The massive emission of greenhouse gas carbon dioxide (CO2) has attracted great attention due to its impact on global warming. Researchers have been working on this project for a long time and found that photocatalytic CO2 reduction has shown great potential in developing cost-effective clean energy resources. However, the efficiency of CO2 photo-reduction is low because of limited light absorption efficiency, undesired charge recombination, and high CO2 activation barrier in thermodynamics and kinetics. In this Perspective, we concentrate on recent advanced strategies to improve CO2 photo-reduction and illustrate the mechanism of CO2 activation and we intend to find the most plausible strategy on solving the problems listed. The mainstream approaches for boosting CO2 photo-reduction efficiency lie in (1) tuning the bandgap of the photocatalysts by incorporating heteroatoms in a photosensitizer causing enhanced light absorption; (2) constructing heterojunctions resulting in effective charge separation; and (3) introducing surface defects, basic sites, and functional groups, as well as increasing the surface area of catalysts contributing to enhanced CO2 adsorption and activation. Moreover, this Perspective will conclude with brief perspectives and recommendations regarding the promising research of converting CO2 into valuable fuels.

CO2 emission has been rapidly growing, which caused the greenhouse effect and global warming nowadays. Solar energy as a clean source has enormous potential to convert CO2 into fuels and value-added chemicals. The products of CO2 conversion such as carbon monoxide (CO), methane (CH4), methanol (CH3OH), formic acid (HCOOH), and other hydrocarbon compounds have been intensively investigated in recent years. Furthermore, various systems with low energy consumption, such as photo-thermal conversion and photo-electrochemical transformation, are established to drive CO2 reduction.1–5 

As shown in Fig. 1, a consensus has been reached that there are three main steps in photocatalytic CO2 reduction, including light absorption, photoinduced charge separation, and transfer, as well as surface reactions.6,7 First, when the energy of incident light is over the bandgap of the semiconductor, electron–hole pairs would be generated. Second, the produced charge would be effectively separated and transferred from the bulk phase of the materials to the surface. Then, the electrons are captured by active sites on the surface where their adjacent absorbed CO2 molecule would be reduced if the conduction band (CB) potential of the semiconductor is more negative than the normal hydrogen electrode (NHE) of CO2. Moreover, the holes can be utilized to oxidize H2O when the valance band of the semiconductor is more positive than the NHE of H2O or consumed by the electron donor of materials. Finally, the formed carbon-based species desorb from the surface.

FIG. 1.

Schematic presentation of CO2 photo-reduction on a semiconductor photo-catalyst combined with co-catalysts for solar-to-fuel conversion. Reproduced with permission from Ran et al., Adv. Mater. 30, 1704649 (2018). Copyright 2018, Wiley-VCH.

FIG. 1.

Schematic presentation of CO2 photo-reduction on a semiconductor photo-catalyst combined with co-catalysts for solar-to-fuel conversion. Reproduced with permission from Ran et al., Adv. Mater. 30, 1704649 (2018). Copyright 2018, Wiley-VCH.

Close modal

In the past few decades, remarkable progress to boost CO2 solar-to-fuel efficiency has been achieved. For example, some novel visible-light semiconductors are utilized to improve light conversion efficiency such as TiO2, C3N48,9 of carbon materials, BiVO4,10,11 Cu2O,12,13 InTaO4,14 CuFeO2,15,16 CdS quantum dots (QDs),17 and carbon quantum dots.18 The bandgap is adjusted and optimized to improve light absorption through ion doping,19 co-catalyst incorporation, and surface plasmon improvement.20,21 Notably, the nanomaterials of the photo-catalysts, including nanorods, nanowires,22 nanotubes,23 and nanobelts,24 are extensively applied in improving CO2 adsorption and activation due to their abundant exposed area and increased efficient sites. Additionally, photo-catalysts possessing ultrathin structures can promote photoinduced charge transfer from the bulk of the photosensitizer to the surface, reducing the possibilities of recombination. The ultrathin structure is also beneficial to the migration of photoinduced electrons from the surface to the exposed active sites, which not only connects with the CO2 molecule but also contributes to CO2 activation and photoconversion.25 Moreover, constructing different junctions on the enhancement of charge separation and migration has been widely utilized in CO2 photo-reduction, such as semiconductor–semiconductor heterojunctions and semiconductor–metal junctions.26 Benefiting from these advantages, the CO2 conversion efficiency and product selectivity would be improved. Unfortunately, CO2 reduction usually involves the hydrogen evolution reaction (HER) at the emergence of water (H2O) vapor, which consumes the active photoelectrons and causes poor efficiency and low selectivity.

The CO2 with a highly symmetrical linear structure is an inert molecule. The dissociation energy of the C=O bond is 750 kJ mol−1.27 Besides, a high Gibbs free energy is required in an endothermic reaction of converting CO2 into CH3OH and CH4, which makes CO2 photo-reduction in the presence of H2O difficult.7 Possible reactions and predicted products with the corresponding standard redox potentials according to acquired thermodynamic data are shown in Table I.28 Generally, acidic solutions contribute to the HER due to the presence of a series of protons. CO2 photo-reduction is prevalent in a basic solution owing to its improved solubility.29 The standard redox potentials acquired from Table I have a reference with a NHE at pH = 7 in an aqueous solution. According to the standard redox potentials, the CO2 photo-conversion usually competes with the HER, which causes decreasing product selectivity and low CO2 solar-to-fuel transformation efficiency. Generally, there are two main mechanisms of CO2 photo-reduction.

  • Usually, the most accepted mechanism is the CO2 reduction via selective generation of the CO2 intermediate formed with −1.9 V negative potential vs NHE in theory.30 This step is the rate-limiting step since an enormous reorganizational energy needs from the linear molecule of CO2 to the radical anion of bent CO2.6 Moreover, in addition to the large activation barrier of CO2, a high overpotential requested for CO2 reduction in thermodynamics causes poor performance of CO2 photo-conversion.

  • Another route of CO2 photo-reduction is the multiple proton-coupled electron transfer (PCET) process31 in which reduced products were made following multi-key elementary steps. As shown in Table I, the various products are generated by the transfer of the protons and electrons, including CO, HCOOH, CH4CH3OH, C2H4, C2H5OH, and H2C2O4. In addition, the high activation barriers caused by large reorganization energies and unstable intermediates can be readily bypassed. However, multi-electrons and protons are required in the process of PCET. For instance, reducing CO2 into CH4 and CH3OH, which are favorable in thermodynamics, requires eight electrons and six electrons, respectively, while converting CO2 into CO and HCOOH, which prefers kinetics, involves two electrons. In summary, it is important to preserve a fair number of protons in solution and enriched electron density around the exposed sites of photo-catalysts in the PCET mechanism. As a result, both thermodynamics and kinetics limitations increase the difficulties in converting CO2 into fuels.

TABLE I.

The basic products of CO2 reduction at the presence of water and the corresponding standard redox potentials according to acquired thermodynamic data with reference to the NHE at pH 7 in aqueous solution, 25 °C, and 1 atm gas pressure. Reproduced with permission from Hong et al., Anal. Methods 5, 1086 (2013). Copyright 2013 Royal Society of Chemistry.

ProductReactionE0 (V vs NHE)Equation
Hydrogen 2H2O + 2e →2OH + H2 −0.41 (1) 
Methane CO2 + 8H+ + 8e → CH4 + 2H2−0.24 (2) 
Carbon monoxide CO2 + 2H+ + 2e → CO + H2−0.51 (3) 
Methanol CO2 + 6H+ +6e → CH3OH + H2−0.39 (4) 
Formic acid CO2 + 2H+ +2e → HCOOH −0.58 (6) 
Ethane 2CO2 + 14H+ + 14e → C2H6 + 4H2−0.27 (6) 
Ethanol 2CO2 + 12H+ + 12e → C2H5OH + 3H2−0.33 (7) 
Oxalate 2CO2 + 2H+ + 2e → H2C2O4 −0.87 (8) 
ProductReactionE0 (V vs NHE)Equation
Hydrogen 2H2O + 2e →2OH + H2 −0.41 (1) 
Methane CO2 + 8H+ + 8e → CH4 + 2H2−0.24 (2) 
Carbon monoxide CO2 + 2H+ + 2e → CO + H2−0.51 (3) 
Methanol CO2 + 6H+ +6e → CH3OH + H2−0.39 (4) 
Formic acid CO2 + 2H+ +2e → HCOOH −0.58 (6) 
Ethane 2CO2 + 14H+ + 14e → C2H6 + 4H2−0.27 (6) 
Ethanol 2CO2 + 12H+ + 12e → C2H5OH + 3H2−0.33 (7) 
Oxalate 2CO2 + 2H+ + 2e → H2C2O4 −0.87 (8) 

To increase the efficiency of CO2 reduction, some significant factors should be considered in developing a highly efficient reaction system. Primarily, H2O is recognized as an ideal proton source and agents of consuming holes in the system of CO2 photo-reduction. However, in some photocatalytic system, the electron donor and sacrificial agents, including triethanolamine (TEOA),32–34 triethylamine (TEA),35,36 and ascorbic acid (AA),37,38 are widely utilized to consume holes to reduce the recombination possibilities with photoelectrons. Moreover, usually, the performance of CO2 photo-reduction is evaluated by the formation rate of the product (turnover number, TON, µmol g−1 h−1) or apparent quantum yield (AQE, %). However, since photocatalytic experimental conditions are complicated, including the designs of the photocatalytic equipment, reaction types (liquid–solid or gas–solid), amount of catalysts used, and light intensity, direct comparison of photo-activities via the generation rate among reaction systems in the literature is unreliable.

In this review, we comprehensively summarize the recent findings of enhanced CO2 photo-reduction efficiency from the parts including improvement of light absorption and effective charge separation as well as the enhancement of CO2 adsorption and activation. Moreover, a series of methodologies employed to tune the bandgap of the semiconductor and surface area, and surface connection modes for CO2 activation are also presented. We hope that this review can provide comprehensive discussions about CO2 photo-reduction and help people optimize their own reaction system for CO2 reduction.

As we discussed previously, the suitable band alignment of the photo-catalyst is ∼1.23 V to reduce CO2 under visible-light, with a similar potential required for water splitting.39 In view of thermodynamics, the bottom of the conduction band of photoactive materials should be more negative than the standard reduction potential for the photo-reduction reactions, while the photo-oxidation process can exist only when the top of the valance band of the semiconductors is more positive than the standard oxidation potential. In all photocatalytic semiconductors, TiO2, C3N4, Cu2O, and CdS are widely applied in CO2 photo-reduction. However, poor charge separation and serious recombination lead to low light-convert efficiency when semiconductor materials are solely utilized in this reaction.

1. Semiconductors doped with non-metals

The proper cation doping can not only narrow the bandgap of the semiconductor with enhanced light absorption at the long wavelength region but also improve the charge separation and migration. For example, introducing S atoms to the g-C3N4 lattice by Cheng’s group would lead to a unique electron structure modification that is a broader valence band, and upshift of the conduction band was determined and could facilitate charge transfer. The HER photoactivity of S doped g-C3N4 is 7.2-fold higher than that of bare g-C3N4.40 Shown et al. reported a carbon-doped SnS2 nanomaterial applied in CO2 photo-reduction, which displayed the enhanced photo-activity of reducing CO2 into CH3CHO compared with undoped SnS2. The carbon-doped SnS2 showed enhanced light absorption with the bandgap of 2.34 eV–2.54 eV with respect to bare SnS2 [Fig. 2(a)]. Moreover, the C element doping not only built microstrain within SnS2 but also altered the electronic structure of the bandgap alignment, which benefited for charge separation and improved carrier lifetime. As a result, carbon-doped SnS2 photo-catalysts showed a higher CH3CHO formation rate of 139.8 µmol g−1 h−1 compared to the generation rate of 0.55 µmol g−1 h−1 over sole SnS2 [Fig. 2(b)]. This carbon-doped SnS2 nanostructure obtained an apparent quantum efficiency of exceeding 0.7%.41 Similarly, Pt loading on carbon decorated indium-oxide (In2O3) utilized by Goodenough et al. exhibited a CH4 evolution rate of 126.6 µmol h−1, accompanying with a CO rate of 27.9 µmol h−1 [Fig. 2(d)]. It was measured that the bandgap of In2O3 nanomaterials was 2.53 eV, while the carbon decorated In2O3 owned the narrow bandgap of 2.48 eV [Fig. 2(c)]. The ultrathin carbon layer induced efficient charge transfer at the interface of the heterojunction, which increased sunlight absorption and meanwhile decreased the carrier recombination.42 It has been generally recognized that the valence band can be widened by homogeneous doping when the decorated atom has low electronegativity compared with the substituted atom, such as introducing the N atom in the TiO2 lattice replacing the O atom. The broadened valence band is contributed to the hole transfer and thus reduced charge recombination. Overall, the dopant concentration should be carefully tuned and controlled since excessive cation doping would cause defect sites acting as charge recombination centers.

FIG. 2.

(a) UV–vis diffuse reflectance and (insets) Tauc plots with both direct and indirect fittings. (b) The acetaldehyde generation rate for SnS2–C and SnS2 samples. (c) Energy bandgap alignment of the carbon decorated indium-oxide (C–In2O3) and indium-oxide (P–In2O3) with respect to the CO2 reduction potentials to carbon species of CO, CH4, HCOOH, HCHO, and CH3OH, regarding the reference of the NHE at pH = 7. (d) The reaction activity of Pt loading on carbon decorated indium-oxide and Pt decorated indium-oxide. (a) and (b) Reproduced with permission from Shown et al., Nat. Commun. 9, 169 (2018). Copyright 2018 Nature Springer. (c) and (d) Reproduced with permission from Pan et al., J. Am. Chem. Soc. 139, 4123 (2017). Copyright 2017 American Chemical Society.

FIG. 2.

(a) UV–vis diffuse reflectance and (insets) Tauc plots with both direct and indirect fittings. (b) The acetaldehyde generation rate for SnS2–C and SnS2 samples. (c) Energy bandgap alignment of the carbon decorated indium-oxide (C–In2O3) and indium-oxide (P–In2O3) with respect to the CO2 reduction potentials to carbon species of CO, CH4, HCOOH, HCHO, and CH3OH, regarding the reference of the NHE at pH = 7. (d) The reaction activity of Pt loading on carbon decorated indium-oxide and Pt decorated indium-oxide. (a) and (b) Reproduced with permission from Shown et al., Nat. Commun. 9, 169 (2018). Copyright 2018 Nature Springer. (c) and (d) Reproduced with permission from Pan et al., J. Am. Chem. Soc. 139, 4123 (2017). Copyright 2017 American Chemical Society.

Close modal

2. Semiconductors doped with metals

a. Noble metals.

The utilization of co-catalysts can lower activation energy and decrease charge recombination for CO2 photo-reduction. Apart from the non-metal dopant, metal atom incorporation can show unique characters and excellent potential applications due to the diversity of the electron configurations and orbits. Wang et al. reported platinum (Pt) nanoparticle (NP) loading on high-quality crystallized titanium oxide displaying enhanced CO2 photocatalytic activity [Fig. 3(a)].43 A highly efficient nanostructured film via facile gas-phase deposition presented a high methane generation rate of 1361 µmol g−1 h−1 during the reaction of CO2 photo-reduction [Fig. 3(b)]. The photo-generated electrons gathered on the co-catalyst Pt at the beginning and then rapidly transferred, which was further captured by the adsorbed CO2 molecule [Fig. 3(c)]. When the average sizes of Pt nanoparticles are no less than 1 nm, it may contribute to proper energy band alignment for effective separation due to the existence of the quantum confinement effect. Moreover, femtosecond time-resolved transient absorption (TA) spectroscopy performed for unmodified TiO2 and Pt decorated TiO2 samples showed that the carrier recombination of the Pt decorated TiO2 film was slower than that of bare TiO2. Finally, Pt incorporation not only reduced the barriers to CO2 activation but also enhanced efficient charge separation. Similarly, Zhang et al. reported a prominent size effect of Pt NPs as a co-catalyst in CO2 photo-reduction. They found that small Pt NPs promote electronic charge transfer, which could enhance both the hydrogen evolution reaction (HER) and carbon dioxide photo-catalytic reduction activity, leading to higher selectivity toward hydrogen over methane. Combined with theoretical calculations, in Pt NPs, the terrace sites were believed to be the active sites for methane generation; meanwhile, the less-coordinated sites were suggested as the active sites for the HER.44 Besides Pt modified TiO2, palladium and platinum co-decorated graphitic carbon nitride were applied to CO2 reduction investigated by calculations.45 During the process of CO2 reduction, g-C3N4 furnished the generation of hydrogen (H*), while the metal atoms solely acted as the active sites contributing to the generation of carbon intermediates. The complete, Pt loading on g-C3N4 preferably produced CH4, which requires a rate-determining activation barrier of 1.16 eV, while (Pd) loading on g-C3N4 selectively generated HCOOH, which possesses a reaction barrier of 0.66 eV. In addition, the atom deposited on g-C3N4 markedly increased the visible-light absorption and achieved the excellent photocatalytic activity in CO2 reduction. Moreover, Xie et al. investigated the effects of co-catalysts, including silver (Ag), rhodium (Rh), gold (Au), Pd, and Pt supported on TiO2 with a system of CO2 photo-reduction in the presence of H2O. Transient photocurrent response data showed that the photocurrent reduced in the sequence of Pt > Pd > Au > Rh > Ag. This indicated that constructing noble metal co-catalyst supported on TiO2 could promote electron extractions from TiO2, dedicating to enhancing the efficiency of photo-generated electrons on CO2 reduction.46 In addition, this agreed well with that researched by Ishitani et al.47 In summary, the noble metal modified catalysts could extract electrons from photosensitive materials and serve as electronic capture centers, contributing to enhanced local electron densities that facilitate CH4 formation.

FIG. 3.

(a) A model and field emission scanning electron microscopy (FESEM) image of Pt nanoparticles supported on columnar TiO2 thin films. (b) CH4 and CO yields on a Pt promoted TiO2 thin film as a function of the photo-reaction time. (c) Schematic diagram of the CO2 photo-reduction mechanism by utilizing Pt promoted TiO2 nanostructured films. (a)–(c) Reproduced with permission from Wang et al., J. Am. Chem. Soc. 134, 11276 (2012). Copyright 2012 American Chemical Society.

FIG. 3.

(a) A model and field emission scanning electron microscopy (FESEM) image of Pt nanoparticles supported on columnar TiO2 thin films. (b) CH4 and CO yields on a Pt promoted TiO2 thin film as a function of the photo-reaction time. (c) Schematic diagram of the CO2 photo-reduction mechanism by utilizing Pt promoted TiO2 nanostructured films. (a)–(c) Reproduced with permission from Wang et al., J. Am. Chem. Soc. 134, 11276 (2012). Copyright 2012 American Chemical Society.

Close modal
b. Transition metals.

In addition to the wonderful performance of noble metal, transition metals such as cobalt (Co), nickel (Ni), and copper (Cu) were also applied to improve the photo-activity of CO2 reduction, which can build doping energy levels not only increasing light absorption and electronic conductivity but also enhancing carrier lifetime. For instance, indium-doped TiO2 formed toward the blue shift of the band edge absorption and caused the enhanced bandgap. The CH4 yield of the indium (In) doped TiO2 nanoparticle possessed the rate of 244 µmol gcat−1 h−1, which exhibited 7.9 times higher than that of bare TiO2 at the reaction temperature of 100 °C.48 Besides TiO2 modification, molybdenum (Mo) merged into graphitic carbon nitride displayed a redshift of light adsorption with the increase in the Mo doping due to photoinduced electron transfer from the dopant band alignment to the g-C3N4 conduction band or from the valence band of g-C3N4 to the dopant band alignment. A dopant energy level formed by Mo doping contributed to decreasing charge recombination and increasing carrier lifetime. As a result, Mo-doped g-C3N4 materials showed the formation rate of CO and CH4 at 887 μmol gcat−1 and 123 µmol gcat−1 under 8 h irradiation compared with undoped g-C3N4 at 400 μmol gcat−1 and 40 µmol gcat−1 for CH4 and CO, respectively.49 In addition to altering electronic structures by elementary doping, the product selectivity can be tuned in some cases. The photo-catalyst of Ni doped ZnCo2O4 reported by Liu et al. was proved to facilitate the CO desorption, while the undoped ZnCo2O4 layer preferred the CH4 desorption confirmed by temperature-programed desorption.50 Moreover, the dopant of active metals with unsaturated species can lower the barrier of CO2 activation. Besides, Huang et al. synthesized activated Co2+ sites supported on g-C3N4, which displayed excellent photo-activity and selectivity for CO formation [Figs. 4(a) and 4(b)]. As shown in Fig. 4(c), the existence of Co2+ sites was indicated by the signal of Co 2p3/2 at 781 eV. Furthermore, the bivalent Co in active Co2+ loading on C3N4 samples was confirmed by Co K-edge x-ray absorption near edge structure (XANES) spectroscopy. Moreover, the generation of CO reached 0.128 µmol mg−1 [Fig. 4(d)] and TON reached 200, with the quantum yield up to 0.40%.51 In addition, Yuan et al. utilized dispersed Cu modified mesoporous TiO2 to selectively reduce CO2 into CH4 while proposing that Cu(0) is efficient in sacrificing holes, while Cu(I) is an active site in facilitating the formation of CH4 from in situ XANES characterization at the Cu K-edge.52 

FIG. 4.

(a) Schematic presentation of the photo-catalyst of Co2+ loading on C3N4. (b) The structure of cobalt complex. (c) X-ray photoelectron spectra of sole C3N4 and C3N4 decorated by different Co2+ concentrations. (d) Amounts of CO as a function of Co2+ loading. (a)–(d) Reproduced with permission from Huang et al., J. Am. Chem. Soc. 140, 16042 (2018). Copyright 2018 American Chemical Society.

FIG. 4.

(a) Schematic presentation of the photo-catalyst of Co2+ loading on C3N4. (b) The structure of cobalt complex. (c) X-ray photoelectron spectra of sole C3N4 and C3N4 decorated by different Co2+ concentrations. (d) Amounts of CO as a function of Co2+ loading. (a)–(d) Reproduced with permission from Huang et al., J. Am. Chem. Soc. 140, 16042 (2018). Copyright 2018 American Chemical Society.

Close modal

3. Constructing heterojunctions

Effective charge separation is another important factor to improve CO2 solar-to-fuel transformation efficiency. Possible reason is that charge recombination within 10−9 s is one to eight orders of magnitude faster than the surface chemical reaction (10−8 s to 10−1 s). Considering photo-generated charge transfer mechanisms and light absorption, the following four types of heterojunctions are summarized, including semiconductor–metal junction with the n-type heterojunction, semiconductor–metal junction with p-type connection, two semiconductors with type II band alignment, and two semiconductors of the Z-scheme structure39 (Fig. 5).

FIG. 5.

Hybrid systems benefited for enhanced charge separation: (a) n-type semiconductor–metal structure, (b) p-type semiconductor–metal heterojunction, (c) two semiconductors with a structure of type II band alignment, and (d) a Z-scheme band alignment of two semiconductors where the redox mediator molecule is used to inhibit electron transfer in the undesired direction. (a)–(d) Reproduced with permission from Stolarczyk et al., ACS Catal. 8, 3602 (2018). Copyright 2018 American Chemical Society.

FIG. 5.

Hybrid systems benefited for enhanced charge separation: (a) n-type semiconductor–metal structure, (b) p-type semiconductor–metal heterojunction, (c) two semiconductors with a structure of type II band alignment, and (d) a Z-scheme band alignment of two semiconductors where the redox mediator molecule is used to inhibit electron transfer in the undesired direction. (a)–(d) Reproduced with permission from Stolarczyk et al., ACS Catal. 8, 3602 (2018). Copyright 2018 American Chemical Society.

Close modal

In the n-type semiconductor–metal junction, electrons transfer from the semiconductor to metal because the position of the Fermi level of the semiconductor is higher than that of metal [Fig. 5(a)]. The band alignment at the heterojunction creates a Schottky barrier for hindering back electron transfer. Kuehnel et al. reported that nickel terpyridine complexes anchored on CdS quantum dots showed a TON of CO2 reduction at 20 and AQE at 0.28 ± 0.04% under the monochromatic light of 400 nm.53 Raziq et al. designed Au-decorated nanocrystalline TiO2 supported on MnOx-decorated g-C3N4. This hybrid material yielded CH4 at 140 µmol g−1 h−1 and H2 at 313 µmol g−1 h−1, which are higher than that of bare g-C3N4.54 

In a p-type semiconductor–metal structure, the electrons can easily transfer from the semiconductor to metal. However, at the same time, the hole cannot migrate due to the barrier between the semiconductor and metal [Fig. 5(b)]. As a result, the efficiency of charge separation improves with the suppressed recombination process. Sekizawa et al. utilized the ruthenium complex supported on p-type semiconductor Fe2O3 to increase the reaction performance of the CO2 reduction. The combination of the Ru complex with a structure of modified Fe2O3 heterojunctions yielding HCOOH, CO, and H2 exhibited a steady photocurrent of 150 µA cm−2 at 0.1 V with the reference of the reversible hydrogen electrode (RHE).55 As mentioned above, a Ru–Re supramolecular decorated p-type semiconductor CuGaO2 was reported by Kumagai et al., which showed a pretty good photo-electrochemical performance of converting CO2 into CO at a low potential of 0.3 V with the reference of Ag/AgCl.56 Meanwhile, the electron injection from the supramolecular complex to CuGaO2 benefited from this hybrid structure. Besides, Ag co-catalyst modified Cu2O was utilized to facilitate electron transfer, and thus, the structure of Ag/Cu2O/ZnO exhibited a CO evolution rate of 9.94 µmol g–1.57 

In the type II heterojunction, the electrons transfer in one direction, and the holes migrate to the opposite. This structure leads to effective charge separation and long carrier lifetime [Fig. 5(c)]. Zhou et al. synthesized boron carbon nitride and CdS hybrid materials to fine-tune the reaction system.58 This optimized photo-catalyst yielded a CO evolution rate of 250 µmol g−1 h−1, which is ten times higher than the activity achieved from the bare BCN semiconductor. Besides, partially oxidized SnS2 atomic layers reported by Jiao et al. delivered a CO generation rate of 12.28 µmol g−1 h−1, which is about 2.6-fold-enhanced photo-activity when compared with SnS2.59 In addition to the sulfide compound, quantum dots (QDs) were widely applied in CO2 photo-reduction. For example, Ou et al. reported the heterojunction of the CsPbBr3 QDs anchored on g-C3N4 nanosheets, displaying the CO evolution rate of 149 µmol g−1 h−1, which was roughly 15-fold-higher compared with sole CsPbBr3 QDs.60 

Finally, the Z-scheme junction as a complementation system effectively promotes charge separation. In this energy band alignment, the photo-generated electrons can inject from the conduction band of semiconductor 1 (SC1) into the valance band of semiconductor 2 (SC2), where a mediator redox pair or material is employed to inhibit electron transfer in the reverse direction. As a result, reduction reactions selectively occur in SC2, while SC1 prefers oxidizing reactions [Fig. 5(d)]. This structure also contributes to decreasing carrier recombination and increasing energy carriers. Iwase et al. designed the hybrid structure of Pt decorated metal sulfide combined with CoOx decorated BiVO4. In this photocatalytic system, reduced graphene oxide acting as an electron mediator was employed to promote efficient charge transfer. This Z-scheme system successfully realized water splitting and CO2 photo-reduction.10 Di et al. constructed a direct Z-scheme junction of g-C3N4/SnS2 in which the transformation of the electron from g-C3N4 to SnS2 built an internal electric field at the interface. This g-C3N4/SnS2 hybrid material exhibited a high CH3OH yield (2.3 µmol g−1 h−1) compared with g-C3N4 (1.2 µmol g−1 h−1) or SnS2 (0.8 µmol g−1 h−1), which can benefit for accelerated charge transfer and improved electron extraction.61 Jiang et al. prepared a hierarchical heterojunction of α-Fe2O3 supported on g-C3N4, which showed the enhanced performance of converting CO2 into CO at a rate of 27.2 µmol g−1 h−1. This is higher than the reduction rate over g-C3N4 (10.3 µmol g−1 h−1).62 Besides, due to the negative conduction band of −0.9 eV (vs NHE in pH = 7), CdS has been widely applied in the hybrid structure of Au@CdS/TiO2,63 CdS/WO3,64 and CdS/CdWO4.65 Moreover, Mo modified In2O3–ZnIn2Se4 nanosheets,66 Ag/Ag3PO4/CeO2,67 and combination of GaN: ZnO, TaON, and TiO2: Ta/N with a binuclear Ru(ii) complex68 were reported in CO2 photo-reduction. In addition, Bi-rich systems such as CdS/BiVO469 and g-C3N4/Ag/AgCl/BiVO470 were also applied to improve photoactivity in CO2 solar-to-fuel transformation.

1. Increase of the surface defects

Defects are indispensable in photocatalytic reactions. Crystallographic defects exist where the periodic arrangement of the atom or molecular is interrupted and broken, which alters the electronic property of catalysts and further influences catalytic performance. Defects in the solid are divided into four types, including point defects, line defects, facet defects, and volume defects according to the dimensionality. In all of defects, point defects, which can be formed by introducing vacancies and impurities, are widely applied in photocatalytic CO2 reductions, which promotes the CO2 adsorption and activation. Moreover, one of the most beneficial vacancies is oxygen vacancy (VO), especially oxygen vacancies in metal oxides. The ultrathin WO3·0.33H2O nanotubes reported by Sun et al. exhibited a series of active VO sites, achieving CO2 photo-conversion into CH3COOH with the rate of 9.4 µmol g−1 h−1.71 As suggested by another photocatalytic system, the photo-generated oxygen defects would be selectively generated over reducible metal oxide hybrids such as the vacancies of α-Zn–Ge–O can be occupied with O of CO2; then, the CO2 molecule could be reduced to the solid carbon atom by accepting photoinduced electrons suggested by Wang et al.72 The CO2 molecules adsorbed at the VO activated sites situated on the (110) facet of TiO2 could be demonstrated by utilizing scanning tunneling microscopy (STM).73 Furthermore, the adsorbed CO2 molecule was identified at a VO site with exhibiting tilted adsorption geometry in which VO sites occupied by the CO2 species present brighter than the bridging hydroxyl features [Figs. 6(a) and 6(b)].

FIG. 6.

(a) Presentation exhibiting an oxygen vacancy (VO), a group of bridging hydroxyl (OHb), and a CO2 molecule adsorbed at a VO activated site. (b) The diamonds and circles represented the CO2 molecule and OHb species, respectively. Reproduced with permission from Lee et al., J. Am. Chem. Soc. 133, 10066 (2011). Copyright 2011 American Chemical Society.

FIG. 6.

(a) Presentation exhibiting an oxygen vacancy (VO), a group of bridging hydroxyl (OHb), and a CO2 molecule adsorbed at a VO activated site. (b) The diamonds and circles represented the CO2 molecule and OHb species, respectively. Reproduced with permission from Lee et al., J. Am. Chem. Soc. 133, 10066 (2011). Copyright 2011 American Chemical Society.

Close modal

Vacancies serving as trapping sites for photoinduced electron–hole pairs are dedicated to the kinetics of carrier separation, which is another benefit of defects. Defective ZnIn2S4 layers were created for the first time by Jiao et al. Utilizing ultrafast transient absorption (TA) spectroscopy confirmed that a large number of zinc vacancy concentrations benefited for a 1.7-fold increase in average carrier lifetime [Fig. 7(a)], resulting in improved carrier separation rates. The electrons trapped by the zinc vacancies were confirmed by electron spin resonance (EPR), as shown in Fig. 7(c). As a result, an evolution rate of 33.2 µmol g−1 h−1 for carbon monoxide formation was obtained at high zinc vacancy within ZnIn2S4 layers [Fig. 7(d)].74 Later, to utilize surface defect, partially oxidized SnS2 layers were properly constructed where the locally oxidized area served as an active site. Both charge-carrier separation and photo-generated electrons situated on Sn atoms around the O atoms would benefit from this defect. Thus, the activation barrier of intermediates was reduced, and intermediates was stabilized. As a consequence, the adequately oxidized SnS2 samples presented a carbon monoxide product rate of 12.28 µmol g−1 h−1.59 

FIG. 7.

Ultrafast TA spectroscopy of (a) the rich Zn (VZn-rich) vacancy of one-unit-cell ZIS layers and (b) the poor Zn vacancy (VZn-poor) of one-unit-cell ZIS layers. (c) EPR spectra of VZn-rich and VZn-poor ZIS layers. (d) The amounts of the CO product after 1 h light irradiation. Inset in (d) reveals the products of 13CO2 photo-conversion for the optimized ZIS layers. Reproduced with permission from Jiao et al., J. Am. Chem. Soc. 139, 7586 (2017). Copyright 2017 American Chemical Society.

FIG. 7.

Ultrafast TA spectroscopy of (a) the rich Zn (VZn-rich) vacancy of one-unit-cell ZIS layers and (b) the poor Zn vacancy (VZn-poor) of one-unit-cell ZIS layers. (c) EPR spectra of VZn-rich and VZn-poor ZIS layers. (d) The amounts of the CO product after 1 h light irradiation. Inset in (d) reveals the products of 13CO2 photo-conversion for the optimized ZIS layers. Reproduced with permission from Jiao et al., J. Am. Chem. Soc. 139, 7586 (2017). Copyright 2017 American Chemical Society.

Close modal

2. Increase in the surface basic sites

In addition, basic sites prepared by utilizing the solid base, including (hydro) oxide, sodium hydroxide (NaOH), magnesium oxide (MgO), and zirconia (ZrO2) on the surface of the photo-catalyst, were applied to improve CO2 adsorption and enhance photo-activity. With respect to CO2 photo-reduction over MgO in the presence of H2 reported by Kohno et al., the unconventional reaction mechanism with IR investigation was proposed in which stable CO2 molecule adsorbed on basic sites of MgO was activated to form the activated CO2 radical under photo-irradiation and then was reduced by H2 to generate the surface formate even without irradiation. Furthermore, CO was selectively generated by C–O bond rupture within surface formate. However, it cannot decompose to generate CO by itself.75 Similarly, the mechanism over ZrO2 agreed well with the MgO reaction process proposed by Kohno et al. Besides, loading a proper amount of MgO on Pt-promoted TiO2 would increase the generation rate of CH4 while decrease the formation rate of H2 and CO in addition to CO2 photo-reduction over MgO- and Pt-promoted TiO2 reported by Xie and co-workers.46 A possible reason is that basic sites preserved by MgO could increase the population of CO2 via chemisorption.

As mentioned above, functionalization to the surface of photo-catalysts with basic groups could improve the adsorption of CO2. For example, the amine groups modified TiO2 surface formed carbamate, which can promote CO2 activation due to high photo-activity compared with the linearly adsorbed CO2. Liao et al. prepared TiO2 modified by the amine functional group through titanium tetrachloride (TiCl4) and monoethanolamine (MEA) decoration following a solvothermal process. As a result, MEA–TiO2 exhibited a higher formation rate for methane (8.61 ppm h−1) and CO (66.75 ppm h−1) than that of sole TiO2 or hydroxyl-functionalized TiO2, which attributed to enhanced CO2 chemisorption and effective charge migration from excited MEA–TiO2 to CO2.76 In addition, the CdS nanocrystals functioned by the Ni complex(ii) with the two ligands of terpyridine promoted CO2 adsorption and activation reported by Kuehnel et al. The reaction mechanism proposed is that nickel complexes were reduced by capturing the electron to generate the nickel(i) compound, which then released a terpy ligand upon anchor with CO2 to form the Ni–CO2 complex contributing to the CO formation.53 Moreover, the CdS quantum dots modified by the Ni complex remained active in a pure aqueous solution and achieved more than 90% selectivity for CO evolution.

3. Increase in the surface area

a. Morphological control and structural modifications.

The rationally designed active sites and facets by tuning the morphology on semiconductors not only benefit CO2 chemisorption but also increase the surface area of catalysts leading to improved photocatalytic performance. The semiconductor of TiO2 is widely utilized in photo-catalysis, especially in understanding the CO2 reaction mechanism over different exposed facets. The ratio of exposed facets between (001) and (101) can be tuned and controlled by changing HF content in the precursor solution during TiO2 nanosheet preparation [Figs. 8(a)–8(c)].77 As they proposed, for HF0 samples, the exposed facets of (101) accumulated a large number of photoinduced electrons and holes with respect to (001) facets. For HF4.5 samples, there is a proper ratio between (001) and (101) facets. Probably, the (001) facet acted as an activated site of oxidation, while the facet of (101) was utilized as a reduction site. As a consequence, HF4.5 samples achieved a large CH4 evolution rate of 1.35 µmol h−1 g−1 compared with HF9.0. Liu et al. synthesized ultrathin ZnGa2O4 nanosheets following a facile solvothermal method. The ultrathin nanosheet with a thickness of 6 nm showed the CH4 formation rate of 6.9 ppm h−1. The nanosheet scaffold also presented a 35% improved conversion when compared with meso-ZnGa2O4 [Fig. 8(d)]. Moreover, the ultrathin nanosheets owned an AQE of 0.035% measured at 280 ± 15 nm.78 Similarly, Wu et al. reported flower-like M0.33WO3 (M = K, Rb, Cs) materials via the solvothermal method, exhibiting excellent photocatalytic CO2 reduction performance at ambient pressure. In the SAED pattern, the main facet exposed is (010) [Figs. 8(e)–8(h)], and CO2 is thermodynamically favored on facet (010) with adsorption energy at −1.00 eV compared to an adsorption energy of −0.38 eV on the (001) facet.79 

FIG. 8.

(a)–(c) Transmission electron microscopy (TEM) images of the samples without HF addition and HF4.5 and HF9. The insets exhibit the partial enlarged images of the samples of HF0, HF4.5, and HF9. (d) TEM images of ultrathin nanosheet-scaffolded ZnGa2O4 microspheres. (e) TEM, (f) element mapping, (g) high resolution transmission electron microscopy (HRTEM) image, and (h) selected area electron diffraction (SAED) pattern of the sample Rb0.33WO3. (a)–(c) Reproduced with permission from Yu et al., J. Am. Chem. Soc. 136, 8839 (2014). Copyright 2014 American Chemical Society. (d) Reproduced with permission from Liu et al., ACS Appl. Mater. Interfaces. 6, 2356 (2014). Copyright 2014 American Chemical Society. (e)–(h) Reproduced with permission from Wu et al., J. Am. Chem. Soc. 141, 5267 (2019). Copyright 2019 American Chemical Society.

FIG. 8.

(a)–(c) Transmission electron microscopy (TEM) images of the samples without HF addition and HF4.5 and HF9. The insets exhibit the partial enlarged images of the samples of HF0, HF4.5, and HF9. (d) TEM images of ultrathin nanosheet-scaffolded ZnGa2O4 microspheres. (e) TEM, (f) element mapping, (g) high resolution transmission electron microscopy (HRTEM) image, and (h) selected area electron diffraction (SAED) pattern of the sample Rb0.33WO3. (a)–(c) Reproduced with permission from Yu et al., J. Am. Chem. Soc. 136, 8839 (2014). Copyright 2014 American Chemical Society. (d) Reproduced with permission from Liu et al., ACS Appl. Mater. Interfaces. 6, 2356 (2014). Copyright 2014 American Chemical Society. (e)–(h) Reproduced with permission from Wu et al., J. Am. Chem. Soc. 141, 5267 (2019). Copyright 2019 American Chemical Society.

Close modal
b. Utilization of porous materials.

Apart from semiconductor catalysts, metal–organic frameworks (MOFs) and zeolites can be extensively developed in photo-catalysis. Various unique properties such as the pore size and electronic characters can be adjusted and tuned by changing the ligand and metal center. Yan et al. reported stable and porous MOFs of dinuclear Eu(iii)2 clusters, displaying the generation of formate with the rate of 321.9 µmol h−1 mmol MOF−1.80 Transient absorption and theoretical calculations were used to explain the charge injection behavior between Ru ligands and Eu(iii)2. Han et al. constructed Ni metal–organic framework monolayers. They proposed the following reaction mechanism: the photoactive agent of [Ru(bpy)3]2+ was photoinduced to generate the [Ru(bpy)3]2+* intermediate, which was then quenched by the electron donor TEOA, accompanying with the production of the reduced [Ru(bpy)3]+. Then, the electrons injected from the [Ru(bpy)3]+ to Ni-MOFs, which effectively promoted CO2 adsorption on unsaturated and exposed Ni sites. Therefore, the Ni-MOF exhibited 97.8% CO selectivity with a high AQE of 1.96%.81 In summary, the exposed metal site of porous MOF materials offers coordinatively unsaturated sites contributing to CO2 adsorption and activation.

On the other hand, zeolites are another type of porous material with a structure consisting of interlocking SiO4 and AlO4 tetrahedra with shared oxygen atoms. This feature benefited large porosity and high adsorption capacities. For example, the titanium oxide photosensitizer within the structure of the three-dimensional channel (Ti-MCM-48) synthesized by Anpo et al. exhibited the reactivity of CH3OH formation at 7.5 µmol g−1 h−1.82 Later, Ti-MCM-41 zeolite with the specific ratio of Si vs Al researched by Jia et al. displayed enhanced CH4 evolution with 93 ppm g−1 h−1 compared to P25 (24 ppm g−1 h−1). Moreover, when Ti-MCM-41 zeolite was decorated by Pt nanoparticles, the photo-activity achieved CH4 evolution at 8835 ppm g−1 h−1 and 93% selectivity.83 However, most of the reported photo-catalysts of zeolite only response to the ultraviolet irradiation of the short wavelength, resulting in poor solar-to-fuel efficiency. Therefore, there is an urgent demand to design and prepare active porous materials of visible-light response for CO2 photo-reduction.

Photocatalytic CO2 reduction not only consumes excessive CO2 on the earth but also achieves efficient conversion from solar to valued fuels. In this paper, strategies for improving the light-convert efficiency in CO2 reduction are summarized from the view of enhanced light harvesting, promoted the CO2 adsorption and activation, in which the bandgap of the semiconductor can be tuned and adjusted to increase light absorption by ion doping or co-catalyst incorporation. The effective charge separation and migration can be realized by constructing the diversity of hybrid heterojunctions, which increases carrier lifetime. The adsorption and activation of CO2 can be enhanced by defect engineering, functional group and basic site modification, and morphology tuning. Although significant progress to improve CO2 photo-conversion efficiency has been achieved, some matters and problems still exist. The relationship of structure and reaction performance is still unclear on the CO2 photo-reaction based on a large number of photo-catalytic systems studied. The insights of reaction barrier and rate-limiting step(s) are very important for the design of the photo-catalysts. Many in situ techniques require elucidating reaction paths of the CO2 reduction from the molecular level of view. On the other hand, adequate desorption and decreased re-oxidation of the product are also key factors to improve light-conversion efficiency for CO2 reduction. In addition, the products of CO2 photo-reduction are limited to CO and CH4, at the same time, the C2 compounds and C3 products are rarely reported. Besides, H2 evolution often competes with the CO2 reduction causing low production selectivity. Until now, the mechanism between product selectivity and materials structure is not clear. More efforts should be devoted to improving product selectivity through calculations and mechanistic study about the reaction process. Constructing hydrophobic interfaces in the CO2 photoreaction system with suppressed H2 evolution to increase product selectivity should be necessary.84 Isotopic trace experiments should be utilized to confirm the real carbon source for CO2 reduction. During the reaction process, the compounds involved in the photocatalytic system including photo-catalysts, solvents, and electron donor agents may possibly decompose. Thus, it is indispensable to illustrate the real carbon source. On the other hand, it is promising to use H2O as a sacrificial agent instead of TEOA and TEA. Many p-type semiconductors pose a relatively negative conduction band, which has a strong driving force to make CO2 reduction, including Cu2O, ZnTe, and GaP, while suffering from photo-dissociation, and have poor stability in the aqueous solution. Hence, the enhancement on the reaction stability of these photo-catalysts is highly desired. For example, people can build a protective layer via constructing hydrophobic interfaces and incorporating proper components of light resistant without altering the electronic properties of photo-catalysts. Finally, in addition to p-type semiconductors, tremendous efforts have been taken to improve CO2 solar-to-fuel efficiency by constructing porous materials with the narrowed bandgap such as the MOF with Cu,85 Co, Ni, and zinc (Zn) active sites.86 Moreover, the synergistic effect of photo-reduction and thermal-driven has been widely identified and become a promising strategy to boost CO2 reduction efficiency. Various materials, such as plasmonic nanoparticles supported on semiconductors, plasmonic nanoparticles, including RuO2/SrTiO3,87 Ag/ZrO2,88 Al/Cu2O,89 Cu/ZnO,21 and Cu–Ru,90 are utilized to promote CO2 photo-conversion efficiency. In summary, the utilization of photo-catalysis to convert carbon dioxide into valuable products is a challenge but very promising research topic. It is hoped that this Perspective will inspire more researchers in this field with enthusiasm and confidence.

The authors acknowledge financial support from the “111 Project” of China (Grant No. B18030), the “100 Young Academic Leaders Training Program” offered by Nankai University and the Tianjin Natural Science Foundation (Grant No. BE122121), and the National Natural Science Foundation of China (Grant Nos. 21971117 and 21522106). The authors also appreciate the National Key R&D Program of China (Grant No. 2017YFA0208000), the Open Funds (Grant No. RERU2019001) of the State Key Laboratory of Rare Earth Resource Utilization, the Tianjin Key Lab for Rare Earth Materials and Applications (ZB19500202), and the Functional Research Funds for the Central Universities, Nankai University (Grant No. 63186005).

1.
C. D.
Windle
and
R. N.
Perutz
,
Coord. Chem. Rev.
256
,
2562
(
2012
).
2.
E. V.
Kondratenko
,
G.
Mul
,
J.
Baltrusaitis
,
G. O.
Larrazábal
, and
J.
Pérez-Ramírez
,
Energy Environ. Sci.
6
,
3112
(
2013
).
3.
Y.
Xu
,
M.
Kraft
, and
R.
Xu
,
Chem. Soc. Rev.
45
,
3039
(
2016
).
4.
G.
Chen
,
G. I. N.
Waterhouse
,
R.
Shi
,
J.
Zhao
,
Z.
Li
,
L. Z.
Wu
,
C. H.
Tung
, and
T.
Zhang
,
Angew. Chem., Int. Ed.
58
,
17528
(
2019
).
5.
L.
Wang
,
W.
Chen
,
D.
Zhang
,
Y.
Du
,
R.
Amal
,
S.
Qiao
,
J.
Wu
, and
Z.
Yin
,
Chem. Soc. Rev.
48
,
5310
(
2019
).
6.
J.
Ran
,
M.
Jaroniec
, and
S.-Z.
Qiao
,
Adv. Mater.
30
,
1704649
(
2018
).
7.
X.
Chang
,
T.
Wang
, and
J.
Gong
,
Energy Environ. Sci.
9
,
2177
(
2016
).
8.
W.-J.
Ong
,
L.-L.
Tan
,
S.-P.
Chai
, and
S.-T.
Yong
,
Dalton Trans.
44
,
1249
(
2015
).
9.
J.
Liu
,
H.
Shi
,
Q.
Shen
,
C.
Guo
, and
G.
Zhao
,
Green Chem.
19
,
5900
(
2017
).
10.
A.
Iwase
,
S.
Yoshino
,
T.
Takayama
,
Y. H.
Ng
,
R.
Amal
, and
A.
Kudo
,
J. Am. Chem. Soc.
138
,
10260
(
2016
).
11.
C.
Kim
,
K. M.
Cho
,
A.
Al-Saggaf
,
I.
Gereige
, and
H.-T.
Jung
,
ACS Catal.
8
,
4170
(
2018
).
12.
L.
Wan
,
Q.
Zhou
,
X.
Wang
,
T. E.
Wood
,
L.
Wang
,
P. N.
Duchesne
,
J.
Guo
,
X.
Yan
,
M.
Xia
,
Y. F.
Li
,
A. A.
Jelle
,
U.
Ulmer
,
J.
Jia
,
T.
Li
,
W.
Sun
, and
G. A.
Ozin
,
Nat. Catal.
2
,
889
(
2019
).
13.
M.
Schreier
,
J.
Luo
,
P.
Gao
,
T.
Moehl
,
M. T.
Mayer
, and
M.
Grätzel
,
J. Am. Chem. Soc.
138
,
1938
(
2016
).
14.
P.-W.
Pan
,
Y.-W.
Chen
,
A. S.
Brichkov
, and
V. V.
Kozik
,
Mod. Res. Catal.
8
,
39
(
2019
).
15.
X.
Yang
,
E. A.
Fugate
,
Y.
Mueanngern
, and
L. R.
Baker
,
ACS Catal.
7
,
177
(
2017
).
16.
U.
Kang
,
S. K.
Choi
,
D. J.
Ham
,
S. M.
Ji
,
W.
Choi
,
D. S.
Han
,
A.
Abdel-Wahab
, and
H.
Park
,
Energy Environ. Sci.
8
,
2638
(
2015
).
17.
Q.
Guo
,
F.
Liang
,
X.-B.
Li
,
Y.-J.
Gao
,
M.-Y.
Huang
,
Y.
Wang
,
S.-G.
Xia
,
X.-Y.
Gao
,
Q.-C.
Gan
,
Z.-S.
Lin
,
C.-H.
Tung
, and
L.-Z.
Wu
,
Chem
5
,
2605
(
2019
).
18.
T.
Yuan
,
F.
Yuan
,
X.
Li
,
Y.
Li
,
L.
Fan
, and
S.
Yang
,
Chem. Sci.
10
,
9801
(
2019
).
19.
Z.
Xiong
and
X. S.
Zhao
,
J. Am. Chem. Soc.
134
,
5754
(
2012
).
20.
Z.
Xu
,
M.
Quintanilla
,
F.
Vetrone
,
A. O.
Govorov
,
M.
Chaker
, and
D.
Ma
,
Adv. Funct. Mater.
25
,
2950
(
2015
).
21.
Z.-j.
Wang
,
H.
Song
,
H.
Pang
,
Y.
Ning
,
T. D.
Dao
,
Z.
Wang
,
H.
Chen
,
Y.
Weng
,
Q.
Fu
,
T.
Nagao
,
Y.
Fang
, and
J.
Ye
,
Appl. Catal., B
250
,
10
(
2019
).
22.
J.
Liu
,
B.
Liu
,
Y.
Ren
,
Y.
Yuan
,
H.
Zhao
,
H.
Yang
, and
S.
Liu
,
J. Mater. Chem. A
7
,
14761
(
2019
).
23.
J.
Di
,
C.
Zhu
,
M.
Ji
,
M.
Duan
,
R.
Long
,
C.
Yan
,
K.
Gu
,
J.
Xiong
,
Y.
She
,
J.
Xia
,
H.
Li
, and
Z.
Liu
,
Angew. Chem., Int. Ed.
57
,
14847
(
2018
).
24.
G.
Yang
,
D.
Chen
,
H.
Ding
,
J.
Feng
,
J. Z.
Zhang
,
Y.
Zhu
,
S.
Hamid
, and
D. W.
Bahnemann
,
Appl. Catal., B
219
,
611
(
2017
).
25.
Y.
Sun
,
S.
Gao
,
F.
Lei
, and
Y.
Xie
,
Chem. Soc. Rev.
44
,
623
(
2015
).
26.
J.
Xiong
,
P.
Song
,
J.
Di
, and
H.
Li
,
Appl. Catal., B
256
,
117788
(
2019
).
27.
L.
Zhang
,
Z.-J.
Zhao
,
T.
Wang
, and
J.
Gong
,
Chem. Soc. Rev.
47
,
5423
(
2018
).
28.
J.
Hong
,
W.
Zhang
,
J.
Ren
, and
R.
Xu
,
Anal. Methods
5
,
1086
(
2013
).
29.
J. L.
White
,
M. F.
Baruch
,
J. E.
Pander
 III
,
Y.
Hu
,
I. C.
Fortmeyer
,
J. E.
Park
,
T.
Zhang
,
K.
Liao
,
J.
Gu
,
Y.
Yan
,
T. W.
Shaw
,
E.
Abelev
, and
A. B.
Bocarsly
,
Chem. Rev.
115
,
12888
(
2015
).
30.
N.
Zhang
,
R.
Long
,
C.
Gao
, and
Y.
Xiong
,
Sci. China Mater.
61
,
771
(
2018
).
31.
G.
Liu
,
C.
Zhen
,
Y.
Kang
,
L.
Wang
, and
H.-M.
Cheng
,
Chem. Soc. Rev.
47
,
6410
(
2018
).
32.
A.
Rosas-Hernández
,
P. G.
Alsabeh
,
E.
Barsch
,
H.
Junge
,
R.
Ludwig
, and
M.
Beller
,
Chem. Commun.
52
,
8393
(
2016
).
33.
Z.
Zhang
,
J.
Huang
,
Y.
Fang
,
M.
Zhang
,
K.
Liu
, and
B.
Dong
,
Adv. Mater.
29
,
1606688
(
2017
).
34.
S.
Wang
,
B. Y.
Guan
, and
X. W. D.
Lou
,
J. Am. Chem. Soc.
140
,
5037
(
2018
).
35.
C.
Yang
,
F.
Mehmood
,
T. L.
Lam
,
S. L.-F.
Chan
,
Y.
Wu
,
C.-S.
Yeung
,
X.
Guan
,
K.
Li
,
C. Y.-S.
Chung
,
C.-Y.
Zhou
,
T.
Zou
, and
C.-M.
Che
,
Chem. Sci.
7
,
3123
(
2016
).
36.
M. J.
Berr
,
P.
Wagner
,
S.
Fischbach
,
A.
Vaneski
,
J.
Schneider
,
A. S.
Susha
,
A. L.
Rogach
,
F.
Jäckel
, and
J.
Feldmann
,
Appl. Phys. Lett.
100
,
223903
(
2012
).
37.
W.
Liu
,
X.
Li
,
C.
Wang
,
H.
Pan
,
W.
Liu
,
K.
Wang
,
Q.
Zeng
,
R.
Wang
, and
J.
Jiang
,
J. Am. Chem. Soc.
141
,
17431
(
2019
).
38.
M. F.
Kuehnel
,
C. E.
Creissen
,
C. D.
Sahm
,
D.
Wielend
,
A.
Schlosser
,
K. L.
Orchard
, and
E.
Reisner
,
Angew. Chem., Int. Ed.
58
,
5059
(
2019
).
39.
J. K.
Stolarczyk
,
S.
Bhattacharyya
,
L.
Polavarapu
, and
J.
Feldmann
,
ACS Catal.
8
,
3602
(
2018
).
40.
G.
Liu
,
P.
Niu
,
C.
Sun
,
S. C.
Smith
,
Z.
Chen
,
G. Q.
Lu
, and
H.-M.
Cheng
,
J. Am. Chem. Soc.
132
,
11642
(
2010
).
41.
I.
Shown
,
S.
Samireddi
,
Y. C.
Chang
,
R.
Putikam
,
P. H.
Chang
,
A.
Sabbah
,
F. Y.
Fu
,
W. F.
Chen
,
C. I.
Wu
,
T. Y.
Yu
,
P. W.
Chung
,
M. C.
Lin
,
L. C.
Chen
, and
K. H.
Chen
,
Nat. Commun.
9
,
169
(
2018
).
42.
Y.-X.
Pan
,
Y.
You
,
S.
Xin
,
Y.
Li
,
G.
Fu
,
Z.
Cui
,
Y.-L.
Men
,
F.-F.
Cao
,
S.-H.
Yu
, and
J. B.
Goodenough
,
J. Am. Chem. Soc.
139
,
4123
(
2017
).
43.
W.-N.
Wang
,
W.-J.
An
,
B.
Ramalingam
,
S.
Mukherjee
,
D. M.
Niedzwiedzki
,
S.
Gangopadhyay
, and
P.
Biswas
,
J. Am. Chem. Soc.
134
,
11276
(
2012
).
44.
C.
Dong
,
C.
Lian
,
S.
Hu
,
Z.
Deng
,
J.
Gong
,
M.
Li
,
H.
Liu
,
M.
Xing
, and
J.
Zhang
,
Nat. Commun.
9
,
1252
(
2018
).
45.
G.
Gao
,
Y.
Jiao
,
E. R.
Waclawik
, and
A.
Du
,
J. Am. Chem. Soc.
138
,
6292
(
2016
).
46.
S.
Xie
,
Y.
Wang
,
Q.
Zhang
,
W.
Deng
, and
Y.
Wang
,
ACS Catal.
4
,
3644
(
2014
).
47.
O.
Ishitani
,
C.
Inoue
,
Y.
Suzuki
, and
T.
Ibusuki
,
J. Photochem. Photobiol., A
72
,
269
(
1993
).
48.
M.
Tahir
and
N. S.
Amin
,
Appl. Catal., B
162
,
98
(
2015
).
49.
Y.
Wang
,
Y.
Xu
,
Y.
Wang
,
H.
Qin
,
X.
Li
,
Y.
Zuo
,
S.
Kang
, and
L.
Cui
,
Catal. Commun.
74
,
75
(
2016
).
50.
K.
Liu
,
X.
Li
,
L.
Liang
,
J.
Wu
,
X.
Jiao
,
J.
Xu
,
Y.
Sun
, and
Y.
Xie
,
Nano Res.
11
,
2897
(
2018
).
51.
P.
Huang
,
J.
Huang
,
S. A.
Pantovich
,
A. D.
Carl
,
T. G.
Fenton
,
C. A.
Caputo
,
R. L.
Grimm
,
A. I.
Frenkel
, and
G.
Li
,
J. Am. Chem. Soc.
140
,
16042
(
2018
).
52.
L.
Yuan
,
S.-F.
Hung
,
Z.-R.
Tang
,
H. M.
Chen
,
Y.
Xiong
, and
Y.-J.
Xu
,
ACS Catal.
9
,
4824
(
2019
).
53.
M. F.
Kuehnel
,
K. L.
Orchard
,
K. E.
Dalle
, and
E.
Reisner
,
J. Am. Chem. Soc.
139
,
7217
(
2017
).
54.
F.
Raziq
,
L.
Sun
,
Y.
Wang
,
X.
Zhang
,
M.
Humayun
,
S.
Ali
,
L.
Bai
,
Y.
Qu
,
H.
Yu
, and
L.
Jing
,
Adv. Energy Mater.
8
,
1701580
(
2018
).
55.
K.
Sekizawa
,
S.
Sato
,
T.
Arai
, and
T.
Morikawa
,
ACS Catal.
8
,
1405
(
2018
).
56.
H.
Kumagai
,
G.
Sahara
,
K.
Maeda
,
M.
Higashi
,
R.
Abe
, and
O.
Ishitani
,
Chem. Sci.
8
,
4242
(
2017
).
57.
F.
Zhang
,
Y.-H.
Li
,
M.-Y.
Qi
,
Z.-R.
Tang
, and
Y.-J.
Xu
,
Appl. Catal., B
268
,
118380
(
2020
).
58.
M.
Zhou
,
S.
Wang
,
P.
Yang
,
C.
Huang
, and
X.
Wang
,
ACS Catal.
8
,
4928
(
2018
).
59.
X.
Jiao
,
X.
Li
,
X.
Jin
,
Y.
Sun
,
J.
Xu
,
L.
Liang
,
H.
Ju
,
J.
Zhu
,
Y.
Pan
,
W.
Yan
,
Y.
Lin
, and
Y.
Xie
,
J. Am. Chem. Soc.
139
,
18044
(
2017
).
60.
M.
Ou
,
W.
Tu
,
S.
Yin
,
W.
Xing
,
S.
Wu
,
H.
Wang
,
S.
Wan
,
Q.
Zhong
, and
R.
Xu
,
Angew. Chem., Int. Ed.
130
,
13758
(
2018
).
61.
T.
Di
,
B.
Zhu
,
B.
Cheng
,
J.
Yu
, and
J.
Xu
,
J. Catal.
352
,
532
(
2017
).
62.
Z.
Jiang
,
W.
Wan
,
H.
Li
,
S.
Yuan
,
H.
Zhao
, and
P. K.
Wong
,
Adv. Mater.
30
,
1706108
(
2018
).
63.
Y.
Wei
,
J.
Jiao
,
Z.
Zhao
,
J.
Liu
,
J.
Li
,
G.
Jiang
,
Y.
Wang
, and
A.
Duan
,
Appl. Catal., B
179
,
422
(
2015
).
64.
J.
Jin
,
J.
Yu
,
D.
Guo
,
C.
Cui
, and
W.
Ho
,
Small
11
,
5262
(
2015
).
65.
X.
Jia
,
M.
Tahir
,
L.
Pan
,
Z.-F.
Huang
,
X.
Zhang
,
L.
Wang
, and
J.-J.
Zou
,
Appl. Catal., B
198
,
154
(
2016
).
66.
Y.
Chao
,
P.
Zhou
,
N.
Li
,
J.
Lai
,
Y.
Yang
,
Y.
Zhang
,
Y.
Tang
,
W.
Yang
,
Y.
Du
,
D.
Su
,
Y.
Tan
, and
S.
Guo
,
Adv. Mater.
31
,
1807226
(
2019
).
67.
W.
Xiong
,
W.
Dai
,
X.
Hu
,
L.
Yang
,
T.
Wang
,
Y.
Qin
,
X.
Luo
, and
J.
Zou
,
Mater. Lett.
232
,
36
(
2018
).
68.
A.
Nakada
,
R.
Kuriki
,
K.
Sekizawa
,
S.
Nishioka
,
J. J. M.
Vequizo
,
T.
Uchiyama
,
N.
Kawakami
,
D.
Lu
,
A.
Yamakata
,
Y.
Uchimoto
,
O.
Ishitani
, and
K.
Maeda
,
ACS Catal.
8
,
9744
(
2018
).
69.
Z.-H.
Wei
,
Y.-F.
Wang
,
Y.-Y.
Li
,
L.
Zhang
,
H.-C.
Yao
, and
Z.-J.
Li
,
J. CO2 Util.
28
,
15
(
2018
).
70.
R. A.
Rather
,
M.
Khan
, and
I. M. C.
Lo
,
J. Catal.
366
,
28
(
2018
).
71.
S.
Sun
,
M.
Watanabe
,
J.
Wu
,
Q.
An
, and
T.
Ishihara
,
J. Am. Chem. Soc.
140
,
6474
(
2018
).
72.
B.
Wang
,
X.
Wang
,
L.
Lu
,
C.
Zhou
,
Z.
Xin
,
J.
Wang
,
X.-k.
Ke
,
G.
Sheng
,
S.
Yan
, and
Z.
Zou
,
ACS Catal.
8
,
516
(
2018
).
73.
J.
Lee
,
D. C.
Sorescu
, and
X.
Deng
,
J. Am. Chem. Soc.
133
,
10066
(
2011
).
74.
X.
Jiao
,
Z.
Chen
,
X.
Li
,
Y.
Sun
,
S.
Gao
,
W.
Yan
,
C.
Wang
,
Q.
Zhang
,
Y.
Lin
,
Y.
Luo
, and
Y.
Xie
,
J. Am. Chem. Soc.
139
,
7586
(
2017
).
75.
Y.
Kohno
,
H.
Ishikawa
,
T.
Tanaka
,
T.
Funabiki
, and
S.
Yoshida
,
Phys. Chem. Chem. Phys.
3
,
1108
(
2001
).
76.
Y.
Liao
,
S.-W.
Cao
,
Y.
Yuan
,
Q.
Gu
,
Z.
Zhang
, and
C.
Xue
,
Chem. - Eur. J.
20
,
10220
(
2014
).
77.
J.
Yu
,
J.
Low
,
W.
Xiao
,
P.
Zhou
, and
M.
Jaroniec
,
J. Am. Chem. Soc.
136
,
8839
(
2014
).
78.
Q.
Liu
,
D.
Wu
,
Y.
Zhou
,
H.
Su
,
R.
Wang
,
C.
Zhang
,
S.
Yan
,
M.
Xiao
, and
Z.
Zou
,
ACS Appl. Mater. Interfaces
6
,
2356
(
2014
).
79.
X.
Wu
,
Y.
Li
,
G.
Zhang
,
H.
Chen
,
J.
Li
,
K.
Wang
,
Y.
Pan
,
Y.
Zhao
,
Y.
Sun
, and
Y.
Xie
,
J. Am. Chem. Soc.
141
,
5267
(
2019
).
80.
Z. H.
Yan
,
M. H.
Du
,
J.
Liu
,
S.
Jin
,
C.
Wang
,
G. L.
Zhuang
,
X. J.
Kong
,
L. S.
Long
, and
L. S.
Zheng
,
Nat. Commun.
9
,
3353
(
2018
).
81.
B.
Han
,
X.
Ou
,
Z.
Deng
,
Y.
Song
,
C.
Tian
,
H.
Deng
,
Y. J.
Xu
, and
Z.
Lin
,
Angew. Chem., Int. Ed.
57
,
16811
(
2018
).
82.
M.
Anpo
,
H.
Yamashita
,
K.
Ikeue
,
Y.
Fujii
,
S. G.
Zhang
,
Y.
Ichihashi
,
D. R.
Park
,
Y.
Suzuki
,
K.
Koyano
, and
T.
Tatsumi
,
Catal. Today
44
,
327
(
1998
).
83.
W.
Jia
,
T.
Liu
,
Q.
Li
, and
J.
Yang
,
Catal. Today
335
,
221
(
2019
).
84.
A.
Li
,
Q.
Cao
,
G.
Zhou
,
B. V. K. J.
Schmidt
,
W.
Zhu
,
X.
Yuan
,
H.
Huo
,
J.
Gong
, and
M.
Antonietti
,
Angew. Chem., Int. Ed.
58
,
14549
(
2019
).
85.
L.
Zeng
,
Z.
Wang
,
Y.
Wang
,
J.
Wang
,
Y.
Guo
,
H.
Hu
,
X.
He
,
C.
Wang
, and
W.
Lin
,
J. Am. Chem. Soc.
142
,
75
(
2020
).
86.
L. Z.
Dong
,
L.
Zhang
,
J.
Liu
,
Q.
Huang
,
M.
Lu
,
W. X.
Ji
, and
Y. Q.
Lan
,
Angew. Chem., Int. Ed.
132
,
2681
(
2020
).
87.
D.
Mateo
,
J.
Albero
, and
H.
García
,
Joule
3
,
1949
(
2019
).
88.
H.
Zhang
,
T.
Itoi
,
T.
Konishi
, and
Y.
Izumi
,
J. Am. Chem. Soc.
141
,
6292
(
2019
).
89.
H.
Robatjazi
,
H.
Zhao
,
D. F.
Swearer
,
N. J.
Hogan
,
L.
Zhou
,
A.
Alabastri
,
M. J.
McClain
,
P.
Nordlander
, and
N. J.
Halas
,
Nat. Commun.
8
,
27
(
2017
).
90.
L.
Zhou
,
J. M. P.
Martirez
,
J.
Finzel
,
C.
Zhang
,
D. F.
Swearer
,
S.
Tian
,
H.
Robatjazi
,
M.
Lou
,
L.
Dong
,
L.
Henderson
,
P.
Christopher
,
E. A.
Carter
,
P.
Nordlander
, and
N. J.
Halas
,
Nat. Energy
5
,
61
(
2020
).