Chemically Modified Nanostructures for Photoelectrochemical Water

Journal of Photochemistry and Photobiology C: Photochemistry Reviews 19 (2014) 3551

Contents lists available at ScienceDirect

Journal of Photochemistry and Photobiology C:Photochemistry Reviews

jo ur nal home p ag e: www.elsev ier .com/ locate / jphotochemrev

Invited Review

Chemically modied nanostructures for photoelectrochemical watersplitting

Gongmina Department ob KLGHEI of EnEngineering, Su

a r t i c l

Article history:Received 17 MReceived in reAccepted 24 OAvailable onlin

Keywords:Photoelectrochemical water splittingPhotoanodeHydrogen generationMetal oxidesChemical modications

bandgap, diffusion distance, carrier lifetime and photostability of semiconductors. Although nanostruc-tured photoelectrodes improve the photoelectrochemical water splitting performance to some extent, byincreasing electrolyte accessible area and shortening minority carrier diffusion distance, nanostructureengineering cannot change their intrinsic electronic properties. More importantly, recent development inchemically modication of nanostructured electrodes, including surface modication with catalyst and

Contents

1. Introd1.1. 1.2. 1.3.

2. Chem2.1. 2.2. 2.3. 2.4.

3. ConclAcknoRefer

CorresponE-mail add

1389-5567/$2http://dx.doi.oplasmonic metallic structures, element doping and incorporation of functional heterojunctions, haveled to signicant enhancements in the efciencies of charge separation, transport, collection and solarenergy harvesting. In this review, we provide an overview of the recent process in photoelectrochemicalwater splitting by using chemically modied nanostructured photoelectrodes. Finally, we also discussthe current challenges and future opportunities in the area of photoelectrochemical water splitting.

2013 Elsevier B.V. All rights reserved.

uction to photoelectrochemical water splitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Principle of PEC water splitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Common electrode materials for PEC water splitting and their limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Morphology engineering and chemical modication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

ically modied nanostructure for PEC water splitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Element doped semiconductor nanostructures for PEC water splitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Metal oxide heterostructures for PEC water splitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Plasmonic materials for PEC water splitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44Surface electrochemical catalyst assisted PEC water splitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

usion and outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47wledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

ences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

ding author. Tel.: +1 8314591952.ress: [email protected] (Y. Li).

0.00 2013 Elsevier B.V. All rights reserved.rg/10.1016/j.jphotochemrev.2013.10.006g Wanga, Yichuan Linga, Hanyu Wanga, Xihong Lua,b, Yat Lia,

f Chemistry and Biochemistry, University of California, Santa Cruz, CA 95064, United Statesvironment and Energy Chemistry, MOE of the Key Laboratory of Bioinorganic and Synthetic Chemistry, School of Chemistry and Chemicaln Yat-Sen University, Guangzhou 510275, Peoples Republic of China

e i n f o

ay 2013vised form 22 October 2013ctober 2013e 1 November 2013

a b s t r a c t

Hydrogen gas is chemical fuel with high energy density, and represents a clean, renewable and carbon-free burning fuel, which has the potential to solve the more and more urgent energy crisis in todayssociety. Inspired by natural photosynthesis, articial photosynthesis to generate hydrogen energy hasattracted a lot of attentions in the eld of chemistry, physics and material. Photoelectrochemical watersplitting based on semiconductors represents a green and low cost method to generate hydrogen fuel.However, solar to hydrogen conversion efciency is quite low, due to some intrinsic limitations such as

36 G. Wang et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 19 (2014) 3551

Gongming Wang received his B.S. degree in chemistryfrom the University of Science & Technology of Chinain 2008. He joined Prof. Yat Lis group at University ofCalifornia, Santa Cruz in the fall of 2008 and received his

1. Introdu

With thtinuously inbelieved toever, the insolar energsupply, it rthe excess tive ways isas alcohols research effphotoactiveeration via

1.1. Principle of PEC water splitting

Water cannot be directly decomposed by light, because itis transparent to visible light, but only with the irradiation

ngth shorter than 190 nm (deep ultraviolet light) [1]. Forchemical water electrolysis, a minimum voltage of 1.23 V isd to split water. This voltage is equivalent to the energynce with a wavelength of 1000 nm. Therefore, if solar

can be effectively used in the electrochemical system, waterg can be achieved under visible light irradiation. The rstal photosynthesis of generating H2 by water splitting wasstrated by Honda and Fujishima in 1972, using semiconduc-nium dioxide (TiO2) as photoanode in a PEC cell [1]. Fig. 1

the conguration of PEC cell with n-type semiconductorhotoanode and a Pt counter electrode. When the semicon-

is contacted with the electrolyte, the charge transfer occursinterface between semiconductor and electrolyte, and lead-surface charging. As a result, electronic bands bend upward.tential barrier created by the band bending is known asoltz barrier, which depends on the nature of the aqueouslyte and the semiconductor electrode [2]. This interfacialial barrier could facilitate the separation of electron and holehich can be photogenerated when TiO2 photoelectrode is

ted with light with photon energy larger than or equal to itsap. The photoexcited electrons transfer to Pt counter elec-nd reduce water to generate H2, while holes diffuse to the

of TiO2 and oxidize water to form O2 [25]. The reactionons on each electrode are shown in the following:

anod

de :

ordinum e, thethannctiolly, t

eV. Horetsplit, volt

hotoetoano

ced wPhD degree in chemistry in 2013. His research focuses onchemically modied nanostructures for energy conver-sion and storage.

Yichuan Ling is currently a 5th year chemistry PhD stu-dent at University of California, Santa Cruz, under thesupervision of Prof. Yat Li. He received his B.S. degreein chemistry from Fudan University, China in 2009. Hisresearch focuses on the synthesis of hematites and III-V semiconductor nanomaterials and investigates theirapplications for solar energy conversion.

Hanyu Wang received her B.E. degree in chemical engi-neering and technology from Qingdao University, and herM.S. degree in chemistry from Shandong University. She isa 4th year chemistry PhD student in Prof. Yat Lis group atUniversity of California, Santa Cruz. Her research focuseson the development of high performance nanostructu-red metal oxide electrodes for photoelectrochemical andmicrobial photoelectrochemical systems.

Xihong received his B.S. degree in applied chemistry fromSun Yat-sen University in 2008, and then earned his doc-tor degree in 2013 under the guidance of Prof. YexiangTong from Sun Yat-sen University. During 20112013, hewas a visiting graduate student in Prof. Yat Lis lab at Uni-versity of California, Santa Cruz. He is now a lecturer inthe school of chemistry and chemical engineering at SunYat-sen University. His research interests center on thedevelopment of nanostructured materials for the appli-cation of supercapacitors, photocatalyst and functionalabsorber materials.

Yat Li received his B.S. and PhD in chemistry from theUniversity of Hong Kong. He was a postdoctoral researchfellow at Harvard University, from 2003 to 2007, under thesupervision of Prof. Charles M. Lieber. He joined UC SantaCruz in 2007 and he is now an associate professor of chem-istry at UC Santa Cruz. His research focuses on the designand synthesis of semiconductor nanostructures, investi-gation of their fundamental properties and exploration oftheir potentials for solar energy conversion and storage.

ction to photoelectrochemical water splitting

e ever-growing of global population, there is a con-creasing energy demand. Renewable solar energy is

be a potential solution to energy sustainability. How-termittent solar irradiation poses a new challenge fory utilization. To obtain continuous and stable powerequires efcient and cost effective methods to storeenergy generated in daytime. One of the most attrac-

storing solar energy in the form of chemical fuels, suchand hydrogen gas. Inspired by natural photosynthesis,orts have been devoted to mimic this process by using

materials. One of the major directions is hydrogen gen-photoelectrochemical (PEC) water splitting.

waveleelectrorequireirradiaenergysplittinarticidemontor titashowsTiO2 pductorat the ing to The poHelmhelectropotentpairs, wirradiaband gtrode asurfaceequati

Photo

Catho

Accminimenergylarger as a fuoreticais 1.23the thewater nation

Fig. 1. PTiO2 pho

Reprodue : H2O + 2h+ 2H+ + 1/2O2 Eox = 1.23 V2H+ + 2e H2 Ered = 0 V

g to the Nernst equation, water electrolysis requires anergy of 1.23 V. To gain the required electrochemical

photoelectrode must absorb light with photon energy 1.23 eV. Fig. 2 shows the solar energy spectrum plotn of number of photons and radiation energy [2]. The-he minimum energy required for PEC water splittingowever, in practice, photons with energy larger than

ical limit is needed due to the energy loss during PECting. The energy losses include electron-hole recombi-age losses at the contacts, and the potential loss due to

lectrochemical water splitting cell based on n-type semiconductorde.

ith permission from Ref. [7].

G. Wang et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 19 (2014) 3551 37

Fig. 2. Number of photons in solar spectrum, versus photon energy.Reproduced with permission from Ref. [2].

the electrode and contact resistance [2]. The estimated energy lossis around 0.8 eV; therefore, the practical voltage required for PECwater splitting is 2.0 eV. Additionally, the water splitting voltageis also depends on the electrode surface overpotential for wateroxidation as well as the band edge positions of the photoelectrode.To achieve of the semtrolysis potmost cases,bandgap of

1.2. Commotheir limitat

Since thelectrode, aied, includi[3953], WTa3N5 [62[6871] andphotoelectrdepending chemical prof common

hydrogen electrode potential [77]. For metal oxides, only largebandgap metal oxides such as TiO2, SrTiO3 and ZnO can strad-dle the water reduction and oxidation potentials [4]. However, thelarge bandgaps limit their absorption of sunlight. For small bandgapmetal oxides such as Fe2O3 and WO3, their conduction band poten-tials are more positive than water reduction potential, and thus,the application of external bias is needed to split water. Besides,the poor conductivity and extremely short carrier diffusion lengthof Fe2O3 result in signicant electron-hole recombination loss [4].Metal chalcogenides such as CdS and CdSe have band structuresthat are favorably aligned with water electrolysis potentials. How-ever, they are not stable in oxidative environment and can be easilyoxidized to metal oxides in aqueous system [13,23,78,79]. Likewise,metal nitrides or metal oxynitrides also suffer from photocorro-sion and photooxidation, which cause gradual reduction of theirphotoactivity [65]. Silicon is small bandgap semiconductor that canabsorb a large portion of solar light. Nevertheless, the valence bandof silicon is not positive enough to oxidize water and the applicationof external bias is needed [3]. Additionally, silicon is photoelectro-chemically instable due to the surface oxidation and large surfaceover-potential [72]. In summary, each semiconductor electrodematerial has its own limitations. Research efforts have been pri-marily focused on addressing these limitations and achieving betterphotoelectrodes.

orph

elopprovi

in te extort e thcom

bsorping l

the nducemomicastru. Checation ca

Fig. 3. The ban

Reproduced wnon-biased water electrolysis, the band edge positionsiconductor electrode should straddle the water elec-entials, E(H+/H2) and E(O2/H2O) [5,6]. Therefore, in

an external bias is still needed to be applied even if the the photoelectrode is larger than 1.23 eV.

n electrode materials for PEC water splitting andions

e rst demonstration of PEC water splitting using TiO2 number of semiconductor materials have been stud-ng TiO2 [820], ZnO [2128], -Fe2O3 [2938], BiVO4O3 [5461], metal nitrides and phosphides such as65] and GaP [66,67], metal oxynitrides such as TaON

silicon such as n-type and p-type silicon [7276]. Theseodes can be used as photoanodes or photocathodes,on their chemical nature, band structure and electro-operties. Fig. 3 shows the relative band edge positions

semiconductors versus vacuum potential and normal

1.3. M

Devcould lengesprovidand shimprovhole relight aincreasmodifyable cobeen d

Cheof nanonaturemodireactiod edge potentials of common semiconductor photoelectrode materials plot against the n

ith permission from Ref. [77].ology engineering and chemical modication

ing chemically modied nanostructured electrodesde new opportunities in addressing the current chal-he PEC water splitting. For instance, nanostructuresremely large semiconductor/electrolyte interfacial areadiffusion distance for minority carriers, which cane charge separation and reduce the loss due to electron-bination [5,80,81]. Nanostructure could also increasetion efciency via reducing surface light reection andight scattering. Moreover, quantum connement couldband structure of semiconductors; for instance, the tun-tion band potential of CdSe and CdSe quantum dots hasnstrated via the control of quantum dot sizes [82,83].l modications could further improve the performancectured photoelectrodes by manipulating their chemicalmical modications including element doping, surfacen of electrochemical catalysts such as oxygen evolutiontalyst (OER) and hydrogen evolution reaction catalyst

ormal hydrogen electrode (NHE) at pH 0 and vacuum potential.


(HER), surface plasmonic effect modication, and formation ofhetero-junctions, have been extensively explored to increase theperformance of nanostructured electrodes for PEC water split-ting. Element doping has been demonstrated to be effective in themodicatiosemiconducof defects be a generaphotoconduHowever, eaffect the tron/hole reof semicond

Surface lize and increported ththin lm opreventing conductor eoxidation oextra photochemical cfor water othe overpoalyst modiof photoelemore, photand Ag nantion of semabsorption,the formattwo semicoband structincrease lig

2. Chemicasplitting

Recent signicantlstructured we will revon the photWe will alsoopportuniti

2.1. Elemensplitting

Elementtronic and Various doptin [103,104nitrogen [16and carbonwe will revsemiconduc

It has bechange theZnO and TiOThe visible of impuritiethe p orbitoxide [111]doped with

[9,108110] and sulfur [10,107,118,119]. For example, carbon-doped TiO2 for PEC water splitting was reported by Khan andcoworkers in 2002. The incorporation of carbon was achieved byame oxidization of Ti foil in natural gas [85]. The band gap of

as re-dopotoce TiOince dataced p

lightbe amped bynninwithotocly hin hargarticl-dopan tom). dueas derogenly, H

usin]. TiOd on2 sa

es cop to ire oni

inserodund stn-do0 nmbserted n. On n enhwiseed in,120]nd eble cnitutoeletionaeir imnO nhermoraten alslightnowt denotoatal oxtoanoxynn of electronic and optical properties of metal oxidetors [8490]. In addition, the controlled introductionsuch as oxygen vacancies has been demonstrated tol method to increase the electrical conductivity andctivity of metal oxides semiconductors [4,12,14,54].xcess dopants could increase density of trap states,charge mobility, resulting in the increasing of elec-combination loss and the reduction of PEC performanceuctor photoelectrodes.modication is another important strategy to stabi-rease PEC performance of photoelectrodes. It has beenat the deposition of metal oxide such as TiO2 and Fe2O3n silicon can stabilize the photoactivity of silicon bysilicon surface oxidation [72]. Moreover, some semi-lectrodes have large surface over-potentials for waterr water reduction, which need to be compensated withvoltage or the application of external bias. Electro-

atalysts have been used to reduce the overpotentialxidation or reduction [35,45,47,9195]. By suppressingtential for water oxidation and reduction, photocat-cation also help to prevent self-oxidation/reductionctrode and improve their photostability [65]. Further-oanodes decorated with plasmonic metal such as Auoparticles have been reported to increase light absorp-iconductor electrodes, via electric eld enhanced light

scattering and hot electron injection [9699]. Finally,ion of heterojunctions (type II junction) by couplingnductor materials with different chemical nature andure could facilitate charge separation at interface andht absorption [43,100,101].

lly modied nanostructure for PEC water

advancement in chemical modication methods hasy improved the performance of various kinds of nano-photoelectrodes for PEC water splitting. In this article,iew these chemical modication methods with a focusoanodes and their applications for PEC water splitting.

give an overview of the current challenges and futurees in the area of PEC water splitting.

t doped semiconductor nanostructures for PEC water

doping is a promising approach to modify the elec-optical properties of the semiconductor electrodes.ants including metal ions such as titanium [34,102],] and iron [105,106] and non-metal elements such as,22], sulfur [10,107], silicon [29,30], boron [9,108110]

[11,85] have been reported. In the following section,iew the doping effect on a number of commonly usedtor photoelectrodes.en demonstrated that non-metal element doping can

optical property of wide bandgap metal oxide such as2 to achieve visible light photoactivity [8,11,16,17,22].light absorption is believed to be due to the formations states in the metal oxide bandgap or the mixing of

al of non-metal dopant and the O 2p orbital of metal. Visible light absorption has been reported for TiO2

carbon [8,11,85,112,113], nitrogen [114117], boron

TiO2 wcarbonthe phpristinever, s(IPCE) enhanvisiblenanotuhigh tethesizethe scatubes, The phsiderabTiO2 thtorial cnanopcarbonsity th(>420 nof TiO2TiO2 w

NitRecentdation[16,17methoing TiOanalysgen (unanowin amm(Fig. 5,the intthe banitrogeof 52tivity osupporof TiO2effect i

Likeobserv[62,68and bacal staof magas phothe raand th[22]. Zhydrotincorption cawhite ZnO nacurrenhas ph

Meas phometal duced from 3.0 to 2.3 eV after doped with carbon. Theed TiO2 absorbed light at wavelength up to 535 nm andurrent density was enhanced by 4 times compared to2 under the same measurement conditions [85]. How-the incident photon-to-current conversion efciency

was not provided in this article, it was not clear if thehotocurrent was due to the increased photoresponse in. Later, Park et al. demonstrated carbon doping on TiO2rrays by annealing in carbon monoxide atmosphere atrature (500-800 C) [8]. TiO2 nanotube arrays were syn-

an electrochemical anodization method. Fig. 4a showsg electron microscopy (SEM) image of the TiO2 nano-

diameter of around 70 nm and wall thickness of 20 nm.urrent density of pristine TiO2 nanotube arrays was con-igher than P25 nanoparticle thin lms and nanoporouslm (Fig. 4b). This was attributed and the improved vec-e transport in one-dimensional structure, compared toe lm and nanoporous lm [8]. More importantly, theed TiO2 electrode exhibited higher photocurrent den-

pristine TiO2 electrode under visible light illuminationThese results clearly supported the bandgap narrowing

to carbon doping. The bandgap energy of carbon-dopedtermined to be 2.2 eV (Fig. 4d) [8].

doping is another extensively studied dopant for TiO2.oang et al. reported visible light driven PEC water oxi-g nitrogen-doped TiO2 nanowire arrays as photoanode2 nanowire arrays were synthesized by a hydrothermal

FTO glass. Nitrogen doping was carried out by anneal-mple in ammonia gas at temperature of 500 C. XPSnrmed the presence of substantial amount of nitro-1.08% atomic %) in TiO2 [16]. The color of pristine TiO2lm was changed from white to yellow after annealinga gas, suggesting visible light absorption after dopingt). The increased visible light absorption is attributed toced energy states of substitutional nitrogen species inructure of TiO2 [16]. The IPCE spectra revealed that theped TiO2 exhibits photoactivity up to the wavelength, corresponding to 2.4 eV; while there was no photoac-ved for pristine TiO2 beyond 420 nm [16]. These resultsitrogen doping can increase visible light photoactivity

the whole, carbon and nitrogen dopings showed similarancing the visible light photoactivity of TiO2 materials., visible light absorption and photoactivity were also

other nitrogen-doped large bandgap semiconductors such as ZnO and Ta2O5 [22,69]. ZnO has similar bandgapdge positions as TiO2. Although ZnO is less chemi-ompared to TiO2, its electron mobility is two ordersde higher than TiO2, which makes it very attractivectrode for PEC water splitting. Yang et al. reportedl synthesis of nitrogen-doped ZnO nanowire arraysplementation as photoanode for PEC water splitting

anowire arrays were synthesized by a seed mediatedal growth and followed by annealing in ammonia to

N as a dopant [22]. The nitrogen dopant concentra-o be controlled by varying the annealing time. Under

illumination (AM 1.5G, 100 mW/cm2), nitrogen dopedires showed an order of magnitude increase in photo-sity. IPCE results conrmed that nitrogen-doped ZnOctivity in visible light up to 420 nm.ynitrides and metal nitrides were also demonstratedodes that absorb visible light. The valence bands ofitrides are more negative than those of oxides due to


Fig. 4. (a) Top ess othree different M KOof carbon dope d) Banat 600 C from

Reproduced w

the hybridition bands resulting inoxides [120more favorwater splittwith band gof 2.07 [62]simple metlm, by anooxide nanotfor nitridatiaverage inn7 nm shownTa3N5 is loccompared t

ing her so prgreese re

is an

Fig. 5. IPCE sparrays (lower,

Reproduced w view of scanning electron microcopy image of TiO2 nanotube arrays with a thickn TiO2 photoanodes including P25, nanoporous TiO2 and TiO2 nanotube arrays, in 1.0d TiO2 nanotube arrays under white light and visible light (>420 nm) illumination. (

(jphh)1/2 versus h plot.


zation of N 2p with O 2p orbitals, while the conduc-consist predominantly of empty orbitals of the metal,

similar energy levels to those of corresponding metal]. Therefore, some oxynitride and nitride materials haveable bandgap energy and band edge positions for PECing. For example, Ta O is a large bandgap metal oxide

narrowthe higdata algood aAll thedoping2 5

ap of 4.2 eV, while Ta3N5 and TaON have a bandgap and 2.5 eV [120], respectively. Feng et al. developed ahod to fabricate highly oriented Ta3N5 nanotube arraydization of tantalum foil to achieve oriented tantalumube arrays followed by annealing in ammonia at 700 Con [62]. The prepared Ta3N5 nanotube arrays have aner pore diameter of 22 nm and wall thickness of about

in Fig. 6a and b. Importantly, the absorption edge ofated at about 600 nm (Fig. 6c), red shifted by 300 nmo that of Ta2O5 nanotube arrays lm [62]. The bandgap

metal oxidemetal oxynof these elepotential somethods in

Elementimprove elhematite hhowever, thand extrem

ectra of pristine TiO2 and nitrogen doped TiO2 nanowire arrays. The insets are digital imagright) and band structure of nitrogen doped TiO2 (right).

ith permission from Ref. [16].f 3 m. (b) The photocurrent density versus measured potentials forH solution. (c) The photocurrent density versus measured potentialsd-gap determination of carbon doped TiO2 nanotube arrays prepared

is due to the substitution of N for O, which results frompotential energies of N 2p orbitals than O 2p. The IPCEove the visible light photoactivity of Ta3N5, which is inment with the UVvis diffuse reectance spectra [62].sults have clearly demonstrated substitutional nitrogen

effective method to increase visible light absorption for

s. The major challenge for nitrogen-doped metal oxides,itrides and nitrides is photostability due to re-oxidationctrodes during PEC water oxidation. We will discuss thelutions to this problem using chemically modication

later section. doping is also an effective and general strategy toectronic properties of semiconductors. For example,as a favorable band gap for visible light harvesting;e poor electrical conductivity and photoconductivityely short diffusion distance for charge carrier limit its

es of N-TiO2 lm (upper, right), side view SEM image of TiO2 nanowire


Fig. 6. (a) Top surfimage of prepa lms electrode with

Reproduced w

applicationet al. reporthin hematTiSi2 nanonarea scaffolefciency oFe2O3/TiSi2[33]. Howethick layer thin lm lia potentialof hematitebe improvebe modulatdopants. n-have been hematite. Fsynthesized(APCVD) anThe silicon-area and shnormal to hematite ldensity comphotocurreenhancemewhich acts its electric[89]. Furthelm for PECby a deposititanium n-

-doptron view of SEM image of prepared Ta3N5 nanotube arrays. (b) SEM image of the bottomred Ta3N5 lm on Ta foil. (c) Diffuse reectance UVvis spectra of Ta3N5 nanotube

different thickness measured in 1 M KOH at an applied bias of 0.5 V.


as photoelectrode [80]. To address this limitation, Linted to use atomic layer deposition method to deposit

siliconas elecite lm (10 nm) on TiSi2nanonet structure [33]. Theet is highly conductive, working as a high surfaced for charge collection. An excellent external quantumf 46% at wavelength of 400 nm was achieved on thenanonet electrode, without any intentional doping

ver, hematite is an indirect bandgap material and ais required for complete light absorption, the hematitemits the amount of light absorption and would be

drawback for this approach. To have a thick layer materials, the charge transport of hematite shouldd. The electrical properties of semiconductors caned by intentionally doped with intrinsic or extrinsicType dopants such as Ti [34], Si [29] and Sn [103]used to increase the electrical conductivity of n-typeor example, silicon doped hematite thin lms were

using atmospheric pressure chemical vapor depositiond ultrasonic spray pyrolysis (USP) methods [29,89].doped hematite nanostructure exhibited large surfaceowed a preferred orientation with (0 0 1) basal planethe substrate [29]. The APCVD grown silicon-dopedm showed orders of magnitude enhanced photocurrentpared to the undoped hematite lm, and achieved a

nt density of 1.45 mA/cm2 at 1.23 V vs. RHE [89]. Thent was attributed to the substitutional silicon doping,as an electron donor for hematite, and thus, improvedal/photoelectrical conductivity and charge transportrmore, Wang et al. reported to use Ti-doped hematite

water splitting [34]. Ti-doped -Fe2O3 was preparedtion-annealing process using FeCl3 as Fe precursor andbutoxide as dopant precursor. Similar to the case of

studies shoby two ordto undoped-Fe2O3 wadensity warange (Fig. tially enhan350-610 nmdoped -Feindicating t

In addiconductorsstrategy to csemiconduOxygen vacoxides. Thiarrayed phonanowire amethod, fobetween 20TiO2 nanowformation ostates are bthe conducwere used t(H-TiO2) saplots reveaTiO2. The d

Nd = (2/e0ace of the prepared Ta3N5 nanotube array lm. The inset is the digitalwith different thickness. (d) IPCE spectra for Ta3N5 nanotube arrayed

ed TiO2, Fe3+ was substituted by Ti4+ that functioned donor for -Fe2O3 lms. Electrochemical impedance

wed that the donor density of hematite was increaseders of magnitude after titanium doping. In comparison

-Fe2O3, the photocurrent onset potential of Ti-dopeds negatively shifted by 0.1-0.2 V and the photocurrents enhanced by two times in the entire potential7a) [34]. The Ti doped -Fe2O3 lm shows substan-ced IPCE values over the entire wavelength range of. There was no photoresponse for both undoped and Ti-2O3 lms, in accordance with the bandgap of hematite,he doping did not narrow the bandgap of -Fe2O3 [34].tion to introducing extrinsic dopants into semi-, Li and coworkers have recently demonstrated a novelreate intrinsic defects, oxygen vacancies, in metal oxidectors to increase their carrier densities [4,12,54,121].ancies function as shallow donors for a number of metals idea was initially demonstrated on TiO2 nanowire-toelectrodes by hydrogen thermal treatment [12]. TiO2rrays were synthesized on FTO glass via a hydrothermalllowed by hydrogen treatment at the temperature0 and 550 C. XPS studies supported that the surface ofires was functionalized with hydroxyl groups and thef Ti3+ (oxygen vacancy) [12]. The oxygen vacancy energyelieved to be located at around 0.75 and 1.18 eV belowtion band of TiO2. Electrochemical impedance studieso estimate the carrier density of hydrogen-treated TiO2mples (Fig. 8d). The positive slopes of MottSchottkyled the n-type properties of TiO2 and hydrogen treatedonor densities were calculated by the equation:

0)[d(1/C2)/dV ]1


Fig. 7. (a) The (b) IPCE spectr

Reproduced w

where e0 is0 is the pethe appliedwere increabe due to thTiO2 (H-TiOphotocurreachieved a vs. Ag/AgClactivity waswas no visibwas observtrum with achieved th1.1%, at 0strated the oxygen vacmance [12][123], photand perform

Li and cocreate oxyggen decienthe preparaimplementa

oxygen decient hematite nanowires (denoted as N-hematite)were obtained though thermal decomposition of FeOOH nanowiresin an oxygen decient atmosphere (N2 + air). The existence of Fe2+

the N-hematite was conrmed by XPS analysis shown in [124lly inMosneali). Th

withemeiO2 agas,

oxidaonclutal dtion couldndu

etal o

formed fumicofacili

by tce, woreosites inFig. 9astantia[124]. the an(Fig. 9csistentenhancwise, Tand N2water

In celemenabsorpwhich semico

2.2. M

Theenhanctwo secould driveninterfation. Mlinear sweep voltammograms of undoped and Ti doped -Fe2O3 lms.a of these two lms collected at 1.02 V vs. RHE, in 1.0 M NaOH solution.


the electron charge, is the dielectric constant of TiO2,rmittivity of vacuum, Nd is the donor density and V is

bias at the electrode. The donor densities of H-TiO2sed by three orders of magnitude, which is believed toe creation of oxygen vacancies. The hydrogen-treated2) nanowire electrodes showed substantially increasednt densities compared to pristine TiO2 electrode, andremarkable photocurrent density of 2.5 mA/cm2 at 0 V

[12]. IPCE studies suggested that the enhanced photo- due to the improved charge collection efciency. Therele light photoactivity, although visible light absorptioned for H-TiO2 nanowires. By integrating the IPCE spec-standard AM 1.5G solar spectrum, the H-TiO2 samplee excellent solar to hydrogen conversion efciency of.6 V vs. Ag/AgCl. Li and coworkers have further demon-hydrogen treatment is a general method to introduceancies into metal oxides and improve their PEC perfor-. Hydrogen treated WO3 [54], ZnO [21,122] and BiVO4oanodes all showed enhanced electrical conductivityance for PEC water splitting.workers also demonstrated an alternative method toen vacancies in metal oxide by annealing them in oxy-t conditions [14,124]. For example, Ling et al. reportedtion of oxygen decient hematite nanowires and theirtion as photoanode for PEC water splitting [124]. The

with small toactivity. Vas WO3/BiVFe2O3/BiVOhydrogen gadvancemeerojunction

Recentlytrochemicaheterojunctan attractivlight with wbandgap enbandgap oftively poora type II Wprovide a sycharge sepaby layer-byture [100]. composite of BiVO4 ardue to the pThe holes mdirectly or The type II cient transfthe chance odensity cur[43]. The cthan both Btocurrent obare WO3 in the entirtransfer at Hong et al. chemical im]. As a result, the donor density of N-hematite was sub-creased, compared to the air annealed hematite (Fig. 9b)t importantly, the small change of oxygen content inng process can signicantly enhance the photocurrente photoactivity of oxygen decient hematite was con-

the bandgap of hematite, suggesting the photocurrentnt was due to improved electrical conductivity. Like-nnealed in oxygen decient condition such as vacuum

also exhibits substantially improved photoactivity fortion [14].sion, the above mentioned studies have demonstratedoping is an effective strategy that can improve the lightproperties and electronic properties of semiconductors,

potentially address some intrinsic limitations of thesectors in the application for PEC water splitting.

xide heterostructures for PEC water splitting

ation of heterojunctions could enable new and/ornction of semiconductor photoelectrodes. For example,nductors form a heterojunction with type II alignmenttate the charge separation. The charge separation ishe internal electric eld formed at the heterojunctionhich helps to suppress the electronhole recombina-ver, large bandgap semiconductor forms heterojunctionbandgap semiconductor can increase visible light pho-arious heterojunction photoelectrode structures suchO4 [43,100], TiO2/SrTiO3 [125], WO3/Fe2O3 [126] and4 [127] have shown enhanced performance for PECeneration. In this section, we will review the recentnt on the design and fabrication of nanostructured het-

for PEC water splitting., Hong et al. and Su et al. reported the synthesis, elec-l and photoelectrochemical properties of WO3/BiVO4ion structure for PEC water splitting [43,100]. WO3 ise photoelectrode material, which can absorb the solaravelength of less than 500 nm, in accordance to theergy of 2.7 eV [5456]. BiVO4 has an even smaller

2.4 eV, however, its photoactivity is limited by the rela- charge transport properties [40,49]. The formation ofO3/BiVO4 heterojunction photoelectrode was able tonergistic effect in improving both light absorption andration. The heterojunction electrodes were fabricated-layer deposition of WO3 and BiVO4 on WO3 nanostruc-Fig. 10a shows the potential energy diagram for thelms. The photoexcited electrons in the conduction bande thermodynamically favorable to inject into the WO3,otential difference between the conduction bands [43].igrate to the semiconductor/electrolyte interface eitherafter transfer from WO3 to BiVO4 for water oxidation.alignment of conduction and valence bands allows ef-er of photoexcited electrons and holes, which reducesf recombination [100]. Fig. 10b shows the photocurrent

ves of bare WO3, BiVO4 and WO3/BiVO4 heterojunctionomposite lm showed higher photocurrent densitiesiVO4 and WO3 lms, and achieved the highest pho-f 0.8 mA/cm2 at 1 V, which is two times higher thanlm. The composite lm exhibited higher IPCE values

e range between 340 and 540 nm, indicating the chargethe interface between WO3 and BiVO4 is efcient [43].also studied the interfacial charge transfer by electro-pedance spectroscopy (EIS) measurements [100]. EIS


Fig. 8. (a) Linear sweep voltammograms of pristine TiO2 nanowire arrays and hydrogen treated TiO2 nanowire arrays at 350, 400 and 450 C in 1.0 M NaOH aqueous solution.(b) IPCE spectra of pristine TiO2 and hydrogen treated TiO2 samples, collected at the incident wavelength range from 300 to 650 nm at a potential of 0.6 V vs. Ag/AgCl. Inset:Magnied IPCE spectra that highlighted in the dashed box at the incident wavelength range from 440 to 650 nm. (c) Simulated solar to hydrogen conversion efcienciesfor the pristine TiO2 and hydrogen treated TiO2 samples as a function of wavelength, by integrating their IPCE spectra collected at 0.6 V vs. Ag/AgCl with a standard AM1.5G solar spectrum (ASTM G-173-03). (d) MottSchottky plots collected at a frequency of 5 kHz in the dark for the pristine TiO2 and hydrogen treated TiO2 samples. Inset:MottSchottky plots of hydrogen treated TiO2 nanowires prepared at 350, 400 and 450 C, collected under the same conditions.

Reproduced with permission from Ref. [12].

Fig. 9. (a) The XPS Fe 2p spectra of A-hematite and N-hematite nanowires, together with their difference spectrum (N-hematite minus A-hematite). The vertical dashed lineshighlight the satellite peaks for Fe3+ and Fe2+. (b) The comparison of MottSchottky plots of A-hematite and N-hematite; the inset is the magnied MottSchottky plot ofNi-hematite. (c) Linear sweep voltammograms collected for A-hematite and N-hematite in 1.0 M NaOH solution under 100 mW/cm2AM 1.5 solar light illumination. (d) IPCEspectra collected at 1.23 V and 1.5 V vs. RHE, respectively.



Fig. 10. (a) The energy diagram of the WO3/BiVO4 heterojunction (at pH 7) and electron transfer. (b) current densitypotential (with respect to normal hydrogen electrodeNHE scale) plots for planar WO3/BiVO4 heterojunction lm in red line, bare planar WO3 lm in blue line, and bare planar BiVO4 lm in green line, illuminated with choppedwhite light (100 mW/cm2) in an aqueous solution of 0.5 M Na2SO4.


spectra were measured under simulated solar light illuminationand presented in Nyquist diagram in the frequency range of 100 kHzto 100 MHz for bare WO3, BiVO4 and composite lms [100]. Thesemicircular arc in the Nyquist plot designates the charge trans-fer kinetics on the working electrode. In the equivalent circuit, Rsis the solution resistance, Q1 is the constant phase element (CPE)for the electrolyte/electrode interface, and Rct is the charge trans-fer resistance across the electrode/electrolyte interface. The resultsshowed that the bare WO3 has excellent charge transfer efciencywith the lowest Rct. Yet, the BiVO4 exhibited the largest Rct, whichsuggests the low activity of bare BiVO4 was due to the poor chargetransfer efciency. Interestingly, the charge transfer rate for thebare WO3 and composite 1 lm is similar. It clearly demonstrated

that the charge transfer of BiVO4 lm can be improved by forminga heterojunction with WO3 [100].

Furthermore, the heterojunction structures can also be used toincrease the visible light absorption of large bandgap metal oxides.For example, small bandgap metal suldes or selenides such as CdSand CdSe quantum dots have been used to sensitize wide bandgapmetal oxides such as ZnO and TiO2 for PEC hydrogen generation[13,14,23,128]. First, these small bandgap quantum dots substan-tially increase the visible light absorption of ZnO and TiO2 basedphotoelectrodes. The type II junction also allows the rapid separa-tion and transfer of photoexcited electron-hole pairs. However, thephotostability is a major concern for CdSe and CdS based sensiti-zers due to their self-oxidation reaction. Sacricial reagents such as

Fig. 11. (a) Entreated, TaON four measuredspectra of eleclight illuminat

Reproduced wergy diagram and charge ow of the TaON/CaFe2O4 heterojunction photoelectrode. (b) Ptreated/CaFe2O4 and TaON-as deposited/CaFe2O4 in 0.6 M NaOH solution under chopped

electrodes. (c) IPCE of TaON-treated electrode (triangle) and TaON-as deposited/CaFe2Otrodes are also presented. The inset shows the interface between TaON and CaFe2O4. (d)ion (>420 nm) at a DC potential of 0.2 V vs. Ag/AgCl with an AC potential frequency range

ith permission from Ref. [129].hotocurrent densities versus potentials of as deposited TaON, TaON light irradiation (>420 nm). The inset is the schematic models for the4 electrode (diamond) at 0.2 V vs. Ag/AgCl. UVvis diffuse reectance

EIS spectra of corresponding photoelectrodes in Nyquist plots under from 100,000 to 0.1 Hz.


S2 were added to prevent the self-oxidation, while the photoan-ode is no longer oxidizing water. Sacrice chemicals to generatehydrogen are certainly not desirable. Therefore, it is necessary todevelop smmetal oxide

The semalso be comwell demoncan separateld induceconcept to pn heteroj[129]. n-Taoxidation ualigned banoxide semiedge positioof bare CaFeand transpojunction coufor CaFe2O4fabricated CaFe2O4 poheterojuncttion and fobetween TaCaFe2O4 heirradiated fton energy by the layesolar light uhetero-intephotocurrelms as a f(>420 nm). sity of 4 Aattributed tducting suband greatlythe TaCl5 trwhich showhigher thanthe direct cwas used aafter CaFe2As a result,1.26 mA/cmthe TaON-tinterfacial cseparation

The heteIPCE valuesThe IPCE varelatively sm[129]. EIS silluminatioheterojunct[129]. Thesation of pnPEC performtransfer.

In summheterojunctgeneration.have been and WO3/B

suitable and matched band edge positions to form type II junctionat the interfaces. Central to the success of heterojunction structureis the semiconductors should have matched bandgap and well

band edge positions.

asmo

mon inter

plas desc

valeturalng fowavee andsitioible pectr

nanoal cooxideent rsplittre arr mctor tion to thructuht h

nanctronrmorng tsolarzers.of plso obnic m

ctingand ar l40] a

The tionld loicond

plaAs tprop

elec coud byas prndu

plan toeneoe platal Ss [1ructunic nd to all bandgap semiconductors to sensitize large bandgaps for water splitting.iconductor materials that form the heterojunction canplementarily doped to create a np junction. It has beenstrated in photovoltaic devices that the pn junctione photoexcited electrons and holes by internal electricd by band bending. Recently, Kim et al. extended thisphotoelectrode for PEC water splitting and designed aunction electrode with n-type TaON and p-type CaFe2O4ON was selected because it is highly active for waternder visible light illumination and it has a reasonablyd edge positions for water splitting. CaFe2O4 is a p-typeconductor with favorable bandgap of 1.9 eV and bandns for water splitting. However, the PEC performance2O4 is extremely low, due to the poor charge separationrt in CaFe2O4 [129]. Therefore, the formation of hetero-ld improve the charge separation and PEC performance. The heterojunction electrode of TaON/CaFe2O4 wasby electrophoretic deposition of TaON powder andwder onto FTO glass substrate, respectively [129]. Theion composite was further treated with TaCl5 solu-llowed by thermal annealing to ensure direct contactON and CaFe2O4. The energy diagram of the n-TaON/p-terojunction is illustrated in Fig. 11a. When light isrom back side of FTO, TaON absorbs the light with pho-above 2.5 eV, and the low energy light will be absorbedr of CaFe2O4. This tandem structure could improve thetilization. Furthermore, the internal electric eld at therface can facilitate charge separation. Fig. 11b shows thent densities of TaON and various type of TaON/CaFe2O4unction of potentials under visible light illuminationThe TaON electrode showed a tiny photocurrent den-/cm2 at 1.23 V vs. RHE. The small photocurrent waso the poor connection between TaON particles and con-strate. The TaCl5 treatment enhanced the connection

improved photocurrent. CaFe2O4 was deposited oneated TaON lm (denoted as: TaON-treated/CaFe2O4),ed a photocurrent of 0.72 mA/cm2 that is three times

that of TaON-treated electrode. Moreover, to ensureontact between TaON and CaFe2O4, TaON-as depositeds the base lm, and TaCl5 treatment was conductedO4 layer was deposited on the TaON-as deposited lm.

this electrode exhibited the highest photocurrent of2 at 1.23 V vs. Ag/AgCl, which is ve times higher thanreated electrode [129]. These results demonstrated theontact between TaON and CaFe2O4 is critical for chargeand transfer.rojunction electrode showed substantially increased

due to the improved charge separation, as expected.lues in the wavelength between 500 and 600 nm wasall due to the inefcient indirect transition of CaFe2O4

pectra collected at 0.2 V vs. Ag/AgCl under visible lightn veried that charge transfer is more efcient in theion electrode, compared to the TaON lms (Fig. 11d)e results have successfully demonstrated that the cre-

heterojunction in photoelectrodes can enhance theirance by improving both light absorption and charge

ary, we have reviewed the design of various types ofion electrodes and their performances for PEC hydrogen

Although various design of heterojunction structuresdeveloped such as ZnO/CdS, Fe2O3/BiVO4, Si/Fe2O3iVO4 [43,100,125127], it is always challenging to nd

aligned

2.3. Pl

Plasstrongsurfacecan betion ofthe narestorionant the sizcompois posssolar smonicchemicbon dithe recwater

Thetransfeconduabsorpogous nanostfor ligmonichot eleFurthechangientire sensititivity was alplasmocondumetal ing ne[139,1posed.interactric ea semexcitedelds. tor is rate ofductorinducetion hsemicoexcitedadditiohomogfor largthe mephotonnanostplasmoand lea[131].nic materials for PEC water splitting

ic metallic nanostructures such as Au, Ag and Cu exhibitaction with resonant photons through an excitation ofmon resonance (SPR) shown in Fig. 12a [130132]. SPRribed as the resonant photon induced collective oscilla-nce electrons, when the frequency of photons matches

frequency of surface electron oscillation against therce of positive nuclei. While the SPR intensity and res-length are material dependent, they can be tuned by

shape of metallic nanostructure. By manipulating then, shape and size of the plasmonic nanostructure, itto design nanostructure that interacts with the entireum [131]. Recently, it has been demonstrated that plas-structure could be applied in the eld of photon drivennversion such as photocatalytic water splitting and car-s reduction [133135]. In this section, we will reviewesearch progress in the surface plasmon-mediated PECing.e three non-mutually exclusive SPR enhanced energyechanisms for improving the performance of semi-photoelectrodes [131]. First, SPR can enhance lightof the electrode [136138]. The mechanism is anal-e dye or quantum dot sensitization, in which metalres anchor to a semiconductor and act as a sensitizerarvesting. As shown in Fig. 12b, the metallic plas-oparticles absorb resonant photons and transfer the

formed in the SPR excitation to semiconductor [131].e, the ability to tune the resonance wavelength byhe size or shape of nanostructure suggests that the

spectrum could be exploited using the plasmonic-metal Furthermore, SPR induced enhancement in photoac-asmonic metal nanoparticle modied semiconductorserved on the system where the semiconductor andetal were separated from each other with a thin non-

spacer that prevents direct charge transfer betweensemiconductor. Therefore, other mechanisms includ-ed electromagnetic induced electron-hole generationnd photon scattering mechanism [141,142] were pro-near eld electromagnetic mechanism is based on the

of the semiconductor with the strong SPR induced elec-calized nearby at the metallic nanostructure. Whenuctor is brought into the proximity of a photon-smonic nanostructure and encounters these intensehe rate of electronhole formation in a semiconduc-ortional to the local intensity of the electric eld, thetronhole formation in some regions of the semicon-ld be increased by signicant enhanced electric eld

SPR [143,144]. Recently, the electrodynamic simula-oved that the rate of electronhole formation in thectors was enhanced in the region of proximity of thesmonic nanostructure, as shown in Fig. 12c [131]. In

the local elds which play a role in the spatially non-us formation of electronhole pairs in semiconductors,smonic nanostructure (larger than 50 nm in diameter)PR is accompanied by an efcient scattering of resonant31,141,142]. The scattering of photons by plasmonicres could increase the average photon path length inanostructures and semiconductor composite electrode,an overall increased light utilization efciency (Fig. 12d)


Fig. 12. (a) No ) nano(b) Mechanism onducphotoexcited A s elecSchematic dia cally ephotons in the

Reproduced w

Plasmonmuch attenact with liincrease thparison to dhave excellvisible lighttrode by intcrystals (Ncomposite and conseqnanocrystalarrays and electrochemquently depa photocatacan be ratiophotonic balayer whichthe SPR abmodied Tition enhancnanoparticlNTPC showby controlli

More imment was photocurre1.23 V vs. R50 times higphotocurrermalized extinction spectra of Ag nanosphere (38 nm), Au (25 nm) and Cu (133 nm of SPR induced charge transfer from plasmonic metallic nanostructure to semicu nanoparticles, permeating into a neighboring TiO2 structure. The color bar show

gram shows the scattering mechanism by metal nanoparticles. The addition of opti composite structure.

ith permission from Ref. [131].ic metal nanostructure modied TiO2 have attractedtion because the metallic structures can strongly inter-ght in the visible and infrared region, which coulde visible light photoactivity of TiO2 [133,145]. In com-ye molecules or quantum dots, metallic nanostructuresent photostability. Recently, Zhang et al. developed a

responsive plasmonic photocatalytic composite elec-egrating Au nanocrystals and TiO2 nanotube photonicTPC) [99]. As shown in Fig. 13a, the design of theelectrode is expected to increase the Au SPR intensityuently boost the hot electron injection from the Aus into the conduction band of TiO2 [99]. TiO2 nanotubeTiO2 nanotube photonic crystals were developed byical anodization method. Au nanocrystals were subse-osited onto the TiO2 nanotube and photonic crystals bylytic reduction method. Importantly, the Au/TiO2 NTPCnally designed so that light with energy matching thendgap of TiO2 is trapped and localized in the photonic

reects and scatters the photon ux. Fig. 13b showssorption of Au modied TiO2 NT and NTPC lms. AuO2 NT and NTPC show a signicant visible light absorp-ement around 556 nm, due to the SPR absorption of Aues. Au nanocrystals with bigger size on the same TiO2ed a red shift of SPR peak, suggesting the SPR is tunableng the size of Au nanoparticles [99].portantly, a signicant photocurrent density enhance-observed for the Au modied TiO2 NTPC with ant density of 150 A/cm2 obtained at the potential ofHE under visible light illumination (>420 nm), which isher than the bare TiO2 NTPC (Fig. 13c). The visible lightnt response demonstrates the electron injection from

SPR metal nerably highmodied Ti(90 A/cm2

the photonnanocrystaanalyses shand Au(590the waveletion of Au photoactiviAu nanocryregion, whiscattering m

While tbeen well standing chelectrochemlimitation ftra of Cu anelectrochemof semicondthe enhancSPR effect iquantum doefcient SP

2.4. Surface

In orderenergy losparticles. The standard spectrum of AM 1.5 solar light is also shown.tor. (c) Optical simulations showing SPR-enhanced electric elds bytric eld intensity normalized by the light source intensity [135]. (d)xcited plasmonic nanoparticles increases the average path length ofanoparticle to TiO2. Moreover, the value is also consid-er than the photocurrent densities obtained for the AuO2 NTs (60 A/cm2) and the large size Au modied NTPC), suggesting the matching of the photonic bandgap ofic crystal structure with the SPR wavelength of the Auls can further improve the PEC performance [99]. IPCEowed that the Au(556)/TiO2 NT, Au(556)/TiO2 NTPC,)/TiO2 NTPC exhibited photoactivity enhancement atngths corresponding to the characteristic SPR absorp-nanocrystals. These results indicated that visible lightty improvement was due to surface plasmonic effect ofstals [99]. IPCE enhancement was also observed in UVch supports the near eld electromagnetic and photonechanism [99].

he concept of SPR-mediated PEC water splitting hasdemonstrated in the previous reports, there are out-allenges remain. For example, the photostability andical stability of the metal nanostructures are potential

or this approach. For example, the SPR absorption spec-d Ag can be easily manipulated, and Ag and Cu are notical stable and may easily be oxidized at the surfaceuctor photoanodes during water oxidation. Moreover,

ement for visible light utilization of semiconductors vias still relatively low compared to conventional dye ort sensitization. To develop photoelectrodes with highlyR induced photoactivity is still remains challenging.

electrochemical catalyst assisted PEC water splitting

to maximize efciency of PEC water splitting, theses during the process should be minimized. The


Fig. 13. (a) SP ce. (b)photonic cryst c) Amillumination o the rnanostructure

Reproduced w

energy los[2]:

Gloss =

Where energy lossface band bthe overpotpotential loin the previtures couldreduce the minimized trolyte, andto reduce thand cathod

ts (Oyed ts Co-

mose, Hrate

Fig. 14. (a) Topdensities in so

Reproduced wR charge carrier transfer under visible light irradiation at Au/TiO2nantoube interfaals (NTPC) and TiO2 nanotubes (NT), with the spectrum of TiO2 being subtracted. (f visible light (wavelength > 420 nm), with 60 s light on/off cycles. (d) IPCE plots ins.


ses can be expressed by the following equation

Gtransport + e(Ua + Uc + IR) [eV]

Gloss is the sum of the energy loss; Gtransport is the

catalysemplosuch aused toLikewito deco during electron/hole transportation through the sur-ending and the bulk of the semiconductor; Ua and Uc areential at the anode and cathode, respectively; IR is thess on the external circuit and the contacts. As discussedous sections, element doping and heterojunction struc-

facilitate charge separation and transport, which couldenergy loss of Gtransport. The potential loss of IR can beby increasing the conductivity of electrodes and elec-

decreasing the contact resistances in the device. Finally,e potential loss on the surface overpotential of anodee, electrochemical catalysts such as oxygen evolution

catalyst-momance. In tphotoelectr

Large ovof major limet al. reportwater oxidsurface of hThe thickne[31]. Fig. 1of bare hem

view SEM image of Co-Pi modied hematite lm. Co-Pi modication is achieved by electrolid line for hematite Fe2O3 (red) and Co-Pi/Fe2O3 (blue), collected at a scan rate of 50 mV

ith permission from Ref. [31]. SPR absorption spectra of the Au nanocrystals on the TiO2 nanotubeperometric It curves at an applied potential of 1.23 V vs. RHE underange of 400-700 nm at 1.23 V vs. RHE for TiO2 and Au modied TiO2

ER) and hydrogen evolution catalysts (HER) should beo modify the photoelectrodes. Various OER catalystsPi [31,45], IrO2 [95,146], and NiOx [35,73] have beendify photoanodes such as WO3, Fe2O3, BiVO4 and ZnO.ER catalysts such as Pt and MoS2 have also been used

photocathodes such as Si and Cu2O [93,147]. All these

died photoelectrodes showed increased PEC perfor-his section, we will review several catalyst-modiedode systems to exemplify the function of catalysts.erpotential for water oxidation is known to be oneitations for hematite photoelectrode. Recently, Zhong

ed Co-Pi catalyst modied hematite photoanode for PECation [31,91]. The Co-Pi catalyst was deposited on theematite lm via electrochemical deposition method.ss of Co-Pi catalyst layer is around 200 nm (Fig. 14a)4b compares the PEC water oxidation performanceatite and Co-Pi modied hematite photoelectrodes.

chemical deposition method. (b) Dark in dashed line and photocurrent/s under 1 sun AM 1.5G light illumination.


Fig. 15. (a) Da (thin electrolyte at p cay cuwith 30 min d rrent dpH 13.6 NaOH urren(red) are also i

Reproduced w

Clearly, thefor PEC watdensity of Cthe lower acatalyst motential of he

Zhong ettoactivity wAs shown inobserved upto steady stis much smat the sameother electrunstable phthey also obelectrode wand photocuIn contrastscan rate inbottleneck was believethick Co-Pidiffusion, anproductive the reductikinetic limiwith small showed lesscan rate dehematite phlar electrochsuch as KPi

Furthermof Ni(OH)2

t mourreurreropoate or. T

thet. Thrk current (dotted) and photocurrent (solid) densities of a Fe2O3 photoanode beforeH 8. Inset is the photocurrent density vs. time at 1.1 V vs. RHE. (b) Photocurrent de

eposition. (c) Top view SEM image of 15 min Co-Pi modied hematite. (d) Photocu (black) electrolytes collected on the hematite with 15 min Co-Pi deposition. Photocncluded for comparison.


presence of Co-Pi catalyst reduced the onset potentialer oxidation by >350 mV. Therefore, the photocurrento-Pi modied hematite was signicantly enhanced inpplied potentials [31]. These results proved that OERdication could suppress the water oxidation overpo-matite.

al. also found that the enhanced photocurrent or pho-as often accompanied by a kinetic limitation [148].

catalysphotocphotocThey pscan rbehavidue tocatalys Fig. 15a inset, a large initial spike of photocurrent wason illumination, followed by a rapid decay until reachate at a lower current density. The steady photocurrentaller than the measured photocurrent in the JV curves

potential. Similar currenttime curves were observed inolytes with different pH values (Fig. 15b), indicating theotocurrent is not related to the electrolyte. Moreover,served the JV curves of Co-Pi modied hematite photo-ere scan rate dependent [148]. Both the cathodic shiftsrrent densities decrease with the decrease of scan rate.

, the JV curves of bare hematite photoelectrode aredependent. These observations suggest that the kineticis likely associated with the Co-Pi modication, whichd to be due to the slow interfacial charge transfer. The

catalyst layer may hinder charge or proton transportd thus, restricting current ow and allowing other non-recombination pathways [148]. Zhong et al. found thaton of Co-Pi loading amount can partially address thetation. Fig. 15c shows the SEM image of hematite lmloading of Co-Pi catalyst. This hematite photoelectrodes cathodic shift and photocurrent enhancement. Thependence of JV curves is also not as signicant as theotoelectrode with high loading of Co-Pi catalyst. Simi-emical behavior was observed in different electrolytes

and NaOH (Fig. 15d) [148].ore, Wang et al. investigated the catalytic effecton hematite for PEC water splitting [35]. The Ni

on hematitof photocuadding holNi2+. By mithe photoccatalytic eff

HER catato suppressmonly usedRecently, oalso been deSi and InP foand low cothe solar co

3. Conclus

In this revarious typdoping, forplasmonic address thephotoanodetures with ldiffusion diby increasiand collectiblack) and after 30 min of Co-Pi deposition (thick colored) in 0.1 M KPirves in pH 8 KPi (blue) and pH 13.6 (black) collected on the hematiteecay curves in pH 8 KPi (blue), pH 8 buffered salt water (green) andt decaying curves measured for bare hematite in pH 8 KPi electrolyte

died hematite showed cathodic shift and enhancednt. Likewise, they also observed the gradual decay ofnt with time and the JV curves are scan rate dependent.sed an alternative explanation for the observation. Thedependent JV curves is a characteristic capacitivehe enhanced photocurrent current was found to be

photocharging effect (Ni2+ converted to Ni3+) of Niis photocharging effect eventually depleted the Ni2+e electrode surface, resulting in the gradual decayrrent. The enhanced photocurrent was stabilized bye scavenger such as glucose, which can regeneratenimizing the loading amount of Ni catalyst to suppressharging effect, they found that Ni has relatively smallect for water oxidation [35].lyst modied photocathodes have also been developed

the overpotential for water reduction. The most com- HER catalyst is platinum, while it is very expensive.ther low cost HER catalysts such as MoS2 and Ni haveveloped and coupled with p-type photocathode such asr water reduction [149151]. The discovery of efcientst OER and HER photocatalysts are critical to improvenversion to hydrogen efciency.

ion and outlook

view, we have highlighted the recent progress in usinges of chemical modication methods, including elementmation of heterojunction, surface modication withmetallic structures and electrochemical catalysts, to

existing limitations of commonly used nanostructureds for PEC water splitting. Semiconductor nanostruc-arge electrolyte accessible surface area and short carrierstance could potentially improve their PEC performanceng the efciencies of charge generation, separationon. Moreover, chemical modications can intrinsically


Fig. 16. (a) Schematic of the proposed articial photosynthesis water splitting device. The photoanode (top) is decorated with oxygen evolving catalysts and absorbs shortwavelength light. The photocathode is decorated with hydrogen evolving catalyst and absorbs the longer wavelength light. Reproduced with permission from Ref. [152]. (b)A dual bandgap n/p PEC band structure conguration with n- and p-type photoelectrodes. Reproduced with permission from Ref. [3].

change the trodes. For of semiconerties of seelectron acstructure cociency by caddition, into form tanof solar lighmetal nanoof semicondinjection, nscattering ecould also bface modiHER catalysoxidation aPEC water s

Althougimproved ttrodes, therequiremenband edge pTherefore, ductor strucombinatiotion to the ultimate goachieve wa

p n/d phoathoThe ve thf thetentn-pein blectred wal ba

for wuid ice ofnerangu

is deestino eleanc

cessthermial sen guld sater

riven a p t

Fig. 17. Generenergy diagram

Reproduced welectronic and optical properties of semiconductor elec-example, element doping could narrow the bandgapductor, and modulate the electronic and optical prop-miconductor, by controlled introduction of dopants asceptors and donors. The formation of heterojunctionuld improve the charge separation and collection ef-reating an internal electric eld at the junction. Integration of various types of semiconductor materialsdem structure can substantially increase the utilizationt. Moreover, electrode surface modied with plasmonicstructure could improve the light collection efciencyuctors by various mechanisms including hot electronear eld electromagnetic enhancement and photonffect. Plasmonic absorption of metallic nanostructurese tuned to cover the entire solar spectrum. Finally, sur-cation with electrochemical catalysts such as OER andts could suppress the surface overpotentials for waternd reduction, and thus, reducing the energy loss duringplitting.h chemically modication methods have substantiallyhe PEC performance of existing semiconductor elec-re is still not a single electrode can satisfy all thets for photoelectrodes in terms of bandgap energy,otentials, cost and photo- and electrochemical stability.there is a need to design and discover new semicon-ctures that can fulll all these requirements, via then of computational and experimental efforts. In addi-development new semiconductor photoelectrodes, theal is to couple the photoanode and photocathode to

bandgaode anphotocment. negatiband otion poa protoduced photoemodiThe ducarriertor/liqinterfacan gethis cociencyin harvthe twperformthe suc

Furmicrobhydrogtion cowastewsolar dsists ofter splitting at non-bias condition. Recently, a dual and Shewan

al working mechanism of a solar MPC. (a) Schematic diagram of a dual-chamber MPC with illustrates the carrier generation and charge transfer and reactions at the interface betw

ith permission from Ref. [153].p PEC device has been proposed by integrating photoan-tocathode, shown in Fig. 16 [152]. The photoanode andde materials are preferably to be earth-abundant ele-conduction band of the photocathode should be morean the water reduction potential, while the valence

photoanode is more positive than the water oxida-ial. The photoanode and photocathode are separated byrmeable membrane, which can separate the gas pro-

oth chambers and balance ion ux in between. Theseodes consists of nanostructures such as nanorod arraysith electrochemical OER and HER catalyst, respectively.ndgap n/p PEC utilizes electron and hole minority chargeater splitting reactions at their respective semiconduc-

nterfaces. The majority carriers will recombine at the photoanode and photocathode. Therefore two photonste one electronhole pair for water splitting [3]. Whileration can achieve non-bias PEC water splitting, its ef-termined by the performance of individual electrodeg solar light as well as the current matching betweenctrodes. Therefore, the continuous improvement in thee of both photoanode and photocathode is essential to

of achieving non-bias and efcient PEC water splitting.ore, hybrid device that combines PEC system and

ystem is emerging as a new research direction foreneration. Interfacing photocatalysis and bioremedia-imultaneously address the need of energy recovery and

treatment. Qian et al. have demonstrated a self-biased, microbial photoelectrochemical cell (MPC), which con-ype cuprous oxide photocathode as solar light absorber

ella oneidensis MR-1 colonized bio-anode [153]. Fig. 17

Cu2O photocathode and Shewanella bioanode. (b) The correspondingeen microbial and semiconductor system.


illustrates the dual chamber MPC device and the correspondingenergy diagram showing carrier generation and charge transfer atthe microbial and semiconductor system. It requires the potentialmatching between the oxidation potential of eletrogenic bacteriaand electroode. In comalone, substMPC devicegenic bactehydrogen gand solar liextended folimitations well as theFurther optand bio-ano

Acknowled

Y.L. ackn(DMR-0847ChancellorScholarship

References

[1] A. Fujishtor elec

[2] T. Bak, generatHydrog

[3] M.G. WN.S. Lew

[4] G.M. Wphotoel(2012) 6

[5] Y. Li, J.Zting bas

[6] R. Abewater sPhotoch

[7] A. KudoChem. S

[8] J.H. Parkaspect r

[9] N. Lu, XTiO2 nacapabili

[10] E.M. Rovisible l55456

[11] Y. Park,TiO2 phand the

[12] G.M. WWang, Jctroche

[13] J. Hensesensitizchemica

[14] H.Y. Watoelectrquantum

[15] G.M. Whydroge82158

[16] S. HoanphotoelNano Le

[17] S. Hoanlight phwith H2134 (20

[18] Y.J. HwTiO2 nadeposit

[19] M. Xu, P.M. Da, H.Y. Wu, D.Y. Zhao, G.F. Zheng, Controlled Sn-doping inTiO2 nanowire photoanodes with enhanced photoelectrochemical conver-sion, Nano Lett. 12 (2012) 15031508.

[20] Y.J. Hwang, A. Boukai, P.D. Yang, High density n-Si/n-TiO2 core/shell nanowirearrays with enhanced photoactivity, Nano Lett. 9 (2009) 410415.

. Cooent annder i360

.Y. Yan

. Li, Nilitting.M. Wantum

ydroge.C. Qiuoping ectrocShi, MterfacemicL. Xiehotoacrays f272.X. Liomposr pho023

.G. Linriven pater-s. Kay, anostr. Sivuematit.K. Zhmpos866

. Sivul. Gratrepare367

.J. Lin, anostr982

.M. Wang, ater o.M. Wudy inl wate.W. Gtivity. Francile phanc013) 5S. Li, Y

-Fehotoel. Sayambe, H. 2 on pfect o.H. Ng,iVO4 pys. Ch.C. Loropert855

Chatctalytight irr. Su, nction

(201. IwassponsMater.K. ZhcombAm. C. Su, L

iVO4 nectrocnic band structure of the semiconductor photocath-parison to the PEC cell and microbial electrolysis cellantially enhanced photocurrent was obtained from the

at zero bias. The synthetic culture medium for electro-ria could be replaced by wastewater, in order to achieveeneration with the only energy sources are waste waterght. Furthermore, this device concept can be furtherr other applications such as carbon xation. The currentfor the hybrid system are the efciency of bioanode, as

photostability and photovoltage of the photocathode.imization of the performances of both photocathodede is necessary.

gements

owledges the nancial support by United State NSF786). GMW thanks the nancial support of UCSCs Dissertation Year Fellowship. XHL thanks the China

Council for nancial support.

ima, K. Honda, Electrochemical photolysis of water at a semiconduc-trode, Nature 238 (1972) 3738.J. Nowotny, M. Rekas, C.C. Sorrell, Photo-electrochemical hydrogenion from water using solar energy. Materials-related aspects, Int. J.. Energy 27 (2002) 9911022.alter, E.L. Warren, J.R. McKone, S.W. Boettcher, Q.X. Mi, E.A. Santori,is, Solar water splitting cells, Chem. Rev. 110 (2010) 64466473.ang, Y.C. Ling, Y. Li, Oxygen-decient metal oxide nanostructures forectrochemical water oxidation and other applications, Nanoscale 46826691.

. Zhang, Hydrogen generation from photoelectrochemical water split-ed on nanomaterials, Laser Photon. Rev. 4 (2010) 517528., Recent progress on photocatalytic and photoelectrochemicalplitting under visible light irradiation, J. Photochem. Photobiol. C-em. Rev. 11 (2010) 179209., Y. Miseki, Heterogeneous photocatalyst materials for water splitting,oc. Rev. 38 (2009) 253278., S. Kim, A.J. Bard, Novel carbon-doped TiO2 nanotube arrays with highatios for efcient solar water splitting, Nano Lett. 6 (2006) 2428.. Quan, J.Y. Li, S. Chen, H.T. Yu, G.H. Chen, Fabrication of boron-dopednotube array electrode and investigation of its photoelectrochemicalty, J. Phys. Chem. C 111 (2007) 1183611842.ckafellow, L.K. Stewart, W.S. Jenks, Is sulfur-doped TiO2 an effectiveight photocatalyst for remediation? Appl. Catal. B-Environ. 91 (2009)2.

W. Kim, H. Park, T. Tachikawa, T. Majima, W. Choi, Carbon-dopedotocatalyst synthesized without using an external carbon precursor

visible light activity, Appl. Catal. B-Environ. 91 (2009) 355361.ang, H.Y. Wang, Y.C. Ling, Y.C. Tang, X.Y. Yang, R.C. Fitzmorris, C.C..Z. Zhang, Y. Li, Hydrogen-treated TiO2 nanowire arrays for photoele-mical water splitting, Nano Lett. 11 (2011) 30263033.l, G.M. Wang, Y. Li, J.Z. Zhang, Synergistic effect of CdSe quantum dotation and nitrogen doping of TiO2 nanostructures for photoelectro-l solar hydrogen generation, Nano Lett. 10 (2010) 478483.ng, G.M. Wang, Y.C. Ling, M. Lepert, C.C. Wang, J.Z. Zhang, Y. Li, Pho-ochemical study of oxygen decient TiO2 nanowire arrays with CdS

dot sensitization, Nanoscale 4 (2012) 14631466.ang, Y.C. Ling, X.H. Lu, H.Y. Wang, F. Qian, Y.X. Tong, Y. Li, Solar drivenn releasing from urea and human urine, Energy Environ. Sci. 5 (2012)219.g, S.W. Guo, N.T. Hahn, A.J. Bard, C.B. Mullins, Visible light drivenectrochemical water oxidation on nitrogen-modied TiO2 nanowires,tt. 12 (2012) 2632.g, S.P. Berglund, N.T. Hahn, A.J. Bard, C.B. Mullins, Enhancing visibleoto-oxidation of water with TiO2 nanowire arrays via cotreatmentand NH3: synergistic effects between Ti3+ and N, J. Am. Chem. Soc.

12) 36593662.ang, C. Hahn, B. Liu, P.D. Yang, Photoelectrochemical properties ofnowire arrays: a study of the dependence on length and atomic layerion coating, ACS Nano 6 (2012) 50605069.

[21] J.Kmu17

[22] XYsp

[23] Gquh

[24] Ydel

[25] J. Inch

[26] S.par14

[27] HCfo38

[28] Ydw

[29] An

[30] Kh

[31] Dco60

[32] KMp74

[33] Yn23

[34] GZhw

[35] Gstca

[36] Hac

[37] RFaen(2

[38] L.ofp

[39] KAOef

[40] YBPh

[41] Mp54

[42] P.cali

[43] J.Zju11

[44] AreJ.

[45] DreJ.

[46] J.ZBelper, Y.C. Ling, C. Longo, Y. Li, J.Z. Zhang, Effects of hydrogen treat-d air annealing on ultrafast charge carrier dynamics in ZnO nanowiresn situ photoelectrochemical conditions, J. Phys. Chem. C 116 (2012)17368.g, A. Wolcott, G.M. Wang, A. Sobo, R.C. Fitzmorris, F. Qian, J.Z. Zhang,trogen-doped ZnO nanowire arrays for photoelectrochemical water, Nano Lett. 9 (2009) 23312336.ang, X.Y. Yang, F. Qian, J.Z. Zhang, Y. Li, Double-sided CdS and CdSe

dot co-sensitized ZnO nanowire arrays for photoelectrochemicaln generation, Nano Lett. 10 (2010) 10881092., K.Y. Yan, H. Deng, S.H. Yang, Secondary branching and nitrogenof ZnO nanotetrapods: building a highly active network for photo-hemical water splitting, Nano Lett. 12 (2012) 407413..B. Starr, H. Xiang, Y. Hara, M.A. Anderson, J.H. Seo, Z.Q. Ma, X.D. Wang,e engineering by piezoelectric potential in ZnO-based photoelectro-al anode, Nano Lett. 11 (2011) 55875593., X.H. Lu, T. Zhai, W. Li, M.H. Yu, C.L. Liang, Y.X. Tong, Enhancedtivity and stability of carbon and nitrogen co-treated ZnO nanorodor photoelectrochemical water splitting, J. Mater. Chem. 22 (2012)14275., C.W. Cheng, X.L. Li, J.P. Liu, C. Guan, Y.Y. Tay, H.J. Fan,ition-graded ZnxCd1xSe@ZnO core-shell nanowire array electrodestoelectrochemical hydrogen generation, J. Phys. Chem. C 116 (2012)807., Y.K. Hsu, Y.C. Chen, L.C. Chen, S.Y. Chen, K.H. Chen, Visible-light-hotocatalytic carbon-doped porous ZnO nanoarchitectures for solarplitting, Nanoscale 4 (2012) 65156519.I. Cesar, M. Gratzel, New benchmark for water photooxidation byuctured -Fe2O3 lms, J. Am. Chem. Soc. 128 (2006) 1571415721.la, F. Le Formal, M. Gratzel, Solar water splitting: progress usinge (-Fe2O3) photoelectrodes, ChemSusChem 4 (2011) 432449.ong, J.W. Sun, H. Inumaru, D.R. Gamelin, Solar water oxidation byite catalyst/-Fe2O3 photoanodes, J. Am. Chem. Soc. 131 (2009)087.a, R. Zboril, F. Le Formal, R. Robert, A. Weidenkaff, J. Tucek, J. Frydrych,zel, Photoelectrochemical water splitting with mesoporous hematited by a solution-based colloidal approach, J. Am. Chem. Soc. 132 (2010)444.S. Zhou, S.W. Sheehan, D.W. Wang, Nanonet-based hematite hetero-uctures for efcient solar water splitting, J. Am. Chem. Soc. 133 (2011)401.ang, Y.C. Ling, D.A. Wheeler, K.E.N. George, K. Horsley, C. Heske, J.Z.Y. Li, Facile synthesis of highly photoactive -Fe2O3-based lms forxidation, Nano Lett. 11 (2011) 35033509.ang, Y.C. Ling, X.H. Lu, T. Zhai, F. Qian, Y.X. Tong, Y. Li, A mechanisticto the catalytic effect of Ni(OH)2 on hematite for photoelectrochemi-r oxidation, Nanoscale 5 (2013) 41294133.ao, C. Liu, H.E. Jeong, P.D. Yang, Plasmon-enhanced photocatalytic

of iron oxide on gold nanopillars, ACS Nano 6 (2012) 234240.king, L.S. Li, M.A. Lukowski, F. Meng, Y.Z. Tan, R.J. Hamers, S. Jin,ost-growth doping of nanostructured hematite photoanodes fored photoelectrochemical water oxidation, Energy Environ. Sci. 600512..H. Yu, F. Meng, Y.Z. Tan, R.J. Hamers, S. Jin, Facile solution synthesisF33H2O nanowires and their conversion to -Fe2O3 nanowires forectrochemical application, Nano Lett. 12 (2012) 724731.a, A. Nomura, T. Arai, T. Sugita, R. Abe, M. Yanagida, T. Oi, Y. Iwasaki, Y.

Sugihara, Photoelectrochemical decomposition of water into H2 andorous BiVO4 thin-lm electrodes under visible light and signicantf Ag ion treatment, J. Phys. Chem. B 110 (2006) 1135211360.

A. Iwase, A. Kudo, R. Amal, Reducing graphene oxide on a visible-lighthotocatalyst for an enhanced photoelectrochemical water splitting, J.em. Lett. 1 (2010) 26072612.ng, W.M. Cai, H. Kisch, Visible light induced photoelectrochemicalies of n-BiVO4 and n-BiVO4/p-Co3O4, J. Phys. Chem. C 112 (2008)4.hai, Y. Murakami, S.Y. Kishioka, A.Y. Nosaka, Y. Nosaka, Efcient photo-c activity of water oxidation over WO3/BiVO4 composite under visibleadiation, Electrochim. Acta 54 (2009) 11471152.L.J. Guo, N.Z. Bao, C.A. Grimes, Nanostructured WO3/BiVO4 Hetero-

lms for efcient photoelectrochemical water splitting, Nano Lett.1) 19281933.e, A. Kudo, Photoelectrochemical water splitting using visible-light-ive BiVO4 ne particles prepared in an aqueous acetic acid solution,. Chem. 20 (2010) 75367542.ong, S. Choi, D.R. Gamelin, Near-complete suppression of surfaceination in solar photoelectrolysis by Co-Pi catalyst-modied W:BiVO4,hem. Soc. 133 (2011) 1837018377..J. Guo, S. Yoriya, C.A. Grimes, Aqueous growth of pyramidal-shapedanowire arrays and structural characterization: application to photo-hemical water splitting, Cryst. Growth Des. 10 (2010) 856861.


[47] S.K. Pilli, T.E. Furtak, L.D. Brown, T.G. Deutsch, J.A. Turner, A.M. Herring, Cobalt-phosphate (Co-Pi) catalyst modied Mo-doped BiVO4 photoelectrodes forsolar water oxidation, Energy Environ. Sci. 4 (2011) 50285034.

[48] S.P. Berglund, D.W. Flaherty, N.T. Hahn, A.J. Bard, C.B. Mullins, Photoelectro-chemical oxidation of water using nanostructured BiVO4 lms, J. Phys. Chem.C 115 (2

[49] H. Ye, Jlysts byphotoel

[50] Y.Q. Liatum ef17594

[51] W.J. LuoYu, Z.G.photoan

[52] F.F. AbdCo-Pi-ca

[53] S.P. BerW into oxidatio

[54] G.M. WHydrogEnviron

[55] X.H. ZhaWei, Y.Xnanostrlight dr

[56] X. Liu, photoel78947

[57] D.D. QinT.E. Malanodes 72973

[58] D.V. EspK.H. Leecose forEnviron

[59] V. CristiphotoelLangmu

[60] J.C. Hill,type W116 (20

[61] C.X. KroAntounhetero-

[62] X.J. Fengnanotub94895

[63] Y.Q. CoBard, Sfor pho14541

[64] Y. Yangnitride asubstra

[65] Y.B. Li, Ttically awater s

[66] J.W. Sungrowthhydroge19306

[67] C. Liu, Jnanowi12 (201

[68] K. MaedsplittingSoc. 133

[69] S. Baneby sonofor phot(2009) 7

[70] M. HigaTaON p134 (20

[71] M. HigatoanodeSci. 4 (2

[72] Y.W. Chtyre, Atwater o

[73] K. Sun, NS.H. Jin,photo-o

[74] N.C. Strandwitz, D.J. Comstock, R.L. Grimm, A.C. Nichols-Nielander, J. Elam,N.S. Lewis, Photoelectrochemical behavior of n-type Si(100) electrodes coatedwith thin lms of manganese oxide grown by atomic layer deposition, J. Phys.Chem. C 117 (2013) 49314936.

[75] J.H. Oh, T.G. Deutsch, H.C. Yuan, H.M. Branz, Nanoporous black silicon photo-thodenvironOh, J. om sil.W. Jin.W. HuuctionemonsP. Liuu/TiO2erform.K. Laromogeblique.A. Whnthesl prop. Zha

hotoelP. Liuissio

008) .M. Wanocry. Irie, hotoca4835U.M. Kng by .B. Chtalysi011) . Serpoping hem. B.D. Ch

n hemPhys. Cesaranodefect oCesaricknehotoan.K. Zhsistedhotoan7591.D. Ho.M. Se. Norsilicon3443. WanT. Lee,lls thr011) . Barrratzelodie.S.A. 1

Badiaimeneith an.M. Ch.L. Tsefects fproacLiu, Fu nan5609. Shi, Knerate pho.H. Zhanocryelectrano LeJ. Honhanc

7811011) 37943802.. Lee, J.S. Jang, A.J. Bard, Rapid screening of BiVO4-based photocata-

scanning electrochemical microscopy (SECM) and studies of theirectrochemical properties, J. Phys. Chem. C 114 (2010) 1332213328.ng, T. Tsubota, L.P.A. Mooij, R. van de Krol, Highly improved quan-ciencies for thin lm BiVO4 photoanodes, J. Phys. Chem. C 115 (2011)17598., Z.S. Yang, Z.S. Li, J.Y. Zhang, J.G. Liu, Z.Y. Zhao, Z.Q. Wang, S.C. Yan, T.

Zou, Solar hydrogen generation from seawater with a modied BiVO4ode, Energy Environ. Sci. 4 (2011) 40464051.i, R. van de Krol, Nature and light dependence of bulk recombination intalyzed BiVO4 photoanodes, J. Phys. Chem. C 116 (2012) 93989404.glund, A.J.E. Rettie, S. Hoang, C.B. Mullins, Incorporation of Mo andnanostructured BiVO4 lms for efcient photoelectrochemical watern, Phys. Chem. Chem. Phys. 14 (2012) 70657075.ang, Y.C. Ling, H.Y. Wang, X.Y. Yang, C.C. Wang, J.Z. Zhang, Y. Li,en-treated WO3 nanoakes show enhanced photostability, Energy. Sci. 5 (2012) 61806187.ng, X.H. Lu, Y.Q. Shen, J.B. Han, L.Y. Yuan, L. Gong, Z. Xu, X.D. Bai, M.. Tong, Y.H. Gao, J. Chen, J. Zhou, Z.L. Wang, Three-dimensional WO3uctures on carbon paper: photoelectrochemical property and visibleiven photocatalysis, Chem. Commun. 47 (2011) 58045806.F.Y. Wang, Q. Wang, Nanostructure-based WO3 photoanodes forectrochemical water splitting, Phys. Chem. Chem. Phys. 14 (2012)911., C.L. Tao, S.A. Friesen, T.H. Wang, O.K. Varghese, N.Z. Bao, Z.Y. Yang,louk, C.A. Grimes, Dense layers of vertically oriented WO3 crystals asfor photoelectrochemical water oxidation, Chem. Commun. 48 (2012)1.osito, R.V. Forest, Y.C. Chang, N. Gaillard, B.E. McCandless, S.Y. Hou,, R.W. Birkmire, J.G.G. Chen, Photoelectrochemical reforming of glu-

hydrogen production using a WO3-based tandem cell device, Energy. Sci. 5 (2012) 90919099.no, S. Caramori, R. Argazzi, L. Meda, G.L. Marra, C.A. Bignozzi, Efcientectrochemical water splitting by anodically grown WO3 electrodes,ir 27 (2011) 72767284.

K.S. Choi, Effect of electrolytes on the selectivity and stability of n-O3 photoelectrodes for use in solar water oxidation, J. Phys. Chem. C12) 76127620.nawitter, L. Vayssieres, S.H. Shen, L.J. Guo, D.A. Wheeler, J.Z. Zhang, B.R., S.S. Mao, A perspective on solar-driven water splitting with all-oxidenanostructures, Energy Environ. Sci. 4 (2011) 38893899., T.J. LaTempa, J.I. Basham, G.K. Mor, O.K. Varghese, C.A. Grimes, Ta3N5e arrays for visible light water photoelectrolysis, Nano Lett. 10 (2010)2.ng, H.S. Park, S.J. Wang, H.X. Dang, F.R.F. Fan, C.B. Mullins, A.J.ynthesis of Ta3N5 nanotube arrays modied with electrocatalyststoelectrochemical water oxidation, J. Phys. Chem. C 116 (2012)14550., Y.C. Ling, G.M. Wang, X.H. Lu, Y.X. Tong, Y. Li, Growth of galliumnd indium nitride nanowires on conductive and exible carbon cloth

tes, Nanoscale 5 (2013) 18201824.. Takata, D. Cha, K. Takanabe, T. Minegishi, J. Kubota, K. Domen, Ver-ligned Ta3N5 nanorod arrays for solar-driven photoelectrochemicalplitting, Adv. Mater. 25 (2013) 125131., C. Liu, P.D. Yang, Surfactant-free, large-scale, solution-liquid-solid

of gallium phosphide nanowires and their use for visible-light-drivenn production from water reduction, J. Am. Chem. Soc. 133 (2011)19309..W. Sun, J.Y. Tang, P.D. Yang, Zn-doped p-type gallium phosphidere photocathodes from a surfactant-free solution synthesis, Nano Lett.2) 54075411.a, M. Higashi, B. Siritanaratkul, R. Abe, K. Domen, SrNbO2N as a water-

photoanode with a wide visible-light absorption band, J. Am. Chem. (2011) 1233412337.rjee, S.K. Mohapatra, M. Misra, Synthesis of TaON nanotube arrayselectrochemical anodization followed by nitridation: a novel catalystoelectrochemical hydrogen generation from water, Chem. Commun.1377139.shi, K. Domen, R. Abe, Highly stable water splitting on oxynitridehotoanode system under visible light irradiation, J. Am. Chem. Soc.12) 69686971.shi, K. Domen, R. Abe, Fabrication of efcient TaON and Ta3N5 pho-s for water splitting under visible light irradiation, Energy Environ.011) 41384147.en, J.D. Prange, S. Duhnen, Y. Park, M. Gunji, C.E.D. Chidsey, P.C. McIn-omic layer-deposited tunnel oxide stabilizes silicon photoanodes forxidation, Nat. Mater. 10 (2011) 539544.. Park, Z.L. Sun, J.G. Zhou, J. Wang, X.L. Pang, S.H. Shen, S.Y. Noh, Y. Jing,

P.K.L. Yu, D.L. Wang, Nickel oxide functionalized silicon for efcientxidation of water, Energy Environ. Sci. 5 (2012) 78727877.

caE

[76] I. fr

[77] DXdd

[78] L.Ap

[79] Gho

[80] Dsyca

[81] J.Zp

[82] L.em(2

[83] Gn

[84] Hp5

[85] S.ti

[86] Xca(2

[87] NdC

[88] WoJ.

[89] I. toef

[90] I. thp

[91] Dasp1

[92] YAJ.Ka 4

[93] XS.ce(2

[94] MGmU

[95] L.Gw

[96] HMefap

[97] J. C9

[98] Xgeid

[99] ZntoN

[100] S.en1 for H2 production by photoelectrochemical water splitting, Energy. Sci. 4 (2011) 16901694.Kye, S. Hwang, Enhanced photoelectrochemical hydrogen productionicon nanowire array photocathode, Nano Lett. 12 (2012) 298302.g, L.J. Guo, L.A. Zhao, X.M. Zhang, H.A. Liu, M.T. Li, S.H. Shen, G.J. Liu,, X.H. Zhang, K. Zhang, L.J. Ma, P.H. Guo, Efcient solar hydrogen pro-

by photocatalytic water splitting: From fundamental study to pilottration, Int. J. Hydrog. Energy 35 (2010) 70877097., G.M. Wang, Y. Li, Y.D. Li, J.Z. Zhang, CdSe quantum dot-sensitizedhybrid mesoporous lms and their enhanced photoelectrochemicalance, Nano Res. 4 (2011) 249258.sen, B.C. Fitzmorris, C. Longo, J.Z. Zhang, Y.P. Zhao, Nanostructurednous CdSe-TiO2 composite visible light photoanodes fabricated by

angle code position, J. Mater. Chem. 22 (2012) 1420514218.eeler, G.M. Wang, Y.C. Ling, Y. Li, J.Z. Zhang, Nanostructured hematite:is, characterization, charge carrier dynamics, and photoelectrochemi-erties, Energy Environ. Sci. 5 (2012) 66826702.ng, Metal oxide nanomaterials for solar hydrogen generation fromectrochemical water splitting, MRS Bull. 36 (2011) 4855., Q. Peng, Y.D. Li, Preparation of CdSe quantum dots with full colorn based on a room temperature injection technique, Inorg. Chem. 4750225028.ang, Y. Li, Nickel catalyst boosts solar hydrogen generation of CdSestals, ChemCatChem 5 (2013) 12941295.Y. Watanabe, K. Hashimoto, Nitrogen-concentration dependence

Documents

Chemically Modified Nanostructures for Photoelectrochemical Water