Formation of Hydrogen Peroxide in Photocatalytic Reactions

Article abstract of DOI:10.1021/jp036237+

To characterize the formation of H₂O₂ in photocatalytic reactions, HCHO and HCOOH have been photocatalytically decomposed by use of TiO₂ and Pd/TiO₂ under irradiation with the UV light emitted from a blacklight blue fluorescent lamp (BL) or germicidal lamp (GL). An enzymatic method has been introduced to accurately measure the H₂O₂ concentration. The photocatalyst irradiated with the UV light emitted from the GL produces H₂O ₂ from water at a larger rate and amount. Deposition of Pd on TiO₂ increases the rate and amount of the formation of H ₂O₂ from water. In the absence of the photocatalyst, no HCOOH is decomposed under irradiations with both the UV lights from the BL and GL, whereas the decomposition becomes possible in the presence of the photocatalyst. H₂O₂ is formed in parallel with the photocatalytic decomposition of HCOOH. In the presence of the photocatalyst, the concentration of the H₂O₂ formed during the decomposition of HCHO is smaller than that formed from water alone. In contrast, this relationship is reversed with the decomposition of HCOOH. This is considered due to the formation of H₂O₂ in the process of the decomposition of HCOOH to carbon dioxide and water. The rates of decomposition of HCHO and HCOOH are directly proportional to their respective rates of formation of H₂O₂, indicating that there is a close relationship between the photocatalytic activity and the capability of the photocatalyst to form H₂O₂.

Full text of DOI:10.1021/jp036237+

11072
J. Phys. Chem. A 2003, 107, 11072-11081
Formation of Hydrogen Peroxide in Photocatalytic Reactions
Fumihide Shiraishi,* Tsugunori Nakasako, and Zhaozhe Hua
Department of Biochemical Engineering and Science, Kyushu Institute of Technology,
Kawazu, Iizuka 820-8502, Japan
ReceiVed: July 31, 2003
To characterize the formation of H
photocatalytically decomposed by use of TiO
a blacklight blue fluorescent lamp (BL) or germicidal lamp (GL). An enzymatic method has been introduced
to accurately measure the H concentration. The photocatalyst irradiated with the UV light emitted from
the GL produces H from water at a larger rate and amount. Deposition of Pd on TiO increases the rate
and amount of the formation of H from water. In the absence of the photocatalyst, no HCOOH is
decomposed under irradiations with both the UV lights from the BL and GL, whereas the decomposition
becomes possible in the presence of the photocatalyst. H is formed in parallel with the photocatalytic
decomposition of HCOOH. In the presence of the photocatalyst, the concentration of the H formed during
the decomposition of HCHO is smaller than that formed from water alone. In contrast, this relationship is
reversed with the decomposition of HCOOH. This is considered due to the formation of H in the process
of the decomposition of HCOOH to carbon dioxide and water. The rates of decomposition of HCHO and
HCOOH are directly proportional to their respective rates of formation of H , indicating that there is a
close relationship between the photocatalytic activity and the capability of the photocatalyst to form H
2
O
2
in photocatalytic reactions, HCHO and HCOOH have been
2
and Pd/TiO under irradiation with the UV light emitted from
2
2 2
O
O
2 2
2
2 2
O
2
O
2
2
O
2
2 2
O
2
O
2
O .
2 2
1
. Introduction
supposed that the researchers were unable to pay much attention
to the presence of a relationship between the hydrogen peroxide
and photocatalytic reaction.
In the decomposition of organic compounds by titanium
dioxide (TiO2) irradiated with UV light, various kinds of reactive
1
The photocatalytic reaction may drastically be promoted or
retarded by doping TiO2 with a certain kind of metal. In the
past two decades, numerous reports have been published on the
photocatalytic activities that were enhanced in the production
of hydrogen and oxygen from a mixture of water and alcohol
by platinum- and/or ruthenium-doped TiO2 photocatalysts.1
The use of RuO2-doped TiO2 powder or RuO2 and Pt-doped
TiO2 particles induced water cleavage by visible light in water
radicals are formed. From the mechanisms of the photocatalytic
reactions described in a number of papers,2-9 it is easy to notice
that among these, only a few kinds of radicals, such as
superoxide radicals and hydroxyl radicals (HO ), are directly
related to the photocatalytic decomposition.
•
8-22
We recently investigated the time courses of the formic acid
(HCOOH) concentration in its photocatalytic decompositions,
where two kinds of enzymes were individually added to the
reaction mixture. Addition of superoxide dismutase (SOD),
catalyzing a conversion of superoxide into hydrogen peroxide
23-26
1
0
photolysis systems.
In addition, many reports are available
on metal-doped or metal-depositing TiO2 photocatalysts to
efficiently decompose toxic compounds. Several of the prepara-
(H2O2), did not affect the time course of the HCOOH concen-
2
7-29
tions showed positive effects,
negative ones.
while others presented
tration. On the other hand, addition of catalase, catalyzing a
decomposition of H2O2 into O2 and H2O, caused a temporary
stop of the photocatalytic reaction. This is probably because
H2O2 was rapidly decomposed by catalase and consequently,
2
7,28,30-32
We also photodeposited palladium (Pd)
on a thin film of TiO2 covering the surface of a glass tube to
examine its performance in the photocatalytic decompositions
3
3
•
of several kinds of organic compounds in water. As a result,
this Pd/TiO2 photocatalyst was found to decompose the organic
compounds, especially formaldehyde (HCHO), at a higher rate
than does the TiO2 photocatalyst. In general, such an increase
in the decomposition rate by depositing a certain metal on TiO2
is explained by an inhibition of the recombination of electrons
with holes by the deposited metal. Under such a condition,
however, it is unclear which chemical species in the system
changes its concentration and why the change results in a change
of the decomposition rate.
the concentration of HO , which is usually produced greatly
and allowed to aggressively react with HCOOH, was signifi-
cantly reduced. This experimental result suggests that H2O2 is
closely related to the photocatalytic decomposition of HCOOH,
or organic compounds.
However, there is little report on the concentration level of
H2O2 formed during the photocatalytic decomposition or on the
time course of the H2O2 concentration. This is considered due
11-17
to low sensitivities of the analytical methods
that have been
used to measure the H2O2 concentration in the photocatalytic
reaction. That is, in these analytical methods, it is difficult to
accurately measure a very low concentration of H2O2 that is
likely encountered in the photocatalytic reaction, so that it is
In the present work, therefore, we examine the formation of
H2O2 in the photocatalytic reactions under various reaction
conditions and then elucidate the relationship between the
photocatalytic activity and H2O2 concentration. To achieve this
goal, we introduce the analytical method using an enzyme,
which has an ability to accurately measure a very low
*
Author to whom correspondence should be addressed. Telephone:
(Japan)-948-29-7827.Fax: (Japan)-948-29-7801.E-mail: fumi@bse.kyutech.ac.jp.
1
0.1021/jp036237+ CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/19/2003

Formation of H2O2 in Photocatalytic Reactions
J. Phys. Chem. A, Vol. 107, No. 50, 2003 11073
concentration of H2O2 that is encountered in the photocatalytic
reaction. First, HCHO is photocatalytically decomposed using
TiO2 and Pd/TiO2 of different activities, and the time courses
of the concentrations of H2O2 formed during the photocatalytic
decompositions are investigated. Throughout a series of the
experiment, two types of low-pressure mercury lamps (blacklight
blue fluorescent lamp and germicidal lamp) are used to
investigate the effect of the wavelength of UV light on the H2O2
formation. Second, the same discussion is carried out on the
photocatalytic decomposition of HCOOH, which is an inter-
mediate in the photocatalytic decomposition of HCHO. Finally,
the role of the H2O2 formed in the photocatalytic reaction is
discussed under estimation of the mechanism of the formation
easily gelatinized after a period of time, it was redissolved prior
to use by addition of a small amount of H2O2 solution.
2
.4. Coating the Inside Surface of a Glass Tube with a
Thin Film of TiO2. The inside surface of a glass tube was
coated with a thin film of TiO2 according to the following
10,34-38
procedure.
(1) A Pyrex glass tube (28.5 mm in inside diameter, 1.8 mm
in wall thickness, and 140 mm long) was ultrasonically washed
in IPA for 3 min, rinsed in boiling distilled water, and dried at
3
73 K.
(2) The coating solution of TiO2 was uniformly applied to
the inside surface of the glass tube with a brush, and this glass
tube was then heated at 673 K for 30 min. The same procedure
was repeated four times.
•
of H2O2 and HO in the photocatalytic decomposition of HCHO.
(
3) After one more application of the coating solution, the
2
. Experimental Section
.1. Reagents and Materials. Palladium(II) nitrate [Pd-
glass tube was heated at 773 K for 1 h. Consequently, a glass
tube was obtained whose inside surface was coated with a
transparent thin film of anatase TiO2.
2
(NO3)2], methanol, 2-propanol (IPA), and HCHO (a 37 (v/v)%
aqueous solution containing 8 (v/v)% methanol as a stabilizer)
were purchased from Wako Pure Chemical Industries, Ltd.
2
.5. Photodeposition of Pd on a Thin Film of TiO2.
Photodeposition of Pd on a thin film of TiO2 was carried out
according to the following procedure, which was constructed
on the basis of the literature.
(Osaka). Titanium tetra-isopropoxide (TIP) was obtained from
33
Katayama Chemicals, Inc. (Osaka), and a 31 (v/v)% aqueous
H2O2 solution was from Mitsubishi Chemical Co., Ltd. (Tokyo).
All these reagents were at least of laboratory reagent grade. The
25
(1) An aqueous solution of 0.4 M methanol and 0.1 M
palladium nitrate was prepared.
6
-W UV lamps used as a light source were a blacklight blue
fluorescent lamp (BL) with wavelengths of 300-380 nm
FL6BL-B; Matsushita Electric Industrial Co., Ltd., Osaka) and
-
5
3
(
2) This solution (1.5 × 10 m ) was poured into the glass
tube coated with a thin film of TiO2, and both the ends of the
glass tube were plugged with rubber stoppers.
(
a germicidal lamp (GL) with a wavelength of 254 nm (GL-6;
Sankyo Denki, Tokyo).
(3) The glass tube was horizontally attached to a rotary
machine and slowly rotated for 5 min at 7 rpm under irradiation
with the UV light emitted from two BLs; consequently, Pd was
photodeposited on the thin film of TiO2.
2
.2. Preparation of Amorphous TiO2 Powder. Amorphous
34
TiO2 powder was prepared as follows.
(
1) TIP and IPA were mixed at a molar ratio of 1:5 (for
example, 29.69 g of TIP/ 31.39 g of IPA) at 278 K for 2 h.
2) To this solution, an aqueous solution of IPA at a molar
(4) After detachment from the rotary machine, the glass tube
was sufficiently washed with distilled water and dried at room
temperature.
2
(
ratio of 5:4 (for example, 31.39 g of IPA and 7.53 g of ultrarpure
water) was slowly added by spending more than 10 min and
the resulting solution, consisting of TIP, IPA, and H2O at a molar
ratio of 1:10:4, was mixed for 4 h; consequently, a white-colored
solution containing amorphous TiO2 fine particles was obtained.
.6. Photocatalytic Reactor System and Its Operating
1
0,36,39
Method. An annular-flow photocatalytic reactor,
as il-
lustrated in Figure 1, is a plastic cylindrical vessel with a UV
lamp fixed in the center, a quartz glass tube (20 mm in inside
diameter and 220 mm long) for protection of the lamp, and a
(3) The white-colored solution was filtered under suction and
2
Pyrex glass tube (photocatalytic surface area, 0.0123 m ) coated
the cake of amorphous TiO2 particles remaining on the filter
was dried at 373 K for 5 h, sometimes breaking into pieces
with a small hammer.
with a thin film of TiO2 or Pd/TiO2. The Pyrex glass tube was
inserted and fixed between the outer surface of the quartz glass
tube and the inside wall of the cylindrical vessel. Since the outer
surface of the Pyrex glass tube was in close contact with the
inside wall of the cylindrical vessel, a reaction mixture was
allowed to flow through an annulus of 2 mm in width between
the outer surface of the Pyrex glass tube and the inner surface
of the Pyrex glass tube. A substantial amount of liquid inside
(
4) After being sufficiently pounded in a mortar, the resulting
TiO2 powder was dried at 373 K for 15 h.
5) The amorphous TiO2 fine powder thus prepared was kept
in a desiccator with silica gel until next use.
(
2
.3. Preparation of a Coating Solution of TiO2. A coating
10,34-38
solution of TiO2 was prepared as follows.
-5
3
the reactor was 7.0 × 10 m .
(
1) The amorphous TiO2 powder was added to a 31% aqueous
The photocatalytic reactor, a mixed-flow vessel, and a
peristaltic pump (RP-1000; EYELA, Tokyo) were connected
-
5
3
H2O2 solution (a volume of 2.0 × 10 m per one gram of the
TiO2 powder) and well mixed at 298 K for 2 h; the TiO2 powder
was thus completely dissolved in the aqueous solution with
generating bubbles.
-4
3
in a loop, as illustrated in Figure 2. A 5.0 × 10 m aqueous
solution of HCHO prepared at a given concentration was poured
into the mixed-flow vessel and then recirculated at a flow rate
of 1.08 × 10 m min in the batch recirculation reactor
system. The reaction was started by switching the UV lamp in
the reactor on. A sample liquid was withdrawn at appropriate
time intervals from the mixed-flow vessel to determine the
HCHO and H2O2 concentrations by their respective analytical
methods described below. The same experiment was carried
(
2) The mixture was let stand until gelatinization at room
temperature.
3) The gelatinized solution was redissolved by addition of a
-3
3
-1
(
-
4
3
3
1% aqueous H2O2 solution (a volume of 1.2 × 10 m per
one gram of TiO2) and mixed at 298 K for 12 h.
4) This solution was let stand for about 50 h until generation
of bubbles stopped.
5) The transparent TiO2 solution of a pale yellow color thus
obtained was used as a coating solution. Since this solution was
(
-4
3
out for a 3.0 × 10 m aqueous solution of HCOOH at a given
concentration, and the HCOOH and H2O2 concentrations were
measured by the analytical methods described below.
(

11074 J. Phys. Chem. A, Vol. 107, No. 50, 2003
Shiraishi et al.
Figure 1. Annular-flow photocatalytic reactor.
of the substrate-specificity of an enzyme’s action.40 That is, an
enzyme can distinguish its own reactant from the others and
catalyzes a designated reaction. For example, in the kit used in
the present work, the peroxidase selectively acts on H2O2 to
cause the coloring reaction:
2
H O + 4-aminoanti-pyrine + phenol f
2 2
red quinoneimine dye + 4H O
2
If the solution becomes colored, therefore, the presence of H2O2
can be identified. In fact, since the coloring reaction occurred
when the analytical reagent was added to the sample solution,
there is no doubt that H2O2 was formed in the photocatalytic
reaction investigated in the present work.
Figure 2. Batch recirculation reactor system.
An enzymatic reaction is more or less influenced by the pH
of the sample solution. Therefore, aqueous solutions containing
H2O2 at the same concentration were analyzed in the pH range
of 3.5-6.5, corresponding to a pH change of the solution during
the HCHO decomposition. As a result, we found that the
measured values of the H2O2 concentration are changed within
only 5%. In the following discussion, therefore, we used the
measured values without any correction.
2
.7. Analytical Method for H2O2. Different types of analyti-
1
1-17
cal methods
have been used to measure the concentration
of H2O2 formed in the photocatalytic decompositions of organic
compounds in water. From the standpoint of the measure-
ment sensitivity, however, these methods based on chemical
reactions seem not to be adequate to accurately measure H2O2
at a very low concentration that is encountered in the photo-
catalytic reaction.
2
.8. Analytical Methods for HCHO and HCOOH. The
HCHO concentration was determined by coloring the sample
solution with an analytical reagent (Formaldehyde Test Wako;
Wako Pure Chemical Industries, Ltd.) and then measuring the
absorption at 550 nm with a spectrophotometer (UVmini-1240;
Shimadzu, Tokyo). The HCOOH concentration was measured
by ion chromatography (DX-100; Dionex Corporation, CA).
We therefore introduced an enzymatic method using a
commercially available analytical reagent (Glucose B Test
Wako; Wako Pure Chemical Industries, Ltd.). This reagent is
originally used to measure a â-D-glucose concentration. Addition
of the analytical reagent to an aqueous solution containing â-D-
glucose first causes an oxidation of glucose to â-D-gluconic acid
by glucose oxidase. At the same time, H2O2 is produced in the
oxidation and successively allowed to react with a coloring dye
by a catalytic action of peroxidase. Consequently, the â-D-
glucose concentration is selectively determined by measuring
the absorbance of the colored solution at 505 nm.
3
. Results and Discussion
3.1. Formation of HCOOH during Decomposition of
HCHO. Figure 3 shows the time courses of the HCHO and
It is readily predictable that if only H2O2 is contained in a
sample solution, addition of the above analytical reagent would
cause only the second enzymatic reaction. Although it is not so
difficult to prepare such an analytical reagent containing only
peroxidase, we used the commercially available analytical
reagent without any modification. The minimum measurable
HCOOH concentrations and pH in the decomposition of HCHO
-
3
at an initial concentration of 0.271mol m by Pd/TiO2 under
irradiation with the UV light emitted from the GL. HCHO is
decomposed toward zero concentration. In parallel to this
reaction, HCOOH is successively formed. Its concentration
initially increases, passing through a maximum and decreasing.
Such a time course of the concentration indicates that HCOOH
is an intermediate in the decomposition of HCHO. However,
-
4
-3
concentration of H2O2 was about 2 × 10 mol m . Another
advantage of introducing the enzymatic method is utilization

Formation of H2O2 in Photocatalytic Reactions
J. Phys. Chem. A, Vol. 107, No. 50, 2003 11075
Figure 3. Time courses of HCHO and HCOOH concentrations and
pH in decomposition of HCHO at an initial concentration of 0.271mol
-
3
m
by Pd/TiO
2
under irradiation with UV light emitted from a
germicidal lamp.
2 2
Figure 4. Formation of H O from water by photocatalyst under
irradiation with UV light from a blacklight blue fluorescent lamp.
since the HCOOH concentration reaches a maximum of about
-
3
0
.04 mol m and this value is rather small compared with the
concentration corresponding to the HCHO decomposed, it is
supposed that the HCOOH formed on the photocatalyst surface
is mostly decomposed on the spot without being released from
the surface. The pH of the reaction mixture is decreased from
2 2
Figure 5. Time courses of HCHO and H O concentrations in
6.5 to 3.8 as a result of the formation of HCOOH.
-3
decompositions of HCHO at an initial concentration of 0.3 mol m
by UV light alone or photocatalyst irradiated with UV light from a
blacklight blue fluorescent lamp. (a) Change in HCHO concentration,
3
.2. Decomposition of HCHO. 3.2.1. Blacklight Blue
Fluorescent Lamp. Figure 4 shows the time courses of the H2O2
concentration in the mixed-flow vessel when water was
recirculated through the reactor with the glass tube coated with
a thin film of TiO2 or Pd/TiO2 under irradiation with the UV
light from the BL. In the case of TiO2 alone, the H2O2
concentration increases until 120 min and then levels off at about
2 2
(b) change in H O concentration.
the BL. The HCHO concentration decreases even in the absence
of the photocatalyst, indicating that HCHO is easily decomposed
by just irradiating with UV light. The decomposition of HCHO
is certainly increased in the presence of the photocatalyst, which
increase is significant with Pd/TiO2. It should be noted that there
is a remarkable difference between the concentrations of H2O2
formed from water alone and from water containing HCHO.
As a result of the decomposition of HCHO, the H2O2 concentra-
tion is lowered by about one-half with TiO2 and one-fourth with
Pd/TiO2. This result clearly indicates that a considerable amount
of H2O2 is steadily consumed throughout the decomposition of
HCHO.
-
3
0
.005 mol m . In the case of Pd/TiO2, the H2O2 concentration
quickly increases from the initial stage and continues to increase
even after 600 min. These results indicate that H2O2 is certainly
produced in the presence of TiO2 and the production rate is
more increased by depositing Pd on TiO2.
Figure 5 shows the time courses of the HCHO and H2O2
concentrations in the mixed-flow vessel when an aqueous
solution of 0.3 mol m-3 HCHO was recirculated through the
photocatalytic reactor under irradiation with the UV light from

11076 J. Phys. Chem. A, Vol. 107, No. 50, 2003
Shiraishi et al.
Figure 6. Formation of H
2 2
O from water by photocatalyst under
irradiation with UV light from a germicidal lamp.
3
.2.2. Germicidal Lamp. Figure 6 shows the time courses of
the H2O2 concentration when water was recirculated through
the reactor under irradiation with the UV light from the GL.
As in the case of the BL, H2O2 is certainly produced in the
presence of the photocatalyst and the capability of the photo-
catalyst to form H2O2 is enhanced by depositing Pd on TiO2.
Figure 7 shows the time courses of the HCHO and H2O2
concentrations in the mixed-flow vessel when an aqueous
solution of 0.3 mol m-3 HCHO was recirculated through the
photocatalytic reactor under irradiation with the UV light emitted
from the GL. As in the case of the BL, the decomposition of
HCHO is caused even in the absence of TiO2 or Pd/TiO2 and
is accelerated in the presence of the photocatalyst, especially
with Pd/TiO2. The H2O2 concentration is lowered compared to
the case of water alone in Figure 6, implying that H2O2 is
steadily consumed during the decomposition of HCHO.
The rate of decomposition of HCHO is clearly proportional
to the rate of formation of H2O2. It should be noted that the
H2O2 concentration for Pd/TiO2 passes through a maximum at
2
40 min and then decreases rapidly from 360 min, where most
HCHO has been decomposed. As discussed later, such a time
course of the H2O2 concentration is closely related to the H2O2
formed in parallel with the decomposition of HCOOH, which
is produced as an intermediate in the decomposition of HCHO.
2 2
Figure 7. Time courses of HCHO and H O concentrations in
decompositions of HCHO at an initial concentration of 0.3 mol m-3
by UV light alone or photocatalyst irradiated with UV light from a
germicidal lamp. (a) Change in HCHO concentration, (b) change in
H O concentration.
2 2
3
.3. Decomposition of HCOOH. A formation mechanism
of H2O2 has been proposed in the photolytic decomposition of
HCOOH.4
1-43
Also, Figure 3 gives a proof for the formation
-
4
3
of HCOOH in the photocatalytic decomposition of HCHO. In
this photocatalytic decomposition, thus, there is a possibility
that H2O2 is formed in parallel with the decomposition of
HCOOH. If this is true, it is clear that this formation more or
less affects the time course of the H2O2 concentration during
the photocatalytic decomposition. To clarify the formation
mechanism of H2O2 in the photocatalytic decomposition of
HCHO, therefore, the photocatalytic reaction was carried out
using HCOOH as an initial reactant.
where a 3.0 × 10 m aqueous solution of HCOOH, at an
-
3
initial concentration of 0.45 mol m , was treated by TiO
2
3
-
5
irradiated with the UV light from the BL and a 5.0 × 10
m
-
3
aqueous solution of HCOOH at 0.45 mol m was added after
720 min. The H O concentration initially increases with the
2
2
decrease in the HCOOH concentration, passing through a
maximum at around 180 min and then decreasing. It is apparent
that the H O concentration temporarily increases just after the
2
2
addition of HCOOH at 720 min, which clearly shows that the
formation of H2O2 occurred in parallel with the photocatalytic
decomposition of HCOOH.
3.3.2. Blacklight Blue Fluorescent Lamp. Figure 9 shows the
time courses of the H2O2 concentration in the mixed-flow vessel
when water was recirculated through the reactor with the glass
tube coated with a thin film of TiO2 or Pd/TiO2 under irradiation
with the UV light from the BL. This experimental result is
fundamentally the same as that in Figure 4. However, we dare
3
.3.1. Formation of H2O2 in Photocatalytic Decomposition
of HCOOH. Although several research groups have studied the
44-46
photocatalytic decomposition of HCOOH,
there is no report
on the formation of H2O2 in this photocatalytic reaction. To
confirm the formation of H2O2, therefore, we investigated if
the H2O2 concentration is changed when the HCOOH concen-
tration is instantaneously increased in the process of the
decomposition of HCOOH. The result is shown in Figure 8,

Formation of H2O2 in Photocatalytic Reactions
J. Phys. Chem. A, Vol. 107, No. 50, 2003 11077
Figure 8. Time courses of HCOOH and H
2
O
2
concentrations in
-
4
3
treatment of a 3.0 × 10 m aqueous solution of HCOOH at an initial
-3
concentration of 0.45 mol m by TiO
2
under irradiation with UV light
-
5
3
from a blacklight blue fluorescent lamp. A 5.0 × 10 m aqueous
-
3
solution of HCOOH at 0.45 mol m was added after 720 min.
2 2
Figure 9. Time courses of H O concentration under irradiation of
water with UV light from a blacklight blue fluorescent lamp.
to show the experimental data because the volume of the
reaction mixture was different between the decompositions of
HCHO and HCOOH. In addition, we used different lots of the
glass tubes coated with TiO2 and Pd/TiO2 and the BLs. In the
case of TiO2 alone, the H2O2 concentration is steadily increased
from the start of the reaction to 600 min and becomes almost
constant at about 0.009 mol m . The rate of formation of H2O2
for Pd/TiO2 is more than three times larger than that for TiO2
and the H2O2 concentration still continues to increase even after
2 2
Figure 10. Time courses of HCOOH and H O concentrations in
-3
decomposition of HCOOH at an initial concentration of 0.45 mol m
by UV light alone or by photocatalysts irradiated with UV light from
a blacklight blue fluorescent lamp. (a) Time courses of HCOOH
-3
2 2
concentration, (b) time courses of H O concentration.
6
00 min.
that HCOOH is certainly decomposed in the presence of the
photocatalyst and the rate of decomposition for Pd/TiO2 is larger
than that for TiO2. The higher the rate of decomposition of
HCOOH, the larger the concentration of the H2O2 formed.
Moreover, the H2O2 concentration measured is larger in water
containing HCOOH than in water alone. In the photocatalytic
decomposition of HCHO, on the other hand, this relation was
reversed. These experimental results clearly show that when
HCOOH is used as a reactant, the formation of H2O2 occurs in
parallel with the decomposition of HCOOH. Since the decom-
Figure 10 shows the time courses of the HCOOH and H2O2
concentrations in the mixed-flow vessel when an aqueous
solution of 0.45 mol m-3 HCOOH was recirculated through
the photocatalytic reactor under irradiation with the UV light
from the BL. In the absence of TiO2 or Pd/ TiO2, there is little
decrease in the HCOOH concentration, indicating that HCOOH
cannot easily be decomposed by just irradiating with the UV
light from the BL, whereas HCHO was decomposed under the
same reaction condition, as shown in Figure 5. It is evident

11078 J. Phys. Chem. A, Vol. 107, No. 50, 2003
Shiraishi et al.
2 2
Figure 11. Time courses of H O concentration under irradiation of
water with UV light from a germicidal lamp.
position of HCHO causes a successive formation of HCOOH
(Figure 3), it is thus considered that H2O2 is simultaneously
formed with the decomposition of HCOOH, while the H2O2
formed is consumed in the oxidation of HCHO to HCOOH,
leading to a lowering in the H2O2 concentration compared with
the concentration of the H2O2 formed from water alone.
3
.3.3. Germicidal Lamp. Figure 11 shows the time courses
of the H2O2 concentration in the mixed-flow vessel when water
was recirculated through the reactor with the glass tube coated
with a thin film of TiO2 or Pd/TiO2 under irradiation with the
UV light from the GL. This experimental result is fundamentally
the same as that in Figure 6. Again, we dare to show the
experimental data because the volume of the reaction mixture
was different between the decompositions of HCHO and
HCOOH and we used different lots of the glass tubes coated
with TiO2 and Pd/TiO2 and the GLs. It is clear that depositing
Pd on TiO2 certainly increases the capability of the photocatalyst
to form H2O2.
Figure 12 shows the time courses of the HCOOH and H2O2
concentrations in the mixed-flow vessel when an aqueous
solution of 0.45 mol m-3 HCOOH was recirculated through
the photocatalytic reactor with a glass tube of TiO2 or Pd/TiO2
under irradiation with the UV light from the GL. The rate of
decomposition of HCOOH with the GL is 1.6 times larger than
that with the BL in the presence of TiO2, and 3.2 times larger
in the presence of Pd/TiO2. There is little decomposition of
HCOOH in the absence of the photocatalyst, as found with the
use of the BL. In contrast, HCOOH is remarkably decomposed
in the presence of the photocatalyst. The rate of decomposition
is larger with Pd/TiO2.
Figure 12. Time courses of HCOOH and H
decomposition of HCOOH at an initial concentration of 0.45 mol m
by UV light alone or by photocatalysts irradiated with UV light from
a germicidal lamp. (a) Time courses of HCOOH concentration, (b) time
2 2
courses of H O concentration.
2 2
O concentrations in
-
3
3.4. Relationship between Rates of HCHO Decomposition
and of H2O2 Formation. Figure 13 shows a plot of the initial
rate of decomposition of HCHO against the initial rate of
formation of H2O2 in the photocatalytic decomposition of
HCHO. These values were calculated from Figures 5 and 7. It
is apparent that there is a proportional relationship between the
two initial rates. That is, the higher the ability to form H2O2,
the higher the ability to decompose HCHO.
ship between the two initial rates, indicating that the higher the
ability to form H O , the higher the ability to decompose
2
2
HCOOH.
Likewise, the initial rate of decomposition of HCOOH is
plotted against the initial rate of formation of H2O2 in the
decomposition of HCOOH in Figure 14. These values
were calculated from Figures 10 and 12. As in the case of
the decomposition of HCHO, there is a proportional relation-
Consequently, the data plots for the decompositions of HCHO
and HCOOH suggest that the photocatalytic activity of TiO2 is
strongly dependent on the ability of the photocatalyst to form
H2O2. It is thus concluded that the higher the ability to form
H2O2, the higher the photocatalytic activity.

Formation of H2O2 in Photocatalytic Reactions
J. Phys. Chem. A, Vol. 107, No. 50, 2003 11079
2 2
Figure 14. Relationship between initial rates of formation of H O
from water and initial rates of decomposition of HCOOH in photo-
catalytic decomposition of HCOOH.
2 2
Figure 13. Relationship between initial rates of formation of H O
from water and initial rates of decomposition of HCHO in photocatalytic
decomposition of HCHO.
•
HO2 is produced via several steps, followed by the formation
of H2O2 (eqs 6 and 8).
3
.5. Estimation of Reaction Mechanism. 3.5.1. Formation
of Radicals in Photocatalytic Reactions. The mechanism of the
formation of radicals from water on the photocatalyst is
generally described as4
It is known that an oxygen molecule is decomposed into
oxygen atoms under irradiation with UV light of a wavelength
7,48
4
9
shorter than 240 nm (eq 12). The oxygen atom thus produced
easily reacts with an oxygen molecule to become ozone (eq
13, which subsequently reacts with water to give H2O2 (eq 14):
hν
-
+
TiO2
9
8 TiO (e + h )
2
(1)
(2)
+
•
+
UV(185 nm)
H O + h f HO + H
•
•
2
O2
8 O + O
(12)
-
+
•
OH + h f HO
(3)
(4)
•
O + O
9
8 O
(13)
(14)
2
3
-
-•
2
O + e f O
2
H O + O
9
8 H O + O
2
3
2
2
2
O2-• + H f HO2
+
•
(5)
(6)
However, the wavelengths of the light sources used in the
present work are above 240 nm, and therefore it is considered
that the formation of H2O2 does not occur according to the
reaction mechanism given by eqs 12-14. In fact, when the BL
was used as a light source, there was little formation of H2O2
by just irradiating the UV light (300-380 nm). Likewise, the
formation of H2O2 was slight when the GL was used.
•
2
•
2
HO + HO f H O + O
2
2
2
O2-• + HO f HO + O2
•
2
•-
2
(7)
•
2
+
HO + H fH O
2
(8)
2
3
.5.3. Decomposition of H2O2. It is considered that the
photocatalytic and photolytic decompositions of H2O2 occur in
parallel with the formation of H2O2. The former decomposition
proceeds according to eqs 9 and 10, and the latter decomposition
proceeds under irradiation with UV light according to eq 11.
Moreover, the experimental result for the UV photolysis of
-
•
-
H O + e f HO + OH
(9)
2
2
•
2
-
•
-
H O + O f HO + OH + O
(10)
(11)
2
2
2
hν
42
•
H2O2 suggested the following mechanism for the photolysis
of H2O2 in an aqueous solution:
H O
98 2HO
2
2
This reaction mechanism indicates that several kinds of radicals
are successively produced as a result of the formation of
electrons and holes on the TiO2 irradiated with UV light.
•
•
HO + H O
9
8 HO + H O
(15)
(16)
2
2
2
2
•
•
HO + HO
9
8 H O + O
3
.5.2. Formation of H2O2. From the above reaction mecha-
2
2
2
nism, it is considered that the formation of H2O2 in the presence
of TiO2 occurs in the following way. First, water molecules
are decomposed by the action of holes on the photocatalytic
Also, the following reaction may proceed:
2HO•
•
surface excited with UV light to produce HO (eq 2). This radical
9
8 H O
(17)
2
2
is consumed for the decomposition of organic compounds if
they are present. On the other hand, the dissolved O2 is
converted to O2 by the action of electrons (eq 4). Finally,
•
3.5.4. Formation Mechanism of HO . The following three
-
•
•
pathways may be possible for the formation of HO :

11080 J. Phys. Chem. A, Vol. 107, No. 50, 2003
Shiraishi et al.
Figure 15. Reaction mechanism proposed for photocatalytic decomposition of HCHO.
Pathway 1: formed as a result of the decomposition of H2O2
by UV light (eq 11).
on TiO2 was caused by promotion of each elementary reaction
owing to a decrease in the recombination rate of electrons with
holes.
Pathway 2: formed as a result of the decomposition of H2O2
by the action of radicals densely existing in the very neighbor-
hood of the photocatalyst surface (eqs 9 and 10).
Pathway 3: formed after several reaction steps including the
photocatalytic decomposition of water (eqs 2 and 3).
To elucidate which pathway among pathways 1-3 is the main
3.6. Significance of the Measurement of H O Concentra-
2
2
tion. There is little report on the time course of the H O2
2
concentration during the photocatalytic reaction. This is probably
because the concentration of the H O formed by the photo-
2
2
catalyst is remarkably reduced by the autodecomposition or the
consumption in the decomposition of organic compounds,
thereby becoming difficult to detect H O . In fact, this was
•
route to the HO formation in the photocatalytic reaction, H2O2
at almost the same concentration level as produced in the
photocatalytic decomposition of HCOOH was added to the
HCOOH solution and a change in the HCOOH concentration
was investigated under irradiation with the UV light from the
GL. Unlike in the presence of photocatalyst, however, HCOOH
was slowly decomposed under this condition. Therefore, the
H2O2 concentration was increased by 10 times and HCOOH
was likewise decomposed. As a result, such a high concentration
of H2O2 made it possible to decompose HCOOH rapidly,
suggesting that the decomposition of HCOOH occurs in the very
neighborhood of the photocatalyst surface and the formation
2
2
confirmed in our experiment. When HCHO was used as a
reactant, the H O concentration was lower than that formed
2
2
from water alone under UV-irradiation, because H O was
2
2
unilaterally consumed in the oxidation of HCHO to HCOOH,
which makes it difficult to accurately measure the H O2
2
concentration by the conventional analytical methods based on
1
1-15,17
chemical reactions.
When HCOOH was used as a
reactant, on the other hand, the H O concentration became
2
2
higher because H O was formed in the process of the
2
2
decomposition of HCOOH. In the photocatalytic decompositions
of most organic compounds, H O would not be formed unless
•
of HO is mainly carried out through pathway 2 or 3. A further
2
2
study is necessary to judge which of pathway 2 and 3 is more
utilized.
HCOOH is produced as an intermediate, leading to a remarkably
low H O concentration and then causing a difficulty in the
2
2
3
.5.5. Mechanism of Photocatalytic Decomposition of HCHO.
measurement of the concentration. However, the use of the
enzymatic method introduced in the present work would enable
us to characterize the formation of H2O2 in the photocatalytic
reaction.
The result of the measurement of H2O2 concentration indicates
that the photocatalytic activity is directly proportional to the
capability of the photocatalyst to form H2O2. This new finding
would be useful to understand the mechanism of the photo-
catalytic reaction more clearly. Moreover, the concentration of
the H2O2 produced in the photocatalytic treatment of water alone
could be used as an index to identify the performances of the
photocatalysts prepared by various methods.
Figure 15 shows a mechanism estimated for the decomposition
of HCHO including the formations of H2O2 and HO . HCHO
is finally decomposed into carbon dioxide and water via the
formation of HCOOH. HO is formed according to the mech-
•
•
anism described above and is directly related to the decomposi-
tions of HCHO and HCOOH. In parallel, formation of H2O2
occurs in the photocatalytic reaction. This is carried out through
two pathways; the inherent pathway of the photocatalyst and
the pathway included in the process of the decomposition of
HCOOH. H2O2 is gradually accumulated in the solution with
the progress of the reaction, while it is decomposed by the
actions of UV light and radicals. Consequently, the increase in
the H2O2 concentration terminates at a certain concentration
level.
4
. Conclusions
An increase in the photocatalytic activity when a certain metal
is deposited is generally explained by attraction of electrons to
the metal particles on TiO2 followed by reduction in the
recombination rate of electrons with holes. Thus, it is concluded
that the increase of the H2O2 concentration by depositing Pd
In the present work, the photocatalytic decompositions of
HCHO and HCOOH were performed by use of TiO2 and Pd/
TiO2 under irradiation with the UV light emitted from a UV
lamp (BL and GL) and the formation of H2O2 in their
photocatalytic decomposition was investigated by introducing

Formation of H2O2 in Photocatalytic Reactions
J. Phys. Chem. A, Vol. 107, No. 50, 2003 11081
a highly accurate method for measurement of H2O2 based on
an enzymatic reaction. As a result, the following conclusions
were drawn.
(18) Sobczynski, A.; Bard, A. J.; Campion, A.; Fox, M. A.; Mallouk,
T.; Webber, S. E.; White, J. M. J. Phys. Chem. 1987, 91, 3316.
(19) Kanno, H.; Yamamoto, Y.; Harada, H. Chem. Phys. Lett. 1985,
1
21, 245.
(
1) The photocatalyst irradiated with the UV light from the
GL produces H2O2 from water at a larger rate and amount.
2) Deposition of Pd on TiO2 increases the rate and amount
of the formation of H2O2 from water.
3) In the absence of the photocatalyst, no HCOOH is
(20) Kawai, T.; Sakata, T. J. Chem. Soc. Chem. Commun. 1980, 102,
694.
(21) Moon, S. C.; Mametsuka, H.; Suzuki, E.; Nakahara, Y. Catal. Today
(
1
998, 45, 79084.
(
22) Duonghong, D.; Borgarello, E.; Gratzel, M. J. Am. Chem. Soc. 1981,
103, 4685.
(23) Ohno, T.; Tanigawa, F.; Fujihara, K.; Izumi, S.; Matsumura, M. J.
Photochem. Photobiol. A: Chem. 1999, 127, 107.
24) Lobedank, J.; Bellmann, E.; Bendig, J. J. Photochem. Photobiol.
(
decomposed under irradiations with both the UV lights from
the BL and GL, whereas the decomposition becomes possible
in the presence of the photocatalyst.
(
A: Chem. 1997, 108, 89.
(25) Borgarello, E.; Kiwi, M.; Pelizzetti, E.; Visca, M.; Gratzel, J. J.
Am. Chem. Soc. 1981, 103, 6324.
(
4) H2O2 is formed in parallel with the photocatalytic
decomposition of HCOOH.
5) In the presence of the photocatalyst, the concentration of
(
26) Borgarello, E.; Kiwi, J.; Gratzel, M.; Pelizzetti, E.; Visca, M. J.
Am. Chem. Soc. 1982, 104, 2996.
27) Lindner, M.; Theurich, J.; Bahnemann, D. W. Water Sci. Technol.
(
the H2O2 formed during the decomposition of HCHO is smaller
than that formed from water alone. In contrast, this relationship
is reversed with the decomposition of HCOOH. This is
considered due to the formation of H2O2 in the process of the
decomposition of HCOOH to carbon dioxide and water.
(
1997, 35, 79.
(28) Tajima, M.; Niwa, M.; Fujii, Y.; Koinuma, Y.; Aizawa, R.;
Kushiyama, S.; Kobayashi, S.; Mizuno, K.; Ohuchi, H. Appl. Catalysis B:
EnVironmental 1997, 14, 97.
(
29) Bockelmann, D.; Goslich, R.; Bahnemann, D. Solar Thermal Energy
Utilization; Springer-Verlag GmbH: Heidelberg, 1992; Vol. 6.
30) Herrmann, J.-M.; Tahiri, H.; Ait-Ichou, Y.; Lassaletta, G.; Gonzalez-
Elipe, A. R.; Fernandez, A. Appl. Catal. B: EnVironmental 1997, 13, 219.
31) Sclafani, A.; Palmisano, L.; Schiavello, M. Res. Chem. Intermed.
(6) The rates of decomposition of HCHO and HCOOH are
(
directly proportional to their respective rates of formation of
H2O2, indicating that there is a close relationship between the
photocatalytic activity and the capability of the photocatalyst
to form H2O2.
(
1
992, 18, 211.
(32) Palmisano, L.; Augugliaro, V.; Sclafani, A.; Schiavello, M. J. Phys.
Chem. 1988, 92, 6710.
(
33) Jin, S.; Shiraishi, F. Chem. Eng. J., in press.
References and Notes
(34) Matsuo, K.; Takeshita, T.; Nakano, K. J. Crystal Growth 1990,
(
1) Jaeger, C. D.; Bard, A. J. J. Phys. Chem. 1979, 83, 3146.
99, 621.
(35) Xu, J.-H.; Shiraishi, F. J. Chem. Technol. Biotechnol. 1999, 74,
1096.
(36) Wang, S.; Shiraishi, F. Eco-Engineering 2002, 14, 9.
(
21.
2) Cho, S.; Choi, W. J. Photochem. Photobiol. A. Chem. 2001, 143,
2
(3) Sabate, J.; Anderson, M. A.; Kikkawa, H.; Edwards, M.; Hill, C.
G. J. Catal. 1991, 127, 167.
(
(
(
37) Fukinbara, S.; Shiraishi, F.; Nakano, K. CELSS J. 2001, 13, 1.
38) Fukinbara, S.; Shiraishi, F. CELSS J. 2001, 13, 11.
39) Shiraishi, F.; Toyoda, K.; Fukinbara, S.; Obuchi, E.; Nakano, K.
(
(
(
4) Matthews, R. W. Water Res. 1986, 20, 569.
5) Matthews, R. W. J. Catal. 1988, 111, 264.
6) Hoffmann, M. T.; Martin, S. T.; Choi, W.; Bahnemann, D. W.
Chem. Eng. Sci. 1999, 54, 1547.
40) Conn, E. E.; Stumpf, P. K.; Bruening, G.; Doi, R. H. Outlines of
biochemistry; John Wiley & Sons: New York, 1987.
41) Sugawara, T.; Kawashima, T.; Sakurada, J.; Sugawara, K. Kagaku
Kogaku Ronbunshu 1999, 25, 842.
42) Guittonneau, S.; De Laat, J.; Dore, M.; Duguet, J. P.; Bonnel, C.
ReV. Sci. de l’Eau 1988, 1, 35.
Chem. ReV. 1995, 95, 69.
7) Augugliaro, V.; Coluccia, S.; Loddo, V.; Marches, L.; Marta, G.;
Palmisano, L.; Schiavello, M. Appl. Catal. B 1999, 20, 15.
(
(
(
(
(
8) Spacek, W.; Bauer, R. Chemosphere 1995, 30, 477.
9) Fox, M. A.; Dulay, M. T. Chem. ReV. 1993, 93, 341.
(
(
(
10) Wang, S.; Shiraishi, F.; Nakano, K. Chem. Eng. J. 2002, 87, 261.
11) Arslan, I.; Balcioglu, A.; Bahnemann, D. W. Dyes Pigments 2000,
(
(
(
43) Legrini, O.; Oliveros, E.; Braun, A. M. Chem. ReV. 1993, 93, 671.
44) Kim, D. H.; Anderson, M. A. EnViron. Sci. Technol. 1994, 28, 479.
45) Kim, D. H.; Anderson, M. A. J. Photochem. Photobiol. A 1996,
4
7, 207.
12) Fallmann, H.; Krutzler, T.; Bauer, R.; Malato, S.; Blanco, J. Catal.
Today 1999, 54, 309.
13) Hugul, M.; Apak, R.; Demirci, S. J. Hazard. Mater. 2000, 77, 193.
14) Lunar, L.; Sicilia, D.; Rubio, S.; Perez-Bendito, S.; Nickel, U. Water
Res. 2000, 34, 1791.
15) Nedoloujko, A.; Kiwi, J. Water Res. 2000, 34, 3277.
16) Sanchez, L.; Peral, J.; Domenech, X. Electrochim. Acta 1985, 41,
981.
17) Rao, M. V.; Rajeshwar, K.; Pai Verneker, V. R.; DuBow, J. J.
Phys. Chem. 1980, 84, 1987.
(
9
4, 221.
(
(
(46) Mazzarino, I.; Piccinini, P. Chem. Eng. Sci. 1999, 54, 3107.
(47) Zhao, M.; Chen, S.; Tao, Y. J. Chem. Technol. Biotechnol. 1995,
64, 339.
(48) Pelizzetti, E.; Carlin, V.; Minero, C.; Gratzel, M. New J. Chem.
(
(
1
1991, 15, 351.
(
(49) Golimowski, J.; Golimowska, K. Anal. Chim. Acta 1996, 325,
111.