Catalytic conversions of chloroolefins over iron oxide nanoparticles: 1. Isomerization of dichlorobutenes in the presence of iron oxide nanopaticles immobilized on silicas with different structures

Article abstract of DOI:10.1007/s11172-005-0421-2

The influence of the support nature and electronic state of iron oxide nanoclusters on the catalytic properties of supported systems was studied for dichlorobutene isomerization. A sample with a Fe content of 2.5 wt.% on the activated silica matrix containing Fe^III and Fe^II ions in the paramagnetic state exhibits the highest activity. The activity of iron on silica gel enhances with the appearance of magnetically ordered nanoclusters of γ-iron oxide formed at the iron content on the catalyst as high as 15 wt.%. An increase in the catalyst activity is favored by the formation of two states of iron (Fe^III and Fe^II) that occurs under the synthesis conditions or during the action of a reactant.

Full text of DOI:10.1007/s11172-005-0421-2

1418
Russian Chemical Bulletin, International Edition, Vol. 54, No. 6, pp. 1418—1424, June, 2005
Catalytic conversions of chloroolefins over iron oxide nanoparticles
1. Isomerization of dichlorobutenes in the presence of iron oxide nanopaticles
immobilized on silicas with different structures
T. N. Rostovshchikova,aꢀ V. V. Smirnov,a M. V. Tsodikov,b O. V. Bukhtenko,b
Yu. V. Maksimov,c O. I. Kiseleva,a and D. A. Pankratova
aDepartment of Chemistry, M. V. Lomonosov Moscow State University,
1 Leninskie Gory, 119992 Moscow, Russian Federation.
Fax: +7 (095) 932 8846. Еꢀmail: rtn@kinet.chem.msu.ru
bA. V. Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences,
29 Leninsky prosp., 119991 Moscow, Russian Federation.
Fax: +7 (095) 230 2224. Eꢀmail: tsodikov@ips.ac.ru
cN. N. Semenov Institute of Chemical Physics, Russian Academy of Sciences,
4 ul. Kosygina, 119991 Moscow, Russian Federation.
Eꢀmail: maksimov@chph.ras.ru
The influence of the support nature and electronic state of iron oxide nanoclusters on the
catalytic properties of supported systems was studied for dichlorobutene isomerization. A sample
with a Fe content of 2.5 wt.% on the activated silica matrix containing FeIII and FeII ions in the
paramagnetic state exhibits the highest activity. The activity of iron on silica gel enhances with
the appearance of magnetically ordered nanoclusters of γꢀiron oxide formed at the iron content
on the catalyst as high as 15 wt.%. An increase in the catalyst activity is favored by the
formation of two states of iron (FeIII and FeII) that occurs under the synthesis conditions or
during the action of a reactant.
Key words: iron oxide, nanoparticles, supports, dichlorobutene isomerization, structure,
catalytic activity.
Iron oxides are widely used as catalysts in the
of nanoparticles.3 The role of structures including two
charge states of the iron ions (FeII and FeIII) was observed
in the Fischer—Tropsch reaction and some other cataꢀ
lytic processes.4,5
Fischer—Tropsch synthesis, dehydrogenation, alkylation,
and other processes.1,2 They are also active in catalytic
reactions involving halogenꢀcontaining hydrocarbons3 in
which the catalytic properties of bulky and nanoscale iron
oxides differ significantly. Isomerization of dichloroꢀ
butenes and benzene alkylation with chloroolefins of the
allyl type on bulky iron oxide are accompanied by the
formation of a large (up to 30%) amount of oligomerizaꢀ
tion products characteristics of chain radical conversions
of chloroolefins in the presence of peroxide initiators.
The same reactions on nanoscale catalysts proceed with
much higher rates and enhanced selectivity with respect
to the primary alkylation products.3 The main pathway of
the catalytic process in a benzene—chloroolefin system
depends on the presence of oxygen. Benzene alkylation
with allyl chloride or dichlorobutene occurs only in the
presence of oxygen. Removal of the latter favors dichloroꢀ
butene isomerization. In the absence of benzene, isomerꢀ
ization also occurs in an oxygenꢀcontaining medium.
A strong effect of oxygen on the rate and direction of the
processes suggests that the iron ions in mixedꢀvalence
states are involved in catalysis on nonstoichiometric phases
The nature of the catalyst surface is determined by the
size of iron oxide particles and their resistance to reducꢀ
tion and depends strongly on the iron to support ratio
and character of particle distribution on the support.
For instance, relatively large iron oxide nanoparticles
(~13—15 nm) obtained at a high (10—15%) iron content
on silica gel are reducible with hydrogen on heating to
metallic iron.6 At the same time, fine iron oxide nanoꢀ
clusters (<5 nm in size), which are formed on the support
at a low (down to 5%) iron content, cannot virtually be
reduced under similar conditions. This is caused by a
stronger interaction between nanoclusters of a smaller
size and the support surface. In addition to the particle
size, the structure of the support, its surface, and porosity
exert a substantial effect on the formation of catalytically
active iron species.
Isomerization of dichlorobutenes seems to be conveꢀ
nient to find specific features of the catalyst effects, beꢀ
cause this process has previously been studied in detail in
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 6, pp. 1376—1382, June, 2005.
1066ꢀ5285/05/5406ꢀ1418 © 2005 Springer Science+Business Media, Inc.

Dichlorobutene isomerization on Fe2O3
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 6, June, 2005
1419
60—120 °С in the presence of oxygen. A tube was loaded with a
weighed sample of the catalyst with the specified metal content
and a measured amount of 3,4ꢀDCB or its solution in diꢀ
chloroethane with the reactant to solvent ratio varying from 1 : 1
to 1 : 9. All experiments were conducted under conditions when
the process rate was independent of the stirring velocity.
The composition of the reaction mixture was analyzed on a
Kristall Lyuksꢀ4000 chromatograph with a flameꢀionization deꢀ
tector on a column packed with the SEꢀ30 stationary phase
using the temperatureꢀprogrammed regime of the column in an
interval of 50—150 °С. According to the data of chromatoꢀ
graphic analysis, the main product of 3,4ꢀDCB isomerization
was transꢀ1,4ꢀDCB. The yield of the cisꢀisomer did not exꢀ
ceed 2%. At high conversions, the reaction products contained
up to 1% byꢀproducts: chloroprene and higherꢀboiling subꢀ
stances.
the presence of the homoꢀ and heterovalent copper comꢀ
plexes,7 iron(III) chloride,8 and nanodimensional metals.9
The isomerization mechanism in the presence of the supꢀ
ported metal oxides is studied in less detail.3
The purpose of this work is to determine the influence
of the nature of silica and the ratio of active component to
support on the catalytic activity of iron oxides in the
isomerization of 3,4ꢀdichlorobutꢀ1ꢀene (3,4ꢀDCB) to
transꢀ1,4ꢀdichlorobutꢀ2ꢀene (transꢀ1,4ꢀDCB).
CH2ClCHClCH=CH2
transꢀCH2ClCH=CHCH2Cl
Silica gel with the standard globular structure and an
activated silica matrix (ASM) were chosen as supports.
The ASM is prepared by the removal of metal cations
from natural vermiculite4,10 and retains the layer strucꢀ
ture typical of mica.
The specific catalytic activity was calculated as the number
of moles of transꢀ1,4ꢀDCB per mole of iron per hour using the
maximum rate of the process determined from a slope of the
curve of transꢀ1,4ꢀDCB accumulation.
Experimental
In the absence of an induction period (a lowꢀrate region) in
the kinetic curves, the results were processed in the coordinates
of a reversible firstꢀorder reaction ln[(A∞ – A0)/(A∞ – A)]—t,
where A∞ and A0 are the equilibrium and initial concentrations
of transꢀ1,4ꢀDCB, and A is its current concentration. Equilibꢀ
rium concentrations A∞ at different temperatures were calcuꢀ
lated from the known equilibrium constants of isomerization.13
The content of transꢀ1,4ꢀDCB in the initial mixture (A0) was
determined by chromatography; this value usually did not exꢀ
ceed 1%. The accuracy of determination of isomerization rate
constants was 15%.
Nanoparticles of Fe2O3 on the KSKꢀ2 silica gel (specific
surface S = 340 m2 g–1) were prepared3 by the impregnation of
the support with a solution of iron nitrate followed by calcinaꢀ
tion for 6 h at 400 °С. The iron content in the catalysts was 3, 5,
10, and 15 wt.%. Samples with the 5% iron content including
the phases Fe3O4 or FexO (x < 1) along with Fe2O3 were obꢀ
tained by temperatureꢀprogrammed reduction with hydrogen in
an interval of 400—500 °С and continuous in situ measurement
of magnetization. Reduction was interrupted each time when a
new oxide phase was formed.6 The main characteristics of the
catalysts (particle size, specific surface, parameter of Mössbauer
spectra) prepared by this method have been described earlier.3,11
The activated silica matrix was prepared by twoꢀstage acid
etching of vermiculite.12 At the first stage, vermiculite was treated
with dilute hydrochloric acid at 50—60 °С, and the second stage
was the treatment with a mixture of 40% sulfuric and concenꢀ
trated nitric acids taken in a ratio of 40 : 60 (v/v). Thus obtained
ASM was characterized by the high specific surface (S =
350 m2 g–1) and uniform size distribution of mesopores
(~8 nm).12 The ASMꢀimmobilized iron oxide with a metal conꢀ
tent of 2.5 wt.% was prepared by the impregnation of the support
with a solution of iron acetylacetonate followed by evacuation
and calcination at 400 °C.
Structural studies of the initial catalysts were carried out
by Xꢀray diffraction and Mössbauer spectroscopy. Xꢀray
phase analysis of ironꢀcontaining phases was conducted on a
DRONꢀ3M diffractometer using filtered СuꢀКα radiation. The
crystallite size was determined from the broadening of diffracꢀ
tion peaks corrected for structure anisotropy. Mössbauer abꢀ
sorption spectra were measured on MS1101E (MNPP "MosTek,"
Russia and Haldor, Germany) electrodynamicꢀtype spectromꢀ
eters at room temperature. Sources of γꢀradiation were 57Со
sources in the matrix of metallic chromium or rhodium with
activity lower than 1.1 GBq. Chemical shifts were determined
relatively to αꢀFe. Spectra were processed by the standard leastꢀ
squares method assuming the Lorentzian line shape.
Results and Discussion
Iron oxide nanoparticles immobilized on silicas. The
Xꢀray diffraction patterns of the samples supported on the
KSKꢀ2 silica gel with the iron content 10 and 15 wt.%
contain a broad halo corresponding to an amorphous to
Xꢀray phase and broadened reflections from the αꢀFe2O3
nanoclusters (d = 0.2762, 0.2714, 0.2528, 0.2292, and
0.2209 nm) with sizes of 13 and 15 nm, respectively. The
spectra of the catalysts based on KSKꢀ2 with the iron
content 3 and 5 wt.% exhibit very weak reflections, most
likely, from the same clusters along with the broad
halo. The ASMꢀsupported catalyst with the iron content
2.5 wt.% is amorphous to Xꢀrays.
The parameters of the Mössbauer spectra of the studꢀ
ied catalysts are presented in Table 1. As follows from the
data presented, the lines of magnetic HFS are the most
intense in the spectra of the KSKꢀ2ꢀsupported catalyst
samples. The presented parameters of components 1 and 2
(FeIII(magn.)) correspond to the spectra of nanoclusters
of αꢀFe2O3 (hematite) with the rhombohedral lattice and
mean size >10 nm.14 This agrees with the data on Xꢀray
diffraction. The presence of two HFS components from
FeIII(magn.) (1 and 2) in the spectra of the samples conꢀ
taining 5 and 10 wt.% iron indicates that the system conꢀ
tains two dimensional modes of hematite nanoclusters.
Reagents and solvents were dried and distilled, and purity
was monitored by chromatography. Isomerization of 3,4ꢀDCB
was carried out in sealed glass tubes with stirring in an interval of

1420
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 6, June, 2005
Rostovshchikova et al.
Table 1. Parameters of the Mössbauer spectra of the starting catalysts at room temperature (δ is the isomeric shift,
∆ is the quadrupole shift in the spectra with magnetic splitting or quadrupole splitting in the spectra with
paramagnetic components, Γ is the linewidth, and Hin is the internal field on the iron nucleus)
Fe
Component
δ 0.03
∆ 0.03
Γ 0.03
Hin 0.1/T
Relative
(wt.%)
content ( 5%)
mm s–1
KSKꢀ2
5
FeIII(magn.), 1;
FeIII(magn.), 2;
FeIII(paramagn.)
FeIII(magn.), 1;
FeIII(magn.), 2;
FeIII(paramagn.);
singlet
FeIII(magn.), 1;
FeIII(magn.), 3;
FeIII(magn.), 4;
FeIII(paramagn.);
singlet
0.35
0.22
0.30
0.38
0.33
0.29
0.54
0.38
0.32
0.39
0.31
0.58
–0.22
–0.26
0.70
–0.21
–0.23
0.74
0.39
0.74
0.53
0.32
0.46
0.64
0.68
0.43
0.73
1.13
1.12
0.32
50.4
47.8
—
50.9
49.3
—
48
15
37
69
24
5
10
—
—
2
15
–0.21
–0.06
–0.05
0.68
50.4
44.9
39.0
—
48
14
13
23
2
—
—
ASM
2.5*
FeIII(paramagn.);
FeII(paramagn.)
FeIII(paramagn.);
FeII(paramagn.)
0.36
1.19
0.35
1.18
0.88
2.07
0.76
1.96
0.60
0.56
0.68
0.33
—
—
—
—
85
15
70
30
After
reaction
* In the starting catalyst.
Narrower lines and a higher internal field on the iron
nucleus (Γ ≈ 0.3—0.4 mm s–1, Hin ≈ 50— 51 T) for comꢀ
ponent 1 can be assigned to larger particles with a rather
narrow size distribution. The parameters of the second
mode with Γ ≈ 0.5—0.7 mm s–1 and Hin ≈ 48—49 T charꢀ
acterize smaller clusters with a broader size distribution of
particles. Since this component is absent from the specꢀ
trum of the sample with the iron content 15 wt.%, we can
assume that of hematite clusters of the same size with the
parameters Hin ≈ 50.4 T and Γ ≈ 0.4 mm s–1 prevail. Along
with component 1 from the magnetic hematite clusters,
the spectrum of the sample with 15 wt.% Fe exhibits
distinctly two HFS components: 3 (FeIII(magn.)) and
4 (FeIII(magn.)). The spectra with these parameters charꢀ
acterize magnetically ordered defect clusters of γꢀoxide
with the spinel structure and average size of ~10 nm.15
The spectra of all the KSKꢀ2ꢀsupported iron samples conꢀ
tain the component FeIII(paramagn.) corresponding to
the FeIII ions in the highꢀspin state in the octahedral
oxygen environment. Probably, this spectral component
characterizes very small (≤5—6 nm) clusters of γꢀ or
αꢀiron oxide with superparamagnetic properties.16 The
data in Table 1 show that the contribution of the
FeIII(magn.) components decreases and, correspondingly,
the relative fraction of small clusters increases with a deꢀ
crease in the iron content. Finally, weak single lines in
the spectra of the samples containing 10 and 15 wt.%
iron can be attributed (by the isomeric shift value) to
superparamagnetic magnetite nanoclusters with a size
of ~6 nm.17—19
Unlike the spectra of the catalysts with KSKꢀ2 as a
carrier, the spectrum of the iron sample on the ASM
support contains only paramagnetic signals from the FeIII
and FeII ions in the highꢀspin state. The presence of the
paramagnetic signal from the FeIII ion measured at room
temperature agrees with the Xꢀray diffraction data, beꢀ
cause it cannot be excluded that this signal corresponds to
very small clusters of γꢀiron oxide amorphous to Xꢀrays.
Weak ferrimagnetism has previously4 been observed in
similar systems by an analysis of the data on magnetic
susceptibility.
Catalytic properties of iron oxides supported on silica
gel. The kinetic curves of transꢀ1,4ꢀDCB accumulation
for the KSKꢀ2ꢀbased catalysts with different composiꢀ
tions at different catalyst and reactant concentrations in a
temperature interval of 80—110 °С are shown in Figs 1
and 2. In some cases, the rate of transꢀ1,4ꢀDCB formaꢀ
tion increased after the induction period had elapsed. An
estimate of the induction time was made by analyzing the
kinetic curves and taking the time needed to achieve the
maximum isomerization rate (Table 2). As follows from
the data in Fig. 1 and Table 2, the induction time inꢀ
creases for the samples with a low content of the catalyst
active phase and with a decrease in the catalyst amount in
the reaction mixture. The induction time decreases with
an increase in the iron content and temperature increase.

Dichlorobutene isomerization on Fe2O3
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 6, June, 2005
1421
Yield of 1,4ꢀDCB (%)
70
lny
2.5
2.0
1.5
1.0
0.5
3
60
1
50
40
30
20
10
0
2
2´
2
3
1
10
20
30
40
50
t/min
500
1000
1500
2000
2500
t/s
Fig. 1. Kinetic curves of transꢀ1,4ꢀDCB accumulation on the
KSKꢀ2ꢀsupported iron oxide with a Fe content of 15 (1),
10 (2, 2´), and 5 wt.% (3); T = 100 (1—3) and 110 °С (2´);
catalyst weight 0.003 (1), 0.005 (2, 2´), and 0.010 g (3);
V3,4ꢀDCB = 0.1 mL.
Fig. 3. Kinetic curves of transꢀ1,4ꢀDCB accumulation at 90 (1),
100 (2), and 110 °С (3) on the catalyst supported on KSKꢀ2
with a Fe content of 10 wt.% in the coordinates lny—t (y =
(A∞ – A0)/(A∞ – A)).
isomerization rate decreases sharply, and the system does
not achieve the equilibrium yield of transꢀ1,4ꢀDCB, beꢀ
ing 77% at 100 °C.13 This is probably related to the fast
catalyst deactivation due to the formation of iron(II) chloꢀ
ride, which is inactive in isomerization.3 When 3,4ꢀDCB
is diluted with the polar chloroorganic solvent (dichloroꢀ
ethane), the induction period also decreases and disapꢀ
pears completely in excess solvent (see Fig. 2).
Thus, the isomerization of 3,4ꢀDCB is a complex proꢀ
cess, which cannot be described, in the general case, by
simple kinetic equations. Only under certain conditions
(at the iron content in the catalyst >10 wt.%, iron content
in the reaction mixture >0.08 mol L–1, and temperature
>90 °С) almost without induction period, the reversible
reaction formally obeys a kinetic equation of the first
order with respect to 3,4ꢀDCB. This is indicated by the
linearized kinetic curves shown in Fig. 3. The apparent
rate constants of the forward and backward reactions along
with the known equilibrium constants13 at different temꢀ
peratures are given in Table 3. Under these conditions,
Yield of 1,4ꢀDCB (%)
70
1
2
3
60
50
40
30
20
10
4
0
60
120
180
240
300
t/min
Fig. 2. Kinetic curves of transꢀ1,4ꢀDCB accumulation on the
KSKꢀ2ꢀsupported iron oxide (3 wt.% Fe) under different condiꢀ
tions: 90 °С, 0.6 mL of a 3,4ꢀDCB—dichloroethane (1 : 9)
mixture (1); 80 °С, 0.6 mL of a 3,4ꢀDCB—dichloroethane (1 : 9)
mixture (2); 90 °С, 0.6 mL of a 3,4ꢀDCB—dichloroethane (1 : 1)
mixture (3); 80 °С, 0.6 mL of 3,4ꢀDCB (4). The catalyst weight
is 0.004 g.
Table 3. Equilibrium constants (Keq)13 and apparent rate conꢀ
stants of the forward and backward reactions of 3,4ꢀDCB isomerꢀ
ization to transꢀ1,4ꢀDCB (k1 and k–1, respectively) at different
temperatures for the KSKꢀ2ꢀsupported catalyst with the Fe conꢀ
tent 10 wt.%
In the presence of the catalyst containing 15 wt.% iron,
no induction period is observed and a high initial rate of
the process is achieved (see Fig. 1). However, further the
Table 2. Duration of the induction period of 3,4ꢀDCB isomerꢀ
ization (τ) at different iron contents in the catalysts supported
on KSKꢀ2 at 100 °С and different catalyst amounts
T/°С
Keq
(k1 + k–1)•104
k1•104
k–1•104
s–1
Fe (wt.%)
Catalyst weight/g
τ+5/min
120
110
100
90
2.70
3.00
3.33
3.85
—
18.0+2.0
9.0+1.0
7.0+1.0
4.0+0.6
12.0+1.0
13.0+1.0
7.0+0.5
5.5+0.5
3.2+0.3
9.5+0.5
5.0+1.0
2.0+0.5
1.5+0.5
0.8+0.3
2.5+0.5
3
5
10
10
10
0.015
0.010
0.005
0.0025
0.0015
25
12
7
10
15
90*
* In dichloroethane.

1422
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 6, June, 2005
Rostovshchikova et al.
Asp/mole of 1,4ꢀDCB (mole Fe)–1 h–1
Yield of 1,4ꢀDCB (%)
50
1
200
150
100
40
2
30
20
10
3
2
4
6
8
10
[Fe] (wt.%)
0
6
12
18
24
t/min
Fig. 4. Specific catalytic activity (Asp) vs. iron content for the
catalysts supported on KSKꢀ2 at 100 °С.
Fig. 5. Kinetic curves of transꢀ1,4ꢀDCB accumulation at 100 °С
on the iron oxides supported on KSKꢀ2 (5 wt.% Fe) with differꢀ
ent degrees of reduction: Fe3O4 (1), Fe2O3 (2), and FeхO (3)
(0.01 g of catalyst, 0.1 mL of 3,4ꢀDCB).
the apparent activation energy of the forward reaction is
53+10 kJ mol–1, and that of the backward reaction is
69+12 kJ mol–1
.
The plot of the specific catalytic activity vs. iron conꢀ
tent in the catalyst is shown in Fig. 4. The specific activity
increases from 90 to 230 moles of 1,4ꢀDCB (mol Fe)–1 h–1
with an increase in the iron content from 3 to 15 wt.%
(see Fig. 4). However, it follows from the data of Xꢀray
diffraction and Mössbauer spectroscopy that the fraction
of large (>10 nm) nanoclusters increases with an increase
in the iron content ≥10 wt.%. Therefore, with the increase
in the iron content a decrease rather than an increase in
the specific activity can be expected. An enhancement of
the catalytic activity with an increase in the iron content
can be explained by a change in the composition of the
catalyst oxide phase found by Mössbauer spectroscopy. In
particular, the starting catalyst can contain the catalytiꢀ
cally more active Fe3O4 clusters3 or cubic γꢀiron oxide
with the spinel structure, which can be reduced with reacꢀ
tants to form mixedꢀcharge states.20
catalyst. Fine iron oxide nanoparticles (≤5—6 nm), preꢀ
vailing at low Fe contents on KSKꢀ2 (3 and 5 wt.%), are
capable of strong interacting with the support. Under these
conditions, the formation of a catalytically active strucꢀ
ture is energetically unfavorable, and the formation of the
active catalyst surface requires a long time. Larger iron
oxide particles (>10 nm) in the catalysts with 10 and
15 wt.% Fe are easily reduced with reactants and, hence,
the catalytically active structure is formed more rapidly. If
the catalyst contains iron ions in two oxidation states as in
magnetite already in the initial state, no induction period
is observed.
For larger iron oxide nanoparticles carried by KSKꢀ2
(10 and 15 wt.% Fe), the reaction with chloroolefins can
proceed until the complete conversion of iron(III) oxide
to iron(II) chloride is achieved.3 As a result, the catalysts
with high iron contents are not catalytically active when
used repeatedly.
To establish the effect of the charge state of a metal on
the activity of the system, the rate of 3,4ꢀDCB isomerizaꢀ
tion was determined in the presence of the catalysts conꢀ
taining other oxides in addition to iron(III) oxide: Fe3O4
or FexO (the parameters of the Mössbauer spectra of the
samples have been presented earlier3). The data obtained
are shown in Fig. 5. As can be seen, the catalyst containꢀ
ing Fe3O4 is the most active. At the same time, the presꢀ
ence of wustite FexO, whose lattice contains mainly FeII
ions, does not increase the catalytic activity. It can be
concluded that an active phase of the catalyst based on
magnetite, whose structure simultaneously contains the
FeII and FeIII ions, is formed by the action of reactants
and/or solvent during the induction period. This result
confirms the assumption that an increase in the activity
with an increase in the iron oxide content on the support
can be caused by the formation of mixedꢀcharge states
already in the starting catalyst or by the action of reacꢀ
tants, for example, involving γꢀiron oxide.
Catalytic properties of iron oxides supported on the
activated silica matrix. For the ASMꢀsupported iron oxide,
the samples with a low iron content (2.5 wt.%) have a
high catalytic activity. Already at 80 °С, the reaction ocꢀ
curs without an induction period, and the equilibrium
yield of transꢀ1,4ꢀDCB equal to ~80% is achieved within
10—15 min. Meanwhile, such a high yield is not achieved
even within several hours for the KSKꢀ2ꢀsupported cataꢀ
lyst with a similar iron content (3 wt.%) (see Fig. 2,
curve 4). For the both supported systems, the dilution of
the reaction mixture with dichloroethane enhances the
process rate and reduces the induction period. The kiꢀ
netic curves of transꢀ1,4ꢀDCB accumulation at 60 °С in
the presence of dichloroethane for two supported cataꢀ
lysts (3 wt.% Fe on KSKꢀ2 and 2.5 wt.% Fe on ASM) are
presented in Fig. 6. As can be seen, the presence of the
iron oxide supported on the ASM decreases the induction
time. The activity of iron oxide on the ASM estimated
from regions of the maximum rate at different temperaꢀ
tures is presented in Table 4. A comparison of these data
As shown (see Table 2), the duration of the induction
time increases with a decrease in the iron content in the

Dichlorobutene isomerization on Fe2O3
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 6, June, 2005
1423
Yield of 1,4ꢀDCB (%)
70
phous to Xꢀrays (lamellas) of silica favors the reduction of
some FeIII ions to FeII, and a possible electron transfer
between them provides a high catalyst activity in redox
steps of the reaction.20
60
1
2
50
40
30
20
10
Thus, with the use of the activated silica matrix for
stabilization of iron oxide nanoparticles a more active
catalyst can be formed. Reasons for high activity of the
catalysts immobilized on silicas remain unclear. The data
of this work confirmed that chloroolefin isomerization
assuming the С—Cl bond cleavage is, most likely, a comꢀ
plex multistep process involving both oxidation—reducꢀ
tion and acid sites of the catalyst via different routes.3,7—9
Evidently, the support nature, particle size, and simultaꢀ
neous presence of two charge iron states in the catalyst are
key factors determining the optimum properties of the
catalyst.
0
40
80
120
160
200
t/min
Fig. 6. Kinetic curves of transꢀ1,4ꢀDCB accumulation at 60 °С
in the presence of the iron oxide supported on ASM (2.5 wt.% Fe,
0.005 g of catalyst, 0.6 mL of a 3,4ꢀDCB—dichloroethane (1 : 9)
mixture) (1) and KSKꢀ2 (3 wt.% Fe, 0.004 g of catalyst, 0.6 mL
of a 3,4ꢀDCB—dichloroethane (1 : 9) mixture) (2).
The authors are grateful to P. A. Chernavskii for synꢀ
thesis of the catalysts by temperatureꢀprogrammed reducꢀ
tion and to G. V. Murav´eva for their Xꢀray diffraction
study.
Table 4. Maximum isomerization rate (V ) and specific catalytic
activity (Asp) of iron oxide immobilized on the ASM in the
isomerization of 3,4ꢀDCB at different temperatures
This work was financially supported by the Russian
Foundation for Basic Research (Project Nos 01ꢀ03ꢀ32783,
03ꢀ03ꢀ33104, and 03ꢀ03ꢀ32029).
T/°С
V/mol L–1 h–1
Asp
/mole of 1,4ꢀDCB (mol Fe)–1 h–1
60
70
80
0.68
1.05
1.73
180
280
470
References
1. H. Hayashi, L. Z. Chen, T. Tago, M. Kishida, and
K. Wakabayshi, Appl. Catal. A: Gen., 2002, 231, 81.
2. C. Kurts, Y. Arita, W. Weiss, W. Ranke, and R. Schlögl,
Topics in Catal., 2001, 14, 111.
3. T. N. Rostovshchikova, O. I. Kiseleva, G. Yu. Yurkov, S. P.
Gubin, D. A. Pankratov, Yu. D. Perfil´ev, V. V. Smirnov,
P. A. Chernavskii, and G. V. Pankina, Vestn. Mosk. Gos.
Univ., Ser. 2. Khim., 2001, 42, 318 [Vestn. Mosk. Univ., Ser.
Khim., 2001, 42 (Engl. Transl.)].
4. M. V. Tsodikov, Ya. R. Katsobashvili, E. V. Slivinskii, Yu. V.
Maksimov, V. V. Matveev, I. P. Suzdalev, O. G. Ellert, and
G. F. Ivanova, Izv. Akad. Nauk SSSR. Ser. Khim., 1986,
2666 [Bull. Acad. Sci. USSR, Div. Chem. Sci., 1986, 35, 2443
(Engl. Transl.)].
5. Yu. V. Maksimov, M. V. Tsodikov, M. A. Perederii, I. P.
Suzdalev, A. I. Nekhaev, V. T. Popov, and O. V. Bukhtenko,
Izv. Akad. Nauk. Ser. Khim., 1997, 86 [Russ. Chem. Bull.,
1997, 46, 81 (Engl. Transl.)].
6. P. A. Chernavskii, V. V. Kiselev, and V. V. Lunin, Zh. Fiz.
Khim., 1992, 66, 2712 [Russ. J. Phys. Chem., 1992, 66 (Engl.
Transl.)].
7. T. N. Rostovshchikova, V. V. Smirnov, and A. I. Kokorin,
J. Mol. Catal. A: Chem., 1998, 129, 141.
8. T. N. Rostovshchikova, V. V. Smirnov, and E. N.
Golubeva, Khim. Fiz. [Chemical Physics], 1998, 17, 63 (in
Russian).
9. L. I. Trakhtenberg, G. N. Gerasimov, V. K. Potapov, T. N.
Rostovshchikova, V. V. Smirnov, and V. Yu. Zufman, Vestn.
Mosk. Gos. Univ., Ser. 2. Khim., 2001, 42, 325 [Vestn. Mosk.
Univ., Ser. Khim., 2001, 42 (Engl. Transl.)].
with the results for the catalysts on silica gel (see Fig. 4)
shows that even at a lower temperature (80 °С) the
ASMꢀsupported catalyst is approximately twofold more
active than the most efficient catalyst on KSKꢀ2 as a
carrier with the 15% iron content. The rate constants of
the forward and backward reactions, which were estiꢀ
mated from the kinetic data, in the presence of iron oxide
on the ASM at 90 °С were 5.7•10–4 and 2.1•10–4 s–1
,
respectively. These values approximately twofold exceed
the isomerization rate constants obtained for the catalysts
with the optimum composition prepared using a KSK
support (see Table 3).
The ASMꢀimmobilized iron oxide is also superior to
the catalyst obtained using KSKꢀ2 by stability to reducꢀ
tion. As can be seen from the data in Table 1, after the
ASMꢀsupported iron oxide was used in isomerization (on
heating the reaction mixture for several hours at 80 °С),
the Mössbauer spectra retain signals from the FeIII and
FeII ions. The parameters of these signals are close to
those observed for the starting sample, and the relative
fraction of the reduced form increases approximately twoꢀ
fold in the spent catalyst. An enhanced stability of the
ASMꢀimmobilized FeIII oxide to complete reduction
compared to the iron oxides on KSKꢀ2 is related, probꢀ
ably, to the strong interaction of the metal ion with layer
silica. The presence of structural defects in particles amorꢀ

1424
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 6, June, 2005
Rostovshchikova et al.
10. V. V. Matveev, Yu. V. Maksimov, I. P. Suzdalev, M. V.
Tsodikov, and I. A. Litvinov, Hyperfine Interact., 1990,
56, 1707.
11. P. A. Chernavskii, G. V. Pankina, and V. V. Lunin, Zh. Fiz.
Khim., 1996, 70, 1016 [Russ. J. Phys. Chem., 1996, 70 (Engl.
Transl.)].
12. M. V. Tsodikov, Doct. Sci. (Chem.) Thesis, A. V. Topchiev
Institute of Petrochemical Synthesis, Russian Academy of
Sciences, Moscow, 1992 (in Russian).
15. Yu. V. Maksimov, M. V. Tsodikov, O. G. Ellert, O. V.
Bukhtenko, and V. V. Matveev, J. Catal., 1994, 148, 119.
16. Yu. V. Maksimov, M. V. Tsodikov, E. A. Trusova, I. P.
Suzdalev, and J. A. Navio, Catal. Lett., 2001, 72, 11.
17. F. Menil, J. Phys. Chem. Sol., 1985, 46, 763.
18. S. Morup, H. Topsoe, and J. Lipka, J. Phys. (Paris), 1976,
37, C6ꢀ287.
19. S. J. Oh, D. C. Cook, and H. E. Townsend, Hyperfine
Interact., 1998, 112, 59.
13. E. M. Asatryan, N. L. Saakyan, I. V. Gul´neva, and G. T.
Martirosyan, Arm. Khim. Zh., 1986, 39, 584 (in Russian).
14. I. P. Suzdalev, Dinamicheskie effekty v gammaꢀrezonansnoi
spektroskopii [Dynamic Effects in GammaꢀResonance
Spectroscopy], Atomizdat, Moscow, 1979, 192 pp. (in
Russian).
20. O. S. Morozova, Yu. V. Maksimov, D. P. Shashkin, P. A.
Shirjaev, V. A. Zhorin, and O. V. Krylov, Appl. Catal., 1991,
78, 227.
Received December 30, 2003;
in revised form April 6, 2005