Radiation induced catalytic dechlorination of hexachlorobenzene on oxide surfaces

Article abstract of DOI:10.1021/jp010386f

Radiation induced catalytic dehalogenation of hexachlorobenzene (HCB) adsorbed to alumina (Al₂O₃), silica (SiO₂), titania (TiO₂), zirconia (ZrO₂), and a commercially available zeolite has been studied using Cobalt-60 (⁶⁰Co) as a radiation source. Solid-particulate samples were irradiated over a dose range of 0-58 kGy, and the chemical changes were monitored using Fourier transform infrared diffuse reflectance (FTIR-DR) and gas chromatography with electron capture detection (GC-ECD). The extent of HCB degradation on the metal oxides was found to increase dramatically in samples evacuated under vacuum, pointing to the competitive scavenging of conduction band electrons by surface adsorbed species, primarily oxygen. Coadsorbed water diminished HCB conversion on all oxides but to a greater degree on alumina. HCB degradation on metal oxides was found to be highly dependent upon the conduction band energy of the support material, thus confirming the occurrence of ultra-band-gap excitation and charge separation in irradiated oxides. Higher yields of dechlorination products were witnessed in alumina and silica samples. Zeolite, titania, and zirconia were also found to be inefficient in promoting radiation induced catalysis. The absence of oxidation products in the irradiated HCB/oxide samples suggests the inaccessibility of holes to undergo interfacial charge transfer with the organic substrate.

Full text of DOI:10.1021/jp010386f

J. Phys. Chem. B 2001, 105, 4715-4720
4715
Radiation Induced Catalytic Dechlorination of Hexachlorobenzene on Oxide Surfaces
George Adam Zacheis,† Kimberly A. Gray,*, † and Prashant V. Kamat*, ‡
Department of CiVil Engineering, 2145 Sheridan Road, Northwestern UniVersity, EVanston Illinois 60208-3109,
and Notre Dame Radiation Laboratory, Notre Dame Indiana 46556-0579
ReceiVed: January 31, 2001
Radiation induced catalytic dehalogenation of hexachlorobenzene (HCB) adsorbed to alumina (Al2O3), silica
(SiO2), titania (TiO2), zirconia (ZrO2), and a commercially available zeolite has been studied using Cobalt-60
(60Co) as a radiation source. Solid-particulate samples were irradiated over a dose range of 0-58 kGy, and
the chemical changes were monitored using Fourier transform infrared diffuse reflectance (FTIR-DR) and
gas chromatography with electron capture detection (GC-ECD). The extent of HCB degradation on the metal
oxides was found to increase dramatically in samples evacuated under vacuum, pointing to the competitive
scavenging of conduction band electrons by surface adsorbed species, primarily oxygen. Coadsorbed water
diminished HCB conversion on all oxides but to a greater degree on alumina. HCB degradation on metal
oxides was found to be highly dependent upon the conduction band energy of the support material, thus
confirming the occurrence of ultra-band-gap excitation and charge separation in irradiated oxides. Higher
yields of dechlorination products were witnessed in alumina and silica samples. Zeolite, titania, and zirconia
were also found to be inefficient in promoting radiation induced catalysis. The absence of oxidation products
in the irradiated HCB/oxide samples suggests the inaccessibility of holes to undergo interfacial charge transfer
with the organic substrate.
Introduction
particles results in charge separation. These charge carriers, in
the conduction and valence bands, recombine or migrate to the
The radiolytic degradation of HCB in contaminated soils
occurs at higher rates and to a greater extent in low organic,
high mineral content soils than in soils having high organic
content.1 To probe the nature of the specific reaction of
contaminants adsorbed to natural mineral soils under gamma-
radiolysis, alumina was selected as a model soil mineral and
the radiolytic transformation of HCB adsorbed to alumina was
followed using diffuse reflectance FTIR and GC-ECD. Because
alumina has a very large band-gap with a conduction band close
to the vacuum level,2 it can only be activated with high-energy
radiation. The results of these experiments showed the occur-
rence of radiation-induced, ultra band-gap charge separation,
followed by the interfacial charge transfer and degradation of
adsorbed HCB.3
Earlier radiolytic studies with silica and zeolites have shown
that ionization events in these solids are similar to those observed
in nonpolar media.4-6 Gamma rays are energetic enough to
create electron-hole pairs in wide band gap oxide particles,
most of which recombine within first few picoseconds to
produce triplet excitons.5 A fraction of these charge carriers are
also trapped at defect sites. There have been numerous studies
using ionizing radiation to produce excitations in metal oxide
particles with subsequent charge transfer to adsorbed mole-
cules.3,7-10 For systems of solids with coadsorbed molecules,
high-energy radiation will generally excite the solid producing
ionization in remote regions of the solid and result in subsequent
transfer of excitation energy to the adsorbate.7,11
particle surface, where they become trapped and/or may
participate in the interfacial oxidation and reduction of adsorbed
species.12 In general, the radiolysis of solids with adsorbate
produces ionic products and free radicals derived from these
ions. In photocatalysis, one employs semiconducting materials
with band-gaps less than 3.5 eV. However, in radiation induced
catalysis, one can use relatively large band-gap materials
(insulators), thus paving the way to use oxides that are present
in greater proportions in natural soils. Radiolysis of EDTA
solutions containing titania colloids led to a 30% increase in
EDTA degradation.13 In addition, this phenomenon may play a
role in the radiolytic processes occurring in radioactive tank
wastes containing large amounts of metal oxides and organic
compounds at US Department of Energy (DOE) sites.14 Finally,
because gamma rays have large penetration depths in compari-
son to UV radiation, it may be possible to engineer radiation
activated catalysts having high surface areas and redox properties
to treat the most recalcitrant pollutants.
A basic understanding of the catalytic properties of repre-
sentative oxides is therefore useful for developing radiolytic
treatment strategies for contaminated soils. We have now carried
out steady-state radiolysis of HCB adsorbed on to silica,
alumina, titania, zirconia, and zeolite in order to establish the
role of different oxide support in catalyzing HCB transformation.
Experimental Section
The principle of radiation induced catalysis is similar to that
of photocatalysis in which ultraband-gap excitation of oxide
Sample Preparation. Fumed alumina and titania samples
were obtained from Degussa (Akron, OH), HCB and powdered
zirconia from Aldrich (Milwaukee, WI), and fumed silica and
powdered zeolite from Sigma (St. Louis, MO). These chemicals
were of the highest purity available and were used as received
* To whom correspondence should be sent.
† Northwestern University (Tel 847-467-4252; Fax 847-491-4011; E-mail
k-gray@nwu.edu).
‡ University of Notre Dame (Tel 219-631-5411; Fax 219-631-8068;
E-mail PKamat@nd.edu or http:/www.nd.edu/∼pkamat).
10.1021/jp010386f CCC: $20.00 © 2001 American Chemical Society
Published on Web 04/27/2001

4716 J. Phys. Chem. B, Vol. 105, No. 20, 2001
Zacheis et al.
TABLE 1: Surface Areas of Supports
support
surface area [m2/g]
Al2O3
SiO2
TiO2
ZrO2
zeolite
100a
255a
50a
50b
200b
a Measured values. b Manufacturer reported.
without further purification. In preparing metal oxide samples,
a desired mass of HCB was added to a given mass of metal
oxide and then thoroughly mixed with acetone and hexane (1:1
v/v). Solvents were then evaporated under nitrogen until the
metal oxide solution was reduced to a paste. Subsequently,
pastes were allowed to dry in an oven overnight (50 °C). All
oxide samples were prepared via the addition of HCB on a mass
to mass basis with a normalized amount of 15 mg HCB per
total surface area of 100 m2 to account for differences in the
surface areas of various supports. Unless otherwise specified
all supports were used as received.
Table 1 above reports the surface area of all the oxide supports
used in the present investigation. In the case of zeolite and
zirconium oxide, surface areas were determined from methylene
blue adsorption experiments.15 This was accomplished through
the Langmuir adsorption isotherm by adding a known amount
of methylene blue solution to various amounts of the support.
The amount of methylene blue required to achieve monolayer
coverage of the support was estimated for determining the
surface area of the support.
Individual vials containing approximately 1.0 g of solid
sample were irradiated to produce samples for GC measure-
ments. These cells were prepared and irradiated under the same
conditions as the FTIR samples. A series of experiments were
performed in the absence of oxygen (under vacuum) and in a
water-saturated atmosphere. Oxygen was removed from the
sample by placing the vial or the holder in a glass cell and
connecting it to vacuum line for 15 min prior to irradiation.
The evacuated cell assembly was kept in the irradiator for the
desired time of exposure. Because four dose periods were used,
a given sample was repeatedly (5 times total) evacuated by
vacuum after FTIR analysis at each dose interval. GC-ECD (gas
chromatograph with Electron Capture Detection) results con-
firmed that no significant HCB desorption from the oxide
samples occurred with the application of the vacuum conditions.
For oxides irradiated under a water-saturated atmosphere,
samples were initially exposed (30 min) to a water-saturated,
nitrogen gas stream within a large glass cell (150 mL).
Gamma Radiolysis. Radiolysis of all samples was performed
using a Shepard-109 60Co source. 60Co is an emitter of high
energy, 1.25 MeV, photons (γ-rays). The Shepard-109 is a
concentric well type source with a measured dose rate of
approximately 243 Gy/min at the time of experimentation. Dose
rates were calculated in our laboratory using Fricke dosimetry.16
FTIR samples were irradiated within diffuse reflectance
sampling cups to prevent variation in the location of samples
with respect to the FTIR infrared beam. For vacuum samples,
a special glass irradiation cell was constructed to hold FTIR
sampling cups. Spectra of individual samples were recorded
before and after radiolysis at different time period of irradiation.
Diffuse Reflectance Fourier Transform Infrared (FTIR)
Measurements. All FTIR experiments were performed using
a Bio-Rad FTS 175, FTIR spectrometer with a diffuse reflec-
tance accessory. Samples were placed within micro sampling
cups for analysis inside the instruments and nitrogen was purged
Figure 1. FTIR-DR spectra of HCB on silica (38.25 mg HCB/g) after
(a) 0, (b) 1, and (c) 3 h of γ-radiation and under a vacuum (-760
mmHg).
through the chamber to prevent interference from CO2. The
diffuse reflectance spectra of all samples were measured in the
region 400-4000 cm-1 at a resolution of 8 cm-1. No further
enhancement in spectral features was observed when the
resolution was increased from 8 to 4 cm-1. The measurements
were limited to an analytical window of 2000-1200 cm-1
.
Extraction and Analysis. All samples were analyzed by
conventional GC techniques for byproduct analysis and quan-
tification. Compounds bound to metal oxides were extracted in
a solution of acetone and hexane 1:1(v/v). An internal standard,
1,3,5-tribromobenzene, was added to adjust responses for
changes in volume. This compound was used due to its structural
similarity to HCB and high response in electron capture detectors
(ECD). After internal standard addition, samples were sonicated
for 15 s to enhance desorption of HCB from the oxide surface.
Resulting suspensions were then centrifuged for 3 min at 890
× g to remove oxide particles from suspension, thus producing
a clean sample for injection into the GC interface.
A method for the detection and quantification of HCB and
reduced forms of HCB using GC-ECD was established in
previous research.1 Identification of products was verified
through the use of a Hewlett-Packard 6890 GC-MS. Mass
spectra were compared to NIST databases for tentative identi-
fication and confirmed with known standards. For GC-MS work,
a J&W Scientific DB-5 column was used with oven program-
ming being identical to that used in GC-ECD work.
Results
60Co Irradiation of HCB/Oxides. In Figure 1, the FTIR
spectrum of HCB adsorbed onto silica (Spectrum a) is compared
to the spectra of irradiated silica/HCB samples (Spectra b and
c) under vacuum. Prior to irradiation (Spectrum a), samples
showed two peaks at 1343 cm-1 and 1290 cm-1, which can be
attributed to the stretching of aromatic CdC bonds in the HCB
molecule.17 This spectrum also showed a weak region (1600-
1450 cm-1) of overtone bands from C-Cl bending bands.17 The
region from 1580 to 1500 cm-1 is often attributed to commonly
adsorbed hydrogen carbonate ions on the surface of alumina
particles.18 Such bands are the result of asymmetric stretching
2-
modes of the CO3 ion.19-21 However, it is important to note
that because C-Cl overtone bands may also exist in this region,
it is not possible to resolve these bands in the present data.
Upon irradiation for an hour (dose of 14.6 kGy) (Figure 1;
Spectrum b), the aromatic CdC stretch band found at 1343 cm-1
is slightly broadened, decreased in intensity, and formed a

Hexachlorobenzene on Oxide Surfaces
J. Phys. Chem. B, Vol. 105, No. 20, 2001 4717
Figure 2. Formation of PeCB on high coverage (a) alumina (9), (b)
silica (b), (c) titania (2), (d) zeolite (1), and (e) zirconia ([) as a
function of time (dose) under an open atmosphere.
Figure 3. Formation of PeCB on high coverage (a) alumina (9), (b)
silica (b), (c) zeolite (1), and (d) zirconia ([) as a function of time
(dose) under a vacuum.
doublet at 1347 and 1332 cm-1. In addition, other aromatic Cd
C stretch band at 1290 cm-1 broadened and decreased in
intensity as well. Furthermore, the formation of a conglomerate
of peaks at 1447, 1429, and 1411 cm-1 could also be seen.
Earlier research in our laboratory with HCB adsorbed to alumina
has shown that these changes in the FTIR spectra arise from
the dechlorinated forms of HCB, primarily pentachlorobenzene
(PeCB) and tetrachlorobenzene (TeCB) isomers.3 The large
conglomerate at ∼1429 cm-1 results from the constructive
interference of peaks related to the presence of dechlorination
byproducts, which have numerous contributing peaks in the
region from 1480 to 1300 cm-1. In a similar manner, changes
occurring in the aromatic CdC bands at 1343 and 1290 cm-1
are consistent with the formation of PeCB on the oxide surface.
Finally, the peak at 1699 cm-1 is attributed to the CdO
stretching of residual acetone used in sample preparation.
Inspection of the 3-h spectrum (Figure 1 Spectrum c) showed
further broadening and decreased intensity of the aromatic
stretch bands at 1343 cm-1 and 1290 cm-1. In addition, the
doublet at 1347 cm-1 and 1332 cm-1 became distinct and
separated after 3 h of irradiation. With increasing radiation dose,
we expect to observe an increased amount of transformation of
HCB on silica surface. Dechlorination byproduct peaks at 1447,
1429, and 1411 cm-1 increased slightly and became more
pronounced. However, little change in the peaks at 1699 cm-1
and 1627 cm-1 was witnessed when compared to the 1 h
irradiated spectrum (Figure 1 Spectrum b).
FTIR spectra obtained for all 5 HCB coated oxide samples
before and after irradiation under an open atmosphere are
included in the Supplementary Information (Figure 1). The 4 h
irradiated samples (58.3 kGy) showed spectral changes similar
to those found in Figure 1 for HCB coated silica. In the case of
zirconia/HCB samples, very little degradation in comparison
to the other surfaces was observed. Even at high irradiation doses
(160.6 kGy), we could not observe any significant transforma-
tions in these samples.
Dechlorination of HCB on Oxides under an Open Atmo-
sphere. In Figure 2, the production of PeCB (extracted and then
quantified by GC-ECD) in samples irradiated under an open
atmosphere is shown. Results are expressed as mg PeCB per
kg oxide (ppm). Overall, the production of PeCB found in the
3 and 4 h irradiated samples was greater in alumina and silica
samples. Silica showed greater overall HCB degradation,
approximately 2-3 times greater than that on alumina. In
contrast, virtually no HCB conversion was found in titania (a
negligible 1.5 ppm conversion following 4 h of irradiation) or
zirconia samples. Likewise, zeolite samples showed only a
marginal HCB conversion. In alumina and silica samples, trace
levels of TeCB were also detected pointing to an advanced
dechlorination step in these samples.
Dechlorination of HCB on Oxides under Evacuated
Conditions. The irradiation of silica and alumina samples under
vacuum (10-3 Torr) resulted in very large increases in PeCB
formation (Figure 3). As compared to the open atmosphere
irradiation, the production of PeCB in evacuated alumina
samples increased by more than an order of magnitude (18×).
Similar enhancement was also observed in evacuated silica/HCB
samples. Evacuation of these samples led to 1.6 times more
degradation on alumina than on silica, a reversal of the results
under an open atmosphere (Figure 2). Although the extent of
PeCB formation on ziolite and zirconia increased by small
amount (2.35 and 6.41 times, respectively) in comparison to
the open atmosphere condition, overall, very limited HCB
conversion was observed relative to that on silica or alumina.
If the adsorption of surface species such as oxygen or water
was the only factor responsible for slower degradation in open
atmosphere irradiated samples, we would have seen a similar
increased degradation efficiency for all supports. Thus, the
choice of oxide as a catalytic support is important in controlling
the radiation induced degradation of HCB.
Elevated levels of HCB degradation under vacuum resulted
in a likewise increase of TeCB production (Figure 4), for
alumina and silica samples. Formation of TeCB was found to
increase rapidly as a function of time. Because of limitations
in GC separation, only one TeCB isomer (1,2,3,4-TeCB) was
quantifiable due to the coelution of 1,2,3,5-TeCB and 1,2,4,5-
TeCB.
Effect of Moisture on HCB Dechlorination. The irradiation
of samples under a water-saturated atmosphere was performed
to determine the effect of surface adsorbed water on HCB
degradation. Prior to irradiation, a water-saturated gas stream
(nitrogen) was passed through a cell containing the oxide
samples for about an hour. Surfaces irradiated under a water-
saturated atmosphere (Figure 5) showed an inhibitive effect
toward to HCB reduction. Unlike the open atmosphere or
evacuated conditions very little degradation occurred at low dose
periods. At higher doses, the degradation differed somewhat
from the trends observed in Figures 2 and 3. As witnessed in
the open atmosphere experiments, HCB degradation in silica
samples was better than other supports. For 4 h irradiated silica/
HCB and alumina/HCB samples, the extent of degradation in

4718 J. Phys. Chem. B, Vol. 105, No. 20, 2001
Zacheis et al.
effective for HCB degradation than the other oxides when the
experiment was carried out under ambient or water saturated
irradiation conditions. In general, the irradiation of samples
under an open atmosphere showed slightly greater PeCB
production on all suports when compared to water saturated
samples. In particular, alumina samples showed the greatest
increase (4.4 ×) in PeCB production when moving from a water
saturated to open atmosphere. (2.0 × 10-3 vs 8.8 × 10-3
,
respectively).
Discussion
High-energy radiation is known to interact directly with the
solid or liquid substrate and induce ionization events. The cross
section for absorption of ionizing radiation is directly propor-
tional to the electron density of the irradiated material. In the
present experiments, the oxide supports constitute the major
fraction (>99%) of the irradiated material. The amount of energy
absorbed directly by HCB in a oxide/HCB sample is negligible
because the concentration and electron density of the oxide
support dominate the irradiated sample. Hence, the incident
gamma radiation is deposited into oxide support via electronic
interactions. This argument is supported by the observation that
negligible HCB transformations occurred in earlier research on
KBr3 and on TiO2 in the present research when coated with
HCB. (Note that these supports do not induce catalytic
transformation of HCB and hence serve as reference.)
Figure 4. Formation of 1,2,3,4-TeCB on high coverage (a) alumina
(9), (b) silica (b), (c) zeolite (1), and (d) zirconia ([) as a function of
time (dose) under a vacuum.
The oxide supports in the present experiments directly absorb
the incident γ-rays and induce charge separation within the oxide
material. The electrons produced in the oxide particle are quickly
thermalized to conduction band energy level or get trapped at
the surface vacancies. Triplet excitons and trapped electrons
can also react with surface bound OH groups to generate H
atoms via radiolytic homolysis of the hydroxyl group (O-H
bond) and dissociative electron attachment to silanol groups.4,5
Surface adsorbed aromatic species may thus interact with the
conduction band electrons or H atoms to undergo reductive
dechlorination. In experiments where silica was pretreated at
higher temperature to drive off surface bound water and
hydroxide ions, subsequent irradiation produced a decreased
amount of ionized products of the adsorbed aromatic compounds
suggesting the necessity of chemisorbed water on the oxide
surface.5 Holes, on the other hand, get trapped or react with H
atoms and surface bound hydroxyl groups to form surface bound
protons and hydroxyl radicals, respectively. On the basis of the
Figure 5. Formation of PeCB on high coverage (a) alumina (9), (b)
silica (b), (c) zeolite (1), and (d) zirconia ([) as a function of time
(dose) under a water saturated atmosphere.
comparison to the open atmosphere irradiation experiments,
decreased by a factor of 4-5. Other oxides also showed similar
decreased HCB degradation efficiency. These results show that
the coadsorbed water on the oxide surface has an inhibitive
effect on the HCB degradation.
Yields of HCB Byproducts. The radiolytic transformations
of solutes can be quantified in terms of G values, defined as
the number of molecules formed or lost per 100 eV of energy
absorbed.22 G values allow for the direct comparison of
pentachlorobenzene (PeCB) yields, at a given dose, between
samples of different supports under different atmospheric
conditions. Table 2 summarizes the G values, or yields, of PeCB
when samples are irradiated under an open atmosphere, water-
saturated nitrogen atmosphere and in a vacuum for 4 h (58.2
kGy). G values are also reported for the formation of 1,2,3,4-
tetrachlorobenzene (TeCB) in a vacuum samples.
G values of dechlorinated products (viz., PeCB and TeCB)
indicate that the surface adsorbed species (e.g., O2 or H2O)
influence the course of HCB degradation on irradiated oxide
samples. Exclusion of these species by evacuation resulted in
maximizing the efficiency of catalytic degradation of HCB on
oxide surfaces. Almost 2 orders of magnitude difference exist
between the G values for samples irradiated on the surface of
alumina under a water-saturated atmosphere and vacuum.
Similarly, large increases were observed in silica (×21) and
zirconia (×16) samples. However, the increase in values of G
for zeolite samples remained relatively small. Silica was more
information above and on HCB reaction on alumina,3
a
mechanism of surface promoted degradation of an adsorbed
contaminant is proposed in Scheme 1.
The degradation process in Scheme 1 is affected by the
presence of surface adsorbed oxygen and excess water. Because
oxygen is a good electron scavenger,23,24 it competes with HCB
for the conduction band electrons. The retarding effect of these
surface adsorbed species can be overcome by carrying out the
irradiation under vacuum. These observations further ascertain
the domination of a reductive process in the dechlorination of
HCB.
The presence of moisture on the surfaces of silica and alumina
clearly had profound effects on the degradation of HCB. Past
work has found that surface adsorbed moisture on titania led to
enhanced electron/hole recombination.25 Given the extreme
hydrophobicity of HCB, we may expect a certain degree of HCB
displacement from oxide surfaces by water molecules, breaking
the monolayer arrangement of HCB and thus, decreasing the
HCB/surface interactions. Indeed, water addition to silica and
alumina has been shown, by NMR studies, to result in an

Hexachlorobenzene on Oxide Surfaces
J. Phys. Chem. B, Vol. 105, No. 20, 2001 4719
TABLE 2: G Values (× 10-3) of Dechlorinated Products Following 4-Hour Irradiation (58.2 kGy)a
experimental
condition
product
alumina
σAl
silica
σSi
zirconia
σZr
zeolite
σZeo
water sat
open atmosphere
vacuum
PeCB
PeCB
PeCB
TeCB
2.0
8.8
176.0
6.0
0.06
0.27
5.5
5.3
20.0
110.9
4.0
0.16
0.63
3.4
0.58
1.8
9.4
0.018
.056
0.29
1.8
2.6
15.0
0.62
0.055
0.081
0.46
vacuum
0.2
0.13
0.21
0.0065
0.020
a Standard deviation (σ) values for average G values are presented. Furthermore, titania/HCB samples irradiated in open atmosphere have a very
low G value of conversion (2.6 × 10-4).
SCHEME 1: Radiation Induced Processes in Large
Band-Gap Oxides
SCHEME 2: Band Edge Positions for Metal Oxides
The band-gaps of both alumina and silica are larger than those
for titania and zirconia. The incident gamma rays are sufficiently
energetic enough to induce charge separation in all these oxide
materials. However, the difference from the reducing property
is evident from the difference in the conduction band energy
levels of these oxides. In contrast to TiO2 and ZrO2, the
conduction band energies of alumina and silica are highly
negative (∼-4.2 V vs NHE). Because the reduction potential
of HCB is -1.2 V vs NHE,38 only silica and alumina substrates
are capable of reducing HCB effectively. Indeed, this relation-
ship between the reduction potential of HCB and the conduction
band energies of the metal oxides is consistent with our
observation of higher yields of dechlorinated products in
irradiated alumina and silica samples. The titania and zirconia
remained ineffective catalysts to induce transformations of HCB.
The energy levels of oxides and the reduction potential of HCB
that illustrate the two different scenarios are summarized in
Scheme 2.
TABLE 3: Band Edge Properties of Metal Oxides
metal oxide band-gap [eV] EVB [V vs NHE] ECB [V vs NHE]
Al2O3
SiO2
TiO2
ZrO2
8.45-9.9a,b
8-8.9a,b,c,f
3.0-3.3a,d
5.0e
4.09b
4.26f
2.97d
4.0e
-4.21b
-4.24f
-0.17d
-1.0e
a Ref 33. b Ref 2. c Ref 34. d Ref 35. e Ref 36. f Ref 37.
increase in the mobility of benzenoid compounds when adsorbed
to these surfaces.26-28 Water adsorption leads to the blocking
of strong adsorption sites, otherwise available for the adsorption
of organics.26 The results shown in Figures 3-5 suggest that
the displacement of HCB by water to be greater on alumina
than on silica.
The water adsorption capacity of metal oxides is influenced
by such factors as dehydration conditions, pore size distribution,
and surface area.29 Because the adsorption of aromatic com-
pounds such as HCB to the surfaces of alumina and silica occurs
via physisorption,26,28 the surface interactions are readily affected
by water adsorption. Past research with chlorinated benzene
compounds in alumina and silica solutions showed a greater
degree of adsorption of these chemicals on silica than
alumina.30-32 In fact, the disparity in organic adsorption to
alumina and silica increased upon chlorine substitution, a result
of increased dipole interactions between the aromatic ring and
the surface of the metal oxides.30 These adsorption experiments
are consistent with the results in the present research, which
demonstrate decreased HCB degradation, presumably due to
decreased adsorption, on the surface of alumina in the presence
of water.
Zeolites are Si-Al-O oxides which, upon irradiation, would
also lead to the production of trapped holes and electrons.4,39
These trapped holes are stabilized within the zeolite or react
with surface absorbed organics. Low reactivity of zeolite in
inducing reductive dechlorination shows that the electrons and
holes that are produced during gamma irradiation undergo quick
recombination or the electrons are scavenged by impurities or
traps present in the zeolite framework. Trapping of electrons
has also been shown to lead to color centers in zeolites,40
a
phenomena witnessed in our zeolite samples. As shown earlier,
the cation clusters play an important role in the stabilization of
electrons in dehydrated zeolites.39,41 The competition between
cation stabilization and electron hydration agrees with the
decrease in HCB degradation on zeolites under a water-saturated
atmosphere compared to samples under an open atmosphere.
Surprisingly, oxidative processes were not discovered on any
of the oxide supports employed in the present investigation.
Extractions of the irradiated oxide/HCB samples using polar/
nonpolar solvent mixtures were conducted to probe for the
formation of HCB oxidation byproducts such as phenol,
chlorophenol, dichlorobenzoquinones, and dichlorohydroquino-
nes. However, GC-ECD and GC-MS analysis of the samples
irradiated for extended time periods (4 h) did not indicate even
trace levels of oxidation products. Even the extended extraction
An important observation in the present study is our
demonstration that the HCB degradation on metal oxides is
dependent upon the energetics of the conduction band of the
metal oxide, especially in evacuated samples. Information
concerning the electronic properties (Table 3) of metal oxides
provides further credence to our argument that the oxides are
capable of catalyzing the radiation induced transformation of
adsorbed substrates.

4720 J. Phys. Chem. B, Vol. 105, No. 20, 2001
Zacheis et al.
(6) Zhang, G. H.; Mao, Y.; Thomas, J. K. J. Phys. Chem. B 1997,
101, 7100-7113.
procedures (solvent mixtures with varying polarity, sonication,
and thermal desorption) did not yield any oxidation products.
Although it is surprising to see no direct evidence for a
radiation induced oxidation process on oxide surfaces, such a
phenomenon is not unprecedented. Schatz et al.37 reported that
the absorption of high-energy electrons by silica suspensions,
led to electron ejection from silica as secondary electrons. They
could observe only the reduction process at the solid/water
interface and not the oxidative process. The absence of
interfacial charge-transfer involving holes at the solid/water
interface was indicative of the fact that the holes get trapped
within the silica particle.
(7) Caffrey, J. J. M.; Allen, A. O. J. Phys. Chem. 1958, 62, 33-37.
(8) Wong, P. K.; Willard, J. E. J. Phys. Chem. 1969, 73, 2226.
(9) Wong, P. K.; Allen, A. O. J. Phys. Chem. 1970, 74, 774-778.
(10) Hentz, R. R.; Perkey, L. M.; Williams, R. H. J. Phys. Chem. 1966,
70, 731-735.
(11) Caulfield, K. J.; Cooper, R.; Boas, J. F. J. Chem. Phys. 1990, 92,
6441-6446.
(12) Linsebigler, A. L.; Lu, G.; Yates, J. T. Chem. ReV. 1995, 95, 735-
758.
(13) Su, Y.; Wang, Y.; Daschbach, J. L.; Fryberger, T. B.; Henderson,
M. A.; Janata, J.; Peden, C. H. F. J. AdV. Oxid. Technol. 1998, 3, 63-69.
(14) Gerber, M. A.; Burger, L. L.; Nelson, D. A.; Ryan, J. L.; Zollars,
R. L. Assessment of Concentration Mechanisms for Organic Wastes in
Underground Storage Tanks at Hanford; Pacific Northwest National
Laboratories: 1992.
Conclusions
(15) Kipling, J. J.; Wilson, R. B. J. Appl. Chem. 1960, 10, 109-113.
(16) Das, S., personal communication.
The results described in this study show the ability of oxides
to act as catalysts in the radiolytic degradation of organic
compounds. Only reductive pathways involving dechlorination
of HCB occurred on oxide surfaces. The degradation yield is
directly influenced by the intrinsic properties of the oxides.
Alumina and silica with conduction band energy more negative
than the reduction potential of HCB were most effective in
inducing the reductive dechlorination. The present study high-
lights the merits of radiolysis in treating recalcitrant contami-
nants such as chlorinated benzene using metal oxides as
catalysts. After the completion of the dechlorination step, one
can consider a bioremediation or an advanced oxidation process
to achieve complete mineralization. The radiation-induced
catalysis may thus be useful as a pretreatment process for
contaminated soils.
(17) Nakanishi, K.; Solomon, P. H. Infrared Absorption Spectrometry;
Holden-Day, Inc.: San Francisco, 1977.
(18) Lee, S.; Bray, P. J. Phys. Chem. Glasses 1962, 3, 37-42.
(19) Frueh, A. J.; Golightly, J. P. Can. Mineral 1967, 9, 51.
(20) Serna, C. J.; White, J. L.; Hem, S. L. Clays Clay Mineral. 1977,
25, 384.
(21) Garcia, J. V.; Ramas, C. J. N. Jb. Mineral. Mh. 1987, 397.
(22) Klassen, N. V. Primary Products in Radiation Chemistry; VCH
Publishers: New York, 1987; Chapter 2.
(23) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W.
Chem. ReV. 1995, 95, 69-96.
(24) Kormann, C.; Bahnemann, D. W.; Hoffman, M. R. EnViron. Sci.
Technol. 1988, 22, 798-806.
(25) Anpo, M.; Chiba, K.; Tomonari, M.; Coluccia, S.; Che, M.; Fox,
M. A. Bull. Chem. Soc. Jpn. 1991, 64, 543-551.
(26) von Fricks, G.; Hausmann, H.; Francke, V.; Gunther, H. J. Org.
Chem. 1997, 62, 5074-5079.
(27) Gunther, H.; Oepen, S.; Ebener, M.; Francke, V. Magn. Reson.
Chem. 1999, 37, S142-S146.
(28) Ebener, M.; von Fricks, G.; Gunther, H. Hel. Chim. Acta 1991,
74, 1296-1304.
Acknowledgment. K.A.G. gratefully acknowledges the
support of the Center for Catalysis and Surface Science at
Northwestern University and the NSF (BES-0000644). P.V.K.
acknowledges the support of the Office of Basic Energy
Sciences of the Department of Energy. This is Contribution No.
NDRL-4259 from the Notre Dame Radiation Laboratory.
(29) Desai, R.; Hussain, M.; Ruthven, D. M. Can. J. Chem. Eng. 1993,
70, 699-706.
(30) Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. EnVi-
ronmental Organic Chemistry; John Wiley & Sons: New York, 1993.
(31) Schwarzenbach, R. P.; Westall, J. EnViron. Sci. Technol. 1981, 15,
1360-1367.
(32) Szecsody, J. E.; Bales, R. C. J. Contam. Hydrol. 1989, 4, 181-
203.
Supporting Information Available: FTIR spectra obtained
for all 5 HCB coated oxide samples before and after irradiation
under an open atmosphere. This material is available free of
charge via the Internet at http://pubs.acs.org.
(33) Strehlow, W. H.; Cook, E. L. J. Phys. Chem. Ref. Data 1973, 2,
163-193.
(34) DiStefano, T. H.; Eastman, D. E. Solid State Comm. 1971, 9, 2259.
(35) Oosawa, Y.; Takahashi, R.; Yonemura, M.; Sekine, T.; Goto, Y.
New J. Chem. 1989, 13, 435-440.
(36) Emeline, A.; Kataeva, G. V.; Litke, A. S.; Rudakova, A. V.;
Ryabchuk, V. K.; Serpone, N. Langmuir 1998, 14, 5011-5022.
(37) Schatz, T.; Cook, A. R.; Meisel, D. J. Phys. Chem. 1998, 102,
7225-7230.
References and Notes
(1) Zacheis, G. A.; Gray, K. A.; Kamat, P. V. EnViron. Sci. Technol.,
in review 2001.
(38) Fukui, K.; Morokuma, K.; Kato, H.; Yonezawa, T. Bull. Chem.
Soc. Jpn. 1963, 36, 217-222.
(2) Mullins, W. M.; Averbach, B. L. Surf. Sci. 1988, 206, 41-51.
(3) Zacheis, G. A.; Gray, K. A.; Kamat, P. V. J. Phys. Chem. B 1999,
103, 2142-2150.
(4) Thomas, J. K. Chem. ReV. 1993, 93, 301-320.
(5) Shkrob, I. A.; Trifunac, A. D. J. Chem. Phys. 1997, 704, 2374-
2385.
(39) Liu, X.; Zhang, G.; Thomas, J. K. J. Phys. Chem. 1995, 99, 10 024-
10 034.
(40) Stamires, D.; Turkevich, J. J. Am. Chem. Soc. 1964, 86, 749-757.
(41) Liu, X.; Zhang, G.; Thomas, J. K. J. Phys. Chem. B 1997, 101,
2182-2194.