sdarticle

Surface Science Reports 63 (2008) 515–582

Contents lists available at ScienceDirect

Surface Science Reports

journal homepage: www.elsevier.com/locate/surfrep

TiO2 photocatalysis and related surface phenomenaAkira Fujishima a,∗, Xintong Zhang b, Donald A. Tryk ca Kanagawa Academy of Science and Technology, 3-2-1 Sakado, Takatsu-ku, Kawasaki-shi, Kanagawa 213-0012, Japanb Center for Advanced Optoelectronic Functional Materials Research, Northeast Normal University, 5268 Renmin Street, Changchun 130024, Chinac Fuel Cell Nanomaterials Center, University of Yamanashi, Takeda 4-3-11, Koufu, Yamanashi 400-8510, Japan

a r t i c l e i n f o a b s t r a c t

Article history:Accepted 1 October 2008editor: Y. Murata

Keywords:Titanium dioxideTitaniaTiO2Self-cleaning surfacesSuperhydrophilic effectAnion dopingWater splittingEnvironmental cleaning

The field of photocatalysis can be traced back more than 80 years to early observations of the chalking oftitania-based paints and to studies of the darkening of metal oxides in contact with organic compoundsin sunlight. During the past 20 years, it has become an extremely well researched field due to practicalinterest in air andwater remediation, self-cleaning surfaces, and self-sterilizing surfaces. During the sameperiod, there has also been a strong effort to use photocatalysis for light-assisted production of hydrogen.The fundamental aspects of photocatalysis on the most studied photocatalyst, titania, are still beingactively researched and have recently become quite well understood. The mechanisms by which certaintypes of organic compounds are decomposed completely to carbon dioxide and water, for example,have been delineated. However, certain aspects, such as the photo-induced wetting phenomenon,remain controversial, with some groups maintaining that the effect is a simple one in which organiccontaminants are decomposed, while other groups maintain that there are additional effects in whichthe intrinsic surface properties are modified by light. During the past several years, powerful tools suchas surface spectroscopic techniques and scanning probe techniques performed on single crystals in ultra-high vacuum, and ultrafast pulsed laser spectroscopic techniques have been brought to bear on theseproblems, and new insights have become possible. Quantum chemical calculations have also providednew insights. Newmaterials have recently been developed based on titania, and the sensitivity to visiblelight has improved. The new information available is staggering, but we hope to offer an overview ofsome of the recent highlights, as well as to review some of the origins and indicate some possible newdirections.

© 2008 Elsevier B.V. All rights reserved.

Contents

1. Introduction........................................................................................................................................................................................................................5162. Historical overview ............................................................................................................................................................................................................5163. Properties of TiO2 materials ..............................................................................................................................................................................................519

3.1. Crystal structures...................................................................................................................................................................................................5193.2. Electronic properties .............................................................................................................................................................................................5203.3. Surface structure studies.......................................................................................................................................................................................5233.4. Surface chemical studies: Interactions with water .............................................................................................................................................5233.5. Surface chemical studies: Interactions with dioxygen and other species .........................................................................................................5273.6. Bulk chemistry—Hydrogen....................................................................................................................................................................................5273.7. Electrochemical properties ...................................................................................................................................................................................5293.8. Photoelectrochemical properties..........................................................................................................................................................................534

4. Fundamentals of photocatalysis........................................................................................................................................................................................5344.1. Mechanisms of photocatalysis ..............................................................................................................................................................................534

4.1.1. Photoelectrochemical basis of photocatalysis ......................................................................................................................................5344.1.2. Time scales ..............................................................................................................................................................................................5384.1.3. Trapping of electrons and holes.............................................................................................................................................................541

∗ Corresponding author. Tel.: +81 (0)44 819 2020; fax: +81 (0)44 819 2038.E-mail address: [email protected] (A. Fujishima).

0167-5729/$ – see front matter© 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.surfrep.2008.10.001

http://www.elsevier.com/locate/surfrep

http://www.elsevier.com/locate/surfrep

mailto:[email protected]

http://dx.doi.org/10.1016/j.surfrep.2008.10.001

516 A. Fujishima et al. / Surface Science Reports 63 (2008) 515–582

4.1.4. Oxidizing species at the TiO2 surface ....................................................................................................................................................5424.1.5. Role of molecular oxygen.......................................................................................................................................................................5454.1.6. Effect of crystal face................................................................................................................................................................................5474.1.7. Remote photocatalysis ...........................................................................................................................................................................549

4.2. Photocatalytic reactions ........................................................................................................................................................................................5514.2.1. Decomposition of gaseous pollutants ...................................................................................................................................................5514.2.2. Decomposition of aqueous pollutants...................................................................................................................................................5524.2.3. Decomposition of liquid and solid films ...............................................................................................................................................5534.2.4. Photocatalytic sterilization ....................................................................................................................................................................553

4.3. Visible-light-induced photocatalysis....................................................................................................................................................................5554.3.1. Non-metal doping...................................................................................................................................................................................5554.3.2. Origin of visible light photoactivity.......................................................................................................................................................5564.3.3. Activity and stability of N-doped TiO2 photocatalysts .........................................................................................................................558

5. Fundamentals of the photo-induced hydrophilic (PIH) effect ........................................................................................................................................5605.1. Overview ................................................................................................................................................................................................................5605.2. Mechanisms of the PIH effect................................................................................................................................................................................562

5.2.1. Decomposition of organic films .............................................................................................................................................................5625.2.2. Reductive mechanism ............................................................................................................................................................................5635.2.3. Oxidative mechanism.............................................................................................................................................................................5635.2.4. Combined redox mechanism .................................................................................................................................................................5635.2.5. Visible-light-induced PIH effect.............................................................................................................................................................564

6. Brief review of applications...............................................................................................................................................................................................5656.1. Self-cleaning surfaces ............................................................................................................................................................................................5656.2. Water purification .................................................................................................................................................................................................5676.3. Air purification .......................................................................................................................................................................................................5686.4. Self-sterilizing surfaces .........................................................................................................................................................................................5696.5. Anti-fogging surfaces.............................................................................................................................................................................................5706.6. Heat transfer and heat dissipation........................................................................................................................................................................5716.7. Anticorrosion applications ....................................................................................................................................................................................5716.8. Environmentally friendly surface treatment .......................................................................................................................................................5726.9. Photocatalytic lithography ....................................................................................................................................................................................5726.10. Photochromism......................................................................................................................................................................................................5746.11. Microchemical systems .........................................................................................................................................................................................574

7. Summary ............................................................................................................................................................................................................................575Appendix. TiO2 film preparation methods ..................................................................................................................................................................576References...........................................................................................................................................................................................................................576

1. Introduction

Photocatalysis is generally thought of as the catalysis of aphotochemical reaction at a solid surface, usually a semiconductor[1–16]. This simple definition, while correct and useful, however,conceals the fact that theremust be at least two reactions occurringsimultaneously, the first involving oxidation, fromphotogeneratedholes, and the second involving reduction, from photogeneratedelectrons. Both processes must be balanced precisely in order forthe photocatalyst itself not to undergo change, which is, after all,one of the basic requirements for a catalyst.It will be seen in this review of the fundamentals and selected

applications of photocatalysis, principally on titanium dioxide,that there is a host of possible photochemical, chemical andelectrochemical reactions that can occur on the photocatalystsurface. The types of reactions occurring, their extent and theirrates depend upon a host of factors that are still in the processof being unraveled. Furthermore, there can indeed be changesthat occur, involving the surface and bulk structure and evendecomposition of the photocatalyst, a fact that appears to stretchthe definition of the term.This topic started its early history asmostly a nuisance involving

the chalking of titania-based paints [17,18] and then graduallytransformed into a highly useful approach to the remediation ofwater and air and then into an approach tomaintain surfaces cleanand sterile. Along theway, it has also transformed into an approachto photolytically split water into hydrogen and oxygen [19–21]and also an approach to perform selective oxidation reactions inorganic chemistry [22].Clearly, with so many varied aspects, photocatalysis is nearly

impossible to review comprehensively. In the present review, we

have tried to put together an overview of some of the morefundamental aspects, which are in their own right extremelyscientifically interesting and which also need to be betterunderstood in order tomake significant progresswith applications.The review will be divided into several sections: 2. Historical

overview; 3. Properties of TiO2 materials; 4. Fundamentals ofphotocatalysis; 5. Fundamentals of the photo-induced hydrophiliceffect; 6. Brief review of applications; 7. Summary, and Appendix(film preparation methods).

2. Historical overview

We will give a brief overview of the early history ofphotocatalysis, which will be based just on papers that we havebeen able to access, which means that we will almost certainlybe ignoring some important papers. The earliest work that wehave been able to find is that of Renz, at the University of Lugano(Switzerland), who reported in 1921 [17] that titania is partiallyreduced during illumination with sunlight in the presence of anorganic compound such as glycerol, the oxide turning from whiteto a dark color, such as grey, blue or even black; he also foundsimilar phenomena with CeO2,Nb2O5 and Ta2O5. For TiO2, thereaction proposed was:

TiO2 + light→ Ti2O3 or TiO. (2.1)

Baur and Perret, at the Swiss Federal Institute of Technology, werethe first to report, in 1924, the photocatalytic deposition of a silversalt on zinc oxide to produce metallic silver [23]. Even at thisearly date, the authors suspected that both oxidation and reduction

A. Fujishima et al. / Surface Science Reports 63 (2008) 515–582 517

were occurring simultaneously. The reaction pathway proposedwas this:

ZnO+ hν → h+ + e− (2.2)

h+ + OH− →14O2 +

12H2O (2.3)

e− + Ag+ → Ag0. (2.4)

Three years later, Baur and Neuweiler proposed simultaneousoxidation and reduction to explain the production of hydrogenperoxide on zinc oxide [24].

2h+ + CH2O+ 2OH− → CO+ 2H2O (2.5)

2e− + 2H+ + O2 → H2O2. (2.6)

It was not until many years later that this was absolutelyconfirmed, however. In 1932, Renz reported the photocatalyticreduction of silver nitrate to metallic silver and gold chloride tometallic gold on anumber of illuminated oxides, including TiO2 andNb2O5, [25] and discussed the results in terms of the Baur redoxmechanism.It has been recognized for quite a long time that titania-based

exterior paints tend to undergo ‘‘chalking’’ in strong sunlight.This means that a non-adherent, white powdery substance tendsto form on the surface, similar to the chalk on a blackboard.This effect was recognized to result from the actual removal ofpart of the organic component of the paint, leaving the titaniaitself exposed. With this background, Goodeve and Kitchener, atUniversity College, London, carried out an excellent study on thephotocatalytic decomposition of a dye on titania powder in air in1938, including absorption spectra and determination of quantumyields (Fig. 2.1) [26]. These authors proposed that titania acts as acatalyst to accelerate the photochemical oxidation and also studieda number of other oxides and speculated on the precisemechanism[27]. In 1949, Jacobsen, at the National Lead Company (USA), alsoattempted to explain the paint chalking phenomenon in terms ofa redox mechanism. He found a correlation between the tendencyof a number of different titania powders to undergo photo-inducedreduction in the presence of organic compounds to their chalkingtendency [18]. The photo-induced reduction was measured as aloss of reflectivity, due to the discoloration of the powder uponreduction, presumably to various oxygen-deficient forms, all theway to Ti2O3. The author proposed a cyclic redox process in whichthe titania was reduced while the organic paint components wereoxidized, followed by re-oxidation of the titania by oxygen fromthe air. The changes experienced by the titania were recognized tobe completely reversible, while those experienced by the organicpaint were recognized to be irreversible, leading to the formationof water-soluble organic acids and CO2. Even though Jacobsen wasapparently unaware of the work of Baur on the redox mechanism,he referred to the 1921 paper of Renz on the photo-reduction ofmetal oxides and proposed the same basic mechanism that hadbeen proposed by Baur; thus, a foundation was laid for later workon the redox mechanism.During the 1950s, the development of photocatalysis shifted to

zinc oxide. In 1953, two studies appeared in which the puzzlingphenomenon of hydrogen peroxide production on zinc oxideilluminated with UV light was studied [28,29], followed by a seriesof follow-up studies in ensuing years [30–34]. In these studies, theoverall reactions and mechanisms were completely clarified, andit became apparent that an organic compound was oxidized whileatmospheric oxygen was reduced. Even in the earliest study, anoverall reaction with phenol to produce catechol was proposed,and the involvement of radical species such as the hydroxyl radical(•OH) was also speculated upon [28]. Thus, the original proposal

Fig. 2.1. Original data of Goodeve and Kitchener showing the photocatalyticdecomposition of a dye (‘‘chlorazol sky blue)’’ adsorbed on anatase powder underUV illumination at 365 nm [26].© 1938, Royal Society of Chemistry.

of Baur and Neuweiler was finally confirmed, with the overallreaction:

RHOH+ H2O+ O2 → H2O2 + R(OH)2. (2.7)

Markham, first at the Catholic University of America and laterat St. Joseph’s College (USA), continued to study photocatalyticreactions on ZnO, and her papers constitute an impressive, yetunderappreciated, body of work [28,30,31,35–39]. This workculminated in a highly intriguing study in which Markham andco-worker Upreti constructed and studied a number of differenttypes of photo-assisted fuel cells, using illuminated ZnO as thephoto-anode with formamide or several alcohols as the organicsubstrates [39]. At the dark cathode (platinum), several differentredox mediators were examined, with atmospheric oxygenultimately being the electron acceptor. The authorsmay have beendiscouraged by the inevitable problem of ZnO photocorrosion,which prevented this system from reaching practical application.It was not until years later that the same basic ideas were re-examined with TiO2. Unfortunately, Markham and Laidler, in theirinitial study in 1953, examined TiO2 but subsequently abandonedit, since it did not produce measurable amounts of hydrogenperoxide [28].It is also interesting to note that Stephens et al. (Wayne

State University), in their study in 1955 of hydrogen peroxideproduction on a large assortment of illuminated semiconductors,but, unfortunately, not TiO2, remarked that ‘‘zinc oxide and theother catalytic solids should not be abandoned as devices forcapturing solar energy in a form capable of transfer to somechemical system’’ [32]. These authors found that CdS was themostactive photocatalyst, exceeding ZnO in activity.In a study reported in 1956 in Nature, Hindson and Kelly

(Defense Standards Laboratory) reported on the effects of variousrot-inhibitors on tent fabrics for use in Australia. They examinedthe effects of fabric strength after one year of exposure to sunlight.They stated: ‘‘The effect of anatase is startling. Fabrics containing3% of this pigment lost 90% in strength’’.In 1958, Kennedy et al., at the University of Edinburgh, studied

the photo-adsorption of O2 on TiO2 in order to try to more fullyunderstand the photocatalytic process [40]. They concluded thatelectrons were transferred to O2 as a result of photoexcitation, andthe resulting reduced form of O2 adsorbed on the TiO2 surface.These authors found a correlation between the ability of the TiO2sample to photocatalytically decompose chlorazol sky blue (thesame dye used earlier by Goodeve and Kitchener) and the ability


to photo-adsorb O2. This phenomenon is certainly important forphotocatalysis and will be commented upon later.During this period, researchers in Russia were active. Photo-

adsorption of O2 on illuminated ZnO was studied by Terenin andSolonitzin at the University of Leningrad (now University of St.Petersburg) [41]. In a very interesting early work, Filimonov, atthe same institution, compared the photocatalytic oxidation ofisopropanol to acetone on ZnO and TiO2 [42] and concluded thatthemechanism on TiO2 involved an overall reduction of O2 to H2O,while the reduction of O2 on ZnO onlywent as far as H2O2. On TiO2,the surface reactions were proposed to be:

TiO∗2 + (CH3)2CHOH→ (CH3)2CO+ TiO+ H2O (2.8)

TiO+12O2 → TiO2. (2.9)

Thus, this mechanism is a more detailed version of the Baur cyclicmechanism. It involves the removal of a surface lattice oxygenatom,whichwould be a kind of reduction process. Thismechanismwill be discussed later also, in Section 3.3.In Japan, at the Kyoto Institute of Technology, an early study

(1964) by Kato and Mashio also found that various types oftitania powders had different photocatalytic activities, specificallyto oxidize hydrocarbons and alcohols, simultaneously producinghydrogen peroxide [43]. Interestingly, these authors found thatanatase powders were more active than the rutile ones.In further work at the University of Edinburgh, McLintock and

Ritchie, using gas-phase adsorption measurements, studied thephotocatalytic oxidation of ethylene and propylene at TiO2 [44].This study is one of the first that we have found that shows thatit is possible to oxidize organic compounds completely to CO2 andH2O:

C2H4 + 3O2 → 2CO2 + 2H2O. (2.10)

The mechanism was proposed to involve the production ofsuperoxide from oxygen:

O2 + e− → O•−2 . (2.11)

Markham et al. had already proposed this reaction to take placeon illuminated ZnO [31]. Work on similar photo-reactions hascontinued into more recent years [45].In an important study for the relationship between photoelec-

trochemistry and photocatalysis, whichwewill come back to later,in Section 3.2, Lohmann, at the Cyanamid European Research Insti-tute, in 1966 published a highly detailed study of the photoelectro-chemical (PEC) behavior of ZnO, both in the presence and absenceof redox couples, including ferro/ferricyanide and methylene blue[46]. He clearly showed that the overall current at the ZnO elec-trode under illumination is the sum of anodic and cathodic cur-rents, the anodic current being a combination of the dissolution ofthe ZnO itself and the oxidation of any redox species present. Thecathodic process was the reduction of O2 to H2O2. This same ap-proach had been introduced in 1938 by Wagner and Traud, at theTechnical University of Darmstadt, to help explain the corrosion ofmetals, coupled with either hydrogen evolution or oxygen reduc-tion [47,48].Another PEC study that wewill mention in this overview is that

of Morrison and Freund, of the Stanford Research Institute, whoalso studied ZnO [49]. These authors also demonstrated in detailthe various situations that arise in the presence and absence ofredox couples. They also showed that oxidation products of someorganic compounds are different in the case of the PEC electrodepoised at the open circuit potential, i.e., with both oxidation andreduction currents balanced, compared to the case of a purelyelectrochemical oxidation. This difference was proposed to be dueto the presence of cathodically generated superoxide. This is one of

Fig. 2.2. Photocurrent vs. potential for illuminated rutile single crystal. SCE refersto the saturated calomel electrode, which is 0.241 V vs. the standard hydrogenelectrode (SHE) (taken from Fujishima et al. [214]).© 1971, Chemical Society of Japan.

the key points in understanding photocatalysis, and wewill returnto it later.During the late 1960s, one of the present authors, at the

University of Tokyo, began to study the photoelectrochemistry oftitania and found that oxygen gas was evolved at potentials verymuch shifted from the thermodynamic expectation, for example,with an onset of ca. −0.25 V vs. the standard hydrogen electrode(SHE), compared to the standard potential of +0.95 V in pH 4.7aqueous buffer (Fig. 2.2) [50,51]. At first, there was skepticism ofthis result, but then it slowly became accepted. One reason thatthis result was difficult to understand is that the photoexcitationprocess converts the photon energy to chemical energy with littleloss, and thus the photogenerated hole has a very high reactivity, sothat it can react directly with either water or quite robust organicand inorganic compounds. Subsequently, a number of studieswerecarried out in which the photoelectrochemical oxidation processon TiO2 was examined for the competitive oxidation of waterto O2 with the oxidation of a variety of inorganic and organicsubstrates [52,53]. Both types of reactions, of course, involve theuse of light energy to get over an energy barrier, either an overalluphill process, as in the case of O2 evolution, or an overall downhillprocess, as in the case of organic oxidations.With the report of the ability to simultaneously generate hydro-

gen gas in 1972 (see Fig. 2.3) [19], the PEC field started to receivemuch wider attention, due to its implications for solar energy con-version [54,55]. From this point, also, photoelectrochemistry be-came closely associatedwith photocatalysis.We shall return to thistopic later, in Section 3.2, and more carefully describe the detailedrelationships.In this overview, we briefly mention some of the early work

of Bard and co-workers at the University of Texas. Frank and Bardwere the first ones to propose that illuminated TiO2 could be usedfor the purification of water via the photocatalytic decompositionof pollutants [56,57]. They suggested that cyanide and sulfite couldbe photocatalytically oxidized to cyanate and sulfate, respectively.In one of these studies, they found that photocatalytic oxidationscould also occur at other illuminated semiconductors, such asZnO, CdS, Fe2O3 and WO3. The most active semiconductor wasfound to be ZnO [57]. These authors expanded this study to along list of inorganic and organic species [58] and speculated thatphotocatalysis could be a useful approach to both environmentalcleanup and photo-assisted organic synthesis. The Bard groupalso suggested that each small illuminated semiconductor particlecould be considered as a PEC cell, with both photo-assisted


Fig. 2.3. Photoelectrochemical cell used in the photolysis of water [19].© 1972, Nature Publishing Group.

oxidation and dark reduction reactions taking place [59]. The Bardgroup also proposed photocatalysis as a way to remove toxicmetals from wastewater [60].For a period of several years, the photocatalysis area continued

to expand as a technology for both the selective oxidation oforganic compounds [22] and the unselective oxidation of organiccompounds for purposes of water purification [1,2,61–64] and,to some extent, also air purification [65–69]. There have alsobeen reviews and listings of references of work on both airand water purification [4,70–72]. For these technologies, it istypically necessary to use powerful ultraviolet (UV) light sources.For passive purification, without special light sources, it becameapparent in the early 1990s that the amount of light present ineither natural sunlight or artificial light was insufficient to processlarge amounts of organic compounds. Therefore, attention wasturned to applications in which a relatively small number of UVphotons could be used to carry out reactions at the TiO2 surface,for example, to decompose thin organic films on solid surfaces orto kill bacteria on surfaces [5,6,73–76]. Thus, the focus turned fromwater purification to passive, self-cleaning, self-sterilizing solidsurfaces, which, with sometimes only slight modification, couldalso be used to purify air. For these types of applications, it wasnecessary to develop ways to coat various materials with TiO2films. Such applications included the self-cleaning glass cover forhighway tunnel lamps, as well as a number of others, which havebeen reviewed previously and which will also be reviewed brieflylater in this article.The large number of applications has also generated a renewed

scientific interest in photocatalysis, and indeed on photo-assistedreactions on semiconducting metal oxides in general. One of theways that we have tracked this activity is by looking at thenumber of citations of the 1972 Nature paper on water photolysis(Fig. 2.4(a)). This number of yearly citations has been climbingsteadily over the past ten years or so and of course is correlatedwith the number of publications appearing on photocatalysis(Fig. 2.4(b)).

Fig. 2.4. (a) Citations per year of the 1972 Nature paper: ‘‘Electrochemicalphotolysis of water at a semiconductor electrode’’ [19]; (b) Numbers of researcharticles appearing on photocatalysis per year: search results in the period of1972–2007with the ‘‘Webof Science’’ (a) by the keyword ‘‘photocataly∗ ’’ (blue bars)and (b) the keywords ‘‘TiO2 AND photocataly*’’ (green bars). (For interpretationof the references to colour in this figure legend, the reader is referred to the webversion of this article.)

3. Properties of TiO2 materials

3.1. Crystal structures

As often described, there are three main types of TiO2structures: rutile, anatase and brookite. The size dependence ofthe stability of various TiO2 phases has recently been reported[77,78]. Rutile is the most stable phase for particles above 35 nmin size [77]. Anatase is the most stable phase for nanoparticlesbelow 11 nm. Brookite has been found to be the most stablefor nanoparticles in the 11–35 nm range, although the Grätzelgroup finds that anatase is the only phase obtained for theirnanocrystalline samples [79,80]. These have different activitiesfor photocatalytic reactions, as summarized later, but the precisereasons for differing activities have not been elucidated in detail.Since most practical work has been carried out with either rutileor anatase, we will focus more attention on these.Rutile has three main crystal faces, two that are quite low

in energy and are thus considered to be important for practicalpolycrystalline or powder materials [81]. These are: (110) and(100) (Fig. 3.1a, b). The most thermally stable is (110), andtherefore it has been the most studied. It has rows of bridgingoxygens (connected to just two Ti atoms). The corresponding Tiatoms are 6-coordinate. In contrast, there are rows of 5-coordinateTi atoms running parallel to the rows of bridging oxygens andalternating with these. As discussed later, the exposed Ti atomsare low in electron density (Lewis acid sites). The (100) (Fig. 3.1b)surface also has alternating rows of bridging oxygens and 5-coordinate Ti atoms, but these exist in a different geometric


Fig. 3.1. Schematic representations of selected low-index faces of rutile: (a) (110); (b) (100); and (C) (001).

relationship with each other. The (001) face (Fig. 3.1c) is thermallyless stable, restructuring above 475 ◦C [81]. There are double rowsof bridging oxygens alternating with single rows of exposed Tiatoms, which are of the equatorial type rather than the axial type.Anatase has two low energy surfaces, (101) and (001)

(Fig. 3.2a, b), which are common for natural crystals [80,82].The (101) surface, which is the most prevalent face for anatasenanocrystals [79], is corrugated, also with alternating rows of5-coordinate Ti atoms and bridging oxygen, which are at theedges of the corrugations. The (001) (Fig. 3.2b) surface is ratherflat but can undergo a (1 × 4) reconstruction [82,83]. The (100)surface is less common on typical nanocrystals but is observedon rod-like anatase grown hydrothermally under basic conditions(Fig. 3.2c) [80]. This surface has double rows of 5-coordinate Tiatoms alternating with double rows of bridging oxygens. It canundergo a (1× 2) reconstruction [84].Recently, the brookite phase, which is rarer and more difficult

to prepare, has also been studied as a photocatalyst (see later). Theorder of stability of the crystal faces is (010) < (110) < (100)(Fig. 3.3) [85].Recently also, the discovery of high-pressure phases of TiO2

was made [86]. These are expected to have smaller band-gaps but similar chemical characteristics [87]. Their existencewas theoretically predicted and then experimentally proven;specifically, a form of TiO2 with the cotunnite structure wasprepared at high temperature and pressure and then quenched inliquid nitrogen. It is the hardest known oxide.There are actually quite a variety of different structures for

compounds with compositions close to TiO2, including those with

excess titanium, such as the Magneli phases, TnO2n−1, where ncan range from 4 up to about 12 and the titanium oxide layeredcompounds, in which there can be as much as several percentexcess oxygen. The oxygen-deficient Magneli phases, which alsoexist for V, Nb, Mo, Re and W, have been known for many years[88–91]. In these compounds, oxygen vacancies are ordered andlead to the slippage of crystallographic planes with respect toeach other; this leads to formation of planes in which, instead ofcorner or edge-shared TiO6 octahedra, there are now face-sharedoctahedra. Fig. 3.4 shows a schematic diagram of this situation. Thecorresponding Ti atoms are then unusually close and can interactelectronically [92]. It has been found recently that laser ablation ofa TiO2 rutile target can produce Magneli-phase nanoparticles [93].There are also quite a number of layered titanate compounds

in which there is an apparent excess of oxygen. For example, thelayered protonic titanate HxTi−2 x/4�x/4O4 •H2O has been preparedand exfoliated into single sheets, termed ‘‘titania nanosheets’’.Fig. 3.5 shows (a) a diagram of the layered structure and (b) TEMand AFM images of single sheets.

3.2. Electronic properties

It was reported in 1942 by Earle that rutile and anatase TiO2in the form of powders are n-type semiconductors and thatthe conductivity decreases with increasing O2 partial pressure attemperatures above 600 ◦C [94]. The effect of O2 was explained onthe basis of an equilibrium involving thermal release of O2 from thelattice. We recognize today that this leads to the creation of Ti3+


Fig. 3.2. Schematic representations of selected low-index faces of anatase: (a) (101); (b) (100); and (C) (001).

sites, which are responsible for the electronic conductivity. Theactivation energy for the electronic conductivity was found to be1.75 eV for unsintered rutile powder and 1.7 eV for sintered rutilepowder. No evidence for ionic conduction was found. Cronemeyerand Gilleo reported in 1951 that rutile single crystals exhibita band-gap energy of 3.05 eV [95]. Absorption spectra werereported for both normal and slightly reduced crystals. For thelatter, the blue color was based on a very broad absorption thatpeaked at 1.8 µm. In the following year, Cronemeyer publisheda very extensive study of the electronic properties of singlecrystal rutile in which the preliminary findings were substantiated[96]. Detailed photoconductivity measurements were made. Darkconductivity and photoconductivity measurements were alsomade on a slightly reduced sample (reduction in H2 at 600 ◦C).Interestingly, therewas found to be amarked hysteresis in the darkconductivity when the sample was raised from room temperatureto 250 ◦C and then cooled back to room temperature. After coolingwith a high applied electric field, the blue color was found to beconcentrated at the negative electrode; the author ascribed this tomovement of oxygen vacancies, but this is not confirmed. It is also

possible that the migration involved interstitial hydrogen. This isan ambiguity that has persisted for many years.Strong reduction of various types of samples was examined,

at various temperatures between 300 and 1150 ◦C. The strongreduction turns the samples blue–black. The activation energyfor electronic conduction had already been reported to be0.07 eV at room temperature, to produce a conductivity of ca.1 �−1 cm−1. The conductivities were found to increase withincreasing reduction time. A ceramic sheet sample heated inhydrogen at 800 ◦C was found to experience a weight loss of0.1%, corresponding to a release of oxygen that would provide3 × 1020 electrons cm−3. Hall effectmeasurements showed a closeagreement between the numbers of carriers and those calculatedon the basis of the weight loss, indicating that all of the electronswere electrically active.Breckenridge and Hosler also published extensive work on the

electrical properties of rutile [97]. The effective electron mass wasfound to be anomalously large, 30–100 times greater than that ofthe free electron. These authors presented convincing argumentsthat the source of electronic conductivity in rutile is Ti3+, which


Fig. 3.3. Schematic representation of the brookite structure (taken from Beltranet al. [85]).© 2006, American Chemical Society.

Fig. 3.4. Schematic representation of the crystallographic shear process to formMagneli phases from rutile (taken from Marezio et al. [632]).© 2000, Elsevier Science.

results from the loss of oxygen, which produces oxygen vacanciesOv. It was proposed that these vacancies (valence +2) can have0, 1 or 2 electrons associated with them, with distinct energies,

Fig. 3.5. Layered lepidocrocite-like protonic titanate: (a) schematic representation;(b) AFM image (taken from Shibata et al. [633] and Sasaki [634], respectively).© 2007, Royal Society of Chemistry; 2007, Ceramic Society of Japan.

the neutral (fully reduced) vacancy being the lowest, followedby the singly reduced and finally the unreduced vacancy. Theobserved optical absorption edge was proposed to correspond tothe transition between the valence band (based on O2−) to theneutral oxygen vacancy, which was considered to be a narrowimpurity band,with ahigh effective electronmass. The band at 3.67eV above the valence band was proposed to be due to Ti4+. Theobserved temperature and oxygen partial pressure dependenceswere fully explained by the equilibrium:

O2− =12O2 + O2+v + 2e

−. (3.1)

The electronic properties of rutile and anatase thin films werestudied by Tang et al. [98]. There were large differences in theelectronic conductivities of the two types of films after reductionby heating in vacuum at either 400 or 450 ◦C. The anatase filmsbecame essentially metallic, with no change in conductivity withtemperature. The rutile films, in contrast, retained measurableactivation energies, 0.076 eV for 400 ◦C and 0.06 eV for 450 ◦C. Thedifference in behavior was considered to be due to the followingproperties for rutile: the average static dielectric constant of ca.100, the effective electronmass of 20m0, and the donor state radiusof ca. 2.6 Å. Since the latter is similar to the distance between Ti4+sites, there is little overlap between donorwave functions. Anatasehas the following properties: static dielectric coefficient of ca. 30,and reduced effective mass of ca. 1 m0, based on an estimateddonor state radius of ca. 15 Å. Based on optical absorption spectra,the band-gap energieswere estimated to be 3.0 eV for rutile and 3.2eV for anatase. Forro et al. reported on the electronic properties of


high purity anatase single crystals and found an activation energyfor electronic conduction of 0.004 eV [99].Recently, Hendry et al. pointed out the problem that the exact

nature of electron transport had not been solved, given the verywide range of values of Hall mobilities (0.01–10 cm2 V−1 s−1)and polaron (electron + accompanying lattice distortion) effectivemasses (8me–190me, where me is the free electron mass) [100].Part of the problem might involve the presence or absenceof dopants, which were found to decrease the mobility. Theseworkers, using THz spectroscopy with undoped rutile singlecrystals, based on Feynman’s analysis [101], found intermediatesize polarons and a mobility of ∼1 cm2 V−1 s−1. Thus, theyconcluded that, in TiO2 films such as those used in dye-sensitizednanocrystalline solar cells, the main limiting factor might beinterparticle contacts. However, this assumes a zero dopant level,which is unrealistic. Clearly, even now, further work needs to bedone to clarify this basic issue.Other recent studies have also focused on the electronic

properties from the standpoint of the dye-sensitized solar cellapplication (see, e.g., work of Aduda et al. [102]). For example, theeffect of the morphology of porous titania films on the electrondrift mobility was studied. Other studies have been focused onthe gas-sensing properties, e.g., for hydrogen gas [103–105]. Innanostructured form, titania has very high sensitivity for H2,increasing greatly in electronic conductivity in the presence oflow concentrations, for example, an increase of three ordersof magnitude upon introduction of 1000 ppm H2 [103]. Themechanism proposed for the increased conductivity was thoughtto involve the adsorption of hydrogen on the titania surface,rather than incorporation into the bulk. The platinum electrodesthat were used to contact the surface in this study were alsopossibly involved, acting to dissociate the H2 molecule, so that Hatoms could be produced and adsorbed more easily. The subject ofhydrogen interactions with titania has been well studied and willbe treated in more detail in Section 3.5.

3.3. Surface structure studies

It is quite difficult to separate work that has been carried outon the surface structure of titania from work that has been carriedout on surface science in general and also on surface chemistry. Thelatter two subjects will be taken up in the two sections that follow.In this section, we will briefly treat the stoichiometric rutile andanatase surfaces. A detailed treatment has been given as part ofDiebold’s extensive review on the surface science [106].Diebold shows how rutile can be cleaved to produce the

commonly shown (110) surface, which is the most stable rutilesurface (see Fig. 3.6). Ramamoorthy et al. carried out theoreticalcalculations on the rutile structure and found the (110) surfaceto be the most stable, based on the fact that it has the leastdangling bonds [81]. The structure shown in Fig. 3.1a is that ofthe unrelaxed bulk and is rather flat, but these authors predictedthat this structure should pucker slightly upon relaxation, with thefive-fold-coordinated Ti atoms depressed by 0.32 a.u. (0.169 Å),and the bridging oxygens also depressed, by 0.15 a.u. (0.079 Å).Vogtenhuber et al., using similar calculations, found that the five-fold Ti atoms were depressed by 0.180 Å, the bridging O atomsby 0.156, the planar O atoms by 0.115 Å and the six-fold Ti atomsby 0.049 Å [107]. An experimental study that made use of surfaceX-ray diffraction found that the five-coordinate Ti was depressedby 0.16 Å, while the six-coordinate Ti was pushed out by 0.12 Å,and the bridging O was depressed by 0.27 Å. A total of seventheoretical studies were compared with the experimental surfaceX-ray diffraction results in the review of Diebold [106]. Most ofthese studies have agreed on the depression of the five-coordinateTi, on the pushing out of the six-coordinate Ti and the depression

Fig. 3.6. Schematic diagram of the cleavage of rutile along the (110) plane (takenfrom Diebold [106]).© 2003, Elsevier Science.

of the bridging O. However, a more recent experimental study thatused LEED found that the bridging O is pushed out by 0.12 Å, incontrast to the earlier experimental study [108]. New theoreticalcalculationsweremostly in agreementwith the LEED study, exceptthat the bridging O position was almost unchanged from theunrelaxed structure [109].The titania surface may undergo significant structural changes

when it is exposed to water. Onishi and co-workers have shownthat single crystal rutile surfaces that have been prepared bystandard methods used for UHV can actually erode and roughenwhen exposed to an aqueous electrolyte [110]. Subsequently,Nakato and co-workers reported on amethod by which atomicallyflat single crystal surfaces could be prepared that were stable inaqueous electrolyte [111]; thismethod involved etching in 20%HF,followed by air-annealing at 600 ◦C.The X-ray crystal truncation rod (CTR) technique has been

used by Zhang et al. to examine the rutile (110) surface in thepresence of pure water and of 1 molal RbOH aqueous solution[112]. Interestingly, the five-coordinate Ti, which had been foundto be significantly depressed in UHV, was found to be depressedto a much smaller degree in pure water (0.051 Å) and only slightlydepressed in 1 m Rb+. This is because the terminal position, whichis empty in UHV, is occupied by a water molecule in aqueoussolution and by a hydroxide ion in the alkaline Rb+ solution. Thesix-coordinate Ti, which had been found to be pushed out in UHV,was found to be depressed to a small degree (0.002 Å in water and0.019 in Rb+ solution). The bridging O, which had been previouslybeen observed to be depressed in the X-ray study and pushed outin the LEED study, was found to be pushed out, by 0.004 Å in waterand 0.010 in Rb+ solution. All of the displacements were smallerthan those found in vacuum, which the authors propose to be duethe fact that either water molecules or Rb+ ions occupy positionsthat would be occupied in the bulk lattice. Further conclusionsfrom this study will be discussed in Section 3.4, in which we treatinteractions of titania with water.Other predictions from the Ramamoorthy work were in regard

to the relative stabilities of the other rutile single crystal surfaces.The order of stability was found to be (110) > (100) >(011) > (001). This calculation is strictly only valid for 0 K andis for vacuum. Based on the results of the X-ray CTR study foraqueous solution, this ordering might be modified slightly, sincethe stabilities are based in part on the presence of dangling bonds,which would of course not be present any longer in the presenceof water.

3.4. Surface chemical studies: Interactions with water

The interactions of titania surfaces with water have been stud-ied extensively and have been reviewed [106,113]. The earlierwork involved conventional surface science methods. The studies


Fig. 3.7. High-resolution electron energy loss spectra of rutile (110) at variousdosages of water, starting from the clean surface at bottom (taken from Hendersonet al. [114]).© 1996, Elsevier Science.

reported during the past five years on this subject have involvedboth theoretical and experimental studies, the latter including alarge number of scanning tunnelingmicroscopy studies. The inter-actions with water are important to understand, because water,either liquid or vapor, is almost always present in photocatalyticreactions. These interactions are especially important for the laterdiscussion of the photo-induced hydrophilic effect.Much of the work that has appeared over the past decade has

been targeted at the question of whether water is adsorbedmolec-ularly or dissociatively. Going back to one of the pioneering workson this subject, Henderson reported a high-resolution electron en-ergy loss spectroscopy (HREELS)-temperature-programmed des-orption (TPD) study that concluded that the adsorption of wateron rutile (110) is molecular on the stoichiometric surface and dis-sociative on the reduced surface, which is conventionally producedby heat treatment, presumably forming oxygen defects [114]. Theprogression of HREELS spectra as a function of water coverage isshown in Fig. 3.7. In the background spectrum at the bottom, thereis a very small peak at 3690 cm−1. This is due to the O–H stretchfor OH groups that are not hydrogen-bonded, often called ‘‘iso-lated’’ OH groups. This vibrational frequency is close to that forOH groups that stick out from the surface of liquid water, withoutbeing hydrogen-bonded to any neighbors, as observed with sum-frequency generation (SFG) spectroscopy [115]. At higher cover-ages, the peaks that appear are shifted to lower wavenumbers,indicating hydrogen bonding. There is no longer any evidence ofthe high wavenumber peak, except at the highest coverage. Thereis also a peak that appears at 1605 cm−1, which is due to theH–O–Hbending mode of liquid water. Thus, it is certain that there are wa-ter molecules adsorbed. However, it is not certain whether thereare also dissociated water molecules present that are hydrogenbonded to neighboring water molecules or to bridging oxygens.There is not much doubt that water dissociates at oxygen

vacancies that are produced by heating in vacuum [113]. The

Fig. 3.8. Schematic diagram of a mixed molecular water-dissociated watermonolayer on the rutile (110) surface (taken from Lindan [127]).© 2003, American Institute of Physics.

main question is whether or not this is true for the non-reduced,stoichiometric surface. Theoretical studies have been divided intothose that predict molecular adsorption [116–122], those thatpredict dissociative adsorption [123–129], and those that also findstability for mixed molecular-dissociative adsorption [122,125–127]. This is a particularly difficult problem, since the energydifferences are rather small. One of the studies that has predicteddissociative adsorption also predicts a mixed layer of molecularand dissociated water at higher coverages [127]. Fig. 3.8 is aschematic diagram taken from this paper that shows how themixed monolayer is arranged; the two structures both includehydrogen bonding between a water molecule adsorbed at a five-coordinate Ti site and an OH group adsorbed at an adjacent 5-coordinate Ti site. There is also a weak interaction of the watermolecule with the bridging oxygen. For reference, the diagramsfor purely dissociative and purelymolecular adsorption are shown.Zhang and Lindan have also calculated a theoretical vibrationalspectrum, which we show along with one of the HREELS spectrafrom Henderson’s work (Fig. 3.9). Even though the simulatedspectrum is significantly shifted upward in wavenumber, the two


Fig. 3.9. (a) A HREELS spectrum for water adsorbed on rutile (111), taken fromHenderson [114]; (b), (c) and (d) show simulated vibrational spectra for the varioustypes of submonolayers and monolayers studied by Zhang and Lindan [127].© 2003, American Institute of Physics.

peaks have an appearance that is similar to the experimentalspectrum. The strong peak at highwavenumber is due to an almostcompletely non-hydrogen-bonded OH, presumably the terminalhydroxyl group, and theweaker, lowerwavenumber peak is due tothe OH group of the water molecule that is bonded to the hydroxylgroup.This type of double-peak structure is rather commonly

observed in experimental infrared spectra for both rutile andanatase powders (Fig. 3.10). The quite sharp peak or peaks athigh wavenumber are due to isolated OH groups, and the broader,lower wavenumber peak or peaks are due to hydrogen-bonded OHgroups. In all of the spectra shown, there is some fine structure.Although not certain, this could be due to the existence of differentcrystal faces, with slightly different geometries for adsorption.Thus, it appears likely that, at least on powders, with coverages

on the order of a monolayer, there could be mixed monolayers.Certainly, there is molecular water, and there must also behydroxyl groups, with the OH group pointing up, normal to thesurface, so that there is little opportunity for hydrogen bonding.However, for powders, there are, of course, a variety of crystalfaces exposed, and distinct situations might be found on each. Thisis expected from the work of Henderson in a comparison of the(110) and (100) surfaces [130]. The latter was found to supportdissociative, while the former was found to support molecularadsorption.Direct evidence for molecular adsorption on stoichiometric

rutile (110) can also be found in STM work that has been targetedat interactions of water with oxygen vacancies. This general topicwill be discussed next.The background of the work on the interaction of water with

oxygen vacancies on rutile (110) has been given by Henderson

Fig. 3.10. A series of three sets of infrared spectra for (a) anatase [635] and (b) [636],(c) [637]) rutile powders acquired at various temperatures and water coverages. In(a), the water coverage decreases with spectrum number, and in (b) and (c), withspectrum letter. In (b) and (c), the original spectrawere obtained in the transmissionmode; all spectra have also been replotted with increasing wavenumber. In (b), themain peaks are listed.© 1988, Royal Society of Chemistry; 1987, American Chemical Society; 1971, RoyalSociety of Chemistry.

[113]. One of the interesting aspects is that Ti3+ sites by themselvesdo not have special reactivity; for example, such sites on rutile(100) and on Ti2O3 are not reactive. Only on the (110) surface arethey reactive.The background of the STM work has also been discussed

in the thorough reviews of Henderson [113] and Diebold [106].For example, the latter discusses the problems of distinguishingbetween oxygen vacancies and hydroxyl groups that have beenproduced as a result of a water molecule reacting with anoxygen vacancy. A number of authors concluded that themedium-brightness spots that they observed in STM between bright rowswere due to oxygen vacancies. Diebold et al. pointed out that thereappeared to be two types of defects that were observable on rutile(110), which they termed ‘‘A’’ and ‘‘B’’ [131]. Between the rowsof 5-coordinate Ti atoms, which appear bright due to their highelectron density, there are darker rows that are due to the bridgingoxygens. The ‘‘A’’ type were observed to be significantly brighterthan the ‘‘B’’ type and were proposed to be oxygen vacancies. TheA defects were removed by scanning the tip at a voltage of +3 V,while the B type remained. The A-type defects were also foundto be quite mobile. Suzuki et al. subsequently reported similarimages and also found that the brighter spotswere removablewitha scan at+3 V [132]. These authors found that the spots were alsoremovable by electron-stimulated desorption. They were also ableto produce additional spots by dosing with atomic hydrogen. Thus,they proposed that the bright spots were due to hydroxyl groupsformed by hydrogen adsorption on bridging oxygens. Brookes et al.also carried out STMmeasurements on rutile (110) and found that,


Fig. 3.11. STM images of rutile (110) showing (a) oxygen vacancies and (b) bridging hydroxyl groups (taken fromWendt et al. [129]).© 2005, Elsevier Science.

when they purposefully dosed the surface with water, it adsorbedwithout dissociation at 150 K and then dissociated at 290 K.Significantly, no terminal features were observed, i.e., there wasno evidence for oxygens or hydroxyls adsorbed on 5-coordinate Tisites,whichwouldhave resulted if thewater haddissociated on thestoichiometric surface. Features were only observed on the dark,bridging oxygen rows. Thus, the authors concluded that the waterhad dissociated at oxygen vacancies.Schaub et al. carried out further work and found that there

were two types of ‘‘A’’ defects, a smaller type and a larger, brighterone [117]. The latter was assigned to an oxygen vacancy and theformer to a hydroxyl group, based on theoretical calculations.These authors continued to support the idea of water beingdissociatively adsorbed only at oxygen vacancies, in line withtheir theoretical calculations. In further work from this group,however, the assignmentsweremodified [129]. In thiswork, it wasrecognized that the surfaces that had been examined earlier weremostly hydroxylated. Special care was taken to achieve extremelylow levels of background water, and thus it became clear that thedarker featureswere the actual oxygen vacancies and themedium-bright features were individual hydroxyl groups that had beenformed via water dissociation (Fig. 3.11).Subsequent work showed even more clearly how a water

molecule moves along a row of 5-coordinate Ti sites and thenreacts with a vacancy, first producing a pair of hydroxyl groupson neighboring bridging oxygens [133]. A second water moleculecan then further catalyze the splitting of the hydroxyl pair inan energetic reaction that can result in one of the protonsjumping several rows away. The initial water-dissociation processis shown in Fig. 3.12. Both the dissociation and splitting processesare also available as movies [134,135]. This paper also correctsthe assignments that had been given in work focused on theinteraction of oxygen (O2)with oxygen defects [136].A paper by Bikondoa et al. appeared in early 2006 [137],

apparently written without the knowledge of the one by Wendtet al., which appeared in 2005 [129]; this paper also clearly showedthe whole situation regarding previously published assignmentsby the various groups. These authors can be credited with havingrecognized, from the beginning, the fact that the bright spotsthat were being observed by various groups were in fact due tohydroxyl groups, either single or double, that had been formed viawater reaction at oxygen vacancies.Additional work has recently appeared on the STM observation

of the reaction of water molecules with oxygen vacancies. Zhanget al. reported that the two bridging OH groups that are producedfrom the reaction are actually not identical [138]. The one that isproduced at the site of the original oxygen vacancy is not observed

Fig. 3.12. STM images of rutile (110) showing the dissociation of a water moleculeat an oxygen vacancy (taken from [133]).© 2006, American Physical Society.

to move, whereas the one that is produced at an adjacent bridgingoxygen due to the addition of a proton, is quite mobile. This resultappears to be consistent with the report of Wendt et al., in whichit was shown that proton can jump several rows away [133].The implication is that the electrons associated with the originalvacancy are rather localized. All the theoretical results have notbeen in agreement with this picture.The work just described on the interactions of water with

vacancies on rutile (110) has a more general implication, inaddition to the obvious one. Some of the studies that have beencarried out over the years that have discussed reactions of oxygenvacancies may have been actually dealing with bridging hydroxylgroups. Henderson had already pointed out this effect in 1996[114].Wenote also thework ofMezhenny et al., whichwas focused on

the question of whether or not UV light produces oxygen vacancieson the rutile (110) surface [139]. This work showed little effectof ordinary intensity levels of UV light, i.e., similar to those thatare present in sunlight, in producing oxygen vacancies on thesurface. It is likely that this work might have also suffered fromunrecognized background levels of water, since they report STMimages that are characterized by the brighter spots that have beenassigned by later workers to bridging hydroxyls. Nevertheless, ifoxygen vacancies had indeed been produced, there would have


been an increase in the number of such hydroxyls, whereasno increase was observed. This result is in contrast to earlierresults obtained with second harmonic generation and X-rayphotoelectron spectroscopy that concluded that large numbers ofdefects were in fact produced and were reactive toward dioxygen[140]. These issues will be discussed at greater length in the nextsection and in Section 4, which deals with the photo-inducedhydrophilic effect.Work on the interaction of water with titania surfaces has also

been carried outwith other techniques. For example, anatase pow-der was examined with gas-chromatography–mass spectrometry(GC–MS) and quantum chemical calculations. Experimental evi-dencewas found forwater dissociation, whichwas consistent withthe theoretical calculations; the latter showed that the adsorptionis molecular on the anatase (101) surface and dissociative on the(001) surface.The previously mentioned X-ray CTR study of Zhang et al.

provides rather clear evidence for molecular adsorption for purewater on the stoichiometric rutile (110) surface and dissociativeadsorption in alkaline aqueous solution (pH 12) [112].

3.5. Surface chemical studies: Interactions with dioxygen and otherspecies

It was realized in 1998 that dioxygen does not only react withoxygen vacancies (or possibly, as discussed above, bridging hy-droxyls) to produce a near-stoichiometric surface at temperaturesabove 600 K, but also, at temperatures below 600 K, it can leave be-hind an oxygen atom adsorbed at a 5-coordinate Ti site [141]. Thisrealization led to doubts concerning previously published workthat had found dissociative water adsorption at rutile (110). It alsoled to a reassessment of what had been an accepted procedure forthe preparation of high quality, clean surfaces. The scheme thatEpling et al. proposed to explain the interactions of O2 with oxygenvacancies and subsequent reactionwithwater is shown in Fig. 3.13.In the same paper, these authors also found evidence to support asimilar end product that resultedwhenwaterwas present initially,so that bridging hydroxyls had already been formed.Henderson et al. reported later that O2 can adsorb at a reduced,

i.e., vacancy-containing, rutile (110) surface without dissociationat temperatures below 150 K [142]. One of the more interestingaspects was the observation that O2 can adsorb, probably as O•−2 ,at a ratio of up to three molecules per oxygen vacancy, whichnecessarily means that it does not have to interact directly withthe vacancy but can reside on an adjacent cation site. Anotherpaper from the same group appeared more recently exploring thereaction of O2 with bridging hydroxyl groups in more detail [143].These authors conclude that the role played byO2 in photocatalysisinvolves specifically this reaction. They also found, in agreementwith their earlier work, that a second monolayer of water blocksthe access of O2 to the bridgingOHgroups, effectively impeding theelectron transfer. On the basis of these results and other studies inwhich superoxidewas generated both on thermally reduced titaniaand on UV-illuminated titania, the authors proposed that bridginghydroxyls are a key intermediate in the photocatalytic process.We agree with this proposal and also propose (see later) that suchbridging hydroxyls can be generated electrochemically.The effect of gas-phase O2 on nanocrystalline titania films has

been studied in terms of the gas-sensing application [144]. Itwas found that the film conductivity decreased in the presenceof O2. This effect could also be related to the effect discussedabove but is more likely to involve the scavenging of bulk trappedelectrons, as discussed in the next section. The authors alsoobserved photo-induced adsorption of O2, a phenomenon that hadbeen described by Kennedy et al. in 1958, as mentioned in thehistorical overview [40].

Fig. 3.13. Schematic diagram of an O2 molecule reacting at an oxygen vacancy onrutile (110) and dissociating, with further reaction with a water molecule (takenfrom Epling et al. [141]).© 1998, Elsevier Science.

We include here a brief mention of surface photochemicalreactions involving O2. The work of Thompson and Yates hasemployed the photodesorption of O2 as a means of monitoringthe arrival of photogenerated holes to the surface of a rutile singlecrystal with exposed (110) face [145]. This process is essentiallythe reverse of the photo-adsorption process just alluded to, whichrequires a trapped electron, creating a partially or fully reducedO2, i.e., O•−2 . The presence of methanol as a hole trapping agentsignificantly decreased the photodesorption. In furtherwork by thesame authors, they proposed a fractal rate law to fit the observedkinetics of the reaction of trapped electrons with trapped holes[146]. It was assumed that the electrons were associated withoxygen vacancies, but this picture may be in some doubt, based onthe ability of trace water to convert these to bridging hydroxyls.

3.6. Bulk chemistry—Hydrogen

In this section, we briefly review the literature on thebulk chemistry of titania. This subject is mostly limited to theincorporation of elemental hydrogen. It also can include theincorporation of lithium or sodium, but this is beyond the presentscope. The subject of hydrogen incorporation can also includeelectrochemically-induced processes; these will be treated in thenext section. The characteristics of hydrogen as a bulk impurityin titania are central to the understanding of its electrical,electrochemical and photoelectrochemical behavior, which is, inturn central to the understanding of the photocatalytic behavior.An early paper on the optical and infrared absorption spectra

of rutile single crystals by Soffer showed evidence for theincorporation of H, already present in the as-received crystal,


Fig. 3.14. (a) H binding site within the rutile structure and (b) proposed diffusion path within the solid structure, along a c-channel (taken from Bates and Perkins [154]).© 1979, American Physical Society.

and D, which was introduced by heating at 900 ◦C in D2; thesewere evidenced by the appearance of IR bands at 3277 (main)and 3322 cm−1 due to an O–H stretch and at 2442 cm−1 for thecorresponding O–D stretch [147]. The author remarked that thebands are unusually narrow for a solid-state O–H stretch and alsodiscussed the possibility of a hydrogen-bonding-related shift.Prior to the work of Hill in 1968 [148], a number of different

studies had suggested that non-stoichiometry in rutile TiO2was associated with increased electrical conductivity, but themechanismwas not clear. It had also been suggested that hydrogenacts as a dopant in rutile. In Hill’s work, heating rutile crystalsin hydrogen below 600 ◦C led to increases in the bulk hydrogenconcentration, measured by IR. Heating above 650 ◦C in vacuumled to decreases in hydrogen concentration, coupled with oxygenloss to produce water. This led to increased conductivity, probablydue to the formation of faults involving Magneli phases. For theheated crystals, the electrical behavior was modeled involvinga series network of two parallel RC circuits, with one beingassociated with an ‘‘exhaustion layer’’, i.e. either a depletion layeror layer that is very low in carriers, as discussed in the next section.Johnson et al. reported further detailed work on the optical and

infrared spectra for H and D incorporated in rutile single crystals[149]. This paper referred to the earlier work of von Hippel et al.that provided presumably more accurate values for the absorptionmaxima: 3276 and 3317 cm−1 for O–H and 2435 and 2463 cm−1for O–D. That paper had proposed that the peak splitting was dueto slightly differing O–Ti distances. Johnson et al., however, deniedthis possibility due to the symmetry of the structure. H and Ddoping was carried out by heating in an atmosphere of H2O orD2O, plus O2 at 850 ◦C, or in some cases, H2 or D2 below 550 ◦C.Conduction-band electrons produced a broad absorption band at1.5 µm. This was remarked to not be due to conventional freeelectron behavior, which should produce no peak. These authorscarried out a detailed analysis of the various possible bindingsites for H or D within the crystal. The same group reported theuse of the IR absorption band in the precise determination of Hand D concentrations in rutile [150]. DeFord and Johnson studiedthe H/rutile system in detail from the viewpoint of theoreticalsemiconductor and thermodynamic properties [151]. Later, theymade measurements of H and D diffusion using the isotopeexchange technique in order to avoid internal electric fields [152].The diffusion was carried out simply by heating the samples in theappropriate atmosphere (see above) for various times and thenmeasuring the H or D concentrations via the IR absorption. Thediffusion coefficients for H varied from ca. 3 × 10−8 cm2 s−1 at350 ◦C to ca. 1.7 × 10−6 cm2 s−1 at 700 ◦C along the c-axis,

compared with 8 × 10−8 cm2 s−1 at 698 ◦C along the a-axis. Thediffusion coefficient estimated for room temperature was 1.8 ×10−13 cm2 s−1, which was in reasonable agreement with the valuevery roughly estimated by Chester and Bradhurst, in the range10−11–10−13 cm2 s−1, based on electrochemical insertion.Bates and Perkins measured the infrared frequencies for H, D

and T in rutile TiO2 and carried out a detailed structural analysis,comparing the results with theory for anharmonic oscillators andalso with that of hydrogen bonding [153]. The agreement with thelatter was poor. Bates et al. later published a much more detailedstudy, including a review of the literature up to 1979 [154]. Theycarried out a detailed analysis of themechanism of diffusion of H inrutile. The binding site within the lattice for the proton is shown inFig. 3.14a, and the proposed path for diffusion in Fig. 3.14b. Thebinding site was later confirmed by Klauer and Wöhlecke usingpolarized Raman [155]. The understanding of this system achievedin this work is excellent. It was concluded that the wavenumbershift of the IR absorption was not due to hydrogen bonding, whichis consistent with the observed sharpness of the band. Instead, itwas proposed to be due to the electrostatic environment withinthe lattice (however, see below). Further work was also publishedon tritium diffusion [156].Peacock and Robertson have carried out quantum chemical

calculations for H in a variety of oxides that are considered ashigh dielectric constant oxide gate materials [157,158]. They findthat the H0 energy level lies above the conduction band in ZnO,TiO2 and SrTiO3, consistent with the fact that H is a shallowdonor in all the three. Work of Park et al. also confirmed theseresults for rutile [159]. Koudriachova et al. have also carried out arecent ab initio quantum chemical calculation on H incorporationin rutile [160]. These authors sought to recheck the binding site,due to the difficulty already mentioned, i.e., the O–H–O bonddistances were not consistent with established rules relating themto vibrational frequencies. The new calculation found a distortionin the cage surrounding the H atom such that the distancesbecame highly consistent with the correlation. They found thatthe H atoms are most favorably located at ordered positions, asshown in Fig. 3.15a. The lattice expands linearly with increasingH incorporation (Fig. 3.15b).It is also appropriate to mention here theoretical calculations

that were carried out on the surface adsorption of hydrogen asH2 [161]. In that work, it was found that up to one monolayer isadsorbed, with all of the bridging oxygens becoming hydroxylatedand the underlying Ti4+ ions being reduced to Ti3+. Wemight notethat this type of surface could in principle be produced thermallyby removing half of the bridging oxygens, followed by exposure to


Fig. 3.15. (a) Illustration of a high stability ordered arrangement of interstitialhydrogen atoms in the rutile structure, with stoichiometry H1/4TiO2 (taken fromKoudriachova et al. [160]). (b) Plot of lattice volume for various numbers ofinterstitial H (open circles) and Li atoms (filled circles) (taken from Koudriachovaet al. [160]).© 1994, American Physical Society.

water. It could also be produced in principle electrochemically (seeSection 3.7).It has been found that titania nanotubes respond to the

presence of hydrogen in the gas phase, as already discussed inthe section on electronic properties. In that work, it was notconsidered that hydrogen could actually be absorbed. However,work of Lim et al. showed that for nanotubes that were preparedhydrothermally, there was a reversible uptake of ca. 2% [162]. Theincorporation led to an increase in the IR absorption (3427 cm−1),which is significantly higher than that for single crystal rutile. Only75% of the uptake was reversible at room temperature with theremaining 25% requiring temperatures up to 130 ◦C to desorb.A relatively detailed study has been carried out on small

rutile crystals by the use of optical and IR absorption and Ramanscattering [163]. The effect of neutron irradiation was also studied.The incorporation of H in minerals is of interest to geologists,because it affects the macroscopic properties and can be a wayby which water is incorporated in minerals that normally do notabsorb water.

Panayotov and Yates recently reported on experiments inwhich they reduced titania pellet samples (Degussa P25) witha source of H atoms [164]. They were able to observe thebroad, featureless background visible-IR spectrum expected forconduction-band electrons. In addition, there were increases inelectronic conductivity. However, there was no discernable IRabsorption at ca. 3280 cm−1 for the internally bound O–H stretchalready discussed above. The activation energy for the diffusion ofH into the solid was estimated to be 0.09 eV.

3.7. Electrochemical properties

The intrinsic electrochemistry of TiO2 has been studiedcontinuously for a long period, since the first report of Boddy in1968 on the oxygen evolution reaction [165]. This work made useof single crystal rutile electrodes. Interestingly, this work containsa figure in which the photocurrent vs. potential behavior is given,one year prior to the reports of one of the present authors [50,166].This will be discussed further, in the next section.Many aspects of the behavior of TiO2 can be explained on the

basis of a semiconductor model, as already discussed. One of theways of characterizing a semiconductor is to measure its flat-bandpotential electrochemically, for example, with capacitance. In ouroriginal work reported in 1969, we described such measurementsand concluded that the flat-band potential EFB for rutile (001) ispH-dependent, with a relationship similar to the following [167]:

TiO(O)+ H+ + e− = TiO(OH) (3.2)

with the resulting Nernst relationship:

EFB = 0.00+FRTln[H+] = −0.0591pH at 25 ◦C. (3.3)

This result has essentially been confirmed by subsequent workers,with the value at pH 0 being +0.01 ± 0.05 V vs. SHE for rutile(001) [169,171,175]. The pH dependence is typical of the behaviorof most oxide semiconductors and has generally been consideredto be due to a surface acid–base equilibrium for these oxides. Thevalue for anatase is more negative:−0.20 V vs. SHE [168].There has been a certain amount of discussion devoted to the

question of exactly how the capacitance measurements should beconducted and how the results should be interpreted, even forsingle crystals. This discussion is interesting, because it displaysthe convergence of an ideal, simple theoretical model with a real,non-ideal complicated material. The model, already discussed,involves an ideal semiconductor with a space charge region whosethickness is dependent upon the potential difference between theFermi level, to which we can assign an electrochemical equivalentEFL and the conduction-band-edge energy, again given on theelectrochemical scale, ECB. The energy difference ξ between ECB andEFL deep within the bulk of the material is dependent upon thecarrier concentration. The space charge capacitance CSC based onthis simple model is given by

1C2SC=2(E − EFB)εε0eND

(3.4)

where ND is the bulk concentration of donors, with the conse-quence that a plot (Mott–Schottky) of C−2 vs. potential yields thecarrier concentration from the slope and the flat-band potentialfrom the intercept with the potential axis. Many workers havefound that these plots are either non-linear or that the interceptgives a result that is not reasonable, for example, more negativethan that given in Eq. (3.3). Such results have been explained invarious ways, including (1) non-uniform depth profile of carriersand (2) deep trap levels. There appears to be some consensus onthe merits of the first explanation.


Fig. 3.16. Mott–Schottky plots for a rutile (001) surface in pH 4.7 buffer: (a) rawdata for a more lightly doped sample (circles), with a solid line showing the fitto Eq. (3.5) (including both space charge CSC and passive layer CPL contributions);(b) calculated intrinsic CSC behavior, after removing the effect of CPL; (c) raw datafor a more heavily doped sample (squares), with a solid line showing the fit toEq. (3.5); (d) calculated intrinsic CSC behavior, after removing the effect of CPL . SCErefers to the saturated calomel electrode,which is 0.241V vs. the standard hydrogenelectrode (SHE) (based on [167]).

A model that can explain the appearance of a steep slope nearthe intercept and a shallower slope at more positive potentialsis that of a thin layer near the surface that has a lower carrierconcentration. This model is reasonable, because, as has beenreported often, the effect can arise as a result of oxidizing etchingtreatments, which could remove electrons from such a surfacelayer, especially if the treatment is conducted for a relativelyshort time. In this case, the overall capacitance behaves as ifthere is a smaller, potential-independent capacitance due to thepassive layer CPL, in series with the potential-dependent spacecharge capacitance. The slope of the upper portion of the curveis somewhat increased from that which is characteristic of thecarrier concentration in the bulk of the material, and the interceptis shifted to the negative. The overall behavior is described by

1C2TOT=2(E − EFB)εε0eND

+

(2(E − EFB)εε0eND

)1/2 (1CPL

)+

(1C2PL

)(3.5)

which is derived by substituting

1CTOT=1CSC+1CPL

(3.6)

into Eq. (3.4). In our original paper, we reported curves with thistype of double slope (see Fig. 3.16) and made a mathematicalcorrection of this type in order to estimate the true flat-bandpotential [167]. Our explanation at that time invoked a Helmholtzlayer capacitance, which was later correctly disputed by Dutoitet al. [169]. Based on our nitric acid etching procedure usedin that work, however, it is quite reasonable to invoke thepassive layer model just described. DeGryse et al. also found faultwith the involvement of the Helmholtz layer and proposed aninhomogeneous carrier concentration [170]. Tomkiewicz proposeda model based on surface states with energies in the middle ofthe band-gap [171]; this model was criticized by Ullman, whoproposed a surface passive layer [172].It should be noted that this same model has been discussed by

Schoonman et al. [173], but the equation given in that paper wasdifferent, in that the cross-term was not included, leading to theerroneous result that the curve is simply raised but has the sameslope. If theMott–Schottky behavior does not fit this simplemodel,it is still possible that it can be fitted with a more complicatedmodel, in which the slope at any given point in the curve canbe related to the carrier concentration at a certain depth; thus acomplete concentration profile can be estimated [174].

Finklea has given an excellent summary of the practicalmethods of obtaining linear Mott–Schottky plots with correctintercepts [175]. The opposite situation can also arise if a surfacelayer exists with a higher concentration of carriers compared tothe bulk. This leads to a shallow slope near the flat-band potential,followed by a steeper slope at more positive potentials.In certain cases, if the carrier concentration is sufficiently

high, the space charge capacitance can approach that of theelectrical double layer, and the Helmholtz capacitance must alsobe considered.It can be argued in general that the behavior of TiO2 is simply

not ideal, and, in each individual case, a full impedance treatmentis necessary to extract the space charge capacitance or even toevaluate whether or not it exists. Similar situations also exist foroxide films; the situation is particularly complicated for oxide filmsgrown on metals, for example, iron and zirconium [176].For certain types of TiO2 electrodes, it may be doubtful that the

behavior can be strictly described with a semiconductor model.For example, the Mott–Schottky plots may exhibit a significantfrequency dispersion, either with or without non-linearities. Thiscan signal the fact that a different type of model might bemore appropriate. In most cases, it is advisable to examine thefull impedance spectrum over a range of potentials in orderto determine whether or not the behavior does in fact followthat expected for a semiconductor. One specific type of behaviorthat has been observed is that of an electrochromic film, inwhich all of the charge injected into the film is associated withincreased absorption of visible light (see Refs. [177–181] and laterdiscussion).The electrochemical impedance spectral (EIS) behavior can

be quite powerful in elucidating the behavior of electroactivematerials, i.e., those that can undergo electron transfer reactionswithin pores and even within the solid material itself. Nogamiexamined the EIS behavior of several single crystal rutile samplesand obtained sets of straight lines, with varying slopes in Bodeamplitude (log |Z | vs. log frequency) plots [182]. He explainedthis behavior on the basis of a disordered layer on the surfaceof the single crystal. The Bode plots show straight lines for thelogarithmof themagnitude of the impedance (absolute value of thecomplex impedance) vs. the logarithm of the frequency. At morepositive potentials, the slopes of the lines are close to 0.7, while,at more negative potentials, the slopes are close to 0.5. For a purecapacitance, the slope should be 1.0, while, for a pure resistance,it should be 0.0. For intermediate cases, there are both resistanceand capacitance distributed through the material. The simplestelectrical case is that of the uniform semi-infinite transmissionline, in which there is a ‘‘ladder’’ of constant resistances andcapacitances (Ref. [183] and references therein). This type ofbehavior has been shown to be followed by a porous mediumwith cylindrical pores. A rate-limiting diffusion process inwhich anelectroactive species diffuses through a uniform medium behaveselectrically in the sameway, and thus there is some ambiguity as towhich process is actually operating. In either case, the slope shouldbe 0.5, as observed by Nogami for more negative potentials.For the general case, in which the slope of the Bode plot can

have an arbitrary value between 0.5 and 1, the same type ofRC ladder is also valid, and the physical picture is also similarto that for the uniform network, but with pores that are non-cylindrical. For example, Wang and Bates have shown that horn-shaped pores can give rise to Bode slopes in this range [184]. Thus,the results of Nogami could be explained by a situation in whichhorn-shaped pores of larger diameter (lower roughness factor)exist at higher potentials; at intermediate potentials, the diametersbecome smaller (higher roughness factor); finally, at the mostnegative potentials, the pores behave as if they are cylindrical andlonger than the penetration length of even the lowest frequencies


Fig. 3.17. Electrochemical impedance spectrum for a nanocrystalline anatase film on a conductive support: (upper left) experimental results from Cao et al. [187]; (upperright, lower right) simulated log of the impedance amplitude vs. log frequency and phase angle vs. log frequency, respectively.© 1995, American Chemical Society.

measured. It should be noted that impedances with arbitraryslopes are often referred to as constant phase elements, meaningsimply that the phase angle, which is 0◦ for a pure resistance, 90◦for a pure capacitance and 45◦ for the uniform transmission line, isconstant but is not one of these standard ones. In any case, it seemsclear thatNogami’s conclusion, i.e., that the frequencydispersion oftheMott–Schottky plots is due to the presence of a disordered layeron the single crystal surface, might be valid. Frequency dispersionis a result of the mixing of resistive (i.e., electronically conducting)character with capacitive (i.e., ionically conducting) character.Furthermore, his proposal that the behavior involves surface

states might also be valid. Wang also has shown that constant-phase angles between 90◦ and 45◦ can be obtained if theresistances and capacitances in the ladder are non-uniform,e.g., with the resistance increasing and the capacitance decreasingin a logarithmic manner [183]. Even a 45◦ angle can result witha non-uniform situation. A non-linear gradient of resistance andcapacitance could well result if there is a similar gradient ofconcentration of Ti3+ sites within the film. Cahan and Chen havediscussed such gradients in the passive oxide film on iron. Theyhave also pointed out that these gradients can naturally give riseto linear log current vs. potential behavior (linear Tafel slope; seebelow) that mimics that for rate-limiting electron transfer [185,186].Cao et al. have reported impedance results for a nanocrystalline

anatase TiO2 film at a single, intermediate potential (0.24V vs. SHE)[187]. In this case, the Bode plot exhibited three regions: capacitivebehavior at high frequency, diffusive behavior at intermediate fre-quencies, and capacitive behavior at lower frequencies (Fig. 3.17).This behavior can be simulated approximately by the equivalentcircuit shown, with the resulting Bode plot shown on the right.Here, we have not attempted a perfect fit, which would have re-quired the introduction of constant-phase elements to simulatethe non-ideal capacitive behavior, which results in phase anglesconsiderably below 90◦. The similar shape of the calculated plot

compared to the experimental result shows that there is some va-lidity for this model. The higher frequency capacitive behavior,with a capacitance of ca. 10 µF cm−2, is typical for the flat surfaceof the film. The lower frequency capacitance (ca. 25 µF cm−2) isproposed to involve a bulk reduction of Ti4+ to Ti3+, as discussedbelow. The diffusion-like processmight involve either actual diffu-sion in pores or combined gradients of resistance and capacitance,or even a combination.Muñoz et al. studied the impedance behavior of titania

nanotube arrays using EIS and also found behavior with similaraspects [188]. The nanotubes were prepared via anodic oxidationof Ti metal; the titania produced in this way is amorphousand can be converted to anatase with heat treatment. Theimpedance of both forms was measured. For the amorphous filmsat intermediate potential (0.0 V vs. Hg/Hg2SO4), there were twocapacitive regions, as in theworkdescribed above,with a transitionregion at intermediate frequencies that could be attributed to aWarburg impedance. At higher potential (+1.0 V), the impedancewas almost purely capacitive, and, at lower potential (−0.5 V),it was capacitive at lower frequencies, with a relatively lowimpedance (higher capacitance) and predominantly resistive athigher frequencies. The capacitive components are discussed interms of a space charge capacitance due to the oxide film at thebase of the nanotubes, the film being in contact with the metal,and a Helmholtz (double layer) capacitance of the pore walls.Interestingly, for the annealed compact film, without nanotubes,the impedance also exhibited two distinct capacitive regions, justas found by Cao et al., with little dependence upon potential. Incontrast, after annealing, the spectra for the nanotube films werealmost entirely capacitive, with a single component.This general behavior is similar to that for other oxides of

transition metals of the early transition series, e.g., MoO3 andWO3. Our own work with amorphous WO3 films clearly showedfrequency regions in which the behavior was more capacitiveand regions in which it was more diffusive [189,190]. There


even appeared to be two types of diffusion, which we attributedto diffusion in pores and diffusion within particles. In recentwork on MoO3 films, impedance results also showed diffusivebehavior at intermediate frequencies and capacitive behavior atlower frequencies; the latter was attributed to an electrochemicalreduction process [191]. With impedance, in which the potentialvariation is usually only a few millivolts, an electrochemicalprocess can behave exactly like a capacitance and is often termedas pseudocapacitance.The use of impedance measurements can clearly yield much

information about TiO2 photocatalytic films. It is clear from acomparison of the reported impedance results that more workis needed in order to fully elucidate the physical significance.Further studies could establish the relationships between the filmnanostructure and the electrochemical behavior.The electrochemical behavior of TiO2 in the more negative

region has been studied by several groups with the morecommonly used cyclic voltammetric (CV) technique, on singlecrystal rutile [192], single crystal anatase [168] and nanocrystallineanatase [187,193]. In all cases, there is a distinctive feature thatinvolves a cathodic voltammetric peak (reduction reaction) asthe potential is swept negatively, followed by an anodic peak(anodic reaction) as the potential is swept back in the positivedirection. Similar voltammetric features have been obtained forlithium ion-containing non-aqueous electrolytes and thus havebeen attributed to the intercalation of lithium [194,195]:

TiO2 + xLi+ + xe− → LixTiO2. (3.7)

The proton counterpart has the same form and has been argued toalso involve intercalation (see below):

TiO2 + xH+ + xe− → HxTiO2. (3.8)

The latter reaction has been associated with the filling of electrontrap sites, with an elevation of the Fermi level. These changes canbe followed spectroscopically, by an increase in the absorption oflight in thewavelength region from 380 to 600 nm. However, therehas been some controversy involved with this idea. Some workersconclude that the coloring of TiO2 films occurs as a result of thefilling of the conduction band, with the absorption of light excitingelectrons from lower to higher energy levels within the CB. Thisis a more physical view, which has been advanced by Fitzmauriceand others [196–201]. Other workers have concluded that thecoloring process does indeed involve reaction (3.8) and that theabsorption of light involves electronic transitions associated withthe Ti3+ ion. This is a more chemical view, which has beenadvanced by Meyer and co-workers [187]. It has been difficultto conclude which is correct, because the absorption spectrumincludes aspects that can be explained in both ways. Specifically,if the electrons are not trapped at specific sites, the absorptionshould exhibit a steadily increasing absorbance with increasingwavelength, as was observed by Panayotov and Yates, as discussedearlier [164]. This is because there aremany, closely spaced energylevels that are available, with the probability being larger to absorba smaller amount of energy. If, on the other hand, the electronsare trapped at specific, relatively well-defined sites, there shouldbe specific, widely spaced energy levels, which would lead toabsorbance peaks. Cao et al. argue that, since there is a broad peakin the absorbance at ca. 1000 nm, which corresponds to a specificabsorption process for Ti3+, the electrons are essentially trapped atthese sites [187].In situ reflectance measurements have been used by Lyon

and Hupp to follow the variation of the conduction-band energyECB with pH over an extremely wide range for nanocrystallineanatase films [202]. They used solutions with standard Hammettacidity functions H0 and H−, from H0 = −10 to H− = +27.These workers found that Nernstian behavior was observed from

Fig. 3.18. Compendium of Tafel plots for oxygen reduction at various types of TiO2electrodes, with curves at the left in alkaline solution (pH 13 or 14) and curves at theright for acid solution (pH 0): (a) anodized Ti in 1 M NaOH (O2) [203]; (b) anodizedTi in 0.05MH2SO4(O2) [203]; (c) rutile (001) in 1M KOH (air) [192]; (d) rutile (001)in 1MNaOH (air) [204]; (e) anatase film in 0.1MNaOH (O2) [206]; (f) rutile (001) in0.1 M NaOH (O2) [206]; (g) anatase (001) in 0.1 M NaOH (O2) [206]; (h) rutile (001)in 1 M NaOH (O2) [205]. In cases in which air was used (curves c and d), the curveshave been multiplied by a factor of 5 to make themmore directly comparable withthose in which pure O2 was used.

H0 = −8 to H− = +23, i.e., 31 log units of proton activityor pH. Quartz crystal microbalance (QCMB) measurements werealso carried out from H0 = −6 to pH = 12. In fact, the QCMBmeasurementswere consistentwith themass corresponding to theuptake of a proton, or to a deuteron in the case of D2O solution. Itwas argued that only intercalation (Eq. (3.8)), rather than surfaceprotonation–deprotonation (Eq. (3.2)) can account for the variationof ECB with proton activity and mass changes.The electrochemical insertion of protons was used later by

Pelouchova et al. as a means of controlling the doping levelof anatase single crystals and the electrochemical behavior ofthe electrode in the presence of a redox couple, in this case,methylviologen [180]. Specifically, the electrochemical dopingincreased the rate of electron transfer to the oxidized form of themethylviologen. There have been no other results reported thus farfor other redox couples, but similar results could be expected.The behavior of the oxygen reduction reaction (ORR) on TiO2

electrodes as a dark reaction has been examined by several groupsover the past 30 years: Clechet et al. used an anodized Ti electrodeandmademeasurements in both acid and alkaline solutions [203];Parkinson et al. [192], Kesselman et al. [204], Tafalla and Salvador[205], and Tsujiko et al. [206] all used rutile single crystal (001)electrodes, and Tsujiko et al. also made measurements with ananatase (001) single crystal [206]. Most of these studies usedalkaline solution but also examined the behavior at other pHvalues. In Fig. 3.18, all curves except one (b) are shown for alkalinesolution, either pH 13 or 14, while curve b is for pH 0.Clechet measured oxygen reduction of anodized Ti in both

acidic and basic solutions and found a large difference in thebehavior [203] (Fig. 3.18, curves a and b). The most strikingdifference is the large difference in the slope. In basic solution, theinverse of the slope (known as the Tafel slope) is quite close to thatexpected for a one-electron transfer as the rate-determining step(rds), i.e.,

dEd log i

=(ln 10)RTαnF

= 0.1183 V decade-1 (3.9)


Table 3.1Exchange current densities (log i0,A cm−2) and Tafel slopes (V decade−1) for the reduction of O2 to O•−2 at various types of TiO2 electrodes.

Electrode type log i0,A cm−2 Slope, V dec−1 Reference

Anodized Ti in 1 M NaOH (O2) −5.2 0.12 Clechet et al. [203]Anodized Ti in 0.05 M H2SO4(O2) −3.4 0.40 Clechet et al. [203]Rutile (001) in 1 M KOH (air) −5.7 0.12 Parkinson et al. [192]; currents were multiplied by a factor of 5.Rutile (001) in 1 M NaOH (air) −10.2 0.06 Kesselman et al. [204]; currents were multiplied by a factor of 5.Anatase film in 0.1 M NaOH (O2) −5.2 0.12 Tsujiko et al. [206]Rutile (001) in 0.1 M NaOH (O2) −9.3 0.12 Tsujiko et al. [206]Anatase (001) in 0.1 M NaOH (O2) −8.0 0.06 Tsujiko et al. [206]Rutile (001) in 1 M NaOH (O2) −8.4 0.10 Tafalla et al. [205]

where i is the current density in A cm−2, E is the potential,α is the transmission coefficient (usually 0.5), F is the Faradayconstant (96 485 Cmol−1), R is the gas constant (8.314 J K−1mol−1,and T is the absolute temperature. This slope is consistent withthe rds being superoxide production (Eq. (2.11)). The standardpotential for this reaction has been reported as −0.284 V vs. SHE[207], as shown at the bottom of the figure. In contrast, the slopefor acid solution is approximately 0.40 V decade−1. There is nosimple interpretation for slopes larger than 0.118 V decade−1; onepossibility is that there is a thin passive surface layer, which wouldlead to an effective transmission coefficient less than 0.5, in thiscase, 0.148. Tsujiko et al. have discussed this dramatic difference inbehavior between acid and base and have concluded that, in basicsolution, there is an adsorption of O2 on the negatively chargedTiO2 surface [206], whereas adsorption does not take place in acid,due to the protonation of the surface at pH lower than the potentialof zero charge.Tafalla and Salvador proposed that the ORR on rutile takes place

with the involvement of surface states in a manner similar tothat proposed by Vandenmolden et al. for outer-sphere electrontransfer to various metal complexes [205]. The model states thatthe first step is the relatively fast production of surface states,assumed for the moment to be essentially Ti3+ in nature, in anelectrochemical step, written here as an equilibrium:

Ti4+ + e− ↔ Ti3+ (3.10)

followed by a rate-determining step that is a purely chemicalreaction with oxygen:

O2 + Ti3+ → O•−2 + Ti4+. (3.11)

This model should result in a Tafel slope of 0.06 V decade−1,however, not 0.12 V decade−1, as observed experimentally. The0.06 V decade−1 slope results from the fact that since theelectron transfer step is sufficiently fast to be considered to be inequilibrium, its potential dependence is described by the Nernstequation:

E = E0 − (RT/F)(ln([Ti3+]/[Ti4+])). (3.12)

The fact that the electron transfer occurs at potentials that are asmuch as 0.4 V positive of the flat-band potential is not consistentwith purely semiconducting behavior. Instead, it can be reasonablyproposed that the electrode is sufficiently highly doped so that itbehaves like a metal, and the band-edge energies become mobileor ‘‘unpinned’’.However, curve b, for pH 0, follows a much shallower slope,

approximately 400 mV decade−1, i.e., an α value of ca. 0.15. Forthe ORR in base, the first ET would involve the production ofsuperoxide (Eq. (2.11)), which has a standard redox potential E0of−0.284 V [207], indicated on the potential axis of the figure. Thecurves appear to show an onset in the region near this potential,while having no apparent correlation with the flat-band potential,which is ca.−0.83 V vs. SHE (also indicated). The intercepts of thestraight lines with 118 mV slopes with the E0 yield the exchange

Fig. 3.19. Comparison of Tafel plots for oxygen reduction on rutile (001) at variouspH values: (a) pH 14 [204]; (b) pH 10.3 [204]; (c) pH 7.2 [204]; (d) pH 13 [206];(e) pH 11.1 [206]; (f) pH 14 [205]; (g) pH 3 [205]. Curves a, b and c, measured inair, were multiplied by 5 to make them more directly comparable with the othercurves, which were measured in pure O2 .

current densities, which are given in Table 3.1. Based on thesedata, one is impressed by the wide variation in the results forsimilar surfaces (rutile (001)) on the one hand and the similarityof the results for seemingly different surfaces, e.g., rutile (001) andanodized Ti.In comparing results for rutile (001) in various pH solutions

from different groups, we see the variations evenmore graphically(Fig. 3.19). Curves a, b and c show a regular positive shift inpotential with decreasing pH [204], all with similar slopes, whichagain can be approximately fitted to the standard 118 mV slope.Curves f and g, from a different group (Tafalla and Salvador [205])also show a positive shift in going from pH 14 to pH 3, but theshift in comparing curve g for pH 3 with curve c for pH 7.2is unexpectedly small. The behavior obtained by Tsujiko et al.recently showsmuch different behavior: in going frompH13 to pH11.1, there is actually a negative potential shift. One must believethat the surface preparation in the more recent work is superior,given the access to advanced surface characterization techniques.This would explain the much lower absolute currents, presumablydue to a nearly atomically smooth surface. The opposite trend inpH behavior is perplexing, however [206].The explanation of the wide variation in electron transfer

rates from titania electrodes to oxygen to produce superoxide(Eq. (2.11)) in alkaline solution may necessarily involve the bulkelectronic properties close to the surface. The doping level cancertainly vary over many orders of magnitude. From the puresemiconductor standpoint, Pleskov and Myamlin, among others,


have summarized this problem, concluding that the ET rate isproportional to the density of electronic states and the electronenergy distribution function [208]. As we have seen, the dopinglevel can continue to increase above the point at which thebehavior is no longer like that of a semiconductor. In this regime,characterized by the presence of surface states, one could expectthat the ET rate must also continue to rise.

3.8. Photoelectrochemical properties

Although the field of semiconductor photoelectrochemistrywas introduced to a large extent by workers at Bell Laboratoriesas part of their work on semiconductor electrochemistry [165,209–211], the field was developed to a very high level by thework of Gerischer on materials such as Ge, GaAs, CdS and ZnO[212]. Our early efforts were very much influenced by this work.These semiconductors were fundamentally interesting, but thepossibilities for practical applications seemed limited due totheir susceptibility to photocorrosion. Our work [19,50,167,213,214], showed that TiO2 is much less susceptible and could thusbe considered for applications such as solar energy conversion.Interestingly, Boddy, as part of a study on anodic oxygen evolution,had already just published a photocurrent vs. potential curve(similar to Fig. 2.2) when our studies appeared, but the fullimplications of the result were not recognized [165].The photoelectrochemistry of TiO2 can be summarized by

referring to Fig. 3.20. It can be seen that the conduction-bandenergy ECB of rutile is essentially coincident with the reversiblehydrogen potential at all pH values, whereas that for anatase ismore negative by 0.20 V. At lower pH values, the ECB for rutileis also coincident with the reversible potential for O2 reductionto hydroperoxyl radical HO2•, but, at higher pH, ECB continues tobecome more negative, while the potential for O2 reduction tosuperoxide radical anion O•−2 remains constant at−0.284 V [207].This implies that this process should becomemore favorable in thealkaline pH region, as already discussed in the previous section.The fact that the ECB for anatase is sufficiently negative to reduceO2 to O•−2 over a wide pH range is consistent with its higherphotocatalytic activity. The valence-band energies EVB for bothrutile and anatase lie at approximately the same potential, which issufficiently positive to produce free •OH radicals (lower pH) or O•

−

radicals (higher pH). However, it is only positive enough to oxidize2-propanol to its radical cation, in principle, in the acidic pH region.The EVB for both materials is more than sufficiently positive

to oxidize water, by approximately 1.8 V, at all pH values. Thepotentials for water oxidation are also much less positive thanthose for the reactions involved in photocatalysis, for example,the production of •OH radicals and the direct hole-mediatedoxidation of organic compounds. Nevertheless, some of the samereactions and intermediatesmay also be involved. Variousworkershave studied the detailed mechanisms of oxygen photoevolutionreaction (OPER) on TiO2 [111,215–221]. The originally proposedmechanism involved the reaction of photogenerated holes withadsorbed hydroxyl groups (OH−) to produce •OH radicals, whichthen coupled to form H2O2, which was further oxidized to O2[215,216]. In order to explain the photo-etching of TiO2, Nakatoand co-workers have proposed alternative pathways involving thereaction of photogenerated holes with bridging Ti–O–Ti moieties,with breakage of the Ti–O bond [111,217–220] (see Section 4.1.4).Recently, Neumann et al. have proposed a further modificationof this model [221]. Qu et al. have carried out quantum chemicalcalculations related to these mechanisms; these workers foundlarge energy barriers to the diffusion of O-adatoms and thusquestioned whether O2 could be evolved as a result of therecombination of O-adatoms [222].

Fig. 3.20. Energy bands on TiO2 as a function of pH. Note that the rutile CB line issuperimposed on the reversible hydrogen potential H+/H2 at all pH.

The photoelectrochemical oxidation of organic compoundshas been studied extensively. In a recent example, Villarealet al. studied the photoelectrochemistry of polycrystalline anataseelectrodes as a model system for photocatalysis and found thatit was possible to distinguish between direct hole-mediatedoxidation and •OH radical-mediated oxidation [223]. For example,methanol, which is less strongly adsorbed, was found to beoxidized via an •OH radical, while formic acid, which is morestrongly adsorbed, was oxidized directly.

4. Fundamentals of photocatalysis

4.1. Mechanisms of photocatalysis

4.1.1. Photoelectrochemical basis of photocatalysisAs described in the Historical Overview, it became recognized

by several researchers that photocatalysis is based on ‘‘back-to-back’’ or short-circuited photoelectrochemical and electrochem-ical reactions, involving electrogenerated electrons and holes. Atthe most global level, these can be written:

hν → e−CB + h+

VB (4.1)

2H2O+ 4h+VB → O2 + 4H+ (4.2)

O2 + 4H+ + 4e−CB → 2H2O. (4.3)

Reactions (4.2) and (4.3) can be designated as the oxygenphotoevolution reaction (OPER) and the oxygen reduction reaction(ORR), respectively. Of course, the latter can occur either, asshown, on the titania surface itself or on a separate electrode.These processes can be examined separately with standardelectrochemical methods and can provide an overall rationale forthe energetics, as seen in Fig. 4.1. The point atwhich the anodic andcathodic (flipped up to view the crossing point) currents are equalis a good indication of the local current that could be expectedunder a given set of conditions. We first showed such crossing-point diagrams in a paper by Inoue et al. to show the short-circuit current for PEC cells [224]. The basic idea has been usedfor many years to quantitatively analyze the energetics of localelectrochemical cells on corroding metals [47,48].The set of results in Fig. 4.2, from Lewis and Hoffman and co-

workers [204] shows that, in 1 M NaOH, the crossing point for the


Fig. 4.1. Energy band diagram showing a photoelectrochemical cell, at open circuit,in which oxygen photoevolution is occurring at a rutile electrode and oxygenreduction is occurring at a platinum electrode. This shows the maximum opencircuit potential that could be expected with a highly active catalyst.

Fig. 4.2. Current–potential curves for (a) oxygen photoevolution on a rutile singlecrystal; (b) hydrogen evolution on the same rutile electrode, where the cathodiccurrent is flipped up to show the point at which it crosses curve A; (c) hydrogenevolution on a roughened platinum electrode, cathodic current and flipped up toshow the crossing point; and (d) oxygen reduction on the roughened platinumelectrode, cathodic and [187] flipped up to show the crossing point (based onKesselman et al. [204]).

OPER on the illuminated rutile TiO2 single crystal surface with theORR also on TiO2 is at ca. 2 µA cm−2, or ca. 25% of the maximumphotocurrent available (Fig. 4.2, curves+B and−B, the latter beingflipped up to show the crossing point). Interestingly, if we showthe hydrogen evolution reaction (HER) on a platinum electrode(curve+C and−C), which is similar to a situation in which a TiO2particle is platinized, but with no oxygen to reduce, the cross-overis slightly higher, ca. 3 µA cm−2, or ca. 37.5% of the maximum.This situation is also essentially the same as that in a complete PECcell and is consistent with our original work on the photo-assistedelectrolysis of water; this was reported for the 1 M NaOH case in1975 [55], as well as with measurements of the flat-band potential[167,169,171].If we now look at what happens when we couple the OPER on

rutile TiO2 with the ORR on platinum, i.e., similar to the situationon the platinized TiO2 particle in the presence of dissolved oxygen,we can see that the crossing point is now near the maximum(ca. 7.5 µA cm−2). This idea was first proposed by Bard and co-workers [225] and has prompted several groups of researchers,principally Heller and Gerischer and co-workers [226], to enhance

the photocatalytic process by adding catalysts for the ORR, such asPt and Pd, to TiO2. Kesselman et al. noted that the extension of theirwork to the case of real photocatalysts was not a simple one, dueto differences, for example, between rutile and anatase [204]. Thisis an important point, which will be taken up again later.This picture can then be related directly to the energy-level

diagram, in which the Fermi levels for the photo-anode (reaction(4.2)) and cathode (reaction (4.3)) are shown as being equal (atthe crossing point in Fig. 4.2), which they must be, in a short-circuited PEC cell (Fig. 4.1). This is helpful in understanding theenergetics and demonstrates that the reactions are expected toproceed spontaneously. Of course, we are ignoring the detailedmechanisms at the moment and will return to these later.Surprisingly, the photocurrent at a single crystal rutile electrode

(Fig. 4.2, curve A), which can be totally consumed in generatingO2, is not significantly affected by the presence of an organiccompound, as shown by Izumi et al. [227] and again by Kesselmanet al. [204]. This is because the photocurrent is essentiallydetermined by the characteristics of the semiconductor. Atincreasingly positive potentials (low Fermi levels), the bandbending becomes sufficient for the photogenerated electrons andholes to be efficiently separated, while, at increasingly negativepotentials, the band bending continues to decrease to the flat-bandcondition, and then, contrary to the expected behavior for an idealdiode (Fig. 2.2, curve 1–3), in the absence of oxygen, there is agap of several tenths of a volt before cathodic current starts. Forthe moment, we are also ignoring the fact that there is actuallysome transient cathodic current as one sweeps the potential in thenegative direction [204], which could be due to the intercalation ofprotons into the TiO2 structure (see later).A series of papers were published in which a microelectrode

was placed at various distances above an illuminated anatase film,whichwas deposited on an indium–tin oxide (ITO) substrate [228–233].Measurements thatwere carried outwith themicroelectrodetechnique showed that, even when an anatase film is activelyoxidizing an electron donor (in this case, ferrocyanide) and evenwhile there is oxygen reduction occurring at a bare section of theITO substrate, there is still a quite significant reduction currentpassing at the anatase surface (Fig. 4.3) [232].At this point, it is possible to propose a tentative working

mechanism for photocatalysis in terms of the balancing of anodicand cathodic reactions. The anodic reaction can be oxygenevolution, hydroxyl radical production or organic oxidation. Thecathodic reaction can either be oxygen reduction, to superoxide,hydrogen peroxide or water, or reduction of the titanium dioxideitself, via hydrogen insertion, for example. On a polycrystallinefilm, certain crystal faces may be favored as reduction sites. Thisis not certain at the present stage of understanding. Nevertheless,it is certain that the oxygen reduction consumes protons, therebymaking the adsorbed water alkaline, as long as there are somecations present. We already know that the potentials at whichORR proceeds are more negative in alkaline solution. Thus, thereaction becomes progressively more difficult. However, there aretwo mitigating factors: (1) the presence of CO2 in the atmosphereand its absorption could buffer the pH; and (2) O2 reduction occursmore easily in alkaline solution, i.e., larger changes in current perunit potential change compared to acid, as already mentioned inSection 3.8.There are two other factors that could help in offsetting the

inhibiting effects of pH changes. First, the trapping of electronsin restricted areas of the TiO2 surface and consequent increasein the local electrical conductivity can lead to an increase in theability of that area to transfer electrons to solution phase species.If the concentration of trapped electrons becomes high enough,the local area can become like a metal. At that point, the fact thatthe conduction band edge should shift to more negative potentials


Fig. 4.3. (a) Schematic diagrams of the microelectrode monitoring system, with the microelectode situated far from the surface (left) and close to the surface (right), whereit can sense the oxygen reduction reaction that is occurring in parallel with ferrocyanide oxidation; (b) current vs. time curves showing the current when the electrode is105 µm from the electrode (left) and when it is 15 µm from the surface; the concentration of ferrocyanide is 5 mM; the microelectrode diameter is 15 µm; supportingelectrolyte, 0.1 M K2SO4; UV light intensity, 8.5 mW cm−2 , electrode potential,+0.70 V vs. SCE (taken from Maeda [232]).© 1999, American Chemical Society.

due to a pH effect becomes meaningless, and the electrons are notrequired to pass over such an energy barrier (Fig. 4.4). At neutralpH, the conduction band energy for anatase should be ca. −0.6 Vvs. SHE, and the Fermi level should be ca. −0.4 V vs. SHE. This isjust negative enough to reduce O2 to O•−2 (E0 = −0.284 V vs.SHE) [207]. We also note here that, for rutile, the energetics arenot as favorable, i.e., the conduction-band energy would be ca.−0.4 V and the Fermi level ca.−0.2 V, which is not quite negativeenough to reduce O2, as discussed in Section 3.7. If the local areain which the electrons are trapped is relatively small and theremaining area, in which oxidation takes place is relatively large,the anodic current density in this remaining area will be small. Insuch a case, the pH change will be smaller than expected, i.e., theconduction-band lowering due to acidification will be minimized,and thus there may be minimal enhancement of the tendency toaccumulate electrons at the surface (see Fig. 4.5). Now, we have apossible scenario inwhich the effect of the natural pH changes thatcould occur in a weakly buffered solution is largely avoided, andit becomes energetically possible to reduce oxygen to superoxidein areas that cover a small fraction of the surface while oxidationof organic molecules or other reactions (e.g., peroxide or oxygenproduction) can occur in areas that cover a large fraction of thesurface.

This model is basically an extension or modification of thatof Gerischer and Heller [226,234–236], in which relatively smallareas of TiO2 particle surfaces are proposed to act as reducingsites. The novel aspect of the present modification is that itprovides a specific mechanism by which small reducing centerscould develop, i.e., more or less randomly selected portions ofthe TiO2 surface are proposed to provide starting points forconductive regions to begin to develop within the bulk of theparticle or film. This can occur as a result of electrochemicalhydrogen insertion (Fig. 4.5). As illumination proceeds, and moreelectrons are produced, the latter can be partially consumed in thegeneration of such bulk conductive regions. The reason that suchstructures would be spontaneously generated is that the electronswould be transported to the surface more efficiently in a moreconductive region (‘‘wire’’). In other words, the first electron thatgoes to the surface and produces an interstitial hydrogen atommakes it more likely that a second electron might be directed tothe same area. The probability depends on the spatial distributionof the electron density for this first bulk trapped electron. Thiseffect could explain why only relatively small surface regionsmight exist. Furthermore, if the electrons are channeled in thisway, they not only avoid electrostatic repulsion, whichwould tendto disperse them, but also localize them and thus protect them


Fig. 4.4. Modified band diagram for photocatalysis, showing that the cathodicreduction of O2 takes place through surface states, with no impediment from theincreasingly alkaline pH at these sites.

from recombination. These processes are being considered morequantitatively and will be reported elsewhere.The insertion of hydrogen atoms would tend to increase

the volume, as already mentioned in Section 3.6, as confirmedtheoretically by Koudriachova et al. [160]. This could help inexplaining results that have been reported recently involvingthe effects of compressive vs. tensile stress and others involvingchanges in volume and roughness (see Section 5).Several specific cases can be envisioned (Fig. 4.5). First, as a

reference case, let us say that only water, either liquid or vapor,is present at the surface (Fig. 4.5a). A photoproduced electron isshown hitting a hydrogen-rich region and being conducted alongit to the surface, where it reacts with water, producing anotherinterstitial hydrogen atom, which is essentially a kind of trappedelectron.If a hole trapping agent such as an alcohol is present at the

surface, the holes will be more effectively trapped, and there willbe less recombination (Fig. 4.5c). In this case, the photoproducedelectrons will have a greater opportunity to react and produceinterstitial hydrogen. If this process continues indefinitely, a largeportion of the TiO2 particle, or even macroscopic solid, can bereduced, so that the blue color produced by the presence of Ti3+ isvisible. This explains the first early report of Renz inwhich is notedthe discoloration of TiO2 samples exposed to organic compoundsand sunlight [17].When oxygen is present aswell, it can be reduced to superoxide

or hydrogen peroxide by the photoproduced electrons. In this case,conventional photocatalysis can take place, since the superoxidecan participate in the photocatalytic reactions (Fig. 4.5d). Whenoxygen is present but no organic compound (Fig. 4.5c), thesituation becomes like that for the photo-induced hydrophiliceffect (see Section 5). The superoxide produced by oxygenreduction and the hydroxyl radicals produced by water oxidationcan both react further, the superoxide via disproportionation, andthe hydroxyl radicals likewise via disproportionation, to produce asurface that has a high coverage of hydrogen peroxide. In this case,again, the hole trapping process is not perfect, and recombinationcan occur, and also some electrons are removed via oxygen

reduction, so the H-rich region does not have an opportunityto grow.The scenario described here could be considered to be an

example of a simple self-organizing system, i.e., a system thatspontaneously organizes itself into an array (in this case, two-dimensional) of two different types of regions with distinctproperties. There are a number of examples of such systemsin chemistry [237]. The ideal system does not assume any pre-existing heterogeneity; this ideal could be approached on a singlecrystal surface. It is quite likely that the pattern formation thatwe observed on a rutile single crystal surface (see Section 5) wasof this type, in which cathodic and anodic regions might haveformed spontaneously. However, on a perfectly homogeneoussurface, the pattern formation may not be robust, and the regionsmight tend to move around, based on statistical fluctuations. Ina real photocatalytic film, heterogeneity does in fact exist andcan affect the pattern of regions that might develop. The processof formation of small conductive regions might involve a kind ofnucleation and subsequent growth, as in the deposition of crystalsor electrodeposition ofmetals. If the nucleation is based on specificconditions that tend to favor it, for example, a particular crystalface or the presence of trace impurities, the pattern could be quiterobust. This is an area that will be developed further.As mentioned above, this is basically a modification of the

Gerischer–Heller model, in which relatively small areas on thesurface trap electrons, which subsequently react with oxygento produce superoxide [226,234–236]. These authors envisionedsingle, stationary surface traps on titania nanoparticles. They arguethat the rate of electron transfer from surface traps to solution-phase oxygen is significantly below the rate at which holes areproduced, and therefore, the number of electrons builds up withinthe particle to levels at which most of the holes are consumed byrecombination. They argue further that, in order to accelerate theoxygen reduction reaction, electrocatalysts such as Pd or Pt couldbe used.The merits of this approach were also demonstrated by

Kesselman et al. in their work with rutile single crystal electrode,inwhich they showed that the oxygen reduction current–potentialbehavior on a Pt-covered rutile electrode surface was indeedincreased to the point where it matched the anodic photocurrent[204]. Aside from the practical problems of this approach, oneof the fundamental chemical problems is that it ignores theneed to produce superoxide as a reaction intermediate in thephotocatalytic process. In fact, Heller and co-workers [226,234–236] and other groups, including our own, have proposedmechanisms for catalysis in which superoxide plays a key role.We propose here that oxygen reduction to superoxide and

hydrogen peroxide can occur at TiO2 at significant rates undercertain conditions, specifically, those under which small near-surface regions become sufficiently highly doped that the electrontransfer approaches the limit predicted by the Marcus theory. Thiscan involve doping with either oxygen vacancies, as in the case ofbulk electrodes, or it can involve hydrogen doping, which can, inprinciple, be induced electrochemically, as already discussed.In summary, this slightly modified Gerischer–Heller model

takes into account the requirement of electroneutrality, i.e., thatboth oxidation and reduction must take place at exactly equalrates. It explains how oxygen reduction can proceed even atsignificant rates without an added catalyst, and even though thelocal pH might tend to slow down the reaction. It also explainshow photogenerated electrons can be channeled to the surfaceefficiently, avoiding recombination with holes. Finally, it helpsin explaining changes in bulk lattice parameters, which wouldotherwise be difficult to explain.


Fig. 4.5. Combined reductive–oxidative model for the photo-induced hydrophilic (PIH) effect, in which reducing and oxidizing regions are spontaneously formed: (a) inthe absence of oxygen, protons are reduced to produce interstitial hydrogen atoms (no organic compound present). (b) in the presence of oxygen, the latter is reduced tosuperoxide or hydrogen peroxide (no organic compound present); (c) in the absence of oxygen, protons are again reduced, while an organic compound is oxidized; (d) inthe presence of oxygen, the latter is again reduced, while an organic compound is oxidized. In (a) and (b), holes react with water to produce a variety of possible hydrophilicmoieties.

Fig. 4.6. Processes occurring on bare TiO2 particle after UV excitation.

4.1.2. Time scalesWhen a TiO2 photocatalyst absorbs a photon with energy equal

to or greater than its band-gap, an electron–hole pair is generated;subsequently, the pair is separated into a free electron and a free

hole. The electron and hole ‘‘walk randomly’’ to the surface ofthe photocatalyst and are trapped there (Fig. 4.6); the trappedelectron e−tr and hole h

+

tr react with acceptor (step c) or donormolecules (step d), respectively, or recombine at surface trapping


Table 4.1Some measured characteristic time for primary processes in TiO2 photocatalysis.

Primary process Characteristic time Sample description References

Charge carrier generationTiO2 + hν → e− + h+ fsCharge trappingh+ → h+tr <200 fs

Nanoporous TiO2 film Tamaki et al. [262]e− → e−tr <150 fsh+s−tr → h+d−tr (relaxation) ∼100 pse−s−tr → e−s−tr (relaxation) ∼500 psCharge recombinatione− + h+tr , h+ + e

−

tr , or e− + h+ 1 µs Nanoporous TiO2 film (water) Yoshihara et al. [246]→ heat (or hν) 25 µs Nanoporous TiO2 film (air) Peiro et al. [238]

Interfacial charge transfer∼300 ps (methanol oxidation) Nanoporous TiO2 film Tamaki et al. [264]∼3 ns (2-propanol oxidation) Nanoporous TiO2 film Tamaki et al. [264]

h+ (or h+tr )+ Red→Red+<2 µs (water oxidation) Degussa P25 powder Yamakata et al. [256]No water oxidation within 80 µs Nanoporous TiO2 film Murai et al. [609]

e−tr + O2 → O−2 <100 ns Nanoporous TiO2 film Yoshihara et al. [246]e− + O2 → O−2 10–100 µs Degussa P25 powder Yamakata et al. [256]e− (or e−tr )+O2 → O−2 ∼10 µs Nanoporous TiO2 film Peiro et al. [238]e− + Pt→ Pt · · · e− 2.3 ps Degussa P25 powder Iwata et al. [265]

Fig. 4.7. Time scales in photocatalysis.

sites (step a). The electron and hole can also be trapped at bulktrapping sites and recombine there with the release of heat (b).The processes of back-reaction between the oxidized donor andelectron, or reduced acceptor and hole, can occur after the initialcharge transfer, especially when the species are strongly adsorbedon the TiO2 surface [238–240]. These processes together with thecharacteristic surface reactions, as illustrated in Fig. 4.6, occur atthe time scales shown in Fig. 4.7 and listed in Table 4.1. Since theband-gap transition is of an indirect nature, deexcitation throughlight emission is not an important process for TiO2 [6,74].The photogenerated charge carriers, either free or trapped,

can be probed by various spectroscopic techniques. Trappedholes and electrons absorb light in the visible and near-infraredspectral regions [238,241–253], whereas free electrons absorb inthe infrared or microwave regions [254–260]. These absorptionsallow one to accurately measure the time scale of photocatalyticprocesses by means of transient spectroscopies, such as transientabsorption (TA) spectroscopy [238,243–249,252,253], transientdiffuse reflectance (TDR) spectroscopy [242,250,251,255,256,258],and time-resolvedmicrowave conductivity (TMRC)measurements[254,259,260]. Typical spectra for these species are shown inFig. 4.8.

Fig. 4.8. Absorption spectra of (a) trapped holes in Refs. [252,253,261], and(b) trapped electrons in Refs. [243,245,252,253]. (c) The spectra of trapped holes,free electrons, and trapped electrons identified by Yoshihara et al. [246].© 2004, American Chemical Society.


In an early study, Serpone et al. compared the TA results of threeTiO2 sols excited by a 30-ps-width UV laser, with particle sizes of2.1, 13.3, and 26.7 nm [247]. They observed that for all three kindsof sols, the transient spectra were fully developed and comprisedspectra of trapped holes and trapped electrons at the end of the30 ps laser pulse. Thus, they estimated that the trapping timesfor electrons and holes should be in the ≤ 1–10 ps range. Later,Yang et al., by means of femtosecond TA spectroscopy, estimatedthe hole trapping within 50 fs and electron trapping within 260 fsfor TiO2 particles of several-nm diameter [261]. The ultrafast holetrapping was interpreted in terms of the difference in the effectivemass of hole (0.8me) and electron (>10me) that results in a 40times faster transit time from the interior to the surface of the TiO2nanoparticle for the hole than that for the electron. Very recently,Tamaki et al. studied the trapping dynamics of electrons andholes in a nanoporous TiO2 film by femtosecond TA spectroscopy[262,263]. They observed that shallowly trapped electrons andholes were generated at a time constant of 100 fs, deeply trappedelectrons at 150 fs, and deeply trapped holes at 200 fs. In addition,they assigned the time constant of 100 ps to trapped holes relaxingfrom shallow traps to deep ones, and the time constant of 500 ps totrapped electrons relaxing from surface shallow traps to bulk deeptraps. These results suggest that in TiO2 photocatalysis the speciesthat lead to surface redox chemistry should be the trapped, ratherthan free, charge carriers.Yoshihara et al. studied the transient adsorption of TiO2

nanoporous film in a wide wavelength range (400–2500 nm)[246]. They used a low-intensity pulsed laser to excite a TiO2film so that less than one electron–hole pair was generated ineach particle, on average, per pulse; this ensured them obtaininginformation of the transient species under conditions close tothose in a real photocatalytic system, i.e., with UV light below the1 mW cm−2 level. They found that, for a nanocrystalline TiO2 film,the half-life for electron–hole recombination is ca. 1 µs for a filmdipped in N2-saturated deuterated water. This result is in contrastto the fast electron–hole recombination (∼10 ns) suggested bythe earlier transient spectroscopic studies on TiO2 colloids andpowders in which high power laser pulses are used [6,259,260],but is consistent with recent studies by Durrant and co-workers[238] and Yamakata et al. [256], in which similar low-intensitylaser pulses were employed for excitation.Yoshihara et al. identified the characteristic absorption of

trapped and free electrons in nanoporous TiO2 films: trappedelectrons showed a broad absorption centered at 770 nm, whereasfree electrons showed a structureless absorption in the near-infrared and infrared region whose intensity increased withwavelength (λ) by a function of λn [246]. They found that trappedelectrons reacted with oxygen more rapidly than free electrons:the absorption of trapped electrons decreased clearly at 100 nsafter excitation when the TiO2 film was immersed in air-saturatedmethanol, whereas no change was observed for free electrons atthe same time scale.Durrant and co-workers studied the photochemical reduction

of oxygen adsorbed on nanoporous TiO2 film by means of TAspectroscopy [238]. They observed that in a nitrogen atmosphere,the TA signal probed at 900 nm decayed with a half-life ofabout 25 µs. The signal, however, exhibited a decay with ahalf-life of about 12 µs in air, indicative of oxygen reductioncompeting directlywith the electron–hole recombination reaction.Yamakata et al. studied the water-induced and oxygen-induceddecay kinetics of free electrons in TiO2 by means of time-resolvedinfrared absorption spectra [256]. They observed that on TiO2(Degussa P25), O2 from the gas-phase captured the electrons andaccelerated the decay rate at a delay time of 10–100µs, and watervapor reacted with holes within 2 µs.Tamaki et al. studied the oxidation of adsorbed alcohols with

trapped holes by means of femtosecond TA spectroscopy (Fig. 4.9)

Fig. 4.9. (a) Absorption spectra of TiO2 nanocrystalline film in the ground state andof trapped hole in the film. (b) Time profiles of transient absorption of TiO2 in air,water, methanol, ethanol, and 2-propanol at 400 nm [264].© 2006, American Chemical Society.

[264]. They recorded the time profile at 400 nm after a laserpulse (160 fs pulse width); at this wavelength, deeply trappedholes contributed 90% of the absorption, while trapped electronscontributed another 10%. In the case of the TiO2 nanoporous film inair, where there is no scavenger for trapped holes, the time profileshowed an additional rise in TA up to 20 ps as aminor contributionafter the instantaneous rise within the time-resolution (200 fs) ofinstrument. They attributed this additional rise to the relaxationfrom shallowly trapped holes to deeply trapped holes. A similarrise was observed for a film immersed in water. The TA did notshow a decay up to 1 ns, either in air or in water, suggestingthat electron–hole recombination and the oxidation of water arenegligible on this time scale. However, in alcohol media, the TAof trapped holes decayed after excitation, clearly indicating thattrapped holes reacted with the alcohols, and the reaction wascompetitive with the trapping of holes. The lifetimes of trappedholes in the films in methanol, ethanol, and 2-propanol were300, 1000, and 3000 ps, respectively; the reaction efficiencieswere close to unity. The decay kinetics of the trapped holes thatreact with alcohols is monoexponential on the nanosecond order,different from the non-exponential recombination kinetics for thefilm in air. This suggests that the hole transfer from trapping sitesto alcohol molecules is the rate-limiting step for the former.There were several studies that dealt with the electron transfer

within Pt/TiO2 photocatalysts. In a femtosecond TDR study, Furubeet al. observed a new decay component of a few ps under 390 nmexcitation in addition to the normal charge recombination kineticsfor Pt/TiO2 photocatalyst; the larger the Pt loading on the TiO2surface, the more efficient the electron migration observed [241].They thus assigned the fast decay process to the migration ofelectrons in TiO2 to Pt. Iwata et al. measured the femtosecond TAspectra of Pt-loaded TiO2 in the spectral region of 0.9–1.5µm. Theyobserved a decay process of 2.3 ps, which represents the transferof electrons from TiO2 to Pt. Despite this fast electron transfer,Iwata et al. suggested that a fraction of the electrons remain in theTiO2 particle, since considerable signal intensity remains even afterhundreds of ps [265]. Yamakata et al. traced the absorption of freeelectrons in TiO2 and Pt/TiO2 particles after a 10-ns-width laserpulse by means of time-resolved IR absorption spectra [256]. Theyobserved that the absorbance on Pt/TiO2 decreased more quicklyin the 0–1 µs range than that on TiO2. The rapid decay on Pt/TiO2suggests that the migration of electrons to Pt clusters occurredwithin 1 µs, and the electrons were equilibrated among the Ptclusters and TiO2 substrates after 1 µs. The three studies coincidewith each other on the fact of fast electron transfer from TiO2 to Ptclusters.The above-mentioned studies focused on TiO2 nanoparticles

and nanoporous films. Further studies need to be conducted onpolycrystalline TiO2 films, which are also important photocatalyst


materials. However, their results are still valuable for us todetermine some important time scales in TiO2 photocatalysis,as summarized in Fig. 4.7 and Table 4.1. These time scales,as summarized in the table, suggest that: (1) the chargerecombination in TiO2 photocatalysts could be a slow processdue to the efficient charge trapping; (2) the interfacial chargetransfer processes are competitive to charge recombination; (3)photocatalytic reactions could be completed within µs to msafter the generation of the electron–hole pair. The last pointsuggests that in the photocatalytic processes under low-intensityUV illumination (1 mW cm−2,∼1015 photons s−1 cm−2), a photonis absorbed by the TiO2 photocatalyst only after the electron–holepair generated by the previously absorbed photon has alreadyundergone reaction or charge recombination.

4.1.3. Trapping of electrons and holesCharge carrier trapping is needed in the photocatalysis

process to suppress recombination and increase the probabilityof interfacial charge transfer [6,74]. As mentioned before, therecombination can be delayed, with a lifetime on the scale ofµs in nanocrystalline TiO2 films [238,246]. This makes the slowcharge transfer of electrons to molecular oxygen competitive withthe recombination. The electron paramagnetic resonance (EPR)technique is frequently used to trace trapped electrons and holesin TiO2 nanoparticles after UV illumination [266–271]. At lowtemperature, trapped electrons can be detected in the form ofTi3+. Adsorption of O2 at surface has been found to remove theTi3+ signal. Trapped holes are also observed and are consideredto be localized deep trap states. However, there is no consensuson the exact nature of the state: several types of species, suchas sub-surface oxygen radicals connected to surface hydroxyls(Ti–O•–Ti–OH) [266,271] surface oxygen radicals (Ti–O–Ti–O•)generated from basic surface hydroxyls [267,270], lattice O•

−

radical, etc. [266,268] are postulated to exist in various cases. In arecent EPR study by Berger et al., a lattice O•

−

radical was observedon the surface of dehydrated 13-nm TiO2 nanocrystals; the signalof O•

−

was completely removed by the addition of a large amountof O2 due to the exchange effect between the O2 molecule and theO•−

species [268].The energetics of electron trap states of various TiO2 materi-

als have been evaluated by a variety of experimental methodolo-gies. For example, for single crystal rutile, trap sites have beenshown by thermoluminescence and thermally stimulated currentmeasurements to exist 0.21–0.87 eV below the conduction-bandedge [272–274]. For nanocrystalline TiO2 film, trap states at about0.5–0.7 eV below the conduction-band edge have been reported[197,201,275]. By time-resolved photoacoustic spectroscopy, elec-tron trapping states in nanocrystalline colloidal titanium dioxidewere determined to exist on average 0.8 eV below the conduction-band edge [276]. Photoluminescence measurements on anatasenanoparticles revealed the presence of four shallow electron trap-ping levels at energies of 0.4, 0.5, 0.64 and 0.86 eV below theconduction-band edge, ascribed to the presence of oxygen vacan-cies [277,278]. Rutile particles show almost the same photolumi-nescence from shallow traps as anatase particles, indicating thesame origin of shallow traps, i.e., the presence of oxygen vacanciesin rutile as well as anatase particles.The energetics of deeply trapped holes has been evaluated

indirectly by several methods. Lawless et al. studied the reactionof TiO2 particles with •OH radicals that were generated by pulseradiolysis [249]. The product of the •OH reaction with the particleswas identified as a deeply trapped hole at the particle surface,whose redox level was approximately 1.3 eV above the valenceband. Nakamura et al. observed a photoluminescence centered at840 nm from the atomically smooth TiO2 (100) surface, ca. 1.5 eV

Fig. 4.10. Schematic diagram of the reaction of trimethylacetic acid (TMA) on arutile (110) single crystal via adsorption at two adjacent 5-coordinate Ti4+ sites,with simultaneous reduction of 6-coordinate Ti4+ sites to Ti3+ and subsequent re-oxidation by O2 to Ti4+ [280].© 2003, American Chemical Society.

above the valence-band edge, and assigned it to the recombinationof conduction-band electronswith deeply trapped holes [219,220].No energetics information for shallowly trapped holes is available,probably due to their short lifetime.Two different types of surface Ti–OH groups have been

proposed to play important roles in trapping charge carriers,forming Ti4+–•OH radicals (hole trapping) or Ti3+–OH groups(electron trapping) [257]. This point has been investigated bothexperimentally and theoretically recently. In an IR spectroscopicstudy, Panayotov et al. observed the depletion of conduction-band electrons in reduced TiO2 powders with water adsorption[279]. Compared with the oxidized TiO2 sample, the reducedsample shows an increase in the background absorbance due to theproduction of free conduction-band electrons. Introducing watervapor into the IR system caused a decrease of the backgroundabsorption. This apparent electron withdrawal was believed to bethe result of the formation of surface hydroxyls by dissociativeabsorption of water on defects on the TiO2 surface. This is an issuethat clearly merits further study.In another study, Henderson et al. sought to identify surface

sites associated with charge transfer and trapping during thephotodecomposition of a trimethyl acetate (TMA) adlayer on therutile (110) surface, using scanning tunneling microscopy (STM),electron energy loss spectroscopy (EELS), and photodesorption(Fig. 4.10) [280]. In ultra-high-vacuum (UHV) conditions, theyobserved the decomposition of a TMA molecule that bridged twoTi4+ sites into isobutene and CO2, with the accumulation of Ti3+cations bound to surface bridging OH groups. The Ti3+ signal,detected as a feature at about 0.8 eV in the EELS spectra,was largelyremoved by postexposure to O2.By means of density functional theory (DFT) calculations on

hydroxylated and reduced rutile TiO2 (110) surfaces, Di Valentinet al. [281] demonstrated the electron trapping nature of bridgingTi–OH groups, which is consistent with Henderson’s study [280],but found no evidence that these groups should also act as holetraps by formation of Ti4+–•OH radicals. As calculated, the oxygenvacancy introduced two localized Ti3+ 3d1 states about 1 eVbelow the conduction-band edge (Fig. 4.11). These states were notremoved upon dissociation of a water molecule and formationof a pair of hydroxyl groups. Instead, the excess electrons ofthe hydroxylated surface were trapped by two Ti4+ ions, whichwere reduced to Ti3+. These two ions were found to be of twotypes: a six-fold-coordinated Ti located between the two bridginghydroxyls, with an energy 1.6 eV below the conduction band(3.4 eV band-gap calculated) denoted as Ti3+br−OH, and a nearby five-fold-coordinated Ti, with an energy 1.2 eV below the conductionband, denoted as Ti3+5c . The O 2p states of the bridging OHgroups were spread within the O 2p band, providing very little


Fig. 4.11. Total and projected density of states for the hydroxylated (top) andreduced (bottom) TiO2 (110) surface, calculated using the B3LYP hybrid functional.The Ti3+ states are localized on: (a) the Ti ion between the two bridging OH groups,Ti3+br−OH; (d) the Ti ion nearest to the oxygen vacancy, Ti

3+br−v; (b),©Ti

3+5c , top view.

The vertical dotted line in the PDOS denotes the position of the Fermi energy [281].© 2006, The American Physical Society.

contribution to the top of the band. This suggests that theseOH groups are not very efficient hole traps. On the contrary,the Ti atom of the Ti–OH group is a good electron trap, as thecorresponding energy level for Ti3+br−OH was found 1.6 eV below theCB by calculation. Once the electron is trapped, the Ti3+–OH canattract and neutralize a hole, i.e., behave as a recombination center.These studies have provided solid evidence for the contribution

of surface Ti–OH groups to electron trapping. For the stoichiomet-ric (unreduced), hydrated TiO2 surface, there are at least two dif-ferent types of surface Ti–OH groups: one is the bridging Ti–OH(actually, considered to be adsorbed protons, i.e., Ti–OH+–Ti, asdiscussed in Section 5.2.4), and the other is the terminal Ti–OH,(considered simply as OH− ions that are adsorbed at 5-coordinateTi4+ sites, as discussed in Section 5.2.4, or, alternatively, as theproduct of the dissociative adsorption of water at these sites (seealso Section 3.4) [282]. The former is more acidic than the latter,which leads to different reactivities. One can consider two path-ways for electron trapping at bridging oxygens (given formally asEq. (3.2)): either the electron is trapped first and the proton trans-fer follows, or vice versa. As outlined above, there is a consensusthat the neutral bridging Ti–OH group is the final product, with theelectron itself trapped at the 6-coordinate Ti4+ site. There is alsoevidence that some electron density exists at the 5-coordinate Ti4+sites (see Sections 3.5 and 4.1.5), since both superoxide and hydro-gen peroxide can adsorb at these sites.More controversial is the role of the more basic terminal

Ti–OH− group, which has been postulated to trap holes bysome workers, but this has been disputed by others (seeSection 4.1.4). At this point, we should note that the role ofhydroxyl groups in general can become complicated, because,

Fig. 4.12. Correlation of O− and Ti3+ EPR signal intensity as a function of UVexposure time at 90 K. The intensity values for the Ti3+ and O− species werenormalized with respect to equal integral intensities of the respective spin centersignals [268].© 2005, American Chemical Society.

under real photocatalytic conditions nanosize TiO2 photocatalystswith highly hydrated surfaces are used. Under these conditions,there are large numbers of hydroxyl groups in a variety of differentenvironments, as shown by the infrared spectra in Fig. 3.10. Morework is still needed to unravel the roles of the various surfaceTi–OH groups in photocatalytic processes.One important point is that, whereas almost all of the

photogenerated holes are trapped in deep or shallow trappingsites, a large fraction of the photogenerated electrons are in anearly free, untrapped state in the interior of the TiO2 particle,as observed by a structureless absorption in the near-IR and IRregion [258]. In a recent EPR and IR study, Berger et al. foundthat even at the low temperature of 90 K, the major fraction,over 90%, of photoexcited electrons remained in the conductionband and was silent in EPR spectra under high-vacuum conditions(Fig. 4.12) [268]. This may explain the slow oxygen reduction onthe irradiated TiO2 surface, since trapped electrons react withmolecular oxygen much faster than free electrons [246]. Thereaction of the trapped electron with O2 will be discussed also inSection 4.1.5.

4.1.4. Oxidizing species at the TiO2 surfaceMany efforts have been devoted to clarifying the oxidizing

species generated at the irradiated TiO2 surface, which isessential for understanding the mechanism of photocatalysis andfor designing photocatalyst systems for environmental cleanupapplications. The ever-clarified oxidizing species include holes,either free or trapped, •OH radicals, O•−2 , and

1O2, among others [6,10,22,283,284]. H2O2 andO2 are also involved in the photocatalyticoxidation processes in various mechanisms [5,6,10].Holes. Holes are the primary oxidizing species in photocatalyticreactions. As mentioned above, photoproduced holes are trappedwithin picoseconds at the TiO2 photocatalyst surface [262]; thisindicates that most of the primary oxidation processes are causedby trapped holes. TA studies have revealed that at least two


Fig. 4.13. Sum-frequency generation (SFG) spectral evolution of methanol on TiO2 (a) with introduction of water vapor (b) and subsequent evacuation (c) [291].© 2004, American Chemical Society.

kinds of different trap sites exist for holes on the surface ofTiO2 nanocrystalline photocatalysts: the deep hole exhibits anabsorption around 520 nm, while the shallow hole exhibits anabsorption around 1200 nm [246]. The shallowly trapped holes areeasily excited thermally into the valence band so as to establishequilibriumwith the free holes [252]. Shallowly trapped holes willtherefore have a comparable reactivity and mobility to free holes.The deeply trapped holes, however, are more or less localized atdeep traps and exhibit lower oxidizing potential [252,277]. Theprecise chemical nature of these is still in question, however.Shallowly trapped holes react very rapidly with chemisorbed

substances; the oxidation process is even competitive to theultrafast charge trapping [264]. Deeply trapped holes, however,prefer to react with more mobile physisorbed substances, dueto their localized nature; the reaction is often a slower process.For instance, NMR studies by Xu and Raftery showed that for theoxidation of ethanol and 2-propanol, the chemisorbed species,i.e., ethoxide or propoxide, were oxidized much more rapidly thanthe H-bonded species [285,286]. Shkrob et al. observed the rapidphoto-oxidation of glycerol by the holes on TiO2 nanoparticles;ca. half of these holes are scavenged within the duration of a3.3-ns fwhm (full width at half maximum), 355-nm excitationlaser pulse; the others are scavenged at a slower rate over 200 nsafter the photoexcitation pulse [287,288]. They suggested that thereaction with chemisorbed and physisorbed glycerol may accountfor the prompt and the slow hole decays, respectively. Tojo et al.studied the one-electron oxidation processes of aromatic sulfidesin TiO2 colloidal solution by laser flash photolysis [289]. They foundthat the 4-methylthiophenylmethanol cation radical showed abroader absorption at 1 µs after excitation than that at 10 µsafter excitation. They assigned the former to the absorption due tosurface-bound cation radicals and the latter to the absorption dueto free cation radicals in solution.One important point is that even for the same substance, the

adsorption states may change, depending on the environment. Bymeans of NMR, Nosaka et al. studied the adsorption of ethanol onTiO2 photocatalysts [290]. When there was a physisorbed waterlayer on the surface of the photocatalyst, ethanol was found toprefer to stay in the mobile water physisorbed water layer, andthus the NMR signal for ethanol was sharp. When the physisorbedwater molecules vaporized completely at elevated temperature,ethanolmolecules reached the surface and reactedwith the surfacetitanols to form ethoxide, resulting in remarkable line broadeningdue to the restricted mobility. With rehydration and recovery ofthe water layer, the ethoxide was hydrolyzed to form ethanolmolecules, which moved back to the physisorbed water layer andproduced sharp NMR signals.In another study related to this, Wang et al. observed the

competitive adsorption between methanol and water on TiO2 film

by means of in situ sum-frequency generation (SFG), a highlysensitive surface characterization technique (Fig. 4.13) [291].When only methanol vapor was in contact with the TiO2 film,two kinds of species, molecular methanol and methoxide, whichwere produced by dissociative chemisorption of methanol, weredetected on the TiO2 surface. When a relatively large amount ofwater vapor (water–methanol ratio, ca. 300) was introduced intothe system, the surface methoxide signal dropped below the SFGdetection limit, but physisorbed methanol remained. Evacuatingthe system containing a large amount of water and a relativelysmall amount of methanol caused the water SFG peak to dropbelow the SFG detection limit. The methanol peak remained.Interestingly, a substantial methoxide SFG peak reappeared. Theseworkers suggested that the oxidation pathway of methanol in thepresence of water should depend on the ratio between water andmethanol.In a purely photoelectrochemical study, Villareal et al. found

that the oxidation of methanol was never mediated by holes,because the conditions were aqueous solution, so that there wassufficient water available to shift the adsorption equilibrium awayfrom methanol. In contrast, they found that formic acid oxidationwas always mediated by holes, due to strong chemisorption of theacid on the TiO2 surface [223].

•OH radicals. Previous product analyses on various photocatalyticreactions led to the postulate that •OH radicals, produced byoxidation of surface hydroxyl or adsorbedwater, play an importantrole in initiating oxidation reactions, especially for substances thatadsorb weakly on the TiO2 surface [6,11]. This oxidation pathwayis sometimes designated as indirect oxidation, in comparison tothe direct oxidation by holes. The presence of •OH radicals atthe irradiated TiO2 surface was demonstrated by spin trappingexperiments with EPR spectroscopy [292–295].Some recent studies, however, suggest that the role of the

•OH radical is probably overestimated. The photocatalytic one-electron oxidation studies by Fox and co-workers [296,297] andTachikawa et al. [298–300] showed that for many substances theprimary oxidation should be initiated by free or trapped holes.A fluorescence probe investigation by Ishibashi et al. showedthat free •OH radicals, detected by terephthalic acid probe, wereproduced in a quantum yield three orders smaller than that oftrapped holes, detected by iodide ions [301]. Nosaka et al. arguedthat the common spin trapping reagents, such as 5,5-dimethyl-1-pyrroline-N-oxide (DMPO), might simultaneously detect bothtrapped holes and surface-adsorbed •OH radicals [302]. Theysuggested that water reacts with the one-electron oxidationproduct of the spin trapping reagent (Eqs. (4.4), (4.5)) and shouldgive the same product as the adduct formation between •OH


Fig. 4.14. Reaction scheme for the oxygen photoevolution reaction on TiO2 (rutile)in contact with an aqueous solution of pH of 1 to about 12 [219,220].© 2004, American Chemical Society.

radicals and the spin trapping reagent (Eq. (4.6)) [302].

DMPO+ h+ → DMPO•+ (4.4)

DMPO•+ + OH− → DMPO–OH • (4.5)•OH+ DMPO→ DMPO–OH • . (4.6)

Similarly, some reactions, whichwere considered to be initiated by•OH radicals previously, could be initiated by trapped holes withthe production of the same products [302]. One such example isthe oxidation of methanol (Eqs. (4.7)–(4.10)):By •OH radicals:

CH3OH+ •OH→ •CH2OH+ H2O (4.7)

•CH2OH→ HCHO+ H+ + e−. (4.8)

By trapped holes:

CH3OH+ h+ → +CH3OH→ •CH2OH+ H+ (4.9)

•CH2OH→ HCHO+ H+ + e−. (4.10)

•OH radicals have been proposed to be generated by theoxidation of either water or adsorbed hydroxide ion withphotogenerated holes [6,215,216,303,304]. Nakato and co-workersand others have found problems with this simple electrochemicalprocess (e.g., failure to explain the 840-nm photoluminescencepeak) [111,217–220,305] and have proposed an alternativepathway in which •OH radicals are produced by a nucleophilicattack of water on a hole trapped at a surface lattice oxygen((4.11)–(4.14)).

[Ti–O–Ti]s + h++ H2O→ [Ti–O • HO–Ti]s + H

+ (4.11)

[Ti–O • HO–Ti]s → [Ti–O–O–Ti]s + H+ (4.12)

[Ti–O–O–Ti]s + H2O→ [Ti–O–OH • HO–Ti]s. (4.13)

When the O–O bond in TiO–OH breaks

[Ti–O–OH • HO–Ti]s → [Ti–O • HO–Ti]s + •OH. (4.14)

A similar scheme is shown in Fig. 4.14, in which the overall wateroxidation process is shown. Quantum mechanical calculationshave also been carried out by Qu and Kroes recently to check theseproposed pathways [222]. These authors examined a large numberof different possible geometries for adsorbed oxygen atoms andperoxide groups. Neumann et al. examined this mechanismexperimentally with electrochemical mass spectroscopic andphotovoltage measurements and found problems with it [221].They proposed a slightly modified pathway for water oxidation.Moreover, there are several reactions of H2O2, produced by

reduction of molecular oxygen or the disproportionation reactionof superoxides, could produce •OH radicals (Eqs. (4.15)–(4.19)) [6,306]. It should be noted that reaction (Eq. (4.17)) only producessurface-adsorbed •OH radicals,whereas reactions (4.18) and (4.19)

Fig. 4.15. Emission spectra observed after the laser pulse irradiation at 355 nmfor TiO2 (P25) suspension in water (�), ethanol (©), and mixture of water–ethanol(M). Copyright The Owner Societies 2004. (b) Schematic illustration of the possibleformation mechanism of singlet 1O2 on UV-excited TiO2 particles.

can produce both surface-adsorbed •OH radicals and free •OHradicals.

O2 + 2e− + 2H+ → H2O2 (4.15)

O•−2 + O•−

2 + 2H+→ H2O2 + O2 (4.16)

H2O2 + e− → •OH+ OH− (4.17)

H2O2 + O•−2 → •OH+ OH−+ O2 (4.18)

H2O2 + hν → •OH+ •OH. (4.19)

O•−2 . Superoxide O•−

2 is less important in initiating oxidationreactions. Although there are several reports that suggested thatO•−2 initiated some type of oxidation reaction, for example, theoxidation of trichloroethylene [307], the examples are quite fewcompared to those concerning holes and •OH radicals. The role ofO•−2 in photocatalytic oxidation processes was discussed mainlywithin the framework of participation in the totalmineralization oforganic substances through reaction with organoperoxy radicals,the production of H2O2 by disproportionation reaction, and theanti-microbial activities [5,6,10]. Recently, Nosaka et al. proposedthat the reaction between O•−2 and a trapped hole would producesinglet oxygen 1O2, a strong oxidant (see below) [283].1O2. Singlet oxygen 1O2 is an important reactive oxygen speciesin atmospheric, biological and therapeutic processes, and is alsoused as a reagent in organic synthesis. The formation of 1O2 inTiO2 photocatalysis was predicted twenty years ago [308], butonly recently 1O2 generated on irradiated TiO2 surface could bedetected. Nosaka et al. first reported the detection of the near-infrared phosphorescence at 1270 nm of 1O2 in a photocatalyticTiO2 aqueous suspension system by means of a gated photoncounting method (Fig. 4.15a) [283]. They suggested that oxidationof O•−2 with holes at the TiO2 surface should be a plausiblemechanism to produce 1O2 (Fig. 4.15b). The lifetime of 1O2 wasdetermined to be 2 µs, which is rather short as compared withthose of •OH radicals (ca. 10 µs) and trapped holes, possibly dueto the fast deactivation of 1O2 at TiO2 surfaces.Daimon and Nosaka measured the quantum yields for 1O2 over

ten commercial TiO2 photocatalysts in air, and found they rangedfrom 0.12 to 0.38, while the lifetimes ranged from 2.0 to 2.5 µs[309]. The short lifetime suggests that the generated 1O2 mainlyremains on the TiO2 surface without being released to the air orsolution. Since the quantum yield is significantly high, 1O2 maycontribute to the oxidation of organic molecules at the irradiatedTiO2 surface.Hirakawa et al. studied the generation of 1O2 by the photocat-

alytic reaction of TiO2 particles, either anatase or rutile, dispersedin solvents [310]. The estimated quantum yield by both types ofTiO2 was about 0.02 in ethanol. The 1O2 emission was completelyquenched by the addition of superoxide dismutase, suggesting that1O2 is produced via oxidation of O•−2 . The

1O2 was found to be


adsorbed at the TiO2 surface and could not oxidize nicotinamideadenine dinucleotide, which has no affinity for the TiO2 surface.However, it can oxidize thewater-soluble protein bovine serum al-bumin, which can be adsorbed on the TiO2 surface. The generationof 1O2 was enhanced in the phospholipid membrane, suggestingthat the phospholipid membrane is an important environment forthe phototoxic reaction mediated by 1O2 generation in TiO2 pho-tocatalysis.Janczyk et al. reported the generation of 1O2 at surface-modified

TiO2 particles [311]. They found that TiO2 particles, after surface-modification with organosilane, could oxidize cyanuric acid, astable material resistant to TiO2 photocatalysis. They concludedthat cyanuric acid was oxidized by 1O2 and suggested an energytransfer pathway for the generation of 1O2 on hydrophobicallymodified TiO2 particles.

4.1.5. Role of molecular oxygenMolecular oxygen plays an indispensable role in TiO2 photo-

catalysis, especially in reactions related to environmental cleanupapplications [5,6,10]. It assists the charge separation in TiO2 bycapturing TiO2 electrons; it generates active species such as O•−2 ,H2O2, 1O2, etc. that participate in reactions; it itself participatesin the reaction and accelerates the mineralization of organic sub-stances; it also helps the TiO2 photocatalyst to maintain stoi-chiometry during photocatalytic reactions. These effects have beeneither postulated or already demonstrated for many years, but re-cent progress in both theoretical simulations and experimentaltechniques have greatly improved our understanding.Improved charge separation. Berger et al. found that molecularoxygen could not only capture the photogenerated electronsto suppress charge recombination, but also improve the photo-stimulated charge separation in TiO2 nanoparticles, i.e., more holescould be trapped within the particles in the presence of molecularoxygen [312]. They designed two experiments to demonstratethis idea: in one experiment, they illuminated dehydrated TiO2nanocrystals (13 nm) and recorded the EPR spectra before andafter the addition of molecular oxygen; in another experiment,they illuminated dehydrated TiO2 nanocrystals and recorded theEPR spectra in the presence of molecular oxygen. They observedthat, in the former experiment, the concentration of trapped holes(O−) was the same as the superoxide, indicative of the depletionof photogenerated electrons with molecular oxygen; thermallystable charge separation could be maintained for quite a long timeafter the formation of superoxide. In the latter experiment, theyobserved that the concentrations of O•−2 andO

−were related to theoxygen pressure during UV excitation, first increasingwith oxygenpressure, and finally saturating at 25 mbar O2 (Fig. 4.16). Thisindicates that either the number of adsorption sites for O•−2 ionson the nanocrystal surface or the surface capacity for hole trappingand O− formation limits the final concentration of EPR-activespecies. Most interestingly, at 140 K, a maximum concentrationof ten O−/O•−2 pairs per particle was obtained after O2 photo-adsorption, but, under high-vacuum conditions, UV excitationresulted in only one e−/O− pair per particle.Ishibashi et al. quantitatively studied the kinetics of pho-

togenerated O•−2 under very low-intensity light illumination(1 µW cm−2) at the air–TiO2 interface as well as the water–TiO2interface by measuring the chemiluminescence from luminol so-lution pipetted onto TiO2 film photocatalysts [313,314]. The totalnumber of steady-state O•−2 on the TiO2 surface was determinedfrom the chemiluminescence intensity: about 1 × 1014 cm−2 atthe air–TiO2 interface and 2× 1014 cm−2 at the water–TiO2 inter-face. These steady-state concentrations were independent of lightintensity (1 µW cm−2–15 mW cm−2), indicating that the amountof O•−2 is restricted by the number of adsorption sites on TiO2. The

Fig. 4.16. Time-resolved accumulation of trapped holes and scavenged electronsupon UV exposure at different oxygen pressures: (a) 0.1 mbar O2 and (b) 10 mbarO2 . (c) UV exposure at P < 10−6 mbar and subsequent addition of O2 . Note thatthe profiles of O− and O•−2 coincide for a given pressure, as the ratio of O

− to O•−2equals unity throughout all experiments [312].© 2005, Wiley-VCH Verlag GmbH & Co.

greater concentration of O•−2 at thewater–TiO2 interface comparedto the air–TiO2 interface was thought to be due to the lower inter-facial energy barrier for the electron transfer in water.As discussed in Section 3.7, the reaction of TiO2 electrons with

molecular oxygen is a process that is still poorly characterized,varying over many orders of magnitude in rate, even for samplesthat are nominally identical. It has been considered as the rate-determining step in photocatalytic processes [234,236], althoughthis is still not certain. The kinetics for oxygen reduction on theTiO2 surface can probably be improved by depositing a noblemetalco-catalyst [226], although this is a far from practical solution.It will be interesting, both fundamentally and practically, to tryto understand the detailed reasons for the wide variations inrate. As discussed in Section 3.8, the driving force for electrontransfer is quite low, as the potential difference between theconduction-band edge of TiO2 and the potential for O2 reductionto superoxide is quite small, which may result in a slow electrontransfer [238]. This driving force is most likely decreased evenfurther, due to the fact that the electron transfer takes place fromsurface electron traps.Generation of active oxygen species. Superoxide (O•−2 ) and hydrogenperoxide are two main products of O2 reduction in TiO2photocatalysis and are involved in the production of other activeoxygen species such as •OH, 1O2, as described in Section 4.1.4.While superoxide is the product of one-electron reduction ofmolecular oxygen, hydrogen peroxide can be produced by eithertwo-electron reduction of molecular oxygen or disproportionationof two superoxides [6]. The generation of superoxide has beenrecognized for a long time based on the EPR [268,269,312,315,316]or chemiluminescence studies [313,314,317,318]. The detectionof surface-peroxo species, however, is difficult, and only recentlywas achieved by Nakamura et al. by means of multiple internalreflection infrared spectroscopy [319].Tilocca et al. studied the adsorption of O2 on hydroxylated

rutile TiO2 (110) using density functional theory total energycalculations and first principles molecular dynamics simulations[320]. Their results suggested that the electron transfer fromTi–OHto O2 does not need to pass through a chemisorbed O2 state.Instead, scavenging by O2 starts as soon as the oxygen moleculecaptures two hydrogen atoms from the bridging hydroxyls(Fig. 4.17, a–c); this produces a hydroperoxide species, whichis eventually chemisorbed onto the five-fold-coordinated surfaceTi (Ti5c) atoms (Fig. 4.17, d–f). The H2O2–Ti5c molecule wasfurther transformed to HO2–Ti5c after transferring a proton back


Fig. 4.17. Selected snapshots from the MD trajectory started with an O2 molecule above the surface, illustrating the formation of H2O2 , its adsorption and furthertransformation. Slab Ti and O atoms are represented as white and dark sticks, while adsorbed O and H atoms are represented as large dark and small white spheres,respectively; dashed lines denote hydrogen bonds [320].© 2005, American Chemical Society.

to a bridging oxygen (Fig. 4.17, g–i). The theoretical calculationssupport the experimental hypothesis of Henderson that electronscavenging could happen betweenweakly bound (physisorbed) O2and bridging hydroxyls [143].Atwood et al. compared the oxygen-centered radicals gener-

ated on surfaces of Degussa P25 powder and two rutile powdersby EPR spectroscopy [321]. They found that stable superoxideradicals are generated on either dehydroxylated P25 powder ordehydrated rutile powders under UV excitation; the signal can per-sist up to several days. On the fully hydrated P25 surface, no signals(O−, O•−2 , •O2H) are observed in the EPR spectrum, even after pro-longed UV irradiation at 100 K. On the partially hydrated P25 sur-face, the dominant signal of O−; and traces of O•−2 and •O2H areobserved; both were thermally unstable and quickly decayed todiamagnetic species such as H2O2 upon warming to 300 K. How-ever, thermally stable O•−2 and •O2H could be observed on rutilepowders, even in the case of full hydration. It was believed that onthe P25 surface, which is mainly anatase, superoxide anions can besufficientlymobile on the fully andpartially hydrated surfaces suchthat disproportionation reactions (forming H2O2) occur. However,on the hydrated rutile surface, some surface sites capable of O•−2stabilization remain.Consistent with Atwood’s study, Goto et al. reported a

preferred generation of hydrogen peroxide on anatase duringthe photocatalytic degradation of 2-propanol in water, while forrutile photocatalysts, the generation of superoxide was preferred[322]. They observed this trend on several anatase and rutilephotocatalysts; the data are shown in Table 4.2.Participation in reactions. Heller suggested that the photocatalyticoxidation of organic pollutants is mostly mediated by molecularoxygen, not by photogenerated holes or •OH radicals [5,235].

Fig. 4.18. Photocatalytic decomposition of several kinds of organic compoundsthrough the tetraoxide mechanism suggested by Schwitzgebel et al. [235].© 1995, American Chemical Society.

Although valence-band holes of TiO2 were required for theoxidation to be initiated, the actual oxidizer was molecularoxygen. O2 was suggested to react with organic radicals, generatedupon the hole or •OH radical reaction with the reactant, toproduce an organoperoxy radical (ROO•), and subsequently thereactant was degraded and mineralized through thermochemicalreactions, probably by the tetraoxide pathway (Fig. 4.18). A similarconclusion was also drawn from photoelectrochemical studies[230,231] and isotope studies [323].Although the existence of the organoperoxy radical was

predicted in photocatalytic processes, its detection was difficultdue to its short lifetime [324–326]. In a recent EPR study, Murphyet al. observed that a number of different organoperoxy speciesare involved in the heterogeneous decomposition of ketones overdehydrated polycrystalline TiO2 [327]. These radical intermediateswere identified as organoperoxy-based species of general formula


Table 4.2Physical properties and estimated production rates of hydrogen peroxide and superoxide ion on several TiO2 powders [322].

TiO2 powder Anatase component (%) Surface area (m2/g) Purity (%) H2O2 (10−7 mol min−1) O•−2 (10−7 mol min−1)

NS-90 <1.0 0.9 >99 0 1.0CR-EL <0.1 7.1 >99 0.1 0.9TIO-3 <1.0 48 >99 0 1.6TIO-2 >99.9 18 98.5 2.9 0.4ST-21 >99.9 56 95 5.3 0.2ST-11 >99.9 71 95 2.7 0.5

Fig. 4.19. High-resolution transmission electron micrograph of the hydrocarbon/TiO2 films that were illuminated with UV light for 3 h. (1) A low-magnificationimage and the corresponding selected area electron diffraction pattern (inset).(b) A high-magnification image. Additional lattice fringes (d = 0.45 nm and 0.69nm) which do not appear in a perfect rutile crystal, were observed in the inset of (a)(marked by arrows) and (b) [333].© 2006, American Institute of Physics.

ROO• and RCO3•. Theywere formed by reaction of photogeneratedcharge carriers with the adsorbed ketones in the presence ofmolecular oxygen. The organoperoxy intermediates are thermallyunstable and decompose at temperatures in the region of180–250 K. This work demonstrates the importance of free-radicalpathways involving both organoperoxy and superoxide radicals forthe photo-oxidation processes over TiO2 photocatalysts.Maintenance of the stoichiometry of TiO2. Surface coordinatively

unsaturated O atoms of TiO2 photocatalysts, especially those atsurface irregularities such as kinks and edges, are labile underUV illumination due to hole trapping, and could be exchangedwith molecular oxygen [328] or abstracted during photocatalytic

reactions [329–337]. Under O2-deficient conditions, these O atomswould gradually be lost during photocatalytic reactions; thisresults in a loss of stoichiometry of the TiO2 photocatalystand also its photocatalytic activity [329–333]. Yoshida et al.recently studied the structural changes of a TiO2 film duringthe photocatalytic oxidation of an organic film in vacuum by aunique in situ TEM/ EELS system [331–333]. The organic filmwas deposited on a 60-nm-thick TiO2 film supported on a Cugrid for TEM observation. During UV irradiation, the organicmaterial was oxidized and decomposed, desorbing from the TiO2surface (Fig. 4.19). At the same time, single crystalline TiO2was transformed into polycrystals, and additional lattice fringes,belonging to substoichiometric forms of TiO2, appeared. The TiO2film gradually became a network of aggregates after one week ofirradiation. A detailed EELS analysis revealed that the structuralchanges were associated with the loss of oxygen atoms in the TiO2crystal lattice [331–333].Muggli and Falconer studied the photocatalytic oxidation of

formic acid and acetic acid in a helium atmosphere [329,330]. Theyobserved the fast deactivation of the TiO2 photocatalyst and therecovery of activity by storage in O2 in the dark. They suggestedthat the loss of surface lattice oxygen caused the deactivation ofthe photocatalyst, and O2 was able to replenish the lattice oxygenvacancies.

4.1.6. Effect of crystal faceMorris Hotsenpiller et al. studied the photochemical reduction

of Ag+ ions in an aqueous solution on oriented TiO2 (rutile)surfaces [338]. They observed the oriented photochemical reactionrates and quantum yields, i.e., the (101), (111), and (001)surfaces had higher Ag-deposition rates than the (100) and (110)

Fig. 4.20. (a) AFM image (1× 1µm2) of ultra-smooth TiO2 (110) surface. (b) Ag formed on as-polished (dark bar) and ultra-smooth (slashed bar) TiO2 surfaces with variousorientations. The Ag/Ti intensity ratio was evaluated by EDX [340].© 2005, The Japan Society of Applied Physics.


Fig. 4.21. (a) A SEM image of n-TiO2 (001) surface of (001) after photoelectrochemical etching under 2.0 V anodic polarization [218]. (b) A SEM image of TiO2 polycrystallineelectrode after photoelectrochemical etching under 1.0 V anodic polarization [344].© 1998, The Electrochemical Society, Inc.

Fig. 4.22. (a) SEM image of a rutile particle on which Pt fine particles weredeposited selectively on (110) face by photocatalytic reactions. (b) SEM image ofan anatase particle (center) on which Pt fine particles (indicated by arrows) weredeposited by photocatalytic reactions. (c) and (d) SEM images of rutile and anataseparticles, respectively, co-deposited with Pt and PbO2 by photocatalytic reactions.© 2002, The Royal Society of Chemistry and the Centre National de la RechercheScientifique [345].

surfaces. In a later study, Lowekamp et al. used polished andannealed surfaces of randomly textured rutile polycrystals tophotochemically reduce Ag+ to Ag metal in an aqueous solution[339]. By correlating the surface orientation of more than 100individual crystallites with the amount of deposited Ag, theyconcluded that the most reactive orientations lie near the (101)surfaces. These studies, however, did not involve atomicallysmooth surfaces, and thus might be open to question (see below).In a recent study, Yamamoto et al. reconsidered the anisotropic

photochemical reactivities of rutile with atomically smooth rutilesingle crystal surfaces (Fig. 4.20) [340]. They observed only littleorientation dependence of the photochemical reactivity for Ag+reduction on the group of as-polished surfaces. In contrast, thephoto-reactivity on the group of ultra-smooth surfaces exhibitedstrong orientation dependence, although their reactivities wereone to two orders of magnitude smaller than those on the as-polished surfaces. The reactivity of ultra-smooth surfaces for Ag+reduction followed the sequence: (101) > (100) > (001) >

(111) > (110) [340]. Note that the Ag reduction reactionis accompanied by water photo-oxidation in their experiments.Thus, the above reactivity sequence should represent the sum ofreduction and oxidation reactions. In another study, Nakamuraet al. measured the flat-band potential of atomically smooth TiO2rutile single crystal surfaces [111]. They found that the flat-bandpotential of the (100) surface is 0.1 V more negative than that ofthe (110) surface, indicative of the higher reduction ability of theformer surface.Nakato et al. observed the anisotropic etching of TiO2 single

crystal surface during anodized polarization in H2SO4 electrolytesolution [217,341,342]. The etching left rectangular holes on therutile (001) surface (Fig. 4.21a) [218]. The holes, 50–200 nm wide,extend in the (001) direction, with the (100) face or equivalentexposed selectively at the walls. For the (110) surface, longrectangular grooves were produced at the surface in the (001)direction, with the (100) face or equivalent exposed at the wallsof the grooves. For the (100) surface, the results were quitesimilar to those for the (110) surface, except that the walls ofthe photoproduced grooves were either parallel or perpendicularto the electrode surface. This also shows that the (100) face orequivalent is exposed at the walls. A similar phenomenonwas alsoobserved on polycrystalline TiO2 electrodes during anodizationpolarization in H2SO4 solution, with uniform rectangular cellscreated on the surfaces of microcrystals (Fig. 4.21b) [343,344].These etching results were considered to be abnormal, since the(110) face is the most stable for rutile, rather than the (100) face,but this could be explained by the difference in hole reactivitywithwater for various crystal faces [218].Matsumura and co-workers studied the crystal-face-dependent

photocatalytic reactivity of rutile and anatase particles [345]. TheTiO2 powder used in their study consisted of 1-µm size rutileand anatase particles with well-developed crystal faces. The rutileparticles showed a tetragonal prism structure with four planes,which are assigned to the (110) faces. Each end of the prism iscapped by four planes, which were assigned to the (011) faces. Theanatase particle had a tetragonal bipyramid structure consisting ofeight (011) faces. Both vertexeswere removed and capped by (001)faces. After photocatalytic oxidation of water on the powder usinghexachloroplatinate ions as the electron acceptors, Pt depositswere observed mostly on the rutile particles, especially on the(110) face, indicating that the reduction sites of rutile particleswere on the (110) face (Fig. 4.22a). When 2-propanol was addedto the solution, Pt was deposited on both the anatase (Fig. 4.22b)and rutile particles. Using the thus-prepared Pt-deposited TiO2powder, Pb2+ ionswere photocatalytically oxidized to PbO2, whichwere deposited on the TiO2 particles (Fig. 4.22, c, d). They observedthe selective deposition of PbO2 on the (011) face of rutile particles,and for anatase particles, PbO2was deposited in a larger amount on


Fig. 4.23. SEM images of (a) an anatase particle after UV irradiation in aqueoussolution containing 1mMH2PtCl6 and 0.52M2-propanol and (b) an anatase particleafter sequential UV irradiation in the same solution as in (a) and then in an aqueoussolution of 0.1 M Pb(NO3)2 [346].© 2003, The Royal Society of Chemistry and the Centre National de la RechercheScientifique.

Fig. 4.24. Active oxygen species and some processes related to remotephotocatalysis. Neutral species including H2O2 , HO2 , and 1O2 are supposed todiffuse farther away from irradiated TiO2 surface than the highly reactive •OH,through air, water, or surface.

the (001) face than on the (011) face. Thismeans that the (011) faceprovides the oxidation site for rutile particles, while for anataseparticles the (001) face is more oxidative than the (011) face.In another study, Matsumura and co-workers etched the

same anatase particles with HF solution [346]. By this process,the edge between two (011) faces was selectively etched andas a result, eight new faces were generated on each particle,which were assigned as (112) faces. The new anatase particleswere photocatalytically deposited with Pt and PbO2 in sequence.Interestingly, most of the Pt was deposited on the (011) faces,while PbO2 was deposited on the newly generated (112) faces, asshown in Fig. 4.23, indicating the reductive function of the (011)faces and the oxidative function of the (112) faces for the anataseparticles. These results suggest that crystal faces could assist in theseparation of electrons and holes, and thus, it may be possible toobtain high photocatalytic activity through control of the surfaceatomic arrangement of TiO2 photocatalysts.Morework needs to be

carried out in which highly well-defined experimental results arecompared with high accuracy quantum mechanical calculations.

4.1.7. Remote photocatalysisRecent studies have shown that photocatalytic reactions could

be initiated at a distance away from the TiO2 photocatalysts. Someexamples include the remote killing of bacteria [347], the remotedecoloration of dye films [348–352], the remote decompositionof organic or polymeric films [350,353–358], and the remoteoxidation of inorganic materials (Cu, C, SiC, etc.) [359,360]. Inthese examples, the impact distance of the TiO2 photocatalyst wasfound to be as long as several millimeters, and the reaction mediacould be air, water, or solid surfaces; in addition, several typesof oxidizing species were involved in the photocatalytic reactions(Fig. 4.24).Tatsuma et al. first observed the remote oxidation ofmethylene

blue film at a distance of 500 µm away from a TiO2 film [352].Later studies showed that aromatic or even aliphatic substancescan be oxidized and decomposed into CO2 remotely by TiO2films (Fig. 4.25), and the impact distance could be up to 2.2 mm[350]. It was clear that the oxidation could not be caused byholes on the TiO2 surface. Similarly, it was hard to believe thathydroxyl radicals generated on the TiO2 surface could travel sucha long distance in air, considering their high activity [350]. Someneutral oxygen species generated on illuminated TiO2 surfaces,such as HO2,H2O2, can diffuse a long distance in air. However,these species are not reactive enough with aliphatic hydrocarbonsto produce CO2. Tatsuma et al. thus postulated that hydroxylradicals should participate in the remote photocatalytic reactionsdescribed above, but a major proportion of these hydroxyl radicalsshould be generated in the gas phase by the UV-photolysis of H2O2[350]. The possible remote oxidation mechanism thus includes(1) the diffusion of H2O2 in the gas phase, (2) the photolysis ofH2O2 into hydroxyl radicals in the gas phase under UV irradiation(λ < 365 nm), and (3) oxygenation and decomposition of targetmaterials by hydroxyl radicals. Also, H2O2 can be generated in thedisproportionation of HO2. Thus, either H2O2 or HO2 can be thediffusing species [350].Kubo and Tatsuma designed several experiments to verify

the proposed double-excitation mechanism. In one of theseexperiments, air was forced to flow through a TiO2-loaded cell thatwas irradiatedwith UV light and an aqueous solution of peroxidasein sequence. By this experiment, they proved the existence ofH2O2 in the gas phase, released from the irradiated TiO2 film[361]. In another work, they observed that H2O2 vapor, whenilluminated with UV light shorter than 365 nm, could oxidizeODS-modified glass [362]. Importantly, the reaction rates of theH2O2–UV system were in close agreement with those for the

Fig. 4.25. Time course of water contact angle of (a) octadecyltriethoxysilane (ODS) modified glass surface and (b) polystyrene-coated glass surface during the remoteoxidation experiment. The polyimide spacer was 50 µm thick and the light source was an Hg–Xe lamp (light intensity 10 mW cm−2) [350].© 2001, American Chemical Society.


Fig. 4.26. Correlations between the remote oxidation ability and the H2O2generation ability for various photocatalysts (�). TiO2 ‘‘A’’ and ‘‘B’’ are anatase sols,with average particle sizes of 20 nm and 7 nm, respectively. Pt (1.0 wt%) andAg (0.0055 wt%) were deposited photocatalytically. The correlation between theoxidation ability and the H2O2 amount for the H2O2–UV reaction is also shown (4)[362].© 2006, American Chemical Society.

photocatalytic remote oxidation, suggesting the similarity of theactive species in both systems (Fig. 4.26). They also compared theremote oxidation of octadecyltriethoxysilane (ODS)-modified glasswith various photocatalysts, and found that the oxidation ratesfollowed the sequence: Pt-loaded TiO2 > TiO2 > ZnO > WO3,the same sequence for the decomposition activity of methyleneblue in conventional photocatalysis [362]. Thus, they suggestedthat the conventional photocatalytic activity could be a measurefor roughly predicting the remote oxidation activity of a givenphotocatalyst [362].Naito et al. studied the remote photocatalysis by means of

single-molecule fluorescence microscopy [348,351]. They foundthat the double-excitation process described above should be

involved in the bleaching of the fluorescence of dye molecules[351]. But they also argued that it is hard to exclude thecontribution of other active oxygen species, such as 1O2. Theydesigned an experiment that used terrylenediimide (TDI) as afluorescent probe for 1O2. The molecule reacts to 1O2 to forma strongly fluorescent TDI diepoxide with blue-shifted emission.They found that 1O2 generated on UV-excited TiO2 film coulddiffuse in air over 2mmdistance, and oxidize TCI molecules at thatdistance (Fig. 4.27) [348].Haick et al. reported another type of remote photocatalytic

oxidation by surface diffusion of oxidizing species [357,358]. Theyprepared microstripes of TiO2 on an Si substrate and observedthe mineralization of aliphatic chains anchored to inert silicondomains within minutes, even when these chains were located asfar as 20 µm away from the TiO2 microstripes [358]. The hydroxylradical is believed to be the oxidizing species in this system. Inanother study, they covered the TiO2 film with a photomask andobserved the decomposition of aliphatic chains in the dark region[357]. Very interestingly, the occurrence of the photocatalyticreaction in the dark region did not cause any decrease in thereaction rate in the illuminated region, even though competitionmay exist in sharing the oxidizing species. Kawahara et al. studiedthe surface diffusion behavior of photogenerated oxidizing specieson TiO2 films and concluded that the diffusion length was as longas 75 µm [356].The observation of remote photocatalysis, especially that

induced by surface diffusion of oxidizing species, may have asignificant effect on the design of photocatalytic materials. Theefficiency of photocatalysis is restricted by recombination. Inthe illuminated state, on the TiO2 surface, electrons, holes, andactive oxygen species maintain a dynamical equilibrium. Whenthere is spillover of an oxidizing species, light can be utilizedmore efficiently for chemical conversion. This implies that ahybrid of TiO2 and an adsorbent (e.g., zeolite, active carbon) mayexhibit performance better than that of pure TiO2. The role of theadsorbent is not restricted to the adsorption of pollutants; it canalso act as a reaction site via the out-diffusion of oxidizing species.

Fig. 4.27. Fluorescence images of single terrylenediimide (TDI) molecules spin-coated on a polymethyl methacrylate (PMMA)-coated coverslip before (a) and after (b) UVirradiation for 5 min (scale bars are 10 µm). The air gap between TDI layer and TiO2 film is 12.5 µm. The bright spots correspond to TDI diepoxide. The UV irradiation areais inside the white circle in the images, in this area TDIs are bleached by direct UV irradiation. (c) Scheme for the fluorescent detection of airborne 1O2 with TDI [348].© 2006, American Chemical Society.


Fig. 4.28. (a) Quantum yield dependence on absorbed photons at various initial 2-propanol concentrations; (b) quantum yield dependence on normalized absorbed photonnumber (Inorm) at various initial 2-propanol concentrations [366].© 1997, American Chemical Society.

The observation of remote photocatalysis may cause concernregarding the safety of TiO2 photocatalysts. However, the reactionrates for remote photocatalysis are lower than those for ordinaryphotocatalysis, and they decrease with distance. For a 50-µmdiffusion distance, the rate of remote photocatalysis via airdiffusion is ca. 2 orders of magnitude lower than that of ordinaryphotocatalysis, while that through surface diffusion is ca. 1 order ofmagnitude lower, as roughly estimated by the published data [350,356,357]. Therefore, the TiO2 surface should not be harmful evenwhen human skin is in close proximity to it, in the low-intensityUV light contained in solar light or interior illumination.Remotephotocatalysis through air diffusion of oxidizing species

may have applications in lithographic techniques. Tatsuma et al.[360,363] have investigated the feasibility of this new techniqueand have observed the patterning of copper plates and organicmonolayer-modified glass, with a resolution of 10µmor better, byplacing a TiO2 film on the surface to be patterned and irradiatingthe film through a photomask. The technique requires neitherphoto-resist nor special equipment other than a light source.

4.2. Photocatalytic reactions

As mentioned above, several kinds of active oxidative species,including free and trapped holes, •OH radicals, O•

−

2 , and1O2, are

involved in the initiation of photocatalytic oxidation reactions [6,11,15,16,283,302,311,317,364]. This offers the TiO2 photocatalyst’ssuperior ability to oxidize almost all kinds of organic and polymermaterials, kill microbes, and mineralize these substances withthe aid of molecular oxygen. These reactions mainly occur on orvery close to the surface of the TiO2 photocatalyst. The remotephotocatalysis also contributes to the oxidation reactions, but asmentioned above, at rates one or two orders of magnitude lower.

4.2.1. Decomposition of gaseous pollutantsUnder UV illumination, TiO2 is able to oxidize gaseous

pollutants in quantum yields ranging from 1% to over unity, whichhas important applications in cleaning indoor and outdoor air [11].Given that there is a sufficient supply of O2 and UV illumination,gaseous organic substrates can be completely degraded into CO2,H2O, and mineral acids on a TiO2 thin film. However, the removalrate is influenced by numerous parameters; these include lightintensity, substrate concentration, O2 partial pressure, humidity,substrate type, and so on. Also, in many cases, the generation ofrelatively stable reaction intermediates lowers the removal rateand even stops the reaction through blocking active sites [365].Air–solid photocatalytic reactions are influenced by substrate

concentration and incident light intensity in a cooperative

way. Ohko et al. measured the quantum yield (QY) of asimple photocatalytic reaction, 2-propanol oxidation, on a TiO2film in ambient air, for a series of 2-propanol concentrations(1–1000 ppmv) and a wide range of light intensity (tens ofnW cm−2–1 mW cm−2) [366]. They found that the QY increasedgradually with decreasing light intensity I , and reached amaximum value when the light intensity was extremely low. Theydesignated the condition for maximum QY as light-limited. Theyalso found thatwhen the 2-propanol concentrationwas decreased,the maximum QY was obtained at lower light intensity. Theyreplotted QY as a function of a normalized light intensity (Inorm),which was defined as the ratio of the number of photons absorbedper second to the number of adsorbed 2-propanol molecules,both normalized to the geometric area. The plots obtained fordifferent initial concentrations of 2-propanol fell on the samecurve, which is a very interesting result (Fig. 4.28). QY increasedas Inorm decreased and finally became constant at ∼28% for Inormvalues below 10−4 s−1. The maximum QY varied from film tofilm, which indicated that there are recombination losses that areintrinsic to each film. As interesting as these results are, they areyet to be adequately explained.The mechanism of photodecomposition of 2-propanol does

not involve chain reactions. Only one photon participates inthe generation of one molecule of acetone. Interestingly, theQY of acetaldehyde degradation, for which there is the possibleinvolvement of radical chain reactions, exhibited the same trendas 2-propanol [367]. The curves for both acetaldehyde and 2-propanol coincided over a wide range of Inorm values (>10−4 s−1),except for the region of Inorm below 10−4 s−1, where the QYvalues for acetaldehyde degradation continued to increase withdecreasing Inorm and reached 180% for the value of Inorm of 3 ×10−5 s−1 due to the involvement of the radical chain pathway.It should be pointed out that, aside from the light-limited con-

dition, over a wide range of experimental conditions, photocat-alytic reactions on TiO2 follow reasonably well-known behavior,i.e., first-order kinetics in the adsorbed concentration of the organiccompound and an order of α in light intensity [6].

r = kΓ Iα. (4.20)

Here, r is the reaction rate, k the first-order rate constant, Γ theconcentration per unit real surface area, and I the light intensity.Under light-rich conditions, the value of α is between 0 and1 [239]. This was explained by considering the recombinationof charge carriers to surface-adsorbed intermediate products[239]. Conversely, the light-limited reaction rate is representedas α = 1. Many organic compounds, as well as noxious gasessuch as H2S and NH3, follow Langmuir adsorption behavior, so


that this simple equation can be converted into the familiarLangmuir–Hinshelwood formby substituting the expression forΓ :

Γ −1 = Γ −10 + (Keqc)−1 (4.21)

where Γ0 is the surface concentration at full coverage, Keq theadsorption equilibrium constant and c the gas-phase or liquid-phase concentration [6,11,15].The O2 concentration also greatly influenced themineralization

of gaseous organic substances. Durrant et al. suggested thatan oxygen concentration of 5% is sufficient to suppress chargerecombination through capture of photogenerated electrons [238].However, the study of Lichtin et al. clearly showed that high O2concentration facilitates the complete mineralization of organicpollutants [368]. Lichtin et al. measured the molar ratio of CO2produced to vaporized CH3OH removed after 15 min of irradiationin different O2 concentrations. They observed that in 20% O2,generation of the stoichiometric yield of CO2 took twice as longas complete removal of CH3OH, while in 100 mol% O2 the timesrequired for complete removal of CH3OH vapor and for formationof CO2 were similar. This supports the Gerischer–Heller proposalthat the photocatalytic oxidation of organic pollutants is mostlymediated by molecular oxygen, as mentioned in Section 4.1.5.Humidity influences the removal rate in a complicated way

and appears to be dependent on the concentration and type oforganic compound. Obee et al. reported that ethylene oxidationrates decreased significantly as the water vapor concentration wasincreased from 1000 ppmv [369]. The influence of water vapor onthe reaction rate derived from the low adsorption of ethylene dueto its low adsorption affinity relative to water. Sano et al. observedthat the removal rate of acetaldehyde was higher in dry air than inhumidified air (50% relative humidity) [370]. However, Einaga et al.found that in the cases of benzene, toluene, and hexane, removalrates decreased with decreasing humidity, due to the increasingamounts of carbon deposits on TiO2 [371]. Blount and Falconeralso observed the same phenomenon in toluene photocatalyticoxidation [372]. Recently, Jo and Park reported that in the caseof very low-level contamination in indoor air (ppb level), there isno appreciable dependency on humidity for removal of volatileorganic compounds (VOCs), which results from the competitiveadsorption of water and the substrate compounds on the TiO2surface [373].The removal rate is also influencedby the type of substrate com-

pound. A comparison of reaction rates between 0.1 to 100 ppmv fortoluene, n-butane, 1,3-dichloroethene, n-hexane, methyl-ethyl-ketone, n-decane, propanal, 1-butanol, trichloroethylene, ethy-lene, ammonia and phosgene showed that the rates, from themostto least reactive, differed by about a factor of fifteen at any particu-lar concentration [11]. Substrates such as trichloroethylene, whosedegradation concerns radical propagation, are often degraded at ahigh rate. The degradation of aromatic substrates is slow and suf-fers photocatalyst deactivation due to the accumulation of reactionintermediates.As oftenmentioned in the literature, photocatalyst deactivation

is a general problem in the removal of gaseous pollutants, dueto the accumulation of intermediate products or products on theTiO2 surface [8]. Some deactivations are reversible; the deactivatedphotocatalyst could recover activity after UV illumination in cleanhumid air [8,372,373]. Some deactivations are irreversible; theseare often observed in the removal of gaseous pollutants containingheteroatoms such as S, N, P, and Si, etc. [374–377]. In such cases,mineral acid products block the active surface sites, and thuswaterwashing is needed to regenerate the deactivated photocatalyst.Loading noble metals can retard the deactivation of TiO2

photocatalysts. Einaga et al. reported that Rh0 could greatlyimprove the catalyst durability in benzene photo-oxidation [378].The role of Rh0 supported on the TiO2 surface is to reduce the

amounts of carbonaceous intermediates and byproducts on thecatalyst surface. Sano et al. reported that Pt–TiO2 could be free fromdeactivation in the removal of acetaldehyde and toluene througha combination of photocatalysis and thermocatalysis [371]. Theycombined Pt–TiO2 photocatalysts with a solar concentrator, bywhich the photocatalyst could be heated to about 200 ◦C. Atthis temperature, Pt could effectively decompose the intermediateproducts and prevent the TiO2 from deactivating.

4.2.2. Decomposition of aqueous pollutantsMost of the organic pollutants in water can be completely de-

composed and mineralized at the surface of UV-excited TiO2 pho-tocatalysts; these include alkanes, haloalkanes, aliphatic alcohols,carboxylic acids, alkenes, aromatics, haloaromatics, polymers, sur-factants, herbicides, pesticides, and dyes [11,365,379–382]. Whileonly UV light and O2 are necessary for the reactions, many factorssuch as light intensity, pH, ions, photocatalysts, kinds and concen-trations of substrates, etc., have a great influence on the efficiencyof themineralization process. As in the gas–solid system, photocat-alytic reactions in water work best at room temperature, and thusno heating is needed.Serpone and co-workers studied the quantum yields of

liquid–solid photocatalytic reactions on TiO2 slurry photocatalysts[239,383–385]. Similarly to the study of Ohko et al., they foundthat maximum quantum yields could be obtained under light-limited conditions, that is, low light intensity and relative highsubstrate concentration [239]. The maximum QY (365 nm) forphenol degradation was measured as 14% using Degussa P25as a photocatalyst at pH 3, and this value varied over a rangeof 3.5%–30% for six types of TiO2 photocatalysts tested [383,384]. Aside from the light-limited condition, over a wide rangeof experimental conditions, liquid–solid photocatalytic reactionson TiO2 can also be described by Eq. (4.20), the same as forgas–solid photocatalytic reactions [239]. The QYs measured inthese conditions varied over the wide range of 0.1%–10%, whichwas several times lower than those of gas–solid photocatalyticreactions [4,11,379]. The low QY should restrict the applicationsof TiO2 photocatalysis in water purification. The intrinsic reasonsfor the lower QY in water vs. air are yet to be explained in detail.Pichat et al. studied the influence of the sintering process on the

photocatalytic removal rate of various organic pollutants in water[386–388].While the increase of sintering temperature is expectedto decrease the recombination rate of charge carriers; it does re-sult in a decrease in surface area. Their purpose was to examinethe net effect of the sintering process on photocatalytic reactions.In one study, they compared the removal rate of three chlorophe-nolic compounds and one chloroaliphatic acid compound on fourTiO2 samples, which were all obtained identically by TiOSO4 ther-mohydrolysis with subsequent calcinations at various tempera-tures [388]. They found that the removal rate increased with thesintering temperatures for the three chlorophenolic compounds,whereas it was the opposite for the aliphatic acid compound. Theysuggested that the hole attack mechanism for carboxylic acids ismuch more sensitive to surface area variation than would be the•OH radical mechanism for cholorophenolic compounds, whichcan react in the near-surface solution phase. Their studies demon-strate suggest the difficulty in finding a high-efficiency photocata-lyst versatile for all types of pollutants in water [386].Generally, loading of noblemetal co-catalysts such as Pt, Au, and

Pd, etc., enables accelerating liquid–solid photocatalytic reactions[389]. These co-catalysts could enhance the charge separationand catalyze the oxygen reduction reaction. Recent studiessuggest that addition of electron receptors, such as hydrogenperoxide, ozone, persulfate, etc., also works well for improvingreaction rates [382,390,391]. These electron receptors are morereducible than molecular oxygen. Moreover, when accepting an


Table 4.3Photodecomposition of various organic compounds on TiO2 film [400].

Samples Stoichiometric formula Melting point (◦C) Light intensity (mW cm−2) Rate constants (µg cm−2h−1)

Octadecane C18H38 28.28 0.8 0.70Stearic acid C18H36O2 69.3 0.8 0.75Glycerol trioleate C57H104O6 −5 1.1 6.6Glycerol C3H8O3 17.8 1.1 31Liquid paraffina – −12.5 1.1 7.6Salad oilb – −5 to−10 1.1 6.2PEG (MW 600 D) (C2H6O2)n 15–25 1.1 17PEG (MW 500kD) (C2H6O2)n 57–65 1.1 12a Including 15–20-carbon hydrocarbons, alkyl naphthalene hydrocarbons.b Including oleic, linoleic, linolenic, palmitic acids and stearic acid.

Fig. 4.29. (a) Pots of weight changes vs. illumination time for the photodecomposition of octadecane under 0.8 mW cm−2 UV illumination. (b) Plots of weight changes vs.illumination time for the photodecomposition of glycerol trioleate (•), glycerol (�), liquid paraffin (+), and salad oil (×) under 1.1 mW cm−2 UV illumination [400].©2000,Elsevier Sciences B.V.

electron, these chemicals will dissociate into highly reactiveradicals in subsequent reactions, which can also participate in thephotocatalytic reactions [382,390,391]:

H2O2 + e− → OH− + •OH (4.22)

O3 + e− + H+ → O2 + •OH (4.23)

S2O2−8 + e−→ SO2−4 + SO4•− . (4.24)

4.2.3. Decomposition of liquid and solid filmsThe photocatalytic decomposition of monolayers, multi-layers,

and thin films of organic substances has been well studied forthe purpose of development of self-cleaning surfaces [5,9,10,392–404]. Minabe et al. compared the decomposition of severalliquid and solid organic films (∼500 nm thick) on TiO2 surfaces,under about 1 mW cm−2 UV illumination (Table 4.3) [400]. Allof the compounds exhibited pseudo-zero-order kinetics in theinitial stage of the reactions. This is reasonable, since the effectiveconcentration of reactant is high and can be approximated by aconstant value. Carbon dioxide was the only detectable gaseousproduct during the experiments. The weight change vs. timebehavior during the initial stages for octadecane and stearicacid yielded similar rates, i.e., thinning rates of 9–11 nm h−1(Fig. 4.29). The decomposition rate (ca. 70 nm h−1) for glyceroltrioleate, however, was much higher than those for octadecaneand stearic acid. Other liquid compounds, including liquid paraffin,salad oil, glycerol, and even polyethylene oxide, etc., exhibiteddegradation rates similar to that for glycerol trioleate. Thus, thedegradation rates of organic films are dependent on their physicalstates, i.e., solid or liquid film. The higher rates for liquid organiccompounds are reasonable, since organic compounds in the liquidstate can remain in more intimate physical contact with the TiO2surface than those in the solid state.

The decomposition rates of oily compounds were found tobe dependent on the square root of light intensity [400]. Evenin a dry atmosphere, the organic compounds were effectivelydecomposed, suggesting that the water that is generated duringthe decomposition can be effectively used. The decompositionwas believed to occur through a mechanism in which tetraoxideintermediates are involved (Fig. 4.18). The hydrocarbon chains canbe attacked at any point along their length, with C–C bond scissionand subsequent decomposition via the resulting aldehydes [235].The decomposition process proceeds at significant rates, with theevolution of CO2 as the only detectable gas-phase product. Thequantum yield for stearic acid decomposition could be over 20%,as estimated by Mills et al. [404].Lee et al. studied the solid-phase photocatalytic reaction on

the carbon black soot/TiO2 interface [353]. They observed thatthe bulk of the soot layer with ∼2 µm thickness was completelyoxidized to CO2 over a 30-h irradiation. The presence of O2 wasessential in the photocatalytic soot oxidation. In another study,Cho et al. studied the solid-phase photocatalytic degradation ofpolyvinylchloride–TiO2 polymer composites [355]. Irradiating thecomposite film for 300 h in air reduced its average molecularweight by two-thirds and weight by 27%. The photocatalyticdegradation of the composite filmwas accompanied by an increasein the FT-IR carbonyl peak intensity, the evolution of CO2 andH2O, and whitening due to visible light scattering from growingcavities. SEM images of the irradiated composite films showed thedevelopment of cavities around the imbedded TiO2 particles andindicated that active oxygen species which were photogeneratedon TiO2 surface desorbed and diffused across a fewmicrometers toreact with the polymer matrix.

4.2.4. Photocatalytic sterilizationTiO2 photocatalysts have been found to kill cancer cells,

bacteria, viruses, and algae under UV illumination [10,405–410].


Fig. 4.30. AFM images of E. coli cells on a TiO2 film: (a) no illumination; (b) illumination for 1 day; (c) illumination for 6 days; light intensitywas 1.0mWcm−2 . The outermostlayer, labeled as ‘‘A’’ in (a), disappeared after 24 h of illumination, with no layer visible in image (b) at the designated positions [414].©2003, Elsevier Science S.A.

Fig. 4.31. Schematic illustration of the three stages in the process of E. coli photokilling on a TiO2 film. In the lower row, part of the cell envelope is magnified [414].©2003,Elsevier Science S.A.

This results in important applications in the disinfection ofair, water, and surfaces with TiO2 photocatalysis. Kikuchi et al.studied the killing of Escherichia coli on TiO2 thin films [347]. Atypical experiment involves placing 150 microlitres of an E. colisuspension, containing 3 × 104 cells, on an illuminated TiO2-coated glass plate (1 mW cm−2 UV light). Under these conditions,there were no surviving cells after only 1 h of illumination. Evenwhen the cells were separated from the TiO2 film by a 50-µmPTFE membrane, with 0.4 µm pores, the cells were still killedremotely: after 4 h, there were none remaining. By contrast, after4 h under UV illumination without a TiO2 film, only 50% of thecells were killed. Further studies showed that various bacteria,E. coli, Staphylococcus aureus, and Pseudomonas aeruginosa, etc.,were killed rapidly on TiO2 surfaces under UVA illumination(320–380 nm, ∼1 mW cm−2) [9]. Kühn et al. observed that thekilling rates of bacteria were dependent on the thickness andstructure of cell walls [411]. Bacteria having thin cell walls, such asE. coli, and P. aeruginosawere killedmuch faster than bacteria suchas Candida albicans that have a thick eukaryotic cell wall [411].Sunada et al. observed that not only bacteria are killed on the

TiO2 surface by photocatalytic action, but also the toxic ingredientof bacteria can be decomposed [412]. If the UV illuminationcontinued for a sufficiently long time, the bacteria were foundto be mineralized completely into CO2, H2O and other mineralsubstances [413–415]. This is a unique property of photocatalyticsterilization compared to that of other antibacterial agents.Besides bacteria, viruses and fungi can also be killed and totallymineralized by photocatalytic action [408,413], but the killing offungi sometimes is much slower than that of bacteria because oftheir chemical stability [413].Sunada et al. studied the photokilling process of E. coli by

means of AFM (Fig. 4.30) [414]. They found that the photokilling

of bacteria on the illuminated TiO2 surface could be divided intothree stages (Fig. 4.31): (1) disordering of the outer membrane ofbacteria cells by reactive species (•OH, H2O2, O•−2 ); (2) disorderingof the inner membrane (the cytoplasmic membrane) and killing ofthe cell; and (3) decomposition of the dead cell. In the first stage,the outer membranes of E. coli cells were decomposed partially bythe reactive species produced by the TiO2 photocatalyst. Duringthis stage, cell viability was not lost very efficiently. The partialdecomposition of the outer membrane, however, changes thepermeability to reactive species. Consequently, reactive specieseasily reach and attack the inner membrane, leading to theperoxidation of the membrane lipid. The structural and functionaldisordering of the cytoplasmicmembranedue to lipid peroxidationled to the loss of cell viability and cell death. If the illuminationcontinued for a sufficiently long time, the dead cells were found tobe decomposed completely.Ordinary TiO2 surfaces do not work well in killing bacteria

under low-intensity UV illumination. However, when there areCu particles dispersed on the TiO2 surfaces, the hybrid surfacesexhibit excellent self-sterilizing properties even under room lightthat contains only a very small amount of UV light (∼1 µW cm−2)(Fig. 4.32) [416]. Although metallic Cu is a kind of antibacterialmaterial that works even in the dark, it is not the only reason forthe observed antibacterial effect of the hybrid films, since someCu-resistant bacteria can be effectively killed on the TiO2–Cu hybridsurface under indoor illumination [416]. Comparative studiesshowed that metallic Ag also enhanced the antibacterial effect ofTiO2 film, butmetallic Pt did not, even though Pt could improve thephotocatalytic performance of TiO2. Based on these observations,Sunada et al. proposed amechanism for the enhanced antibacterialperformance of TiO2–Cu hybrid film that emphasizes the effectof cuprous ions [416]. Under UV illumination, cuprous ions may


Fig. 4.32. Changes in survival of copper-resistant E. coli cells on a Cu/TiO2film under dark condition (•) and under UV illumination at light intensities of7 µW cm−2 (�) and 1 µW cm−2 (N) [416].© 2003, American Chemical Society.

exist on the TiO2 surface through the oxidation of metallic copperwith photogenerated holes. The cuprous ion may play two rolesin killing bacteria: (1) it increases the yield of hydroxyl radicalthrough a photo-Fenton process by reaction with photogeneratedH2O2; and (2) it diffuses into the cytoplasmic membrane of thebacteria and accelerates the lethal effect after the outer membraneof the bacteria is destroyed by oxidizing species.

4.3. Visible-light-induced photocatalysis

Due to their wide band-gaps, pristine anatase and rutile TiO2photocatalysts mainly absorb ultraviolet photons. However, solarlight only contains a small amount of ultraviolet photons (about5%), and room light lamps emit mainly visible photons. Therefore,starting one decade ago, efforts have been devoted to extendingthe spectral response of pure TiO2 material (mainly anatase TiO2)to visible light. These efforts have included doping TiO2 withmetal impurities [417–425], coupling TiO2 with narrow band-gapsemiconductors [426,427], preparing oxygen-deficient TiO2 [428–433], and doping TiO2 with non-metal atoms (anion doping) [434–462,221,463–487]. Especially in recent years, non-metal dopinghas become the most attractive; various non-metallic elements

such as N [434–464], C [221,465–474], S [475–481], B [482], P[483], and F [484–487], were reported to shift the absorptionthreshold of TiO2 into the visible range and provide the TiO2material with photocatalytic activity under visible illumination.Various mechanisms have been postulated to explain the shift ofabsorption threshold and the photocatalytic activity of non-metal-doped TiO2 photocatalysts [435,449,450,452,459]. These will besummarized in the sections below.

4.3.1. Non-metal dopingN-doped TiO2. In a paper published in 1986, Sato describedthat impurity NH4Cl shifted the absorption threshold of TiO2into the visible light region upon calcination [434]. The as-calcined material, assigned to NOx-doped TiO2, showed higherphotocatalytic activity for oxygen isotope equilibration and theoxidation of carbon monoxide and ethane than pristine TiO2 inthe visible light region. This work, however, was not noticed forquite a long time. Fifteen years after Sato’s paper, Asahi et al.reported visible light photocatalysis with N-doped TiO2 [435].They prepared the photocatalysts by two methods: one was bysputtering a TiO2 target in an N2/Ar gas mixture; the other wasby calcining TiO2 powder in an NH3/Ar atmosphere. Bothmethodsproduced yellowish TiO2 photocatalysts that noticeably absorbedlight at less than 500 nm and showed activity under visibleillumination for decomposition of acetaldehyde and decolorationof methylene blue solution (Fig. 4.33). Nitrogen atoms werebelieved to substitute for lattice oxygen atoms, as evidenced by thepeak at 396 eV in the N 1s XPS spectra.Asahi’s work is considered as a breakthrough in visible light

photocatalysis. After that, researchers started to consider thedoping TiO2 with non-metal atoms in order to obtain visible lightphotoactivity, with N-doped TiO2 being the most studied system.There have been a great volume of publications that deal withthe preparation of N-doped TiO2 by physical or chemical methods,including sol-gel [437,442,445,448,449,452,461], sputtering [440,453], ion implantation [455,456], and plasma-enhanced chemicalvapor depositionmethod [454]. Either substitutional or interstitialnitrogen is assigned to the observed photocatalytic activityunder visible illumination. Various spectroscopic techniques [449,451,452,457,462,463], surface analysis techniques [455,456], aswell as theoretical calculations [435,447,449,450,456] have beenapplied to N-doped TiO2 in order to elucidate the origin of thevisible light photoactivity and to develop visible-light-responsivephotocatalysts with higher activity.C-doped TiO2. Khan et al. reported in 2002 that the controlledcombustion of Ti metal in a natural gas flame could produce

Fig. 4.33. (a) Optical absorption spectra of TiO2−xNx and TiO2 films. (b) N 1s XPS spectra of TiO2−xNx and TiO2 films. (c) CO2 evolution as a function of irradiation time (lightat zero) during the photodegradation of acetaldehyde gas under UV irradiation and visible irradiation [435].© 2001, The American Association for the Advancement of Science.


C-doped TiO2 with an absorption threshold that was extendedinto the visible spectral range [465]. The preparation methodwas similar to that first used by Fujishima and Honda in theearly 1970s for the photoelectrochemical splitting of water [54].Khan et al. claimed that carbon substituted for some of thelattice oxygen atoms in their sample. Besides the lower band-gap energy than rutile (2.32 vs. 3.00 eV) estimated from theabsorption spectrum, the C-doped TiO2 material was reportedto photoelectrochemically split water with the unbelievably highphotoconversion efficiency of 8.35% under Xe lamp illumination[465]. Their work was critically commented on by several groups[221,466–468]; the main criticism was the high photoconversionefficiency. It should be noted that, in their paper Khan et al.did not provide any evidence of carbon doping or photoactivityunder visible illumination. A later study by Bard and co-workersshowed that TiO2 nanotube arrays annealed under controlled COgas flow could split water under visible illumination, but they didnot try to estimate the photoconversion efficiency [470]. Recently,Grimes and co-workers have actively investigated carbon-dopedTiO2 nanotubes for photoelectrochemical water splitting and havereported efficiencies as high as 6.8% [488–490].Irie et al. reported the preparation of carbon-doped TiO2

by oxidative heating of TiC powder [469]. Either anatase orrutile type C-doped TiO2 was prepared through control of theheating temperature. The prepared materials were yellowish withnoticeable shifts of the absorption threshold to the visible spectralregion. The carbon dopant showed a peak at 281.8 eV in theC 1s XPS spectra, which was assigned to substituted carbon. Asimilar peak was also observed on the C-doped TiO2 samplesprepared by reactive sputtering [473] and ion-assisted electron-beam deposition methods [474]. The photocatalytic activity of C-doped TiO2 under visible illuminationwas quiteweak, as evaluatedby the decomposition of gaseous 2-propanol [469].S-doped TiO2. Umebayashi et al. reported that sulfur doping wasunable to extend the absorption of TiO2 into the visible region[475–477]. They prepared S-doped TiO2 either by oxidative heatingof TiS2 powder [475,476] or by sulfur-ion implantation [477]. Ineither case, a red-shift of the absorption threshold was observedcompared to pristine TiO2, and the prepared materials were ableto generate photocurrent and to decolor methylene blue solutionunder visible illumination. Umebayashi et al. explained theobserved photoactivity in the visible spectrum to the substitutionof oxygen atoms with sulfur, which caused the narrowing of theband-gap [477].Ohno et al. found that S-doped TiO2 could be easily prepared by

a chemically modified sol-gel process [478–480]. They preparedthe precursor powder by mixing titanium isopropoxide withthiourea; heating the powder produced yellow-colored S-dopedTiO2. The oxidation states of sulfur in Ohno’s samples, however,were S4+ and/or S6+, inconsistent with the results of Umebayashiet al. [479]. Sakthivel and Kisch followed the experiments of Ohnoet al.; however, they did not find any evidence of sulfur doping.Instead, they claimed the formation of N-doped TiO2 [442]. Thisdisagreement suggests the complexity of the chemistry involvedin the preparation, and the possible misreading of characterizationresults.Others. There are several reports that deal with doping TiO2 withB [482], P [483], or F [484–487] atoms by the sol-gel method.The doped TiO2 materials are all yellowish, similar to N-, C-, orS-doped TiO2, and have been reported to exhibit activities fordecomposition of organic substances.TiO2−x, the oxygen-deficient form of TiO2, has also been

studied extensively as a visible-light-responsive photocatalyst[428–433]. Nakamura et al. reported the preparation of TiO2−xthrough the H2 plasma treatment of TiO2 powders [428]. TheTiO2−x photocatalysts showed weak absorption in the region

of 400–500 nm, and were able to oxidize nitrogen oxidesunder visible illumination. Justicia et al. prepared TiO2−x filmsby metal-organic chemical vapor deposition, using titaniumtetraisopropoxide as a Ti precursor and N2 as the carrier gas [430].Their preparation condition was quite similar to several chemicalpreparations of non-metal-doped TiO2 materials, i.e., a reducingatmosphere was involved in the preparation [470,471].

4.3.2. Origin of visible light photoactivityAlthough there has been much published work on non-metal-

doped TiO2 materials, the origin of visible light photoactivityof these materials is still in debate [435,449,450,452,459]. Thecontroversies are focused on two issues: (1) the origin ofthe absorption in the visible light region; and (2) in whatstate the non-metal atoms are doped in the TiO2 lattice. As arepresentative, N-doped TiO2 has been examined intensively byvarious experimental and theoreticalmethods. In particular, recentworks that combine density functional theory calculations withsurface analysis techniques and spectroscopic characterizationhave shed light on the origin of the visible light activity of N-dopedTiO2 [449,452,456]. These also aid the understanding of other non-metal-doped TiO2 materials.Origin of absorption in the visible light region. In their paper in

2001, Asahi et al. suggested that the N 2p level could mix withthe valence band of TiO2, which is mainly made up of O 2p, andthis resulted in the narrowing of the band-gap and photocatalyticactivity in the visible light region (Fig. 4.33) [435]. They calculatedthe densities of states of the substitutional doping of C, N, F, P, orS or O in the anatase TiO2 crystal by the full-potential linearizedaugmented plane wave (FLAPW) formalism in the framework ofthe local density approximation (LDA). Their calculation supportedthe idea of band-gap narrowing with N doping.The model of band-gap narrowing was soon challenged by

several experimental and theoretical studies on N-doped TiO2materials [437–441,443,447,459]. As commented by Serpone, itis hard to believe that a low level of doping (<2 atomic %) canrigidly shift up the valence band [459]. Irie et al. reported thatthe quantum yields of gaseous 2-propanol decomposition on N-doped TiO2 powders under visible illumination were several timeslower than those under UV illumination [438]. This observationcould not be explained by the band-gap narrowing model. Instead,it suggested that nitrogen doping led to the formation of localizedmidgap states above the valence-band edge (Fig. 4.34). Torreset al. studied the photoelectrochemical behavior of N-doped TiO2for water oxidation [440]. They observed that UV light gives ahigher probability than visible light for the electrons and holes toleave the electrode and contribute to a photocurrent and oxidizingwater, which is consistent with the conclusion of Irie et al. [438].Nakamura et al. observed that an N-doped TiO2 photoelectrodecould oxidize SCN− and Br− under UV illumination; however,no oxidation was observed under visible illumination [439]. Thisstudy also points out the possibility that the dopednitrogen speciesgives rise to a midgap level slightly above the top of the valenceband. A similar conclusion was drawn by Tachikawa et al. intransient spectroscopy studies on N-doped TiO2 [457], as well asC- and S-doped TiO2 [481].Lin et al. used the spin-polarized plane-wave pseudopotential

method, based on density functional theory, to calculate theelectronic-band structures and the optical absorption spectra ofnitrogen-doped TiO2 [450]. According to their calculation results,substitutional nitrogen should give rise to localized N 2p acceptorstates above the valence band, and absorption in the regionbetween 400 and 500 nm in the calculated spectrum. Moreover,the calculation reveals that even for high nitrogen concentration(12.5% doping), the N 2p states are still localized, lying slightlyabove the top of the O 2p valence band. Their conclusion is in


Fig. 4.34. Various schemes illustrating the possible band-gap electronic structureand excitation processes of visible-light-responsive TiO2 materials: (a) pure TiO2;(b) band-gap narrowing model for non-metal-doped TiO2; (c) oxygen-deficientTiO2; (d) localized midgap level model for non-metal-doped TiO2; (e) oxygenvacancy levels and non-metal-doped midgap levels are considered together.

Fig. 4.35. Valence-band spectra for Ar+-reduced and N-implanted rutile TiO2(110), acquired with a photon energy of 325 eV. Dashed lines denote the referencespectrum from stoichiometric TiO2 (110) [456].© 2006, American Institute of Physics.

contrast to the spin-restricted FLAPW calculation of Asahi et al.[435], but is consistent with spin-polarized calculations [447] andseveral experimental results [437–441]. They also found that atleast 20% N doping is needed to mix the N 2p state with the O 2pvalence band, but in practice such a high doping level would resultin the formation of TiN [448,462].The formation of localized midgap states is further proven

by recent UV-photoelectron spectroscopy (UPS) studies on N-implanted TiO2 single crystals [455,456]. Batzill et al. dopedrutile and anatase single crystal surfaces by low energy nitrogenion implantation (N. 3 at.%); the single N 1s peak at a bindingenergy of 396 eV suggests that the doping is of a substitutionaltype [455]. Ultraviolet photoemission studies show that N dopinginduces localized N 2p states within the band-gap just abovethe valence band, but no evidence of band-gap narrowing isobserved in this study. For the rutile (110) surface, N doping shiftsthe edge up by 0.4 eV; for the anatase surface, the N-derivedstates span the region in the band-gap from the valence bandmaximum up to the higher lying Ti 3d defect states [455]. In

later synchrotron-based photoemission studies, Nambu et al. alsoreached the same conclusion on the N-doped rutile (110) surface[456]. They compared the valence-band spectra of Ar+-reducedand N-implanted rutile TiO2 (110), and found that both spectracontained new states just below the Fermi level EF that could beassigned to the occupied 3d states of Ti3+, but for the N-implantedTiO2, they observed extra features above the top of the O 2p band,which shifted the edge by ∼0.5 V towards EF (Fig. 4.35) [456].Considering that the conduction-band edge is little influenced bynitrogen doping [442], the N-derived states should result in a red-shift of the absorption edge of N-doped TiO2 materials.The doping of nitrogen facilitates the formation of oxygen

vacancies in TiO2 materials, which is observed as midgapstates below the conduction band [440,449,451,452,455,456]. ThisexplainswhyNdoping does not change the n-type semiconductingTiO2 to p-type. In the case of the rutile (110) surface, N doping ledto a 1 × 2 type surface reconstruction, triggered by the increasedformation of oxygen vacancies [455]. EPR and UPS studies, andDFT calculations revealed the existence of attractive interactionsbetween the dopants and the O vacancies [449,452,456]. Thepresence of nitrogen dopants was found to reduce the formationenergy of oxygen vacancies [451,452]. At the same time, theexistence of O vacancies stabilized the N impurities [449,452,456].When oxygen vacancies and N impurities are together, there is anelectron transfer from the higher energy 3d band of Ti3+ to thelower energy 2p band of the nitrogen impurities [449,452,456].Oxygen vacancies also contribute to the absorption in the

visible light region (Fig. 4.36) [450]. This point was often neglectedin studies on non-metal-doped TiO2 [459,460]. Lin et al. calculatedthe absorption spectra of oxygen-deficient TiO2 and found that theentire absorption edge of oxygen-deficient TiO2 is red-shifted byabout 20 nm with the dominant visible light absorption occurringabove 500 nm (Fig. 4.36) [450].Doping states of nitrogen atoms. As the direct evidence of

nitrogen doping, XPS studies on N-doped TiO2 materials often givetwo kinds of N 1s peaks, one at about 396 eV, and the other one atabout 400 eV [452]. Sato observed a peak at 400 eV for TiO2 powderheated in the presence of ammonium chloride and assigned thepeak to an NOx species [434]. Asahi et al., however, observed anadditional peak at 396 eV for a TiO2 film prepared by the sputteringmethod in addition to the peak at 400 eV [435]. They assignedthe 396 eV peak to substitutional N doping, and argued thatonly substitutional N is responsible for the observed photoactivityunder visible illumination, since the 400 eV peak is also observablefor pristine TiO2 powders. In later studies, the peak at about 400eVwas frequently observed, in particular for the samples preparedby chemical methods [438,442,444,452,461]. In some cases, onlythe peak at 400 eV was observed for N-doped TiO2 materialsthat exhibited photocatalytic activity under visible illumination[442,452,461]. It was thus believed that at least two kinds ofnitrogen doping are responsible for the visible light photoactivity.The 396 eV peak is generally assigned to substitutional nitrogendoping; while the 400 eV peak is believed to related to interstitialnitrogen doping.Di Valentin et al. studied N-doped TiO2 powder prepared by the

sol-gel method with EPR combined with DFT calculations [449].They observed the presence of two slightly different nitrogenspecies in the EPR spectra. On the basis of comparisonwith the DFTcalculations, one of the two species was assigned to substitutionalnitrogen atoms, and the other was assigned to interstitial nitrogenin the form of NO (Fig. 4.37). In another study, Reyes-Garciaet al. studied N-doped TiO2 photocatalysts with 15N solid stateNMR and EPR characterization [461]. Their samples were preparedby high temperature nitridation of TiO2 nanoparticles in thepresence of urea, or by a sol-gel preparation with various N-containing chemicals as nitrogen sources. Interstitial nitrogen


Table 4.4Preparation and characteristics of N-doped TiO2 films.

Author (Ref.) Preparation method Description Photocatalytic activity

Asahi et al. [435] Sputtering deposition in N2/Ar, Orannealing TiO2 powders in NH3 at873K

Absorption onset: 500 nm; N1s peaksat 396 eV and 400 eV; nitrogenconcentration: 1–1.4 atomic %.

Decompose methylene blue and acetaldehyde invisible irradiation, with similar activity in UVirradiation as non-doped TiO2 .

Burda et al. [446] Nitriding TiO2 nanoparticles withtriethyl amine.

Absorption onset: 600 nm, N1s peak at401.3 eV

Decompose methylene blue in solution by visiblelaser.

Sakthivel Sakthivel et al. [443] Hydrolyzing TiCl4 in water withaddition of ammonia, post-annealingat 673 K.

Absorption onset: 520 nm; a weak N1speak at 404 eV

Degradation of 4-chlorophenol, benzene andacetaldehyde with visible light (λ > 455 nm),

Irie et al. [438] Annealing TiO2 powders in NH3 N-substituted for oxygen: 0.5–1.9% Degradation of gaseous 2-propanol under visiblelight irradiation; UV light activity was weaker fordoped samples than non-doped. Visible lightactivity was several times lower than UV lightactivity.

Mrowetz et al. [441] One sample (a) was prepared byBurda’s method, the other sample (H)was prepared similarly to Asahi’s andIrie’s works.

Increased absorption in the visiblelight region for both sample A andsample H.

Both samples A and H showed no apparentactivity for degradation of HCOO− and NH3 · H2Oin visible irradiation; sample A was bleachedwith UV irradiation.

Diwald et al. [444] Annealing rutile (110) in NH3 at 873 K. N1s peaks at 399.6 and 396.7 eV;increased absorption in the range of2.4–3.0 eV.

Photodeposition of Ag from AgNO3 solution invisible irradiation.

Maeda et al. [454] PECVD in NH3/Ar using Ti(OiPr)4 asprecursor; post-annealing over 673 K.

N1s peaks at 396 eV and 399.3 eV,Area ratio of the two peaks is 4:1.

Degradation of stearic acid molecules undervisible irradiation.

Kitano et al. [453] RF Magnetron sputtering in N2/Ar N-substituted for oxygen: 2–16.5%;photo-response up to 550 nm.

Degradation of 2-propanol solution with visibleirradiation; 6% N-substitution is the best.

Yates et al. [491] Chemical vapor deposition in NH3/N2 ,using TiCl4 and ethyl acetate asprecursor.

Nitrogen atomic concentration: 1.5–5% No visible light activity for degradation of stearicacid molecules; weak UV-activity.

atoms were observed in their studies, and these were assigned tonitrates (NOx).Di Valentin et al. calculated the electronic-band structures for

substitutional and interstitial N-doped anatase TiO2 [449]. For bothof these N-doped model systems, the formation of localized statesin the band-gap was predicted (Fig. 4.37). Substitutional nitrogenstates lie just above the valence band, while interstitial nitrogenstates lie higher in the gap [449]. The localized nature of theN-induced states has the consequence that the hole generatedby visible irradiation is less mobile than that generated by UVirradiation. In particular, the interstitial N impurities, which giverise to the higher energy states in the gap,might behave as strongerhole trapping sites, reducing the direct oxidation power of thesample in the photocatalytic process [449]. The calculations ofDi Valentin et al. also suggested that there is a cost to inducesubstitutional nitrogen formation starting from the interstitial one.In an excess of oxygen and nitrogen, interstitial nitrogen doping isdefinitely preferred. However, under highly reducing conditions,as is the case after annealing at high temperature, substitutional

nitrogen species in parallelwith oxygen vacancies could be favored[449].

4.3.3. Activity and stability of N-doped TiO2 photocatalystsAsahi reported that nitrogen doping offered visible light (VL)-

photocatalytic activity to TiO2, without any loss of UV-activity, forthe decoloration of methylene blue solution and degradation ofgaseous acetaldehyde [435]. However, many studies after Asahi’swork showed that nitrogen doping only offeredmodest VL-activityto TiO2, but at the same time lowered the UV-activity (Table 4.4)[438,442,443,463,464]. Some studies even showed almost no VL-activity for N-doped TiO2 (Table 4.4) [441,491]. It thus seems thatthe preparation history, which dictates the types and level ofnitrogen doping and the concentration of oxygen vacancies, greatlyinfluences the photocatalytic activity.Mrowetz argued that the often-used probe, methylene blue,

is not a good indicator for photocatalytic activity [441]. Methy-lene blue solution could be decolored via several pathways, in-cluding oxidation with holes or hydroxyl radicals, reduction with

Fig. 4.36. (a) Calculated optical absorption spectra for various N concentrations in the polycrystalline TiO2: (I) undoped TiO2; (II) 12.5% nitrogen doped; (III) 6.2% nitrogendoped; and (IV) 3.1% nitrogen doped. (b) Calculated optical absorption spectra of polycrystalline TiO2 with different O vacancy contents: (I) undoped TiO2; (II) 12.5% oxygenvacancies; (III) 6.2% oxygen vacancies; (IV) 3.1% oxygen vacancies [450].© 2005, American Chemical Society.


Fig. 4.37. Schematic sketch and electronic-band structure for (a) substitutional and(b) interstitial N-doped anatase TiO2 . The calculated value of the band-gap is alsoreported [449].© 2005, American Chemical Society.

conduction-band electrons, sensitized photodegradation, and ad-sorption on catalyst surfaces [492–495]. Thus, they studied theVL-activity of N-doped TiO2 by the oxidation of HCOO− anions inwater [441]. They used two samples: the first, designated ‘‘A’’, wasprepared by Burda’s method, that is, nitriding TiO2 nanoparticleswith triethyl amine [445,446]; the second, designated ‘‘H’’, wasprepared by annealing TiO2 powders in NH3 with a post-annealingin air. They found that sample A suffered bleaching under UV irra-diation. They inferred that the reaction between the TiO2 nanopar-ticles and triethyl amine at room temperature led to the formationof surface organotitanium complexes, rather than to substitutionalN-doped TiO2, the former being readily degraded by UV light [441].Sample H, although doped with nitrogen atoms, did not show ap-preciable VL-activity for the degradation of HCCO− and NH+4 ions[441]. Moreover, no appreciable hydroxyl radicals were detectedwith ESR for H under visible irradiation. Their results are contrastto those of Asahi [435] and Irie [438], who used similar methodsto prepare N-doped TiO2 materials. One possible reason is that thepost-annealing in Mrowetz’s work might re-oxidize the surface ofsample H, which has a detrimental influence on VL-activity.Livraghi et al. studied the VL-activity of N-doped TiO2 bymeans

of EPR spectroscopy [452]. They found that nitrogen dopants formeither diamagnetic (N−b ) or paramagnetic (N

•

b ) bulk centers. Both

types of N•b centers give rise to localized states in the band-gap of the oxide. The diamagnetic N−b species are expected tobe more abundant than the paramagnetic N•b ones, since theyare energetically favored. Irradiation with visible light at 437 nmpromotes electrons from these states to the conduction band andthus leads to the increase of paramagnetic N•b centers, as shownin Fig. 4.38. When the irradiation at 437 nm is performed in anoxygen atmosphere, the increase of the N•b signal is accompaniedby the simultaneous appearance of new EPR lines corresponding toa surface superoxide O•−2 radical species (Fig. 4.38), suggesting thatthe photoexcited electrons are available for chemical interactionsat the surface [452]. The holes, formed by charge separation,remain localized on the N•b centers, which limits the VL-activity ofN-doped TiO2 [452].Emeline et al. measured the QYs of phenol degradation over a

commercial VL-responsive TiO2 photocatalyst [458]. They found,under optimum conditions, where recombination losses wereminimal, that the QYs at 365 nm and 436 nm were 12% and8%, respectively. The QY of phenol degradation over Degussa P25was 14% at 365 nm measured by the same experimental system[458]. Emeline’s work was done under ‘‘light-limited’’ conditions,which is by far different from the ‘‘light-rich’’ conditions of actualphotocatalytic systems. However, this suggested that the VL-activity of N-doped TiO2 could be comparable to the UV-activity,provided that the recombinative losses introduced by oxygenvacancies were suppressed and the absorption of visible light wereimproved.Intense efforts have been devoted to the development of prepa-

rationmethods for highly efficientN-dopedTiO2materials in termsof increasing both the VL-absorption and the activity. Maeda andWatanabe developed a plasma-enhanced chemical vapor deposi-tion method to prepare N-doped TiO2 [454], which ensured thesubstitutional doping of nitrogen, in contrast to the sol-gel methodor annealing in NH3, which often produced predominantly inter-stitially doped nitrogen [434,441,452,461]. Their film photocata-lysts were able to decompose stearic acid molecules under visibleirradiation, suggesting promising applications in VL-activated self-cleaning materials [454].Kitano et al. developed a RF magnetron sputtering technique

to prepare N-doped TiO2 films [453]. By their method, nitrogenatoms were able to substitute oxygen in the range of 2–16.5%.Such a high-level doping was not achieved previously and resultedin a steep absorption edge in the visible light region [453]. Afilm photocatalyst with a nitrogen concentration of 6% (strongabsorption up to 500 nm) exhibited the highest reactivity forthe photocatalytic oxidation of 2-propanol in water under visible

Fig. 4.38. (a) EPR signal intensity of N•b (substitutional and interstitial N) and O•−

2 in various conditions. I0 is the N•

b intensity in the non-irradiated sample. ‘‘Blue’’ indicates437-nm irradiation, while ‘‘Green’’ indicates 500-nm irradiation. (b) Sketch of the proposed mechanism for the processes induced by visible light irradiation of the N-dopedsample in O2 atmosphere [452]. N−b and N

•

b are nitrogen dopants forming paramagnetic and diamagnetic bulk centers, respectively.© 2006, American Chemical Society.


irradiation (λ > 450 nm). Higher nitrogen concentration wouldintroduce large amounts of oxygen vacancies in the films and resultin lower VL-activity [453].The stability of N-doped TiO2 is also an important issue

concerning actual applications. It was reported that N-doped TiO2exhibited better thermal stability than oxygen-deficient TiO2, as aresult of the interaction between the nitrogen dopants and oxygenvacancies [437]. The latter lost VL-absorption due to oxidationupon heating at 473 K [437]. However, Kitano et al. observedthat N-doped TiO2 lost VL-absorption gradually upon heatingat 573 K, accompanied by a decrease in nitrogen concentration[453]. Kitano et al. also examined the stability of N-doped TiO2film under irradiation [453]. They found that the film afterphotoelectrolysis of water under visible irradiation showed adecreased N concentration at the surface layer, indicating that thesurface of the filmwas oxidized during photoelectrolysis. These arechallenges for the further development of these types of materials.

5. Fundamentals of the photo-induced hydrophilic (PIH) effect

5.1. Overview

Our group first reported, along with co-workers from the TOTOCorp., in 1997 on the phenomenon that we termed the ‘‘light-induced amphiphilic surface’’ [496]. When a titania film wasilluminated with UV light, the contact angle for water decreasedto near 0◦, and the same occurred also with organic liquids. Weexpected that there would be several applications for this neweffect, including self-cleaning surfaces and anti-fogging mirrors.With friction force microscopy, it was observed that there wasa spontaneous formation of a pattern of two different types ofareas with hydrophobic and hydrophilic properties (Fig. 5.1). Wehypothesized that the presence of two types of surfaces at the tensof nanometer scale was responsible for the high affinity for bothwater and organic liquids, hence the term ‘‘amphiphilic surface’’.The following year, we published a more detailed article

and included a proposed concept for the mechanism, basedin part on the observed infrared spectrum, which showed thereversible growth and decay of a peak (3695 cm−1) assigned to theformation of hydroxyl groups, presumably at siteswhere there hadbeen dissociative adsorption of water at oxygen vacancies [514].Even though the ideas have become much more detailed in themeantime, the same basic idea is still valid. The large numberof studies that have been described in Section 3.4 have led toa consensus that water is adsorbed molecularly under neutralpH conditions on the stoichiometric rutile (110) surface. Thus,the interaction with water is not particularly strong, and theintroduction of hydroxyl groups would be expected to lead to asurface with a stronger interaction with water.The effect of enhanced hydrophilicity was found to slowly

revert to the normal, hydrophobic state during storage of the filmin the dark, this process requiring varying amounts of time. It wasfound that ultrasonication of the film accelerated the reconversionto the initial state; with low-intensity agitation (0.6 mW cm−2),the process was found to take ca. 140 min, while, with high-intensity agitation, it required only 20 min [497]. Subsequent UVillumination served to make the surface hydrophilic again.Most of the experiments that were carried out involved pure

anatase TiO2 films, but it was found that the addition of 10–30wt%SiO2 to a TiO2 film improved the hydrophilic properties [498]. Thepresence of SiO2 helped the films to retain higher surface area andto suppress the thermal conversion of anatase to rutile during filmheat treatment.One of the first mechanisms proposed included a step in which

bridging oxygens, such as those on the rutile (110) surface, werethought to be removed from the surface during UV illumination

Fig. 5.1. (a) Friction force microscopic (FFM) image (5 × 5 µm) for a rutile (110)surface before UV illumination; (b) FFM image (5×5µm) of the same surface afterillumination; (c) a medium scale FFM image (1000 nm × 1000 nm) of the framedarea in (b); (d) a higher-magnification topographic image (240 × 240 nm) of theframed area in (c). For (d), the sample stage was rotated 45◦ from the position in (c)(taken fromWang et al. [514]).© 1998, John Wiley & Sons.

(see later), and thus it was of interest to examine the influence ofthe crystal face exposed. Our study with single crystals of rutilefound that the (110) and (100) faces had similar characteristics,i.e., both became hydrophilic at about the same rate under UVillumination [499,500]. However, the (001) surface was foundto be much slower in becoming hydrophilic. Under a pure O2atmosphere, it was found that, for all three surfaces, (110), (100)and (001), the effect was partly suppressed; indeed, with the(001) surface, the effect was almost completely suppressed. X-ray photoelectron spectra showed that the (110) and (100) facesacquired a small peak in the Ti 2p spectrum that was attributableto Ti3+ sites, while the (001) face did not. Thus, the idea of UVinduction of oxygen vacancies was supported. However, otherpossible explanations exist, as discussed later.The photo-induced hydrophilic (PIH) effect was initially

thought to involve simply the photocatalytic removal of hydro-carbon contamination films, but this was found to be an inade-quate explanation. One of the experiments that underscored thisinvolved the comparison of TiO2 and SrTiO3 (501). Thesewere bothprepared as polycrystalline films on glass. Bothwere found to pho-tocatalytically decompose methylene blue, but the SrTiO3 film didnot become hydrophilic under illumination (Fig. 5.2).In an effort to continue improving the characteristics of the PIH

effect, variations were examined. One of the most promising onesinvolved the deposition of a thin film of WO3 on the TiO2 surface[427]. An amorphous film was found to be more effective than acrystalline film. With the former, a PIH conversion was observedto reach completion after 100 min at the surprisingly low UV lightintensity of 1 µW cm−2.The amphiphilic effect is unique for TiO2: no other oxide that

has been examined thus far exhibits this behavior. In furtherexploration of this effect, it was found that, the oleophilic behaviorof TiO2 films was actually reconverted to oleophobic after about10 h of continuous UV illumination at 1 mW cm−2 [502], whilethe hydrophilic character remained. It was speculated that this


Fig. 5.2. Comparison of the photocatalytic and PIH behavior of TiO2 and SrTiO3: (left panel) time dependence of absorbance under UV irradiation ofmethylene blue adsorbedon TiO2 (solid diamonds), on SrTiO3 (open squares), and on an uncoated glass substrate (open triangles); (right panel) time dependence of water contact angle under UVirradiation of TiO2 (solid diamonds), and SrTiO3 (open squares). UV irradiation was 1 mW cm−2 under ambient conditions (295 K; relative humidity, 60%, air) (taken from[501].

behavior resulted from a redistribution of the nanometer-scalehydrophobic (oleophilic) domains with time.The effect of repeated cycles of UV illumination followed

by ultrasonication was studied for rutile (110) and (001) films[503]. Interestingly, although the (001) surface was found to bequite slow compared to the (110) surface in reaching hydrophilicbehavior, taking about three times as long, it exhibited anincreasingly rapid effect with cycling; by the third cycle, it reacheda fully hydrophilic state after less than 50 min under a UV lightintensity of 40 mW cm−2. This was thought to be the result ofrestructuring of the surface to expose more bridging oxygens.The influence of electrochemical potential was examined, and

it was found that more positive potentials were effective inaccelerating the PIH effect [504]. This effect reached a plateau inthe 0.4–0.8 V vs. Ag/AgCl range in aqueous pH 6.86 buffer. As thepotential was held at increasingly negative potentials, starting atabout −0.2 V, the PIH effect began to be suppressed, with thesuppression increasing with negative potential. This phenomenonled to a modification of the mechanism; it was concluded thatphotogenerated holes are trapped at bridging oxygens, causingthem to be ejected from the surface, thus allowing a watermolecule to dissociate at the resulting oxygen vacancy. Thismechanism will be discussed in more detail below. It shouldbe noted that the type of film studied in this work begins toabsorb light at longer wavelengths (monitored at 700 nm) as thefilm is maintained at potentials below about −0.8 V vs. Ag/AgCl,which is an indication that electrons are building up, either in theconduction band or in electronic states close to the conductionband. In this region, it becomes more difficult to separate thephotogenerated electrons and holes, since the film behaves morelike a metal and less like a semiconductor.The electrochemical potential was also found to have an

effect even in the dark, but this effect was rather weak forconventional polycrystalline anatase films [505]. However, foramorphous titania films, the effect was much more obvious. In thepotential region below −0.6 V vs. Ag/AgCl, the amorphous filmbegan to absorb light (monitored at 700 nm). At this potential,the contact angle also began to decrease in the dark for theamorphous film. At more negative potentials, the effect was morepronounced: at −1.0 V, the contact angle approached 0◦ after20 min. In contrast, for the polycrystalline film, even at a potentialof −1.2 V, the contact angle only approached 12◦ after 20 min.The photo-response of the amorphous film to UV light was foundto be very small, with only a small µA cm−2 photocurrent being

Fig. 5.3. Trace of water contact angle (WCA) vs. time for an anatase film sample,showing reversible lowering of WCA under UV illumination and increasing WCAunder visible light illumination (taken from [506]).© 2002, Elsevier Science.

generated over a wide range of potential under illumination at360 nm and an intensity of 50 mW cm−2, whereas the anatasefilm generated about 1.5 mW cm−2 photocurrent at 0.8 V. Thehydrophilic effect for the amorphous film was thought to involveEq. (3.2), thus generating surface hydroxyl groups,which should bemore hydrophilic. This is still not a complete explanation, however,because, if the effect just involves the surface, the anatase filmshould also produce nearly the same degree of hydrophilicity.There is clearly someother factor involved, for example, roughness.It was discovered that visible light illumination, without

UV, is actually effective in reconverting anatase films from ahydrophilic state back to a semi-hydrophobic state (Fig. 5.3)[506]. This effect was found to result from the visible-light-induced heating of the film. The highly reversible alternation ofthe hydrophilic and hydrophobic character is one of the mostconvincing demonstrations that this effect is not simply due tothe build-up and destruction of contamination layers. These resultswere also confirmed by Stevens et al. [507].The PIH effect has also been examined on other metal oxides,

for example, ZnO [508]. This oxide also exhibits a pronounced


effect. Initially, the ZnO surface is much more hydrophobic thanthe TiO2 surface, exhibiting a water contact angle of ca. 110◦. Thecontact angle for these films also increases during storage in thedark in a manner similar to that for TiO2. It was proposed that theZnO surface can also be induced by UV light to generate oxygenvacancies. This point will be discussed later. As with TiO2, ZnO alsoreconverts to a hydrophobic state at different rates according to thetype of atmosphere. For example, the conversion is fastest in pureO2. This is thought to be due to a competition between O2 and H2Ofor reaction with oxygen vacancies.A number of other metal oxides were also examined for their

PIH activity and photocatalytic activity [509]. The following oxideswere found to exhibit the photocatalytic effect: TiO2, SrTiO3 andZnO, with SnO2 exhibiting only a slight effect. The following werefound to exhibit a PIH effect: TiO2, SnO2, ZnO, WO3 and V2O5.Other oxides studied were: CeO2, CuO, MoO3, Fe2O3, Cr2O3 andIn2O3. All of the latter group exhibited neither effect. These resultsare also one of the most convincing with regard to distinguishingthe two effects. As already noted, SrTiO3 is the clearest exampleof an oxide that exhibits the photocatalytic effect but not PIH;in this work, there are two examples of oxides that exhibit PIHbut not photocatalysis: WO3 and V2O5. Similar to TiO2, theseoxides are able to lose oxygen, creating oxygen vacancies to thepoint where the structure can change, as in the Magneli phases.However, an exception to this correlation is MoO3, which is alsoknown to be able to convert to Magneli phases. This explanationrequires modification, because the loss of substantial amounts ofoxygen can also lead to metallic conduction, which suppressesphotoeffects that normally occur in semiconductors.In an effort to make the PIH measurements more precise, Sakai

et al. used the reciprocal of the water contact angle to quantifythe conversion rates from hydrophobic to hydrophilic and back[510]. In this work, a correlation was found between stronglyadsorbed water, as observed as a component in the O 1s XPSspectra, and the reciprocal of the contact angle. Temperature-programmed desorption was also used to measure the amountsand types of adsorbed water. The reconversion of the surface tohydrophobic was found to exhibit a clear transition between a fastand a slow process, the latter being thought to involve a structuralreconstruction process.The influence of tensile and compressive stress on the PIH effect

is surprising [511]. Tensile stress applied to a TiO2 film enhancesthe effect, while compressive stress suppresses the effect. Thisinteresting phenomenon was explained in terms of the intrinsicstress created when the surface undergoes the reconstructiondiscussed above. Related to this is a study in which it was observedthat a TiO2 film became significantly rougher under illuminationand then reverted to its smooth texture in the dark, as reported byWatanabe and coworkers [521].Another related study, carried out with in situ grazing angle

incidence X-ray diffraction, showed that illumination of a rutilesingle crystal in the presence of various alcohols led to a reversibleexpansion of the lattice [512]. One possible explanation is theelectrochemical intercalation of protons Eq. (3.8), leading to thetype of lattice expansion shown in Fig. 3.15. A more recent studycarried out with the X-ray truncation rod technique in aqueoussolution without an organic compound present to trap holes, ledto a disordering of the single crystal surface [513]. This result wasconsidered to be consistent with the type of domain formationreported by Wang et al. in our early papers on the PIH effect [496,514].More recent work has focused on newways to enhance the PIH

effect. For example, Gu et al. found that nanotexturing of the TiO2surface can lead to an extremely long-lived hydrophilic character,lastingmore than one year, in comparison to the conventional film,which becomes completely hydrophobic within 50 days [515]. The

texturing was carried out by first forming a layer of polystyrenespheres (250 nm), followed by intrusion of a suspension of 15-nm TiO2 particles, with the polystyrene spheres finally beingremoved by heat treatment in air (Fig. 5.4). Other recent work,including that aimed at enhancing the visible light sensitivity, willbe discussed later.

5.2. Mechanisms of the PIH effect

Some indications have already been given regarding the pos-sible mechanisms for the PIH effect. These will now be discussedin more detail: they include the original mechanism, which essen-tially says that the PIH effect and photocatalysis are identical; thereductivemechanism, whichwas originally proposed, the later ox-idativemechanism, and finally, a newcombined redoxmechanism,which is again similar to photocatalysis but includes both.

5.2.1. Decomposition of organic filmsThe initial thought that has occurred to almost everyone

when learning about the PIH effect is that it is a simpleprocess of photocatalytic removal of organic contamination films,which are notorious and ubiquitous. Our group first assumedthis mechanism, but later evidence, as outlined above, made itincreasingly clear that the two effects, while very closely relatedindeed, are separate.Several groups during the past several years have revisited this

issue. White et al. used the rutile (110) surface and found that thepresence of oxygen vacancies, produced by heating the crystal, didnot produce anoticeable effect on the TPDbehavior forwater [516].Both the stoichiometric and the slightly reduced surfaces werehighly hydrophilic. After the surfaces are coveredwith amonolayerof trimethylacetic acid, they become highly hydrophobic, asevidenced by TPD curves that indicate the presence of liquid ratherthan adsorbed water. After these surfaces are illuminated with UVlight for 300 s in the presence of 5 × 10−6 Torr O2, they becamemuch more hydrophilic and again exhibited the behavior of theclean surface, i.e., with first-layer, second-layer and multi-layerwater TPD peaks.The main conclusions of this work were that (1) the original

TiO2 surface is highly hydrophobic, without UV illumination,because it is clean; (2) the presence or absence of oxygen vacancieshas no discernable effect on the apparent interaction with water;and (3) the photocatalytic removal of an organic monolayer leadsto a recovery of hydrophilic properties.A second major work that advocated the contamination layer

mechanism was that of Zubkov et al. [517]. This paper showsconvincingly the effect of additions of hexane to the gas phaseabove the sample. For example,with 120ppmhexanepresent in a 1atmO2 atmosphere, there is an induction time of about 150 s; up tothat time, the contact angle is relatively constant at 20◦, and thenit drops suddenly to the range in which it is difficult to measurethe contact angle (0◦–7◦), as shown in Fig. 5.5. However, it is alsosignificant that, in the absence of hexane, the clean crystal initiallyshows a contact angle of ca. 25◦ (see curve for 0 ppmhexane). Afterthe initiation of UV illumination, it requires only ca. 12 s to reachthe 0◦–7◦ range. This result is further substantiated by results thatare shown for the effect of the presence vs. absence of O2 (Fig. 5.6).There is no doubt that contamination films cause the surface to

be hydrophobic; this issue is important and must be kept in mindin the proper design of experimental measurements in this area.In very recent work, it has become increasingly evident that thereare actually two distinct effects, the first being the destruction oforganic contaminants, which lowers the contact angle from higherranges (e.g., 60◦) to the 20◦–35◦ range and a separate effect thatlowers the value down to the 0◦–5◦ range [518]. Consistent withthis is the result already cited by Zubkov et al., as well as the recentwork of Hennessy et al. [513].


Fig. 5.4. Nanostructured TiO2 surface (a) microscopic image and (b) WCA vs. time for the conventional TiO2 (upper curve) and nanotextured (lower curve) surfaces.© 2004, American Institute of Physics.

Fig. 5.5. Traces of WCA vs. time under various atmospheres: 1 atm O2 containingadded hexane (taken from Zubkov et al. [517]).© 2005, American Chemical Society.

5.2.2. Reductive mechanismIn our papers prior to 2001, we focused on what could be called

a reductive mechanism, in which UV light was thought to produceoxygen vacancies, at which water then dissociated, creating twohydroxyl groups. There are several papers in which it is reportedthat UV light can create oxygen defects, but, as already mentioned,Mezhenny et al. showed the lack of an effect of reasonable lightintensities on the appearance of rutile (110) surfaces with STM[139]. Thus, it seems clear that this mechanism would have to bemodified. One modification that is sensible is simply to recognizethat an electron with the reducing power of hydrogen, or evenslightly less, can directly reduce a bridging oxygen on rutile (110)to produce a single OH group (reaction (3.2)). This product isvery similar, if not identical to that produced via reaction ofa thermally produced oxygen vacancy with water. The degreeof sameness or difference should in fact be carefully examined,but, to a first approximation, it should be identical. Our reporton results obtained with cathodic polarization enhancement ofhydrophilicity [505] seems to support this, but further work needsto be carried out.As White et al. point out, even in the case in which there are

indeed oxygen vacancies present on the surface, they do not haveameasurable effect on the hydrophilic properties [516]. Intuitively,we expect that the presence of bridging hydroxyls should leadto an enhanced interaction with bulk water, and this suppositionappears to be valid from the experiments just cited with cathodicpolarization [505]. However, the effect with anatase film is not

pronounced and is clearly not the dominant one in the PIH effect.Thus, it is clear that the mechanism needs to be modified.

5.2.3. Oxidative mechanismThe report that positive polarization has a strong effect on the

attainment of a hydrophilic surface demonstrates the importanceof photogenerated holes in any broadly useful mechanism forthe PIH effect [504]. The mechanism that was proposed in thatpaper again invoked the production of oxygen vacancies, and thisindeed appears to be a weakness, for the same reasons alreadydiscussed. In this case, we cannot argue that the direct productof the reaction of a vacancy with water could be the same asthat obtained electrochemically, but there is almost certainly astructure that is produced photochemically that could be similarto the one obtained electrochemically. This point also needs to beexamined in further research. The recent work of Nakato and co-workers on the photoproduction of peroxo species on the titaniasurface is important in this regard [219,220,519,520].

5.2.4. Combined redox mechanismThe mechanism already given for the general photocatalytic

processes (see Section4.1.1 and Fig. 4.5) is also proposed for the PIHeffect, because it combines both reductive and oxidative elements,both of which are important. Specifically, Fig. 4.5a, which involveswater as an electron donor, without an organic present, may beuseful. Panel A includes water as an electron acceptor, while panelB includes O2 as an electron acceptor. Either may be appropriateunder specific conditions. From the result of Zubkov et al. (Fig. 5.6),however, it can be seen that O2 is necessary to accelerate thePIH process. It is important to keep in mind, however, that O2accelerates the recovery of the original hydrophobic behavior inthe dark.As mentioned earlier, this type of mechanism might be

consistent with a report in which the effect of tensile stresswas discussed [511]. The presence of localized areas in whichhydrogen is inserted electrochemically is consistent with the ideaof volume expansion, as discussed theoretically by Koudriachovaet al. (see Fig. 3.15a, b) [160]; volume expansion can certainlylead to compressive stress, which would be alleviated by tensilestress. Volume expansion was also indicated in the work ofMatsushige and co-workers [512], inwhich in situ X-ray diffractionmeasurements were made on a rutile single crystal under UVillumination in the presence of organic hole trapping agents. Thisidea is also consistent with the report ofWatanabe and co-workerson increased surface roughness during illumination [521]. Morerecently, an in situ X-ray truncation rod study showed an increasein apparent roughness during UV illumination in the absence of a


Fig. 5.6. Traces of WCA vs. time under inert atmosphere (no O2) and under 10% O2(taken from Zubkov et al. [517]).© 2005, American Chemical Society.

hole trapping agent [513]. Thiswas suggested to be consistentwiththe original reports ofWang et al., in which surface nanostructureswere observed on a rutile single crystal during UV illumination(Fig. 5.1). It is likely that the presence of hole trapping agent in theearlier X-ray study [512] allowed the regions of proton injection toincrease in size and thus to be observed easily.The development of a general mechanism for photocatalysis

and a specific one for the PIH effect is an ongoing endeavor. Atthis point, we would like to recommend that a focused effort beundertaken to understand the possible reactions occurring on theTiO2 rutile (110) surface, both experimentally and theoretically,that might lead to PIH effects. As already mentioned, there areseveral candidates for surface entities that might enhance thehydrophilic properties, including bridging OH groups, as in thereductive mechanism [496,505,514], and surface-trapped holesof various types, as in the oxidative mechanism [504]. On theoxidative side, the intermediates involved with the O2 evolutionreaction could also be important [111,218–220]. The attack of asurface-trapped hole on adsorbed water can in principle producean adsorbed •OH radical in an electrochemical reaction, or canlead to a bond breakage, which might be more likely in the caseof a bridging oxygen. We have attempted to assemble as many aspossible of the reasonable unitary reactions, i.e., a single reactionevent, including electron transfer, proton transfer, bond breakingand bondmaking (Fig. 5.7, Table 5.1.). Such unitary reactionsmightalso be useful in building more complicated mechanisms. Wepresent these as a possible starting point for further guidelines instimulating both experimental and theoretical studies that wouldtry to finally be directed at achieving a deep understanding of themost studied oxide surface, rutile (110). If good agreement canbe achieved between experiment and theory on such a surface, itshould also be a great impetus toward the achievement of similarunderstanding of other important rutile and anatase surfaces.

5.2.5. Visible-light-induced PIH effectSeveral reports have appeared concerning the development of

visible-light-induced hydrophilicity, encouraged by the boom instudies on visible light photocatalysis [454,473,474,522]. Thesestudies focusing on N, or C-doped TiO2 films and showed quitescattered results on the rate of hydrophilic conversion due todifferences in the preparation methods of the films. Some studiesstated that the hydrophilic conversion rates were slow and thesuperhydrophilic state were not obtainable even in reasonablylong period (over 100 h) of visible irradiation [473,522], but some

Table 5.1Possible unitary reactions that might be involved in the PIH effect.

The locations of each of themoieties are indicated by the letters in the titles for eachreaction, referring to Fig. 5.7. The listed reactions are a sampling of the possible onesfor the TiO2 rutile (110) surface.

studies stated that the superhydrophilic state could be obtained forN-, or C-doped TiO2 films within hours of visible irradiation [454,


Fig. 5.7. Schematic top view of the rutile (110) surface, showing the structures thatare used in Table 5.1 regarding the possible reactions involved in the PIH effect. Thered rectangle corresponds to the designation ‘a’ in Table 5.1, the blue rectangle to‘b’, and the green rectangle to ‘c’. (For interpretation of the references to colour inthis figure legend, the reader is referred to the web version of this article.)

474]. However, all of the studies have reached the same conclusion,i.e., the rate of hydrophilic conversion is much slower undervisible irradiation than UV irradiation even when the absorbedphotons are in same number for both cases, the same as whatwas observed in visible light photocatalysis. Also, it appears thatfilms that are more photocatalytically active undergo, hydrophilicconversion more easily and rapidly under the visible irradiation[454,473,474,522].

6. Brief review of applications

6.1. Self-cleaning surfaces

The TiO2 surface can decompose organic contamination withthe aid of ultraviolet light. This observation suggests the applica-tion of TiO2 photocatalysis to a novel ‘‘self-cleaning’’ technique,i.e., a surface coated with TiO2 can maintain itself clean underultraviolet illumination (Fig. 6.1) [5,9,10,392]. This technique isobviously of great value, since it can utilize freely available so-lar light or waste ultraviolet emission from fluorescent lamps,save maintenance costs, and reduce the use of detergents. Watan-abe, Hashimoto and Fujishima first demonstrated this self-cleaningconcept on a titania-coated ceramic tile in 1992 [392]. A simi-lar idea was conceived independently by Heller [5]. One of thefirst commercialized products using this technique was the self-cleaning cover glass for highway tunnel lamps [9,14,523]. This typeof lamp, which is often a sodium lamp in Japan, emits UV light ofabout 3 mW cm−2 at the position of the cover glass. This UV lightis of no use for lighting, but it is sufficient to decompose the con-tamination from exhaust compounds. As a result, the cover glasscan maintain transparency for long-term use.The efficacy of self-cleaning surfaceswas found to be dependent

on the relative rates of decontamination vs. contamination.The TiO2 photocatalyst can maintain the surface clean onlywhen the photocatalytic decontamination rate is greater thanthat of contamination. Wang et al., however, observed that theself-cleaning effect of TiO2 surfaces could be enhanced whenwater flow, such as natural rainfall, was applied to the surface[514]. They attributed this enhancing phenomenon of water

Fig. 6.1. Schematic diagram of the decontamination process occurring on thesuperhydrophilic self-cleaning surface [15].©2006, Elsevier Science Ltd.

flow to the superhydrophilic property of TiO2 surface, i.e., waterpenetrated the molecular-level space between the stain andthe superhydrophilic TiO2 surface [14,15]. In other words, thisphenomenon has effectively removed the limitation of the self-cleaning function of TiO2 photocatalysis set by the number ofincident photons. Even though the number of photons maybe insufficient to decompose the adsorbed stain, the surfaceis maintained clean when water is supplied there. Thus, theysuggested that the best use of self-cleaning TiO2 surfaces shouldbe exterior construction materials, since these materials could beexposed to abundant sunlight and natural rainfall [9,14,15]. Suchmaterials, including tiles, glass, aluminium siding, plastic films,tent materials, cement, etc., have already been commercialized inJapan since the late 1990s and in other countries in recent years[15,524]. Some examples are shown in Fig. 6.2.As estimated by TOTO, Ltd., the pioneer of self-cleaning

technology, a building in a Japanese city covered with ordinarytiles should be cleaned at least every five years to maintaina good appearance, while that covered with self-cleaning tilesshould remain clean over a span of twenty years without anymaintenance. Therefore, self-cleaning technology can lead to largedecreases in maintenance costs. We have been examining theefficacy of self-cleaning technology at the Photocatalyst Museumof our Institute. Fig. 6.3 shows the results of one test on a self-cleaning tent material produced by Taiyo Kogyo Co. The tentmaterial, made from PVC film, was coated with TiO2 photocatalystmaterial on the left half. The test was started on July 22nd, 2004;on that day, both the non-coated and TiO2-coated halves hadthe same white appearance, as shown in Fig. 6.3a. The photoshown in Fig. 6.3b was taken on April 23rd, 2007. The TiO2-coated part remains white in the photo, whereas the non-coatedpart, however, is dark grey. Note that the tent was placed in alocation surrounded by tall buildings, where little direct sunlightillumination could reach the surface. Nevertheless, even undersuch conditions, the self-cleaning effect is obvious. On clear days,contaminants accumulate on the whole surface, but during a rain,these are washed off the TiO2-coated part of the surface.In Japan, several thousand tall buildings have been coveredwith

TOTO’s self-cleaning tiles. The Central Japan International Airport,which opened in February, 2005 near Nagoya, Japan, used over20 000m2 of self-cleaning glassmanufactured by theNippon SheetGlass Co. Self-cleaning tent materials have been widely applied forstorage structures, business facilities, bus and train stations, sportscenters, sunshades in parks and at the seaside. The PanaHomeCompany, one of the major house manufacturers in Japan, hasmarketed ‘‘eco-life’’-type houses since 2003; self-cleaning tiles andwindows, in addition to a solar cell-covered roof-top, are utilized.New applications of self-cleaning technology are also under

examination. Zhang et al. recently reported on antireflectiveTiO2–SiO2 self-cleaning coatings that may find application forthe cover glass of solar cells [525,526]. Conventional TiO2-basedcoatings are highly reflective due to the large refractive index of the


Fig. 6.2. Applications of self-cleaning exterior building materials. (a) Picture of the MM Towers, in Yokohama, coated with self-cleaning tiles (courtesy of TOTO). (b) Pictureof the Matsushita Denso building covered with self-cleaning glass (courtesy of Nippon Sheet Glass). (c) Picture of the self-cleaning sound-proof wall (courtesy of Sekisui).(d) Eco-life-type houses using self-cleaning tiles and glass (courtesy of PanaHome). (e) Self-cleaning roof of a train station in Motosumiyoshi (courtesy of Taiyo Kogyo).

Fig. 6.3. Outdoor exposure test for a PVC tent material manufactured by Taiyo Kogyo done in the Photocatalyst Museum, KAST. The left half part of the tent material wascoated with TiO2 . (a) Picture taken in July 22, 2004. (b) Picture taken in April 23, 2007.

material, and thus the loss of transmittance by TiO2 can be largerthan that caused by contamination layers. Zhang et al. preparednanoporous TiO2–SiO2 coatings that have a low refractive index;

the coating enhanced the transmission of the glass to exceed 97%for visible light, greater than the 90–92% usual for glass (Fig. 6.4)[525,526].


Fig. 6.4. Transmission spectra of a bare glass and a glass coated with TiO2–SiO2composite film. Themaximum transmittance for the coated glass could be over 99%[525].© 2005, American Chemical Society.

TiO2-coated textiles are another promising application of self-cleaning technology. Recent studies by Kiwi and co-workers haveshown that coffee and wine stains on TiO2-treated cotton orsynthetic textiles can be decolorized under sunlight [527,528].

6.2. Water purification

As noted in our historical overview, the earliest reports ofthe cleaning properties of TiO2 were those of Frank and Bard in1977, in which they found that suspensions of TiO2 powder incontaminated water were able to photocatalyze the conversion ofcyanide to cyanate, thus detoxifying the water [56,57].One of the advantages of TiO2 photocatalysis for water

decontamination is that only the TiO2 photocatalyst (immobilizedor suspended) and UV light, either from solar light or artificiallight sources, are needed and thus its cost can be lower than otherkinds of advanced oxidation techniques (UV/O3,UV/H2O2, photo-Fenton). Moreover, no toxic intermediate products are generatedin the photocatalytic decontamination process; this makes it veryattractive for cleaning the water environment, even for cleaningdrinking water [379,380,382,529].However, it is generally accepted that TiO2 photocatalysis

is feasible only for the treatment of wastewater that containscontaminants at low tomedium pollutant concentrations, becauseof its relatively low efficiency and the limited flux of ultravioletphotons [382,530]. Malato and co-workers studied the solardecontamination of industrial [390], agricultural [531,532] (3,4),and municipal wastewater [533] at the pilot-plant scale, usingcompound parabolic collector (CPC) reactors and TiO2 slurryphotocatalysts. They observed the complete mineralization oforganic matter at concentrations of ca. 50 mg L−1 within hoursunder sunlight. However, the decontamination rate was generallylower for TiO2 photocatalysis than for the photo-Fenton processin their studies [534,535]. Herrmann and co-workers reporteda solar photo-reactor design based on the multi-step cascadefalling-film principle to ensure good exposure to sunlight and goodoxygenation of the effluent to be treated [536]. Theseworkers usedan immobilized TiO2 photocatalyst that was prepared by coatingnon-woven paper with anatase paste using aqueous SiO2 colloidas a binder material in order to avoid the filtration procedurerequired in slurry photo-reactor. This film photo-reactor wasfound to be as efficient as the CPC slurry photo-reactor for 4-chlorophenol, formetanate, and a mixture of pesticides in terms oftotal degradation under identical solar exposure.One interesting application of TiO2 photocatalysis is to remove

endocrine disruptor chemicals (EDC) in the aqueous environment

[537–540]. These chemicals, including natural hormones, dioxins,biphenol-A, etc., in the aquatic environment have been implicatedas health hazards for both humans and wildlife. They are assumedto disrupt normal endocrine functions through interaction withsteroid hormone receptors, even at very low concentrations.Conventional biological methods to remove these EDCs requirelong periods of time, and chemical oxidation methods are ingeneral not economical because of the low concentrations of theEDCs. Nakashima et al. designed a photocatalytic reactor usingTiO2-modified-PTFE mesh sheets as photocatalysts and utilizedthis to treat the water discharged from the Kitano sewage-treatment plant, which is located next to the Tama River nearTokyo [540]. Concentrations of estron in the dischargedwaterwere140 ng L−1. Under UV illumination (1.2 mW cm−2), about 90%of the initial estron was decomposed in a very short time, withgood reproducibility (Fig. 6.5). Thus, TiO2 photocatalysis can beapplied in the treatment of sewage effluent as a safe method forremoving natural and synthetic estrogens. This approach shouldalso be suitable for the removal of other low-level EDCs in theaquatic environment.Hashimoto and co-workers recently reported the application

of photocatalytic technology to the hydroponic cultivation oftomatoes [14]. In conventional hydroponics, tomato plants areplanted in an organic culture medium. The culture solution,containing nitric acid, phosphoric acid, potash, etc., flows throughthe culture medium in a circulated mode and provides nutrientsto the plants. However, organic substances are released intothe culture medium during circulation, and as a result, plantpathogens can be propagated in the system and can cause diseasesin the plants. The researchers connected a photocatalytic-water-treatment tank to the circulation system and used sunlight todecompose organic substances in solution. Themethod is effective,as evidenced by the apparent decrease in the total organic carbonin solution and the increased yields of tomatoes.Other important water-purification applications of TiO2 pho-

tocatalysis include water disinfection [406,408,541–545], re-mediation of metal contamination [546,547], and oxidation ofarsenite [548]. For developing countries, solar photocatalytic dis-infection appears to be a promising process to produce drinkingwater, which could help in improving public health. Rincón et al.reported that real lake water contaminated with E. coli K12 at alevel of 106 CFU mL−1 could be disinfected completely within 3 hon a clear day in summer, using a CPC reactor and a TiO2 slurrycatalyst [549]. In comparison, solar light alone is unable to effec-tively disinfect the contaminatedwater. Themechanism for photo-catalytic disinfection is still under debate (see Section 4.2.4) [408].Various active oxygen species, such as hydroxyl radicals, superox-ide, hydrogen peroxide, singlet oxygen, etc., may participate in thedisinfection process. In addition, the self-repair action of microbescomplicates the disinfection process [408].There have been several types of water-purification facilities on

the market for various purposes. One of these is designed to re-move volatile organic compounds (VOC), such as trichloroethylene,tetrachloroethylene, 1,3-dichlorobenzene, and dichloromethane,etc., from underground water or soil. These substances are widelyused as extracting solvents in industrial processes and in dry clean-ing and are released into the environment in large quantities. Thefacility basically involves suctioning up underground water froma deep well dug in the polluted area, spraying the water to evap-orate the VOCs into the gas phase, and decomposing the gaseousVOCs with a photocatalytic air-cleaning unit. VOCs at concentra-tion levels ranging from several hundreds to several thousands ofparts per million can be decomposed almost completely to carbondioxide and hydrochloric acidwith such a facility,with amaximumtreatment capability of 5 kg VOCs per day for a 15-kW facility.


Fig. 6.5. (a) Concentration change of estron (E1) in water discharged from a sewage-treatment plant during photocatalysis. The measurement was repeated three times(UV light intensity, 1.2 mW cm−2; water temperature, 15 ◦C) [540]. ©2003, Elsevier Science Ltd. (b) Picture of a photocatalytic reactor for treatment of sewage water. Thereactor consists of two 34-L columns, 100 pieces of TiO2-coated PTFE mesh sheet, and 12 black-light (40 W) lamps which gave a light intensity of about 1.2 mW cm−2 .

Fig. 6.6. (a) Photographic images of a honeycomb-type air-cleaning filter (left) and a three-dimensional porous ceramic air-cleaning filter (right); (b) Schematic diagram ofthe microscopic structure of the air-cleaning filter; the substrate materials (ceramics, paper, fabrics) are coated with a composite of TiO2 and adsorbent.

6.3. Air purification

One of the most important applications of TiO2 photocatal-ysis is to decontaminate, deodorize, and disinfect indoor air.Low-concentration volatile organic compounds (VOCs), such asformaldehyde, and toluene, emitted from interior furnishings andconstruction materials, may lead to the ‘‘sick building syndrome’’and other diseases. Besides, the indoor air of public facilities, whichis normally contaminated with bacteria, fungi, etc., threatens thehealth of users. Conventional air-purification systems often adaptfilter-type components for the cleaning of polluted air. Pollutantsare accumulated in the active carbon filters, and the filter becomessaturated with adsorbed substances and loses its function after acertain period of use. Treatment of the used air filters may causethe risk of secondary pollution. Photocatalytic air-cleaning filters,however, can decompose the organic substances instead of accu-mulating them and, as a result, exhibit better performance thanconventional ones [9,15,373,550–552]. Moreover, the photocat-alytic filter can kill the bacteria floating in indoor air, which is alsoimportant for indoor air purification [553,554].Fig. 6.6 shows images of two types of photocatalytic filters used

in air-purification devices. The filters feature either a honeycomb-

type construction or a three-dimensional porous structure forminimumpressure drop. TiO2 nanoparticles are coated on the bodyof the filter with active carbon, zeolites, etc., as co-adsorbents.The co-adsorbents facilitate the adsorption of VOCs on the filter,and ensure that no intermediate gaseous compounds diffuse intothe ambient atmosphere. The adsorbed substances diffuse on thesurface of the adsorbent until they reach the TiO2 and then undergophotodecomposition. The oxidizing species on the photocatalystcan also diffuse to the surface of the adsorbent and participatein oxidation reactions there. UV light, illuminated on the filtersurface from black-light lamps, is generally at the level of severalmW cm−2; this ensures the rapid mineralization of VOCs thattypically exist in concentrations of about 1 ppm or even lower.After long-term use, the filter may be poisoned by HNO3 or H2SO4formed during removal of ammonia or hydrogen sulfide. In sucha case, the poisoned filter can be regenerated by simply washingwith water.The available air-cleaning devices based on photocatalytic

filters include air cleaners, air conditioners, air-cleaning unitsfor refrigerators, etc. Among these application groups, the aircleaner is the largest. There are more than thirty companies


Fig. 6.7. Usage of TiO2-based photocatalytic material on roadway surfaces to convert nitrogen oxides (NOx) to nitrate: (left) application of the coating; and (right) finishedroadway, with the coated surface showing a lighter color (courtesy of Fujita Road Construction Co., Ltd.).

manufacturing photocatalyst-type air cleaners in Japan, with overonehundredproduct types available on themarket. The consumersof photocatalyst-type air cleaners include hospitals, institutions forthe elderly, pet stores, offices, smoking rooms, sports clubs, hotels,schools, restaurants, food manufacturers, or in any location whereindoor air purification is necessary. In addition to air cleaners,deodorizing sheets that are composed of TiO2 photocatalyst andactive carbon adsorbents are also available on the market. Thesheet is able to remove malodorous substances in kitchens,refrigerators, vehicles, and shoes through the adsorption functionof the activated carbon. Placing the used sheet in sunlight readilydecomposes the adsorbed pollutants by the action of the TiO2photocatalyst and regenerates the sheet.Nitrogen oxides (NOx) exhausted from automobiles have be-

come a serious source of air pollution in urban areas. PhotocatalyticTiO2 has the capability of removing nitrogen oxides [555,556]. Ithas been suggested that the nitrogen oxide (NO) in the air is oxi-dized when the catalyst is exposed to light. Then, through the in-termediate stage of nitrogen dioxide (NO2), it is finally convertedto nitric acid, which must then be stored. At the nitrogen dioxidestage, part of the gas may escape from the photocatalyst surface,but, with an adsorbent like activated carbon mixed with the cata-lyst, this gas may be effectively trapped [9].The range of concentrations of air pollutants that can be

efficiently removed is from 0.01 ppm to 10 ppm. These concentra-tions correspond to the range from those in the ordinary environ-mental atmosphere to those in highway tunnels. Some companiesin Japan are now considering covering roads with TiO2 photocata-lyst, and removing the NOx exhausted from automobiles with sun-light. TiO2-coated road bricks, which are used to cover pavement,have become available on the market. Interestingly, TiO2 coatingsfor roadways are being prepared bymixing colloidal TiO2 solutionswith cement. Nitric acid, formed during the oxidation of nitrogenoxides, reacts with the cement to form calcium nitrate, which canbe washed off by rainwater. This method, which has been termed‘‘photo-road’’ technology, has been tested in more than fourteenplaces in Japan over the past nearly ten years (see Fig. 6.7). Oneof the places tested is on the 7th belt highway in Tokyo, withan area of about 300 m2. Nitrogen oxide removed from this test-ing area was estimated to be 50–60 mg per day, correspondingto the amount of nitrogen oxide discharged by 1000 automobilespassing by.More recently, this idea has been developed to a great extent

in Europe. In particular, the Italian cement company Italcementihas created several demonstration projects, for example, in Romeand Paris, to check the actual photocatalytic effect of a TiO2-coatedhighway in the conversion ofNOx and SOx tomore environmentallybenign forms, i.e., NO−3 and SO

2−4 [557].

6.4. Self-sterilizing surfaces

TiO2 photocatalysts can kill bacteria on their surfaces, andtherefore self-sterilizing surfaces can be prepared [9,10,530]. Suchmaterials are particularly useful for hospitals and elder carefacilities, where the control of surface and airborne bacteria andviruses is important. In cooperation with TOTO Ltd., Fujishimaet al. developed a type of antibacterial tile made by coveringordinary tiles with a TiO2–Cu composite coating [9,530]. In a testexperiment, the bacteria on these tiles were killed completely inone hour under illumination equal to the brightness of the surfaceof a study desk (UV light: ∼4 µW cm−2). On the basis of theseresults, such tiles were tested on the floor and walls of a hospitaloperating room where sterile conditions are crucial [9,530]. Afterinstalling the tiles, the bacterial counts on the walls decreased tonegligible levels in a period of one hour. Surprisingly, the bacterialcounts in the surrounding air also decreased significantly. TOTOLtd. has marketed the antibacterial tiles in Japan. Its users includehospitals, hotels, and restaurants, among others.One advantage of photocatalytic sterilizing surfaces is that

they operate in a passive fashion, that is, without the need forelectrical power or chemical reagents, only light and oxygen beingrequired. Unlike chemical antibacterial agents, TiO2 surfaces arenon-poisonous and will not result in environmental pollution.It is thus expected in the near future that self-sterilizing TiO2materials may have many medical applications. At least, hospitalroom walls, medical instruments, and hospital uniforms can allbe coated with self-sterilizing TiO2 materials. In a recent study,Ohko et al. coated flexible silicone catheters with a TiO2 thinfilm [558]. The film, as shown in Fig. 6.8, strongly adhered tothe body of the catheter: it was able to sustain repeated bendingwithout cracking and peeling observed macroscopically; it wasable to pass an adhesive tape test, and it was resistant to scratchingwith a 3H pencil. The TiO2-coated silicone catheter exhibited goodself-sterilizing properties under 1 mW cm−2 UV illumination, asshown in Fig. 6.9, and was shown to be safe in an experimentusing cultured cells and animal experiments [559]. Clinical studiesshowed that the antibacterial capability of TiO2-coated catheterswas superior to that of conventional catheters, and suggestedthe promising clinical use of the self-sterilizing catheter as anintermittent (self-)catheterization catheter for neurogenic bladderpatients [559]. The self-sterilizing catheter can be disinfected ina specially designed portable sterilizer box, or by direct sunlight.The intra-corporeal use of TiO2-coated catheters is also underexamination. In this case, metallic Ag is deposited on the TiO2surface to enhance the antibacterial action under dark conditions.


Fig. 6.8. Scanning electronmicrographs of (a) the surface and (b) the cross-section of the TiO2-coated silicone catheter. There is a thin intermediate SiO2 layer between TiO2and catheter substrate [558].© 2000, John Wiley & Sons, Inc.

Fig. 6.9. (a) Survival rate of E. coli inside several kinds of catheters underultraviolet-A irradiation. The catheters were filled with a 0.1 mL aliquot of E. colisuspension, as follows: N, control 1, original silicone catheters, no UVA irradiation;�, control 2, TiO2-coated catheters, noUVA irradiation;�, original silicone catheters,with UVA irradiation (1mW cm−2);×, TiO2-coated catheters, with UVA irradiation(1 mW cm−2) [559].© 2007, The Japanese Urological Association.

6.5. Anti-fogging surfaces

The fogging of the surfaces of mirrors and glass occurs whenmoist air cools down on these surfaces, forming many waterdroplets. These droplets, ranging from micrometer to millimetersize, tend either to scatter light or simply to reflect or refract itrandomly. In both cases, visual clarity is impaired drastically. Ina serendipitous discovery, Watanabe and co-workers found thata TiO2–SiO2 surface could become extremely hydrophilic underUV illumination [500,560]. The result of this property is thatwater spreads evenly across the surface. If the amount of water isrelatively small, the water layer becomes very thin and evaporatesquickly. If the amount of water is larger, it forms a sheet-like layerthat also has high visual clarity (Fig. 6.10). The first commercial

application of this phenomenon has been for automobile side-viewmirrors. The TOTO Ltd., in Japan currently is supplying TiO2-treatedglass for such mirrors to major automobile manufacturers [9,14].In addition, it has marketed adhesive TiO2-coated plastic films orTiO2-containing sprays for the conversion of existing mirrors.The uses of anti-fogging technology, however, are not limited to

mirrors and glass. One example in our daily life is air conditioners.When an air conditioner is operating to cool during the summer,air that is cooled on the heat exchanger of the air conditioneroften has moisture condensing out of it, except in an extremelydry climate. For an ordinary plate heat exchanger, the condensedwater droplets fog the fins and large water droplets often form,often bridging between the adjacent fins; this can reduce the heatexchange efficiency and increase the resistance to air flow. Coatingthe heat exchanger surface with superhydrophilic TiO2, however,can hinder the fogging and the formation of water bridging, andremove the condensed water more effectively. This has beenproven by the recent studies of Takata et al. [561]. They observedthat the volume of water flowing off superhydrophilic TiO2-coatedfins was 20% greater than that from ordinary fins, when humidair (RH 90%, 40 ◦C) was passed through the cooled fins (surfacetemperature: 15 ◦C).As the properties of superhydrophilic films continue to be

improved, there will be many other possible applications ofthe anti-fogging phenomenon. Research activities have beenfocusing on: (1) increasing the rate of superhydrophilic conversion[427,515]; (2) prolonging the superhydrophilic states of thefilms [515,526,562]; and (3) activating the films at longerwavelengths [473,522]. A rough TiO2 surface morphology hasthe advantages of fast superhydrophilic conversion and longpersistence of the superhydrophilic state, but it can do harm tofilm transparency. A mixed film of TiO2 and SiO2 (or Al2O3) has

Fig. 6.10. Photographs of fogged surfaces of usual glass (a) and TiO2-coated glass after sufficient UV illumination (b), typically > 150 min at an intensity of 1 mW cm−2[500].©1999, Elsevier Science Ltd.


the advantages of long persistence of the superhydrophilic state,but an excessive SiO2–TiO2 ratio can have a negative effect onthe rate of superhydrophilic conversion. To develop a desirableanti-fogging film, one obviously should strike a balance betweenfilm composition and surface roughness. Extending the responsewavelength of TiO2 materials into the visible light region isconsidered important for the design and actual applications ofsuperhydrophilic anti-fogging materials, but until now, only veryfew reports have appeared on the superhydrophilic conversionof visible-light-responsive TiO2 films (N-, or C-doped TiO2) undervisible illumination (see Section 5.2.5).

6.6. Heat transfer and heat dissipation

Superhydrophilic surfaces may improve the heat transferaccompanied by liquid–vapor-phase transition. Takata et al.studied the heat transfer properties of superhydrophilic TiO2surfaces. They found that a TiO2-coated Cu cylinder exhibitedexcellent heat transfer characteristics in the nucleated boilingregime, and its critical heat flux (CHF), a condition where liquidcannot rewet the heater wall because of the rate of vaporproduction impeding the liquid flow back to the hot surface,was about two times higher than that of the uncoated surface[561,563]. They also compared the falling-film evaporation ona superhydrophilic surface with that on an ordinary surface[564]. In this work, water was sprayed on the heated surfacethrough nozzles to form a water film. They found that on thesuperhydrophilic surface, a stable water film was formed formuch smaller water flow rates than those on an ordinary surface.Since the heat transfer rate increases for thinner water films, thesuperhydrophilic surface is expected to improve the performanceof falling-film evaporators, which are widely used in seawaterdesalination units andmilk and juice concentrators. They observedthat the superhydrophilic surface performed 40 times betterthan the ordinary surface in low water flow and low heat flux.For high water flow, a thick water film formed on both thesuperhydrophilic and the ordinary surface, and thus no differencein heat transfer rate was observed. These workers suggested thatthe superhydrophilic surface could be an ideal heat transfer surfaceand would be applicable to various heat transfer phenomena thatare affected by surface wettability [561,563,564].Hashimoto et al. reported recently the cooling of the exterior

walls of buildings with a falling-water film, with the intent tohelp resolve the urban heat island problem for major cities and toreduce the electricity consumption by air conditioners in summer[14,565]. In their experiments, water was continuously sprinkledonto the surfaces of buildings that were covered with TiO2 films.Due to the superhydrophilic property of the material, water wasable to spread on the surfaces in the form of a thin water film(∼0.1 mm). The evaporation of the water film can effectively coolthe building surface and also the surrounding air. They observedon a clear day in the middle of summer that the temperaturedrops were 15 ◦C on window glass and 40–50 ◦C on black-roof-tilesurfaces. The cooling of the building surfaces is expected to resultin the reduction of electricity consumption by air conditioners.As estimated by Hashimoto et al., this reduction could be tento several tens of percentage for a real house. Therefore, TiO2-coated building materials could contribute much to energy-savingtechnologies in the future (Fig. 6.11). The water film is so thin thatonly a small amount of water is required to cover the buildingsurfaces. Moreover, natural rainwater can be collected and storedin specially designed reservoirs for this purpose so that the cost ofthe water can be reduced. The water film also helps in maintainingthe building surface clean by the self-cleaning effect.

Fig. 6.11. Energy-saving system using solar light and stored rainwater [14].

6.7. Anticorrosion applications

In general, metals are protected from corrosion in severalways, for example, by a corresponding passive metal compoundlayer, by a paint layer, or by a sacrificial metal coating (with aless positive corrosion potential). Of these, the passive layer andthe paint layer are effective only when the coating covers themetal surface perfectly. As for the sacrificial metal coating, itslifetime is limited, because the metal coating dissolves gradually.Recently, the application of photocatalytic technology in corrosionprevention for metals has been examined [566–574].In cooperation with the Koyo Electrical Construction Co.,

Ohko et al. studied the anticorrosion effects of a TiO2 coatingfor Type 304 stainless steel [566]. Under UV irradiation, TiO2injected electrons into the steel and as a result protected it fromcorrosion, while the photogenerated holes decomposed organiccontaminants to provide a self-cleaning function. Interestingly,the coating remained effective even if perforated with pinholes.Later, the same group observed that TiO2, when coupled withWO3, maintained an anticorrosion effect even in the dark fora period of time after UV irradiation ceased [567,570,571]. Thisphenomenon can be explained by the energy storage ability ofWO3, as illustrated in Fig. 6.12. Under UV irradiation, electrons inthe valence band of TiO2 are excited to the conduction band. Theexcited electrons are injected into the metal so as to maintain itspotential more negative than the corrosion potential. The excesselectrons can be accumulated in the ‘‘electron pool’’ of WO3 dueto the lower conduction-band edge of WO3 compared to TiO2. Thereduced WO3 then reacts with protons adsorbed on the surfaceor sodium cations in the electrolyte solution to form metastabletungsten bronze (MxWO3, M:H+,Na+; x ≤ 1) so that the reductiveenergy generated at the irradiated semiconductor can be storedas in a battery. After the UV light is turned off, the metastabletungsten bronze gradually releases electrons into the metal via aself-discharge process, and as a result the metal remains protectedfromcorrosion in the dark until the discharge process is completed.Similarly, MoO3 also exhibited an electron pool function whencomposited with TiO2 [572].Research activities are continuing in order to extend the

anticorrosion period of TiO2-based coatings in the dark [573,574]. Recent studies have shown that phosphotungstic acid (PWA)exhibits a longer self-discharge time than WO3 [574]. The TiO2-PWA coating can maintain an anticorrosion effect over 12 h inelectrolyte solution, after being illuminated with 1 mW cm−2 UVlight for 1 h. This is a very promising result for the development ofpractical semiconductor coatings for metal corrosion protection.


Fig. 6.12. Schematic diagram of an energy storage-type photocatalytic anticorro-sion system, which can protect metal from corrosion even in the dark [567].© 2001, American Chemical Society.

6.8. Environmentally friendly surface treatment

In conventional technology for manufacturing printed circuitboards (PCBs), insulating resin materials are first etched withstrong oxidants such as permanganate and hexavalent chromiumto prepare rough surfaces followed by electroless plating of acopper layer on the whole surface and removal of the unwantedcopper after applying a temporary mask. The surface treatmentprocedure is of vital importance for the strong adherence of thecopper layer to the resin substrate. However, the use of thoseoxidants can cause serious environmental problems. Moreover,the roughness of the etched surfaces is often of the orderof micrometers, which can lead to technical problems in thefabrication of fine copper wires. Honma and co-workers recentlyreported a novel environmentally friendly surface treatmenttechnique for PCB manufacture by treatment of resin surfaceswith TiO2 photocatalyst, replacing conventional etchants [575].They found that an aqueous TiO2 suspension was able to oxidizean epoxy resin surface within a depth of 30–50 nm under UVillumination but did not affect surface roughness appreciably.Palladium and tin catalysts for the electroless deposition of copperpenetrated into the surface-modified layer and electroless copperplating was initiated at the bottom of this layer. As a result, thecopper layer showed excellent adhesion (1.15 kN m−1) to theepoxy surface, even better than that obtained by conventionaltechnology. By using this new surface treatment technique, Honmaet al. succeeded in fabricating fine patterns of copperwith lines andspaces of 10 µm/10 µm, as shown in Fig. 6.13.Honma and co-workers also applied this new surface treatment

technique to modify the ABS resin surface, which is often etchedwith hexavalent chromium in conventional technology [576]. The1, 2-polybutadiene distributed in theABSmatrixwas preferentiallyreformed by photocatalytic oxidation and induced hydrophilicityon the reformed surface. The ABS resin surfacewas reformed downto the depth of a few µm and the catalyst (Pd and Sn) penetratedinto the surface to the extent of 30–40 nm. Metal layers depositedon the modified ABS surface showed adhesive strengths of about1.0 kN m−1. They concluded that the superior adhesion strengthwas derived from a nano-anchor effect by the deposition of metalinside the 30–40 nm reformed resin layer.

6.9. Photocatalytic lithography

The active oxygen species such as the hydroxyl radical or singletO2 generated during the photocatalytic process are so active thatthey can oxidize not only organic materials but also inorganicmaterials such as copper, silicon, carbon, and SiC etc. [353,359,360]. The oxidation can occur either on the TiO2 surface or at adistance away from the surface. Therefore, it is possible to developa novel photocatalytic lithographic method based on the above-mentioned oxidation reactions to fabricate surface patterns with

Fig. 6.13. SEMmapping of 10µm/10µmL/S copper wiring. (a)–(c) PCB substrateswere treated by TiO2 photocatalyst. (d)–(f) PCB substrates were etched withpermanganate etching process. (c) and (f) are reflectionmode images. As shown, thephotocatalytic treatment did not change the surface morphology of PCB much, andcould be used to prepare finer copperwires than those obtained by the conventionaletching method [575].© 2006, Surface Finishing Society of Japan.

composition/wettability contrast. This novel method would havethe advantages of simplicity and low cost: it does not requireexpensive photo-resist materials and special equipment.Tatsuma and co-workers first examined the concept of

photocatalytic lithography based on the remote oxidation effect ofTiO2 [360]. They placed a TiO2-coated quartz substrate in contactwith a glass plate modified with an ultrathin organic layer, orsilicon, copper, or silver plate, separated by a small gap. The TiO2was irradiated with UV light in air through a photomask. As aresult, two-dimensional images corresponding to the photomaskwere obtained; those images were based on the contrast of non-oxidized to oxidized surfaces. In a following study, Kubo et al.examined the resolution of their novel lithography technique andfound that the resolution of remote oxidation-based photocatalyticlithography deteriorated slightly due to the lateral diffusion ofactive oxygen species in the air [353] They obtained a resolution of10µmby optimizing the TiO2 film/photomask assembly (Fig. 6.14)and suggested that the theoretical resolution should approach thescale of the light wavelength, namely the submicrometer level,provided that a highly collimated light beam and a highly uniform,non-light scattering TiO2 film were available [353].Lee and Sung combined photocatalytic lithography and selec-

tive atom layer deposition (ALD) to prepare patterned ZrO2 thinfilms on Si substrates [577]. In their study, a patterned TiO2 thinfilm (line width, 400 nm; line space, 580 nm) on quartz was usedas the lithographic tool. The TiO2 film was placed on an octadecyl-siloxane self-assembled monolayer (SAM)-modified Si substrate,and the assembly was irradiated with 254-nm UV light. The SAMwas selectively removed, retaining the dimensions of the TiO2 pat-tern without noticeable line spreading. By means of the ALD tech-nique, a ZrO2 thin filmwas selectively deposited in the regionwith-out the SAM, with a 430-nm linewidth, through reaction with thesilanol groups of the Si substrate.


Fig. 6.14. (a) Micrograph of the Cu plate surfaces patterned by the photocatalytic lithography. (b) Micrographs of masked TiO2 film (upper) and heptyldecafluorode-cyltrimethoxysilane (HDFS)-coated glass plates patterned by photocatalytic lithography (below) [360].© 2004, American Chemical Society.

Fig. 6.15. Micrographs of a drop of polystyrene microsphere suspension pipetted on a preheated superhydrophobic–superhydrophilic pattern, which were taken at (a) 5 s,(b) 5 min, (c) 8 min, and (d) 9 min after pipetting the drop. Polystyrene microspheres were selectively deposited in the superhydrophilic stripes. The scale bar in pictures is400 µm [585].© 2007, American Chemical Society.

Photocatalytic lithography has also been applied to preparechemical affinity/wettability patterns on surfaces [578–580].Spencer and co-workers irradiated an alkanethiolmonolayer on anAu substrate through a TiO2 film and a filter with a transmittancegradient [580]. Thus, these workers were able to prepare aconcentration gradient of alkanethiol on the Au surface, sincethe oxidized alkanethiol molecule was removed from the surface.Adsorption of other types of alkanethiol molecules led to theformation of chemical affinity/wettability gradients on the Ausurface.In addition to remote oxidation-based photocatalytic lithogra-

phy, there have also been some efforts to prepare chemical affin-ity/wettability patterns directly on TiO2 surfaces. The patternedTiO2 surfaces were used to assemble colloidal crystal films [581]

and to site-selectively immobilize semiconductor quantum dots[582]. There have been several studies concerning the pattern-ing of superhydrophobic TiO2 films by photocatalytic lithogra-phy [583,584]. The resulting superhydrophobic–superhydrophilicsurface patterns were shown to guide the evaporation of a la-tex suspension (Fig. 6.15) or to be filled site-selectively withelectroless-platedmetal or polymer hydrogel, due to the extremelylarge wettability contrast in the pattern [585,586] In these cases,the diffusion of holes in TiO2 and the surface diffusion of active oxy-gen species should restrict the lithographic resolution to the mi-crometer scale, as suggested by the studies of Haick and Paz [357,358] A recent study by Kobayashi et al., however, showed that afeature with 30-nm linewidth was able to be produced in a stearicacid monolayer deposited on a TiO2 substrate by photocatalytic


Fig. 6.16. (a) Photochromism of the Ag–TiO2 film. Absorption spectra of Ag deposited on the TiO2 film by ultraviolet light irradiation (300–400 nm, 1mW cm−2 , 15min) andafter visible light irradiation (>400 nm, 50mW cm−2 , 1 h). Corresponding photographs are also shown. (b) Multicolor photochromism of the Ag–TiO2 film. A–D, Differentialabsorption spectra of the Ag–TiO2 film after visible light irradiation (10 mW cm−2 , 5 min) using a xenon lamp with bandpass filters [588].© 2003, Nature Publishing Group.

lithography, aided by a local electrical field applied by a conduc-tive AFM cantilever [587]. The positive bias (10 V) applied to TiO2substrate lowered the local charge recombination, which led to therapid decomposition of the stearic acid molecules at the site towhich bias was applied.

6.10. Photochromism

Photochromic materials, which change their colors reversiblyin response to light, can be applied to smart windows, displaysand memories. Conventional photochromic materials respondin a monochromatic way, so that multicolor photochromismhas required several different materials or filters combinedappropriately. If multicolor photochromism could be achievedwith a simple material, photochromic devices would find a greaternumber of applications, including a rewritable color copy paperor electronic paper and a high-density multi-wavelength opticalmemory. Ohko et al. recently reported the interesting multicolorphotochromism of TiO2 films loaded with silver nanoparticles[588–591]. The film color, initially brownish-grey, changed undermonochromatic visible light to almost the same color as that ofthe light (Figs. 6.16 and 6.17); the apparently uniform Ag–TiO2film can take on almost any color. The color reverted to brownish-grey under ultraviolet light, and these processes were found to berepeatable.Ohko et al. proposed the following model for the multicoloring

process [588]: on irradiation with monochromatic visible light,the corresponding Ag nanoparticles adsorb the light, and theelectrons thus excited are accepted by O2, resulting in oxidationof the Ag nanoparticles to Ag+ and an absorption decreaseappears at the corresponding wavelength, i.e., the color of theexcitation wavelength. The most important role of the TiO2 filmis the repeatable generation of Ag nanoparticles with diversesizes and shapes due to the ‘‘molding effect’’ of the nanoporousTiO2 film; this resulted in different plasmon resonance absorptionwavelengths of the Ag particles in thewhole visible light spectrum.Later studies by Kawahara et al. showed that not only TiO2 but

Fig. 6.17. A thick multicolored Ag–TiO2 film. Each spot (6 mm diameter) wasirradiated successively with a blue, green, red, or white light.© 2004, American Chemical Society [591].

also ZnO films were able to work as the substrate material for Ag-basedmulticolor photochromism [589]. SiO2 and ITO, however, didnot work. These workers suggested that both TiO2 or ZnO playan important role in the charge separation between the excitedelectrons and the Ag+. Actually, a fraction of the excited electronson Ag were transferred to oxygen molecules via TiO2 or ZnO andnon-excited Ag.The color image is bleached gradually in air by ambient white

light, which is detrimental to practical applications. Naoi et al.showed that a stable color image can be obtained by eithermodifying the Ag nanoparticles with thiols [592], or washing thephotogenerated Ag+ ions off the film [591]. Interestingly, the colorimage fixed by washing off the Ag+ ions could be maintained as a‘‘latent image’’, that is, the image can be recovered after a cycle ofwhite light bleaching andUV coloration due to the ‘‘molding effect’’of the nanoporous TiO2 film [591].

6.11. Microchemical systems

Microchemical systems using microchannels as reaction fieldsprovide many technical advantages over conventional technolo-gies for chemical synthesis as well as analysis. Coating the in-ner walls of microchannels with TiO2 provides photocatalyticreactivity to microchemical systems. Possible applications in-clude the photocatalytic degradation of organic pollutants [590,


593], photocatalytic synthesis of organic compounds [594], site-selective introduction of functional metal nanoparticles [595],photo-controllable wettability of channels [596,597], and self-cleaning microchannels, among others.Takei et al. employed a TiO2-modified microchannel chip for

the photocatalytic synthesis of L-pipecolinic acid from L-lysine[594]. They observed that the conversion rate in the chip was 70times larger than that in a cuvette with almost the same selectivityand enantiomeric excess [594]. They suggested that the largesurface area of themicrochannels, togetherwith the short diffusiondistance, should improve the efficiency of photocatalytic reactionswithin the channels.Castellana et al. reported a simple method to introduce

noble metal nanoparticles inside sealed microfluidic channelsthat were coated with a TiO2 photocatalyst [595]. Under UVillumination, metal ions, for example, Ag+, were reduced fromsolution to the area covered with TiO2 to form an Ag nanoparticlefilm. In cases in which photomasks were used, a patterned Agfilm was readily prepared. The Ag nanoparticle film could bebiofunctionalized to capture proteins with good discriminationbetween addressed locations and background. Their methodappears to be quite general, and films of Au, Pd and Cu couldalso be patterned. The ability to address individual ligands atopnanoparticle films inside microfluidic devices can be combinedwith such technologies as transmission surface plasmon resonancespectroscopy or surface-enhanced fluorescence. This allows forthe development of powerful lab-on-a-chip devices with label-free detection or fluorescence detectionwith enhanced sensitivity.Moreover, the metal nanoparticle films can be used to immobilizepatches of oriented enzymes inside microfluidic devices, whichhave promising applications in enzymatic microreactors.Kim and co-workers applied photocatalytic reactions to control

the wettability of microchannels, by which pressure barriers, so-called Laplace valves, were introduced tomicrochannels at desiredpositions [596]. They modified the inner walls of the microchan-nels with TiO2 nanoparticles; afterwards they modified the TiO2nanoparticles with a hydrophobic self-assembled monolayer. Thewalls of the microchannel became superhydrophobic (water con-tact angle > 150◦) after these procedures. When UV illuminationwas applied to the superhydrophobicmicrochannel, TiO2 nanopar-ticles decomposed the hydrophobic monolayer, which led to thegradual decrease of the water contact angle to 0◦, the so-calledsuperhydrophilic state. Since the water contact angle of the TiO2-modified microchannel wall was adjustable between 150◦ and 0◦,the pressure barrierwas found to be tunable over awide range; thelatter was determined by the Laplace pressure P = −2γ cos θ/r ,where γ is the surface tension of the liquid, r the channel radius,and θ the contact angle. They demonstrated the preparation of4-step wettability-based Laplace valves having different pressurebarriers in one channel (Fig. 6.18), and also the construction of abatch operation system consisting of two sub-nL dispensers and areaction chamber.Thus, given the number of promising approaches already

reported, it appears likely that photocatalysis will find increasingapplication in microchemical and microfluidic devices.

7. Summary

We have tried to provide an overview of the field ofphotocatalysis from its very beginning in 1921 [17] through itsdevelopments in fundamental studies, both experimental andtheoretical, which have been strongly tied to applications. Ofcourse, it is impossible to do justice to this vast field in a reviewof even this length, and new work is emerging every day. Thisis the nature of the field. We have also tried to focus on someof the aspects of the field that would not be treated in a general

Fig. 6.18. (a) Irradiation time for preparation of the 4-step Laplace valve in onechannel and (b) fluorescence microscope images of fluorescein aqueous solution inthe photopatterned and tuned channel [596].© 2007, The Royal Society of Chemistry.

review or in a pure surface science-oriented review, for example,electrochemistry and photoelectrochemistry.The early work in the field can be inspirational and still teach

us much. For example, the work of Baur in the 1920s shows us thepower of imagination in the development of science. His idea of thepairing of oxidation and reduction reactions is still a central one[23,24,598]. The work of Goodeve and Kitchener in 1938 could bepublished todaywith only slightmodification [26], but the fact thatit was done at that time should inspire us to even greater efforts.Fundamental studies in spectroscopy have led to a much better

understanding of the processes occurring in TiO2 in its variousforms, including single crystals, powders and films. There is somuch information available now that it is a challenge to makesense of it all. One of the points that we have learned is that thenature of the sample itself, i.e., its preparation, handling, history,and chemical and physical characteristics must be specified inas great detail as possible, because all of these are important indetermining the observed behavior.The dual use of experimental studies with single crystal stud-

ies and theoretical studies on the same surfaces is becomingan extremely powerful approach. It is hoped that this approachwill be extended and deepened, with increasingly sophisticatedtechniques, until a convergence can be reached even for photocat-alytically related reactions on the rutile (110) surface, which weventure to say is still not completely understood. As a convergenceis achieved for this surface,we can becomemore confident that ourexperimental and theoretical techniques can be applied to othersurfaces. Of course, this is not to say that we should wait to studythe other surfaces, but the point is that we cannot be truly confi-dent until we can obtain convergence for thismost studied system.The work of Henderson has provided an excellent example [113,114,130,141–143,280,516,599–608].The recent STMwork on reactions of water and oxygenwith the

rutile (110) surface is truly impressive and provides an exampleof what can be learned at the atomic scale [129,133,137]. Therealization that tracewater, even in ultra-high vacuum, can react soreadily with oxygen vacancies to produce OH groups should causeus to examine carefully anywork inwhich vacancies are discussed,


at least on the TiO2 surface. Similar approaches can no doubt leadto further fundamental insight into the basic phenomena that areinvolved with photocatalysis.As important as the surface is,wemust not forget about the bulk

material, even in the form of a nanoparticle. This is even a morechallenging aspect, since we do not have a real-space techniquelike STM that can probe beneath the surface at the atomic scale.It will be necessary to develop ingenious new techniques to try todo this. The roles of sub-surface oxygen vacancies, before and afterreacting with water, as well as other bulk structures will need tobe elucidated fully in order to understand in detail the trapping ofelectrons and holes.After examining carefully the controversial aspects of the

photo-induced hydrophilic effect, we have reached the conclusionthat there is an aspect of this effect that does not involve simplythe cleaning of the surface, as shown by recentwork [513,518]. Theprecise nature of the effect has not been elucidated even now, butwe propose that the surface species are basically the same onesinvolved with conventional photocatalysis.We have indicated some of the fascinating new applications of

photocatalysis. These are not to be considered so very separatefrom the fundamental work. Both tend to reinforce each other.We believe that people working in both fundamental and appliedaspects should try very hard to understand what the others aredoing. This will undoubtedly lead to advances in both areas.

Appendix. TiO2 film preparation methods

For reference, a list of photocatalytic film preparation methodshas been assembled in Table A.1.

Table A.1Preparation of TiO2 active surface layers.

Method Author Year Ref.

Sol-gel Negishi et al. 1995 [76]Paz and Heller 1997 [395]Tada et al. 1997 [610]

Spray-pyrolysis Wang et al. 1998 [611]Ohko et al. 2001 [566]

Electrophoretic deposition Fernandez et al. 1995 [612]Yanagida et al. 2005 [613]

Chemical bath deposition Gao et al. 2003 [614]CVD Goossens 1998 [615]

Ding et al. 2001 [616]O’Neill 2003 [617]

RF Magnetron sputtering Zeman et al. 2002 [618]Song et al. 2003 [619]Kitano et al. 2005 [620]Song et al. 2006 [621]Takeuchi et al. 2006 [622]Kitano et al. 2006 [453]

Electron-beam evaporation Yang et al. 2004 [623]Yang et al. 2004 [624]

Vacuum evaporation Miyata et al. 2006 [625]Plasma spray Berger-Keller et al. 2003 [626]

Lee et al. 2003 [627]Anodic oxidation Mor et al. 2006 [628]

Macak et al. 2007 [629]Thermal oxidation Fujishima et al. 1975 [54]Inorganic–organic graded film Takami et al. 2002 [630]Silica gel substrate Kobayakawa et al. 1998 [631]

References

[1] E. Pelizzetti, N. Serpone (Eds.), Homogeneous andHeterogeneous Photocatal-ysis, D. Reidel Publishing Company, Dordrecht, 1986.

[2] N. Serpone, E. Pelizzetti (Eds.), Photocatalysis — Fundamentals and Applica-tions, John Wiley & Sons, New York, 1989.

[3] P.V. Kamat, Chem. Rev. 93 (1993) 267.[4] D.F. Ollis, H. Al-Ekabi (Eds.), Photocatalytic Purification and Treatment ofWater and Air, Elsevier, Amsterdam, 1993.

[5] A. Heller, Acc. Chem. Res. 28 (1995) 503–508.[6] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995)69–96.

[7] A. Mills, S. LeHunte, J. Photochem. Photobiol. A: Chem. 108 (1997) 1–35.[8] J. Peral, X. Domènech, D.F. Ollis, J. Chem. Tech. Biotech. 70 (1997) 117–140.[9] A. Fujishima, K. Hashimoto, T. Watanabe, TiO2 Photocatalysis: Fundamentalsand Applications, BKC, Inc., Tokyo, 1999.

[10] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C 1 (2000) 1–21.[11] D.F. Ollis, C. R. Acad. Sci. Paris, Serie IIC, Chim. 3 (2000) 405–411.[12] D.A. Tryk, A. Fujishima, K. Honda, Electrochim. Acta 45 (2000) 2363–2376.[13] A. Fujishima, D.A. Tryk, in: A.J. Bard, M. Stratmann, S. Licht (Eds.),

Encyclopedia of Electrochemistry, Vol. 6: Semiconductor Electrodes andPhotoelectrochemistry, Weinheim, 2002.

[14] K. Hashimoto, H. Irie, A. Fujishima, Japan. J. Appl. Phys. 44 (2005) 8269–8285.[15] A. Fujishima, X. Zhang, C. R. Chimie 9 (2006).[16] A. Fujishima, X. Zhang, D.A. Tryk, Internat. J. Hydrogen Energy 32 (2007)

2664–2672.[17] C. Renz, Helv. Chim. Acta 4 (1921) 961–968.[18] A.E. Jacobsen, Ind. Eng. Chem. 41 (1949) 523–526.[19] A. Fujishima, K. Honda, Nature 238 (1972) 37–38.[20] T.Watanabe, A. Fujishima, K. Honda, in: T. Ohta (Ed.), Solar-Hydrogen Energy

Systems, Oxford, 1979, pp. 137–169.[21] A. Fujishima, Innovative Hydrogen Production fromWater (UNESCO), 2000.[22] M.A. Fox, M.T. Dulay, Chem. Rev. 93 (1993) 341–357.[23] E. Baur, A. Perret, Helv. Chim. Acta 7 (1924) 910–915.[24] E. Baur, C. Neuweiler, Helv. Chim. Acta 10 (1927) 901–907.[25] C. Renz, Helv. Chim. Acta 15 (1932) 1077–1084.[26] C.F. Goodeve, J.A. Kitchener, Trans. Faraday Soc. 34 (1938) 570–579.[27] C.F. Goodeve, J.A. Kitchener, Trans. Faraday Soc. 34 (1938) 902–908.[28] M.C. Markham, K.J. Laidler, J. Phys. Chem. 57 (1953) 363–369.[29] T.R. Rubin, J.G. Calvert, G.T. Rankin,W.MacNevin, J. Am. Chem. Soc. 75 (1953)

2850–2853.[30] M.C. Markham, M.C. Hannan, S.W. Evans, J. Am. Chem. Soc. 76 (1954)

820–823.[31] M.C. Markham,M.C. Hannan, R.M. Paternostro, C.B. Rose, J. Am. Chem. Soc. 80

(1958) 5394–5397.[32] R.E. Stephens, B. Ke, D. Trivich, J. Phys. Chem. 59 (1955) 966–969.[33] J.G. Calvert, K. Theurer, G.T. Rankin, W.M. MacNevin, J. Am. Chem. Soc. 76

(1954) 2575–2578.[34] C.B. Vail, J.P. Holmquist, L. White, J. Am. Chem. Soc. 76 (1954) 624–625.[35] M.G. Markham, J. Barry, M. Iava, J. Haddad, J. Phys. Chem. 61 (1957)

1665–1668.[36] J.C. Kuriacose, M.C. Markham, J. Phys. Chem. 65 (1961) 2232–2236.[37] M.C. Markham, J.C. Kuriacose, J. DeMarco, C. Giaquinto, J. Phys. Chem. 66

(1962) 932–936.[38] A. Nergararian, M.C. Markham, J. Phys. Chem. 67 (1963) 2679–2682.[39] M.C. Markham, M.C. Upreti, J. Catal. 4 (1965) 229–238.[40] D.R. Kennedy, M. Ritchie, J. Mackenzie, Trans. Faraday Soc. 54 (1958).[41] A. Terenin, Y. Solonitzin, Discuss. Faraday Soc. 28 (1959) 28–35.[42] V.N. Filimonov, Doklady Akademii Nauk SSSR 154 (1964) 922–925.[43] S. Kato, F. Mashio, Kogyo Kagaku Zasshi (J. Chem. Soc. Japan., Indust. Chem.

Sect.) 67 (1964) 1136–1140.[44] I.S. McLintock, M. Ritchie, Trans. Faraday Soc. 61 (1965) 1007–1016.[45] C. Kormann, D.W. Bahnemann, M.R. Hoffmann, Environ. Sci. Technol. 22

(1988) 798–806.[46] F. Lohmann, Ber. Bunsenge. Phys. Chem. 70 (1966) 87–92.[47] C. Wagner, W. Traud, Z. Elektrochem. 44 (1938) 391–454.[48] C. Wagner, W. Traud, F. Mansfeld, Corrosion 62 (2006) 843–855.[49] S.R. Morrison, T. Freund, J. Chem. Phys. 47 (1967) 1543–1551.[50] A. Fujishima, K. Honda, S. Kikuchi, Kogyo Kagaku Zasshi (J. Chem. Soc. Japan.,

Indust. Chem. Sect.) 72 (1969) 108–113.[51] A. Fujishima, K. Honda, J. Chem. Soc. Japan. 74 (1971) 355 (in Japanese).[52] T. Inoue, T. Watanabe, A. Fujishima, K. Honda, Chem. Lett. (1977) 1073–1076.[53] A. Fujishima, T. Inoue, K. Honda, J. Am. Chem. Soc. 101 (1979) 5582–5588.[54] A. Fujishima, K. Kohayakawa, K. Honda, J. Electrochem. Soc. 122 (1975)

1487–1489.[55] A. Fujishima, K. Kohayakawa, K. Honda, Bull. Chem. Soc. Jpn. 48 (1975)

1041–1042.[56] S.N. Frank, A.J. Bard, J. Am. Chem. Soc. 99 (1977) 303–304.[57] S.N. Frank, A.J. Bard, J. Phys. Chem. 81 (1977) 1484–1488.[58] S.N. Frank, A.J. Bard, J. Am. Chem. Soc. 99 (1977) 4667–4675.[59] B. Kraeutler, A.J. Bard, J. Am. Chem. Soc. 100 (1978) 2239–2240.[60] H. Reiche, W.W. Dunn, A.J. Bard, J. Phys. Chem. 83 (1979) 2248–2251.[61] E. Pelizzetti, M. Borgarello, E. Borgarello, N. Serpone, Chemosphere 17 (1988)

499–510.[62] R.I. Bickley, T. Gonzalez-Carreno, J.S. Lees, L. Palmisano, R.J.D. Tilley, J. Solid

State Chem. 92 (1991) 178–190.[63] R.W. Matthews, in: D.F. Ollis, H. Al-Ekabi (Eds.), Photocatalytic Purfication

and Treatment of Water and Air, Amsterdam, 1993, pp. 121–138.[64] A. Mills, R.H. Davies, D. Worsley, Chem. Soc. Rev. 22 (1993) 417.[65] L.A. Dibble, G.B. Raupp, Environ. Sci. Technol. 26 (1992) 492–495.[66] J. Peral, D.F. Ollis, J. Catal. 136 (1992) 554–565.[67] L.A. Phillips, G.B. Raupp, J. Molec. Catal. 77 (1992) 297–311.[68] G.B. Raupp, C.T. Junio, Appl. Surf. Sci. 72 (1993) 321–327.[69] T.N. Obee, R.T. Brown, Environ. Sci. Technol. 29 (1995) 1223–1231.


[70] D.M. Blake, Bibliography of Work on the Photocatalytic Removal ofHazardous Compounds fromWater and Air, Golden, Colorado, USA, NationalRenewable Energy Laboratory, 1994.

[71] D.M. Blake, Bibliography of Work on the Photocatalytic Removal ofHazardous Compounds fromWater and Air Update Number 1 to June, 1995,Golden, Colorado, USA, National Renewable Energy Laboratory, 1995.

[72] D.M. Blake, Bibliography of Work on the Photocatalytic Removal ofHazardous Compounds from Water and Air Update Number 2 to October,1996, Golden, Colorado, USA, National Renewable Energy Laboratory, 1996.

[73] I. Sopyan, S. Murasawa, K. Hashimoto, A. Fujishima, Chem. Lett. (1994) 1994.[74] A.L. Linsebigler, G. Lu, J.J.T. Yates, Chem. Rev. 95 (1995) 735–758.[75] H. Matsubara, M. Takada, S. Koyama, K. Hashimoto, A. Fujishima, Chem. Lett.

(1995) 767–768.[76] N. Negishi, T. Iyoda, K. Hashimoto, A. Fujishima, Chem. Lett. (1995) 841–842.[77] H. Zhang, J.F. Banfield, J. Phys. Chem. B 104 (2000) 3481–3487.[78] M.R. Ranade, A. Navrotsky, H.Z. Zhang, J.F. Banfield, S.H. Elder, A. Zaban, P.H.

Borse, S.K. Kulkarni, G.S. Doran, H.J. Whitfield, Proc. Nat. Acad. Sciences 99(2002) 6476–6481.

[79] V. Shklover, M.K. Nazeeruddin, S.M. Zakeeruddin, C. Barbe, A. Kay, T. Haibach,W. Steurer, R. Hermann, H.U. Nissen, M. Grätzel, Chem. Mater. 9 (1997)430–439.

[80] S.D. Burnside, V. Shklover, C. Barbe, P. Comte, F. Arendse, K. Brooks,M.Grätzel,Chem. Mater. 10 (1998) 2419–2425.

[81] M. Ramamoorthy, D. Vanderbilt, R.D. King-Smith, Phys. Rev. B 49 (1994)16721.

[82] R. Hengerer, B. Bolliger,M. Erbudak,M. Grätzel, Surf. Sci. 460 (2000) 162–169.[83] Y. Liang, S. Gan, S.A. Chambers, E.I. Altman, Phys. Rev. B 63 (2001) 235402.[84] N. Ruzycki, G.S. Herman, L.A. Boatner, U. Diebold, Surf. Sci. 529 (2003)

L239–L244.[85] A. Beltran, L. Gracia, J. Andres, J. Phys. Chem. B 110 (2006) 23417–23423.[86] L.S. Dubrovinsky, N.A. Dubrovinskaia, V. Swamy, J. Muscat, N.M. Harrison,

R. Ahuja, B. Holm, B. Johansson, Nature 410 (2001) 653–654.[87] M. Mattesini, J.S. de Almeida, L. Dubrovinsky, N. Dubrovinskaia, B. Johansson,

R. Ahuja, Phys. Rev. B 70 (2004) 115101-9.[88] S. Andersson, A. Magnéli, Naturwissenschaften 43 (1956) 495–496.[89] L.A. Bursill, B.G. Hyde, Prog. Solid State Chem. 7 (1972) 177–253.[90] Y. Le Page, P. Strobel, J. Solid State Chem. 47 (1983) 6–15.[91] Y. Le Page, P. Strobel, J. Solid State Chem. 44 (1982) 273–281.[92] J.B. Goodenough, Phys. Rev. 117 (1960) 1442.[93] M. Hirasawa, T. Seto, T. Orii, N. Aya, H. Shimura, Appl. Surf. Sci. 197–198

(2002) 661–665.[94] M.D. Earle, Phys. Rev. 61 (1942) 56–62.[95] D.C. Cronemeyer, M.A. Gilleo, Phys. Rev. 82 (1951) 975.[96] D.C. Cronemeyer, Phys. Rev. 87 (1952) 876.[97] R.G. Breckenridge, W.R. Hosler, Phys. Rev. 91 (1953) 793.[98] H. Tang, K. Prasad, R. Sanjines, P.E. Schmid, F. Levy, J. Appl. Phys. 75 (1994)

2042–2047.[99] L. Forro, O. Chauvet, D. Emin, L. Zuppiroli, H. Berger, F. Lévy, J. Appl. Phys. 75

(1994) 633.[100] E. Hendry, F. Wang, J. Shan, T.F. Heinz, M. Bonn, Phys. Rev. B 69 (2004)

081101-4.[101] R.P. Feynman, Phys. Rev. 97 (1955) 660–665.[102] B.O. Aduda, P. Ravirajan, K.L. Choy, J. Nelson, Int. J. Photoenergy 6 (2004)

141–147.[103] O.K. Varghese, D. Gong, M. Paulose, K.G. Ong, C.A. Grimes, Sens. Actuators B

93 (2003) 338–344.[104] G.K.Mor,M.A. Carvalho, O.K. Varghese,M.V. Pishko, C.A. Grimes, J. Mater. Res.

19 (2004) 628–634.[105] Abu S. Zuruzi, Noel C. MacDonald, M. Moskovits, A. Kolmakov, Angew. Chem.

Internat. Ed. 46 (2007) 4298–4301.[106] U. Diebold, Surf. Sci. Rep. 48 (2003) 53–229.[107] D. Vogtenhuber, R. Podloucky, A. Neckel, S.G. Steinemann, A.J. Freeman, Phys.

Rev. B 49 (1994) 2099.[108] R. Lindsay, A. Wander, A. Ernst, B. Montanari, G. Thornton, N.M. Harrison,

Phys. Rev. Lett. 94 (2005) 246102-4.[109] V. Swamy, J. Muscat, J.D. Gale, N.M. Harrison, Surf. Sci. 504 (2002) 115–124.[110] H. Uetsuka, A. Sasahara, H. Onishi, Langmuir 20 (2004) 4782–4783.[111] R. Nakamura, N. Ohashi, A. Imanishi, T. Osawa, Y. Matsumoto, H. Koinuma,

Y. Nakato, J. Phys. Chem. B 109 (2005) 1648–1651.[112] Z. Zhang, P. Fenter, N.C. Sturchio, M.J. Bedzyk, M.L. Machesky,

D.J. Wesolowski, Surf. Sci. 601 (2007) 1129–1143.[113] M.A. Henderson, Surf. Sci. Rep. 46 (2002) 1–308.[114] M.A. Henderson, Surf. Sci. 355 (1996) 151–166.[115] M.S. Shultz, S. Baldelli, C. Schnitzer, D. Simonelli, J. Phys. Chem. B 106 (2002)

5313–5324.[116] E.V. Stefanovich, T.N. Truong, Chem. Phys. Lett. 299 (1999) 623–629.[117] R. Schaub, P. Thostrup, N. Lopez, E. Laegsgaard, I. Stensgaard, J.K. Nørskov,

F. Besenbacher, Phys. Rev. Lett. 87 (2001) 266104.[118] W. Langel, Surf. Sci. 496 (2002) 141–150.[119] L.A. Harris, A.A. Quong, Phys. Rev. Lett. 93 (2004) 086105-4.[120] A.V. Bandura, D.G. Sykes, V. Shapovalov, T.N. Troung, J.D. Kubicki,

R.A. Evarestov, J. Phys. Chem. B 108 (2004) 7844–7853.[121] L.A. Harris, A.A. Quong, Phys. Rev. Lett. 95 (2005) 029602-1.[122] H. Perron, J. Vandenborre, C. Domain, R. Drot, J. Roques, E. Simoni,

J.J. Ehrhardt, H. Catalette, Surf. Sci. 601 (2007) 518–527.[123] P.J.D. Lindan, N.M. Harrison, J.M. Holender, M.J. Gillan, Chem. Phys. Lett. 261

(1996) 246–252.

[124] J. Goniakowski, M.J. Gillan, Surf. Sci. 350 (1996) 145–158.[125] P.J.D. Lindan, C. Zhang, Phys. Rev. B 72 (2005) 075439-7.[126] P.J.D. Lindan, N.M. Harrison, M.J. Gillan, Phys. Rev. Lett. 80 (1998) 762.[127] C. Zhang, P.J.D. Lindan, J. Chem. Phys. 118 (2003) 4620–4630.[128] P.J.D. Lindan, C. Zhang, Phys. Rev. Lett. 95 (2005) 029601-1.[129] S. Wendt, et al., Surf. Sci. 598 (2005) 226–245.[130] M.A. Henderson, Langmuir 12 (1996) 5093–5098.[131] U. Diebold, J. Lehman, T. Mahmoud, M. Kuhn, G. Leonardelli, W. Hebenstreit,

M. Schmid, P. Varga, Surf. Sci. 411 (1998) 137–153.[132] S. Suzuki, K.-I. Fukui, H. Onishi, Y. Iwasawa, Phys. Rev. Lett. 84 (2000) 2156.[133] S. Wendt, J. Matthiesen, R. Schaub, E.K. Vestergaard, E. Laegsgaard,

F. Besenbacher, B. Hammer, Phys. Rev. Lett. 96 (2006) 066107-4.[134] S. Wendt, et al. http://www.phys.au.dk/spm/movies/hydroxylation.mpg,

2005.[135] S. Wendt, et al. http://www.phys.au.dk/spm/movies/Double_OHbr_splitting.

mpg, 2005.[136] R. Schaub, E. Wahlstrom, A. Rønnau, E. Lægsgaard, I. Stensgaard,

F. Besenbacher, Science 299 (2003) 377–379.[137] O. Bikondoa, C.L. Pang, R. Ithnin, C.A. Muryn, H. Onishi, G. Thornton, Nature

Mater. 5 (2006) 189–192.[138] Z. Zhang, O. Bondarchuk, B.D. Kay, J.M. White, Z. Dohnalek, J. Phys. Chem. B

110 (2006) 21840–21845.[139] S.Mezhenny, P.Maksymovych, T.L. Thompson, O. Diwald, D. Stahl, S.D.Walck,

J.T. Yates, Chem. Phys. Lett. 369 (2003) 152–158.[140] A.N. Shultz, W. Jang, W.M. Hetherington, D.R. Baer, L.-Q. Wang,

M.H. Engelhard, Surf. Sci. 339 (1995) 114–124.[141] W.S. Epling, C.H.F. Peden, M.A. Henderson, U. Diebold, Surf. Sci. 412–413

(1998) 333–343.[142] M.A. Henderson, W.S. Epling, C.L. Perkins, C.H.F. Peden, U. Diebold, J. Phys.

Chem. B 103 (1999) 5328–5337.[143] M.A. Henderson, W.S. Epling, C.H.F. Peden, C.L. Perkins, J. Phys. Chem. B 107

(2003) 534–545.[144] A. Rothschild, A. Levakov, Y. Shapira, N. Ashkenasy, Y. Komem, Surf. Sci.

532–535 (2003) 456–460.[145] T.L. Thompson, J.T. Yates, J. Phys. Chem. B 109 (2005) 18230–18236.[146] T.L. Thompson, J.T. Yates, J. Phys. Chem. B 110 (2006) 7431–7435.[147] B.H. Soffer, J. Chem. Phys. 35 (1961) 940–945.[148] G.J. Hill, J. Phys. D: Appl. Phys. (1968) 1151.[149] O.W. Johnson, W.D. Ohlsen, P.I. Kingsbury, Phys. Rev. 175 (1968) 1102.[150] O.W. Johnson, J. DeFord, J.W. Shaner, J. Appl. Phys. 44 (1973) 3008–3012.[151] J.W. DeFord, O.W. Johnson, J. Appl. Phys. 44 (1973) 3001–3007.[152] O.W. Johnson, S.H. Paek, J.W. DeFord, J. Appl. Phys. 46 (1975) 1026–1033.[153] J.B. Bates, R.A. Perkins, Phys. Rev. B 16 (1977) 3713.[154] J.B. Bates, J.C. Wang, R.A. Perkins, Phys. Rev. B 19 (1979) 4130.[155] S. Klauer, M. Wöhlecke, Europhys. Lett. 20 (1992) 439–443.[156] J.V. Cathcart, R.A. Perkins, J.B. Bates, L.C. Manley, J. Appl. Phys. 50 (1979)

4110–4119.[157] J. Robertson, P.W. Peacock, Thin Solid Films 445 (2003) 155–160.[158] P.W. Peacock, J. Robertson, Appl. Phys. Lett. 83 (2003) 2025–2027.[159] D.J. Park, H.H. Nahm, C.H. Park, J. Korean Phys. Soc. 49 (2006) S473; S476.[160] M.V. Koudriachova, S.W. de Leeuw, N.M. Harrison, Phys. Rev. B 70 (2004)

165421-5.[161] J. Leconte, A. Markovits, M.K. Skalli, C. Minot, A. Belmajdoub, Surf. Sci. 497

(2002) 194–204.[162] S.H. Lim, J. Luo, Z. Zhong, W. Ji, J. Lin, Inorg. Chem. 44 (2005) 4124–4126.[163] G. Bromiley, A. Shiryaev, Phys. Chem. Minerals 33 (2006) 426–434.[164] D.A. Panayotov, J.J.T. Yates, Chem. Phys. Lett. 436 (2007) 204–208.[165] P.J. Boddy, J. Electrochem. Soc. 115 (1968) 199–203.[166] A. Fujishima, K. Honda, S. Kikuchi, J. Chem. Soc. Japan 72 (1969) 108 (in

Japanese).[167] A. Fujishima, A. Sakamoto, K. Honda, Seisan Kenkyu 21 (1969) 450.[168] L. Kavan, M. Graetzel, S.E. Gilbert, C. Klemenz, H.J. Scheel, J. Am. Chem. Soc.

118 (1996) 6716–6723.[169] E.C. Dutoit, F. Cardon, W.P. Gomes, Ber. Bunsenge. Phys. Chem. 80 (1976)

475–481.[170] R. De Gryse, W.P. Gomes, F. Cardon, J. Vennik, J. Electrochem. Soc. 122 (1975)

711–712.[171] M. Tomkiewicz, J. Electrochem. Soc. 126 (1979) 1505–1510.[172] D.L. Ullman, J. Electrochem. Soc. 127 (1980) 1321–1322.[173] J. Schoonman, K. Vos, G. Blasse, J. Electrochem. Soc. 128 (1981) 1154–1157.[174] E.H. Rhoderick, R.H. Williams, Metal-Semiconductor Contacts, Clarendon

Press, Oxford, 1988.[175] H.O. Finklea, J. Electrochem. Soc. 129 (1982) 2003–2008.[176] A. Goossens,M. Vazquez, D.D.Macdonald, Electrochim. Acta 41 (1996) 35–45.[177] D.S. Ginley, M.L. Knotek, J. Electrochem. Soc. 126 (1979) 2163–2166.[178] L.A. Harris, R. Schumacher, J. Electrochem. Soc. 127 (1980) 1186–1188.[179] P. Clechet, C. Martelet, R. Olier, J.P. Thomas, M. Fallavier, J. Electrochem. Soc.

130 (1983) 1795–1796.[180] H. Pelouchova, P. Janda, J. Weber, L. Kavan, J. Electroanalyt. Chem. 566 (2004)

73–83.[181] A. Ghicov, H. Tsuchiya, R. Hahn, J.M. Macak, A.G. Munoz, P. Schmuki,

Electrochem. Commun. 8 (2006) 528–532.[182] G. Nogami, J. Electrochem. Soc. 132 (1985) 76–81.[183] J.C. Wang, J. Electrochem. Soc. 134 (1987) 1915–1920.[184] J.C. Wang, J.B. Bates, Solid State Ion. 18&19 (1986) 224–228.[185] B.D. Cahan, C.T. Chen, J. Electrochem. Soc. 129 (1982) 700–705.[186] B.D. Cahan, C.T. Chen, J. Electrochem. Soc. 129 (1982) 474–480.

http://www.phys.au.dk/spm/movies/hydroxylation.mpg









http://www.phys.au.dk/spm/movies/Double_OHbr_splitting.mpg












[187] F. Cao, G. Oskam, P.C. Searson, J.M. Stipkala, T.A. Heimer, F. Farzad, G.J. Meyer,J. Phys. Chem. 99 (1995) 11974–11980.

[188] A. Muñoz, Q. Chen, P. Schmuki, J. Solid State Electrochem. 11 (2007)1077–1084.

[189] J.J. Kim, D.A. Tryk, T. Amemiya, K. Hashimoto, A. Fujishima, J. Electroanalyt.Chem. 435 (1997) 31–38.

[190] J.J. Kim, D.A. Tryk, T. Amemiya, K. Hashimoto, A. Fujishima, J. Electroanalyt.Chem. 433 (1997) 9–17.

[191] B. Wiecek, U. Twardoch, J. Phys. Chem. Solids 65 (2004) 263–268.[192] B. Parkinson, F. Decker, J.F. Juliaõ, M. Abramovich, H.C. Chagas, Electrochim.

Acta 25 (1980) 521–525.[193] F. Fabregat-Santiago, I. Mora-Sero, G. Garcia-Belmonte, J. Bisquert, J. Phys.

Chem. B 107 (2003) 758–768.[194] L. Kavan, K. Kratochvilova, M. Grätzel, J. Electroanalyt. Chem. 394 (1995)

93–102.[195] L. Kavan, M. Grätzel, J. Rathousky, A. Zukalb, J. Electrochem. Soc. 143 (1996)

394–400.[196] B. O’Regan,M. Grätzel, D. Fitzmaurice, J. Phys. Chem. 95 (1991) 10525–10528.[197] G. Redmond, D. Fitzmaurice, M. Grätzel, J. Phys. Chem. 97 (1993) 6951–6954.[198] G. Redmond, D. Fitzmaurice, J. Phys. Chem. 97 (1993) 1426–1430.[199] G. Rothenberger, D. Fitzmaurice, M. Grätzel, J. Phys. Chem. 96 (1992)

5983–5986.[200] G. Boschloo, D. Fitzmaurice, J. Phys. Chem. B 103 (1999) 7860–7868.[201] G. Boschloo, D. Fitzmaurice, J. Phys. Chem. B 103 (1999) 2228–2231.[202] L.A. Lyon, J.T. Hupp, J. Phys. Chem. B 103 (1999) 4623–4628.[203] P. Clechet, C. Martelet, J.R. Martin, R. Olier, Electrochim. Acta 24 (1979)

457–461.[204] J.M. Kesselman, G.A. Shreve, M.R. Hoffmann, N.S. Lewis, J. Phys. Chem. 98

(1994) 13385–13395.[205] D. Tafalla, P. Salvador, Ber. Bunsenges. Phys. Chem. 91 (1987) 475–479.[206] A. Tsujiko, H. Itoh, T. Kisumi, A. Shiga, K. Murakoshi, Y. Nakato, J. Phys. Chem.

B 106 (2002) 5878–5885.[207] M.R. Tarasevich, A. Sadkowski, E. Yeager, B.E. Conway, J.O’M. Bockris,

E. Yeager, S.U.M. Khan, R.E. White (Eds.), Kinetics and Mechanisms ofElectrode Processes, New York, 1983, pp. 301–398.

[208] Y.V. Pleskov, Y.Y. Gurevich, Semiconductor Photoelectrochemistry, Consul-tants Bureau, New York, 1986.

[209] J.F. Dewald, J. Phys. Chem. Solids 14 (1960) 155–161.[210] J.F. Dewald, Bell Syst. Tech. J. 39 (1960) 615–639.[211] W.H. Brattain, P.J. Boddy, J. Electrochem. Soc. 109 (1962) 574–582.[212] H. Gerischer, J. Electrochem. Soc. 113 (1966) 1174–1182.[213] A. Fujishima, K. Honda, Seisan Kenkyu 22 (1970) 478.[214] A. Fujishima, K. Honda, Bull. Chem. Soc. Jpn. 44 (1971) 1148–1150.[215] P. Salvador, C. Gutierrez, J. Phys. Chem. 88 (1984) 3696–3698.[216] P. Salvador, J. Phys. Chem. 89 (1985) 3863–3869.[217] Y. Nakato, H. Akanuma, Y. Magari, S. Yae, J.I. Shimizu, H. Mori, J. Phys. Chem.

B 101 (1997) 4934–4939.[218] T. Kisumi, A. Tsujiko, K. Murakoshi, Y. Nakato, J. Electroanalyt. Chem. 545

(2003) 99–107.[219] R. Nakamura, Y. Nakato, J. Am. Chem. Soc. 126 (2004) 1290–1298.[220] R. Nakamura, T. Okamura, N. Ohashi, A. Imanishi, Y. Nakato, J. Am. Chem. Soc.

127 (2005) 12975–12983.[221] B. Neumann, P. Bogdanoff, H. Tributsch, S. Sakthivel, H. Kisch, J. Phys. Chem.

B 109 (2005) 16579–16586.[222] Z.w. Qu, G.J. Kroes, J. Phys. Chem. B 110 (2006) 23306–23314.[223] T.L. Villarreal, R. Gomez, M. Neumann-Spallart, N. Alonso-Vante, P. Salvador,

J. Phys. Chem. B 108 (2004) 15172–15181.[224] T. Inoue, C.Weber, A. Fujishima, K. Honda, Bull. Chem. Soc. Jpn. 53 (1980) 234.[225] B. Kraeutler, A.J. Bard, J. Am. Chem. Soc. 100 (1978) 4317–4318.[226] C.-M.Wang, A. Heller, H. Gerischer, J. Am. Chem. Soc. 114 (1992) 5230–5234.[227] I. Izumi, W.W. Dunn, K.O. Wilbourn, F.-R.F. Fan, A.J. Bard, J. Phys. Chem. 84

(1980) 3207–3210.[228] H. Sakai, R. Baba, K. Hashimoto, A. Fujishima, J. Electroanal. Chem. 379 (1994)

199–205.[229] H. Sakai, R. Baba, K. Hashimoto, A. Fujishima, A. Heller, J. Phys. Chem. 99

(1995) 11896–11900.[230] K. Ikeda, H. Sakai, R. Baba, K. Hashimoto, A. Fujishima, J. Phys. Chem. B 101

(1997) 2617–2620.[231] K. Ikeda, K. Hashimoto, A. Fujishima, J. Electroanal. Chem. 437 (1997)

241–244.[232] H. Maeda, K. Ikeda, K. Hashimoto, K. Ajito, M. Morita, A. Fujishima, J. Phys.

Chem. B 103 (1999) 3213–3217.[233] K. Ikeda, H. Sakai, R. Baba, K. Hashimoto, A. Fujishima, Chem. Lett. (1995)

979–980.[234] H. Gerischer, A. Heller, J. Phys. Chem. 95 (1991) 5261–5267.[235] J. Schwitzgebel, J.G. Ekerdt, H. Gerischer, A. Heller, J. Phys. Chem. 99 (1995)

5633–5638.[236] H. Gerischer, A. Heller, J. Electrochem. Soc. 139 (1992) 113–118.[237] G. Nicolis, I. Prigogine, Self-Organization in Nonequilibrium Systems: From

Dissipative Structures to Order through Fluctuations, JohnWiley & Sons, NewYork, 1977.

[238] A.M. Peiro, C. Colombo, G. Doyle, J. Nelson, A.Mills, J.R. Durrant, J. Phys. Chem.B 110 (2006) 23255–23263.

[239] A.V. Emeline, V.K. Ryabchuk, N. Serpone, J. Phys. Chem. B 109 (2005)18515–18521.

[240] Y. Wang, K. Hang, N.A. Anderson, T. Lian, J. Phys. Chem. B 107 (2003)9434–9440.

[241] A. Furube, T. Asahi, H. Masuhara, H. Yamashita, M. Anpo, Chem. Phys. Lett.336 (2001) 424–430.

[242] A. Furube, T. Asahi, H. Masuhara, H. Yamashita, M. Anpo, J. Phys. Chem. B 103(1999) 3120–3127.

[243] P.V. Kamat, I. Bedja, S. Hotchandani, J. Phys. Chem. 98 (1994) 9137–9142.[244] A. Henglein, Chem. Rev. 89 (1989) 1861–1873.[245] U. Kölle, J. Moser, M. Grätzel, Inorg. Chem. 24 (1985) 2253–2258.[246] T. Yoshihara, R. Katoh, A. Furube, Y. Tamaki, M. Murai, K. Hara, S. Murata,

H. Arakawa, M. Tachiya, J. Phys. Chem. B 108 (2004) 3817–3823.[247] N. Serpone, D. Lawless, R. Khairutdinov, E. Pelizzetti, J. Phys. Chem. 99 (1995)

16655–16661.[248] G. Rothenberger, J. Moser, M. Graetzel, N. Serpone, D.K. Sharma, J. Am. Chem.

Soc. 107 (1985) 8054–8059.[249] D. Lawless, N. Serpone, D. Meisel, J. Phys. Chem. 95 (1991) 5166–5170.[250] D.P. Colombo Jr., R.M. Bowman, J. Phys. Chem. 100 (1996) 18445–18449.[251] D.P. Colombo Jr., R.M. Bowman, J. Phys. Chem. 99 (1995) 11752–11756.[252] D.W. Bahnemann, M. Hilgendorff, R. Memming, J. Phys. Chem. B 101 (1997)

4265–4275.[253] D. Bahnemann, A. Henglein, J. Lilie, L. Spanhel, J. Phys. Chem. 88 (1984)

709–711.[254] C. Colbeau-Justin, M. Kunst, D. Huguenin, J. Mater. Sci. 38 (2003) 2429–2437.[255] A. Yamakata, T.-A. Ishibashi, H. Onishi, Chem. Phys. Lett. 333 (2001) 271–277.[256] A. Yamakata, T. Ishibashi, H. Onishi, J. Phys. Chem. B 105 (2001) 7258–7262.[257] S.H. Szczepankiewicz, A.J. Colussi, M.R. Hoffmann, J. Phys. Chem. B 104 (2000)

9842–9850.[258] A. Yamakata, T.-A. Ishibashi, K. Takeshita, H. Onishi, Topics Catal. V35 (2005)

211–216.[259] S.T. Martin, H. Herrmann, M.R. Hoffmann, J. Chem. Soc. Faraday Trans. 90

(1994) 3323–3330.[260] S.T. Martin, H. Herrmann, W. Choi, M.R. Hoffmann, J. Chem. Soc. Faraday

Trans. 90 (1994) 3315–3322.[261] X. Yang, N. Tamai, Phys. Chem. Chem. Phys. 3 (2001) 3393–3398.[262] Y. Tamaki, A. Furube, M. Murai, K. Hara, R. Katoh, M. Tachiya, Phys. Chem.

Chem. Phys. 9 (2007) 1453–1460.[263] Y. Tamaki, A. Furube, R. Katoh, M. Murai, K. Hara, H. Arakawa, M. Tachiya,

C. R. Chimie 9 (2006) 268–274.[264] Y. Tamaki, A. Furube, M. Murai, K. Hara, R. Katoh, M. Tachiya, J. Am. Chem.

Soc. 128 (2006) 416–417.[265] K. Iwata, T. Takaya, H. Hamaguchi, A. Yamakata, T. Ishibashi, H. Onishi,

H. Kuroda, J. Phys. Chem. B 108 (2004) 20233–20239.[266] D.C. Hurum, A.G. Agrios, K.A. Gray, T. Rajh, M.C. Thurnauer, J. Phys. Chem. B

107 (2003) 4545–4549.[267] O. Micic, Y. Zhang, K.R. Cromack, A. Trifunac, M. Thurnauer, J. Phys. Chem. 97

(1993) 13284–13288.[268] T. Berger, M. Sterrer, O. Diwald, E. Knozinger, D. Panayotov, T.L. Thompson,

J.T. Yates, J. Phys. Chem. B 109 (2005) 6061–6068.[269] Y. Nakaoka, Y. Nosaka, J. Photochem. Photobiol. A: Chem. 110 (1997)

299–305.[270] O.I. Micic, Y. Zhang, K.R. Cromack, A.D. Trifunac, M.C. Thurnauer, J. Phys.

Chem. 97 (1993) 7277–7283.[271] R.F. Howe, M. Grätzel, J. Phys. Chem. 89 (1985) 4495–4499.[272] R.B. Lauer, J.R.R. Addiss, A.K. Ghosh, J. Appl. Phys. 42 (1971) 3508–3512.[273] J.R.R. Addiss, A.K. Ghosh, F.G. Wakim, Appl. Phys. Lett. 12 (1968) 397–400.[274] A.K. Ghosh, F.G. Wakim, R.R. Addiss, Phys. Rev. 184 (1969) 979.[275] G.K. Boschloo, A. Goossens, J. Phys. Chem. 100 (1996) 19489–19494.[276] S. Leytner, J.T. Hupp, Chem. Phys. Lett. 330 (2000) 231–236.[277] N. Serpone, D. Lawless, R. Khairutdinov, J. Phys. Chem. 99 (1995)

16646–16654.[278] N.D. Abazovic, M.I. Comor, M.D. Dramicanin, D.J. Jovanovic, S.P. Ahrenkiel,

J.M. Nedeljkovic, J. Phys. Chem. B 110 (2006) 25366–25370.[279] D.A. Panayotov, J.J.T. Yates, Chem. Phys. Lett. 410 (2005) 11–17.[280] M.A. Henderson, J.M. White, H. Uetsuka, H. Onishi, J. Am. Chem. Soc. 125

(2003) 14974–14975.[281] C. Di Valentin, G. Pacchioni, A. Selloni, Phys. Rev. Lett. 97 (2006) 166803-4.[282] H.P. Boehm, Discuss. Faraday Soc. 52 (1971) 264–275.[283] Y. Nosaka, T. Daimon, A.Y. Nosaka, Y. Murakami, Phys. Chem. Chem. Phys. 6

(2004) 2917–2918.[284] T. Tachikawa,M. Fujitsuka, T.Majima, J. Phys. Chem. C 111 (2007) 5259–5275.[285] W. Xu, D. Raftery, J. Phys. Chem. B 105 (2001) 4343–4349.[286] S. Pilkenton, S.-J. Hwang, D. Raftery, J. Phys. Chem. B 103 (1999)

11152–11160.[287] I.A. Shkrob,M.C. Sauer, D. Gosztola, J. Phys. Chem. B 108 (2004) 12512–12517.[288] I.A. Shkrob, M.C. Sauer, J. Phys. Chem. B 108 (2004) 12497–12511.[289] S. Tojo, T. Tachikawa, M. Fujitsuka, T. Majima, Phys. Chem. Chem. Phys. 6

(2004) 960–964.[290] A.Y. Nosaka, T. Fujiwara, H. Yagi, H. Akutsu, Y. Nosaka, Langmuir 19 (2003)

1935–1937.[291] C.y. Wang, H. Groenzin, M.J. Shultz, J. Am. Chem. Soc. 126 (2004) 8094–8095.[292] G. Riegel, J.R. Bolton, J. Phys. Chem. 99 (1995) 4215–4224.[293] H. Noda, K. Oikawa, H. Ohya-Nishiguchi, H. Kamada, Bull. Chem. Soc. Jpn. 67

(1994) 2031–2037.[294] C.D. Jaeger, A.J. Bard, J. Phys. Chem. 83 (1979) 3146–3152.[295] M.A. Grela, M.E.J. Coronel, A.J. Colussi, J. Phys. Chem. B 100 (1996)

16940–16946.[296] R.B. Draper, M.A. Fox, Langmuir 6 (1990) 1396–1402.[297] R.B. Draper, M.A. Fox, J. Phys. Chem. 94 (1990) 4628–4634.


[298] T. Tachikawa, S. Tojo, M. Fujitsuka, T. Majima, Langmuir 20 (2004)2753–2759.

[299] S. Tojo, T. Tachikawa, M. Fujitsuka, T. Majima, Chem. Phys. Lett. 384 (2004)312–316.

[300] T. Tachikawa, S. Tojo, M. Fujitsuka, T. Majima, J. Phys. Chem. B 108 (2004)5859–5866.

[301] K. Ishibashi, A. Fujishima, T.Watanabe, K. Hashimoto, J. Photochem. PhotobiolA: Chem. 134 (2000) 139–142.

[302] Y. Nosaka, S. Komori, K. Yawata, T. Hirakawa, A.Y. Nosaka, Phys. Chem. Chem.Phys. (2003) 4731–4735.

[303] P. Salvador, J. Electrochem. Soc. 128 (1981) 1895–1900.[304] C. Gutierrez, P. Salvador, J. Electrochem. Soc. 133 (1986) 924–929.[305] Y. Murakami, E. Kenji, A.Y. Nosaka, Y. Nosaka, J. Phys. Chem. B 110 (2006)

16808–16811.[306] O. Legrini, E. Oliveros, A.M. Braun, Chem. Rev. 93 (1993) 671–698.[307] S.J. Hwang, C. Petucci, D. Raftery, J. Am. Chem. Soc. 120 (1998) 4388–4397.[308] G.Munuera, A. Navio, V. Rives-Arnau, J. Chem. Soc., Faraday Trans. 1 77 (1981)

2747–2749.[309] T. Daimon, Y. Nosaka, J. Phys. Chem. C 111 (2007) 4420–4424.[310] K. Hirakawa, T. Hirano, Chem. Lett. 35 (2006) 832.[311] J. Janczyk, E. Krakowska, G. Stochel, W. Macyk, J. Am. Chem. Soc. 128 (2006)

15574–15575.[312] T. Berger, M. Sterrer, O. Diwald, E. Knözinger, Chem. Phys. Chem. 6 (2005)

2104.[313] K.-I. Ishibashi, A. Fujishima, T. Watanabe, K. Hashimoto, J. Phys. Chem. B 104

(2000) 4934–4938.[314] K. Ishibashi, Y. Nosaka, K.Hashimoto, A. Fujishima, J. Phys. Chem. B102 (1998)

2117–2120.[315] J.M. Coronado, A.J. Maira, J.C. Conesa, K.L. Yeung, V. Augugliaro, J. Soria,

Langmuir 17 (2001) 5368–5374.[316] M. Anpo, N. Aikawa, Y. Kubokawa, M. Che, C. Louis, E. Gianello, J. Phys. Chem.

89 (1985) 5689–5694.[317] Y. Nosaka, M. Nakamura, T. Hirakawa, Phys. Chem. Chem. Phys. 4 (2002)

1088–1092.[318] T. Hirakawa, H. Kominami, B. Ohtani, Y. Nosaka, J. Phys. Chem. B 105 (2001)

6993–6999.[319] R. Nakamura, A. Imanishi, K. Murakoshi, Y. Nakato, J. Am. Chem. Soc. 125

(2003) 7443–7450.[320] A. Tilocca, C. DiValentin, A. Selloni, J. Phys. Chem. B 109 (2005) 20963–20967.[321] A.L. Attwood, D.M. Murphy, J.L. Edwards, T.A. Egerton, R.W. Harrison, Res.

Chem. Intermed. 29 (2003) 449–465.[322] H. Goto, Y. Hanada, T. Ohno, M. Matsumura, J. Catal. 225 (2004) 223–229.[323] X. Li, W.S. Jenks, J. Am. Chem. Soc. 122 (2000) 11864–11870.[324] A.L. Attwood, J.L. Edwards, C.C. Rowlands, D.M. Murphy, J. Phys. Chem. A 107

(2003) 1779–1782.[325] C.A. Jenkins, D.M. Murphy, J. Phys. Chem. B 103 (1999) 1019–1026.[326] A.R. Gonzalez-Elipe, M. Che, J. Chim. Phys. 79 (1982) 535.[327] E. Carter, A.F. Carley, D.M. Murphy, Chem. Phys. Chem. 8 (2007) 113–123.[328] P. Pichat, H. Courbon, R. Enriquez, T.T.Y. Tan, R. Amal, Res. Chem. Intermed.

33 (2007) 239–250.[329] D.S. Muggli, J.L. Falconer, J. Catal. 187 (1999) 230–237.[330] D.S. Muggli, J.L. Falconer, J. Catal. 191 (2000) 318–325.[331] K. Yoshida, J. Yamasaki, N. Tanaka, Nanotechnol. (2004) S349.[332] K. Yoshida, J. Yamasaki, N. Tanaka, Appl. Phys. Lett. 84 (2004) 2542–2544.[333] K. Yoshida, T. Nanbara, J. Yamasaki, N. Tanaka, J. Appl. Phys. 99 (2006)

084908-8.[334] M.T. Chen, C.F. Lien, L.F. Liao, J.L. Lin, J. Phys. Chem. B 107 (2003) 3837–3843.[335] T.L. Thompson, D.A. Panayotov, J.T. Yates, I. Martyanov, K. Klabunde, J. Phys.

Chem. B 108 (2004) 17857–17865.[336] J. Zhuang, C.N. Rusu, J.T. Yates, J. Phys. Chem. B 103 (1999) 6957–6967.[337] S.A. Larson, J.A. Widegren, J.L. Falconer, J. Catal. 157 (1995) 611–625.[338] P.A. Morris Hotsenpiller, J.D. Bolt, W.E. Farneth, J.B. Lowekamp, G.S. Rohrer, J.

Phys. Chem. B 102 (1998) 3216–3226.[339] J.B. Lowekamp, G.S. Rohrer, P.A. Morris Hotsenpiller, J.D. Bolt, W.E. Farneth, J.

Phys. Chem. B 102 (1998) 7323–7327.[340] Y. Yamamoto, K. Nakajima, T. Ohsawa, Y. Matsumoto, H. Koinuma, Japan. J.

Appl. Phys. 44 (2005) L511–L514.[341] Y. Nakato, H. Akanuma, J.-i. Shimizu, Y. Magari, J. Electroanalyt. Chem. 396

(1995) 35–39.[342] A. Tsujiko, T. Kisumi, Y. Magari, K. Murakoshi, Y. Nakato, J. Phys. Chem. B 104

(2000) 4873–4879.[343] T. Sugiura, S. Itoh, T. Ooi, T. Yoshida, K. Kuroda, H. Minoura, J. Electroanalyt.

Chem. 473 (1999) 204–208.[344] T. Sugiura, T. Yoshida, H. Minoura, Electrochem. Solid-State Lett. 1 (1998)

175–177.[345] T. Ohno, K. Sarukawa, M. Matsumura, New J. Chem. 26 (2002) 1167–1170.[346] T. Taguchi, Y. Saito, K. Sarukawa, T. Ohno, M. Matsumura, New J. Chem. 27

(2003).[347] Y. Kikuchi, K. Sunada, T. Iyoda, K. Hashimoto, A. Fujishima, J. Photochem.

Photobiol. A: Chem. 106 (1997) 51–56.[348] K. Naito, T. Tachikawa, S.C. Cui, A. Sugimoto, M. Fujitsuka, T. Majima, J. Am.

Chem. Soc. 128 (2006) 16430–16431.[349] J.S. Park, W. Choi, Chem. Lett. (2005) 1630.[350] T. Tatsuma, S.-i. Tachibana, A. Fujishima, J. Phys. Chem. B 105 (2001)

6987–6992.[351] K. Naito, T. Tachikawa, M. Fujitsuka, T. Majima, J. Phys. Chem. B 109 (2005)

23138–23140.

[352] T. Tatsuma, S.-i. Tachibana, T. Miwa, D.A. Tryk, A. Fujishima, J. Phys. Chem. B103 (1999) 8033–8035.

[353] M.C. Lee, W. Choi, J. Phys. Chem. B 106 (2002) 11818–11822.[354] J.S. Park, W. Choi, Langmuir 20 (2004) 11523–11527.[355] S. Cho, W. Choi, J. Photochem. Photobiol. A: Chem. 143 (2001) 221–228.[356] K. Kawahara, Y. Ohko, T.T.a.A. Fujishima, Phys. Chem. Chem. Phys. 5 (2003)

4764–4766.[357] H. Haick, Y. Paz, Chem. Phys. Chem. 4 (2003) 617–620.[358] H. Haick, Y. Paz, J. Phys. Chem. B 105 (2001) 3045–3051.[359] Y. Ishikawa, Y. Matsumoto, Y. Nishida, S. Taniguchi, J. Watanabe, J. Am. Chem.

Soc. 125 (2003) 6558–6562.[360] T. Tatsuma, W. Kubo, A. Fujishima, Langmuir 18 (2002) 9632–9634.[361] W. Kubo, T. Tatsuma, Analyt. Sci. 20 (2004) 591–593.[362] W. Kubo, T. Tatsuma, J. Am. Chem. Soc. 128 (2006) 16034–16035.[363] W. Kubo, T. Tatsuma, A. Fujishima, H. Kobayashi, J. Phys. Chem. B 108 (2004)

3005–3009.[364] A.Y. Nosaka, E. Kojima, T. Fujiwara, H. Yagi, H. Akutsu, Y. Nosaka, J. Phys.

Chem. B 107 (2003) 12042–12044.[365] M.L. Sauer, D.F. Ollis, J. Catal. 163 (1996) 215–217.[366] Y. Ohko, K. Hashimoto, A. Fujishima, J. Phys. Chem. A 101 (1997) 8057–8062.[367] Y. Ohko, D.A. Tryk, K. Hashimoto, A. Fujishima, J. Phys. Chem. B 102 (1998)

2699–2704.[368] N.N. Lichtin, M. Avudaithai, Environ. Sci. Technol. 30 (1996) 2014–2020.[369] T.N. Obee, S.O. Hay, Environ. Sci. Technol. 31 (1997) 2034–2038.[370] T. Sano, N. Negishi, K. Takeuchi, S. Matsuzawa, Solar Energy 77 (2004)

543–552.[371] H. Einaga, S. Futamura, T. Ibusuki, Appl. Catal. B: Environ. 38 (2002) 215–225.[372] M.C. Blount, J.L. Falconer, Appl. Catal. B: Environ. 39 (2002) 39–50.[373] W.-K. Jo, K.-H. Park, Chemosphere 57 (2004) 555–565.[374] J. Peral, D.F. Ollis, J. Molec, Catal. A: Chem. 115 (1997) 347–354.[375] N. González-García, J.A. Ayllón, X. Doménech, J. Peral, Appl. Catal. B: Environ.

52 (2004) 69–77.[376] R.-D. Sun, A. Nakajima, T. Watanabe, K. Hashimoto, J. Photochem. Photobiol.

A: Chem. 154 (2003) 203–209.[377] J.A. Moss, S.H. Szczepankiewicz, E. Park, M.R. Hoffmann, J. Phys. Chem. B 109

(2005) 19779–19785.[378] H. Einaga, T. Ibusuki, S. Futamura, Environ. Sci. Technol. 38 (2004) 285–289.[379] D.F. Ollis, E. Pelizzetti, N. Serpone, Environ. Sci. Technol. 25 (1991)

1522–1529.[380] D. Bahnemann, Solar Energy 77 (2004) 445–459.[381] J.M. Herrmann, Topics Catal. 34 (2005) 49–65.[382] J. Blanco-Galvez, P. Fernandez-Ibanez, S. Malato-Rodriguez, J. Solar Energy

Eng. 129 (2007) 4–15.[383] A. Salinaro, A.V. Emeline, J. Zhao, H. Hidaka, V.K. Ryabchuk, N. Serpone, Pure

Appl. Chem. 71 (1999) 321–335.[384] N. Serpone, A. Salinaro, Pure Appl. Chem. 71 (1999) 303–320.[385] N. Serpone, J. Photochem. Photobiol. A: Chem. 104 (1997) 1–12.[386] R. Enríquez, A.G. Agrios, P. Pichat, Catal. Today 120 (2007) 196–202.[387] A.G. Agrios, P. Pichat, J. Photochem. Photobiol. A: Chem. 180 (2006) 130–135.[388] R. Enriquez, P. Pichat, J. Environ. Science Health A 41 (2006) 955–966.[389] S. Sakthivel, M.V. Shankar, M. Palanichamy, B. Arabindoo, D.W. Bahnemann,

V. Murugesan, Water Res. 38 (2004) 3001–3008.[390] M. Kositzi, A. Antoniadis, I. Poulios, I. Kiridis, S.Malato, Solar Energy 77 (2004)

591–600.[391] T.E. Agustina, H.M. Ang, V.K. Vareek, J. Photochem. Photobiol. C: Photochem.

Rev. 6 (2005) 264–273.[392] T. Watanabe, K. Hashimoto, A. Fujishima, 1st International Conference on

TiO2 Photocatalytic Purification and Treatment of Water and Air, 1992.[393] S. Sitkiewitz, A. Heller, Nouv. J. Chem. 20 (1996) 233–241.[394] Y. Paz, Z. Luo, L. Rabenberg, A. Heller, J. Mater. Res. 10 (1996) 2842.[395] Y. Paz, A. Heller, J. Mater. Res. 12 (1997) 2759.[396] P. Sawunyama, L. Jiang, A. Fujishima, K. Hashimoto, J. Phys. Chem. B 101

(1997) 11000–11003.[397] P. Sawunyama, L. Jiang, A. Fujishima, K. Hashimoto, Langmuir 15 (1999)

3551–3556.[398] V. Romeas, P. Pichat, C. Guillard, T. Chopin, C. Lehaut, Ind. Eng. Chem. Res. 38

(1999) 3878–3885.[399] V. Roméas, P. Pichat, C. Guillard, T. Chopin, C. Lehaut, New J. Chem. 23 (1999)

365–374.[400] T. Minabe, D.A. Tryk, P. Sawunyama, Y. Kikuchi, K. Hashimoto, A. Fujishima,

J. Photochem. Photobiol. A: Chem. 137 (2000) 53–62.[401] J.T. Remillard, J.R. McBride, K.E. Nietering, A.R. Drews, X. Zhang, J. Phys. Chem.

B 104 (2000) 4440–4447.[402] E. Puzenat, P. Pichat, J. Photochem. Photobiol. A: Chem. 160 (2003) 127–133.[403] A. Mills, A. Lepre, N. Elliott, S. Bhopal, I.P. Parkin, S.A. O’Neill, J. Photochem.

Photobiol. A: Chem. 160 (2003) 213–224.[404] A. Mills, J. Wang, J. Photochem. Photobiol. A: Chem. 182 (2006) 181–186.[405] J.R. Peller, R.L. Whitman, S. Griffith, P. Harris, C. Peller, J. Scalzitti, J.

Photochem. Photobiol. A: Chem. 186 (2007) 212–217.[406] C.McCullagh, J.M.C. Robertson, D.W. Bahnemann, P.K.J. Robertson, Res. Chem.

Intermed. 33 (2007) 359–375.[407] A. Fujishima, J. Ohtsuki, T. Yamashita, S. Hayakawa, Photochemistry and

Photobiology (Proceedings of the 8th AnnualMeeting of the Japanese Societyfor Photomedicine and Photobiology) 1986, 45.

[408] T. Matsunaga, R. Tomoda, T. Nakajima, H. Wake, FEMS Microbiol. Lett. 29(1985) 211–214.

[409] R. Cai, K. Hashimoto, Y. Kubota, A. Fujishima, Chem. Lett. (1992) 427–430.


[410] R. Cai, K. Hashimoto, K. Itoh, Y. Kubota, A. Fujishima, Bull. Chem. Soc. Jpn. 64(1991) 1268–1273.

[411] K.P. Kühn, I.F. Chaberny, K. Massholder, M. Stickler, V.W. Benz, H.-G. Sonntag,L. Erdinger, Chemosphere 53 (2003) 71–77.

[412] K. Sunada, Y. Kikuchi, K. Hashimoto, A. Fujishima, Environ. Sci. Technol. 32(1998) 726–728.

[413] E.J. Wolfrum, J. Huang, D.M. Blake, P.C. Maness, Z. Huang, J. Fiest, W.A. Jacoby,Environ. Sci. Technol. 36 (2002) 3412–3419.

[414] K. Sunada, T.Watanabe, K. Hashimoto, J. Photochem. Photobiol. A: Chem. 156(2003) 227–233.

[415] W.A. Jacoby, B.C. Maness, E.J. Wolfrum, D.M. Blake, J.A. Fennell, Environ. Sci.Technol. 32 (1998) 2650–2653.

[416] K. Sunada, T. Watanabe, K. Hashimoto, Environ. Sci. Technol. 37 (2003)4785–4789.

[417] J.-M. Hermann, M.N. Mozzanega, P. Pichat, J. Photochem. 22 (1983) 333.[418] J.-M. Herrmann, J. Disdier, P. Pichat, Chem. Phys. Lett. 108 (1984) 618–622.[419] P. Pichat, J.-M. Hermann, J. Disdier, M. Mozzanega, H. Courbon, Stud. Surf. Sci.

Catal. 19 (1984) 319.[420] N. Serpone, D. Lawless, J. Disdier, J.-M. Herrmann, Langmuir 10 (1994)

643–652.[421] M. Anpo, Y. Ichihashi, M. Takeuchi, H. Yamashita, Res. Chem. Intermed. 24

(1998) 143–149.[422] S. Ikeda, N. Sugiyama, B. Pal, G. Marcí, L. Palmisano, H. Noguchi, K. Uosaki, B.

Ohtani, Phys. Chem. Chem. Phys. 3 (2001) 267–273.[423] M. Anpo, M. Takeuchi, J. Catal. 216 (2003) 505–516.[424] S. Karvinen, P. Hirva, T.A. Pakkanen, J. Molec. Struct.: THEOCHEM 626 (2003)

271–277.[425] J. Zhu, Z. Deng, F. Chen, J. Zhang, H. Chen, M. Anpo, J. Huang, L. Zhang, Appl.

Catal. B: Environ. 62 (2006) 329–335.[426] H. Tada, T. Mitsui, T. Kiyonaga, T. Akita, K. Tanaka, Nature Mater. 5 (2006)

782–786.[427] M. Miyauchi, A. Nakajima, K. Hashimoto, T. Watanabe, Adv. Mater. 12 (2000)

1923–1927.[428] I. Nakamura, N. Negishi, S. Kutsuna, T. Ihara, S. Sugihara, K. Takeuchi, J. Mol.

Catal. A: Chem. 162 (2000) 205–212.[429] I. Nakamura, S. Sugihara, K. Takeuchi, Chem. Lett. 29 (2000) 1276–1277.[430] I. Justicia, P. Ordejón, G. Canto, J.L. Mozos, J. Fraxedas, G.A. Battiston, R.

Gerbasi, A. Figueras, Adv. Mater. 14 (2002) 1399–1402.[431] F.M. Hossain, G.E. Murch, L. Sheppard, J. Nowotny, Solid State Ion. 178 (2007)

319–325.[432] T. Ihara, M. Miyoshi, M. Ando, S. Sugihara, Y. Iriyama, J. Mater. Sci. V36 (2001)

4201–4207.[433] H. Kikuchi, M. Kitano, M. Takeuchi, M.Matsuoka, M. Anpo, P.V. Kamat, J. Phys.

Chem. B 110 (2006) 5537–5541.[434] S. Sato, Chem. Phys. Lett. 123 (1986) 126–128.[435] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293 (2001)

269–271.[436] T. Morikawa, R. Asahi, T. Ohwaki, K. Aoki, Y. Taga, Japan. J. Appl. Phys. Part 2

40 (2001) L561-–L563.[437] T. Ihara, M. Miyoshi, Y. Iriyama, O. Matsumoto, S. Sugihara, Appl. Catal. B:

Environ. 42 (2003) 403–409.[438] H. Irie, Y. Watanabe, K. Hashimoto, J. Phys. Chem. B 107 (2003) 5483–5486.[439] R. Nakamura, T. Tanaka, Y. Nakato, J. Phys. Chem. B 108 (2004) 10617–10620.[440] G.R. Torres, T. Lindgren, J. Lu, C.G. Granqvist, S.E. Lindquist, J. Phys. Chem. B

108 (2004) 5995–6003.[441] M. Mrowetz, W. Balcerski, A.J. Colussi, M.R. Hoffmann, J. Phys. Chem. B 108

(2004) 17269–17273.[442] S. Sakthivel, M. Janczarek, H. Kisch, J. Phys. Chem. B 108 (2004) 19384–19387.[443] S. Sakthivel, H. Kisch, Chem. Phys. Chem. 4 (2003) 487–490.[444] O. Diwald, T.L. Thompson, T. Zubkov, E.G. Goralski, S.D. Walck, J.T. Yates, J.

Phys. Chem. B 108 (2004) 6004–6008.[445] J.L. Gole, J.D. Stout, C. Burda, Y. Lou, X. Chen, J. Phys. Chem. B 108 (2004)

1230–1240.[446] C. Burda, Y. Lou, X. Chen, A.C.S. Samia, J. Stout, J.L. Gole, Nano Lett. 3 (2003)

1049–1051.[447] C. Di Valentin, G. Pacchioni, A. Selloni, Phys. Rev. B 70 (2004) 085116–4.[448] H. Ozaki, S. Iwamoto, M. Inoue, Chem. Lett. 34 (2005) 1082–1083.[449] C. Di Valentin, G. Pacchioni, A. Selloni, S. Livraghi, E. Giamello, J. Phys. Chem.

B 109 (2005) 11414–11419.[450] Z. Lin, A. Orlov, R.M. Lambert, M.C. Payne, J. Phys. Chem. B 109 (2005)

20948–20952.[451] Y. Nakano, T. Morikawa, T. Ohwaki, Y. Taga, Appl. Phys. Lett. 86 (2005)

132104–3.[452] S. Livraghi, M.C. Paganini, E. Giamello, A. Selloni, C.D. Valentin, G. Pacchioni,

J. Am. Chem. Soc. 128 (2006) 15666–15671.[453] M. Kitano, K. Funatsu, M. Matsuoka, M. Ueshima, M. Anpo, J. Phys. Chem. B

110 (2006) 25266–25272.[454] M. Maeda, T. Watanabe, J. Electrochem. Soc. 153 (2006) C186–C189.[455] M. Batzill, E.H. Morales, U. Diebold, Phys. Rev. Lett. 96 (2006) 026103–4.[456] A. Nambu, J. Graciani, J.A. Rodriguez, Q. Wu, E. Fujita, J.F. Sanz, J. Chem. Phys.

125 (2006) 094706–8.[457] T. Tachikawa, Y. Takai, S. Tojo, M. Fujitsuka, H. Irie, K. Hashimoto, T. Majima,

J. Phys. Chem. B 110 (2006) 13158–13165.[458] A.V. Emeline, X. Zhang, M. Jin, T. Murakami, A. Fujishima, J. Phys. Chem. B 110

(2006) 7409–7413.[459] N. Serpone, J. Phys. Chem. B 110 (2006) 24287–24293.[460] V.N. Kuznetsov, N. Serpone, J. Phys. Chem. B 110 (2006) 25203–25209.

[461] E.A. Reyes-Garcia, Y. Sun, K. Reyes-Gil, D. Raftery, J. Phys. Chem. C 111 (2007)2738–2748.

[462] H. Chen, A. Nambu, W. Wen, J. Graciani, Z. Zhong, J.C. Hanson, E. Fujita,J.A. Rodriguez, J. Phys. Chem. C 111 (2007) 1366–1372.

[463] C. Belver, R. Bellod, S.J. Stewart, F.G. Requejo, M. Fernández-García, Appl.Catal. B: Environ. 65 (2006) 309–314.

[464] S.-K. Joung, T. Amemiya, M. Murabayashi, K. Itoh, Chem. Eur. J. 12 (2006)5526–5534.

[465] S.U.M. Khan, M. Al-Shahry, W.B. Ingler Jr., Science 297 (2002) 2243–2245.[466] A. Fujishima, Science 301 (2003) 1673a.[467] C. Hägglund, M. Grätzel, B. Kasemo, Science 301 (2003) 1673b.[468] K.S. Lackner, Science 301 (2003).[469] H. Irie, Y. Watanabe, K. Hashimoto, Chem. Lett. 32 (2003) 772–773.[470] J.H. Park, S. Kim, A.J. Bard, Nano Lett. 6 (2006) 24–28.[471] S. Sakthivel, H. Kisch, Angew. Chem. Internat. Ed. 42 (2003) 4908–4911.[472] C. DiValentin, G. Pacchioni, A. Selloni, Chem. Mater. 17 (2005) 6656–6665.[473] H. Irie, S. Washizuka, K. Hashimoto, Thin Solid Films 510 (2006) 21–25.[474] S.-W. Hsu, T.-S. Yang, T.-K. Chen, M.-S. Wong, Thin Solid Films 515 (2007)

3521–3526.[475] T. Umebayashi, T. Yamaki, S. Tanaka, K. Asai, Chem. Lett. 32 (2003) 330–331.[476] T. Umebayashi, T. Yamaki, H. Itoh, K. Asai, Appl. Phys. Lett. 81 (2002) 454–456.[477] T. Umebayashi, T. Yamaki, S. Yamamoto, A. Miyashita, S. Tanaka, T. Sumita, K.

Asai, J. Appl. Phys. 93 (2003) 5156–5160.[478] T. Ohno, T. Mitsui, M. Matsumura, Chem. Lett. 32 (2003) 364–365.[479] T. Ohno, M. Akiyoshi, T. Umebayashi, K. Asai, T. Mitsui, M. Matsumura, Appl.

Catal. A: Gen. 265 (2004) 115–121.[480] K. Takeshita, A. Yamakata, T.-A. Ishibashi, H. Onishi, K. Nishijima, T. Ohno,

J. Photochem. Photobiol. A: Chem. 177 (2006) 269–275.[481] T. Tachikawa, S. Tojo, K. Kawai, M. Endo, M. Fujitsuka, T. Ohno, K. Nishijima,

Z. Miyamoto, T. Majima, J. Phys. Chem. B 108 (2004) 19299–19306.[482] W. Zhao, W. Ma, C. Chen, J. Zhao, Z. Shuai, J. Am. Chem. Soc. 126 (2004)

4782–4783.[483] L. Lin, W. Lin, Y. Zhu, B. Zhao, Y. Xie, Chem. Lett. 34 (2005) 284–285.[484] J.C. Yu, J.G. Yu, W.K. Ho, Z.T. Jiang, L.Z. Zhang, Chem. Mater. 14 (2002)

3808–3816.[485] K. Nukumizu, J. Nunoshige, T. Takata, J.N. Kondo, M. Hara, H. Kobayashi,

K. Domen, Chem. Lett. 32 (2003) 196–197.[486] D. Li, H. Haneda, N.K. Labhsetwar, S. Hishita, N. Ohashi, Chem. Phys. Lett. 401

(2005) 579–584.[487] D. Li, H. Haneda, S. Hishita, N. Ohashi, Chem. Mater. 17 (2005) 2596–2602.[488] G.K. Mor, K. Shankar, M. Paulose, O.K. Varghese, C.A. Grimes, Nano Lett. 5

(2005) 191–195.[489] G.K. Mor, H.E. Prakasam, O.K. Varghese, K. Shankar, C.A. Grimes, Nano Lett. 7

(2007) 2356–2364.[490] O.K. Varghese, C.A. Grimes, Solar Energy Mater. Solar Cells 92 (2008)

374–384.[491] H.M. Yates, M.G. Nolan, D.W. Sheel, M.E. Pemble, J. Photochem. Photobiol. A:

Chem. 179 (2006) 213–223.[492] S.-K. Lee, A. Mills, Chem. Commun. (2003) 2366–2367.[493] A. Mills, J. Wang, J. Photochem. Photobiol. A: Chem. 127 (1999) 123–134.[494] K. Vinodgopal, D.E. Wynkoop, P.V. Kamat, Environ. Sci. Technol. 30 (1996)

1660–1666.[495] H. Kisch, W. Macyk, Chem. Phys. Chem. 3 (2002) 399–400.[496] R. Wang, K. Hashimoto, A. Fujishima, M. Chikuni, E. Kojima, A. Kitamura, M.

Shimohigoshi, T. Watanabe, Nature 388 (1997) 431–433.[497] N. Sakai, R. Wang, A. Fujishima, T. Watanabe, K. Hashimoto, Langmuir 14

(1998) 5918–5920.[498] M. Machida, K. Norimoto, T. Watanabe, K. Hashimoto, A. Fujishima, J. Mater.

Sci. 34 (1999) 2569–2574.[499] R. Wang, N. Sakai, A. Fujishima, T. Watanabe, K. Hashimoto, J. Phys. Chem. B

103 (1999) 2188–2194.[500] T. Watanabe, A. Nakajima, R. Wang, M. Minabe, S. Koizumi, A. Fujishima,

K. Hashimoto, Thin Solid Films 351 (1999) 260–263.[501] M. Miyauchi, A. Nakajima, A. Fujishima, K. Hashimoto, T. Watanabe, Chem.

Mater. 12 (2000) 3–5.[502] A. Nakajima, S. Koizumi, T. Watanabe, K. Hashimoto, Langmuir 16 (2000)

7048–7050.[503] A Nakajima, S.-I. Koizumi, T. Watanabe, K. Hashimoto, J. Photochem.

Photobiol. A: Chem. 146 (2001) 129–132.[504] N. Sakai, A. Fujishima, T.Watanabe, K. Hashimoto, J. Phys. Chem. B 105 (2001)

3023–3026.[505] N. Sakai, A. Fujishima, T. Watanabe, K. Hashimoto, J. Electrochem. Soc. 148

(2001) E395–E398.[506] M. Miyauchi, N. Kieda, S. Hishita, T. Mitsuhashi, A. Nakajima, T. Watanabe,

K. Hashimoto, Surf. Sci. 511 (2002) 401–407.[507] N. Stevens, C.I. Priest, R. Sedev, J. Ralston, Langmuir 19 (2003) 3272–3275.[508] R.-D. Sun, A. Nakajima, A. Fujishima, T. Watanabe, K. Hashimoto, J. Phys.

Chem. B 105 (2001) 1984–1990.[509] M. Miyauchi, A. Nakajima, T. Watanabe, K. Hashimoto, Chem. Mater. 14

(2002) 2812–2816.[510] N. Sakai, A. Fujishima, T.Watanabe, K. Hashimoto, J. Phys. Chem. B 107 (2003)

1028–1035.[511] T. Shibata, H. Irie, K. Hashimoto, J. Phys. Chem. B 107 (2003) 10696–10698.[512] T. Horiuchi, H. Ochi, K. Kaisei, K. Ishida, K. Matsushige, Materials Research

Society, Structure-Property Relationships of Oxide Surfaces and Interfaces II(MRS Proceedings Volume 751) 2002 Z3.47.1-Z3.47-4.


[513] D.C. Hennessy, M. Pierce, K.-C. Chang, S. Takakusagi, H. You, K. Uosaki,Electrochim. Acta 53 (2008) 6173–6177.

[514] R. Wang, K. Hashimoto, A. Fujishima, M. Chikuni, E. Kojima, A. Kitamura,M. Shimohigoshi, T. Watanabe, Adv. Mater. 10 (1998) 135–138.

[515] Z.Z. Gu, A. Fujishima, O. Sato, Appl. Phys. Lett. 85 (2004) 5067–5069.[516] J.M.White, J. Szanyi, M.A. Henderson, J. Phys. Chem. B 107 (2003) 9029–9033.[517] T. Zubkov, D. Stahl, T.L. Thompson, D. Panayotov, O. Diwald, J.T. Yates, J. Phys.

Chem. B 109 (2005) 15454–15462.[518] X. Yan, R. Abe, T. Ohno, M. Toyofuku, B. Ohtani, Thin Solid Films 516 (2008)

5872–5876.[519] A. Imanishi, H. Suzuki, K. Murakoshi, Y. Nakato, J. Phys. Chem. B 110 (2006)

21050–21054.[520] A. Imanishi, T. Okamura, N. Ohashi, R. Nakamura, Y. Nakato, J. Am. Chem. Soc.

(2007).[521] K.-I. Katsumata, A. Nakajima, T. Shiota, N. Yoshida, T. Watanabe,

Y. Kameshima, K. Okada, J. Photochem. Photobiol. A: Chem. 180 (2006)75–79.

[522] H. Irie, S.Washizuka, Y.Watanabe, T. Kako, K. Hashimoto, J. Electrochem. Soc.152 (2005) E351–E356.

[523] H. Honda, A. Ishizaki, R. Sama, K. Hashimoto, A. Fujishima, J. Illuminat. Eng.Soc. Winter (1998) 1998.

[524] L. Cassar, MRS Bulletin 29 (2004) 328–331.[525] X.T. Zhang, O. Sato, M. Taguchi, Y. Einaga, T. Murakami, A. Fujishima, Chem.

Mater. 17 (2005) 696–700.[526] X. Zhang, A. Fujishima, M. Jin, A.V. Emeline, T. Murakami, J. Phys. Chem. B 110

(2006) 25142–25148.[527] A. Bozzi, T. Yuranova, I. Guasaquillo, D. Laub, J. Kiwi, J. Photochem. Photobiol.

A: Chem. 174 (2005) 156–164.[528] A. Bozzi, T. Yuranova, J. Kiwi, J. Photochem. Photobiol. A: Chem. 172 (2005)

27–34.[529] J.M. Herrmann, Top. Catalysis 34 (2005) 49–65.[530] T.N. Rao, A. Fujishima, D.A. Tryk, in: A.J. Bard, M. Stratmann, S. Licht

(Eds.), Encyclopedia of Electrochemistry, in: Semiconductor Electrodes andPhotoelectrochemistry, vol. 6, Weinheim, 2002.

[531] I. Oller, W. Gernjak, M.I. Maldonado, L.A. Pérez-Estrada, J.A. Sánchez-Pérez,S. Malato, J. Hazard. Mater. 138 (2006) 507–517.

[532] M.H. Pérez, G. Peñuela, M.I. Maldonado, O. Malato, P. Fernández-Ibáñez,I. Oller, W. Gernjak, S. Malato, Appl. Catal. B: Environ. 64 (2006) 272–281.

[533] M. Kositzi, I. Poulios, S. Malato, J. Caceres, A. Campos, Water Res. 38 (2004)1147–1154.

[534] I. Oller,W. Gernjak, M.Maldonado, P. Fernández-Ibáñez, J. Blanco, J. Sánchez-Pérez, S. Malato, Environ. Chem. Lett. 3 (2005) 118–121.

[535] M.I. Maldonado, P.C. Passarinho, I. Oller, W. Gernjak, P. Fernández, J. Blanco,S. Malato, J. Photochem. Photobiol. A: Chem. 185 (2007) 354–363.

[536] C. Guillard, J. Disdier, C. Monnet, J. Dussaud, S. Malato, J. Blanco,M.I. Maldonado, J.-M. Herrmann, Appl. Catal. B: Environ. 46 (2003) 319–332.

[537] Y. Ohko, I. Ando, C. Niwa, T. Tatsuma, T. Yamamura, T. Nakashima, Y. Kubota,A. Fujishima, Environ. Sci. Technol. 35 (2001) 2365–2368.

[538] Y. Ohko, K.-i. Iuchi, C. Niwa, T. Tatsuma, T. Nakashima, T. Iguchi, Y. Kubota,A. Fujishima, Environ. Sci. Technol. 36 (2002) 4175–4181.

[539] T. Nakashima, Y. Ohko, D.A. Tryk, A. Fujishima, J. Photochem. Photobiol. A-Chem. 151 (2002) 207–212.

[540] T. Nakashima, Y. Ohko, Y. Kubota, A. Fujishima, J. Photochem. Photobiol. A:Chem. 160 (2003) 115–120.

[541] T. Saito, T. Iwase, J. Horie, T. Morioka, J. Photochem. Photobiol. B: Biol. 14(1992) 369–379.

[542] Z. Huang, P.-C. Maness, D.M. Blake, E.J. Wolfrum, S.L. Smolinski, W.A. Jacoby,J. Photochem. Photobiol. A: Chem. 130 (2000) 163–172.

[543] A.-G. Rincón, C. Pulgarin, Appl. Catal. B: Environ. 51 (2004) 283–302.[544] P. Fernández, J. Blanco, C. Sichel, S. Malato, Catal. Today 101 (2005) 345–352.[545] E.F. Duffy, et al., Solar Energy 77 (2004) 649–655.[546] M.I. Litter, Appl. Catal. B: Environ. 23 (1999) 89–114.[547] M.R. Prairie, L.R. Evens, B.M. Stange, S.L. Martinez, Environ. Sci. Technol. 27

(1993) 1776–1782.[548] M.A. Ferguson, M.R. Hoffmann, J.G. Hering, Environ. Sci. Technol. 39 (2005)

1880–1886.[549] A.-G. Rincón, C. Pulgarin, Solar Energy 77 (2004) 635–648.[550] P. Pichat, J. Disdier, C. Hoang-Van, D.Mas, G. Goutailler, C. Gaysse, Catal. Today

63 (2000) 363–369.[551] C.H. Ao, S.C. Lee, Appl. Catal. B: Environ. 44 (2003) 191–205.[552] C.H. Ao, S.C. Lee, J. Photochem. Photobiol. A: Chem. 161 (2004) 131–140.[553] J.-H. Kim, G. Seo, D.-L. Cho, B.-C. Choi, J.-B. Kim, H.-J. Park, M.-W. Kim,

S.-J. Song, G.-J. Kim, S. Kato, Catal. Today 111 (2006) 271–274.[554] S.A. Grinshpun, A. Adhikari, T. Honda, K.Y. Kim, M. Toivola, K.S. Ramchander-

Rao, T. Reponen, Environ. Sci. Technol. 41 (2007) 606–612.[555] T. Ibusuki, K. Takeuchi, J. Mol. Catal. 88 (1994) 93–102.[556] H. Ichiura, T. Kitaoka, H. Tanaka, Chemosphere 51 (2003) 855–860.[557] Italcementi, http://www.italcementigroup.com/ENG/Research+and+

Innovation/Innovative+Products/Innovative+Products.htm, 2008.[558] Y. Ohko, Y. Utsumi, C. Niwa, T. Tatsuma, K. Kobayakawa, Y. Satoh, Y. Kubota,

A. Fujishima, J. Biomed. Mater. Res. 58 (2001) 97–101.[559] Y. Sekiguchi, Y. Yao, Y. Ohko, K. Tanaka, T. Ishido, A. Fujishima, Y. Kubota,

Internat. J. Urology 14 (2007) 426–430.[560] A. Fujishima,D.A. Tryk, T.Watanabe, K. Hashimoto, Internat. Glass Rev. (1998)

114–116.[561] Y. Takata, S. Hidaka, J.M. Cao, T. Nakamura, H. Yamamoto, M. Masuda, T. Ito,

Energy 30 (2005) 209–220.

[562] M. Miyauchi, H. Tokudome, J. Mater. Chem. 17 (2007) 2095–2100.[563] Y. Takata, S. Hidaka, M. Masuda, T. Ito, Internat. J. Energy Res. 27 (2003)

111–119.[564] Y. Takata, S. Hidaka, J.M. Cao, K. Tanaka, M. Masuda, T. Ito, T. Watanabe,

M. Shimohigoshi, Thermal Science Eng. 8 (2000) 33–41.[565] H. Irie, K. Sunada, K. Hashimoto, Electrochemistry 72 (2004) 807–812.[566] Y. Ohko, S. Saitoh, T. Tatsuma, A. Fujishima, J. Electrochem. Soc. 148 (2001)

B24–B28.[567] T. Tatsuma, S. Saitoh, Y. Ohko, A. Fujishima, Chem. Mater. 13 (2001)

2838–2842.[568] Y. Ohko, S. Saitoh, T. Tatsuma, A. Fujishima, Electrochemistry 70 (2002)

460–462.[569] H. Park, K.Y. Kim, W. Choi, J. Phys. Chem. B 106 (2002) 4775–4781.[570] P. Ngaotrakanwiwat, T. Tatsuma, S. Saitoh, Y. Ohko, A. Fujishima, Phys. Chem.

Chem. Phys. 5 (2003) 3234–3237.[571] T. Tatsuma, S. Saitoh, P. Ngaotrakanwiwat, Y. Ohko, A. Fujishima, Langmuir

18 (2002) 7777–7779.[572] Y. Takahashi, P. Ngaotrakanwiwat, T. Tatsuma, Electrochim. Acta 49 (2004)

2025–2029.[573] P. Ngaotrakanwiwat, T. Tatsuma, J. Electroanalyt. Chem. 573 (2004)

263–269.[574] P. Ngaotrakanwiwat, S. Saitoh, Y. Ohko, T. Tatsuma, A. Fujishima, J.

Electrochem. Soc. 150 (2003) A1405–A1407.[575] T. Bessho, K. Inoue, K. Ishikawa, I. Koiwa, H. Honma, J. Surface Finishing Soc.

Japan. 57 (2006) 157.[576] M. Sugimoto, K. Tashiro, T. Bessho, I. Koiwa, H. Honma, J. Surface Finishing

Soc. Japan. 57 (2006) 162.[577] J.P. Lee, M.M. Sung, J. Am. Chem. Soc. 126 (2004) 28–29.[578] W. Kubo, T. Tatsuma, Appl. Surf. Sci. 243 (2005) 125–128.[579] H. Notsu,W. Kubo, I. Shitanda, T. Tatsuma, J. Mater. Chem. (2005) 1523–1527.[580] N. Blondiaux, S. Zurcher, M. Liley, N.D. Spencer, Langmuir 23 (2007)

3489–3494.[581] Z.-Z. Gu, A. Fujishima, O. Sato, Angew. Chem. Internat. Ed. 114 (2002)

2171–2174.[582] R.S. Dibbell, G.R. Soja, R.M. Hoth, D.F. Watson, Langmuir 23 (2007)

3432–3439.[583] K. Tadanaga, J. Morinaga, A. Matsuda, T. Minami, Chem. Mater. 12 (2000)

590–592.[584] X.T. Zhang, O. Sato, A. Fujishima, Langmuir 20 (2004) 6065–6067.[585] X. Zhang, M. Jin, Z. Liu, D.A. Tryk, S. Nishimoto, T. Murakami, A. Fujishima,

J. Phys. Chem. C 111 (2007) 14521–14529.[586] X. Zhang, H. Kono, Z. Liu, S. Nishimoto, D.A. Tryk, T. Murakami, H. Sakai,

M. Abe, A. Fujishima, Chem. Commun. (2007) 4949–4951.[587] K. Kobayashi, Y. Tomita, Y. Maeda, Japan. J. Appl. Phys. 44 (2005) 5829–5831.[588] Y. Ohko, T. Tatsuma, T. Fujii, K. Naoi, C. Niwa, Y. Kubota, A. Fujishima, Nature

Mater. 2 (2003) 29–31.[589] K. Kawahara, K. Suzuki, Y. Ohko, T. Tatsuma, Phys. Chem. Chem. Phys. 7 (2005)

3851–3855.[590] R. Gorges, S. Meyer, G. Kreisel, J. Photochem. Photobiol. A: Chem. 167 (2004)

95–99.[591] K. Naoi, Y. Ohko, T. Tatsuma, J. Am. Chem. Soc. 126 (2004) 3664–3668.[592] K. Naoi, Y. Ohko, T. Tatsuma, Chem. Commun. (2005) 1288–1290.[593] S. Teekateerawej, J. Nishino, Y. Nosaka, J. Photochem. Photobiol. A: Chem. 179

(2006) 263–268.[594] G. Takei, T. Kitamori, H.-B. Kim, Catal. Commun. 6 (2005) 357–360.[595] E.T. Castellana, S. Kataoka, F. Albertorio, Cremer, Anal. Chem. 78 (2006)

107–112.[596] G. Takei, M. Nonogi, A. Hibara, T. Kitamori, H.-B. Kim, Lab Chip 7 (2007)

596–602.[597] H. Nagai, T. Irie, J. Takahashi, S.-i. Wakida, Biosens. Bioelectr. 22 (2007)

1968–1973.[598] E. Baur, A. Rebmann, Helv. Chim. Acta 4 (1921) 256–262.[599] J.M.White, J. Szanyi, M.A. Henderson, J. Phys. Chem. B 108 (2004) 3592–3602.[600] J.M. White, M.A. Henderson, J. Phys. Chem. B 109 (2005) 12417–12430.[601] M.A. Henderson, J.M. White, H. Uetsuka, H. Onishi, J. Catal. 238 (2006)

153–164.[602] H. Uetsuka, H. Onishi,M.A. Henderson, J.M.White, J. Phys. Chem. B 108 (2004)

10621–10624.[603] H. Uetsuka, M.A. Henderson, A. Sasahara, H. Onishi, J. Phys. Chem. B 108

(2004) 13706–13710.[604] M.A. Henderson, J. Phys. Chem. B 108 (2004) 18932–18941.[605] M.A. Henderson, J. Phys. Chem. B 109 (2005) 12062–12070.[606] C.L. Perkins, M.A. Henderson, J. Phys. Chem. B 105 (2001) 3856–3863.[607] M.A. Henderson, J. Phys. Chem. B 101 (1997) 221–229.[608] M.A. Henderson, Surf. Sci. 319 (1994) 315–328.[609] M. Murai, Y. Tamaki, A. Furube, K. Hara, R. Katoh, Catal. Today 120 (2007)

214–219.[610] H. Tada, M. Tanaka, Langmuir 13 (1997) 360–364.[611] H. Wang, T. Wang, P. Xu, J. Mater. Sci.: Mater. Electr. 9 (1998) 327–330.[612] A. Fernández, G. Lassaletta, V.M. Jiménez, A. Justo, A.R. González-Elipe,

J.M. Herrmann, H. Tahiri, Y. Ait-Ichou, Appl. Catal. B: Environ. 7 (1995) 49–63.[613] S. Yanagida, A. Nakajima, Y. Kameshima, N. Yoshida, T. Watanabe, K. Okada,

Mater. Res. Bull. 40 (2005) 1335–1344.[614] Y. Gao, Y. Masuda, Z. Peng, T. Yonezawa, K. Koumoto, J. Mater. Chem. 13

(2003) 608–613.[615] A. Goossens, E.-L. Maloney, J. Schoonman, Chem. Vapor Deposition 4 (1998)

109–114.

http://www.italcementigroup.com/ENG/Research+and+Innovation/Innovative+Products/Innovative+Products.htm











[616] Z. Ding, X. Hu, P.L. Yue, G.Q. Lu, P.F. Greenfield, Catal. Today 68 (2001)173–182.

[617] S.A. O’Neill, I.P. Parkin, R.J.H. Clark, A. Mills, N. Elliott, J. Mater. Chem. 13(2003) 56–60.

[618] P. Zeman, S. Takabayashi, J. Vac. Sci. Technol. A 20 (2002) 388–393.[619] P.K. Song, M. Yamagishi, H. Odaka, Y. Shigesato, Japan. J. Appl. Phys. 42 (2003)

L1529–L1531.[620] M. Kitano, M. Takeuchi, M. Matsuoka, J.M. Thomas, M. Anpo, Chem. Lett. 34

(2005) 616–617.[621] P.K. Song, Y. Irie, Y. Shigesato, Thin Solid Films 496 (2006) 121–125.[622] M. Takeuchi, T. Yamasaki, K. Tsujimaru, M. Anpo, Chem. Lett. 35 (2006)

904–905.[623] M.-C. Yang, T.-S. Yang, M.-S. Wong, Thin Solid Films 469–470 (2004) 1–5.[624] T.-S. Yang, C.-B. Shiu, M.-S. Wong, Surf. Sci. 548 (2004) 75–82.[625] T. Miyata, S. Tsukada, T. Minami, Thin Solid Films 496 (2006) 136–140.[626] N. Berger-Keller, G. Bertrand, C. Filiatre, C. Meunier, C. Coddet, Surf. Coatings

Technol. 168 (2003) 281–290.

[627] C. Lee, H. Choi, C. Lee, H. Kim, Surf. Coatings Technol. 173 (2003) 192–200.[628] G.K. Mor, O.K. Varghese, M. Paulose, K. Shankar, C.A. Grimes, Solar Energy

Mater. Solar Cells 90 (2006) 2011–2075.[629] J.M. Macak, M. Zlamal, J. Krysa, P. Schmuki, Small 3 (2007) 300–304.[630] K. Takami, T. Adachi, A. Nakajima, K. Hashimoto, T. Watanabe, Surface

Coatings Internat. B: Coatings Trans. 85 (2002) 155–158.[631] K. Kobayakawa, C. Sato, Y. Sato, A. Fujishima, J. Photochem. Photobiol. A:

Chem. 118 (1998) 65–69.[632] M. Marezio, A. Gauzzi, F. Licci, E. Gilioli, Physica C: Superconductivity 338

(2000) 1–8.[633] T. Shibata, N. Sakai, K. Fukuda, Y. Ebina, T. Sasaki, Phys. Chem. Chem. Phys. 9

(2007) 2413–2420.[634] T. Sasaki, J. Ceram. Soc. Japan. 115 (2007) 9–16.[635] C. Morterra, J. Chem. Soc., Faraday Trans. 1 (84) (1988) 1617–1637.[636] Y. Suda, T. Morimoto, Langmuir 3 (1987) 786–788.[637] P. Jones, J.A. Hockey, Trans. Faraday Soc. 67 (1971) 2669–2678.

Documents

sdarticle