You are on page 1of 9

FULL PAPER

DOI: 10.1002/chem.201102807

Site-Selected Doping of Upconversion Luminescent Er3 + into SrTiO3 for


Visible-Light-Driven Photocatalytic H2 or O2 Evolution

Jinwen Shi,[a, b] Jinhua Ye,*[b] Lijing Ma,[a] Shuxin Ouyang,[b] Dengwei Jing,[a] and
Liejin Guo*[a]

Abstract: A series of upconversion lu- transfer from the high-energy excited tion in the presence of the correspond-
minescent erbium-doped SrTiO3 states of Er3 + with B-site occupancy to ing sacrificial reagents. The results gen-
(ABO3-type) photocatalysts with differ- the host SrTiO3 and thus enhanced the erally suggest that the introduction of
ent initial molar ratios of Sr/Ti have band-to-band transition of the host upconversion luminescent agents into
been prepared by a facile polymerized SrTiO3. Consequently, the erbium- host semiconductors is a promising ap-
complex method. Er3 + ions, which doped SrTiO3 species with B-site occu- proach to simultaneously harnessing
were gradually transferred from the A pancy showed higher photocatalytic ac- low-energy photons and maintaining
to the B site with increasing Sr/Ti, ena- tivity than those with A-site occupancy redox ability for photocatalytic H2 and
bled the absorption of visible light and for visible-light-driven H2 or O2 evolu- O2 evolution and that the site occupan-
the generation of high-energy excited cy of doped elements in ABO3-type
states populated by upconversion pro- perovskite oxides greatly determines
Keywords: energy conversion ·
cesses. The local internal fields arising the photocatalytic activity.
erbium · perovskite · photocatalysis ·
from the dipole moments of the dis-
photochemistry
torted BO6 octahedra promoted energy

Introduction cally restrained by wide bandgaps.[1c] Accordingly, metal cat-


ions with d10 (such as Cu + and Ag + ),[5] d10 s2 (such as Sn2 + ,
Photocatalytic water splitting under solar-light irradiation Pb2 + , and Bi3 + ),[5a, 6] or partially filled dn (such as Cr3 + , Rh3 +
has been acknowledged as a promising way to realize the , and Ni2 + )[7, 1l] electronic configurations have usually been
ideal conversion of solar energy into hydrogen since the dis- incorporated into the host materials of oxide semiconduc-
covery of the Fujishima–Honda effect in 1972.[1] Corre- tors (such as TiO2,[7a,b,e] SrTiO3,[5c, 7b–h,j,k] and NaTaO3[6e, 7m]) to
spondingly, the exploitation of such low-cost photocatalysts tailor the electronic structures by establishing new energy
with high stability and high efficiency under the irradiation bands or localized impurity levels between the bandgaps,
of visible light, which accounts for around 43 % of the in- and thus to extend the light-response range.[1c] However, the
coming solar energy,[1l] becomes the prerequisite but remains redox ability of the photogenerated charge carriers was in-
a big challenge.[1g, 2] Oxide semiconductors (such as TiO2,[1k] evitably reduced and therefore most photocatalysts pre-
SrTiO3,[3] and NaTaO3[4]) showed unparalleled stability and pared by this strategy failed in overall water splitting and
high activity for photocatalytic H2 and/or O2 evolution, espe- had the sole ability to extract either H2 or O2 from H2O in
cially for overall water splitting, under UV irradiation,[1k, 3, 4] solutions of the corresponding sacrificial reagents under visi-
but the ability to utilize visible light has often been intrinsi- ble-light irradiation.[5c,e, 7b,d–f,k] Consequently, it is highly desir-
able to mediate the incompatibility between the use of low-
[a] J. Shi, L. Ma, Dr. D. Jing, Prof. Dr. L. Guo
energy photons and the preservation of the driving force for
International Research Center for Renewable Energy redox reactions. Fortunately, upconversion luminescent
State Key Laboratory of Multiphase Flow agents (such as Eu3 + ,[8] Nd3 + ,[9] Ho3 + ,[10] Er3 + ,[11] and Tm3 +
in Power Engineering (MFPE) [12]
), in which a high-energy photon can be emitted by ab-
Xi’an Jiaotong University (XJTU)
sorbing two or more low-energy photons,[13] may open up
28 West Xianning Road, Xi’an, Shaanxi 710049 (P.R. China)
Fax: (+ 86) 29-8266-9033 a new avenue to resolve this issue. It has been reported that
E-mail: lj-guo@mail.xjtu.edu.cn several composite or doped photocatalysts have been
[b] J. Shi, Prof. Dr. J. Ye, Dr. S. Ouyang formed by coupling upconversion luminescent agents (such
Photocatalytic Materials Center (PCMC) as Er3 + [11d,f–i] and Tm3 + [12]) with host semiconductors (such
National Institute for Materials Science (NIMS) as TiO2,[11d,f–h, 12b–d] CdS,[12a] and Bi2WO6[11i]) and applied for
1-2-1 Sengen, Tsukuba, Ibaraki 305-0047 (Japan)
Fax: (+ 81) 29-859-2301
effective organic decomposition under the irradiation of
E-mail: jinhua.ye@nims.go.jp light beyond the absorption edge of host materials.[11d,f–i, 12]
Supporting information for this article is available on the WWW The host materials could absorb the high-energy photons,
under http://dx.doi.org/10.1002/chem.201102807. which were obtained from the upconversion luminescent

Chem. Eur. J. 2012, 18, 7543 – 7551  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 7543
agents under the excitation of low-energy photons that tuning the initial molar ratios of strontium and titanium,
could not be harvested by the host materials themselves due thereby aiming to improve the activity of photocatalysts for
to wide bandgaps, to enable the band-to-band transition. visible-light-driven H2 or O2 evolution and to reveal the re-
Meanwhile, the host materials maintained the redox ability lationship between site occupancy and activity.
of charge carriers in the valence band (VB) and conduction Herein, a series of upconversion luminescent erbium-
band (CB). Nevertheless, to the best of our knowledge, pho- doped SrTiO3 with the same erbium content, but different
tocatalysts based on this strategy for visible-light-driven H2 initial molar ratios of Sr/Ti (designated as SrxTiy–Er, Sr/Ti/
and/or O2 evolution have not yet been explored. Er = x:y:0.02) were prepared by means of a facile polymer-
In this study, upconversion luminescent erbium-doped ized complex (PC) method and applied to visible-light-
SrTiO3 was considered as a promising photocatalyst for visi- driven H2 or O2 evolution in solutions of the corresponding
ble-light-driven H2 and/or O2 evolution for the following sacrificial reagents. For comparison purposes, samples with-
reasons. On the one hand, early in 1976, SrTiO3 was proved out erbium (designated as SrxTiy) and Er2Ti2O7 were also
to be an excellent photocatalyst for overall water splitting prepared by varying the initial molar ratios of the metal pre-
under UV irradiation due to its high stability and suitable cursors. The Er2O3 and SrO used in this work were commer-
band-edge positions,[3a] and, since then, has been extensively cial materials. The physical mixture of Sr1.00Ti1.00 and
investigated by optimizing, for example, the reactant solu- Er2Ti2O7 with a molar ratio of 0.99/0.01 (i.e., Sr/Ti/Er =
tion,[3c] co-catalyst,[3b,d–f,h] preparation method,[3g] and de- 0.99:1.01:0.02) was denoted as Sr1.00Ti1.00 + Er2Ti2O7, and
ACHTUNGREfects.[3i] SrTiO3 was also regarded as an ideal host material the physical mixture of Sr1.00Ti1.00 and Er2O3 with a molar
for preparing upconversion luminescent materials (e.g., ratio of 0.99:0.01 (i.e., Sr/Ti/Er = 0.99:0.99:0.02) was denoted
erbium-doped SrTiO3[14]) because of a low maximum as Sr1.00Ti1.00 + Er2O3.
phonon energy (around 550 cm1),[14, 15] which can suppress
the nonradiative decay by a multiphonon relaxation process
to prolong the lifetimes of photogenerated high-energy ex- Results and Discussion
cited states and thus further enhance the upconversion pro-
cess.[16] On the other hand, Er3 + is the first reported upcon- The powder X-ray diffraction (PXRD) patterns employed
version luminescent ion[13] and still attracts considerable at- to determine the crystal structures and phase compositions
tention due to the rich 4f levels that facilitate 4f!4f intra- of samples are shown in Figure 1 and Figure S1 in the Sup-
band transitions to generate equally spaced, long-lived excit- porting Information. First, all the samples of SrxTiy (Fig-
ed states and then emit light in the UV, visible-light, and ure S1a in the Supporting Information) and SrxTiy–Er (Fig-
near-infrared (NIR) regions,[11h,i] and hence it has been ap- ure S1b in the Supporting Information) could be indexed to
plied to photocatalytic organic decomposition as described
above.[11d,f–i] Previous research confirmed that a portion of
photogenerated high-energy excited states of Er3 + could be
pumped by visible light of different wavelengths (such as
450,[11f,g] 486.5/488,[11d, 17] 542.4/548.8,[11c] 632.8,[11a,b] and
652.2 nm[11e]) to generate upconversion luminescence in the
UV region. Consequently, as for erbium-doped SrTiO3, the
energy stored in such photogenerated high-energy excited
states of Er3 + was no less than the bandgap of the host
SrTiO3 and therefore might transfer radiatively or nonradia-
tively from Er3 + to and excite the host SrTiO3,[11d,f,g,i, 18] in
which charge carriers were anticipated to be generated by
the band-to-band transition and finally reacted with H2O
and sacrificial reagents to produce H2 or O2, respectively. In
addition, earlier studies revealed that the ABO3-type perov-
skite oxide is highly susceptible to the substitution of differ-
ent metal cations at the A and/or B sites,[7g,i, 19] and that the
site occupancy of the metal cation dopant could be con-
trolled by adjusting the initial molar ratios of the A- and B-
site cation precursors if the ionic size of the dopant fell in
between those of the A- and B-site cations.[6e, 7g, 20] Such site
occupancy greatly determined the physicochemical proper-
ties and photocatalytic activities of the as-prepared materi-
als due to a different chemical environment between the A
and B sites. On account of the intermediate size of Er3 +
compared with Sr2 + and Ti4 + ,[21] selective site occupancy Figure 1. PXRD patterns of the strongest diffraction peaks of samples of
was expected to be achieved on erbium-doped SrTiO3 by a) SrxTiy and b) SrxTiy–Er.

7544 www.chemeurj.org  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2012, 18, 7543 – 7551
Photocatalytic H2 or O2 Evolution
FULL PAPER
pure tausonite SrTiO3 with cubic-phase perovskite structure
(JCPDS No. 00-035-0734). In contrast to SrxTiy–Er,
Sr1.00Ti1.00 + Er2Ti2O7 (Figure S1c in the Supporting Infor-
mation) with comparable erbium content shows the charac-
teristic peak of cubic-phase Er2Ti2O7 (JCPDS No. 00-054-
0181) in addition to those of main-phase SrTiO3, and
Sr1.00Ti1.00 + Er2O3 (Figure S1c in the Supporting Informa-
tion) with comparable erbium content exhibits the charac-
teristic peak of cubic-phase Er2O3 (JCPDS No. 00-043-1007)
in addition to those of main-phase SrTiO3. It could be de-
duced that the SrxTiy–Er samples are different to the physi-
cal mixtures of SrTiO3 and erbium oxides. Secondly, the
(110) peaks of all the SrxTiy samples (Figure 1a) appear in
the same position, whereas those of the SrxTiy–Er samples
(Figure 1b) gradually shift to lower angles with increasing
Sr/Ti. It was demonstrated that such a sequential peak shift
in SrxTiy–Er could not be induced only by the change in Sr/
Ti, but by the synergistic effect of the change in Sr/Ti and
the introduction of erbium. Thirdly, compared with that of
Sr1.00Ti1.00 with a Sr/Ti ratio of 1, the (110) peak of
Sr0.99Ti0.99–Er, also with a Sr/Ti ratio of 1, appears at
almost the same position, whereas that of Sr0.98Ti1.00–Er
with Sr/Ti less than 1 is shifted to a higher angle and those
of erbium-containing samples with Sr/Ti more than 1 (i.e.,
Sr1.00Ti0.98–Er, Sr1.02Ti0.98–Er, and Sr1.05Ti0.98–Er) are Figure 2. FTIR spectra of a) Sr0.98Ti1.00, Sr1.00Ti1.00, Sr1.00Ti0.98, and
shifted to lower angles to varying extents (Figure 1b). It is SrO and b) samples of SrxTiy–Er.
widely accepted that such peak shifts reflect changes in the
crystal lattices of host materials and can be caused by incor-
porating alien ions with sizes different to those of the host The FTIR spectra used to analyze the local structures of
ions.[7ll] Considering the sequence of Shannon effective ionic samples are shown in Figure 2 and Figure S2 in the Support-
radii (i.e., rACHTUNGRE(Ti4+) < rACHTUNGRE(Er3+) < rACHTUNGRE(Sr2+)),[21] the substitution of ing Information. The spectra of SrxTiy (Figure 2a) and
Er3 + for Sr2 + at the A site of SrTiO3 would result in the SrxTiy–Er (Figure 2b) with Sr/Ti not more than 1 are consis-
shrinkage of crystal lattices and thus shift the peak to tent with the FTIR spectra reported for SrTiO3.[22] Neverthe-
a higher angle, whereas the substitution of Er3 + for Ti4 + at less, for SrxTiy and SrxTiy–Er with Sr/Ti more than 1, two
the B site of SrTiO3 would cause the expansion of crystal new bands emerge at 1450 and 860 cm1, which gradually in-
lattices and hence the peak would shift to a lower angle. In crease with increasing Sr/Ti. These two bands are also de-
addition, on account of the counterbalance of the change in tected in the spectrum of SrO (Figure 2a), but not in those
crystal lattices, the simultaneous substitution of Er3 + for of Er2O3 and Er2Ti2O7 (Figure S2 in the Supporting Informa-
both Sr2 + and Ti4 + at the A and B sites of SrTiO3 would lead tion), and thus they can be ascribed to the vibration of
to little peak shift. Consequently, it can be inferred that the bonds in strontium oxides. It was also suggested that no
Er3 + ions in Sr0.98Ti1.00–Er are mainly located at the A strontium oxides exist in SrxTiy and SrxTiy–Er with Sr/Ti
site, and that those in Sr1.00Ti0.98–Er, Sr1.02Ti0.98–Er, and not more than 1 (i.e., Sr0.98Ti1.00, Sr1.00Ti1.00,
Sr1.05Ti0.98–Er are mainly located at the B site, and that Sr0.98Ti1.00–Er, and Sr0.99Ti0.99–Er), and that small
those in Sr0.99Ti0.99–Er are located at both the A and B amounts of strontium oxides are highly dispersed on the sur-
sites, thereby revealing the gradual transfer of Er3 + from the faces of samples in SrxTiy and SrxTiy–Er with Sr/Ti more
A to the B site as the ratio of Sr/Ti increases. In addition, than 1 (i.e., Sr1.00Ti0.98, Sr1.00Ti0.98–Er, Sr1.02Ti0.98–Er,
compared with Sr1.02Ti0.98–Er, Sr1.05Ti0.98–Er shows little and Sr1.05Ti0.98–Er) due to an excess of strontium relative
(110) peak shift, which indicates that almost all the Er3 + to titanium for the formation of stoichiometric SrTiO3 or of
ions in Sr1.02Ti0.98–Er and Sr1.05Ti0.98–Er are located at erbium-doped SrTiO3 with B-site occupancy. The absence of
the B site, whereas Sr1.00Ti0.98–Er exhibits a (110) peak the corresponding PXRD peaks confirmed the low content
shift to a higher angle, thereby demonstrating that some of and high dispersion of strontium oxides on the surfaces of
the Er3 + ions in Sr1.00Ti0.98–Er are still located at A site. SrxTiy and SrxTiy–Er with Sr/Ti more than 1. On account of
Based on the PXRD analysis, it can be concluded that the high flexibility of strontium entering or exiting the crys-
erbium-doped SrTiO3 was successfully prepared by a facile tal lattices, the compositions of erbium-doped SrTiO3 could
PC method and that selective site occupancy was achieved achieve the balance automatically by self-adjusting the con-
by varying the initial molar ratios of Sr/Ti. tent of strontium in the crystal lattices with the excess stron-
tium forming strontium oxides on the surfaces.

Chem. Eur. J. 2012, 18, 7543 – 7551  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 7545
J. Ye, L. Guo et al.

These spectra mainly exhibit six emission bands correspond-


ing to the radiative decay of the 4fn !4fn intraband transi-
tion from excited states of Er3 + .[14] The bands at around 520,
550, 655, 790, 850, and 980 nm are associated with the intra-
band transitions of 2H = !4I = , 4S = !4I = , 4F = !4I = , 4I = !
11
2
15
2
3
2
15
2
9
2
15
2
9
2
4
I = , 4S = !4I = , and 4I = !4I = , respectively. In addition, for
15
2
3
2
13
2
11
2
15
2

the SrxTiy–Er samples, the fine bands at around 550 nm


(Figure 3b) display subtle changes with increasing Sr/Ti,
which implies the variance of local structures and crystal
fields around Er3 + and of the physicochemical properties of
erbium-doped SrTiO3, and thus also confirmed the transfer
of Er3 + from the A to the B site.[13] In addition, the upcon-
version PL emission spectra of SrxTiy–Er recorded under
972.9 nm excitation, by taking Sr0.98Ti1.00–Er as an exam-
ple, verify the upconversion luminescent ability of erbium-
doped SrTiO3 (Figure S5b in the Supporting Information).
Similarly, the bands at around 520, 550, and 655 nm corre-
spond to the intraband transitions of 2H = !4I = , 4S = !4I = , 11
2
15
2
3
2
15
2

and 4F = !4I = , respectively.[14, 24]


9
2
15
2

The optical properties of the samples are characterized by


the UV/Vis spectra shown in Figure 4 and Figure S6 in the

Figure 3. Spectra recorded on Raman spectrometers of a) Sr1.00Ti1.00,


Sr1.00Ti1.00 + Er2Ti2O7, Sr1.00Ti1.00 + Er2O3, and Sr0.99Ti0.99–Er and
b) samples of SrxTiy–Er. The intensities of the signals are normalized ac-
cording to the strongest peak in individual spectra and the abscissae are
expressed as Raman shift and wavelength in (a) and (b), respectively.

The spectra of samples recorded on Raman spectrometer


are shown in Figure 3 and Figures S3 and S4 in the Support-
ing Information. All the SrxTiy samples show similar signals
(Figure S3 in the Supporting Information) and are consistent
with the Raman spectra reported for SrTiO3[23] but com-
pletely different to those of the SrxTiy–Er samples (Fig- Figure 4. UV/Vis spectra of Er2O3, Er2Ti2O7, Sr1.00Ti1.00, and
ure 3b). Moreover, it is surprising that the signals of both Sr0.99Ti0.99–Er.
Sr1.00Ti1.00 + Er2Ti2O7 and Sr1.00Ti1.00 + Er2O3 are not
similar to those of SrxTiy–Er samples but to those of SrxTiy
(Figure 3a). It was also confirmed that the SrxTiy–Er sam- Supporting Information. Sr1.00Ti1.00 (Figure 4) and all the
ples are not physical mixtures of SrTiO3 and erbium oxides SrxTiy–Er samples (Figure S6a in the Supporting Informa-
but SrTiO3 with erbium doping, which remarkably modified tion) show intense absorption in the UV region and a steep
the physicochemical properties of SrTiO3. In fact, according absorption edge at around 390 nm, which indicates that such
to the Raman spectra of Sr0.98Ti1.00–Er recorded by using absorption could not be ascribed to the transitions of local-
two lasers with different wavelengths as excitation sources ized impurity levels but to the band-to-band transitions of
(Figure S4 in the Supporting Information), both spectra the host SrTiO3,[3a, 7l] in which, with a bandgap of around
show the same shape but with a relative displacement along 3.2 eV, the valence band maximum (VBM) and conduction
the abscissa when the abscissa is expressed as Raman shift band minimum (CBM) are composed of oxygen 2p and tita-
(Figure S4a in the Supporting Information), whereas both nium 3d orbitals, respectively,[7i] and are little affected by the
spectra not only show the same shape but also appear at the orbitals of Er3 + .[25] Sr1.00Ti1.00 (Figure 4) exhibits no ab-
same position when the abscissa is expressed as wavelength sorption bands in the visible-light and NIR regions, in
(Figure S4b in the Supporting Information). Thus, the spec- which, in contrast, all the erbium-containing samples
tra of the SrxTiy–Er samples should be assigned as photolu- (Figure 4 and Figure S6b in the Supporting Information) ex-
minescence (PL) signals stemming from the luminescent hibit several sharp and isolated absorption bands, thereby
erbium doped in the host SrTiO3 and are in accordance with demonstrating the effective extension of the light-response
the PL emission spectra recorded on the fluorescence spec- range. The bands at around 450, 490, 520, 550, 655, and
trophotometer (Figure S5a in the Supporting Information). 790 nm were attributed to the intraband transitions of Er3 +

7546 www.chemeurj.org  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2012, 18, 7543 – 7551
Photocatalytic H2 or O2 Evolution
FULL PAPER
from the ground-state level (GSL) of 4I = to the excited- 15
2
phology and particle size (e.g., Sr0.98Ti1.00–Er forms aggre-
state levels (ESLs) of 4F = + 4F = , 4F = , 2H = , 4S = , 4F = , and 4I = ,
5
2
3
2
7
2
11
2
3
2
9
2
9
2
gates composed of irregular particles with a size of around
respectively,[24] and are consistent with the PL emission 100 nm, see Figure S7 in the Supporting Information) and
bands of the SrxTiy–Er samples. Nevertheless, the accurate comparable Brunauer–Emmett–Teller (BET) surface areas
band positions vary to some extent in different erbium-con- (e.g., Sr0.98Ti1.00–Er, 4.34 m2 g1; Sr1.02Ti0.98–Er,
taining samples (Figure 4 and Figure S6b in the Supporting 4.26 m2 g1), thereby eliminating the effect of textural struc-
Information), which indicates different chemical environ- tures on activity. By combining all the above analyses on the
ments around Er3 + . In particular, for the SrxTiy–Er samples crystal and textural structures, local structures, and optical
(Figure S6b in the Supporting Information), the bands at properties, it can be deduced that such a trend in activity es-
around 450, 490, and 550 nm gradually shift to higher wave- sentially originates from the transfer of Er3 + in SrTiO3 from
lengths with increasing Sr/Ti. By combining the PXRD and the A to the B site, which indicates that the substitution at
PL analyses, it can be deduced that such a band shift stems the B site is much more effective than that at the A site for
from the transfer of Er3 + from the A to the B site of visible-light-driven H2 or O2 evolution on erbium-doped
SrTiO3. In addition, bands corresponding to the interband SrTiO3. Accordingly, on account of the complete transfer of
transitions between the Er3 + levels and the VB or CB are Er3 + from the A to the B site, the highest activity for H2
not observed for all SrxTiy–Er samples.[26] evolution was achieved on Sr1.02Ti0.98–Er with a rate of
Figure 5 shows the visible-light-driven photocatalytic ac- 46.23 mmol h1 gcat1 (Figure 5 and Figure S8a and Tables S1–
tivity of samples for H2 (or O2) evolution by using Na2S/ S3 in the Supporting Information) and an apparent quantum
efficiency of 0.30 % (see the Supporting Information and
Table S2 therein), and the highest activity for O2 evolution
was also obtained on Sr1.02Ti0.98–Er with an initial rate of
44.23 mmol h1 gcat1 (Figure 5 and Figure S8c and Table S2 in
the Supporting Information) and an apparent quantum effi-
ciency of 0.40 % (see the Supporting Information and
Table S2 therein). The turnover number (TON), which is
generally employed to judge whether the reactions proceed-
ed photocatalytically,[1c] was calculated according to Equa-
tions (1) and (2).

Figure 5. Visible-light-driven photocatalytic activity for H2 and O2 evolu- Number of H2 evolved  2


TON ¼ ð1Þ
tion on different samples. The “Amount of O2 evolved” value corre- Number of Er atoms in photocatalyst
sponds to the amount of O2 evolved during the first 3 h of the photocata-
lytic reaction and gcat1 represents “per gram photocatalyst”. Number of O2 evolved  4
TON ¼ ð2Þ
Number of Er atoms in photocatalyst

Na2SO3 (or AgNO3) as the sacrificial reagent. No H2 (or O2) The TON for H2 (or O2) evolution on Sr1.02Ti0.98–Er
was detected when the reaction proceeded free of photoca- was 18.41 (or 5.03) after reaction for 29.73 h (or 23.35 h),
talyst or in the dark (e.g., Sr1.02Ti0.98–Er, see Table S1 in thereby confirming that the reaction proceeded photocata-
the Supporting Information) while keeping other conditions lytically. Sr1.02Ti0.98–Er also showed activity for H2 evolu-
the same. These control experiments guaranteed that neither tion under the irradiation of visible light with longer wave-
photolysis of an aqueous solution of Na2S/Na2SO3 (or lengths, longer even than 700 nm (Table S1 in the Support-
AgNO3) nor mechanocatalysis of metal oxides contributed ing Information). In addition, the lower activity on
to the evolution of H2 (or O2).[1c] Neither Sr1.00Ti1.00 nor Sr1.05Ti0.98–Er compared with Sr1.02Ti0.98–Er was as-
Er2O3 showed activity for H2 (or O2) evolution although cribed to the increasing amounts of strontium oxides, which,
SrTiO3 can split water into H2 and O2 stoichiometrically as proven by the FTIR analysis, are highly dispersed over
under UV irradiation[3] and only a small amount of H2 (or the surfaces of SrxTiy–Er with a ratio of Sr/Ti of more than
O2) was detected from Er2Ti2O7. It was thus shown that nei- 1, and, with increasing Sr/Ti, gradually cover the active sites
ther Er2O3 nor Er2Ti2O7 are competent for visible-light- of the SrxTiy–Er photocatalysts and lead to a deterioration
driven H2 (or O2) evolution although they could absorb visi- of the photocatalytic reaction. In addition, the PXRD analy-
ble light due to the intraband transitions of Er3 + . With the sis (Figure S1b,d in the Supporting Information) and the
SrxTiy–Er samples, no H2 was evolved with Sr0.98Ti1.00–Er, time courses of visible-light-driven H2 or O2 evolution (Fig-
but on increasing the ratio of Sr/Ti from 0.98/1.00 to 1.02/ ure S8 and Figure S14a in the Supporting Information) con-
0.98, H2 evolution gradually increased, and with a further in- firmed the good stability of the SrxTiy–Er samples in the
crease in Sr/Ti from 1.02/0.98 to 1.05/0.98, was then de- photocatalytic systems (see the Supporting Information). Fi-
pressed; the evolution of O2 coincidently exhibited the same nally, the H2 evolution on Sr1.02Ti0.98–Er under visible-
trend. All the SrxTiy–Er samples show similar surface mor- light irradiation was even higher than that on Sr1.00Ti1.00

Chem. Eur. J. 2012, 18, 7543 – 7551  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 7547
J. Ye, L. Guo et al.

under UV/Vis irradiation (see Table S3 and Figure S9 in the nonradiative recombination due to concentration quench-
Supporting Information). ing.[16] Consequently, the photoluminescence process, espe-
As a result, upconversion luminescent erbium-doped cially the upconversion luminescent process, is greatly en-
SrTiO3 samples with B-site occupancy were validated to be hanced by virtue of radiative recombination, as evidenced
novel, stable, and efficient photocatalysts for visible-light- by the significantly enhanced PL signals from the SrxTiy–Er
driven H2 or O2 evolution, and showed distinct advantages samples compared with Sr1.00Ti1.00 + Er2Ti2O7 and
over the commercially available visible-light-active photoca- Sr1.00Ti1.00 + Er2O3 (Figure 3). The photogenerated high-
talysts of WO3 and nitrogen-doped TiO2 as well as over the energy excited states during the upconversion process can
standard photocatalyst of Degussa P25 (see Figure S10 in be populated by different mechanisms, such as excited-state
the Supporting Information). It was found that the following absorption (ESA), energy transfer upconversion (ETU), co-
three factors, namely, the host SrTiO3 with the ability to operative upconversion, and the photon avalanche effect.[13]
split water into H2 and O2 stoichiometrically,[3] the upconver- As reported by Guo et al., ESA and ETU processes are the
sion luminescent Er3 + with the ability to harness visible possible upconversion mechanisms in erbium-doped
light, and the appropriate initial molar ratio of Sr/Ti with SrTiO3.[14] The energy stored in a portion of the photogener-
the ability to modulate the site occupancy of Er3 + in SrTiO3, ated high-energy excited states of Er3 + is no less than the
are indispensable for constructing such photocatalysts. bandgap of the host SrTiO3 and therefore could be trans-
Sr1.00Ti1.00 showed no activity due to a bandgap in ferred radiatively or nonradiatively from Er3 + to the host
SrTiO3 that is too wide for visible light to use.[5a–c, 7b–h,j,k,m] SrTiO3 to drive the band-to-band transition. That is, the host
Both Er2O3 and Er2Ti2O7 can absorb visible light due to the SrTiO3 could use visible light indirectly by virtue of the up-
intraband transitions of Er3 + . However, the highly localized conversion luminescent Er3 + to generate charge carriers
ESLs of Er3 + rather than dispersed energy bands restrain (i.e., holes (h + ) and electrons (e) in the VB and CB, re-
the mobility of the photogenerated excited states,[7l, 27] most spectively; Step II in Figure 6),[11d,f,g,i, 18] which finally react
of which would recombine radiatively or nonradiatively in with H2O and the sacrificial reagents, respectively, to pro-
the bulk and thus cannot separate and migrate to the sur- duce H2 or O2 (Step III in Figure 6). The ability to produce
face to take part in redox reactions, thereby resulting in in- both H2 and O2 on erbium-doped SrTiO3 testifies to the suc-
ferior activity of Er2O3 and Er2Ti2O7. Meanwhile, the high cessful preservation of the redox potentials of charge carri-
contents of Er3 + in Er2O3 and Er2Ti2O7 would lead to the re- ers, although even under UV/Vis irradiation, this material
combination of photogenerated excited states mainly by cannot simultaneously split water into H2 and O2 due to
way of a nonradiative path due to severe concentration other unoptimized conditions, such as the reactant solu-
quenching.[16] The main processes in the photocatalytic reac- tion[3c] and co-catalyst[1m, 3b,d–f,h] (see the details following Fig-
tions of the SrxTiy–Er samples are schematically illustrated ure S9 in the Supporting Information).
in Figure 6. The doping of Er3 + in the host SrTiO3 enables However, as described above, not all the SrxTiy–Er sam-
ples were efficient for visible-light-driven photocatalytic H2
or O2 evolution and erbium-doped SrTiO3 with B-site occu-
pancy was more active than those with A-site occupancy.
This result is different to those obtained with chromium-
doped SrTiO3 reported by Wang et al. and bismuth-doped
NaTaO3 reported by Kanhere et al.[6e, 7g] Chromium-doped
SrTiO3 with A-site occupancy was more active than that
with B-site occupancy for visible-light-driven H2 evolution
due to the preservation of the valence state of Cr3 + and
thus the avoidance of Cr6 + , which usually behaves as a re-
combination center of photogenerated charge carriers and
Figure 6. Schematic illustration of the main processes of the photocatalyt-
therefore reduces the photocatalytic activity.[7g,l] Bismuth-
ic reaction on SrxTiy–Er samples. Step I: Generation of high-energy ex- doped NaTaO3 with equal A- and B-site occupancy was
cited states in Er3 + by the absorption of low-energy photons. Step II: more active than those with either A- or B-site occupancy
Band-to-band excitation of SrTiO3 by energy transferred from high- for visible-light-driven organic decomposition because of the
energy excited states in Er3 + . Step III: Redox reactions (see the detailed
optimized electronic structures for bandgap narrowing and
reaction mechanisms in the Supporting Information).
thus visible-light absorption but not due to the valence state
of the dopant (Bi3 + ).[6e] With regard to the SrxTiy–Er sam-
visible-light absorption and the excitation of Er3 + (Step I in ples, on the one hand, the site occupancy would not alter
Figure 6). Although the photogenerated excited states ini- the valence state of Er3 + in view of its high stability[28] and it
tially remain on the highly localized ESLs of Er3 + and thus has a limited effect on the electronic structures of erbium-
possess a mobility that is too low to participate in the redox doped SrTiO3 in view of the analogous optical properties
reaction directly, the low content and homogeneous distribu- apart from the subtle change of the 4fn !4fn intraband tran-
tion of Er3 + in the crystal lattice of the host SrTiO3 prevent sitions of Er3 + , as evidenced by the PL and UV/Vis analyses.
cluster formation of Er3 + and hence greatly suppress the On the other hand, the energy transfer from Er3 + to the

7548 www.chemeurj.org  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2012, 18, 7543 – 7551
Photocatalytic H2 or O2 Evolution
FULL PAPER
host SrTiO3 is the key step in generating charge carriers in amount of H2 was detected during the visible-light-driven
the VB and CB of SrTiO3 as discussed above, and thus de- photocatalytic reaction (Figure S14d in the Supporting Infor-
termines the photocatalytic activity for H2 or O2 evolution. mation). Detailed investigations on SrxTiy–Ln samples are
On the basis of the discrepant activity of the SrxTiy–Er sam- in progress and will be reported soon. Finally, it is generally
ples, it was speculated that the site occupancy of Er3 + plays suggested that site-selected doping of appropriate upconver-
a pivotal role in determining the energy transfer efficiency. sion luminescent agents into host semiconductors is an effec-
Goldschmidt proposed the tolerance factor (t) for ABO3- tive strategy for the design of photocatalysts for visible-
type perovskites.[29] The t value of an ideal perovskite should light-driven H2 or O2 evolution.
be equal to 1, deviation from which could be used to predict
the degree of distortion of perovskites with different A and/
or B ions.[30] Substitution of Er3 + for Sr2 + and/or Ti4 + in Conclusion
SrTiO3 resulted in t values less than 1 (see the Supporting
Information), which indicates the distortion of erbium- A series of upconversion luminescent erbium-doped SrTiO3
doped SrTiO3, in which the BO and AO bonds are under photocatalysts with different initial molar ratios of Sr/Ti
compression and tension, respectively, the BOB bond have been prepared by a facile PC method. Er3 + ions were
angle is bent away from 1808, and the number and length of gradually transferred from the A to the B site with increas-
the shortest AO bonds were reduced.[19a] Correspondingly, ing ratio of Sr/Ti. The light-response range was effectively
with the B ions displaced from the centers of gravity of the extended from the UV (for SrTiO3) to the visible-light (for
six surrounding oxygen ions, the BO6 octahedra are also dis- erbium-doped SrTiO3) region due to intraband transition
torted, thereby generating dipole moments.[31] It is widely absorption of Er3 + and the high-energy excited states of
accepted that the local internal fields arising from the dipole Er3 + were populated by upconversion processes. Compared
moments of BO6 octahedra can significantly promote photo- with those with A-site occupancy, Er3 + ions with B-site oc-
excitation[31] and the separation and migration of photogen- cupancy are located in distorted BO6 octahedra with dipole
erated charge carriers.[1b, 25, 31, 32] Er3 + ions with B-site occu- moments and thus subject to much stronger local internal
pancy are located in such distorted BO6 octahedra, whereas fields, which promotes energy transfer from the high-energy
Er3 + ions with A-site occupancy are not. It can therefore be excited states of Er3 + to the host SrTiO3 and facilitates the
reasonably deduced that, compared with those with A-site generation of charge carriers in the VB and CB of the host
occupancy, Er3 + ions with B-site occupancy are subject to SrTiO3. Consequently, erbium-doped SrTiO3 with B-site oc-
much stronger local internal fields, which promote the cupancy instead of A-site occupancy are validated as being
energy transfer from Er3 + to the host SrTiO3 and hence im- novel, stable, and efficient photocatalysts for visible-light-
proves the photocatalytic activity. Also, in contrast to those driven H2 or O2 evolution. In addition, upconversion lumi-
with A-site occupancy, erbium-doped SrTiO3 with B-site oc- nescent holmium- and neodymium-doped SrTiO3 are also
cupancy has a shorter interatomic distance of ErO, which active in visible-light-driven photocatalytic H2 evolution. It
also facilitates the energy transfer from Er3 + to the host is generally suggested that the site-selected doping of upcon-
SrTiO3 through the oxygen atom.[33] In addition, with regard version luminescent agents into host semiconductors is an
to Er2Ti2O7, concentration quenching of the photogenerated effective strategy in the design of photocatalysts for visible-
excited states and no site occupancy of Er3 + in the BO6 oc- light-driven H2 or O2 evolution.
tahedra collectively led to its low activity.
Upconversion luminescent lanthanide (Ln) elements
other than erbium were also employed to prepare SrxTiy–
Experimental Section
Ln (Ln = Er, Ho, Nd, or Eu) samples (see the Supporting In-
formation). The preliminary results for Sr1.00Ti0.98–Ln in- Preparation of samples: All reagents were of analytical purity and used
dicated that the crystal structures and local structures, which as received from Wako Pure Chemical Industries, Ltd, except titanium
were determined by PXRD (Figure S11 in the Supporting isopropoxide (Kojundo Chemical Laboratory Co., Ltd, 5N). Samples
Information) and FTIR (Figure S12 in the Supporting Infor- were prepared by a PC method.[34] Samples were typically prepared as
follows: Initially, anhydrous citric acid (CA, 0.0400 mol), titanium iso-
mation) analyses, respectively, are similar and that the opti- propoxide (TTIP, 0.0100 mol), SrCl2·6H2O (0.0098 mol), Er-
cal properties, characterized by the UV/Vis spectra (Fig- ACHTUNGRE(CH3COO)3·4H2O (0.0002 mol), and dehydrated ethylene glycol (EG,
ure S13 in the Supporting Information), are different due to 0.1600 mol), in a molar ratio of 4.00:1.00:0.98:0.02:16.00, were dissolved
the variance of the 4f!4f intraband transitions in different in dehydrated methanol (50.0 mL) in sequence. The resulting clear solu-
tion was then heated, with stirring, at 130 8C for 24 h in a fume hood to
lanthanide elements with diverse 4f levels. Both
promote polymerization and remove excess solvent. The glassy resin ob-
Sr1.00Ti0.98–Ho and Sr1.00Ti0.98–Nd, like Sr1.00Ti0.98–Er, tained was then pyrolyzed in a burnout furnace at 200, 300, and 500 8C
display several sharp and isolated absorption bands in the for 6, 6, and 3 h, respectively. The resulting powder was finally calcined
visible-light region and, correspondingly, exhibit photocata- at 1100 8C for 6 h in air to facilitate the crystallization. The as-prepared
lytic activity for visible-light-driven H2 evolution (Fig- sample was named as Sr0.98Ti1.00–Er. Accordingly, a series of samples
were prepared by varying the initial molar ratios of Sr/Ti (i.e., x/y =
ure S14a–c in the Supporting Information). However, for 0.98:1.00, 0.99:0.99, 1.00:0.98, 1.02:0.98, and 1.05:0.98) while keeping the
Sr1.00Ti0.98–Eu, no obvious absorption bands were evident erbium content the same and thus the samples were denoted as SrxTiy–
in the visible-light region and, accordingly, only a trace Er.

Chem. Eur. J. 2012, 18, 7543 – 7551  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 7549
J. Ye, L. Guo et al.

Characterization of materials: The crystal structures and phase composi- Basic Research Program of China (973, No. 2009CB220000). This work
tions of the samples were determined by powder X-ray diffraction analy- was also sponsored by a scholarship from the State Scholarship Fund of
sis (RINT-2000, RIGAKU, Japan) with a q/2q Bragg–Brentano geometry. the China Scholarship Council (CSC) and by an award of an Internation-
Diffraction data in the 2q range from 10.0 to 80.08 were collected (40 kV, al Joint Graduate School (IJGS) Fellowship under the “NIMS-XJTU
30 mA; CuKa radiation, l = 1.541841 ; nickel filter) with a scan step size Joint Graduate School Program” from NIMS. J.S. thanks P. Guo from
of 0.028 and a counting time of 0.3 s. The divergence, scattering, and re- XJTU for help with PL analysis, Z. Mei, N. Zhang, and P. Li from NIMS
ceiving slits were 28, 18, and 0.30 mm, respectively. The BET surface for performing supplementary experiments, and especially the anony-
areas of the samples were determined by N2 adsorption at 77 K on a sur- mous reviewers for valuable comments on an earlier draft of this paper.
face area analyzer (Gemini-2360, Micromeritics, USA) and an accelerat-
ed surface area and porosimetry analyzer (ASAP 2020, Micromeritics,
USA) after degassing the samples at 300 8C for 5 h. The surface morphol-
[1] a) F. E. Osterloh, Chem. Mater. 2008, 20, 35 – 54; b) Y. Inoue, Energy
ogies and particle sizes of the samples were observed on a field-emission
Environ. Sci. 2009, 2, 364 – 386; c) A. Kudo, Y. Miseki, Chem. Soc.
scanning electron microscope (FESEM: JSM-6700F, JEOL, Japan) with
Rev. 2009, 38, 253 – 278; d) R. Abe, J. Photochem. Photobiol. C 2010,
an accelerating voltage of 5 kV. The optical properties of the samples
11, 179 – 209; e) X. B. Chen, S. H. Shen, L. J. Guo, S. S. Mao, Chem.
were measured by using a UV/Vis spectrophotometer (UV-2500, Shimad-
Rev. 2010, 110, 6503 – 6570; f) M. Kitano, M. Hara, J. Mater. Chem.
zu, Japan) in diffuse reflectance (DR) mode with BaSO4 as reference.
2010, 20, 627 – 641; g) K. Maeda, K. Domen, J. Phys. Chem. Lett.
The local structures of the samples were investigated by FTIR spectros-
2010, 1, 2655 – 2661; h) A. Kudo, MRS Bull. 2011, 36, 32 – 38; i) F. E.
copy (IRPrestige-21, Shimadzu, Japan) with a diffuse reflectance accesso-
Osterloh, B. A. Parkinson, MRS Bull. 2011, 36, 17 – 22; j) S. H. Shen,
ry (DRS-8000A). Specimens were prepared by well dispersing the sam-
J. W. Shi, P. H. Guo, L. J. Guo, Int. J. Nanotechnol. 2011, 8, 523 – 591;
ples (5.0 wt %) into KBr powder and then pressing them into pellets. The
k) A. Fujishima, K. Honda, Nature 1972, 238, 37 – 38; l) Z. G. Zou,
FTIR spectra were collected between 1800 and 400 cm1 with 50 scans
J. H. Ye, K. Sayama, H. Arakawa, Nature 2001, 414, 625 – 627; m) K.
and a resolution of 4 cm1. The photoluminescence spectra of the samples
Maeda, K. Teramura, D. L. Lu, T. Takata, N. Saito, Y. Inoue, K.
were recorded on a PTI QuantaMaster 40 steady-state fluorescence spec-
Domen, Nature 2006, 440, 295 – 295; n) K. Maeda, M. Higashi, D. L.
trophotometer with both excitation and emission slits set to 8 nm. Two
Lu, R. Abe, K. Domen, J. Am. Chem. Soc. 2010, 132, 5858 – 5868;
laser micro-Raman spectrometers (NRS-1000, JASCO, Japan; LabRAM
o) A. Iwase, Y. H. Ng, Y. Ishiguro, A. Kudo, R. Amal, J. Am. Chem.
HR 800, Horiba/Jobin Yvon, France) with backscattering configuration
Soc. 2011, 133, 11054 – 11057.
were employed to study both the local structures and photoluminescence
[2] N. S. Lewis, D. G. Nocera, Proc. Natl. Acad. Sci. USA 2006, 103,
properties of the samples. Spectra recorded on Raman spectrometer
15729 – 15735.
from the NRS-1000 (dispersive) spectrometer were recorded with an ex-
[3] a) M. S. Wrighton, A. B. Ellis, P. T. Wolczanski, D. L. Morse, H. B.
position time of 0.1 s and 10 scans by using a frequency-doubled, diode-
Abrahamson, D. S. Ginley, J. Am. Chem. Soc. 1976, 98, 2774 – 2779;
pumped Nd:YAG laser (1 mW, 532 nm) as the excitation source and
b) K. Domen, S. Naito, M. Soma, T. Onishi, K. Tamaru, J. Chem.
spectra from the LabRAM HR 800 (Fourier transform) spectrometer
Soc. Chem. Commun. 1980, 543 – 544; c) F. T. Wagner, G. A. Somor-
were recorded with an exposition time of 10 s and one scan using an Ar +
jai, Nature 1980, 285, 559 – 560; d) K. Domen, S. Naito, T. Onishi, K.
laser (20 mW, 514.532 nm) as the excitation source.
Tamaru, J. Phys. Chem. 1982, 86, 3657 – 3661; e) K. Domen, A.
Evaluation of photocatalytic activity: The photocatalytic activities of the Kudo, T. Onishi, J. Catal. 1986, 102, 92 – 98; f) K. Domen, A. Kudo,
samples were evaluated by photocatalytic H2 (or O2) evolution under T. Onishi, N. Kosugi, H. Kuroda, J. Phys. Chem. 1986, 90, 292 – 295;
visible-light irradiation. The photocatalytic reaction was performed in g) A. Kudo, A. Tanaka, K. Domen, T. Onishi, J. Catal. 1988, 111,
a 432 mL (or 453 mL for O2 evolution) Pyrex glass cell, which was con- 296 – 301; h) S. Ikeda, K. Hirao, S. Ishino, M. Matsumura, B. Ohtani,
nected to a closed gas circulation system and had a flat round side- Catal. Today 2006, 117, 343 – 349; i) T. Takata, K. Domen, J. Phys.
window for external light incidence. A 300 W xenon arc lamp Chem. C 2009, 113, 19386 – 19388.
(CERMAX LX-300, ILC Technology, USA) with a UV cut-off filter (L- [4] H. Kato, K. Asakura, A. Kudo, J. Am. Chem. Soc. 2003, 125, 3082 –
42, Hoya, Japan; l > 420 nm) was used as the light source. As shown in 3089.
Table S2 and Figure S15 in the Supporting Information, the average in- [5] a) H. Irie, Y. Maruyama, K. Hashimoto, J. Phys. Chem. C 2007, 111,
tensity of incident light was measured with a spectroradiometer (USR- 1847 – 1852; b) D. F. Wang, T. Kako, J. H. Ye, J. Am. Chem. Soc.
40, Ushio, Japan) and was equal to 412.01 mW cm2 (or 603.30 mW cm2 2008, 130, 2724 – 2725; c) D. F. Wang, T. Kako, J. H. Ye, J. Phys.
for O2 evolution), and the irradiated area was 12.57 cm2. Photocatalyst Chem. C 2009, 113, 3785 – 3792; d) O. Palasyuk, A. Palasyuk, P. A.
(0.30 g) was added to a 270 mL aqueous solution of 0.35 mol L1 Na2S Maggard, Inorg. Chem. 2010, 49, 10571 – 10578; e) Z. G. Yi, J. H. Ye,
and 0.25 mol L1 Na2SO3 (or of 0.185 mol L1 AgNO3 for O2 evolution) in N. Kikugawa, T. Kako, S. X. Ouyang, H. Stuart-Williams, H. Yang,
the cell and a suspension was formed by using a magnetic stirrer with J. Y. Cao, W. J. Luo, Z. S. Li, Y. Liu, R. L. Withers, Nat. Mater. 2010,
a constant rotational velocity. For the loading of 1.0 wt % (or 0.0 wt % 9, 559 – 564; f) Y. P. Bi, S. X. Ouyang, N. Umezawa, J. Y. Cao, J. H.
for O2 evolution) of the platinum co-catalyst onto the surface of the as- Ye, J. Am. Chem. Soc. 2011, 133, 6490 – 6492; g) S. X. Ouyang, J. H.
prepared photocatalysts by an in situ photodeposition method, an appro- Ye, J. Am. Chem. Soc. 2011, 133, 7757 – 7763; h) J. W. Shi, J. H. Ye,
priate amount of an aqueous solution of H2PtCl6·6H2O was also added. Q. Y. Li, Z. H. Zhou, H. Tong, G. C. Xi, L. J. Guo, Chem. Eur. J.
Prior to light irradiation, the system was evacuated below 2.0 kPa (or 2012, 18, 3157 – 3162.
1.0 kPa for O2 evolution) to eliminate air and then back-filled with [6] a) H. G. Kim, D. W. Hwang, J. S. Lee, J. Am. Chem. Soc. 2004, 126,
a small amount of argon gas beneficial for circulation. The temperature 8912 – 8913; b) J. W. Tang, Z. G. Zou, J. H. Ye, Angew. Chem. 2004,
of the system during the photocatalytic reaction was maintained at 116, 4563 – 4566; Angew. Chem. Int. Ed. 2004, 43, 4463 – 4466; c) Y.
around 20 8C by the thermostatic circulating of water. The evolved gases Hosogi, Y. Shimodaira, H. Kato, H. Kobayashi, A. Kudo, Chem.
were detected in situ by using an online gas chromatograph (GC-8AIT, Mater. 2008, 20, 1299 – 1307; d) Z. Y. Zhang, Q. P. Lin, S. T. Zheng,
Shimadzu, Japan), which was connected to the system and equipped with X. H. Bu, P. Y. Feng, Chem. Commun. 2011, 47, 3918 – 3920; e) P. D.
a thermal conductivity detector (TCD) and an MS-5A column with Kanhere, J. W. Zheng, Z. Chen, J. Phys. Chem. C 2011, 115, 11 846 –
argon as carrier gas. 11 853.
[7] a) E. Borgarello, J. Kiwi, M. Gratzel, E. Pelizzetti, M. Visca, J. Am.
Chem. Soc. 1982, 104, 2996 – 3002; b) H. Kato, A. Kudo, J. Phys.
Chem. B 2002, 106, 5029 – 5034; c) T. Ishii, H. Kato, A. Kudo, J. Pho-
Acknowledgements tochem. Photobiol. A 2004, 163, 181 – 186; d) R. Konta, T. Ishii, H.
Kato, A. Kudo, J. Phys. Chem. B 2004, 108, 8992 – 8995; e) R. Niish-
This work was supported by the National Natural Science Foundation of iro, H. Kato, A. Kudo, Phys. Chem. Chem. Phys. 2005, 7, 2241 –
China (NSFC, Nos. 50821064, 51121092, and 20906074) and the National 2245; f) J. W. Liu, G. Chen, Z. H. Li, Z. G. Zhang, J. Solid State

7550 www.chemeurj.org  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2012, 18, 7543 – 7551
Photocatalytic H2 or O2 Evolution
FULL PAPER
Chem. 2006, 179, 3704 – 3708; g) D. F. Wang, J. H. Ye, T. Kako, T. [16] A. Patra, C. S. Friend, R. Kapoor, P. N. Prasad, J. Phys. Chem. B
Kimura, J. Phys. Chem. B 2006, 110, 15824 – 15830; h) S. W. Bae, 2002, 106, 1909 – 1912.
P. H. Borse, J. S. Lee, Appl. Phys. Lett. 2008, 92, 104107; i) W. Wei, [17] a) H. Xu, Z. Dai, Z. Jiang, Eur. Phys. J. D 2001, 17, 79 – 83; b) N. N.
Y. Dai, H. Jin, B. B. Huang, J. Phys. D 2009, 42, 055401; j) K. Iwashi- Zu, H. G. Yang, Z. W. Dai, Phys. B 2008, 403, 174 – 177.
na, A. Kudo, J. Am. Chem. Soc. 2011, 133, 13272 – 13275; k) H. Yu, [18] a) J. Wang, R. H. Li, Z. H. Zhang, W. Sun, R. Xu, Y. P. Xie, Z. Q.
S. X. Ouyang, S. C. Yan, Z. S. Li, T. Yu, Z. G. Zou, J. Mater. Chem. Xing, X. D. Zhang, Appl. Catal. A 2008, 334, 227 – 233; b) J. Wang,
2011, 21, 11347 – 11351; l) J. W. Shi, J. H. Ye, Z. H. Zhou, M. T. Li, Y. P. Xie, Z. H. Zhang, J. Li, X. Chen, L. Q. Zhang, R. Xu, X. D.
L. J. Guo, Chem. Eur. J. 2011, 17, 7858 – 7867; m) X. Zhou, J. Y. Shi, Zhang, Sol. Energy Mater. Sol. Cells 2009, 93, 355 – 361; c) R. Xu, J.
C. Li, J. Phys. Chem. C 2011, 115, 8305 – 8311. Li, J. Wang, X. F. Wang, B. Liu, B. X. Wang, X. Y. Luan, X. D.
[8] J. Silver, M. I. Martinez-Rubio, T. G. Ireland, G. R. Fern, R. With- Zhang, Sol. Energy Mater. Sol. Cells 2010, 94, 1157 – 1165.
nall, J. Phys. Chem. B 2001, 105, 9107 – 9112. [19] a) J. B. Goodenough, J. S. Zhou, Chem. Mater. 1998, 10, 2980 – 2993;
[9] a) R. M. Macfarlane, F. Tong, A. J. Silversmith, W. Lenth, Appl. b) M. A. Pena, J. L. G. Fierro, Chem. Rev. 2001, 101, 1981 – 2017.
Phys. Lett. 1988, 52, 1300 – 1302; b) X. Zhang, E. Daran, C. Serrano, [20] M. H. Lin, H. Y. Lu, Mater. Sci. Eng. A 2002, 335, 101 – 108.
F. Lahoz, J. Lumin. 2000, 87 – 9, 1011 – 1013; c) J. Fernndez, M. [21] R. D. Shannon, Acta Crystallogr. Sect. A 1976, 32, 751 – 767.
Sanz, A. Mendioroz, R. Balda, J. P. Chaminade, J. Ravez, L. M. [22] J. T. Last, Phys. Rev. 1957, 105, 1740 – 1750.
Lacha, M. Voda, M. A. Arriandiaga, J. Alloys Compd. 2001, 323, [23] F. A. Rabuffetti, H. S. Kim, J. A. Enterkin, Y. M. Wang, C. H.
267 – 272. Lanier, L. D. Marks, K. R. Poeppelmeier, P. C. Stair, Chem. Mater.
[10] J. Silver, E. Barrett, P. J. Marsh, R. Withnall, J. Phys. Chem. B 2003, 2008, 20, 5628 – 5635.
107, 9236 – 9242. [24] H. Lin, G. Meredith, S. B. Jiang, X. Peng, T. Luo, N. Peyghambarian,
[11] a) J. Silver, M. I. Martinez-Rubio, T. G. Ireland, G. R. Fern, R. With- E. Y. B. Pun, J. Appl. Phys. 2003, 93, 186 – 191.
nall, J. Phys. Chem. B 2001, 105, 948 – 953; b) J. Silver, M. I. Marti- [25] Z. Y. Zhao, Q. J. Liu, J. Phys. D 2008, 41, 085417.
nez-Rubio, T. G. Ireland, R. Withnall, J. Phys. Chem. B 2001, 105, [26] A. W. Xu, Y. Gao, H. Q. Liu, J. Catal. 2002, 207, 151 – 157.
7200 – 7204; c) H. L. Xu, Z. K. Jiang, Chem. Phys. 2003, 287, 155 – [27] D. F. Wang, J. H. Ye, H. Kitazawa, T. Kimura, J. Phys. Chem. C
159; d) J. Wang, G. Zhang, Z. H. Zhang, X. D. Zhang, G. Zhao, F. Y. 2007, 111, 12848 – 12854.
Wen, Z. J. Pan, Y. Li, P. Zhang, P. L. Kang, Water Res. 2006, 40, [28] E. G. Reut, A. I. Ryskin, Phys. Status Solidi A 1973, 17, 47 – 57.
2143 – 2150; e) H. G. Yang, Z. W. Dai, Z. W. Sun, J. Lumin. 2007, [29] V. M. Goldschmidt, Naturwissenschaften 1926, 14, 477 – 485.
124, 207 – 212; f) G. F. Feng, S. W. Liu, Z. L. Xiu, Y. Zhang, J. X. Yu, [30] P. M. Woodward, Acta Crystallogr. Sect. B 1997, 53, 44 – 46.
Y. G. Chen, P. Wang, X. J. Yu, J. Phys. Chem. C 2008, 112, 13692 – [31] H. Nishiyama, H. Kobayashi, Y. Inoue, Chemsuschem 2011, 4, 208 –
13699; g) Y. H. Ling, W. F. Jiang, X. M. Wu, X. D. Bai, J. Nanosci. 215.
Nanotechnol. 2009, 9, 714 – 717; h) W. L. Wang, Q. K. Shang, W. [32] a) M. Kohno, S. Ogura, K. Sato, Y. Inoue, Chem. Phys. Lett. 1997,
Zheng, H. Yu, X. J. Feng, Z. D. Wang, Y. B. Zhang, G. Q. Li, J. 267, 72 – 76; b) S. Ogura, M. Kohno, K. Sato, Y. Inoue, Phys. Chem.
Phys. Chem. C 2010, 114, 13 663 – 13 669; i) Z. J. Zhang, W. Z. Wang, Chem. Phys. 1999, 1, 179 – 183; c) J. Sato, H. Kobayashi, Y. Inoue, J.
W. Z. Yin, M. Shang, L. Wang, S. M. Sun, Appl. Catal. B 2010, 101, Phys. Chem. B 2003, 107, 7970 – 7975; d) H. Kadowaki, J. Sato, H.
68 – 73. Kobayashi, N. Saito, H. Nishiyama, Y. Simodaira, Y. Inoue, J. Phys.
[12] a) C. H. Li, F. Wang, J. A. Zhu, J. C. Yu, Appl. Catal. B 2010, 100, Chem. B 2005, 109, 22995 – 23000; e) J. H. Wang, Z. G. Zou, J. H.
433 – 439; b) W. P. Qin, D. S. Zhang, D. Zhao, L. L. Wang, K. Z. Ye, J. Phys. Chem. Solids 2005, 66, 349 – 355.
Zheng, Chem. Commun. 2010, 46, 2304 – 2306; c) T. G. Li, S. W. Liu, [33] a) P. Boutinaud, E. Pinel, M. Oubaha, R. Mahiou, E. Cavalli, M.
H. P. Zhang, E. H. Wang, L. J. Song, P. Wang, J. Mater. Sci. 2011, 46, Bettinelli, Opt. Mater. 2006, 28, 9 – 13; b) Y. Inaguma, T. Tsuchiya, T.
2882 – 2886; d) Z. X. Li, F. B. Shi, T. Zhang, H. S. Wu, L. D. Sun, Katsumata, J. Solid State Chem. 2007, 180, 1678 – 1685.
C. H. Yan, Chem. Commun. 2011, 47, 8109 – 8111. [34] M. Kakihana, M. M. Milanova, M. Arima, T. Okubo, M. Yashima,
[13] F. Auzel, Chem. Rev. 2004, 104, 139 – 173. M. Yoshimura, J. Am. Ceram. Soc. 1996, 79, 1673 – 1676.
[14] H. Guo, N. Dong, M. Yin, W. P. Zhang, L. R. Lou, S. D. Xia, J.
Alloys Compd. 2006, 415, 280 – 283.
[15] a) A. Patra, C. S. Friend, R. Kapoor, P. N. Prasad, Chem. Mater. Received: September 8, 2011
2003, 15, 3650 – 3655; b) E. Pinel, P. Boutinaud, R. Mahiou, J. Alloys Revised: February 3, 2012
Compd. 2004, 380, 225 – 229. Published online: April 24, 2012

Chem. Eur. J. 2012, 18, 7543 – 7551  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 7551

You might also like