Characterization of adsorbed species on TiO2 after photocatalytic oxidation of toluene

Article abstract of DOI:10.1006/jcat.2001.3174

Heterogeneous photocatalytic oxidation (PCO) is a promising method for removing organic pollutants present in low concentrations in waste gas streams. Strongly bound intermediates formed that were much less reactive than toluene during the initial step of PCO of toluene on TiO₂ at room temperature. Temperature-programmed hydrogenation (TPH) was an effective technique to characterize these intermediates. Titania was a hydrogenation catalyst at elevated temperatures, and the intermediates hydrogenated to toluene and benzene, which were more weakly bound. Platinum was added to TiO₂ to increase the hydrogenation rate through spillover of hydrogen so that the intermediates were removed at lower temperatures, and the platinum also increased the PCO rate at room temperature. More than one intermediate might be on the surface, and the intermediates had intact aromatic rings. Similarities in TPH spectra showed that toluene reacted through benzaldehyde, but benzaldehyde oxidized quickly to some less reactive intermediate. Benzaldehyde adsorbed in more than one form on TiO₂, and these forms had different reactivities during PCO. Benzoic acid was not the less reactive intermediate formed during benzaldehyde PCO.

Full text of DOI:10.1006/jcat.2001.3174

Journal of Catalysis 200, 21–33 (2001)
doi:10.1006/jcat.2001.3174, available online at http://www.idealibrary.com on
Characterization of Adsorbed Species on TiO after Photocatalytic
2
Oxidation of Toluene
1
M. Catherine Blount and John L. Falconer
Department of Chemical Engineering, University of Colorado, Boulder, Colorado 80309-0424
Received June 21, 2000; revised January 18, 2001; accepted January 18, 2001; published online April 11, 2001
Transient reaction of a monolayer of toluene at room
During the initial step of photocatalytic oxidation (PCO) of temperature was used to measure O2 consumption and CO2
toluene on TiO2 at room temperature, strongly bound intermedi- formation. The intermediates that form on the surface were
ates form that are much less reactive than toluene. Temperature-
programmed hydrogenation (TPH) is shown to be an effective
characterized by temperature-programmed hydrogenation
(
TPH) because hydrogenation is expected to saturate the
technique to characterize these intermediates. Titania is a hydro-
genation catalyst at elevated temperatures, and the intermediates
hydrogenate to benzene and toluene, which are more weakly bound.
Platinum was added to TiO2 to increase the rate of hydrogenation
through spillover of hydrogen so that the intermediates were re-
moved at lower temperatures, and the Pt also increased the rate
of PCO at room temperature. More than one intermediate may be
intermediates so that they are more weakly held and des-
orb. Though H2 was used for most TPH experiments, some
experiments used a D2 flow to help identify intermediates.
Platinum was added to the TiO2 since it has been reported
to increase the rate of PCO (19, 23–27), but also to take
advantage of spillover of H atoms onto the TiO2 to more
on the surface, and the intermediates have intact aromatic rings. effectively remove the PCO intermediates from the TiO2
Similarities in TPH spectra indicate that toluene reacted through surface. For Pt loading of 0.2% , less than 0.4% of the TiO2
benzaldehyde, but benzaldehyde oxidized quickly to some less re- surface is covered with Pt, and thus the photocatalytic pro-
active intermediate. Benzaldehyde adsorbed in more than one form
on TiO2, and these forms had different reactivities during PCO.
Benzoic acid is not the less reactive intermediate formed during
cesses are expected to be similar to those on TiO2 alone. Fu
et al. (19) observed that Pt increased both the rate of ben-
zene consumption at 343 K and the ratio of CO2 produced
to benzene consumed.
Most previous studies of gas-phase PCO of BETX com-
pounds on TiO2 and Pt/TiO2 have focused on PCO rates
rather than on the identification of adsorbed intermediates
and surface processes, which are poorly understood (18,
benzaldehyde PCO.
�c 2001 Academic Press
Key Words: photocatalytic oxidation; toluene.
INTRODUCTION
Heterogeneous photocatalytic oxidation (PCO) is a 19, 23–31). Ibusuki and Takeughi (8) reported that dur-
promising technique for the removal of organic pollutants ing PCO of toluene in dry air on TiO2, a small amount
present in low concentrations in waste gas streams. During of benzaldehyde formed in addition to CO2. Larson and
PCO, the catalyst absorbs UV light, which excites electrons Falconer (9) observed that toluene reacts quickly on TiO2 to
from the valence band into the conduction band. The re- form strongly bound intermediates that react more slowly
sulting electron/hole pairs can then migrate to the surface to form CO2 and H2O. They observed that the interme-
and initiate redoxreactionswith adsorbed organics. Though diates formed during PCO of toluene reacted differently
the objective of PCO is complete oxidation, many organics from benzaldehyde, benzyl alcohol, and m-cresol, indicat-
react to form partial oxidation products (1–22). Less reac- ing that these compounds are probably not the intermedi-
tive, strongly bound compoundsform duringPCO of BETX ates formed during toluene PCO. The less reactive, strongly
(
(
benzene, ethylbenzene, toluene, and xylenes) compounds bound intermediates either thermally decomposed or oxi-
7–19). Accumulation of these partial oxidation products dized during temperature-programmed desorption (TPD)
can poison the catalyst surface by blocking reaction sites. or oxidation (TPO) (9).
Toluene, a potential human carcinogen and common air
Jacoby et al. (10) used water to extract adsorbed interme-
pollutant, was used in this study as a model aromatic to diates from a TiO2 catalyst following steady-state benzene
characterize surface intermediates that form during PCO.
PCO. Their analysis identified phenol, benzoquinone
and/or hydroquinone, and malonic acid as possible inter-
mediates. Since they only reacted several monolayers of
toluene, these species may be the primary intermediates.
1
To whom correspondence should be addressed. Fax: 303 492-4341.
E-mail: john.falconer@colorado.edu.
21
0021-9517/01 $35.00
Copyright �c 2001 by Academic Press
All rights of reproduction in any form reserved.

2
2
BLOUNT AND FALCONER
D’Hennezel et al. (14) observed similar species during a wa- thin-film, annular photoreactor with a 1-mm annular spac-
ter extraction following humidified benzene PCO on TiO2. ing was used to ensure high gas flow rates over the catalyst.
Recent studies have used extraction and FTIR to study This increased mass transfer and rapidly flushed gas-phase
PCO of a continuous flow of toluene on TiO2 (12–14). These products from the reactor. A quartz reactor, which had a
studiesfound that surface speciesformed duringPCO ofun- porous quartz frit to support the catalyst and allow gas flow,
humidified toluene deactivate the catalyst. Mendez-Roman was used for the powdered TiO2 experiment. The catalysts
and Cardona-Martinez (12) observed benzaldehyde and used were Degussa P25 TiO2 (75% anatase/25% rutile with
2
benzoic acid on the surface following toluene PCO and pro- a BET surface area of 50 m /g) and 0.2% (wt) Pt/TiO2. Ap-
posed that toluene reacts to form benzaldehyde, which then proximately 30 mg of catalyst coated the inside of the Pyrex
reacts to benzoic acid, which deactivates the surface. Au- photoreactor. A cylindrical quartz furnace wrapped with
gugliaro et al. (13) detected benzaldehyde, benzene, benzyl Kanthal wire surrounded the reactor. The heating wire was
alcohol, CO2, and trace amounts of benzoic acid and phenol spaced so that six near-UV lamps (GE F4T5-BLB), posi-
during PCO of a continuous flow of toluene on TiO2 using tioned evenly around the reactor and approximately 2.5 cm
GC and extraction. The catalyst deactivated after 20 h of from the catalyst film, could illuminate the catalyst. These
PCO of unhumidified toluene. Benzene production was a 4-W bulbs generate light in the 300–500 nm wavelength
transient reaction product that appearing during the first 4 h range (near-UV) with a maximum intensity at 356 nm (33).
of reaction. Carbon dioxide was always associated with ben- The light intensity at the catalyst surface was measured
2
zene formation and was not detected at later reaction times, as 0.3 mW/cm . A temperature controller used feedback
when only benzaldehyde was detected in the gas phase. from the thermocouple to maintain a constant heating rate
The authors proposed that toluene reacts with adsorbed hy- during temperature-programmed experiments. The reactor
droxyl radicals to form benzaldehyde during the first steps effluent was analyzed by a quadrupole mass spectrometer
of PCO, and that benzaldehyde then reacts with another hy- (Balzers QMA 125), which was housed in a turbomolecular-
droxyl radical to form benzoic acid, which decomposes to pumped vacuum chamber located downstream of the reac-
benzene and CO2. They proposed that the catalyst deacti- tor. A heated, stainless steel capillary was used to divert
vated during PCO of unhumidified toluene because surface some of the effluent into the vacuum chamber for anal-
hydroxyls were consumed (32). In contrast to other stud- ysis. Computer-controlled data acquisition simultaneously
ies, they proposed that benzaldehyde was the unreactive monitored and recorded selected masssignals, temperature,
intermediate responsible for deactivation during PCO of and elapsed time. The mass spectrometer signals were cali-
unhumidified toluene. None of these more recent studies brated by injecting known quantities of gases or liquids into
quantified their results; thus, strongly bound intermediates the appropriate gas stream composition, and mass signals
or gas-phase products may have been missed.
Transient reaction techniques are useful for studying
were corrected for cracking in the mass spectrometer.
The catalyst was heated to 723 K in a flowing O2/He mix-
PCO because the adsorption, surface reaction, and desorp- ture before each experiment to obtain a reproducible sur-
tion steps can be separated in time. In the current study, face. Organics were injected upstream of the reactor and
an adsorbed layer of toluene was oxidized in the absence allowed to evaporate into the flowing gas and adsorb onto
of gas-phase toluene, and less than a monolayer was re- the cooled catalyst. Because benzoic acid is solid at room
acted so that the intermediates detected are more likely temperature, it was heated to 460 K, and the resulting ben-
to be the result of toluene PCO and not side reactions. zoic acid vapor contacted TiO2. Because toluene is weakly
Toluene is weakly adsorbed on TiO2 and desorbs intact adsorbed on TiO2 and slowly desorbs at room tempera-
at relatively low temperatures (9); thus unreacted toluene ture, the reactor was cooled to 273 K following toluene
can be readily separated from intermediates that are more adsorption on TiO2 to ensure that gas-phase toluene was
strongly bound to the surface. Consequently, temperature- not present during PCO and that accurate measurements
programmed experiments are well suited to characterize of adsorbed and reacted toluene could be made. The reac-
the intermediates. Photocatalytic oxidation using labeled tor was cooled to 253 K for Pt/TiO2 experiments because
13
toluene (C6H5� CH3) provided information on the oxida- toluene desorbed more readily from the Pt/TiO2 catalyst
tion of ring and methyl carbons on Pt/TiO2. Benzaldehyde than from the TiO2 catalyst. Benzaldehyde PCO was per-
PCO was also studied because it is a possible intermedi- formed at room temperature. For experiments with less
ate during toluene PCO (8, 10, 12–16). The TPH spectra of than saturation coverage of benzaldehyde, only a fraction
other proposed intermediates (8, 12, 14) (benzyl alcohol, of the amount of organic needed for saturation coverage
benzoic acid, and m-cresol) were also investigated.
was injected into the flow gas for adsorption. An aluminum
shroud shielded the reactor from the near-UV lights at the
start of PCO experiments, and the lights warmed up for ap-
proximately 15 min before the shroud was removed. The
EXPERIMENTAL METHODS
The apparatus used for PCO and the temperature- intensity of the lights increased by approximately 5% dur-
programmed experiments was described previously (3). A ing the first 10 min of operation, so the warm-up period

PHOTOCATALYTIC OXIDATION OF TOLUENE ON TiO2
23
allowed the lights to reach a constant output before illumi-
nating the catalyst. The PCO was carried out in 0.2% O2
in He so that the rate of O2 consumption could be mea-
sured.
Species adsorbed on the catalyst surface were ana-
lyzed bytemperature-programmed oxidation (TPO), temp-
erature-programmed desorption (TPD), and temperature-
programmed hydrogenation (TPH). During TPO, a 20%
3
O2/He mixture (100 standard cm /min) flowed over the
catalyst surface as the temperature was increased at a con-
stant rate of 1 K/s. During TPD and TPH, pure He and
H2, respectively, flowed over the catalyst at the same rate.
Most of the TPH experiments were carried out in 100%
H2 flow to obtain high rates of hydrogenation of the in-
termediates, but some TPH experiments on TiO2 used a
3
5
% D2/He flow (100 standard cm /min) and the various
deuterated products were detected. Heating was stopped at
23 K because of difficulties in maintaining a constant heat-
FIG. 1. Oxygen consumption and CO2 formation during 20-min PCO
of toluene adsorbed to saturation coverage on TiO2 at 273 K.
7
ing rate at higher temperatures and to avoid irreversible
changes in the catalyst at higher temperatures. The catalyst
was held at 723 K until all signals returned to their original
values. An interrupted TPD to 348 K was performed fol-
lowing toluene PCO but before TPH to remove unreacted
toluene so that the amount of unreacted toluene during
PCO could be determined, and any toluene formed dur-
ing TPH could be distinguished from unreacted toluene.
Temperature-programmed desorption showed that toluene
was much more weakly held than the other species formed,
so only toluene desorbed during TPD. During some TPH
experiments, the effluent gas flowed through a bed of the
polymer resin Tenax TA (from Scientific Instrument Ser-
The aromatics reacted and their initial coverages on the
catalyst surface are given in Table 1.
During TPD of adsorbed toluene when no reaction was
performed, toluene desorbed in a single peak with a max-
imum at 370 K. Saturation coverage of adsorbed toluene
at 273 K was determined to be � 80 �mol/g TiO2. After
2
0 min of PCO of toluene on TiO2, only � 22 �mol of unre-
acted toluene/g of TiO2, or 27% of the originally adsorbed
toluene, desorbed during TPD (Fig. 2). That is, 73% of the
toluene monolayer oxidized during 20 min of PCO in the
low-intensity light used. Few gas-phase products formed
during 20 min of PCO, and the amounts that desorbed dur-
ing TPD were small, so the rest of the toluene apparently
reacted to form surface intermediates that did not desorb
by 723 K. Following toluene PCO, a second small toluene
peak, with a maximum at 610 K (Fig. 2), desorbed in ad-
dition to unreacted toluene. A small amount of benzene
also desorbed with two maxima at high temperatures. In
addition, CO2 formed in small quantities over a broad tem-
perature range, and CO desorbed in a high-temperature
peak.
2
vices, total area of 35 m /g) adsorbent to trap desorbed
organics for headspace analysis with a GC/MS (Hewlett-
Packard Model 5890 GC equipped with a HP Model 5791
mass-selective detector). Tenax was used because it has a
high affinity for organics but a low affinity for water. The
products from several TPHs were collected on the same
resin in repeat experiments to obtain higher concentrations
for GC/MS analysis.
RESULTS
Photocatalytic Oxidation of Toluene on TiO2
TABLE 1
Toluene is weakly adsorbed on TiO2, and desorbs intact
at relatively low temperatures. Because the toluene des-
orption rate was finite at 300 K, the catalyst was cooled to
Oxygen Consumed and CO Produced during 20-min PCO
2
Percent
Adsorbed
aromatic
saturation
coverage
O2 consumed
(�mol/g catalyst)
CO2 produced
(�mol/g catalyst)
2
73 K before PCO. As reported by Larson and Falconer (9),
when a monolayer of toluene on TiO2 was illuminated with
UV at 273 K, O2 was immediately consumed. As shown
in Fig. 1, the rate of O2 consumption quickly reached a
maximum and then decayed. Carbon dioxide was the only
gas-phase species detected during PCO (Fig. 1), and its rate
of formation was much slower than the rate of O2 uptake.
Table 1 shows the formation and uptake amounts for PCO.
TiO2
Toluene
Benzaldehyde
100
60
68 � 4
38 � 7
2 � 0.4
9 � 2
Pt/TiO2
135� 8
171
Toluene
Benzaldehyde
100
60
6
38

2
4
BLOUNT AND FALCONER
TABLE 3
Amounts Desorbed during TPD (No PCO) and TPO on TiO2
Amount desorbed (�mol/g catalyst)
Adsorbed
aromatic
TPO Total
Toluene Benzene
CO
CO2
4
CO2
C
Toluene
Benzaldehyde
81 � 8
571
175
a
3 � 1
5 � 2
19 � 16 16 � 8
89
a
Benzaldehyde adsorbed to 15% of saturation coverage.
genate them; a more hydrogenated species is expected to
be more weakly bound to TiO2. To distinguish products of
hydrogenation from the unreacted toluene, an interrupted
TPD was first performed to 348 K after PCO to remove un-
reacted toluene. Toluene had a significant desorption rate
FIG. 2. Temperature-programmed desorption spectra after 20-min
PCO of toluene, which was adsorbed to saturation coverage on TiO2 at at 348 K, and the TiO was held at that temperature un-
2
2
73 K.
til all of the unreacted toluene desorbed. Figure 3 shows
the TPH spectra obtained after a 20-min PCO of toluene
and an interrupted TPD to remove unreacted toluene. Note
that although TPD removed all the unreacted toluene, the
largest peak during TPH is toluene, and it forms at much
higher temperatures than adsorbed toluene desorbs from
TiO2. In addition to toluene, benzene and a small amount of
methylcyclohexane (MCH, not shown for clarity) desorbed
during TPH. The reactor effluent was trapped using a bed
of Tenax adsorbent for several of these experiments. Subse-
quent GC/MS headspace analysis of the Tenax confirmed
the presence of benzene and toluene. No other organics
were detected in identifiable quantities. Neither CO2 nor
CO adsorbs significantly on Tenax and so neither was de-
tected.
The species that were not removed by TPD were oxi-
dized to CO2 and H2O during a TPO that followed the
TPD. A comparison to the amount of toluene that des-
orbed during TPD without PCO indicates that the combi-
nation of TPD and TPO removed all the adsorbed species.
Carbon dioxide formed at temperatures as low as 350 K
during TPO; however, most of the surface species remain-
ing after TPD were oxidized to CO2 at temperatures above
4
�
50K. Temperature-programmed desorption removed only
20% of the strongly bound intermediates that formed dur-
ing PCO. Apparently, TiO2 is a sufficiently good thermal
oxidation catalyst at high temperature that all the interme-
diates were oxidized during TPO. Tables 2 and 3 show the
amount of each desorbing species during TPD and the sub-
sequent TPO for the indicated adsorbed aromatic. The total
carbon was calculated by summing the carbon contained in
the desorbing species during TPD and TPO.
Temperature-Programmed Hydrogenation
of Toluene on TiO2
Because the intermediates are strongly bound to the sur-
face and do not desorb by 723 K, TPH was used to hydro-
TABLE 2
Amounts Desorbed during TPD after 20-min PCO on TiO2
Amount desorbed (�mol/g catalyst)
Adsorbed
aromatic
Benzal-
TPO Total
Toluene Benzene dehyde CO CO2 CO2
C
Toluene
Benzaldehyde
25a
4
4
16
27
12
12
48
339
449
577
682
b
7
FIG. 3. Temperature-programmed hydrogenation spectra after
20-min PCO of toluene, which was adsorbed to saturation coverage
on TiO2 at 273 K. After PCO, an interrupted TPD to 338 K removed
unreacted toluene prior to TPH.
a
Note: � 3 �mol/g TiO2 toluene was from the decomposition of
strongly bound intermediates.
b
Benzaldehyde adsorbed to 60% of saturation coverage.

PHOTOCATALYTIC OXIDATION OF TOLUENE ON TiO2
25
TABLE 4
Amounts Formed during Temperature-Programmed Experiments after 20-min PCO
Amount desorbed (�mol/g catalyst)
Adsorbed
aromatic
TPH
MCH
Other
organics
TPO
(CO2)
TPD
Toluene
Benzene
CO
CO2
Total C
TiO2
3 � 1
1
Toluene
22 � 3
29 � 5
8 � 2
20 � 7
8 � 2
79 � 35
533
806
Benzaldehydea
55
17
1
53
9
243
Pt/TiO2
27 � 6
10
Toluene
Benzaldehyde
25 � 8
25 � 5
8 � 1
20
38
88
16 � 3
19 � 1
695
569
b
35
21
20
20
a
Benzaldehyde adsorbed to 70% of saturation coverage on TiO2.
Benzaldehyde adsorbed to 30% of saturation coverage on Pt/TiO2.
b
The PCO and subsequent temperature-programmed ex- (34), where Vo represents oxygen vacancies, and
periment results were reproducible on the same cata-
lyst, even after extensive use of the catalyst. A repeat of
the experiment shown in Fig. 3 performed approximately
TiO2 + H2 → Ti3O5 + H2O
1H = 8.9 kJ/mol [2]
6
months later showed that the amounts desorbed and re- (35). Luo and Falconer (36) observed that during TPH of
action rates were approximately the same for PCO and the oxidized TiO2, H2O from TiO2 reduction desorbed at 600–
temperature-programmed experiments.
620 K, which may indicate that TiO2 is not reduced until
Table 4 lists the amounts of the desorbing species de- 600 K. A TPO after TPH successfully reoxidized their cata-
tected during interrupted TPD, TPH, and the subsequent lyst.
TPO following reaction of the indicated aromatic. Total car-
bon was calculated by adding the amount of carbon con- Photocatalytic Oxidation of Benzaldehyde on TiO2
tained in the desorbing species during interrupted TPD,
TPH, and subsequent TPO. Approximately 22 �mol of
toluene/g of TiO2 desorbed during the interrupted TPD.
Because benzaldehyde has been suggested as the inter-
mediate that forms during toluene PCO on TiO2 (8, 12–15),
the same experiments were carried out for a submonolayer
This amount is approximately the same as the amount of
of adsorbed benzaldehyde. A submonolayer was used so
unreacted toluene desorbed during a normal TPD follow-
that the amounts of desorbing species were similar to the
ing 20-min PCO (Table 2), and this suggests that all the
amounts desorbed during TPD and TPH after toluene
unreacted toluene was removed during interrupted TPD.
PCO. Though the amount of benzaldehyde adsorbed cor-
Although only � 3 �mol of toluene/g of TiO2 formed at
responds to less than a monolayer, because benzaldehyde
elevated temperatures during TPD from decomposition
is strongly adsorbed the local coverage may not be less than
of strongly bound intermediates following 20-min PCO,
a monolayer. That is, most or all of the benzaldehyde may
3
0 �mol of toluene/g of TiO2 formed during TPH after
have adsorbed at the top portion of the reactor rather than
being distributed evenly over the entire catalyst surface.
Temperature-programmed desorption of benzaldehyde
adsorbed to 15% of saturation coverage on TiO2 was per-
formed. The toluene, benzene, and CO desorption peaks
and desorption amounts (Table 3) were similar to those
seen during TPD following toluene PCO (Fig. 2). Much
of the benzaldehyde did not desorb during TPD, and thus
TPO was used to measure the amount of benzaldehyde
adsorbed.
2
0-min PCO and interrupted TPD. Thus, the stronglybound
surface species that formed during PCO readily hydro-
genated to toluene during TPH. Surface species remain-
ing after TPH were oxidized to CO2 during TPO. Only 80
�
mol of CO2/g of TiO2 desorbed during TPO after TPH
whereas 339 �mol of CO2/g of TiO2 formed during TPO
after TPD; therefore, at least 80% of the strongly bound in-
termediate was hydrogenated to gas-phase products during
TPH.
Following TPH, the catalyst had a blue color indicat-
ing reduction of TiO2. The catalyst returned to its white
color during the subsequent TPO. Past studies (34–36) have
shown that TiO2 can be reduced in H2. Two mechanisms for
reduction have been proposed:
Toluene, benzene, CO, CO2, and unreacted benzalde-
hyde desorbed during the TPD following 20-min PCO of
benzaldehyde at 60% of saturation coverage. The forma-
tion of toluene and benzene was similar to the formation
of these species during TPD of benzaldehyde without re-
action. Carbon monoxide, however, desorbed at a lower
temperature during TPD following benzaldehyde PCO,
1H = 183 kJ/mol [1] and CO2 did not form until high temperatures following
Ti + O2� + H2
4
+
3
+
→
H2O + Vo + Ti + e�

2
6
BLOUNT AND FALCONER
benzaldehyde PCO, whereas CO2 formed over a broad tem-
perature range following toluene PCO. Thus, much of the
CO and CO2 formed during TPD after benzaldehyde PCO
appears to be from the decomposition of benzaldehyde ox-
idation products.
TABLE 5
Amount Formed during TPH and TPO (No PCO)
Amount desorbed (�mol/g catalyst)
Adsorbed
aromatic
TPH
TPO
Toluene Benzene MCH CO CO2 CO2 Total C
Temperature-Programmed Hydrogenation
of Benzaldehyde on TiO2
TiO2
Benzaldehydea
25
5
1
8
2
29
251
Though not much benzaldehyde or benzaldehyde de-
composition products desorbed during TPD of adsorbed
benzaldehyde (20% of saturation coverage), benzaldehyde
readily hydrogenated to toluene during TPH, as shown in
Fig. 4. Desorption amounts are given in Table 5. In addi-
tion to toluene, small amounts of benzene, CO2, CO, and
MCH (not shown for clarity) formed. Toluene formed in
two overlapping peaks centered at 550 K and also in a high-
temperature peak.
Pt/TiO2
Benzyl alcohol
47
54
14
1
2
2
23
90
2
127
13
>360
>517
1070
m-Cresol
Benzaldehyde
b
121
10
a
Benzaldehyde adsorbed to 20% of saturation coverage on TiO2.
Benzaldehyde adsorbed to 55% of saturation coverage on Pt/TiO2.
b
725 K. A small amount of benzene also formed, and CO
During TPH following benzaldehyde PCO, toluene, ben- formed in a broad peak. The lower temperature toluene
zene, CO, CO2, and a small amount of benzaldehyde des- peak was similar to the toluene peak observed during TPH
orbed, as shown in Fig. 5. Desorption amounts are shown after toluene PCO.
in Table 4. Toluene formed in a broad peak at 550 K
Benzoic acid hydrogenated mainly to benzaldehyde at
and a high-temperature peak. The TPH after benzalde- high temperatures. Smaller amounts of toluene and ben-
hyde PCO is similar to the TPH of adsorbed benzaldehyde, zene also formed at high temperatures as well as CO2, CO,
but more CO, CO2, and high-temperature benzene and and CH4. The TPH spectra of benzoic acid are significantly
toluene formed during TPH after benzaldehyde PCO. All different from the TPH spectra after toluene PCO and after
the species formed during TPH after benzaldehyde PCO benzaldehyde PCO, since not much benzaldehyde desorbed
(
Fig. 5) are comparable to the respective species formed at high temperature after those PCOs.
during TPH after toluene PCO (Fig. 3).
During TPH of adsorbed m-cresol, a small toluene peak
formed with a maximum at 520 K, and a large toluene peak
formed at 730 K. Both toluene peaks were at the same
temperatures as the toluene observed in the TPH follow-
ing toluene PCO. Carbon monoxide desorbed in a high-
temperature peak similar to that observed duringTPH after
toluene PCO. A small amount of benzene also formed.
Temperature-Programmed Hydrogenation of Benzyl
Alcohol, Benzoic Acid, and Cresol on TiO2
During TPH of adsorbed benzyl alcohol, toluene formed
in a large peak centered at 525 K and a smaller peak at
FIG. 4. Temperature-programmed hydrogenation spectra of ben-
zaldehyde, which was adsorbed to 20% of saturation coverage on TiO2
at 300 K.
FIG. 5. Temperature-programmed hydrogenation after 20-min PCO
of benzaldehyde, which was adsorbed to 70% of saturation coverage on
TiO2 at 300 K.

PHOTOCATALYTIC OXIDATION OF TOLUENE ON TiO2
27
Temperature-Programmed Hydrogenation
with Deuterium on TiO2
During TPH using 5% D2, both deuteration of the
strongly bound species and deuterium exchange of hydro-
gen on the adsorbed molecules took place. During a TPH
in 5% D2 of adsorbed toluene when no PCO had been per-
formed, toluene with 0–6 D atoms desorbed, but no MCH
formed. Thus, deuterium exchanged with hydrogen on the
adsorbed toluene. Figure 6 shows the mass peaks for the
deuterated toluene and benzene products that formed dur-
ing TPH after a 20-min PCO of toluene and an interrupted
TPD to remove unreacted toluene. Masses 91, 92, 93, and
9
4 correspond to deuterated and undeuterated toluene and
mass 79 corresponds to benzene with 1 deuterium. The mass
spectra indicate that toluene with 0–6 deuterium atoms and
benzene with 0–3 deuterium atoms (undeuterated and ben-
zene with 2–3 D not shown) form. Deuterium is added to
the strongly bound species by either hydrogenation or ex-
change. Although 92 amu is the molecular weight of un-
deuterated toluene, fragmentation in the mass spectrom-
eter results in 91 amu as the main mass detected for un-
deuterated toluene. Likewise, toluene with 1 D is assumed
to be detected at mass 92, and toluene with 2 and 3 D is
assumed to appear at masses 93 and 94, respectively. The
more highly deuterated benzene and toluene species and
undeuterated benzene form in smaller amounts and are not
shown in Fig. 6. Because the carrier gas contained only 5%
D2, less toluene formed than formed in the TPH experi-
ment using pure H2 carrier gas. Experiments with 5% D2
show that only � 25% as much toluene and � 70% as much
benzene formed as with pure H2. In addition, the peaks in
Fig. 6 are much broader than those in Fig. 3 and are shifted
FIG. 7. Temperature-programmed hydrogenation spectra in 5%
D2/He of benzaldehyde, which was adsorbed to 45% of saturation cov-
erage on TiO2 at 300 K.
to higher temperature because of the lower pressure of D2.
Toluene and benzene form at slightly higher temperatures
as the amount of deuterium in the molecules increases.
For comparison, Fig. 7 shows the TPH of adsorbed ben-
zaldehyde in 5% D2. The more highly deuterated toluene
and benzene products and undeuterated benzene signals
are not shown. During TPH of benzaldehyde, two toluene
species desorb in overlapping peaks. A more highly deuter-
ated toluene (masses 93 and 94) containing 0–6 D atoms
desorbed with a maximum at 555 K, and a less deuterated
toluene (masses 91 and 92) containing 0–2 D atoms des-
orbed with a maximum at 605 K.
Temperature-programmed hydrogenation in 5% D2 after
2
0-min benzaldehyde PCO was also performed. The same
initial coverage of benzaldehyde was used for the adsorbed
benzaldehyde spectra and for the PCO of benzaldehyde.
As with the adsorbed benzaldehyde spectra, two toluene
peaks formed, but the amount in each peak was signifi-
cantly different. After PCO, less of both toluene species
desorbed, as expected because less benzaldehyde was on
the catalyst surface after PCO. More of the less deuterated
toluene (higher-temperature toluene peak) desorbed than
the more deuterated toluene (lower-temperature toluene
peak).
Photocatalytic Oxidation of Toluene on Pt/TiO2
The PCO of toluene was similar on Pt/TiO2 to unpla-
tinized TiO2, but Pt increased the PCO rate. Upon illumi-
nation, O2 was immediately consumed, and a small amount
of CO2 was produced (Fig. 8). Table 1 shows the amount
of O2 consumed and CO2 produced in 20 min. No other
gas-phase products were detected. The maximum O2 con-
2
FIG. 6. Temperature-programmed hydrogenation spectra in 5% D2/
He. After 20-min PCO of toluene, which was adsorbed to saturation cov-
erage on TiO2 at 273 K, an interrupted TPD to 338 K removed unreacted
toluene prior to TPH.
sumption rate on Pt/TiO was approximately twice and the

2
8
BLOUNT AND FALCONER
cyclohexane (MCH) formed in a large peak with a maxi-
mum at 410 K and a smaller peak at 575 K. Toluene and ben-
zene formed in two high-temperature peaks, and both CO2
and CO formed at high temperature. Additionally, ethy-
lene and methane (not shown for clarity) formed during
TPH; neither of these species formed during TPH after
PCO on TiO2. Desorption amounts are given in Table 4.
Temperature-programmed hydrogenation on Pt/TiO2 was
effective in hydrogenating 95% of the strongly bound in-
termediates off the surface, so the amount formed during
the subsequent TPO was small.
Photocatalytic Oxidation of Benzaldehyde on Pt/TiO2
The PCO of benzaldehyde on Pt/TiO2 is similar to PCO
on TiO2 (9). Oxygen was immediately consumed (Fig. 10),
but the O2 uptake decayed more rapidly than the O2 uptake
FIG. 8. Carbon dioxide formation during 20-min PCO of toluene during toluene PCO. Carbon dioxide formed immediately
1
3
(
C6H5- CH3) adsorbed to saturation coverage on Pt/TiO2 at 253 K.
during PCO of benzaldehyde on Pt/TiO and then quickly
2
decayed, and a small amount of benzene was detected dur-
maximum CO2 production rate on Pt/TiO2 was approxi- ing the initial stages (Fig. 10). Benzene production (total
mately 1.5 times the respective rates on unplatinized TiO2. of 1 �mol/g catalyst) was negligible after about 8 min. No
1
3
When C6H5- CH3 was used during PCO on Pt/TiO2, benzene was detected during toluene PCO on Pt/TiO or
2
12
CO2 was produced from both ring carbon ( CO2) and the benzaldehyde PCO on TiO (9). Oxygen uptake and CO2
2
1
3
methyl carbon ( CO2), as shown in Fig. 8. At the beginning formation amounts are given in Table 1.
12
of the PCO, CO2 was preferentially produced, and the
12
13
Temperature-Programmed Hydrogenation
of Benzaldehyde on Pt/TiO2
CO2/ CO2 ratio was 7.5 at 5 min. After 12 min, however,
the 12CO2/ CO2 ratio was 3.5, indicating that the methyl C
13
12
13
was preferentially oxidized since the CO2/ CO2 ratio was
less than the stoichiometric ratio of 6.
During TPH of adsorbed benzaldehyde at approximately
5
5% coverage, benzaldehyde hydrogenated to form mainly
MCH in a peak with a maximum at � 450 K. Small amounts
of CO, toluene, benzene, and CO2 also formed during TPH.
The amounts desorbed are given in Table 5.
Temperature-Programmed Hydrogenation
of Toluene on Pt/TiO2
The TPH spectra following 20-min PCO of toluene on
The TPH spectra following 20-min PCO of benzalde-
Pt/TiO2 and an interrupted TPD are shown in Fig. 9. Methyl- hyde at approximately 30% coverage is shown in Fig. 12.
FIG. 9. Temperature-programmed hydrogenation after interrupted
TPD and 20-min PCO of toluene, which was adsorbed to saturation cov-
erage at 253 K on Pt/TiO2.
FIG. 10. Oxygen consumption and CO2 and benzene formation dur-
ing20-min PCO ofbenzaldehyde, which wasadsorbed to 60% ofsaturation
coverage at 300 K on Pt/TiO2.

PHOTOCATALYTIC OXIDATION OF TOLUENE ON TiO2
29
DISCUSSION
Multiple Forms of Benzaldehyde Adsorption
and Different Reactivities
Adsorbed benzaldehyde hydrogenated mainly to MCH,
which formed in an asymmetrical peak during TPH on
Pt/TiO2. After only 30 s of PCO, some of the benzalde-
hyde reacted to surface species that hydrogenate to ben-
zene and toluene at high temperatures. Figure 11 shows the
MCH spectra during TPH of adsorbed benzaldehyde and
after 30-s, 2-min, and 20-min PCO on Pt/TiO2. Note that the
benzaldehyde coverage differs for the MCH spectra; ben-
zaldehyde was adsorbed to 30% of saturation coverage for
TPH of adsorbed benzaldehyde and after 20-min PCO, and
benzaldehyde was adsorbed to 10% of saturation coverage
for TPH after 30-s and 2-min PCO. The appearance of two
FIG. 11. Methylcyclohexane fomation during TPH of adsorbed ben- distinct MCH peaks after only 30 s of PCO indicates that
zaldehyde, which was adsorbed to 10–30% coverage at 300 K, and after
benzaldehyde adsorbs in more than one form or on more
3
0-s, 2-min, and 20-min benzaldehyde PCO.
than one type of site. The asymmetrical nature of the MCH
peak observed during TPH of adsorbed benzaldehyde is
consistent with a large MCH peak at 435 K overlapping a
smaller MCH peak at 490 K.
Although the main hydrogenation product during
TPH of adsorbed benzaldehyde was MCH, only a small
amount of MCH desorbed during TPH following ben-
zaldehyde PCO. In addition, a significant amount of
toluene and benzene formed at high temperatures. Carbon
dioxide formed throughout the TPH, and CO was the
dominant product and formed in two high-temperature
peaks. During TPH following a 10-s PCO of benzaldehyde,
methylcyclohexane desorbed in three overlapping peaks
with apparent maxima at 420, 480, and 520 K (Fig. 11). Two
high-temperature toluene peaks and two benzene peaks
were also seen.
Nagao and Suda (37) concluded that chlorobenzene ad-
sorbs both parallel to the TiO2 surface through a �-bond
interaction and perpendicular to the surface through Cl in-
teraction with surface hydroxyls. Benzaldehyde may also
adsorb with its ring both parallel to the surface and perpen-
dicular to the surface through its aldehyde group. Whereas
adsorbed benzaldehyde hydrogenated mainly to MCH on
Pt/TiO2, the main hydrogenation productswere toluene and
benzene on TiO2. Two overlapping toluene peaks formed
during TPH of benzaldehyde on TiO2 (Fig. 4), and two dis-
tinct toluene peaks formed during TPH in 5% D2 of ben-
zaldehyde on TiO2 (Fig. 7). The formation of two toluene
peaks indicates that benzaldehyde adsorbs in two forms or
on two types of reaction sites on TiO2 catalysts. The lower
temperature toluene peak was more highly deuterated than
the higher temperature peak. Parallel adsorption of ben-
zaldehyde would provide easier access for exchange and
addition of deuterium, resulting in the formation of the
more deuterated toluene observed during TPH in 5% D2.
Likewise, deuteration of benzaldehyde adsorbed perpen-
dicular to the surface would be hindered so that this type
of benzaldehyde would form the less deuterated toluene.
On Pt/TiO2, the MCH peaks for both forms of ben-
zaldehyde decreased significantly with PCO; thus, both
forms of adsorbed benzaldehyde are reactive. The lower-
temperature MCH peak decreased more quickly than
the higher-temperature peak, and after 20 min of PCO
the lower-temperature peak appeared as a shoulder to the
higher-temperature peak (Fig. 11). Therefore, the form of
adsorbed benzaldehyde that hydrogenated to the lower-
temperature MCH is more reactive than the benzalde-
hyde that formed the higher-temperature MCH peak. The
FIG. 12. Temperature-programmed hydrogenation after 20-min PCO
of benzaldehyde, which was adsorbed to 30% of saturation coverage at
3
00 K on Pt/TiO2.

3
0
BLOUNT AND FALCONER
different forms of benzaldehyde on TiO2 also react at dif- ble for catalyst deactivation. On TiO2, the toluene that
ferent rates. The more deuterated toluene decreased faster formed during TPH after toluene PCO (Fig. 3) is similar
than the less deuterated toluene with time during benzalde- to the lower-temperature toluene peak that formed during
hyde PCO. The benzaldehyde that hydrogenated to the TPH of adsorbed benzaldehyde. Likewise, MCH forma-
more deuterated toluene oxidized faster.
tion on Pt/TiO2 is similar for TPH of adsorbed benzalde-
During benzaldehyde PCO, CO2 formation almost im- hyde (Fig. 11) and TPH after toluene PCO (Fig. 9). The
mediately reached a maximum and then quickly fell high-temperature hydrogenation products formed during
(
Fig. 10). Thus, a large amount of CO2 formed as the more toluene PCO on Pt/TiO2 (Fig. 9) did not form during TPH of
reactive benzaldehyde was oxidized, and the CO2 forma- adsorbed benzaldehyde, but similar high-temperature hy-
tion rate decreased as the more reactive benzaldehyde was drogenation products form during TPH after benzaldehyde
consumed and oxidation of the less-reactive benzaldehyde PCO (Fig. 12). The surface species hydrogenated during the
became significant. As seen in Fig. 11 after 30 s of PCO, TPH after benzaldehyde PCO on TiO2 (Fig. 5) and Pt/TiO2
slightly more of the lower-temperature MCH, which corre- (Fig. 12) appear to be the same surface species on the cata-
sponds to the more reactive benzaldehyde, formed than the lyst surface after toluene PCO (Figs. 3 and 9, respectively).
higher-temperature MCH, which corresponds to the less Thus, toluene reacts to form benzaldehyde, which then re-
reactive benzaldehyde. After 20 min of PCO, most of the acts to form less reactive, strongly bound surface species.
more reactive benzaldehyde had oxidized, but a significant
Because toluene adsorbs parallel to the TiO2 surface,
amount of the less reactive benzaldehyde remained on the benzaldehyde that is adsorbed parallel to the catalyst sur-
surface and was hydrogenated to the higher-temperature face is expected to form from toluene oxidation. The mass
MCH. Thus, after 20 min, most of the CO2 formed was spectra and desorption temperatures of the toluene that
from oxidation of the less reactive benzaldehyde, and the formed during TPH in 5% D2 after toluene PCO (Fig. 6)
CO2 formation rate was much lower than at the beginning are comparable to those of the more deuterated toluene
of the PCO.
that formed at lower temperatures during TPH in 5% D2
of adsorbed benzaldehyde (Fig. 7). Thus, toluene appears to
react predominantly to form benzaldehyde that is adsorbed
Intermediates in Toluene PCO
Toluene quickly reacted to form some intermediate, as parallel to the catalyst surface, and little of the toluene
shown by the immediate O2 uptake (Fig. 1 ). This inter- forms benzaldehyde that is adsorbed perpendicular to the
mediate reacted more slowly to form CO2 and H2O. Much catalyst surface.
less CO2 formed than O2 consumed, and since CO2 adsorp-
Larson and Falconer (9) concluded that benzaldehyde
tion coverage on TiO2 is low (38), the excess O2 was incor- was not the main intermediate formed during toluene PCO.
porated into surface organic species. The PCO of labeled They found that benzaldehyde reacted to CO at a rate
2
13
toluene (C6H5- CH3) showed that the ring carbon was oxi- 10timesfaster than toluene reacted to CO . However, much
2
dized in the initial stages of toluene PCO. At slightly longer more benzaldehyde was adsorbed on TiO than toluene
2
reaction times, methyl carbon oxidized preferentially.
in their study, and benzaldehyde PCO was performed at
The TPH spectra following toluene PCO provide infor- room temperature while toluene PCO was performed at
mation about the adsorbed intermediates (Figs. 3 and 9). 273 K. In addition, toluene must first react to benzaldehyde;
Benzene and toluene are weakly adsorbed on TiO2 (9) and thus, benzaldehyde is expected to react faster than toluene
Pt/TiO2, so their formation during TPH and toluene PCO during PCO. They observed that significant amounts of
is reaction limited. As expected for a fully hydrogenated unreacted benzaldehyde desorbed during TPO after ben-
species, MCH does not significantly adsorb on TiO2 at room zaldehyde PCO on TiO , but only a small amount of ben-
2
temperature; it is much more weakly bound than toluene, zaldehyde desorbed during TPO after toluene PCO. Our
and its appearance during TPH is also reaction limited. The TPH experiments show that benzaldehyde quickly oxi-
formation of MCH and toluene during TPH suggests that at dized to less reactive species. Benzaldehyde formed dur-
least one intermediate has a methyl group (or an oxidized ing toluene PCO would also quickly react to less reactive
methyl group) attached to an aromatic ring. Approximately species, and little benzaldehyde would be on the surface
6
0% of the strongly bound surface species on TiO2 and 75% during toluene PCO. Thus, the desorption of a small amount
on Pt/TiO2 were hydrogenated to toluene and MCH during of benzaldehyde during TPO after toluene PCO observed
TPH following 20-min PCO of toluene, so most of the ad- by Larson and Falconer (9) is consistent with benzaldehyde
sorbed intermediates still had an intact aromatic ring with a as an intermediate in toluene PCO.
methyl or oxidized methyl group attached. The complexity
of the TPH spectra on TiO2 and Pt/TiO2 indicates that more zyl alcohol was dramatically different from toluene PCO.
than one intermediate may be on the surface.
Benzyl alcohol reacted to form benzaldehyde and CO , and
Benzaldehyde appears to be an intermediate in toluene more benzaldehyde than CO formed. In addition, CO for-
Larson and Falconer (9) observed that the PCO of ben-
2
2
2
PCO, but not the less reactive intermediate responsi- mation was delayed and did not reach a maximum during

PHOTOCATALYTIC OXIDATION OF TOLUENE ON TiO2
31
6
00 s of illumination. Thus, because gas-phase benzalde- D’Hennezel et al. (14) observed small amounts of hydrox-
hyde was not observed during toluene PCO, and because ybenzoic acid, hydroxybenzyl alcohol, and hydroxyben-
CO2 formation was higher at longer reaction times, benzyl zaldehyde following extraction of surface species on TiO2
alcohol does not appear to be an intermediate in toluene after PCO of a continuous flow of toluene. During toluene
PCO. The TPH spectra of benzyl alcohol had similarities PCO on several cation-exchanged zeolites, aromatic con-
to the TPH spectra following toluene PCO (Fig. 3) and densation products formed (40). The amount of these con-
the TPH spectra of adsorbed benzaldehyde (Fig. 4). Ben- densation products was dependent on the acidity of the
zaldehyde probably hydrogenated to benzyl alcohol, which zeolite, and only trace amounts were observed for the less
is strongly adsorbed and hydrogenated further to form acidic zeolites. No further analysis to determine the reac-
toluene.
tivity of these species during PCO was performed, but an
Larson and Falconer (9) also studied the PCO and TPO of intermediate similar to the above species observed during
cresol since cresols form during PCO of toluene in aqueous extraction may be the less reactive intermediate formed
solution. Fujihira et al. (16) proposed that cresols form from during toluene PCO.
hydroxyl attack on the ring. Hydroxyl radicals are thought
to be the reactive species during PCO in the aqueous phase
39), but their role in gas-phase PCO is not well under-
Effect of Pt on Photocatalytic Oxidation
(
The addition of Pt increased the PCO rate for both
toluene and benzaldehyde PCO. Table 6 shows the max-
imum areal CO2 formation and O2 consumption rates
on TiO2 and Pt/TiO2. The addition of Pt approximately
doubled the maximum areal O2 consumption rates for both
toluene and benzaldehyde, but the CO2 formation rates
were only � 1.5 times higher. Thus, Pt increased the initial
PCO rate of toluene and benzaldehyde to less reactive in-
termediates, and it also increased the rate of oxidation of
the intermediates to CO2, but not as much.
stood. No reaction was detected for adsorbed cresol upon
illumination on TiO2, and subsequent TPO was the same as
observed without PCO (9). During TPO of cresol, approx-
imately half of the cresol desorbed intact, and the catalyst
turned black. The catalyst color returned to white after ex-
tended heating in O2 at 723 K. Because neither cresol nor a
color change was observed during TPO after toluene PCO,
cresol does not appear to be an intermediate in toluene
PCO.
Benzoic acid has been proposed as the intermediate re-
sponsible for catalyst deactivation during PCO of a con-
tinuous flow of toluene on TiO2 (12–14). The authors pro-
posed that benzaldehyde or benzyl alcohol oxidizes to form
benzoic acid (12–14). Traces of benzoic acid were detected
after toluene PCO using extraction (12–14); however, no
mass balance or further analysis of the surface after ex-
traction to confirm that all of the surface species had been
removed and identified was reported for any of the stud-
ies. Mendez-Roman and Cardona-Martinez (12) also ob-
served benzoic acid on the surface following toluene PCO
using FTIR. Augugliaro et al. (13) observed that significant
amounts of CO2 and benzene formed during PCO of ben-
zoic acid on anatase. These results suggest that benzoic acid
is reactive on TiO2 and is not the strongly bound interme-
diate responsible for catalyst deactivation. Because the low
volatility of benzoic acid made evaporation and calibration
difficult, we did not perform benzoicacid PCO. DuringTPH
of adsorbed benzoic acid, a large amount of benzaldehyde
desorbed along with toluene and benzene at high temper-
atures. Only small amounts of benzaldehyde desorbed af-
ter toluene PCO on TiO2. Thus, though small amounts of
benzoic acid may be on the surface during toluene PCO,
benzoic acid does not appear to be the less reactive inter-
mediate responsible for catalyst deactivation.
Falconer and Magrini-Bair (41) observed that Pt slowed
TiO2 deactivation that resulted during acetaldehyde PCO
at high temperatures. They suggested that oxygen spillover
from the Pt particles onto the TiO2 surface oxidized the
acetaldehyde reaction products and slowed deactivation.
Similarly, oxygen spillover from Pt to TiO2 may be the rea-
son the toluene and benzaldehyde PCO rates are higher on
Pt/TiO2 than on TiO2.
In addition to CO2 formation, a small amount of benzene
formed during benzaldehyde PCO on Pt/TiO2 (Fig. 10).
Larson and Falconer (9) did not observe benzene forma-
tion during benzaldehyde PCO on TiO2. The benzaldehyde
PCO rate is higher on Pt/TiO2 than on TiO2, and only a
small amount of benzene formed during PCO. Thus, ben-
zene may have formed during benzaldehyde PCO on TiO2
without being detected.
TABLE 6
Areal Rates of CO2 Formation and O2 Consumption
4
Maximum rates � 10 (�mol/g catalyst/s)/(�mol/g)
TiO2
Pt/TiO2
Adsorbed
aromatic
Jacoby et al. (10) extracted adsorbed intermediates fol-
lowing PCO of a continuous flow of benzene on TiO2.
Their analysis identified phenol, hydroquinone and/or ben-
zoquinone, and malonic acid as possible intermediates.
O2 uptake CO2 formation O2 uptake CO2 formation
Toluene
Benzaldehyde
19
23
0.5
6
35
47
0.65
9

3
2
BLOUNT AND FALCONER
Hydrogenation Reactions on TiO2
Hydrogenation Reactions on Pt/TiO2 and Spillover
during TPH
Though TiO2 is not a typical hydrogenation catalyst,
strongly bound surface species were effectively hydro-
For a Pt/TiO2 catalyst, spillover (49) from the Pt par-
genated during TPH. Little of the strongly bound surface ticle to the TiO2 surface is important in temperature-
species desorbed or decomposed during TPD, and thus the programmed experiments. During TPH, strongly bound
information provided by TPD was severely limited. During surface species hydrogenate to more weakly bound species
TPH on TiO2, however, 80% of the adsorbed intermediates that desorb more readily. Because of spillover of H onto the
were removed from the surface.
TiO2 surface, more complete hydrogenation was achieved
Most studies on TiO2 in H2 flow found that TiO2 is a on Pt/TiO2, and hydrogenation occurred at lower tempera-
poor hydrogenation catalyst (42 – 48). Halliday et al. (42) tures.
reported that 1,3-butadiene was hydrogenated to butene
Ionnides and Verykios (50) explored the interaction of
and some unidentified, strongly bound species; no butane benzene and toluene with Rh dispersed on TiO2 using
was observed. By replacing H2 with D2 in the flow gas, they TPD and TPH. Their TPD experiments indicated that
concluded that self-hydrogenation of 1,3-butadiene formed the aromatics were almost exclusively adsorbed on the
the butene. In experiments with propadiene in H2 on TiO2 support. When Rh/TiO2 was exposed to H2 at room tem-
(
43), they observed that at temperatures below 400 K, iso- perature, adsorbed benzene readily hydrogenated to cy-
merization of the propadiene was the main reaction. At clohexane, which desorbed at low temperature since it
higher temperatures, unidentified strongly bound species was weakly adsorbed on the surface. On supports with-
formed as well as a small amount of propene. They con- out Rh, the aromatic was not hydrogenated in H2 flow.
cluded that TiO2 is not an efficient catalyst for hydrogena- Thus, Rh provided a site for adsorption and dissociation
tion ofdienesand alkynesbecause theyreact mainlyto form of H2, and this spillover hydrogen hydrogenated benzene
strongly bound species that remain on the surface. Propane adsorbed on the support. Toluene exhibited similar behav-
and isobutane formed propene and isobutene, respectively, ior in our experiments. Platinum resulted in complete hy-
when flowed in H2 over TiO2 at 750 K (43). About 45% drogenation of strongly adsorbed species to MCH at tem-
of the isobutane formed strongly bound surface species peratures lower than that at which hydrogenation on TiO2
that did not desorb. Cyclopentane in H2 flow over TiO2 re- occurred.
sulted in a mixture containing 25% cyclopentane and 25%
cyclopentadiene.
Vannice and Poondi (45) examined the hydrogenation of
a continuous flow of benzaldehyde over reduced Pt/TiO2
Vannice and Poondi (45) saw no hydrogenation of ben- between 333 and 493 K. Benzyl alcohol was the pri-
zaldehyde on TiO2 (Degussa P25) from 473–723 K. Haffad mary product during benzaldehyde hydrogenation, but the
et al. (46), however, observed benzaldehyde hydrogenation amount of toluene and benzene formed increased with
on TiO2 (formed from ex-isopropylate) in both pure He and higher conversions. Neither cyclohexane nor MCH was
pure H2 flow at 573 K. In pure He flow, benzaldehyde con- detected. In our experiments, in which a monolayer
verted to toluene and a small amount of benzene. They sug- of benzaldehyde was on the surface with no gas-phase
gested that the benzaldehyde reacts with surface hydrox- benzaldehyde present, no benzyl alcohol was observed,
yls to form toluene, which desorbs, and a surface benzoate and a significant amount of MCH formed. Vannice and
species. A small amount of toluene evolved even after an Poondi (45) proposed that benzaldehyde hydrogenated
extended reaction period, suggesting the regeneration of to toluene by the pathway PhCHO + H2 → PhCH2OH +
some surface hydroxyl species and the disappearance of H2 → PhCH3 + H2O. Toluene is weakly adsorbed on
surface benzoate species. Haffad et al. (46) suggested that Pt/TiO2, and during continuous-flow hydrogenation of ben-
benzaldehyde hydrogenation on TiO2 in H2 follows the re- zaldehyde, toluene may be displaced from the surface by
action PhCHO → PhCH2OH → PhCH3 + H2O, where Ph gas-phase benzaldehyde before the toluene can hydro-
is an aromatic ring.
genate to MCH. During TPH of a monolayer of ben-
We also observed formation of a small amount of toluene zaldehyde, toluene was not displaced. In addition, Vannice
and benzene during TPD at approximately 573 K during and Poondi (45) studied benzaldehyde hydrogenation over
TPD of adsorbed benzaldehyde. In their reaction scheme, Pt/TiO2 catalysts reduced in H2 flow at 473 and 773 K, and
an adsorbed alkoxy forms toluene by hydrogen transfer and they found that the Pt/TiO2 catalyst reduced at 773 K was
leaves an adsorbed benzoate on the surface, which decom- more active for benzaldehyde hydrogenation. In our exper-
poses to benzene and CO2. In our study, 3 times as much iments, the Pt/TiO2 catalyst wasoxidized at the start ofTPH,
CO2 as benzene desorbed during TPD, so although some of and the catalyst turned a dark blue color characteristic of
the CO2 may be the result of benzoate decomposition, a sig- reduced TiO2 by the end of the TPH. Differences in the
nificant portion of CO2 must result from the decomposition reduction state of the Pt/TiO2 catalysts may also result in
of another surface species.
the differences in benzaldehyde hydrogenation observed.

PHOTOCATALYTIC OXIDATION OF TOLUENE ON TiO2
33
CONCLUSIONS
15. Fujihira, M., Satoh, Y., and Osa, T., J. Electroanal. Chem. 126, 277
1981).
(
1
1
6. Fujihira, M., Satoh, Y., and Osa, T., Nature 293, 206 (1981).
7. Lichtin, N. N., and Sadeghi, M., J. Photochem. Photobiol. A 113, 81
(
Temperature-programmed hydrogenation was effective
at hydrogenating strongly bound intermediates formed dur-
ing photocatalytic oxidation (PCO) to more weakly bound
species that desorbed during TPH. These strongly bound
intermediates did not desorb to a significant extent during
TPD. The intermediates formed during toluene PCO have
a ring structure with a methyl (or oxidized methyl) group
attached. Platinization of TiO2 increased the toluene PCO
rate. Both the ring carbon and the methyl carbon of toluene
oxidized during PCO.
1998).
8. Gratson, D. A., Nimlos, M. R., and Wolfrum, E. J., in ”Proc. of the
8th Annual Meeting of the Air and Waste Management Assoc.,” June
1
8
18–23, 1995, San Antonio, TX.
1
2
2
9. Fu, X., Zeltner, W. A., and Anderson, M. A., Appl. Catal. B 6, 209
(
1995).
0. Hwang, S.-J., Petucci, C., and Raftery, D., J. Am. Chem. Soc. 120, 4388
1998).
1. Driessen, M. D., and Grassian, V. H., J. Phys. Chem. 102, 1418
(1998).
(
2
2
2. Fan, J., and Yates, J. T., J. Am. Chem. Soc. 118, 4686 (1996).
3. Fu, X., Clark, L. A., Zeltner, W. A., and Anderson, M. A., J. Pho-
tochem. Photobiol. A 97, 181 (1996).
Benzaldehyde appears to adsorb in two forms on TiO2,
and these two forms have different reactivities during PCO.
Similarities in TPH spectra after benzaldehyde and toluene
PCO on both TiO2 and Pt/TiO2 indicate that toluene mayre-
2
2
4. Falconer, J. L., and Magrini-Bair, K., J. Catal. 179, 171 (1998).
5. Izumi, I., Fan, F. R. F., and Bard, A. J., J. Phys. Chem. 85, 218 (1981).
act through benzaldehyde, which then reacts to form some 26. St. John, M. R., Furgala, A. J., and Sammells, A. F., J. Phys. Chem. 87,
8
01 (1983).
strongly bound, less reactive intermediate, and this less re-
active intermediate does not appear to be benzoic acid.
When toluene is oxidized, only one type of adsorbed ben-
zaldehyde forms.
2
7. Takahama, K., Sato, T., Yokoyama, M., and Hirao, S., Nippon Kagaku
Kaishi 7, 613 (1994).
8. Obee, T. N., Environ. Sci. Technol. 30, 3578 (1996).
9. Blanco, J., Avila, P., Bahamonde, A., Alvarez, E., Sanchez, B., and
Romero, M., Catal. Today 29, 437 (1996).
2
2
3
3
3
3
3
3
0. Obee, T. N., and Brown, R. T., Environ. Sci. Technol. 29, 1223
ACKNOWLEDGMENTS
(
1995).
1. Avila, P., Bahamonde, A., Blanco, J., Sanchez, B., Cardona, A. I., and
Romero, M., Appl. Catal. B 17, 75 (1998).
We gratefully acknowledge support from the National Renewable
Energy Laboratory (NREL Contract XCK-5-143118-05) and the National
Science Foundation, Grant CTS-9714403. We also thank Dr. Kim Magrini-
Bair at NREL for assistance with the GC/MS analysis and for preparing
the Pt/TiO2 catalyst.
2. Martra, G., Coluccia, S., Marchese, L., Augugliaro, V., Loddo, V.,
Palmisano, L., and Schiavello, M., Catal. Today 53, 695 (1999).
3. Nimlos, M. R., Wolfrum, E. J., Brewer, M. L., Fennell, J. A., and
Bintner, G., Environ. Sci. Technol. 30, 3102 (1996).
4. Haerudin, H., Bertel, S., and Kramer, R., Faraday Trans. 94, 1481
(1998).
REFERENCES
5. Matsuda, S., and Kato, A., Appl. Catal. 8, 149 (1983).
1
2
3
. Walker, A., Formenti, M., Meraiudeau, P., and Teichner, S. J.,
J. Catal. 50, 237 (1977).
. Ait-Ichou, I., Formenti, M., Pommier, B., and Teichner, S. J.,
J. Catal. 91, 293 (1985).
. Larson, S. A., Widegren, J. A., and Falconer, J. L., J. Catal. 157, 611 40. Panov, A. G., Larsen, R. G., Totah, N. I., Larsen, S. C., and Grassian,
36. Luo, S. C., and Falconer, J. L., Catal. Lett. 57, 89 (1999).
37. Nagao, M., and Suda, Y., Langmuir 5, 42 (1989).
38. Luo, S., and Falconer, J. L., J. Catal. 185, 393 (1999).
39. Turchi, C. S., and Ollis, D. F., J. Catal. 122, 178 (1990).
(
1995).
V., J. Phys. Chem. B 104, 5706 (2000).
4
5
. Blake, N. R., and Griffin, G. L., J. Phys. Chem. 92, 5697 (1988).
41. Falconer, J. L., and Magrini-Bair, K., J. Catal. 179, 171 (1998).
. Nimlos, J., Jacoby, W. A., Blake, D. M., and Milne, T. A., Environ. 42. Halliday, M. M., Kemball, C., and Leach, H. F., J. Chem. Soc., Faraday
Sci. Technol. 27, 732 (1993). Trans. 1 70, 1743 (1974).
. Sopyan, I., Murasawa, S., Hashimoto, K., and Fujishima, A., Chem. 43. Halliday, M. M., Kemball, C., Leach, H. F., and Scurrell, M. S., in
6
Lett. 4, 723 (1994).
“Proceedings, 6th International Congress on Catalysis, London. 1976”
(G. C. Bond, P. B. Wells, and F. C. Tompkins, Eds.). The Chemical
Society, London, 1977.
7
8
9
. Peral, J., and Ollis, D. F., J. Catal. 136, 554 (1992).
. Ibusuki, T., and Takeughi, K., Atmos. Environ. 20, 1711 (1986).
. Larson, S. A., and Falconer, J. L., Catal. Lett. 44, 57 (1997).
44. Kemball, C., Ann. N.Y. Acad. Sci. 213, 90 (1973).
45. Vannice, M. A., and Poondi, D., J. Catal. 169, 166 (1997).
1
0. Jacoby, W. A., Blake, D. M., Fennell, J. A., Boulter, J. E., Vargo,
L. M., George, M. C., and Dolberg, S. K., J. Air Waste Manage. 46. Haffad, D., Kameswari, U., Bettahar, M. M., Chambellan, A., and
Assoc. 46, 891 (1996).
Lavalley, J. C., J. Catal. 172, 85 (1997).
1
1
1. Sitkiewitz, S., and Heller, A., New J. Chem. 20, 233 (1996).
2. Mendez-Roman, R., and Cardona-Martinez, N., Catal. Today 40, 353
47. Harrison, D. L., Nicholls, D., and Steiner, H., J. Catal. 7, 359
(1967).
(
1998).
48. Dixon, G. M., Nicholls, D., and Steiner, H., in “Proceedings, 3rd
International Congress on Catalysis, Amsterdam, 1964.” Wiley, New
Work, 1965.
49. Conner, W. C., Jr., and Falconer, J. L., Chem. Rev. 95, 759 (1995).
50. Ionnides, T., and Verykios, X. E., J. Catal. 143, 175 (1993).
1
3. Augugliaro, V., Coluccia, S., Loddo, V., Marchese, L., Martra, G.,
Palmisano, L., and Schiavello, M., Appl. Catal. B 20, 15 (1999).
4. d’Hennezel, O., Pichat, P., and Ollis, D. F., J. Photochem. Photobiol.
A 118, 197 (1998).
1