Kinetics of photocatalytic reactions under extremely low-intensity UV illumination on titanium dioxide thin films

Article abstract of DOI:10.1021/jp972002k

The photocatalytic decomposition efficiency of gaseous 2-propanol was studied using a titanium dioxide thin film under very weak UV light; the incident UV light intensity was 36 nW-45 μW·cm^-2. Under such lowintensity UV illumination, the value of the quantum yield (QY) increased gradually with decreasing number of absorbed photons and finally saturated (28%) for a number of absorbed photons less than 4 × 10¹¹ quanta·cm^-2 ·s^-1 for an initial 2-propanol concentration of 1000 ppmv. Thus, purely light-limited conditions were reached. For lower initial concentrations, the QY values decreased, but the same maximum QY value as that for 1000 ppmv was also approached with decreasing light intensity. We discuss these results in terms of the normalized absorbed photon number (I_norm/s^-1), a parameter that we have defined as the ratio of the number of absorbed photons ([photon]ab) to the number of adsorbed 2-propanol molecules ([M]ad). When all of the experimental QY values were plotted as a function of I_norm, all of the points appeared on a single line for a wide range of initial 2-propanol concentrations. On the basis of these results, we conclude that either ·OH radicals or 2-propanol molecules must be able to diffuse at least ca. 11 nm on the titanium dioxide surface in order to react with each other. We also conclude that the maximum QY value of 28% represents the intrinsic charge-separation efficiency for this photocatalyst.

Full text of DOI:10.1021/jp972002k

J. Phys. Chem. A 1997, 101, 8057-8062
8057
Kinetics of Photocatalytic Reactions under Extremely Low-Intensity UV Illumination on
Titanium Dioxide Thin Films
Yoshihisa Ohko,† Kazuhito Hashimoto, and Akira Fujishima*
Department of Applied Chemistry, Faculty of Engineering, The UniVersity of Tokyo,
7-3-1 Hongo, Bunkyo-ku, Tokyo 113
ReceiVed: June 18, 1997X
The photocatalytic decomposition efficiency of gaseous 2-propanol was studied using a titanium dioxide thin
film under very weak UV light; the incident UV light intensity was 36 nW-45 µW‚cm-2. Under such low-
intensity UV illumination, the value of the quantum yield (QY) increased gradually with decreasing number
of absorbed photons and finally saturated (28%) for a number of absorbed photons less than 4 × 1011
quanta‚cm-2‚s-1 for an initial 2-propanol concentration of 1000 ppmv. Thus, purely light-limited conditions
were reached. For lower initial concentrations, the QY values decreased, but the same maximum QY value
as that for 1000 ppmv was also approached with decreasing light intensity. We discuss these results in terms
of the normalized absorbed photon number (Inorm/s-1), a parameter that we have defined as the ratio of the
number of absorbed photons ([photon]ab) to the number of adsorbed 2-propanol molecules ([M]ad). When all
of the experimental QY values were plotted as a function of Inorm, all of the points appeared on a single line
for a wide range of initial 2-propanol concentrations. On the basis of these results, we conclude that either
•OH radicals or 2-propanol molecules must be able to diffuse at least ca. 11 nm on the titanium dioxide
surface in order to react with each other. We also conclude that the maximum QY value of 28% represents
the intrinsic charge-separation efficiency for this photocatalyst.
Introduction
achieved under 0.5 mW‚cm-2 irradiance for 1000 ppmv gaseous
acetaldehyde decomposition using TiO2-containing paper12 and
It is well-known that when photons with energies greater than
the bandgap are absorbed by semiconductors such as ZnO, CdS,
and TiO2, electron-hole pairs are generated and are then
transferred to the surface, where they are trapped and can react
with surface-adsorbed molecules. TiO2 in particular has
beneficial characteristics, such as its chemical and physical
stability, as well as the strong oxidizing power of the photo-
generated holes, with which most organic compounds can be
oxidized to carbon dioxide at ambient temperature and pressure.
Therefore, the photocatalytic destruction of harmful and toxic
materials using TiO2 powders and films has great promise for
the purification of air and water.1-3 Traditionally, high-pressure
mercury and xenon lamps have been used in most photocatalytic
research for the purpose of high-rate photodecomposition and
also to compensate for the low photocatalytic efficiency of TiO2
itself.4-6 In recent years, since Heller et al. reported the
oxidative stripping of oil sheens from the surface of sea water
with buoyant photocatalyst-coated ceramic microbubbles,7 the
field of environmental purification has been devoting increasing
attention to photocatalysis using sunlight and black-light-type
fluorescent lamps as UV light sources, i.e., at the several
mW‚cm-2 level, which is safe for human exposure.8-11
In contrast, we have been devoting our attention to a unique
type of passive-type purification system for indoor working and
living environments. These systems incorporate deodorizing,
antibacterial, and self-cleaning functions under lower level
illumination from room light. In connection with this concept,
many different types of TiO2 materials that exhibit higher
photocatalytic activities than P-25 powder, which is known for
its highly photocatalytic activity, have been prepared.12-15 For
example, quantum yields (QY) of close to 100% have been
semitransparent TiO2 thin films formed on glass.13,14 In
addition, we have found that Escherichia coli cells can be
completely killed on TiO2-coated glass tiles even under room
light.15 Using these materials, we can also investigate the
kinetics of photocatalytic reactions under even lower-intensity
UV illumination, e.g., at nW‚cm-2 levels. Using this approach,
new aspects of photocatalysis, distinct from those associated
with the higher-intensity UV illumination employed in many
other reports, can be expected to be discovered.
In the present work, we have examined the stationary
photocatalytic decomposition of dilute gas-phase 2-propanol on
a TiO2 thin film under extremely low-intensity UV illumination,
from as little as 10 nW‚cm-2 and ranging up to 10 µW‚cm-2
,
for the first time. We have paid particular attention to the
dependence of the QY values on the number of absorbed
photons ([photon]ab) and on the number of 2-propanol molecules
([M]ad) adsorbed on the surface of the film. Consequently, we
found that the ratio of [photon]ab/[M]ad uniquely determines QY,
even with a wide range of different initial concentrations, i.e.,
1-1000 ppmv. We have discussed the diffusion length of •OH
radicals on the TiO2 surface and the charge-separation efficiency
of the TiO2 film on the basis of these results.
The reasons that 2-propanol was chosen as a reactant for these
experiments were that (i) it is efficiently photodecomposed to
acetone, which undergoes further reactions at a much slower
rate, (ii) acetone can be detected sensitively using gas chroma-
tography (GC), (iii) a single photon is considered to participate
in the generation of each acetone molecule, and (iv) its self-
oxidation is negligible. The latter aspect is distinct from the
aldehydes, which are decomposed via free radical chain reactions
involving reduced oxygen species,16,17 and therefore, the reaction
dynamics cannot be accounted for in a simple fashion. It should
also be noted that, under very low-intensity illumination
* To whom correspondence should be addressed.
† Research Fellow of the Japan Society for the Promotion of Science.
X Abstract published in AdVance ACS Abstracts, September 15, 1997.
S1089-5639(97)02002-1 CCC: $14.00 © 1997 American Chemical Society

8058 J. Phys. Chem. A, Vol. 101, No. 43, 1997
Ohko et al.
conditions, the adsorption of the reactant on the TiO2 surface
is essentially at equilibrium. We are currently examining mass
transport effects at higher UV light intensities.
Experimental Section
TiO2 thin films were prepared on soda lime glass by a
conventional spin-coating process, using a commercial TiO2
anatase aqueous sol (Ishihara Sangyo Kaisha, Ltd., STS-21, 20
nm particle diameter, 50 m2‚g-1 surface area). A 7 cm × 7
cm piece of glass was spun at 1500 rpm for 10 s. After air-
drying, a second coating was applied in the same manner over
the coated gel. The resulting sample was calcined at 450 °C
for 1.5 h in air. The thickness of the semitransparent film was
about 1.7 µm, according to cross-sectional observation by atomic
Figure 1. Inverse plots of the gaseous concentration (C/mg‚m-3) and
the weight of adsorbed 2-propanol (M/mg) on the TiO2 thin film used
for analysis of the Langmuir-type isotherm.
force microscopy. The weight of the film was 0.40 mg‚cm-2
.
The absorbed photon flux was estimated as follows. The
The roughness factor of the film was estimated to be about 150
cm2/cm2 , by measuring the amount of adsorbed cyanine dye
(1-(2-carboxyethyl)-2-[7-[1-(2-carboxyethyl)-1,3-dihydro-3,3-
dimethyl-2H-indol-2-ylidene]-1,3,5-heptatryenyl]-3,3-dimethyl-
3H-indolium hydroxide, inner salt, NK3422, Nippon Kankoh-
Shikiso Kenkyusyo Co., Ltd.) on the sample surface. Scanning
electron micrographic observation showed that the film consisted
of particles with diameters of about 50-100 nm due to particle
growth during the sintering step.
incident photon flux (I0) can be expressed as follows18
I0 ) Ia + It + If + Ib
(1)
where Ia is the intensity of light absorbed in the film, It is the
transmitted intensity, If is the forward-scattered intensity, and
Ib is the backscattered intensity. I0 was measured directly using
the UV power meter. However there was some degree of
nonuniformity of intensity over the illuminated area because
of passage of the UV light though a 1-m glass fiber light pipe.
Therefore I0 was estimated in a manner analogous to that
described by Hill et al.19 for the measurement of Ib. This
involves the establishment of a contour map of light intensity
and subsequent calculation of the average intensity over the
illuminated area. The full-strength intensity of I0, after passing
through the light pipe, was 45 µW‚cm-2. The sum of It and If
was determined using a conventional integrating sphere (Shi-
madzu ISR-260). The modified method of Bolton et al.20 was
not used, because Ib was measured separately (see below). The
sample was placed at the sample position of the integrating
sphere with a piece of glass plate similar to that used as a support
for the TiO2 film as the reference. The apparent transmission
factor, (It + If)/I0, reading of the instrument was 0.30 at 365
nm. Ib from the sample was estimated using a method similar
to that of Hill et al.19 Ib at normal incidence was extrapolated
by measuring Ib as a function of the incident light angle. The
factor Ib/I0 evaluated following this methodology was 0.05. The
factor Ia/I0 for the sample TiO2 film was calculated to be 0.65.
The UV absorption by 2-propanol on the film (365 nm) was
considered to be negligible.
The amount of physically adsorbed water on the TiO2 surface
was estimated using a differential thermobalance (Shikuu Riko
TGD 7000RH) for the TiO2 powder (Ishihara Sangyo Kaisha,
Ltd., ST-21), the commercial powder corresponding to that
dispersed in STS-21 sol, which was used after it was subjected
to the same heat treatment that was used to prepare the film
(see above).
The amount of 2-propanol adsorbed on the TiO2 sample was
estimated by measuring the decrease in the balance of 2-pro-
panol gas concentration in the glass vessel containing the TiO2
film sample versus that in a separate empty glass vessel. The
concentrations were measured 1 h after a given concentration
gas mixture was introduced into each glass vessel.
An O2 (20%)-N2 gas mixture, which was passed through a
16 °C water humidifier in order to adjust the relative humidity
to 50%, was used to fill the 1-L Pyrex glass photocatalytic
reaction vessel. For purposes of preparing the gas mixtures
containing various concentrations of 2-propanol, the nonhu-
midified O2-N2 mixture gas was first saturated with 2-propanol
(Kosou Kagaku Yakuhin) by passage through a 2-propanol
liquid reservoir at room temperature, resulting in a concentration
of 5 vol %, as measured using GC (see below) after dilution.
Measured quantities of the 2-propanol-saturated gas were then
injected into the 1-L vessel using a syringe.
For the photocatalytic decomposition of gaseous 2-propanol,
the TiO2 thin film was illuminated with a Hg-Xe lamp (Hayashi
Tokei, Luminar Ace 210). To obtain monochromatic UV light,
a 365 nm band-pass filter (fwhm ) 2 nm, Kenko, BP-W1-365)
was used. To control the intensity, poly(ethylene terephthalate)
sheets were used as neutral density filters. The UV intensity
was measured using a UV power meter (TOPCON UVR-1) that
had previously been corrected against a thermopile meter (No.
30198E6, The Eppley Laboratory, Inc.). Because the detection
limit of the power meter was 1 µW‚cm-2, we used extrapolated
values below 10 µW‚cm-2 incident UV light. Illumination was
conducted at room temperature after equilibrium between
gaseous and adsorbed 2-propanol on the TiO2 thin film was
achieved, as evidenced by a constant 2-propanol concentration.
The 2-propanol concentration was measured using a GC
(Shimadzu Model GC-8A) equipped with a 5-m PEG1000
column and a flame ionization detector, using N2 as the carrier
gas. The CO2 concentration was measured using the same GC,
equipped with a 2-m Porapak-Q column, with a methanizer and
a flame ionization detector, also using N2 as the carrier gas.
The detection limits for both acetone and CO2 were ap-
proximately 0.1 ppmv.
All of the data were collected using the same TiO2 thin film
in order to avoid variations in activity caused by the difference
of the surface area and thickness. The photocatalytic activity
of the film was able to be fully regenerated by illumination
with 5 mW‚cm-2 UV light in fresh humid air for 120 min.
Results
Figure 1 shows the experimental data for the adsorption
isotherm in the form of an inverse plot, the weight of adsorbed
2-propanol on the TiO2 thin film (M/mg) vs the gaseous

2-Propanol Photocatalytic Decomposition Efficiency
J. Phys. Chem. A, Vol. 101, No. 43, 1997 8059
Figure 2. Concentration changes of gaseous 2-propanol, acetone, and
CO2 as a function of time in the decomposition of gaseous 2-propanol
(incident UV intensity, 45 µW‚cm-2; initial 2-propanol concentration,
100 ppmv; b 2-propanol, O acetone, 2 carbon dioxide).
Figure 4. Dependence of QY on absorbed photons (initial 2-propanol
concentration: b 1000 ppmv, 4 100 ppmv, 2 10 ppmv, O 1 ppmv).
Error bars for 100 ppmv are omitted for clarity.
the slope, i.e., the exponent R in the R ) KIR relation, is ca.
0.7-0.8. It is well-known in heterogeneous photocatalysis that,
under light-rich conditions, the reaction rate is often character-
ized by an R value of 0.5 due to domination by the second-
order-dependent carrier-recombination process.21,22 Conversely,
the light-limited reaction rate is represented as R ) 1.23,24 Thus
the present experimental regime is in a transition region between
the two asymptotic values. R values of ca. 0.7 were also
reported in the photodecomposition of formic acid by Hill et
al.19 and of acetone by Ollis et al.25 under several mW‚cm-2
UV irradiation. For 365 nm light, an incident UV light intensity
of 1 µW‚cm-2 corresponds to a photon flux of 1.8 × 1012
quanta‚cm-2‚s-1, of which the TiO2 film absorbs 1.2 × 1012
quanta‚cm-2‚s-1
. We calculated the initial rates using the
Figure 3. Dependence of the acetone generation rate on the absorbed
photons (initial 2-propanol concentration: b 1000 ppmv, 4 100 ppmv,
2 10 ppmv, O 1 ppmv).
conventional least-squares method over the first 1 h and used
the least-squares statistics to calculate the error ranges, based
on a 50% confidence level. Only for the plots in which
2-propanol was exponentially decomposed, i.e., at the highest
light intensity, the initial amount of acetone generated in 15
min was used as the initial rate.
2-propanol concentration (C/mg‚m-3). These data were ana-
lyzed in terms of a Langmuir-type isotherm, which is described
as follows
Figure 4 shows semilog plots of I versus the apparent QY
values for acetone generation. The apparent QY values were
calculated using the following equation:
1/M ) 1/µ + 1/(µTC)
(2)
where µ is the maximum weight of molecules in an adsorbed
monolayer, for a given TiO2 sample, and T is the adsorption
binding constant. The values of µ and T were ca. 0.33 mg and
0.000 63 m3‚mg-1, respectively. These results will be used in
the discussion of the QY values for 2-propanol decomposition
in the next section.
Figure 2 shows a typical experimental data set for the
concentration changes of gaseous 2-propanol, acetone, and CO2
as a function of time in the decomposition of gaseous 2-propanol
(incident UV intensity, 45 µW‚cm-2; initial 2-propanol con-
centration, 100 ppmv). After 60 min, when equilibrium between
gaseous and adsorbed 2-propanol on the TiO2 thin film had been
achieved, UV light illumination was initiated. The amount of
generated acetone was equivalent to that of the decomposed
2-propanol, and the generation of CO2 and other stable
intermediates was not detected, within experimental error.
Under the present experimental conditions, i.e., 1-1000 ppmv
initial 2-propanol concentration and 36 nW-45 µW‚cm-2
incident UV light intensity, only acetone was generated stoi-
chiometrically.
number of generated acetone molecules
QY )
(3)
number of absorbed photons
Error bars associated with the QY values were evaluated on
the basis of the total error of measuring the acetone concentra-
tion and extrapolating the value of the UV light intensity.
However, good repeatability, within 7% , was obtained even
for the lowest I values. The QY values increased gradually
with decreasing I and finally saturated for the highest initial
2-propanol concentration (1000 ppmv) for I less than 4 × 1011
quanta‚cm-2‚s-1, and thus a purely light-limited condition (R
) 1) was reached, where the maximum QY value was 27.8 (
2.5%. This maximum QY value also appeared to be approached
for lower initial concentrations. Moreover, the curve shapes
for the different initial 2-propanol concentrations are similar
regardless of the initial concentration. For a given QY value,
increasing the initial 2-propanol concentrations by a factor of
10 leads to an increased value of I. However, the increases
became smaller with increasing concentration: 10 times, 6 times,
and only 2 times for 1-10 ppmv, 10-100 ppmv, and 100-
1000 ppmv, respectively.
Figure 3 shows log-log plots of the acetone generation rates
(R) versus the number of absorbed photons (I). R decreased
with decreasing I, in essentially a linear fashion, with slopes
that were almost constant regardless of the initial 2-propanol
concentration. The higher the initial 2-propanol concentration,
the higher the value for the same UV intensity. The value of
Discussion
The mechanisms for the photocatalytic decomposition of
2-propanol to acetone has been described.16,26 The photocata-

8060 J. Phys. Chem. A, Vol. 101, No. 43, 1997
Ohko et al.
lytic processes prior to the initiation of 2-propanol decomposi-
tion are well-known:
hν f h+ + e-
(4)
h+ + OH-(s) f •OH, or h+ + H2O f •OH + H+ (5)
•-
e- + O2(s) f O2
(6)
(7)
O2•- + H+ f HO2
•
•
The generated OH radical reacts with 2-propanol, abstracting
its hydrogen atom to form a radical:
CH3CH(OH)CH3 + •OH f CH3C‚(OH)CH3 + H2O (8)
This CH3C•(OH)CH3 radical is decomposed to acetone through
several reaction pathways. One of these is the so-called current-
doubling reaction, expressed by
CH3C•(OH)CH3 f CH3COCH3 + H+ + e-(C.B.) (9)
where e- (C.B.) represents an electron in the TiO2 conduction
band.27 Under ambient conditions, however, O2 can attack the
radical, producing an unstable peroxo radical, which decomposes
to acetone:
CH3C•(OH)CH3 + O2 f CH3COO•(OH)CH3 (10)
-
CH3COO•(OH)CH3 f CH3COCH3 + H+ + O2 (11)
Figure 5. Schematic diagram of a series of photocatalytic processes
along the time axis. The time interval of excitation by photons in this
figure is assumed to be that for an experimental condition of 1 µW‚cm-2
incident UV intensity and 1000 ppmv initial 2-propanol concentration.
•
In the same manner, HO2 radicals generated via eqs 6 and 7
may also react with the CH3C•(OH)CH3 radical
•
CH3C•(OH)CH3 + HO2 f CH3COOH(OH)CH3 (12)
oxidation occurs in on the order of 100 ns. Electron-hole
recombination proceeds in the 10-100 ns range.34,35 Electron
transfer to O2 molecules is usually assumed to proceed more
slowly20,35 but still much faster than the time interval of the
overall reaction frequency. These processes are schematically
summarized in Figure 5. As seen in Figure 5, the charge-
separation efficiency must be independent of I. Therefore, the
maximum QY value of 28% can be considered to represent the
intrinsic charge-separation efficiency of this TiO2 film. The
remaining 72% must be converted into heat in a charge-
recombination process, as shown below.
CH3COOH(OH)CH3 f CH3COCH3 + H2O2 (13)
Note that there is no chain reaction involved in the above
processes. Overall, only one photon participates in generating
one molecule of acetone, and therefore QY values were
calculated using eq 3. Considering that no radical chain
reactions are involved, the maximum QY value obtained (28%)
is very high. We have previously reported that QY values for
the decomposition of gaseous acetaldehyde ranged up to nearly
100%,12-14 but this reaction involves chain reactions.28
In the present experiments, in which very low-intensity UV
light was used, the frequency of the reaction of each adsorbed
molecule with photoproduced reactive species such as OH is
h+ + e- f heat
(14)
•
very low. Let us estimate the frequency, for example, for the
case of a light intensity of 1 µW‚cm-2 and a gaseous 2-propanol
concentration of 1000 ppmv. From the Langmuir adsorption
isotherm in Figure 1, the area occupied by each molecule is
estimated to be 0.3 nm2. One µW‚cm-2 of 365 nm UV light
corresponds to 2 × 1012 quanta‚cm-2‚s-1. Therefore on the
average, each 2-propanol molecule can encounter a photogen-
On a photoirradiated TiO2 surface, various types of active
oxygen species exist. These species react with each other,
forming stable products, as follows:
•
•OH + HO2 f H2O + O2
(15)
(16)
•
•OH + •OH f H2O2
erated OH radical only once every ∼103 s if a single photon
always generates a •OH radical. Conversely, the time scale of
one series of photocatalytic processes including charge-separa-
tion and charge-transfer processes is very short. The time scale
of electron-hole pair generation is on the order of 100 fs.29,30
Thus, hole-trapping and electron-trapping at the TiO2 surface,
in other words, the charge-transfer processes to the surface, are
completed in the picosecond-nanosecond region.31,32 For
example, electron migration to the TiO2 surface is estimated to
be 0.83 ns using D ) 0.02 cm2‚s-1 in a 100-nm-TiO2 particle.33
•OH radicals are formed in on the order of 10 ns, and 2-propanol
Therefore, when the 2-propanol concentration is low, these
reactions predominate. Conversely, with decreasing light
intensity, these recombination reactions proceed less efficiently,
and the QY values for 2-propanol decomposition increase.
•
Because •OH radicals are much more reactive than HO2
•
radicals, the steady-state OH concentration is probably much
•
•
less than that of HO2 . Moreover, the reaction rates of OH
with HO2• and with •OH were estimated to be 1.1 × 10-10 and
1.8 × 10-12 cm3‚mol-1‚s-1, respectively.36 Therefore, we can

2-Propanol Photocatalytic Decomposition Efficiency
J. Phys. Chem. A, Vol. 101, No. 43, 1997 8061
TABLE 1: Correlation of the Initial 2-Propanol
Concentration, Its Adsorbed Amount, and Absorbed Photons
at a QY Value of 15%
initial
concentration of
2-propanol/ppmv
adsorbed amount
of 2-propanol/
adsorbed photons/
molecules‚cm-2
quanta‚cm-2‚s-1
1000
100
10
4.4 × 1016
9.6 × 1015
1.1 × 1015
1.1 × 1014
1 × 1013
3 × 1012
5 × 1011
5 × 1010
1
assume that reaction 15 is the main recombination process on
the TiO2 surface.
Next, let us consider the variation of the QY-absorbed photon
curves on the reactant concentration in Figure 4. When the
reactant concentration in the gas phase was decreased, the curve
shifted to the lower light intensity direction. If we assume that
both incident photon and reactant molecule flux arriving at the
TiO2 surface react immediately, the QY values should be
determined by the ratio of the light intensity to the gas
concentration. In other words, when the gas concentration
increases by a factor of 10, the same QY value should be
obtained with a factor of 10 higher light intensity. As can be
seen in Figure 4, however, the experimental results do not
support this model, showing that the factor correlating the QY
values with the photon flux is not the flux of 2-propanol
molecules under the present experimental conditions.
It is reasonable to consider that the adsorbed amount of
2-propanol is more important in determining the QY values.
For example, in Table 1 are shown the initial 2-propanol
concentrations (ppmv), the amounts of adsorbed 2-propanol
molecules on TiO2 (molecules‚cm-2), and the numbers of
absorbed photons (quanta‚cm-2‚s-1) at a QY of 15% , as
determined from Figure 4. The adsorbed amounts of 2-propanol
were estimated from the Langmuir isotherm in Figure 1. A
good proportionality between the amount of adsorbed 2-propanol
molecules and the number of absorbed photons is suggested in
this table. Even over the wide range of QY values shown in
Figure 4, this relationship is satisfied. In other words, QY values
appear to be determined by the ratio between the number of
adsorbed 2-propanol molecules ([M]ad) and the number of
absorbed photons ([photon]ab).
Figure 6. Plots of QY values vs log Inorm (/s-1), a parameter that is
defined as the ratio of the number of absorbed photons to the number
of adsorbed 2-propanol molecules (initial 2-propanol concentration: b
1000 ppmv, 4 100 ppmv, 2 10 ppmv, O 1 ppmv).
•
possible diffusion length of either OH radicals or 2-propanol
may be at least 11 nm. However, when the average 2-propanol
•
intermolecular distance becomes greater than the OH radical
diffusion length, the QY values can decrease against totally Inorm
in Figure 6.
In the present study, we used a pure anatase TiO2 sintered
thin film, which has sufficient surface adsorbed water and
oxygen10 for electron-hole pairs to transfer at the TiO2 interface.
However, some lattice defects or doping transition metals exist
in the bulk of TiO2. These behave as trapping sites or
recombination sites for electron-hole pairs.34,37,38 Thus the
variable amounts of these in different samples could influence
the QY values vs Inorm in Figure 6. In addition, the variation
of the amounts of surface hydroxyl groups, water, and oxygen
molecules can also influence the charge-transfer process at the
•
surface, and thus the OH radical diffusion distance.
Conclusion
From the present kinetic study of the photocatalytic decom-
position efficiency of gaseous 2-propanol using purely anatase
TiO2 sintered thin film under very weak UV light, it can be
concluded for the first time that QY values are determined by
the ratio of the number of adsorbed 2-propanol molecules to
the number of absorbed photons. This phenomenon indicates
Here we define the normalized photon number (Inorm) as
[photon]ab divided by [M]ad and have replotted the QY values
as a function of Inorm in Figure 5. Despite the wide range of
different initial concentrations of 2-propanol, the plots fall on
the same curve. The value of QY increases as Inorm decreases
and finally becomes constant at 28% for Inorm values below 10-4
(s-1). Because [photon]ab is defined as the number of photons
being absorbed in 1 s, Inorm has the dimension of s-1 in Figure
5.
•
that either OH radicals or 2-propanol can diffuse on the TiO2
surface at least ca. 11 nm. The decrease in QY is attributed to
•
•
increases in the rates of reaction for OH radicals with HO2
•
radicals and OH with itself, relative to that with 2-propanol.
The maximum QY value of 28% represents the intrinsic charge-
separation efficiency of this sample. It is interesting that we
can apply this reaction dynamics for such a wide 2-propanol
concentration range. We believe that these findings can become
significant models for photocatalysis involving more complex
reactions, for example, in the case of reactants that are easily
oxidized via radical chain reactions.
This result indicates that either reactive species (•OH) or
reactant (2-propanol) diffuses on the TiO2 surface and the
decomposition reaction efficiency is determined by the collision
probability of these species. Moreover, it is suggested that the
oxidation rate of 2-propanol by •OH (eq 7) is much faster than
the deactivation rates of •OH (eqs 14 and 15). In other words,
in the region where the QY value is constant with respect to
Acknowledgment. We express gratitude to Dr. D. A. Tryk
for careful reading of the manuscript. We are grateful to the
Ministry of Education, Science, and Culture for financial
support. We are grateful also to Ishihara Sangyo Kaisha, Ltd.,
for suggesting the method for the roughness factor measurement
of the TiO2 thin film.
•
Inorm, the OH produced by one photon always reacts with
2-propanol, not with either HO2• or •OH. From Figure 6, for 1
ppmv, i.e., of the lowest initial 2-propanol concentration in this
study, we might also expect that QY approaches 28%. On the
basis of this concentration, an intermolecular distance of
adsorbed 2-propanol of ca. 11 nm can be calculated from the
adsorption isotherm in Figure 1, making use of the surface area
of this film (roughness factor of 150 cm2‚cm-2) Therefore the
References and Notes
(1) Photocatalysis Fundamentals and Applications; Serpone, N., Pel-
izzetti, E., Eds.; Wiley-Interscience: Amsterdam, 1989.
(2) Photocatalytic Purification and Treatment of Water and Air; Ollis,
D. E., Al-Ekabi, H., Eds.; Elsevier: Amsterdam, 1993.

8062 J. Phys. Chem. A, Vol. 101, No. 43, 1997
Ohko et al.
(3) See, e.g.: Abstracts of papers; The First International Conference
on Advanced Oxidation Technologies for Water and Air Remediation,
London, Ontario, Canada, June 25-30, 1994; Elsevier: Amsterdam, 1994.
(4) Fujishima, A.; Honda, K. Nature 1972, 238, 37.
(5) Bickley, R. I.; Stone, F. S. J. Catal. 1973, 31, 389.
(6) Izumi, I.; Dunn, W. W.; Wilbourn, K. O.; Fan, F. F.; Bard, A. J.
J. Phys. Chem. 1980, 84, 3207.
(21) D’Oleivera, J.; Al-Sayyed, G.; Pichat, P. EnViron. Sci. Technol.
1990, 24, 990.
(22) Sitkiewitz, S.; Heller, A. New J. Chem. 1996, 20, 233.
(23) Egerton, T. A.; King, C. J. J. Oil Colour Chem. Assoc. 1979, 62,
386.
(24) Okamoto, K.; Yamamoto, Y.; Taneka, H.; Itaya, A. Bull. Chem.
Soc. Jpn. 1985, 58, 2023.
(7) Rosenberg, I.; Brock, J. R.; Heller, A. J. Phys. Chem. 1992, 96,
3423.
(8) Matthews, R. W. In Photocatalytic Purification and Treatment of
Water and Air; Ollis, D. E., Al-Ekabi, H., Eds.; Elsevier: Amsterdam, 1993.
(9) Aguado, M. A.; Anderson, M. A. Sol. Energy Mater. Sol. Cells
1993, 28, 345.
(25) Peral, J.; Ollis, D. F. J. Catal. 1992, 136, 554.
(26) Harvey, P. R.; Rudham, R.; Ward, S. J. Chem. Soc., Faraday Trans.
1983, 1, 1381.
(27) Yamagata, S. Ph.D. Thesis, University of Tokyo, 1989.
(28) Sopyan, I.; Watanabe, M.; Murasawa, S.; Hashimoto, K.; Fujishima,
A. J. Photochem. Photobiol. A 1996, 98, 79.
(10) Raupp, G. R.; Junio, C. T. Appl. Surf. Sci. 1993, 72, 321.
(11) Takeda, N.; Torimoto, T.; Sampath, S.; Kuwabata, S.; Yoneyama,
H. J. Phys. Chem. 1995, 99, 9986.
(12) Matsubara, H.; Takada, M.; Koyama, S.; Hashimoto, K.; Fujishima,
A. Chem. Lett. 1995, 9, 767.
(13) Sopyan, I.; Murasawa, S.; Hashimoto, K.; Fujishima, A. Chem. Lett.
1994, 723.
(14) Sopyan, I.; Watanabe, M.; Murasawa, S.; Hashimoto, K.; Fujishima,
A. Chem. Lett. 1996, 69.
(15) Kikuchi, Y.; Sunada, K.; Iyoda, T.; Hashimoto, K.; Fujishima, A.
J. Photochem. Photobiol., in press.
(16) Ikeda, K.; Hashimoto, K.; Fujishima, A. J. Phys. Chem. B 1997,
101, 2617.
(17) Schwitzgebel, J.; Ekerdt, J. G.; Gerischer, H.; Heller, A. J. Phys.
Chem. 1995, 99, 5633.
(18) Augugliaro, V.; Palmisano, L.; Schiavello, M. AIChE. J. 1991, 37,
1096.
(29) Colombo, D. P. J.; Roussel, K. A.; Saeh, J.; Skinner, D. E.; Cavaleri,
J. J.; Bowman, R. M. Chem. Phys. Lett. 1995, 232, 207.
(30) Skinner, D. E.; Colombo, D. P.; Cavaleri, J. J.; Bowman, R. M. J.
Phys. Chem. 1995, 99, 7853.
(31) Serpone, N.; Lawless, D.; Khairutdinov, R.; Pelizzetti, E. J. Phys.
Chem. 1995, 99, 16655.
(32) Lepore, G. P.; Langford, C. H.; Vichova, J.; Vlcek, A. J. J.
Photochem. Photobiol. A 1993, 75, 67.
(33) Gratzel, M. In Photocatalysis Fundamentals and Applications;
Serpone, N., Pelizzetti, E., Eds.; Wiley-Interscience: Amsterdam, 1989.
(34) Martin, S. T.; Herrmann, H.; Choi, W. Y.; Hoffmann, M. R. J.
Chem. Soc., Faraday Trans. 1994, 90, 3315.
(35) Martin, S. T.; Herrmann, H.; Hoffmann, M. R. J. Chem. Soc.,
Faraday Trans. 1994, 90, 3323.
(36) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F. J.; Kerr,
J. A.; Tore, J. J. Phys. Chem. Ref. Data 1989, 18, 932.
(37) Sclafani, A.; Palmisano, L.; Davi, E. J. Photochem. Photobiol. A
1991, 56, 113.
(19) Aguado, M. A.; Anderson, M. A.; Hill, C. G. J. Mol. Catal. 1994,
89, 165.
(20) Sun, L. Z.; Bolton, J. R. J. Phys. Chem. 1996, 100, 4127.
(38) Palmisano, L.; Augugliaro, V.; Sclafani, A.; Schiavello, M. J. Phys.
Chem. 1988, 92, 6710.