Effect of diffusional film on formation of hydrogen peroxide in photocatalytic reactions

Article abstract of DOI:10.1021/jp047873q

We recently found that photocatalytic reactions produce H₂O ₂ at a concentration level of ppm (parts per million) from water. For some reason, however, other researchers have paid little attention to the formation of H₂O₂ in this reaction. Of course, one of the reasons is due to low sensitivities of the conventional methods employed to measure the H₂O₂ concentration. However, we suppose that there is another reason which makes it difficult to measure it. To elucidate this reason, therefore, an effect of linear liquid velocity on the photocatalytic formation of H₂O₂ from water is experimentally investigated, and an explanation for the experimental result is made using a mathematical model that takes into consideration the formation and decomposition of H₂O₂ and its diffusion through a film in the neighborhood of the photocatalyst surface into a bulk liquid. This effect is also discussed on the formation of H₂O₂ from an aqueous solution of formic acid. As a result, it is concluded that the diffusion film is responsible for the reduction in the H₂O₂ concentration.

Full text of DOI:10.1021/jp047873q

J. Phys. Chem. A 2004, 108, 10491-10496
10491
Effect of Diffusional Film on Formation of Hydrogen Peroxide in Photocatalytic Reactions
Fumihide Shiraishi* and Chinami Kawanishi
Department of Biochemical Systems Engineering, Faculty of Computer Science and Engineering,
Kyushu Institute of Technology, 680-4, Kawazu, Iizuka 820-8502, Japan
ReceiVed: May 18, 2004; In Final Form: September 14, 2004
We recently found that photocatalytic reactions produce H2O2 at a concentration level of ppm (parts per
million) from water. For some reason, however, other researchers have paid little attention to the formation
of H2O2 in this reaction. Of course, one of the reasons is due to low sensitivities of the conventional methods
employed to measure the H2O2 concentration. However, we suppose that there is another reason which makes
it difficult to measure it. To elucidate this reason, therefore, an effect of linear liquid velocity on the
photocatalytic formation of H2O2 from water is experimentally investigated, and an explanation for the
experimental result is made using a mathematical model that takes into consideration the formation and
decomposition of H2O2 and its diffusion through a film in the neighborhood of the photocatalyst surface into
a bulk liquid. This effect is also discussed on the formation of H2O2 from an aqueous solution of formic acid.
As a result, it is concluded that the diffusion film is responsible for the reduction in the H2O2 concentration.
1. Introduction
invoked experimentally by Matthews12 and then commented by
Turchi and Ollis13 from the standpoint of photocatalytic reactor
design. However, no detailed procedure for mathematical
analysis is available so far.
Let us consider the following reaction that produces H2O2
from H2O on the photocatalyst surface:
A photocatalytic reaction is constituted by the formation of
several kinds of radicals.1-3,4 The mechanism of this reaction
includes the formation of hydrogen peroxide, H2O2.5-7 In fact,
it is known that the photocatalytic reactions using TiO2,8,9 TiO2-
CoIITSP,9 ZnO,8 and so forth produce H2O2. We also have
recently found that the photocatalytic reactions using TiO2 and
Pd-TiO2 produce H2O2 at a concentration level of ppm (parts
per million) from water,10 and the initial rates of formation of
H2O2 are in proportion to the initial rates of decomposition of
formaldehyde (HCHO) and formic acid (HCOOH), i.e., the
photocatalytic activities. For example, Pd-TiO2 decomposed
several organic compounds and produced H2O2 more quickly
than did TiO2.11 Thus, it is clear that the photocatalytic activity
is closely related to the formation of H2O2.
k1
1
2
H2O + O2
98 H2O2
(1)
Assuming that the dissolved oxygen concentration in the water
is sufficiently high and kept constant because of continuous
supply from the air in a batch-recirculation reactor, the rate of
formation of H2O2 on the photocatalyst surface, rH O ,i (mol m-2
-
2
2
surface min-1), is expressed as
For some reason, however, other researchers have paid little
attention to the formation of H2O2 in the photocatalytic reaction.
Of course, this is first due to the fact that the conventional
methods employed to measure the concentration of a trace of
H2O2 formed during the photocatalytic reaction were insensitive.
Fortunately, we could successfully measure the time course of
the H2O2 concentration by use of the enzyme peroxidase that
catalyzes a coloring reaction.10 On the other hand, we suppose
that there is another reason which makes it difficult to measure
the H2O2 concentration in the photocatalytic reaction.
In the present work, therefore, we first investigate experi-
mentally an effect of linear flow velocity on the photocatalytic
formation of H2O2 from water and then attempt to explain the
experimental result using a mathematical model that takes into
consideration the formation and decomposition of H2O2 and its
diffusion through a film formed in the neighborhood of the
photocatalyst surface into a bulk liquid. Similarly, a formation
of H2O2 from an aqueous solution of HCOOH is investigated.
k1KHC
H2O,i
rH O ,i
)
(2)
2
2
1 + KHCH O,i
2
where k1 (mol m-3 min-1) is the kinetic constant, KH is the
adsorption equilibrium constant (mol-1 m3), and CH O,i is the
2
concentration of H2O at the photocatalyst surface (mol m-3).
For convenience, we will express the concentration of a given
chemical substance at the photocatalyst surface in units of moles
per volume. Since the concentration of H2O is extremely high,
eq 2 is simplified to
r
H O ,i = k1
(3)
2
2
which means that the formation of H2O2 proceeds with zeroth-
order kinetics.
On the other hand, when conducting the photocatalytic
reaction in an aqueous solution, a diffusion film is commonly
formed in the neighborhood of the photocatalyst surface, as
illustrated in Figure 1. Reactant molecules must diffuse through
the film to the photocatalyst surface, which lowers the rate of
the photocatalytic reaction. Therefore, the overall rate of the
photocatalytic reaction must be described by taking into
consideration the effect of film-diffusional resistance. Moreover,
2. Theory
2.1. Photocatalytic Formation of H2O2 from H2O. A mass-
transfer limitation to the photocatalytic reaction has first been
* Corresponding author. Telephone: +948-29-7827. Fax: +948-29-
7801. E-mail: fumi@bse.kyutech.ac.jp.
10.1021/jp047873q CCC: $27.50 © 2004 American Chemical Society
Published on Web 10/27/2004

10492 J. Phys. Chem. A, Vol. 108, No. 47, 2004
Shiraishi and Kawanishi
coefficient involved in eq 8 from the experimental data. As
described later, we conducted a series of the photocatalytic
reaction experiments in a batch-recirculation reactor, as de-
scribed elsewhere,10,15 where the reactant mixture was recircu-
lated at a constant flow velocity between the annular-flow
reactor and perfectly mixed flow vessel. In the following,
therefore, we will describe the method for determination of
kinetic parameters and mass-transfer coefficient on the assump-
tion of using the experimental data obtained in the batch-
recirculation reactions.
From eq 8, it is obvious that the initial rate of formation of
H2O2 from H2O is written as
dC
k1 - k2C
H2O2,b
|
|
|
|
|
|
H2O2,b
rH O ,L t)0
|
)
) (S/VL)
)
2
2
dt
1 + k2/kL
t)0
t)0
k1
(S/VL)
(9)
1 + k2/kL
Figure 1. Distribution of hydrogen peroxide concentration in neighbor-
hood of photocatalyst surface.
Since the mass-transfer coefficient increases toward the infinity
as the recirculation flow rate (or linear velocity) increases, eq
9 becomes
the decomposition of H2O2 proceeds in parallel with its
formation in the very neighborhood of the photocatalyst surface
where various kinds of radical species exist at high concentra-
tions. Consequently, we simply describe the formation and
decomposition of H2O2 as
rH O ,L t)0
|
) (S/VL)k1
(10)
2
2
which indicates that the initial rates of formation of H2O2
measured at different linear velocities become constant at a
sufficiently large linear velocity. Consequently, the value of k1
can be determined from the measured values.
Similarly, when H2O2 is added to H2O to decompose under
irradiation with UV light recirculating the solution at a suf-
ficiently high linear velocity, eq 8 gives the rate of the
decomposition as
k1
1
2
H2O + O2 {
\
} H2O2
(4)
k2
When the system is at equilibrium, the rate of formation of H2O2
on the photocatalyst surface, rH O ,i is identical to the rate of
2
2
diffusion of H2O2 through the film, NH O (mol m-2-surface h-1),
2
2
dC
so that14
H2O2,b
-rH O ,L ) -
) (S/VL)(k2CH O ,b - k1) (11)
2 2
2
2
dt
NH O ) kL(CH O ,i - CH O ,b) ) k1 - k2CH O ,i ) rH O ,i (5)
2
2
2
2
2
2
2
2
2 2
which gives the following relationship between the initial rate
of decomposition of H2O2 and the initial concentration of H2O2.
where CH O ,i and CH O ,b express the concentrations of H2O2
2
2
2 2
(mol m-3) at the photocatalyst surface and in the bulk liquid,
respectively, and kL is the mass-transfer coefficient (m h-1).
Solving eq 5 with respect to CH O ,i leads to
dC
|
H2O2,b
|
-rH O ,L t)0
|
) -
) (S/VL)(k2CH O ,b0 - k1)
2 2
|
2
2
dt
t)0
2
2
(12)
C
+ k1/kL
Thus, the initial rates of photocatalytic decomposition of H2O2
are measured at different initial concentrations of H2O2, and
these values and the value of k1 already determined are applied
to eq 12. As a result, the value of k2 is obtained.
H2O2,b
CH O ,i
)
(6)
2
2
1 + k2/kL
Therefore, the rate of formation of H2O2 on the photocatalyst
surface is described as a function of only CH O ,b by
Furthermore, solving eq 9 for kL leads to
2
2
rH O ,L t)0 k2
|
2
2
k1 - k2C
H2O2,b
kL )
(13)
r
)
(7)
(S/VL)k1 - r
|
H2O2,i
H2O2,L t)0
1 + k2/kL
Therefore, the values of k1 and k2 determined above and the
Consequently, when a reaction mixture with a liquid volume
of VL (m3) is treated by the photocatalyst with an external surface
of S (m2-surface) irradiated with UV light, the rate of photo-
values of rH O ,L t)0 measured at different linear velocities, uz
|
2
2
(m h-1), are applied to eq 13. The values of kL thus calculated
are plotted against their respective linear velocities. In general,
the relationship between kL and uz is expressed as16,17
catalytic reaction on a liquid volume basis, rH O ,L (mol m-3
-
2
2
liquid min-1), is given as
kL ) auzb
(14)
dC
k1 - k2C
H2O2,b
H2O2,b
rH O ,L
)
) (S/VL)
(8)
Thus, eq 14 is applied to the plotted data in order to determine
the parameter values by the least-squares method.
2
2
dt
1 + k2/kL
2.2. Method for Determination of Kinetic Parameters and
Mass-Transfer Coefficient. Before calculating the time course
of the H2O2 concentration in the photocatalytic reaction, it is
necessary to determine the kinetic parameters and mass-transfer
3. Experimental Section
3.1. Materials and Reagents. Titanium tetraisopropozide
(TIP) was purchased from Katayama Chemicals Co., Ltd.

Effect of Diffusional Film
J. Phys. Chem. A, Vol. 108, No. 47, 2004 10493
Figure 2. Annular-flow photocatalytic reactor.
(Tokyo). Isopropyl alcohol (IPA) was obtained form Wako Pure
Chemical Industries Ltd. (Osaka). A 31% (v/v) aqueous H2O2
solution was a product of Mitsubishi Chemical Co., Ltd.
(Tokyo). Glucose B test Wako (Wako Pure Chemical Industries
Ltd.) was used to measure the H2O2 concentration. Two kinds
of 6 W UV lamps were used as a light source; black light blue
fluorescent lamp (FL6BL-B; Matsushita Denki, Tokyo) and
germicidal lamp (GL-6; Sankyo Denki, Tokyo). The wavelength
distributions of UV lights emitted from these lamps are available
elsewhere.18,19
3.2. Operational Method of the Photocatalytic Reactor.
Amorphous titanium dioxide was prepared from TIP and used
to coat the surface of a glass tube according to the procedure
described elsewhere.20,21 As illustrated in Figure 2, the photo-
catalytic reactor has an annular structure,21-23 consisting of a
light source protected by a quartz glass tube (24 mm in outer
diameter and 220 mm long) centered along the main axis of a
plastic cylinder and a Pyrex glass tube (28 mm in inner diameter,
140 mm long, and 1.8 mm in thickness), whose inner surface
is coated with a thin film of photocatalyst (surface area 0.0123
m2), which is inserted and fixed so as to closely contact with
the outer surface of the plastic cylinder. A reactant mixture was
allowed to flow through the annulus of 2 mm in width between
the quartz and Pyrex glass tubes. The inlet volume of the reactor
was about 7 × 10-5 m3.
Figure 3. Batch-recirculation reactor system.
was measured at 505 nm with a spectrophotometer (UV-240;
Shimazu, Tokyo) to determine the H2O2 concentration.
The photocatalytic reactor was circularly connected with a
mixed-flow vessel and a peristaltic pump (RP-1000; EYELA,
Tokyo), as shown in Figure 3. Distilled water of 3.0 × 10-4
m3 with or without formic acid was poured into the mixed-
flow vessel and circulated at a constant flow rate. The light
source was switched on to start the photocatalytic reaction and
a small amount of reaction mixture was withdrawn at appropriate
intervals from the mixed-flow vessel. All the experiments were
conducted by placing the experimental apparatus in a constant-
temperature chamber to keep the liquid temperature constant
(21 °C) and to shut out all unnecessary light.
4. Results and Discussion
4.1. H2O System. 4.1.1. Relationship between the Initial Rate
of Formation of H2O2 and Linear Velocity. Table 1 summarizes
the kinetic scheme of the photocatalytic reaction described in a
number of papers.5-7 There exist two elementary reactions that
produce H2O2. We experimentally elucidated that the photo-
catalyst can produce H2O2 from distilled water. However, the
rate of its formation sooner or later slowed and became constant,
suggesting that the decomposition of H2O2 takes place in parallel
with its decomposition.
3.3. Measurement of H2O2 Concentration. Glucose B Test
Wako (3.0 × 10-6 m3) was added to the reaction mixture (2.0
× 10-6 m3) withdrawn from the mixed-flow vessel to cause a
coloring reaction. After 20 min, the absorbance of the solution
It is considered that this decomposition is caused by the highly
active radicals formed during the photocatalytic reaction as well

10494 J. Phys. Chem. A, Vol. 108, No. 47, 2004
Shiraishi and Kawanishi
TABLE 1: Formation of Radicals in a Photocatalytic
Reaction
germicidal lamp. The initial rate of formation of H2O2 initially
increases linearly with the linear velocity and then plateaus. The
Reynolds number in the present system lies in the range of
24
66.7-233 for the linear velocity of 2-7 m min-1, indicating
that the liquid flow in the photocatalytic reactor is a laminar
flow. A comparison at the same linear velocity indicates that
the germicidal lamp, emitting more powerful UV light with a
shorter wavelength, can produce H2O2 more rapidly than does
the black light blue fluorescent lamp. The initial rate of
formation of H2O2 with the black light blue fluorescent lamp
plateaus in the region of lower linear velocity because of its
lower rate of formation.
Since the thickness of the film formed in the very neighbor-
hood of the photocatalyst surface decreases with the increase
of the linear velocity,17 we suppose that the film must have
strongly affected the amount of H2O2 produced. In the following,
therefore, this will be undertaken to explain using the math-
ematical model described above.
4.1.2. Determination of Kinetic Parameters and Mass-
Transfer Coefficient. Figure 4, where the germicidal lamp was
employed as a light source, clearly indicates that the initial rate
of formation of H2O2 becomes almost constant above a certain
value of linear velocity. From the initial rate of formation in
this region, the value of k1 was determined to be 6.24 × 10-6
mol m-2 min-1. When decomposition of H2O2 in an aqueous
solution by UV light is carried out at a sufficiently large linear
velocity, the initial rate of decomposition is expressed by eq
12, which takes into consideration the decomposition of H2O2
in the film. Therefore, H2O2 at an initial concentration of 3.21
× 10-2 mol m-3 was decomposed using the photocatalyst
irradiated with the UV light from the germicidal lamp. The result
is shown in Figure 5. From these experimental data, the initial
as the UV light, which terminates increasing concentration of
H2O2 at a certain low level. As a result, it becomes difficult to
monitor the time course of the H2O2 concentration in the
photocatalytic reaction. According to our measurement,21 the
concentration of H2O2 formed is on the order of 10-4-10-3
mol m-3 at the maximum. We were able to accurately measure
the concentration of H2O2 at such a low level by introducing
an enzymatic method. Another advantage of using this enzy-
matic method is the substrate specificity of the enzyme, in which
the enzyme acts on only H2O2. Thus, the true concentration can
be measured. However, would the low sensitivities of the
measurement methods be only a reason for the fact that little
measurement of H2O2 has been done in the past photocatalytic
reaction studies? A comparison shows that most of the photo-
catalytic reaction experiments reported by other researchers have
been performed at low linear velocities compared with our
experiment, so that we suppose that this experimental condition
may cause a difficulty in the measurement of H2O2 concentration
in the photocatalytic reaction. In the following, therefore, the
effect of linear velocity on the photocatalytic reaction will be
investigated both experimentally and theoretically.
Figure 5. Time courses of hydrogen peroxide concentration in
decomposition of hydrogen peroxide by UV lights.
rate of decomposition, r
H O ,L|t)0, was determined to be 2.93 ×
2
2
10-4 mol m-3 min-1. Furthermore, this value and the value of
k1 determined above were applied to eq 12, so that k2 was
determined to be 4.17 × 10-4 mol m-2 min-1. The values of k1
and k2 and the experimental data in Figure 4 were further applied
to eq 13 to obtain the value of kL for each linear velocity. The
values of kL thus determined are plotted against their respective
linear velocities, as shown in Figure 6. Equation 14 was applied
to the data in Figure 6 and the parameters included in the
Figure 4. Effect of linear velocity on initial rate of formation of
hydrogen peroxide.
Figure 4 shows the relationship between the initial rate of
formation of H2O2 and the linear velocity when the surface of
the photocatalyst immersed in the distilled water was irradiated
with the UV light from a black light fluorescent lamp and

Effect of Diffusional Film
J. Phys. Chem. A, Vol. 108, No. 47, 2004 10495
Figure 6. Effect of linear velocity on mass-transfer coefficients
determined from experimental data of photocatalytic reactions.
Figure 7. Comparison between initial rates of formation of hydrogen
peroxide from water and an aqueous solution of formic acid (light
source: germicidal lamp).
TABLE 2: Kinetic Constants and Equations of
Mass-Transfer Coefficient
on the photocatalyst irradiated with the UV light from the
germicidal lamp is three times larger throughout the whole
reaction time. It should be noted that the parameter values in
the equations of mass-transfer coefficient for the germicidal lamp
and black light blue fluorescent lamp are almost the same. This
result is reasonable because the same reactor with either black
light blue fluorescent lamp or germicidal lamp was used to carry
out the photocatalytic reactions under the same liquid flow
condition so that the mass-transfer coefficients must be funda-
mentally the same.
4.1.3. Comparison of the Calculated Results. The kinetic
parameters and mass-transfer coefficients determined above
were applied to eq 9 in order to determine the relationship
between the kinetic parameters and mass-transfer coefficients.
The calculated results for the germicidal lamp and black light
blue fluorescent lamp are shown by solid and broken lines,
respectively, in Figure 4. The calculated lines are in good
agreement with their respective experimental values, indicating
that there is a close relationship between the film formed in the
neighborhood of the photocatalyst surface and the observed
result that the initial rate of formation of H2O2 initially increases
with the linear velocity and then plateaus.
black light blue
germicidal lamp
fluorescent lamp
k1 (mol m-2 min-1
k2 (mol m-2 min-1
)
)
6.24 × 10-6
1.71 × 10-3 u0.621
1.87 × 10-6
1.85 × 10-3 u0.538
4.17 × 10-4
1.29 × 10-4
kL (m min-1
)
equation were determined by the least-squares method. As a
result, the following equation was obtained.
kL ) 1.71 × 10-3u0.621
(15)
The same procedure was applied to the experimental data in
Figure 4, where the black light blue fluorescent lamp was used
as a light source, and the value of k1 was determined to be 1.87
× 10-6 mol m-2 min-1. From the experimental result for the
decomposition of H2O2 at an initial concentration of 2.63 ×
10-2 mol m-3 (Figure 5), the value of rH O ,L t)0
was determined
|
2
2
to be 6.17 × 10-5 mol m-3 min-1. Furthermore, this value and
the value of k1 determined above was applied to eq 12 and the
value of k2 was determined to be 1.29 × 10-4 mol m-2 min-1
.
The values of k1 and k2 and the data in Figure 4 were further
applied to eq 13 in order to determine the value of kL for each
linear velocity. The values of kL thus determined are plotted
against their respective linear velocities, as shown in Figure 6.
Equation 14 was therefore applied to the data in Figure 6 and
the parameters involved in the equation were determined by
the least-squares method. As a result, the following equation
was obtained.
4.2. Formic Acid-H2O System. In a previous paper,21 we
have revealed that H2O2 is also formed in the process of the
photocatalytic decomposition of formic acid. Therefore, the
H2O2 concentration becomes higher in the photocatalytic
treatment of H2O containing formic acid than in the photocata-
lytic treatment of H2O alone. Figures 7 and 8 show the
relationships between the initial rate of formation of H2O2 and
the linear velocity when H2O containing formic acid at an initial
concentration of 0.45 mol m-3 was photocatalytically treated
using the germicidal lamp and black light blue fluorescent lamp,
respectively. For comparison, each figure also includes the initial
rate of formation of H2O2 in the photocatalytic treatment of H2O
alone. It is clear that despite the kind of the light source, the
initial rate of formation is remarkably increased in the presence
of formic acid. This is explained as follows. The intermediate
product HCOOH is known to be a scavenger of the hole h+ in
the recombination reaction described as h+ + e- f heat and
kL ) 1.85 × 10-3u0.538
(16)
The parameter values and mass-transfer coefficients for the
two light sources are compared in Table 2. The kinetic constant
for the formation of H2O2, k1, with the germicidal lamp is 3.3
times larger than that with the black light blue fluorescent lamp.
Similarly, the kinetic constant for the decomposition of H2O2,
k2, with the germicidal lamp is 3.2 times larger than that with
the black light blue fluorescent lamp. In eq 8, the denominator
on the right-hand side can be approximately written as 1 + k2/
kL = 1, and therefore, the apparent rate of formation of H2O2
causes the competitive reaction described as h+ + HCOO-
HCOO•, leading to the formation of a larger amount of H2O2
f

10496 J. Phys. Chem. A, Vol. 108, No. 47, 2004
Shiraishi and Kawanishi
(linear velocity) in the photocatalytic reaction using a batch
recirculation reactor to elucidate the effect of film formed in
the neighborhood of the photocatalyst on the photocatalytic
reaction. As a result, we obtained the following conclu-
sions.
(1) The rate of formation of H2O2 in the photocatalytic
reaction apparently increases with the increase of the linear
velocity.
(2) This is probably because the thickness of the film becomes
thin with the increase of the linear velocity and the H2O2 formed
on the photocatalyst surface can easily move through the film
to the bulk liquid.
(3) This phenomenon can successfully be explained by using
a mathematical model for the overall photocatalytic reaction
that takes into consideration the diffusion of H2O2 through the
film.
(4) One of the reasons other researchers could not detect H2O2
in the photocatalytic reaction is that the reaction experiments
were performed under the conditions where H2O2 formed on
the photocatalyst was so quickly decomposed in the film that
the H2O2 concentration in the bulk liquid did not become
high.
(5) As a result, it is concluded that a diffusion film formed
in the neighborhood of the photocatalyst surface is responsible
for a reduction in the H2O2 concentration.
Figure 8. Comparison between initial rates of formation of hydrogen
peroxide from water and an aqueous solution of formic acid (light
source: black light blue fluorescent lamp).
according to the subsequent reactions described as e- + O2 f
-•
O2 f ‚‚‚‚ f H2O2.
References and Notes
Interestingly, the initial rate of formation of H2O2 for the
germicidal lamp is lower than that for the black light blue
fluorescent lamp in the region of low linear velocity, rapidly
increasing with the increase of linear velocity and finally
becoming higher. Considering the fact that the germicidal lamp
emits more powerful UV light than does the black light blue
fluorescent lamp, the experimental result is evidently contradic-
tory. This can be explained as follows. When the linear velocity
is low and the film is relatively thick, the H2O2 produced on
the photocatalyst surface must move slowly by diffusion through
the film to the bulk liquid. As a result, H2O2 is exposed for a
longer time to a highly reactive circumstance in the film,
crowded by a larger amount of radicals produced under
irradiation with the UV light from the germicidal lamp, and
the majority of H2O2 is quickly decomposed in the film. In the
region of high linear velocity, however, the thickness of the
film is thin and the H2O2 produced can quickly move out of
the film or is withdrawn from the film into the bulk liquid by
the recirculation flow passing through the neighborhood of the
photocatalyst surface, thereby increasing the H2O2 concentration
in the bulk liquid. Even when the film becomes very thin due
to a higher linear liquid velocity and the system is reaction-
controlled, H2O2 is decomposed by the radicals that are
continuously produced on the photocatalyst surface and by the
UV light in the bulk liquid. Consequently, the H2O2 concentra-
tion in the bulk liquid takes a constant value at a point where
the rate of formation of H2O2 is equal to the rate of decomposi-
tion of H2O2.
(1) Hoffmann, M. T.; Martin, S. T.; Choi, W.; Bahnemann, D. W.
Chem. ReV. 1995, 95, 69.
(2) Jaeger, C. D.; Bard, A. J. J. Phys. Chem. 1979, 83, 3146.
(3) Wang, C.; Rabani, J.; Bahnemann, D. W.; Dohrmann, J. K. J.
Photochem. Photobiol. A:; Chem. Eng. J. 2002, 148, 169.
(4) Matthews, R. W. J. Catal. 1988, 111, 264.
(5) Sabate, J.; Anderson, M. A.; Kikkawa, H.; Edwards, M.; Hill, C.
G. J. Catal. 1991, 127, 167.
(6) Cho, S.; Choi, W. J. Photochem. Photobiol. A. Chem. 2001, 143,
221.
(7) Augugliaro, V.; Coluccia, S.; Loddo, V.; Marches, L.; Marta, G.;
Palmisano, L.; Schiavello, M. Appl. Catal., B 1999, 20, 15.
(8) Rao, M. V.; Rajeshwar, K.; Pai Verneker, V. R.; DuBow, J. J.
Phys. Chem. 1980, 84, 1987.
(9) Hong, A. P.; Bahnemann, D. W.; Hoffmann, M. R. J. Phys. Chem.
1987, 91, 2109.
(10) Fukinbara, S.; Shiraishi, F.; Nagasue, H. CELSS J. 2000, 12, 9.
(11) Sha, J.; Shiraishi, F. Chem. Eng. J. 2004, 97, 203.
(12) Matthews, R. W. J. Phys. Chem. 1987, 91, 3328.
(13) Turchi, C. S.; Ollis, D. F. J. Phys. Chem. 1988, 92, 6853.
(14) Shiraishi, F. Computational methods for analysis of immobilized
enzyme reactions: from reaction kinetics to reactor-design methods;
Corona: Tokyo, 1997.
(15) Wang, S.; Shiraishi, F.; Nakano, K. J. Chem. Technol. Biotechnol.
2002, 77, 805.
(16) Shiraishi, F.; Hiromitsu, M.; Hasegawa, T.; Kasai, S. J. Chem.
Technol. Biotechnol. 1996, 66, 405.
(17) Wang, S.; Shiraishi, F. Eco-Engineering 2002, 14, 9.
(18) Fukinbara, S.; Shiraishi, F.; Nakano, K. CELSS J. 2001, 13, 1.
(19) Fukinbara, S.; Shiraishi, F. CELSS J. 2001, 13, 11.
(20) Matsuo, K.; Takeshita, T.; Nakano, K. J. Cryst. Growth 1990, 99,
621.
(21) Shiraishi, F.; Nakasako, T.; Hua, Z. J. Phys. Chem. A 2003, 107,
11072.
(22) Shiraishi, F.; Toyoda, K.; Fukinbara, S.; Obuchi, E.; Nakano, K.
Chem. Eng. Sci. 1999, 54, 1547.
(23) Wang, S.; Shiraishi, F.; Nakano, K. Chem. Eng. J. 2002, 87,
261.
5. Conclusion
In the present work, we investigated the relationship between
the rate of formation of H2O2 and the recirculation flow rate
(24) Bird, R. B.; Stewart, W. E.; Lightfoot, E. N. Transport phenomena;
John Wiley & Sons: New York, 1960.