Catalytic Oxidation of 1,2-Dichlorobenzene over Supported Transition Metal Oxides

Article abstract of DOI:10.1006/jcat.2000.2895

The catalytic oxidation of 1,2-dichlorobenzene has been systematically investigated over a series of transition metal oxides (i.e., Cr₂O₃, V₂O₅, MoO₃, Fe₂O₃, and Co₃O₄) supported on TiO₂ and Al₂O₃. The activity of the different catalysts for this reaction depends on the nature of the transition metal oxide used, with Cr₂O₃-and V₂O₅-based catalysts being the most active ones. With the exception of the cobalt oxide catalysts, the TiO₂-supported systems were more active than the corresponding Al₂O₃-supported ones, indicating that the metal oxide-support interactions are significant in this reaction. Experiments conducted in the presence of water indicate an inhibiting effect for the V₂O₅- and Cr₂O₃-based catalysts and a promoting effect for Co₃O₄/TiO₂. The Fe₂O₃- and MoO₃-based catalysts were unaffected by the presence of water. Competitive adsorption between the surface species and water is suspected to be the reason for the inhibition, while the promoting effect can be attributed to the reaction of water with surface Cl^-. In situ FTIR studies indicate the presence of carboxylates (i.e., acetates and formates), maleates, and phenolates on the surfaces of all catalysts studied under reaction conditions. These surface species were reactive in the presence of gas-phase oxygen and are potential intermediates for the oxidation of 1,2-dichlorobenzene.

Full text of DOI:10.1006/jcat.2000.2895

Journal of Catalysis 193, 264–272 (2000)
doi:10.1006/jcat.2000.2895, available online at http://www.idealibrary.com on
Catalytic Oxidation of 1,2-Dichlorobenzene over Supported
Transition Metal Oxides
Sundaram Krishnamoorthy, Juan A. Rivas, and Michael D. Amiridis1
Department of Chemical Engineering, University of South Carolina, Columbia, South Carolina 29208
Received December 22, 1999; revised April 11, 2000; accepted April 17, 2000
catalytic reduction with NH3, are also active for the oxi-
dation of PCDD/PCDFs (2–6). In contrast, noble metal-
based catalysts, which show high activity for the oxidation
of nonchlorinated organics, undergo poisoning in the pres-
ence of Cl (7, 8).
The catalytic oxidation of 1,2-dichlorobenzene has been system-
atically investigated over a series of transition metal oxides (i.e.,
Cr2O3, V2O5, MoO3, Fe2O3, and Co3O4) supported on TiO2 and
Al2O3. The activity of the different catalysts for this reaction de-
pends on the nature of the transition metal oxide used, with Cr2O3-
and V2O5-based catalysts being the most active ones. With the ex-
ception of the cobalt oxide catalysts, the TiO2-supported systems
were more active than the corresponding Al2O3-supported ones, in-
dicating that the metal oxide–support interactions are significant in
this reaction. Experiments conducted in the presence of water in-
dicate an inhibiting effect for the V2O5- and Cr2O3-based catalysts
and a promoting effect for Co3O4/TiO2. The Fe2O3- and MoO3-based
catalysts were unaffected by the presence of water. Competitive ad-
sorption between the surface species and water is suspected to be
the reason for the inhibition, while the promoting effect can be at-
tributed to the reaction of water with surface Cl . In situ FTIR
studies indicate the presence of carboxylates (i.e., acetates and for-
mates), maleates, and phenolates on the surfaces of all catalysts
studied under reaction conditions. These surface species were re-
active in the presence of gas-phase oxygen and are potential inter-
c
Due to the high toxicity of PCDD/PCDFs, laboratory
studies in this area usually employ model compounds
such as chlorobenzenes and chlorophenols to predict the
dioxin destruction behavior of different catalysts (9–15).
Our previous work on the oxidation of 1,2-dichlorobenzene
(o-DCB) over vanadia supported on Al2O3 and TiO2 has
indicated that chlorine abstraction is probably the first step
of the reaction (16) and that the rate-determining step in-
volves a single vanadia site (10). The rate of the reaction
was also affected significantly by the nature of the sup-
port, with V2O5/TiO2 being more active than V2O5/Al2O3.
The same effect has also been observed in the case of the
oxidation of methanol (17) and butane (18) over similar
catalysts.
In thispaper, we present the resultsofour detailed kinetic
and in situ FTIR studies of o-DCB oxidation over several
different transition metal oxides supported on Al2O3 and
TiO2. The performance of these catalysts was compared
with that of the vanadia-based catalysts studied in our pre-
vious work (10, 16), in an attempt to understand the role of
the different transition metal oxides, and to obtain a better
picture of the reaction mechanism. Activity measurements
were also conducted in the presence of water, since water
vapor is an integral part of the flue gas emanating from the
incinerators and could affect the catalyst performance. Fi-
nally, in situ FTIR studies were performed both in the pres-
ence and in the absence of water, in order to understand the
nature and reactivity of the various surface species formed
during the reaction.
mediates for the oxidation of 1,2-dichlorobenzene.
2000 Academic
Press
Key Words: PCDD/PCDF; dichlorobenzene; oxidation; transi-
tion metal oxides.
INTRODUCTION
The destruction of polychlorinated aromatics, such as
polychlorinated dibenzodioxins (PCDDs) and dibenzofu-
rans (PCDFs), has received a lot of attention recently,
due to a worldwide concern about the toxicity of these
compounds (1). Thus, stringent limits (e.g., 0.1 ng of I-
TEQ/Nm3) have been imposed on the emissions of such
compounds in many countries (2). One of the principal
sources of PCDD/PCDF emissions, solid waste incinera-
tors, is becoming the method of choice for waste disposal in
urban areas. Catalytic oxidation is the preferred method for
the destruction of PCDD/PCDFs from these incinerators.
V2O5/TiO2-based catalysts, employed in most inciner-
ators for the control of NOx emissions via its selective
EXPERIMENTAL
Catalyst Preparation
Allthe catalystsused in thisstudywere prepared byincip-
ient wetness impregnation. Catapal-G Al2O3 and Kemira-
95 TiO2 were used as the supports. Chromium nitrate
1 To whom correspondence should be addressed. E-mail: amiridis@
engr.sc.edu.
264
0021-9517/00 $35.00
c
Copyright
2000 by Academic Press
All rights of reproduction in any form reserved.

CATALYTIC OXIDATION OF 1,2-DICHLOROBENZENE
265
nonahydrate (Aldrich), cobalt nitrate hexahydrate temperature. The saturated o-DCB/N2 stream was mixed
(Aldrich), iron nitrate hexahydrate (Aldrich), and ammo- with O2 and an additional stream of N2 to achieve the de-
nium molybdate (Aldrich) were used as the precursors for sired concentrations of 600 ppmv o-DCB and 10% O2. The
chromium, cobalt, iron, and molybdenum oxides, respec- reactant stream was then preheated and introduced into
tively. Vanadium oxalate, prepared by the addition of oxalic the reactor. The volumetric flow rate through the reactor
acid (Mallinkrodt) to an aqueous solution of vanadium was maintained at 445 scm3/min (space velocity of approx-
oxide (Strem) in ammonium hydroxide (Aldrich), was used imately 25,000 h 1). Prior to each experiment, the catalyst
as the precursor for V2O5. The metal oxide loadings were sample was pretreated in situ in O2 at 773 K for 2 h. For
calculated such that all catalysts contained approximately measurements in the presence of water, an additional sat-
the same amount of transition metal on a molar basis (see urator maintained at room temperature was connected to
Table 1).
The impregnated catalyst samples were first dried
the second N2 stream.
The analysis of o-DCB in the reactant and product
overnight in a vacuum oven at 353 K. The dried samples streams was performed on-line with an SRI 8610 gas chro-
were then slowly heated to 773 K in 6 h and calcined at this matograph equipped with a 18 -in. silica gel packed col-
temperature for an additional 2 h. A slightly higher calcina- umn and a flame ionization detector. In order to pro-
tion temperature of 793 K was used for the vanadia samples. tect the chromatographic column from damage due to
At these conditions, all precursors decomposed to form the HCl formed during the reaction, a scrubber containing
1
corresponding oxides. The stoichiometries of these oxides an alkaline material (ALCOA Selexsorb SPCL in.) was
are believed to be as follows: Cr2O3, Co3O4, MoO3, Fe2O3, fitted to the exit of the reactor. Initial meas8urements
and V2O5 (19, 20).
demonstrated that the alkaline scrubber becomes saturated
The catalysts were characterized via inductively coupled with o-DCB within a few minutes, while HCl continues
plasma (ICP) spectroscopy (Galbraith Laboratories) and to react with this material. The results presented in this
BET surface area measurements (Pulse Chemisorb 2000). paper were obtained under steady-state conditions (i.e.,
The results of these analyses are shown in Table 1.
after the o-DCB breakthrough was completed). Further-
more, experiments were conducted without the scrubber,
and the o-DCB conversions observed were identical to the
ones observed with the scrubber attached to the reactor
outlet.
Activity Measurements
Activity measurements were carried out in a quartz,
single-pass flow reactor which incorporated 500 mg of cata-
lyst in the form of 80–120 mesh size particles. Experiments
carried out with different particle sizes have demonstrated
the absence of any internal diffusional limitations in the
system. The reaction temperature was monitored using a
thermocouple projecting into the catalyst bed. o-DCB was
introduced into the vapor phase using a saturator. Nitro-
gen was flowed through the saturator maintained at room
In Situ FTIR Spectroscopy
Infrared spectra were collected using a Nicolet 740 FTIR
spectrometer equipped with an MCT-B detector cooled by
liquid nitrogen. The spectra were collected in the single-
1
beam mode with a 2 cm resolution. Experiments were
conducted in a stainless steel IR cell (NaCl windows), with
a path length of 10 cm. The catalyst sample was in the form
of a thin self-supported disk, approximately 1 cm in diam-
eter and 16 mg in weight. The disk was placed in a sample
holder and kept in the middle of the cell. Typical concentra-
tions used for the FTIR experiments were 700 ppmv o-DCB
and 5% O2, with He being the carrier gas. Prior to each ex-
periment, the sample was oxidized in situ at 623 K in a 10%
O2 in He mixture for 2 h. Spectra of clean surfaces taken
in O2 (for reaction studies) or He (for adsorption studies)
were used as the backgrounds for the different experiments.
Gas-phase subtraction of o-DCB was also performed on all
spectra.
TABLE 1
Composition and Surface Area of Catalysts Studied
Catalysts
TiO2
V2O5/TiO2
Cr2O3/TiO2
Fe2O3/TiO2
Co3O4/TiO2
MoO3/TiO2
Al2O3
V2O5/Al2O3
Cr2O3/Al2O3
Fe2O3/Al2O3
Co3O4/Al2O3
MoO3/Al2O3
% MxOy
Loadingb
SAc
a
—
5.8
5.2
5.0
4.5
8.9
—
5.6
4.5
4.7
4.7
9.0
—
6.4
6.8
6.3
6.0
6.2
—
6.2
5.9
5.9
6.3
6.3
85
77
87
89
97
94
140
140
146
139
133
139
RESULTS AND DISCUSSION
Activity Measurements
a Metal oxide loading (wt% MxOy).
b Metal loading ( 10 4 mol/g catalyst).
c BET surface area (m2/g).
The o-DCB conversions observed with different tran-
sition metal oxides supported on TiO2 are shown as

266
KRISHNAMOORTHY, RIVAS, AND AMIRIDIS
was observed between the fresh and spent samples). Fur-
thermore, X-ray diffraction patterns of the spent samples
do not reveal the formation of any new crystalline phases.
Results obtained in the presence of water (discussed in sub-
sequent sections) suggest that the low activity of this sample
can be attributed to the slow removal of Cl from the sur-
face of the Co3O4/TiO2 catalyst and, hence, the blockage of
the active sites by Cl .
Activity measurements were also conducted at different
temperatures with the Al2O3-supported catalysts and the
results are shown in Fig. 2. Once again the transition metal
oxide-containing catalysts exhibited much higher activities
than the Al2O3 support, indicating that these oxides provide
more active sites for the oxidation of o-DCB. A comparison
between Figs. 1 and 2 shows that the TiO2-supported cata-
lysts exhibit higher activities than the corresponding Al2O3-
supported ones. Similar results regarding the effect of the
support have been previously reported for the oxidation of
methanol over V2O5 (17), Cr2O3, and MoO3 catalysts (25),
and for the oxidation of benzene (30), but-1-ene (31), and
furan (32) over vanadia catalysts. The activities of the dif-
ferent transition metal oxides were in general in the same
order as with the titania-supported catalysts. The exception
was MoO3/Al2O3, which in this case was found to be at the
lower end of the activity scale.
FIG. 1. o-DCB conversion as a function of temperature for different
TiO2-supported catalysts. (᭹) 5.2 wt% Cr2O3/TiO2, (ᮀ) 5.8 wt% V2O5/
TiO2, (+) 8.9 wt% MoO3/TiO2, (᭡) 5.0 wt% Fe2O3/TiO2, (♦) 4.5 wt%
Co3O4/TiO2, (᭿) TiO2 (600 ppmv o-DCB, 10% O2, 25,000 h 1).
functions of temperature in Fig. 1. Cr2O3, V2O5, MoO3, and
Fe2O3 exhibited relatively high activities for the reaction,
with the first two being more active than the others. The
TiO2 support also exhibits some activity for the oxidation
of o-DCB. However, the addition of the transition metal
oxides—with the exception of Co3O4—increased the activ-
ity of TiO2 appreciably, indicating that the transition metal
oxides provide more active sites. A blank reactor experi-
ment indicated no significant gas-phase oxidation of o-DCB
at temperatures up to 773 K.
Our results are in general agreement with previous re-
ports regarding the oxidation activity of these transition
metal oxides. Wachs and co-workers, for example, have
shown that the turnover frequency for the oxidation of
methanol over V2O5/TiO2 is one order of magnitude higher
than that over MoO3/TiO2 (21). Studies conducted by other
researchers on the oxidation of methanol (22–24) and
1-butene (25) have also shown similar behavior for Cr2O3,
V2O5, and MoO3 catalysts. Finally, Larrson and co-workers
have shown that TiO2-supported iron oxide is more active
than cobalt oxide dispersed over the same support for the
oxidation of toluene (26).
Effect of water. Activity measurements were also con-
ducted in the presence of 1.0 and 5.0% water vapor, since
water is present in significant quantities (usually around
10% ) in the flue gas of different combustion processes and
has been found to affect the rate of other emission control
reactions over similar catalysts (33). Three different types
of behavior were observed during our experiments with 1%
water vapor. In the case of Cr2O3 (Fig. 3) and V2O5 cata-
lysts, the results indicate that the presence of water inhibits
The observed steady-state activity of the Co3O4/TiO2 cat-
alyst was significantly lower than the activity of the other
catalysts tested. This result appears to contradict results re-
ported previously by Aida et al. (27) and Subbanna et al.
(28), who indicated that supported cobalt oxide catalysts
are highly active for the oxidation of methylene chloride
and polychlorinated biphenyls (PCBs). Kießling et al. (29)
reported the formation of volatile CoCl2 during the oxida-
tion of CH2Cl2 over LaCoO3 and attributed an observed
low steady-state activity to the loss of Co via this route.
However, in our case, analysis of the spent Co3O4/TiO2 cat-
alyst does not render support to such a hypothesis under
FIG. 2. o-DCB conversion as a function of temperature for different
Al2O3-supported catalysts. (᭹) 4.5 wt% Cr2O3/Al2O3, (ᮀ) 5.6 wt% V2O5/
Al2O3, (᭡) 4.7 wt% Fe2O3/Al2O3, (♦) 4.7 wt% Co3O4/Al2O3, (+) 9.0 wt%
our reaction conditions (i.e., no change in the Co content MoO3/Al2O3, (᭿) Al2O3 (600 ppmv o-DCB, 10% O2, 25,000 h 1).

CATALYTIC OXIDATION OF 1,2-DICHLOROBENZENE
267
FIG. 5. Effect of water on the activity of a 4.5 wt% Co3O4/TiO2
catalyst for the oxidation of o-DCB. (ᮀ) 600 ppmv o-DCB, 10% O2;
(᭿) additional 1% H2O; and (᭹) additional 5% H2O.
FIG. 3. Effect of water on the activity of supported Cr2O3 catalysts for
the oxidation of o-DCB. (ᮀ, ᭿) 5.2 wt% Cr2O3/TiO2, (4, ᭡, ᭹) 4.5 wt%
Cr2O3/Al2O3 ((ᮀ, 4) 600 ppmv o-DCB, 10% O2; (᭿, ᭡) additional 1%
H2O; and (᭹) additional 5% H2O).
The overall observed effect of water depends on the rela-
tive importance of these two mechanisms. Cr2O3 and V2O5
catalysts are the most active ones under dry conditions,
which suggests that surface Cl removal may not be kinet-
ically significant in these cases. In addition, these catalysts
are active at lower temperatures, at which the adsorption
of water is an important issue. An increase in the water
concentration from 1 to 5% further reduces the activity of
Cr2O3/Al2O3 as can be observed in Fig. 3. This is consistent
with the competitive adsorption mechanism, which appears
to be significant under these conditions, explaining the ob-
served reduction in the activity.
In the case ofthe Fe2O3 and MoO3 catalysts, neither ofthe
two proposed mechanisms appears to be important. This is
further supported by the fact that no change in activity is
observed with an increase in the water concentration from
1 to 5% . The fact that these catalysts operate at higher tem-
peratures than Cr2O3 and V2O5 may explain the reduced
importance of water adsorption in these cases.
the oxidation of o-DCB. In contrast, in the case of MoO3
(Fig. 4) and Fe2O3 catalysts, aswellaswith Co3O4/Al2O3, the
reaction was not affected by the presence of water. Finally,
the activity of the Co3O4/TiO2 catalyst (Fig. 5) increased in
the presence of water.
Water can affect the rate of o-DCB oxidation in two dif-
ferent ways. Via a competitive adsorption mechanism it can
block a number of active sites, and hence, inhibit the rate
of the reaction. Our previous work with V2O5/TiO2 has in-
deed shown that at temperatures in the range of interest in
the current study, a significant amount of water is adsorbed
on the active vanadia sites (33). Water can also facilitate
the removal of Cl present on the catalyst surface by the
following reaction:
Cl + H2O → HCl + OH .
Finally, the increased activity of the Co3O4/TiO2 catalyst
can be attributed to the removal of surface Cl through its
reaction with water. An increase in the water concentration
from 1to 5% did not result in anyfurther increase in activity,
suggesting that most of the Cl is effectively removed from
the catalyst surface at the 1% water level. This hypothesis
is further supported by elemental analysis results obtained
with the spent catalysts tested both in the presence and in
the absence of water. These catalysts contained 0.25 and
1.20 wt% Cl , respectively.
Product selectivity. Results from previous studies on
the oxidation of chlorobenzenes suggest the formation of
by-products such as polychlorinated benzenes and poly-
chlorinated biphenyls over Pt-based and CuCl2-based cata-
lysts, respectively (9, 34). In contrast, when vanadia-based
catalysts were used for the oxidation of chlorinated
benzenes, CO and CO2 were the only carbon-containing
FIG. 4. Effect of water on the activity of supported MoO3 catalysts for
the oxidation of o-DCB. (ᮀ, ᭿, ᭹) 8.9 wt% MoO3/TiO2, (4, ᭡) 9.0 wt%
MoO3/Al2O3 ((ᮀ, 4) 600 ppmv o-DCB, 10% O2; (᭿, ᭡) additional 1%
H2O; and (᭹) additional 5% H2O).

268
KRISHNAMOORTHY, RIVAS, AND AMIRIDIS
products detected (10, 14). Similar results were obtained of a main feature at 1577 cm 1 with adjoining shoulders at
with Cr2O3-based catalysts for the oxidation of 1,2- approximately 1550 and 1612 cm 1. Also seen in all cases
1
dichlorobenzene (35), chlorobenzene (36), chloromethane is a peak at 1437 cm along with a doublet at 1389 and
(37), and trichloroethylene (38). During our current investi- 1376 cm 1, although the former appears to be very weak
gation CO and CO2 were the only carbon-containing prod- at the lower temperature. Additional peaks of low inten-
ucts detected through GC analysis. The ratio of the amounts sity are observed at 1455, 1360, 1322, and 1290 cm 1. Fi-
1
of CO to CO2 formed was approximately 30 : 70 in the case nally, a doublet is observed at 1234 and 1217 cm in the
of Fe2O3, 40 : 60 in the case of Co3O4 and V2O5, and 10 : spectra collected at temperatures below 623 K. These spec-
90 in the case of Cr2O3. From a practical standpoint, the tra are similar to the ones obtained with the V2O5/Al2O3
formation of CO or any other partial oxidation by-product catalyst in our previous studies (16). The strong peak at
1
does not represent a problem in this case, due to the very 1577 cm (assymmetric –COO stretching) and the dou-
1
low concentration of PCDD/PCDFs (ppb level) in the flue blet at 1389 (–CH bending) and 1376 cm (symmetric
gas of the incinerators and their extremely high toxicity –COO stretching) can be assigned to surface carboxy-
equivalence.
lates of the formate type, on the basis of previous stud-
Carbon balances in our study were closed to 5% . There ies of the adsorption of phenol on Cr2O3 (39), n-butane
was, however, a trace amount of organic residue formed on on MgCr2O4 (40), and dichloromethane on Al2O3 (41, 42).
the wall of the reactor. This was identified to be a poly- The shoulder at 1550 cm 1 (asymmetric –COO stretching)
chlorinated phenol, but apparently the selectivity toward along with peaks at 1437 (symmetric –COO stretching)
this compound is very low, and hence, its formation did not and 1360 cm 1 (–CH3 bending) can be assigned to carboxy-
significantly affect the closure of the carbon balances.
lates of the acetate type (40, 43–46). Both types of carboxy-
lates are formed on the transition metal site but could easily
migrate to the alumina support. As can be seen in the spec-
tra of Fig. 6, the formation of acetates becomes favored
over the formates at higher temperatures, due to the lower
stability of the latter. The total amount of surface carboxy-
In Situ FTIR Spectroscopic Studies
Dryconditions. In situ FTIR spectra ofthe Cr2O3/Al2O3
catalyst collected after 5 min on-stream at different temper-
atures are shown in Fig. 6. All the spectra show the presence
1
lates (as indicated by the broad feature at 1550–1580 cm
incorporating the main peaks of both acetate and formate
species) initially increases with temperature up to 623 K,
and then decreases at the higher temperature (i.e., 673 K).
This behavior is the result of the increased oxidation activ-
ity of the catalyst at elevated temperatures (which leads to
a higher formation rate of carboxylates) combined with the
lower stability of these species at the higher temperatures
(leading to their further oxidation to carbon oxides). The
result is the maximum in intensity of the carboxylate peaks
seen in Fig. 6.
1
The peak at 1322 cm has previously been assigned
to surface maleate species (47). The doublet at 1234 and
1217 cm 1 is characteristic of the C–O stretching vibrations
of phenates (47, 48), whereas the peak at approximately
1290 cm 1 has previously been assigned to a C–O stretch-
ing vibration of a phenolic group on Cr2O3 (39, 49). Charac-
teristic peaks of these phenate-type species were observed
only in spectra collected at lower temperatures, indicating a
higher reactivity of these species at elevated temperatures.
In fact, spectra collected at different time intervals indicate
that these species are also formed at the higher tempera-
tures but react quickly under these conditions.
Finally, no peaks corresponding to C–Cl vibrations were
observed in any of the spectra of Fig. 6, which has also
been the case for V2O5/Al2O3 (16). This indicates that Cl
abstraction is the first step in the activation of o-DCB. A
similar first step for the oxidation of chlorinated organics
has also been suggested by Hatje et al. (50) on the basis
of studies of the adsorption of chlorobenzene on Pt–Y and
FIG. 6. In situ FTIR spectra of the Cr2O3/Al2O3 catalyst collected
after 5 min on-stream at (a) 523 K, (b) 573 K, (c) 623 K, and (d) 673 K
(700 ppmv o-DCB, 5% O2, balance He).

CATALYTIC OXIDATION OF 1,2-DICHLOROBENZENE
269
Pd–Y catalysts. Finocchio et al. (51) have further suggested
that organic compounds are first activated at their weakest
point (i.e., the C–H bond for hydrocarbons and the C–Cl
bond for chlorinated hydrocarbons). All these results are
in agreement with our assessment that the aromatic ring
remains intact during the adsorption process and that the
resulting adsorbed species is probably bonded through the
dechlorinated carbon atoms.
Similar mechanisms have also been proposed by van den
Brink et al. (41) and Clet et al. (52) for the adsorption of
CH2Cl2 and CCl4, respectively, on Al2O3. According to van
den Brink et al., during the adsorption of CH2Cl2, a Cl
atom is displaced and the remaining group is attached to
the alumina surface. This is quickly followed by the ab-
straction of the second Cl atom. The Cl atoms can either
replace surface hydroxyl groups and react with the Al3+
ions to form aluminum chloride, or pick up protons to form
HCl. A similar mechanism has also been proposed by Clet
et al.
When the spectrum of the Cr2O3/Al2O3 sample was col-
lected at 573 K in the absence of O2 (Fig. 7), similar peaks
were observed as in the case where O2 was present, sug-
gesting that the adsorbed partially oxidized species can be
formed via reaction with surface oxygen. The only differ-
FIG. 8. (Left) In situ FTIR spectra of the Cr2O3/Al2O3 catalyst col-
lected in (a) 700 ppmv o-DCB, 5% O2 in He at 673 K after 30 min on-
stream, followed by flushing with O2 for (b) 1 min, (c) 5 min, (d) 20 min,
(e) 60 min, and (f) 90 min. (Right) In situ FTIR spectra of the Cr2O3/Al2O3
catalyst collected in (a) 700 ppmv o-DCB, 5% O2 in He at 673 K after
30 min on-stream, followed by flushing with He for (b) 1 min, (c) 5 min,
(d) 20 min, (e) 60 min, and (f) 90 min.
ence is that the intensity of most of the peaks is lower in the
absence of O2.
The stability/reactivity of the different adsorbed species
formed on Cr2O3/Al2O3 at 673 K was examined next. A
sample was initially exposed to the reacting gas mixture for
30 min, in order to allow for sufficient amounts of the differ-
ent surface species to be formed. At that point, the flow of
the reactants was stopped, and the sample was exposed to
either He or a 5% O2 in He mixture with spectra collected at
different time intervals. These spectra are shown in Fig. 8.
As can be seen in the figure, the intensities of the FTIR
peaks corresponding to the different surface species are re-
duced significantly regardless of whether He or 5% O2/He
is used. A reduction in intensity of the different peaks under
He flow indicates the relatively low stability of these species
at 673 K. Since the disappearance rate is more pronounced
in the presence of O2, we can conclude that all these surface
species react with O2 toward the final products, i.e., CO and
CO2. In addition, the spectrum collected after 120 min in
FIG. 7. In situ FTIR spectra of the Cr2O3/Al2O3 catalyst collected
after 5 min on-stream at 573 K. (a) 700 ppmv o-DCB, balance He;
(b) 700 ppmv o-DCB, 5% O2, balance He.

270
KRISHNAMOORTHY, RIVAS, AND AMIRIDIS
studied. The first step of this mechanism is the removal of
a Cl atom from the aromatic ring and its subsequent sub-
stitution with a surface oxygen, via what is probably a nu-
cleophilic mechanism. The removal of Cl is likely affected
by the nature of the support. The removal of the second Cl
atom and the formation of a second carbon–surface oxy-
gen bond follow the first step. This second step leads to the
formation of an adsorbed form of dihydroxybenzene. Since
no peaks corresponding to C–Cl vibrations were observed
in any of the spectra, the abstraction of the second Cl atom
is likely to be a fast step. The dihydroxybenzene species
can react further, undergoing ring cleavage and leading to
the formation of formates, acetates, and maleates. All of
these surface species react with oxygen to form CO, CO2,
and H2O. Our FTIR studies suggest that the rates of oxida-
tion of the acetates and formates are almost one order of
magnitude higher than that of the maleates.
Effect of water. In situ FTIR experiments were also per-
formed to study the effect of water on the concentration of
the different surface species. During these experiments, a
steady state was obtained first by flowing the reactants in
the absence of water over the catalyst for a certain period
FIG. 9. In situ FTIR spectra collected after 5 min on-stream at 673 K:
(a) Co3O4/Al2O3, (b) Fe2O3/Al2O3, (c) V2O5/Al2O3, and (d) Cr2O3/Al2O3
(700 ppmv o-DCB, 5% O2 in He).
O2 revealed the presence of a peak at 1378 cm 1, which was
found to be very stable on the catalyst surface. A similar
peak has previously been assigned to a coke deposit (53).
A comparison of spectra collected at 673 K after 5 min
on-stream for the different transition metal oxide catalysts
studied is attempted in Fig. 9. The positions of the IR peaks
corresponding to the different partially oxidized species
were almost identical in all these spectra. The only sig-
nificant difference is that the peak corresponding to the
symmetric –COO stretching of acetates was present at
1457 cm 1 for the V2O5, Fe2O3, and Co3O4 systems and was
shifted to 1437 cm 1 for the Cr2O3 system. This result may
be interpreted to suggest that the surface acetate species is
primarily associated with Cr2O3 in this system and with the
Al2O3 support in all the other cases. The doublet at 1236 and
1219 cm 1 corresponding to surface phenolates is observed
only in the case of the Co3O4/Al2O3 and Fe2O3/Al2O3 cata-
lysts, which can be attributed to the lower activity of these
catalysts at 673 K as compared to the other two transition
metal oxides. This doublet was also present in the spectra of
the other transition metal oxides, when these spectra were
collected at lower temperatures.
FIG. 10. In situ infrared spectra of the V2O5/Al2O3 catalyst collected
at 623 K in (a) 700 ppmv o-DCB and 5% O2 in He for 120 min, followed
by (b) 700 ppmv o-DCB, 1.4% water, and 5% O2 in He for 120 min, and
The spectra shown in Fig. 9 suggest that a similar mecha-
nism operates over all the supported transition metal oxides (c) 700 ppmv o-DCB, 5% O2 in He for 90 min.

CATALYTIC OXIDATION OF 1,2-DICHLOROBENZENE
271
of time. This was followed by the introduction of 1.4% wa- conditions. Furthermore, there was an apparent change in
ter vapor to the reaction stream. Spectra were collected at the distribution of surface formates and acetates, with the
different time intervals as the system approached a new former being significantly decreased in the presence of wa-
steady state. Subsequently the flow of water was stopped ter. Spectra collected following water removal from the gas
and a new set of spectra was collected. The choice of the phase (Fig. 11c) indicate the reestablishment of the pheno-
operating temperatures was such that all catalysts exhibited late peaks, and an increase in the amount of surface acetates
similar activity.
at the expense of the formates. A new peak was also ob-
Results obtained with V2O5/Al2O3 at 623 K are shown served at 1404 cm 1, which along with a shoulder at approx-
in Fig. 10. These results indicate a decrease in the concen- imately 1540 cm 1 (probably covered by the intense acetate
tration of the surface carboxylates in the presence of water. peak at 1550 cm 1) has previously been assigned to surface
This can be attributed to the formation of surface hydroxyls carbonates of the bidentate type (51). These changes point
on the vanadia sites, which partially prevents the formation to a potential irreversible change of the Cr2O3/Al2O3 sys-
of carboxylates, thus decreasing their surface concentra- tem upon exposure to water vapor.
tion. Such an explanation is consistent with the competitive
adsorption mechanism utilized to rationalize the observed
water effect on the catalyst activity. This effect is reversible,
CONCLUSIONS
since the removal of water from the reacting mixture leads
A systematic kinetic and in situ FTIR investigation of
to a new steady-state spectrum very similar to the one ob-
the oxidation of o-DCB has been carried out over sup-
tained prior to the introduction of water.
ported Cr2O3, V2O5, Fe2O3, MoO3, and Co3O4 catalysts.
Similar results were also observed with the other Al2O3-
Our activity measurements show that the transition met-
supported catalysts. In the case of the Cr2O3/Al2O3 catalyst
als are the active sites for this reaction. Cr2O3 and V2O5
at 523 K (Fig. 11), the presence of water resulted in the com-
catalysts exhibited the highest activity among the different
plete removal of the surface phenolates present under these
oxides tested. The activity was also affected by the nature
of the support (i.e., MxOy/TiO2 > MxOy/Al2O3), which may
indicate that the metal–oxygen–support bond is critical for
this reaction.
Our in situ FTIR studies indicate the presence of several
partialoxidation products(i.e., formates, acetates, maleates,
and phenolates) on the surface of the different transition
metal oxide catalysts examined. All these species can re-
act further to form CO, CO2, and H2O under certain re-
action conditions. Furthermore, all these partial oxidation
products can be formed even in the absence of gas-phase
oxygen, indicating that surface oxygen is involved in their
formation. No evidence could be found in any of the spectra
collected over the different catalysts for the presence of a
surface species containing a C–Cl bond, indicating that Cl
abstraction is the first step of the reaction. The presence of
the same species over all the catalysts studied indicates that
a similar mechanism operates in all cases.
The addition of water to the reaction stream had differ-
ent effects on the various catalysts. The activity of Cr2O3
and V2O5 catalysts was inhibited due to the competition
between o-DCB and water for adsorption onto the catalyst
sites. In contrast, the presence of water had a promoting ef-
fect on the Co3O4/TiO2 catalyst, which can be attributed to
the more efficient removal of Cl from the surface. Finally,
the addition of water had no significant effect on the activ-
ity of the MoO3 and Fe2O3 catalysts, indicating that neither
competitive adsorption nor the removal of surface Cl is
kinetically important in these cases. In situ FTIR studies
in the presence of water indicated a decrease in the con-
FIG. 11. In situ infrared spectra of the Cr2O3/Al2O3 catalyst collected
at 523 K in (a) 700 ppmv o-DCB and 5% O2 in He for 35 min, followed
centration of the different surface species, consistent with
a competitive adsorption effect.
by (b) 700 ppmv o-DCB, 1.4% water, and 5% O2 in He for 60 min, and
(c) 700 ppmv o-DCB, 5% O2 in He for 45 min.

272
KRISHNAMOORTHY, RIVAS, AND AMIRIDIS
ACKNOWLEDGMENTS 27. Aida, T., Higuchi, R., and Niiyama, H., Chem. Lett. 12, 2247
The authors acknowledge the experimental assistance of Mr. John
(1990).
28. Subbanna, P., Greene, H., and Desai, F., Environ. Sci. Technol. 22, 557
(1988).
Church and Mr. Brian Frushour, the thoughtful suggestions of Dr. Bert
Chandler, and the financial support of the National Science Foundation
(CTS- 9624433 and EEC-9732227).
29. Kießling, D., Schneider, R., Kraak, P., Haftendorn, M., and Wendt, G.,
Appl. Catal. B 19, 143 (1998).
30. Mori, K., Inomata, M., Miyamoto, A., and Murakami, Y., J. Chem.
Soc., Faraday Trans. 1 80, 2655 (1984).
REFERENCES
31. Mori, K., Miyamoto, A., and Murakami, Y., J. Chem. Soc., Faraday
Trans. 1 82, 13 (1986).
1. Johnson, J., Environ. Sci. Technol. 29, 24A (1995).
2. Hiraoka, M., Sakai, S., Sakagawa, T., and Hata, Y., Organohalogen
Compds. 31, 446 (1997).
32. Mori, K., Miyamoto, A., and Murakami, Y., J. Catal. 95, 482 (1985).
33. Amiridis, M. D., Wachs, I. E., Deo, G., Jehng, J., and Kim, D. S.,
J. Catal. 161, 247 (1996).
3. Olie, K., Vermeulen, P. L., and Hutzinger, O., Chemosphere 6, 445 34. Lago, R. M., Green, M. L. H., Tsang, S. C., and Odlyha, M., Appl.
(1977).
Catal. B 8, 107 (1996).
4. Hagenmaier, H., Tichaczek, K. H., and Brunner, H., Organohalogen
Compds. 3, 65 (1990).
35. Lester, G. R., Proc. 1989 Annu. AWMA Meeting, Anaheim 96, A
(1989).
5. Carlsson, K. B., Chemosphere 25, 135 (1992).
6. Ok, G., Hanai, Y., and Katou, T., Chemosphere 26, 2167 (1993).
36. Padilla, A. M., Corella, J., and Toledo, J. M., Appl. Catal. B 22, 107
(1999).
7. Freidel, I. M., Frost, A. C., Herbert, K. J., Meyer, F. S., and Summers, 37. Solymosi, F., Rasko, J., Papp, E., Oszko, A., and Bansagi, T., Appl.
J. C., Catal. Today 17, 367 (1993). Catal. A 131, 55 (1995).
8. Mendyka, B., Musialik-Piotrowska, A., and Syczewska, K., Catal. To- 38. Butt, J. B., Spivey, J. J., and Agarwal, S. K., in “Catalyst Deactivation”
day 11, 597 (1992).
(B. Delmont and G. F. Froment, Eds.). Stud. Surf. Sci. Catal. 88, 19
9. van den Brink, R. W., Louw, R., and Mulder, P., Appl. Catal. B 16, 219
(1994).
(1998).
39. Suprunov, V. E., and Ivanov, A. A., React. Kinet. Catal. Lett. 33, 75
10. Krishnamoorthy, S., Baker, J. P., and Amiridis, M. D., Catal. Today 40,
(1987).
39 (1998).
40. Busca, G., Catal. Today 27, 457 (1996).
11. Weber, R., Sakurai, T., and Hagenmaier, H., Appl. Catal. B 20, 249–256 41. van den Brink, R. W., Mulder, P., Louw, R., Sinquin, G., Petit, C., and
(1999).
Hindermann, J.-P., J. Catal. 180, 153 (1998).
42. Haber, J., Machej, T., Derewinski, M., Janik, R., Krysciak, J., and
Sadowska, H., Z. Phys. Chem. 197, 97 (1996).
43. Escribano, V. S., Busca, G., and Lorenzelli, V., J. Phys. Chem. 94, 8939
(1990).
12. Atalay, S., and Alpay, H. E., Ind. Eng. Chem. Res. 26, 2212 (1987).
13. Ide, Y., Kashiwabara, K., Mori, T., and Hara, M., Chemosphere 32, 189
(1996).
14. Jones, J., and Ross, J. R. H., Catal. Today 35, 97 (1997).
15. Hutchings, G. J., Heneghan, C. S., Hudson, I. D., and Taylor, S. H., 44. Chintawar, P. S., and Greene, H. L., J. Catal. 165, 12 (1997).
Nature 384, 341 (1996).
16. Krishnamoorthy, S., and Amiridis, M. D., Catal. Today 51, 203
(1999).
45. Kuznetsov, V. A., Geri, S. V., and Gorokhovat-skii, Y. B., Kinet. Katal.
18, 710 (1977).
46. Finocchio, E., Willey, R. J., Busca, G., and Lorenzelli, V., J. Chem. Soc.,
Faraday Trans. 93, 175 (1997).
17. Deo, G., and Wachs, I. E., J. Catal. 146, 323 (1994).
18. Wachs, I. E., Jehng, J.-M., Deo, G., Weckhuysen, B. M., Guliants, 47. Busca, G., Ramis, G., and Lorenzelli, V., “New Developments in Se-
V. V., Benziger, J. B., and Sundaresan, S., J. Catal. 170, 75 (1997).
19. Vuurman, M. A., and Wachs, I. E., J. Phys. Chem. 96, 5008 (1992).
20. Tung, H.-C., Yeh, C.-T., and Hong, C.-T., J. Catal. 122, 211 (1990).
lective Oxidation,” p. 825. Elsevier, Amsterdam, 1990.
48. Xu, B. Q., Yamaguchi, T., and Tanabe, K., Mater. Chem. Phys. 19, 291
(1988).
21. Wachs, I. E., Deo, G., Vuurman, M. A., Kim, D. S., and Hu, H., in “Pro- 49. Miyata, H., Ohno, T., and Hatayama, F., J. Chem. Soc., Faraday Trans.
ceedings, 10th International Congress on Catalysis, Budapest, 1992” 91, 3505 (1995).
(L. Guczi, F. Solymosi, and P. Tetenyi, Eds.), p. 71. Akad| miai Kiad, 50. Hatje, U., Hagelstein, M., and Forster, H., in “Zeolites and Re-
Budapest, 1993.
lated Microporous Materials: State of the Art 1994” (J. Weitkamp,
H. G. Karge, H. Pfeifer, and W. Holderich, Eds.), Vol. 84, Elsevier,
Amsterdam, 1995.
22. Kim, D. S., Tatibouet, J. M., and Wachs, I. E., J. Catal. 136, 209 (1992).
23. Kim, D. S., and Wachs, I. E., J. Catal. 142, 166 (1993).
24. Hu, H., and Wachs, I. E., J. Phys. Chem. 99, 10911 (1995).
25. Ramani, N. C., Sullivan, D. L., Ekerdt, J. G., Jehng, J.-M., and Wachs,
I. E., J. Catal. 176, 143 (1998).
51. Finocchio, E., Busca, G., Lorenzelli, V., and Willey, R. J., J. Catal. 151,
204 (1995).
52. Clet, G., Goupil, J.-M., and Cornet, D., Bull. Soc. Chim. Fr. 134, 223
(1997).
26. Larrson, P., Andersson, A., Wallenberg, L. R., and Svensson, B.,
J. Catal. 163, 279 (1996).
53. Datka, J., Sarbak, Z., and Eischens, R. P., J. Catal. 145, 544 (1994).