Catalytic and photocatalytic oxidation of ethylene on titania-based thin-films

Article abstract of DOI:10.1021/es991250m

The influence of reaction temperature on the catalytic and photocatalytic oxidation of ethylene on unplatinized and platinized versions of a TiO₂/ZrO₂ mixed-oxide thin-film catalyst was studied. A half order model of reaction kinetics was used to examine ethylene conversion data obtained using a recirculating tubular reactor. When operated in the presence of UV, both catalysts displayed considerable increases in the reaction rate constant at 70°C compared to 30°C. No difference was seen between platinized and unplatinized versions at these temperatures. However, only the platinized catalyst performed better at 107°C compared to its performance at 70°C. At 107°C in the presence of UV, the reaction rate constant determined for the platinized catalyst was a factor of 2 larger than that realized using the unplatinized catalyst. Catalytic reaction rate constants for the platinized catalyst were 1 (107°C) to 2 (70°C) orders of magnitude lower than corresponding photocatalytic reaction rate constants. Comparisons between the thin-film catalysts and a previously reported particulate catalyst revealed a fundamental difference in photocatalytic behavior, chiefly due to insufficiently usage of the internal bulk of the particulate catalyst.

Full text of DOI:10.1021/es991250m

Environ. Sci. Technol. 2000, 34, 5206-5210
the hydrodynamic pressure drop problems encountered
Catalytic and Photocatalytic
Oxidation of Ethylene on
Titania-Based Thin-Films
when using unsupported particulate catalysts. Thin-film
photocatalytic reactors have been previously reported for
oxidation of various compounds in both the liquid (6-8)
and gas phase (9-13). In this study, ethylene conversion
data collected under dry conditions using a multipass,
recirculating tubular reactor were analyzed using a half order
model of reaction kinetics. Calculated reaction rate constants
were then used to compare the behavior of the two catalyst
formulations under different reaction conditions. The results
obtained with supported thin-films were compared to
previously reported results using similarly prepared par-
ticulate forms of the catalyst (2).
M I C H A E L E . Z O R N , D E A N T . T O M P K I N S ,
W A L T E R A . Z E L T N E R , A N D
M A R C A . A N D E R S O N *
Water Chemistry Program, University of WisconsinsMadison,
6
60 North Park Street, Madison, Wisconsin 53706
Experimental Section
The effect of reaction temperature on the catalytic and
photocatalytic oxidation of ethylene on unplatinized and
platinized versions of a TiO2/ZrO2 mixed-oxide thin-film catalyst
is presented in this study. Ethylene conversion data
collected using a recirculating tubular reactor are analyzed
using a half order model of reaction kinetics. When
operated in the presence of UV, both catalysts exhibit
significant (95% confidence) increases in the reaction rate
constant at 70 °C compared to 30 °Csno difference is
observed between unplatinized and platinized versions at
these two temperatures. However, only the platinized
catalyst performs better at 107 °C compared to its performance
at 70 °C. Also, at 107 °C in the presence of UV, the
reaction rate constant obtained for the platinized catalyst
is a factor of 2 larger than that obtained using the
unplatinized catalyst. Catalytic reaction rate constants for
the platinized catalyst were 1 (107 °C) to 2 (70 °C)
orders of magnitude lower than corresponding photocatalytic
reaction rate constants. Comparisons between the thin-
film catalysts studied here and a previously reported
particulate catalyst show a fundamental difference in
photocatalytic behavior, mainly due to insufficient utilization
of the internal bulk of the particulate catalyst.
Catalyst Preparation and Characterization. The mixed-oxide
catalyst material used in this study was prepared using a
sol-gel processing approach (1) in which titania and zirconia
sols were prepared separately prior to combining (13). Three
layers of active catalyst material were deposited onto
borosilicate glass cylinders or “rings” (4 mm o.d. × 3 mm i.d.
× 12 mm long) that were prerinsed with high purity water.
The glass rings were prepared by cutting thin-wall glass tubing
(Kimble Glass, Deerfield, IL) to the required length. After
applying each layer, the rings were dried in air at 100 °C for
1
3
h. The coated rings were then fired in air in a furnace at
50 °C for 3 h using a 3 °C per minute ramp rate. Physical
properties of this catalyst can be estimated from measure-
ments on the unsupported particulates. Fu et al. (1) have
reported the specific surface area and porosity of similarly
particulates as 250 m2
-1
prepared TiO
respectively.
2
/ ZrO
2
g
and 55%,
Platinization was achieved by impregnating rings coated
as above with an aqueous chloroplatinic acid solution (8 wt
% H
PtCl
sodium borohydride (NaBH
M sodium hydroxide. Transmission electron micrographs
2
PtCl
was subsequently reduced with a solution of 0.1 M
, Aldrich, Milwaukee, WI) in 0.01
6
, Aldrich, Milwaukee, WI) (2, 14). The adsorbed
H
2
6
4
5
(
up to 3.3 × 10 magnification) of similarly prepared
particulates do not indicate the presence of individual
platinum particles on the titania catalyst at a platinum loading
level of 0.3 wt % (2, 14). Rather, it is likely that the platinum
is uniformly dispersed over the surface of the catalyst. The
platinized rings employed in this study did not exhibit any
noticeable dark color; however, a faint dark color was
observed with rings that were nominally loaded with five
times the amount of platinum as those employed herein.
To determine the mass of catalyst employed in this study,
Introduction
2 4
The removal of ethylene (C H , a plant hormone that
promotes the ripening of fruits and vegetables and accelerates
the aging of flowers) on titania-based materials has previously
been reported by a number of researchers (1-5). The effects
of temperature (2-5), water vapor concentration (2-4), and
ethylene concentration (3, 4) have been presented in these
studies. In addition, Fu et al. (1) reported enhanced photo-
an ammonium sulfate [(NH
burgh, PA]/ sulfuric acid (H
employed (15-17). For each sample, 4 g (NH
to 10 mL of hot concentrated H SO . After the (NH
completely dissolved in the H SO , five catalyst-coated glass
4
)
2
SO
SO ) digestion procedure was
SO was added
SO
4
4
, Fisher Scientific, Pitts-
2
4
catalytic performance of mixed-metal oxide (TiO
2
/ SiO
2
and
4
)
2
4
TiO / ZrO ) particles compared to pure titania for degrading
2
2
2
4
4
)
2
ethylene. Fu et al. (2) also reported enhanced degradation
of ethylene at elevated temperatures through the addition
of platinum to pure titania particles. The increased reactivity
was attributed to an increase in conventional heterogeneous
2
4
rings were added to the solution. The samples were covered
and carefully boiled for 1 h. After removing the glass rings
from the solution, appropriate dilutions were performed to
result in a 1% acid solution. Visual inspection of the digested
rings suggests that the digestion process completely removed
all of the catalyst coating. Elemental analyses (Ti, Zr, and Pt)
were subsequently performed using inductively coupled
plasma-mass spectrometry (ICP-MS) at the Soil and Plant
Analysis Laboratory at the University of WisconsinsMadison.
The average total mass of catalyst deposited on each ring
was determined to be 0.12 mg, with an average composition
by weight of 10% zirconia and 1.4% platinum (platinized
version only), with the balance titania.
(
nonphoto) catalytic reactions occurring on the Pt/ TiO
2
catalyst.
In this study, the effect of reaction temperature on the
catalytic and photocatalytic oxidation of ethylene on un-
platinized and platinized mixed-oxide TiO / ZrO thin-films
2 2
was examined. Coating a thin film of catalyst onto a support
allows for more efficient distribution of light and can decrease
*
Corresponding author phone: (608)262-2674; fax: (608)262-0454;
e-mail: nanopor@facstaff.wisc.edu.
5
2 0 6
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 24, 2000
10.1021/es991250m CCC: $19.00

2000 Am erican Chem ical Society
Published on Web 11/01/2000

or removal of smaller sample volumes (e.g., 100 µL) from the
recycling loop.
At the beginning of an experiment, the UV lights were
energized (photocatalytic experiments only), the pump was
started, the reactor temperature was allowed to reach the set
temperature, the recycling loop was broken between the
pump outlet and the inlet of the reactor, and the loop was
flushed with room air for 10-15 min. Next, a gas sampling
-
5
bag filled with ethylene at a concentration of 4.3 × 10 mol
-
1
L
(1055 ppm) with the balance ultra zero air (<0.1 ppm
total hydrocarbons, <5 ppm water vapor, AGA Specialty Gas,
Cleveland, OH) was connected at the reactor inlet, and the
ethylene mixture was pumped through the reactor. [Although
such a high ethylene concentration is not indicative of most
horticultural applications (with the possible exception of
apple storage and banana ripening facilities), it was necessary
to allow for a sufficient number of samples to be obtained
for all experiments. In addition, these reaction conditions
were chosen to allow comparison to work performed by Fu
et al. (2)]. After 3 min, the gas sampling bag was detached,
the loop was closed, and sampling was initiated. The
recirculating flow rate maintained by the pump was 200 mL
FIGURE 1. Experimental set up. 1stemperature controlled chamber
with four UV light sources; 2silluminated section of glass tube
(photocatalytic experiments only) randomly packed with eight
catalyst-coated glass rings; 3snonilluminated sections of glass
tube (not packed with catalyst-coated glass rings); 4sstainless
steel tubing; 5ssampling port (septum); 6sViton pump tubing; 7s
peristaltic pump.
Reactor Apparatus. The catalytic and photocatalytic
oxidation experiments employed a recycling loop as shown
in Figure 1. The recycling loop consisted of a 27 mL (1.5 cm
i.d. × 15 cm long) glass tube that held eight catalyst-coated
glass rings packed randomly in the center, a 30 mL (1.5 cm
i.d. × 17 cm long) empty glass tube, a sampling port (septum),
stainless steel tubing, and a peristaltic pump with Viton (Cole
Palmer, Vernon Hills, IL) pump tubing. For all experiments,
the total mass of catalyst (W) was 1.0 mg, and the total void
volume of the empty loop, the “reservoir” (Vres), was 72.5 mL.
By suitable adjustment of a cooling air stream and/ or a heat
tape, the average reactor air temperature (measured by
placing a J-type thermocouple in a glass insert located at the
center of the packed bed) was maintained at average
temperatures of 30 °C, 70 °C, or 107 °C (( 2 °C).
-
1
min , except where otherwise noted. Gas samples (250 µL)
were withdrawn from the recycling loop at regular intervals
by inserting a gastight syringe through the sampling port.
Values of initial ethylene concentration (CA0) ranged between
-
5
-1
-5
2
.9 × 10 mol L and 4.1 × 10 mol L-1.
Gas Chrom atography. The reactor contents were ana-
lyzed using a Hewlett-Packard 5890 Series II gas chromato-
graph (GC). The column effluent was routed through a
thermal conductivity detector (TCD) and flame ionization
detector (FID) connected in seriesssince measurement of
thermal conductivity is a nondestructive detection technique,
both detectors produced a response from a single sample.
The FID (maintained at 300 °C) was used to monitor the
concentration of ethylene, while the TCD (maintained at
For the photocatalytic experiments, four 4 W fluorescent
light bulbs (F4T5BL, WIKO Brand, Bulb Direct Co., Pittsford,
NY) were symmetrically spaced at a radius of 2.5 cm from
the axial centerline of the reactor. These bulbs produce a
strong peak centered at ca. 365-370 nm and provided an
2
00 °C) was used to monitor the concentration of carbon
dioxide produced during oxidation of ethylene. The con-
centration of water vapor, also a product of the reaction, was
not monitored during the experiments. Helium flowing at 30
-
1
-
2
mL min was used as the carrier gas. A 250 µL injection was
performed using a gastight syringe. The GC column was a
Porapak R packed column (6 feet long × 1/ 8 in. o.d., 80/ 100
mesh, Alltech Associates, Deerfield, IL) that was maintained
at 40 °C for 4 min.
average measured light irradiance of 3.1 mW cm . Light
irradiance levels were measured at the center of the reactor
housing (after removal of the catalyst-containing glass tube)
using a light meter (Model IL 1400 with Super-Slim probe in
the 250-400 nm wavelength range, International Light,
Newburyport, MA) and were obtained by rotating the light
probe at 45 degree increments and averaging the individual
measurements (individual values ranged from 2.8 to 3.4 mW
Kinetics. To allow estimation of kinetic rate constants,
-
1
A
the concentration of ethylene (C , in units of mol L ) was
measured over time. The data generated were then fit using
a general power law model (including powers of 0, 0.5, 1, 1.5,
and 2) and a Langmuir-Hinshelwood-Hougen-Watson
-
2
cm , with 6.2% relative standard deviation, n ) 10). Based
on separate (unpublished) studies with ethylene using
multiple reactor geometries, it appears that the catalyst does
not absorb all of the UV radiation present in this reactor. In
addition, it has previously been shown that a thin-film loading
(
LHHW) model of reaction kinetics. The half order model
and LHHW model both provided adequate fits to the data;
however, the half order model provided a slightly better fit
in all cases. A linearized form of the half order kinetic model
can be given as
-
2
of 0.3-0.5 mg cm of Degussa P-25 titania was required for
-
2
complete absorption of UV radiation at 5.3 mW cm (18).
In the experiments reported here, the catalyst loading was
approximately 1 order of magnitude below this level.
Prior to performing any degradation studies, leak-free
operation and tubing compatibility were verified by intro-
ducing ethylene into the recycling loop at room temperature
with the UV lights turned off and monitoring the concentra-
tion over extended time periods. In experiments with other
analytes, the effect of sample withdrawal (250 µL per sample,
see below) on analyte concentration was shown to be minimal
with the recycling loop employed in this study. After removal
of 15 samples from the recycling loop maintained at room
temperature and with the UV lights turned off, the measured
decrease in analyte concentration was only ca. 4%. In future
studies, the effect of sample removal may be decreased even
further by removal of a minimum number of samples and/
1
/ 2
1/ 2
Wt
CA
) CA0
- k
(1)
2
Vres
where k is the reaction rate constant (in units of mol L1/ 2
1
/ 2
-
1
-1
g
s ), W is the catalyst mass (in units of g), t is time (in
units of s), and Vres is the reservoir volume (in units of L)ssee
Hill (19) for a detailed discussion of kinetic modeling. The
reaction rate constant can be determined by performing a
1
/ 2
least-squares regression analysis of C
A
versus t, where the
reaction rate constant is the negative slope multiplied by
2Vres/ W.
Results and Discussion
Effect of Tem perature with Unplatinized TiO
Film s. The degradation of ethylene was measured over time
2 2
/ZrO Thin-
VOL. 34, NO. 24, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5 2 0 7

TABLE 1. Half Order Reaction Rate Constants (in units of
1
/2 1/2 -1 -1
mol L g s ) for Thin-Films
catalyst
30 °C
70 °C
107 °C
a
-4
-4
-4
-4
-6
-4 b
unplatinized
1.80 × 10
d
3.22 × 10
3.35 × 10
4.03 × 10
3.34 × 10
6.62 × 10
3.60 × 10
b
a
-4
-5
platinized
platinized
1.67 × 10
c
a
Photocatalytic oxidation (with UV irradiation). Average of four
separate experim ents (relative standard deviation ) 1.4%). Includes1
one replicate with a recirculating volum etric flow rate of 100 m L m in-
(
not significantly different at 95% confidence from replicates perform ed
-1
at 200 m L m in , the flow rate used for all other experim ents reported
in this study). See text for individual values. Catalytic oxidation (without
UV irradiation). Ethylene degradation not observed.
c
d
FIGURE 3. Fit of the ethylene conversion data obtained with
photocatalysis at 30 °C(0), 70 °C(O), and 107 °C()) with a platinized
TiO /ZrO thin-film catalyst to a 1/2 order kinetic model.
2 2
07 °C (95% CI: 3.28 × 10 to 3.51 × 10 mol1/ 2 L1/ 2 g-1 s-1);
-
4
-4
1
however, the reaction rate constant calculated at 70 °C is not
significantly different from the reaction rate constant cal-
culated at 107 °C. Thus, with the unplatinized catalyst,
increasing the reaction temperature from 70 °C to 107 °C
does not appear to have a significant effect on the reaction
rate constant. The same phenomenon was also recently
observed with the photocatalytic oxidation of acetone vapor
2 2
on TiO / ZrO thin-films (13).
Using GC/ TCD, the production of carbon dioxide was
monitored to calculate a mass balance based on the following
reaction:
FIGURE 2. Fit of the ethylene conversion data obtained with
photocatalysis at 30 °C (0), 70 °C (O), and 107 °C ()) with an
C H + 3O f 2CO + 2H O
(2)
unplatinized TiO
2
/ZrO
2
thin-film catalyst to a 1/2 order kinetic model.
2
4
2
2
2
The average number of moles of carbon dioxide produced
per mol of ethylene degraded (mineralization ratio) was 1.9
at 30 °C, 1.9 at 70 °C, and 2.0 at 107 °C, indicating that the
ethylene was completely mineralized to carbon dioxide and
water. In addition, no other peaks were detected with the
GC/ FID during periods of photocatalytic oxidation.
at three different temperatures (30 °C, 70 °C, and 107 °C) in
an initially dry (<5 ppm water vapor) feed stream with an
2 2
unplatinized TiO / ZrO thin-film catalyst (10 wt % zirconia)
in the presence of UV light. A linearized half order model
was used to calculate a kinetic rate constant at each
temperature, as listed in Table 1. Reaction rate constants
were determined by performing least-squares regression
analyses, as shown in Figure 2.
2 2
Effect of Tem perature with Platinized TiO /ZrO Thin-
Film s. The degradation of ethylene was also measured over
time in an initially dry feed stream at three different
temperatures (30 °C, 70 °C, and 107 °C) with a platinized
To assess the reproducibility of the experimental proce-
dure and to explore the possibility of any mass transfer
limitations from the bulk flow to the catalyst surface at the
2 2
TiO / ZrO thin-film catalyst (10 wt % zirconia and 1.4 wt %
platinum) in the presence of UV light. A linearized half order
model was again used to calculate a kinetic rate constant at
each temperature, as listed in Table 1. Reaction rate constants
were determined by performing least-squares regression
analyses, as shown in Figure 3. As with the unplatinized
catalyst, the reaction rate constant calculated at 30 °C is
-
1
recirculating flow rate used in this study (200 mL min ), the
reaction rate constant at 107 °C was calculated from multiple
experiments. Three experiments were performed at a flow
-
1
rate of 200 mL min , while one was performed at 100 mL
-
1
-4
min . The individual rate constants were 3.40 × 10 , 3.31
10 , 3.30 × 10 , and 3.37 × 10 (flow rate ) 100 mL
min ) mol . The reaction rate constant measured
at 100 mL min was not significantly different from the three
-
4
-4
1/ 2 1/ 2 -1 -1
-4
×
-4
-4
1/ 2
significantly lower (95% CI: 1.56 × 10 to 1.77 × 10 mol
-1
L
g
s
1/ 2 -1 -1
L
g
s ) than the reaction rate constants calculated at 70
-1
-4
-4
1/ 2 1/ 2 -1 -1
°
C (95% CI: 2.96 × 10 to 3.75 × 10 mol
L
g
s
) and
-
1
reaction rate constants measured at 200 mL min (average
-4
-4
1/ 2 1/ 2
g-1
-1
1
07 °C (95% CI: 6.21 × 10 to 7.02 × 10 mol
L
s ).
-
4
-5
1/ 2
)
3.34 × 10 , 95% confidence interval ) 1.3 × 10 mol
However, in contrast to the unplatinized catalyst, the reaction
rate constant calculated at 70 °C is also significantly lower
than the reaction rate constant calculated at 107 °C. In fact,
the reaction rate constant at 107 °C is nearly double the
reaction rate constant at 70 °C.
1
/ 2 -1 -1
L
g
s ). Although not a rigorous experimental investiga-
tion, this result suggests that mass transfer limitations were
-
1
not present at 200 mL min . The average reaction rate
-
4
constant at 107 °C using all four experiments was 3.34 × 10
1
/ 2 1/ 2 -1 -1
mol , with a relative standard deviation of 1.4%.
L
g
s
Comparisons can also be made between the unplatin-
ized and platinized catalyst at each temperature. At 30 °C
and 70 °C, the reaction rate constants calculated using the
unplatinized catalyst are not significantly different from the
reaction rate constants calculated using the platinized
catalyst. However, at 107 °C, the platinized catalyst provides
a significant increase (a factor of 2) in the reaction rate
constant compared to the unplatinized catalyst.
The improved photocatalytic activity of platinized catalyst
materials at relatively low platinum loadings (0.1-1 wt %,
similar to that employed herein) has previously been
observed, and explanations have been proposed (2, 14, 21).
The general mechanism for the process of photocatalysis
Rigorous statistical analyses were performed to determine
whether the difference in reaction rate constants (i.e., the
difference in slopes) at the three temperatures is statistically
significantssee Draper and Smith (20) for a complete
discussion of the significance of slopes. Confidence intervals
(
that provide lower and upper limits within which the true
value is contained) were calculated for each of the reaction
rate constants to evaluate significance. The reaction rate
constant calculated at 30 °C (95% confidence interval (CI):
-
4
-4
1/ 2 1/ 2 -1 -1
1
.77 × 10 to 1.83 × 10 mol
L
g
s ) is significantly
lower than the reaction rate constants calculated at 70 °C
95% CI: 3.07 × 10 to 3.37 × 10 mol
-
4
-4
1/ 2 1/ 2 -1 -1
(
L
g
s ) and
5
2 0 8
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 24, 2000

involves the illumination of a semiconductor catalyst surface
with light of energy greater than the semiconductor band
gap, resulting in the promotion of electrons from the valence
band to the conduction band. This leaves behind positive
holes in the valence band. It is possible that the electrons
and holes will simply recombine (i.e., return of excited
electrons to the valence band), effectively decreasing the rate
of the photocatalytic process. Alternatively, the electrons can
be scavenged by an oxidant (typically molecular oxygen),
and the positive holes can be scavenged by oxidizable
compounds (e.g., water vapor to form reactive hydroxyl
radicals and/ or possible direct oxidation of hydrocarbons,
although this latter possibility is somewhat controversial),
thus enabling overall reactions to occur. [Note that even
though the ethylene/ air mixture that is introduced to the
reaction system is dry, some water is physisorbed to the thin
film catalyst at the start of all experiments. In addition, water
vapor is generated during the oxidation of ethylene, as shown
in eq 2.] It has been theorized that the improved photo-
catalytic performance observed with platinized catalyst
materials in aqueous systems can be attributed to an
enhanced ability to collect photogenerated electrons on the
islands of platinum that form at the catalyst surface, thereby
resulting in a more efficient reduction of oxygen as compared
to unplatinized catalyst materials (2, 14, 21, 22). In addition,
it has been reported that molecular oxygen undergoes
dissociative adsorption on platinum metal (14, 23). Although
the actual mechanism is not known, the presence of these
highly reactive dissociated species, especially at higher
temperatures, coupled with the more efficient utilization of
photogenerated electrons may explain the observed en-
hancement in reaction rate constant with the platinized thin-
films at 107 °C but not at 30 °C or 70 °C.
TABLE 2. Half Order Reaction Rate Constants (in units of
1
/2 1/2 -1 -1
a
mol L g s ) for Particulates
catalyst
30 °C
70 °C
107 °C
unplatinizedb
platinizedb
platinizedd
1.34 × 10
e
-6
-7
2.56 × 10
6.71 × 10
4.49 × 10
-6
-6
-6
2.56 × 10
1.99 × 10
1.99 × 10
-6
-5
-5
c
c
8.14 × 10
a
Calculated from data presented in ref 2. b Photocatalytic oxidation
with UV irradiation). Calculated from a fractional conversion of 1.0
c
(
(
i.e., 100% degradation of ethylene). Hence, reaction rate constant m ay
be underestim ated. d Catalytic oxidation (without UV irradiation).
e
Ethylene degradation not observed.
For comparison, half order reaction rate constants can
be calculated at each of the temperatures studied herein
using the data presented by Fu et al. (2). Rate constants can
be calculated using reported fractional conversion (f ) values
A
according to the following equation
1/ 2
2
F C
V A0
k )
(1 -
x1 - fA)
(3)
W
where fractional conversion is calculated as
CA
fA ) 1 -
CA0
(4)
The half order reaction rate constants for photocatalytic and
catalytic oxidation of ethylene using unplatinized and
platinized particulates are provided in Table 2. A number of
similarities and differences can be observed between the
thin-film and particulate systems. First, the photocatalytic
reaction rate constants calculated using unplatinized and
platinized particulates follow a similar pattern as described
above with thin-films. With the unplatinized catalyst, an
increase in reaction rate constant is observed between 30 °C
and 70 °C; however, no difference is observed between 70
The effect of conventional heterogeneous catalysis alone
(
i.e., with the UV lights turned off) was also examined at each
of the three temperatures (30 °C, 70 °C, and 107 °C) with the
platinized TiO / ZrO thin-film catalyst used above. As before,
2
2
a linearized half order model was used to calculate kinetic
rate constants at 70 °C and 107 °Csconventional catalysis
was not observed at 30 °C (see Table 1). Catalytic reaction
rate constants calculated with the platinized catalyst were 1
°
C and 107 °C. With the platinized catalyst, increased reaction
rate constants are observed at both 70 °C and then again at
07 °C. The behavior differs somewhat in that the rate
1
constants observed with the platinized particulates are slightly
lower at 30 °C and slightly higher at 70 °C as compared to
the unplatinized particulates. With the thin-films, no dif-
ference is observed between unplatinized and platinized
versions at these two temperatures. With the platinized
catalyst, no catalytic oxidation was observed with either the
particulates or thin films at 30 °C.
(
107 °C) to 2 (70 °C) orders of magnitude lower than
corresponding photocatalytic reaction rate constants (both
are significant at a 95% confidence level). It is evident that,
although conventional catalysis occurs to some degree at
temperatures above 30 °C, the addition of UV light greatly
enhances the oxidation of ethylene to carbon dioxide and
water.
A major difference between the two systems is that the
photocatalytic reaction rate constants calculated with the
thin-films are about 2 orders of magnitude higher than the
corresponding reaction rate constants calculated with par-
ticulates. Actually, the observed differences in reaction rates
between thin-films and particulates might be even greater
if measured with equivalent levels of light irradiance, since
the UV radiation is not completely utilized in the reactor
apparatus used here. In addition, note that the level of water
vapor, the other major factor controlling the reaction rate
constant in these systems, is similar (less than 5 ppm initially)
in both studies. A likely explanation for this result is that the
UV light incident on the outside of the particle bed is quickly
absorbed by the catalyst. Therefore, only a small percentage
of particulate catalyst is actually photoactivated. In contrast,
the UV light is better distributed and more efficiently utilized
in the thin-film system, resulting in photoactivation of a much
larger percentage of the catalyst. This would explain the larger
reaction rate constants listed in Table 1 for the thin-films.
If this proposed explanation is true, the reaction rate
constants calculated with conventional catalysis alone should
be nearly equivalent between particulates and thin-films since
As with the unplatinized photocatalytic experiments, the
production of carbon dioxide using the platinized catalyst
was monitored using GC/ TCD. The average mineralization
ratio was 1.9 at 30 °C, 1.9 at 70 °C, and 2.0 at 107 °C.
Com parison to Particulates. Similar experiments were
previously performed by Fu et al. (2) using unplatinized (TiO
2
)
and platinized (0.3 wt % platinum, balance TiO ) particulate
2
catalysts. These data were collected using a single pass (i.e.,
noncirculating) fixed-bed tubular reactor (Pyrex, 2.4 mm
inner diameter, 11 cm length) that was packed with 0.4 g of
catalyst particles. Ethylene at an initial concentration of 1.59
-
5
-1
×
10 mol L was passed through the reactor at a volumetric
-
3
-1
-1
flow rate (F
V
) of 1 × 10 L s (60 mL min ). As with the
experiments performed in this study, the water vapor
concentration was also less than 5 ppm. The geometric
configuration of the reactor and the four 4 W fluorescent
light bulbs (F4T5BLB, GE Brand, Bulb Direct, Pittsford, NY)
employed by Fu et al. (2) was similar to that employed in this
study. Light irradiance measured for that system [1.65 × 10
einsteins s or 5.9 mW cm (1, 24)] was about a factor of
greater than that employed here.
-7
-
1
-2
2
VOL. 34, NO. 24, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5 2 0 9

the internal bulk of the particulate catalyst is accessible to
both heat and reactants. As shown in Tables 1 and 2, the
catalytic reaction rate constants calculated with thin-films
and particulates are relatively similar. The rate constants at
However, as mentioned, a likely cause is that the majority
of the particulates is not efficiently illuminated. As shown in
the present study, when the catalyst is efficiently illuminated,
as with the thin-films, heterogeneous catalytic oxidation plays
a much smaller role in the overall reaction (at the temper-
atures investigated in this study).
7
0 °C are within 10%, and the rate constants at 107 °C are
within a factor of 2 (thin-films > particulates). The particulate
catalytic reaction rate constant at 107 °C may be slightly
underestimated since it is calculated based on a fractional
conversion of 1.0.
For purposes of illustration, it is possible to calculate
reaction rate constants using the data presented by Fu et al.
Literature Cited
(
1) Fu, X.; Clark, L. A.; Yang, Q.; Anderson, M. A. Environ. Sci. Technol.
1996, 30, 647-653.
(2) Fu, X.; Clark, L. A.; Zeltner, W. A.; Anderson, M. A. J. Photochem.
Photobiol. A: Chem. 1996, 97, 181-186.
(
2) that are adjusted to account only for the particulate catalyst
(
3) Obee, T. N.; Hay, S. O. Environ. Sci. Technol. 1997, 31, 2034-
mass that is illuminated. The most meaningful comparison
of this type is for the unplatinized samples where photo-
catalysis is the only process occurringsthe platinized versions
include both photocatalytic as well as catalytic effects.
Assuming spherical particles with an average diameter of 1
mm (reported range: 0.5-1.4 mm), illumination of the entire
outer surface of each particle (a reasonable assumption
considering that the size of the particles is on the same order
as the inner diameter of the reactor, 2.4 mm), and a UV
penetration depth ranging between 1.0 and 4.5 µm [a 1 µm
penetration depth is obtained from data obtained in our
2
038.
(
4) Tibbitts, T. W.; Cushman, K. E.; Fu, X.; Anderson, M. A.; Bula,
R. J. Adv. Space Res. 1999, 22, 1443-1451.
(5) Sirisuk, A.; Hill, C. G., Jr.; Anderson, M. A. Catal. Today 1999,
5
4, 159-164.
(
(
(
6) Matthews, R. W. J. Phys. Chem. 1987, 91, 3328-3333.
7) Matthews, R. W. J. Catal. 1988, 111, 264-272.
8) Porter, J. F.; Yue, P. L. In Photocatalytic Purification and
Treatment of Water and Air; Ollis, D. F., Al-Ekabi, H., Eds.; Elsevier
Science Publishers B.V.: Amsterdam, 1993; pp 759-762.
(9) Alberici, R. M.; Jardim, W. E. Appl. Catal. B: Environ. 1997, 14,
5-68.
5
(
(
10) Jacoby, W. A.; Nimlos, M. R.; Blake, D. M.; Noble, R. D.; Koval,
C. A. Environ. Sci. Technol. 1994, 28, 1661-1668.
11) Jacoby, W. A.; Blake, D. M.; Fennell, J. A.; Boulter, J. E.; Vargo,
L. M.; George, M. C.; Dolberg, S. K. J. Air Waste Manage. Assoc.
1996, 46, 891-898.
laboratory using TiO thin films (25), and 4.5 µm is obtained
2
from the literature for Degussa P25 particles (26)], a semi-
quantitative estimate for the percentage of catalyst that is
actually illuminated can be set between 0.3 and 1.3%. This
estimate is obtained by subtracting the volume of a sphere
with a diameter of 1 mm (representing the total catalyst
particle volume) from the volume of a sphere with a diameter
of either 0.9955 or 0.9990 mm (representing the portion of
the catalyst particle volume that is not activated by UV light)
and dividing this difference by the total catalyst particle
volume. Dividing the unplatinized particulate reaction rate
constants listed in Table 2 by 0.003 and by 0.013 (to account
only for the portion of catalyst that is illuminated assuming
a penetration depth of 1.0 and 4.5 µm, respectively) yields
corrected values that bound the unplatinized thin-film
reaction rate constants measured in this study, as listed in
Table 1. More specifically, the corrected unplatinized par-
(12) Raupp, G. B.; Nico, J. A.; Annangi, S.; Changrani, R.; Annapragada,
R. AIChE J. 1997, 43, 792-801.
13) Zorn, M. E.; Tompkins, D. T.; Zeltner, W. A.; Anderson, M. A.
Appl. Catal. B: Environ. 1999, 23, 1-8.
14) Fu, X.; Zeltner, W. A.; Anderson, M. A. Appl. Catal. B: Environ.
1995, 6, 209-224.
(
(
(15) Barksdale, J. Titanium: Its Occurrence, Chemistry, and Technol-
ogy, 2nd ed.; Ronald Press Company: New York, 1966; p 139.
(
16) Norris, J. D. Encyclopedia of Analytical Science, Vol. 9; Academic
Press: San Diego, CA, 1995; p 5237.
(
17) Jackson, N. B.; Wang, C. M.; Luo, Z.; Schwitzgebel, J.; Ekerdt,
J. G.; Brock, J. R.; Heller, A. J. Electrochem. Soc. 1991, 138, 3660-
3664.
(18) Jacoby, W. A.; Blake, D. M.; Noble, R. D.; Koval, C. A. J. Catal.
995, 157, 87-96.
1
-
4
(19) Hill, C. G., Jr. An Introduction to Chemical Engineering Kinetics
and Reactor Design; Wiley: New York, 1977; Chapters 3 and 6.
ticulate reaction rate constants range between 1.0 × 10
-
4
1/ 2 1/ 2 -1 -1
and 4.5 × 10 mol at 30 °C and between 2.0
L
g
s
(
20) Draper, N. R.; Smith, H. Applied Regression Analysis, 2nd ed.;
-
4
-4
1/ 2 1/ 2 -1 -1
×
10 and 8.5 × 10 mol
L
g
s
at 70 °C and 107 °C.
Wiley: New York, 1981; p 25.
This is excellent agreement considering the above-mentioned
assumptions and the less than rigorous estimation method
employed for this comparison.
(21) Fox, M. A.; Dulay, M. T. Chem. Rev. 1993, 93, 341-357.
(22) Izumi, I.; Dunn, W. W.; Wilbourn, K. O.; Fan, F. F.; Bard, A. J.
J. Phys. Chem. 1980, 84, 3207-3210.
(
23) Boudart, M.; Dj e´ ga-Mariadassou, G. Kinetics of Heterogeneous
Catalytic Reactions; Princeton University Press: Princeton, NJ,
The difference between thin-films and particulates with
regard to light distribution affects one of the major conclu-
sions drawn by Fu et al. (2): that “the role of increasing
reaction temperatures is mainly to increase the rate of
heterogeneous catalytic oxidation”. It was stated that at
temperatures above 60 °C, “conventional heterogeneous
catalytic oxidation may become the predominant reaction
1
984; p 23.
(24) Yamazaki-Nishida, S.; Nagano, K. J.; Phillips, L. A.; Cervera-
March, S.; Anderson, M. A. J. Photochem. Photobiol. A: Chem.
1
993, 70, 95-99.
25) Miller, L. Ph.D. Dissertation, University of Wisconsin-Madison,
998.
26) Peral, J.; Ollis, D. F. J. Catal. 1992, 136, 554-565.
(
(
1
on an illuminated Pt/ TiO
reached because at relatively high temperatures (i.e., above
0 °C), the rates of reaction with platinized particulates are
2
surface”. This conclusion was
Received for review November 8, 1999. Revised manuscript
received September 5, 2000. Accepted September 11, 2000.
7
very similar with and without the use of the UV lights (i.e.,
the effect of photocatalysis is observed to be very small).
ES991250M
5
2 1 0
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 24, 2000