Efficient nanomaterials for harvesting clean fuels from electrochemical and photoelectrochemical CO₂ reduction

Abstract

The excessive utilization of fossil fuels accompanied by large amounts of anthropogenic CO₂ emissions have led to adverse global environmental changes and a growing global energy crisis. Hence, converting CO₂ into high-value chemical fuels, such as CO, CH₄, HCOOH, and CH₃OH, through catalysis is one of the most attractive topics in energy conversion. Among various approaches, electrochemical (EC) and photoelectrochemical (PEC) reduction are considered to be promising methods. Over the past decades, research in the area of CO₂ EC and PEC reduction has been growing quickly. Herein, highly efficient nanostructured metals, metal alloys, metal oxides, metal-free electrodes, and photoelectrodes developed in recent years are discussed in detail in this review. Additionally, the strategies and mechanisms for improving the faradaic efficiency (FE), current density, and stability of catalysts are also discussed. The challenges and future perspectives for CO₂ EC and PEC reduction are also discussed.

---

Yang Yang

Yang yang is a PhD candidate under the supervision of Prof. Liwu Zhang in the Department of Environmental Science and Engineering at Fudan University. He obtained his Master's degree from the School of Environmental Science and Engineering at Nanjing University of Information Science & Technology in 2017. His research is focused on environmental materials development and CO2 reduction.

Saira Ajmal

Saira Ajmal is a PhD candidate under the supervision of Prof. Liwu Zhang in the School of Environmental Science and Engineering at Fudan University. She obtained her Master's degree from the School of Chemistry at the University of Sargodha Pakistan in 2015. Her research is focused on CO2 electrochemical reduction.

Xiuzhen Zheng

Dr Xiuzhen Zheng is a postdoctoral research fellow in the Department of Environmental Science and Engineering at Fudan University. She received her PhD degree from the State Key Laboratory of Photocatalysis on Energy and Environment from the Fuzhou University in 2015. Currently, she is conducting her postdoctoral studies under the supervision of Prof. Liwu Zhang, with a focus on enhancing the performance of artificial photosynthesis with the incorporation of photonic nanostructures, and further studying the impact of light manipulation in these systems.

Liwu Zhang

Prof. Liwu Zhang joined the Department of Environmental Science and Engineering at Fudan University in November 2014. Prior to his faculty appointment, he did his postdoctoral work under the supervision of Prof. Jeremy J. Baumberg (FRS) at the Nanophotonics Centre in the Cavendish Laboratory and collaborated with Dr Erwin Reisner from the Department of Chemistry at the University of Cambridge. Before that, he was an Alexander von Humboldt Research Fellow under the supervision of Prof. Detlef Bahnemann at Hannover University. He received his PhD in Chemistry from Tsinghua University under the supervision of Prof. Yongfa Zhu. He is a recipient of the Marie Curie Intra-European Research Fellow award. He has over a decade of experience in environmental nanotechnology and a strong background in photocatalysis and has published more than 50 highly cited papers in this field. His work has been cited >3000 times, and he has an h-index of 23.

---

1 Introduction

Due to the increasing global population, the demand for energy resources is also growing quickly. As early as 2012, researchers reported that a large portion of the worldwide energy requirement (80%) is still fulfilled by fossil fuels, including petroleum, nature gases, and coal.¹ However, the use of such resources results in an increased concentration of greenhouse gases (GHS) and many associated environmental problems, including climate change, acid rain, depletion of the ozone layer, sea level rises, and global warming.^2,3 According to the studies of the Intergovernmental Panel on Climate Change (IPCC), the CO₂ gas emitted by the burning of fossils fuel among all other anthropogenic gases is supposed to be the chief source of climate change and global warming.^4,5 Thus in an ideal scenario, there should be a balance between CO₂ production and consumption, but unfortunately this has been disrupted.⁶ It has been reported that the atmospheric level of CO₂ reached up to ∼403.95 ppm by July 2017 (Data from the NOAA research) and was predicted to rise and reach nearly 600 ppm by 2100.^7,8

How to remove the excessive CO₂ is one of the biggest current challenges for mankind. CO₂ also has wide applications in agriculture, food, as an inert gas, in fire extinguishers, in the medical field, as a refrigerant, and in other fields, but these cannot remove CO₂ from the air.⁹ In order to solve this problem, converting CO₂ into another useful fuel is an attractive proposition.^10,11 Up to now, a number of routes have been established for CO₂ conversion, such as biological,^12,13 chemical reorganizing,^14–17 mineralization,^18–20 photochemical, EC methods,^16,21,22 and PEC methods.^23–26

Also, effective energy storage of non-fossil fuel sources is required as part of the long-term solution of mankind's fuel requirements.^27,28 With this in mind, various fuels have been projected as alternative fuels. One of the emerging fuels proposed as the next generation energy source is hydrogen (H₂). However, the production of H₂ fuel requires a high amount of energy intensive labor along with the need for special materials to avoid H₂ diffusion and leakage. The use of H₂ gas as a fuel also carries a risk of explosion, so wide safety precautions are usually required.^29,30 Bioelectrochemical systems (BESs) have recently been proposed as a new and sustainable technology for CO₂ reduction.³¹ However, these often lack effective bacteria, and furthermore, the bacteria are easily broken by the reduction production. Additionally, the membrane is easily polluted, which also increases the cost. Photochemical (PC), electrochemical (EC), and photoelectrochemical (PEC) methods have attracted great attention for CO₂ reduction due to their energy-saving benefits and wide range of product, including carbon monoxide (CO), methane (CH₄), ethylene (C₂H₄), methanol (CH₃OH), formic acid (HCOOH), formaldehyde (HCHO), and other high alcohols, hydrocarbons, and oxygenates.^{14,16,21,22,32–35} The driving force of the photochemical reduction of CO₂ is only solar energy, and consequently it is regarded as an artificial model for photosynthesis. However, the current low solar energy conversion efficiency, charge separation, and transportation efficiency have limited its application. For the EC reduction of CO₂, the driving force is an external applied bias, which consumes a lot of energy, and so this is a limiting factor to its wider take up. However, the reduction products and process can be tuned by the applied external bias, which are considered benefits. For the PEC reduction of CO₂, the driving force is the synergy of solar energy and an external bias, which thus combines the advantages of both the EC and PC processes.^32,36 However, the CO₂ molecule is very stable due to its linear and centrosymmetric structure.^32,37 Additionally, the C–O bond length in CO₂ is about 116.3 pm, which is shorter than that of a C–O single bond and most other C–O multiply-bonded functional groups. It is thus difficult to break the C–O bond to form new carbon-based materials. Although a number of photocatalysis systems, electrodes, and photoelectrodes have been developed for CO₂ reduction, most of them still suffer from the low energy conversion efficiency, low selectivity, a low FE, poor stability, and incapability to suppress the competing hydrogen evolution reaction (HER). Therefore, how to design and fabricate the required photocatalysts, electrodes, and photoelectrodes with high selectivity, high efficiency, and good stability is still a big challenge for CO₂ reduction.

Up to now, several excellent reviews about CO₂ photochemical reduction have been published.^38–42 These reviews involve the design and fabrication of various photocatalysts as well as descriptions of the CO₂ reduction process. Chang et al. summarized the strategies for enhancing the adsorption and activation of CO₂ as well as the factors determining the reaction pathways and selectivity for the reduction products.⁴² There are also some reviews in the literature about EC reduction covering various metal electrodes, metal oxide electrodes, sulfide catalysts, and carbon catalysts, etc., as well as the crucial factors affecting the electroreduction performance.^10,43,44 There are also some reviews available on CO₂ PC, and EC and PEC reduction reactions employing metal complex catalysts.^45–50 In addition, there are a few reviews involving PEC reduction.^49,51 However, as far as we know, there are hardly any reviews entirely devoted to both EC and PEC reduction of CO₂ by metals, metallic alloys, and inorganic semiconductors in inorganic salt aqueous solutions. Therefore, herein we specifically focus on emerging nanomaterials and nanoarchitectures as efficient EC and PEC electrodes with a high FE of the desired CO₂ reduction products. Moreover, the current review summarizes the variety of electrocatalysts that have been commonly employed in the past 5 years for CO₂ reduction through either the EC or PEC method. We summarize different groups of electrodes, including metal nanostructures (Cu, Au, Ag), metal alloys (Cu–Ni, Cu–Pd, Cu–Sn, Cu–Zn, Cu–Ag, Cu–In, and others), metal oxides (CuO_x, SnO₂, CeO₂, etc.), chalcogenides (MoS₂, WS₂, etc.), and metal-free electrodes (C, Si, etc.). Besides, we divided the photoelectrodes into three categories: (1) phosphides (GaP, InP, etc.); (2) metal oxides (Cu_xO_y, TiO₂, ZnO, etc.); and (3) metal-free (C₃N₄, Si) photoelectrodes for CO₂ reduction. The reaction mechanisms and catalysis pathways are also discussed in this review.

2 Basic concepts and mechanism of CO₂ reduction

The CO₂ molecule has a linear structure and is highly stable thermodynamically as well as kinetically. It exhibits a highest oxidation state of +4 and is the ultimate product of many biological and chemical processes.³⁷ For the reduction of CO₂ into other useful compounds, a high amount of energy is required because of its low energy levels. As CO₂ is a linear molecule and centrosymmetric, the molecule has no electrical dipole. The length of carbon–oxygen bonds is 116.3 pm, which is smaller than that of a C–O single bond, and even shorter than most other C–O multiply-bonded functional groups. Its particular molecular structure and weak electrophile nature makes it difficult to transfer to other carbon compounds. The reduction of CO₂ into renewable lower products involves multistep electron pathways.²⁴ Among the various methods of CO₂ reduction, the EC and PEC methods have advantages as the reaction rate and products could be controlled by tuning the external bias overpotential.⁵² The reduction of CO₂ forms various products, such as carbon monoxide (CO), methanol (CH₃OH), ethanol (C₂H₅OH), ethylene (C₂H₄), methane (CH₄), formate (HCOO⁻), and formic acid (HCOOH).^53,54 The production of final products is greatly influenced by the nature of the catalyst used and the applied potential.⁵⁵

CO₂ adsorption and activation are the key steps in both EC and PEC reduction processes. The pathways of EC and PEC reduction of CO₂ are also significantly affected by the applied potential. According to the different reaction mechanisms and pathways, we can get various products, such as CO, CH₄, CH₃OH, HCOOH, HCHO, C₂H₆, other carbohydrates. One of the challenges is to suppress the evolution of H₂ (HER), which always occurs during the CO₂ reduction process. The equilibrium potentials of the different products are shown in Table 1. For the pathway involving 2 electrons, the possible reduction products are H₂, CO, and HCOOH. However, the standard redox potentials for CO and HCOOH are more negative than that of H₂, which means that the HER will compete with the CO₂ reduction. Also, 6 and 8 electrons are needed for the formation of CH₃OH and CH₄, respectively, which would be more difficult than that of H₂ evolution. Additionally, the Gibbs free energy for the formation of CH₃OH and CH₄ by the reaction of CO₂ and H₂O is about 702 and 818 kJ mol⁻¹, respectively.⁵⁶ Meanwhile, the Gibbs free energy change for converting H₂O to H₂ and 1/2O₂ is about 273 kJ mol⁻¹, which will occur more easily than CO₂ reduction.⁵⁷ The solubility of CO₂ in H₂O is about 0.33 mol L⁻¹ at 25 °C under 1 atm, which would further make it difficult to adsorb and activate CO₂.^42,58 It has been found in some research studies that the nature of the products formed is mainly governed by the kinetics. Furthermore, that the kinetics greatly depend on the reaction temperature, pressure, reactant concentration, electrolyte, and so on. For enhancing the surface-adsorbed CO₂ concentration, researchers usually design some porous materials, including a 3D inverse opal structure, nanotubes, hollow structures and so forth. Designing suitable devices can also enhance the catalyst surface CO₂ concentration and enable forming different products. Ampelli et al. designed a novel EC setup by using metal-doped conjugated microporous polymers for CO₂ reduction, where their design had the ability to assemble the electrode materials to form a multilayered composite to guarantee good proton mobility and electron conductivity.^59,60 Compared with conventional cells, this electrolyte-less configuration exhibited differences in the productivity and product types, which may indicate a different reaction mechanism. More discussion of this can be found in the following sections. Hence, how to suppress the evolution of H₂ and improve the selectivity is still the biggest challenge in PEC and EC CO₂ reduction.

Table 1 The main products from CO₂ reduction and the corresponding reduction potentials with reference to NHE at pH = 7 in aqueous solution, 25 °C, and 1 atm gas pressure

Products	Reaction	E^0 (V vs. NHE)	Eqn
H2	2H^+ + 2e^− → H2	−0.41	(1)
CO	CO2 + 2H^+ + 2e^− → CO + H2O	−0.51	(2)
HCOOH	CO2 + 2H^+ + 2e^− → HCOOH	−0.58	(3)
HCHO	CO2 + 4H^+ + 4e^− → HCHO + H2O	−0.48	(4)
CH3OH	CO2 + 6H^+ + 6e^− → CH3OH + H2O	−0.38	(5)
CH4	CO2 + 8H^+ + 8e^− → CH4 + 2H2O	−0.24	(6)
CH3CH2OH	2CO2 + 12H^+ + 12e^− → CH3CH2OH + 3H2O	−0.33	(7)
CH3CH3	2CO2 + 14H^+ + 14e^− → CH3CH3 + 4H2O	−0.27	(8)
—	CO2 + e^− → CO2˙^−	−1.9 V (pH = 0)	(9)
O2	H2O + 2h^+ → 1/2O2 + 2H^+	0.82	(10)

---

For photocatalytic CO₂ reduction, the oxidation and reduction reactions often occur on different sites of the catalyst; hence, the products are mixed. In addition, the driving force for the photocatalytic reduction of CO₂ is only from solar energy, while the solar-to-fuel energy conversion efficiency is very low, which has limited its application to date. For the EC reduction of CO₂, the driving force is from an external applied bias, and the reaction can thus be more easily manipulated. For the PEC reduction of CO₂, the driving force is from the synergy between solar energy and an external bias, which thus combines the advantages of PC and EC methods. However, their reaction pathways are similar, and are usually affected by the experiment conditions, including the properties of the catalysts, electrolyte species, reaction temperature, and pressure, etc.⁴² Among the various factors, the property of the catalysts (electrodes and photoelectrodes) often plays a key role in the products distribution. For EC CO₂ reduction, taking the metal electrodes as an example, the reaction pathways may be different by using different types of electrodes, which could then lead to a distinct distribution of products. According to the previous investigations,⁶¹ In, Sn, Pb, and Cd have a high overpotential for H₂ evolution, but have a negligible ability to adsorb the CO intermediate, resulting in the preferred formation of HCOOH on these electrodes; while, Au, Ag, and Zn have a medium overpotential for H₂ evolution and a weak ability to adsorb the CO intermediate, resulting in the preferred formation of CO; and Al, Si, V, Fe, Mn, Co, Mo, and Pt, etc. have a low overpotential for H₂ evolution and a strong ability to adsorb CO, resulting in the formation of H₂. The distribution of products on the Cu electrode is very wide, including CO, H₂, CH_4, CH₃OH, HCOOH, and C₂H₄ and other hydrocarbons. The external bias applied on the electrode also plays an important role in the products distribution, which has been widely investigated by many groups, especially on the Cu electrode.^62–64 At the initial potential (around 0.8 to 0.95 V vs. SHE), HER is the main reaction. Until the applied potential reaches −0.95 V vs. the standard hydrogen electrode (SHE), the rate of HER start to decrease, and meanwhile CO and formate begin to be generated. The generation rate of CO, HCOO, C₂H₄, and CH₄ increase significantly with the negative shift of the applied potential (from −1.12 V to −1.35 V vs. SHE). When the applied external bias is more negative than −1.35 V, C₂H₄ and CH₄ become the predominant products. However, it is worth noting that different electrodes may show different responses to the applied potential during CO₂ reduction.⁴² More detailed influencing factors in the EC reduction of CO₂ have been discussed by Gong's group and by Zhong et al.^10,42,43Additionally, Centi and coauthors reported that a suitable device also can change the reaction pathway and products.^65–67 They designed a new strategy for enhancing the electrode surface concentration of CO₂ by adding a CO₂ adsorption component to a gas diffusion layer (GDL) type electrode, which showed a significantly enhanced activity and selectively for >C₁ products. This report provided a new approach to generate >C₁ carbon products.⁶⁵

For the PEC reduction of CO₂, the conduction and valence band potential, exposed facets, external optical properties, and potential, etc. play important roles in the products distribution, FE, and conversation efficiency. The band-edge position of some typical semiconductor and redox potentials versus (vs.) NHE of CO₂ reduction at pH = 7 are shown in Fig. 1.³⁸ From Table 1 and Fig. 1, we can see that all the candidate semiconductors listed here cannot carry out the one electron reduction of CO₂ to CO₂˙⁻ (about −1.9 V vs. NHE, pH = 0), because they do not have enough driving force, as their conduction band edge are below the potential of CO₂ to CO₂˙⁻. Therefore, another mechanistic route has been proposed, which is through a series of multiple proton-coupled electron transfers to form CO₂˙⁻ radicals. However, it still needs to overcome high activation barriers. It has been reported that the higher concentration of available protons and the partial electron density on the surface of catalysts would promote the reaction.⁴² In addition, a suitable heterojunction will benefit the CO₂ reduction because the heterojunction can facilitate the transportation of electrons and holes, which can lower the overpotential toward CO₂ reduction and improve the charge-carrier separation. Moreover, the surface state may also be modified due to the formation of a heterojunction, which may promote the adsorption and activation of CO₂. Except for the driving force for CO₂ reduction, the mechanism and pathways of CO₂ reduction are generally similar for EC reduction and PEC reduction processes.

Fig. 1 Band-edge positions of some typical semiconductor photocatalysts relative to the energy levels of the redox couples involved in the reduction of CO₂.³⁸ Copyright 2016 Royal Society of Chemistry.

3 Electrochemical reduction of CO₂

Among all the various approaches for reducing CO₂ into low carbon chemicals for fuel purposes, the methodology of EC reduction is the simplest one that can work on a huge scale under green conditions.^50,68 The electricity produced from renewable sources, such as from solar and wind power, is used to drive the EC reduction processes.⁶⁹ CO₂ has high stability and very complicated kinetics from multiple electron transformations, which are a hindrance to overcoming the high energy barrier.⁷⁰ For CO₂ EC reduction, a variety of electrocatalysts, including metals, metals oxides, metallic nanostructures, and molecular complexes, have been explored.^71–75 As their activities for huge-scale application are still not active enough, the biggest challenge during CO₂ EC reduction reaction is to improve the low activity, poor efficiency, selectivity, and yield of products.^32,76 In the field of CO₂ EC reduction, a lot of the research efforts have focused on the development of the best electrocatalyst in order to overcome the above challenges. So in the following section of the review, we summarize several groups of electrocatalysts, such as metallic nanostructures, metallic alloys, metal oxides, and metal-free electrodes, for CO₂ reduction.

3.1 Metallic nanostructures

In the early 1980s, introductory research in the field of CO₂ reduction was carried out with the catalytic performance of bulky metallic catalysts. For this, some single metallic elements were examined. The products obtained after the CO₂ EC reduction depend upon various factors, like the kind of materials used as the cathode, the electrolyte, and also on the voltage, temperature, and pressure. In order to improve the CO₂ EC reduction efficiency, nanostructured metallic materials have attracted much attention because of their versatile properties. These metallic nanomaterials have a high surface area and contain a large portion of edges or low coordinated sites. Due to these properties, they show different catalytic behaviors compared to the bulk materials.^43,77,78 This section describes the recent progress in various metallic nanostructures for CO₂ EC reduction.

3.1.1 Silver nanostructures. For EC CO₂ reduction, Ag and its composites have been widely used due to their several advantages, like relatively low material cost, high catalytic activity, and stability. Ag nanostructures co-catalyzed by ionic liquids have the ability to reduce CO₂ into CO with a good FE of 96% at an overpotential of less than 0.2 V. The shortcomings of this system, however, are its moisture sensitivity and the high cost of ionic liquids, which would prevent the scalability of the catalysts.^79–81 For Ag nanoparticles (Ag NPs) electrodes, Salehi-Khojin et al. demonstrated that size has influences the catalytic activity of Ag NPs. By decreasing the size, there is an increase in activity but this happens to some extent at particular sizes of about less than 5 nm, where this trend is changed dramatically. As compared to bulk Ag, the rate of reduction is 10 times higher than for 5 nm Ag NPs. The authors ascribed that smaller particles have more binding energies as compared to bulk materials, as the intermediates formed have more stability against further reduction. However, when the particle size becomes too small, the intermediates bind more strongly to the surface, making the product release more difficult. So in an ideal scenario, the electrocatalyst should have an appropriate intermediate binding strength, neither very strong nor very weak.⁸² Recently, Lu et al. synthesized a Ag nanoporous electrode by applying the dealloying arc melting technique.⁷⁹ The resulting electrode showed an enhanced EC surface area with a larger number of active surface sites as compared to polycrystalline Ag. Due to these properties, Ag nanoporous electrodes have the ability to produce CO with a high FE of 92% and a current density of 18 mA cm⁻² at −1.6 V (vs. RHE) in 0.5 M KHCO₃ electrolyte. The structural morphology and electrocatalytic performance of nanoporous Ag is shown in Fig. 2. Although Ag nanoporous electrodes have promising ability toward CO₂ EC reduction, this process does not have much potential for large-scale applications due to the high cost of the arc melting method, which normally requires a high temperature.⁷⁹ Hsieh et al. fabricated a coral-like Ag nanostructured electrocatalyst by using chloride anions through an oxidation–reduction method, showing a very high catalytic activity toward CO₂ reduction.⁸³ The electrode also showed a higher selectivity for CO production and an FE of 95%. In this work, the authors demonstrated that the presence of chloride anions on the surface of the catalysts played a significance role toward CO₂ reduction. The presence of chloride ions on the surface of Ag nanocorals enhanced the specific activity 32 times as compared to polycrystalline Ag. Kim et al. reported that the introduction of a small organic molecule on the surface of a Ag nanoparticle make it more efficient toward CO₂ EC reduction. In order to control the catalytic activity toward CO₂ reduction, the authors performed density functional theory (DFT) calculations. They prepared three types of Ag nanoparticles by using different surface capping agents including oleylamine, oleic acid, and dodecanethiol having amine, carboxyl, and thiol functional groups. These three modified nanoparticles showed different catalytic activity toward CO₂ reduction into CO. Ag nanoparticles capped with amines showed the highest FE (94.2% ± 1.5%) due to an exceptional suppression of HER. Ag nanoparticles having a thiol showed poor selectivity for CO production but a prominent increase in activity was found for both HER and CO₂ reduction. The presence of functional groups on the surface of Ag is responsible for enhancing its catalytic activity toward CO₂ reduction.⁸⁴

Fig. 2 (a) The highly internal surface morphology of a nanopore of a Ag electrocatalyst (b), (c) SEM and TEM images of nanoporous Ag (for b and c, the scale bar is 500 and 2 nm, respectively). (e)–(g) Electrocatalytic activity of nanoporous Ag and polycrystalline Ag at −0.60 V, and np–Ag at −0.50 V, and at −0.40 V verses RHE, total current density and FE verses time on the left axis and right axis, respectively.⁷⁹ Copyright 2014 Nature Publishing Group.

Very recently, Daiyan et al. fabricated a three-dimensional (3D) nanoporous Ag electrode termed Ag foam by using a simple electrodeposition method.⁸⁵ Ag foam has the capability toward the selective reduction of CO₂ into CO with a high FE of 94.7% and a current density of 10.8 mA cm⁻² at −0.99 V vs. RHE. Due to its unique 3D nanoporous structure, Ag foam has shown amazing activities with a higher FE and current density. In this particular work, the electrocatalytic activity of Ag foam was investigated in a 0.1 M solution of KHCO₃ for CO₂ EC reduction, with the Ag foam directly used as the working electrode without any further treatment. The comparison between Ag foam and other high performance Ag electrodes showed that it had the highest FE efficiency among the others, including a high-cost organometallic Ag electrode (90% FE at −1.1), polycrystalline Ag foil (90% FE at −1.1 V), immobilized Ag nanoparticle (80% FE at −0.75 V), and nanostructured Ag foil (83.5% FE at −1.04 V).⁸⁵ From this research, it was concluded that the 3D porous Ag foam electrode has more EC active sites, and that these dramatically increase the contact between the electrolyte and electrode. We can thus conclude that the highly porous structure leads to the more effective EC reduction of CO₂.

3.1.2 Gold nanostructures. To date, Au has attracted much attention due to its ability of converting CO₂ into CO with high selectivity and efficiency; however, for huge-scale application, the biggest hindrance is its high cost and relative scarcity.⁸⁶ Recently, researchers have focused on nanostructured Au catalysts, which have shown increased catalytic activity and higher conversion efficiency as compared to polycrystalline or single crystalline surfaces.^87,88 Peterson et al. used Au NPs with different sizes of 4, 6, 8, and 10 nm for CO₂ EC reduction, where the smaller sized particle of 4 nm showed the highest CO partial mass activity (J_CO) of 14 A g⁻¹ in 0.5 M KHCO₃ at −0.9 V verses RHE, and produced the overall largest mass activity.⁸⁹ The particles with a smaller size produced a larger amount of H₂, where nanoparticles of 8 nm size showed the highest FE. In this work, the authors estimated the size-dependent ratio of edges, facets, and surface corner sites of the Au NPs by using structural models. It was also observed by DFT modeling that the excessive evolution of H₂ on smaller nanoparticles is because of the stronger binding of H* at the underneath coordinating corner positions, whereas the greater evolution of CO on larger nanoparticles is due to the stronger *COOH binding at the edge position. Au NPs with 8 nm size exhibited the highest CO FE because of having the optimal amount of edge sites against the corner sites. The edge sites of Au NPs have more selectivity for the reduction of CO₂ into CO as compared to corner sites. A more recent study reported that the formation of H₂ occurs because the corner sites prefer to reduce a proton from water.⁹⁰ Peterson et al. suggested that Au nanowires would have more edge sites as compared to Au NPs, whereby they synthesized 2 nm Au nanowires (NWs) with different lengths of 15, 100, and 500 nm.⁹¹ The longer NWs had a higher friction of edge sites as compared to smaller ones, whereby the former showed a higher FE, mass activity, and lower onset overpotential for CO₂ EC reduction. In 0.5 M KHCO₃ electrolyte at 0.35 V verses RHE, the Au NWs with 500 nm length produced a high mass activity of approximately 6.5 A g⁻¹, which was larger than the previously reported values for Au NPs of 4–8 nm size at the same voltage.⁸⁹ Besides, Mistry et al. also studied the influence of Au nanoparticles on the products from CO₂ reduction.⁹² They found that the H₂ : CO ratio can be tailored by changing the size of the catalyst particles. Koh et al., for the first time, synthesized Au nanostructured islands by a single-step oxygen plasma-induced method, in which thermal annealing simultaneously occurred to prevent the formation of gold oxide.⁹³ The physical and chemical properties, including the crystal structure and oxidative state of the nanostructured Au electrode, was explored by X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS). The CO₂ EC reduction was performed in an acetonitrile–water electrolyte solution. They showed a higher selectivity for CO with a 95% FE and enhanced current density. As compared to polycrystalline Au, the production rate, onset potential, and FE were significantly improved by the nanostructured Au island electrode. The authors further demonstrated that, in the presence of 1-butyl-3-methylimidazolium tetrafluoroborate ([BMIM] BF₄), the CO₂ reduction activity was improved on Au islands as compared to that without [BMIM] BF₄. CO and H₂ were confirmed as the main products by gas chromatography, and the formic acid concentration was found to be below the detection limited. For bare Au, Au islands, and Au with ([BMIM] BF₄), the FE and CO rate of production plotted against voltage (vs. NHE) are illustrated in Fig. 3.

Fig. 3 (a) Schematic representation of the fabrication of Au islands prepared by oxygen plasma treatment (b) (c) linear sweep voltammograms (LSV) for the bare Au and Au islands in acetonitrile with or without 0.1 M [BMIM]BF₄ and of bare Au and a Au islands electrode prepared under various oxygen plasma treatment times (d), (e) FE and products rate of CO measured on the bare Au and Au islands (60 min) in acetonitrile with or without 0.1 M [BMIM]BF₄.⁹³ Copyright 2015 American Chemical Society.

Sargent et al. described Au nanoneedles with a high current density for CO₂ EC reduction,^21,94 and further predicted, by utilizing DFT calculations, that the berries could be reduced by the adsorption of K⁺ ions during CO₂ reduction at various Au facets. The authors explored the CO₂ reduction activity of Au needles, particles, and rods in 0.5 M KHCO₃. In the range of −0.30 V to −0.50 V, they found that on Au needles, the voltammetry curve displayed a distinct reduction peak, while Au rods and particles showed plane current–voltage curves. The FE for CO was approximately >95% on Au needles, while on Au rods and particles it was about 25% and 3%. The Au needles showed great stability during an 8 h electrocatalytic process. The Tafel analysis showed a slope of 96 mV dec⁻¹ for Au particles, 80 mV dec⁻¹ for Au rods, and 42 mV dec⁻¹ for Au needles. The significantly smaller Tafel slope indicated a faster kinetics of first electron transfer and confirmed that the Au needles have a higher activity for CO₂ reduction. Additionally, Hall et al. designed a Au inverse opal (Au–IO) film for CO₂ reduction, which showed high activity for CO at modest overpotentials and a high selectivity relative to hydrogen evolution (HER).⁹⁵ They demonstrated that the origin of hydrogen suppression in Au–IO films was derived from the generation of diffusional gradients within the pores of the electrodes rather than from changes in surface faceting or in the Au grain size. As a result, the activity for HER decreased by 10-fold with increasing the porous film thickness, while the CO evolution activity was largely unchanged. For the optimal mesostructured electrodes, 99% selectivity for CO could be obtained at an overpotential as low as 0.4 V.

3.1.3 Copper nanostructures. Copper (Cu) is one of most studied metal electrocatalysts for CO₂ reduction because of its unique properties. It has the ability to reduce CO₂ into a variety of hydrocarbon products that can be used as fuels.^96–98 Reske and coworkers explored the size effect of Cu NPs on the catalytic activity and products selectivity for CO₂ EC reduction.⁹⁹ An unexpected selectivity and activity variations were observed on Cu NPs with a size ranging from 2 to 15 nm. The Cu NPs showed a significantly higher catalytic activity as compared to Cu foil. Cu NPs with a size less than 5 nm showed dramatically increased catalytic selectivity and activity for H₂ and CO production. The smallest particle with a size of 1.9 nm exhibited a 100% increase in current density (−48 mA cm⁻²) compared to a Cu foil electrode. It has been reported that Cu nanoparticles (Cu NPs) with a smaller size have more under-coordinated sites. The results from electrocatalytic reduction experiments in CO₂-saturated 0.1 M KHCO₃ at −1.1 V vs. RHE revealed that the product distribution was highly size dependent. The products on bulk Cu were CH₄ in a large amount along with H₂ and C₂H₄ in smaller quantities together with a very few traces of CO. The Cu NPs on the other hand produced H₂ with 60–70% FE, CO with a smaller FE of 20–30%, and C₂H₆ and CH₄ with FEs of less than 10% and 10–15%, respectively. It was hypothesized by the authors that Cu NPs with a high amount of surface under-coordinated sites favorably evolved H₂.

In order to improve the catalytic activity of the Cu electrode, different kinds of Cu nanostructured have been fabricated, including nanoparticles of zero-dimensional (0D), one-dimensional (1D), and two-dimensional (2D) overlayers and three-dimensional (3D) foams. Their effects on CO₂ EC reduction have been explored systematically. An enhanced catalytic activity and selectivity were observed on 0D, 2D, and 3D nanoparticles.⁴³ Recently, Wang et al. fabricated a more innovative electrocatalyst of Cu nanowires for CO₂ reduction.¹⁰⁰ The EC reduction of CO₂ was performed in an electrocatalytic cell with a separating membrane between the cathode and anode in 0.1 M KHCO₃.¹⁰⁰ Different characterization techniques were used for gas and liquid phase samples, including gas chromatography (GC) with a mass spectrometer (MS) detector and nuclear magnetic resonance (NMR) spectroscopy. The catalytic activity of Cu nanowires was reported in comparison with different previously reported Cu nanomaterials. The authors demonstrated that Cu nanowires produced by EC reduction processes have a highly advanced catalytic ability for CO₂ reduction. Under 0.3 V overpotential, the current density could reach 1 mA cm⁻² with an FE of 60% for CO production. Recently, Yang et al. reported ultrathin (diameter ∼20 nm) 5-fold twinned copper nanowires (Cu NWs) for CO₂ reduction.¹⁰¹ They showed a higher selectivity for CH₄ with a 55% FE over other hydrocarbons at −1.25 V vs. RHE, and this was attributed to the presence of a high density of wedge edge sites owing to the twin boundaries. The authors further explored wrapping the wires with graphene oxide, which was found to be a useful strategy for preserving both the morphology and reaction selectivity of the Cu NWs. Fig. 4 illustrates the morphological and electrocatalytic characterization of a Cu NWs electrocatalyst.

Fig. 4 TEM images of Cu NWs: (a) bare, (b) loaded with carbon black. (c) SEM image of Cu NWs/CB dispersed on glassy carbon black. Electrocatalytic and morphological characterization of Cu NW catalysts: (d) rapid formation of C₂H₄ with a concurrent decrease in CH₄ activity after the first Coulomb of electrolysis, (e) TEM images of Cu NWs after passing a given amount of charge (f), (g) analogous morphological and electrocatalytic characterization for the RGO–Cu NW catalysts showing preservation of the selectivity and morphology.¹⁰¹ Copyright 2017 American Chemical Society.

Robert and coworkers fabricated copper nanocubes for CO₂ electrochemical reduction.¹⁰² The Cu nanocubes showed a higher selectivity for ethylene at a very low overpotential. CO₂ could be reduced into multiple hydrocarbon over a Cu nanocube electrode by favoring C–C combination instead of completely reducing it into methane. This could be an important step toward developing catalysts with selectivity for multicarbon products. Very recently, Buonsanti et al. explored the activity and product selectivity of Cu nanocrystals (NCs) toward CO₂ reduction with assessing also the effects of size and shape simultaneously.¹⁰³ Through colloidal-based chemistry, Cu NCs of sphere and cube shaped were fabricated. They found that cube shaped Cu NCs were intrinsically more active as compared to sphere shaped Cu NCs. Fig. 5(a) illustrates the TEM images and XRD pattern of Cu cubes and Cu spheres. The FE of Cu NCs for different products in comparison with Cu foil is shown in Fig. 5(b). It was found that Cu cubes with sizes of 24 nm and 44 nm and the smallest sphere sized 7.5 nm exhibited 80% higher selectivity toward the CO₂ reduction reaction and 41% FE for ethylene.

Fig. 5 (a) TEM images of Cu cubes (A–C) and Cu spheres (D–E). (F) Typical XRD patterns of Cu cubes and spheres. (b) The bar graph exhibits the FE at −1.1 V vs. RHE for each product in the different sizes of Cu NC cubes, and on Cu foil, with glassy carbon subtracted.¹⁰³ Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

Moreover, for enhancing the CO₂ reduction activity, in addition to controlling the size of the NP grains, the boundary density has also been proposed as a useful approach for increasing the low coordination sites. A large atomic disorder has been suggested to enhance the catalytic activity of many reactions, particularly in the field of organic chemistry and thermal processes.¹⁰⁴ Kanan et al.¹⁰⁵ fabricated Cu NPs on carbon nanotubes (Cu/CNTs) with different average grain boundaries (GBs), which were quantified by transmission electron microscopy (TEM). At potentials ranging from −0.3 V to −0.5 V vs. RHE, the specific reduction activity of CO was linearly proportional to the fraction of surfaces comprising GBs of NPs. The products were ethanol (25–35% FE), acetate (10–30% FE), and small amounts of C₂H₄, C₂H₆, and propanol (<10% FE). Their work showed that at a moderate overpotential, the CO reduction activity on Cu NPs was correlated to the density of GBs. Bell et al. attributed the higher selectivity of CuNPs for C₂H₄ due to the high density of Cu grain boundaries.¹⁰⁶ It has also been revealed by several other research groups that the EC activity is directly related to the grain boundary density on the electrode surface.^107–109

3.1.4 Other metallic nanostructures. Some other metallic nanostructures have also been investigated, including In,¹¹⁰ Zn,⁷⁵ Pd, Pb, Sn,¹¹¹ and Pt,^112,113 for CO₂ EC reduction. Palladium (Pd) nanostructures have been paid most attention as it was reported by Gao et al.¹¹⁴ that CO₂ EC reduction is greatly influenced by the size of the Pd nanostructures. Pd NPs with a smaller size showed a higher FE and partial current density for CO compared to larger ones. Recently, the work of Rosen et al. on zinc (Zn) dendrite nanostructures developed a great interest among researchers toward various Zn electrodes,⁷⁵ especially as Zn can meet the demands for large-scale application because of its low cost and relative abundance. Zn dendrite nanostructures have shown 80% FE for CO. Zhang and coworkers explored a new Zn electrocatalyst with a layer of nanoparticles that showed a high catalytic activity toward CO₂ reduction into formate with an FE of 87%, which was much higher than that of bulk Zn foil.¹¹⁵ It has been reported that nanoparticles of Pt with graphene as nanocomposites exhibited excellent electrocatalytic properties for CO₂ EC reduction.¹¹⁶

3.2 Metal alloys

For the EC reduction of CO₂, the biggest challenge is the fabrication of a catalyst with high activity and selectivity at low cost. During the reduction processes of CO₂, the binding strength of the intermediates play a vital role in the reaction activity and product selectivity.⁵⁵ Nowadays, metal alloys have drawn great attention because they can boost the electrocatalytic activity of metallic catalysts. They were also found to have the ability to improve the stability of the key intermediates.^117,118 Cu has been demonstrated as one of the most promising candidates for CO₂ reduction. However, the products selectivity and durability still need to be improved.¹¹⁹ Alloying is a promising approach to enhance the products selectivity and the FE and to lessen the overpotential. Various alloys of Cu have been reported, such as Cu–Ni, Cu–Sn, Cu–Pb, Cu–Zn, Cu–Cd, Cu–In, and Cu–Ag, for the reduction of CO₂ in 0.05 M KHCO₃ aqueous solution.¹²⁰

Recently, Cu–Pt monodispersed nanocrystals (NCs) with different molar ratios were fabricated by Guo et al.¹²¹ The change in the molar concentration of the precursor Cu and Pt had no noticeable influence on the morphology of the Cu–Pt NCs. The authors observed that the composition-dependent activity of Cu–Pt NCs with increasing the Cu content in the Cu–Pt NCs facilitated the reduction of CO₂, leading to enhanced catalytic activity. Due to the composition-dependent catalytic activity, the Cu–Pt (3 : 1) NCs exhibited the highest activity and FE for CH₄ at room temperature in 0.5 M KHCO₃. On the basis of experimental analysis, the authors proposed a possible mechanism for CO₂ EC reduction at the Cu–Pt NCs. Also, DFT calculations for the reduction of CO₂ to CH₄ have been clearly reported by Peterson et al.¹²² According to the proposed reaction mechanism and DFT calculations, the key step is the protonation of adsorbed CO* to form HCO* because of Pt, which shows high activity for HER. The equation: CO* + H⁺ + e⁻ → HCO* was proposed for the reduction of CO to CHO. This step shows the selectivity of CO₂ electroreduction on Cu, which ultimately leads to the formation of CH₄ due to the synergistic effect between Cu and Pt. The probable reaction mechanism can be expressed by the equation: CO* + H⁺/H₂O + e⁻ → HCO* → CH₄.

Zhao and coworkers fabricated Au₃Cu alloy NCs by a phase stabilization method,¹²³ which showed enhanced catalytic activity and selectivity toward CO₂ EC reduction as compared to nanocrystals of monometallic Au. The Au₃Cu alloy NCs endorsed the production of other hydrocarbons too, like CH₄ and C₂H₄, with a 30% FE, as Au NPs favored the production of CO. The authors proposed that this methodology of fabricating alloys could provide a more convenient route to obtaining different alloy NCs. Very recently, Sarfraz et al. reported an electrocatalyst comprising Cu and Sn with a great potential for CO₂ EC reduction¹²⁴ and showing high activity and selectivity over a wide range of potentials. This bimetallic electrode was fabricated by the deposition of Sn on oxide-derived copper (OD–Cu). It showed a maximum selectivity for CO with an FE greater than 90% at a potential range from −0.4 to −0.8 V vs. RHE. The current density, FE comparison, and SEM images of the Cu–Sn before and after CO₂ reduction are shown in Fig. 6. The elemental analysis showed that the presence of Sn with Cu enhanced its selectivity for CO, as monometallic Cu can promote H₂ production. Wang et al.¹¹⁸ reported carbon-based palladium–copper (PdCu/C) alloy nanoparticles (<5 nm) as an efficient electrocatalyst for CO₂ EC reduction. Due to the synergistic effects, CuPd exhibited excellent activity toward CO₂ reduction and a higher selectivity for CO. Among the different catalysts of PdCu/C and Pd/C, Pd₈₅Cu₁₅/C showed the highest catalytic activity in CO₂-saturated 0.1 M KHCO₃ solution and the highest FE of 86% with a current density 6.9 mA cm⁻² and mass activity of 4.5 A g⁻¹ for CO production at −0.89 V vs. RHE. X-ray absorption fine structure (EXAFS) analysis and temperature programmed desorption-mass spectrometry (TPD-MS) results revealed that the higher selectivity of Pd₈₅Cu₁₅/C for CO is due to it having low coordination sites and an optimum ratio of copper element. The authors demonstrated that bimetallic Pd–Cu alloys have a great potential for selective CO₂ EC reduction, and that a variety of bimetallic alloy nanoparticle catalysts can be fabricated in order to achieve a higher selectivity for CO₂ reduction. The Cu alloys exhibited a distinct behavior for CO₂ EC reduction as compared to pure metals. The results obtained utilizing different combinations of Cu with various metals opened up a new direction of research for developing Cu-based electrocatalysts.

Fig. 6 (a) FE representation of 0D-Cu (b) Cu–Sn, and (c) Sn deposited on the Sn sheet at applied potentials; the FE is indicated as an average and calculated at the steady-state current and product concentration. (d), (e) SEM images of Cu–Sn before and after 1 h CO₂ reduction at −0.6 V vs. RHE.¹²⁴ Copyright 2016 American Chemical Society.

Besides Cu-based alloys, there are some other metallic alloys applied in CO₂ reduction. Recently, Kim et al.¹²⁵ fabricated Sn–Zn alloys with various compositions by an electrodeposition method. The Sn–Zn electrocatalyst showed the highest selectivity for formic acid (HCOOH) formation. Among all the studied alloys in that work, Sn_0.91Zn_0.09 exhibited the maximum production and conversion efficiencies for HCOOH. The authors demonstrated that the alloying of Zn on Sn resulted in a modification of its surface morphology and electronic surface, which led to an enhanced activity and selectivity for CO₂ EC reduction. It was hypothesized by the authors that Zn–Sn can be further enhanced by increasing the surface area, but this requires new strategies. Very recently, Park et al.¹²⁶ reported the CO₂ EC reduction to formate by using tin-lead (Sn–Pb) alloys can lead to a novel electrocatalyst. Such Sn–Pb alloy electrocatalysts were prepared with different compositions by electrodepositing on the surface of carbon paper. The results obtained by performing electroreduction in an H-type cell at −0.2 V vs. Ag/AgCl showed a 16% FE and 25% partial current density for formate production, which are both higher values than for the pure metallic electrode. The maximum FE (79.8%) and partial current density (45.7 mA cm⁻²) was exhibited by an alloy with the surface composition of Sn_56.3Pb₄₃. The results of cyclic voltammetry (CV) and X-ray photoelectron spectroscopy (XPS) analyses revealed that, on the surface of the alloys, Sn was oxidized to SnO_x, which enhanced the catalytic activity and electrical conductivity of the Sn–Pb alloys toward CO₂ EC reduction to HCOO⁻.

Jedidi et al. successfully prepared Cu–In alloy surfaces by the reduction of mixed metal oxides of CuInO₂, which showed good activity for electrocatalytic CO₂ reduction.¹²⁷ They combined experimental results and density functional theory calculations (DFT) to demonstrate that In site occupation occurs on Cu step sites and that these In sites could hinder the adsorption of H and CO. These synergistic effects accounted for an improved selectivity for CO and HCOOH products. Some differences with the research reported by Jedidi et al. and Rasul et al. showed that the In was preferentially located on the edge sites rather than on the corner or flat sites.¹²⁸ Additionally, DFT calculations also showed that the d-electron nature of Cu remained intact, but the adsorption properties of neighboring Cu was perturbed by the presence of In. Moreover, the prepared Cu–In bimetallic electrocatalyst showed nearly 95% FE of CO under a −0.7 V vs. RHE condition. Recently, Larrazábal and coauthors prepared Cu–In nanoalloy electrocatalysts by an in situ reduction of CuInO₂⁻ and In₂O₃-supported Cu nanoparticles.¹²⁹ This study showed that In(OH)₃ played a key role in CO formation, which indicated that catalysts with controlled interfaces would be beneficial for the desired reaction.

3.3 Chalcogenides

3.3.1 Metal oxides. Along with metallic electrodes, different kinds of metal oxides, such as FeO_x, Cu₂O SnO₂, CeO₂, TiO₂, Co₃O₄, and Ga₂O₃, have been investigated as potential electrocatalysts for CO₂ reduction as well.^130–132 Tin oxide (SnO₂) has been investigated as an effective electrocatalyst for CO₂ reduction. Meyer and coworkers fabricated NPs of SnO₂ with a 5 nm average diameter utilizing a hydrothermal method. To increase the surface area and facilitate the CO₂ reduction process, the NPs were loaded onto carbon black and graphene. The SnO₂ NPs with carbon black showed a high selectivity for formate at −1.8 V with a current density of 6.2 mA cm⁻² during electrocatalytic reduction in 0.1 M NaHCO₃ aqueous solution. By replacing carbon black with graphene a maximum FE of 93% with 10 mA cm⁻² current density for formate production was observed.¹³³ Bocarsly and coworkers explored the mechanism of CO₂ EC reduction on the Sn/SnO_x catalysts by using in situ attenuated total reflectance infrared spectroscopy (ATR-IR).¹³⁴ They proposed that the reduction of CO₂ is preceded by a two-electron reduction of the native SnO₂ to Sn(II) oxyhydroxide. Then, a surface-bound tin bicarbonates intermediate was formed, which was further reduced to formate. The authors elucidated the role of the carbonates and surface oxides, and proposed that the SnO₂ nanoparticles are more active for CO₂ EC reduction.¹³⁴ Recently, Li et al. developed 3D mesoporous SnO₂ nanosheets on carbon cloth as an efficient electrocatalyst toward CO₂ reduction,¹³⁵ which also showed a higher selectivity and efficiency toward formate production, with 87% FE and 45 mA cm⁻² current density at a 0.88 V overpotential. The catalysts exhibited higher stability and flexibility due to the porous hierarchical structure. Fig. 7 illustrates the structural morphology and electrocatalytic performance of the mesoporous SnO₂ electrocatalyst. The research on 3D hierarchical structured electrodes provides a new option for the development of highly efficient electrocatalysts.¹³⁵

Fig. 7 (a), (b) SEM images of SnS₂ nanosheets on carbon cloth (CC) and porous SnO₂ nanosheets on CC. (c) XRD patterns of SnS₂ and SnO₂ nanosheets. (d) Current density of the SnO₂/CC electrode in CO₂ saturated 0.5 M NaHCO₃ solution at different applied potentials. (e) The corresponding FE for CO, H₂, and formate (f) Tafel slope of the SnO₂/CC electrode.¹³⁵ Copyright 2017 Wiley-VCH Verlag GmbH & Co. KGaA.

Another transition metal oxide, MoO₂, has drawn attention for the EC reduction of CO₂. Oh et al.¹³⁶ reported the highest catalytic activity of MoO₂ toward CO₂ EC reduction in acetonitrile with 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF6) as an organic solvent. The highest activity was observed at −20 °C and the addition of a small amount of water promoted the performance. After that, the same group reported an enhanced catalytic activity of MoO₂ in several ionic liquids based on imidazolium acting as cocatalysts and an electrolyte simultaneously. The catalytic activity of MoO₂ toward CO₂ reduction was enhanced by the addition of BMIN-PF₆ in acetonitrile. The improvement in catalytic activity and the change in product selectivity proved the combined effect of the ionic liquid acting as both an electrolyte and as a cocatalyst.¹³⁷

Some other metal oxide electrodes have been reported by various research groups. Kim et al.¹³⁸ reported a cuprous oxide (Cu₂O) electrode for the highly effective conversion of CO₂ to lower hydrocarbon fuels and also reported its higher selectivity for ethylene (C₂H₄). In this study, the authors explored the formation of oxygen vacant structures during the electrolysis processes, which led to a higher production of C₂H₄. For the selective conversion of CO₂ into larger oxygenate molecules, the oxide-derived metallic layer and surface oxide coverage were the key factors. Great efforts have been made in order to obtain higher carbon chain products.^139–142 Centi and coworkers investigated the behavior of Fe, Co, Cu (mono-, bi-, and trimetallic) carbon nanotube (CNT) electrocatalysts for CO₂ reduction and obtained C₁–C₃ products. The research showed that the inhibition of surface reactivity toward H₂ formation plays a key role in improving the performance and for realizing selective electrocatalysts.¹⁴² Recently, they deposited ultrafine Cu NPs (less than 3 nm) on the CNTs, which showed a high activity for C₂–C₃ products (ethanol, acetic acid, isopropanol, and acetone) due to the interfacial catalytic sites between Cu NPs and CNTs.¹³⁹ Genovese et al. investigated the critical step of C–C bond formation by comparing the reactivity in tests with CO₂ and CO, in which formic acid or formaldehyde was initially added.¹⁴⁰ The results demonstrated that dissolved CO₂ was needed for acetic acid formation, which showed a different pathway toward formic acid.

Sekimato et al. developed Sn- or Si-doped gallium oxide (Ga₂O₃) cathodes for the selective conversion of CO₂ into HCOOH and showed a 80% FE for HCOOH.¹⁴³ The DFT results showed that the higher selectivity of Ga₂O₃ for HCOOH production occurred because of the stable intermediate on the surface. This study revealed that compound catalysts could be more efficient catalysts for CO₂ EC reduction. Ramesha et al.¹⁴⁴ used TiO₂ nanostructures for CO₂ EC reduction, with methanol observed as the primary product. Ullah et al.¹⁴⁵ reported an indium/ruthenium oxide (Ir/Ru oxide) electrocatalyst for CO₂ reduction into various hydrocarbons with FEs of 85% and 96% for ethanol at 295 K and 277 K, respectively. It was demonstrated that for CO₂ conversion into hydrocarbons, it can be considered as a highly efficient electrode.

3.3.2 MoS₂. Transition metal dichalcogenides (TMDs) have attracted significant interest owing to their fascinating mechanical, electrical, and optical properties.¹⁴⁶ MoS₂, MoSe₂, and WS₂ have been widely investigated as efficient electrocatalysts for CO₂ reduction. Salehi-Khojin et al.¹⁴⁷ first explored the electrocatalytic activity of MoS₂ for selectively reducing CO₂ into CO at a 54 mV lower overpotential in a mixture of 96% water and 4% EMIM-BF₄. This system showed a much improved FE of 98% and a current density of ≈65 mA cm⁻² for CO compared to Ag and Au NPs. In 2016, the same authors compared the electrocatalytic activity of MoS₂, MoSe₂, WS₂, and WSe₂ for the four different TMD materials with the same size. Among all the WSe₂ forms, nanoflakes exhibited an exceptional current density of ≈19 mA cm⁻², and a 24% FE for CO and a TOF of 0.28 s⁻ at η = 54 mV (ref. 1) in a 50 : 50 vol% ionic liquid/water electrolyte. At η = 0.65 V, the current density for WSe₂ reached 330 mA cm⁻² with an FE of ≈85%. The DFT calculations revealed that on TMD edges, the CO was stabilized, and the formation of COOH was exergonic unlike on transition metals surfaces, indicating that the conversion of CO₂ into CO was kinetically favorable.¹⁴⁸ Recently, the same research group reported that the catalytic activity of TMD materials can be further enhanced by adding an appropriate dopant.¹⁴⁹ MoS₂ doped with 5% Nb exhibited the lowest onset potential of 31 mV. These catalysts showed a TOF for CO one order of magnitude higher compared to the pristine form at an overpotential range of 50−150 mV and two orders of higher magnitude than that of Ag nanoparticles at the potential range of 100−650 mV. The DFT calculations revealed that low concentrations of Nb near the Mo edge atoms could enhance the TOF of CO formation by modifying the binding energies of the intermediates to MoS₂ edge atoms. This work indicated that embedding a dopant into the atomic structure of the catalysts could open a promising route to enhance the catalytic performance of TMD materials.

Nørskov et al.¹⁵⁰ explored the active edge sites of MoS₂, MoSe₂, and Ni-doped MoS₂ by using DFT calculations. These edges showed the unique property of the selective binding of COOH and CHO to S or Se atoms and CO to the metal atom. The DFT calculations suggested a significant improvement in CO₂ reduction activity over the transition metals. Furthermore, the same authors found that the metal-doped MoS₂ provided two different linear scaling relationships. This was due to a combination of dopant metal sites for *CO binding and S binding for *COOH, *CHO, and *COH. This synergistic result enhanced the CO₂ reduction performance compared to pristine Mo metal.¹⁵¹ Han et al.¹⁵² reported a Mo–Bi bimetallic chalcogenide (BMC) as an efficient electrocatalyst for CO₂ reduction into methanol. The Mo–Bi BMC showed a higher FE of 71.2% with a current density of 12.1 mA cm⁻² at −0.7 V vs. SHE for methanol. The higher catalytic activity of the electrode resulted from the excellent synergistic effect of Mo and Bi, where the formation of CO from CO₂ was favored by Bi and Mo, which further promoted the production of methanol by the hydrogenation of CO. In the future research into CO₂ fixation, more efforts should be devoted to metal chalcogenides.

3.4 Metal-free catalysts

Renewable metal-free electrodes could be a next generation catalyst for CO₂ EC reduction because of their distinct properties, such as low cost and flexibility.¹⁵³ Recently, nanostructured carbon materials, like S, N-doped carbon nanofibers, carbon nanotubes, fullerenes, and graphene, have been studied for CO₂ EC reduction.⁵⁵ These nanostructured carbon materials have different properties, including high stability, outstanding conductivity, and mechanical strength, as well as all being cheap and environmentally friendly. Kumar et al.¹⁵⁴ reported the catalytic activity of polyacrylonitrile-based heteroatomic carbon nanofibers (CNFs) as a low-cost metal-free electrocatalyst for CO₂ EC reduction. This metal-free electrocatalyst showed a higher efficiency for CO₂ EC reduction as compared to noble metal catalysts (e.g., Ag) at a low overpotential of 0.17 V, and also exhibited remarkable selectivity for CO production. Fig. 8 illustrates the structural characterization and electrocatalytic performance of a CNFs electrocatalyst.¹⁵⁴ The authors found that the nitrogen (N) in the CNFs did not directly participate in the CO₂ reduction processes, and that the oxidized carbon atoms were responsible for CO₂ reduction. The authors proposed a possible reaction mechanism for the CO₂ electroreduction in the electrolyte 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIM-BF₄) ionic liquid; whereby during the redox cycling process, the oxidized carbon atoms are initially reduced. Then, an intermediate EMIM–CO₂ complex is formed, which when dissolved the CO₂ molecules coordinate with EMIM cations. These further adsorb on the reduced carbon atoms, which is accompanied by CO₂ reduction to CO. Finally, the reduced carbon atoms are reoxidized to their initial state and the produced CO is released. So the metal-free catalyst CNFs possess a high catalytic proficiency due to redistribution of the spin density and charge between the doped atoms of nitrogen with the nearby carbon atoms. Recently, N-doped carbon nanotubes (NCNTs) as a metal-free electrocatalyst were developed by Wu et al.¹⁵⁵ These NCNTs showed a high efficiency, stability, and selectivity for the reduction of CO₂ into CO. The authors used liquid chemical vapor deposition (CVD) for fabrication of the NCNTs, which had a multiwall bamboo-shaped structure with a uniform distribution of N atoms of almost 5 atom%. The N content with the configuration of graphitic, pyrrolic, and pyridine are 1.24, 1.1, and 1.5 atom%, respectively. To understand the mechanism for CO₂ reduction and the role of the N dopant, the binding energies of CO and COOH were calculated by density field theory (DFT) simulations and a computational hydrogen electrode model. The pyridine N has strong binding with the COOH and at the same time eases the desorption of CO, so the higher selectivity toward CO is due to the N of pyridine. The NCNTs showed an 80% FE toward CO at −0.26 V. The authors proposed that for the commercial reduction of CO₂, NCNTs could be a good alternative to high-cost noble metallic electrocatalysts.

Fig. 8 (a) SEM image of a CNFs mat-entangled fibers and high resolution SEM images of an individual fiber (scale bar, 5 nm and 200 nm). (b) Atomic force microscopic image of an individual fiber (scale bar, 500 nm). (c) TEM image (scale bar, 5 nm) of the edge of the fiber surface. (d) Electrocatalytic performance of the CNF CVs for CO₂ reduction. (e) The absolute current density for CO₂ reduction at different electrodes (bulk Ag, Ag nanoparticles, and CNF) in pure EMIM-BF₄ electrolyte.¹⁵⁴ Copyright 2013 Nature Publishing Group.

More recently, Wu and coworkers introduced N-doped graphene quantum dots (NGQDs) as a metal-free electrocatalyst for the reduction of CO₂ into more useful products.¹⁵⁶ The NGQDs were fabricated from graphene oxide (GO) cutting and exfoliating, followed by N doping at elevated temperature and pressure in dimethylformamide (DMF) solvent. The CO₂ EC reduction was performed in 1 M KOH electrolyte, and it was shown that the overall FE efficiency increased from 18% to 90% at potentials varying from −0.26 to −0.75 V. At −0.26 V, the reduction products of CO and HCOO⁻ were observed and other different hydrocarbons and multicarbon oxygenates, including methane (CH₄), C₂H₄, C₂H₅OH, n-propanol (n-C₃H₇OH), and acetate (CH₃COO), were also formed at more negative potentials with a maximum FE of 31% for C₂H₄ as the major hydrocarbon product and 26% for multicarbon oxygenates. The structure characterization revealed that the doping of heteroatom N was responsible for the improved selectivity and activity of NGQDs. Xu et al. synthesized a series of N-doped carbon nanotube materials with tailored nitrogen contents and varying fractions of pyrrolic, pyridinic, and quaternary nitrogen species. They also studied the CO₂ reduction performance on these catalysts, which showed excellent performances with a 90% current efficiency for CO and stability over 60 h. Their research showed that the most active nitrogen chemical state for CO₂ reduction is quaternary nitrogen, followed by pyridinic and pyrrolic nitrogen, which provide researchers a strategy for designing nanocarbon catalysts for CO₂ reduction.¹⁵⁷ Recently, Ju et al. further investigated the metal (containing Mn, Fe, Co, Ni, and Cu)-nitrogen-doped carbon catalysts for CO₂ reduction. They studied their intrinsic catalytic reactivity, CO turnover frequencies, and CO faradaic efficiencies, which show that the Ni–N–C catalysts can rival Au- and Ag-based catalysts. They also provided reactivity–selectivity descriptors combining density functional theory (DFT) calculations and experimental data, which led to a greater atomic scale understanding of potential-dependent CO and hydrocarbon selectivity from the M–N_x moieties (M = Mn, Fe, Co, Ni, and Cu).¹⁵⁸

Bandosz et al. investigated polymer-derived S-doped and dual S, N-doped nanoporous carbon materials as metal-free electrocatalysts for CO₂ EC reduction.¹⁵⁹ The S, N-doped electrocatalyst showed a higher selectivity for CO and CH₄ production with FEs of 13.8% and 0.18% at −0.99 V vs. RHE, respectively. The authors demonstrated that the high catalytic performances of the heteroatom-doped carbon materials were due to the positively charged sites of the carbon matrix next to the electronegative nitrogen atoms of pyridine and sulfur atoms in the thiophene configuration. The CO production was promoted by stabilization of the intermediates. The catalytic selectivity for the production of CH₄ was improved due to the basicity of the surface and ultra-microspores. Jhong et al. developed a metal-free N-doped carbon catalyst for CO₂ EC reduction comprised of carbon nitride and multiwall carbon nanotube composites.¹⁶⁰ The morphology and electrocatalytic performance are shown in Fig. 9. This material exhibited a higher activity and selectivity for CO and H₂ production, with FEs of 98% and 2%, respectively. Also, compared to Ag nanoparticles-based catalysts, this material showed a maximum current density of 90 mA cm⁻². The mass activity and energy efficiency of up to 48% were also higher than the Ag nanoparticles reference.

Fig. 9 (a) Reconstructed 3D view of the electrode obtained from MicroCT data. (b) SEM cross-sectional image of the CN/MWCNT catalyst layer on the microporous layer of the gas diffusion electrode. (c) SEM images of the CN/MWCNT electrocatalyst layer (loading 2.39 mg cm⁻²). Electrocatalytic performance of CN/MWCNT and Ag catalysts: (d) partial current density for CO and H₂versus the cathode potential, (e) FE efficiency for CO and H₂versus the cathode potential.¹⁶⁰ Copyright 2017 John Wiley and Sons.

4 PEC reduction of CO₂

The PEC reduction of CO₂ has attracted extensive interest due to its high efficiency compared to the PC and EC approaches. However, reduction of the protons or H₂O leads to a decrease in the CO₂ conversion efficiency, production selectivity, and FE during the CO₂ reduction process. Moreover, the stability and improper band gap of the semiconductor are also potential problems, making it suffer from a limited light irradiation range and high overpotential. Therefore, various semiconductors have been designed and employed as photocathodes to convert CO₂ to other useful fuels. The PEC cells for CO₂ reduction can be categorized into three typical types, which are shown in Fig. 10.³⁸ The semiconductors can be used as photocathodes, photoanodes, or both photocathodes and photoanodes. In this review, we mainly divide semiconductor photoelectrodes into the following three categories: (1) phosphides; (2) metal oxides; and (3) metal-free photoelectrodes.

Fig. 10 Schematics of three typical two-compartment PEC cells separated by proton-exchange membranes for CO₂ reduction: (a) semiconductor as a photocathode; (b) semiconductor as a photoanode; (c) semiconductor as both a photocathode and photoanode.³⁸ Copyright 2016 Royal Society of Chemistry.

4.1 Phosphides

Since Halmann¹⁶¹ first reported P-GaP as an efficient photocathode for the reduction of CO₂, the reduction of CO₂ by the PEC method has been investigated by many researchers. Many utilized P-GaP under 365 nm wavelength illumination with an applied overpotential of −1.4 V vs. SCE, but the current efficiency was very low. Later, an FE of 80% was obtained with a P-GaP cathode at −1.0 V (vs. a standard silver electrode, NSE) in 0.5 M Na₂CO₃ solution under 8.5 atm pressure of CO₂, with a 150 W Xe lamp, as reported by Aurian-Blajeni et al.¹⁶² After many efforts, Barton and coauthors designed a program and used P-GaP to reduce CO₂ to methanol, achieving nearly a 100% FE in acidified aqueous pyridine solution, which was greater than 300 mV below the standard potential of −0.52 V vs. SCE at a pH of 5.2.¹⁶³ In 2015, Kronawitter et al. uncovered the mechanism of the high FE reported by Aurian-Blajeni etc. through experimental and theoretical calculations.¹⁶⁴ They demonstrated that a partially dissociated surface layer containing Ga–OH, H₂O and P–H species was produced when it was exposed to water, and the strongly surface hydrides contributed greatly to the low H₂ evolution efficiency. Recently, GaP as an attractive photoelectrode has been investigated by other researchers as well.^165–167 Carter's group did a lot of work on pyridine-catalyzed CO₂ reduction with GaP photoelectrodes,^166–171 and also clarified the role of pyridinium in the pyridine-catalyzed CO₂ reduction on GaP photocathodes.^168,169 In their research system, they proposed an energetically viable pathway for pyridinium reduction on the P-GaP photocathode, which led to the formation of adsorbed species that react to form adsorbed dihydropyridine, which played the role of the active catalyst. Recently, they used density function theory (DFT) calculations and experiments to demonstrate that surface hydrides were transferred to adsorbed pyridine, accompanied by aqueous proton addition, to form the active catalyst adsorbed dihydropyridine, which in turn transfers hydrides to CO₂, leading to its reduction.¹⁷¹

Another phosphide, InP, has a nearly ideal band gap for the efficient conversion of solar energy to electrical or chemical energy, and also has a good photoelectric stability.¹⁷² In 1984, Parkinson et al. combined the P–InP photoelectrode with a biological catalyst to influence the first step: a two-electron reduction of CO₂ to formic acid.¹⁷² Kaneco et al. reported that the maximum current efficiency for CO (41.5%) was obtained at −2.4 V for the P–InP photocathode in a methanol-based electrolyte, which was thus more effective than P–Si and P-Ga-As.¹⁷³ Later, their group also investigated metal-modified P–InP photoelectrodes in a LiOH/methanol-based electrolyte and improved the current efficiency for CO to 80.4%.¹⁷⁴ Arai et al. modified the P–InP with an electropolymerized ruthenium complex in water and successfully obtained HCOO⁻ during the CO₂ reduction process.¹⁷⁵ Sato et al. combined p-type InP with a Ru complex polymer hybrid photocatalyst for the PEC reduction of CO₂ to HCOO⁻.¹⁷⁶

Some other efforts to further improve the production selectively, FE, and materials stability of phosphides have also been reported.^{163,177–179} Zeng et al. used a TiO₂-passivated GaP photocathode for CO₂ reduction under 532 nm wavelength illumination, which showed a high FE for methanol.¹⁷⁹ They also investigated the influence of the TiO₂ layer thickness on the PEC performance, where no enhancement was observed for TiO₂ layer thicknesses above 10 nm. Optical microscopy and atomic force microscopy images of bare GaP and TiO₂-passivated GaP are shown in Fig. 11. Recently, Qiu et al. reported that the TiO₂-passivated InP photocathode's FE for methanol could be improved to 4.79%, which was 5.7 times higher than that for the bare InP at −0.6 V vs. NHE.¹⁸⁰ In addition, copper was also deposited on TiO₂, and acted as a cocatalyst to further improve the selectivity and yield of methanol production by up to 8-fold with an FE of 8.7%.¹⁸⁰ Zeng et al. deposited TiO₂ on the InP surface and used this for CO₂ reduction. The FE for CO reached 99% with a potential of +0.78 V in the ionic liquid electrolyte (1-ethyl-3-methylimidazolium tetrafluoroborate, [EMIM]BF₄).¹⁸¹

Fig. 11 (a) Optical microscope image, (b) atomic force microscope image, and (c) surface topography of the bare GaP surface after an 8 h reaction at −0.5 V overpotential. (d) Optical microscope image, (e) atomic force microscope image, and (f) surface topography of 5 nm TiO₂ on the GaP surface after an 8 h reaction.¹⁷⁹ Copyright 2014 American Chemical Society.

4.2 Metal oxides

Most metal oxides can be used as a semiconductor, and have thus attracted great attention and been widely studied in the photocatalysis, electrocatalysis, and photoelectrocatalysis research fields. Among these, TiO₂, ZnO, and Cu_xO_y, have been widely used as photoelectrodes for the PEC reduction of CO₂.

4.2.1 TiO₂. TiO₂ and modified TiO₂ used as photocatalysts for the photoreduction of CO₂ to value fuels have been investigated extensively. However, the research reports about TiO₂ and modified TiO₂ photoelectrodes for the reduction of CO₂ are very few in the literature. Li et al. fabricated CdSeTe NSs/TiO₂ with a flaky structure and used this for the PEC reduction of CO₂ for the first time.¹⁸² The current density of the CdSeTe NSs/TiO₂ sample could reach up to 4.5 mA cm⁻² at −0.8 V vs. NHE, which is about 14.5 times higher than that of TiO₂ without the decoration of CdSeTe NSs. The production of methanol could reach up to 1166.77 μmol L⁻¹ after 5 h. Guaraldo et al. prepared a new Si/TiO₂/Pt junction for the PEC reduction of CO_2, which showed 96.5% FE for carbon-based production, including methanol, methane, and acetone at −0.8 V vs. Ag/AgCl.¹⁸³ Additionally, for the Si/TiO₂/Pt junction sample, the product rates of methanol, ethanol, and acetone were 0.88, 2.60, and 0.049 mmol L⁻¹ at 90 min, respectively. Cheng et al. used TiO₂ nanotubes as s photoanode and Cu foam combined with Pt-modified reduced graphene oxide (Pt-RGO) as the cathode for CO₂ reduction in a PEC cell.¹⁸⁴ The carbon conversion rate could reach as high as 4340 nmol (h⁻¹ cm⁻²), which is the highest ever reported.^{176,185–187} Moreover, Cheng et al. also fabricated Pt-modified reduced graphene oxide (Pt-RGO) on Ni foam as a cathode and Pt-modified TiO₂ nanotubes (Pt-TNT) on Ti foil as a photoanode for CO₂ conversion.^185,188 The influence of the GO layer on the CO₂ reduction activity was investigated. The SEM and AFM images are shown in Fig. 12. The carbon atom conversion rate could reach 1250 nmol (h⁻¹ cm⁻²) and the highest current density could reach 4 mA cm⁻² at a constant voltage of 2 V. Ampelli et al. reported a similar PEC cell in 2010, as shown in Fig. 12(a), which provided a useful ideal for PEC CO₂ reduction.¹⁸⁹

Fig. 12 (a) View and schematic diagram of the PEC system (a₁, view of the lab-scale PEC device; a₂, image of a photo/electrocatalytic disc; a₃, scheme of a PEC device for CO₂ reduction), (b) SEM image of TiO₂ nanotubes, (c) SEM cross-sectional image of TiO₂, (d) SEM image of Pt-modified TiO₂ nanotubes (Pt-TNT), (e) and (f) AFM images and section analysis of graphene oxide (GO): (e) single layer and (f) double layers, (g) and (h) AFM images and section analysis of Pt-modified reduced graphene oxide (Pt-RGO): (g) single layer and (h) double layers. (a).¹⁸⁹ Copyright 2010 Royal Society of Chemistry; (e), (f), (g), and (h).¹⁸⁸ Copyright 2014 American Chemical Society; (b), (c), and (d).¹⁸⁵ Copyright 2014 Elsevier.

In most cases, TiO₂ is used as the photoanode, while CO₂ can only be reduced on the cathode, and thus the products greatly depend on the material used for the cathode. Yamamoto et al. reported TiO₂ as a photoanode in a methanol electrolyte for the PEC reduction of CO₂, where the main product on the Pb cathode was methyl formate. When Ag was used as the cathode, the main product was carbon monoxide.¹⁹⁰ Zhou et al.¹⁹¹ fabricated a solar-driven CO₂ reduction cell, consisting of a GaAs/InGaP/TiO₂/Ni photoanode for the oxygen-evolution reaction (OER) and a Pd/C/Ti mesh cathode for CO₂ reduction in 2.8 M KHCO₃ solution. At the operational current density of 8.5 mA cm⁻², the cathode exhibited <100 mV overpotential and >94% FE for the reduction of CO₂ (g) to formate. The anode exhibited a 320 ± 7 mV overpotential for the OER in 1.0 M KOH (aq), while the bipolar membrane exhibited a ∼480 mV voltage loss with minimal product crossovers and >90% and >95% selectivity for protons and hydroxide ions, respectively.

There are also some reports using TiO₂ as a photocathode. Xu et al. designed a PEC cell containing a multifunctionalized TiO₂ photocathode and Co–Pi/W:BiVO₄ anode.¹⁹² On the surface of the TiO₂ film, Pd nanoparticles were introduced to catch protons, while molecules of Eosin Y disodium salt were used to absorb solar energy, and four different amine ligands were immobilized to capture and activate the CO₂.¹⁹² On the optimized sample, the maximum formation rate and highest selectivity for methanol were 43.6 μM cm⁻² h⁻¹ and 100%, respectively. Some methods to further improve its efficiency are also available.^192–196 A heterojunction structure and cocatalysts are efficient methods to improve the CO₂ conversion efficiency and the products FE.

4.2.2 ZnTe and ZnO. Jang et al. reported a new ZnTe–ZnO/Zn photocathode for the first time for PEC CO₂ reduction.¹⁹⁷ This new photoelectrocatalyst showed stable CO formation at −0.2 to −0.7 V vs. RHE without the need for a sacrificial reagent. However, the FE of carbon-based products is very low. Hence, the authors designed a Au coupled ZnTe/ZnO electrode for photoelectrocatalytic CO₂ reduction, which also exhibited excellent efficiency and a higher selectivity of CO compared to bare ZnTe/ZnO, as shown in Fig. 13.¹⁹⁸ Besides, they also designed a system composed of a Au-decorated triple-layered ZnO@ZnTe@CdTe core–shell nanorod array photocathode and a CH₃NH₃PbI₃ perovskite solar cell in tandem.¹⁹⁹ This system provided an example of a photocathode–photovoltaic tandem device operating under sunlight without external bias for selective CO₂ reduction. Very recently, Xiong and coauthors reported a combined spectroelectrochemical and transient absorption (TA) spectroscopic investigation of ZnTe photoelectrodes for CO₂ reduction.²⁰⁰ Their research as a mechanistic investigation provided a rationalization for the observed photocurrent response, with the charge transport and accumulation at the surface becoming competitive with relaxation into deep-lying inactive states at increasing cathodic potentials. Li et al. designed a PEC cell for the selective conversion of CO₂ to CO by a mesoporous Pd–Cu bimetallic cathode as well as by a hierarchical carbon nanotube and Co₃O₄-decorated ZnO nanowire array photoanode, with the schematic model, FE of CO and IPCE spectra all shown in Fig. 14.²⁰¹ The maximum FE for CO was about 75% at −1.2 V vs. RHE with a high IPCE of 16% at 340 nm.

Fig. 13 Time-profiled production of CO and H₂ for the ZnTe/ZnO–NWs (ZOZT) (a and b, respectively) and the ZnTe/ZnO–NWs (ZOZT–Au) (c and d, respectively) in CO₂-purged 0.5 M KHCO₃ electrolytes under simulated 1 sun illumination at various applied potentials: −0.1, −0.3, −0.5, and −0.7 V_RHE; (e) schematic illustration of the Au coupled ZOZT; (f) faradaic efficiencies of CO and H₂ for the bare ZOZT (A) and Au coupled ZOZT (B) at different applied potentials in CO₂-saturated KHCO₃ electrolyte for 3 h. The values above the top of each column present the selectivity of CO from CO₂ reduction relative to H₂ from H₂O.¹⁹⁸ Copyright 2015 Royal Society of Chemistry.

Fig. 14 (a) Schematic model of a PEC device for the selective conversation of CO₂ to CO; cathode: mesoporous Pd₇Cu₃ alloy; photoanode: ternary hierarchical CNTs-ZnO-Co₃O₄ NW composite; (b): FEs for CO over the samples; (c) IPCE spectra of M-PdCu/T-ZnO at a high FE_CO bias region.²⁰¹ Copyright 2016 Royal Society of Chemistry.

4.2.3 Cu₂O and CuO. Recently, copper-based materials have attracted great attention for CO₂ reduction due to their low toxicity, high abundance, and wide production distribution. Cu₂O with a direct band gap of about 2.0 eV can utilize visible light and can suppress the HER and so have been widely used in CO₂ reduction.⁵⁶ However, the poor stability of Cu₂O is one of the limiting factors for their large-scale utilization in CO₂ reduction. Hence, a lot of efforts have been made to improve their stability, such as by coating with suitable protective layers or cocatalysts. Ghadimkhani et al.²⁰² and Rajeshwar et al.²⁰³ both reported a CuO–Cu₂O semiconductor nanorod array as a photoelectrode for the PEC reduction of CO₂ to methanol, achieving about a 95% FE. Besides, Homayoni et al.²⁰⁴ designed a new continuous flow PEC reactor (CFPR) for CO₂ reduction. They used the hybrid p-type CuO/Cu₂O nanorod array as a photocathode, achieving an FE of about 75–96%, where the main products were alcohols, including ethanol, isopropanol, and methanol. There are also some researches focusing on Cu/Cu₂O photoelectrodes.^205–208 Ba and coauthors found that the CO₂ conversion efficiency over a p-type Cu/Cu₂O electrode was much higher than that over an n-type Cu/Cu₂O electrode that had a similar morphology to the p-type Cu/Cu₂O.²⁰⁵ Brito et al. discovered that the pH played a key role in the products distribution (including methanol, ethanol, formaldehyde, acetaldehyde, acetone) over the Cu/Cu₂O electrode.²⁰⁶ Besides, Schreier et al. utilized the covalent immobilization of a molecular catalyst on a mesoporous-TiO₂-modified Cu₂O photocathode for CO₂ reduction, which showed significantly enhanced CO evolution efficiency at a large current density and with high selectivity.²⁰⁹

Although the coating of a protective layer can improve the stability and CO₂ conversion efficiency, the active sites for the preferential reduction of CO₂ on the Cu₂O surface may be covered by the protective layers. As a result, a large amount of H₂ may be generated, resulting in a decreased selectivity for carbonaceous products. Chang et al. used Cu₂O as the dark cathode and TiO₂ as the typical photoanode for CO₂ reduction, which showed an 87.45% FE and 92.65% selectivity for all carbonaceous products at a low potential of 0.75 V vs. RHE.²¹⁰ This strategy could significantly improve the stability of Cu₂O and its selectivity for carbonaceous products.

CuO has also been reported as a photoelectrode for PEC CO₂ reduction, where it has been frequently combined with other materials to enhance the performance.^{209,211–215} Kang et al. used a CuFeO₂/CuO electrode to convert CO₂ to formate, with over 90% selectivity under simulated solar light (100 mW cm⁻²) at +0.9 V vs. RHE.²¹⁵ Moreover, a strategy connecting the CuFeO₂/CuO photocathode to a Pt anode was further developed, which could produce formate over 1 week at a solar-to-formate energy conversion efficiency of about 1% (selectivity of >90%) without any external bias. The PEC cell and performances are shown in Fig. 15. Yang et al. also fabricated a CuFeO₂/CuO photoelectrode for the PEC reduction of CO₂, where the FE for acetate reached 80% at −0.4 V vs. Ag/AgCl under visible light irradiation.²¹⁴ Meanwhile, the selectivity for CO₂ production could be tuned by controlling the composition ratio of mixed Fe–Cu oxide catalysts. This article reported a material with the ability to efficiently couple C–C bonds and generate acetate on such a low-cost, earth-abundant catalyst during CO₂ reduction.

Fig. 15 (a) Comparison of linear sweep voltammograms for CuFeO₂/CuO, CuFeO₂, and CuO electrodes. (b) Time-profiled formate production with CuFeO₂/CuO electrodes at +0.15 V_RHE in bicarbonate (0.1 M) solutions (PCE-1). (c) Time-profiled formate production with open-circuited CuFeO₂/CuO, CuFeO₂, and CuO electrodes in CO₂-purged bicarbonate solution (PEC-2). No potential biases were applied. (d) Time-profiled changes in cell voltage (E_cell), simultaneous production of formate, and solar-to-formate (STF) energy conversion efficiencies (PEC-3 and 4). (e) Changes in E_cell and formate production with a wired CuFeO₂/CuO and Pt couple in CO₂-purged bicarbonate (0.1 M) solution (PEC-4).²¹⁵ Copyright 2015 Royal Society of Chemistry.

4.3 Metal-free

Metal-free photoelectrodes have attracted great attention due to their many advantages, such as low cost, environmental friendliness and high charge-carrier mobility. Graphitic, silicon (Si), and C₃N₄ as typical metal-free photoelectrodes have been widely used in various fields. Tahir et al. reported a review on graphene and its unique structure, band structure engineering, physical and chemical properties, applications in solar energy conversation (including the PEC reduction of CO₂), energy storage, environment purification, and biosensors.²¹⁶ However herein, we do not discuss the PEC performance for CO₂ reduction by using a graphene photoelectrode.

Graphene-like carbon nitride (C₃N₄), as a metal-free semiconductor, has also attracted great attention for CO₂ reduction. However, most of the reports about C₃N₄ are about its use as a photocatalyst, and there are few works on C₃N₄ film as a photoelectrode for CO₂ reduction.^217–220 Recently, Azofra and coauthors investigated the role of corrugation in C₃N₄ during CO₂ conversion by a DFT method.²²¹ Besides, Sagara et al. fabricated B-doped g-C₃N₄ as a photoelectrocatalyst for CO₂ reduction, whose photocurrent response was about 5 times higher than that of pure g-C₃N₄.²²² In addition, they also investigated the effects of a cocatalyst (including Au, Ag, and Rh) on the CO₂ reduction performance. Although they have done lots of work toward improving the CO₂ reduction performance, the selectivity, FE, and productivity are still unsatisfactory. Hence, efforts are still needed to further improve the performance of C₃N₄ for CO₂ reduction.

Si is another good candidate as a photoelectrode for CO₂ reduction. Recently, a mini review of silicon materials for solar-to-chemical conversion have involved some PEC CO₂ reduction processes.²²³ Zhang et al. designed a PEC cell composed of a nanoporous Ag cathode and a Ni-coated Si photoanode for CO₂ conversion to CO.²²⁴ The photocurrent density could reach up to about 10 mA cm⁻² with a 70% FE for CO at 2.0 V vs. RHE. Song et al. coated a nanoporous Au thin film on Si photoelectrodes for CO₂ reduction, which exhibited an FE of 96% for CO production at 480 mV overpotential.^225,226 The schematic of electron transfer, CO₂ reduction, and the FE of CO are shown in Fig. 16. Additionally, Wang et al. designed a monolithically integrated GaN nanowire/Si solar cell photocathode for CO₂ reduction, where the FE for CH₄ (19%) production was more than 30 times higher than that for CO.²²⁷ Urbain et al. recently designed a bias-free reactor device for CO₂ conversion. In this system, a Si heterojunction solar cell with Ni foam catalyst was used as a photoanode to facilitate the OER, while a Cu foam coated with nanosized Zn flakes was used as a cathode for the CO₂ reduction reaction.²²⁸ A stable and tunable H₂-to-CO ratio between 5 and 0.5 along with an 85% FE and 39.4 mA cm⁻² current density for CO was obtained in their work. In order to understand the reaction mechanism of CO₂ reduction, Sekimoto and coauthors carried out ¹³CO₂-labeling experiments on a GaN-Si-based tandem photoelectrode in KCl and KHCO₃ solutions, respectively.²²⁹ The results showed that almost all the products (including HCOOH, CH₄, and C₂H₄) were converted form ¹³CO₂ when KCl was used as the electrolyte. However, HCOOH from ¹²CO₂ was observed when KHCO₃ was used, indicating that HCO₃⁻ partly contributed to CO₂ reduction.

Fig. 16 (a) Si photoelectrode with Au mesh cocatalysts: schematic of the photoelectrode design for CO₂ reduction, where mesh-type cocatalysts form on the Si photoelectrode with a buried p–n junction. The mesh-type cocatalysts allow independent wiring on top of the catalyst surface for the RA treatment. Ohmic contact was made on the backside of the photoelectrode to collect photoexcited holes; (b) faradaic efficiency for CO production from the RA-treated Au thin film. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

5 Summary and perspectives

To date, many efforts have been made to design highly selective, highly conversation efficient, and energy-saving electrocatalysts and photoelectrocatalysts for CO₂ reduction. The main products and FE for the EC and PEC reduction of CO₂ are summarized and shown in Tables 2 and 3.

Table 2 The main products and their FE over different electrodes

Electrocatalyst	Electrolyte	Overpotential	Major products	FE	Ref.
Ag nanoporous	0.5 M KHCO3	<500 mV vs. RHE	CO	92%	79
Ag nanocorals	0.1 M KHCO3	0.37 V vs. RHE	CO	95%	83
Ag nanoparticles	0.5 M KHCO3	−0.75 and −1.0 V vs. RHE	CO	94.2%	84
Ag foam	0.1 M KHCO3	−0.99 V vs. RHE	CO	94.7%	85
Au nanoparticles (4 nm, 8 nm)	0.5 M KHCO3	−0.9 V vs. RHE	CO, H2	Up to 90%	89
Au nanowires (500 nm)	0.5 M KHCO3	−0.35 V vs. RHE	CO	94%	91
Au nanoneedles	0.5 M KHCO3	−0.30 V to −0.50 V vs. RHE	CO	≥95%	21 and 94
Cu nanoparticles	0.1 M KHCO3	−1.1 V vs. RHE	CO, H2	60–70%, 20–30%	92
Cu nanowires	0.1 M KHCO3	0.3 V vs. RHE	CO	60%	100
Ultrathin Cu nanowires (diameter 20 nm)	0.1 M KHCO3	−1.25 V vs. RHE	CH4	55%	101
Cu nanocrystals	0.1 M KHCO3	−1.1 V vs. RHE	H2, CO, CH4, C2H4	41% (ethylene)	102
Zn nanoparticles	0.5 M NaHCO3	−0.93 V to −2.13 V vs. RHE	HCOO^−	87%	115
Zn dendrite nanostructures	0.5 M NaHCO3	0.9 V to −1.1 V vs. RHE		∼80%	75
Au3Cu alloys	0.1 M KH2PO4/0.1 M K2HPO4	−1.6 V vs. Ag/AgCl	CO	30%	123
Cu–Sn bimetallic alloys	0.1 M KHCO3	−0.6 V vs. RHE	H2, CO HCOOH, CH3COOH	≥90% (CO)	124
PdCu/C nanoparticle alloys	0.1 M KHCO3	−0.89 V vs. RHE	CO	86%	118
Sn–Pb alloys	0.5 M KHCO3	−0.2 V vs. Ag/AgCl	HCOO^−	79.8%	126
Graphene (GN)/Cu2O	0.5 M NaHCO3	−1.7 V vs. Ag/AgCl	Ethanol (predominant product)	9.95%	130
SnO2 nanoparticles	0.1 M NaOH	−1 to −1.8 V vs. SCE	HCOO^−	93%	133
3D SnO2 nanoparticles	0.5 M NaHCO3	−1.6 V vs. Ag/AgCl	H2, HCOO^−	87%	135
Ga2O3	3.0 M KCl & 5.0 M NaOH	−1.6 to −2.0 V vs. Ag/AgCl	HCOOH	80%	143
N-doped graphene quantum dots (NGQDs)	1 M KOH	−0.20 to −1.10 V vs. RHE	CO, HCOO^−	90%	156
Carbon nanoporous (CNF), (CPNS)	0.1 M KHCO3	−0.99 V vs. RHE	CO, CH4	13.8%, 0.18%	159
N-doped carbon	1.0 M KCl	−1.46 V vs. Ag/AgCl	CO, H2	90%, 2%	160

---

Table 3 The main products and corresponding FE for the PEC reduction of CO₂ over various photoelectrodes

Photoelectrocatalyst	Electrolyte	Conditions	Products	FE	Ref.
P-GaP	0.1 M acetate buffer containing 10 mM pyridinium	200 W Hg–Xe arc light source, pH = 5.2, External bias: −0.52 V vs. SCE	Methanol	∼100%	163
InP	Methanol	Xenon lamp: 1.5 mA cm^−2, External bias: −2.4 V vs. Ag/AgCl	CO	∼41.5%	173
Metal-modified P–InP	LiOH/methanol	Xenon lamp: 1.5 mA cm^−2	CO	80.4%	174
InP/TiO2/Cu	0.5 M KCl	Under 532 nm illumination, External bias: −0.6 V vs. NHE	Methanol, H2	8.7%, 76.9%	180
CdSeTe NSs/TiO2	0.1 M KHCO3	500 W xenon lamp (λ ≥ 420 nm, 100 mW cm^−2), External bias: −0.8 V vs. NHE	Methanol, Methane	87.95% (selectivity: 74.2%), 40.79%, (Selectivity: 25.81%)	182
Si/TiO2/Pt	0.1 M NaHCO3	125 W (250–600 nm, 21 mW cm^−2), External bias: −0.8 V	Methanol, Methane, Acetone	Total FE: 96%	183
Cathode: Pt-RGO/Cu foam, Photoanode: TiO2 nanotube	Catholyte: 0.5 M NaHCO3, Anolyte: 0.5 M H2SO4	UV Xe lamp, Potential between cathode and anode is about 2 V	Ethanol, Acetic acid, Propanoic acid	Total conversation rate: 4340 nm (h^−1 cm^−2)	184
Cathode: Pb or Ag, Photoanode TiO2	Catholyte: 0.3 M KOH in methanol, Anolyte: 3.0 M KHCO3 in water	UV LED 365 nm, Potential: 2 V vs. Ag QRE	Pb: formic acid, CO and H2, Ag: formic acid, CO and H2	Pb: 80%, 9.5%, 5.1%, Ag: 13%, 61%, 3.3%	190
Cathode: Pd/C/Ti mesh, Photoanode: GaAs/InGaP/TiO2/Ni	Catholyte: 2.8 M KHCO3, Anolyte: 1.0 M KOH	<100 mV overpotential, 1 atm CO2 (g)	Formate	>94%	191
ZnTe–ZnO/Zn	0.5 M KHCO3	500 W Hg lamp (>420 nm, 400 mW cm^−2), Potential: −0.7 V vs. RHE	CO, H2	CO: 22.9%, H2: 60.7%	197
Au–ZnTe/ZnO nanowire array	0.5 M KHCO3	100 mW, Potential: −0.7 V vs. RHE	CO, H2	CO: 58%, H2: 31.8%	198
Cathode: PdCu, Photoanode: CNTs–ZnO–Co3O4 nanowire		AM 1.5, Potential: −1.2 V vs. RHE	CO	75%	201
Cu2O/CuO nanorode array	0.1 M NaHCO3	AM 1.5 (70 mW cm^−2), Potential: −0.2 V vs. SHE	Methanol	95%	202 and 203
p-type CuO/Cu2O nanorode array	0.1 M NaHCO3	AM 1.5 (70 mW cm^−2), Potential: −0.3 V vs. SHE	Alcohols	2 h, Ethanol: 52%, Isopropanol: 40%, Methanol: 4%	204
Cu/Cu2O	0.1 M KHCO3	Blue LED, 435–450 nm, Potential: −2.0 V vs. Ag/AgCl	C2H4	Cu/Cu2O (p-type, nanobelt arrays): 32.69%, Cu/Cu2O (p-type, stone-like): 30.60%, Cu/Cu2O (n-type, stone-like): 26.55%	205
Cu/Cu2O	0.1 M NaCO3/NaHCO3 with different PH	AM 1.5, 100 mW cm^−2, +0.2 V vs. Ag/AgCl, PH = 9	Methanol, Ethanol, Formaldehyde, Acetaldehyde, Acetone	80% of CO2 is converted	206
CuFeO2/CuO	0.1 M KHCO3	AM 1.5, 100 mW cm^−2, Potential: +0.9 V vs. RHE	Formate	90%	215
CuO/CuFeO2 (Fe : Cu = 1.3)	0.1 M NaHCO3	AM 1.5, 100 mW cm^−2, Potential: −0.4 V vs. Ag/AgCl	Acetate	80%	214
Cathode: Cu2O, Photoanode: TiO2 nanorod array	0.1 M KHCO3	AM 1.5, 100 mW cm^−2, Potential: +0.75 V vs. RHE	CH4, CO, CH3OH, H2	CH4: 54.63%, CO: 30.03%, CH3OH: 2.79%, H2: 6.94%	210
B-doped C3N4	0.5 M NaHCO3	AM 1.5, 100 mW cm^−2, Potential: −0.4 V vs. Ag/AgCl	Ethanol, CO	Ethanol: 78%, CO: 1.2%	222
Cathode: Ag porous, Photoanode: Ni-coated Si	Catholyte: 0.5 M KHCO3, Anolyte: boric acid, potassium chloride and potassium hydroxide in DI water	300 W Xe lamp with 400 nm filter, Potential: 2.0 V vs. RHE	CO	70%	224
Si photoelectrode with RA-treated Au mesh cocatalyst	0.2 M KHCO3	1 sun, 100 mW cm^−2, Potential: −0.59 V vs. RHE	CO	96%	226
Cu/GaN/n^+–p Si	0.5 M KHCO3	300 W Xe lamp, Potential: −1.4 V vs. Ag/AgCl	CH4	19%	227

---

Although great success has been achieved for CO₂ reduction, it is still a big challenge to design and fabricate efficient catalysts. In particular, the low energy conversation efficiency has limited its practical application significantly. Generally, the CO₂ adsorption and activation ability of catalysts are the crucial factors for the subsequent reduction steps and suppress the H₂ evolution. In order to improve the adsorption and activation ability of electrodes and photoelectrodes, various strategies have been developed, such as controlling the structure and morphology (exposing the active facets), improving the surface area (using porous materials), the active sites (surface defects), cocatalysts, and heterojunctions. In this review, we classified and presented a detailed discussion about the typical electrodes and photoelectrodes for CO₂ reduction, which could help the investigators to design a strategy for improving the CO₂ conversation efficiency and product selectivity by selecting suitable materials. For developing EC or PEC electrodes with a high energy conversion efficiency, high FE, and good stability for CO₂ reduction, whereby several suggestions are proposed:

(1) For selecting the suitable electrodes: the overpotential, CO₂ adsorption and activating ability, conductivity, stability, economics, etc. should be carefully evaluated.

(2) To enhance the adsorption and activation performances of CO₂: designing a porous structure or inverse opal structures may be a good strategy for improving the products selectivity and CO₂ adsorption. Additionally, introducing surface defects or cocatalysts may be another option to maximize the adsorption and activation of CO₂ molecules.

(3) For further understanding the mechanism and pathways of CO₂ reduction: in situ characterization techniques are very urgently needed and are necessary to understand the reaction mechanism and pathways of CO₂ reduction at the molecular level, such as in situ attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR),²³⁰ attenuated total reflection surface-enhanced infrared absorption spectroscopy (ATR-SEIRAS),²³¹in situ Raman,²³²in situ scanning transmission electron microscopy (STEM), and in situ X-ray photoelectron spectroscopy (XPS). In addition, density functional theory (DFT) calculations are also helpful to understand the pathway of CO₂ reduction, especially in guiding researchers to finding the most suitable materials.

Specifically, for PEC electrodes, because light-active semiconductors are involved in this system, several more suggestions are proposed:

(1) For PEC electrodes, the suitable band gap and band-edge potential, are important factors for CO₂ reduction. As the photoconversion efficiency of most semiconductors is very low; hence, more efficient semiconductors with superior light-harvesting ability, good charge generation ability, and excellent charge separation and transportation efficiency should be designed.

(2) Enhancing the semiconductor stability and selectivity of CO₂ reduction products are urgent and necessary. As we have mentioned in this review, introducing surface defects, surface functional group, a cocatalyst, or doping will benefit the CO₂ reduction reaction and semiconductor stability in some way.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors gratefully acknowledge financial support from Ministry of Science and Technology of the People's Republic of China (2016YFE0112200), National Natural Science Foundation of China (No. 21507011 and No. 21677037).

References

1. L. Suganthi and A. A. Samuel, Renewable Sustainable Energy Rev., 2012, 16, 1223–1240.
2. P. Nejat, F. Jomehzadeh, M. M. Taheri, M. Gohari and M. Z. A. Majid, Renewable Sustainable Energy Rev., 2015, 43, 843–862.
3. T. A. Faunce, W. Lubitz, A. W. Rutherford, D. R. MacFarlane, G. F. Moore, P. D. Yang, D. G. Nocera, T. A. Moore, D. H. Gregory, S. Fukuzumi, K. B. Yoon, F. A. Armstrong, M. R. Wasielewski and S. Styring, Energy Environ. Sci., 2013, 6, 695–698.
4. C. L. A. D. L. Hartmann, S. Brönnimann, F. J. Dentener, E. J. Dlugokencky, D. R. Easterling, A. Kaplan and B. J. Soden, IPCC (2013), Climate Change 2013, in The Physical Science Basis, Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, WMO/UNEP, Cambridge, 2014.
5. R. K. Pachauri, M. R. Allen, V. R. Barros, J. Broome, W. Cramer, R. Christ, J. A. Church, L. Clarke, Q. Dahe and P. Dasgupta, Contribution of working groups I, II and III to the fifth assessment report of the intergovernmental panel on climate change, Climate change 2014: synthesis report, IPCC, 2014.
6. G. A. Olah, A. Goeppert and G. K. Prakash, J. Org. Chem., 2016, 74, 487.
7. X. D. Xu and J. A. Moulijn, Energy Fuels, 1996, 10, 305–325.
8. M. Mikkelsen, M. Jorgensen and F. C. Krebs, Energy Environ. Sci., 2010, 3, 43–81.
9. A. Guais, G. Brand, L. Jacquot, M. Karrer, S. Dukan, G. Grevillot, T. J. Molina, J. Bonte, M. Regnier and L. Schwartz, Chem. Res. Toxicol., 2011, 24, 2061–2070.
10. J. Gong, L. Zhang and Z. J. Zhao, Angew. Chem., Int. Ed., 2017, 56, 11326–11353.
11. S. A. Montzka, E. J. Dlugokencky and J. H. Butler, Nature, 2011, 476, 43–50.
12. M. Mondal, S. Khanra, O. N. Tiwari, K. Gayen and G. N. Halder, Environ. Prog. Sustainable Energy, 2016, 35, 1605–1615.
13. T. S. Wong, J. Bioprocess. Biotech., 2014, 4, 100155–100170.
14. X. Chang, T. Wang, P. Zhang, Y. Wei, J. Zhao and J. Gong, Angew. Chem., Int. Ed. Engl., 2016, 55, 8840–8845.
15. S. Tan, S. J. Kim, J. S. Moore, Y. Liu, R. S. Dixit, J. G. Pendergast, D. S. Sholl, S. Nair and C. W. Jones, ChemCatChem, 2016, 8, 214–221.
16. A. Wuttig and Y. Surendranath, ACS Catal., 2015, 5, 4479–4484.
17. F. Studt, I. Sharafutdinov, F. Abild-Pedersen, C. F. Elkjaer, J. S. Hummelshoj, S. Dahl, I. Chorkendorff and J. K. Norskov, Nat. Chem., 2014, 6, 320–324.
18. J. Highfield, H. Lim, J. Fagerlund and R. Zevenhoven, RSC Adv., 2012, 2, 6535–6541.
19. J. W. Ko, S. W. Kim, J. Hong, J. Ryu, K. Kang and C. B. Park, Green Chem., 2012, 14, 2391–2394.
20. W. J. Bao, H. Q. Li and Y. Zhang, Ind. Eng. Chem. Res., 2010, 49, 2055–2063.
21. M. Liu, Y. Pang, B. Zhang, P. De Luna, O. Voznyy, J. Xu, X. Zheng, C. T. Dinh, F. Fan, C. Cao, F. P. de Arquer, T. S. Safaei, A. Mepham, A. Klinkova, E. Kumacheva, T. Filleter, D. Sinton, S. O. Kelley and E. H. Sargent, Nature, 2016, 537, 382–386.
22. Z. Cao, D. Kim, D. Hong, Y. Yu, J. Xu, S. Lin, X. Wen, E. M. Nichols, K. Jeong, J. A. Reimer, P. Yang and C. J. Chang, J. Am. Chem. Soc., 2016, 138, 8120–8125.
23. K. Li, X. An, K. H. Park, M. Khraisheh and J. Tang, Catal. Today, 2014, 224, 3–12.
24. R. J. Lim, M. Xie, M. A. Sk, J. M. Lee, A. Fisher, X. Wang and K. H. Lim, Catal. Today, 2013, 233, 169–180.
25. A. J. Martin, G. O. Larrazabal and J. Perez-Ramirez, Green Chem., 2015, 17, 5114–5130.
26. J. Shen, R. Kortlever, R. Kas, Y. Y. Birdja, O. Diaz-Morales, Y. Kwon, I. Ledezma-Yanez, K. J. Schouten, G. Mul and M. T. Koper, Nat. Commun., 2015, 6, 8177–8185.
27. S. C. Roy, O. K. Varghese, M. Paulose and C. A. Grimes, ACS Nano, 2010, 4, 1259–1278.
28. P. L. Staddon and M. H. Depledge, Environ. Sci. Technol., 2015, 49, 8269–8270.
29. J. Baxter, Z. X. Bian, G. Chen, D. Danielson, M. S. Dresselhaus, A. G. Fedorov, T. S. Fisher, C. W. Jones, E. Maginn, U. Kortshagen, A. Manthiram, A. Nozik, D. R. Rolison, T. Sands, L. Shi, D. Sholl and Y. Y. Wu, Energy Environ. Sci., 2009, 2, 559–588.
30. N. Hazari, E. Iglesia, J. A. Labinger and D. A. Simonetti, Acc. Chem. Res., 2012, 45, 653–662.
31. M. Villano, F. Aulenta, C. Ciucci, T. Ferri, A. Giuliano and M. Majone, Bioresour. Technol., 2010, 101, 3085–3090.
32. A. S. Agarwal, Y. Zhai, D. Hill and N. Sridhar, ChemSusChem, 2011, 4, 1301–1310.
33. M. Aresta, A. Dibenedetto and A. Angelini, Chem. Rev., 2014, 114, 1709–1742.
34. C. Liu, J. J. Gallagher, K. K. Sakimoto, E. M. Nichols, C. J. Chang, M. C. Chang and P. Yang, Nano Lett., 2015, 15, 3634–3639.
35. O. Martin, A. J. Martin, C. Mondelli, S. Mitchell, T. F. Segawa, R. Hauert, C. Drouilly, D. Curulla-Ferre and J. Perez-Ramirez, Angew. Chem., Int. Ed., 2016, 55, 6261–6265.
36. B. Kumar, M. Llorente, J. Froehlich, T. Dang, A. Sathrum and C. P. Kubiak, Annu. Rev. Phys. Chem., 2012, 63, 541–569.
37. B. Mondal, F. Neese and S. Ye, Inorg. Chem., 2015, 54, 7192–7198.
38. S. Xie, Q. Zhang, G. Liu and Y. Wang, Chem. Commun., 2016, 52, 35–59.
39. W. G. Tu, Y. Zhou and Z. G. Zou, Adv. Mater., 2014, 26, 4607–4626.
40. Y. Ma, X. Wang, Y. Jia, X. Chen, H. Han and C. Li, Chem. Rev., 2014, 114, 9987–10043.
41. 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., 2015, 115, 12888–12935.
42. X. X. Chang, T. Wang and J. L. Gong, Energy Environ. Sci., 2016, 9, 2177–2196.
43. Z. L. Wang, C. L. Li and Y. Yamauchi, Nano Today, 2016, 11, 373–391.
44. J. Qiao, Y. Liu, F. Hong and J. Zhang, Chem. Soc. Rev., 2014, 43, 631–675.
45. N. Elgrishi, M. B. Chambers, X. Wang and M. Fontecave, Chem. Soc. Rev., 2017, 46, 761–796.
46. H. Takeda, C. Cometto, O. Ishitani and M. Robert, ACS Catal., 2016, 7, 70–88.
47. Y. Yamazaki, H. Takeda and O. Ishitani, J. Photochem. Photobiol., C, 2015, 25, 106–137.
48. W. Kim, B. A. McClure, E. Edri and H. Frei, Chem. Soc. Rev., 2016, 45, 3221–3243.
49. W. H. Wang, Y. Himeda, J. T. Muckerman, G. F. Manbeck and E. Fujita, Chem. Rev., 2015, 115, 12936–12973.
50. E. S. Sanzpérez, C. R. Murdock, S. A. Didas and C. W. Jones, Chem. Rev., 2016, 116, 11840–11876.
51. J. Wu, Y. Huang, W. Ye and Y. Li, Adv. Sci., 2017, 4, 1700194–1700222.
52. J. Kim, D. Summers and K. Frese, J. Electroanal. Chem. Interfacial Electrochem., 1988, 245, 223–244.
53. H.-R. M. Jhong, S. Ma and P. J. Kenis, Curr. Opin. Chem. Eng., 2013, 2, 191–199.
54. J. Albo, M. Alvarez-Guerra, P. Castaño and A. Irabien, Green Chem., 2015, 17, 2304–2324.
55. D. D. Zhu, J. L. Liu and S. Z. Qiao, Adv. Mater., 2016, 28, 3423–3452.
56. X. Li, J. Q. Wen, J. X. Low, Y. P. Fang and J. G. Yu, Sci. China Mater., 2014, 57, 70–100.
57. M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. Mi, E. A. Santori and N. S. Lewis, Chem. Rev., 2010, 110, 6446–6473.
58. K. Hara, A. Kudo, T. Sakata and M. Watanabe, J. Electrochem. Soc., 1995, 142, L57–L59.
59. C. Ampelli, C. Genovese, M. Errahali, G. Gatti, L. Marchese, S. Perathoner and G. Centi, J. Appl. Electrochem., 2015, 45, 701–713.
60. C. Ampelli, C. Genovese, B. Marepally, G. Papanikolaou, S. Perathoner and G. Centi, Faraday Discuss., 2015, 183, 125–145.
61. M. Azuma, K. Hashimoto, M. Hiramoto, M. Watanabe and T. Sakata, J. Electroanal. Chem., 1989, 260, 441–445.
62. M. Gattrell, N. Gupta and A. Co, J. Electroanal. Chem., 2006, 594, 1–19.
63. J. J. Kim, D. P. Summers and K. W. Frese Jr, J. Electroanal. Chem., 1988, 245, 223–244.
64. Y. Hori, A. Murata and R. Takahashi, ChemInform, 1990, 21, 2309–2326.
65. B. C. Marepally, C. Ampelli, C. Genovese, T. Saboo, S. Perathoner, F. M. Wisser, L. Veyre, J. Canivet, E. A. Quadrelli and G. Centi, ChemSusChem, 2017, 10, 4442–4446.
66. C. Ampelli, G. Centi, R. Passalacqua and S. Perathoner, Catal. Today, 2016, 259, 246–258.
67. C. Ampelli, C. Genovese, B. C. Marepally, G. Papanikolaou, S. Perathoner and G. Centi, Faraday Discuss., 2015, 183, 125.
68. K. P. Kuhl, T. Hatsukade, E. R. Cave, D. N. Abram, J. Kibsgaard and T. F. Jaramillo, J. Am. Chem. Soc., 2017, 136, 14107–14113.
69. K. P. Kuhl, T. Hatsukade, E. R. Cave, D. N. Abram, J. Kibsgaard and T. F. Jaramillo, J. Am. Chem. Soc., 2014, 136, 14107–14113.
70. R. Kortlever, J. Shen, K. J. Schouten, F. Calle-Vallejo and M. T. Koper, J. Phys. Chem. Lett., 2015, 6, 4073–4082.
71. C. Costentin, M. Robert and J. M. Saveant, Chem. Soc. Rev., 2013, 44, 2423–2436.
72. S. Lin, C. S. Diercks, Y. B. Zhang, N. Kornienko, E. M. Nichols, Y. Zhao, A. R. Paris, D. Kim, P. Yang, O. M. Yaghi and C. J. Chang, Science, 2015, 349, 1208–1213.
73. Y. Liu, S. Chen, X. Quan and H. Yu, J. Am. Chem. Soc., 2015, 137, 11631–11636.
74. Q. Lu, J. Rosen and F. Jiao, ChemCatChem, 2015, 46, 2.
75. J. Rosen, G. S. Hutchings, Q. Lu, R. V. Forest, A. Moore and F. Jiao, ACS Catal., 2015, 5, 4586–4591.
76. J. M. Saveant, Chem. Rev., 2008, 108, 2348–2378.
77. J. H. Koh, H. S. Jeon, M. S. Jee, E. B. Nursanto, H. Lee, J. H. Yun and B. K. Min, J. Phys. Chem. C, 2015, 119, 883–889.
78. Q. Lu and F. Jiao, Nano Energy, 2016, 29, 439–456.
79. Q. Lu, J. Rosen, Y. Zhou, G. S. Hutchings, Y. C. Kimmel, J. G. Chen and F. Jiao, Nat. Commun., 2014, 5, 3242.
80. B. A. Rosen, A. Salehikhojin, M. R. Thorson, W. Zhu, D. T. Whipple, P. J. Kenis and R. I. Masel, Science, 2011, 334, 643–644.
81. N. Hoshi, M. Kato and Y. Hori, J. Electroanal. Chem., 1997, 440, 283–286.
82. A. Salehi-Khojin, H. R. M. Jhong, B. A. Rosen, W. Zhu, S. C. Ma, P. J. A. Kenis and R. I. Masel, J. Phys. Chem. C, 2013, 117, 1627–1632.
83. Y. C. Hsieh, S. D. Senanayake, Y. Zhang, W. Q. Xu and D. E. Polyansky, ACS Catal., 2015, 5, 5349–5356.
84. C. Kim, T. Eom, M. S. Jee, H. Jung, H. Kim, B. K. Min and Y. J. Hwang, ACS Catal., 2017, 7, 779–785.
85. R. Daiyan, X. Y. Lu, Y. H. Ng and R. Amal, ChemistrySelect, 2017, 2, 879–884.
86. D. R. Kauffman, D. Alfonso, C. Matranga, H. Qian and R. Jin, J. Am. Chem. Soc., 2012, 134, 10237–10243.
87. D. Gao, J. Wang, H. Wu, X. Jiang, S. Miao, G. Wang and X. Bao, Electrochem. Commun., 2015, 55, 1–5.
88. H. Kim, H. S. Jeon, M. S. Jee, E. B. Nursanto, J. P. Singh, K. Chae, Y. J. Hwang and B. K. Min, ChemSusChem, 2016, 9, 2097–2102.
89. W. Zhu, R. Michalsky, O. Metin, H. Lv, S. Guo, C. J. Wright, X. Sun, A. A. Peterson and S. Sun, J. Am. Chem. Soc., 2013, 135, 16833–16836.
90. H. J. Feng, Y. M. Yang, Y. M. You, G. P. Li, J. Guo, T. Yu, Z. X. Shen, T. Wu and B. G. Xing, Chem. Commun., 2009, 15, 1984–1986.
91. W. Zhu, Y. J. Zhang, H. Zhang, H. Lv, Q. Li, R. Michalsky, A. A. Peterson and S. Sun, J. Am. Chem. Soc., 2014, 136, 16132–16135.
92. H. Mistry, R. Reske, Z. Zeng, Z. J. Zhao, J. Greeley, P. Strasser and B. R. Cuenya, J. Am. Chem. Soc., 2014, 136, 16473–16476.
93. J. H. Koh, H. S. Jeon, M. S. Jee, E. B. Nursanto, H. Lee, Y. J. Hwang and B. K. Min, J. Phys. Chem. C, 2015, 119, 883–889.
94. T. Saberi Safaei, A. Mepham, X. Zheng, Y. Pang, C. T. Dinh, M. Liu, D. Sinton, S. O. Kelley and E. H. Sargent, Nano Lett., 2016, 16, 7224–7228.
95. A. S. Hall, Y. Yoon, A. Wuttig and Y. Surendranath, J. Am. Chem. Soc., 2015, 137, 14834–14837.
96. Y. Kwon, Y. W. Lum, E. L. Clark, J. W. Ager and A. T. Bell, ChemElectroChem, 2016, 3, 1012–1019.
97. Y. Shen, Q. Fang and B. Chen, Environ. Sci. Technol., 2015, 49, 67–84.
98. D. Ren, N. T. Wong, A. D. Handoko, Y. Huang and B. S. Yeo, J. Phys. Chem. Lett., 2016, 7, 20–24.
99. R. Reske, H. Mistry, F. Behafarid, B. R. Cuenya and P. Strasser, J. Am. Chem. Soc., 2014, 136, 6978–6986.
100. D. Raciti, K. J. Livi and C. Wang, Nano Lett., 2015, 15, 6829–6835.
101. Y. Li, F. Cui, M. B. Ross, D. Kim, Y. Sun and P. Yang, Nano Lett., 2017, 17, 1312–1317.
102. F. S. Roberts, K. P. Kuhl and A. Nilsson, Angew. Chem., Int. Ed., 2015, 54, 5179–5182.
103. A. Loiudice, P. Lobaccaro, E. A. Kamali, T. Thao, B. H. Huang, J. W. Ager and R. Buonsanti, Angew. Chem., Int. Ed., 2016, 55, 5789–5792.
104. K. D. Yang, C. W. Lee, J. H. Jang, T. R. Ha and K. T. Nam, Nanotechnology, 2017, 28, 352001–352010.
105. X. Feng, K. Jiang, S. Fan and M. W. Kanan, ACS Cent. Sci., 2016, 2, 169–174.
106. Y. Kwon, Y. Lum, E. L. Clark, J. W. Ager and A. T. Bell, ChemElectroChem, 2016, 3, 1012–1019.
107. Y. Chen, C. W. Li and M. W. Kanan, J. Am. Chem. Soc., 2012, 134, 19969–19972.
108. A. Verdaguer-Casadevall, C. W. Li, T. P. Johansson, S. B. Scott, J. T. McKeown, M. Kumar, I. E. Stephens, M. W. Kanan and I. Chorkendorff, J. Am. Chem. Soc., 2015, 137, 9808–9811.
109. C. W. Li, J. Ciston and M. W. Kanan, Nature, 2014, 508, 504–507.
110. Z. M. Detweiler, J. L. White, S. L. Bernasek and A. B. Bocarsly, Langmuir, 2014, 30, 7593–7600.
111. L. Sun, G. K. Ramesha, P. V. Kamat and J. F. Brennecke, Langmuir, 2014, 30, 6302–6308.
112. K. Dhar and C. Cavallotti, J. Phys. Chem. A, 2014, 118, 8676–8688.
113. C. Shi, C. P. O'Grady, A. A. Peterson, H. A. Hansen and J. K. Norskov, Phys. Chem. Chem. Phys., 2013, 15, 7114–7122.
114. D. Gao, H. Zhou, J. Wang, S. Miao, F. Yang, G. Wang, J. Wang and X. Bao, J. Am. Chem. Soc., 2015, 137, 4288–4291.
115. T. T. Zhang, H. X. Zhong, Y. L. Qiu, X. F. Li and H. M. Zhang, J. Mater. Chem. A, 2016, 4, 16670–16676.
116. M. Baghayeri, Sens. Actuators, B, 2017, 240, 255–263.
117. D. A. Torelli, S. A. Francis, J. C. Crompton, A. Javier, J. R. Thompson, B. S. Brunschwig, M. P. Soriaga and N. S. Lewis, ACS Catal., 2016, 6, 2100–2104.
118. Z. Yin, D. Gao, S. Yao, B. Zhao, F. Cai, L. Lin, P. Tang, P. Zhai, G. Wang and D. Ma, Nano Energy, 2016, 27, 35–43.
119. C. S. Chen, A. D. Handoko, J. H. Wan, L. Ma, D. Ren and B. S. Yeo, Catal. Sci. Technol., 2014, 5, 161–168.
120. M. Watanabe, M. Shibata and A. Kato, J. Electrochem. Soc., 1991, 138, 3382–3389.
121. X. Guo, Y. Zhang, C. Deng, X. Li, Y. Xue, Y. M. Yan and K. Sun, Chem. Commun., 2015, 51, 1345–1348.
122. A. A. Peterson, F. Abild-Pedersen, F. Studt, J. Rossmeisl and J. K. Norskov, Energy Environ. Sci., 2010, 3, 1311–1315.
123. W. G. Zhao, L. N. Yang, Y. D. Yin and M. S. Jin, J. Mater. Chem. A, 2014, 2, 902–906.
124. S. Sarfraz, A. T. Garcia-Esparza, A. Jedidi, L. Cavallo and K. Takanabe, ACS Catal., 2016, 6, 2842–2851.
125. I. Choi, H.-Y. Kim, S. H. Ahn, S. J. Hwang, S. J. Yoo, H. Y. Kim, J. Choi, H. J. Park, J. H. Jang and S.-K. Kim, J. Nanosci. Nanotechnol., 2016, 16, 10470–10474.
126. S. Y. Choi, S. K. Jeong, H. J. Kim, I.-H. Baek and K. T. Park, ACS Sustainable Chem. Eng., 2016, 4, 1311–1318.
127. A. Jedidi, S. Rasul, D. Masih, L. Cavallo and K. Takanabe, J. Mater. Chem. A, 2015, 3, 19085–19092.
128. S. Rasul, D. H. Anjum, A. Jedidi, Y. Minenkov, L. Cavallo and K. Takanabe, Angew. Chem., Int. Ed., 2015, 127, 2174–2178.
129. Z. B. Hoffman, T. S. Gray, K. B. Moraveck, T. B. Gunnoe and G. Zangari, ACS Catal., 2017, 7, 5381–5390.
130. R. A. Geioushy, M. M. Khaled, A. S. Hakeem, K. Alhooshani and C. Basheer, J. Electroanal. Chem., 2017, 785, 138–143.
131. N. Kumari, N. Sinha, M. A. Haider and S. Basu, Electrochim. Acta, 2015, 177, 21–29.
132. Y. F. Li, W. P. Zhang, X. Shen, P. F. Peng, L. B. Xiong and Y. Yu, Chin. J. Catal., 2015, 36, 2229–2236.
133. S. Zhang, P. Kang and T. J. Meyer, J. Am. Chem. Soc., 2014, 136, 1734–1737.
134. M. F. Baruch, J. E. Pander, J. L. White and A. B. Bocarsly, ACS Catal., 2015, 5, 3148–3156.
135. F. Li, L. Chen, G. P. Knowles, D. R. MacFarlane and J. Zhang, Angew. Chem., Int. Ed., 2017, 56, 505–509.
136. Y. Oh, H. Vrubel, S. Guidoux and X. Hu, Chem. Commun., 2014, 50, 3878–3881.
137. Y. Oh and X. Hu, Chem. Commun., 2015, 51, 13698–13701.
138. D. Kim, S. Lee, J. D. Ocon, B. Jeong, J. K. Lee and J. Lee, Phys. Chem. Chem. Phys., 2015, 17, 824–830.
139. B. C. Marepally, C. Ampelli, C. Genovese, F. Tavella, L. Veyre, E. A. Quadrelli, S. Perathoner and G. Centi, J. CO₂ Util., 2017, 21, 534–542.
140. C. Genovese, C. Ampelli, S. Perathoner and G. Centi, Green Chem., 2017, 19, 2406–2415.
141. C. Ampelli, C. Genovese, M. Errahali, G. Gatti, L. Marchese, S. Perathoner and G. Centi, J. Appl. Electrochem., 2015, 45, 701–713.
142. C. Genovese, C. Ampelli, S. Perathoner and G. Centi, J. Catal., 2013, 308, 237–249.
143. T. Sekimoto, M. Deguchi, S. Yotsuhashi, Y. Yamada, T. Masui, A. Kuramata and S. Yamakoshi, Electrochem. Commun., 2014, 43, 95–97.
144. G. K. Ramesha, J. F. Brennecke and P. V. Kamat, ACS Catal., 2014, 4, 3249–3254.
145. N. Ullah, I. Ali, M. Jansen and S. Omanovic, Can. J. Chem. Eng., 2015, 93, 55–62.
146. X. Cui, G.-H. Lee, Y. D. Kim, G. Arefe, P. Y. Huang, C.-H. Lee, D. A. Chenet, X. Zhang, L. Wang and F. Ye, Nat. Nanotechnol., 2015, 10, 534–540.
147. M. Asadi, B. Kumar, A. Behranginia, B. A. Rosen, A. Baskin, N. Repnin, D. Pisasale, P. Phillips, W. Zhu, R. Haasch, R. F. Klie, P. Král, J. Abiade and A. Salehi-Khojin, Nat. Commun., 2014, 5, 4470–4477.
148. M. Asadi, K. Kim, C. Liu, A. V. Addepalli, P. Abbasi, P. Yasaei, P. Phillips, A. Behranginia, J. M. Cerrato and R. Haasch, Science, 2016, 353, 467–470.
149. P. Abbasi, M. Asadi, C. Liu, S. Sharifi-Asl, B. Sayahpour, A. Behranginia, P. Zapol, R. Shahbazian-Yassar, L. A. Curtiss and A. Salehi-Khojin, ACS Nano, 2016, 11, 453–460.
150. K. Chan, C. Tsai, H. A. Hansen and J. K. Nørskov, ChemCatChem, 2014, 6, 1899–1905.
151. X. Hong, K. Chan, C. Tsai and J. K. Nørskov, ACS Catal., 2016, 6, 4428–4437.
152. X. Sun, Q. Zhu, X. Kang, H. Liu, Q. Qian, Z. Zhang and B. Han, Angew. Chem., Int. Ed., 2016, 55, 6771–6775.
153. X. W. Mao and T. A. Hatton, Ind. Eng. Chem. Res., 2015, 54, 4033–4042.
154. B. Kumar, M. Asadi, D. Pisasale, S. Sinha-Ray, B. A. Rosen, R. Haasch, J. Abiade, A. L. Yarin and A. Salehi-Khojin, Nat. Commun., 2013, 4, 2819–2826.
155. J. Wu, R. M. Yadav, M. Liu, P. P. Sharma, C. S. Tiwary, L. Ma, X. Zou, X. D. Zhou, B. I. Yakobson, J. Lou and P. M. Ajayan, ACS Nano, 2015, 9, 5364–5371.
156. J. Wu, S. Ma, J. Sun, J. I. Gold, C. Tiwary, B. Kim, L. Zhu, N. Chopra, I. N. Odeh, R. Vajtai, A. Z. Yu, R. Luo, J. Lou, G. Ding, P. J. Kenis and P. M. Ajayan, Nat. Commun., 2016, 7, 13869–13875.
157. J. Xu, Y. Kan, R. Huang, B. Zhang, B. Wang, K. H. Wu, Y. Lin, X. Sun, Q. Li and G. Centi, ChemSusChem, 2016, 9, 1085–1089.
158. W. Ju, A. Bagger, G.-P. Hao, A. S. Varela, I. Sinev, V. Bon, B. R. Cuenya, S. Kaskel, J. Rossmeisl and P. Strasser, Nat. Commun., 2017, 8, 944–953.
159. W. L. Li, M. Seredych, E. Rodriguez-Castellon and T. J. Bandosz, ChemSusChem, 2016, 9, 606–616.
160. H. M. Jhong, C. E. Tornow, B. Smid, A. A. Gewirth, S. M. Lyth and P. J. Kenis, ChemSusChem, 2017, 10, 1094–1099.
161. M. Halmann, Nature, 1978, 275, 115–116.
162. B. Aurian-Blajeni, M. Halmann and J. Manassen, Sol. Energy Mater. Sol. Cells, 1983, 8, 425–440.
163. E. E. Barton, D. M. Rampulla and A. B. Bocarsly, J. Am. Chem. Soc., 2008, 130, 6342–6344.
164. C. X. Kronawitter, M. Lessio, P. Zhao, C. Riplinger, A. Boscoboinik, D. E. Starr, P. Sutter, E. A. Carter and B. E. Koel, J. Phys. Chem. C, 2015, 119, 17762–17772.
165. E. S. Brown, S. L. Peczonczyk, Z. J. Wang and S. Maldonado, J. Phys. Chem. C, 2014, 118, 11593–11600.
166. M. Lessio, C. Riplinger and E. A. Carter, Phys. Chem. Chem. Phys., 2016, 18, 26434–26443.
167. T. P. Senftle, M. Lessio and E. A. Carter, Chem. Mater., 2016, 28, 5799–5810.
168. M. Lessio and E. A. Carter, J. Am. Chem. Soc., 2015, 137, 13248–13251.
169. M. Lessio, T. P. Senftle and E. A. Carter, ACS Energy Lett., 2016, 1, 464–468.
170. T. P. Senftle, M. Lessio and E. A. Carter, Chem. Mater., 2016, 28, 5799–5810.
171. M. Lessio, J. M. Dieterich and E. A. Carter, J. Phys. Chem. C, 2017, 121, 17321–17331.
172. B. A. Parkinson and P. F. Weaver, Nature, 1984, 309, 148–149.
173. S. Kaneco, H. Katsumata, T. Suzuki and K. Ohta, Chem. Eng. J., 2006, 116, 227–231.
174. S. Kaneco, H. Katsumata, T. Suzuki and K. Ohta, Appl. Catal., B, 2006, 64, 139–145.
175. T. Arai, S. Sato, K. Uemura, T. Morikawa, T. Kajino and T. Motohiro, Chem. Commun., 2010, 46, 6944–6946.
176. S. Sato, T. Arai, T. Morikawa, K. Uemura, T. M. Suzuki, H. Tanaka and T. Kajino, J. Am. Chem. Soc., 2011, 133, 15240–15243.
177. S. Kaneco, Y. Ueno, H. Katsumata, T. Suzuki and K. Ohta, Chem. Eng. J., 2009, 148, 57–62.
178. E. Barton Cole, P. S. Lakkaraju, D. M. Rampulla, A. J. Morris, E. Abelev and A. B. Bocarsly, J. Am. Chem. Soc., 2010, 132, 11539–11551.
179. G. T. Zeng, J. Qiu, Z. Li, P. Pavaskar and S. B. Cronin, ACS Catal., 2014, 4, 3512–3516.
180. J. Qiu, G. T. Zeng, M. A. Ha, M. Y. Ge, Y. J. Lin, M. Hettick, B. Y. Hou, A. N. Alexandrova, A. Javey and S. B. Cronin, Nano Lett., 2015, 15, 6177–6181.
181. G. Zeng, J. Qiu, B. Hou, H. Shi, Y. Lin, M. Hettick, A. Javey and S. B. Cronin, Chem.–Eur. J., 2015, 21, 13502–13507.
182. P. Li, J. Zhang, H. Wang, H. Jing, J. Xu, X. Sui, H. Hu and H. Yin, Catal. Sci. Technol., 2014, 4, 1070–1077.
183. T. T. Guaraldo, J. F. de Brito, D. Wood and M. V. B. Zanoni, Electrochim. Acta, 2015, 185, 117–124.
184. J. Cheng, M. Zhang, J. Z. Liu, J. H. Zhou and K. F. Cen, J. Mater. Chem. A, 2015, 3, 12947–12957.
185. J. Cheng, M. Zhang, G. Wu, X. Wang, J. Zhou and K. Cen, Sol. Energy Mater. Sol. Cells, 2015, 132, 606–614.
186. P. Li, J. Xu, H. Jing, C. Wu, H. Peng, J. Lu and H. Yin, Appl. Catal., B, 2014, 156, 134–140.
187. P. Li, H. Jing, J. Xu, C. Wu, H. Peng, J. Lu and F. Lu, Nanoscale, 2014, 6, 11380–11386.
188. J. Cheng, M. Zhang, G. Wu, X. Wang, J. H. Zhou and K. F. Cen, Environ. Sci. Technol., 2014, 48, 7076–7084.
189. C. Ampelli, G. Centi, R. Passalacqua and S. Perathoner, Energy Environ. Sci., 2010, 3, 292–301.
190. T. Yamamoto, H. Katsumata, T. Suzuki and S. Kaneco, ECS Trans., 2017, 75, 31–37.
191. X. Zhou, R. Liu, K. Sun, Y. Chen, E. Verlage, S. A. Francis, N. S. Lewis and C. Xiang, ACS Energy Lett., 2016, 1, 764–770.
192. Y. Xu, Y. Jia, Y. Zhang, R. Nie, Z. Zhu, J. Wang and H. Jing, Appl. Catal., B, 2017, 205, 254–261.
193. Y. H. Yang, R. R. Xie, H. Li, C. J. Liu, W. H. Liu and F. Q. Zhan, Trans. Nonferrous Met. Soc. China, 2016, 26, 2390–2396.
194. I. Ganesh, P. P. Kumar, I. Annapoorna, J. M. Sumliner, M. Ramakrishna, N. Y. Hebalkar, G. Padmanabham and G. Sundararajan, Appl. Surf. Sci., 2014, 293, 229–247.
195. M. R. Hasan, S. B. Abd Hamid and W. J. Basirun, Appl. Surf. Sci., 2015, 339, 22–27.
196. L. Wang, Y. Jia, R. Nie, Y. Zhang, F. Chen, Z. Zhu, J. Wang and H. Jing, J. Catal., 2017, 349, 1–7.
197. J. W. Jang, S. Cho, G. Magesh, Y. J. Jang, J. Y. Kim, W. Y. Kim, J. K. Seo, S. Kim, K. H. Lee and J. S. Lee, Angew. Chem., Int. Ed., 2014, 53, 5852–5857.
198. Y. J. Jang, J. W. Jang, J. Lee, J. H. Kim, H. Kumagai, J. Lee, T. Minegishi, J. Kubota, K. Domen and J. S. Lee, Energy Environ. Sci., 2015, 8, 3597–3604.
199. Y. J. Jang, I. Jeong, J. Lee, J. Lee, M. J. Ko and J. S. Lee, ACS Nano, 2016, 10, 6980–6987.
200. X. Xiong, M. Forster, J. D. Major, Y. Xu and A. J. Cowan, J. Phys. Chem. C, 2017, 121, 22073–22080.
201. M. Li, P. Li, K. Chang, H. Liu, X. Hai, H. Zhang and J. Ye, Chem. Commun., 2016, 52, 8235–8238.
202. G. Ghadimkhani, N. R. de Tacconi, W. Chanmanee, C. Janaky and K. Rajeshwar, Chem. Commun., 2013, 49, 1297–1299.
203. K. Rajeshwar, N. R. de Tacconi, G. Ghadimkhani, W. Chanmanee and C. Janaky, ChemPhysChem, 2013, 14, 2251–2259.
204. H. Homayoni, W. Chanmanee, N. R. de Tacconi, B. H. Dennis and K. Rajeshwar, J. Electrochem. Soc., 2015, 162, E115–E122.
205. X. Ba, L. L. Yan, S. Huang, J. G. Yu, X. J. Xia and Y. Yu, J. Phys. Chem. C, 2014, 118, 24467–24478.
206. J. F. de Brito, A. R. Araujo, K. Rajeshwar and M. V. B. Zanoni, Chem. Eng. J., 2015, 264, 302–309.
207. J. F. De Brito, A. A. da Silva, A. J. Cavalheiro and M. V. B. Zanoni, Int. J. Electrochem. Sci., 2014, 9, 5961–5973.
208. C. Janaky, D. Hursan, B. Endrodi, W. Chanmanee, D. Roy, D. Liu, N. R. de Tacconi, B. H. Dennis and K. Rajeshwar, ACS Energy Lett., 2016, 1, 332–338.
209. M. Schreier, J. Luo, P. Gao, T. Moehl, M. T. Mayer and M. Gratzel, J. Am. Chem. Soc., 2016, 138, 1938–1946.
210. X. Chang, T. Wang, P. Zhang, Y. Wei, J. Zhao and J. Gong, Angew. Chem., Int. Ed., 2016, 55, 8840–8845.
211. N. C. D. Nath, S. Y. Choi, H. W. Jeong, J. J. Lee and H. Park, Nano Energy, 2016, 25, 51–59.
212. E. Kecsenovity, B. Endrődi, Z. Pápa, K. Hernádi, K. Rajeshwar and C. Janáky, J. Mater. Chem. A, 2016, 4, 3139–3147.
213. B. Blackburn, I. Hassan, C. Zhang, C. Blackman, K. Holt and C. Carmalt, J. Nanosci. Nanotechnol., 2016, 16, 10112–10116.
214. X. Yang, E. A. Fugate, Y. Mueanngern and L. R. Baker, ACS Catal., 2017, 7, 177–180.
215. U. Kang, S. K. Choi, J. H. Dong, M. J. Sang, W. Choi, S. H. Dong, A. Abdel-Wahab and H. Park, Energy Environ. Sci., 2015, 8, 2638–2643.
216. A. A. Tahir, H. Ullah, P. Sudhagar, M. A. M. Teridi, A. Devadoss and S. Sundaram, Chem. Rec., 2016, 16, 1591–1634.
217. S. Zhou, Y. Liu, J. M. Li, Y. J. Wang, G. Y. Jiang, Z. Zhao, D. X. Wang, A. J. Duan, J. Liu and Y. C. Wei, Appl. Catal., B, 2014, 158, 20–29.
218. Y. M. He, L. H. Zhang, M. H. Fan, X. X. Wang, M. L. Walbridge, Q. Y. Nong, Y. Wu and L. H. Zhao, Sol. Energy Mater. Sol. Cells, 2015, 137, 175–184.
219. Y. He, L. Zhang, B. Teng and M. Fan, Environ. Sci. Technol., 2015, 49, 649–656.
220. J. C. Wang, H. C. Yao, Z. Y. Fan, L. Zhang, J. S. Wang, S. Q. Zang and Z. J. Li, ACS Appl. Mater. Interfaces, 2016, 8, 3765–3775.
221. L. M. Azofra, D. R. MacFarlane and C. H. Sun, Phys. Chem. Chem. Phys., 2016, 18, 18507–18514.
222. N. Sagara, S. Kamimura, T. Tsubota and T. Ohno, Appl. Catal., B, 2016, 192, 193–198.
223. D. Zhang, J. Shi, W. Zi, P. Wang and S. Liu, ChemSusChem, 2017, 10, 4324–4341.
224. Y. Zhang, W. Luc, G. S. Hutchings and F. Jiao, ACS Appl. Mater. Interfaces, 2016, 8, 24652–24658.
225. K. D. Yang, Y. Ha, U. Sim, J. An, C. W. Lee, K. Jin, Y. Kim, J. Park, J. S. Hong, J. H. Lee, H. E. Lee, H. Y. Jeong, H. Kim and K. T. Nam, Adv. Funct. Mater., 2016, 26, 233–242.
226. J. T. Song, H. Ryoo, M. Cho, J. Kim, J. G. Kim, S. Y. Chung and J. Oh, Adv. Energy Mater., 2016, 7, 1601103–1601111.
227. Y. Wang, S. Fan, B. AlOtaibi, Y. Wang, L. Li and Z. Mi, Chemistry, 2016, 22, 8809–8813.
228. F. Urbain, P. Tang, N. M. Carretero, T. Andreu, L. G. Gerling, C. Voz, J. Arbiol and J. R. Morante, Energy Environ. Sci., 2017, 10, 2256–2266.
229. T. Sekimoto, H. Hashiba, S. Shinagawa, Y. Uetake, M. Deguchi, S. Yotsuhashi and K. Ohkawa, J. Phys. Chem. C, 2016, 120, 13970–13975.
230. N. J. Firet and W. A. Smith, ACS Catal., 2016, 7, 606–612.
231. S. Zhu, B. Jiang, W.-B. Cai and M. Shao, J. Am. Chem. Soc., 2017, 139, 15664–15667.
232. D. Ren, Y. Deng, A. D. Handoko, C. S. Chen, S. Malkhandi and B. S. Yeo, ACS Catal., 2015, 5, 2814–2821.

---

Footnote

† These authors have contributed equally to this work.

---