Catalytic isomerization of allylbenzene on organomagnesium clusters

Article abstract of DOI:10.1007/s11172-005-0378-1

Catalytic isomerization of allylbenzene to form trans-β-methylstyrene quantitatively occurs on the magnesium - anthracene cluster adduct under mild conditions. A low-stability organomagnesium compound, presumably of the cluster nature and active in catalytic allylbenzene isomerization, is formed by cocondensation of magnesium and allylbenzene vapors at the liquid nitrogen temperature. The products of low-temperature solid-phase reactions of magnesium with hydrocarbons containing aromatic rings exhibit high catalytic activity in the allyl isomerization of allylbenzene.

Full text of DOI:10.1007/s11172-005-0378-1

Russian Chemical Bulletin, International Edition, Vol. 54, No. 5, pp. 1185—1188, May, 2005
1185
Catalytic isomerization of allylbenzene on organomagnesium clusters
ꢀ
D. A. Potapov, L. A. Tjurina, and V. V. Smirnov
Department of Chemistry, M. V. Lomonosov Moscow State University,
1 Leninskie Gory, 119992 Moscow, Russian Federation.
Fax: +7 (095) 932 8846. Eꢀmail: smirnov@kinet.chem.msu.ru, tyurina@kinet.chem.msu.ru
Catalytic isomerization of allylbenzene to form transꢀβꢀmethylstyrene quantitatively ocꢀ
curs on the magnesium—anthracene cluster adduct under mild conditions. A lowꢀstability
organomagnesium compound, presumably of the cluster nature and active in catalytic
allylbenzene isomerization, is formed by cocondensation of magnesium and allylbenzene vaꢀ
pors at the liquid nitrogen temperature. The products of lowꢀtemperature solidꢀphase reactions
of magnesium with hydrocarbons containing aromatic rings exhibit high catalytic activity in
the allyl isomerization of allylbenzene.
Key words: catalysis, isomerization, allylbenzene, magnesium, metal vapor synthesis, magꢀ
nesium—anthracene adduct.
Magnesium compounds exhibit no catalytic activity in
hydrocarbon transformations under conventional condiꢀ
tions, although some data indicate the activity of magneꢀ
siumꢀcontaining products of metal vapor synthesis (MVS)
in hydrocarbon conversion. For instance, the study of the
magnesium cocondensates with propylene1,2 revealed
deuterioꢀhydrogen exchange. On heating the magnesium
cocondensates with hexene, anthracene, indene, and
1ꢀmethylindene the isomerization of the starting olefins
and products of their autohydrogenation occurs.3,4 In
many cases, the product : magnesium ratio is higher than
unity, indicating the catalytic character of the processes.
The catalytic isomerization and hydrogenation of unsatꢀ
urated hydrocarbons are characteristic of strong acids or
systems based on transition metals. It was assumed4,5 that
the catalytic activity of samples containing only magneꢀ
sium and hydrocarbon is caused by the formation of acꢀ
tive magnesium clusters.
It has previously6—8 been shown that the Grignard
reagent cluster containing the Mg4 core can be formed in
magnesium—halohydrocarbon systems at low temperaꢀ
tures. The quantumꢀchemical calculations9,10 show that
Mg4 is the most stable of the small magnesium clusters. It
has recently been shown that cluster compounds are also
formed by the reactions of magnesium with some hydroꢀ
carbons. For instance, derivatives containing the Mg4 clusꢀ
ter appear in the magnesium—anthracene system. Their
formation was proved by MALDI TOFꢀMS mass specꢀ
trometry.11
catalytic conversion of hydrocarbons in the presence of
preliminarily prepared individual organomagnesium clusꢀ
ter derivatives has not been proved reliably to date.
The purpose of the present work is to carry out the
catalytic isomerization of olefins (using allylbenzene as
an example) in the presence of the preliminarily prepared
organomagnesium cluster adduct and to reveal conditions
for the reaction to proceed quantitatively.
Experimental
Cocondensates of magnesium with organic compounds were
obtained by the simultaneous precipitation of magnesium and
organic reactant vapors on a vacuumꢀactivated reactor surface
cooled from outside with liquid nitrogen. Reactors of two types
were used: "immerse" (A) and "spectral" (B). In reactor A, conꢀ
densation is conducted on the internal surface of the vacuum
reactor cooled from outside with liquid nitrogen. Reactor B is a
glass vacuum vessel applied for IR spectroscopic studies of the
cocondensates. The vessel has a potassium bromide window,
and a polished copper unit cooled with liquid nitrogen is placed
into the reactor. Reagents are condensed on this unit. When
IR spectra are recorded, a beam from a spectrometer is directed
by a system of mirrors through the KBr window and a condenꢀ
sate film onto the copper unit, reflected, again passed through
the KBr window, and directed to the spectrometer detector.12
Cocondensation was carried out under a pressure of ~10–4 Torr
inside the reactors. Magnesium was evaporated from a quartz
evaporator at 400—450 °C. The evaporation temperatures of
anthracene (AN) and allylbenzene (AB) were selected to ensure
the AN/Mg ratio of 50—200 and AB/Mg ratio of 10—110 during
condensation. Reagents Mg, AN, and AB were condensed with
rates of (1.6—4.0)•10–5, (0.9—8.5)•10–3, and (0.1—3.8)•10–3
mol h–1, respectively. The condensation duration in reactors A
and B was 0.5—1.5 h and 10—30 min, respectively.
All these data agree with our previous4,5 supposition
that the catalytic activity of the MVS products involving
magnesium is caused by the presence of magnesium clusꢀ
ters with unsaturated coordination sites. At the same time,
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 5, pp. 1155—1158, May, 2005.
1066ꢀ5285/05/5405ꢀ1185 © 2005 Springer Science+Business Media, Inc.

1186
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 5, May, 2005
Potapov et al.
When cocondensation of Mg and AN vapor was completed,
reactor A was warmed to room temperature to form a magneꢀ
sium—anthracene adduct. The moment of reaction cessation
was determined from the time necessary for the complete decolꢀ
oration of the condensate.11 After the end of the reaction, a
certain amount of allylbenzene was introduced into the condenꢀ
sate. The liquid phase was stored for 0.5—2 h, recondensed into
an ampule, and analyzed by GCꢀMS and GLC.
IR spectra were recorded with an Infralyum FTꢀ801 specꢀ
trometer (LyumeksꢀSibir´). GCꢀMS spectra were obtained on a
Finnigan MATꢀ212 spectrometer. A Kristall Lyuks 4000 instruꢀ
ment was used for GLC.
The cocondensation of Mg and AB vapors was carried out in
the reactors of both types. When reactor A was used, after
cocondensation to 80 K, the reactor was heated to a specified
temperature within 200—298 K, and this temperature was mainꢀ
tained constant for 5—100 min. To stop the reaction, excess
(over magnesium) water amount was introduced into the sysꢀ
tem, the reactor was warmed to room temperature, and the
liquid phase was recondensed into an ampule and analyzed by
GCꢀMS and GLC. When Mg and AB vapors were condensed in
reactor B, the IR spectra of the cocondensate film were recorded
at specified temperatures in an interval of 80—160 K. Special
testing experiments showed that olefins are not transformed in
the absence of Mg and in the presence of the preliminarily
obtained film of "bulk" metal.
autohydrogenation of the starting olefin. Nonvolatile
products of olefin condensation, which are formed, most
likely, in parallel, remain in the precipitate along with the
metal when organic compounds are evaporated.
Dihydroanthracene is formed along with the allylꢀ
benzene conversion in the catalytic autohydrogenation of
excess anthracene on the adduct.4 The yields of the prodꢀ
ucts at different ratios of AB, AN, and Mg—AN adduct
are presented in Table 1. The yield of transꢀβꢀmethylꢀ
styrene is controlled only by excess AB introduced into
the reaction and, most likely, can be much higher than
that presented in Table 1. Since the reaction ceases within
2—3 min, a very high catalytic activity of the Mg—AN
adduct in isomerization under mild conditions can be
implied. The activity in autohydrogenation is much lower,
and the complete anthracene conversion is not achieved.
Note that the yield of 9,10ꢀdihydroanthracene calculated
for magnesium increases with an increase in the AN/Mg
ratio.
It was of interest to find out whether isomerization
can occur directly in the cocondensate of Mg and AB.
Available published data suggest that the cocondensation
of magnesium with substituted benzenes produces unꢀ
stable cluster adducts similar in nature to the anthracene
adducts.13 These adducts are also active, most likely, in
isomerization.
The Mg content in the obtained samples was analyzed by
complexonometric titration with EDTA in the presence of the
indicator Eriochrome black T.
The cocondensation of the magnesium and AB vapors
at 80 K produces a brown film. The analysis of the liquid
phase obtained after melting of the film shows that
it contains 89—95% transꢀβꢀmethylstyrene and small
amounts (2—5%) of nꢀpropylbenzene and cisꢀβꢀmethylꢀ
styrene.
Thus, the complete conversion of AB to thermodyꢀ
namically more stable isomers and autohydrogenation
products occurs in the Mg—AB system. It can be assumed
that, as in the case of the Mg—AN adduct, the catalysts of
this reaction are the magnesium cluster derivatives formed
by the reaction of small magnesium clusters with AB.
Results and Discussion
The Mg—AN adduct is stable in vacuo at room temꢀ
perature.11 This adduct catalyzes the transformation of
AB at room temperature, and the complete conversion
occurs within several minutes. The main product is the
most thermodynamically stable AB isomer transꢀβꢀmeꢀ
thylstyrene; in addition, two byꢀproducts, viz., cisꢀβꢀmeꢀ
thylstyrene and nꢀpropylbenzene, are formed in small
quantities (0.1—3.0%). The latter is the product of
Table 1. Composition of the reaction products of allylbenzene (AB) with the preliminarily prepared magnesium—anthracene adduct
(MAA) at 300 K
AB : AN* : Mg
AB : AN : MAA**
Conversion
(%)
Yield/mol (mol Mg)–1
(in parentheses, mol (mol MAA)–1
)
AB
AN
transꢀβꢀ
Methylstyrene
cisꢀβꢀ
Methylstyrene***
9,10ꢀDihydroꢀ
anthracene
17.2 : 95 : 1
46.3 : 70 : 1
145.7 : 50 : 1
317.5 : 200 : 1
483 : 120 : 1
68 : 379 : 1
184 : 279 : 1
584 : 199 : 1
1272 : 799 : 1
1932 : 479 : 1
100
100
100
100
100
17
31.5
54
16
32
17 (68)
0.2 (0.8)
0.5 (2)
2.4 (9.6)
1.7 (6.8)
3 (12)
16 (64)
22 (88)
27 (108)
32 (128)
39 (156)
45.8 (183)
143.3 (580)
315.8 (1263)
480 (1920)
* Anthracene.
** Calculated from the empirical formula of MAA С14H10Mg4 (see Ref. 4).
*** Along with nꢀpropylbenzene.

Catalytic allylbenzene isomerization
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 5, May, 2005
1187
C (%)
The Mg—AB system was studied by IR spectroscopy
at 80—160 K. The spectrum of the condensate at
80—120 K is almost identical to the spectrum of AB. No
new distinct bands that could be ascribed to the organoꢀ
magnesium compound were observed. This is probably
related to the fact that the adduct is formed in the presꢀ
ence of excess AB. The strong bands in the IR spectra of
AB prevent recording the spectra of the adduct. At 120 K,
bands characteristic of transꢀβꢀmethylstyrene appear in
the IR spectra at 2733, 1490—1480, and 1284 cm–1 i.e.,
isomerization starts already at this temperature long beꢀ
fore melting of the solid sample.
T = 100 K
T = 150 K
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
5
10
15
t/min
The stepwise curve describing the kinetics of
βꢀmethylstyrene accumulation is observed in the temꢀ
perature range from 150 to 160 K. The typical kinetic
curve is presented in Fig. 1. It is seen that the AB converꢀ
sion is ~0.5% within the 7ꢀmin reaction in the Mg—AB
mixture at 120 K. The subsequent warming of the sample
to 150 K increases the conversion to 3.0—3.8% within
2—3 min. The further storage of the reaction mixture
does not increase the yield of the product. A similar time
dependence of the product yield at 150—160 K was obꢀ
tained in all experiments. Therefore, this reaction proꢀ
ceeds in the stepwise kinetic regime typical of solidꢀphase
processes. Unfortunately, the cocondensate films become
nontransparent on further warming to temperatures above
160 K, which prevents recording the IR spectra. This did
not allow us to obtain a set of several "steps" in the kinetic
curve and determine the activity distribution of reaction
sites of the solidꢀphase process. However, there remains
nearly no doubt about the stepwise character of catalytic
isomerization in the solid phase. The reason for this beꢀ
havior can lie in the environmental and structural differꢀ
ences between intermediate compounds (precursors of
the reaction products) stabilized in the solid matrix.14
Isomerization is accompanied by a change in the geomꢀ
etry of molecules and, therefore, requires a certain rotaꢀ
Fig. 1. Conversion of allylbenzene (C) in the cocondensate with
magnesium vs. reaction duration. Ratio allylbenzene : magneꢀ
sium = 21.5 : 1.
tional (and, perhaps, some translational) mobility of molꢀ
ecules. As the temperature increases, the mobility of some
AB molecules increases and their reactivity enhances. At
the same time, species, which exist in less favorable conꢀ
figurations or are more rigidly fixed by surrounding molꢀ
ecules, do not react.12
The studies at temperatures higher than 160 K were
carried out in an immersionꢀtype reactor (Table 2). The
reaction during allylbenzene matrix melting (at
230—235 K) occurs with quantitative yield. Under these
conditions, as in the case of the process in the presence of
the magnesium—anthracene adduct, the yield calculated
for magnesium is determined only by excess AB. The
addition of AB to the cocondensate warmed above 240 K
when its melting was completed provides no additional
quantities of transꢀβꢀmethylstyrene. This indicates the
decay of a catalytically active compound. A reason for the
decay is, most likely, the aggregation of magnesium clusꢀ
ters to a compact metal. The formation of organoꢀ
magnesium compounds unstable in solution at room temꢀ
Table 2. Conversion of allylbenzene (AB) and the yield of transꢀβꢀmethylstyrene for the Mg—AB systems
Reaction temperature/K
Conversion
of allylbenzene (%)
Allylbenzene/Mg
ratio
Yield of transꢀβꢀmethylstyrene
/mol (mol Mg)–1
80
120
130
150
0
25, 52, 110
0
<0.1
0.5
0.8
6.8
<1
1.5
3.8
13
47
34
21.5
52
210
220
230—235
15
100
58.5
11
8.9
9.4
54
81
109
—
52.2
78.4
103.2
0
230—235*
0
* Magnesium is absent from the condensate film.

1188
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 5, May, 2005
Potapov et al.
4. G. B. Barkovskii, L. A. Tjurina, and V. V. Smirnov, Vestn.
MGU, Ser. 2, Khimiya, 1993, 34, 55 [Vestn. Mosk. Univ., Ser.
Khim., 1993, 34 (Engl. Transl.)].
5. L. A. Tjurina, V. V. Smirnov, and D. A. Potapov, Kinet.
Katal., 2003, 44, 640 [Kinet. Catal., 2003, 44 (Engl. Transl.)].
6. L. A. Tjurina, V. V. Smirnov, and I. P. Beletskaya, Mendeleev
Commun., 2002, 12, 108.
7. L. A. Tjurina, V. V. Smirnov, G. B. Barkovskii, E. N.
Nikolaev, S. E. Esipov, and I. P. Beletskaya, Organometalꢀ
lics, 2001, 20, 2449.
8. V. V. Smirnov, L. A. Tjurina, and I. P. Beletskaya, in Grignard
Reagents. New Developments, Ed. H. G. Richey, Wiley and
Sons, Chichester, 2000, p. 395.
perature has previously been observed for the magnesium
cocondensates with alkylbenzenes.13 It is most likely that
catalytically active organomagnesium clusters are formed
by the reactions of magnesium with all derivatives conꢀ
taining aromatic rings. However, unlike the magneꢀ
sium—anthracene cluster adduct,11 the magnesium adꢀ
ducts with alkylbenzenes and AB are less stable and decay
when the matrix melts. Nevertheless, due to their high
catalytic activity, they induce the complete conversion
of AB, which is present in the system in excess, even
within the short lifetime under the conditions of high
mobility of the reactants.
9. J. Akola, K. Rytkonen, and M. Mannnien, Eur. Phys. J., D,
2001, 16, 21.
This work was financially supported by the Russian
Foundation for Basic Research (Project Nos 03ꢀ03ꢀ32ꢀ
314 and 05/0332909).
10. A. Kohn, F. Weigend, and R. Ahlrichs, Phys. Chem. Chem.
Phys., 2001, 3, 711.
11. L. A. Tjurina, V. V. Smirnov, and I. P. Beletskaya, J. Mol.
Catal., 2002, 395.
12. Cryochemistry, Eds M. Moskovits and G. A. Ozin, John Wiley
and Sons, New York, 1976.
References
13. G. B. Barkovskii, L. A. Tjurina, and V. V. Smirnov, Vestn.
MGU, Ser. 2, Khimiya, 1991, 32, 464 [Vestn. Mosk. Univ.,
Ser. Khim., 1991, 32 (Engl. Transl.)].
14. Ya. S. Lebedev, Kinet. Katal., 1978, 19, 367 [Kinet. Catal.,
1978, 19 (Engl. Transl.)].
1. I. C. Hisatsune, J. Catal., 1982, 74, 9.
2. I. C. Hisatsune, J. Catal., 1982, 74, 18.
3. V. V. Smirnov, I. P. Beletskaya, L. A. Tjurina, G. B.
Barkovskii, A. N. Kashin, and I. G. Tarkhanova, Kinet.
Katal., 1998, 39, 885 [Kinet. Catal., 1998, 39 (Engl.
Transl.)].W
Received April 6, 2004;
in revised form January 25, 2005