Self-Doped Ti3+Enhanced Photocatalyst for Hydrogen Production under
Visible Light
Fan Zuo,Le Wang,Tao Wu,Zhenyu Zhang,Dan Borchardt,and Pingyun Feng* Department of Chemistry,Uni V ersity of California,Ri V erside,California92521
Received May5,2010;E-mail:pingyun.feng@ucr.edu
Abstract:Through a facile one-step combustion method,partially reduced TiO2has been synthesized.Electron paramagnetic resonance(EPR)spectra confirm the presence of Ti3+in the bulk of an as-prepared sample.The UV-vis spectra show that the Ti3+here extends the photoresponse of TiO2from the UV to the visible light region,which leads to high visible-light photocatalytic activity for the generation of hydrogen gas from water.It is worth noting that the Ti3+sites in the sample are highly stable in air or water under irradiation and the photocatalyst can be repeatedly used without degradation in the activity.
Driven by increasing energy needs,decreasing fossil fuel resources, and environmental concerns of nuclear energy,the search for clean and renewable energy is attracting massive research interest.Utilizati
on of solar energy to produce hydrogen gas from water has long been considered the ultimate solution.Since the discovery in1971that TiO2 could act as a photochemical water-splitting catalyst,1over100 photocatalysts have been reported.2Because of its abundance,non-toxicity,and stability,TiO2has been extensively studied.However, for practical applications,pure TiO2is not a good candidate,because it is only active under ultraviolet(UV)irradiation in order to overcome the band gap of3.2eV for anatase phase.Therefore,band gap engineering is required if we want to use TiO2as a water-splitting catalyst under visible light irradiation.Initially,cations such as Al, Nd,Sb,Ag,Ru,V,Cr,Mn,and Fe were used as dopants to introduce states into the TiO2band gap.3However,problems such as thermal instability,increased carrier recombination centers,and the need for an expensive ion-implantation facility pose significant limitations for this strategy.4It was once proposed that doping nitrogen into TiO2is a better choice compared to doping with cations or other anions.4 However,later studies both theoretically and experimentally have raised questions about this N-doping strategy and its suitability as the most efficient method.5Furthermore,the reported activity for the photore-duction of water to hydrogen is quite low.6
Reduced TiO2(TiO2-x),which contains the Ti3+or oxygen vacancy, has been demonstrated to exhibit visible light absorption.7It was believed that the introduced localized oxygen vacancy states with energ
ies0.75to1.18eV below the conduction band minimum of TiO2 are lower than the redox potential for hydrogen evolution,which,in combination with the low electron mobility in the bulk region due to this localization,makes the photocatalytic activity of the reduced TiO2 negligible.8However,theoretical calculations show that a high vacancy concentration could induce a vacancy band of electronic states just below the conduction band.3The relevant experiments also prove the improved activity of reduced TiO2under visible light.Therefore,these results demonstrate that it is possible to fabricate visible-light responsive TiO2by introducing Ti3+.
The reported methods to produce TiO2-x include heating TiO2under vacuum or reducing ,H2),chemical vapor deposition,and high energy particle(laser,electron,or Ar+)bombardment.5For practical application,these strategies have a number of limitations such as multiple steps,harsh synthesis conditions,or expensive facilities. Furthermore,the surface oxygen defects are usually not stable enough in air as the Ti3+is easily oxidized and is even susceptible to oxidation by dissolved oxygen in water.9,10Therefore,developing a simple and economic strategy to synthesize a stable reduced TiO2photocatalyst is still a great challenge,which may be one reason why very limited studies have been reported for TiO2-x photocatalyic activity,especially photocatalytic water splitting.
Here we report a one-step method to synthesize reduced TiO2, which exhibits extremely high stability and is active for photo-catalytic hydrogen production from water.
By combustion of an ethanol solution(10.0g,99.5%ethanol and2.5g,37.1%hydrochloric acid)of titanium(IV)isopropoxide (2.00g,98+%)and2-ethylimidazole(1.80g,98%)at500°C in air and annealing for5h,blue powders(sample)are obtained. During the combustion,the imidazole will react with oxygen and form CO,CO2,NO,NO2,etc.The Ti(IV)could be reduced to Ti(III) by the reducing gas(CO and NO).
The powder X-ray diffraction analysis(Figure S1)shows that the as-produced sample is a mixture of anatase phase and rutile phase TiO2.
To test for the presence of Ti3+,low temperature electron para-magnetic resonance(EPR)spectra were recorded(Figure1A).The as-synthesized sample gave rise to a very strong EPR signal,while no signal was seen for the commercial Degussa P25(a mixture of anatase and rutile TiO2,Figure S2).Anisotropic powder pattern g-values of g x)g y)1.975and g z)1.944were obtained from a simulation that yielded a near perfectfit to the data.The observed g-values are characteristics of a paramagnetic Ti3+center in a distorted rhombic oxygen ligandfield.11The EPR data also indicate that there is no Ti3+ present on the surface of the sample.It is believed that surface Ti3+ would tend to adsorb atmospheric O2,which would be reduced to O2
-Figure1.(A)Experimental(solid line,measured under75K)and simulated (dashed line)EPR spectra for sample.(B)UV-visible diffuse reflectance spectra for commercial anatase TiO2(solid line)and sample(dash
line).
Published on Web08/05/2010
10.1021/ja103843d 2010American Chemical Society 118569J.AM.CHEM.SOC.2010,132,11856–11857
and shows an EPR signal at g≈2.02.12The absence of such a peak in Figure1A would indicate that only the rhombic Ti3+is present in the bulk,which is a key factor in the observed excellent stability of our sample.Furthermore,P25shows no EPR signal,meaning neither pure anatase nor pure rutile can account for the EPR peaks observed in the sample.From this information we conclude that Ti3+is pres
ent in the sample and that only rhombic Ti3+exists in the bulk.Surface analysis of the sample using X-ray photoelectron spectroscopy(XPS) shows no Ti3+peaks and further confirms the conclusion that Ti3+ exists in the bulk.
Figure1B shows the UV-visible absorption spectra for the sample and the commercial anatase.The spectrum of the sample shifts to a longer wavelength revealing a decrease in the band gap. Meanwhile,the absorbance in the visible range is enhanced compared to the stoichiometric anatase.This phenomenon is consistent with the assumption that an electronic band is located just below the conduction band of pure TiO2.
A theoretical simulation using PWscf package13was executed to support the existence of a Ti3+induced electronic band and to further
understand its likely influence on the band structure of oxygen-deficient TiO2.The calculations,based on a plane-wave pseudopotential density functional theory(DFT)approach,were performed on1×1×2and 2×2×1anatase supercell with one O atom removed from each system in order to simulate different concentrations of Ti3+.The plane-wave basis set with an energy cutoff of30Ryd was satisfactory for an ultrasoft pseudopotential with PBE(Perdew-Burke-Ernzerhof) exchange correlations
to capture the properties of anatase phase. Brillouin-zone integration was computed with k points in a Monkhorst-Pack(10,10,5)grid.We see a miniband rising up closely below the conducting band minimum(Figure S4).It is found that the width of the band is related to the concentration of the Ti3+or oxygen vacancy, since the width increases as the concentration of oxygen vacancy increased from1per32to1per16oxygen atoms.Very similar results have also been reported by Figueras.3The above calculations proved that the Ti3+inside the bulk TiO2is responsible for the band gap narrowing.Furthermore,the presence of the vacancy band has been reported as an extra benefit for light absorption.The high concentration of oxygen vacancy could break the selection rule for indirect transitions, resulting in an enhanced absorption for photon energy below the direct band gap,3which has been observed in our UV-visible spectra. The photocatalytic activity of the sample for water reduction was studied using the system supplied from Trustteck Co.,Inc.After loading 0.300g of sample with1%Pt(0.003g),the photocatalyst was placed into a120mL25%methanol(as a sacrificial agent)aqueous solution in a closed-gas circulation system.A Xe lamp(300W)with a400 nm cut-onfilter was used to ensure that only visible light(>400nm) illuminated the photocatalyst.Figure2shows a typical time course of H2evolution.This photocatalytic reaction exhibits a stable H2release rate of∼15µmol/h/0.300g.Even after illumination for200h,the activity is still maintained with no noticeable decrease observed, demonstrating the excellent stability of the sample.To understand the solar energy
conversion efficiency of the sample,the average external quantum efficiency(EQE)in the range of400nm-455nm was measured and found to be0.79%.We further measured the EQE with a420nm band-passfilter,which gave an EQE of0.35%,consistent with the above result.The commercial anatase TiO2has also been studied for comparison.Although it exhibits high activity under UV light,no apparent H2peak appears under visible light(>400nm) illumination for the anatase TiO2,providing strong evidence for extending the photocatalytic activity to the visible light range through our strategy.
To exclude the possible influence of the nonmetal dopants such as nitrogen,we replace the2-ethylimidazole with urea and keep other experimental parameters unchanged.The elemental analyses (Table S1)prove the presence of the N in the sample from urea. However,this sample shows almost no H2production activity under visible light illumination(<0.1µmol/h/0.300g).Also,the reported C-doped TiO2visible light water-splitting reaction requires a photoeletrochemical reaction system and voltage bias,14which are not necessary for our Ti3+sample.Therefore,the visible-light photocatalytic activity of our sample is not due to C-or N-doping. In conclusion,we have developed a simple one-step method to synthesize Ti3+-doped TiO2.The as-prepared reduced TiO2exhibits high stability in air and water with light irradiation.Experimental data show good conversion efficiency in the visible light region (>400nm).Both theoretical calculations and experimental results support that it is the introduced Ti3+th
at accounts for the extension of the photocatalytic activity from the UV to the visible light region. The present study demonstrates a simple and economical method for narrowing the band gap and for the development of a highly active photocatalyst under visible light irradiation. Acknowledgment.We are grateful for support of this work by the NSF(DMR-0907175).
Supporting Information Available:The synthetic procedure, facilities information,XRD patterns,more EPR spectra,and elemental analysis.This material is available free of charge via the Internet at
References
(1)Fujishima,A.;Honda,K.Bull.Chem.Soc.Jpn.1971,44,1148.
(2)Osterloh,F.E.Chem.Mater.2008,20,35.
(3)Justicia,I.;Ordejon,P.;Canto,G.;Mozos,J.L.;Fraxedes,J.;Battiston,
G.A.;Gerbasi,R.;Figueras,A.Ad V.Mater.2002,14,1399.
reported(4)Asahi,R.;Morikawa,T.;Ohwaki,T.;Aoki,K.;Taga,Y.Science2001,
293,269.
(5)Thompson,T.L.;Yates,J.T.Chem.Re V.2006,106,4428.
(6)Kim,H.G.;Hwang,D.W.;Lee,J.S.J.Am.Chem.Soc.2004,126,8912.
(7)Sasikala,R.;Shirole,A.;Sudarsan,V.;Sakuntala,T.;Sudakar,C.;Naik,
R.;Bharadwaj,S.R.Int.J.Hydrogen Energy2009,34,3621.
(8)Cronemeyer,D.C.Phys.Re V.1959,113,1222.
(9)Teleki,A.;Pratsinis,S.E.Phys.Chem.Chem.Phys.2009,11,3742.
(10)Komaguchi,K.;Maruoka,T.;Nakano,H.;Imae,I.;Ooyama,Y.;Harima,
Y.J.Phys.Chem.C2010,114,124.
(11)Khomenko,V.M.;Langer,K.;Rager,H.;Fett,A.Phys.Chem.Miner.
1998,25,338.
(12)Anpo,M.;Che,M.;Fubini,B.;Garrone,E.;Giamello,E.;Paganini,M.C.
Top.Catal.1999,8,189.
(13)Baroni,S.;Dal Corso,  A.;de Gironcoli,S.;Giannozzi,P.
(14)Khan,S.U.M.;Al-Shahry,M.;Ingler,W.B.Science2002,297,2243.
JA103843D
Figure2.Time course of evolved H2under visible light(>400nm) irradiation.(a)Reaction for4h;(b)evacuate and continue reaction for another4h;(c)illuminate for200h,evacuate system,and continue reaction for4h.(b)Sample;(1)Anatase.
J.AM.CHEM.SOC.9VOL.132,NO.34,201011857
C O M M U N I C A T I O N S