Role of excess electrons in TiO2 nanoparticles coated with Pt in reduction reactions studied in radiolysis of aqueous solutions

Article abstract of DOI:10.1021/jp045989g

Surface modification of TiO₂ nanocrystallites in aqueous solution was carried out by reduction of chloroplatinic acid with TiO ₂ electrons (e_TiO2^-). The TiO₂ electrons had been generated through electron injection from 2-propanol radicals, hydrogen atoms, and hydrated electrons produced by γ or pulse radiolysis prior to the addition of the chloroplatinic acid. Upon addition of the latter, Pt⁰ clusters are produced by multistep reduction of the chloroplatinic acid on the TiO₂ surface. The present work concerns the material balance and chemical kinetics of the Pt₀ mediated reduction of water and H⁺ to molecular hydrogen by the TiO ₂ electron. Pt⁰ showed no effect on the rate of oxygen reduction, compared to bare TiO₂. The effect of surface Pt ⁰ on increasing photocatalytic yields, claimed by a number of laboratories, is apparently due to electron capture by the Pt⁰ rather than electron removal by oxygen.

Full text of DOI:10.1021/jp045989g

J. Phys. Chem. B 2004, 108, 20291-20295
20291
Role of Excess Electrons in TiO2 Nanoparticles Coated with Pt in Reduction Reactions
Studied in Radiolysis of Aqueous Solutions
†
‡
,‡
Zorica Kasarevic-Popovic, David Behar, and Joseph Rabani*
Department of Physical Chemistry and Accelerator Unit, The Hebrew UniVersity of Jerusalem,
Jerusalem 91904, Israel, and Institute of Nuclear Sciences Vinca, Belgrade, Serbia and Montenegro
ReceiVed: September 6, 2004; In Final Form: October 12, 2004
Surface modification of TiO
acid with TiO electrons (eTiO2 ). The TiO
-propanol radicals, hydrogen atoms, and hydrated electrons produced by γ or pulse radiolysis prior to the
2
nanocrystallites in aqueous solution was carried out by reduction of chloroplatinic
-
2
2
electrons had been generated through electron injection from
2
0
addition of the chloroplatinic acid. Upon addition of the latter, Pt clusters are produced by multistep reduction
of the chloroplatinic acid on the TiO surface. The present work concerns the material balance and chemical
kinetics of the Pt mediated reduction of water and H to molecular hydrogen by the TiO
2
0
+
electron. Pt0
. The effect of surface Pt on increasing
2
0
showed no effect on the rate of oxygen reduction, compared to bare TiO
2
photocatalytic yields, claimed by a number of laboratories, is apparently due to electron capture by the Pt0
rather than electron removal by oxygen.
Introduction
Wang, Heller, and Gerischer12 have shown that catalyzed oxygen
reduction by surface metal increases photocatalytic oxidation
yields by preventing accumulation of electrons on the TiO2
Semiconductor-metal compositions have been widely used
to enhance photocatalytic efficiencies, in particular platinizations
of TiO2 were used to improve photocatalytic yields and increase
particles, thus inhibiting the rate of electron-hole recombination.
0
1
Incorporation of Pd in the surface increased the quantum
detoxification rates. Graetzel has noted that free diffusion of
efficiency of the photoassisted oxidation of aqueous 2,2-
charge carriers permits electron transfer within a time of 0.1 ps
and thus effectively competes with electron-hole recombination.
It has been recently shown (in the ZnO/Pt nanocrystallites, which
has similar energy properties to TiO2/Pt), that excess electrons
reside exclusively on the Pt. Thus, it is not surprising that the
TiO2/Pt system shows enhanced photocatalytic degradation of
1
2
dichloropropionate 3-fold at 0.01 wt % Pd.
Reduction reactions of excess electrons in TiO2 nanocrys-
tallites produced by pulse radiolysis have been studied earlier.13
Rates of reductions of several substrates were determined and
it was found that the rates depend on the TiO2 particle size.
2
2
+
-
-
-
-
3
,4
2 3 2 3
Cu , ClO , ClO , NO , and NO showed decay of the
biphenol compared to bare TiO2, such as in oxidation of
5
6
TiO2 electron predominately by a single pseudo-first-order
process. On the other hand, hydrogen peroxide and oxygen show
a multiexponential decay.
ammonia to N2 and CO photooxidation. Photoreduction was
occasionally used for preparation of Pt-TiO2. The interaction
of Pt and TiO2 was investigated by means of UV-vis absorption
spectra, photoluminescence emission spectra, and X-ray pho-
toelectron emission spectroscopy. Pt , Pt , and Pt on the
surface of Pt/TiO2, and the Ti3 species existing in its lattice
were proposed to form a defect energy level. It has been claimed
that the overall yield of trapped holes can be considerably
increased when small platinum islands are present on the TiO2
surface, which act as efficient electron scavengers competing
The purpose of the present work is to study the effect of
0
2+
4+
+
surface coated Pt on TiO2 electron reduction of oxygen, H
+
and water. We have studied the reactions of the TiO2 electrons
in the TiO2/Pt system after the generation of the electrons had
10,11
been completed, rather than during photolysis.
This enabled
the elucidation of the reaction mechanism, radiation yields and
kinetic rate constants, as well as separation between the
-
+
7,8
+
with the undesired e /h recombination. Platinum (or nitro-
gen) may also effectively convert titania into a semiconductor
photocatalyst active in the visible. A number of works have
reduction of H and water.
9
Experimental Section
been published concerning photochemical generation of hydro-
gen in Pt modified TiO2 surface.1
0,11
Comparison of the
Colloidal TiO2 solutions were prepared by hydrolysis of TiCl4
14
photocatalytic activity for H2 generation between Au-TiO2 and
Pt-TiO2 in illuminated aqueous 5 M C2H5OH suspension
showed that the overall activity of Au samples was generally
about 30% lower than that of Pt samples.10 The H2 yield was
found to depend on the metal content on TiO2, with a maximum
in the ranges 0.3-1 wt % Pt, and that the rate of H2 production
was strongly dependent on the initial pH of the suspension.
(Aldrich 99.9%). As reported previously, 25 mL of TiO2 were
introduced very slowly under a stream of argon and vigorous
stirring into 700 mL of 0.1 M HCl solution cooled at 0 to 4 °C.
After 30 min stirring at 0 °C, the solution was dialyzed against
HCl solution at pH3 (pore size cutoff 6000-8000). TiO2
concentrations were determined by spectrophotometric measure-
-
1
-1
ments at 215 nm using ꢀ ) 6050 M cm (in terms of TiO2
molecules). Concentrations in the range of 0.16-0.28 M were
obtained as small particles (average diameter of about 1 nm)
and kept in a refrigerator at 2-5 °C. By heating such suspension
at 45 °C for 5 days, the size of the particles increased to a
*
To whom correspondence should be addressed. E-mail: rabani@
vms.huji.ac.il.
†
‡
Institute of Nuclear Sciences Vinca.
The Hebrew University of Jerusalem.
1
0.1021/jp045989g CCC: $27.50 © 2004 American Chemical Society
Published on Web 12/02/2004

2
0292 J. Phys. Chem. B, Vol. 108, No. 52, 2004
Kasarevic-Popovic et al.
H O f e (2.7), OH (2.7), H (0.6), H (0.45), H O (0.7)
-
•
•
diameter of about 5 nm. The pH was adjusted with HCl. The
results are given for the 5 nm diameter, unless otherwise stated.
2
aq
2
2
2
(1)
Pulse Radiolysis. Linear electron beam accelerator (Varian
•
•
7715) operated at 5 MeV and 200 mA current with 1.5 µs pulse
In the presence of 2-propanol, the OH and H radicals abstract
duration was used. Experiments were carried out in a 1 cm
quartz cell (3 cm light path) and temperature (22 ( 2) °C at
hydrogen from the alcohol to produce the 2-propanol radical
•
(
CH3)2C OH as shown in reactions 2 and 3
-
5
the maximum dose, yielding about 1.5 × 10 M reducing
•
•
-
•
•
OH + (CH ) CHOH f H O + (CH ) C OH
(2)
(3)
radicals per pulse (assuming Geaq + GOH + GH ) 6, where
Gx stands for the number of X molecules produced by 100 eV
absorbed radiation). Dosimetry was carried out by measuring
3 2
2
3 2
•
•
H + (CH
)
CHOH f H
+ (CH
)
C OH
3
2
2
3
2
•
-
the 480 nm absorption of (SCN) 2 produced in 5 mM N2O
saturated SCN solution at neutral pH, using ꢀ(SCN) ) 7600
M
-
-
2 6
In the absence of oxygen and H PtCl , all of the reducing
2
-1
-1
-
radicals react with TiO2 nanocrystallites to produce the stable
cm . The absorption changes of the reactions of eTiO2
-
eTiO2 radical (reactions 4-6)
were monitored at 600 nm.
Preparation of Pt Coated TiO2. A total of 5 mL of TiO2
0.16-0.2 M) containing 0.2 M 2-propanol were Ar saturated
eaq- + TiO f e
-
(4)
(5)
(6)
2
TiO2
(
in a test tube sealed with a rubber septum and irradiated either
•
-
-
H + TiO f e
by γ rays or electron beam to produce about 1 mM eTiO . Under
2
TiO2
2
-
these conditions, the eTiO2 is very stable, but upon addition of
•
-
chloroplatinic acid, a fast reaction takes place, reducing the
platinum ions to the metallic form. An amount of H2PtCl6
solution ([H2PtCl6] < [eTiO2-]) saturated with argon was injected
through the septum to reach the desired concentration of the
reduced platinum. Excess electrons remaining at the end of the
reduction process were destroyed by exposure to oxygen.
(CH ) C OH + TiO f e
3
2
2
TiO2
-
Pulse Radiolysis Experiments. Reaction of eTiO with
H2PtCl6. The reduction of H2PtCl6 by eTiO2 (reaction 7) was
determined by pulse radiolysis of argon saturated solutions
containing 0.2 M TiO colloid of big particles, or 0.28 M TiO2
2
-
2
Slow Kinetics. For measurements of reaction rates in the time
range of several hours, solutions of TiO2 colloid coated or
uncoated with platinum were transferred into a quartz cell
connected to a vacuum valve and argon was bubbled through a
sidearm with the aid of a capillary polyethylene tubing for at
least 20 min. After bubbling the valve was closed under a strong
flow of argon and the cell was taken for electron beam
irradiation receiving 30-40 consecutive pulses of about 30 Gy
colloid of small particles, in the presence of 0.2 M (CH3)2CHOH
at pH (2.9-3.0). H2PtCl6 concentrations were varied between
-
4
(0.4-1.25) × 10
M
-
TiO2
0
e
+ H PtCl f f f Pt
(7)
2
6
We were unable to separate the various reduction stages,
0
although the final product is Pt . Under our experimental
-
conditions, >95% of the hydrated electrons produced by the
each to load the TiO2 with excess eTiO2 . Following the
-
pulse reacted with the TiO2 colloid to produce eTiO2 , rather
irradiation the cell was placed in a thermostated holder (25 °C)
connected to a Hewlett-Packard diode array spectrophotometer
for the kinetic measurements.
than with H2PtCl6. This is obvious when considering the rate
-
8
9
-1 -1
constant of eaq with TiO2 (2.8 × 10 and 1.1 × 10 M
s
1
4
for the big and small particles, respectively), compared to 2
Hydrogen Determination. In each experiment, three test
tubes containing 5 mL of argon saturated TiO2 solutions coated
with platinum were bubbled with oxygen to destroy any traces
1
0
-1 -1
-
15
×
10
M
s
for the reaction of eaq with H2PtCl6, and
taking into account the relative concentrations of TiO2 and H2-
PtCl6. Similarly, H radicals will react mainly via reaction 3
rather than reactions 5 or 8: k3 ) 7.5 × 10 M
•
-
of eTiO2 and subsequently deaerated with argon. These solutions
8
-1 -1 15
s
compared
were irradiated again in a 137Cs γ source for 2-3 h, receiving
a total dose of (1.0-1.7) kGy. The irradiation was carried on
together with three reference samples receiving exactly the same
dose. The reference samples contained aqueous 2-propanol (0.2
M) in the absence of TiO2/Pt at pH 2.5. The reference samples
served for determination of the hydrogen yields by comparison
between the hydrogen signals in the reference and in the TiO2/
Pt solutions. The test tubes were left at room temperature for
to
•
+
0
H + H PtCl f f f H + Pt
(8)
2
6
k5 ) 10 M s-1 14 and k8 ) 7 × 10 M s . (CH3)2C OH
6
-1
9
-1 -1 15
•
radical may react with H2PtCl6 in addition to the reaction with
TiO2 and might interfere with the kinetic measurements of
-
reaction 7. It was observed that the initial absorption of eTiO2
1
7-20 h to ensure completion of the reaction producing
at 600 nm decreased as the concentration of the chloroplatinic
acid increased, from 0.022 in the absence of H2PtCl6 down to
hydrogen. The test tubes were vigorously shaken before injecting
a given volume from the gas phase to the HP-5890 gas
chromatograph (TCD detector, Porapak column from Supelco
and argon as the gas carrier). Calibration of peak areas versus
injected volumes of H2/Ar mixtures showed linear response to
H2.
0.01 at the highest concentration of the chloroplatinic acid used.
This observation confirms the competition between TiO2 and
•
H PtCl for the (CH ) C OH radical. Nevertheless, in all the
2
6
3 2
experiments carried out to determine k , a vertical increase of
7
the absorptions was observed at 600 nm after the pulse, so that
-
reactions 4-6 presenting the formation of eTiO2 were com-
pletely separated in time from the decay via reaction 7. The
Results and Discussion
full decay of reaction 7 took place in the 10-40 ms time scale,
-
The primary radicals and molecules produced in water upon
electron-beam pulse or γ irradiation are shown on the right-
hand side of eq 1. The numbers in parentheses represent the
respective G values
whereas the formation of eTiO2 was completed within the sub
ms time scale. A minor residual decay in the 1 to 2 s region
was also observed. A typical example of the absorption decay
at 600 nm is shown in Figure 1. Similar pulse experiments were

Excess Electrons in TiO2 Nanoparticles
J. Phys. Chem. B, Vol. 108, No. 52, 2004 20293
-
Figure 1. Decay of eTiO2 in the presence of H
2
PtCl
6
Followed at 600
nm after pulse radiolysis of argon saturated solution containing 0.2 M
-
5
TiO
2
(big particles), 9.5 × 10 M H
2 6 3 2
PtCl , 0.2 M (CH ) CHOH at
pH 3.0. The line represents first-order rate law.
-
+
Figure 3. Decay of eTiO2 by reaction with H . Absorbance induced
by 30 pulses (total 900 Gy) in 0.2 M TiO
of 0.2 M 2-propanol at pH 2.4. Big TiO
2
saturated Ar, in the presence
particles coated with 2.5 ×
2
-
4
0
1
0
M Pt .
Figure 2. Pseudo first-order rate constant of reaction 7 as a function
2
of [H PtCl
6 2
]. 32 Gy per pulse, 0.2 M TiO (big particles), 0.2 M
2
2
-propanol at pH 3.0, Ar saturated Straight line corresponds to k )
6
-1 -1
.5 × 10 M
s
.
-
+
Figure 4. Catalyzed reaction of eTiO2 with H . Ar saturated solutions
0
of TiO
as in Figure 3. [TiO
2-propanol] ) 0.2 M.
2
coated with Pt . The rate Constant calculated from plots such
carried on with several chloroplatinic concentrations, showing
linear dependence of the decay rate on [H2PtCl6], yielding k7
)
1
0
-4
2
] ) 0.2 M, big particles, [Pt ] ) 2 × 10 M,
[
6
-1 -1
2.5 × 10 M
s
for the big particles (Figure 2) and 6.1 ×
6
-1 -1
-
0 M
s
for the small ones.
TABLE 1: Reaction of eTiO2 with Oxygen in 0.2 M TiO
2
0
Coated with Different Concentrations of Pt in Air Saturated
Comparison of Reactivity of Bare TiO2 Colloid Particles
-
4
Solutions (O
2
) 2.4 × 10 M) Containing 0.2 M
-
with Pt Coated Particles. Reaction of eTiO with Oxygen. The
reaction of eTiO2 with oxygen was measured in air saturated
solutions at pH 3.2 and in the presence of 0.2 M 2-propanol.
The decay of eTiO2 at 600 nm was followed after pulse
irradiating solutions of TiO2 coated with Pt and compared to
the decay of eTiO2 in the absence of Pt . The predominant part
of the initial absorbance decays by a biexponential rate law, in
the presence as well as in the absence of Pt (reaction 9).
Second-order rate constants of the fast and the slow decays
2
(CH ) CHOH at pH 3.2
3
2
-
0
average k9s M-1 s-1
average k9f M-1 s-1
Pt M
-
4
7
0
(2.7 ( 0.3) × 10
(1.9 ( 0.4) × 10
-
-
5
4
7
0
4.8 × 10
(2.0 ( 0.6) × 10
7
1
2
.1 × 10
.3 × 10
(1.7 ( 0.5) × 10
-
0
-4
4
7
(2.1 ( 0.7) × 10
(1.9 ( 0.7) × 10
0
many days. We observed that eTiO2-, produced by a burst of 50
pulses, 30 Gy each on TiO2 coated with Pt , disappears within
several hours. Figure 3 presents a typical decay of eTiO
0
(reaction 9f and 9s, respectively) have been measured at different
-
0
2
concentrations of Pt and compared to solutions without the
platinum coating. The average results of 4-7 repetitive mea-
surements at each concentration are summarized in Table 1,
-
4
produced in deaerated 0.2 M TiO2 coated with 2 × 10
M
0
Pt . The catalytic reaction obeys the same first order rate law
throughout the decay process and reaches its end after practically
all of the electrons (>90%) have been reacted, with no
0
showing unequivocally that under our conditions, Pt has no
catalytic effect on the reduction of oxygen by eTiO2-
0
observable changes in the Pt .
-
+
0
-
eTiO2 + O (+H ) f HO + TiO
(9)
At a given Pt concentration, the decay rate of eTiO increased
linearly with H concentration. This is shown in Figure 4. The
linear dependence suggests that hydrogen is produced by
reductions of H and to a lesser extent by reduction of H2O.
The reaction with H2O is introduced in order to rationalize the
intercept observed in Figure 4. The Pt mediated reduction of
2
2
2
2
+
7
-1
An average value of the fast decay k9 ) (1.9 ( 0.6) × 10 M
-
1
+
s
is calculated.
-
+
Reaction of eTiO2 with H . The typical blue color of colloidal
-
0
eTiO2 in acid solution is known to be stable under argon for

20294 J. Phys. Chem. B, Vol. 108, No. 52, 2004
Kasarevic-Popovic et al.
0
number of Pt atoms per TiO2 nanoparticle is considerably
0
smaller: The formal concentration of Pt (Figure 5) is in the
-
5
-4
range 5 × 10 - 2.5 × 10 M, meaning that on the average
0
each colloid particle is loaded with only 0-3 atoms of Pt
depending on the Pt concentration) and most TiO2 surface is
0
(
not covered with platinum. A significant fraction of the TiO2
nanocrystallites has no Pt . If the decay of eTiO involves Pt0
0
-
2
and electrons on the same nanocrystallite, only partial decay of
-
-
eTiO2 is expected, contrary to the observation that the eTiO2
absorbance decays nearly to zero by a single-exponential
process. Our observations can be rationalized by aggregation
of the TiO2 nano-particles as a three dimensional network,16
which enables electron hopping between the nanocrystallites
0
0
to Pt sites. In addition, migration of Pt through the aqueous
suspension, in line with the Ostwald ripening, enables the Pt0
catalyst activity at a number of TiO2 aggregates.
Figure 5. Catalyzed decay of eTiO2- as a function of Pt0 catalyst
0
GC Measurements. H2 yields developed from reactions 1, 3,
12, and 13 have been determined after irradiating TiO2 solutions
in the presence as well as in the absence of platinum, at acid
pH containing 0.2 M (CH3)2CHOH. Blank argon saturated
aqueous solutions containing only (CH3)2CHOH at the same
pH were irradiated. In the blank solution, H2 originates from
concentration. Pt coated on TiO
from plots such as Figure 3, [TiO
0.2 M, Ar saturated. The line corresponds to k ) 0.52 M
2
at pH 2.5. The rate constant taken
2
] ) 0.2 M, big particles, [2-propanol]
-
1
-1
)
s .
+
H apparently involves adsorbed hydrogen atoms (reaction 10
and 11)
-
+
three sources: (i) Reaction of eaq (G ) 2.7) with H to give
-
+
0
•
0
•
•
eTiO2 + H (Pt ) f H (Pt ) + TiO
(10)
(11)
H followed by reaction 3; (ii) primary H atoms (G ) 0.6)
reacting according to (3); (iii) molecular yield of H2 (G ) 0.45).
The total yield in the blank solution sums up to G ) 3.75. The
evolution of H2 in TiO2 solutions in the absence of platinum is
expected to be lower: At the pH of the experiments (pH 2.7),
2
-
0
•
0
-
eTiO2 + H O(Pt ) f H (Pt ) + TiO + OH
2
2
-
•
0
+
+
-
there is competition between H and TiO2 for the eaq . Taking
eTiO2 + H (Pt ) + H f H
(12)
(13)
2
-
+
10
-1 -1 15
the rate constant of eaq with H as 2.2 × 10
M
s
and
-
8
-1 -1 14
eaq with TiO2 as 2.6 × 10 M s , 50% of the solvated
•
0
•
0
H (Pt ) + H (Pt ) f H
+
2
electrons react with H to produce H2 (G ) 1.3). The rest of
-
the solvated electrons produce eTiO , which in the absence of
2
0
0
It should be emphasize that the presence of Pt is essential for
reaction 10 and 11 to take place. The adsorbed hydrogen radicals
from reaction 10 and 11 produce molecular hydrogen by
reactions 12 and 13, rather than hydrogen abstraction from
Pt do not produce H . Since practically all H atoms produce
2
H according to reaction 3, the calculated yield of H is 1.3 +
2
2
0.45 + 0.6 ) 2.35. The experimental result from six different
solutions showed an average yield of 2.4 ( 0.3, in very good
agreement with the expected value.
(
CH3)2CHOH (reaction 3). This will be further discussed later.
2
0
-1 -1
0
Our results yield average values k10 ) 1.3 × 10 [Pt ] M
s
When Pt is also present, the H yield is increased due to
2
0
-1
•
and k11) 0.1[Pt ] s .
reactions 10-13. Furthermore, the OH radicals reacting with
2-propanol produce (CH3) C OH radicals, which eventually
•
Kinetic measurements at different platinum concentrations
2
-
-
and at a given pH show that the rate of eTiO decay increases
linearly with Pt concentration. This is shown in Figure 5.
reduce TiO to form e
according to reaction 6. This further
2
2
TiO2
0
-
increases the expected total H yield. Had e
2
TiO2
reacted with
0
+
The linear increase is expected if the added Pt increases the
H /water producing free H atoms, the total molecular hydrogen
yield would have increased to at least 3.75 + 2.7 ) 6.45. It
should be noted that the expected yield in the case of hydrogen
abstraction from the 2-propanol by the hydrogen atom inter-
mediate must be even higher, because generation of free H atoms
0
number of catalytic sites, which may be aggregates of Pt settled
on the TiO2 surface or separate Pt atoms. The possibility that
increasing Pt concentration enhances the catalysis by increasing
0
0
the surface area of existing aggregates can be ruled out, as in
such a case where the rate is expected to increase with the 2/3
power of the concentration rather than the observed linear
dependency. This conclusion is also in agreement with an earlier
-
from eTiO2 would have initiated a chain reaction producing H .
However, the energy of eTiO2 is not sufficient for generation
2
-
of free hydrogen atoms, and consequently, adsorbed hydrogen
0
0
observation that the Pt aggregates grow up to a constant average
atoms recombine on the surface of Pt to produce molecular
0
11
size, which does not depend on the Pt concentration.
hydrogen (reaction 13). In such mechanism taking into consid-
It is worthwhile to consider the distribution of eTiO2 and Pt0
on the TiO2 particles. The average number of TiO2 molecules
per particle in solutions of big particles (diameter 5 nm) is
estimated as about 2000. This value is derived from the
-
+
-
eration the competition between H and TiO for e , a total
2
aq
hydrogen yield of 4.7 is predicted. The average of six different
0
2
TiO /Pt solutions at pH 2.3 showed a yield of (4.8 ( 0.6), in
excellent agreement with the value predicted above on the basis
of adsorbed hydrogen atom recombination.
3
equation: n ) πd FA/6m, where n is the number of TiO2
molecules in a particle, d is the particle diameter in cm, F is
3
Conclusions
the density of TiO2 taken as 4 g/cm , A stands for Avogadro’s
number, and m is the molecular weight of TiO2. This calculation
The reaction of eTiO2- with water and hydrogen ions is easily
-
4
-
leads to a concentration of about 10 M of colloid particles in
separated in time from the generation of eTiO2 by pulse
-
the 0.2 M TiO2 solutions. Thus, the initial concentration of eTiO
radiolysis. Thus, we have been able to study the kinetics of the
electron decay as a function of pH. The pseudo-first-order
kinetics and the first-order rate dependency on H and Pt0
2
produced by the electron beam (about 10-3 M) corresponds to
an average of 10 electrons per TiO2 nanoparticle. The average
+

Excess Electrons in TiO2 Nanoparticles
J. Phys. Chem. B, Vol. 108, No. 52, 2004 20295
References and Notes
(1) Graetzel, M. Heterogeneous Photochemical Electron Transfer; CRC
Press: Boca Raton, FL, 1989.
concentrations show that the rate determining step of the
catalytic reduction is the transfer of the TiO2 excess electron
0
+
from the Pt cluster to H and H2O. The results enable the
(
2) Wood, A.; Giersig, M.; Mulvaney, P. J. Phys. Chem. B 2001, 105,
810.
(3) Chiang, K.; Lim, T. M.; Tsen, L.; Lee, C. C. Appl. Catal. A 2004,
61, 225.
4) Crittenden, J. C.; Liu, J. B.; Hand, D. W.; Perram, D. L. Water
+
elucidation of the separate reaction rate constants for H and
water.
8
2
Pt0 as well as other metals are well-known to enhance
(
photocatalytic oxidations of organic solutes. Wang, Heller, and
Res. 1997, 31, 429.
(5) Choi, W.; Lee, J.; Kim, S.; Hwang, S.; Lee, M. C.; Lee, T. K. J.
Ind. Eng. Chem. 2003, 9, 96.
(6) Einaga, H.; Harada, M.; Futamura, S.; Ibusuki, T. J. Phys. Chem.
B 2003, 107, 9290.
Gerischer12 have shown that catalyzed oxygen reduction by
0
surface Pd increases photocatalytic oxidation yields by prevent-
ing accumulation of electrons on the TiO2 particles, thus
(
7) Li, F. B.; Li, X. Z. Chemosphere 2002, 48, 1103.
inhibiting the rate of electron-hole recombination. On the other
(8) Bahnemann, D. W.; Hilgendorff, M.; Memming, R. J. Phys. Chem.
8
hand, Bahnemann, Hilgendorff, and Memming attributed the
B 1997, 101, 4265.
(
9) Kisch, H.; Macyk, W. ChemPhysChem 2002, 3, 399.
enhanced oxidation to the metal effect on increasing charge
separation. The present work supports the latter conclusion,
although the role of platinum may differ from that of palla-
dium.
(
10) Bamwenda, G. R.; Tsubota, S.; Nakamura, T.; Haruta, M. J.
Photochem. Photobiol. A-Chem. 1995, 89, 177.
(11) Kiwi, J.; Graetzel, M. J. Phys. Chem. 1984, 88, 1302.
(
12) Wang, C. M.; Heller, A.; Gerischer, H. J. Am. Chem. Soc. 1992,
14, 5230.
13) Gao, R. M.; Safrany, A.; Rabani, J. Radiat. Phys. Chem. 2003, 67,
1
(
2
5.
(
Acknowledgment. We express our gratitude to Eran Gilead
for maintenance of the electron beam system and to Yair Ozery
for assistance with gas chromatography measurements. This
research was supported by the European Commission RTD
framework 6th program and by the Berman Foundation.
14) Gao, R. M.; Safrany, A.; Rabani, J. Radiat. Phys. Chem. 2002, 65,
599.
(
15) Neta, P.; Huie, R.; Ross, A. B. J. Phys. Chem. Ref. Data 1988, 17,
1027.
(
16) Wang, C.; B o¨ ttcher, C.; Bahnemann, D. W.; Dohrmann, J. K. J.
Nanoparticle Res. 2004, 6, 119.