CO2 Reduction Using Water as an Electron Donor over Heterogeneous Photocatalysts Aiming at Artificial Photosynthesis.

Photocatalytic and photoelectrochemical CO2 reduction of artificial photosynthesis is a promising chemical process to solve resource, energy, and environmental problems. An advantage of artificial photosynthesis is that solar energy is converted to chemical products using abundant water as electron and proton sources. It can be operated under ambient temperature and pressure. Especially, photocatalytic CO2 reduction employing a powdered material would be a low-cost and scalable system for practical use because of simplicity of the total system and simple mass-production of a photocatalyst material.In this Account, single particulate photocatalysts, Z-scheme photocatalysts, and photoelectrodes are introduced for artificial photosynthetic CO2 reduction. It is indispensable to use water as an electron donor (i.e., reasonable O2 evolution) but not to use a sacrificial reagent of a strong electron donor, for achievement of the artificial photosynthetic CO2 reduction accompanied by ΔG > 0. Confirmations of O2 evolution, a ratio of reacted e- to h+ estimated from obtained products, a turnover number, and a carbon source of a CO2 reduction product are discussed as the key points for evaluation of photocatalytic and photoelectrochemical CO2 reduction.Various metal oxide photocatalysts with wide band gaps have been developed for water splitting under UV light irradiation. However, these bare metal oxide photocatalysts without a cocatalyst do not show high photocatalytic CO2 reduction activity in an aqueous solution. The issue comes from lack of a reaction site for CO2 reduction and competitive reaction between water and CO2 reduction. This raises a key issue to find a cocatalyst and optimize reaction conditions defining this research field. Loading a Ag cocatalyst as a CO2 reduction site and NaHCO3 addition for a smooth supply of hydrated CO2 molecules as reactant are beneficial for efficient photocatalytic CO2 reduction. Ag/BaLa4Ti4O15 and Ag/NaTaO3:Ba reduce CO2 to CO as a main reduction reaction using water as an electron donor even in just water and an aqueous NaHCO3 solution. A Rh-Ru cocatalyst on NaTaO3:Sr gives CH4 with 10% selectivity (Faradaic efficiency) based on the number of reacted electrons in the photocatalytic CO2 reduction accompanied by O2 evolution by water oxidation.Visible-light-responsive photocatalyst systems are indispensable for efficient sunlight utilization. Z-scheme systems using CuGaS2, (CuGa)1-xZn2xS2, CuGa1-xInxS2, and SrTiO3:Rh as CO2-reducing photocatalyst, BiVO4 as O2-evolving photocatalyst, and reduced graphene oxide (RGO) and Co-complex as electron mediator or without an electron mediator are active for CO2 reduction using water as an electron donor under visible light irradiation. These metal sulfide photocatalysts have the potential to take part in Z-scheme systems for artificial photosynthetic CO2 reduction, even though their ability to extract electrons from water is insufficient.A photoelectrochemical system using a photocathode is also attractive for CO2 reduction under visible light irradiation. For example, p-type CuGaS2, (CuGa)1-xZn2xS2, Cu1-xAgxGaS2, and SrTiO3:Rh function as photocathodes for CO2 reduction under visible light irradiation. Moreover, introducing a conducting polymer as a hole transporter and surface modification with Ag and ZnS improve photoelectrochemical performance.

Key References

.[1] Ag-cocatalyst for an effective active site for CO2 reduction and Ag/BaLa4Ti4O15 photocatalyst for CO2 reduction to form CO as a main reduction product using water as an electron donor even in an aqueous solution.

.[2] Ag/NaTaO3 photocatalyst doped with alkaline earth cations for CO2 reduction to CO with 90% of selectivity in an aqueous solution with a basic salt for enhancement of hydrated CO2 molecules supply.

.[3] Z-scheme system composed of CuGaS2 as a reducing photocatalyst and RGO–(CoO/BiVO4) as an O2-evolving photocatalyst for CO2 reduction to CO using water as an electron donor under visible light irradiation in an aqueous powder suspension system.

.[4] Employing (CuGa)0.5ZnS2 prepared by a flux method in Z-scheme and photoelectrochemical systems with tuning a reactant solution for efficient and stable CO2 reduction to form CO with 10–20% selectivity using water as an electron donor under visible light.

Introduction

Carbon dioxide capture storage and utilization technology (CCSU) has been encouraged, because CO2 emission control is a critical issue in the world. Ideally, CO2 fixation should be realized utilizing renewable energies, such as solar energy, as follows:

hydrogenation of CO2 using solar hydrogen

biological CO2 fixation

electrochemical CO2 reduction utilizing a photovoltaic cell

photocatalytic and photoelectrochemical CO2 reduction directly utilizing solar light

There are advantages and disadvantages to each reaction. Hydrogenation of CO2 can produce various beneficial chemical compounds with high CO2 conversion efficiency through a thermal catalytic process on an industrial scale. Much knowledge toward CO2 hydrogenation has been accumulated in C1 chemistry so far. The hydrogen should be supplied from solar hydrogen production by water splitting with no consumption of fossil resources and no CO2 emission but not from steam reforming of fossil resources. However, the CO2 conversion process requires high temperature and pressure to operate the catalytic process. Biological CO2 fixation is based on natural photosynthesis by plants. Natural photosynthesis involves almost no energy loss for absorbed photon energy conversion. However, the solar energy conversion efficiency is limited because a plant absorbs only a part of the solar spectrum as indicated by its green color. Electrochemical CO2 reduction is also interesting from the viewpoint of electrocatalysis. The reduction products and the selectivity change with electrode materials even under the same electrolysis conditions. However, electrolyzers and batteries are indispensable for the electrochemical system in addition to a photovoltaic cell. Photocatalytic and photoelectrochemical CO2 reduction utilizing solar energy in an aqueous solution is one of the ideal chemical reactions for artificial photosynthesis, because solar energy is directly converted and stored as chemical products. Artificial photosynthesis can be operated under ambient temperature and pressure to produce solar fuels and chemicals and can exceed natural photosynthesis in solar energy conversion efficiency. Especially, a powder-based photocatalyst is attractive because it can be employed for a low-cost and scalable system aimed at artificial photosynthesis.[5]

In this Account, we introduce several types of CO2 reduction systems, mainly based on particulate photocatalyst materials, using water as an electron donor. Key points for evaluation of photocatalytic and photoelectrochemical CO2 reduction are also discussed.

Overview of Photocatalytic and Photoelectrochemical CO2 Reduction Systems for Artificial Photosynthesis

Figure shows various types of photocatalytic and photoelectrochemical systems for artificial photosynthesis.[6,7] The first is a single-particulate photocatalyst system via one-photon excitation, in which photocatalytic reduction by photogenerated electrons and photocatalytic oxidation by photogenerated holes proceed on one particle (Figure a).[6−9] Photocatalysts of semiconductor materials have a band structure in which a conduction band (CB) is separated from a valence band (VB) with a band gap (BG). The thermodynamic relationship between the band structure of a photocatalyst and the redox potential for the objective reaction is important. The equilibrium potentials relative to the normal hydrogen electrode (NHE) at pH 7 and 298 K for CO2 reduction and water splitting are as follows:The conduction band minimum and valence band maximum should locate at more negative and positive levels than redox potentials of objective reactions such as water splitting and CO2 reduction, respectively. When the energy of the incident photon is larger than that of the band gap, electrons and holes are photogenerated in the conduction band and the valence band, respectively. The photogenerated electrons reduce water and CO2 to generate H2 and CO2 reduction products such as CO, while the photogenerated holes oxidize water to form O2. The O2 evolution is a key issue for photocatalytic CO2 reduction using water as an electron donor. Moreover, since CO2 reduction competes with water reduction, selective CO2 reduction is also challenging from the viewpoints of not only thermodynamics but also kinetics. Therefore, the catalytic ability of photocatalyst surface is also a key issue.

Figure 1

Artificial photosynthetic CO2 reduction based on a powdered photocatalyst by (a) a single-particulate system, (b) a Z-scheme system, (c) a photoelectrochemical system using a photocathode, and (d) a photoelectrochemical system combining a photocathode and a photoanode.

The second is a Z-scheme system via a two-photon excitation process consisting of a reducing photocatalyst, an oxidizing photocatalyst, and an electron mediator (Figure b).[10−13] This system mimics natural photosynthesis by a plant. Various photocatalysts that are active for either photocatalytic reduction of water and CO2 reduction or photocatalytic oxidation of water can be employed to make a Z-scheme system. From this viewpoint, it is meaningful to test photocatalytic CO2 reduction using sacrificial electron donors such as organic compounds and S2– in order to find potential CO2-reducing photocatalysts in a part of the Z scheme system, though the sacrificial reaction becomes a downhill reaction (ΔG < 0).

The third is a photoelectrochemical cell.[7] n-Type and p-type semiconductors may function as O2-evolving photoanodes and photocathodes to give H2 and reduction products of CO2, respectively. The photoelectrochemical cell can be constructed by combining a photoelectrode of a working electrode with a counter electrode (Figure c) or combining a photoanode and a photocathode working via two-photon excitation (Figure d). External bias can be applied between the photoanode and photocathode to enhance the photoelectrochemical reaction. However, the external bias should be smaller than the theoretical voltage of electrolysis of an objective reaction to achieve artificial photosynthesis from light energy conversion.

In the following sections, several types of the photocatalytic and photoelectrochemical systems shown in Figure are introduced.

Single Particulate Photocatalysts with Wide Band Gaps for CO2 Reduction Using Water as an Electron Donor (Figure a)

Ag Cocatalyst for CO Formation by Photocatalytic CO2 Reduction

CO2 reduction over metal oxide photocatalysts has extensively been investigated. Although TiO2 has widely been studied for photocatalytic CO2 reduction, those reports involve critical issues such as lack of quantification of O2 and small amounts of reduction products such as CH4 due to low activities. Ishitani et al. reported that CH4 could come from contaminants adsorbed on TiO2.[14] In contrast, Sayama and Arakawa have reported that a ZrO2 photocatalyst (BG = 5.0 eV) produced CO, H2, and O2 in stoichiometric amounts in an aqueous medium.[15] Moreover, loading a Cu cocatalyst and adding a bicarbonate ion enhanced the photocatalytic CO2 reduction. This is the first report to demonstrate photocatalytic CO2 reduction using water as an electron donor over a particulate photocatalyst. However, the major reduction product was H2 and the selectivity for CO formation (CO/(H2 + CO)) was about 12%. In such a background, the author found a highly active Ag cocatalyst for photocatalytic CO2 reduction to form CO with highly active photocatalysts for water splitting.

BaLa4Ti4O15 (BG = 3.9 eV) photocatalyst with a layered perovskite structure was first chosen because NiO/BaLa4Ti4O15 efficiently split water.[16] The particle is plate shaped in which an edge plane and a basal plane are reduction and oxidation site, respectively, as shown in Figure A. The separation of the reduction site from the oxidation site is beneficial for an uphill reaction, because a back reaction of a downhill reaction is suppressed. Ag was found to be a highly active cocatalyst for photocatalytic CO2 reduction to form CO as shown in Table .[1] To compare photocatalytic CO2 reduction abilities, not only the production rate [mol h–1] but also the selectivity for CO2 reduction are essential values. The selectivity is calculated according to eq .The selectivity and the production rate based on the number of reacted electrons are similar to Faradaic efficiency and partial current density, respectively, in an electrochemical reaction. It is noteworthy that CO is the main reduction product with about 70% selectivity, rather than H2, even in an aqueous medium (Table ). A small amount of HCOOH was also obtained. It is reasonable that Ag functions as an efficient cocatalyst to form CO judging from its electrocatalysis in aqueous CO2 solution.[17,18] The high conduction band level of BaLa4Ti4O15 should be important to get an enough driving force for CO2 reduction and high energy potential of photogenerated electrons applied to the Ag cocatalyst. A liquid-phase reduction method gives higher activity for CO formation than photodeposition and impregnation methods for the Ag cocatalyst loading. There is concern that the Ag cocatalyst may efficiently reduce O2 produced by water splitting. However, O2 reduction on the Ag cocatalyst is suppressed more or less, because the reaction is conducted under CO2 flow conditions smoothly removing the O2 from the reaction system.

Figure 2

(A) SEM images of Ag/BaLa4Ti4O15 before and after photocatalytic CO2 reduction, and the proposed mechanism. (B) Photocatalytic CO2 reduction using water as an electron donor under UV light irradiation over Ag(2 wt %)/BaLa4Ti4O15.[1] Photocatalyst, 0.3 g; reactant solution, water (360 mL); flow gas, CO2 (1 atm); light source, 400 W high-pressure mercury lamp; reaction cell, inner irradiation quartz cell. Reproduced with permission from ref (1). Copyright 2011 American Chemical Society.

Table 1

Effect of Cocatalyst on CO2 Reduction Using Water as an Electron Donor under UV Light Irradiation over BaLa4Ti4O15 Photocatalyst[1]a

		activity [μmol h–1]
cocatalyst (wt %)	loading method	H2	O2	CO	HCOOH	CO selectivity (%)	e–/h+
none		5.3	2.4	0	0	0	1.1
NiOx (0.5)	impregnationb	58	29	0.02	0	0.03	1.0
Ru (0.5)	photodeposition	84	41	0	0	0	1.0
Cu (0.5)	photodeposition	96	45	0.6	0	0.6	1.1
Au (0.5)	photodeposition	110	51	0	0	0	1.1
Ag (1.0)	photodeposition	10	7.0	4.3	0.3	30	1.0
Ag (1.0)	impregnation	8.2	5.7	5.2	0.2	38	1.2
Ag (1.0)	impregnation + H2 reduction	5.6	8.7	8.9	0.3	60	0.9
Ag (1.0)	liquid-phase reduction	5.6	12	19	0.4	76	1.0

Photocatalyst, 0.3 g; reactant solution, water (360 mL); flow gas, CO2 (1 atm); light source, 400 W high-pressure mercury lamp; reaction cell, inner irradiation quartz cell.

Treated with H2 reduction and subsequent oxidation.

Figure A shows SEM images of Ag-cocatalyst before and after photocatalytic CO2 reduction and a reaction mechanism. BaLa4Ti4O15 is a plate-like particle with layered perovskite structure. Ag particles of ∼10 nm diameter are loaded on both edge and basal plane by the liquid-phase reduction as prepared. After photocatalytic CO2 reduction, the number of the Ag cocatalyst particles on the edge increases while Ag particles on the basal plane disappear, because Ag on the basal plane dissolves by photooxidation and is subsequently photodeposited on the edge by photoreduction during the photocatalytic CO2 reduction.

Figure B shows time courses of CO, H2, and O2 evolution by photocatalytic CO2 reduction over Ag/BaLa4Ti4O15. The time courses demonstrate not only activity and durability but also other important points to evaluate photocatalytic CO2 reduction as discussed below.

It is important to see if O2 evolves in a stoichiometric amount when the photocatalytic reaction is conducted using water as an electron donor for light energy conversion without any strong sacrificial electron donors. CO2 is reduced by photogenerated electrons on a photocatalyst, while photocatalytic oxidation of water by photogenerated holes simultaneously proceeds as the counterpart as shown in Figures a and 2B. It is also important to see if the ratio of reacted electrons to holes estimated from products is unity according to eq .Unity means that reduction and oxidation products are obtained in a stoichiometric amount. If the e–/h+ is not unity, side reactions or noncatalytic but quantitative reactions such as reduction or oxidation of the photocatalyst itself may proceed. In addition, it is necessary to pay attention to whether some products are not detected by the measurement technique employed. The O2 evolution with at unity e–/h+ ratio is satisfied for the present photocatalytic CO2 reduction over Ag/BaLa4Ti4O15 as shown in Table and Figure B.

Photocatalytic reaction must proceed by irradiation the energy of which is larger than the band gap energy. The band gap of BaLa4Ti4O15 is 3.9 eV, which corresponds to about 320 nm light. This photocatalyst works with use of a quartz reaction cell with a suitable UV lamp, while the activity is negligible using a Pyrex reaction cell. This result indicates that the photoresponse of the BaLa4Ti4O15 photocatalyst is reasonable.

Turnover number defined by eq is also an important indicator to consider if the reaction proceeds photocatalytically.Turnover number (TON) indicates how many atoms or molecules react on one active site. TON based on the number of reacted electrons is often used for photocatalysis accompanied by redox reactions according to eq .If the TON is too small, we cannot guarantee that it is a photocatalytic reaction because not catalytic but quantitative reactions on the surface of photocatalyst cannot be excluded. In a heterogeneous photocatalyst, the molar quantity of the active site is often replaced with the molar quantity of an employed photocatalyst, because it is difficult to estimate the number of actual active sites on the surface of a photocatalyst. In some cases, the molar quantities of atoms on the surface, dopant, and cocatalyst are used for the denominator. The TON should be above unity to prove that the reaction proceeds catalytically. Photocatalytic CO2 reduction over Ag/BaLa4Ti4O15 proceeds steadily under UV light irradiation and TON to photocatalyst and cocatalyst reach 1.6 and 7.7, respectively, at 7 h being above unity as shown in Figure B.

Products of CO and HCOOH among others must originate from CO2. However, contaminants on the photocatalyst and some carbon materials constituting the photocatalyst system may become a carbon source.[14,19] Therefore, confirmation of the carbon source of the obtained products is necessary. One approach is an isotope experiment using 13CO2. Another is a control experiment using an inert gas to confirm that carbon products are not obtained. The reactant solution conditions (i.e., pH) of the control experiment should be similar to those of CO2 reduction. When 13CO2 is flowed for a photocatalytic reaction over Ag/BaLa4Ti4O15, 13CO is obtained while 12CO is not. In addition, CO is not obtained when Ar instead of CO2 is supplied. These two results prove CO2 is the carbon source.

Thus, it is concluded by confirmations of O2 evolution, the ratio of reacted e– to h+ estimated from obtained products, the TON, and the carbon source that the CO2 reduction photocatalytically proceeds using water as an electron donor over Ag/BaLa4Ti4O15.

Effect of HCO3– in Water on Photocatalytic CO2 Reduction

La or alkaline earth metal doped NaTaO3 (BG = 4.1 eV) with a perovskite structure is also a unique photocatalyst. The doped NaTaO3 has a surface nanostep structure in which a reduction site is separated from an oxidation site.[20,21] While NiO/NaTaO3 with dopant splits water efficiently but does not reduce CO2, Ag/NaTaO3:Ba gives CO with about 50% selectivity under UV light upon flowing CO2 into pure water.[2] Moreover, with addition of a basic salt into the reactant solution, CO formation rate drastically increases and the selectivity reaches about 90% even in an aqueous solution (Figure A). The enhancement of CO2 reduction with salt addition is due to efficient supply of hydrated CO2 molecule reactant and pH control.

Figure 3

(A) Photocatalytic CO2 reduction using water as an electron donor under UV light irradiation over Ag/NaTaO3:Ba. Reactant solution, NaHCO3(aq) (360 mL); flow gas, CO2 (1 atm); light source, 400 W high-pressure mercury lamp; reaction cell, an inner irradiation quartz cell. (B) Proposed mechanism of photocatalytic CO2 reduction in the presence of NaHCO3.[2] Reproduced with permission from ref (2). Copyright 2017 Wiley.

A proposed mechanism of photocatalytic CO2 reduction over Ag/NaTaO3:Ba in the presence of a basic additive is shown in Figure B. It was confirmed that not HCO3– or CO32– but a hydrated CO2 molecule is a reactant in photocatalytic CO2 reduction as in electrochemical CO2 reduction. HCO3– functions as a buffer for supply of hydrated CO2 molecules. After the CO2 adsorbs on the Ag-cocatalyst to make CO2•–(ad), CO evolves through path A (hydrogenation of CO2•–(ad)) or path B (reduction of COOH(ad)). Water is photooxidized to form O2 on the photocatalyst surface. Thus, adding a basic salt is key for efficient photocatalytic CO2 reduction with smooth supply of hydrated CO2 molecules.

Ag Cocatalyst-Loaded Photocatalysts for Single Particulate Photocatalytic CO2 Reduction to Form CO

Various metal oxide photocatalysts with different components and crystal structure have been developed for CO2 reduction based on loading Ag cocatalyst and adding NaHCO3 strategies from our group as shown in Table , for example, CaTa4O11,[22] LaTa7O19,[22] and KCaSrTa5O15[23,24] photocatalysts. In addition, many metal oxide photocatalysts with wide band gaps have been reported for CO2 reduction such as La2Ti2O7,[25] CaTiO3,[26] SrTiO3:Al,[27] Ga2O3:Zn,[28] and ZnGa2O4/Ga2O3[29] with the Ag cocatalyst from other groups. Substitution of elements is also a beneficial approach to develop new photocatalysts for CO2 reduction as well as for water splitting. For example, KCaSrTa5O15 (BG = 4.1 eV) has a tungsten bronze structure, which is similar to a defect type of perovskite structure (A1–BO3). K, Ca, and Sr at an A site in KCaSrTa5O15 can be replaced with various other cations. SrKNaTa5O15 and K2RETa5O15 (RE = rear earth metal) obtained by the substitution are also active for photocatalytic CO2 reduction.[30−32]

Table 2

Single Particulate Photocatalysts with Wide Band Gaps for CO2 Reduction Using Water as an Electron Donor[1,2,22−24]a

					activity [μmol h–1]
photocatalyst	BG [eV]	crystal structure	Ag cocatalyst (wt %, loading method)	additive	H2	O2	CO	CO selectivity (%)
CaLa4Ti4O15	3.9	layered perovskite	Ag (1.0, LPR)	none	3.2	6.6	9.3	72
SrLa4Ti4O15	3.8	layered perovskite	Ag (1.0, LPR)	none	4.8	5.8	7.1	56
BaLa4Ti4O15	3.9	layered perovskite	Ag (1.0, LPR)	none	5.6	12	19	76
K4Nb6O17	3.4	layered	Ag(3.0, LPR)	NaHCO3	11	9	8	42
NaTaO3	4.0	perovskite	Ag(1.0, PD)	none	32	16	1.4	4.2
NaTaO3:Ba	4.1	perovskite	Ag(3.0, LPR)	NaHCO3	24	76	125	84
NaTaO3:Sr	4.1	perovskite	Ag(2.0, LPR)	NaHCO3	28	102	176	86
NaTaO3:Ca	4.1	perovskite	Ag(2.0, LPR)	NaHCO3	15	84	148	91
AgTaO3	3.4	perovskite	none	NaHCO3	27	15	4.2	13
KCaSrTa5O15	4.1	tungsten bronze	Ag(0.5, Imp)	NaHCO3	15	46	97	87
K3Ta3B2O12	4.0	tungsten bronze like	Ag(2.0, PD)	NaHCO3	55	32	16.7	23
SrTa2O6	4.4	CaTa2O6	Ag(3.0, LPR)	NaHCO3	95	86	87	48
BaTa2O6	4.1	CaTa2O6 as main phase	Ag(2.0, LPR)	NaHCO3	30	16	7	19
LaTa7O19	4.1	laminate	Ag(1.0, Imp)	NaHCO3	9	17	25	74
CaTa4O11	4.5	laminate	Ag(1.0, Imp)	NaHCO3	31	30	35	53

Photocatalyst, 0.3–1.5 g; reactant solution, water (350–360 mL); flow gas, CO2 (1 atm); light source, 400 W high-pressure mercury lamp; reaction cell, inner irradiation quartz cell. PD, photodeposition; LPR, liquid-phase reduction; Imp, impregnation.

Rh–Ru Cocatalyst for CH4 Formation by Photocatalytic CO2 Reduction

Although many photocatalysts have been developed for CO2 reduction as mentioned above, obtained products are limited to two-electron reduction products such as CO and HCOOH. Therefore, it is challenging to demonstrate CO2 reduction to form CH4, an eight-electron reduction product, using water as an electron donor. Rh–Ru/NaTaO3:Sr(1%) continuously produces CH4, H2, and O2 under UV irradiation.[33] The selectivity for CH4 formation based on the number of reacted electrons is about 10%. The e–/h+ ratio estimated from obtained products is 1.1, and TON based on CH4 formation with Rh and Ru cocatalysts is 2.0. No CH4 is obtained under Ar rather than CO2 flow. These results prove that CH4 is obtained by photocatalytic CO2 reduction using water as an electron donor over the Rh–Ru/NaTaO3:Sr(1%).

Z-Scheme CO2 Reduction Using Water as an Electron Donor under Visible Light Irradiation (Figure b)

It is a key issue to construct visible light responsive CO2 reduction system using water as an electron donor for efficient sunlight utilization beyond the wide band gap photocatalysts. In this section, visible light responsive photocatalysts for CO2 reduction in the presence of a sacrificial electron donor (Table ) and application of those photocatalysts to Z-scheme systems for CO2 reduction using water as an electron donor under visible light (Table ) are introduced.

Table 3

Sacrificial CO2 Reduction Using Metal Sulfide Photocatalysts under Visible Light Irradiation[19]a

				activity [μmol h–1]
metal sulfide	crystal structure	BG, EG [eV]	electron donor	H2	CO	HCOOH
CuGaS2	chalcopyrite	2.3	K2SO3	11	0.25	trace
(AgInS2)0.22–(ZnS)1.56	wurtzite	2.3	Na2S + K2SO3	16	0.01	0
(AgInS2)0.1–(ZnS)1.8	wurtzite	2.6	Na2S	23	0.06	0.10
Ag2ZnGeS4	stannite	2.5	Na2S	38	0	0.14
ZnS:Ni(0.1%)	wurtzite + zinc blend	2.3	Na2S	22	trace	4.0
ZnS:Pb(1.0%)	wurtzite + zinc blend	2.4	Na2S	47	0.02	0.96
(ZnS)0.9–(CuCl)0.1	zinc blende	2.9	Na2S	140	0.01	0
ZnGa0.5In1.5S4	layered	2.7	Na2S	14	0.01	0

Photocatalyst, 0.2–0.3 g; reactant solution, 0.05–0.1 mol L–1 Na2S or 0.1 mol L–1 K2SO3(aq) (120–150 mL) or both; gas, CO2 (1 atm); light source, 300 W Xe lamp (λ > 420 nm); irradiation area, 33 cm2. BG = band gap, EG = energy gap.

Table 4

Z-Scheme Photocatalyst Systems for CO2 Reduction Using Water as an Electron Donor under UV or Visible Light Irradiation[3,4,19,43,46]a

					activity [μmol h–1]
entry	reducing photocatalyst	O2-evolving photocatalyst	mediator	additive (mmol L–1)	H2	O2	CO	CO selectivity (%)	e–/h+
1	CuGaS2	RGO–TiO2	RGO	none	28.8	11.2	0.15	0.5	1.29
2	CuGaS2	RGO–(CoOx/BiVO4)	RGO	none	3.1	1.3	0.04	1.3	1.21
3	Cu0.8Ag0.2GaS2	RGO–(CoOx/BiVO4)	RGO	NaHCO3 (1)	4.0	1.6	0.03	0.7	1.26
4	CuGa0.8In0.2S2	RGO–(CoOx/BiVO4)	RGO	NaHCO3 (1)	3.5	1.6	0.04	1.1	1.11
5	(CuGa)0.5ZnS2	RGO–(CoOx/BiVO4)	RGO	NaHCO3 (1)	3.5	1.9	0.4	11	1.04
6	(CuGa)0.5ZnS2	RGO–(CoOx/BiVO4)	RGO	NaHCO3 (10)	12.0	6.4	1.8	13	1.08
7	(CuGa)0.5ZnS2	RGO–(CoOx/BiVO4)	RGO	KHCO3 (10)	8.1	4.6	2.1	20	1.11
8	(CuGa)0.5ZnS2	RGO–(CoOx/BiVO4)	RGO	NaHCO3 (100)	8.9	3.5	3.2	26	1.73
9	[Ru(dpbpy)]/(CuGa)0.3Zn1.4S2	BiVO4	Co[(tpy)2]3+/2+	NaHCO3 (250)	1.7	0.8	2.7	56	3.00
10	SrTiO3:Rh	BiVO4	none	none	8.7	4.0	0.018	0.2	1.09
11	Au/SrTiO3:Rh	BiVO4	none	none	3.5	1.9	0.031	0.9	0.93

Photocatalyst, 0.1–0.4 g; reactant solution, water (120–150 mL); flow gas, CO2 (1 atm); light source, 300 W Xe lamp (λ > 300 nm for TiO2 and λ > 420 nm for BiVO4 systems); irradiation area, 33 cm2.

Visible-Light Responsive Metal Sulfide Photocatalysts for CO2 Reduction Using Sacrificial Electron Donor

Metal sulfide photocatalysts are active for not only water reduction but also CO2 reduction under visible light using a sacrificial electron donor. For example, CdS is active for sacrificial CO2 reduction to form CO in an aqueous solution containing a sacrificial reagent.[34,35] Metal sulfides with various crystal structures have also been developed for sacrificial CO2 reduction under visible light irradiation as shown in Table .[19] CuGaS2 and ZnS:Ni photocatalysts are highly active for CO and HCOOH formation, respectively. However, these CO2 reductions are not artificial photosynthesis because strong sacrificial electron donors are used. Since they cannot oxidize water into O2 because of self-photooxidation (photocorrosion), single particulate overall water splitting and CO2 reduction accompanied by O2 evolution by water oxidation as shown in Figure a is difficult. Construction of Z-scheme systems is a beneficial approach to employ metal sulfide photocatalysts showing CO2 reduction activity combined with an O2-evolving photocatalyst as shown in Figure b.

Z-Scheme System Employing RGO as a Solid-State Electron Mediator (Figure A(a))

A Z-scheme system consisting of CuGaS2 as a reducing photocatalyst, TiO2 as an O2-evolving photocatalyst, and reduced graphene oxide (RGO) as a solid-state electron mediator is active for not only water splitting[36] but also CO2 reduction to form CO (Table , entry 1).[19] The carbon source for the CO2 reduction product should carefully be checked, because RGO is a carbon material. 13CO formed under 13CO2 flow, indicating that flowed CO2 was the carbon source. However, 12CO was obtained in addition to the 13CO. Moreover, a small amount of CO formed even under Ar gas instead of CO2. So, a part of CO formed by Z-scheme CO2 reduction, whereas other CO formed by photooxidation of RGO on TiO2. The Z-scheme system works only under UV light because of limitations of TiO2. When visible light responsive RGO–(CoO/BiVO4) is employed instead of RGO–TiO2, Z-scheme CO2 reduction to form CO proceeds using water as an electron donor under visible light in an aqueous suspension (Table entry 2).[3] CO is not obtained under Ar flow in the Z-scheme system composed of RGO–(CoO/BiVO4) unlike that using RGO–TiO2. The inhibition of RGO oxidation is due to less oxidation power of holes photogenerated in the valence band of BiVO4 than that of TiO2.

Making a solid solution based on CuGaS2 with p-type character is beneficial to developing a reducing photocatalyst, because the band structure is tunable by a change in the composition of the solid solution.[6,37] For example, solid solutions of CuGaS2 with CuInS2 can absorb longer wavelengths of visible light than CuGaS2, because In 5s5p orbitals of CuInS2 lower the conduction band consisting of Ga 4s4p orbitals of CuGaS2 resulting in band gap narrowing. Red-powdered CuGa0.8In0.2S2, which absorbs visible light up to 600 nm functions as a CO2-reducing photocatalyst in the Z-scheme system (Table , entry 4). Making a (CuGa)1–Zn2S2 solid solution between CuGaS2 and ZnS improves CuGaS2 performance, though the band gap does not become narrower than that of CuGaS2.[38] The Z-scheme system using (GuGa)0.5ZnS2 prepared by a solid-state reaction (SSR) combined with RGO–(CoO/BiVO4) shows higher water splitting and CO2 reduction activities than that using CuGaS2 prepared by SSR (Figure B). When the (CuGa)0.5ZnS2 particle is prepared by a flux method, fine particles of (CuGa)0.5ZnS2 with a few hundreds of nanometers in size are obtained, while the particle size when prepared by conventional SSR is about 1 μm.[39] When the fine particulate (CuGa)0.5ZnS2 is applied to a Z-scheme system, photocatalytic water splitting and CO2 reduction are much enhanced (Figure B).[4] The Z-scheme CO2 reduction activity strongly depends on the reactant solution conditions (Table , entries 5–8). Addition of a basic salt not only stabilizes but also enhances Z-scheme CO2 reduction because of efficient supply of hydrated CO2 to the photocatalyst surface. We stress that the selectivity for CO formation in the Z-scheme CO2 reduction reaches 10–20% even using bare metal sulfide without surface modification. Although a Ag cocatalyst is effective for CO2 reduction to form CO over wide band gap metal oxides as mentioned in section , Ag on a metal sulfide does not enhance CO formation in the Z-scheme CO2 reduction at the present stage, probably due to poisoning of the Ag surface by sulfurization. Therefore, further highly selective CO2 reduction is expected by introducing a suitable active site and surface modification of the metal sulfide photocatalyst for Z-scheme CO2 reduction.

Figure 4

(A) Various types of Z-scheme photocatalysts for CO2 reduction using water as an electron donor. (B) Z-scheme CO2 reduction under visible light irradiation using CuGaS2 or (CuGa)0.5ZnS2 prepared by a SSR or a flux method combined with RGO–(CoO/BiVO4). Reproduced with permission from ref (4). Copyright 2022 American Chemical Society. (C) Z-scheme CO2 reduction under visible light irradiation using [Ru(dpbpy)]/(CuGa)0.3Zn1.4S2, BiVO4, and [Co(tpy)2]3+/2+. Reproduced with permission from ref (43). Copyright 2018 The Royal Society of Chemistry. Photocatalyst, 0.1–0.4 g; reactant solution, NaHCO3(aq) (120–150 mL); flow gas, CO2 (1 atm); light source, 300 W Xe lamp (λ > 420 nm); irradiation area, 33 cm2.

Z-Scheme System Employing a Co-Complex as an Electron Mediator (Figure A(b))

Metal complexes have been widely examined as selective CO2-reducing catalysts in electrochemistry, coordination chemistry, and photochemistry.[40,41] Recently, hybrid systems combining a metal complex catalyst with semiconductor photocatalyst materials have been studied for highly selective CO2 reduction in photoelectrochemical and photocatalytic systems.[42] For example, Z-scheme CO2 reduction under visible light has been demonstrated using [Ru(dpbpy)]-loaded (CuGa)0.3Zn1.4S2, BiVO4, and a Co-complex as an electron mediator (Figure C).[43] CO evolves as a main reduction product with introduction of the highly active Ru-complex catalyst for CO2 reduction on (CuGa)0.3Zn1.4S2. HCOOH is also produced in the reaction. The catalytic activity of a metal complex is usually inhibited in the presence of O2. Therefore, it is notable that CO2 reduction and simultaneous O2 evolution proceed even using a metal complex catalyst with a semiconductor photocatalyst in an aqueous solution, though the amount of O2 is small compared with a stoichiometric amount.

Z-Scheme System Driven by Interparticle Electron Transfer without an Electron Mediator (Figure A(c))

SrTiO3:Rh shows high sacrificial H2 evolution activity, though it does not oxidize water into O2.[44] However, SrTiO3:Rh can be employed to construct a Z-scheme system working via interparticle electron transfer with BiVO4 without an electron mediator (Figure A(c)).[45,46] The Z-scheme system reduces CO2 to CO accompanied by H2 and O2 under visible light. Loading Ag or Au cocatalyst on SrTiO3:Rh improves the CO evolution activity (Table , entries 10, 11). The suitable pH is around 4, because SrTiO3:Rh and BiVO4 particles aggregate well with each other to get good contact between the particles, resulting in smooth electron transfer from BiVO4 to SrTiO3:Rh via interparticle electron transfer. It is notable that the Z-scheme CO2 reduction proceeds using just photocatalyst powders, water, and CO2 because of self-pH-adjustment by dissolved CO2.

CO2 Reduction on p-Type Cu(I)-Containing Metal Sulfide Photocathodes under Visible Light Irradiation (Figure c,d)

A photoelectrochemical CO2 reduction system is also interesting to construct an artificial photosynthesis system. Photoelectrochemical measurement is generally conducted in a 3-electrode system or a 2-electrode system connected to a potentiostat and a power supply (Figure A). Scientifically intrinsic information on the working electrode, for example, an absolute electrode potential, is obtained with the 3-electrode system using a reference electrode. The 2-electrode system is useful for evaluation of cell performance such as open circuit voltage, short circuit current, and energy conversion efficiency. It is meaningless in a photoelectrochemical cell if an externally applied voltage is larger than the theoretical voltage of electrolysis, for example, 1.23 V for water splitting. Applying no external bias is ideal. To compare the performance of a photoelectrode, a current–potential curve is usually measured using the 3-electrode system. In addition, analysis of products by bulk electrolysis is also indispensable, as well as measurement of photocurrent to examine the Faradaic efficiency, that is, electrochemical selectivity. The Faradaic efficiency reveals if the photocurrent is due to desired redox reactions. Moreover, not only the Faradaic efficiency but also a partial photocurrent density (i.e., rate of production) are important to see how fast a certain product is formed. Incident photon to current conversion efficiency (IPCE) and solar energy conversion efficiency are also important.

Figure 5

(A) Two-electrode and three-electrode systems for photoelectrochemical CO2 reduction. (B) Photoelectrochemical CO2 reduction under visible light irradiation over a (CuGa)0.5ZnS2 powder-based photocathode. Electrolyte, 0.1 mol L–1 KHCO3(aq); flow gas, CO2 (1 atm); light source, 300 W Xe lamp (λ > 420 nm); applied bias, 0.1 V vs RHE (−0.5 V vs Ag/AgCl (pH 6.9)). Reproduced with permission from ref (4). Copyright 2022 American Chemical Society.

The photoelectrochemical cell can employ p-type semiconductors as a photocathode even photocorrosive materials. For example, visible light responsive CuGaS2,[47] (CuGa)0.5ZnS2,[4,38] Cu0.8Ag0.2GaS2,[48,49] and Cu2ZnGeS4[50] function as a CO2-reducing photocathodes. The bare (CuGa)0.5ZnS2 photocathode reduces CO2 to CO with high stability under visible light with application of an external bias (Figure B).[4] Faradaic efficiencies for CO and H2 formation are 20% and 80%, respectively, being almost 100% of total Faradaic efficiency. It is stressed that high CO formation is observed even without cocatalyst and surface modification on the photocathode.

Surface modification with CdS and ZnS of an n-type semiconductor and loading of a cocatalyst improve the performance of p-type Cu0.8Ag0.2GaS2,[49] Cu2ZnGeS4,[51] and (CuGa1–In)1–Zn2S2 solid solution[52] photocathodes. Introduction of an electrically conducting polymer such as polypyrrole (PPy) or poly(3,4-ethylenedioxythiophene) (PEDOT) as hole transporter also improves a photocathode composed of a powdered material, because electric contact between the powders and the substrate electrode such as FTO is usually poor.[53,54] PPy-modified CuGaS2 gives higher cathodic photocurrent for water and CO2 reduction than a bare CuGaS2 photocathode. Moreover, the 2-electrode system combining a PEDOT–CuGaS2 photocathode and a CoO/BiVO4 photoanode with visible light response also reduces CO2 to CO using water as an electron donor under application of a small bias and simulated sunlight irradiation.

Conclusions and Perspectives

Artificial photosynthesis is ideal green chemistry and technology to convert and store solar energy to chemical products as an uphill reaction. Solar water splitting to produce H2 is representative of artificial photosynthesis. Solar water splitting using a powder-based photocatalyst on a large scale (100 m2) has been demonstrated.[5] It will accelerate the industrial application of solar hydrogen production in the near future. In contrast to solar water splitting, artificial photosynthetic CO2 utilization using photocatalysts is still at the stage of basic research. However, recent and rapid progress of this research area is hopeful. A variety of photocatalyst and photoelectrode systems for CO2 utilization has been extensively developed using homogeneous and heterogeneous photocatalyst materials. This Account focused on photocatalytic and photoelectrochemical systems based on particulate photocatalysts for CO2 reduction as an artificial photosynthesis system working under UV and visible light.

Highly active photocatalysts for water splitting such as BaLa4Ti4O15 (BG = 3.9 eV) and doped NaTaO3 (BG = 4.1 eV) were able to be applied to CO2 reduction, because they have sufficiently high conduction bands and enough potential for water oxidation to form O2. The O2 evolution ability and a suitable cocatalyst working as a reaction center for CO2 reduction are indispensable for photocatalytic CO2 reduction using water as an electron donor. Ag and Rh–Ru cocatalysts were developed for CO and CH4 formation, respectively. Moreover, the photocatalytic activity was increased with optimization of reaction conditions such as tuning of the reactant solution. Metal sulfide photocatalysts with a high conduction band and visible light response are attractive for CO2 reduction, though they cannot oxidize water. This means that the metal sulfide photocatalyst itself cannot use water as an electron donor to achieve an uphill reaction. However, CuGaS2, (CuGa)1–Zn2S2, and CuGa1–InS2 metal sulfide materials were able to be employed as a CO2-reducing photocatalysts to make a Z-scheme photocatalyst system to achieve photocatalytic CO2 reduction using water as an electron donor under visible light irradiation. p-Type metal sulfides CuGaS2, (CuGa)1–Zn2S2, and Cu1–AgGaS2 were able to be applied to a photocathode for photoelectrochemical CO2 reduction, even if their powdered materials were employed.

Strategies to design photocatalytic and photoelectrochemical systems for CO2 reduction using water as an electron donor under visible light irradiation become clearer as mentioned above. Therefore, it is expected that more efficient photocatalyst and photoelectrode systems can be developed with further extensive study. We believe that photocatalyst and photoelectrode systems for solar CO2 utilization can be a practical use in the future as well as solar hydrogen production by water splitting.