Photodynamics and surface characterization of TiO2 and Fe2O3 photocatalysts immobilized on modified polyethylene films

Article abstract of DOI:10.1021/jp011339q

Polyethylene block-copolymer films containing negative anhydride groups were used to immobilize TiO₂, Fe₂O₃, and Fe³⁺ photocatalysts. The kinetics of the mineralization of azo-dye Orange II and chlorophenols on copolymer-TiO₂, copolymer-Fe₂O₃, and copolymer-Fe³⁺ have been tested under optimized experimental conditions. In the case of copolymer-TiO₂, the degradation kinetics for the model organic compounds were about the same as those observed with TiO₂ suspensions containing about a 27 times higher amount of TiO₂ per unit volume. The surface of the derivatized copolymer semiconductor catalysts was studied by infrared attenuated total reflection spectroscopy. The spectroscopic data provided evidence for a TiO₂ interaction with the negatively charged conjugated carboxylic groups of the copolymer, leading to an asymmetric-stretching band of -COO-Ti⁴⁺ at the position expected for metal carboxylates. In the case of Fe₂O₃ and Fe³⁺, the asymmetric-stretching carboxylate bands are ascribed to the carboxylate bands of -COO-Fe₂O₃ and -COOO-Fe³⁺. Evidence is presented by X-ray photoelectron spectroscopy for the existence of two oxidation states of Ti and Fe after the photocatalytic degradation of Orange II. This observation is consistent with light-induced interracial charge transfer (redox processes) taking place at the metal-oxide copolymer surface. The nature of the latter processes is presented in detail during this study.

Full text of DOI:10.1021/jp011339q

12046
J. Phys. Chem. B 2001, 105, 12046-12055
Photodynamics and Surface Characterization of TiO2 and Fe2O3 Photocatalysts Immobilized
on Modified Polyethylene Films
†
⊥
†
‡
§
|
M. R. Dhananjeyan, E. Mielczarski, K. R. Thampi, Ph. Buffat, M. Bensimon, A. Kulik,
⊥
,†
J. Mielczarski, and J. Kiwi*
Department of Chemistry, Laboratory of Photonics and Interfaces, Institute of Electron Microscopy,
Department of CiVil Engineering, and Department of Physics, Institute of Atomic Engineering,
Swiss Federal Institute of Technology, CH-1015 Lausanne, Switzerland, and Laboratory LEM/UMR 7569,
CNRS INPL-ENSG, BP40, 54501 VandoeuVre les Nancy Cedex, France
ReceiVed: April 10, 2001; In Final Form: August 16, 2001
2
Polyethylene block-copolymer films containing negative anhydride groups were used to immobilize TiO ,
2 3
O
, and Fe3 photocatalysts. The kinetics of the mineralization of azo-dye Orange II and chlorophenols
+
Fe
on copolymer-TiO
conditions. In the case of copolymer-TiO
about the same as those observed with TiO
per unit volume. The surface of the derivatized copolymer semiconductor catalysts was studied by infrared
attenuated total reflection spectroscopy. The spectroscopic data provided evidence for a TiO interaction with
the negatively charged conjugated carboxylic groups of the copolymer, leading to an asymmetric-stretching
3+
2
, copolymer-Fe
2
O
3
, and copolymer-Fe have been tested under optimized experimental
, the degradation kinetics for the model organic compounds were
suspensions containing about a 27 times higher amount of TiO
2
2
2
2
band of -COO-Ti4 at the position expected for metal carboxylates. In the case of Fe
+
O
3
3
and Fe , the
3+
2
asymmetric-stretching carboxylate bands are ascribed to the carboxylate bands of -COO-Fe
2
O and -COOO-
3+
Fe . Evidence is presented by X-ray photoelectron spectroscopy for the existence of two oxidation states of
Ti and Fe after the photocatalytic degradation of Orange II. This observation is consistent with light-induced
interfacial charge transfer (redox processes) taking place at the metal-oxide copolymer surface. The nature
of the latter processes is presented in detail during this study.
Introduction
adequate long-term catalytic stability, and (c) preclude semi-
conductor leaching during the degradation process. Derivatized
polyethylene has been selected because polyethylene is the most
inert polymer material from DuPont after Teflon.
During the past decade, heterogeneous photocatalysis has
been increasingly used in the preparation of fine chemicals and
in water and air purification. More than 1360 references have
appeared during the last 10 years in internationally reviewed
The catalytic performance and characterization of these new
materials will be explored during the degradation of a model
azo-dye (Orange II) and some common chlorocarbons. Non-
biodegradable azo-dye Orange II is widely used in the textile
industry. It is resistant to bacterial treatment; therefore, biological
journals where TiO21-10 and Fe2O311 are used as photocatalysts.
However, two major obstacles hinder the TiO2 photocatalytic
performance of these semiconductors in suspension: (a) the
separation of the semiconductor catalyst after the treatment and
3
degradation cannot be employed to abate this compound. The
(b) the low quantum efficiency of these processes. This study
4
TiO2-mediated degradation of Orange II and other azo-dyes,
addresses the first issue, namely, the anchoring of TiO2, Fe2O3,
and Fe3 on copolymers to avoid the costly separation of the
suspensions after treatment. Supported titania-coated glass has
been reported during the past decade and led to catalyst
+
5
6
such as Rose Bengal and Eosin, is a field of active research.
The removal of chlorophenols on TiO2 semiconductor
suspensions under light irradiation has been widely reported
1
2,13
1,2,7-9
dissolution
during the degradation of pollutants. More
during the past decade
because they are EPA priority
recently, Nafion-Fe catalytic membranes have been shown to
be effective as photocatalysts and resistant to the attack of OH
radicals with adequate kinetics and stability during long-term
operation.14 However, Nafion is too expensive to be used as a
catalyst support in large-scale applications. The present study
presents a new copolymer15,16 with TiO2 or Fe2O3 immobilized
on its surface that is able to (a) withstand reactive radical attack
during light irradiation in an oxidative media, (b) maintain
pollutants. The degradation of 2,4-dichlorophenol (2,4-DCP)
mediated by TiO2 suspensions under light has been recently
10,11
reported.
Because of the long time needed for biological
degradation, 2,4-DCP goes unabated through biological treat-
ment stations and gives rise to a serious environmental problem.
During this study, the catalyst performance will be shown to
depend on the suitable combination of (a) the type of the
semiconductor applied on the copolymer surface, (b) the oxidant
leading to the most effective separation, and (c) the radiation
frequency and intensity of the lamp source. Both O2 and H2O2
have been used as oxidants for the reactions. In the case of
copolymer-Fe2O3, H2O2 is the only effective oxidant, whereas
in the case of copolymer-TiO2, O2 was able to carry out the
oxidation reaction.
†
Department of Chemistry, Laboratory of Photonics and Interfaces, Swiss
Federal Institute of Technology.
⊥
Laboratory LEM/UMR 7569, CNRS INPL-ENSG.
Institute of Electron Microscopy, Swiss Federal Institute of Technology.
Department of Civil Engineering, Swiss Federal Institute of Technology.
‡
§
|
Department of Physics, Institute of Atomic Engineering, Swiss Federal
Institute of Technology.
1
0.1021/jp011339q CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/08/2001

TiO2 and Fe2O3 Photocatalysts
J. Phys. Chem. B, Vol. 105, No. 48, 2001 12047
Experimental Section
measurement was performed with a takeoff angle close to 90°.
High charging effects on the samples were observed at 2 and 4
eV. The carbon 1s line (284.6 eV) was used as a reference to
correct the charging effects. The characteristic peaks of Ti 2p,
Fe 2p, O 1s, and C 1s, were recorded in the appropriate range
of the binding energies in order to obtain a good signal-to-noise
ratio for the best resolution of the recorded spectra. The recorded
lines were smoothed by a polynomial fit and background
Chemicals. Orange II, 2-chlorophenol, 4-chlorophenol, 2,4-
DCP, and H2O2 were Fluka p.a. and used as received. Triple-
distilled water was used in all of the experiments. Titania TiO2
2
P-25 Degussa (50 m /g) was deposited on polyethylene-maleic
anhydride block-copolymer films (DuPont Co.; 30 µm thick)
by the procedure outlined as follows.1
5,16
Preparation of the Semiconductor-Loaded Copolymer.
The immobilized anhydride-derivatized polyethylene-TiO2 was
prepared in the following way. First, the polyethylene film was
thoroughly washed with water. Then, the membrane was dipped
in a TiO2 P-25 suspension (5 g/L) for several hours. This titania
suspension was heated for 1 h at 75 °C in the presence of the
polyethylene copolymer, and subsequently, the copolymer-TiO2
film was taken out and dried at 100 °C for 1 h. The copolymer
film was then washed to eliminate the loosely bound TiO2
particles. The final copolymer film had a titania loading of 1.5
17
subtraction according to Shirley and then fitted using a curve-
fitting program provided with a Gaussian and Lorentzian curve
generator.
Infrared Attenuated Total Reflection (IRATR) Fourier
Transform Spectroscopy. This technique was applied to study
the surfaces of the unloaded and loaded copolymer-TiO2,
3+
copolymer-Fe2O3, and copolymer-Fe . The reflection spectra
were recorded on a Bruker IFS 55 FTIR spectrophotometer
equipped with a mercury cadmium telluride detector and an
internal reflection attachment from Harrick Co. The detailed
description of the applied ATR technique can be found in ref
2
mg spread on a 48 cm area of the copolymer (100 mg). The
titania deposition could be reproduced on the copolymer film
within (15% accuracy by weight of the deposited TiO2. The
hematite-copolymer samples were prepared similarly using
1
8. The polyethylene samples were contacted with a ZnSe
reflection element with dimensions of 50 × 20 × 3 mm and an
incident angle (Θ) of 45°. The depth of penetration (dp) of
incident radiation, defined as the distance required for the
3
+
hematite instead of TiO2. For the anchoring of Fe , FeCl3
(Fluka) was used as the starting material.
-1
electric-field amplitude to fall to e , is calculated from
Photoreactor and Irradiation Procedures. The irradiation
vessels were 60 mL cylindrical Pyrex flasks (cutoff 290 nm)
each containing 40 mL of a reagent solution. Copolymer strips
of 48 cm were placed immediately behind the wall of the
2
2 1/2
21
d ) λ /2 (sin Θ - n
)
(1)
p
1
∏
2
The refraction index used was n2 ) 1.4 for polyethylene and
n1 ) 2.4 for the reflection of the elements. This allowed for an
estimation of 2.3 µm for the penetration depth at a wavenumber
reaction vessel. The titania suspensions as well as the copolymer-
TiO2 strips were irradiated in the UV region using a medium-
pressure 125 W mercury lamp (360° radiation field at ∼2.5 ×
-
1
15
of 1700 cm . The reflection spectra were recorded for both
sides of each copolymer sample, showing no differences.
High-Resolution Inductively Coupled Plasma Spectrom-
etry (ICPS). The samples were acidified with nitric acid and
diluted in ultrapure water. The experiments were directed toward
1
0
photons/s) from Photochemical Reactors Limited, Berk-
shire, U.K. The main irradiation wavelength of the mercury lamp
was 366 nm, and the light of <290 nm was cut by the batch
reactor Pyrex wall.
The irradiation of polymer-Fe2O3 was carried out in the
visible region in the cavity of a Hanau Suntest solar simulator
air-cooled at 46 °C. The Suntest lamp had a wavelength
distribution with about 7% of the photons between 290 and 400
nm. The profile of the photons emitted between λ ) 400-800
nm followed the solar spectrum. The Suntest solar simulator
4
+
the determination of the Ti ion in the irradiated solutions.
The ion beam in the plasma of the instrument (Micromass, U.K.)
was directed through the sampling interface and then accelerated
into the mass analyzer. The collector assembly is a dual-detector
system provided with a Faraday cup for the high-beam current
coupled with an electron-multiplier amplifier for the low-
intensity signals. To overcome the overlap interference prob-
lems, measurements were referenced to a mass resolution of
3500.
2
had an intensity of 80 mW/cm within the wavelength range of
2
2
90 and 800 nm. The radiant flux in mW/cm was measured
with a power meter from LSI Corp., Yellow Springs, CO.
3+
Photolysis of Solutions. The detection of Fe in solution
3+
Transmission Electron Microscopy (TEM) and Diffraction
Measurements. A Philips EM 430 Twin (300 kV and 2.3 Å
point resolution) instrument was used for TEM and diffraction
measurements. The samples were prepared by ultramicrotomy
using a diamond knife (45° edge) to cut them into thin slices,
and then the samples were subsequently embedded in an Embed
was carried out by complexing the Fe ion in solution with
thiocyanate and the Fe2+ ions with phenanthroline. Spectro-
photometric analyses of the solutions were carried out with a
Hewlett-Packard 8452 diode array spectrophotometer. The total
organic carbon (TOC) was monitored via a Shimadzu 500
instrument equipped with an automatic sample injector. Dye
degradation was followed by a Varian HPLC provided with a
812 epoxy resin. Different cutting geometries were tried for the
softer polymer oxide loaded samples.
9
065 diode. A Phenomenex C-18 inverse-phase column was
Gas Adsorption Studies. Gas physisorption was carried out
using a Sorptomatic 1990 Micropore unit. The adsorption
isotherms were reproducible to 2%. The experiments were
performed at the liquid-nitrogen boiling temperature (77 K).
The surface area of the samples was evaluated with the help of
a 19190 Micropore computer system.
used, and the dye signals were detected at 229 and 254 nm.
The gradient solution was regulated with a buffer consisting of
ammonium acetate and methanol. The peroxide concentration
in the solution was followed by Merkoquant paper at levels
between 0.5 and 25 mg/L of H2O2 in the aqueous solution.
X-ray Photoelectron Spectroscopy (XPS). The XPS spectra
were obtained with a multidetection electron-energy analyzer
Results and Discussion
(VSW FAT mode CL 150). The monochromatic magnesium
radiation source (1253.6 eV) was operated at 15 kV and 10
mA. The XPS resolution between 0.9 and 1.0 eV was deter-
mined by the use of a gold standard. The operating pressure of
Decoloration and Degradation of Orange II on Copoly-
mer-TiO2 Membranes. Figure 1 presents the UV-vis spectra
of the copolymer alone and copolymer-TiO2. The increase in
the absorbance on the copolymer is due to the deposited TiO2.
-8
the spectrometer analyzer chamber was close to 10 mbar. The

12048 J. Phys. Chem. B, Vol. 105, No. 48, 2001
Dhananjeyan et al.
2
Figure 1. Absorption spectra of the copolymer and of the TiO -coated
copolymer.
Without the presence of copolymer-TiO2, the decoloration of
Orange II (0.1 mM) by H2O2 (10 mM) under mercury-lamp
irradiation did not proceed. The results indicate that catalytic
sites at the surface of titania, whether in a suspended or an
immobilized form, are responsible for the abatement of Orange
II. In the dark, no decrease in the Orange II absorbance was
observed.
Figure 2a shows the result of the decoloration of Orange II
using titania suspensions and that of copolymer-TiO2 in the
presence of H2O2 under medium-pressure mercury light (400
W) and in the dark. The peroxide in the solution decreased from
initially 340-25 mg/L after 6 h.
Figure 2b shows that the decrease in TOC under the
irradiation of Orange II on copolymer-TiO2 proceeds to a level
of 8 mg of C/L. The shoulder that sets in after 4 h of degradation
in Figure 2b does not correspond to copolymer decomposition
under light irradiation resulting from the H2O2 present, as
indicated by control experiments. The shoulder is due to the
nondegradable intermediates generated during the degradation
of Orange II under light. The upper trace in Figure 2b shows
the formation of nondegradable intermediates in the dark,
precluding further decomposition of Orange II. This is in
4,14,16,18-20
agreement with previous work with TiO2 suspensions.
The reduction in TOC is seen to be slower than the Orange
decoloration shown in Figure 2a. Figure 2b shows that the
decrease in TOC of Orange II under light irradiation is higher
when TiO2 suspensions are used. It should be noted that the
concentration of titania is about 27 times higher in the
suspension when compared to the amount anchored on the
polymer.
Figure 2. (a) Absorbance of Orange II (0.1 mM) during the
decoloration at a pH of 6 in the presence of H
medium-pressure mercury-lamp irradiation. The supported catalyst was
copolymer-TiO , and the suspensions of the TiO catalyst were 1 g/L.
b) TOC decrease of Orange II solutions up to 10 h irradiation time
2 2
O (10 mM) under
2
2
(
with the same copolymer and experimental conditions as those used in
part a.
Orange II has been reported to act as the photosensitizer of
the surface bound TiO2. The low-energy edge of Orange II has
been reported to be 565 nm or 2.3 eV.19 The photosensitization
of Orange II on copolymer-TiO2 leads to a charge transfer with
the concomitant quenching of Orange II* and the formation of
state of the dye to the conduction band of TiO2 at -0.1 eV is
thermodynamically possible. This is shown in Scheme 1.
IR and XPS Characterization of Copolymer and Copoly-
mer-TiO2 before and after Photodegradation of Orange II.
IR spectra of the copolymer and the loaded copolymer-TiO2
polymer before and after the photodegradation of Orange II are
presented in Figure 3. The band assignments are listed in Table
•
+
Orange II (eq 2). The latter species is documented in the
19,20
literature.
On thermodynamic grounds, charge transfer from
Orange II + copolymer-TiO + hν f
2
1
and indicate that the TiO2 is immobilized on the copolymer
[
Orange II*...TiO ]copolymer f
2
through carboxyl groups. The reflection spectrum of copolymer-
Orange II•+ + copolymer-TiO + e
-
(2)
cb
-1
TiO2 shows the growth of bands at 1541 and 1415 cm (Figure
2
3b). These bands correspond to the asymmetric- and symmetric-
Orange II in the excited state to TiO2 is possible. On the basis
of cyclic voltammetry, an oxidation potential of 0.76 V (NHE)
was reported for Orange II.20 Because the low-energy absorption
edge of Orange II is 2.3 eV, the excited-state standard potential
of Orange II can be estimated [from the energy gap between
the ground sate and the excited state (∆G°)] of Orange II to be
stretching vibrations of the carboxylate groups. The position of
the bands characteristic for metal carboxylates reported in Figure
22-26
3 have been well documented in the literature.
Con-
-
1
comitantly, the band at 1793 cm due to the anhydride
carboxylic groups disappears (Figure 3a). The frequency dif-
ference (∆ν) between the asymmetric- and symmetric-stretching
vibrations of the carboxylate group can be correlated with the
2
1
-1.54 eV. Therefore, the electron transfer from the excited

TiO2 and Fe2O3 Photocatalysts
J. Phys. Chem. B, Vol. 105, No. 48, 2001 12049
SCHEME 1
coordination mode of the carboxylate ligand.24,25 The calculated
∆
ν parameter is 126 cm-1 and indicates the formation of a
bidentate ligand between polymer carboxyl groups and the
2
5,26
surface titanium atoms, according to the literature.
The
distinction between chelation or bridging surface complex
formation is difficult in this case without other supporting
information.
Figure 3. Attenuated total reflection infrared spectra of (a) copolymer
Other observed differences in spectra a-c in Figure 3 consist
-1
alone, (b) copolymer-TiO
2
2
before use, and (c) copolymer-TiO after
of the shift from 1300 to 1280 cm of the broad band resulting
from the coupled vibration of C-O and OH groups. Further, a
lower intensity for the OH deformation band was observed at
Orange II degradation at a pH of 6 with visible-light irradiation.
conclusion that, during photodegradation, titanium changes its
oxidation state. Another finding (Table 2) is that the surface of
-
1
9
09 cm . These observations support the conclusion that a
carboxyl-titanium complex is formed when TiO2 particles are
loaded on the copolymer. It should also be noted that there are
still a significant number of carboxylic groups left on the
the TiO particle on the copolymer is more strongly hydrox-
ylated, as reflected in the OH-surface component observed at
531.6 eV compared to the initial P-25 Degussa sample. The
intensity of the surface hydroxyl groups is very high compared
2
-
1
copolymer, as shown by the band at 1714 cm , that are not
involved in binding with the titanium atoms from the loaded
with that of the oxygen signal in TiO at 529.5 eV. Through
2
-
1
particles. A new band was also observed at 1640 cm ,
indicating water adsorption on the loaded TiO2 particles. This
suggests that water adsorbs on TiO2 though they also strongly
interact (chemisorb) with the copolymer surface. The strong
these surface hydroxyl groups, Orange II interacts with TiO2
and undergoes photocatalytic decomposition.
IR and XPS Characterization of Copolymer-Fe O before
2
3
and after Degradation of Orange II. IRATR spectra of the
-
1
increase in absorption below 850 cm (Figure 3b) is charac-
teristic of the vibrations of TiO2. The higher frequency vibration
loaded Fe O polymer before and after the photodegradation
2
3
of Orange II and the band assignments are reported in Table 1.
Fe O is immobilized on the copolymer through carboxyl
4
+
2-
is due to the displacement of Ti relative to the O ions along
2
3
the c axis24 immobilized on the copolymer surface.
-1
groups. The corresponding band at 1520 cm is very weak in
the spectrum of the nonloaded copolymer. The reflection spectra
of polymer-Fe2O3 samples show new bands at 1567 cm and
a stronger band at 1520 cm , and the latter two bands are
assigned to the asymmetric-stretching vibrations of the car-
boxylate groups. Close to the strong -CH2 deformation band
at 1462 cm , a weak shoulder at about 1415 cm is also
observed. This shoulder is due to the symmetric-stretching
vibrations of the carboxylate groups. At the same time, the band
at 1793 cm assigned to carboxyl anhydride almost disappears.
These observations indicate that an iron-carboxylate surface
complex is bonding the Fe2O3 particles to the copolymer surface.
Two bands at 1567 and 1520 cm are observed and assigned
to the asymmetric-stretching vibrations of the carboxylate
groups, suggesting two conformations of carboxylate groups as
The reflection spectrum of copolymer-TiO2 after Orange II
degradation (Figure 3c) shows features almost identical to those
found for copolymer-TiO2 before use (Figure 3b). The more
-
1
-
1
-1
intensive bands at 1017 and 1046 cm are assigned to the C-O
stretching vibration of the intermediates produced during
photocatalytic degradation. The amount of the adsorbed inter-
mediates on copolymer-TiO2 is very low, suggesting that a
decrease in catalyst efficiency with time will also be very low
or will not take place at all. Another assignment of the two
low-intensity bands is related to the stretching vibration of the
C-O groups of the surface carboxylate groups. With the latter
assignment, any significant adsorption of the intermediates
during a catalytic reaction is excluded, suggesting a good
catalytic efficiency of the copolymer-TiO2 catalyst.
-
1
-1
-1
-
1
28
The samples investigated by IR were also characterized by
XPS, taking the TiO2 P-25 Degussa powder as reference sample.
The results obtained are presented in parts a and b of Table 2.
Analysis of the results led to a few interesting observations.
The most interesting are probably the presence of two oxidation
states (Figure 4) of surface titanium species, two doublets of
Ti 2p with an intensity ratio of 1:2, and a splitting of about 5.7
reported for the calcium-oleate surface complex. The forma-
2
+
3+
tion of mixed valance Fe -Fe complexes should also be
considered during carboxylate surface complex formation. The
calculated ∆ν parameters for the observed asymmetric- and
-
1
symmetric-carboxylate vibrations are 152 and 105 cm ,
respectively. According to the literature,
24,25
values of ∆ν lower
-
1
than 200 cm are indicative of chelation or bridging formation.
Both of them could be suggested as responsible for the strong
attachment of the Fe2O3 particles to copolymer surfaces.
Similarly, as for the copolymer-TiO2 sample, a new band
27
eV in samples after the photodegradation of Orange II. The
4+
major component is Ti , typical of TiO2 powder, but also
3
+
∼
35% of the total titanium surface atoms are in the Ti state
-
1
(
see Figure 4 and Table 2b). This observation supports the
at 1640 cm
is observed, indicating a significant water

12050 J. Phys. Chem. B, Vol. 105, No. 48, 2001
Dhananjeyan et al.
, and Copolymer-Fe3
+
TABLE 1: Assignment of Absorbance Bands on the Copolymer, Copolymer-TiO
2
, Copolymer-Fe
2
O
3
before and after Degradation of Orange IIa
polymer
with TiO
polymer
with Fe
polymer
with Fe3
+
2
2
O
3
virgin
polymer
polymer
with TiO
after Orange II
decomposition
polymer
2 3
with Fe O
after Orange II
decomposition
polymer
after Orange II
decomposition
with Fe3
+
assignment
OH stretching
2
-
∼3340 w broad ∼3340 w broad 3386
3386
3473 vs,
3473 vs,
3363 vs
i
3
363 vs
2
915 vs
847 vs
i
i
i
i
i
i
i
CH
stretching
CH asymmetric
2
asymmetric
2
i
i
i
i
2
stretching
1
1
-
793 w
714 s
-
-
-
i
i
-
-
i
i
-
carboxyl (anhydrite)
carboxyl group
adsorbed water
carboxylate asymmetric
stretching
i
i
i
i
1640
i
i
i
1
541
1541
1567, 1520
1567, 1520
1594, 1560,
523
1594, 1560,
1523
1
metal complex
1
1
-
473 vs
462 vs
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
CH
CH
2
deformation
deformation
2
1415
carboxylate symmetric
stretching
1
1
1
∼
379 w
368 w
352 w
1300 w
broad
-
-
-
-
-
i
i
-
CH
CH
3
deformation
deformation
i
i
i
i
i
3
i
i
i
i
i
CH deformation
coupled CO stretching
and OH in-plane
deformation
1280
1280
1285
1285
1300, 1250
1300, 1250
-
1046, 1017
i
1046, 1017
1048, 1016
i
1016
1047, 1018
i
1047, 1018
i
CO stretching of
carboxylate group
OH out-of-plane
deformation
9
09
i
i
below 850 vs below 850 vs
below 850 vs
below 780 vs
below 780 vs
829 broad
829 broad
absorbance due to
Me-O stretching
of loaded catalysts
C-C skeletal vibration
7
19 vs
i
i
i
i
i
i
a
According to refs 22-26. Key to abbreviations: (-) band not observed, (i ) same position as in previous column, (vs) very strong, (s) strong,
(
w) weak.
TABLE 2: Binding Energy of Ti 2p, O 1s, and C 1s
Elements and Atomic Surface Concentration of Detected
Elements for TiO Catalysts
2
(a) Binding Energy (eV)
Ti 2p1/2
Ti 2p3/2
O 1s
sample
powder
polymer-TiO
before use
C 1s
TiO
2
458.3
458.5
464.0
464.2
533.1 531.6 529.5 284.6
533.4 531.9 529.7 284.6
2
polymer-TiO
after use in
2
458.5 458.0 464.2 463.6 533.4 531.9 529.8 284.6
degradation
of Orange II
(b) Atomic Surface Concentration (%)
Ti
O
sample
Ti4
+
Ti3+
H O
2 ads
OHsurf
O
oxides
C
Figure 4. XPS spectrum of copolymer-TiO
2
after degradation of
, whereas
Orange II (Ti 2p line). Component A is characteristic for TiO
2
TiO
polymer-TiO2
before use
2
powder
10.1
4.7
3.3
1.7
7.0
9.7
21.6
12.7
57.8
71.1
3
+
component B is due to Ti surface species.
polymer-TiO
after use in
degradation
of Orange II
2
2.25
1.25
1.4
10.7
9.6
74.5
adsorption on the loaded Fe2O3 particles. The strong increase
-1
in absorption below 780 cm is characteristic for the vibrations
24
of Fe2O3 immobilized on the copolymer surface.
The reflection spectrum of copolymer-Fe2O3 after Orange
II degradation shows spectral features similar to those of the
copolymer-Fe2O3 before use. There is a difference in the
1048 cm- suggest the adsorption of intermediates produced
during the photocatalytic degradation of Orange II. The amount
of the adsorbed intermediates on the copolymer-Fe2O3 catalyst
is very low; therefore, it is expected that the decrease of the
catalytic efficiency with time will also be very low.
1
-1
spectral region at about 1000 cm in which the band assignment
is somewhat difficult. Assuming that the low-intensity bands
-1
at 1016 and 1045 cm are due to the C-O stretching vibration
of an iron-carboxylate surface complex, the observation of a
The XPS results of the copolymer-Fe2O3 samples before and
after use in the Orange II photodegradation and also of the Fe2O3
-
1
new band at 985 cm and the disappearance of the band at

TiO2 and Fe2O3 Photocatalysts
J. Phys. Chem. B, Vol. 105, No. 48, 2001 12051
TABLE 3. Binding Energy of Ti 2p, O 1s, and C 1s
Elements and Atomic Surface Concentration of Detected
TABLE 4: Binding Energy of Fe 2p, O 1s, and C 1s
Elements and Atomic Surface Concentration of Detected
Elements for Polymer-Fe3 Catalysts
+
Elements for Fe
2
O
3
Catalysts
(
a) Binding Energy (eV)
(a) Binding Energy (eV)
O 1s
O 1s
sample
Fe 2p1/2 Fe 2p3/2
C 1s
sample
Fe 2p
711.2
C 1s
polymer-Fe3
before use
+
+
532.6
533.1
531.3
531.8
529.7
530.2
284.6
Fe
polymer-Fe
before use
polymer-Fe
2
O
3
powder
710.9
711.3
724.6 532.8 531.3 529.6 284.6
725.0 532.6 531.1 529.4 284.6
2
O
3
3
polymer-Fe3
712.0
284.0
2
O
711.1
724.8 531.1 531.6 529.8 284.6
after use in
degradation
of Orange II
after use in
degradation
of Orange II
(b) Atomic Surface Concentration (%)
(b) Atomic Surface Concentration (%)
O
Fe 2p
O
sample
Fe 2p
4.2
H
2
O
ads
OHsurf
O
oxides
C
sample
Fe2
+
Fe3+
H
2
O
ads
OHsurf
O
oxides
C
polymer-Fe3
before use
+
2.9
4.0
17.3
4.4
3.0
71.4
Fe O
2 3
powder
10.7
6% 44%
1.4
58% 42%
7.8
46% 54%
polymer-Fe3
after use in
degradation
of Orange II
+
0.9
6.3
86.0
5
4.4
7.9
9.6
6.8
17.5
0.9
59.8
76.5
70.7
polymer-Fe
before use
polymer-Fe
2
O
3
3
11.6
0
2
O
after use in
degredation
of Orange II
14.8
the symmetric-stretching vibrations of the carboxylate groups.
-
1
Concomitantly, the band at 1793 cm assigned to carboxyl
anhydride almost disappears. These observations indicate that
an iron-carboxylate surface complex is formed. The bands at
reference powder are shown in parts a and b of Table 3. The
most striking difference is the nearly 5 times higher intensity
of the Fe 2p line for the copolymer-Fe2O3 sample after Orange
II degradation compared to the unused copolymer-Fe2O3. This
is seen in Table 3b, where 10.7, 1.4, and 7.8 represent the total
concentration of Fe on the surface for the Fe2O3 powder and
copolymer-Fe2O3 before and after use, respectively. Con-
comitantly, the O 1s line undergoes very strong variations. The
unused copolymer-Fe2O3 shows very strong oxygen compo-
nents at 532.6 and 531.1 eV, characteristic for the adsorbed
water and the surface OH groups, while the oxygen component
at 529.4 eV due to Fe2O3 is seen to be almost negligible. This
indicates that the Fe2O3 surface is strongly hydrated after loading
the iron oxide on the copolymer. After the copolymer-Fe2O3
mediates Orange II photodegradation, the opposite is observed.
The O 1s oxide component at 529.8 eV is the dominating one,
whereas the OH component is much lower with no presence of
adsorbed water (component at ∼533 eV). The photocatalytic
degradation performed in acidic conditions removes the thick
hydroxylate layer, which causes the attenuation of the photo-
electron escaping the Fe2O3 particles.
-
1
1560 and 1523 cm have positions similar to those observed
in the copolymer-Fe2O3 samples, and the peak assignment
follows what has been reported in the previous section. The
-
1
additional band at 1594 cm is due to the formation of a
carboxylate complex with iron bonded to the OH group. The
higher position of asymmetric vibration indicates that a more
covalent bonding is taking place. The formation of mixed-
2
+
3+
valence Fe -Fe complexes could also be considered. The
strongest bands due to the asymmetric and symmetric stretching
of carboxylate groups are observed at 1594 and 1415 cm ,
respectively. The calculated ∆ν parameter is 179 cm according
-
1
-1
to the literature and indicates the formation of a surface bridging
2
5,26
complex.
For the lower intensity asymmetric bands at 1560
-
1
-1
and 1523 cm , the calculated ∆ν values are 145 and 108 cm ,
indicating chelation or bridging. It should be noted that only
one band of the symmetric-stretching vibration bands of the
-COO groups is observed in the reflection spectrum. Other
symmetric-vibration bands overlap with the strong CH2 defor-
-
-
1
mation bands of the copolymer at 1460 cm . This makes the
The Fe 2p1/2 line of the copolymer-Fe2O3 sample is very
broad and can be deconvoluted into two components with line
positions at 710.0 and 711.6 eV. These two components
interpretation of the ∆ν parameter ambiguous. The band position
-1
at about 1460 cm is frequently reported as being characteristic
of the asymmetric-stretching vibrations of the iron and titanium-
correspond to the Fe2 and Fe species, respectively. For the
+
3+
27
25,26
carboxylate complexes produced in solution.
Assuming this
-
1
sample after Orange II degradation, the intensity ratio of the
possibility, the estimated value of ∆ν is 63 cm , indicating
bridging iron-carboxylate complex formation for the copolymer-
Fe catalyst. The copolymer-Fe samples show a band at
1640 cm , indicating a significant water adsorption similar to
2+
3+
Fe /Fe species was found to be 46:54. Powder Fe2O3 used
as a reference showed an intensity ratio of 56:44. Hence, the
catalyst after Orange II degradation is significantly richer in
more positive iron ions than in the reference Fe2O3 powder.
This observation supports the conclusion that Fe2O3 particles
participate in oxidation-reduction reactions during Orange II
photodegradation.
3
+
3+
-
1
that of copolymer-Fe2O3. Compared to the spectrum of a
-
1
copolymer-Fe2O3 sample, a new broad band at about 829 cm
-
1
and strong bands at ∼3473 and 3363 cm are also observed.
The latter bands indicate the formation of iron oxide-hydroxide
IR and XPS Characterization of Copolymer-Fe3 before
and after Degradation of Orange II. IRATR spectra of
copolymer-Fe before and after Orange II degradation and
the respective band assignments are listed in Table 1. A
reflection spectrum of a polymer-Fe3 sample shows the
appearance of new bands at 1594, 1560, and 1523 cm . These
bands are assigned to the asymmetric-stretching vibrations of
+
24
species by the copolymer-loaded sample. The reflection
3+
spectrum of copolymer-Fe after Orange II degradation shows
few differences compared with the spectrum of a freshly
prepared sample. A more defined band is found at about 1594
3
+
+
-1
cm . This corresponds to a covalent surface carboxylate
-
1
complex.
The XPS results of copolymer-Fe3 samples before and after
+
-
1
the carboxylate groups. The new band at 1415 cm is due to
Orange II degradation are presented in parts a and b of Table

12052 J. Phys. Chem. B, Vol. 105, No. 48, 2001
Dhananjeyan et al.
Figure 5. Decrease in TOC of a solution of Orange II (0.2 mM) at a
pH of 3 under Suntest irradiation in the presence of H . [Traces:
1) copolymer-Fe in the dark and H (10 mM), (2) copolymer-
Fe under light and H (10 mM), (3) Fe suspensions of 25
mg/L and H (1 mM), (4) Fe suspensions of 75 mg/L and H
1 mM), (5) Fe suspensions of 25 mg/L and H (10 mM), (6)
Fe suspensions of 75 mg/L and H (10 mM).]
2 2
O
(
O
2 3
2 2
O
O
2 3
O
2 2
2 3
O
O
2 2
2
O
3
2 2
O
(
2
O
3
2 2
O
O
2 3
2 2
O
Figure 7. (a) Degradation of 2,4-dichlorophenol (0.5 mM) on
copolymer-TiO
irradiation at a pH of 6. Solutions were continuously purged with O
during irradiation. The TiO content of the suspension was 1 g/L. (b)
Decrease in TOC of a solution of 4-chlorophenol (0.65 mM) at a pH
of 3 under Suntest light irradiation. [Traces: (1) copolymer-Fe in
the dark and H (10 mM), (2) copolymer-Fe under light and
(10 mM), (3) Fe suspensions of 25 mg/L and H (1 mM),
(4) Fe suspensions of 75 mg/L and H (1 mM), (5) Fe
suspensions of 25 mg/L and H (10 mM), (6) Fe suspensions of
2
membranes under medium-pressure mercury-light
2
2
Figure 6. Decrease in TOC for a solution of Orange II (0.2 mM) in
the presence of H under Suntest irradiation. [Traces: (1) copolymer-
2 3
O
2
O
2
2
O
2
2 3
O
3
+
3+
Fe and H
2
O
2
(10 mM) in the dark, (2) copolymer-Fe and H
2
O
2
H
2
O
2
O
2 3
2 2
O
3
+
(
(
10 mM) under light, (3) Fe (0.2 mM) and H
O
2 2
(1 mM) under light,
(1 mM) under light, (5) Fe (0.2 mM)
O
2 3
O
2 2
2 3
O
3
+
3+
4) Fe (0.5 mM) and H
(10 mM) under light, (6) Fe (0.5 mM) and H
under light.]
2
O
2
O
2 2
2 3
O
3
+
and H
2
O
2
2
O
2
(10 mM)
2 2
75 mg/L and H O (10 mM).]
2
a 150 m /g surface area, the photodegradation results follow
what is expected from the concentrations of the catalyst in
4
. The most striking difference found is in the intensity of the
3+
3+
2+
Fe 2p line observed for copolymer-Fe after use. The intensity
of the latter band is about 5 times lower compared to the unused
solution. For the suspensions in Figure 5, the Fe to Fe redox
cycle on copolymer-Fe O is suggested as
2
3
3
+
copolymer-Fe . The low intensity of the Fe 2p line observed
in copolymer-Fe3 after Orange II degradation makes it
difficult to describe the changes in the Fe oxidation state after
use. However, there is a significant difference in the O 1s lines
between these two samples. The unused sample shows a much
higher surface density of two oxygen components. Particularly,
the oxygen-containing peaks related to OHsurf (at 533 eV) and
to the oxide (at 530 eV) are observed to be weak after Orange
II degradation.
Degradation of Orange II on Polymer-Fe2O3 and Fe -
Polymer Films under Suntest Light Irradiation. Figure 5
shows the degradation of Orange II on copolymer-Fe2O3 in
the dark and under light irradiation. No degradation was
observed in the dark, but under light irradiation, ∼90%
degradation was completed in ∼7 h. The abatement rate for
the copolymer-Fe2O3-mediated process was seen to be com-
parable to a solution containing Fe2O3 (75 mg/L) in the presence
of H2O2 (10 mM). For the other concentrations of hematite with
+
Orange II + copolymer-Fe O + hν f
2
3
[
Orange II*...Fe O ]copolymer f
2 3
Orange II•+ + copolymer-Fe O + e
-
cb
(3)
2
3
Similar processes for Orange II degradation on Fe2O3 powders
2
4
have been recently reported. No absorption changes were
observed on copolymer-Fe2O3 after six repetitive degradation
cycles. This confirms the stable anchoring between iron oxide
3
+
3
+
2+
and the copolymer film. No Fe or Fe was detected in
solution by thiocyanate or phenanthroline after the runs when
using copolymer-Fe2O3. The BET surface area of the copoly-
2
mer-Fe2O3-loaded copolymer was 2.77 m /g compared to that
2
of the copolymer alone (1.34 m /g).
3+
Figure 6 presents the results for copolymer-Fe -mediated
degradation of Orange II under different experimental condi-
tions. The experimental results follow the trend shown previ-

TiO2 and Fe2O3 Photocatalysts
J. Phys. Chem. B, Vol. 105, No. 48, 2001 12053
2
,4-DCP under mercury-light irradiation at an initial pH of 6.
Under light irradiation, the mineralizations in the titania
suspensions and the copolymer-TiO2 films are seen to proceed
at comparable rates. This shows the advantage of using
copolymer-TiO2 instead of TiO2 suspensions because the fixed
titania avoids the separation of the catalyst in suspension at the
end of the treatment. In the dark, no mineralization was
observed, as indicated by the upper traces of Figure 7a. No H2O2
was used as an electron acceptor because the continuous purging
with O2 was sufficient to enhance the charge separation at the
TiO2 surface
O2)ads + e f (O2•-)
-
(5)
(
cb
ads
The pH during 2-chlorophenol mineralization decreased from
.0 to 4.0 resulting from the production of HCl as shown below
6
HO-C H -Cl + 6O f 6CO + H O + 2HCl (6)
6
3
2
2
2
2
H2O2 generated in the solution was observed to remain at a
level of ∼0.5 mg/L. The formation of peroxide in TiO2
suspensions under light has been widely reported in the
Figure 8. Repetitive mineralization of Orange II (0.1 mM) during the
decoloration at a pH of 6 on copolymer-TiO
2
under medium-pressure
light irradiation in the presence of H (0.10 mM).
2
O
2
7
literature. Figure 7a shows that the mineralization kinetics of
2
,4-DCP on titania suspensions or with copolymer-TiO2 are
ously in Figure 5 for copolymer-Fe2O3 catalysts. The photo-
sensitization of Orange II on copolymer-Fe3+ would involve
a charge-transfer interaction with the quenching of Orange II
about the same. The copolymer had a loading of 1.5 mg of TiO2
on a polymer surface area of 48 cm . The amount of titania in
2
the suspension was 40 mg/40 mL, which makes it ∼27 higher
than the titania bound on the copolymer surface. The poor
performance observed in suspensions could be ascribed to the
screening effect of the incoming light by the titania suspension.
This is not the case for the transparent copolymer-TiO2. The
photo-oxidation of halocarbons and other organic pollutants on
TiO2 thin films supported on glass or silica paper has been
recently reported and is an area of current interest. Recent work
3
+
Orange II + Fe -copolymer + hν f
3+
[
Orange II*...Fe ]copolymer f
Orange II•+ + Fe -copolymer (4)
2+
The lack of mineralization in the dark compared to light-induced
processes in Figure 7bis due to the slow conversion kinetics of
3
+
2+ 2,14
30
31
Fe to Fe .
by Heller and Brock, Fujishima et al., Pichat and
3
2
33
Degradation of Chlorophenols by Copolymer-TiO2 and
Copolymer-Fe2O3. Figure 7a presents the mineralization of
Ceulemans, and Zahraa et al. have addressed the problem
of the deposition of TiO2 on suitable inorganic supports.
Figure 9. Atomic force microscopy showing the copolymer-Fe3+ sample before and after catalyzing the degradation of Orange II.

12054 J. Phys. Chem. B, Vol. 105, No. 48, 2001
Dhananjeyan et al.
To check the stability of copolymer-TiO2 during the reaction,
ICPS was carried out to determine the amount of TiO2 released
as a fine particulate into the solution (which is equivalent to
the copolymer-TiO2 samples was observed to be consider-
able: 658 Å after use and 568 Å before use. This was the only
change observed in the catalysts after repetitive recycling.
the amount of Ti4 ions). The concentration of Ti ions in
solution after eight recyclings of the copolymer-TiO2 catalyst
varied between 0.14 and 0.50 ppm of Ti4 ion. This is far below
the upper limit set by the European Directive (for details see
+
4+
Conclusions
+
This study indicates that the azo-dye Orange II and chloro-
carbons can be decolored and degraded via photocatalysis on
TiO and Fe O immobilized on modified copolymer films. The
2 2 3
4+
ref 34) allowing up to 5 ppm of Ti in aqueous solutions.
Furthermore, Ti has been reported to be nontoxic up to very
4+
high levels.1
,3,5
degradation of Orange II is shown to depend on the type of the
model organic compound used, the physical characteristics of
the bound semiconductor, the type of scavenger chosen for the
separation of the charges on the semiconductor surface, and the
irradiation source. No catalyst deposited on the polymer surface
leaches out during degradation. IR spectroscopy data also
provides evidence that the conjugated carboxylic groups of the
The fact that we have practically not found any
titania in solution after the catalytic run confirms that, for
copolymer-TiO2, TiO2 is chemically bonded to the surface of
the polymer via carboxylic groups. IR data provided evidence
that the conjugated carboxylic groups of the maleic anhydride
4+
strongly interact with Ti . The structure of this bonding would
4
+
involve two Ti ions binding to one carboxylic group. Such a
4
+
maleic anhydride strongly interact with Ti . The bands
observed by IR also show the high stability of the polymer
surface. XPS analysis confirms the intervention of supported
Fe2O3 during redox processes on the copolymer surface during
Orange II photodegradation. By XPS, it was not possible to
detect the absorption of intermediates on the catalyst surface
after photocatalysis, indicating an adequate catalytic activity of
the new materials. The performance of the immobilized photo-
catalysts was shown to be comparable to semiconductor
suspensions with a higher semiconductor loading per unit
volume of solution. This effect is due to the absence of the
screening effect between the catalyst particles when they are
dispersed on thin films.
structure has been worked out for 2,2′-bipyridylruthenium(II)
complexes chelating the TiO2 surface.29 Copolymer-TiO2 was
observed to be stable in acidic as well as basic media up to a
pH of 9.
Figure 7b shows the photodegradation of 4-chlorophenol on
copolymer-Fe2O3. Experiments with copolymer-Fe2O3 films
in the dark show a reduction in TOC of ∼10%. However, under
light irradiation, degradation of 4-chlorophenol reached ∼90%
within 7 h. Experiments varying the amount of Fe2O3 in the
suspensions and the concentration of H2O2 in solution are shown
in traces 3-6. The trend in Figure 7b followed the results in
Figure 7a. Very few studies are available on the iron oxide-
1
,2,7,14,18,35
mediated degradation of halocarbons.
Cunningham and
Sedlak36 have investigated the degradation of oxalic acid on
hematite and found that the process was rather inefficient.
Acknowledgment. We thank KTI/CTI TOP NANO 21 for
financial support under Grant No. 4823. The authors also thank
J. Lambert for his technical assistance with the XPS part of
this study.
Catalytic Nature of the Photodegradation of Orange II
on Copolymer-TiO2. Figure 8 presents the cyclic repetitive
decoloration of Orange II on copolymer-TiO2. At the end of
each cycle, the polymer was washed and Orange II (0.2 mM)
and H2O2 were added into the solution. The results in Figure 8
confirm the photocatalytic nature of Orange II degradation on
References and Notes
4+
(1) Halmann, M. Photodegradation of Water Pollutants; CRC Press:
Boca Raton, FL, 1996.
supported catalyst. No measurable amounts of Ti were found
after each cycle in the solution.
(
2) Bauer, R.; Waldner, G.; Fallmann, H.; Hager, S.; Klare, M.; Malato,
S.; Maletzky, P. Catal. Today 1999, 53, 131.
3) Pitter, P.; Chudoba, J. Biodegradability of Organic Substances in
the Aquatic EnVironment; CRC Press: Boca Raton, FL, 1990.
4) Lucarelli, L.; Nadtochenko, V.; Kiwi, J. Langmuir 1999, 16, 1102.
(5) Acher, J.; Rosenthal, I. Water Res. 1997, 1111, 557.
6) Zhang, F.; Zhao, J.; Hidaka, H.; Pelizzetti, E.; Serpone, N. Appl.
Catal., B 1998, 15, 147.
7) Ollis, F. D.; Al-Ekabi, H. Photocatalytic Purification of Water and
TEM and BET Areas. Electron microscopy of copolymer-
TiO2 revealed that the TiO2 particle size more frequently
observed was in the range of 20-30 nm. The diffraction pattern
of TiO2 showed anatase as the main phase and rutile as the
secondary phase. The ratio between the two phases corresponds
to the known ratio in Degussa P-25 of 80:20. The electron
diffraction was carried out to confirm the Degussa P-25
crystallographic structure of the titania bound on the copolymer.
The particles of TiO2 on the copolymer were seen to form a
protecting compact layer. No significant change in the BET
(
(
(
(
Air; Elsevier: Amsterdam, The Netherlands, 1993.
(8) Mills, A.; Morris, S. J. Photochem. Photobiol., A 1993, 71, 285.
(
9) Hofstadler, K.; Bauer, R.; Novalic, S.; Heisler, G. EnViron. Sci.
Technol. 1994, 28, 670.
(
10) Minero, C.; Pelizzetti, E.; Sega, M.; Vincenti, M. EnViron. Sci.
2
surface area of the loaded (1.45 m /g) compared to the unloaded
Technol. 1996, 29, 2226.
2
(11) Korman, C.; Bahneman, D.; Hoffmann, M. J. Photochem. Photo-
biol., A 1989, 48, 161.
(
1.34 m /g) copolymer film was observed.
Atomic Force Microscopy of Copolymer Samples. The
(12) Stafford, U.; Gray, K. A.; Kamat, P. V. J. Phys. Chem. 1994, 98,
histograms of fresh copolymer-Fe3 samples are shown in the
upper part of Figure 9, and the samples after mediating the
photodegradation of Orange II are in the lower part of this figure.
These histograms refer to the selected areas shown to the left
in Figure 9. The average roughness for the unused sample was
+
6343.
(13) Al-Ekabi, H.; Serpone, N. J. Phys. Chem. 1988, 92, 5276.
14) Fernandez, J.; Bandara, J.; Lopez, A.; Buffat, Ph.; Kiwi, J. Langmuir
(
1
999, 15, 185.
15) Thampi, K. R. EPFL Patent Application, 2000.
(16) Dhananjeyan, M. R.; Kiwi, J.; Thampi, K. R. Chem. Commun. 2000,
443.
17) Shirley, A. Phys. ReV. 1979, B5, 4709.
18) Harrick, N. J. Internal Reflection Spectroscopy; Interscience
(
1
3
85 Å, and for the sample after use, the value found was 881
(
(
Å. The increase in roughness (root-mean-square value) of the
used samples indicates that the catalyst has become more porous.
For the copolymer-Fe2O3 catalyst, no meaningful difference
was found for the roughness because a value of 551 Å after
photodegradation was observed versus a roughness factor of
Publishers: New York, 1987.
(19) Vinogpodal, K.; Kamat, P. J. Photochem. Photobiol., A 1994, 10,
767.
1
(
(
20) Bandara, J.; Kiwi, J. New J. Chem. 1999, 23, 717.
21) Turro, N. J. Modern Molecular Photochemistry; University Science
5
07 Å for samples before use. The change in the roughness for
Books: Sausalito, CA, 1991.

TiO2 and Fe2O3 Photocatalysts
J. Phys. Chem. B, Vol. 105, No. 48, 2001 12055
(
22) Bellamy, L. J. Infrared Spectra of Complex Molecules; Chapman
(31) Kikuchi, Y.; Sunada, K.; Hashimoto, K.; Fujishima, A. J. Photo-
chem. Photobiol., A 1997, 106, 51 and references therein.
(32) Ceulemans, E.; Pichat, P. Abstract, Chimie, Soleil, Energie, et
Environnement, Saint-Avold, France, Feb 2000; p 22.
(33) Zahraa, O.; Chen, H.; Dorion, C.; Bouchy, M. Abstract, Chimie,
Soleil, Energie, et Environnement, Saint-Avold, France, Feb 2000;
p 42.
&
Hall: New York, 1975.
(
(
(
(
23) Colthup, J. J. Opt. Soc. Am. 1950, 40, 397.
24) Bandara, J.; Mielzcarski, J.; Kiwi, J. Langmuir 1999, 15, 7680.
25) Deacon, G. B.; Philips, R. J. Coord. Chem. ReV. 1980, 33, 227.
26) Mehrotra, R. C.; Bohra, R. Metal Carboxylates; Academic Press:
New York, 1983.
27) Handbook of X-ray Photoelectron Spectroscopy; Mullenberg, G.
E., Ed.; Perkin-Elmer Corp.: Eden Prairie, MN, 1985.
(
(34) Valandro, V.; Betti, R. J. EnViron. Pathol. Toxicol. Oncol. 1997,
16, 163.
(
28) Mielczarski, J. A.; Mielczarski, E. J. Phys. Chem. 1995, 99, 3206.
(35) Lausanne Workshop on Advanced Oxidation Technologies, Lau-
sanne, Switzerland, Oct 1998.
(29) Nazeeruddin, Md. K.; Zakeeruddin, S. M.; Humphry-Baker, R.;
Shklover, V.; Fisher, C.-H.; Gr a¨ tzel, M. Inorg. Chem. 1999, 38, 6298.
30) Heller, A.; Brock, J. U.S. Patent 4,997,576, March 5, 1991.
(36) Cunningham, J.; Sedlak, P. J. Photochem. Photobiol., A 1994, 77,
255.
(