Decomposition of CCl3F over vanadium oxides and [MgV xOy]MgO shell/core-like particles

Article abstract of DOI:10.1016/j.jcat.2004.02.026

Decomposition of gaseous CCl₃F over in-house-prepared V ₂O₃ (HP-V₂O₃), commercial V ₂O₃, VO₂, V₂O₅, a mechanical mixture of the HP-V₂O₃/Aerogel-prepared MgO (AP-MgO), and [MgV_xO_y]MgO shell/core-like particles at 280°C has been investigated with FTIR spectroscopy. The reaction of CCl ₃F with the HP-V₂O₃ proceeds vigorously. Intermediate compounds including CCl₄, CCl₂F₂, CCl₂O, and final gaseous products CClF₃, CO₂ were detected. No chloride/fluoride was found on the sample surface, suggesting the formation of volatile vanadium-halogen products as well. Vanadium-chloride product was found to react with the surface of the bare AP-MgO leading to deposition of vanadium-chloride species. Activities of VO₂ and V ₂O₃ were found to be qualitatively similar to the HP-V₂O₃. The reaction of CCl₃F with V ₂O₅ proceeds much slower, with only CCl₂O and CO₂ gaseous products observed. Activation of the AP-MgO with vanadium-containing species leads to formation of [MgV_xO _y]MgO shell/core-like particles. The MgO crystalline phase was the only one detected in the [MgV_xO_y]MgO samples. In the reaction with CCl₃F, the [MgV_xO_y]MgO samples retain some properties characteristic for the bare AP-MgO (presence of an induction period for V/Mg=1 mol%, lower activity toward CCl₄ than toward CCl₂F₂ for V/Mg=1, 10 mol%), and some properties characteristic for the HP-V₂O₃/AP-MgO mechanical mixture (accumulation of chlorine but no fluorine on the [MgV_xO _y]MgO, V/Mg=10 mol% sample surface).

Full text of DOI:10.1016/j.jcat.2004.02.026

Journal of Catalysis 224 (2004) 340–346
www.elsevier.com/locate/jcat
Decomposition of CCl F over vanadium oxides and [MgV O ]MgO
3
y
x
shell/core-like particles
∗
Igor N. Martyanov and Kenneth J. Klabunde
Department of Chemistry, Kansas State University, 111 Willard Hall, Manhattan, KS 66506, USA
Received 13 March 2003; revised 12 September 2003; accepted 16 February 2004
Available online 21 April 2004
Abstract
Decomposition of gaseous CCl F over in-house-prepared V O (HP-V O ), commercial V O , VO , V O , a mechanical mixture of
3
2
3
2
3
2
3
2
2 5
◦
the HP-V O /Aerogel-prepared MgO (AP-MgO), and [MgV Oy]MgO shell/core-like particles at 280 C has been investigated with FTIR
2
3
x
spectroscopy. The reaction of CCl F with the HP-V O proceeds vigorously. Intermediate compounds including CCl , CCl F , CCl O,
3
2
3
4
2
2
2
and final gaseous products CClF , CO were detected. No chloride/fluoride was found on the sample surface, suggesting the formation of
3
2
volatile vanadium–halogen products as well. Vanadium–chloride product was found to react with the surface of the bare AP-MgO leading
to deposition of vanadium–chloride species. Activities of VO and V O were found to be qualitatively similar to the HP-V O . The
2
2
3
2 3
reaction of CCl F with V O proceeds much slower, with only CCl O and CO gaseous products observed. Activation of the AP-MgO with
3
2
5
2
2
vanadium-containing species leads to formation of [MgV Oy]MgO shell/core-like particles. The MgO crystalline phase was the only one
x
detected in the [MgV Oy]MgO samples. In the reaction with CCl F, the [MgV Oy]MgO samples retain some properties characteristic for
x
3
x
the bare AP-MgO (presence of an induction period for V/Mg = 1 mol%, lower activity toward CCl than toward CCl F for V/Mg = 1,
4
2 2
1
0 mol%), and some properties characteristic for the HP-V O /AP-MgO mechanical mixture (accumulation of chlorine but no fluorine on
2 3
the [MgV Oy]MgO, V/Mg = 10 mol% sample surface).
x

2004 Elsevier Inc. All rights reserved.
Keywords: MgO; V O ; V O ; CCl F; Magnesium oxide; Vanadium oxide; Freon; Decomposition; Nanoparticles; Catalysis
2
3
2
5
3
1
. Introduction
mixture was suggested to be the reason for the reaction ac-
celeration.
The positive catalytic influence of transition metal oxide
additives on AP-MgO performance with respect to CCl3F
destructive adsorption could be expected from earlier stud-
ies. Indeed, coverage of alkaline-earth particles with a shell
of transition metal oxides considerably improved AP-MgO
and CaO sorption abilities [1,9]. In the case of CCl4 destruc-
tion over AP-MgO, the effectiveness for reaction promotion
increased in the following order [1]:
About a decade ago alkaline-earth metal oxides were
found to be active in numerous destructive adsorption
processes including but not limited to detoxification of
chlorine- [1–5] and phosphorous- [6,7] containing organic
compounds as well as flue [8] gases. The properties of these
oxides have been thoroughly studied [9–13] and the advan-
tages of their nanoparticulate form have been well docu-
mented.
Recently, nanoparticulate Aerogel-prepared MgO (AP-
MgO) was found to be active [14,15] in the destruction of
ozone-damaging [16,17] CCl2F2 and CCl3F gases. Interest-
ingly, these reactions begin very slowly but drastically accel-
erate over time. Accumulation of halogens on the AP-MgO
Sc < none < Cr < Ni < Cu < Ti < Zn < Fe < Co
<
Mn < V.
In this regard investigation of the interaction of vanadium-
activated AP-MgO with CCl3F could be of particular interest
for a number of reasons. First, among all transition metals
tested, the vanadium additive allowed the highest activity to
be achieved. Second, upon the reaction of the vanadium ad-
surface followed by formation of a MgO ClyFz intimate
x
*
ditive with CCl F, formation of relatively mobile and conse-
Corresponding author. Fax: (785) 532 6666.
3
E-mail address: kenjk@ksu.edu (K.J. Klabunde).
quently more easily detected vanadium halogen-containing
0021-9517/$ – see front matter  2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2004.02.026

I.N. Martyanov, K.J. Klabunde / Journal of Catalysis 224 (2004) 340–346
341
species can be expected. Third, a difference in the chemistry
of chlorine- and fluorine-containingcompounds could be ex-
plored, which may help in elucidating a reaction mechanism.
Having this in mind, studies of the reactivity of vanadium
oxides and vanadium-activated AP-MgO shell/core-like par-
ticles toward trichlorofluoromethane decomposition were
carried out herein and compared with activity of the bare
AP-MgO.
samples contained on average V/Mg = 1 or 10% by molar
percentage. Though no water was added, the amount of H2O
and OH groups held on the AP-MgO surface was sufficient
to provide complete [(CH3)2CHO]3VO hydrolysis.
2.2. Sample characterization
A scanning electron microscope (SEM, Hitachi Science
System) was used for obtaining sample images and deter-
mination of their elemental composition. The samples were
pressed into a tungsten mesh and necessarily exposed to air
prior to introduction in the SEM vacuum chamber. Energy
dispersive X-ray analysis (EDXA) with 20 keV electron ac-
celeration voltage allowed determination of surface elemen-
tal composition of samples. An element was considered to
be present on a sample surface if its concentration exceeded
the ca. 0.5 mol% detection limit.
2
. Experimental
2
.1. Reagents and materials
CCl3F, CF2Cl2 of “pure” grade, CCl4, V2O3, VO2,
V2O5, [(CH3)2CHO]3VO, and CO2 (Aldrich), toluene, and
methanol (Fisher) were used as received.
Surface area of the samples was derived with the multi-
point BET method from nitrogen adsorption isotherms mea-
sured at 77 K on a NOVA-1200 (Quantachrome Corpo-
ration). Before surface area measurements, samples were
2
.1.1. AP-MgO preparation
Preparation of Aerogel-prepared MgO (AP-MgO) has
been described earlier [9]. In short this includes formation
of Mg(OCH3)2 via reaction of Mg with CH3OH, dilution of
Mg(OCH3)2 with toluene, formation of Mg(OH)2 with wa-
ter addition, supercritical drying of Mg(OH)2 suspension,
and conversion of Mg(OH)2 into AP-MgO via heat treat-
◦
additionally degassed at 500 C under dynamic vacuum
−
2
(10 Torr) for 1 h.
◦
A Sintag diffractometer working in a step mode (0.02
step, 1 s/step accumulation time) was employed for deter-
mining powder X-ray diffractograms (XRD) of the samples.
All recordings were carried out in open air with no additional
sample pretreatment.
◦
ment under dynamic vacuum at 500 C overnight.
2
.1.2. In-house-prepared vanadium oxide (HP-V2O3)
preparation procedure
Initially two mixtures were prepared. The first one was
obtained via addition of 1.1 mL of water to 20 mL of toluene
kept under continuous ultrasound sonication (Fisher Sci-
entific). The second one was obtained via dissolution of
2.3. Kinetic experiments
All experiments were carried out in a stainless-steel IR-
cell [18] (350 mL) connected to the vacuum line, power
supply, temperature controller, and equipped with KBr crys-
tal windows. An FTIR spectrometer (Mattson company) was
used to record IR spectra at 1-min intervals with 1 cm res-
olution.
4
3
.72 mL of [(CH3)2CHO]3VO into 20 mL of toluene. About
0 min after sonication was started the second mixture was
−1
slowly added to the first one. The resultant suspension was
kept under continuous sonication for 60 min in an open
beaker. Later the toluene was removed under dynamic vac-
−
2
uum (10 Torr) at room temperature overnight followed by
2.3.1. Experiments with vanadium oxides (HP-V2O3, and
commercial V2O3, VO2, V2O5), mechanical
HP-V2O3/AP-MgO mixture, and [MgV Oy]MgO
◦
high-temperaturevacuum treatment at 300 C for 30 min and
at 500 C for 1 h.
◦
x
shell/core-like particles
2
.1.3. Mechanical HP-V2O3/AP-MgO mixture preparation
Preliminary estimated quantities of the HP-V2O3 and
About 0.1 g of sample was pressed into tungsten (or in
the case of V2O5, VO2, and V2O3 into molybdenum) mesh,
inserted into the IR cell followed by sample pretreatment un-
the AP-MgO were carefully mixed to produce a composi-
−5
◦
tion with a ratio of V to Mg moles equal to 0.1 (V/Mg =
der dynamic vacuum ca. 10 Torr at 500 C for 1 h. Later,
◦
1
0 mol%).
the sample was cooled to 280 C (if not stated otherwise),
1
.5 Torr of CCl3F was introduced, the cell was closed, and
2
.1.4. [MgV Oy]MgO shell/core-like particles preparation
the spectra recording procedure was started. No oxygen was
fed in the IR cell. The IR beam passed only through the gas
phase; thus, changes in the gas phase were followed.
x
procedure
Initially two mixtures were prepared. The first mixture
was made by addition of 0.2 g of AP-MgO to 5 mL of
toluene kept under continuous ultrasound sonication. The
second one contained 12 or 118 µL of [(CH3)2CHO]3VO
dissolved in 5 mL of toluene. The next steps were similar to
those used for HP-V2O3 preparation. With 12 or 118 µL the
2.3.2. Experiments with the sample consisting of separated
spots of the AP-MgO and the HP-V2O3
Preliminary estimated amounts of the HP-V2O3 and the
AP-MgO were pressed into a tungsten mesh as separate

342
I.N. Martyanov, K.J. Klabunde / Journal of Catalysis 224 (2004) 340–346
spots in different locations on the grid. Total amounts of each
compound, their ratio, and further procedures were similar
to those used in the kinetic experiments with the mechanical
HP-V2O3/AP-MgO mixture.
2
.3.3. Quantitative measurements
Current CCl3F, CCl4, and CCl2F2 concentrations were
calculated from areas under characteristic peaks of IR spec-
tra followed by comparison with corresponding calibration
curves. In the case of CClF3, and CCl2O, no calibration
curves were available, so their concentrations are shown
in arbitrary units (A.U.). Due to the difficulties of taking
into account CO2 localized outside the IR cell, CO2 kinetic
curves are also shown in A.U. In most cases superimpo-
sition of absorption peaks originating from different com-
pounds could be neglected, although the measured value of
Fig. 2. Gas-phase FTIR spectra recorded in situ at various times elapsed
after CCl3F was brought into contact with the HP-V2O3 maintained at
T = 280 C.
−1
the CCl3F peak at 846 cm was corrected for the presence
of CCl2O.
◦
3
. Results and discussion
3
.1. Vanadium oxides
XRD patterns of different vanadium oxides used in the
present study are shown in Fig. 1. Comparison of the XRD
pattern of the in-house-prepared vanadium oxide with com-
mercially available V2O3 revealed their structural similar-
ities so that in-house-prepared vanadium oxide was desig-
nated as HP-V2O3. Surface areas of the samples derived
from nitrogen adsorption measurements were HP-V2O3
2
2
2
(
(
5 m /g), V2O3 (1.6 m /g), VO2 (2.7 m /g), and V2O5
2
4.0 m /g).
Fig. 2 presents the gas-phase IR spectra recorded dur-
ing the first 10 min of the reaction of CCl3F with HP-V2O3
Fig. 3. Time evolution of different gaseous compound concentrations af-
◦
ter introduction of CCl F to the HP-V O maintained at T = 280 C. The
3
2 3
course of the reaction was followed by FTIR.
◦
at T = 280 C. Absorption bands of the reaction products
pertaining to CCl3F, CCl2F2, CCl4, CCl2O, and CO2 are
−1
readily noticeable. Two absorption bands around 1830 cm
attributed to phosgene are followed by a weak unidentified
−1
doublet at ca. 1880 cm . It is worth noting that the sim-
ilar unidentified product was observed during reaction of
chlorobenzene with V2O5 [19], suggesting that this com-
pound does not contain fluorine. Since the spectrum of this
−1
unknown compound (at least in the region near 1880 cm
)
is similar to phosgene (with some shift to higher wavenum-
bers) and the concentrations of phosgene and the unknown
product correlate with each other it can be suggested that the
unidentified product is a result of CCl2O interaction with a
vanadium-containing gas.
Fig. 3 shows the evolution of CCl3F, CCl4, CCl2F2,
CClF3, CCl2O, and CO2 concentrations with time. Just
Fig. 1. XRD patterns of various vanadium oxide powders tested in the re-
action with CCl F. Vertical scale is presented in counts per second (CPS)
3
with base lines deliberately shifted for clarity.
4
min is enough to eliminate CCl3F, CCl4, and CCl2F2

I.N. Martyanov, K.J. Klabunde / Journal of Catalysis 224 (2004) 340–346
343
(a)
Fig. 4. Time evolution of different gaseous compound concentrations after
◦
introduction of CCl F to the HP-V O maintained at T = 180 C.
3
2 3
Table 1
The presence of Mg, V, Cl, F, and O elements detected at sample surfaces
by EDX analysis
Mg V Cl F O
Before reaction
After reaction
with CCl F
3
with CCl F
3
HP-V O
− + − − +
+ − − − +
− + − − +
+ + + − +
2
3
Spot of AP-MgO
with HP-V O
(
2
3
spot situated near by)
Shell/core [MgV Oy]MgO
V/Mg = 10%
+ + − − +
+ + + − +
x
The “+” refers to a positive identification, “−” refers to the situation when
the concentration of an element was below 0.5 at% detection limit.
(b)
Fig. 5. SEM images of the HP-V O before (a) and after (b) the reaction.
gases. Conversion of CCl2O into CO2 proceeds later. CO2,
CClF3, and traces of HCl were the only final products de-
tected.
2 3
qualitatively similar to HP-V2O3. However, the behavior of
V2O5 was completely different. Though the surface area of
V2O5 was higher than that of commercial V2O3, V2O5 was
much less active toward CCl3F with no CCl4, CCl2F2, and
CClF3 formed (Fig. 6).
To see the beginning of the process in more detail, the
◦
same reaction was carried out at 180 C (Fig. 4) instead of
◦
2
80 C. Even at this lower temperature, still no induction
period typical for the bare AP-MgO [15] was observed. Ini-
tially, decomposition of CCl3F leads to equal amounts of
CCl4 and CCl2F2. Soon, however, the CCl4 and CCl2F2 con-
centrations reach maximal values and begin to fall with CCl4
disappearing faster than CCl2F2.
+
Activity of V2O5 [19–22] and various VxOy posi-
tively charged clusters [23,24] participating in the reaction
with CCl4 were studied earlier. In the case of V2O5, for-
mation of CO2, CCl2O, and VOCl3 was observed. It was
found that while CO2 concentration gradually increased and
CCl2O passed through a maximum, the VOCl3 concentra-
tion reached a limiting value and did not change later on as
the reaction progressed [21].
The experiments with vanadium oxide clusters are of spe-
cial interest since the reactivity of vanadium in different oxi-
dation states was explored. In particular, it was found that the
clusters containing vanadium in low oxidation states were
considerably more active and reacted in a way different from
the clusters where the oxidation state of vanadium was +5.
Elemental analysis of the HP-V2O3 samples before and
after the reaction revealed no presence of even trace amounts
of chlorine or fluorine moieties (Table 1), whereas SEM im-
ages gave evidence of surface etching (Fig. 5). These find-
ings indicate that vanadium–chlorine/fluorine species form-
◦
ing during the reaction are volatile at 280 C and leave the
surface of the sample.
To gain some information about the influence of vana-
dium oxidation state on sample reactivity, experiments with
commercially available V2O3, VO2, and V2O5 were car-
ried out. The V2O3 and VO2 samples demonstrated activity

344
I.N. Martyanov, K.J. Klabunde / Journal of Catalysis 224 (2004) 340–346
Fig. 7. Time evolution of different compound concentrations after mechan-
Fig. 6. Time evolution of different compound concentrations after exposure
◦
ical HP-V O/AP-MgO mixture (V/Mg = 10 mol%) was brought into con-
of the commercial V O maintained at T = 280 C to CCl F. The reaction
2
2
5
3
◦
tact with CCl F; T = 280 C.
conditions are the same as in Fig. 2.
3
+
For example, the V3O6 cluster (formal vanadium oxida-
major active component of the HP-V2O3/AP-MgO mechan-
ical mixture.
1
tion state is +4 ) reacted with abstraction of chlorine atom,
3
To trace vanadium movements, an experiment with sepa-
rated spots (different locations on the grid) of the AP-MgO
and the HP-V2O3 was carried out. Elemental analysis re-
vealed (Table 1) that some vanadium as well as chlorine but
no fluorine was deposited on the AP-MgO. This is proba-
bly due to the higher reactivity of vanadium–chlorine com-
pounds, which would be volatile and move to the surface of
the MgO, exchanging chlorine for oxygen at reactive sites.
Vanadium–fluorine compounds would be less volatile (com-
+
+
V3O6 + CCl4 → V3O6Cl + CCl3,
(1)
+
whereas the V3O7 cluster containing vanadium atoms in
the +5 oxidation state was much less active and reacted with
phosgene formation,
+
+
V3O7 + CCl4 → V3O6Cl2 + CCl2O.
(2)
+
Note that cluster V3O7 cannot accept a chlorine atom
because it cannot provide electrons for an additional vanadi-
um–chlorine bond. Instead, that cluster prefers to exchange
oxygen for two chlorine atoms. In this respect the clusters
containing vanadium in an oxidation state lower than +5
have certain advantages since they can easily provide elec-
trons for initiating bond formation.
◦
pare, for example, boiling points of VOF3 (480 C) with
◦
analogous VOCl3 (126.7 C)) [25] and perhaps more ther-
modynamically stable and therefore would be less likely
to interact with the MgO surface. Since the atomic ratio
of chlorine to vanadium deposited was Cl/V = 3.2 ± 0.4,
VCl3 might be considered as a possible volatile vanadium–
chlorine-containing compound involved in the reaction.
These results seem to be in good agreement with the
present work. In accordance with the literature data [23],
the reaction of CCl3F with bulk V2O5 proceeds slowly with
CCl2O to be the only detectable intermediate product. The
samples containing vanadium in +3 oxidation state are con-
siderably more active with CCl4, CCl2F2, and CCl2O as
intermediate species.
3
.3. [MgV Oy]MgO shell/core-like particles
x
Activation of the AP-MgO with the vanadium addi-
tive results in a substantial surface area decrease: AP-MgO
2
2
(
530 m /g), [MgV Oy]MgO (V/Mg = 1 mol%) (359 m /g),
x
2
3
.2. Mechanical HP-V2O3/AP-MgO mixture
[MgV Oy]MgO (V/Mg = 10 mol%) (110 m /g). Since de-
x
position of the vanadium compound on the AP-MgO was
done in a way similar to the preparation of HP-V2O3, for-
mation of a V2O3 phase on the top of the AP-MgO might
be expected. However, as it is evident from the XRD data
presented in Fig. 8, no new crystalline phase forms upon
heating of the AP-MgO with the vanadium species. This re-
sult is not unusual in view of earlier studies where vanadium
oxides were found to react with MgO at elevated temper-
atures forming numerous compounds [26]. Having this in
mind and due to the higher vanadium concentration at the
The course of the reaction of CCl3F with a mechanical
mixture of HP-V2O3/AP-MgO having V/Mg = 10% molar
◦
ratio and kept at 280 C is shown in Fig. 7. As this sample
contained a lower amount of the HP-V2O3, the reaction pro-
gressed slower than with the pure HP-V2O3 (Fig. 3). At the
same time the evolution of different compound concentra-
tions was quite similar to the HP-V2O3 alone with exception
that no CClF3 was observed. From these considerations we
conclude that in the early stages of reaction, HP-V2O3 is the

I.N. Martyanov, K.J. Klabunde / Journal of Catalysis 224 (2004) 340–346
345
Fig. 8. XRD patterns of the AP-MgO and the [MgV Oy ]MgO shell/
core-like samples with different vanadium contents.
x
Fig. 9. Time evolution of different gaseous compound concentrations af-
ter [MgV Oy ]MgO shell/core-like sample (V/Mg = 1 mol%) was brought
x
◦
into contact with CCl F; T = 280 C.
3
MgO surface, vanadium-activated (heat treated) magnesium
oxide particles are denoted herein as [MgV Oy]MgO.
x
The formation of an intimate mixture of the AP-MgO
and vanadium oxide results in the appearance of a new
compound, [MgV Oy]MgO (Figs. 9 and 10), with some
x
properties similar to and some properties different from the
properties of the bare AP-MgO, HP-V2O3, or their mechan-
ical mixture. Indeed, [MgV Oy]MgO (V/Mg = 1 mol%)
x
demonstrates the presence of an induction period typical for
the AP-MgO [15], but not for the HP-V2O3. Interaction of
CCl3F with the [MgV Oy]MgO results in the appearance of
x
CCl4, and CCl2F2 (Fig. 9). Although these intermediates are
typical for both the AP-MgO [15] and the HP-V2O3 (Figs. 3
and 4), the lesser stability of CCl2F2 as compared to CCl4
makes the [MgV Oy]MgO sample resemble the AP-MgO
x
rather than the HP-V2O3. The ability for the accumulation of
halogens is yet another way to stress the similarities and the
differences in the properties of the [MgV Oy]MgO and the
AP-MgO, HP-V2O3, as well as the AP-MgO/HP-V2O3 me-
chanical mixture. As can be seen from the results of Table 1,
Fig. 10. Time evolution of different compound concentrations after
x
[
MgV Oy ]MgO shell/core-like sample (V/Mg = 10 mol%) was brought
x
◦
into contact with CCl F; T = 280 C.
3
the [MgV Oy]MgO demonstrates an ability to accumulate
x
chorine but no fluorine, whereas the AP-MgO accumulates
both chorine and fluorine [15]. The AP-MgO/HP-V2O3 me-
chanical mixture accumulates chlorine but no fluorine and
none of the halogens deposit on the HP-V2O3 surface.
Based on the experimental data and the consideration
given above, the following conclusions regarding the process
not eliminate it entirely. Thus, by analogy with the bare
AP-MgO [15] it is reasonable to suggest that the presence
of the vanadium additive accelerates accumulation of halo-
gens (mainly chlorine) on the surface of MgO and is fol-
lowed by formation of a MgVxOyClz intimate mixture. The
MgVxOyClz mixture is more active than the initial MgVxOy
so that acceleration of the reaction takes place.
of decomposition of CCl3F over [MgV Oy]MgO core/shell-
x
like particles can be drawn. Addition of vanadium moieties
to the shell of AP-MgO particles leads to substantial activa-
tion of the AP-MgO. However, in the case of small amounts
of vanadium (V/Mg = 1 mol%), the role of vanadium as
an activator is mainly seen in its influence on the induc-
tion period of the reaction. Indeed, in comparison with the
bare AP-MgO, introduction of such a small amount of vana-
dium shortens the induction period of the reaction but does
Overall, the following processes (not balanced) are
thought to be responsible for the observed behavior of
[MgVxOy]MgO samples. First, the reaction of the vanadium
containing part of the shell, [MgVxOy], with CCl3F leads to
chlorine accumulation:
[
MgV Oy]MgO + CCl3F
x
→ [MgV Oy][MgV OyClz]MgO + VFy + CO2.
(3)
x
x

346
I.N. Martyanov, K.J. Klabunde / Journal of Catalysis 224 (2004) 340–346
Since the [MgV OyClz] is more active than [MgV Oy],
the observed reaction acceleration takes place. Higher ac-
loading level additionally enhances due to the accumulation
of chorine.
x
x
tivity of [MgV Oy] with respect to the bare MgO [15] ex-
x
plains the shorter induction period of the reaction. Later,
Acknowledgment
[
MgV OyClz] may react with CCl3F giving CCl4:
x
The support of the Army Research Office (DAAD 19-01-
0619) is acknowledged with gratitude.
[
MgV Oy][MgV OyClz]MgO + CCl3F
x
x
1
→
[MgV Oy][MgV OyClz]MgO + VFy + CCl4.
(4)
x
x
Fast reaction of [MgV OyClz] with CCl3F and CCl4 should
x
References
lead to progressive substitution of surface oxygen by chorine
until complete oxygen exhaustion. At this point, vanadium
may leave the surface in the forms of volatile VClx as in the
processes used for extraction of vanadium from ores [27]:
[
[
1] Y. Jiang, S. Decker, C. Mohs, K.J. Klabunde, J. Catal. 180 (1998) 24.
2] O. Koper, I. Lagadic, K.J. Klabunde, Chem. Mater. 9 (1997) 838.
[3] O. Koper, E.A. Wovchko, J.A. Glass Jr., J.T. Yates Jr., K.J. Klabunde,
Langmuir 11 (1995) 2054.
[
MgV Oy][MgV OyClz]MgO + CCl3F/CCl4
[4] Y.-X. Li, H. Li, K.J. Klabunde, Environ. Sci. Technol. 28 (1994) 1248.
x
x
[
[
[
5] O. Koper, Y.-X. Li, K.J. Klabunde, Chem. Mater. 5 (1993) 500.
6] Y.-X. Li, K.J. Klabunde, Langmuir 7 (1991) 1388.
7] Y.-X. Li, O. Koper, M. Atteya, K.J. Klabunde, Chem. Mater. 4 (1992)
→
[MgV Oy][MgCl ]MgO + VClx + VFy + CO2. (5)
x
2
In the presence of bare MgO, VClx reacts with it creating
active [MgV OyClz]:
3
23.
8] J.V. Stark, D.G. Park, I. Lagadic, K.J. Klabunde, Chem. Mater. 8
1996) 1905.
9] S. Utamapanya, K.J. Klabunde, J.R. Schlup, Chem. Mater. 3 (1991)
75.
x
[
[
(
[
MgV Oy][MgCl ]MgO + VClx
x
2
1
→
[MgV Oy][MgCl ][MgV OyClz]MgO.
(6)
x
2
x
[
10] K.J. Klabunde, J. Stark, O. Koper, C. Mohs, D.G. Park, S. Decker, Y.
Jiang, I. Lagadic, D. Zhang, J. Phys. Chem. 100 (1996) 12142.
11] R.M. Morris, K.J. Klabunde, Inorg. Chem. 22 (1983) 682.
12] H. Itoh, S. Utamapanya, J.V. Stark, K.J. Klabunde, J.R. Schlup, Chem.
Mater. 5 (1993) 71.
In excess of CCl3F, conversion of [MgV Oy]MgO into
x
[
[
MgF2 as the most thermodynamically stable magnesium
compounds and volatile vanadium–halogens can be ex-
pected:
[
13] O. Koper, I. Lagadic, A. Volodin, K.J. Klabunde, Chem. Mater. 9
(
1997) 2468.
[
MgV Oy]MgO + CCl3F → MgF + VClx + VFy
x
2
[14] I.V. Mishakov, A.F. Bedilo, V.V. Chesnokov, S.V. Filimonova, I.N.
Martyanov, K.J. Klabunde, A.M. Volodin, V.N. Parmon, in prepara-
tion.
+
CCl4 + CO2.
(7)
[
[
[
15] I.N. Martyanov, K.J. Klabunde, in preparation.
16] M.J. Molina, F.S. Rowland, Nature 249 (1974) 810.
17] G.J. Wolff, in: J.I. Kroschwitz, M. Howe-Grant (Eds.), Encyclopedia
of Chemical Technology, vol. 1, Wiley, New York, 1994, p. 715.
18] P. Basu, T.H. Ballinger, J.T. Yates Jr., Rev. Sci. Instrum. 59 (1988)
1321.
4
. Conclusions
In-house prepared V2O3 (HP-V2O3) was found to be
[
[
[
19] S.B. Geyer, L.R. Brock, J.W. Keister, T.C. DeVore, Electrochem. Soc.
Proc. 97 (39) (1997) 298.
20] G. Mink, I. Bertoti, T. Szekely, React. Kinet. Catal. Lett. 27 (1) (1985)
the most active in the decomposition of CCl3F as com-
pared with the commercial V2O5, VO2, and V2O3. The
HP-V2O3 is consumed in the reaction so that besides CO2
and some CClF3, the reaction products include vanadium–
halogen species that leave the surface of the HP-V2O3 under
the reaction conditions studied.
Formation of the volatile vanadium–halogen species fol-
lowed by their recapture by bare MgO is thought to be the
reason behind the catalytic influence of vanadium additives
on the performance of the AP-MgO in the destructive ab-
33.
[21] G. Mink, I. Bertoti, C. Battistoni, T. Szekely, React. Kinet. Catal.
Lett. 27 (1) (1985) 39.
[22] G. Mink, I. Bertoti, I.S. Pap, T. Szekely, C. Battistoni, E. Karmazsin,
Thermochim. Acta 85 (1985) 83.
[23] R.C. Bell, K.A. Zemski, A.W. Castleman Jr., J. Phys. Chem. 103
(1999) 1585.
[24] R.C. Bell, K.A. Zemski, A.W. Castleman Jr., J. Phys. Chem. 103
(1999) 2992.
[25] R.C. Weast (Ed.), Handbook of Chemistry and Physics, fifty fourth
sorption of CCl3F. As a result, the [MgV Oy]MgO samples
demonstrate higher activities and shorter induction periods
as compared to the bare AP-MgO. As the reaction proceeds,
x
ed., CRC Press, Cleveland, 1974.
[
[
26] H. Oshima, J. Am. Ceram. Soc. 63 (1980) 504.
27] S.A. Amirova, V.V. Pechkovskii, R. Kurmaev, Kh. Tsvetn. Metal. 36
(1963) 57.
the activity of the [MgV Oy]MgO with a low vanadium
x