Heterogeneous photocatalytic oxidation of trichloroethylene and toluene mixtures in air: Kinetic promotion and inhibition, time-dependent catalyst activity

Article abstract of DOI:10.1006/jcat.1996.0299

Photocatalyzed degradation of trace level trichloroethylene (TCE) and toluene in air were carried out over near-UV-illuminated titanium dioxide (anatase) powder in a flow reactor using a residence time of about 5-6 ms. Concentration ranges for TCE and toluene were 0-800 mg/m³. TCE photooxidation was very rapid under our experimental conditions, and ～100% conversion was achieved for TCE concentration examined up to 753 mg/m³ as a single air contaminant. Initial photodegradation rates for toluene in humidified air were fitted by a Langmuir-Hinshelwood rate form. Toluene photooxidation sole contaminant rates were of the same order of magnitude as reported previously for m-xylene and acetone. No significant intermediates were detected by GC/FID during TCE or toluene photooxidation reactions. Humidification had significant influence: Toluene photooxidation rate increases with the water concentration up to about 1650 mg/m³ 20% relative humidity) and decreases thereafter. The presence of sufficient TCE (225-753 mg/m³) promoted the toluene photooxidation reaction rate to achieve 90-100% toluene conversion in 5.6 ms. When feed toluene levels measured below ～90 mg/m³. Higher toluene feed levels "quenched" this TCE promotion effect and also depressed TCE conversion very strongly, but the toluene conversion fell just to the toluene-only levels observed in single contaminant experiments. Photooxidation kinetics of TCE and toluene mixtures in air are thus shown to exhibit strong promotion and inhibition behavior vs that expected from the single-species kinetic degradation data. A previously suggested chlorine radical oxidation of TCE is modified to rationalize the TCE enhancement of the toluene photooxidation rate and the corresponding toluene "quench" of TCE destruction. Time-dependent catalyst activation (by TCE) and deactivation (by toluene or toluene oxidation products and, eventually, even by TCE products) were observed. Carboxylate formation and carboxylic acid accumulation postulated by previous investigators could be a major cause of such catalyst deactivation.

Full text of DOI:10.1006/jcat.1996.0299

JOURNAL OF CATALYSIS 163, 1–11 (1996)
ARTICLE NO. 0299
Heterogeneous Photocatalytic Oxidation of Trichloroethylene and
Toluene Mixtures in Air: Kinetic Promotion and Inhibition,
Time-Dependent Catalyst Activity
1
2
Yang Luo and David F. Ollis
Department of Chemical Engineering, North Carolina State University, Raleigh, North Carolina 27695-7905
Received March 28, 1995; revised March 27, 1996; accepted March 29, 1996
hydes, alcohols, light hydrocarbons, aromatics, and chlo-
Photocatalyzed degradation of trace level trichloroethylene rinated solvents. Photocatalytic oxidation of air contami-
TCE) and toluene in air were carried out over near-UV-illuminated nants over titanium dioxide has been studied by various
(
titanium dioxide (anatase) powder in a flow reactor using a resi-
dence time of about 5–6 ms. Concentration ranges for TCE and
toluene were 0–800 mg/m . TCE photooxidation was very rapid
research groups over the last two decades. Early photo-
catalysis explorations included oxidations of alkanes (1–
3
3
), isopropanol (1–5), methane (6), secondary alcohols (7),
under our experimental conditions, and � 100% conversion was
ammonia photooxidation (8), and the decarboxylation of
acetic and formic acids (9). Studies providing kinetic rate
expressions are those of isobutane partial oxidation (10),
3
achieved for TCE concentration examined up to 753 mg/m as a
single air contaminant. Initial photodegradation rates for toluene in
humidified air were fitted by a Langmuir–Hinshelwood rate form.
Toluene photooxidation sole contaminant rates were of the same
order of magnitude as reported previously for m-xylene and ace-
1
-butanol photooxidation (11a), and conversions of ace-
tone, 1-butanol, formaldehyde, and m-xylene photo oxida-
tone. No significant intermediates were detected by GC/FID dur- tion (11b). All of these oxygenate and hydrocarbon studies
ing TCE or toluene photooxidation reactions. Humidification had showed slow to moderate rates of conversion, correspond-
significant influence: Toluene photooxidation rate increases with ing to small apparent quantum yields, e.g., of 1–2% for ace-
3
the water concentration up to about 1650 mg/m 20% relative hu-
tone (11b).
midity) and decreases thereafter. The presence of sufficient TCE
225–753 mg/m ) promoted the toluene photooxidation reaction
rate to achieve 90–100% toluene conversion in 5.6 ms. When feed
In sharp contrast, trichloroethylene photocatalyzed con-
3
(
version has been found to be near or at 100% even in short
residence time (<100 ms) reactors (see summary below).
The present research was undertaken to explore if and to
what degree the (possible chain) rapid reaction oxidation
mechanism of TCE could enhance the photocatalytic oxi-
dation of an aromatic, toluene, which is an important air
3
toluene levels measured below � 90 mg/m . Higher toluene feed
levels “quenched” this TCE promotion effect and also depressed
TCE conversion very strongly, but the toluene conversion fell just
to the toluene-only levels observed in single contaminant exper-
iments. Photooxidation kinetics of TCE and toluene mixtures in
air are thus shown to exhibit strong promotion and inhibition be- contaminant from both an ozone formation and noncancer-
havior vs that expected from the single-species kinetic degradation risk point of view (12).
data. A previously suggested chlorine radical oxidation of TCE is
Trichloroethylene Studies
modified to rationalize the TCE enhancement of the toluene pho-
tooxidation rate and the corresponding toluene “quench” of TCE
destruction. Time-dependent catalyst activation (by TCE) and de-
activation (bytoluene or toluene oxidation productsand, eventually,
even by TCE products) were observed. Carboxylate formation and
carboxylic acid accumulation postulated by previous investigators
Trichloroethylene has been examined recently in gas-
phase photooxidations by several research groups due to
its unusually high photocatalytic quantum yield. Dibble
and Raupp systematically explored trichloroethylene pho-
tooxidation (13–15) to determine the kinetics of conversion
of trace (0–100 ppm) trichloroethylene in air over illumi-
nated TiO2 using both a fixed-bed reactor and a fluidized
bed reactor. The observed reaction rate was a positive to
zero apparent order in trichloroethylene and oxygen vapor-
phase mole fractions and zero to negative order in water va-
por mole fraction, depending on the concentration of each
species. Trace water presence was found necessary to main-
tain photocatalytic catalyst activity for extended periods of
time, but higher water levels were strongly inhibitory.
could be a major cause of such catalyst deactivation.
�c 1996 Academic
Press, Inc.
1
. INTRODUCTION
Gas–solid photocatalysis has a favorable technical po-
tential to treat a range of air contaminants including alde-
1
Current address: International Paper, Erie Research Center, 1540 E.
Lake Rd. Erie, PA 16533.
2
To whom correspondence should be addressed.
1
0021-9517/96 $18.00
Copyright �c 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

2
LUO AND OLLIS
Nimlos et al. combined mass spectrometry (MS) and sion is less than 10% , which suggests that the external mass
Fourier transform infrared (FTIR) spectroscopy to de- transfer is not controlling the reaction rate.
tect directly the important by-products and intermedi-
In a recent kinetic study of trichloroethylene photooxi-
ates of trichloroethylene photocatalyzed conversions over dation, Jacoby constructed a TiO2 coated annular photore-
TiO2 (16). They observed substantial gas-phase chlorinated actor and quantitatively explored the byproducts and in-
species including phosgene, molecular chlorine, hydrogen termediates of the trichloroethylene photooxidation (19).
chloride, and dichloroacetyl chloride. Their finding of ap- He observed that the trichloroethylene reaction can pro-
preciable levels of both phosgene and molecular chlorine, ceed either through the CHCl2CClO dichloroacetyl chlo-
known in World War I as chemical warfare agents, has ride (DCAC) intermediate to the final products, CO, CO2,
stirred concern over photocatalysis by-products in air con- HCl, COCl2, and trace Cl2, or through direct photooxida-
taining chlorinated olefins.
tion to form COCl2, HCl, and CO2. Up to 85% of the re-
Holden et al. studied photochemical destruction of acted trichloroethylene appears as DCAC in the product
trichloroethylene in air (17). Trichloroethylene and its mixture, indicating the primary importance of the path-
photooxidation reaction products were quantified by gas way through this intermediate. Both trichloroethylene and
chromatography, coupled gas chromatography–mass spec- DCAC reaction rates can be described by a Langmuir–
troscopy, proton nuclear magnetic resonance spectroscopy, Hinshelwood rate function. Apparent quantum yields up to
infrared spectroscopy, and wet chemical analysis pro- 4 (400% ) were observed, suggesting a chain reaction mech-
cedures. Their results showed that the destruction of anism. Jacoby concluded that trichloroethylene destruction
trichloroethylene was complete at slow flow rates (100 was predominantly via a chain reaction that occurs exclu-
ml/min); the major product isolated was dichloroacetyl sivelyon the TiO2 surface. (We reviewand use thisproposed
chloride. At slower flow rates (25 or 10 ml/min), dichloro- chain mechanism in our discussion of (toluene + TCE) re-
acetyl chloride was also destroyed to yield a mixture sults.) For this annular photoreactor, a mass transport lim-
of carbon monoxide, carbon dioxide, chlorine, and phos- ited region (Re < 100) and a surface reaction limited region
gene. Small amounts of trichloroethylene epoxide were (Re > 200) were determined.
also detected. These flow rates correspond to approximate
reactor residence times of 2.77, 11.1, and 27.7 min, respec-
Toluene Studies
tively.
Ibusuki and Takeuchi investigated the photooxidation of
Anderson’s group examined the dependencies of trace 80 ppm toluene in air over UV irradiated TiO at am-
2
trichloroethylene photocatalyzed degradation rate on light bient temperature (20). The main product of the reaction
intensity, feed composition (TCE, O2, H2O), and temper- at 10 min residence time was CO , with a trace (<2 ppm)
2
ature in a bed reactor packed with TiO2 pellets (18). A of benzaldehyde as the only detected gas-phase interme-
monochloroacetate intermediate was identified on the TiO2 diate. Water appears to be required for reaction: No reac-
pellet surface via FTIR. The reaction rate is first order tion products or decrease of toluene concentration were
with respect to the light intensity, when the reactor is oper- detected in a toluene (80 ppm)–dry air system, and only a
�
5
ated with feed mole fractions of 3.7 � 10 TCE, 0.02, O2, slight reactant decrease was noted in a toluene (80 ppm)–
�
3
� 6
2
.3 � 10 H2O, and at a space time of 4.3 � 10 g s/mol; NO (80 ppm)–dry air mixture. These authors observed that
2
space time is defined as the amount of catalyst divided the batch photoreactor CO concentration at 10 min resi-
2
by the inlet TCE molar flowrate. The apparent quantum dence time increased linearly with increasing relative hu-
yield, calculated as the mineralization rate of TCE divided midity over the range 0–60% ; thus, water vapor enhanced
by the incident photon rate, was 4–17% depending upon the photooxidation rate.
space time. The trichloroethylene reaction rate was inde-
Suzuki et al. studied toluene photocatalytic deodoriza-
pendent of trichloroethylene concentration in the ranges tion ofair on TiO -coated honeycomb ceramicmonolith in a
2
of 37–450 ppm, oxygen mole fraction 0.01–0.2, and water recirculation reactor system (21). The odor compounds ex-
vapor mole fraction 0.001–0.028. The ratio (moles of CO2 amined individually included acetaldehyde, isobutyric acid,
produced)/(moles of TCE degraded) increased to 2, corre- toluene, methyl mercaptan, hydrogen sulfide, and trimethy-
sponding to the complete mineralization of trichloroethy- lamine. They reported that these odor compounds disap-
lene to final products CO2 and HCl, when a temperature peared via pseudo-first-order photooxidation kinetics. The
�
� 7
g
of 64 C was used with space time greater than 2.7 � 10
photocatalytic monolith has been examined as an on-board
s/mol. Temperature had no effect on trichloroethylene re- car air purifier in new and used vehicles; its effectiveness has
action rate in the range of 23 to 62 C. These findings suggest been evaluated through both chemical analysis and human
�
that the complete mineralization of TCE can be obtained panel tests (22) and a reactor model developed to include
simply by increasing the temperature of the photoreactor concentration, momentum, and illumination fields (23).
and the space time. They show that the trichloroethylene
Most gas–solid photocatalytic oxidations discussed in the
reaction rate is independent of flowrate when the conver- literature were exclusively limited to the single-component

HETEROGENEOUS PHOTOCATALYTIC OXIDATION
3
TABLE 1
of (i) a TCE chain mechanism, (ii) a new solid phase, and
iii) water influences.
(
Photocatalytic Sensitization by TCE (24)
Compounds
Isooctane (0.115% )
Time for 90% conversion
2. EXPERIMENTAL
The photocatalyzed degradation of trace level tri-
22.5
7
550
300
270
155
720
380
Isooctane + 400 ppm TCE
chloroethylene and toluene mixtures in air was carried out
using near-UV-illuminated titanium dioxide (anatase) pow-
der in a flow reactor, designed previously by Peral and Ollis
to study photocatalytic oxidation kinetics of acetone, 1-
butanol, butyraldehyde, formaldehyde, and m-xylene (25).
The catalyst used was P25 TiO2 (Degussa), characterized
by the manufacturer as having a primary particle diameter
Methylene chloride (0.75% )
Methylene chloride + 669 ppm TCE
Methylene chloride + 0.373% TCE
Methylene chloride + 0.752% TCE
Chloroform (2.3% )
Chloroform + 0.206% TCE
2
of 30 nm, a surface area of 50 m /g, and a crystal struc-
pollutant system; these provide convenient cases for un- ture of primarily anatase. The catalyst particles were spher-
derstanding the kinetics of photocatalytic oxidation. For ical and nonporous, with stated purity of >99.5% TiO2.
commercial photocatalysis application, a contaminated air Stated impurities included: Al2O3 (<3% ), HCl (<0.3% ),
stream will often contain more than one contaminant. SiO2 (<0.2% ), and Fe2O3 (<0.01% ). This catalyst was used
Therefore, multicomponent system studies can help us un- as supplied, without pretreatment. The trichloroethylene
derstand photocatalytic oxidation under more realistic pro- and toluene used to prepare individual gas mixtures were
cess conditions and ultimately provide a stronger photore- of HPLC grade, supplied by Aldrich and Fisher Scientific,
actor design basis.
respectively.
Berman and Dong studied the individual photocatalytic
Gas concentrations were analyzed by gas chromatogra-
oxidation for trichloroethylene, perchloroethylene, vinyl phy (Perkin–Elmer Sigma 1) operating with a flame ion-
chloride, 1,2-dichloroethane, methyl chloroform, hexanes, ization detector (FID). An Alltech C-5000 column with
methylene chloride, carbon tetrachloride, isooctane, AT-1000 Chromosorb W-AW 80/100 packaging was used
methanol, ethanol, acetone, trioxane, benzene, and toluene for trichloroethylene and toluene analysis.
in air in a continuous-flow reactor (24). They first reported
A 100-W black light (UVP) lamp was used to provide
that the individual isooctane, methylene chloride, and vertical near-UV illumination. The incident light intensity
chloroform photocatalytic decomposition rates could be in- for the lamp was determined via ferrioxalate liquid acti-
�
7
2
creased substantially with the addition of trichloroethylene nometry to be 5.0 � 10 E/cm min. All experiments were
�
to the system. Table 1 summarizes their published exper- performed at 22–24 C.
imental results for an air stream containing 3% water and
.115% isooctane or 0.75% methylene chloride or 2.3% uniformly over the surface of the porous fritted glass plate,
In a typical experiment, 0.1 g TiO2 powder was spread
0
�
chloroform at 30 C. They interpreted this rate promotion as providing a 3.2-mm-deep TiO2 powder layer. The 20.11-
�
4
2
caused by trichloroethylene providing Cl� atoms to initiate liter gas reservoir was filled to 1.013 � 10 N/m with UHP
some (unstated) oxidative chain reactions of other com- air. A suitable amount of water and contaminant, each in
pounds, but gave no mechanism, kinetic rate equation, or liquid form, were each injected into the reservoir through
other multicomponent data to rationalize such an increase. a sample port. Following complete vaporization, the reser-
While these authors termed the effect a “photosensitiza- voir was filled with additional UHP air up to a final pressure
5
2
tion,” we prefer and use here a “chlorine promotion” effect, (typically 2.18 � 10 N/m ). Two streams flowing directly
based on the presumed involvement of chlorine redicals.
from the UHP air tank and the air–water vapor contam-
We present below our initial study of both individual inant reservoir were mixed continuously and passed into
and simultaneous photocatalyzed conversion of TCE and the photoreactor. A mass flow controller (Linde FM4574)
toluene. Our individual reactant results show conventional was used to provide the desired mass flow rates for both air
Langmuir–Hinshelwood rates for toluene and the expected and contaminant streams. The resultant contaminant and
high rates (100% conversion) for TCE in a 5-ms residence humidified air mixture was passed through the reactor for a
time downflow powder bed reactor. The simultaneous con- considerable time until the feed and reactor exit gas concen-
version displays considerable surprise, including a chlorine trations were identical. Then the light was turned on, and
promotion of toluene, which can reach 90–100% toluene gas samples were taken every 15 to 30 min. The exit gas
conversion at low toluene concentrations, and conversely concentration was carefully monitored until it reached the
an efficient toluene “quench” of TCE conversion as the near-steady-state value. This photo-steady state was usually
3
toluene level is raised above 120 mg/m . Finally, we explore reached in 3–7 h for different contaminant concentrations.
alternate rationalizations for this novel behavior in terms Between the kinetic runs, the TiO2 catalyst was regenerated

4
LUO AND OLLIS
through UV illumination with dry air passing through
the reactor.
3
. RESULTS
TCE and Toluene Individual Photooxidation
The purpose of this research is to study trichloroethylene
and toluene mixture photooxidation in a powder bed flow
reactor. It is essential that we first look into toluene and
trichloroethylene single-component photooxidation kinet-
ics separately.
As the literature reported, trichloroethylene photo-
degradation rate is very rapid, supposedly due to a Cl�-
initiated chain reaction nature, which results in an unusu-
ally high quantum yield. In the powder bed flow reactor,
we confirmed these high rate reports, finding that � 100%
TCE conversion was achieved for all gas-phase concentra-
3
FIG. 1. Toluene photooxidation (Cw = 1619.4 mg/m ).
3
If our rate form assumption is valid, the ln C0/C/(C0� C)
vs 1/(C0� C) plot should be linear with a negative intercept,
since the apparent adsorption constant K is always positive.
Figure 1 indicates a linear plot of the data; thus the
Langmuir–Hinshelwood kinetic form combined with inte-
gral rate law analysis can be used to describe the toluene
photooxidation rate (as found earlier for m-xylene). The k
and K values calculated from the slope and intercept of the
ln C0/C/(C0� C) vs 1/(C0� C) plot are
tions examined, up to 753 mg/m in air at 20% relative
humidity.
At the same relative humidity, with gas-phase toluene
3
concentrations of 80–550 mg/m , pseudo-steady conver-
sions of 20 to 8% , respectively, were observed in the powder
bed flow reactor.
To analyze the toluene photooxidation data, we con-
sider the plug-flow integral analysis of Peral and Ollis
for the downflow powder bed reactor (25). If we assume
that toluene photooxidation also follows the Langmuir–
Hinshelwood rate form, the gas-phase toluene concentra-
tion can be described as:
k = 3.14 g/liter min
3
K = 0.00463 m /mg.
dC
dZ
kKC
V
= �
,
[1]
Table 2 compares these toluene results with the values
for m-xylene, acetone, and 1-butanol (25) photooxidation
in the powder bed flow reactor. The toluene, m-xylene, and
acetone have similar order magnitude of photodegrada-
tion rate constant and apparent binding constant, while the
alcohol is most reactive.
For toluene concentration below 160 mg/m , the oxida-
tion rate is approximately first order, whereas in the range
of 160–550 mg/m , the oxidation rate is between zero and
1 + KC
where V is the linear gas velocity, C is the gas-phase toluene
concentration, k is the apparent reaction rate constant, K is
the adsorption constant, Z is the vertical position in the
TiO2 catalyst layer (measured from the top). The apparent
reaction rate constant k is a function of light intensity and
thuscatalyst bed depth. Equation [2]can be used to evaluate
the reaction rate constant variation with Z:
3
3
first order. The rate equation is
k(Z) = k0e�
�ꢀZ
,
[2]
3
162Ctol
rate =
,
[5]
where k0 is the rate constant at surface (I0 = 5 � 10 E/cm2
�
7
1 + 0.00463Ctol
min), � is a constant (� = 0.5–1.0), ꢀ is the extinction coeffi-
�
1
cient of TiO2, taken as (10211 cm ). Inserting Eq. [2] into
Eq. [1], we have
TABLE 2
Photooxidation Rate Comparison in Powder Bed Reactor
�
�ꢀZ
dC
dZ
k0e
KC .
V
= �
[3]
Concentration
k
K
1 + KC
3
3
Contaminants
(mg/m )
(g/liter min) (m /mg) Investigator
Integrating Eq. [3] and assuming complete light absorption
in the TiO2 layer gives
Toluene
m-Xylene
Acetone
80–550
1–260
1–260
1–260
3.14
1.30
7.75
0.0046
0.0066
0.0064
0.0011
This study
(2, 25)
(2, 25)
ln C0/C
C0 � C
k0 K
1
1-Butanol
49.2
(2, 25)
=
� K.
[4]
�ꢀV C0 � C

HETEROGENEOUS PHOTOCATALYTIC OXIDATION
5
Toluene Photooxidation in the Presence
of Trichloroethylene
We showed above that toluene photooxidation follows
the Langmuir–Hinshelwood form. Trichloroethylene pho-
tooxidation kinetic data have also to be fitted by such
rate forms, with further modification to include strong
water inhibition (13–20). Conventionalcompetitive adsorp-
tion analysis for toluene and trichloroethylene in a two-
contaminant mixture in air would predict that the photooxi-
dation rate for both compounds should decrease.
We performed three sets of experiments at fixed water
concentration in the feed stream.
3
(
1) Fixed trichloroethylene concentration at 753 mg/m ,
3
varied toluene concentration from 80 to 550 mg/m ;
2) fixed trichloroethylene concentration at 226 mg/m ,
FIG. 2. Toluene reaction rate vs water concentration.
3
(
3
varied toluene concentration from 80 to 550 mg/m ;
(
trichloroethylene concentration from 220 to 760 mg/m .
3
3) fixed toluene concentration at 134 mg/m , varied
where the concentration unit is milligrams per cubic meter
and the rate unit is milligrams per cubic meter per minute.
No intermediates were detected by gas chromatography
3
Experimental results for these mixed reactant feeds are
with a flame ionization detector. Therefore, we believe that plotted in Figs. 3–8. No intermediates were detected by
toluene degraded to form primarily carbon dioxide and GC–FID for our toluene and trichloroethylene mixtures af-
water. This view is consistent with the findings of Ibusuki ter photocatalyzed oxidation. In all three experiment sets,
and Takeuchi, who reported that only a trace benzaldehyde we observed a region where the toluene conversion in-
(
less than 1 ppm) appeared in the gas phase during photo- creased markedly with the addition of trichloroethylene
catalytic conversion of 80 ppm toluene in humidified air; and another where the trichloroethylene conversion de-
carbon dioxide was always the dominant product (20).
creased stronglywith increasingtoluene levels, compared to
the conversions achieved for single component toluene or
trichloroethylene photooxidations, respectively. When the
trichloroethylene/toluene ratio was increased, the toluene
conversion increased dramatically at a certain threshold,
and at the same time, trichloroethylene conversion also in-
creased. The photooxidation rates behaved similarly.
Toluene and trichloroethylene oxidation conversion
Water Influence on Toluene Photooxidation
The influence of water in photooxidation depends upon
the contaminant. Water vapor strongly inhibits isopropanol
(
4, 5, 20), trichloroethylene (4, 5, 13–15, 20), and acetone
oxidation (2); enhances toluene oxidation (1); has no sig-
nificant effect on 1-butanol oxidation (2); and increases
(
Fig. 3) decreases with increasing toluene concentration
3
m-xylene oxidation rate up to 1500 mg/m and decreases
3
at a fixed trichloroethylene level of 753 mg/m . When the
thereafter (2).
3
toluene concentration is below 100 mg/m , both toluene
Figure 2 shows the measured toluene oxidation rate vs
water concentration in the feed air stream. No toluene pho-
todegradation proceeded in the total absence of water in
the toluene–air mixture, as also observed by Ibusuki and
Takeuchi (20). Toluene oxidation rate was enhanced by
3
water concentrations up to 2000–3000 mg/m , which corre-
sponded to about 23–40% relative humidity and was some-
3
what inhibited at 60% rh (6100 mg/m ). For water concen-
3
tration between 0 and 6000 mg/m , the water influence for
toluene oxidation rate can be correlated as
7
10.7CH O
2
rate = 1
. [6]
+ 5.325 � 10�
4
CH O + 1.9241 � 10
� 7
C
2
H2O
2
As with the previous equation [5], the concentration unit is
milligrams per cubic meter and the rate unit is milligrams
per cubic meter per minute.
3
FIG. 3. Toluene, TCE conversions (Ctce = 753.09 mg/m ).

6
LUO AND OLLIS
3
FIG. 4. Toluene, TCE conversions (Ctce = 225.93 mg/m ).
3
FIG. 6. Toluene conversion vs time (Ctol = 534.57 mg/m ).
and trichloroethylene underwent almost complete pho- when the toluene concentration was above 160 mg/m3,
3
todegradation. At toluene concentrationsabove 160mg/m , the toluene conversion dropped sharply to near its single-
trichloroethylene was hardly converted, and only a slight component value, and the trichloroethylene conversion was
toluene conversion increase in the mixture was observed, remarkably low, typically less than 5% .
compared to the toluene single-component photooxi-
dation.
Figure 5 compares toluene oxidation conversions and
rates for three cases: toluene single component, toluene
3
3
A survey of air stripper effluents for remediation with 225.93 mg/m , and 753 mg/m trichloroethylene, re-
of trichloroethylene-contaminated ground water found spectively. Trichloroethylene presence enhances toluene
trichloroethylene concentrations to be frequently around conversion, for all toluene concentrations below about
3
3
1
00–200 mg/m (26), although some sample sites might 200 mg/m , and essentially total toluene conversion is
3
3
have trichloroethylene concentrations up to 6000 mg/m . found at toluene levels below 90 mg/m . Above 200 mg
3
3
Figure 4 shows toluene and trichloroethylene oxidation toluene/m , the presence of 89–753 mg/m trichloroethy-
conversions at fixed trichloroethylene concentration of lene was without influence on toluene conversions.
3
2
26mg/m . Toluene and trichloroethylene conversionsdrop
dramatically when the toluene concentration is higher than TiO Catalyst Activity
2
3
8
8 mg/m , corresponding to a trichloroethylene/toluene
Time-dependent TiO2 catalyst activity was observed
during powder bed flow reactor experiments. In the
trichloroethylene single-component system, TiO2 short-
term activity increased with photooxidation reaction time
mass ratio near 3.
For trichloroethylene concentration varying from 220 to
3
3
7
60 mg/m and toluene concentration below 88 mg/m , both
trichloroethylene and toluene were completely converted;
(
Fig. 6).
3
FIG. 5. Toluene conversions with or without TCE.
FIG. 7. TCE conversion vs time (Ctce = 753.09 mg/m ).

HETEROGENEOUS PHOTOCATALYTIC OXIDATION
7
however, the TiO2 catalyst deactivated considerably and
could no longer be regenerated simply under UV illumi-
nation and dry air flow. This result probably indicates the
eventual accumulation of some adsorbed species, which is
not easily oxidizable.
4
. DISCUSSIONS
TCE and Toluene Reaction Pathways
It is well accepted that when TiO2 particle is illuminated
by wavelength <370 nm, the valence band electrons of the
TiO2 can be excited to the conduction band, creating highly
�
+
reactive electron (e ) and hole (h ) pairs. Those electrons
and holes migrate to the TiO2 solid surface and are trapped
at different sites. The first electron transfer products for
3
FIG. 8. Toluene, TCE conversions vs time (Ctce = 753.09 mg/m ),
3
Ctol = 150.57 mg/m ).
�
�
gas–solid photocatalysis are the oxygen species O2 or O
on the surface. The photogenerated holes may be trapped
by hydroxyl ions or water on the surface to form hydroxyl
radicals (1–5, 2, 13–15, 19, 20, 25–31). The initial reaction
steps for the above mechanisms may be summarized as the
following:
Catalyst deactivation was observed during toluene pho-
tooxidation. For toluene photooxidation with or without
trichloroethylene presence, TiO2 activity decreased with
3
time (Fig. 7 (no TCE), Fig. 8 (113.6 mg/m TCE)). In the
+
�
presence of trichloroethylene, the TiO2 catalyst could be re-
generated via UV illumination with flowing dry air through
the reactor for longer than 3 h. An experiment was de-
signed to compare catalyst activity decline of fresh and
regenerated catalyst in the presence of TCE. Figure 9 shows
toluene conversion for fresh catalyst (solid circle, increas-
ing conversion with time, approaching 100% after 150 min);
TiO2 + h� ↔ h + e
+
�
h + OH → OH�
4+
�
3+
4+
Ti + e → Ti
Ti + O2 → Ti + O .
3+
�
2
A number of homogeneous photochemical oxidations of
catalyst regenerated after 35 h continuous illumination toluene have been carried out in the presence of NO, NO2,
(
open circle, conversion = 85–93% after 270 min); and cat- and air to simulate smog formation conditions. The homo-
alyst regenerated after seven successive use/regeneration geneous reaction is believed to be initiated by OH radi-
cycles (solid diamond). At each intervening time the cata- cal generated via the photodecomposition of nitrous acid.
lyst was illuminated for approximately 5 h. The fresh cata- Two reaction mechanisms have been proposed, based re-
lyst and catalyst regenerated after long-term (35 h) contin- spectively on the addition of OH radical to aromatic rings
uous illumination gave similar toluene conversions. After and the OH radical abstraction of hydrogen from the alkyl
seven successive use and regeneration cycles (diamond). group (6, 32–34). The yield for each homogeneous pathway
was reported by several authors (35–38). The abstraction
pathway yield ranged from 0.08 to 0.16, and that for addi-
tion varied from 0.92 to 0.84.
Ibusuki and Takeuchi (20) studied trace (80 ppm)
toluene-photocatalyzed oxidation in air with and without
8
0ppm NO2 presence. Without NO2 presence, onlyCO2 and
trace benzaldehyde (<1 ppm concentration) were noted as
the products. They reported that the CO2 formation rate in-
creased linearly with rising relative humidity and that trace
benzaldehyde yield decreased with increasing relative hu-
midity. They concluded that the increasing gas-phase H2O
�
concentration probably increased the surface OH concen-
tration and thus surface �OH levels.
The trichloroethylene homogeneous photooxidation re-
action has been studied extensively. Although the exact
reaction mechanism is still under investigation, it seems
agreed that trichloroethylene undergoes a chain reaction
that resultsin a quantum efficiency greater than 1. Sanhueza
FIG. 9. Catalyst activity test for toluene conversion (Ctce = 753.09
mg/m , Ctol = 88.48 mg/m ).
3
3

8
LUO AND OLLIS
et al. (39, 40) proposed a chain reaction mechanism that in-
Jacoby carried out a detailed study of the photocatalyzed
volved Cl� atoms attacking trichloroethylene to initiate the destruction of trichloroethylene in air in a TiO2 surface
oxidation reaction. The Cl� atoms required for the TCE (19). He proposed two reaction pathways for TCE pho-
photooxidation reaction were produced by an initiation re- tooxidation. In pathway A, the reaction proceeds through
action involving trichloroethylene (41).
the DCAC intermediate, and in pathway B, trichloroethy-
The existence of quantum yields greater than 1.0 for the lene is oxidized directly to products. Experimentally, Jacoby
photocatalyzed destruction of TCE leads to the supposi- observed that up to 85% of reacted trichloroethylene was
tion that the photocatalyzed reaction must be (primarily) a converted through the DCAC pathway. He further con-
chain reaction also. The observation of dichloroacetylchlo- cluded by addition of ethane as a “quenching” agent that
ride (CHCl2CCl2(O)) as a major intermediate suggests that trichloroethylene oxidation was taking place exclusively on
a Sanhueza-like chain mechanism may occur on illuminated the TiO2 surface, i.e., no gas-phase chain reaction was oc-
photocatalyst.
The Sanhueza et al. mechanism:
curing.
The chain reaction notion thus appears for two basic rea-
sons:
CCl2CHCl + Cl� → CHCl2CClCl�
(
i) high observed quantum yields, exceeding unity and
CHCl2CClCl� + O2 → CHCl2CCl2OO�
(
ii) high intermediate levels of DCAC, known to be pro-
2CHCl2CCl2OO� → 2CHCl2CCl2O� + O2
rd eu ac ce t di o inn . gas-phase photochemistry via a Cl�-driven chain
CHCl2CCl2O� → CHCl2CCl(O) + Cl�
We discuss below a potential chain reaction for toluene/
TCE mixtures, which could rationalize our observed pro-
motion and inhibition effects.
or a minor pathway may take place
CHCl2CCl2O� → CHCl2� + CCl2.
Catalyst Deactivation
The Cl� radicals required on the TiO2 surface may arise
from the hydroxyl radicals or oxygen atoms produced by
TiO2. The production of oxygen atoms on TiO2 has often
been invoked in the gas–solid heterogeneousphotocatalytic
oxidation of organics (42), and oxygen atoms formed on the
TiO2 surface could react with trichloroethylene.
Photocatalyst deactivation during alcohol oxidation was
first reported by Cunningham and Hodnett (7), who ex-
amined 2-butanol, 2-propanol, and 1-butanol over illumi-
nated TiO2 and ZnO at alcohol concentration ranges 160–
3
2
0000 mg/m . Catalyst deactivation profiles were similar
Similarly the hydroxyl radical is known to be produced on
TiO2 wet surface under UV illumination (31), and hydroxyl
radical reactions with trichloroethylene could release chlo-
rine radicals, a` la the gas-phase reactions of Kleingienst et al.
for different alcohol concentrations. The authors suggested
that the CO2 product formed during reaction absorbed on
the catalyst surface and competed for the catalyst active
sites.
(
41). In either case, for the photocatalyzed trichloroethy-
However, since all photocatalyzed oxidations of carbona-
ceous reactants produce CO2, but only certain reaction sys-
tems exhibit catalyst deactivation phenomena, it appears
more likely that some less general inhibitory products or re-
calcitrant intermediates are formed during reaction and re-
main absorbed on the catalyst surface. These intermediates
are presumably responsible for catalyst deactivation. For
example, Blake and Griffin proposed a 1-butanol photooxi-
dation mechanism in which they speculated that butanoic
acid was gradually formed from adsorbed butyraldehyde
during photooxidation reaction, in order to rationalize the
slow appearance of carboxylate IR bands (10). They fur-
ther stated that the presumed butanoic acid was strongly
absorbed on the TiO2 surface and that such acid accumula-
tion was responsible for the slow catalyst deactivation ob-
served. The catalyst deactivation could be enhanced during
each dark period, presumably by the dark, facile oxida-
tion of the remaining adsorbed butyraldehyde to butanoic
acid. They suggested that on the TiO2 surface, a hole attack
on the adsorbed carboxylate caused photocatalyzed decar-
boxylation, thereby regenerating the catalyst activity in the
lene conversion, the likely source for the Cl� atoms is the re-
action of trichlorethylene with photocatalytically produced
radicals to produce the observed intermediate, dichloroac-
etaldehyde chloride (19):
CCl2CHCl + �OH → �CCl2CHClOH
�
CCl2CHClOH + O2 → �OOCCl2CHClOH
2
�OOCCl2CHClOH → 2�OCCl2CHClOH + O2
�OCCl2CHClOH → CHClOHCCl(O) + Cl�.
A possible route for the reaction of the intermediate DCAC
to produce a second observed intermediate, phosgene,
Cl2CO, is given by Jacoby (19):
OH� + CHCl2CCl(O) → �CCl2Cl(O) + H2O
O2 + �CCl2CCl(O) → �OOCCl2CCl(O)
2
�OOCCl2CCl(O) → 2�OCCl2CCl(O) + O2
�
OCCl2CCl(O) → CCl2O + CClO�
CClO� → Cl� + CO.

HETEROGENEOUS PHOTOCATALYTIC OXIDATION
9
absence of further reactant alcohol:
Propagation
�
+
TCE + Cl� → CHCl CCl �
RCOO + h → RCOO� → CO2 + R�
2
2
CHCl2CCl2� + O2 → CHCl2CCl2OO�
→
gaseous products.
2
CHCl2CCl2OO� → 2CHCl2CCl2O� + O2
CHCl2CCl2O� → CHCl2CCl(O) + Cl�
In our experiments with simultaneous toluene and
trichloroethylene photocatalyzed oxidation, no intermedi-
ates were found by GC with a FID detector. Toluene pho-
tooxidation takes place on the TiO2 surface and follows the
Langumir–Hinshewood kinetics form. A used TiO2 cata-
lyst was suspended in methanol to extract any adsorbed
oxygenates, and the resultant solution was examine by
UV spectrum. This spectrum verified that some methanol-
extractable intermediates are formed during toluene pho-
tooxidation and remain absorbed on the TiO2 surface. Fur-
ther mass spectrum analysis (NCSU) indicated the major
absorbed intermediate to be benzoic acid. We suggest that
this absorbed benzoic acid is the oxidation product of ben-
zaldehyde, which would be consistent with both Ibusuki’s
Termination (example)
Cl� + H2O → �OH + HCl (presence of H2O)
2
Cl� → Cl2 (absence of H2O)
Chain transfer
Cl� + � CH3 → � CH2� + HCl
This example mechanism is consistent with the following
experimental observations:
(1) For TCE/air feeds, 100% conversion has been
observation of trace benzaldehyde gas-phase appearance achieved in 5–6 ms at all concentrations studied (up to 753
3
during toluene oxidation and Blake and Griffin’s claim mg/m ), which suggests a long reaction chain.
that adsorbed butyraldehyde can be converted to the cor-
responding carboxylic acid.
(2) With low toluene addition to the TCE/air system, Cl�
activates C6H5CH3 by chain transfer and thus increases the
Like Blake and Griffin’s claim that adsorbed carboxylic toluene rate and conversion dramatically; sufficient Cl� re-
acid inhibited butanol conversion, we suggest that accu- mains to maintain a TCE chain length sufficient to give
mulated benzoic acid inhibits toluene and trichloroethy- 100% conversion.
lene photooxidation rates. This view also helps to explain
(3) With higher toluene addition to the TCE/air system,
the TiO2 catalyst activity decreases during toluene and toluene consumes sufficient Cl� to terminate the chain re-
trichloroethylene photooxidation in high toluene (>200 action, therefore sharply inhibiting TCE reaction.
3
mg/m ) atmospheres.
The mechanism of further conversion of the aromatic
intermediates is unknown on the TiO2 surface, but sugges-
Coulpling of TCE and Toluene Oxidations
tions may again be found from gas-phase chemistry. For
To explain our toluene/TCE results, we need to postulate example, Atkinson et al. postulated the following steps to
a coupling between the stoichimetric toluene conversion explain the gas-phase �-dicarbonyl reaction (36).
and the assumed chain reaction model for TCE destruction.
This coupling must rationalize
(
i) toluene rate enhancement by TCE at low toluene
level and
(
ii) toluene quenching of TCE conversion at high
toluene levels.
A useful example for discussion is a network that includes a
chain mechanism for TCE conversion and a chain transfer
step for toluene involvement:
Toluene (nonchain sequence)
TiO2 + OH� + h� → �OH
�
OH + �CH3 → �CH2� + H2O
�
CH2 + (O , �OH) → CO2
2
TCE (chain) initiation
The �-dicarbonyl gas-phase yield is about 25–29% for
TCE + TiO2 + h� → Cl� + ���
toluene. From GC–MS and MS–MS analysis, a variety

1
0
LUO AND OLLIS
of other ring cleavage products are identified, including: motion and inhibition behavior vs that expected from indi-
CH3COCOCH == CH2, CHOCOCH == CH2, CH3COCH == vidual single-reactant kinetic degradation data. The pres-
CH2, CH3COCH == CHCH == CH2, CHOC(OH) == CHCHO, ence of sufficient trichloroethylene enhanced the toluene
and CH3COCH == CHCH == CHCHO.
photooxidation rate and allowed achievement of 90–100%
Similar ring cleavage processes may exist on the illumi- toluene conversion, when toluene feed levels were below
3
nated TiO2 surface for toluene due to the existence of the about 120 mg/m . These results constitute the first report
radical OH� and initiator Cl� and the presence of molec- of (near) total conversion at short residence times of an
ular oxygen, O2. The �-dicarbonyl or other unsaturated aromatic hydrocarbon, which presently represents a very
ring cleavage products formed can be further photooxi- important air emission problem. A global mechanism is
dized rapidly to the final products, CO2 and H2O. This ki- suggested that incorporates the individual nonchain and
netic possibility may rationalize why no appreciable gas- chain mechanisms thought to occur with toluene and TCE
phase intermediates and reaction products were detected individual conversions, respectively.
by GC/FID. It is also consistent with earlier observations by Time-dependent catalyst activation and deactivation
Teichner and Formenti (42) that photocatalyzed oxidation were observed. At first, a deactivated catalyst could be re-
of small olefins produces primarily carbon dioxide.
Other mechanistic possibilities are noted:
generated under UV illumination and dry air flow. Eventu-
ally, catalyst deactivation becomes irreversible. While car-
boxylate formation and carboxylic acid accumulation could
be a major cause of deactivation, the full picture here has
yet to be elucidated.
(
i) The TiO2 surface may become sufficiently chlori-
nated to yield a new surface, or even bulk phase; Primet
et al. (1) have shown that carbon tetrachloride reacts eas-
ily with surface oxygen atoms at very modest temperatures
�
ACKNOWLEDGMENTS
(
150 C), and such new phases may have desirable catalytic
properties, e.g., forming Cl radicals, and reducing or elimi-
nating undesired electron/hole recombinations.
This work was supported by NASA Research Grant NAG 2-684. We are
pleased to acknowledge NASA’s strong interest in multicomponent con-
(
ii) The influence of water remains unclear. Complete versions as a primary motivator for our exploration of the TCE enhanced
oxidation of toluene.
oxidation of TCE consumes water, while toluene conver-
sion to CO2 produces water. Dibble and Raupp (14) found
that excess water inhibits TCE conversion strongly. Thus,
the high rate to low rate transition seen on increasing
toluene could be a switch from a “trace water” system to an
REFERENCES
1
2
3
. Primet, M., Basset, J., Mothien, M. V., and Prettre, M., J. Phys. Chem.
79, 2868 (1970).
“
excess water” system. Further studies are clearly needed
to sort out these and yet other possibilities.
. Djeghri, N., Formenti, M., Juillet, F., and Teichner, S. J., Faraday
Discuss. Chem. Soc. 58, 185 (1974).
. Djeghri, N., and Teichner, S. J., J. Catal. 62, 99 (1980).
5
. CONCLUSIONS
4. Bickley, R. I., and Stone, F. S., J. Catal. 31, 389 (1973).
5
6
. Bickley, R. I., Munuera, G., and Stone, F. S., J. Catal. 31, 398 (1973).
. Gratzel, M., Thampi, K. R., and Kiwi, J., J. Phys. Chem. 93, 4128
(
Trichloroethylene and toluene in humidified air could
be individually photocatalytically oxidized by near-UV
illumination and TiO2 catalyst in a powder bed flow re-
actor with residence time only about 5.6 ms. Trichloroethy-
lene photooxidation was very rapid under our experimen-
tal conditions, and � 100% conversion was achieved for
all trichloroethylene concentrations examined up to 753
1989).
. Cunningham, J., and Hodnett, B. K., J. Chem. Soc. Faraday Trans. 1
7, 2777 (1980).
7
8
9
7
. Mozzanega, H., Herrmann, J., and Pichat, P., J. Phys. Chem. 183, 2251
(1979).
. Sato, S., J. Phys. Chem. 87, 3531 (1983).
10. Childs, L. P., and Ollis, D. F., J. Catal. 67, 35 (1981).
1
1
1a. Blake, N. R., and Griffin, G. L., J. Phys. Chem. 92, 5697 (1988).
1b. Peral, J., and Ollis, D. F., “The First International Conference on TiO2
Photocatalytic Purification and Treatment of Water and Air, London,
Ontario, Canada,” p. 741. Elsevier, Amsterdam/New York 1993.
3
mg/m . The toluene photodegradation rate was successfully
fitted by Langmuir–Hinshelwood form. Toluene oxidation
rates were of the same order magnitude as m-xylene and
acetone. No toluene photooxidation intermediateswere de-
tected; we believe that toluene was eventually converted
to carbon dioxide. Water vapor presence was essential for
toluene oxidation, as no toluene degradation was observed
in the complete absence of water vapor in the feed stream.
Toluene oxidation rate increased with water concentration
1
2. Ponder, W. H., “EPA’s Pollution Prevention Research Program in
the Air and Energy Engineering Laboratory’s Organics Control
Branch,” Presentation, Air and Energy Engineering Research Lab-
oratory. United States Environmental Protection Agency, Washing-
ton, DC, 1994.
13. Dibble, L. A., Ph.D. Dissertation, Arizona State University, 1989.
4. Dibble, L. A., and Raupp, G. B., Catal. Lett. 4, 345 (1990).
15. Dibble, L. A., and Raupp, G. B., Environ. Sci. Technol. 26, 492 (1992).
1
3
up to 1650 mg/m (20% relative humidity) and decreased
1
6. Nimlos, M. R., Jacoby, W. A., Blake, D. M., and Milne, T. A., “The
First International Conference on TiO2 Photocatalytic Purification
and Treatment of Water and Air, London, Ontario, Canada,” p. 387.
Elsevier, Amsterdam/New York, 1993.
thereafter.
The photooxidation kinetics of trichloroethylene and
toluene mixtures in humidified air exhibited strong pro-

HETEROGENEOUS PHOTOCATALYTIC OXIDATION
11
1
1
7. Holden, W., Marcellino, A., Valic, D., and Weedon, A. C., “The First
International Conference on TiO2 Photocatalytic Purification and
Treatment of Water and Air, London, Ontario, Canada,” pp. 393.
Elsevier, Amsterdam/New York, 1993.
8. Anderson, M. A., Yamazaki-Nishida, S., and Cervera-March, S., “The
First International Conference on TiO2 Photocatalytic Purification
and Treatment of Water and Air, London, Ontario, Canada,” p. 405.
Elsevier, Amsterdam/New York, 1993.
28. Gonzalez-Elipe, A. R., Munuera, G., and Soria, J., J. Chem. Soc.
Faraday Trans. 1 75, 748 (1979).
29. Primet, M., Pichat, P., and Mathieu, J., J. Phys. Chem. 75, 1216 (1971).
30. Primet, M., Pichat, P., and Mathieu, J., J. Phys. Chem. 75, 1221 (1971).
31. Anpo, M., Shima, T., and Kubokawa, Y., Chem. Lett. 1799 (1985).
32. Hoshino, M., Akimoto, H., and Okuda, M., Bull. Chem. Soc. Japan
51, 718 (1978).
33. Kenley, R. A., Davenport, J. E., and Hendry, D. G., J. Phys. Chem.
82, 1095 (1978).
34. Atkinson, R., Chem. Rev. 85, 69 (1985).
1
2
2
9. Jacoby, W. A., Ph.D. Dissertation, University of Colorado, 1993.
0. Ibusuki, T., and Takeuchi, K., Atmos. Environ. 20, 1711 (1986).
1. Suzuki, K., Satoh, S., and Yoshida, T., Denki Kagaku 59, 512
35. Gery, M. W., Fox, D. L., Stockburger, L., and Weathers, W. S., Int. J.
Chem. Kinet. 17, 931 (1985).
(
1991).
2
2. Suzuki, K., “The First International Conference on TiO2 Photocata-
lytic Purificaton and Treatment of Water and Air, London, Ontario,
Canada,” p. 421. Elsevier, Amsterdam/New York, 1993.
36. Atkinson, R., Carter, W. P. L., and Winer, A. M., J. Phys. Chem. 87,
1605 (1983).
37. Kenley, R. A., Davenport, J. E., and Hendry, D. G., J. Phys. Chem.
85, 2740 (1981).
38. Perry, R. A., Atkinson, R., and Pitts, J. N. J., J. Phys. Chem. 81, 296
(1977).
39. Sanhueza, E., and Heicklen, J., J. Phys. Chem. 79, 7 (1975).
40. Sanhueza, E., Hisatsune, J., and Heicklen, J., Chem. Rev. 76, 801
(1976).
2
2
3. Luo, Y., Ph.D. Dissertation, North Carolina State University, 1994.
4. Berman, E., and Dong, J., “The Third International Symposium
Chemical Oxidation: Technology for the Nineties, Vanderbilt Uni-
versity, Nashville, Tennessee,” Vol. 3, p. 133. Technomic Publishing,
1
993.
2
5. Peral, J., and Ollis, D. F., J. Catal. 136, 554 (1992).
2
6. Miller, R., and Fox, R., “The First International Conference on TiO2
Photocatalytic Purification and Treatment of Water and Air, London,
Ontario, Canada,” p. 573. Elsevier, Amsterdam/New York, 1993.
7. Munuera, G., Rives-Arnau, V., and Saucedo, A., J. Chem. Soc. Fara-
day Trans. 1 75, 736 (1979).
41. Kleingienst, T. E., Shepson, P. B., Nero, C. M., and Bufalini, J. J.,
Int. J. Chem. Kinet. 21, 863 (1989).
42. Teichner, S. J., and Formenti, M., “Heterogeneous Photocatalysis,
Photoelectrochemistry Photocatalysis and Photoreactors.” Reidel,
Dordrecht, 1985.
2