Photosensitization of crystalline and amorphous titanium dioxide by platinum(IV) chloride surface complexes

Article abstract of DOI:10.1002/1521-3765(20010504)7:9<1862::AID-CHEM1862>3.0.CO;2-G

Anatase, rutile, and amorphous titania powders were surface-modified by grinding with PtCl₄ and H₂[PtCl₆]. Only the anatase modification afforded hybrid photocatalysts capable of degradation of 4-chlorophenol (4-CP) with visible light, with sufficient stability towards decomplexation. Grinding with K₂[PtCl₄] produced materials of only low photocatalytic activity. Most efficient photocatalysts contained up to 2 wt % of Pt^IV. At higher surface loading the excess fraction of the complex is desorbed into the aqueous solution. Scavenging experiments with benzoic acid and tetranitromethane revealed that hydroxyl radicals are produced by the primary reduction of oxygen by conduction band electrons generated through electron injection from a postulated surface platinum(III) complex. It is proposed that the latter is formed from a charge-transfer ligand-to-metal (CTLM) excited state through homolysis of the Pt-Cl bond. Accordingly, the primary oxidation of 4-CP may occur by adsorbed chlorine atoms, the intermediary existence of which was demonstrated by scavenging experiments with phenol.

Full text of DOI:10.1002/1521-3765(20010504)7:9<1862::AID-CHEM1862>3.0.CO;2-G

FULL PAPER
Photosensitization of Crystalline and Amorphous Titanium Dioxide by
Platinum(iv) Chloride Surface Complexes
Wojciech Macyk and Horst Kisch*[a]
Abstract: Anatase, rutile, and amor-
phous titania powders were surface-
modified by grinding with PtCl4 and
H [PtCl ]. Only the anatase modifica-
tion afforded hybrid photocatalysts ca-
pable of degradation of 4-chlorophenol
surface loading the excess fraction of the
complex is desorbed into the aqueous
solution. Scavenging experiments with
benzoic acid and tetranitromethane re-
vealed that hydroxyl radicals are pro-
duced by the primary reduction of oxy-
gen by conduction band electrons gen-
erated through electron injection from a
postulated surface platinum(iii) com-
plex. It is proposed that the latter is
formed from a charge-transfer ligand-to-
metal (CTLM) excited state through
homolysis of the Pt�Cl bond. Accord-
2
6
(
4-CP) with visible light, with sufficient
stability towards decomplexation.
Grinding with K [PtCl ] produced ma-
ingly, the primary oxidation of 4-CP may
occur by adsorbed chlorine atoms, the
intermediary existence of which was
demonstrated by scavenging experi-
ments with phenol.
2
4
Keywords: amorphous materials ´
photochemistry ´ platinum ´ sol ±
gel processes ´ titanium
terials of only low photocatalytic activ-
ity. Most efficient photocatalysts con-
IV
tained up to 2 wt% of Pt . At higher
Introduction
procedure affording high specific surface areas of 160 ±
2
� 1
200 m g . According to extended X-ray absorption fine
Semiconductor photocatalysis of chemical reactions by titania
has progressed from basic research and reached technical
applications like solar purification of air and water from
organic and inorganic pollutants.[1] Although titania absorbs
only about 3% of solar light, this is high enough to induce
reasonable photodegradation rates of these compounds,
present only in trace amounts. However, for this and other
applications a more efficient utilization of solar light is
desirable. Accordingly, many efforts have been made to
sensitize titania for visible light induced photocatalytic
reactions. Recently we have used Na [PtCl ] and other metal
structure (EXAFS) results, isolated PtCl4 molecules are
homogeneously distributed in an almost exclusively amor-
phous matrix of titania.
It was postulated that the photocatalytic activity of PtIV/
AMM-Ti in the visible originates from local excitation of the
platinum(iv) chloride chromophore affording an adsorbed
III
chlorine atom and Pt as an intermediate species. Charge
separation was assumed to proceed through electron injection
III
from Pt into the conduction band of titania, followed by
reduction of oxygen, while the surface-trapped chlorine atom
may oxidize 4-CP to the radical cation. Degradation of the
2
6
chlorides like AuCl and RhCl to prepare amorphous micro-
latter should follow the mechanism as proposed for unmodi-
3
3
[
2±4]
[5±7]
porous titania (AMM-Ti) modified in the bulk.
These
fied titania.
This mechanistic proposal was based, inter
[8]
hybrid photocatalysts catalyzed the photomineralization of
alia, on analogy with the photochemistry of chlorocopper(ii)
4
-chlorophenol (4-CP) with visible light. Apparent disappear-
and chloroplatinum(iv) complexes in aqueous solution. The
broad CTLM bands extending from the UV to the visible
�
3
ance quantum yields reached 1.3 Â 10
at 546 nm and
�
3
increased to 4.5 Â 10 at 366 nm. The most active photo-
region overlap with ligand field bands at approximately
IV
4� x
catalysts contained 3% Pt . Photoelectrochemical experi-
350 nm and are characteristic for [PtCl (H O) ]
(x ˆ 4 ± 6)
x
2
6� x
species.[
9, 10]
[PtCl6] in aqueous solutions exposed both
2�
ments proved the semiconductor nature of the powders; the
action spectrum of the photocurrent extended down to about
6
to UV or visible light irradiation undergoes photoaqua-
tion. The reaction is initiated by photodissociation of a
chlorine atom [Eq. (1)], followed by chain reactions.
[
4]
00 nm. These novel materials were prepared by a sol ± gel
[
a] Prof. Dr. H. Kisch, Dipl.-Chem. W. Macyk
Institut für Anorganische Chemie
Universität Erlangen-Nürnberg
hn
PtIVCl
2�
![PtIIICl
2�
5
[
6
]
]
‡ Cl
(1)
Egerlandstrasse 1, 91058 Erlangen (Germany)
Fax : (‡49)9131-8527363
Chloride substitution takes place with quantum yields of 13.4
to 0.87 in the wide range of irradiation wavelength from 270 to
E-mail: kisch@anorganik.uni-erlangen.de
1862
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-1862 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

1862±1867
4
50 nm.[10] The corresponding thermal reaction requires
[
11]
several days for completion.
Although the proposed mechanism for these amorphous
titania photocatalysts, doped in the bulk, seemed reasonable,
it raised some important questions. For instance, unexpectedly
no reduction of PtCl to elemental platinum occurred during
4
the photodegradation, a reaction well known for unmodified
titania irradiated in the presence of reducing agents like
[
12, 13]
citrate, acetate, or methanol.
This process finds applica-
tion in metallic platinum deposition on TiO surfaces, but also
2
can be useful for platinum recovery from diluted solutions of
IV
Pt . Furthermore, it was unknown whether an amorphous
4
Figure 1. Diffuse reflectance spectra of 3% PtCl /P25 (dashed line), P25
titania phase and volume doping are necessary to obtain the
visible light activity or whether crystalline phases and surface
doping would accomplish the same. In the following we
therefore report on the photocatalytic activity of the two
(solid line), and PtCl (dotted line).
4
region, as also observable in the spectrum of solid PtCl . Since
4
crystalline modifications, anatase (P25) and rutile (TiO -R),
surface modification of P25 with K [PtCl ] resulted in a
2
2
4
and one amorphous phase, (TiO -a), all surface-doped with
material at least 50% less active in 4-CP degradation (l >
2
platinum(iv) chloride complexes. Mechanistic experiments in
the presence of scavengers for electrons and OH radicals
provide basic information on the initial stages of the photo-
degradation.
455 nm) than the analogous PtCl /P25 catalyst, we conclude
that the oxidation state of the photoactive platinum species is
4
‡iv, in accordance with the previously reported bulk-doped
IV
Pt /AMM-Ti photocatalyst.
To investigate the stability of the modified samples with
respect to desorption of the platinum component, aqueous
suspensions were stirred in the dark for five days. Formation
Results and Discussion
4� x
of [PtCl (H O) ]
(x ˆ 4 ± 6) complexes was monitored by
measuring the strong absorbance at 259 nm. In order to have
x
2
6� x
Preparation and stability of surface-modified titania: Surface-
modified photocatalysts were prepared by grinding the two
2�
predominantly the [PtCl
]
6
complex present, as indicated by
crystalline titania powders P25 and TiO -R and the amor-
the absorption spectrum, the filtrates were acidified with HCl
(2m) before measurement. For P25 materials with Pt
contents below or equal to 2% only less than 1% of the total
platinum amount appeared in solution, while 11 and 50%
2
IV
phous TiO -a with various amounts of PtCl and H [PtCl ].
2
4
2
6
2
� 1
The specific surface area of 50 m g of unmodified P25
2
� 1
increases to 65 m g for 1%PtCl /P25. According to trans-
4
mission electron micrography, the morphology closely resem-
bles that of P25 as indicated by the presence of 0.2 to
were found for 1% PtCl /TiO -R and 1% PtCl /TiO -a,
4 2 4 2
respectively. The good stability in the former case suggests
that chemisorption has taken place during or after grinding,
most likely by formation of Ti O Pt bonds. This type of
1
micrometer aggregates, composed of small particles with an
[
14]
� �
average diameter of 20 nm. The diffuse reflectance spec-
I
trum of 3% PtCl /P25 is shown in Figure 1. Compared to the
bonding was also postulated for [Pt (bpy)] complexes (bpy ˆ
4
2,2'-bipyridine).[ When H
15]
[PtCl ] was employed, the result-
2 6
spectrum of P25 a long tail-absorption extends into the visible
ing materials suffered stronger desorption than PtCl -modi-
4
fied P25, suggesting a smaller fraction of chemisorbed
complexes. It is expected, that chloride substitution by surface
Abstract in German: Verreiben von Anatas, Rutil und
amorphem Titandioxid mit PtCl und H [PtCl ] ergibt ober-
4
2
6
�
[16]
2�
[
Ti] � O
substitution in [PtCl (H O) ]
6� x
in [PtCl6]
should be slower than water
flächenmodifizierte Hybridphotokatalysatoren. Diese sind nur
im Falle der Anatasmodifikationen stabil genug, um einen
effizienten Abbau von 4-Chlorphenol mit sichtbarem Licht zu
ermöglichen. Die aktivsten Katalysatoren besitzen bis zu 2
4
� x
(x ˆ 4, 5), the species
x
2
[17]
generated from PtCl4.
The results are also consistent with
[11]
the reactivity of platinum(iv) chlorides in thermal
and
photoaquation[
10]
reactions. The amount of chemisorbed
IV
wt% Pt , bei höherer Beladung wird der überschüssige
complexes can be increased upon irradiation with visible
light as demonstrated for 1% H [PtCl ]/P25. The initial
Komplex in die wäûrige Lösung desorbiert. Als Mechanismus
der photoinduzierten Ladungstrennung wird vorgeschlagen,
daû Lichtabsorption des Oberflächenplatin(iv)komplexes über
einen Charge-Transfer-Ligand-to-Metal (CTLM)-Zustand zur
Homolyse der Pt�Cl-Bindung führt. Das dabei gebildete
2
6
amount of dissolved platinum complex present in an aqueous
suspension decreased from about 4.5% to below 3% after
irradiation with l > 400 nm for 30 min.
Chloratom oxidiert 4-Chlorphenol, während die gleichzeitig
entstandene Platin(iii)zwischenstufe ein Elektron in das Lei-
tungsband von Titandioxid injiziert. Letzteres reduziert an-
schlieûend Sauerstoff zu Superoxid, welches schlieûlich zur
Bildung von OH-Radikalen führt. Abfangversuche mit Ben-
zoesäure, Tetranitromethan bzw. Phenol bestätigen das Auf-
treten dieser intermediären Radikale bzw. Chloratome.
Photodegradation of 4-chlorophenol: Figure 2 displays the
photodegradation curves of 4-CP for four typical catalysts
upon irradiation with visible light (l > 455 nm) in the
irr
presence of air. Whereas the rutile and amorphous titania
based samples, 1% PtCl /TiO -R and 1% PtCl /TiO -a, were
4
2
4
2
almost inactive, the anatase-containing powders PtCl /P25
4
and H [PtCl ]/P25 induced initial decomposition rates even
2
6
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-1863 $ 17.50+.50/0
1863

H. Kisch and W. Macyk
FULL PAPER
ductor-catalyzed reaction exhibits the maximum rate when
the platinum(iv) content is about 2 wt%.
Surprisingly, no photoreduction of Pt to Pt was observ-
able under the experimental conditions mentioned above,
IV
0
irrespective of whether visible or UV light (l > 335 nm) was
irr
employed. This differs significantly from UV irradiation of
2
�
unmodified TiO in the presence of aqueous [PtCl ] which
2
6
0
[18±20]
produces Pt even in the absence of any reducing agent.
Reduction to Pt seems also unlikely, since in the presence of
air, adsorbed [PtCl4]2 is easily oxidized to the Pt complex.
II
�
IV
[21]
Only when 4-CP was replaced by methanol in high concen-
tration (50 vol%) was elemental platinum formed upon UV
or long-term visible irradiation, according to a well-known
reaction.[ The resulting gray powder induced an initial rate
of degradation of 4-CP about 70% lower than that of
]4 /TiO
/P25, * 1%
4
� x
Figure 2. Photodegradation of 4-CP catalyzed by various [PtCl
x
2
-
modified materials upon irradiation at lirr > 455 nm. 1% PtCl
]/P25, ^ 3% PtCl
-a, ~ 1% PtCl
[PtCl /TiO /TiO -R.
2
&
H
2
6
4
2
4
13]
untreated 1% PtCl /P25.
Figure 4 shows results of photooxidation of 4-CP in sunlight
in various systems based on PtCl /TiO materials. The highest
rate of 4-CP degradation was observed in the case of undoped
P25 even though it is active only in UV light. This
phenomenon will be discussed later.
4
5
0 ± 60% higher than those observed for amorphous bulk-
IV
doped Pt /AMM-Ti under the same experimental condi-
tions.
[
2, 3]
4
2
Similar to the latter, mineralization occurred as
indicated by formation of carbon dioxide and hydrochloric
acid; accordingly the pH value changed from approximately 7
to 2.7. The catalytic nature with respect to platinum chloride
follows from the apparent turnover number of 10 as calcu-
lated for a single run from the amount of 4-CP degraded,
divided by the amount of platinum present in 1% PtCl /P25.
4
Making the reasonable assumption that for mineralization of
one 4-CP molecule more than one photocatalytic cycle is
required, the true turnover number should be much higher.
When after complete disappearance of 4-CP a new 4-CP
portion was added to the suspension up to the same initial
concentration, the rate decreased by about 60%. The lower
rate may be due to a loss of catalyst through multiple sample
withdrawing and catalyst deactivation.
The influence of the platinum content on the reaction rate
of surface-doped P25 was investigated with a series of n%
Figure 4. 4-CP degradation upon solar irradiation of * 1% PtCl
/P25,
4
~
3
% PtCl
4
/P25, & P25 suspensions.
PtCl /P25 catalysts (n ˆ 0.5, 1.0, 1.5, 2.0, 2.5, 3.0). The rate of
4
degradation of 4-CP increased with increasing platinum
content up to about 3 wt% (Figure 3). However, this is an
Mechanism: Since the photodegradation of 4-CP catalyzed by
crystalline surface-doped P25 completely corresponds to the
apparent value, since a considerable part is due to the known
IV
homogeneous reaction of the dissolved platinum complex[4]
results observed for the bulk-doped amorphous titania Pt /
AMM-Ti, the same mechanism may operate in both reactions.
It was proposed that visible light CTLM excitation of the
formed by desorption (Figure 3). Accordingly the semicon-
III
platinum(iv) chloride surface complex affords a Pt species
and a weakly bound chlorine atom.[
2±4]
Chloroplatinum(iii)
complexes have been generated in homogeneous solution
through photolysis of [PtCl6]2
� [22, 23]
and at the surface of P25
powders through oxidation of K [PtCl ] by valence band holes
2
4
or OH radicals.[ For the latter case an average lifetime of a
few microseconds was evaluated from diffuse reflectance flash
photolysis. For the present system it is assumed that subse-
24]
III
quent electron injection from Pt into the conduction band of
IV
titania regenerates Pt . This mechanistic pathway is termed
indirect sensitization. It is supported by the fact that mod-
ification of the bulk-doped amorphous titania by PtBr
4
afforded less active photocatalysts.[ However, the alterna-
tive, direct electron transfer sensitization by electron injection
from the excited surface platinum complex without preceding
Pt�Cl cleavage cannot be excluded at the present. It would
3]
Figure 3. Amount of 4-CP decomposed after irradiation with visible light
IV
correspond to the well-established sensitization by bipyridyl
for 6.5 h (
&
) and desorption of Pt complexes after five days in the dark (*)
/P25.
II
II
as function of increasing surface loading; 0 ± 3% PtCl
4
complexes of Ru and Fe . In this case electron injection
1864
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-1864 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

Photosensitization of TiO2
1862±1867
occurs from a CTML excited state.[25±28] Although an analo-
gous mechanistic pathway cannot be ruled out, it seems rather
unlikely since the CTLM states of chloroplatinum(iv) com-
plexes in solution are known to undergo bond cleavage rather
than electron transfer. Furthermore, unlike ruthenium and
relatively stable yellow C(NO2)3 anion was identified
�
through its typical absorbance at 350 nm.[
35±37]
In contrast, a
fast formation of salicylic acid occurred when the latter
experiment was conducted with UV light (l > 335 nm). In this
case light absorption by the titania matrix generates valence
band holes which oxidize water or surface hydroxyl groups to
the OH radical, whereas the conduction band electrons still
reduce tetranitromethane. These experiments clearly demon-
strate that upon excitation with visible light the OH radical is
formed by the reductive pathway. Furthermore, the unexpect-
ed observation that in the solar degradation experiments
undoped P25 induced a slightly faster reaction than the two
modified samples (Figure 4), may be now rationalized. Upon
UV excitation additional OH radicals are formed by water
oxidation through light absorption of the titania matrix. This
effect should be more efficient in unmodified P25, since no
platinum surface complex can act as a light filter as in
modified P25. Thus, the beneficial role of the modified surface
in the visible spectral region apparently does not counter-
balance its negative influence in the UV.
V
iron, the resulting higher oxidation state, in this case Pt , is
known only from complexes with more electronegative
�
[29]
ligands, like in [PtF ] . Irrespective of the detailed mech-
6
anism, the electron injected into the titania conduction band is
expected to reduce oxygen to superoxide, which in turn is
transformed into hydrogen peroxide and OH radicals
[
30±34]
(Scheme 1).
This mechanistic proposal is corroborated
When sensitization of titania proceeds through the indirect
electron transfer mechanism, some of the chlorine atoms are
expected to undergo addition to the substrate. To test for this
possibility, phenol was used instead of 4-CP in the standard
irradiation experiment employing visible light (l > 455 nm).
In this reaction at least 3% of the total Cl content was
converted to CP, proving the existence of intermediary
chlorine atoms (Figure 6). This side reaction may explain
Scheme 1. Proposed mechanism of TiO
complexes. CB and VB denote conduction and valence band edge,
respectively.
sensitization by [PtIVCl
]
4� x
2 x
by experiments in the presence of scavengers for OH radicals
and electrons. Thus, when the standard photodegradation
(
l > 400 nm, 1% PtCl /P25) was performed in the presence of
4
�
2
benzoic acid (10 m), substantial amounts of salicylic acid
were detected. Although this suggests that the OH radicals
are formed through the proposed O reduction pathway, it
2
does not exclude an oxidative pathway via an hitherto
unknown strongly oxidizing surface platinum complex. To
answer this question and to have a more simple system, 4-CP
was omitted in the further experiments. As depicted in
Figure 5, upon irradiation with visible light salicylic acid
formation was observed only when oxygen was the electron
�
2
scavenger. When the latter was tetranitromethane (10 m)
only trace amounts of salicylic acid were detectable, and the
Figure 6. Phenol degradation (*) and 4-CP formation (
irr > 455 nm). [Pt]total ˆ 25mm, [Cl]total ˆ 100mm.
&
4
) at 1% PtCl /P25
(
l
the partial deactivation of the photocatalyst observed upon
multiple use. Free chlorine atoms are slightly weaker oxidants
[
38a]
[38b]
than OH radicals (E8
ˆ 2.6 V,
E8OH/OH
�
ˆ 2.8 V
)
Cl/Cl�
[9, 39]
and therefore should oxidize phenol
and 4-CP but not
water or surface hydroxyl groups. Thus, in agreement with the
scavenging experiments we propose as a working hypothesis
that upon excitation with visible light with the hybrid photo-
catalyst, 4-CP is oxidized to the corresponding phenoxyl
radical and a proton by intermediate chlorine atoms, gener-
ated by homolytic Pt�Cl bond cleavage in the CTLM state of
Figure 5. Formation of salicylic acid upon irradiation of 1% PtCl
aerated suspension in the presence of benzoic acid (lirr > 400 nm, ); in Ar-
saturated suspension in the presence of C(NO >
(lirr > 400 nm, ~ and lirr
4
/P25 in
&
a chloroplatinum(iv) surface complex and by OH radicals
2
)
4
3
35 nm, *). For details see text.
originating from the reduction of oxygen (Scheme 1). Further
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-1865 $ 17.50+.50/0
1865

H. Kisch and W. Macyk
FULL PAPER
oxidation proceeds via hydroquinone, benzoquinone, and addi-
surface areas have been determined with a Gemini 2370 instrument
according to Brunauer± Emmett ± Teller theory.
[
3, 4, 6, 34, 40]
tional intermediates finally to CO , H O, and HCl.
2
2
Photodegradation procedure and product analysis: The photocatalytic
degradation of 4-CP was carried out in a jacketed cylindrical 15 mL quartz
The results summarized in Figure 2 for photodegradation
with visible light demonstrate the strong influence of the
titania modification. Only anatase afforded photocatalysts of
satisfying activity, whereas the rutile and amorphous phase
based materials were almost inactive. Accordingly, calcination
cuvette attached to an optical train. Irradiation was performed with an
� 6
Osram XBO 150 W xenon arc lamp (I
s
o
(400 ± 520 nm) ˆ 2  10 Einstein
�
1
� 2
cm ) installed in a light-condensing lamp housing (PTI, A1010S).
Appropriate cut-off filters were placed in front of the cuvette. Running
water was circulated through the jacket to ensure constant temperature of
the reaction mixture, which was stirred magnetically.
of TiO -a at 300 ± 3508C produced an anatase-containing
2
material, which after surface modification showed activity
similar to that of the modified P25 photocatalyst. This distinct
difference in the degradation with visible light may be
partially due to the smaller fraction of chemisorbed platinum
complexes present at the rutile and amorphous phase surface.
Furthermore, the much higher activity of anatase relative to
rutile resembles the difference generally observed for these
two crystal modifications in photooxidation processes driven
by UV light. It was assumed that faster recombination of
electron-hole pairs in the case of rutile, due to its poorer
2
In the standard experiment, a pure or modified TiO suspension (14 mL,
� 1
� 4
� 1
0
.5 gL ) containing 4-CP (2.5 Â 10 molL ) was sonicated for 15 min and
then transferred to the cuvette. During an illumination run, the reaction
solution (approximately 1.0 mL) was sampled at given time intervals. The
samples were filtered through a Millipore membrane filter (0.22 mm) and
then subjected to HPLC analysis.
For carbon dioxide determination a IR gas cuvette (10 cm) was inserted
into the system in such way that circulation of the reaction atmosphere
through the quartz and IR cuvettes was enabled with help of a small gas
2
pump (Crouzet SK0251). The CO concentration was determined through
�
1
the strong absorption at 2362 cm . The initial 4-CP concentration was ten
times larger than in the standard photodegradation experiment. In the case
adsorption properties for O , is responsible for the lower
2
� 1
of 1% PtCl
4
/P25 (0.5 gL ) approximately 45 mmol CO
2
(>22% of
[
41]
photocatalytic activity. A further reason could be the higher
theoretical amount) was found after 29 h of irradiation with visible light
(>455 nm) and a further >80 mmol (>38%) after 7 h of UV irradiation
(>320 nm).
electron mobility of anatase as compared to rutile.[
42±44]
The
latter difference can also rationalize the much higher activity
of the surface-modified anatase materials upon excitation
with visible light, when it is postulated that electron injection
and oxygen reduction proceed at different surface sites. In this
case recombination of the injected electron with the chlorine
atom competes with electron trapping at the oxygen adsorp-
tion site. In the modified rutile materials electron trapping
may be too slow to compete with recombination.
Solar illuminations were carried out simultaneously on a May afternoon in
�
1
Erlangen (49.58 N Lat.). The catalysts (0.5 gL ) were suspended in 4-CP
�
4
� 1
solution (40 mL, 2.5 Â 10 molL ) in Pyrex glass reactors with magnetic
stirring. HPLC analysis was performed as described above.
Scavenging of OH radicals: Benzoic acid (10 2 m) was added to the
standard photodegradation experiment. Samples were withdrawn as
described above and salicylic acid was detected through the strong
�
[
45]
fluorescence signal at 400 ± 420 nm when excited at 300 nm; shape and
maximum of emission was identical with an authentic sample. When
necessary, argon was bubbled for 40 min prior and during irradiation.
Experiments in the absence of 4-CP were conducted analogously. When
�
2
C(NO
2
)
4
(10 m) was used as an electron scavenger, after filtration the
Cl and the solvent was
Experimental Section
nonionic substances were extracted with CH
2
2
evaporated. The residue was dissolved in water and subjected to
quantitative analysis by fluorescence spectroscopy. Formation of
Materials: 4-Chlorophenol (4-CP, Fluka) was purified by distillation at low
pressure. The fraction boiling at 355 K (p ˆ 5 Torr) was collected. Triply
�
[35±37]
C(NO
2
)
3
anion was proved by UV spectroscopy at 350 nm.
Salicylic
distilled water was used during all manipulations. The modified TiO
2
acid was also found as the result of a much slower thermal oxidation of
powders 1%K
prepared by grinding various amounts of
Degussa), or H [PtCl ] (Fluka), respectively, in an agate mortar with
Degussa P25 (ca. 70% anatase, 30% rutile), TiO -R (Aldrich, rutile); the
2
[PtCl
4
]/TiO
2
, x% PtCl
4
/TiO
2
, and x% H
2
[PtCl
6 2
]/TiO were
benzoic acid with H
2
O
2
in aqueous solution.
K
2
[PtCl
4
]
(Merck), PtCl
4
Scavenging of Cl atoms: Phenol (10 3 m) was used instead of 4-CP in the
standard irradiation experiment, and disappearance of phenol and
formation of CP were monitored by HPLC as described above.
IV
�
(
2
6
2
procedure was reproducible within Æ10% as indicated by the correspond-
ing photodegradation rates of 4-CP. The amorphous TiO
2
-a, which was
4 2
Pt -desorption experiments: Aqueous suspensions of x% PtCl /TiO and
prepared by a sol ± gel procedure described elsewhere,[ was also ground
3]
2 6 2
1% H [PtCl ]/TiO (5 mg in 2.5 mL of water) were sonicated for 15 min
with PtCl
formed from TiO
4
. As proven by X-ray diffraction analysis anatase can be also
-a by heating at 620 K for 17 h under air. The platinum
and then kept in the dark for five days. After filtration through the
Millipore filter, the filtrate (1.8 mL) was added to HCl (0.6 mL, 2m). The
resulting solution was left for another six days for thermal equilibration.
Thereafter the absorbance at 259 nm was compared with that of standard
2
contents (given in wt% of Pt) of 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 wt% were
calculated under the assumption of complete adsorption.
�
1
4
solutions of PtCl in HCl (0.5m); the extinction coefficient at 259 nm was
4
Irradiation of 1% PtCl /P25 (0.5 gL ) in methanol/water (1:1 v/v) with
3
� 1
3
� 1
� 1
3
� 1
IV
0
determined as 11.8 Â 10 mol dm cm as compared to 400 mol dm cm
UV light (l > 320 nm) resulted in reduction of Pt to Pt and significant
darkening of the catalyst. In visible light (l > 455 nm) this process was
observed only after long irradiation times (ca. 20 h).
II
found for the corresponding Pt complex.
Instruments: Diffuse reflectance spectra of the solids were recorded on a
Shimadzu UV-3101PC UV/Vis-NIR scanning spectrophotometer equipped
with a diffuse reflectance accessory. Samples were spread onto an Al O
2 3
plate, the background reflectance of which was measured prior to
measurement. Reflectance was converted by the instrument software to
F(R
1
) values according to the Kubelka ± Munk theory. Electronic absorp-
Acknowledgement
tion spectra were recorded on the same apparatus, fluorescence spectra on
a Perkin Elmer LS 50B luminescence spectrometer. An HPLC (Kroma-
System 2000 controller, Kontron 320 autosampler with 430A photodiode
array detector) was used to determine concentrations of 4-CP. A reverse-
phase C-18 column (Kontron, Spherisorb 5 ODS 2) and water ± methanol
eluent (40/60, v/v) were employed. Absorbances at 280 and 220 nm were
measured for the detection of 4-CP and phenol, respectively. Specific
W. M. thanks Deutsche Forschungsgemeinschaft for a fellowship within the
Graduiertenkolleg ªHomogener und Heterogener Elektronentransferº.
Specific surface areas determinations and transmission electron microg-
raphy experiments were gratefully performed by Mr. Peter Widlok and Dr.
G. Frank, respectively. We are indebted Degussa AG for a gift of H
This work was supported by the Bavarian State program BayForrest.
2 6
[PtCl ].
1866
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-1866 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

Photosensitization of TiO2
1862±1867
[
[
[
[
1] A. Fujishima, K. Hashimoto, T. Watanabe, TiO
2
Photocatalysis.
[24] R. B. Draper, M. A. Fox, Langmuir 1990, 6, 1396.
[25] T. Ohno, D. Haga, K. Fujihara, K. Kaizaki, M. Matsumura, J. Phys.
Chem. B 1997, 101, 6415.
[26] E. Vrachnou, N. Vlachopoulos, M. Grätzel, J. Chem. Soc. Chem.
Commun. 1987, 868.
[27] E. Vrachnou, M. Grätzel, A. J. McEvoy, J. Electroanal. Chem. 1989,
258, 193.
[28] S. Ferrere, B. A. Gregg, J. Am. Chem. Soc. 1998, 120, 843.
[29] S. E. Livingstone, in Comprehensive Inorganic Chemistry, Vol. 3, 1973.
[30] R. B. Cundall, R. Rudham, M. S. Salim, J. Chem. Soc. Faraday Trans. 1
1976, 72, 1642.
Fundamentals and Applications, 1st ed., BKC, Tokyo, 1999.
2] H. Kisch, L. Zang, C. Lange, W. F. Maier, C. Antonius, D. Meissner,
Angew. Chem. 1998, 110, 3201; Angew. Chem. Int. Ed. 1998, 37, 3034.
3] L. Zang, C. Lange, W. F. Maier, I. Abraham, S. Storck, H. Kisch, J.
Phys. Chem. B 1998, 102, 10765.
4] L. Zang, W. Macyk, C. Lange, W. F. Maier, C. Antonius, D. Meissner,
H. Kisch, Chem. Eur. J. 2000, 6, 379.
[
5] H. Al-Ekabi, N. Serpone, J. Phys. Chem. 1988, 92, 5726.
6] H. Al-Ekabi, N. Serpone, E. Pelizzetti, C. Minero, M. A. Fox, R. B.
Draper, Langmuir 1989, 5, 250.
[
[
7] a) U. Stafford, K. A. Gray, P. V. Kamat, J. Phys. Chem. 1994, 98, 6343;
b) J. Theurich, M. Lindner, D. W. Bahnemann, Langmuir 1996, 12,
[31] I. Izumi, F.-R. F. Fan, A. J. Bard, J. Phys. Chem. 1981, 85, 218.
[32] M. Fujihira, Y. Satoh, T. Osa, Bull. Chem. Soc. Jpn. 1982, 55, 666.
[33] R. W. Matthews, J. Chem. Soc. Faraday Trans. 1 1984, 80, 457.
[34] K. Okamoto, Y. Yamamoto, H. Tanaka, M. Tanaka, A. Itaya, Bull.
Chem. Soc. Jpn. 1985, 58, 2015.
6368.
[
8] J. Sykora, M. Pado, M. Tatarko, M. Izakovic, J. Photochem. Photobiol.
A: Chem. 1997, 110, 167.
[
9] L. E. Cox, D. G. Peters, Inorg. Chem. 1970, 9, 1927.
[35] K. D. Asmus, H. Henglein, Ber. Bunsen-Ges. Phys. Chem. 1968, 68,
348.
[36] H. Henglein, Ber. Bunsen-Ges. Phys. Chem. 1982, 86, 241.
[37] P. Boarini, V. Carassiti, A. Maldotti, R. Amadelli, Langmuir 1998, 14,
2080.
[38] a) S. D. Malone, J. F. Endicott, J. Phys. Chem. 1972, 76, 2223; b) W.
Latimer, Oxidation Potentials, 2nd ed., Prentice-Hall, Englewood
Cliffs, 1952.
[
10] L. E. Cox, D. G. Peters, E. L. Wehry, J. Inorg. Nucl. Chem. 1972, 34,
97.
11] C. M. Davidson, R. F. Jameson, Trans. Faraday Soc. 1965, 61, 2462.
12] a) J. P. Bucher, J. Buttet, J. J. van der Klink, M. Grätzel, E. Newson,
T. B. Truong, J. Mol. Catal. 1987, 43, 213; b) B. Kraeutler, A. J. Bard, J.
Am. Chem. Soc. 1978, 100, 4317.
2
[
[
[
13] T. N. Angelidis, M. Koutlemani, I. Poulios, Appl. Catal. B: Environ.
1
998, 16, 347.
[39] a) K. Hasegawa, P. Neta, J. Phys. Chem. 1978, 82, 854; b) G. Grabner,
G. Li, R. M. Quint, R. Quint, N. Getoff, J. Chem. Soc. Faraday Trans.
1991, 87, 1097.
[
14] The measured particles size and specific surface area are in agreement
with values reported by the producer of P25: Degussa Tech. Bull. 1984,
5
6, 8.
15] R. Palmans, A. J. Frank, J. Phys. Chem. 1991, 95, 9438.
16] pK values of 2.9 and 12.7 have been reported for TiO . M. Herrmann,
[40] J. Cunningham, G. Al-Sayyed, P. Sedlak, J. Caffrey, Catal. Today 1999,
53, 145.
[41] N. S. Lewis, M. L. Rosenbluth, in Photocatalysis ± Fundamentals and
Applications (Eds.: N. Serpone, E. Pelizzetti), Wiley, New York, 1989.
[42] L. Forro, O. Chauvet, D. Emin, L. Zuppiroli, H. Berger, F. L e vy, J.
Appl. Phys. 1994, 75, 633.
[
[
a
2
H. P. Boehm, Z. Anorg. Chem. 1969, 368, 73.
17] The pK
18] J. S. Curran, J. Domenech, N. Jaffrezic-Renault, R. Philippe, J. Phys.
Chem. 1985, 89, 957.
O)] was reported as 3.8.[11]
�
[
[
a
value for [PtCl
5
(H
2
[43] H. Tang, K. Prasad, R. Sanjin eÁ s, P. E. Schmid, F. L e vy, J. Appl. Phys.
[
[
19] J.-M. Herrmann, J. Disdier, P. Pichat, J. Phys. Chem. 1986, 90, 6028.
20] E. Borgarello, N. Serpone, G. Emo, R. Harris, E. Pelizzetti, C. Minero,
Inorg. Chem. 1986, 25, 4499.
1994, 75, 2042.
[44] A. Shiga, A. Tsujiko, S. Yae, Y. Nakato, Bull. Chem. Soc. Jpn. 1998, 71,
2119.
[
21] Results observed for TiO
2
from Bayer, which contains predominantly
[45] N. Huang, M. Xu, C. Yuan, R. Yu, J. Photochem. Photobiol. A: Chem.
1997, 108, 229.
the anatase form of titania. G. Burgeth, Diploma Thesis, Friedrich-
Alexander-Universität Erlangen-Nürnberg, 1999.
[
22] R. C. Wright, G. S. Laurence, J. Chem. Soc. Chem. Commun. 1972,
1
32.
23] W. L. Waltz, J. Lilie, A. Goursot, H. Chermette, Inorg. Chem. 1989, 28,
247.
[
Received: June 23, 2000
Revised: November 10, 2000 [F2563]
2
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-1867 $ 17.50+.50/0
1867