maeda, K. (Oxy)Nitrides with d0-electronic configuration as photocatalysts and 59. Kato, H. & Kudo, A. Water splitting into H2 and O2 on alkali tantalate photocatalysts ATaO3
photoanodes that operate under a wide range of visible light for overall water splitting. (A = Li, Na, and K). J. Phys. Chem. B 105, 4285–4292 (2001).
Phys. Chem. Chem. Phys. 15, 10537–10548 (2013). 60. Schwarze, M. et al. Quantification of photocatalytic hydrogen evolution. Phys. Chem.
32. Maeda, K. et al. Photocatalyst releasing hydrogen from water. Nature 440, 295 (2006). Chem. Phys. 15, 3466–3472 (2013).
33. Maeda, K., Lu, D. & Domen, K. Direct water splitting into hydrogen and oxygen under 61. Serpone, N. Relative photonic efficiencies and quantum yields in heterogeneous
visible light by using modified TaON photocatalysts with d0 electronic configuration. photocatalysis. J. Photochem. Photobiol. A 104, 1–12 (1997).
Chem. Eur. J. 19, 4986–4991 (2013). 62. Maeda, K. et al. Efficient overall water splitting under visible-light irradiation on (Ga1-xZnx)
34. Wang, Q. et al. Oxysulfide photocatalyst for visible-light-driven overall water splitting. (N1-xOx) dispersed with Rh–Cr mixed-oxide nanoparticles: effect of reaction conditions
Nat. Mater. 18, 827–832 (2019). on photocatalytic activity. J. Phys. Chem. B 110, 13107–13112 (2006).
35. Oshima, T. et al. An artificial Z-scheme constructed from dye-sensitized metal oxide 63. Abe, R., Sayama, K. & Sugihara, H. Development of new photocatalytic water splitting
nanosheets for visible light-driven overall water splitting. J. Am. Chem. Soc. 142, into H2 and O2 using two different semiconductor photocatalysts and a shuttle redox
8412–8420 (2020). mediator IO3−/I−. J. Phys. Chem. B 109, 16052–16061 (2005).
36. Tanaka, A., Teramura, K., Hosokawa, S., Kominami, H. & Tanaka, T. Visible light-induced 64. Wang, Q. et al. Particulate photocatalyst sheets based on carbon conductor layer for
water splitting in an aqueous suspension of a plasmonic Au/TiO2 photocatalyst with efficient Z-scheme pure-water splitting at ambient pressure. J. Am. Chem. Soc. 139,
metal co-catalysts. Chem. Sci. 8, 2574–2580 (2017). 1675–1683 (2017).
37. Wang, X. et al. A metal-free polymeric photocatalyst for hydrogen production from water 65. Hisatomi, T., Minegishi, T. & Domen, K. Kinetic assessment and numerical modeling of
under visible light. Nat. Mater. 8, 76–80 (2009). photocatalytic water splitting toward efficient solar hydrogen production. Bull. Chem.
To our knowledge, this work is the first report of visible-light H2/O2 evolution using Soc. Jpn. 85, 647–655 (2012).
an organic semiconductor. 66. Herrmann, J.-M. Heterogeneous photocatalysis: fundamentals and applications to the
38. Zhang, G., Lan, Z. A., Lin, L., Lin, S. & Wang, X. Overall water splitting by Pt/g-C3N4 removal of various types of aqueous pollutants. Catal. Today 53, 115–129 (1999).
photocatalysts without using sacrificial agents. Chem. Sci. 7, 3062–3066 67. Maeda, K. et al. Photocatalytic hydrogen evolution from hexaniobate nanoscrolls and
(2016). calcium niobate nanosheets sensitized by ruthenium(II) bipyridyl complexes. J. Phys.
39. Nishioka, S., Shibata, K., Miseki, Y., Sayama, K. & Maeda, K. Visible-light-driven Chem. C. 113, 7962–7969 (2009).
nonsacrificial hydrogen evolution by modified carbon nitride photocatalysts. 68. Doménech, J. & Prieto, A. Stability of ZnO particles in aqueous suspensions under
Chin. J. Catal. 43, 2316–2320 (2022). UV illumination. J. Phys. Chem. 90, 1123–1126 (1986).
40. Zhao, G., Huang, X., Fina, F., Zhang, G. & Irvine, J. T. S. Facile structure design based on 69. Oshima, T., Lu, D., Ishitani, O. & Maeda, K. Intercalation of highly dispersed metal
C3N4 for mediator-free Z-scheme water splitting under visible light. Catal. Sci. Technol. 5, nanoclusters into a layered metal oxide for photocatalytic overall water splitting.
3416–3422 (2015). Angew. Chem. Int. Ed. 54, 2698–2702 (2015).
41. Martin, D. J., Reardon, P. J., Moniz, S. J. & Tang, J. Visible light-driven pure water splitting 70. Oshima, T., Yokoi, T., Eguchi, M. & Maeda, K. Synthesis and photocatalytic activity of
by a nature-inspired organic semiconductor-based system. J. Am. Chem. Soc. 136, K2CaNaNb3O10, a new Ruddlesden–Popper phase layered perovskite. Dalton Trans. 46,
12568–12571 (2014). 10594–10601 (2017).
42. Fabian, D. M. et al. Particle suspension reactors and materials for solar-driven water 71. Maeda, K. et al. Characterization of Rh–Cr mixed-oxide nanoparticles dispersed on
splitting. Energy Environ. Sci. 8, 2825–2850 (2015). (Ga1-xZnx)(N1-xOx) as a cocatalyst for visible-light-driven overall water splitting. J. Phys.
43. Pinaud, B. A. et al. Technical and economic feasibility of centralized facilities for solar Chem. B 110, 13753–13758 (2006).
hydrogen production via photocatalysis and photoelectrochemistry. Energy Environ. Sci. 72. Kronik, L. & Shapira, Y. Surface photovoltage spectroscopy of semiconductor structures:
6, 1983–2002 (2013). at the crossroads of physics, chemistry and electrical engineering. Surf. Interface Anal.
44. Zhao, Z. et al. Electronic structure basis for enhanced overall water splitting 31, 954–965 (2001).
photocatalysis with aluminum doped SrTiO3 in natural sunlight. Energy Environ. Sci. 12, 73. Melo, M. A. Jr et al. Surface photovoltage measurements on a particle tandem
1385–1395 (2019). photocatalyst for overall water splitting. Nano Lett. 18, 805–810 (2018).
This work explains how aliovalent dopants remove lifetime killers in SrTiO3. 74. Cheng, Y. et al. Effect of charge selective contacts on the quasi Fermi level splitting
45. Osterloh, F. E. Photocatalysis versus photosynthesis: a sensitivity analysis of devices of CuGa3Se5 thin film photocathodes for hydrogen evolution and methylviologen
for solar energy conversion and chemical transformations. ACS Energy Lett. 2, 445–453 reduction. EES Catal 1, 74–83 (2023).
(2017). 75. Ma, L., Liu, M., Jing, D. & Guo, L. Photocatalytic hydrogen production over CdS: effects
This perspective explains how the thermodynamics of the photocatalytic reaction of reaction atmosphere studied by in situ Raman spectroscopy. J. Mater. Chem. A 3,
dictate the photocatalyst design. 5701–5707 (2015).
46. Ohno, T., Bai, L., Hisatomi, T., Maeda, K. & Domen, K. Photocatalytic water splitting 76. Mu, C. et al. In situ characterization techniques applied in photocatalysis: a review.
using modified GaN:ZnO solid solution under visible light: long-time operation and Adv. Mater. Interfaces 10, 2201842 (2022).
regeneration of activity. J. Am. Chem. Soc. 134, 8254–8259 (2012). 77. Nakato, Y., Ueda, T., Egi, Y. & Tsubomura, H. Decomposition potentials of crystalline
47. Goto, Y. et al. A particulate photocatalyst water-splitting panel for large-scale solar silicon as related to the photocurrent stability of p–n junction silicon semiconductor
hydrogen generation. Joule 2, 509–520 (2018). electrodes. J. Electrochem. Soc. 134, 353–358 (1987).
48. Schröder, M. et al. Hydrogen evolution reaction in a large-scale reactor using a carbon 78. Maeda, K., Abe, R. & Domen, K. Role and function of ruthenium species as promoters
nitride photocatalyst under natural sunlight irradiation. Energy Technol. 3, 1014–1017 with TaON-based photocatalysts for oxygen evolution in two-step water splitting under
(2015). visible light. J. Phys. Chem. C 115, 3057–3064 (2011).
49. Hisatomi, T., Maeda, K., Takanabe, K., Kubota, J. & Domen, K. Aspects of the water 79. Xu, P., Milstein, T. J. & Mallouk, T. E. Flat-band potentials of molecularly thin metal oxide
splitting mechanism on (Ga1-xZnx)(N1-xOx) photocatalyst modified with Rh2-yCryO3 nanosheets. ACS Appl. Mater. Interfaces 8, 11539–11547 (2016).
cocatalyst. J. Phys. Chem. C 113, 21458–21466 (2009). 80. Han, R., Melo, M. A., Zhao, Z., Wu, Z. & Osterloh, F. E. Light intensity dependence of
50. Nishioka, S. et al. A zinc-based oxysulfide photocatalyst SrZn2S2O capable of reducing photochemical charge separation in the BiVO4/Ru–SrTiO3:Rh direct contact tandem
and oxidizing water. Dalton Trans. 48, 15778–15781 (2019). photocatalyst for overall water splitting. J. Phys. Chem. C. 124, 9724–9733 (2020).
51. Maeda, K. et al. GaN:ZnO solid solution as a photocatalyst for visible-light-driven overall 81. Jia, Q., Iwase, A. & Kudo, A. BiVO4–Ru/SrTiO3:Rh composite Z-scheme photocatalyst
water splitting. J. Am. Chem. Soc. 127, 8286–8287 (2005). for solar water splitting. Chem. Sci. 5, 1513–1519 (2014).
To our knowledge, this work is the first report of reproducible water splitting into 82. Sasaki, Y., Nemoto, H., Saito, K. & Kudo, A. Solar water splitting using powdered
H2 and O2 under visible light. photocatalysts driven by Z-schematic interparticle electron transfer without an electron
52. Schröder, V., Emonts, B., Janßen, H. & Schulze, H. P. Explosion limits of hydrogen/oxygen mediator. J. Phys. Chem. C 113, 17536–17542 (2009).
mixtures at initial pressures up to 200 bar. Chem. Eng. Technol. 27, 847–851 (2004). 83. Nishiyama, H. et al. Photocatalytic solar hydrogen production from water on a 100-m2
53. Zhou, P., Yu, J. & Jaroniec, M. All-solid-state Z-scheme photocatalytic systems. Adv. scale. Nature 598, 304–307 (2021).
Mater. 26, 4920–4935 (2014). 84. Wang, Q. et al. Scalable water splitting on particulate photocatalyst sheets with a solar-
54. Schneider, J. & Bahnemann, D. W. Undesired role of sacrificial reagents in photocatalysis. to-hydrogen energy conversion efficiency exceeding 1%. Nat. Mater. 15, 611–615 (2016).
J. Phys. Chem. Lett. 4, 3479–3483 (2013). This work achieves >1% STH conversion efficiency through photocatalytic water splitting.

0123456789();:
Primer
85. Ji, S. M. et al. Photocatalytic hydrogen production from natural seawater. J. Photochem. 105. Abe, R., Sayama, K., Domen, K. & Arakawa, H. A new type of water splitting system
Photobiol. A 189, 141–144 (2007). composed of two dierent TiO2 photocatalysts (anatase, rutile) and a IO3–/I– shuttle redox
86. Maeda, K., Masuda, H. & Domen, K. Effect of electrolyte addition on activity of (Ga1−xZnx) mediator. Chem. Phys. Lett. 344, 339–344 (2001).
(N1−xOx) photocatalyst for overall water splitting under visible light. Catal. Today 147, 106. Abe, R., Sayama, K. & Arakawa, H. Significant effect of iodide addition on water splitting
173–178 (2009). into H2 and O2 over Pt-loaded TiO2 photocatalyst: suppression of backward reaction.
87. Sayama, K. Production of high-value-added chemicals on oxide semiconductor Chem. Phys. Lett. 371, 360–364 (2003).
photoanodes under visible light for solar chemical-conversion processes. ACS Energy 107. Kim, W., Tachikawa, T., Majima, T. & Choi, W. Photocatalysis of dye-sensitized TiO2
Lett. 3, 1093–1101 (2018). nanoparticles with thin overcoat of Al2O3: enhanced activity for H2 production and
88. Guan, X. et al. Efficient unassisted overall photocatalytic seawater splitting on GaN-based dechlorination of CCl4. J. Phys. Chem. C. 113, 10603–10609 (2009).
nanowire arrays. J. Phys. Chem. C. 122, 13797–13802 (2018). 108. Nishioka, S. et al. Excited carrier dynamics in a dye-sensitized niobate nanosheet
89. Maeda, K. & Domen, K. Development of novel photocatalyst and cocatalyst materials photocatalyst for visible-light hydrogen evolution. ACS Catal. 11, 659–669 (2021).
for water splitting under visible light. Bull. Chem. Soc. Jpn. 89, 627–648 (2016). 109. Saupe, G. B., Mallouk, T. E., Kim, W. & Schmehl, R. H. Visible light photolysis of hydrogen
90. Maeda, K. Metal-complex/semiconductor hybrid photocatalysts and photoelectrodes iodide using sensitized layered metal oxide semiconductors: the role of surface chemical
for CO2 reduction driven by visible light. Adv. Mater. 31, e1808205 (2019). modification in controlling back electron transfer reactions. J. Phys. Chem. B 101,
91. Nakada, A., Kumagai, H., Robert, M., Ishitani, O. & Maeda, K. Molecule/semiconductor 2508–2513 (1997).
hybrid materials for visible-light CO2 reduction: design principles and interfacial
engineering. Acc. Chem. Res. 2, 458–470 (2021). Acknowledgements
92. Nakada, A. et al. Effects of interfacial electron transfer in metal complex–semiconductor S.N. acknowledges support by a Grant-in-Aid for Research Activity Start-up (JP21K20555).
hybrid photocatalysts on Z-scheme CO2 reduction under visible light. ACS Catal. 8, F.E.O. acknowledges support from the US Department of Energy, Office of Science, Office of
9744–9754 (2018). Basic Energy Sciences (Award Number DOE-SC0015329). X.W. acknowledges support from the
93. Bak, T., Li, W., Nowotny, J., Atanacio, A. J. & Davis, J. Photocatalytic properties of TiO2: National Natural Science Foundation of China (22032002 and U1905214). T.E.M. acknowledges
evidence of the key role of surface active sites in water oxidation. J. Phys. Chem. A 119, support from the Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences,
9465–9473 (2015). and Energy Biosciences, Department of Energy (contract DE-SC0019781). K.M. acknowledges
94. Nowotny, J. et al. Defect chemistry and defect engineering of TiO2-based financial support by a Grant-in-Aid for Scientific Research (B) (JP22H01862), a Grant-in-Aid
semiconductors for solar energy conversion. Chem. Soc. Rev. 44, 8424–8442 (2015). for Transformative Research Areas (A) “Supra-ceramics” (JP22H05148) and a Core-to-Core
95. Nishioka, S. et al. Homogeneous electron doping into nonstoichiometric strontium Program (JPJSCCA20200004) (JSPS).
titanate improves its photocatalytic activity for hydrogen and oxygen evolution.
ACS Catal. 8, 7190–7200 (2018). Author contributions
96. Vequizo, J. J. M. et al. Crucial impact of reduction on the photocarrier dynamics of SrTiO3 Introduction (S.N., F.E.O., X.W., T.E.M. and K.M.); Experimentation (S.N., F.E.O., X.W., T.E.M.
powders studied by transient absorption spectroscopy. J. Mater. Chem. A 7, 26139–26146 and K.M.); Results (S.N., F.E.O., X.W., T.E.M. and K.M.); Applications (S.N., F.E.O., X.W., T.E.M. and
(2019). K.M.); Reproducibility and data deposition (S.N., F.E.O., X.W., T.E.M. and K.M.); Limitations
97. Ohtani, B. & Takashima, M. Happy photocatalysts and unhappy photocatalysts: electron and optimizations (S.N., F.E.O., X.W., T.E.M. and K.M.); Outlook (S.N., F.E.O., X.W., T.E.M. and
trap-distribution analysis for metal oxide-sample identification. Catal. Sci. Technol. 12, K.M.); Overview of the Primer (S.N., F.E.O., X.W., T.E.M. and K.M.).
354–359 (2022).
98. Ellis, A. B., Kaiser, S. W., Bolts, J. M. & Wrighton, M. S. Study of n-type semiconducting Competing interests
cadmium chalcogenide-based photoelectrochemical cells employing polychalcogenide The authors declare no competing interests.
electrolytes. J. Am. Chem. Soc. 99, 2839–2848 (1977).
99. Ishikawa, A. et al. Oxysulfide Sm2Ti2S2O5 as a stable photocatalyst for water oxidation and Additional information
reduction under visible light irradiation (λ ≦ 650 nm). J. Am. Chem. Soc. 124, 13547–13553 Peer review information Nature Reviews Methods Primers thanks Yingtang Zhou, Ling Piao,
(2002). Kamalakannan Kailasam and the other, anonymous, reviewer(s) for their contribution to the
100. Lyu, H. et al. An Al-doped SrTiO3 photocatalyst maintaining sunlight-driven overall water peer review of this work.
splitting activity for over 1000 h of constant illumination. Chem. Sci. 10, 3196–3201
(2019). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
101. Nandal, V. et al. Unveiling charge dynamics of visible light absorbing oxysulfide published maps and institutional affiliations.
for efficient overall water splitting. Nat. Commun. 12, 7055 (2021).
102. Li, H. et al. One-step excitation overall water splitting over a modified Mg-doped Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
BaTaO2N photocatalyst. ACS Catal. 12, 10179–10185 (2022). article under a publishing agreement with the author(s) or other rightsholder(s); author self-
103. Kato, H., Sasaki, Y., Shirakura, N. & Kudo, A. Synthesis of highly active rhodium-doped archiving of the accepted manuscript version of this article is solely governed by the terms
SrTiO3 powders in Z-scheme systems for visible-light-driven photocatalytic overall water of such publishing agreement and applicable law.
splitting. J. Mater. Chem. A 1, 12327–12333 (2013).
104. Nishioka, S. et al. Surface-modified, dye-sensitized niobate nanosheets enabling an
efficient solar-driven Z-scheme for overall water splitting. Sci. Adv. 8, eadc9115 (2022). © Springer Nature Limited 2023

0123456789();:

NatRevMethodPrimers - 2023 3 42

Uploaded by

Copyright:

Available Formats

NatRevMethodPrimers - 2023 3 42

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NatRevMethodPrimers - 2023 3 42

Uploaded by

Copyright:

Available Formats

nature reviews methods primers https://doi.org/10.

Primer Check for updates

Photocatalytic water splitting

With the goal of achieving large-scale H2 production from renewable Introduction

resources, water splitting into H2 and O2 using semiconductor Experimentation

for reproducibility and reliable methods for conducting rigorous

Nature Reviews Methods Primers | (2023) 3:42 1

Introduction maximum of metal oxide photocatalysts is, in general, formed by the

Nature Reviews Methods Primers | (2023) 3:42 2

a One-step photoexcitation system b Two-step photoexcitation system

Photocatalyst O2 evolution photocatalyst H2 evolution photocatalyst

Nature Reviews Methods Primers | (2023) 3:42 3

1972: Photo-assisted electrolysis of water Fig. 2 | Key technological milestones in overall

H2O 1997: Z-Scheme water splitting under UV light

2016: 1% of solar energy conversion by H2O

Nature Reviews Methods Primers | (2023) 3:42 4

Nature Reviews Methods Primers | (2023) 3:42 5

Gas chromatograph Sampling

Bubbling Irradiation Sampling

Gas (H2 + O2)

Nature Reviews Methods Primers | (2023) 3:42 6

Nature Reviews Methods Primers | (2023) 3:42 7

Nature Reviews Methods Primers | (2023) 3:42 8

Nature Reviews Methods Primers | (2023) 3:42 9

h+ 0 500 1,000 1,500 2,000

Light power (mW cm–2)

Applications seawater splitting by a rhodium/chromium mixed oxide/GaN:ZnO

Nature Reviews Methods Primers | (2023) 3:42 10

Nature Reviews Methods Primers | (2023) 3:42 11

Nature Reviews Methods Primers | (2023) 3:42 12

Outlook In summary, although further research is needed to progress

Nature Reviews Methods Primers | (2023) 3:42 13

Nature Reviews Methods Primers | (2023) 3:42 14

Nature Reviews Methods Primers | (2023) 3:42 15

You might also like