Oxygen photoevolution on a tantalum oxynitride photocatalyst under visible-light irradiation: How does water photooxidation proceed on a metal-oxynitride surface?

Article abstract of DOI:10.1021/jp0501289

The mechanism of water photooxidation (oxygen photoevolution) on a TaON photocatalyst was studied on the basis of our previous studies on the mechanism of this reaction on TiO₂ and N-doped TiO₂. We have confirmed that photocatalytic O₂ evolution occurs from an aqueous TaON suspension in the presence of Fe³⁺, as reported. In-situ MIR-IR experiments have indicated that the TaON surface is slightly oxidized under visible-light irradiation, indicating that the oxygen photoevolution on TaON actually occurs on a thin Ta-oxide overlayer. The in-situ MIR-IR experiments have also shown that a certain surface peroxo species, tentatively assigned to adsorbed HOOH, is formed as an intermediate of the O₂ photoevolution reaction. Studies on the effect of addition of reductants to the electrolyte on the IPCE have shown that photogenerated holes at the TaON surface cannot oxidize reductants such as SCN^-, Br^-, methanol, ethanol, 2-propanol, and acetic acid, though they can oxidize H₂O into O ₂. Detailed considerations of these results have strongly suggested that the water photooxidation reaction on TaON proceeds by our recently proposed new mechanism, that is, the reaction is initiated by a nucleophilic attack of a water molecule (Lewis base) on a surface-trapped hole (Lewis acid). ? 2005 American Chemical Society.

Full text of DOI:10.1021/jp0501289

8920
J. Phys. Chem. B 2005, 109, 8920-8927
Oxygen Photoevolution on a Tantalum Oxynitride Photocatalyst under Visible-Light
Irradiation: How Does Water Photooxidation Proceed on a Metal-Oxynitride Surface?
Ryuhei Nakamura, Tomoaki Tanaka, and Yoshihiro Nakato*
DiVision of Chemistry, Graduate School of Engineering Science, Osaka UniVersity, Toyonaka,
Osaka 560-8531, Japan, and Core Research for EVolutional Science and Technology (CREST),
Japan Science and Technology Agency (JST)
ReceiVed: January 8, 2005; In Final Form: February 23, 2005
The mechanism of water photooxidation (oxygen photoevolution) on a TaON photocatalyst was studied on
the basis of our previous studies on the mechanism of this reaction on TiO2 and N-doped TiO2. We have
confirmed that photocatalytic O2 evolution occurs from an aqueous TaON suspension in the presence of
Fe3+, as reported. In-situ MIR-IR experiments have indicated that the TaON surface is slightly oxidized
under visible-light irradiation, indicating that the oxygen photoevolution on TaON actually occurs on a thin
Ta-oxide overlayer. The in-situ MIR-IR experiments have also shown that a certain surface peroxo species,
tentatively assigned to adsorbed HOOH, is formed as an intermediate of the O2 photoevolution reaction.
Studies on the effect of addition of reductants to the electrolyte on the IPCE have shown that photogenerated
holes at the TaON surface cannot oxidize reductants such as SCN-, Br-, methanol, ethanol, 2-propanol, and
acetic acid, though they can oxidize H2O into O2. Detailed considerations of these results have strongly
suggested that the water photooxidation reaction on TaON proceeds by our recently proposed new mechanism,
that is, the reaction is initiated by a nucleophilic attack of a water molecule (Lewis base) on a surface-trapped
hole (Lewis acid).
Introduction
for the water photooxidation (oxygen photoevolution) in the
presence of a sacrificial oxidizing reagent such as Ag+. They
also reported15-18 that tantalum nitride (Ta3N5) exhibited the
high photocatalytic activity for the water photooxidation as well,
though its valence band was mainly composed of N2p orbitals
and located at 1.6 V versus NHE.
Photoinduced oxidation reactions of water and organic
compounds at the surface of nanocrystalline metal-oxide
photocatalysts such as TiO2 have been attracting keen attention
from the points of view of solar energy conversion (water
splitting)1,2 as well as environmental cleaning (photocatalytic
decomposition of waste materials and harmful compounds).3
Such metal oxides have high oxidation power under illumination
and are chemically stable, nontoxic, and relatively inexpensive.
Unfortunately, however, most of the metal oxides have too large
band gaps to efficiently absorb the solar light. This is because
the valence band of the metal oxides mainly consists of O2p
orbitals with deep energies of ca. 3 eV versus NHE.
Recent studies have shown that the shortcoming of the metal
oxides can be overcome by introducing midgap levels above
the top of the O2p valence band through doping with other
elements such as nitrogen,4-7 sulfur,8,9 carbon,10-12 and so forth.
For example, the nitrogen doping of TiO2 (anatase) produced
an N-induced midgap level ca. 0.75 eV above the top of the
O2p valence band,7 resulting in the extension of the photoactive
region up to 550 nm. The carbon doping of TiO2 is also reported
to extend the photoactive region for the water photooxidation
up to 535 nm, thus resulting in a large increase in the water
splitting efficiency under white-light illumination.10
The reported studies clearly show that the introduction of an
(N-induced) midgap level or a new valance band (composed of
hybridized N2p and O2p orbitals) above the O2p valence band
is effective to extend the photoactive region to the visible light.
However, problems will arise on the stability and efficiency
for such photocatalysts, because it is expected that metal-N
bonds in N-doped metal oxides or metal oxynitrides are weak
compared with metal-O bonds in metal oxides and that
photogenerated holes are trapped at higher N2p-induced levels
at the surface, resulting in a large decrease in the oxidation
power compared with the O2p valence-band holes. In addition,
the N-doping will introduce more or less distortion in metal-
oxide crystals, leading to their destabilization. It is thus quite
important to clarify how the water photooxidation proceeds at
the metal-oxynitride surface, why photogenerated holes with
a character of N2p holes can oxidize water, and why the metal
oxynitride is stable under the water photooxidation reaction.
In the present work, we have studied the mechanism of the
water photooxidation (oxygen photoevolution) at the surface
of TaON on the basis of our previous studies on the mechanisms
of this reaction at the surface of TiO219-21 and N-doped TiO2.7
The present results have strongly demonstrated that the water
photooxidation on TaON is initiated by our recently proposed
new mechanism of a nucleophilic attack of a water molecule
on a surface hole (a Lewis acid-base mechanism) and not by
the conventional mechanism of oxidation of surface OH group
by the hole (an electron-transfer type mechanism). The nucleo-
Another promising candidate of visible-light responsive
photocatalysts is metal oxynitrides, such as TaON, which have
the valence bands composed of hybridized N2p and O2p
orbitals. Domen et al. reported13-16 that TaON absorbed the
visible light up to ca. 530 nm, with the valence band at 2.2 V
versus NHE, and exhibited a high quantum efficiency of 34%
* Author to whom correspondence should be addressed. E-mail:
nakato@chem.es.osaka-u.ac.jp.
10.1021/jp0501289 CCC: $30.25 © 2005 American Chemical Society
Published on Web 03/30/2005

Water Photooxidation on a TaON Photocatalyst
J. Phys. Chem. B, Vol. 109, No. 18, 2005 8921
philic-attack mechanism has explained why the photogenerated
holes with a character of N2p holes can oxidize water.
Experimental Section
TaON powder was prepared by nitridation of commercial
Ta2O5 powder (Wako, purity: 99.8%) according to the literature:
15
Namely, Ta2O5 powder was initially heated in an alumina
tube reactor under a flow of N2 gas. After the temperature rose
to 800 °C, the N2 gas was replaced by an NH3 gas (purity:
99.999%) and the powder was kept at 800 °C under the NH3
flow of a flow rate of 10 mL/min for 10 h. Then, the NH3 gas
was replaced by a N2 gas again and the powder was cooled to
room temperature over a period of time of 2-3 h. As a reference
sample, Ta3N5 powder was also prepared by similar thermal
nitridation of Ta2O5 powder under a flow of NH3 gas at 800 °C
for 10 h, though in this case the flow rate was increased to
1000 mL/min.
Experiments of photocatalytic reactions were carried out in
a Pyrex reaction vessel attached to a closed gas circulation
system, with which a gas chromatograph (Shimadzu, GC14-B)
was directly connected. An evacuation pump was also connected
with this line via a cold trap. The photocatalytic oxidation of
water into O2 was performed in 200 mL of a 2 mM FeCl3
aqueous solution with 200 mg of TaON powder. The air in the
reaction vessel and the gas circulation system was removed by
repeated cycles of evacuation and introduction of Ar gas, until
the air (N2 and O2) became undetectable by the gas chromato-
graph. After 150 Torr of Ar gas was introduced in the reaction
system, the sample solution was irradiated from the horizontal
side of the vessel with a 300-W Xe lamp (Ushio) as the light
source through a UV cutoff filter (Irie Seisakusho, L-42,
transparent in λ g 420 nm), together with a water filter (5-cm
long) and an IR-cut filter (Irie Seisakusho, IRA-25S) to avoid
solution heating. The evolved gas (O2 and N2) was detected by
the gas chromatograph with a TCD detector.
Particulate TaON film electrodes for photocurrent measure-
ments were prepared by adopting the low-temperature annealing
method reported for TiO2 films,22 to avoid oxidation at the
TaON surface at elevated temperatures. Namely, TaON powder
(1 g) was suspended into pure water (5 mL) by stirring. The
resultant colloidal solution of TaON was then spin-coated at
2000 rpm onto transparent conductive oxide (F-doped SnO2 or
FTO, 20 Ω/square) films, on which an ethanol solution of 50
mM KOH had, in advance, been applied with a spin coater at
1000 rpm to get strong adhesion between TaON and FTO under
low-temperature annealing. The TaON-coated glass was then
dried in air and heated at 150 °C for 15 min with an electric
furnace. The above TaON coating was repeated twice, and
finally the TaON film was heated at 150 °C for 30 min. A
copper wire was attached on an edge of the F-SnO2 film with
silver paste, and the whole part, except a 1.0 × 1.0 cm2 TaON-
film area, was covered with epoxy resin for insulating.
Photocurrents for TaON film electrodes were measured as a
function of wavelength (λ), using a 350-W xenon lamp (Ushio)
as the light source together with a monochromator (Bunkoh-
Keiki Co., M10-T) with a bandwidth of 10 nm. The incident
photon to current efficiency at λ, IPCE (λ), was calculated from
the photocurrent by an equation
Figure 1. XRD patterns of (a) TaON, (b) Ta2O5, and (c) Ta3N5 powder.
Pt plate was used as the counter electrode, and a Ag/AgCl/sat.
KCl was used as the reference electrode. The electrolyte
solutions under experiments were bubbled with a nitrogen gas
to remove dissolved O2.
In-situ multiple internal reflection FTIR (MIR-IR) absorption
spectra were measured by the same experimental setup as used
in a previous paper.20 Ten microliters of aqueous TaON slurry
(10 mg/mL) was spread with a plastic pipet on a multiple
internal reflection element (prism) made of a diamond single
crystal and was dried in air. Spectra were recorded at the
resolution of 4 cm-1 and the accumulation of 400 scans, using
an FTIR spectrometer (Bio-Rad FTS 575C) with a deuterated
triglycine sulfate (DTGS) detector. The IR light was reflected
about nine times at the diamond/TaON solution interface. The
light source for photocatalytic reactions was a 200-W Hg-Xe
lamp (Hypercure 200 UV, Yamashita Denso). A UV cut filter
(λ g 420 nm) and an IR-cut filter were used.
Results
1. Photocatalytic Experiments. Figure 1 shows XRD
patterns for TaON and Ta3N5 powder prepared in the present
work, together with that for Ta2O5 for reference. The XRD
pattern for TaON contained no peak assigned to Ta2O5 and
Ta3N5, implying that the prepared TaON sample contained no
Ta2O5 nor Ta3N5. XPS analysis of the N1s region of TaON
showed the strong peak at 397 eV, which was reported14 to be
assigned to the Ta-N bond. The surface atomic ratio of nitrogen
to oxygen (RN/O) for TaON was estimated from the relative peak
areas of N1s (397 eV, Ta-N bond) and O1s (531 eV, Ta-O
bond) corrected for the sensitivity factor. The RN/O for the TaON
powder was 0.7 and a little smaller than unity. After annealing
at a high temperature of 400 °C in air for 2 h, the amount of
nitrogen further decreased and RN/O reached 0.4.
Figure 2 shows the time course of photocatalytic O2 evolution
from TaON (200 mg) suspended in 200 mL of a 2 mM Fe3+
aqueous solution (pH 2.3) under visible-light (λ g 420 nm)
irradiation. O2 gas was evolved, together with a trace amount
of N2, under visible-light irradiation, confirming that TaON
powder could oxidize water into oxygen, as reported by Domen
et al.13,15 The rate of O2 evolution decreased with the reaction
time owing to a decrease in the Fe3+ concentration, as is
understood from Figure 2 that the amount of evolved O2
approaches the total amount of O2 (100 µmol) evolved by
IPCE (λ) ) 1240 × j(λ)/λI0(λ) × 100
(1)
where λ is the wavelength of incident light in a unit of nm, j(λ)
is the photocurrent density in mA cm-2 under illumination at
λ, and I0(λ) is the incident light intensity in mW cm-2 at λ. A

8922 J. Phys. Chem. B, Vol. 109, No. 18, 2005
Nakamura et al.
Figure 2. Photocatalytic O2 and N2 evolution from an aqueous
suspension of TaON powder with 2 mM FeCl3 (pH 2.3) under visible-
light (λ g 420 nm) irradiation. The curve marked by open triangles is
for the O2 evolution when methanol (0.5 wt %) was added to the above
suspension.
Figure 4. Effect of addition of 6 M MeOH, 0.5 mM I-, 0.5 mM SCN-,
and 0.5 mM Br- to the electrolyte on the IPCE vs λ for a TaON thin
film electrode in 0.1 M HClO4. The curve marked by closed squares
indicates the IPCE vs λ in 0.1 M HClO4 with no added reductant. The
electrode potential was 0.7 V vs Ag/AgCl.
as an electron acceptor strongly depended on the pH of the
suspension, the O2 evolution largely deceasing with the decrease
in pH. However, in the present work, the IPCE for the TaON
film in an alkaline solution of pH 12.7 was nearly the same as
that in an acidic solution of pH 1.1, both being measured at a
potential of 0.7 V versus Ag/AgCl.
Figure 4 shows the effect of addition of a reductant such as
I-, SCN-, Br-, and methanol to the electrolyte on the IPCE
versus λ for a freshly prepared TaON film electrode. The IPCE
was increased largely by addition of I-, whereas no increase in
the IPCE was observed for the addition of SCN-, Br-, and
methanol. We also observed no increase in the IPCE by the
addition of ethanol, 1-propanol, and acetic acid (though not
shown in the figure). In general, when photogenerated holes
react with such reactants, they produce much less surface
reaction intermediates than when the holes react with water.7,23
The surface reaction intermediates act as effective surface carrier
recombination centers especially for nanosized particulate TaON
electrodes with no band bending. Therefore, when photogener-
ated holes react with a reactant, a large increase in the IPCE
should be observed compared with the case of the water
photooxidation owing to much less surface carrier recombina-
tion. The difference in the enhancement of the IPCE in Figure
4 can thus be regarded as reflecting the difference in the
reactivity of reductants with photogenerated holes. We can then
conclude that photogenerated holes in the valence band at the
TaON surface can efficiently oxidize I- but cannot oxidize
SCN-, Br-, methanol, ethanol, 1-propanol, and acetic acid.
As to the stability of the particulate TaON film electrode,
we confirmed that the spectral shape in the action spectrum of
the anodic photocurrent (the IPCE vs λ for the O2 photo-
evolution, Figure 3) and the effect of addition of reductants on
the IPCE (Figure 4) were not altered after prolonged visible-
light irradiation in 0.1 M HClO4 for 20 h, though the amount
of the anodic photocurrent for TaON slightly decreased. Since
it was reported15 that the oxygen photoevolution from the TaON
photocatalyst proceeded without any noticeable decrease in the
activity, this decrease in the photocurrent is probably due to
slight peeling off of TaON particles from the FTO glass substrate
under solution stirring.
Figure 3. Photocurrent action spectra (IPCE vs wavelength) for
particulate TaON and Ta2O5 thin film electrodes, observed in 0.1 M
HClO4 at an applied potential of 0.7 V vs Ag/AgCl. The broken line
indicates an observed UV-vis diffuse reflectance spectrum of TaON
powder.
stoichiometric reduction of Fe3+. Interestingly, the O2 evolution
was completely suppressed when methanol (0.5 wt %) was
added to the aqueous TaON suspension (Figure 2), demonstrat-
ing that the photooxidation of methanol proceeded predomi-
nantly over the water photooxidation, or in other words, that
the evolved O2 came from the water photooxidation.
Figure 3 shows the action spectrum of the anodic photocurrent
(or the IPCE vs wavelength) for a freshly prepared particulate
TaON film electrode in 0.1 M HClO4 (pH 1.1) at a potential of
0.7 V versus Ag/AgCl. The IPCE for a Ta2O5 film electrode is
also included for reference. The action spectrum of the anodic
photocurrent for TaON agreed well with its absorption spectrum
(broken line in Figure 3), extending up to ca. 530 nm. Since
the O2 evolution was confirmed by the photocatalytic experiment
with a TaON suspension, as shown in Figure 2, the anodic
photocurrents can be attributed to the photooxidation of H2O
to O2. It was reported15 that the efficiency of the O2 photo-
evolution from a TaON suspension in the presence of Ag+ ions

Water Photooxidation on a TaON Photocatalyst
J. Phys. Chem. B, Vol. 109, No. 18, 2005 8923
Figure 6. Photoinduced MIR-IR spectra for a TaON particulate film
in contact with pure water (pH 7), recorded in 10, 20, 30, and 40 min
after the start of visible-light (λ g 420 nm) irradiation. A TaON before
irradiation was taken as the reference.
Figure 5. MIR-IR spectra for (a) Ta2O5 and (b) TaON particulate
films in air.
2. In Situ MIR-IR Investigations of Surface Reaction
Intermediates. We reported in previous papers20,24 that the
multiple internal reflection FTIR (MIR-IR) spectroscopy was
sensitive enough to detect surface reaction intermediates of
submonolayer amounts and could be applied to in-situ monitor-
ing of photocatalytic reactions on particulate TiO2 films in
contact with aqueous solutions. By using a diamond single
crystal as a multiple internal reflection element (IRE), we can
measure IR absorptions in a low-frequency region from 1500
to 650 cm-1 with a high spectral sensitivity, in which surface
reaction intermediates of the water photooxidation exhibit
absorption bands because of the O-O stretching vibration
mode.20,24 In this work, we have applied this method to the
photocatalytic reaction at TaON particulate films under visible-
light irradiation.
Figure 5 shows MIR-IR spectra for particulate Ta2O5 and
TaON films before irradiation for reference in later discussion.
Spectrum a is for a particulate Ta2O5 film under dry air. It
exhibits weak peaks at 839 and 942 cm-1 and a strong broad
shoulder below 780 cm-1. The latter broad shoulder can be
assigned25-28 to phonon modes for crystalline Ta2O5 such as
Ta-O and Ta-O-Ta stretching vibrations, while the former
two peaks can be assigned25-28 to phonon modes for oxygen-
deficient crystalline phases such as TaO and TaO2 included as
impurities in the Ta2O5 film. Spectrum b is for a TaON film,
measured after nitridation of Ta2O5 powder at 800 °C under an
NH3 atmosphere. It exhibits new peaks at 782 and 827 cm-1
together with a strong broad shoulder below 780 cm-1. Although
no information is at present available about vibrational spectra
for TaON itself, by referring to vibrational spectra reported for
well-studied oxynitrides such as AlON,29 GaON,29 and SiON,30-33
the former two peaks may be assigned to N-Ta-O phonon
modes of nonstoichiometric TaON phases present as impurities,
while the latter shoulder may be assigned to phonon modes for
crystalline TaON. The presence of the nonstoichiometric TaON
phases may be supported by a result of XPS analysis that the
RN/O for TaON powder prepared in this work is 0.7, a little
smaller than unity, as mentioned before.
Figure 6 shows in-situ MIR-IR absorption spectra for
photoinduced species at a particulate TaON film in contact with
pure water of pH 7 under visible-light (λg420 nm) irradiation.
The spectrum for the TaON film before irradiation was taken
as the spectral reference. The 782 and 827 cm-1 bands, assigned
above to the N-Ta-O phonon modes, appeared as depressions
(or negative peaks), and the (negative) peak height increased
with the irradiation time up to 30 min. On the other hand, a
new broad band appeared in the region of 1350-900 cm-1. The
negative 782 and 827 cm-1 bands and the broad band in 1350-
900 cm-1 both became much weaker when the TaON surface
was in advance oxidized by heating at 400 °C until the surface
atomic N/O ratio decreased from 0.7 to 0.4. Thus, the photo-
induced (negative and positive) bands in Figure 6 can be
regarded to be due to the photoinduced surface oxidation in
the TaON powder.
To confirm the above conclusion, we did isotopic labeling
experiments by using H218O as a solvent, instead of H216O.
Spectrum a in Figure 7 shows an MIR-IR spectrum for a
particulate TaON film in contact with pure H218O, whereas
spectrum b is that for TaON in contact with pure H216O. Both
spectra a and b show essentially the same spectral changes, that
is, the appearances of the negative bands at 782 and 827 cm-1
and the broad band in the range of 1350-900 cm-1, but the
latter broad band in H218O is considerably distorted compared
with that in H216O, namely, the broad band in H218O (spectrum
a) is largely depressed in intensity on the high-frequency side
compared with that for H216O (spectrum b). This result clearly
indicates that the broad band in the range of 1350-900 cm-1
is due to certain oxidized surface species formed by reaction of
photogenerated holes with H2O at the TaON surface under
visible-light irradiation, in harmony with the aforementioned
conclusion.
We mentioned earlier that O2 gas was evolved from an
aqueous suspension of TaON powder under visible-light ir-
radiation when 2 mM Fe3+ was added to the suspension as a
sacrificial electron acceptor (Figure 2). Figure 8 shows the effect

8924 J. Phys. Chem. B, Vol. 109, No. 18, 2005
Nakamura et al.
free HOOH (H16O16OH) in an aqueous HOOH solution is also
included for reference. The photoinduced MIR-IR spectrum
in H216O with 10 mM Fe3+, spectrum a in Figure 8, shows a
new weak absorption band at 872 cm-1, in addition to the
negative bands at 782 and 827 cm-1 and the broad band in the
range of 1350-900 cm-1 that were observed without Fe3+
(Figure 7b). Since Fe3+ acts as an efficient electron acceptor, it
is likely that the 872 cm-1 band is due to a surface intermediate
of the water photooxidation reaction. We can mention that the
872 cm-1 band in H216O is located at nearly the same
wavenumber as that for the O-O stretching band, ν(O-O), of
free HOOH (H16O16OH) peaked at 877 cm-1 34,35 in spectrum
c of Figure 8. However, almost no peak assigned to the bending
band, δ(O-O-H), of free HOOH, peaked at 1380 cm-1 34,35
,
was observed in spectrum a of Figure 8. This implies that the
872 cm-1 band can be assigned to the ν(O-O) band of a certain
peroxo species at the TaON surface, which may tentatively
assigned to adsorbed H2O2, since it is expected that the δ(O-
O-H) mode of adsorbed HOOH decreases in intensity by
interaction with surface Ta atoms.
The above assignment of the 872 cm-1 band has been
supported by 18O-isotopic labeling experiments. Spectrum b in
Figure 8 clearly shows that the photoinduced MIR-IR spectrum
in H218O with 10 mM Fe3+ also shows a new weak absorption
band at 851 cm-1, at a position slightly deviated from 872 cm-1
in H216O, together with a shoulder at 872 cm-1. It is reported20,36
experimentally that the 18O isotopic shift of the ν(O-O) mode
for surface peroxo species between 16O-16O and 18O-16O is
Figure 7. Photoinduced MIR-IR spectra for a TaON film in contact
with (a) H218O and (b) H216O, observed after 20-min visible-light
irradiation.
19 cm-1, while that between 16O-16O and 18O-18O is 42 cm-1
.
Thus, the observed shift of 21 cm-1 (872 f 851 cm-1) in the
present work can be reasonably assigned to the oxygen isotopic
shift of the ν(O-O) mode of surface peroxo species between
16O-16O and 18O-16O.
Discussion
Domen et al. recently reported13,15 that tantalum oxynitride
(TaON) functioned as a visible-light (λ < 520 nm) driven
photocatalyst for the oxidation of H2O into O2 in the presence
of an appropriate electron acceptor. The prominent feature of
TaON is that the photocatalytic water oxidation occurs, though
its valence band should be composed of hybridized N2p and
O2p orbitals. In the present work, we have confirmed that the
photocatalytic O2 evolution occurs from a TaON suspension in
the presence of Fe3+ (Figure 2). However, we have found that
the TaON surface is slightly oxidized under visible-light
irradiation (Figures 6 and 7), indicating that the oxygen
photoevolution on TaON actually occurs on a thin Ta-oxide
overlayer. We have also found that (1) a certain surface peroxo
species is formed in the course of the O2 photoevolution reaction
(Figure 8) and (2) TaON cannot oxidize reductants such as
SCN-, Br-, methanol, ethanol, 2-propanol, and acetic acid
(Figure 4) in spite that it can oxidize H2O into O2.
Now, let us consider how the water photooxidation proceeds
at the surface of TaON under visible-light irradiation, on the
basis of our recently reported mechanism for this reaction on
the TiO2 (rutile) surface under UV irradiation.19-21 The key of
the mechanism is that the water photooxidation reaction is not
initiated by the oxidation of surface OH group (Ti-OHs) with
photogenerated holes (h+),
Figure 8. Photoinduced MIR-IR spectra for a TaON film in contact
with (a) H216O and (b) H218O, both containing 10 mM Fe3+, observed
after 10-min visible-light irradiation. Spectrum (c) is an MIR-IR spectra
for an aqueous solution of 10% HOOH.
Ti-OHs + h+ f [Ti •OH]s+
(2)
of addition of 10 mM Fe3+ on the photoinduced MIR-IR
spectra for a particulate TaON film. An MIR-IR spectrum of
but by a nucleophilic attack of a H2O molecule (Lewis base) to

Water Photooxidation on a TaON Photocatalyst
J. Phys. Chem. B, Vol. 109, No. 18, 2005 8925
SCHEME 1: Reaction Scheme for the Oxygen Photoevolution Reaction on TiO2 (Rutile) in Contact with an Aqueous
Solution of pH of 1 to about 1219,20
a surface-trapped hole (Lewis acid) at surface lattice oxygen
(Ti-O-Tis), accompanied by bond breaking.
in other words, the oxidation of the TaON surface, with a release
of N2 gas.
[Ti-O-Ti]s + h+ + H2O f [Ti-O• HO-Ti]s + H+ (3)
2[Ta-N• HO-Ta]s + 2h+ f 2[Ta-O-Ta]s + 2H+ + N2
(6)
The O2 evolution occurs via formation of a surface peroxo
species, as shown in Scheme 1. This conclusion was obtained
by in-situ spectroscopic studies such as MIR-IR absorption20
and photoluminescence (PL) measurements19,20 and was sup-
ported by mechanistic studies using atomically flat single-crystal
TiO2 (rutile) electrodes.21 This conclusion is also supported by
the following reported results: (1) ESR measurements at low
temperature37-39 showed that photogenerated holes on TiO2
produced no •OH radical but Ti-O• radicals. (2) Recent
ultraviolet photoelectron spectroscopic (UPS) studies indi-
cated40-42 that the O2p level for terminal hydroxyl group (Ti-
OHs) on TiO2 (rutile) was far below the top of the O2p valence
band of TiO2 (rutile). The latter fact indicates that reaction 2 is
energetically impossible. Reaction 2 is an electron-transfer-type
reaction, whereas reaction 3 is a Lewis acid-base-type reaction,
and therefore their energetics and kinetics are quite different
from each other.
The following reactions of photogenerated holes with water
on a thin Ta-oxide layer will proceed by a similar mechanism
to that on TiO2. Namely, the O2 photoevolution is initiated by
the nucleophilic attack of H2O molecules to surface-trapped
holes at surface lattice O atoms (Ta-O-Tas).
[Ta-O-Ta]s + h+ + H2O f [Ta-O• HO-Ta]s + H+ (7)
[Ta-O• HO-Ta]s + h+ f [Ta-O-O-Ta]s + H+ (8)
[Ta-O-O-Ta]s + H2O + 2h+ f
[Ta-O-Ta]s + O2 + 2H+ (9)
This mechanism is in agreement with the in-situ MIR-IR
measurements for surface intermediates on TaON powder. We
mentioned in the preceding section that the weak absorption
band that peaked at 872 cm-1, assigned to the ν(O-O) band of
the surface peroxo species (weakly adsorbed HOOH), appeared
in the course of the O2 photoevolution reaction (Figure 8a). The
18O isotopic labeling experiments (Figure 8b) revealed that the
surface peroxo species, though it was produced in H218O,
contained both O18 and 16O (Figure 8b). This result clearly shows
that the surface peroxo species is formed by the abovementioned
reaction scheme (reactions 7 and 8), as is seen from the
following consideration. Once TaON is immersed with H218O,
it is expected that surface Ta-16OH group on TaON is readily
converted into Ta-18OH, whereas the surface lattice oxygen
still remains in the form of [Ta-16O-Ta] even in contact with
H218O, since it is reported44 that oxygen exchange between water
and surface lattice oxygen in metal oxides in general hardly
occurs at room temperature in the dark. The incorporation of
16O into the surface peroxo species is therefore possible only
when water molecules (H218O) attack the surface lattice oxygen
existing in the form of the surface-trapped holes [Ta-16O‚‚‚
Now, it will be reasonable to assume that the water photo-
oxidation at the surface of TaON under visible-light irradiation
proceeds by a similar mechanism to that on TiO2 (rutile). As
mentioned earlier, the oxygen photoevolution on TaON actually
occurs at the surface of a thin Ta-oxide overlayer (Figure 8).
The initial formation of such a surface thin Ta-oxide layer is
expected to proceed as follows. The valence band of TaON is
constituted by hybridized N2p and O2p orbitals, in which the
contribution of the N2p orbitals is much larger than the O2p
orbitals at the top of the valence band.43 Thus, photogenerated
holes will initially be trapped at surface lattice N atoms, resulting
in a surface-trapped hole which may be represented as [Ta-
N‚‚‚Ta]+ (reaction 4).
s
[Ta-N-Ta]s + h+ f [Ta-N‚‚‚Ta]+
(4)
s
In this surface-trapped hole, the Ta-N bond becomes weaker
by trapping a hole (i.e., one electron is missing in the bonding
orbital). Once this happens, the hole-trapped site acts, because
of its positive charge, as an active site for a nucleophilic reagent
(Lewis acid). In an aqueous solution, a H2O molecule acts as a
nucleophilic reagent (Lewis base) and hence attacks the surface-
trapped hole, accompanied by the bond breaking and attachment
of OH group, similarly to reaction 3.
Ta]+ , such as reaction 7.
s
The aforementioned mechanism (reactions 7-9) is further
supported by the observed effect of the addition of reductants
to the electrolyte on the IPCE. Figure 3 shows that TaON can
oxidize H2O into O2 by visible-light irradiation of up to ca. 530
nm, at which the light absorption starts, whereas Figure 4 shows
that photogenerated holes on TaON cannot oxidize reductants
such as SCN-, Br-, and alcohols such as methanol and ethanol,
which have the redox potentials for the oxidation much more
negative than water.
[Ta-N‚‚‚Ta]+s + H2O f [Ta-N• HO-Ta]s + H+ (5)
Further reactions of [Ta-N• HO-Ta]s with holes and water
finally result in the formation of surface Ta-O-Ta bonds, or

8926 J. Phys. Chem. B, Vol. 109, No. 18, 2005
Nakamura et al.
Figure 9. Schematic illustration of the energy bands for TaON at pH 016 together with state density distributions for one-electron-transfer redox
couples estimated from reported reorganization energies (λ).46 The minus sign with a circle, electron; the plus sign with a circle, hole; Eeq, the
equilibrium redox potentials for one-electron-transfer redox couples, reported.45
This discrepancy can be understood if we take into account
that the photooxidation of reductants such as SCN- and Br-
may indeed occur by an electron-transfer mechanism, but the
water photooxidation proceeds not by the electron-transfer
mechanism but by the Lewis acid-base mechanism (reactions
7-9). Figure 9 shows the energy band diagram for TaON,16
together with the equilibrium redox potentials (Eredoxeq) for the
Finally, it will be important to consider how methanol or other
alcohols are photooxidized on TaON under visible-light irradia-
tion. We mentioned earlier that the addition of methanol to the
electrolyte caused no increase in the IPCE (Figure 4), indicating
that photogenerated holes at the TaON surface cannot oxidize
methanol. On the other hand, the photocatalytic O2 evolution
was completely suppressed by the addition of methanol to an
aqueous TaON suspension (Figure 2), indicating that the
photooxidation of methanol proceeds predominantly over the
water photooxidation at the TaON surface. The discrepancy can
be explained because the methanol does not react directly with
photogenerated holes but reacts with some surface intermediates
of the water photooxidation reaction, such as [Ta-O• HO-
Ta]s. This argument also gives strong support to the validity of
the abovementioned Lewis acid-base mechanism for the water
photooxidation, because, if the water photooxidation was
initiated by a mechanism of direct oxidation of surface Ta-
OH by photogenerated holes, more easily oxidized methanol
(or adsorbed methanol such as Ta-OCH3) would certainly be
oxidized by the holes.
eq
reductants as one-electron-transfer redox couples. The Eredox
for I-/I• (1.35 V45) is much more negative than the top of the
s
valence band of TaON, EV , lying at 2.2 V, and thus we can
expect that I- is efficiently oxidized by holes, in harmony with
the experimental result (Figure 4). The Eredoxeq for SCN-/SCN•
(1.64 V45) and Br-/Br• (2.00 V45) are more negative than EV ,
s
though both SCN- and Br- are not oxidized by photogenerated
holes (Figure 4). This may be explained by taking into account
the presence of the large reorganization energies (λ) for small
ions in an aqueous solution. In fact, the state density distributions
for the occupied levels, estimated from reported λ values46 for
SCN- and Br-, have little overlap with the top of the valence
band of TaON at which photogenerated holes lie, as shown in
Figure 9.
Conclusions
eq
As to the case of the water oxidation, the Eredox for (H2O/
s
•OH), lying at 2.38 V,45 is more positive than EV (2.2 V) and
The present work has confirmed that the photocatalytic O2
evolution occurs from a TaON suspension in the presence of
Fe3+. The in-situ MIR-IR experiments have indicated that the
TaON surface is oxidized under visible-light irradiation, indicat-
ing that the oxygen photoevolution on TaON actually occurs
on a thin Ta-oxide overlayer. The in-situ MIR-IR experiments
have also shown that a certain surface peroxo species, which
may be tentatively assigned to adsorbed HOOH, is formed as
an intermediate of the O2 photoevolution reaction. The experi-
ments on the effect of addition of reductants to the electrolyte
on the IPCE have shown that photogenerated holes at the TaON
surface cannot oxidize reductants such as SCN-, Br-, methanol,
ethanol, 2-propanol, and acetic acid, though it can oxidize H2O
into O2. All these results give strong support to the validity of
our recently proposed mechanism for the water photooxidation
that it is initiated by a nucleophilic attack of a water molecule
(Lewis base) on the surface-trapped hole (Lewis acid).
much more positive than those for SCN- and Br-. This result
clearly indicates that the electron-transfer type oxidation of H2O
by photogenerated holes is energetically impossible, or in other
words, the water photooxidation on TaON cannot be explained
by this mechanism. On the other hand, in the mechanism of
the nucleophilic attack of H2O molecules (Lewis base) to
surface-trapped holes at surface lattice O atoms (Lewis acid),
the reaction rate is not directly governed by the redox potential
Eredoxeq for H2O/•OH. The reaction rate will rather be governed
by the Gibbs energy of formation of intermediate radicals such
as [Ta-O• HO-Ta]s which includes the distortion energy of
the surface TaON (or Ta-oxide) lattice. Thus, the above-
mentioned contradiction can be regarded as giving strong
support to the validity of our previously reported mechanism
of the Lewis acid-base mechanism for the water photooxidation
on metal oxides or oxynitrides.

Water Photooxidation on a TaON Photocatalyst
J. Phys. Chem. B, Vol. 109, No. 18, 2005 8927
(23) Shiga, A.; Tsujiko, A.; Yae, S.; Nakato, Y. Bull. Chem. Soc. Jpn.
1998, 71, 2119.
(24) Nakamura, R.; Imanishi, A.; Murakoshi, K.; Nakato, Y. J. Am.
References and Notes
(1) Fujishima, A.; Honda, K. Nature 1972, 238, 37.
(2) Sato, S.; White, M. Chem. Phys. Lett. 1980, 72, 83.
(3) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W.
Chem. ReV. 1995, 95, 69.
(4) Sato, S. Chem. Phys. Lett. 1986, 123, 126.
(5) Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Science
2001, 293, 269.
(6) Lindgren, T.; Mwabora, J. M.; Avendann˜o, E.; Jonsson, J.; Hoel,
A.; Granqvist, C.-G.; Lindquist, S.-E. J. Phys. Chem. B 2003, 107, 5709.
(7) Nakamura, R.; Tanaka, T.; Nakato, Y. J. Phys. Chem. B 2004, 108,
10617.
(8) Umebayashi, T.; Yamaki, T.; Itoh, H.; Asai, K. Appl. Phys. Lett.
2002, 81, 454.
(9) Ohno, T.; Mitsui, T.; Matsumura, M. Chem. Lett. 2003, 32, 364.
(10) Khan, S. U. M.; Al-Shahry, M.; Ingler, W. B., Jr. Science 2002,
297, 2243.
(11) Sakthivel, S.; Kisch, H. Angew. Chem., Int. Ed. 2003, 42, 4908.
(12) Irie, H.; Watanabe, Y.; Hashimoto, K. Chem. Lett. 2003, 32, 772.
(13) Hitoki, G.; Takata, T.; Kondo, J. N.; Hara, M.; Kobayashi, H.;
Domen, K. Chem. Commun. 2002, 1698.
(14) Hara, M.; Chiba, E.; Ishikawa, A.; Takata, T.; Kondo, J. N.; Domen,
K. J. Phys. Chem. B 2003, 107, 13441.
(15) Hara, M.; Hitoki, G.; Takata, T.; Kondo, J. N.; Kobayashi, H.;
Domen, K. Catal. Today 2003, 78, 555.
(16) Chun, W.-J.; Ishikawa, A.; Fujisawa, H.; Takata, T.; Kondo, J. N.;
Hara, M.; Kawai, M.; Matsumoto, Y.; Domen, K. J. Phys. Chem. B 2003,
107, 1798.
(17) Hitoki, G.; Ishikawa, A.; Takata, T.; Kondo, J. N.; Hara, M.;
Domen, K. Chem. Lett. 2002, 736.
Chem. Soc. 2003, 125, 7443.
(25) Zhang, J.-Y.; Boyd, I. W. Appl. Surf. Sci. 2002, 186, 40.
(26) Fang, Q.; Zhang, J.-Y.; Wang, Z. M.; Wu, J. X.; O’Sullivan, B. J.;
Hurley, P. K.; Leedham, T. L.; Davies, H.; Audier, M. A.; Jimenez, C.;
Senateur, J.-P.; Boyd, I. W. Thin Solid Films 2003, 428, 248.
(27) Boyd, I. W.; Zhang, J.-Y. Microelectron. Reliab. 2000, 40, 649.
(28) Ono, H.; Hosokawa, Y.; Shinoda, K.; Koyanagi, K.; Yamaguchi,
H. Thin Solid Films 2001, 381, 57.
(29) Franchy, R. Surf. Sci. Rep. 2000, 38, 195.
(30) Bermudez, V. M. Thin Solid Films 1999, 347, 195.
(31) Ay, F.; Aydinli, A. Opt. Mater. 2004, 26, 33.
(32) Scarlete, M.; McCourt, N.; Butler, I. S.; Harrod, J. F. Chem. Mater.
2001, 13, 655.
(33) Pandey, R. K.; Patil, L. S.; Bange, J. P.; Patil, D. R.; Mahajan, A.
M.; Patil, D. S.; Gautam, D. K. Opt. Mater. 2004, 25, 1.
(34) Engdahl, A.; Nelander, B.; Karlstro¨m, G. J. Phys. Chem. A 2001,
105, 8393.
(35) Pettersson, M.; Tuominen, S.; Ra¨sa¨nen, M. J. Phys. Chem. A 1997,
101, 1166.
(36) Lin, W.; Frei, H. J. Am. Chem. Soc. 2002, 124, 9293.
(37) Howe, R. F.; Gra¨etzel, M. J. Phys. Chem. 1987, 91, 3906.
(38) Micic, O. I.; Zhang, Y.; Cromack, K. R.; Trifunac, A. D.; Thurnauer,
M. C. J. Phys. Chem. 1993, 97, 7277.
(39) Micic, O. I.; Zhang, Y.; Cromack, K. R.; Trifunac, A. D.; Thurnauer,
M. C. J. Phys. Chem. 1993, 97, 13284.
(40) Brookes, I. M.; Muryn, C. A.; Thornton, G. Phys. ReV. Lett. 2001,
87, 266103.
(41) Muryn, C. A.; Hardman, P. J.; Crouch, J. J.; Raiker, G. N.;
Thornton, G.; Law, D. S. L. Surf. Sci. 1991, 215-242, 747.
(42) Henderson, M. A. Surf. Sci. Rep. 2002, 46, 1.
(43) Fang, C. M.; Orhan, E.; de Wijs, G. A.; Hintzen, H. T.; de Groot,
R. A.; Marchand, R.; Saillard, J.-Y.; de With, G. J. Mater. Chem. 2001,
11, 1248.
(44) Sato, S. J. Phys. Chem. 1987, 91, 2895.
(45) Wardman, P. J. Phys. Chem. Ref. Data 1989, 18, 1637.
(46) Burg, K. V.; Delahay, P. Chem. Phys. Lett. 1981, 78, 287.
(18) Ishikawa, A.; Takata, T.; Kondo, J. N.; Hara, M.; Domen, K. J.
Phys. Chem. B 2004, 108, 11049.
(19) Kisumi, T.; Tsujiko, A.; Murakoshi, K.; Nakato, Y. J. Electroanal.
Chem. 2003, 545, 99.
(20) Nakamura, R.; Nakato, Y. J. Am. Chem. Soc. 2004, 126, 1290.
(21) Nakamura, R.; Ohashi, N.; Imanishi, A.; Osawa, T.; Matsumoto,
Y.; Koinuma, H.; Nakato, Y. J. Phys. Chem. B 2005, 109, 1648.
(22) Ito, S.; Takeuchi, T.; Katayama, T.; Sugiyama, M.; Matsuda, M.;
Kitamura, T.; Wada, Y.; Yanagida, S. Chem. Mater. 2003, 15, 2824.