Role of Interfaces in Two-Dimensional Photocatalyst for WaterSplittingTongming Su,†,‡,§ Qian Shao,∥,§ Zuzeng Qin,*,† Zhanhu Guo,*,§ and Zili Wu*,‡

†School of Chemistry and Chemical Engineering, Guangxi University, Nanning 530004, China‡Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United States§Integrated Composites Laboratory (ICL), Department of Chemical & Biomolecular Engineering, University of Tennessee, Knoxville,Tennessee 37996, United States∥College of Chemical and Environmental Engineering, Shandong University of Science and Technology, Qingdao 266590, China

ABSTRACT: Hydrogen generation from the direct splitting of waterby photocatalysis is regarded as a promising and renewable solutionfor the energy crisis. The key to realize this reaction is to find anefficient and robust photocatalyst that ideally makes use of the energyfrom sunlight. Recently, due to the attractive properties such asappropriate band structure, ultrahigh specific surface area, and moreexposed active sites, two-dimensional (2D) photocatalysts haveattracted significant attention for photocatalytic water splitting. ThisReview attempts to summarize recent progress in the fabrication andapplications of 2D photocatalysts including graphene-based photo-catalysts, 2D oxides, 2D chalcogenides, 2D carbon nitride, and someother emerging 2D materials for water splitting. The constructionstrategies and characterization techniques for 2D/2D photocatalystsare summarized. Particular attention has been paid to the role of 2D/2D interfaces in these 2D photocatalysts as the interfacesand heterojunctions are critical for facilitating charge separation and improving photocatalysis efficiency. We also critically discusstheir stability as photocatalysts for water splitting. Finally, we highlight the ongoing challenges and opportunities for the futuredevelopment of 2D photocatalysts in this exciting and still emerging area of research.

KEYWORDS: photocatalyst, interface, hydrogen, two-dimensional, water splitting

1. INTRODUCTION

With the development of modern society, environmentalpollution and energy shortage have become the focus ofworld attention. A majority of the global energy supplies aregenerated from fossil fuel, which gives rise to environmentalpollution and climate change.1−3 Therefore, the developmentof clean and renewable energy is the key way to meet theincreasing global energy requirement and to resolve theenvironmental problems caused by the overuse of largeamounts of fossil fuels. Hydrogen possesses the highest energycontent per weight among the combustion fuels and producesonly water as the product. Hence, hydrogen is regarded as anultraclean, powerful, environmentally friendly, and promisingalternative for meeting future fuel needs.4

Given the natural abundance of water and sunlight, theproduction of hydrogen from water by using sunlight has beenproven to be a regenerative, eco-friendly, and inexhaustibleapproach to solve both the energy crisis and energy-relatedenvironmental pollution. In the photocatalysis process, a stableand efficient photocatalyst is the critical factor to achieve a highyield of hydrogen. Because TiO2 electrodes were used forphotocatalytic water splitting by Fujishima and Honda in1972,5 a wide variety of photocatalysts have been developed for

solar energy conversion in the past decades, such as TiO2, CdS,C3N4, and so on.6−10 Up to now, numerous semiconductorphotocatalysts were exploited and used for photocatalytic watersplitting. According to the composition, the photocatalysts canbe generally categorized into three types: metal oxides, metalchalcogenides, and metal-free photocatalysts. Despite the rapiddevelopment of these photocatalysts, they still face severalsignificant challenges: (1) Many semiconductors, especiallymetal oxides, can only absorb the ultraviolet light due to theirwide band gap.11 (2) Some semiconductors are not suitable foroverall water splitting due to their improper band position andonly exhibit either water reduction or oxidation activity.12 (3)During the migration of photogenerated charge carries to thesurface reactive sites, the charge recombination occurs easily inthe bulk and on the surface of photocatalysts.13 (4) Most of thereaction active sites in the bulk semiconductor cannot beexposed to the surface and therefore cannot be used forphotocatalytic reaction.14

Received: October 9, 2017Revised: January 29, 2018Published: January 30, 2018

Review

pubs.acs.org/acscatalysisCite This: ACS Catal. 2018, 8, 2253−2276

© XXXX American Chemical Society 2253 DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

To address these challenges, development of new and moreefficient photocatalysts is needed and has been activelyexplored in the field. Among the new photoactive materials,two-dimensional (2D) materials have recently attracted muchattention. 2D materials represent an emerging class of materialsthat possess sheet-like structures with the thickness of onlysingle or few atoms.15 The effort was sparked by the discoveryof graphene in 2004, a single-layer carbon material withexcellent electrical, thermal, and mechanical properties.16 Sincethen, various graphene-like 2D photocatalysts have becometopical subjects in the photocatalysis field. 2D photocatalystsshow different physical and chemical properties compared withtheir bulk counterparts. 2D structures with exotic electronicproperties and a high specific surface area can be producedfrom layered materials, which are characterized by strong in-plane bonds and a weak van der Waals force between thelayers.17 Because of the unusual structural, physical, andchemical properties of 2D materials, the fabrication of a few-layer or single-layer 2D structure has aroused wide interest aspromising photocatalysts with several merits: (1) The band gapand the light absorption of 2D semiconductor can be adjustedby tuning the number of layers.18,19 (2) The recombination ofelectrons and holes in the bulk can be reduced due to theultrathin nature.20,21 (3) Specific surface area of semi-conductors is greatly improved, and most of the active sitescan be exposed on the surface and take part in thephotocatalytic reaction.22 Among many 2D photocatalystswith just a few layers or single-layer structure, graphene-basedphotocatalysts, 2D oxides, 2D chalcogenides, 2D graphiticcarbon nitride (g-C3N4), and other 2D semiconductors beginto draw great attention in photocatalysis recently. The band gapenergies, band position, and thickness of some 2D photo-catalysts are shown in Figure 1.23−34

Although 2D photocatalysts are regarded as the promisingcandidates to convert the solar energy into chemical energy asthe form of hydrogen, there are several hurdles that limit theirapplications, such as the following: (1) The exciton bindingenergy within 2D photocatalysts was greatly increased due tomuch smaller electron screening than the bulk material, whichis unfavorable for the photocatalytic performance.35 (2) Some2D semiconductors are not stable in air or aqueous solution,the thin-layer 2D semiconductor can be flocked together oroxidized by photogenerated holes during the reaction, leadingto the decrease of the photocatalytic activity.36 (3) Although

the recombination of electron−hole pairs is less than that of thebulk semiconductors, it still exists on the 2D photocatalysts.37

(4) The oxidation and reduction potential of some 2Dsemiconductors are insufficient for the overall water splitting.37

To overcome these problems, a variety of strategies havebeen developed to improve the photocatalytic efficiencies of 2Dphotocatalysts, such as doping with a metal or nonmetalelement, inducing defects, and coupling with metal orsemiconductors.21,38−41 In fact, the photoactivity of photo-catalysts not only depends on their properties, such as crystalstructure, band structure, and electron affinity, but also on theinterface between the photocatalyst and cocatalyst. Therefore,to achieve efficient conversion of water into hydrogen, thespatial integration of photocatalyst and cocatalyst to form theintimate interface is of great importance. The intimate interfacecan optimize the light absorption of photocatalysts andpromote the separation of charge carriers. Commonly, thelarger contact area on the interface can provide sufficient chargetransfer and trapping channels for the separation of photo-generated electron−hole pairs.42

As compared with 0D−1D, 1D−1D, 0D−2D, and 1D−2Dinterfaces, the 2D−2D coupled interfaces have attracted wideattention in photocatalysis because of their special advantages.17

First, the formation of an intimate interface between twosemiconductors is in favor of the exciton dissociation, thusenhancing the photocatalytic quantum efficiency.43 Second, it isfacile and efficient to form the intimate interface between 2Dsemiconductors, even when the 2D semiconductors have highlattice mismatch.44,45 Third, the large lateral size with highsurface area leads to large contact area in 2D/2D photo-catalysts, which promotes the charge transfer and the separationof electron−hole pairs.46 Fourth, the band potential can bematched for the overall water splitting by integrating ahydrogen evolution photocatalyst and an oxygen evolutionphotocatalyst. Thus, the oxidation power and reduction powerof the semiconductors can be balanced for overall watersplitting.47,48 Finally, the formation of 2D/2D heterostructure isbeneficial to the improvement of catalyst stability due to thealleviation of photocorrosion and agglomeration.23,49 Withthese advantages of the 2D/2D interface, plenty of 2D/2Dstructures were fabricated in recent years to improve thephotocatalytic performance of photocatalysts.Some excellent reviews have been published in the past few

years on 2D materials that have focused on the synthesis

Figure 1. Band gap energies, band position, and thickness of several 2D photocatalysts related to the redox potentials of water splitting.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2254

methods, properties, catalytic applications, and routes to tunetheir electronic states and active sites.22,50−65 In addition, somecomprehensive reviews with respect to the classification,modification, computational screening, and application of 2Dphotocatalysts were also reported.14,17,36,37,66−77 However, noreview has been centered on the role of interfaces in 2Dmaterials for photocatalysis, including the interface types in 2Dphotocatalysts, construction strategies, characterization techni-ques of various 2D/2D interfaces, and the function of differentcomponents in 2D/2D photocatalyst systems. Therefore, asystematic review focusing on the rational design andconstruction of 2D/2D interface in 2D photocatalysts isnecessary for readers to better understand the latest progress inthe field of 2D photocatalysts. This Review will focus on the 2Dphotocatalysts for photocatalytic water splitting, including thegraphene-based photocatalysts, 2D oxides, 2D chalcogenides,2D graphitic carbon nitride, and a few other 2D semi-conductors. The 2D materials we reviewed here refer to lowdimensional materials with the thickness ranging from singlelayer to a few nanometers with the basal plane dominating thetotal surface area. The photocatalysis principles, synthesis, andstability of these 2D materials will be briefly reviewed. Besides,the construction strategies and characterization techniques of2D/2D photocatalysts are also summarized. More attention willbe paid to the types and the role of interfaces in these 2Dphotocatalysts for water splitting. Finally, the ongoingchallenges and opportunities for the future development of2D photocatalysts in this exciting yet still emerging area ofresearch will be proposed.

2. BASIC PRINCIPLES OF 2D PHOTOCATALYSTS FORWATER SPLITTING2.1. Photocatalysis Principles for Water Splitting on

2D Photocatalysts. Figure 2 shows a schematic illustration of

the basic principle of overall water splitting on a semiconductorphotocatalyst. Under light irradiation, electrons in the valenceband are excited into the conduction band, leaving holes in thevalence band. The excited electrons can cause the H+ reductionreaction to generate H2, while the holes can cause H2Ooxidation reaction to form O2. To realize overall water splittingon the semiconductor, the bottom of the conduction bandmust be more negative than the reduction potential of H+/H2(0 V vs NHE at pH = 0), and the top of the valence band mustbe more positive than the oxidation potential of H2O/O2 (1.23V vs NHE at pH = 0). Furthermore, the band gap of thesemiconductor must exceed the free energy (1.23 eV) of watersplitting. During the photocatalytic water splitting reaction, the

recombination of the photogenerated electron−hole pairs canoccur in the bulk and on their way to the surface of the bulkphotocatalyst.78 The recombination of charge carriers will lowerthe efficiency of the photocatalysts.To effectively separate the photogenerated electron−hole

pairs, the proportion of exposed surface of photocatalyst shouldbe increased and the distance of charge transfer should beshortened to accelerate water splitting. In this respect, 2Dphotocatalysts can reduce the recombination of charge carriersin the bulk, and the electrons and holes can directly transfer tothe surface active sites due to their ultrathin nature. Generally,with the transition from 3D to 2D, the band gap increasesbecause of the quantum confinement effect, which causes ablue-shift of the optical absorption of 2D photocatalysts.79 Onthe other hand, with a larger band gap, the valence band andconduction band will enlarge in an opposite direction. The shiftof the valence band and conduction band will enhance thereducing power and oxidizing power of the photocatalysts,respectively. In addition, the Coulomb interactions (excitonbinding energy) are greatly enhanced in two-dimensionalsystems as a result of spatial confinement and reducedCoulomb screening.80,81 Consequently, the electron−holepairs can be bounded by the Coulomb force and go againstthe photocatalytic reaction. Therefore, strategies need to bedeveloped to enhance the photocatalytic performance of 2Dphotocatalysts, such as the construction of 2D/2D hetero-structures.

2.2. Types of Interfaces in 2D Photocatalysts. Despitethe potential merits of 2D structure, the recombination ofseparated carriers can still exist on 2D photocatalysts. Inaddition, the excitonic effect, which is caused by the Coulombinteractions between photogenerated electrons and holes, playsan important role in the excitation properties of semi-conductors. Because photocatalysis is a typical photoexcitationprocess, the excitonic effect should be taken into account in thephotocatalytic reaction. The bound electron−hole pairs areformed with the generation of excitons, which adversely affectthe separation of electrons and holes and thus decrease thephotocatalytic activity. Several strategies have been used topromote the separation of the electron−hole pairs in thephotocatalysts, such as the construction of metal/2D, doped2D, 2D/2D structure, and so on, among which 2D/2Dstructure may be the most promising.17 The intimate interfacesin the 2D/2D photocatalysts will accelerate the charge carriertransfer between them, thus suppressing the recombination ofelectron−hole pairs. Two main ways to construct the 2D/2Dsemiconductor interface are the intraplane interface and theinterplane interface, as shown in Figure 3.Hybrid photocatalysts, consisting of at least two 2D

materials, are considered as a promising system to promotethe separation of photogenerated electrons and holes, and toextend the light absorption toward visible-light region. These2D/2D photocatalytic systems can be classified as the Z-

Figure 2. Schematic illustration of the basic principles of photo-catalytic water splitting. CB, conduction band; VB, valence band; Eg,band gap.

Figure 3. Intraplane and interplane interface models of the 2D/2Dphotocatalyst for photocatalytic water splitting.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2255

scheme system (such as α-Fe2O3/g-C3N482), the Schottky

junction system (such as GeH/graphene83), the Type I (suchas g-C3N4/ZnIn2S4

49) and Type II (such as g-C3N4/N−La2Ti2O7

84) heterostructure system, as shown in Figures 4 and5.

The Z-scheme system, inspired by natural photosynthesis ingreen plants, is an effective approach for photoinduced chargeseparation and water splitting.85 This system uses twosemiconductors to achieve band alignment. In this way, thephotocatalyst that has either water reduction or oxidationactivity can be applied to one side of the Z-scheme system toachieve the separation of photoinduced electron−hole pairs andoverall water splitting.41,82,85 For example, by combining theLa- and Rh-codoped SrTiO3 (possess reduction activity) withthe Mo-doped BiVO4 (possess oxidation activity), photo-catalytic water splitting simultaneously into H2 and O2 wasrealized.86 For the Z-scheme system, as shown in Figure 4, theelectrons in the CB of A semiconductor can transfer to the VBof B semiconductor to combine with holes, while the electronsin the CB of B semiconductor are used to reduce H2O into H2,and the holes in the CB of A semiconductor are used to oxidizeH2O into O2.Construction of Schottky junction by fabricating the 2D/2D

interface is another effective way to suppress the recombinationof photogenerated electron−hole pairs.87,88 For this system,electrons can flow from the photocatalyst to the cocatalystthrough the interface (from the higher to the lower Fermi level)to align the Fermi energy levels.89 Thus, the cocatalyst hasexcess negative charges, while the semiconductor possessespositive charges near the interface. Consequently, a spacecharge layer is formed at the interface, and the CB and VB ofthe semiconductor are bent “upward”. This space charge layerleads to the formation of a Schottky junction between the

semiconductor and cocatalyst.87 As a result, the Schottkyjunction can serve as an electron trap to efficiently capture thephotoinduced electrons and the photocatalytic activity can beenhanced.87,89,90

2D/2D heterostructured photocatalysts have been widelyapplied to the photocatalytic water splitting reaction. Herein,we discuss these two main types of the heterostructure system(i.e., Type I and Type II; Figure 5). In Type I, the holes andelectrons will accumulate on the B semiconductor because theCB (VB) of the B semiconductor is more positive (negative)than that of the A semiconductor.91 Type II has similar bandalignment as the Z-scheme system, but with different electrontransfer directions. For the Type II system, the CB electrons ofB semiconductor transfer to the CB of A and the VB holes of Atransfer to the VB of B; while the CB electrons of the Asemiconductor transfer to the VB of B in the Z-scheme system.The p−n junction is a typical Type II system, which iscomposed of alternate p- and n-type photocatalysts.92,93

Different electronic structures between p- and n-type materialresult in the accumulation of photoexcited electrons at thecoupling interface of the n-type side and holes in the p-typecomponent. Thus, concentration gradients of electrons andholes are formed at the interface, which promotes thetransportation of charge carriers and suppresses the recombi-nation of photogenerated electrons and holes, and thus favorsan efficient solar-energy conversion.94

So far, many types of 2D junctions were developed to modifythe optical property, electronic property, and the stability of 2Dphotocatalysts, such as the metal−2D, doped-2D, 1D−2D,2D−2D, and so on. In this Review, we will focus on the Z-scheme, Schottky junction, Type I and Type II heterostructuresystems with the 2D/2D interfaces.

2.3. Construction of 2D/2D Photocatalysts. Thestrategy for constructing the 2D/2D interfaces plays animportant role on the photocatalytic performance of the 2Dphotocatalysts, because the quality of interface is determined byconstruction method. To date, many effective strategies havebeen explored to fabricate the 2D/2D intimate interface in the2D photocatalysts, such as the ultrasonic absorption,95 thehydrothermal method,96 the electrostatic self-assembly,97 andthe chemical vapor deposition.98

One of the simplest ways to construct the 2D/2Dphotocatalyst is to disperse two different 2D components inthe solution under stirring or sonication to form a mixture. Themixture was dried in an oven to evaporate the solvents, andthen the 2D/2D photocatalyst was obtained. For example, 2D/2D CdS/MoS2 and g-C3N4/N-doped La2Ti2O7 2D layeredcomposites were fabricated by ultrasonic dispersion anddrying.84,95,99 However, these 2D/2D interfaces were formedby the weak interaction between two 2D components by usingthis method, the 2D components can be separated easily duringthe photocatalysis process. For instance, after four recycles,almost 35% of H2-production activity loss was observed on 2D/2D CdS/MoS2 (construction by ultrasonic absorption), whichmight be ascribed to the slow falloff of MoS2 from CdS.95

The hydrothermal method is a practical way to construct a2D/2D intimate interface between the 2D components. Inbrief, one 2D component was mixed with the precursor of theother 2D component first, then the solution was transferred toan autoclave for the hydrothermal reaction. Thereafter, the 2D/2D photocatalysts were collected after drying. For example, the2D/2D p-MoS2/n-rGO photocatalyst was synthesized by thehydrothermal method, and many intimate interfaces were

Figure 4. Schematic diagram of charge transfer for Z-scheme andSchottky junction system.

Figure 5. Schematic diagram of charge transfer for Type I, IIheterostructure system.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2256

formed between the p-MoS2 and the n-rGO. The intimate p-MoS2/n-rGO junctions greatly improve the generation ofelectron−hole pairs and suppress their recombination.96 g-C3N4/ZnIn2S4,

49 g-C3N4/CaIn2S4,91 TiO2-g-C3N4,

23 et al. werealso successfully prepared by the hydrothermal method.Hydrothermal is an energy-saving and environmentally friendlymethod, because the reaction takes place in closed-systemconditions. Additionally, the hydrothermal process is kineticallyslow at any given temperature and thus easy to be controlled. Amicrowave-hydrothermal process was developed to increase thekinetics of crystallization.100 2D/2D TiO2/graphene wassynthesized by a simple microwave-hydrothermal process andshowed an enhanced photocatalytic H2-production activitycompared with TiO2.

101

Electrostatic self-assembly is also a viable strategy to fabricatethe intimate 2D/2D interfaces in 2D photocatalysts.97,102 Forthis method, surface charge modification plays a significant rolein engineering the 2D/2D photocatalysts with intimateinterfacial contact. To attain the electrostatic self-assembly,the surface charges on different 2D photocatalysts need to betuned by charge modification to obtain the opposite charges(i.e., positive and negative charges). Notably, the zeta potentialvalue can be used to determine the charge of the photocatalyst.For example, to form the 2D/2D g-C3N4/rGO by electrostaticself-assembly, the g-C3N4 was protonated by concentratedH2SO4 and HNO3 under mild ultrasonication to obtain thepositively charged surface, the zeta potential value wasmeasured to be +37.2 mV. Then the protonated g-C3N4 wasmixed with negatively charged rGO (−32.4 mV), with theassistance of ultrasonication and agitation, the g-C3N4/rGOwith the intimate interface was obtained. Moreover, the π-stacking interactions between the sp2 lattices of g-C3N4 and thesp2 graphene lattices as well as the hydrogen-bondinginteractions between the nitrogen-contained groups in g-C3N4are also beneficial for the electrostatic self-assembly. Thehydrogen production rate of the g-C3N4/rGO (557 μmol g−1

h−1) was much higher than that of g-C3N4 (158 μmol g−1h−1)due to the intimate interface and the introduced rGO.97 Inaddition, the construction of the 2D/2D g-C3N4/rGO byelectrostatic self-assembly also facilitates the photocatalyticreduction of carbon dioxide to methane (13.93 μmol g−1),higher activity when compared with that of rGO/pure g-C3N4(8.29 μmol g−1) without the modification of surface charge ong-C3N4.

102

Chemical vapor deposition (CVD) is an effective techniqueto construct the 2D/2D heterostructures with intimateinterfaces, including the intraplane and interplane interfa-

ces.98,103 Generally, mixed gas molecules are injected into areaction chamber which is set at a certain temperature. Whenthe mixed gases come into contact with the substrate within thereaction chamber, a reaction occurs that creates a material filmon the substrate surface. The temperature of the substrate is akey factor for the quality of the obtained 2D material. Forexample, the intraplane 2D/2D WS2/MoS2,

98 MoS2/MoSe2104

and interplane MoS2/WS2,103 MoS2/WSe2

105 were fabricatedby CVD method and showed good optoelectronic andphotovoltaic performance. Although CVD is a powerfultechnique for the construction of 2D/2D materials, the gaseousbyproducts of the process are usually very toxic. Besides, the 2Dmaterials have to stay on the substrate for further use, which isnot efficient in the photocatalytic reaction. Moreover, it is still agreat challenge to synthesize the 2D materials on a large scaleby using the CVD method.

2.4. Characterization of Interfaces in 2D/2D Photo-catalysts. The quality of the interface in 2D/2D photocatalystsdirectly determines the activity and stability of the 2D/2Dphotocatalysts. Therefore, it is very important to investigate thequality and interaction between the 2D components. Powerfulcharacterization techniques have been developed to character-ize the interface in 2D materials, such as the transmissionelectron microscopy (TEM), X-ray photoelectron spectroscopy(XPS), X-ray absorption spectroscopy (XAS), conductiveatomic force microscopy (CAFM), and Kelvin probe forcemicroscopy (KPFM).TEM is a useful technique to examine the size, exposed

facets, crystallinity, phase, crystal orientation, and exposedcrystal facets of materials. Low-magnification TEM images canbe used to explore the lateral size and roughly measure thethickness of the nanosheet. Furthermore, the high-resolutionTEM (HRTEM) image is a practical way to confirm the crystalorientation and the exposed crystal facets of crystallinenanosheet by measuring the lattice distance. Generally, differentcrystalline materials have different crystal orientations andlattice distances, and no lattice fringes can be observed fromamorphous nanosheet. On the basis of this distinction,HRTEM can be used to explore the interfaces between the2D/2D materials. Take the BiOI/La2Ti2O7 heterojunction asan example, the low-magnification TEM images show that BiOInanosheets with a smaller size are randomly dispersed on thesurface of the larger La2Ti2O7 nanosheet (Figure 6a). Thelattice fringes of La2Ti2O7 and BiOI can be clearly observed inthe HRTEM image (Figure 6b), and the measured latticedistances of 0.276 and 0.302 nm correspond to the (002) planeof La2Ti2O7 and (012) plane of BiOI, respectively. Thus, an

Figure 6. TEM and HRTEM images of BiOI/La2Ti2O7. Reproduced with permission from ref 106. Copyright 2016 Royal Society of Chemistry.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2257

obvious interface was observed between the La2Ti2O7 andBiOI.106 The lateral interface between MoS2 and ZnIn2S4 canbe observed in the 2D/2D MoS2/Cu-ZnIn2S4 photocatalyst.

42

Besides, the interface is easily detected in the 2D/2D g-C3N4/MoS2 and g-C3N4/ZnIn2S4, because obvious lattice fringes canbe observed on MoS2 or ZnIn2S4 but not on g-C3N4.

49,107 It isworth noting that the electron irradiation of TEM from highvoltage might decompose or change the structure of ultrathin2D nanomaterials. Moreover, the normal HRTEM is unable todirectly probe and localize different atoms in 2D nanomaterialsto find a clearer interface. Scanning transmission electronmicroscopy (STEM) should be a good choice to separate atomsin 2D materials especially after the introduction of aberration-corrected optics. Especially, annular dark-field scanning trans-mission electron microscope (ADF-STEM) is able to obtainthe atomic resolution images, because the contrast of the atomsis proportional to the atomic number of the atoms. Forexample, the ADF-STEM was used to identify the differentphases of MoS2 in the single-layer MoS2, the atomically sharpphase interface between the 2H phase and 1T phase can beclearly observed in the ADF-STEM image.108

XPS is a surface-sensitive and spectroscopic technique toidentify the elemental composition of a material, as eachelement has its own characteristic binding energy peaks. Inaddition, the peak shape for each element is relevant to theelectron configuration within the atoms. Therefore, thechemical composition, and electronic state of a material canbe determined by analyzing the position and shape of XPSpeaks for each element. Intriguingly, XPS can also be used to

identify the different phases and quantitatively calculate theconcentration of each phase in ultrathin 2D materials. As atypical example, the concentration of 2H and 1T phases in theexfoliated MoS2 nanosheet can be determined by XPS on thebasis of the intensity and location of the element peaks.109 Inmost cases, the formation of the interface in the 2D/2Dmaterials will cause a change in the electron configurationbecause of the interaction between the 2D materials.Consequently, the interaction intensity between the 2Dmaterials can be examined by XPS. For example, the bindingenergy of Mo in 2D/2D MoS2/g-C3N4 is lower than pure Moin pure MoS2 due to the strong chemical interaction (electroncoupling) between the MoS2 layers and the conjugated g-C3N4layers.110

X-ray absorption spectroscopy (XAS) is an effective andnondestructive spectroscopic technique to explore the struc-tural characteristics of materials at the atomic scale. The X-rayabsorption spectrum includes two regimes: X-ray absorptionnear-edge spectroscopy (XANES) and extended X-rayabsorption fine-structure spectroscopy (EXAFS). XANES canbe used to characterize the coordination chemistry andoxidation state of an atom, while EXAFS is used to determinethe distances, coordination number, and neighboring species ofan atom.111,112 XAS is also a feasible way to study theinteractions between different components in 2D hybridmaterials. As a typical example, Qiao et al. used the highlysensitive synchrotron-based XANES to disclose the interactionsbetween the few-layer phosphorene nanosheet (FPS) and CdS.When the FPS was combined with the CdS (1P-CS), an

Table 1. Summary of 2D Photocatalysts for Photocatalytic Hydrogen Evolution

photocatalyst (thickness) synthesis method cocatalyst, sacrificial agent light source H2 production ratequantumefficiency ref

black TiO2 (2 nm) solvothermaltreatment

2 mL 10−5 mol L−1

H2PtCl6, 10 vol %methanol

300 W Xe-lamp 400 μmol h−1 - 118

single-atom Rh/TiO2 (0.7 nm) thermal annealing 10 vol % methanol 500 W Xe-lamp 51 μmol h−1 - 1192H MoS2 (1.4 nm) exfoliation/

thermalannealing

0.01 M lactic acid 200 W Hg-lamp (>400 nm) 0.3 mmol g−1 h−1 2.57% at400 nm

87

CdS (∼4 nm) sonication 0.5 M Na2S−Na2SO3 300 W Xe-lamp (>420 nm) 41.1 mmol g−1 h−1 1.38% at420 nm

120

C3N4 (0.5 nm) solvothermalexfoliation

3 wt % Pt, 10 vol %triethanolamine

300 W Xe-lamp (>420 nm) 8510 μmol g−1 h−1 5.1% at420 nm

27

O-g-C3N4 (∼0.64 nm) thermal oxidationetching

3 wt % Pt, 10 vol %triethanolamine

300 W Xe-lamp (>400 nm) 8874 μmol g−1 h−1 13.7% at420 nm

28

g-C3N4 (∼2 nm) sonication 3 wt % Pt, 10 vol %triethanolamine

300 W Xe-lamp (>420 nm) 93 μmol h−1 3.75% at420 nm

121

carbon nitride (∼3.6 nm) sonication 3 wt % Pt, 10 vol %methanol

(>420 nm) 53 μmol h−1 8.57% at420 nm

122

black BiOCl (3 nm) hydrothermal 10 vol % triethanolamine 300 W Xe-lamp (>420 nm) 2.51 μmol h−1 - 123Fe-BiOCl (3.7 nm) hydrothermal 0.1 M Na2SO4 300 W Xe-lamp (>420 nm) 3.54 μmol h−1 - 30C-BN (3−4 nm) thermal annealing 1 wt % Pt, 10 vol %

triethanolamine300 W Xe-lamp 5.6 μmol h−1 0.54% at

405 nm31

CoOOH (1.5 nm) sonication 0.5 M Na2SO3 300 W Xe-lamp 1200 μmol g−1 h−1 6.9% at420 nm

124

O-doped ZnIn2S4 (6 nm) hydrothermal 0.25 M Na2SO3/0.35 M Na2S

300 W Xe-lamp (>420 nm) 2120 μmol g−1 h−1 - 125

SnNb2O6 (∼3.0 nm) hydrothermal 0.3 wt % Pt, 10 vol % lacticacid

300 W Xe-lamp (>400 nm) 13.2 μmol h−1 0.43% at420 nm

32

HNb3O8 (1.3 nm) proton-exchange/exfoliation

10 vol % triethanolamine 125 W Hg-lamp ∼95 μmol h−1 - 126

HNbWO6 (1.8−2.0 nm) intercalationassisted

exfoliation

1 wt % Pt, 10 vol %triethanolamine

300 W Xe-lamp (300−700nm)

158.9 μmol h−1 - 33

Rh/calcium niobate (∼2.8−3.0nm)

thermal annealing 10 vol % methanol 500 W Xe lamp 384.8 μmol h−1 65% at300 nm

127

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2258

apparent shift of the P K edge toward low photon energy wasobserved compared with that of pure FPS. In addition, the S Ledge for 1P-CS shifted toward high photon energy comparedwith that of CdS. Therefore, the strong electronic couplingbetween the FPS and the CdS was confirmed. This stronginteraction can be attributed to the coordination bonds formedbetween the Cd atoms in CdS and the P atoms in FPS with alone pair of electrons.113 The XAS can also be used to analyzethe interlayer C−N interaction between two monolayers of g-C3N4.

114

Atomic force microscopy (AFM) is generally utilized tomeasure the thickness of 2D materials and provide thetopographic information; however, it is not useful at character-izing the interface in 2D/2D materials. Conductive atomic forcemicroscopy (CAFM), a variation of AFM, is an effective way toexplore the conductivity of 2D material with a high spatialresolution. The interface between two 2D materials can beclearly observed from their conductivity mapping. For example,with the significant differences in the electric conductivity of 1Tand 2H phases of MoS2, AFM conductivity was measured tomap the geographical distribution of 2H and 1T-like phases in asingle-flake MoS2. The conductive 1T-like phase of MoS2 wasobviously observed in the basal plane while the less-conductive2H phase was formed predominately at the edge portion, thusthe intimate interfaces formed between the different MoS2phases can be confirmed. The intimate 1T/2H interface

facilitates the charge separation and is favorable for thehydrogen evolution from water splitting.87 Of note, theCAFM method requires the use of a conductive sample.The surface potential of a catalyst can be altered by the

interfacial interaction between the 2D materials and is notaffected by the thickness of the materials but by the interfacestructure. Therefore, the detection and control of the potentialgradient on the catalyst surface are imperative to expose theinterface structure and enhance the photocatalytic performanceof the photocatalyst. Kelvin probe force microscopy (KPFM) isa powerful technology to study the surface potential and revealthe interfacial interaction between 2D materials.115−117 As atypical example, on the basis of the surface potential differencebetween the calcium niobate (CNO) and nickel oxide (NiO),Ida et al. used the KPFM to study the surface potential of the2D/2D CNO/NiO. They found that the surface potential ofthe CNO/NiO junction part is higher than that of NiO andlower than that of CNO. They further confirmed that thephoto-oxidation sites are on the CNO/NiO junction parts,while the photoreduction sites are on the nonjunction partsand/or their edges. Furthermore, the ultrathin (ca. 2 nm)junction structure of 2D CNO/NiO was verified by low energyion scattering spectroscopy (LEIS).117

Table 2. Summary of 2D/2D Photocatalysts for Photocatalytic Hydrogen Evolution

photocatalyst synthesis method cocatalyst, sacrificial agent light sourceH2 production

ratequantumefficiency ref

p-MoS2/n-rGO hydrothermal 50 vol % ethanol 300 W Xe-lamp 24.8μmol g−1 h−1

13.6% at400 nm

96

pGCN-5 wt % rGO electrostatic self-assembly 10 vol % triethanolamine 300 W Xe-lamp(>420 nm)

557μmol g−1 h−1

- 97

h-BN/graphene incipient wetness/thermalannealing

3 wt % Pt, 10 vol %triethanolamine 300 W Xe-lamp 0.57 μmol h−1 0.3% at380 nm

128

MoS2-graphene/ZnIn2S4 hydrothermal 0.25 M Na2SO3−Na2S 300 W Xe-lamp (>420 nm)

4169μmol g−1 h−1

- 129

TiO2/MoS2 hydrothermal 10 vol % methanol 300 W Xe-lamp 2145μmol g−1 h−1

6.4% at360 nm

130

TiO2/graphene microwave-hydrothermal 25 vol % methanol 350 W Xe-lamp 736μmol g−1 h−1

3.1% at365 nm

101

TiO2-g-C3N4 hydrothermal 3 wt % Pt, 10 vol % triethanolamine 300 W Xe-lamp 18.2mmol g−1 h−1

5.3% at380 nm

23

N−La2Ti2O7/g-C3N4 sonication 2 wt % Pt, 20 vol % methanol Xe-lamp 500mW cm−2

430μmol g−1 h−1

2.1% at400 nm

84

CdS/MoS2 ultrasonic adsorption 0.5 M Na2S−Na2SO3 300 W Xe-lamp (>400 nm)

1.75mmol g−1 h−1

- 95

CdS/MoS2 sonication 20 vol % ethanol 300 W Xe-lamp (>420 nm)

140mmol g−1 h−1

66% at420 nm

99

1T′ MoS2/2H MoS2 exfoliation/thermalannealing

0.01 M lactic acid 200 W Hg-lamp (>400 nm)

1.5mmol g−1 h−1

12.9% at400 nm

87

ZnIn2S4/MoSe2 electrostatic self-assembly 10 vol % lactic acid 300 W Xe-lamp (>400 nm)

64544mol g−1 h−1

- 131

g-C3N4/Graphene impregnation/thermalannealing

1.5 wt % Pt, 25 vol % methanol 350 W Xe-lamp (>400 nm)

451μmol g−1 h−1

2.6% at400 nm

132

g-C3N4/MoS2 impregnation/sulfidation 10 vol % lactic acid 300 W Xe-lamp (>420 nm)

20.6 μmol h−1 2.1% at420 nm

110

g-C3N4/ZnIn2S4 hydrothermal 20 vol % triethanolamine 300 W Xe-lamp (>420 nm)

2.78mmol g−1 h−1

7.05% at420 nm

49

g-C3N4/CaIn2S4 hydrothermal 1.0 wt % Pt, 0.5 M Na2S−Na2SO3 3 W UV-LEDs420 nm

102μmol g−1 h−1

- 91

g-C3N4/α-Fe2O3 thermal annealing 3 wt % Pt, 10 vol % triethanolamine 300 W Xe-lamp (>400 nm)

31.4mmol g−1 h−1

44.35% at420 nm

82

SiC/graphene in situ vapor−solidreaction

0.1 M Na2S 300 W Xe-lamp (>400 nm)

41.4μmol g−1 h−1

- 133

MoS2/Bi12O17Cl2 vigorous stirring/reflux 0.3 M ascorbic acid 300 W Xe-lamp (>420 nm)

33mmol g−1 h−1

36% at420 nm

134

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2259

3. ROLE OF INTERFACES IN 2D PHOTOCATALYSTSFOR WATER SPLITTING

Up to now, many 2D photocatalysts have been successfullyprepared and used for the photocatalytic hydrogen evolutionfrom water where the construction of 2D/2D interface plays animportant role on the photocatalytic performance of 2Dphotocatalysts. Table 1 and Table 2 summarize the recentprogress of 2D and 2D/2D photocatalysts for photocatalytichydrogen evolution, respectively. The role of interface in 2Dphotocatalysts for water splitting will be discussed in detail inthe following section.3.1. Graphene-Based Photocatalysts. 3.1.1. Overview of

Graphene-Based Photocatalysts for Photocatalytic WaterSplitting. Graphene, single-layer carbon with 2D hexagonalpacked lattice structure, has drawn great attention in scientificand engineering fields since its discovery in 2004.16 Because ofmany advantages of graphene, such as large specific surfacearea, excellent electrical conductivity, and high flexibility,graphene has been widely used as a cocatalyst for photo-catalysts.71,135 For example, graphene can be used as electronacceptor,97 photosensitizer,136 or electron transport bridge129 inphotocatalysts. Besides, the conduction-band potential of thesemiconductor can be shifted by combining with thegraphene.137 Moreover, the graphene with functional groupscan also be used as photocatalysts for water splitting.138

However, the H2 evolution rate and the stability of thegraphene as photocatalysts still need to be developed. Thecoupling of the photocatalysts with graphene is a promisingstrategy to promote the photocatalytic water splitting. Thecontact area and the interfacial interaction between grapheneand photocatalysts play an important role in enhancing theefficiency of photocatalytic H2 evolution. Therefore, theconstruction of the 2D/2D interface between the grapheneand photocatalysts has been widely studied.3.1.2. Synthesis of Graphene-Based Photocatalysts. Since

the discovery of graphene, many methods have been exploredfor the synthesis of graphene. Such as mechanical exfoliation,chemical exfoliation, chemical and thermal reduction, andchemical vapor deposition. Many excellent reviews havesummarized the synthesis of graphene systematically.139−141

For the graphene-based photocatalysts, graphene was usuallyused as a cocatalyst for the photocatalysts, in order to combinethe graphene with the 2D photocatalysts for the formation of2D/2D photocatalysts, some effective approaches have beendeveloped, such as thermal annealing,128 hydrothermalmethod,96 and electrostatic self-assembly.97

3.1.3. Role of Interface in Graphene-Based Photocatalysts.Graphene has been widely used as an electron acceptor andelectron storage center to promote the separation of electron−hole pairs for the photocatalysts becasuse of its high electronstorage capacity and excellent electron conductivity.135 Besides,the Fermi level of graphene is lower than the conduction bandof most photocatalysts, and thus, the electrons in the CB of thephotocatalysts can be easily transferred from the photocatalyststo the graphene, which enhances the separation of the electronsand holes.102,135 For example, in the 2D/2D g-C3N4/rGO, theelectrons can be effectively transferred from g-C3N4 to rGO,due to the lower Fermi level of rGO (−0.08 eV vs NHE)relative to the CB of g-C3N4 (−0.842 eV vs NHE), thusrestrained the charge recombination. Furthermore, theinterfacial interaction on the intimate 2D/2D interface betweenthe rGO and g-C3N4 can effectively promote the charge transfer

and separation. Consequently, the enriched electron density onthe rGO favors the photocatalytic hydrogen evolution fromwater. The H2 production rate (557 μmol g−1 h−1) from 2D/2D g-C3N4/rGO is much higher than that of g-C3N4 (158 μmolg−1 h−1).97

Graphene can act not only as the electron acceptor andelectron storage but also as the electron transport bridgebetween the photocatalyst and the cocatalyst.129,142−144 As atypical example, for the 2D ternary MoS2-graphene-ZnIn2S4,both the MoS2 and the ZnIn2S4 are connected to the graphene,and the ZnIn2S4, graphene, and MoS2 act as the light-harvestingsemiconductor, electron transport bridge, and hydrogenevolution reaction catalyst, respectively (Figure 7).129 In this

ternary photocatalyst, the CB of ZnIn2S4 is more negative thanthe graphene/graphene·− (G/G·−) redox potential; thus, theelectrons generated from the CB of ZnIn2S4 can be transferredeffectively to graphene due to the thermodynamic driving force.Moreover, the G/G·− redox potential is more negative than theCB of MoS2 nanosheets, which is beneficial for the electrontransfer from the electron-rich graphene·− to the CB of MoS2.In addition, the 2D/2D intimate interface also contributed tothe improvement of the interfacial charge transfer, andtherefore, the charge carrier recombination can be suppressed.With the effective electron transport bridge and the largenumber of active sites on the 2D MoS2, the H2 evolution ratefrom MoS2-graphene-ZnIn2S4 was 22.8 times higher than thatof pure ZnIn2S4 under visible light.129 Very recently, usinggraphene as the electron transport bridge has also beenexplored in the [ZnTMPyP]4+-MoS2/RGO-TEOA photo-catalytic system and the Au-nanoprism/reduced grapheneoxide/Pt-nanoframe system.142,143

Interestingly, the CB potential of a semiconductor can beshifted to be more negative position when coupled with thegraphene.137,145 For example, in the graphene/Bi2WO6 system,the chemical interaction between graphene and Bi2WO6 led toa shift of the Fermi level and the CB potential of Bi2WO6decreased from +0.09 to −0.3 V, thus the graphene/Bi2WO6composite had a more negative CB potential than the potentialof H+/H2. Moreover, the chemical bonding between grapheneand Bi2WO6 facilitated the electron transportation andsuppressed the recombination of photogenerated chargecarriers, thus further enhancing the H2 production rate.137 Asimilar result was reported by Gao et. al, in the graphene-

Figure 7. (a) Schematic illustration for photocatalytic hydrogengeneration over MoS2-graphene/ZnIn2S4; (b) schematic illustrationfor charge carriers in MoS2-graphene/ZnIn2S4. Reproduced withpermission from ref 129. Copyright 2016 Elsevier.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2260

Bi2WO6 composite, the CB potential was shifted to a lowerposition to enhance the degradation rate of rhodamine B.145

Besides, the graphene can also be used as the photosensitizer toextend the spectral responsive range of wide band gapsemiconductors to visible light, thus further increasing thephoton capture.71,136

Reduced graphene oxide nanosheet (rGO) can be changedto be a n-type semiconductor by doping with heteroatoms, suchas nitrogen.96 The n-type nitrogen-doped rGO can beincorporated with a p-type semiconductor to form the p−njunctions for enhancing the separation of the electron−holepairs. As a typical example, p-type MoS2 nanoplatelets werecoupled with the n-type nitrogen-doped rGO (n-rGO)nanosheets to form the nanoscale p−n junctions. In thisMoS2/n-rGO system, the rGO was able to suppress theaggregation of the MoS2 nanoplatelets effectively, and a largepercentage of edge sites could be created on the thin MoS2.Furthermore, the contact area was increased in the 2D/2DMoS2/n-rGO compared to the 0D/2D MoS2/n-rGO. In the2D/2D MoS2/n-rGO, the rGO increases the energy conversionefficiency as a passive charge extraction layer. Besides, with theformation of the nanoscale p−n junction, the space charge layercreated a built-in electric field and suppressed recombination ofthe electrons and holes. The p−n junction changed the role ofrGO from passive to active, further enhancing the chargeseparation. Consequently, the MoS2/n-rGO with the p−njunction exhibited much better photocatalytic performancethan the p-MoS2/undoped-rGO.

96

Ideal graphene is a carbon material with no obviouselectronic band gap; thus, it is unable to absorb the lightenergy and used directly as a photocatalyst.146 To developgraphene as an efficient photocatalyst, it is imperative toengineer the band gap of graphene with semiconductiveproperties. Doping with heteroatoms, such as B and N, is afeasible way to open band gap in the graphene.147,148 Especially,in-plane doping with h-BN domains in graphene can produceunique planar heterostructures and tunable electronic proper-ties.149 Very recently, atomic-level in-plane decoration of h-BNdomains in graphene (BN-G) was successfully fabricated bycontrolling the doping sequence of heteroatoms and the oxygencontent of the graphene precursor.128 The surface structure andcatalytic activity of graphene were greatly affected by the in-plane decoration of h-BN domains. The band gap of BN-G wascalculated to be 2.8 eV based on the band structure and DOSaround the Fermi level of BN-G. However, the band gaps ofgraphene (G), B doped-graphene (B-G), N doped-graphene(N-G), and B, N codoped-graphene (B, N-G) were quitenarrow (0−1.4 eV), whereas the band gap of bulk h-BN (4.8eV) was too large (Figure 8). As a result, the BN-G showed a10-fold increase in H2 evolution rate compared with those of N-G, B-G, and B,N-G.128

3.2. 2D Oxide Photocatalysts. 3.2.1. Overview of 2DOxide for Photocatalytic Water Splitting. Oxide semi-conductors are the most widely studied photocatalysts in thefield of the water splitting. For example, because of low cost,high chemical stability, and nontoxicity, TiO2 has beenextensively used as photocatalyst in the past decades. However,the application of oxide photocatalysts is limited by therecombination of electrons and holes in the long travel distanceto the surface of photocatalysts.150,151 Therefore, strategies areneeded to solve these problems. In the ultrathin 2Dphotocatalysts, the migration distance of electrons and holesis shortened to reduce the recombination. Besides, more

exposed surface atoms result in the generation of more danglingbonds on the surface atoms, which arouses more reaction activesites to direct contact with the reactant.152,153 However, theband gap of 2D oxides was widened due to the quantumconfinement effect.23 It is noteworthy that a new two-dimensional phase of TiO2 formed on the surface of rutileTiO2 (011) shows a reduced band gap from 3 eV (bulk TiO2)to 2.1 eV (new surface 2D phase), which endows the TiO2 toabsorb visible light.154 However, the new 2D phase of TiO2normally covers only 10−60% of the surface, photoexcitedelectrons in this surface of 2D TiO2 may diffuse into the bulkand can be scavenged at other surfaces not covered by the newTiO2 surface.2D oxides other than TiO2 have also been explored as

photocatalysts including ZnO,155 In2O3,156 Cu2O,

157 WO3,158

etc. Element doping in ultrathin 2D nanomaterials is moreefficient in the improvement of the photocatalytic activity thandoping in the bulk materials.14 This occurs because most of thedopants in ultrathin 2D nanosheets are located very close to thesurface and can directly participate in the surface photocatalyticreaction. For example, in the Rh-doped KCa2Nb3O10, almost allthe dopants are present within the bulk of the photocatalystand have an indirect effect on the surface reaction. In contrast,when the thickness of the Rh-doped Ca−Nb−O nanosheet(exfoliation from Rh-doped KCa2Nb3O10) is reduced to 1 nm,most of the dopants in the nanosheet present on the surface.Thus, most of the dopants can be expected to directlyparticipate in the catalytic reaction and cause a significantimprovement in photocatalytic activity. As a result, the H2production rate of the Rh-doped Ca−Nb−O nanosheet (384.8μmol·h−1) was 165 times as high as that of the parent Rh-dopedKCa2Nb3O10 (2.3 μmol·h−1), the quantum efficiency of Ca−Nb−O nanosheet was 65% at 300 nm.127

Despite the advantages of 2D oxides, the rapid recombina-tion of electrons and holes still exists in the 2D oxides, and thewidened band gap decreases the light absorption of the 2Doxides. In addition, the aggregation can occur in the process ofthe photocatalytic reaction. Construction of the 2D/2Dheterostructure between the 2D oxides and other cocatalyst isa viable way to relieve the aggregation and increase the opticalabsorption of 2D oxides.23 Furthermore, with the intimate 2D/

Figure 8. (A) Atomic structure model, (B) band structure, and (C)total and partial electronic density of states (TDOS and PDOS,respectively) of BN-G. The Fermi level is set at 0 eV. (D) Calculatedband gaps of G, B-G, N-G, B,N-G, BN-G, and h-BN. Reproduced withpermission from ref 128. Copyright 2016 American Chemical Society.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2261

2D interface, the transfer rate of charge carriers can be greatlyimproved and separated in the 2D/2D photocatalysts.101,130

3.2.2. Synthesis of 2D Oxide Photocatalysts. 2D oxidenanosheets can be prepared by a hydrothermal method, whichis simple and easy to operate. In addition, the crystal structureand the morphology of photocatalyst can be controlled byadjusting the types of the precursor, temperature, and reactiontime. For instance, 2D TiO2 nanosheet with exposed (001)facets was synthesized by the hydrothermal method, and theexposed (001) facets could be tailored by adding hydrofluoricacid.130,159,160 However, it should be noted that the hydro-fluoric acid is a contact poison and extremely corrosive, andthus, it must be handled with extreme care. In addition to TiO2,2D La2Ti2O7 nanosheets were also successfully synthesized bythe hydrothermal method.161

Exfoliation of a layered metal oxide composed of singlecrystal layers is also a feasible way to obtain 2D oxide.Generally, the layered precursors of 2D oxides are firstfabricated by a conventional solid-state calcination method athigh temperature and then is converted into the protonatedform by acid-exchange processing in an acid solution (e.g.,HCl). The protons are substituted with sufficient organic ion(e.g., tetrabutylammonium ion) to expand the interlayerspacing. Finally, the 2D oxide nanosheets can be obtaineddue to the decrease of the electrostatic interaction between thehost layers. For example, Ti1.82O4,

119 Ca2Nb3O10,162 and

TaO3163 nanosheets were fabricated successfully by exfoliation.

However, the acid-exchange processing and organic ionexchange processing take a long time with the entirepreparation process for around 2 weeks. In addition, only afew metal oxides have suitable layered precursor that can beexfoliated to 2D oxide. Moreover, the synthesis of 2D oxidenanosheets on a large scale by exfoliation is also a greatchallenge.Another effective way to prepare the 2D oxides is the self-

assembly wet chemistry method. The key of this method is theformation of lamellar structures by the molecular self-assemblybetween the oligomers and surfactant in solution, with hydratedinorganic oligomers confining between the bilayer reversemicelles.164 In this way, well-crystallized ultrathin TiO2nanosheets can be obtained after hydrothermal treatment ofthe solution and the removal of the organic templates.165 Theultrathin 2D TiO2 nanosheets possess a dominant anatasephase and a trace of rutile phase, high specific surface area and athickness around 3 nm. However, the band gap of the TiO2nanosheets was shifted to 3.65 eV, larger than that of the P25nanoparticles (3.20 eV), which could be attributed to theconfined atomic-scale thickness. Except for 2D TiO2 nanosheet,2D ZnO, Co3O4, WO3, Fe3O4, and MnO2 can also besynthesized by this wet chemistry method.164 The advantageof this approach is that the synthesis of 2D metal oxidenanosheets is not restricted by the limited numbers of layeredmaterials, and it is possible to synthesize homogeneousnanomaterials in large quantities. However, it is difficult toobtain single monolayer oxide by the wet chemistry method,because this approach still needs to reach thermodynamicequilibrium even if the growth along the thickness dimension isconfined by micelles.3.2.3. Role of Interface in 2D Oxide Photocatalysts.

Cocatalysts have been widely used to accelerate the rate ofhydrogen evolution over 2D oxide photocatalysts.70,166 Theinterface state between photocatalyst and cocatalyst is of greatimportance for the separation efficiency of the photoinduced

electron−hole pairs and the photocatalytic activity of thephotocatalyst. Therefore, construction of 2D nanojunctionsbetween 2D oxide photocatalysts and 2D cocatalysts is afeasible strategy to enhance the photocatalytic hydrogengeneration rate of 2D oxide photocatalysts. For example, the2D/2D MoS2/TiO2 photocatalysts show high photocatalyticH2 evolution activity, which is about 36.4 times higher than thatof pure TiO2 nanosheets.

130 Furthermore, the 2D/2D MoS2/TiO2 exhibits H2 evolution rate much higher than noble metalloaded TiO2 photocatalysts (such as Pt, Rh, Ru, Pd, and Au-loaded TiO2). These results can be attributed to intimate andlarger contact area between the TiO2 and MoS2 (2D/2Dcontact interface) than that of the noble metal loaded TiO2(point contact interface). Moreover, the effective chargetransfer from excited TiO2 to MoS2 due to the suitable bandalignment is also a key factor for the enhancement ofphotocatalytic activity of TiO2 (Figure 9). Similarly, the 2D/

2D TiO2/MoS2 was also prepared for the photocatalyticreduction of CO2 into CH3OH, and the fast electrontransferred from TiO2 to MoS2 could minimize chargerecombination and thus improved the photocatalytic activity.159

Because of the high specific surface area and superiorelectron conductivity of graphene, it is considered as anefficient cocatalyst for hydrogen production.167 When graphenewas used as the cocatalyst for TiO2 nanosheets, it could act asan electron acceptor to efficiently suppress the recombinationof electrons and holes.101 In the 2D/2D TiO2/graphenesystem, the 2D π-conjugation structure of graphene promotesthe transfer of photogenerated electrons. Additionally, thepotential (−0.08 V vs SHE, pH = 0) of graphene/graphene•− islower than the CB level (−0.24 V) of anatase TiO2, and thephotoexcited electrons in the conduction band of TiO2 can betransferred to graphene. Because the potential of graphene(−0.08 V) is higher than the reduction potential of H+/H2(0V), the protons can be efficiently reduced on the graphene(Figure 10).101 Moreover, the intimate interaction betweenTiO2 nanosheets and graphene sheets with the interplanejunction will further facilitate the separation of electron−holepairs. Therefore, the photocatalytic activity of the TiO2 can begreatly improved. With optimal graphene content of 1.0 wt %,the hydrogen production rate was 736 μmol h−1 g−1 with aquantum efficiency of 3.1%, 41 times higher than that of TiO2nanosheets. The 2D/2D TiO2/graphene system was alsoeffective for the photocatalytic degradation of organic pollutantand CO2 conversion.

168,169

Hybridizing 2D oxide photocatalysts with 2D organicsemiconductors (such as g-C3N4) to form the Type IIheterojunction greatly enhances the photocatalytic H2

Figure 9. Illustration and energy diagram of the charge transfer in the2D/2D MoS2/TiO2 nanojunction. Reproduced with permission fromref 130. Copyright 2016 American Chemical Society.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2262

production owing to the increasing interface area and interfacialcharge transfer rate. A Type II 2D/2D interface can be formedbetween the anatase TiO2 nanosheets (TiO2-A, 5−6 mono-layers) and g-C3N4 nanosheets (∼3 monolayers).23 With thesurface-to-surface heterojunction, the transfer rate of the chargecarriers, donor density, light adsorption, and the life of chargecarriers was enhanced. Besides, a compromised band gap couldbe achieved between the TiO2-A (3.3 eV) and g-C3N4 (2.8 eV),leading to a reduced band gap of TiO2-A to 2.91 eV, and aType II 2D/2D intimate interface was formed between theTiO2-A and g-C3N4, as shown in Figure 11. The quantum

efficiency of 2D/2D TiO2-A/g-C3N4 (5.3%) at 380 nm wasmuch higher than that of TiO2-A(1.28%) and g-C3N4(1.07%).

23 Type II band alignment can also be obtainedin the 2D/2D g-C3N4/NLTO composite photocatalyst(approximately 2 nm thick g-C3N4 and 7 nm thick N dopedLa2Ti2O7 nanosheets, NLTO).

84 The g-C3N4/NLTO exhibiteda high apparent quantum efficiency of 2.1% for photocatalyticH2 evolution at 400 nm, which was 4 times as high as that ofNLTO photocatalyst (0.5%).Another typical Type II heterojunction is the p−n junction

structure. When a 2D p-type oxide semiconductor is combinedwith a 2D n-type oxide semiconductor, the surface potential ofthe semiconductor will be changed due to the diffusion of freecarriers at the 2D/2D interface to balance the Fermi levels ofboth materials, which will affect the photocatalytic performanceof the photocatalyst. Taking 2D/2D n-type calcium niobate/p-type nickel oxide (CNO/NiO) as an example, the surfacepotential of the 2D CNO/NiO junction part is higher than that

of the NiO and lower than that of the CNO. Separated reactionsites can be obtained due to the potential gradient in the CNO/NiO photocatalyst. For example, the CNO/NiO junction partsare photo-oxidation sites, while the nonjunction parts and/ortheir edges are photoreduction sites.117 In addition, photo-catalytic H2 production rate can be affected by the exposedsurfaces of the 2D p−n junction photocatalysts. For instance,the photocatalytic H2 production rate for the CNO/NiOjunction (surface: CNO) was higher than that of NiO/CNOjunction (surface: NiO), which was attributed to the lowerphotocatalytic activity of the surface of NiO than that of CNOsurface.117

3.2.4. Stability of the 2D Oxide Photocatalysts. Some 2Doxides are stable under ambient conditions at room temper-ature, such as TiO2. However, some 2D oxides are unstable,such as Cu2O. The oxidation of Cu2O itself to CuO isthermodynamically more favorable than water oxidation, andthe Cu2O can also be reduced to metallic Cu in thephotochemical reaction.170 Besides, the stability of 2D oxidesis different due to their diverse structure. For example, single-layer MoO2 and WO2 are stable in the structure of honeycomb(H), and unstable in the structure of centered honeycomb (T).Single-layer TiO2, CoO2, and NbO2 are unstable in both the Hand T structures.171 In addition, the 2D oxides can still bestacked together through the van der Waals force to form thebulk counterparts.51 Formation of the 2D oxide/2D cocatalystheterostructure is an effective way to relieve the aggregation ofthe 2D oxides.23 In addition, the stability of oxides can beimproved by reducing the surface energies.172

3.3. 2D Chalcogenide Photocatalysts. 3.3.1. Overviewof 2D Chalcogenides for Photocatalytic Water Splitting. 2Dchalcogenide has recently emerged as a fascinating new class ofmaterial for photocatalysis. The reduction of the dimension ofchalcogenides from 3D to 2D result in novel electronic andmechanical properties.173,174 For 2D chalcogenides, most of thecatalytically active sites are exposed for catalytic reaction andthe ultralarge fraction of low-coordinated atoms is alsobeneficial to solar light harvesting.175,176 Besides, the bandgap of the chalcogenides can be adjusted according to thenumber of layers. For example, the band gap of MoS2 could beenlarged from 1.29 to 1.90 eV when bulk MoS2 exfoliates tomonolayer MoS2.

177 The relative position of the CB and VBcan be also tuned by altering the thickness of 2D chalcogenides,such as GaS.18 Furthermore, with the transformation from bulkto monolayer, the band gap of chalcogenides can be changedfrom indirect to direct band gap,177−179 or vice versa.180 Many2D chalcogenides have been experimentally demonstrated asphotocatalysts for water splitting, such as MoSe2,

181 MoS2,87

WS2,182 SnS2,

183 ReS2,184 etc. Besides, some 2D chalcogenides,

such as HfS2,185 TcSe2,

186,187 etc. were also predicted aspromising photocatalysts for water splitting. Interestingly, somemonochalcogenides show high photocatalytic water splitting inthe form of both 3D phase and 2D phase.95,120,188,189 Incontrast, the 2D layered transition-metal dichalcogenides(TMDs) such as 2D MoS2 shows high activity for photo-catalytic water splitting under visible light irradiation, while thebulk MoS2 has no photocatalytic activity due to the narrowband gap.87 Therefore, the construction of the 2D structure isof significance for the application of TMDs in the photo-catalytic water splitting.Despite the potential advantages of 2D chalcogenides for

photocatalysis, the hydrogen yield from water splitting is stilllow due to the ultrafast recombination of the photogenerated

Figure 10. Proposed mechanism for photocatalytic H2 productionover TiO2/graphene system.

Figure 11. Proposed band gap structure and photocatalyticmechanism for photogeneration of H2 over 2D/2D TiO2-A/g-C3N4photocatalysts.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2263

holes and electrons. Moreover, the photocorrosion driven byphotoexcited holes can occur on some 2D chalcogenides.Besides, 2D chalcogenides tends to restack during the course ofreaction and blocks the surface sites, making them inaccessiblefor the water splitting and leading to the reduction inactivity.190 To overcome these problems, strategies such asdoping with metal and nonmetal, combining with cocatalyst,the formation of 2D/2D junction have been employed tomodify their electronic property and improve their stability.Among them, the 2D/2D junction is the most promising wayto enhance their photocatalytic activity by reducing thephotocorrosion, regulating electron property and enhancingthe photochemical stability.3.3.2. Synthesis of 2D Chalcogenide Photocatalysts. Many

2D chalcogenide nanosheets have been successfully synthe-sized. For monochalcogenides, sonication120 and hydrothermalmethod are able to achieve the ultrathin 2D structures. Forexample, inorganic−organic hybrid CdS-amine nanosheetssynthesized via hydrothermal method were used as the startingmaterial to obtain the CdS nanosheets.120 The H2-productionrate was remarkably enhanced with the CdS nanosheet as thephotocatalyst in comparison to the nanoparticle counterpart.However, L-cysteine must be added to the solution as astabilizing agent to avoid the quick aggregation of the as-prepared sheets, and the stability of the CdS nanosheetdispersion was affected by the pH value.120,191

Quite different from monochalcogenides, TMDs consist oftwo close-packed chalcogenide planes sandwiching a transition-metal layer but have a similar layered structure to that ofgraphite and therefore can also exfoliate into monolayer bychemical and physical methods. Up to now, many strategieshave been developed to fabricate the monolayer TMDsincluding mechanical exfoliation,192 liquid-phase exfoliation,193

chemical intercalation exfoliation,87,194,195 lithium molten saltmethod,196 and chemical vapor deposition197 etc. Mechanicalexfoliation, the so-called “Scotch Tape Method”, the scotchtape-based mechanical exfoliation is the easiest and fastest wayto obtain the pristine, highly crystalline, atomically thin layers of2D TMDs.192 However, the adhesive tape procedure is clearlynot a scalable process. The liquid-phase exfoliation can be usedfor the large-scale production of monolayer chalcogenide,however, rigorous solvent inhibits its applicability. Furthermore,the liquid-phase exfoliation method is an aqueous solutionprocessing and therefore has significant economic and environ-mental disadvantages over the use of organic solvents or ionicliquids.For the chemical intercalation exfoliation, intercalation of

external ions or molecules is a critical step in the production of2D chalcogenides in the solution exfoliation of their layeredbulk counterparts. The neighboring layers can be weakened bythe interlayer expansion. Lithium intercalation exfoliation is aneffective route to obtain the monolayer chalcogenide with highcrystallinity. However, a long time (for example, 3 days at 100°C) is typically needed for the lithiation process.87 Besides,assisted sonication is usually required for solution exfoliation.Therefore, it is hard to obtain large sizes and high quality 2Dchalcogenides by solution-phase routes due to the high-energydriving input during the sonication. To produce high-qualitysingle-layer molybdenum disulfide sheets with unprecedentedlylarge flake size, potassium or sodium naphthalenide was used asthe intercalation agent, followed by sonicating in a low-powersonic bath (60 W) for 30 min, the flake size of obtained MoS2was up to 400 μm2.198 To make the exfoliation easier, recently,

high-quality monolayers with very large size 2D TMD weresuccessfully synthesized by a total single-crystal exfoliationmethod.194 In this method, homogeneous monolayers TMDwith submillimeter scale sizes and high crystallinity can beobtained by only simple manual shaking within several seconds.However, the complex procedure and high risk of using metalliclithium as a reagent are still a big disadvantage for chemicalintercalation exfoliation.Lithium molten salt method can be used to obtain the 2D

TMD without using the rigorous solvent and metallic lithium.In this method, lithium molten salt plays as the fluxing mediumand controls the TMD phases. For instance, (NH4)2MoS4 andLiOH were employed as the starting materials, and differentphases of MoS2 precursors can be obtained by controlling thecalcination temperature. The 2H-phase MoS2 precursors wereobtained at 400 °C while 1T-phase MoS2 precursors weregained at 1000 °C, and these precursors can exfoliate into 2H-and 1T-phase MoS2 monolayers.196 Supercritical carbondioxide (sc-CO2) has also been used to assist the exfoliationof layered materials, owing to its excellent performance, forexample, high diffusion coefficients, outstanding wetting ofsurfaces, and low interfacial tension.199

With the advantages of high crystallinity, simple operationand low cost, chemical vapor deposition (CVD) is regarded as apromising technique to grow various 2D TMD directly ondielectric substrates.200 Many 2D TMDs, such as MoS2,

201,202

WS2,203 MoSe2,

204 ZrS2,205 etc. have been successfully

synthesized. More importantly, CVD is a practical method toconstruct the 2D lateral and vertical TMD heterostruc-tures.104,206,207 Despite the advantages, the CVD method stillhas some drawbacks. Ultrathin 2D TMD obtained by the CVDmethod are always grow on the substrates, and it is needed tobe transferred to other substrates for further application. Forphotocatalytic water splitting reaction, freestanding 2D nano-sheets are more efficient compared to the 2D nanosheetsdeposited on the substrate. Therefore, the wet chemical route ismore appealing for making 2D TMD photocatalysts comparedto the CVD method.

3.3.3. Role of Interface in 2D Chalcogenide Photo-catalysts. Combining the 2D chalcogenide nanosheets withother 2D semiconductors is an effective strategy to enhance thephotocatalytic activity of the photocatalyst for water splitting.For example, when CdS nanosheets were coupled with theultrathin MoS2 nanosheets, the photogenerated electrons in theCB of CdS can rapidly transfer to MoS2 edge defect sites due tothe matched band alignments and intimate 2D/2D couplinginterfaces, as shown in Figure 12. Furthermore, the surface 2DMoS2 on 2D CdS as cocatalysts can also greatly accelerate theirH2-evolution kinetics.95 However, this 2D/2D CdS/MoS2photocatalyst is unstable, with almost 35% H2 production

Figure 12. Schematic diagram of photocatalytic hydrogen evolutionover 2D/2D CdS/MoS2.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2264

activity loss after four recycles, which might be attributed to theslow falloff of MoS2 nanosheets from CdS. Similarly, a few-layered MoS2/CdS van der Waals 2D/2D heterostructure wasfabricated by bubble exfoliation of MoS2 nanosheets into few-layered flakes and further in situ stacking with the ultrathinCdS. By using the few-layered MoS2/CdS photocatalyst, the H2evolution rate was 140 mmol g−1 h−1 with an apparent quantumyield of 66% at 420 nm.99 The nonlayered CdS epitaxiallygrown on the 2D layered MoS2 substrate to form a new quasivertical heterostructure with the clean interface was alsosuccessfully prepared.208 In the 2D/2D MoS2/CdS photo-catalyst system, 2D CdS act as a photosensitizer to absorb lightand subsequently generate electron−hole pairs, while MoS2 isthe cocatalyst offering active sites for H2 evolution with lowactivation potential.Different phases of the TMDs show different physical and

chemical properties and thus can be utilized to build intraplaneheterojunctions for improving the photocatalysis. For example,metastable metallic 1T and quasi-metallic 1T′ phases MoS2possess higher electric conductivity, while 2H phase MoS2 is asemiconductor with narrow direct band gap (∼2 eV). Althoughthe 2H phase MoS2 could be considered as a visible lightphotocatalyst, the electron transfer in the MoS2 plane is slow,which led to the rapid recombination of photogeneratedelectrons and holes. It is possible that if the 2H phase MoS2 andthe 1T/1T′ phases MoS2 can be formed in the same plane, therecombination of photogenerated electrons and holes will berestrained due to fast electron transfer from the 2H phase MoS2to 1T/1T′ phases MoS2. On the basis of this assumption,multiphasic single-layer MoS2 was constructed and acts as anefficient photocatalyst for hydrogen evolution from waterreduction.87 The 2H phase MoS2 was assumed to beresponsible for the absorption of visible light and chargecarrier generation, and photogenerated carriers were separatedthrough the heterojunctions between 1T′ MoS2 and 2H MoS2.The H2 evolution occurs over both basal and edge sites of 1T′MoS2 as well as edge sites of 2H MoS2, resulting in 4 timeshigher HER rate than that of pure 2H MoS2. The formation ofheterojunctions at the 2H and 1T′ phases MoS2 interfaces isillustrated on Figure 13.87 Du et al. also reported that electrons

were preferably transferred from the 2H to 1T′ phases MoS2but the 1T′ MoS2 structure showed comparable hydrogenevolution reaction activity to the 2H MoS2 structure underelectrochemical condition.209

In addition to the formation of intraplane heterostructure inthe same TMD material, the intraplane heterojunctionsconstructed by different TMD layers are also possible. Dueto the same honeycomb lattice structure and matched lattice

constant MX2(M = Mo, W; X = S, Se),210 the intraplaneheterojunctions can be constructed by seamless stitching ofdifferent MX2. For example, intraplane WS2/MoS2 hetero-structures with high crystallinity and subnanometer interfacescan be constructed by ambient pressure chemical vapordeposition. In these WS2/MoS2 heterostructures, MoS2 nano-sheets were laterally joint to the WS2 peripheral edges topredominantly exhibit triangular geometry and possessed thesingle-layer nature with a height of 0.8 nm. Due to the higherFermi level of WS2 than that of MoS2, when the intraplaneheterostructure was formed between WS2 and MoS2, theelectrons will diffuse from WS2 to MoS2, which makes thepotential of WS2 region higher than that of MoS2 region.Therefore, the WS2/MoS2 in-plane heterostructures can serveas quasi p−n junctions, where WS2 is quasi n-side and MoS2 isquasi p-side. The built-in potential or the built-in electrical fieldwithin this heterojunction structure will cause chargeredistribution and separate electrons and holes in two differentregions. Then photovoltaic responses and light emissionefficiency in WS2/MoS2 heterostructures would be domi-nated.211 However, the photocatalytic water splitting activity ofthese WS2/MoS2 heterostructures has not been studied yet.

3.3.4. Stability of 2D Chalcogenide Photocatalysts. Due tothe higher thermal dynamic stability of oxide over sulfide, some2D chalcogenides can be oxidized by photoexcited holes evenin the presence of sacrificial agent (such as lactic acid), whichwill reduce the photocatalytic activity of the photocatalyst, suchas CdS.212,213 2D chalcogenides are more stable in the Na2S−Na2SO3 solution, since the generated S

2− ions in Na2S−Na2SO3can protect the 2D chalcogenide from photocorrosion by theabundance of holes left behind.214 Besides, the restack of 2Dchalcogenides occurred during the reaction is also a bigproblem for the stability of photocatalysts. For example, theGaS sheets appear to restack during the course of reaction andblock the surface reaction active sites of the photocatalyst,leading to a decrease of the active sites for the photocatalytichydrogen evolution and consequently a decrease of the activityfrom 887 μmol h−1 g−1 (the first cycle) to 566 μmol h−1 g−1

(the third cycle). Notably, no photocorrosion was observed onGaS.18 Doping the 2D chalcogenide lattice with a donor atomis a good way to increase their stability.215 Chemicalmodification via functional groups is also identified as a feasibleway for stabilization of the 2D chalcogenide.216,217 For instance,the metastable 1T-MoS2 can be transformed into the stablephase after a crossover coverage of functionalization.217

However, the covering of the surface sites with these functionalgroups might decrease the number of active sites for HER andthus the applicability of this method to photocatalysis is yet tobe determined.

3.4. 2D Graphitic Carbon Nitride Photocatalysts.3.4.1. Overview of 2D Graphitic Carbon Nitride forPhotocatalytic Water Splitting. Graphitic carbon nitride (g-C3N4) has attracted dramatic interest in the field of visible-light-induced photocatalytic hydrogen production since it was firstused as photocatalyst for water splitting by Wang et al. in2009.173 The structure of g-C3N4 is two-dimensional frame-works consisting of tri-s-triazine connected via tertiary amines(Figure 14). g-C3N4 possesses a band gap of ∼2.7 eV,corresponding to a wavelength of ∼460 nm, which allows itto be a visible-light-response photocatalyst. Furthermore, thebottom of CB of the g-C3N4 is more negative than thereduction potential of H2O to H2, and the top of the VB ismore positive than the oxidation potential of the H2O to O2.

Figure 13. Schematic of the formation of heterojunctions at the 2Hand 1T′ phases MoS2 interfaces. Reproduced with permission from ref87. Copyright 2016 American Chemical Society.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2265

Therefore, g-C3N4 can be used as the photocatalyst for theoverall water splitting under visible light. Note that, the bandgap of the g-C3N4 will increase when the bulk g-C3N4 exfoliatesto monolayer due to the quantum confinement effect, as shownin Figure 14.19,27,218

Although g-C3N4 has the appropriate band gap and bandstructure for both water reduction and oxidation, thephotocatalytic activity of g-C3N4 is usually restricted by poorefficiency due to the rapid recombination of photogeneratedelectron−hole pairs. For bare g-C3N4, the H2 production ratewas only a few micromoles per hour per gram, and thethermodynamic driving force of g-C3N4 for O2 evolution wasmuch smaller than that for H2 evolution.173 In the g-C3N4,nitrogen atoms would be the preferred oxidation sites for H2Oto form O2, whereas the carbon atoms provided the reductionsites for H+ to H2.

173 Therefore, one approach to enhance thephotocatalytic activity is to maximize the exposed surface C andN atoms by fabricating the 2D g-C3N4. However, the problemof fast recombination of electron−hole pairs and excitoniceffects still inhibit the photocatalytic activity of the g-C3N4.

219

Therefore, many strategies are applied to improve thephotocatalytic activity of g-C3N4, such as structural modifica-tion,27 metal and nonmetal doping,220−222 copolymerization,223

and coupling with other semiconductors or cocatalysts.224

Among these strategies, combining 2D g-C3N4 with other 2Dmaterials is an effective way to enhance the photocatalyticperformance due to the 2D framework of g-C3N4. The 2D/2Dstructure with the largest and intimate contact interface in two2D materials promotes the electron transfer between them.Furthermore, electronic properties of g-C3N4 can be tuned andthus changes the exciton dissociation in the g-C3N4.

43 Hencethe photogenerated electrons and holes can be separated moreefficiently and then increase the photocatalytic activity of the2D g-C3N4.3.4.2. Synthesis of 2D Graphitic Carbon Nitride Photo-

catalysts. In general, g-C3N4 can be synthesized by the thermalpolymerization of nitrogen-rich precursors such as urea,225

thiourea,226 cyanamide,227 dicyandiamide,228 melamine,229

semicarbazide hydrochloride,230 and so forth. In order toenhance the photocatalytic activity, several synthetic methodsof g-C3N4 are modified to alter the optical and electronicproperties of g-C3N4. For example, the optical absorption of g-C3N4 semiconductor can be extended to visible light region upto 750 nm by copolymerization of dicyandiamide withbarbituric acid, and the band gap of the g-C3N4 decreaseswith the increasing of the barbituric acid content.231 Inaddition, the absorption edge of g-C3N4 can red shift towardto visible light by the introduction of nitrogen defects.Nitrogen-deficient g-C3N4 can be synthesized by thermalpolymerization at high temperature,232 hydrogen reduction,233

hydrothermal process,234 and the thermal polymerization withthe assistance of the alkali (such as KOH, NaOH, Ba(OH)2).

235

The nitrogen-deficient structures of g-C3N4 not only broadened

the visible-light absorption of g-C3N4 but also promoted theseparation of electron−hole pairs. Furthermore, the band gapof g-C3N4 is dependent on the number of nitrogen defects,which can be easily tuned by the alkali/precursor ratio.235

To obtain the 2D g-C3N4 nanosheets, several exfoliationmethods have been used to delaminate the bulk g-C3N4, such asthermal oxidation exfoliation,20 ultrasonic exfoliation,236 liquid-phase exfoliation,237 and so on. For the thermal oxidationexfoliation, g-C3N4 nanosheets can be synthesized bycalcinating bulk g-C3N4 in the air atmosphere at 500−550 °Cdue to the thermal oxidation etching effect.238 This is aneffective way to obtain the g-C3N4 nanosheets with high specificsurface area. However, the polymeric melon chain is not stableenough against the oxidation of O2 at high temperature, whichleads to the low yield of g-C3N4 nanosheets.19 Ultrasonicexfoliation has been widely used for liquid exfoliation of layeredmaterials, and it has been demonstrated to be a simple andscalable method to delaminate the bulk g-C3N4.

239 Theexfoliation efficiency of g-C3N4 using ultrasonic depends onthe ultrasonic power and solvent. Nonetheless, the 2D g-C3N4disintegrates into submicron-sized nanosheets due to the high-energy driving input, and the exfoliation efficiency was still nothigh enough. Liquid-phase exfoliation through redox reactionsand intercalation is a high-efficiency approach for theexfoliation of g-C3N4. For example, g-C3N4 nanosheets with athickness of 1.4−2.3 nm can be achieved by H2SO4 liquid-phaseexfoliation.240 Single-layer g-C3N4 was obtained by using theHNO3 (98%).

241 In addition, HCl242 and H3PO4243 can also be

applied to delaminate the g-C3N4.3.4.3. Role of Interfaces in 2D Graphitic Carbon Nitride for

Water Splitting. An effective strategy for improving thephotocatalytic activity of the g-C3N4 is coupling with a 2Dmaterial that possesses excellent electronic conductivity andreaction active sites, such as graphene. 2D/2D graphene/g-C3N4 composites were synthesized by a combined impregna-tion chemical reduction strategy.132 Because of the high specificsurface area (2600 m2·g−1) and excellent mobility of chargecarriers (200 000 cm2 V−1 s−1), the graphene part can act aselectronic conductive channels to efficiently separate thephotogenerated charge carriers, thus enhancing the visible-light photocatalytic H2-production activity of g-C3N4. With theoptimal graphene content of 1.0 wt %, the H2-production rateof the graphene/g-C3N4 is about 3 times as high as that of pureg-C3N4, and the apparent quantum efficiency reached 2.6%.To gain more insights into the interface between the

electronically active graphene and optically active g-C3N4,theoretical calculation based on the hybrid density functionaltheory was carried out by Du et al.244 The results show that thegraphene/g-C3N4 interface displays strong interlayer electroncoupling. A gap (70 meV) was opened for a g-C3N4 supportedgraphene layer, and the visible light response of g-C3N4 wasenhanced in the presence of graphene. Moreover, the chargedensity is redistributed by forming triangular-shaped electron-rich and hole-rich regions (the interlayer charge transfer fromgraphene to g-C3N4) within the graphene layer (Figure 15).With the open gap in graphene, electrons can be directlyexcited from the VB of graphene to the CB of g-C3N4. Thecharge transfer at the graphene/g-C3N4 interface was predictedto significantly enhance the electron conductivity in a g-C3N4layer, which was beneficial to the photocatalysis.RGO is also a good cocatalyst for g-C3N4 in photocatalysis.

Zhang et al. found that the band structure of g-C3N4 could bewell modulated by coupling with RGO.245 The mechanism for

Figure 14. Crystal structure and band gap of g-C3N4.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2266

the enhanced photocatalysis of the g-C3N4/RGO compositeswith various concentrations of O atom was investigated bydensity functional theory calculations.92 The interactionbetween g-C3N4 and RGO varies with the concentration ofO atom, and the band gap of g-C3N4/RGO is decreased anddependent on the concentration of O atom. It is worth notingthat the g-C3N4/RGO composites with appropriate concen-tration of O atom are good visible light harvesting semi-conductors. Most importantly, with a higher O concentration ing-C3N4/RGO, a type-II staggered band alignment can beobtained in the g-C3N4-RGO interface, leading to the highhydrogen-evolution activity at the negatively charged O atoms(active sites) in the RGO.92

Combining g-C3N4 with a non-noble-metal cocatalyst thatpossesses plenty of active sites for hydrogen evolution such asTMDs is attested to be an effective way for photocatalytic watersplitting. For example, taking advantage of the analogous 2Dlayered structures of the g-C3N4 and MoS2, g-C3N4/MoS2 withlayered nanojunction was synthesized by gas-phase sulfidationtechnique.110 The mesoporous structure of g-C3N4 facilitatesand stabilizes the high dispersion of MoS2 over the g-C3N4surface and the formation of the thin MoS2/g-C3N4 intimatejunction. The organic−inorganic hybrid with graphene-likethin-layered g-C3N4/MoS2 heterojunctions increase the acces-sible area between the planar interface of the MoS2 and the g-C3N4 layers. Furthermore, the barrier for electron transportfrom the g-C3N4 to the MoS2 was diminished, thus facilitatingfast electron transfer across the interface by the electrontunneling effect. In addition, the thin layers of the cocatalystcan lessen the light blocking effect and improve the lightutilization by g-C3N4. Moreover, the band alignment of g-C3N4and MoS2 in favor of the directional migration of photo-generated electrons from g-C3N4 to MoS2, while keepingadequate chemical potential to reduce the H+ to H2 at theactive sites of MoS2 (Figure 16). With the presence of thelayered g-C3N4/MoS2 interface, the hydrogen-evolution rate

over 0.5 wt % MoS2/g-C3N4 reaches 20.6 μmol h−1, even about

4 times higher than that of 0.5 wt % Pt/g-C3N4 (4.8 μmol h−1).

Integration of 2D g-C3N4 with different 2D semiconductorsto form the Type I system is another effective approach toincrease the utilization of g-C3N4 for visible light photocatalytichydrogen generation. For example, combining the 2D g-C3N4with 2D ternary metal sulfide (such as g-C3N4/ZnIn2S4, g-C3N4/CaIn2S4) has been demonstrated as a successful way toimprove the photocatalytic activity for water splitting byenhancing the light absorption and charge separation.49,91 Forthe g-C3N4/ZnIn2S4, the 2D/2D interface between 2D g-C3N4and 2D ZnIn2S4 provides larger contact areas and much fastercharge transfer rate than the 2D/0D structure of g-C3N4nanosheet@ZnIn2S4 microsphere, as shown in Figure 17.49

Moreover, abundant intimate and high-speed charge transfernanochannels are formed in the 2D/2D interfacial junctions,thus enhancing the separation and migration efficiency of thephotogenerated electrons and holes. The H2 evolution rate of2D/2D g-C3N4/ZnIn2S4 is 69.5, 15.4, 8.2, and 1.9 times higherthan that of pure g-C3N4 nanosheet, pure ZnIn2S4 microsphere,2D/0D g-C3N4 nanosheet@ZnIn2S4 microsphere and pureZnIn2S4 nanoleaf, respectively. In this Type I system, thephotogenerated electrons and holes were excited from the g-C3N4 and then transferred to the CB and VB of ZnIn2S4,respectively, and the photocatalytic H2 generation mainlyoccurred on the surface of ZnIn2S4. Similar results werereported on the 2D/2D CaIn2S4/g-C3N4.

91

Formation of the solid-state Z-scheme structure is apromising protocol to separate the photogenerated electron−hole pairs and enhance the photocatalytic H2 production. Veryrecently, the 2D/2D α-Fe2O3/g-C3N4 Z-scheme system wasfabricated and used for hydrogen evolution.82 This 2D/2D α-

Figure 15. Charge transfer at the graphene/g-C3N4 interface: (a) topand (b) side view of the three-dimensional charge density differenceplots. Green and blue balls represent C and N atoms, respectively.Yellow and light blue surfaces represent, respectively, chargeaccumulation and depletion in the space with respect to isolatedgraphene and g-C3N4. Reproduced with permission from ref 244.Copyright 2016 American Chemical Society.

Figure 16. Schematic illustration for the charge transfer in the MoS2/g-C3N4 heterostructures.

Figure 17. Schematic illustration of contact interfaces for (a) 2D/0Dheterojunction and (b) 2D/2D heterojunction. (c) The proposedmechanism of charge transfer for photocatalytic hydrogen evolutionover g-C3N4/ZnIn2S4. Reproduced with permission from ref 49.Copyright 2018 Elsevier.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2267

Fe2O3/g-C3N4 system exhibits a significantly enhancedquantum efficiency up to 44.35% at 420 nm. The tightinterface and band alignment between 2D α-Fe2O3 and 2D g-C3N4 lead to the formation of Z-scheme structure, and play apivotal role in enhancing the photocatalytic activity of α-Fe2O3/g-C3N4 toward H2 evolution. For this 2D/2D α-Fe2O3/g-C3N4Z-scheme system, the photogenerated electrons at the CB of α-Fe2O3 can directly move to recombine with photoinduced holesat the VB of 2D g-C3N4. Electrons generated at the CB of 2D g-C3N4 can promptly transfer to the surface to reduce the H+ intoH2, while holes generated in the valence band of 2D α-Fe2O3can be used in the oxidation of TEOA or oxygen evolutionreaction, thus significantly improves the photocatalytic watersplitting performance. Moreover, the 2D α-Fe2O3/g-C3N4shows potential for overall water splitting.The construction of order−disorder interfaces is an effective

strategy to accelerate exciton dissociation in the bulk g-C3N4.43

On the order−disorder interfaces of g-C3N4, the electrons wereinjected into the ordered chains and holes were blocked in thedisordered chains, resulting in the exciton dissociation and theseparation of electrons and holes. Besides, the recombination ofphotoinduced electron−hole pairs can also be suppressed byconstructing heterojunctions between two different g-C3N4 (1-g-C3N4/2-g-C3N4) synthesized from different precursors.246,247

However, the construction of order−disorder and 1-g-C3N4/2-g-C3N4 interfaces in the 2D g-C3N4 has not been reported.According to the merits of 2D materials, the separationefficiency of electrons and holes might be greatly improved ifthese two kinds of heterostructures can be constructed as theform 2D/2D interface in g-C3N4.3.4.4. Stability of 2D Graphitic Carbon Nitride Photo-

catalysts. g-C3N4 is identified as the most stable allotropecompared with various carbon nitrides under ambientconditions.173 With the structure of tri-s-triazine, it possesseshigh thermal stability and chemical stability. g-C3N4 is stable upto 600 °C in the air and insoluble in acid, base, and organicsolvents. g-C3N4 is stable under light irradiation in solution (pH= 0−14) due to the strong covalent bonds between carbon andnitride atoms.9 Therefore, g-C3N4 is a stable and promisingphotocatalyst in the reaction of the water splitting.3.5. Other 2D Photocatalysts and 2D/2D Composite

Photocatalysts. 3.5.1. 2D Photocatalysts for Overall WaterSplitting. Photocatalytic overall water splitting to produce H2and O2 is a promising technology to convert the solar energyinto chemical energy. To make overall water splitting feasible,the band gap of the semiconductor has to straddle thereduction potential of H+/H2 and the oxidation potential ofH2O/O2 as shown in Figure 2. Many semiconductors possesssuitable band gap for overall water splitting, such as TiO2,

248

C3N4,249 and Zn2GeO4.

250 However, owing to the fastrecombination of photogenerated electron−hole pairs and alack of surface redox active sites, it is difficult to achieve watersplitting with high-efficiency on a single bulk semiconductor.With more exposed active sites, 2D semiconductor should bemore efficient than their bulk counterparts for the photo-catalytic overall water splitting. Furthermore, some 2Dsemiconductors show photocatalytic overall water splittingactivity while their bulk counterparts were inactive. Forexample, the monolayer γ-Ga2O3 can photocatalyze overallwater splitting into hydrogen and oxygen while bulk γ-Ga2O3samples were inactive for oxygen evolution.251 However, thismonolayerγ-Ga2O3 is a wide-bandgap photocatalyst, which can

only absorb ultraviolet light. Thus, ultrathin 2D visible-responsephotocatalysts are worthy of exploitation.Modification of photocatalyst with redox cocatalysts is an

effective way to enhance the overall water splitting activity.252

For example, Mn3O4−GaN:ZnO-Rh/Cr2O3,252 SrTiO3−

NiO,253 and TiO2−Pt254 have been successfully prepared andused for photocatalytic overall water splitting. However,because of the fast recombination of the electrons and holesin the bulk semiconductor, the photocatalytic efficiency is stilllow. Recently, 2D semiconductor modified with cocatalyst foroverall water splitting was reported. For instance, 2D Pt/g-C3N4 synthesized by in situ deposition method is able tophotocatalytic pure water into H2 and O2 with an apparentquantum efficiency of 0.3% at 405 nm.255 In this system, bothPt0 and PtOx was formed on the g-C3N4, and the Pt0 waseffective for H2 evolution while PtOx was able to accelerate O2evolution. Notably, pure 2D g-C3N4 showed no activity, andthe Pt0 and PtOx are indispensable for the overall watersplitting. Pt and Co codeposited carbon nitride with acrystalline poly(triazine imide) framework was also reportedfor the overall water splitting, and the apparent quantum yieldreached 2.1% at 380 nm.256 The efficiency of overall watersplitting can also be improved by altering the reaction pathway.Compared to the typical four-electron pathway for watersplitting, a high-efficiency two-electron pathway was studied byKang and his co-worker.257 Carbon nanodot-carbon nitride(CDots-C3N4) nanocomposite was used as the photocatalyst inthis two-electron pathway. With the optimum CDots/C3N4ratio, the quantum efficiency reached 16% at 420 ± 20 nm, andshowed an overall solar energy conversion efficiency of 2.0%.Nevertheless, the solar-to-hydrogen efficiency still needs to beimproved for large-scale production.Construction of heterostructure in photocatalysts is another

efficient strategy to enhance the photocatalytic performance foroverall water splitting.258 Fabrication of phase junction,259 TypeII heterostructure,260 and Z-scheme system86 have beenreported for overall water splitting. Among them, Z-schemephotocatalytic system, with the H2 production photocatalystand the O2 production photocatalyst, is the most promisingapproach for overall water splitting. With a staggered alignmentof band structures in the Z-scheme system (Figure 4), therecombination of the electron−hole pairs can be greatlyreduced, thus allowing the reduction and oxidation of waterto take place on the H2 production photocatalyst and O2production photocatalyst, respectively. Many Z-scheme photo-catalytic system, such as Au-TiO2/SrTiO3−Rh,

261 BiVO4/CaFe2O4,

262 and SrTiO3:La,Rh/Au/BiVO4:Mo,86 have beendeveloped for overall water splitting. However, the overall watersplitting efficiency is still in need of improvement. Given theadvantages of the 2D/2D interface as described in the previoussections, construction of the 2D/2D Z-scheme system shouldbe an advisible approach for charge separation and theenhancement of overall water splitting. Theoretical calculationbased on density functional theory (DFT) has been studied on2D Z-scheme heterostructure to confirm its feasibility for theoverall water splitting. For example, MoSe2/graphene/HfS2 andMoSe2/N-doped graphene/HfS2 with graphene (N-dopedgraphene) as the redox mediator were predicted to be afeasible 2D Z-scheme photocatalyst for the overall watersplitting.263 Furthermore, a 2D/2D MoSe2/HfS2 and GeS/WS2without redox mediators is also verified by DFT calculations asa direct Z-scheme system for photocatalytic overall watersplitting.263,264 The internal electric field induced by the

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2268

electron transfer at the 2D/2D Z-scheme interface can suppressthe recombination of photogenerated electrons and holes; thus,the hydrogen evolution and oxygen evolution reaction can takeplace at the different part of the Z-scheme system.To date, there are few experimental reports on the 2D/2D Z-

scheme for overall water splitting, which might be due to thedifficulty in the synthesis of the 2D H2 (or O2) productionphotocatalyst and the construction of 2D/2D Z-schemeheterostructure. Very recently, 2D/2D Black Phosphorus/Bismuth Vanadate (BP/BiVO4) Z-scheme heterostructure wasfabricated and tested for overall water splitting.265 The BP andBiVO4 nanosheets were easily combined by electrostaticinteractions owing to the 2D structures, and the interfacialinteraction promoted the separation of the charge carriers. Inthis system, under visible light irradiation, the photogeneratedelectrons in the CB of BiVO4 rapidly combine with thephotogenerated holes in the VB of BP owing to their closeband positions. Consequently, the photogenerated electrons inthe CB of BP can be used for the reduction reaction, while thephotogenerated holes in the VB of BiVO4 be used for theoxidation reaction (Figure 18). The optimum H2 and O2

production rates on BP/BiVO4 were 160 and 102 μmol g−1

h−1, respectively, under irradiation of light (>420 nm) with anapparent quantum efficiency of 0.89% (420 nm). Apparently,more efficient construction technology for 2D and 2D/2Dphotocatalysts should be exploited to enhance the efficiency ofthe photocatalytic overall water splitting.3.5.2. Recent Emerging 2D Photocatalysts. Since the

discovery of 2D graphene, numerous new-types of 2D materialshave been synthesized. Except for the aforementionedmentioned ones, a few 2D materials were recently reportedand showed potential as photocatalysts such as GeH,266

graphene oxide,267 HNb3O8,126 HNbWO6,

33 SnNb2O6,32

ZnIn2S4,42 Cu2ZnSnS4,

268 SiC,133 CoOOH,124 etc. Theyshowed promising prospect to convert the solar energy intochemical energy by photocatalytic water splitting.Germanane (GeH), which was synthesized by an ion-

exchange approach, exhibits photocatalytic activity for watersplitting under visible light.83 GeH possesses a band gap of 1.58eV, and thus has a strong light absorption in the whole visibleregion. In addition, the suitable energy potential (CB = −0.313eV, VB = 1.267 eV vs NHE) endows it to reduce the H+ intoH2. Furthermore, after the cycling photocatalytic experiments,the H2 evolution rate remained constant, and the originalstructures and morphologies of GeH were retained. Theseresults indicate that the GeH samples are stable for photo-

catalytic applications. However, it shows relatively lowphotocatalytic activity due to the fast recombination rate ofthe photogenerated electrons and holes. To address thisproblem, a 2D/2D GeH/graphene heterostructure was studiedon the basis of hybrid density functional calculations.266 Whengraphene was intimately combined with the GeH monolayer, adriving force was generated due to the interaction betweencarbon and hydrogen atoms, which led to the electron transferfrom the graphene to GeH monolayer, while holes moved inthe opposite way. Thus, a built-in electric field is induced (Ebi).With the semimetal nature of graphene and the semiconductornature of GeH, the space charge region of the GeH/grapheneheterointerface leads to the formation of a Schottky barrier.The calculated Schottky barrier for holes to diffuse fromgraphene to GeH is only 0.83 eV (Figure. 19). Consequently,photoexcited holes and electrons can be separated effectively atthe GeH/graphene interfaces, which can lead to the improvedphotocatalytic performance.266

Ultrathin HNb3O8 nanosheets with a thickness of about 1.30nm show a widened band gap (3.68 eV) compared with its bulkcounterpart (3.59 eV) because of the quantum size effect.Benefiting from the 2D structure, the separation of thephotogenerated carriers was greatly enhanced. Thus, thephotocatalytic hydrogen evolution activity of the HNb3O8nanosheet was improved by 4 times higher than that of bulkHNb3O8.

126 Besides, ultrathin HNbWO6 nanosheets can beobtained by the exfoliation method based on an acid−basereaction and ion intercalation strategy.33 The as-preparedmonolayer HNbWO6 suspensions are stable and highlydispersed and efficient in photocatalytic hydrogen production.The apparent quantum yield of 1.0 wt % Pt loaded monolayerHNbWO6 was about 6 times higher than that of Pt loadedrestacked-HNbWO6.β-CoOOH nanosheets (1.5 nm thickness) exfoliated from

bulk β-CoOOH possess a band gap of 2.4 eV and caneffectively absorb visible light up to 500 nm.124 β-CoOOHnanosheets can photocatalytic hydrogen evolution from waterdirectly in a rate of 1200 μmol g−1 h−1 and the quantumefficiencies at 380, 420, and 450 nm were 10.7%, 6.9%, and

Figure 18. Photocatalytic overall water splitting over 2D/2D BP/BiVO4 Z-scheme system.

Figure 19. (a) Top and (b) side view of the three-dimensional chargedensity difference at the GeH/graphene heterointerface. Yellow andlight blue isosurfaces represent charge accumulation and depletion. (c)Band alignment at a GeH/graphene interface. Reproduced withpermission from ref 248. Copyright 2015 Royal Society of Chemistry.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2269

2.3%, respectively. In the process of photocatalytic hydrogenevolution over ultrathin β-CoOOH nanosheets, the low-coordinated surface Co ion strengthens the hybridization ofCo 3d states and S p states of the SO3

2− hole scavenger. Thus,the hole transfer rate between Co ions and SO3

2− and theelectron transfer between the surface catalytic sites and H+ ionscan be facilitated by the strong electron state hybridization,leading to the enhancement of photocatalytic hydrogenevolution activity.3.5.3. Theoretically Predicted 2D Photocatalysts. Except

for experimental efforts, theoretical modeling has also beendevoted to predicting effective 2D photocatalysts. Based on theextensive density functional theory calculations, many new 2Dphotocatalysts such as phosphorene,269 BiOI,270 C2N,

271 metalphosphorus trichalcogenides,272 zinc-blende,273 among others,with suitable band gap were predicted to be promising for thephotocatalytic water splitting.In 2014, monolayer black phosphorus was synthesized using

sticky-tape technique and was called “phosphorene”.269

Phosphorene is a direct and narrow band gap semiconductor,possesses an adjustable electronic band gap (0.3 to 2.29 eV),and exhibits visible to infrared photon absorption capabilities.Thus, phosphorene was predicted as an efficient photocatalystfor visible light water splitting.68,274 The phosphorene nano-sheets have a greater number of active sites for surface atoms,reduced electron−hole recombination rate and faster chargecarrier mobility than bulk BP. However, at ambient condition,the phosphorene can only trigger the hydrogen evolution fromwater, but is not suitable for the oxidation of water into O2since the VB of phosphorene is not more positive than theredox potential of O2/H2O. Notably, the phosphorene shows afavorable band position for overall water splitting in pH = 8.0solutions. The water splitting process on phosphorene isenergetically favorable, and the water oxidation and reductionprocess will take place separately on different sides ofmonolayer phosphorene.274

Single- and few-layer bismuth oxyhalides were predicted aspromising photocatalysts for solar water splitting.270,275 Thetheoretical calculation results show that the freestanding single-layer BiOI has a high dynamic stability and possesses a suitableband gap for the overall water splitting. The single-layer andfew-layer BiOI exhibits an indirect band gap, the band gap ofsingle-layer BiOI is predicted to be 2.28 eV, and the band gapdecreases very slightly with the increase in thickness. Mostimportantly, the band gap of few-layer BiOI is almostinsensitive to the layer thickness, which is due to the weakinterlayer interactions. From single-layer to a few layers (1−5layers), the BiOI can harvest the major portion of solar light(400−500 nm), demonstrating its potential as a visible lightphotocatalyst for water splitting. However, the CB positions arelocated just slightly above the reduction potential of water toH2, which might not have enough thermodynamic driving forcefor the hydrogen reduction reaction. However, this band edgepositions can be tuned by applying the compressive strain onsingle- and few-layer BiOI.270 The BiOCl and BiOBr were alsopredicted to be promising photocatalysts for water splitting.275

Notably, ultrathin BiOCl nanosheets have been experimentallydemonstrated as efficient photocatalysts for photocatalytic H2production from water under visible light irradiation.123

Monolayer C2N with a band gap of 2.47 eV is predicted to bea promising photocatalyst for water splitting based on densityfunctional calculations.271 Moreover, monolayer C2N has highstructural stability due to the high-frequency phonon modes,

which is close to that of graphene. In addition, C2−xSixN andC2−xGexN monolayers obtained by isoelectronic substitutionsat the C-site of C2N possess a reduced band gap and anenhanced absorption of visible light. Furthermore, the bandedge positions are even more favorable than the pristine C2Nfor the separation of electron−hole pairs and the photocatalyticwater splitting.271 On the basis of first-principles calculations,the C2N combined with the C3N4 was demonstrated to be apromising photocatalyst by forming the type II band alignmentwith relatively big chemical potential differences.276

Monolayer metal phosphorus trichalcogenides were demon-strated as promising photocatalysts by first-principles compu-tations.272,277 Take MnPSe3 as an example, with the band gapof 2.32 eV, it shows strong absorption in visible-light spectrum.Furthermore, MnPSe3 can be used to photocatalytic watersplitting into H2 and O2 simultaneously due to its suitable bandedge location. Besides, the exfoliation of bulk MnPSe3 is viablein experiments and it can form a freestanding monolayer. Thehigh carrier mobility of monolayer MnPSe3 (up to 625.9 cm2

V−1 s−1) is in favor of the transfer of charge carriers to reactivesites for the photocatalytic water splitting. With these features,monolayer MnPSe3 is a promising candidate as photocatalystfor water splitting. Additionally, FePS3, NiPS3, CdPS3, ZnPS3,FePSe3, and MnPS3 were also predicted to be potentialphotocatalysts for water splitting.272,277

MXenes, a new family of 2D transition-metal carbides andcarbonitrides, were first reported in 2011278 and have beenwidely studied for various applications.279 Recently, MXene waspredicted to be a promising photocatalyst for water splittingand CO2 reduction based on density functional theory.280,281

Particularly, 2D Zr2CO2 and Hf2CO2 MXene are potentialphotocatalysts for water splitting due to their suitable bandedges for the redox reaction of water splitting.281 In addition,they possess a narrow band gap (1.76 eV for Zr2CO2, and 1.79eV for Hf2CO2) and directionally anisotropic carrier mobility,thus they exhibit good optical absorption performance in visiblelight and the migration and separation of photogeneratedelectron−hole pairs can be effectively promoted. Meanwhile,these two MXenes also exhibit good stability in water.Furthermore, the adsorption of H2O and the formation of H2on the MXene are energetically favorable, which is in favor ofphotocatalytic hydrogen evolution.281 MXenes are also verifiedto be cocatalysts for photocatalysts to promote the separationof the charge carriers.65,282,283 Ti3C2Tx, Ti2CTx, and Nb2CTxwere demonstrated as efficient cocatalysts for TiO2 photo-catalyst. By formation of the Schottky barrier at the MXene/TiO2 interface, the separation of photogenerated charge carrierscan be promoted.90

4. CONCLUSIONS AND PERSPECTIVEIn summary, we have reviewed a few major types of 2Dmaterials-based photocatalysts for water splitting and criticallyassessed the role of interfaces in the 2D/2D photocatalysts.The 2D/2D interface plays an important role in thephotocatalytic water splitting activity of the photocatalystsdue to several reasons. First, the combination of 2Dsemiconductors with other 2D cocatalysts creates a largeintimate interface, which is beneficial for the separation ofelectron−hole pairs. Second, the formation of heterostructurejunctions with band alignment can be used to promote theseparation and transportation of electron−hole pairs betweenthe 2D semiconductors and 2D cocatalysts. Third, theutilization of sunlight is enhanced due to the broadened

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2270

absorption regime caused by the synergistic effect of the 2Dsemiconductors and 2D cocatalysts. Last but not least, theformation of intimate interface also enhances the stability of thephotocatalysts due to the alleviation of photocorrosion andagglomeration. Numerous examples with different types of 2D/2D heterostructures have shown that forming 2D/2D intimateinterface is a promising route to enhance the photocatalyticefficiency of the 2D semiconductors.Despite the encouraging progress made in the 2D photo-

catalysts, there are also enormous challenges in the photo-catalytic water splitting over these 2D materials.First, the synthetic method of 2D nanosheet from nonlayered

or layered precursor should be improved to control the numberof layers and obtain large-scale production of 2D photocatalyst.Despite many synthetic routes exist, such as hydrothermalmethod, ultrasonic exfoliation, ion intercalated exfoliation etc.,they are not feasible for large-scale production, especially forthe nonlayered semiconductors. Some layered semiconductorsare easier to delaminate; however, organic solvents or metalliclithium are usually needed for intercalation exfoliation, which isdangerous and has a harmful effect on the environment.Chemical vapor deposition is an efficient method to synthesize2D materials, yet a substrate is necessary for this process.Generally, freestanding 2D nanosheets are more efficient forthe photocatalytic reaction in an aqueous solution. Therefore,more effective synthetic routes need to be developed to satisfythe high efficiency, the large-scale production, and the stabilityof 2D photocatalysts.Second, many 2D photocatalysts are not stable in the water

splitting reaction due to the self-oxidation, photocorrosion oragglomeration. Therefore, strategies should be developed toenhance the stability of 2D photocatalysts, such as structuraldistortion, surface functionalization, and so on. Controlling thestructural distortion of the 2D photocatalysts to minimize thesurface energy is an effective method to improve their stability.The severe agglomeration of 2D photocatalysts can occurduring the water splitting reaction period, which can decreasethe number of surface active sites and limit their practicalapplication. Therefore, effective strategies should be exploitedto stabilize the 2D photocatalysts, such as the construction of2D/2D heterojunction by a layer-by-layer assembly orelectrostatic self-assembly techniques.Third, due to the fast recombination of electron−hole pairs,

sacrificial reagents (such as CH3OH, lactic acid, triethanol-amine, etc.) are usually needed to obtain the high photoactivity,which goes against with the practical application of 2Dphotocatalysts. Although the formation of 2D/2D interfaceby coupling two 2D photocatalysts can help to separate thephotogenerated charge carriers to a certain degree, moreefficient separation of electron−hole pairs is still highlydesirable. Similar to the interface engineering of bulkphotocatalysts, multiple interfaces among 2D, 1D, and 0Dshould be explored beyond the 2D/2D interfaces tosignificantly improve the charge carrier separation efficiency.Finally, the mechanism of photocatalytic enhancement by the

construction of the 2D/2D interfaces still remains to beexplored. Some explanations of the enhanced activity of the2D/2D photocatalysts have been proposed, but some are still incontroversy. Therefore, a fundamental understanding of thecharge transport process in the 2D/2D interface is in great needto help improve the fabrication of more efficient interfaces.First-principles calculations and advanced in situ techniquessuch as pump−probe spectroscopy on model 2D/2D interfaces

hold the promise to illuminate functions of each of these 2Dparts and the mechanism of charge transport between the 2D/2D interface.With the fast development of synthesis science and in situ/

operando characterizations, there is no doubt that the above-mentioned hurdles will be overcome and more highly efficientand robust 2D-based photocatalysts will emerge in the years tocome.

■ AUTHOR INFORMATION

Corresponding Authors*E-mail for Z.W.: wuz1@ornl.gov. Tel: +1(865)576-1080.*E-mail for Z.Q.: qinzuzeng@gxu.edu.cn.*E-mail for Z.G.: zguo10@utk.edu.

ORCIDZuzeng Qin: 0000-0001-5941-8792Zhanhu Guo: 0000-0003-0134-0210Zili Wu: 0000-0002-4468-3240NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was supported by the Center for NanophaseMaterials Sciences, which is a DOE Office of Science UserFacility. TMS thanks financial support from China ScholarshipCouncil. Notice: This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 withthe U.S. Department of Energy. The United States Govern-ment retains and the publisher, by accepting the article forpublication, acknowledges that the United States Governmentretains a nonexclusive, paid-up, irrevocable, worldwide licenseto publish or reproduce the published form of this manuscript,or allow others to do so, for United States Governmentpurposes. The Department of Energy will provide public accessto these results of federally sponsored research in accordancewith the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).

■ REFERENCES(1) Dincer, I. Renewable Sustainable Energy Rev. 2000, 4, 157−175.(2) Wang, J.; Feng, L.; Tang, X.; Bentley, Y.; Hook, M. Futures 2017,86, 58−72.(3) Su, T.-m.; Qin, Z.-z.; Ji, H.-b.; Jiang, Y.-x.; Huang, G. Environ.Chem. Lett. 2016, 14, 99−112.(4) Christoforidis, K. C.; Fornasiero, P. ChemCatChem 2017, 9,1523−1544.(5) Fujishima, A.; Honda, K. Nature 1972, 238, 37−38.(6) Zhou, X.; Liu, N.; Schmidt, J.; Kahnt, A.; Osvet, A.; Romeis, S.;Zolnhofer, E. M.; Marthala, V. R. R.; Guldi, D. M.; Peukert, W.;Hartmann, M.; Meyer, K.; Schmuki, P. Adv. Mater. 2017, 29, 1604747.(7) Gopannagari, M.; Kumar, D. P.; Reddy, D. A.; Hong, S.; Song, M.I.; Kim, T. K. J. Catal. 2017, 351, 153−160.(8) Ida, S.; Ishihara, T. J. Phys. Chem. Lett. 2014, 5, 2533−2542.(9) Cao, S.; Yu, J. J. Phys. Chem. Lett. 2014, 5, 2101−2107.(10) Acar, C.; Dincer, I.; Naterer, G. F. Int. J. Energy Res. 2016, 40,1449−1473.(11) Lee, Y. Y.; Jung, H. S.; Kang, Y. T. J. CO2 Util. 2017, 20, 163−177.(12) Hisatomi, T.; Domen, K. Curr. Opin. Electrochem. 2017, 2, 148−154.(13) Hisatomi, T.; Kubota, J.; Domen, K. Chem. Soc. Rev. 2014, 43,7520−7535.(14) Luo, B.; Liu, G.; Wang, L. Nanoscale 2016, 8, 6904−6920.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2271

(15) Tan, C.; Cao, X.; Wu, X.-J.; He, Q.; Yang, J.; Zhang, X.; Chen,J.; Zhao, W.; Han, S.; Nam, G.-H.; Sindoro, M.; Zhang, H. Chem. Rev.2017, 117, 6225−6331.(16) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang,Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Science 2004, 306,666−669.(17) Low, J.; Cao, S.; Yu, J.; Wageh, S. Chem. Commun. 2014, 50,10768−10777.(18) Kouser, S.; Thannikoth, A.; Gupta, U.; Waghmare, U. V.; Rao,C. N. R. Small 2015, 11, 4723−4730.(19) Dong, F.; Li, Y.; Wang, Z.; Ho, W.-K. Appl. Surf. Sci. 2015, 358,393−403.(20) Niu, P.; Zhang, L.; Liu, G.; Cheng, H.-M. Adv. Funct. Mater.2012, 22, 4763−4770.(21) She, X.; Liu, L.; Ji, H.; Mo, Z.; Li, Y.; Huang, L.; Du, D.; Xu, H.;Li, H. Appl. Catal., B 2016, 187, 144−153.(22) Sun, Y.; Gao, S.; Lei, F.; Xie, Y. Chem. Soc. Rev. 2015, 44, 623−636.(23) Gu, W.; Lu, F.; Wang, C.; Kuga, S.; Wu, L.; Huang, Y.; Wu, M.ACS Appl. Mater. Interfaces 2017, 9, 28674−28684.(24) Liu, Y.; Liang, L.; Xiao, C.; Hua, X.; Li, Z.; Pan, B.; Xie, Y. Adv.Energy Mater. 2016, 6, 1600437.(25) Liu, C.; Wang, L.; Tang, Y.; Luo, S.; Liu, Y.; Zhang, S.; Zeng, Y.;Xu, Y. Appl. Catal., B 2015, 164, 1−9.(26) Zhang, Z.; Huang, J.; Zhang, M.; Yuan, Q.; Dong, B. Appl.Catal., B 2015, 163, 298−305.(27) Han, Q.; Wang, B.; Gao, J.; Cheng, Z.; Zhao, Y.; Zhang, Z.; Qu,L. ACS Nano 2016, 10, 2745−2751.(28) She, X.; Wu, J.; Zhong, J.; Xu, H.; Yang, Y.; Vajtai, R.; Lou, J.;Liu, Y.; Du, D.; Li, H.; Ajayan, P. M. Nano Energy 2016, 27, 138−146.(29) Wang, Q.; Wang, W.; Zhong, L.; Liu, D.; Cao, X.; Cui, F. Appl.Catal., B 2018, 220, 290−302.(30) Mi, Y.; Wen, L.; Wang, Z.; Cao, D.; Xu, R.; Fang, Y.; Zhou, Y.;Lei, Y. Nano Energy 2016, 30, 109−117.(31) Huang, C.; Chen, C.; Zhang, M.; Lin, L.; Ye, X.; Lin, S.;Antonietti, M.; Wang, X. Nat. Commun. 2015, 6, 7698.(32) Zhou, C.; Zhao, Y.; Shang, L.; Shi, R.; Wu, L.-Z.; Tung, C.-H.;Zhang, T. Chem. Commun. 2016, 52, 8239−8242.(33) Liu, Y.; Xiong, J.; Luo, S.; Liang, R.; Qin, N.; Liang, S.; Wu, L.Chem. Commun. 2015, 51, 15125−15128.(34) Ran, J.; Wang, X.; Zhu, B.; Qiao, S.-Z. Chem. Commun. 2017, 53,9882−9885.(35) Singh, N.; Jabbour, G.; Schwingenschlogl, U. Eur. Phys. J. B2012, 85, 392.(36) Li, Q.; Li, X.; Wageh, S.; Al-Ghamdi, A. A.; Yu, J. Adv. EnergyMater. 2015, 5, 1500010.(37) Li, Y.; Li, Y.-L.; Sa, B.; Ahuja, R. Catal. Sci. Technol. 2017, 7,545−559.(38) Ran, J.; Ma, T. Y.; Gao, G.; Du, X.-W.; Qiao, S. Z. EnergyEnviron. Sci. 2015, 8, 3708−3717.(39) Tonda, S.; Kumar, S.; Kandula, S.; Shanker, V. J. Mater. Chem. A2014, 2, 6772−6780.(40) Ren, Y.; Wang, C.; Qi, Y.; Chen, Z.; Jia, Y.; Xu, Q. Appl. Surf. Sci.2017, 419, 573−579.(41) Ma, D.; Shi, J.-W.; Zou, Y.; Fan, Z.; Ji, X.; Niu, C. ACS Appl.Mater. Interfaces 2017, 9, 25377−25386.(42) Yuan, Y.-J.; Chen, D.; Zhong, J.; Yang, L.-X.; Wang, J.; Liu, M.-J.; Tu, W.-G.; Yu, Z.-T.; Zou, Z.-G. J. Mater. Chem. A 2017, 5, 15771−15779.(43) Wang, H.; Sun, X.; Li, D.; Zhang, X.; Chen, S.; Shao, W.; Tian,Y.; Xie, Y. J. Am. Chem. Soc. 2017, 139, 2468−2473.(44) Li, B.; Huang, L.; Zhong, M.; Li, Y.; Wang, Y.; Li, J.; Wei, Z.Adv. Electron. Mater. 2016, 2, 1600298.(45) Li, X.; Lin, M.-W.; Lin, J.; Huang, B.; Puretzky, A. A.; Ma, C.;Wang, K.; Zhou, W.; Pantelides, S. T.; Chi, M.; Kravchenko, I.;Fowlkes, J.; Rouleau, C. M.; Geohegan, D. B.; Xiao, K. Sci. Adv. 2016,2, e1501882.(46) Yuan, L.; Yang, M.-Q.; Xu, Y.-J. Nanoscale 2014, 6, 6335−6345.

(47) Iwashina, K.; Iwase, A.; Ng, Y. H.; Amal, R.; Kudo, A. J. Am.Chem. Soc. 2015, 137, 604−607.(48) Kobayashi, R.; Tanigawa, S.; Takashima, T.; Ohtani, B.; Irie, H.J. Phys. Chem. C 2014, 118, 22450−22456.(49) Lin, B.; Li, H.; An, H.; Hao, W.; Wei, J.; Dai, Y.; Ma, C.; Yang,G. Appl. Catal., B 2018, 220, 542−552.(50) Deng, D.; Novoselov, K.; Fu, Q.; Zheng, N.; Tian, Z.; Bao, X.Nat. Nanotechnol. 2016, 11, 218−230.(51) Tan, C.; Cao, X.; Wu, X.-J.; He, Q.; Yang, J.; Zhang, X.; Chen,J.; Zhao, W.; Han, S.; Nam, G.-H.; et al. Chem. Rev. 2017, 117, 6225−6331.(52) Xu, M.; Liang, T.; Shi, M.; Chen, H. Chem. Rev. 2013, 113,3766−3798.(53) Chhowalla, M.; Shin, H. S.; Eda, G.; Li, L.-J.; Loh, K. P.; Zhang,H. Nat. Chem. 2013, 5, 263−275.(54) Tan, C.; Zhang, H. Chem. Soc. Rev. 2015, 44, 2713−2731.(55) Butler, S. Z.; Hollen, S. M.; Cao, L.; Cui, Y.; Gupta, J. A.;Gutierrez, H. R.; Heinz, T. F.; Hong, S. S.; Huang, J.; Ismach, A. F.;Johnston-Halperin, E.; Kuno, M.; Plashnitsa, V. V.; Robinson, R. D.;Ruoff, R. S.; Salahuddin, S.; Shan, J.; Shi, L.; Spencer, M. G.; Terrones,M.; Windl, W.; Goldberger, J. E. ACS Nano 2013, 7, 2898−2926.(56) Duan, X.; Wang, C.; Pan, A.; Yu, R.; Duan, X. Chem. Soc. Rev.2015, 44, 8859−8876.(57) Zhang, H. ACS Nano 2015, 9, 9451−9469.(58) Fiori, G.; Bonaccorso, F.; Iannaccone, G.; Palacios, T.;Neumaier, D.; Seabaugh, A.; Banerjee, S. K.; Colombo, L. Nat.Nanotechnol. 2014, 9, 768−779.(59) Huang, X.; Tan, C.; Yin, Z.; Zhang, H. Adv. Mater. 2014, 26,2185−2204.(60) Naguib, M.; Mochalin, V. N.; Barsoum, M. W.; Gogotsi, Y. Adv.Mater. 2014, 26, 992−1005.(61) Sun, Y.; Gao, S.; Xie, Y. Chem. Soc. Rev. 2014, 43, 530−546.(62) Buscema, M.; Island, J. O.; Groenendijk, D. J.; Blanter, S. I.;Steele, G. A.; van der Zant, H. S. J.; Castellanos-Gomez, A. Chem. Soc.Rev. 2015, 44, 3691−3718.(63) Yazyev, O. V.; Chen, Y. P. Nat. Nanotechnol. 2014, 9, 755−767.(64) Rasool, H. I.; Ophus, C.; Zettl, A. Adv. Mater. 2015, 27, 5771−5777.(65) Gao, G.; O’Mullane, A. P.; Du, A. ACS Catal. 2017, 7, 494−500.(66) Singh, A. K.; Mathew, K.; Zhuang, H. L.; Hennig, R. G. J. Phys.Chem. Lett. 2015, 6, 1087−1098.(67) Ong, W.-J. Front. Mater. 2017, 4, 11.(68) Rahman, M. Z.; Kwong, C. W.; Davey, K.; Qiao, S. Z. EnergyEnviron. Sci. 2016, 9, 709−728.(69) Zhang, N.; Zhang, Y.; Xu, Y.-J. Nanoscale 2012, 4, 5792−5813.(70) Xiang, Q.; Yu, J.; Jaroniec, M. Chem. Soc. Rev. 2012, 41, 782−796.(71) Xiang, Q.; Cheng, B.; Yu, J. Angew. Chem., Int. Ed. 2015, 54,11350−11366.(72) Zhang, N.; Yang, M.-Q.; Liu, S.; Sun, Y.; Xu, Y.-J. Chem. Rev.2015, 115, 10307−10377.(73) Marschall, R. Adv. Funct. Mater. 2014, 24, 2421−2440.(74) Dong, X.; Cheng, F. J. Mater. Chem. A 2015, 3, 23642−23652.(75) Reza Gholipour, M.; Dinh, C.-T.; Beland, F.; Do, T.-O.Nanoscale 2015, 7, 8187−8208.(76) Chen, H.; Wang, L. Beilstein J. Nanotechnol. 2014, 5, 696−710.(77) Peng, W.; Li, Y.; Zhang, F.; Zhang, G.; Fan, X. Ind. Eng. Chem.Res. 2017, 56, 4611−4626.(78) Qu, Y.; Duan, X. Chem. Soc. Rev. 2013, 42, 2568−2580.(79) Hu, J.; Guo, Z.; McWilliams, P. E.; Darges, J. E.; Druffel, D. L.;Moran, A. M.; Warren, S. C. Nano Lett. 2016, 16, 74−79.(80) Chernikov, A.; Berkelbach, T. C.; Hill, H. M.; Rigosi, A.; Li, Y.;Aslan, O. B.; Reichman, D. R.; Hybertsen, M. S.; Heinz, T. F. Phys.Rev. Lett. 2014, 113, 076802.(81) Zhu, B.; Chen, X.; Cui, X. Sci. Rep. 2015, 5, 9218.(82) She, X.; Wu, J.; Xu, H.; Zhong, J.; Wang, Y.; Song, Y.; Nie, K.;Liu, Y.; Yang, Y.; Rodrigues, M.-T. F.; Vajtai, R.; Lou, J.; Du, D.; Li, H.;Ajayan, P. M. Adv. Energy Mater. 2017, 7, 1700025.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2272

(83) Liu, Z.; Lou, Z.; Li, Z.; Wang, G.; Wang, Z.; Liu, Y.; Huang, B.;Xia, S.; Qin, X.; Zhang, X.; Dai, Y. Chem. Commun. 2014, 50, 11046−11048.(84) Cai, X.; Zhang, J.; Fujitsuka, M.; Majima, T. Appl. Catal., B2017, 202, 191−198.(85) Maeda, K. ACS Catal. 2013, 3, 1486−1503.(86) Wang, Q.; Hisatomi, T.; Jia, Q.; Tokudome, H.; Zhong, M.;Wang, C.; Pan, Z.; Takata, T.; Nakabayashi, M.; Shibata, N.; et al. Nat.Mater. 2016, 15, 611−615.(87) Peng, R.; Liang, L.; Hood, Z. D.; Boulesbaa, A.; Puretzky, A.;Ievlev, A. V.; Come, J.; Ovchinnikova, O. S.; Wang, H.; Ma, C.; Chi,M.; Sumpter, B. G.; Wu, Z. ACS Catal. 2016, 6, 6723−6729.(88) Ran, J.; Zhang, J.; Yu, J.; Jaroniec, M.; Qiao, S. Z. Chem. Soc. Rev.2014, 43, 7787−7812.(89) Ran, J.; Gao, G.; Li, F.-T.; Ma, T.-Y.; Du, A.; Qiao, S.-Z. Nat.Commun. 2017, 8, 13907.(90) Wang, H.; Peng, R.; Hood, Z. D.; Naguib, M.; Adhikari, S. P.;Wu, Z. ChemSusChem 2016, 9, 1490−1497.(91) Jiang, D.; Li, J.; Xing, C.; Zhang, Z.; Meng, S.; Chen, M. ACSAppl. Mater. Interfaces 2015, 7, 19234−19242.(92) Xu, L.; Huang, W.-Q.; Wang, L.-L.; Tian, Z.-A.; Hu, W.; Ma, Y.;Wang, X.; Pan, A.; Huang, G.-F. Chem. Mater. 2015, 27, 1612−1621.(93) Moniz, S. J.; Shevlin, S. A.; Martin, D. J.; Guo, Z.-X.; Tang, J.Energy Environ. Sci. 2015, 8, 731−759.(94) Zhao, M.; Huang, F.; Lin, H.; Zhou, J.; Xu, J.; Wu, Q.; Wang, Y.Nanoscale 2016, 8, 16670−16676.(95) Ma, S.; Xie, J.; Wen, J.; He, K.; Li, X.; Liu, W.; Zhang, X. Appl.Surf. Sci. 2017, 391, 580−591.(96) Meng, F.; Li, J.; Cushing, S. K.; Zhi, M.; Wu, N. J. Am. Chem.Soc. 2013, 135, 10286−10289.(97) Pu, C.; Wan, J.; Liu, E.; Yin, Y.; Li, J.; Ma, Y.; Fan, J.; Hu, X.Appl. Surf. Sci. 2017, 399, 139−150.(98) Gong, Y.; Lin, J.; Wang, X.; Shi, G.; Lei, S.; Lin, Z.; Zou, X.; Ye,G.; Vajtai, R.; Yakobson, B. I.; Terrones, H.; Terrones, M.; Tay, B. K.;Lou, J.; Pantelides, S. T.; Liu, Z.; Zhou, W.; Ajayan, P. M. Nat. Mater.2014, 13, 1135.(99) Iqbal, S.; Pan, Z.; Zhou, K. Nanoscale 2017, 9, 6638−6642.(100) Sungpanich, J.; Thongtem, T.; Thongtem, S. J. Nanomater.2014, 2014, 8.(101) Xiang, Q.; Yu, J.; Jaroniec, M. Nanoscale 2011, 3, 3670−3678.(102) Ong, W.-J.; Tan, L.-L.; Chai, S.-P.; Yong, S.-T.; Mohamed, A.R. Nano Energy 2015, 13, 757−770.(103) Tongay, S.; Fan, W.; Kang, J.; Park, J.; Koldemir, U.; Suh, J.;Narang, D. S.; Liu, K.; Ji, J.; Li, J.; Sinclair, R.; Wu, J. Nano Lett. 2014,14, 3185−3190.(104) Duan, X.; Wang, C.; Shaw, J. C.; Cheng, R.; Chen, Y.; Li, H.;Wu, X.; Tang, Y.; Zhang, Q.; Pan, A.; Jiang, J.; Yu, R.; Huang, Y.;Duan, X. Nat. Nanotechnol. 2014, 9, 1024−1030.(105) Furchi, M. M.; Pospischil, A.; Libisch, F.; Burgdorfer, J.;Mueller, T. Nano Lett. 2014, 14, 4785−4791.(106) Ao, Y.; Wang, K.; Wang, P.; Wang, C.; Hou, J. Dalton Trans.2016, 45, 7986−7997.(107) Li, J.; Liu, E.; Ma, Y.; Hu, X.; Wan, J.; Sun, L.; Fan, J. Appl. Surf.Sci. 2016, 364, 694−702.(108) Kappera, R.; Voiry, D.; Yalcin, S. E.; Branch, B.; Gupta, G.;Mohite, A. D.; Chhowalla, M. Nat. Mater. 2014, 13, 1128.(109) Eda, G.; Yamaguchi, H.; Voiry, D.; Fujita, T.; Chen, M.;Chhowalla, M. Nano Lett. 2011, 11, 5111−5116.(110) Hou, Y.; Laursen, A. B.; Zhang, J.; Zhang, G.; Zhu, Y.; Wang,X.; Dahl, S.; Chorkendorff, I. Angew. Chem., Int. Ed. 2013, 52, 3621−3625.(111) Sun, Z.; Liu, Q.; Yao, T.; Yan, W.; Wei, S. Science ChinaMaterials 2015, 58, 313−341.(112) Koningsberger, D. C.; Prins, R. X-ray Absorption: Principles,Applications, Techniques of EXAFS, SEXAFS, and XANES; John Wileyand Sons: New York, 1988.(113) Ran, J.; Zhu, B.; Qiao, S.-Z. Angew. Chem., Int. Ed. 2017, 56,10373−10377.

(114) Zhao, D.; Chen, J.; Dong, C.-L.; Zhou, W.; Huang, Y.-C.; Mao,S. S.; Guo, L.; Shen, S. J. Catal. 2017, 352, 491−497.(115) Ahmad, M.; Agarwal, K.; Kumari, N.; Mehta, B. R. Appl. Phys.Lett. 2017, 111, 023904.(116) Sharma, I.; Mehta, B. R. Appl. Phys. Lett. 2017, 110, 061602.(117) Ida, S.; Takashiba, A.; Koga, S.; Hagiwara, H.; Ishihara, T. J.Am. Chem. Soc. 2014, 136, 1872−1878.(118) Yan, B.; Zhou, P.; Xu, Q.; Zhou, X.; Xu, D.; Zhu, J. RSC Adv.2016, 6, 6133−6137.(119) Ida, S.; Kim, N.; Ertekin, E.; Takenaka, S.; Ishihara, T. J. Am.Chem. Soc. 2015, 137, 239−244.(120) Xu, Y.; Zhao, W.; Xu, R.; Shi, Y.; Zhang, B. Chem. Commun.2013, 49, 9803−9805.(121) Yang, S.; Gong, Y.; Zhang, J.; Zhan, L.; Ma, L.; Fang, Z.; Vajtai,R.; Wang, X.; Ajayan, P. M. Adv. Mater. 2013, 25, 2452−2456.(122) Ou, H.; Lin, L.; Zheng, Y.; Yang, P.; Fang, Y.; Wang, X. Adv.Mater. 2017, 29, 1700008.(123) Ye, L.; Jin, X.; Leng, Y.; Su, Y.; Xie, H.; Liu, C. J. Power Sources2015, 293, 409−415.(124) He, S.; Huang, Y.; Huang, J.; Liu, W.; Yao, T.; Jiang, S.; Tang,F.; Liu, J.; Hu, F.; Pan, Z.; Liu, Q. J. Phys. Chem. C 2015, 119, 26362−26366.(125) Yang, W.; Zhang, L.; Xie, J.; Zhang, X.; Liu, Q.; Yao, T.; Wei,S.; Zhang, Q.; Xie, Y. Angew. Chem. 2016, 128, 6828−6832.(126) Xiong, J.; Wen, L.; Jiang, F.; Liu, Y.; Liang, S.; Wu, L. J. Mater.Chem. A 2015, 3, 20627−20632.(127) Okamoto, Y.; Ida, S.; Hyodo, J.; Hagiwara, H.; Ishihara, T. J.Am. Chem. Soc. 2011, 133, 18034−18037.(128) Li, M.; Wang, Y.; Tang, P.; Xie, N.; Zhao, Y.; Liu, X.; Hu, G.;Xie, J.; Zhao, Y.; Tang, J.; Zhang, T.; Ma, D. Chem. Mater. 2017, 29,2769−2776.(129) Yuan, Y.-J.; Tu, J.-R.; Ye, Z.-J.; Chen, D.-Q.; Hu, B.; Huang, Y.-W.; Chen, T.-T.; Cao, D.-P.; Yu, Z.-T.; Zou, Z.-G. Appl. Catal., B 2016,188, 13−22.(130) Yuan, Y.-J.; Ye, Z.-J.; Lu, H.-W.; Hu, B.; Li, Y.-H.; Chen, D.-Q.;Zhong, J.-S.; Yu, Z.-T.; Zou, Z.-G. ACS Catal. 2016, 6, 532−541.(131) Yang, M.-Q.; Xu, Y.-J.; Lu, W.; Zeng, K.; Zhu, H.; Xu, Q.-H.;Ho, G. W. Nat. Commun. 2017, 8, 14224.(132) Xiang, Q.; Yu, J.; Jaroniec, M. J. Phys. Chem. C 2011, 115,7355−7363.(133) Zhou, X.; Gao, Q.; Li, X.; Liu, Y.; Zhang, S.; Fang, Y.; Li, J. J.Mater. Chem. A 2015, 3, 10999−11005.(134) Li, J.; Zhan, G.; Yu, Y.; Zhang, L. Nat. Commun. 2016, 7,11480.(135) Low, J.; Yu, J.; Ho, W. J. Phys. Chem. Lett. 2015, 6, 4244−4251.(136) Min, S.; Hou, J.; Lei, Y.; Ma, X.; Lu, G. Appl. Surf. Sci. 2017,396, 1375−1382.(137) Sun, Z.; Guo, J.; Zhu, S.; Mao, L.; Ma, J.; Zhang, D. Nanoscale2014, 6, 2186−2193.(138) Putri, L. K.; Tan, L.-L.; Ong, W.-J.; Chang, W. S.; Chai, S.-P.Applied Materials Today 2016, 4, 9−16.(139) Bhuyan, M. S. A.; Uddin, M. N.; Islam, M. M.; Bipasha, F. A.;Hossain, S. S. Int. Nano Lett. 2016, 6, 65−83.(140) Lee, H. C.; Liu, W.-W.; Chai, S.-P.; Mohamed, A. R.; Lai, C.W.; Khe, C.-S.; Voon, C. H.; Hashim, U.; Hidayah, N. M. S. ProcediaChem. 2016, 19, 916−921.(141) Lee, H. C.; Liu, W.-W.; Chai, S.-P.; Mohamed, A. R.; Aziz, A.;Khe, C.-S.; Hidayah, N. M. S.; Hashim, U. RSC Adv. 2017, 7, 15644−15693.(142) Lou, Z.; Fujitsuka, M.; Majima, T. J. Phys. Chem. Lett. 2017, 8,844−849.(143) Yuan, Y.-J.; Chen, D.; Zhong, J.; Yang, L.-X.; Wang, J.-J.; Yu,Z.-T.; Zou, Z.-G. J. Phys. Chem. C 2017, 121, 24452−24462.(144) Xiang, Q.; Cheng, F.; Lang, D. ChemSusChem 2016, 9, 996−1002.(145) Gao, E.; Wang, W.; Shang, M.; Xu, J. Phys. Chem. Chem. Phys.2011, 13, 2887−2893.(146) Castro Neto, A. H.; Guinea, F.; Peres, N. M. R.; Novoselov, K.S.; Geim, A. K. Rev. Mod. Phys. 2009, 81, 109−162.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2273

(147) Putri, L. K.; Ng, B.-J.; Ong, W.-J.; Lee, H. W.; Chang, W. S.;Chai, S.-P. ACS Appl. Mater. Interfaces 2017, 9, 4558−4569.(148) Chen, L.-C.; Yeh, T.-F.; Lee, Y.-L.; Teng, H. Appl. Catal., A2016, 521, 118−124.(149) Zhang, J.; Xie, W.; Xu, X.; Zhang, S.; Zhao, J. Chem. Mater.2016, 28, 5022−5028.(150) Ni, M.; Leung, M. K. H.; Leung, D. Y. C.; Sumathy, K.Renewable Sustainable Energy Rev. 2007, 11, 401−425.(151) Linsebigler, A. L.; Lu, G.; Yates, J. T. Chem. Rev. 1995, 95,735−758.(152) Guan, M.; Xiao, C.; Zhang, J.; Fan, S.; An, R.; Cheng, Q.; Xie,J.; Zhou, M.; Ye, B.; Xie, Y. J. Am. Chem. Soc. 2013, 135, 10411−10417.(153) Lu, G.; Linsebigler, A.; Yates, J. T. J. Phys. Chem. 1994, 98,11733−11738.(154) Tao, J.; Luttrell, T.; Batzill, M. Nat. Chem. 2011, 3, 296−300.(155) Wang, J.; Xia, Y.; Dong, Y.; Chen, R.; Xiang, L.; Komarneni, S.Appl. Catal., B 2016, 192, 8−16.(156) Lei, F.; Sun, Y.; Liu, K.; Gao, S.; Liang, L.; Pan, B.; Xie, Y. J.Am. Chem. Soc. 2014, 136, 6826−6829.(157) Gao, S.; Sun, Y.; Lei, F.; Liu, J.; Liang, L.; Li, T.; Pan, B.; Zhou,J.; Xie, Y. Nano Energy 2014, 8, 205−213.(158) Wang, J.; Liu, C.-j. ChemBioEng Rev. 2015, 2, 335−350.(159) Tu, W.; Li, Y.; Kuai, L.; Zhou, Y.; Xu, Q.; Li, H.; Wang, X.;Xiao, M.; Zou, Z. Nanoscale 2017, 9, 9065−9070.(160) Yang, H. G.; Liu, G.; Qiao, S. Z.; Sun, C. H.; Jin, Y. G.; Smith,S. C.; Zou, J.; Cheng, H. M.; Lu, G. Q. J. Am. Chem. Soc. 2009, 131,4078−4083.(161) Meng, F.; Hong, Z.; Arndt, J.; Li, M.; Zhi, M.; Yang, F.; Wu, N.Nano Res. 2012, 5, 213−221.(162) Ebina, Y.; Sasaki, T.; Watanabe, M. Solid State Ionics 2002, 151,177−182.(163) Fukuda, K.; Nakai, I.; Ebina, Y.; Ma, R.; Sasaki, T. Inorg. Chem.2007, 46, 4787−4789.(164) Sun, Z.; Liao, T.; Dou, Y.; Hwang, S. M.; Park, M.-S.; Jiang, L.;Kim, J. H.; Dou, S. X. Nat. Commun. 2014, 5, 3813.(165) Sheng, L.; Liao, T.; Kou, L.; Sun, Z. Materials Today Energy2017, 3, 32−39.(166) Wen, J.; Xie, J.; Chen, X.; Li, X. Appl. Surf. Sci. 2017, 391, 72−123.(167) Xiang, Q.; Yu, J. J. Phys. Chem. Lett. 2013, 4, 753−759.(168) Sun, J.; Zhang, H.; Guo, L.-H.; Zhao, L. ACS Appl. Mater.Interfaces 2013, 5, 13035−13041.(169) Tu, W.; Zhou, Y.; Liu, Q.; Tian, Z.; Gao, J.; Chen, X.; Zhang,H.; Liu, J.; Zou, Z. Adv. Funct. Mater. 2012, 22, 1215−1221.(170) Zhang, L.; Jing, D.; Guo, L.; Yao, X. ACS Sustainable Chem.Eng. 2014, 2, 1446−1452.(171) Ataca, C.; Sahin, H.; Ciraci, S. J. Phys. Chem. C 2012, 116,8983−8999.(172) Zheng, Z.; Huang, B.; Wang, Z.; Guo, M.; Qin, X.; Zhang, X.;Wang, P.; Dai, Y. J. Phys. Chem. C 2009, 113, 14448−14453.(173) Wang, X.; Maeda, K.; Thomas, A.; Takanabe, K.; Xin, G.;Carlsson, J. M.; Domen, K.; Antonietti, M. Nat. Mater. 2009, 8, 76−80.(174) Yoon, Y.; Ganapathi, K.; Salahuddin, S. Nano Lett. 2011, 11,3768−3773.(175) Sun, Y.; Sun, Z.; Gao, S.; Cheng, H.; Liu, Q.; Lei, F.; Wei, S.;Xie, Y. Adv. Energy Mater. 2014, 4, 1300611.(176) Xie, J.; Zhang, J.; Li, S.; Grote, F.; Zhang, X.; Zhang, H.; Wang,R.; Lei, Y.; Pan, B.; Xie, Y. J. Am. Chem. Soc. 2013, 135, 17881−17888.(177) Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Phys. Rev.Lett. 2010, 105, 136805.(178) Wang, H.; Yuan, H.; Sae Hong, S.; Li, Y.; Cui, Y. Chem. Soc.Rev. 2015, 44, 2664−2680.(179) Zhang, Y.; Chang, T.-R.; Zhou, B.; Cui, Y.-T.; Yan, H.; Liu, Z.;Schmitt, F.; Lee, J.; Moore, R.; Chen, Y.; Lin, H.; Jeng, H.-T.; Mo, S.-K.; Hussain, Z.; Bansil, A.; Shen, Z.-X. Nat. Nanotechnol. 2014, 9, 111−115.

(180) Li, X.; Lin, M.-W.; Puretzky, A. A.; Idrobo, J. C.; Ma, C.; Chi,M.; Yoon, M.; Rouleau, C. M.; Kravchenko, I. I.; Geohegan, D. B.;Xiao, K. Sci. Rep. 2015, 4, 5497.(181) Gupta, U.; Naidu, B.; Maitra, U.; Singh, A.; Shirodkar, S. N.;Waghmare, U. V.; Rao, C. APL Mater. 2014, 2, 092802.(182) Mahler, B.; Hoepfner, V.; Liao, K.; Ozin, G. A. J. Am. Chem.Soc. 2014, 136, 14121−14127.(183) Sun, Y.; Cheng, H.; Gao, S.; Sun, Z.; Liu, Q.; Liu, Q.; Lei, F.;Yao, T.; He, J.; Wei, S.; Xie, Y. Angew. Chem., Int. Ed. 2012, 51, 8727−8731.(184) Rahman, M.; Davey, K.; Qiao, S.-Z. Adv. Funct. Mater. 2017,27, 1606129.(185) Singh, D.; Gupta, S. K.; Sonvane, Y.; Kumar, A.; Ahuja, R.Catal. Sci. Technol. 2016, 6, 6605−6614.(186) Jiao, Y.; Zhou, L.; Ma, F.; Gao, G.; Kou, L.; Bell, J.; Sanvito, S.;Du, A. ACS Appl. Mater. Interfaces 2016, 8, 5385−5392.(187) Xiong, R.; Yang, H.; Peng, Q.; Sa, B.; Wen, C.; Wu, B.; Sun, Z.Res. Chem. Intermed. 2017, 43, 5271−5282.(188) Bao, N.; Shen, L.; Takata, T.; Domen, K. Chem. Mater. 2008,20, 110−117.(189) Jang, J. S.; Joshi, U. A.; Lee, J. S. J. Phys. Chem. C 2007, 111,13280−13287.(190) Kouser, S.; Thannikoth, A.; Gupta, U.; Waghmare, U. V.; Rao,C. Small 2015, 11, 4723−4730.(191) Xu, Q.; Zeng, J.; Wang, H.; Li, X.; Xu, J.; Wu, J.; Xiao, G.; Xiao,F.-X.; Liu, X. Nanoscale 2016, 8, 19161−19173.(192) Li, H.; Lu, G.; Wang, Y.; Yin, Z.; Cong, C.; He, Q.; Wang, L.;Ding, F.; Yu, T.; Zhang, H. Small 2013, 9, 1974−1981.(193) Jawaid, A.; Nepal, D.; Park, K.; Jespersen, M.; Qualley, A.;Mirau, P.; Drummy, L. F.; Vaia, R. A. Chem. Mater. 2016, 28, 337−348.(194) Peng, J.; Wu, J.; Li, X.; Zhou, Y.; Yu, Z.; Guo, Y.; Wu, J.; Lin,Y.; Li, Z.; Wu, X.; Wu, C.; Xie, Y. J. Am. Chem. Soc. 2017, 139, 9019−9025.(195) Maitra, U.; Gupta, U.; De, M.; Datta, R.; Govindaraj, A.; Rao,C. N. R. Angew. Chem., Int. Ed. 2013, 52, 13057−13061.(196) Chang, K.; Hai, X.; Pang, H.; Zhang, H.; Shi, L.; Liu, G.; Liu,H.; Zhao, G.; Li, M.; Ye, J. Adv. Mater. 2016, 28, 10033−10041.(197) Lee, Y.-H.; Zhang, X.-Q.; Zhang, W.; Chang, M.-T.; Lin, C.-T.;Chang, K.-D.; Yu, Y.-C.; Wang, J. T.-W.; Chang, C.-S.; Li, L.-J.; Lin, T.-W. Adv. Mater. 2012, 24, 2320−2325.(198) Zheng, J.; Zhang, H.; Dong, S.; Liu, Y.; Nai, C. T.; Shin, H. S.;Jeong, H. Y.; Liu, B.; Loh, K. P. Nat. Commun. 2014, 5, 2995.(199) Zhou, P.; Xu, Q.; Li, H.; Wang, Y.; Yan, B.; Zhou, Y.; Chen, J.;Zhang, J.; Wang, K. Angew. Chem., Int. Ed. 2015, 54, 15226−15230.(200) Sercombe, D.; Schwarz, S.; Pozo-Zamudio, O. D.; Liu, F.;Robinson, B. J.; Chekhovich, E. A.; Tartakovskii, I. I.; Kolosov, O.;Tartakovskii, A. I. Sci. Rep. 2013, 3, 3489.(201) Najmaei, S.; Liu, Z.; Zhou, W.; Zou, X.; Shi, G.; Lei, S.;Yakobson, B. I.; Idrobo, J.-C.; Ajayan, P. M.; Lou, J. Nat. Mater. 2013,12, 754−759.(202) Chen, J.; Tang, W.; Tian, B.; Liu, B.; Zhao, X.; Liu, Y.; Ren, T.;Liu, W.; Geng, D.; Jeong, H. Y.; Shin, H. S.; Zhou, W.; Loh, K. P. Adv.Sci. 2016, 3, 1500033.(203) Okada, M.; Sawazaki, T.; Watanabe, K.; Taniguch, T.; Hibino,H.; Shinohara, H.; Kitaura, R. ACS Nano 2014, 8, 8273−8277.(204) Wang, X.; Gong, Y.; Shi, G.; Chow, W. L.; Keyshar, K.; Ye, G.;Vajtai, R.; Lou, J.; Liu, Z.; Ringe, E.; Tay, B. K.; Ajayan, P. M. ACSNano 2014, 8, 5125−5131.(205) Zhang, M.; Zhu, Y.; Wang, X.; Feng, Q.; Qiao, S.; Wen, W.;Chen, Y.; Cui, M.; Zhang, J.; Cai, C.; Xie, L. J. Am. Chem. Soc. 2015,137, 7051−7054.(206) Yu, Y.; Hu, S.; Su, L.; Huang, L.; Liu, Y.; Jin, Z.; Purezky, A. A.;Geohegan, D. B.; Kim, K. W.; Zhang, Y.; Cao, L. Nano Lett. 2015, 15,486−491.(207) Li, M.-Y.; Shi, Y.; Cheng, C.-C.; Lu, L.-S.; Lin, Y.-C.; Tang, H.-L.; Tsai, M.-L.; Chu, C.-W.; Wei, K.-H.; He, J.-H.; Chang, W.-H.;Suenaga, K.; Li, L.-J. Science 2015, 349, 524−528.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2274

(208) Zheng, W.; Feng, W.; Zhang, X.; Chen, X.; Liu, G.; Qiu, Y.;Hasan, T.; Tan, P.; Hu, P. A. Adv. Funct. Mater. 2016, 26, 2648−2654.(209) Gao, G.; Jiao, Y.; Ma, F.; Jiao, Y.; Waclawik, E.; Du, A. J. Phys.Chem. C 2015, 119, 13124−13128.(210) Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.;Strano, M. S. Nat. Nanotechnol. 2012, 7, 699−712.(211) Chen, K.; Wan, X.; Xie, W.; Wen, J.; Kang, Z.; Zeng, X.; Chen,H.; Xu, J. Adv. Mater. 2015, 27, 6431−6437.(212) Kudo, A.; Miseki, Y. Chem. Soc. Rev. 2009, 38, 253−278.(213) Wang, Z.; Peng, J.; Feng, X.; Ding, Z.; Li, Z. Catal. Sci. Technol.2017, 7, 2524−2530.(214) Chang, K.; Li, M.; Wang, T.; Ouyang, S.; Li, P.; Liu, L.; Ye, J.Adv. Energy Mater. 2015, 5, 1402279.(215) Enyashin, A. N.; Yadgarov, L.; Houben, L.; Popov, I.;Weidenbach, M.; Tenne, R.; Bar-Sadan, M.; Seifert, G. J. Phys. Chem. C2011, 115, 24586−24591.(216) Voiry, D.; Goswami, A.; Kappera, R.; SilvaCecilia de CarvalhoCastro, e.; Kaplan, D.; Fujita, T.; Chen, M.; Asefa, T.; Chhowalla, M.Nat. Chem. 2015, 7, 45−49.(217) Tang, Q.; Jiang, D.-e. Chem. Mater. 2015, 27, 3743−3748.(218) Xu, J.; Zhang, L.; Shi, R.; Zhu, Y. J. Mater. Chem. A 2013, 1,14766−14772.(219) Wang, H.; Jiang, S.; Chen, S.; Zhang, X.; Shao, W.; Sun, X.;Zhao, Z.; Zhang, Q.; Luo, Y.; Xie, Y. Chem. Sci. 2017, 8, 4087−4092.(220) Wang, H.; Wang, B.; Bian, Y.; Dai, L. ACS Appl. Mater.Interfaces 2017, 9, 21730−21737.(221) Ding, K.; Wen, L.; Huang, M.; Zhang, Y.; Lu, Y.; Chen, Z. Phys.Chem. Chem. Phys. 2016, 18, 19217−19226.(222) Gao, J.; Wang, Y.; Zhou, S.; Lin, W.; Kong, Y. ChemCatChem2017, 9, 1708−1715.(223) Chuang, P.-K.; Wu, K.-H.; Yeh, T.-F.; Teng, H. ACSSustainable Chem. Eng. 2016, 4, 5989−5997.(224) Pan, Z.; Zheng, Y.; Guo, F.; Niu, P.; Wang, X. ChemSusChem2017, 10, 87−90.(225) Zhou, M.; Hou, Z.; Zhang, L.; Liu, Y.; Gao, Q.; Chen, X.Sustain. Energy Fuel 2017, 1, 317−323.(226) Huang, H.; Xiao, K.; Tian, N.; Dong, F.; Zhang, T.; Du, X.;Zhang, Y. J. Mater. Chem. A 2017, 5, 17452−17463.(227) Shi, X.; Fujitsuka, M.; Lou, Z.; Zhang, P.; Majima, T. J. Mater.Chem. A 2017, 5, 9671−9681.(228) Chen, X.; Chen, H.; Guan, J.; Zhen, J.; Sun, Z.; Du, P.; Lu, Y.;Yang, S. Nanoscale 2017, 9, 5615−5623.(229) Hong, Y.; Li, C.; Fang, Z.; Luo, B.; Shi, W. Carbon 2017, 121,463−471.(230) Wang, Y.; Bayazit, M. K.; Moniz, S. J. A.; Ruan, Q.; Lau, C. C.;Martsinovich, N.; Tang, J. Energy Environ. Sci. 2017, 10, 1643−1651.(231) Zhang, J.; Chen, X.; Takanabe, K.; Maeda, K.; Domen, K.;Epping, J. D.; Fu, X.; Antonietti, M.; Wang, X. Angew. Chem., Int. Ed.2010, 49, 441−444.(232) Niu, P.; Liu, G.; Cheng, H.-M. J. Phys. Chem. C 2012, 116,11013−11018.(233) Niu, P.; Yin, L.-C.; Yang, Y.-Q.; Liu, G.; Cheng, H.-M. Adv.Mater. 2014, 26, 8046−8052.(234) Hong, Z.; Shen, B.; Chen, Y.; Lin, B.; Gao, B. J. Mater. Chem. A2013, 1, 11754−11761.(235) Yu, H.; Shi, R.; Zhao, Y.; Bian, T.; Zhao, Y.; Zhou, C.;Waterhouse, G. I.; Wu, L. Z.; Tung, C. H.; Zhang, T. Adv. Mater. 2017,29, 1605148.(236) Zhao, H.; Yu, H.; Quan, X.; Chen, S.; Zhang, Y.; Zhao, H.;Wang, H. Appl. Catal., B 2014, 152, 46−50.(237) Cheng, F.; Wang, H.; Dong, X. Chem. Commun. 2015, 51,7176−7179.(238) She, X.; Liu, L.; Ji, H.; Mo, Z.; Li, Y.; Huang, L.; Du, D.; Xu,H.; Li, H. Appl. Catal., B 2016, 187, 144−153.(239) Wang, Y.; Zhang, L.; Shen, L.; Ge, S.; Yu, J.; Yan, M.Microchim. Acta 2017, 184, 2587−2596.(240) Xu, Z.; Zhuang, C.; Zou, Z.; Wang, J.; Xu, X.; Peng, T. NanoRes. 2017, 10, 2193−2209.

(241) Wu, T.; Wang, J.; Liang, W.; Zang, X.; Wang, C.; Wu, Q.;Wang, Z. Microchim. Acta 2017, 184, 2171−2180.(242) Wang, Y.; Li, L.; Wei, Y.; Xue, J.; Chen, H.; Ding, L.; Caro, J.;Wang, H. Angew. Chem., Int. Ed. 2017, 56, 8974−8980.(243) Shi, L.; Chang, K.; Zhang, H.; Hai, X.; Yang, L.; Wang, T.; Ye,J. Small 2016, 12, 4431−4439.(244) Du, A.; Sanvito, S.; Li, Z.; Wang, D.; Jiao, Y.; Liao, T.; Sun, Q.;Ng, Y. H.; Zhu, Z.; Amal, R.; Smith, S. C. J. Am. Chem. Soc. 2012, 134,4393−4397.(245) Zhang, Y.; Mori, T.; Niu, L.; Ye, J. Energy Environ. Sci. 2011, 4,4517−4521.(246) Dong, F.; Zhao, Z.; Xiong, T.; Ni, Z.; Zhang, W.; Sun, Y.; Ho,W.-K. ACS Appl. Mater. Interfaces 2013, 5, 11392−11401.(247) Zhang, J.; Zhang, M.; Sun, R. Q.; Wang, X. Angew. Chem. 2012,124, 10292−10296.(248) Li, R.; Weng, Y.; Zhou, X.; Wang, X.; Mi, Y.; Chong, R.; Han,H.; Li, C. Energy Environ. Sci. 2015, 8, 2377−2382.(249) Zhang, G.; Lan, Z.-A.; Wang, X. Chem. Sci. 2017, 8, 5261−5274.(250) Yan, S.; Wan, L.; Li, Z.; Zou, Z. Chem. Commun. 2011, 47,5632−5634.(251) Zhang, X.; Zhang, Z.; Liang, J.; Zhou, Y.; Tong, Y.; Wang, Y.;Wang, X. J. Mater. Chem. A 2017, 5, 9702−9708.(252) Maeda, K.; Xiong, A.; Yoshinaga, T.; Ikeda, T.; Sakamoto, N.;Hisatomi, T.; Takashima, M.; Lu, D.; Kanehara, M.; Setoyama, T.;et al. Angew. Chem. 2010, 122, 4190−4193.(253) Townsend, T. K.; Browning, N. D.; Osterloh, F. E. ACS Nano2012, 6, 7420−7426.(254) Maeda, K.; Murakami, N.; Ohno, T. J. Phys. Chem. C 2014,118, 9093−9100.(255) Zhang, G.; Lan, Z.-A.; Lin, L.; Lin, S.; Wang, X. Chem. Sci.2016, 7, 3062−3066.(256) Lin, L.; Wang, C.; Ren, W.; Ou, H.; Zhang, Y.; Wang, X. Chem.Sci. 2017, 8, 5506−5511.(257) Liu, J.; Liu, Y.; Liu, N.; Han, Y.; Zhang, X.; Huang, H.; Lifshitz,Y.; Lee, S.-T.; Zhong, J.; Kang, Z. Science 2015, 347, 970−974.(258) Yeh, T.-F.; Teng, C.-Y.; Chen, S.-J.; Teng, H. Adv. Mater. 2014,26, 3297−3303.(259) Wang, X.; Xu, Q.; Li, M.; Shen, S.; Wang, X.; Wang, Y.; Feng,Z.; Shi, J.; Han, H.; Li, C. Angew. Chem., Int. Ed. 2012, 51, 13089−13092.(260) Lv, X.-J.; Zhou, S.; Huang, X.; Wang, C.; Fu, W.-F. Appl. Catal.,B 2016, 182, 220−228.(261) Wang, S.; Gao, Y.; Qi, Y.; Li, A.; Fan, F.; Li, C. J. Catal. 2017,354, 250−257.(262) Srinivasan, N.; Sakai, E.; Miyauchi, M. ACS Catal. 2016, 6,2197−2200.(263) Fu, C.-F.; Luo, Q.; Li, X.; Yang, J. J. Mater. Chem. A 2016, 4,18892−18898.(264) Ju, L.; Dai, Y.; Wei, W.; Li, M.; Huang, B. Appl. Surf. Sci. 2018,434, 365−374.(265) Zhu, M.; Sun, Z.; Fujitsuka, M.; Majima, T. Angew. Chem., Int.Ed. 2018, DOI: 10.1002/anie.201711357.(266) Jin, H.; Dai, Y.; Ma, X.-C.; Yu, L.; Wei, W.; Huang, B.-B. RSCAdv. 2015, 5, 52264−52268.(267) Latorre-Sanchez, M.; Lavorato, C.; Puche, M.; Fornes, V.;Molinari, R.; Garcia, H. Chem. - Eur. J. 2012, 18, 16774−16783.(268) Wang, L.; Wang, W.; Sun, S. J. Mater. Chem. 2012, 22, 6553−6555.(269) Liu, H.; Neal, A. T.; Zhu, Z.; Luo, Z.; Xu, X.; Tomanek, D.; Ye,P. D. ACS Nano 2014, 8, 4033−4041.(270) Wang, J.; Zhang, M.; Meng, J.; Li, Q.; Yang, J. RSC Adv. 2017,7, 24446−24452.(271) Kishore, M. R. A.; Ravindran, P. ChemPhysChem 2017, 18,1526−1532.(272) Zhang, X.; Zhao, X.; Wu, D.; Jing, Y.; Zhou, Z. Adv. Sci. 2016,3, 1600062.(273) Tong, C.-J.; Zhang, H.; Zhang, Y.-N.; Liu, H.; Liu, L.-M. J.Mater. Chem. A 2014, 2, 17971−17978.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2275

(274) Sa, B.; Li, Y.-L.; Qi, J.; Ahuja, R.; Sun, Z. J. Phys. Chem. C 2014,118, 26560−26568.(275) Zhang, X.; Li, B.; Wang, J.; Yuan, Y.; Zhang, Q.; Gao, Z.; Liu,L.-M.; Chen, L. Phys. Chem. Chem. Phys. 2014, 16, 25854−25861.(276) Wang, H.; Li, X.; Yang, J. ChemPhysChem 2016, 17, 2100−2104.(277) Liu, J.; Li, X.-B.; Wang, D.; Lau, W.-M.; Peng, P.; Liu, L.-M. J.Chem. Phys. 2014, 140, 054707.(278) Naguib, M.; Kurtoglu, M.; Presser, V.; Lu, J.; Niu, J.; Heon, M.;Hultman, L.; Gogotsi, Y.; Barsoum, M. W. Adv. Mater. 2011, 23,4248−4253.(279) Naguib, M.; Mochalin, V. N.; Barsoum, M. W.; Gogotsi, Y.Adv. Mater. 2014, 26, 992−1005.(280) Zhang, X.; Zhang, Z.; Li, J.; Zhao, X.; Wu, D.; Zhou, Z. J.Mater. Chem. A 2017, 5, 12899−12903.(281) Guo, Z.; Zhou, J.; Zhu, L.; Sun, Z. J. Mater. Chem. A 2016, 4,11446−11452.(282) Peng, C.; Yang, X.; Li, Y.; Yu, H.; Wang, H.; Peng, F. ACSAppl. Mater. Interfaces 2016, 8, 6051−6060.(283) Su, T.; Peng, R.; Hood, Z. D.; Naguib, M.; Ivanov, I. N.; Keum,J. K.; Qin, Z.; Guo, Z.; Wu, Z. ChemSusChem 2018, DOI: 10.1002/cssc.201702317.

ACS Catalysis Review

DOI: 10.1021/acscatal.7b03437ACS Catal. 2018, 8, 2253−2276

2276