Synthesis and properties of homogeneous models of the Re2O7/Al2O3 metathesis catalyst

Article abstract of DOI:10.1016/S1381-1169(00)00284-3

Soluble complexes, including aluminum coordinated with a perrhenate and a 4-substituted-2-6-di-tert-butylphenoxy ligand or a CH₂- or S-ortho bridged-4,4'-di-tert-butyldiphenoxy ligand, were synthesized and characterized. The reaction of (ArO)₂Al-Cl with AgReO₄ resulted in new complexes (ArO)₂Al-OReO₃ (A), which showed a low activity in the metathesis of 2-pentene. Complexes (ArO)₂Al-R (R = isobutyl, ethyl) reacted with Re₂O₇ in tetrahydrofuran or dioxane, generating type B complexes including oligomeric linkages such as O₃Re-[Al(OAr)-O]₂-ReO₃. Some of these complexes, unstable even under an argon atmosphere, were identified after stabilization with a Lewis base like quinuclidine or bipyridine. Before their stabilization, they exhibited a rather high activity in the metathesis of 2-pentene, with turnover frequency of 100/min, and functional olefins (diallylether, diallylsulfide). Upon reaction with (ArO)₂Al-R in an ether solvent, A complexes might be converted into B complexes. The high activity of B complexes was associated with the presence of Al-O-Al linkages that are known to induce a powerful bidendate Lewis acidity.

Full text of DOI:10.1016/S1381-1169(00)00284-3

Ž
.
Journal of Molecular Catalysis A: Chemical 161 2000 125–140
www.elsevier.comrlocatermolcata
Synthesis and properties of homogeneous models of the
Re2O7rAl2O3 metathesis catalyst
Gwenael Doledec, Dominique Commereuc)
Institut Franc¸ais du Petrole, 1 et 4 aÕenue de Bois-Preau, Rueil-Malmaison Cedex F-92852, France
´
´
Received 15 March 2000; received in revised form 15 May 2000; accepted 22 May 2000
Abstract
The aim of this work was to synthesize and to study homogeneous models of the rhenium oxide on alumina metathesis
Ž
.
catalyst. A series of aluminium complexes, ArO 2 Al–Y, has been synthesized, where ArO is a 4-substituted-2,6-di-tert-
butylphenoxy ligand, or ArO is a –CH2– or –S-ortho bridged-4,4X-di-tert-butyldiphenoxy ligand, and Y is an alkyl or
Ž
.
2Ž
.
Ž
.
Ž .
chloride ligand. The reaction of ArO 2 Al–Cl with AgReO4 led to new complexes ArO 2 Al–OReO3 A . These complexes
Ž
.
Ž
Ž
.
exhibit a low activity in the metathesis of 2-pentene. Complexes ArO 2 Al–R RsiBu, Et react with Re2O7 in THF or
w
.
x
dioxane, giving type B complexes including oligomeric linkages like O3Re– Al OAr –O –ReO3. They show a fairly high
2
activity in the metathesis of 2-pentene, with TOF values as high as 100 miny1. Complexes of type A may be converted into
Ž
.
complexes of type B upon reaction with ArO 2 Al–R in an ether solvent. The high activity of B complexes is tentatively
related to the presence of Al–O–Al linkages that are known to induce a powerful bidentate Lewis acidity. q 2000 Elsevier
Science B.V. All rights reserved.
Keywords: Metathesis; Catalysis; Aluminium; Aryloxy; Rhenium
w x
1. Introduction
is involved at a given time in the metathesis 1 .
Several studies demonstrated that the acidity of the
alumina surface has a great importance on the activ-
The metathesis reaction occurs in the presence of
both homogeneous and heterogeneous catalysts. The
rhenium heptoxide on alumina heterogeneous cata-
lyst is active even at low temperatures, allowing, in
principle, comparisons with homogeneous catalysts.
The activity of the catalyst increases with the amount
of rhenium and a high rhenium loading is required to
obtain a reasonably active catalyst. Besides, it has
been demonstrated that less than 1% of the rhenium
Ž
ity even if the nature of this acidity Lewis or
.
w
x
Bronsted is discussed 2–6 .
¨
There are only few rhenium complexes that are
active for the olefin metathesis reaction. The methyl-
w
x
trioxorhenium complex of Herrmann 7,8 requires
activation by a Lewis acid and alkylating cocatalyst.
The alkylidene complexes of Schrock are active
provided that an electron-attracting alkoxy ligand is
w
x
coordinated to the rhenium 9–17 . Some kind of
acidity is generally required in a metathesis catalyst,
either a ligand in an homogeneous catalyst or the
carrier itself in an heterogeneous catalyst.
)
Corresponding author.
E-mail address: dominique.commereuc@ifp.fr
Ž
.
D. Commereuc .
1381-1169r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S1381-1169 00 00284-3

(
)
126
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
We have been interested to synthesize homoge-
under reflux was added a solution of 2,6-di-tert-
Ž
.
neous models of the hypothetical surface species
present in the heterogeneous catalyst. Such com-
plexes should include an aluminium–perrhenate
bridge, which is present at the surface of the alumina
carrier. The ligands that we selected to complete the
aluminium coordination sphere were substituted ary-
loxy ligands that exhibit well known advantages:
changing the substituents of the aryloxy allows to
modify the acidity of the aluminium and its steric
crowding.
This paper reports the synthesis and the character-
isation of soluble complexes including aluminium
coordinated with a perrhenate and a 4-substituted-
2,6-di-tert-butylphenoxy ligand or a CH2– or S-or-
tho bridged-4,4X-di-tert-butyldiphenoxy ligand. Some
of these complexes have been found to have a high
activity for the metathesis of internal alkenes without
butyl-4-methylphenol 17.46 g, 79.2 mmol in 60 ml
pentane with a canula. The addition was done in 30
min. The medium was stirred under reflux for 70
Ž
min then allowed to cool at room temperature 70
min . Pentane was removed under vacuum to give
20.55 g of a white powder. H NMR showed that
98% of 1a and 2% of phenol were obtained hs
.
1
Ž
.
98% .
1
Ž
.
Ž
.
H NMR C6 D6 : ds7.17 ppm s, Har ; 2.28
3
Ž
.
Ž
ppm s, H-Me ; 1.70 ppm t-hpt, Js6.5–7.1 Hz,
t
3
.
Ž
.
Ž
H-b ; 1.61 ppm s, H- Bu ; 0.76 ppm d, Js6.5
3
.
Ž
.
Hz, H-g ; 0.46 ppm d, Js7.1 Hz, H-a .
(
)
2.1.1.2. Bis 2,6-di-tert-butylphenoxy isobutylalumi-
(
)
nium 1b . This compound was prepared as de-
scribed for the synthesis of 1a. 1H NMR showed that
Ž
95% of 1b and 5% of phenol were obtained hs
w
x
the addition of any cocatalyst 18 . As far as we
know, no complex of this kind has been described in
the literature. A preliminary account of this work has
.
69% .
1
3
Ž
.
Ž
X
H NMR C6 D6 : ds7.30 ppm d, Js7.8 Hz
3
.
Ž
.
Har ; 6.91 ppm dd, Js7.8 Hz Har ; 1.79 ppm
w
x
3
t
been published 19 .
Ž
.
Ž
.
t-hpt, Js6.5–7.2 Hz, H-b ; 1.57 ppm s, H- Bu ;
3
3
Ž
.
Ž
0.73 ppm d, Js6.5 Hz, H-g ; 0.43 ppm d, Js
.
7.2 Hz, H-a .
2. Experimental
(
)
2.1.1.3. Bis 4-bromo-2,6-di-tert-butylphenoxy -
All operations were performed under an inert
atmosphere of argon using standard Schlenk line or
glove box techniques. Toluene, THF, dioxane and
diethyl ether were distilled from sodium benzo-
phenone. Pentane and heptane were purified by dis-
tillation from LiAlH4.
Phenols, alkylaluminium and rhenium oxide were
commercially available and used as received. Olefins
were purified by distillation from sodium. Elemental
analysis was performed by Mikroanalytisches Labor
(
)
isobutylaluminium 1c . This compound was pre-
pared as described for the synthesis of 1a. H NMR
showed that 98% of 1c and 2% of phenol were
1
Ž
.
obtained hs72% .
1H NMR C6 D6 :ds7.40 ppm s, Har ; 1.50
Ž
.
Ž
.
3
Ž
.
.
Ž
ppm t-hpt, Js6.5–7.2 Hz, H-b ; 1.36 ppm s,
t
3
Ž
.
H- Bu ; 0.51 ppm d, Js6.5 Hz, H-g ; 0.15 ppm
3
Ž
.
d, Js7.2 Hz, H-a .
Elemental analysis: exp.: C 58.66%, H 7.65%, Br
24.5%, Al 4.31%; th.: C 58.90%, H 7.57%, Br
24.5%, Al 4.13% for 1c.
Ž
.
Pascher Germany . Nuclear magnetic resonance
Ž
.
NMR spectra were recorded on a Bruker CXP 200.
Gas chromatography analyses were done on a HP
5890 series II.
(
)
2.1.1.4. Bis 2,6-di-tert-butyl-4-methoxyphenoxy -
(
)
isobutylaluminium 1d . This compound was pre-
1
2.1. Synthesis of bisaryloxyaluminium compounds
pared as described for the synthesis of 1a. H NMR
spectrum showed that 97% of 1d and 3% of phenol
(
)
(
)
2.1.1. Alkylbis aryloxy aluminium: ArO 2 Al–R
Ž
.
were obtained hs70% .
1H NMR C6 D6 : ds7.03 ppm s, H-arom ;
Ž
.
Ž
.
3
(
)
Ž
.
Ž
2.1.1.1. Bis 2,6-di-tert-butyl-4-methylphenoxy -
3.51 ppm s, H-OMe ; 1.84 ppm t-hpt, Js6.6–7.2
t
3
(
)
. Ž . Ž
Hz, H-b ; 1.58 ppm s, H- Bu ; 0.80 ppm d, Js6.6
3
isobutylaluminium 1a . To a solution of triisobutyla-
luminium 10 ml, 39.6 mmol in 30 ml pentane
Ž
.
. Ž .
Hz, H-g ; 0.47 ppm d, Js7.2 Hz, H-a .

(
)
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
127
(
)
2.1.1.5. Bis 4-bromo-2,6-di-tert-butylphenoxy -
allowed to warm at room temperature. The resulting
precipitate was filtered to afford 4.2 g of a white
(
)
ethylaluminium 1e . This compound was prepared
as described for the synthesis of 1a. The H NMR
1
Ž
.
powder hs69% .
1
X
X
4
Ž
.
Ž
.
Ž
showed that 100% of 1e was obtained hs79% .
H NMR C6 D6 : ds7.22 ppm AA BB , Js
1
2
Ž
.
Ž
.
.
.
Ž
H NMR C6 D6 : ds7.51 ppm s, Har ; 1.34
2.2 Hz y0 ds23.3 Hz, H-arom ; 4.07 ppm d, Js
t
3
X
2
Ž
Ž
.
Ž
.
.
Ž
.
ppm s, H- Bu ; 0.86 ppm t, Js8.0 Hz, H-b ; 0.27
13.6 Hz Hp ; 3.46 ppm d, Js13.6 Hz Hp ; 3.10
3
3
Ž
.
Ž
ppm q, Js8.0 Hz, H-a .
ppm s, CH3–MeOtBu ; 2.28 ppm t-hpt Js6.6–7.2
.
Ž
.
Ž
.
Hz, H-b ; 2.23 ppm s, H-Me ; 1.57 ppm s, H-tBu ;
3
Ž
.
Ž
1.27 ppm d, Js6.6 Hz, H-g ; 1.01 ppm s, tBu-
(
)
2.1.1.6. Bis 2,4,6-tri-tert-butylphenoxy ethylalumi-
nium 1f . This compound was prepared as described
for the synthesis of 1a. H NMR showed that 100%
3
.
Ž
.
MeOtBu ; 0.54 ppm d, Js7.2 Hz, H-a .
( )
1
X
Ž
.
.
of 1f was obtained hs57% .
(
2.1.1.9. 2,2 -Methylene-bis 6-tert-butyl-4-methylphe-
1
Ž
Ž
.
H NMR C6 D6 : ds7.48 ppm s, Har ; 1.63
)
(
) ( )
noxy ethylaluminium diethylether 1j . This com-
t
t
Ž
.
Ž
.
ppm s, H- Bu-ortho ; 1.39 ppm s, H- Bu-para ;
pound was prepared as described for the synthesis of
3
3
Ž
.
Ž
0.81 ppm t, Js7.9 Hz, H-b ; 0.33 ppm q, Js7.9
Ž
.
1h hs85% .
1
X
X
4
.
Hz, H-a .
Ž
.
Ž
H NMR C6 D6 : ds7.16 ppm AA BB , Js
2
.
Ž
2.2 Hz y0 ds26.2 Hz, H-arom ; 4.10 ppm d, Js
X
2
X
.
Ž
.
13.6 Hz Hp ; 3.50 ppm d, Js13.6 Hz Hp ; 3.45
(
2.1.1.7. 2,2 -Methylene-bis 6-tert-butyl-4-methylphe-
) (
3
Ž
.
Ž
ppm q, Js7.0 Hz CH2–Et2O ; 2.24 ppm s, H-
)
(
)
noxy isobutylaluminium diethylether 1h . To a so-
t
3
.
Ž
.
Ž
Me ; 1.56 ppm s, H- Bu ; 1.40 ppm t, Js8.0 Hz
Ž
.
lution of triisobutylaluminium 5 ml, 19.8 mmol in
3
20 ml diethylether was added a solution of 2,2X-
.
Ž
.
H-b ; 0.73 ppm t, Js7.0 Hz, CH3–Et2O ; 0.40
3
Ž
.
ppm d, Js8.0 Hz, H-a .
Ž
. Ž
methylene-bis 6-tert-butyl-4-methylphenol 6.74 g,
.
19.8 mmol in 75 ml diethylether with a canula. The
addition was done in 5 min at room temperature. The
medium was stirred for 2 h, a precipitate was then
formed. Ether was removed to give a yellow solid
that was recrystallised in diethylether. An amount of
(
)
(
)
2.1.2. Bis aryloxy choloroaluminium: ArO 2 AlCl,
Et2 O
(
)
2.1.2.1. Bis 2,6-di-tert-butyl-4-methylphenoxy -
) (
Ž
.
6.15 g of white crystals was obtained hs63% .
1
X
X
4
(
)
chloroaluminium diethylether 2a . An amount of
Ž
.
Ž
H NMR C6 D6 : ds7.19 ppm AA BB , Js2
2
Ž
.
1.81 g 9.56 mmol of anhydrous stannous chloride
.
Ž
Hz y0 ds28.0 Hz, H-arom ; 4.12 ppm d, Js14
X
2
Ž
.
was added to a solution of 1a 5.0 g, 9.56 mmol in
50 ml diethylether. The medium was stirred for 5
days at room temperature.
Ether was removed under vacuum and 50 ml
toluene was added to the residue. The resulting
yellow suspension was filtered on celite. After the
removal of the solvent, the residue is crystallised in
.
Ž
.
Ž
Hz Hp ; 3.49 ppm d, Js14 Hz Hp ; 3.49 ppm q,
3
.
Ž
.
.
Js7.0 Hz CH2–Et2O ; 2.24 ppm s, H-Me ; 2.22
3
Ž
.
Ž
ppm t-hpt, Js6.5–7.0 Hz, H-b ; 1.57 ppm s,
t
3
Ž
.
Ž
H- Bu ; 1.24 ppm d, Js6.5 Hz, H-g ; 0.74 ppm
3
3
Ž
.
t, Js7.0 Hz, CH3–Et2O ; 0.50 ppm d, Js7.0
.
Hz, H-a .
Elemental analysis: exp.: C 73.82%, H 9.92%, Al
5.30%; th.: C 74.96%, H 9.94%, Al 5.43% for 1h.
Ž
.
ether to afford 3.75 g of a white solid hs68% .
1H NMR C6 D6 : ds7.16 ppm s, H-arom ;
Ž
.
Ž
.
3
Ž
.
Ž
3.91 ppm q, Js7.0 Hz CH2–Et2O ; 2.26 ppm s,
X
(
2.1.1.8. 2,2 -Methylene-bis 6-tert-butyl-4-methylphe-
t
3
.
Ž
.
Ž
H-Me ; 1.52 ppm s, H- Bu ; 0.62 ppm t, Js7.0
)
(
) ( )
noxy isobutylaluminium methyl -tert-butylether 1i .
.
Hz CH3–Et2O .
X
Ž
To a solution of 2,2 -methylene-bis 6-tert-butyl-4-
methyl-phenol , 4.05 g 11.88 mmol , in 30 ml
methyl-tert-butylether MTBE cooled at y308C was
added a solution of triisobutylaluminium 3 ml, 11.88
.
Ž
.
Ž
.
( )
2.1.2.2. Bis 2,6-di-tert-butylphenoxy chloroalumi-
nium diethylether 2b . This compound was pre-
pared as described for the synthesis of 2a with 1b
Ž
(
) ( )
.
mmol in 20 ml methyl-tert-butylether with a canula.
Ž .
starting material. hs62% .
The addition was done in 25 min. The medium was

(
)
128
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
1
3
Ž
.
Ž
H NMR C6 D6 : ds7.30 ppm d, Js7.7 Hz
solvent was removed and the resulting white solid
was washed with pentane to afford 38.0 g of an
X
3
.
Ž
.
Ž
Har ; 6.86 ppm dd, Js7.7 Hz Har ; 3.87 ppm q,
3
t
.
Ž
.
Ž
.
Js7.0 Hz CH2–Et2O ; 1.49 ppm s, H- Bu ; 0.59
orange powder hs97% .
3
1
X
X 4
Ž
.
Ž
.
Ž
ppm t, Js7.0 Hz CH3–Et2O .
H NMR C6 D6 : ds7.14 ppm AA BB , Js2.3
3
.
Ž
Hz y0 ds47.6 Hz, Har ; 4.19 ppm q, Js6.7 Hz
.
Ž
.
Ž
CH2–Et2O ; 2.03 ppm s, H-Me ; 1.41 ppm s,
(
)
2.1.2.3. Bis 4-bromo-2,6-di-tert-butylphenoxy -
chloroaluminium diethylether 2c . This compound
was prepared as described for the synthesis of 2a
t
3
.
Ž
.
H- Bu ; 1.07 ppm t, Js6.7 Hz, CH3–Et2O .
Elemental analysis: exp.: C 61.73%, H 7.84%, Cl
6.87%, S 6.91%, Al 5.86%; th.: C 63.33%, H 7.77%,
Cl 5.47%, S 6.50%, Al 5.47% for 2f.
(
) ( )
Ž
.
with 1c or 1e starting material hs40% .
1
Ž
.
Ž
.
H NMR C6 D6 : ds7.52 ppm s, Har ; 3.74
3
Ž
.
Ž
ppm q, Js7.0 Hz CH2–Et2O ; 1.30 ppm s, H-
t
3
.
Ž
.
Bu ; 0.53 ppm t, Js7.0 Hz CH3–Et2O .
2.2. Synthesis of aryloxyaluminium perrhenate com-
plexes
(
)
2.1.2.4. Bis- 2,4,6-tri-tert-butylphenoxy chloroalu-
(
) ( )
minium diethylether 2d . This compound was pre-
2.2.1. AgReO4 path
pared as described for the synthesis of 2a with 1f or
Ž
.
1g starting material hs53% .
(
)
2.2.1.1. Bis 2,6-di-tert-butyl-4-methylphenoxy alu-
1
Ž
.
Ž
.
H NMR C6 D6 : ds7.47 ppm s, Har ; 3.87
(
)
minium perrhenate, quinuclidine A-1 . Silver per-
rhenate 0.62 g, 1.7 mmol was added to a solution
3
Ž
.
Ž
ppm q, Js7.0 Hz CH2–Et2O ; 1.55 ppm s, H-
Ž
.
t
t
.
Ž
.
Ž
Bu-ortho ; 1.37 ppm s, H- Bu-para ; 0.57 ppm t,
Ž
.
of 2a 1.00 g, 1.7 mmol in 60 ml diethylether under
an argon stream at room temperature. The medium
was stirred in the dark for 24 h. Then it was filtered
and ether was removed. The residue was crystallised
with cold diethylether to afford 0.70 g of an orange
powder. 1H NMR analysis showed that 86% of a
3
.
Js7.0 Hz CH3–Et2O .
X
(
2.1.2.5. 2,2 -Methylene-bis 6-tert-butyl-4-methylphe-
noxy chloroaluminium diethylether 2e . To a solu-
tion of 2,2 -methylene-bis 6-tert-butyl-4-methyl-
phenol 27.15 g, 79.7 mmol in 100 ml diethylether
was added a solution of chlorodiethylaluminium 10
ml, 7.97 mmol in 15 ml diethylether. The addition
was done in 5 min. The medium was stirred for 1.5
h. The solvent was removed and the resulting white
grey solid was washed with pentane to afford 37.1 g
of an orange powder.
)
(
) ( )
X
Ž
. Ž
.
Ž
new product and 14% of phenol are obtained hs
Ž
.
54% .
1H NMR C6 D6 : ds7.13 ppm s, H-arom ;
2.22 ppm s, H-Me ; 1.46 ppm s, H- Bu .
Stabilisation with quinuclidine: to a solution of
the above product 0.5 g, 0.8 mmol in 10 ml
.
Ž
.
Ž
.
t
Ž
.
Ž
.
Ž
.
Ž
pentane was added a solution of quinuclidine 0.15
1
X
X
4
.
Ž
.
Ž
g, 1.3 mmol in 10 ml pentane at room temperature.
H NMR C6 D6 : ds7.14 ppm AA BB , Js
2
.
Ž
2.1 Hz y0 ds19.1 Hz, Har ; 4.02 ppm d, Js14.0
A precipitate was immediately formed. After filtra-
tion and drying, the product was washed with pen-
X
2
.
Ž
.
Hz Hp ; 3.58 ppm d, Js14.0 Hz Hp ; 3.49 ppm
3
Ž
.
Ž
.
Ž
.
q, Js6.8 Hz CH2–Et2O ; 2.21 ppm s, H-Me ;
tane to afford 0.5 g of a grey powder hs86% .
Elemental analysis after stabilisation with quinu-
clidine: exp.: C 50.08%, H 6.57%, N 1.65%, Al
2.79%, Re 21.3%; th.: C 53.73%, H 7.19%, N
1.69%, Al 3.26%, Re 22.5% for A-1.
1.59 ppm s, H- Bu ; 0.83 ppm t, 3Js6.8 Hz,
t
Ž
.
Ž
.
CH3–Et2O .
X
(
2.1.2.6. 2,2 -Thia-bis 6-tert-butyl-4-methylphe-
)
(
) ( )
noxy chloroaluminium diethylether
2f . To
a
X
Ž
(
)
solution of 2,2 -thia-bis 6-tert-butyl-4-methyl-
phenol
2.2.1.2. Bis di-tert-butyl-2,6-phenoxy aluminium per-
.
Ž
.
(
)
28.58 g, 79.7 mmol in 100 ml
rhenate, quinuclidine A-2 . This compound was pre-
pared as described for the synthesis of A-1 with 2b
starting material. NMR analysis showed that 97% of
diethylether was added a solution of chlorodi-
Ž
.
ethylaluminium 10 ml, 7.97 mmol in 20 ml
diethylether. The addition was done in 5 min. The
medium was stirred under argon for 1.5 h. The
Ž
a new product and 3% of phenol are obtained hs
.
59% .

(
)
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
129
1
3
Ž
.
Ž
H NMR C6 D6 : ds7.25 ppm d, Js7.7 Hz
For ratio AlrRes1 and 2, the initial proportion of
X
3
.
Ž
.
Ž
Ž
.
Har ; 6.85 ppm dd, Js7.7 Hz Har ; 1.42 ppm s,
Al OAr R was increased.
2
t
.
H- Bu .
Stabilisation with quinuclidine:the above new
product was stabilised with quinuclidine in a manner
2.2.2.1. 1aqRe2 O7
Ž
THF solÕent. A solution of Re2O7 1.00 g, 2.06
Ž
.
analogous to that described for A-1 hs89% to
give A-2.
Elemental analysis after stabilisation with quinu-
clidine: exp.: C 50.34%, H 7.05%, N 1.82%, Al
3.31%, Re 21.5%; th.: C 52.61%, H 6.94%, N
1.75%, Al 3.38%, Re 23.3% for A-2.
.
mmol in 20 ml THF was cooled at y788C. A
Ž
.
solution of 1a 1.08 g, 2.06 mmol in 20 ml THF
was added through a canula. The medium was al-
lowed to warm at room temperature. At y358C, a
black precipitate appeared. The medium was stirred
for 1 h at room temperature then THF was removed
under vacuum and the residue was dried for 2 h. The
residue was extracted with pentane to afford 1.41 g
of a brown-red oil that was used in metathesis.1H
NMR mainly showed phenol.
(
)
2.2.1.3. Bis 4-bromo-2,6-di-tert-butyl-phenoxy -
(
) (
)
aluminium perrhenate diethylether A-3 . This
compound was prepared as described for the synthe-
sis of A-1 with 2c starting material. NMR analysis
indicated that we obtained 96% of a new product
that seemed to be a dimeric species of A-3 according
to integration and 4% of phenol. We also noticed the
presence of a coordinated ether and two free ether in
solution in C6 D6.
Elemental analysis after stabilisation with bipyri-
Ž
.
dine B-1 : exp.: C 47.46%, H 5.66%, N 3.17%, Al
4.33%, Re 26.4%; th.: C 47.53%, H 5.36%, N
Ž
.
3.82%, Al 3.68%, Re 25.4% for B-1 see Table 2 .
Ž
Dioxane solÕent. To a solution of Re2O7 1.00 g,
2.06 mmol in 25 ml dioxane was added a solution
.
1H NMR C6 D6 : ds7.49 ppm s, H-arom ;
Ž
.
Ž
.
Ž
.
of 1a 1.08 g, 2.06 mmol in 25 ml dioxane through
a canula. A black precipitate immediately formed.
The addition was done slowly for 35 min and the
medium was stirred at room temperature for 2 h.
Dioxane was removed under vacuum and the result-
ing residue was extracted with pentane to afford 1.38
7.46 ppm s, H-arom ; 3.64 ppm q, 3Js6.9 Hz
Ž
.
Ž
3
.
Ž
CH2-coord. Et2O ; 3.25 ppm q, Js7.1 Hz CH2-
t
3
.
Ž
.
Ž
free Et2O ; 1.28 ppm s, H- Bu ; 1.07 ppm t, Js
3
.
Ž
7.1 Hz CH3-free Et2O ; 0.52 ppm t, Js6.9 Hz
.
CH2-coord. Et2O .
1
g of a brown-red oil that was used in metathesis. H
X
NMR showed mainly the presence of phenol.
(
2.2.1.4. 2,2 -Methylene-bis 6-tert-butyl-4-methylphe-
noxy aluminium perrhenate diethylether A-4 . This
compound was prepared as described for the synthe-
sis of A-1 with 2e starting material. H NMR indi-
Stabilisation with quinuclidine: after the extrac-
)
(
) (
)
Ž
.
tion with pentane, 0.23 g 2.1 mmol of quinuclidine
was added. A precipitate was then formed. After
filtration, we obtained 0.40 g of a ochre powder that
was washed with pentane.
1
cated that we obtained 98% of A-4 and 2% of 2e
Ž
.
hs56% .
1
X
X
4
Elemental analysis after stabilisation with quinu-
Ž
.
Ž
H NMR C6 D6 : ds7.09 ppm AA BB , Js2
2
Ž
.
clidine B-3 : exp.: C 53.42%, H 7.53%, N 2.44%,
.
Ž
Hz y0 ds18.8 Hz, H-arom ; 3.85 ppm d, Js14
X
2
Al 3.51%, Re 21.3%; th.: C 55.08%, H 7.56%, N
.
Ž
.
Ž
Hz Hp ; 3.57 ppm d, Js14 Hz Hp ; 3.43 ppm q,
3
Ž
.
2.27%, Al 2.91%, Re 20.1% for B-3 see Table 2 .
.
Ž
.
Js7.0 Hz CH2–Et2O ; 2.18 ppm s, H-Me ; 1.49
ppm s, H- Bu ; 0.77 ppm t, 3Js7.0 Hz, CH3–
t
Ž
.
.
Ž
Et2O .
2.2.2.2. 1bqRe2 O7
THF solÕent. This compound was prepared as
described for the synthesis of B-1 with 1b starting
material.
Elemental analysis: exp.: C 45.64%, H 5.70%, Al
4.02%, Re 26.8%; th.: C 47.01%, H 5.84%, Al
3.91%, Re 27.0% for A-4.
Elemental analysis after stabilisation with quinu-
Ž
.
2.2.2. Re2 O7 path
clidine B-2 : exp.: C 44.87%, H 6.68%, N 2.08%,
Unless otherwise stated, the following syntheses
are described for an initial atomic ratio AlrRes0.5.
Al 3.31%, Re 28.2%; th.: C 44.56%, H 6.28%, N
Ž
.
2.08%, Al 4.00%, Re 27.6% for B-2 see Table 2 .

(
)
130
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
1
Dioxane solÕent. This compound was prepared as
described for the synthesis of B-3. The crude product
was used directly in metathesis.
material in THF. H NMR analysis of the two last
compounds showed mainly phenol but also a new
t
product with a signal at 1.49 ppm as for the Bu
Ž
.
signal of the A-4 complex cf. AgReO4 path .
2.2.2.3. 1cqRe2 O7
THF solÕent. This compound was prepared as
described for the synthesis of B-1 with 1c starting
material. The crude product was used directly in
metathesis.
2.2.3. Oligomerisation of AgReO4 path complexes
(
)
type A complexes
To a solution of A-1 2.71g, 3.8 mmol in 35 ml
Ž
.
Ž
dioxane was added a solution of 1a 2.00 g, 3.8
mmol in 35 ml dioxane. The medium was stirred for
.
2.2.2.4. 1dqRe2 O7
2 h. The dioxane was removed to afford 5.71 g of a
brown-red powder looking like products obtained
with the Re2O7 path. The product was used as such
in metathesis.
THF solÕent. This compound was prepared as
described for the synthesis of B-1 with 1d starting
material. The crude product was used directly in
metathesis.
Elemental analysis after stabilisation by quinucli-
Dioxane solÕent. This compound was prepared as
described for the synthesis of B-3 with 1d starting
material. The crude product was used directly in
metathesis.
Ž
.
dine complex AB : exp.: C 54.34%, H 7.46%, N
1.73%, Al 4.49%, Re 19.0%; th.: C 55.27%, H
Ž
7.40%, N 1.45%, Al 4.19%, Re 19.3% for AB see
.
Section 3.4 .
2.2.2.5. 1fqRe2 O7
2.3. Olefin metathesis
THF solÕent. This compound was prepared as
described for the synthesis of B-1 with 1f starting
material, but with a molar ratio AlrRes1.
Elemental analysis after stabilisation with quinu-
2.3.1. Non-functional olefins
In a double wall glass reactor, 0.10 g of the
aluminium perrhenate complex was introduced the
initial AlrRe ratio used in the synthesis of the
complex had no influence over its activity in
Ž
.
clidine B-4 : exp.: C 52.80%, H 6.66%, N 2.53%,
Ž
Al 4.01%, Re 22.0%; th.: C 53.64%, H 6.60%, N
Ž
.
2.91%, Al 4.21%, Re 19.4% for B-4 see Table 2 .
Dioxane solÕent. This compound was prepared as
described for the synthesis of B-3 with 1f starting
material, but with a molar ratio AlrRes1 and 2.
Elemental analysis after stabilisation with bipyri-
dine: ratio AlrRes1 B-5 : exp.: C 54.79%, H
6.76%, N 2.80%, Al 4.15%, Re 20.1%; th.: C
52.21%, H 6.25%, N 3.04%, Al 4.40%, Re 20.2%
.
metathesis . The complex was dissolved in 40 ml
toluene. The medium was stirred for half an hour at
258C under argon. Distilled olefin, 1.0 ml, was added.
The evolution of the reaction was followed by taking
samples with time. These samples were hydrolysed
with an aqueous solution of 10% nitric acid. The
organic phase was analysed by gas chromatography.
Ž
.
Ž
.
Ž
for B-5 see Table 2 .
.
Ratio AlrRes2 B-6 : exp.: C 56.43%, H 7.13%,
N 2.48%, Al 4.52%, Re 16.2%; th.: C 54.83%, H
6.81%, N 2.56%, Al 4.93%, Re 17.0% for B-6 see
2.3.2. Functional olefins
Ž
An amount of 0.10 g of the aluminium perrhenate
complex was introduced in a Schlenk then dissolved
in 5 ml toluene. The medium was stirred under argon
for half an hour at 258C. Distilled olefin, 1.0 ml, was
added. The evolution of the reaction was followed by
taking samples with time. These samples were hy-
drolysed with an aqueous solution of 10% nitric acid.
The organic phase was analysed by gas chromatogra-
.
Table 2 .
2.2.2.6. 1hqRe2 O7 . This compound was prepared
as described for the synthesis of B-3 with 1h starting
material in dioxane.
2.2.2.7. 1jqRe2 O7 . This compound was prepared as
described for the synthesis of B-1 with 1j starting
Ž
.
phy toluene was used as an internal standard .

(
)
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
131
3. Results and discussion
3.2. Aryloxy complexes of aluminium
(
)
3.2.1. Alkyl-bis aryloxy aluminium complexes
3.1. Strategy for the synthesis of complexes including
aluminium perrhenate
(
)
AoO 2 Al–R
We have chosen the method described by Schreve
w
x
et al. 22 which is the fastest:
2ArOHqAIR3 heptane Al OAr Rq2RH g
Two reaction pathways were considered to achieve
the synthesis of complexes comprising an aluminium
Ž
.
Ž .
2
D
w x
perrhenate bonding. Herrmann 7 and Romao et al.
˜
OAr
R
w x
8
have described the reaction between rhenium
2,6-Di-tert-butyl-4-methyl
phenoxy
2,6-Di-tert-butyphenoxy
4-Bromo-2,6-di-tert-butyl
phenoxy
2,6-Di-tert-butyl-4-methoxy
phenoxy
4-Bromo-2,6-di-tert-butyl
phenoxy
isoButyl
1a
oxide and tetramethylin to synthesize methyltriox-
orhenium. We have transposed this reaction to an
isoButyl
isoButyl
1b
1c
Ž
.
alkyl-bis aryloxy -
aluminium compound according to the following
reactions, where the alkyltrioxorhenium formed
would appear here as a by-product:
isoButyl
ethyl
1d
1e
1f
2,4,6-Tri-tert-butylphenoxy
ethyl
After refluxing in heptane or pentane and recrys-
tallisation, we obtained the expected products with a
Ž
purity of 95–98% the other product being the start-
.
ing phenol with yields between 57% and 98%. As
far as we know, 1c, 1d and 1f complexes have not
Ž
been described in the literature until now the methyl
w
x.
analog of 1c was described but not isolated 23 . All
of the complexes were characterized with H NMR.
The metathetical exchange between silver per-
rhenate and chlorotrimethylsilane was described by
1
Complex 1c was also characterized by elemental
w
x
Schmidt and Schmidbaur 20,21 . By analogy, we
have performed the following sequence of reactions:
analysis.
X
Ž
The 2-2 -methylene-bis 6-tert-butyl-4-me-
.
thyl-phenol exhibited
a specific behaviour.
When this work started, the corresponding com-
plexes 1h and 1j were not described in the literature.
w
x
w x
Only titanium 24–27 and silicon 28 complexes of
These two reaction paths first require the synthe-
X
Ž
.
sis of bis aryloxy aluminium-alkyl or -chloride pre-
cursors.
Ž
.
Scheme 1. 2-2 -Methylene-bis 6-tert-butyl-4-methyl-phenoxy
complexes of titanium.

(
)
132
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
X
Ž
.
2-2 -methylene-bis 6-tert-butyl-4-methyl-phenol
to the expected product. We obtained a white powder
soluble only in hot toluene, maybe an oligomer
arising from an intermolecular reaction. Thus, we
modified the Ittel method using diethylether as a
solvent:
were described.
In our hands, the reaction of 2-2X-methylene-
Ž
.
bis 6-tert-butyl-4-methyl-phenol with a trialkylalu-
minium complex using the Ittel method did not lead
(
)
(
)
We were then able to isolate the complexes 1h
and 1j 99% pure after crystallisation in ether in good
3.2.2. Bis aryloxy chloroaluminium ArO 2 Al–Cl
complexes
Ž
.
yields 63% and 85% . They were characterized by
NMR and elemental analysis. During this work, a
Taiwan team published a study of those compounds
Using the method of Healy, involving the reaction
Ž
.
of methyl- and ethyl-bis aryloxy -aluminium com-
Ž .
plexes and either tin II chloride or trimethyltin
chloride 31 , with the bulkier isobutyl-bis
w
x
w x
that confirmed our results 29,30 .
Studies of titanium complexes showed that
the protons of the CH2 bridge present an AB
Ž
.
aryloxy aluminium, trimethyltin chloride was inac-
Ž .
tive, while tin II chloride allowed to isolate the
Ž
.
Ž
signal ds3.14 and 3.52 ppm . NOE allowed to
expected products but with a slower kinetics 5–8
.
assign the resonance in high fields to Hax pointing
days of reaction .
Ž
.
toward the Ti centre Scheme 1 . The signal in
Concerning monophenols, we could isolate the
low fields was assigned to Heq. We noticed that
products after crystallisation in ether in good yields
Ž
Ž
.
1h and 1j also present an AB signal d;3.5
53–68% except the para-bromo-substituted ary-
X
.
Ž .
loxy complex 2c hs40% . Monophenols did not
and 4.1 ppm, Hp, Hp . We propose that the
Ž
.
same attributions as for titanium complexes may be
react with Sn nBu Cl or with AlEt2Cl even under
3
done.
reflux:
Et2O
Al OAr 2 RqSnCl
Al OAr 2 Cl,Et OqRSnCl
Ž .
Ž
.
2
2
3days
OAr
R
2,6-Di-tert-butyl-4-methyl phenoxy
2,6-Di-tert-butylphenoxy
4-Bromo-2,6-di-tert-butyl phenoxy
4-Bromo-2,6-di-tert-butyl phenoxy
2,4,6-Tri-tert-butylphenoxy
isoButyl
2a
2b
2c
2c
2d
isoButyl
isoButyl
ethyl
ethyl

(
)
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
133
As concerns the complexes 1h and 1j de-
best way was starting from the phenol itself
and diethylaluminium chloride. The complex 2e and
the analogous S-bridged complex 2f were obtained
in good yields.
X
Ž
rived from 2-2 -methylene-bis 6-tert-butyl-4-methyl-
.
Ž
phenol , SnCl2 reacted very slowly 80% yield
.
after 15 days , SnMe3Cl reacted faster, but the
3.3. Aryloxy aluminium perrhenate complexes
ysis matched correctly except for the carbon analy-
sis. A first explanation is that, despite the addition of
a stabilising ligand, the complex was still unstable.
Such a phenomenon has already been observed by
Herrmann: quinuclidine adducts of several allyltriox-
orhenium complexes are less stable than the uncoor-
(
3.3.1. The AgReO4 path conducting to type A com-
)
plexes
3.3.1.1. Synthesis and characterisation of the com-
Ž
.
w x
plexes. The bis aryloxy chloroaluminium complexes
2a and 2b reacted with AgReO4 in the dark, in
diethylether or methylene dichloride solvent. After
filtration and removal of the solvent, a powder was
isolated which was unstable even under an inert
atmosphere. 1H NMR of these crude products showed
that they were composed of various amounts of
phenol and another compound.
dinated allyl starting complexes 32 . Another expla-
nation is the presence of trace amounts of AgCl even
in the stabilised complexes that we were not able to
crystallise pure without decomposition.
When the complex 2c including a para-bromo-
substituted aryloxy ligand was reacted with AgReO4
in diethylether in the dark, the complex was identi-
fied as a dimeric species A-3 by analogy with the
complexes described above and according to NMR.
In the same reaction, the complex 2d including a
para-tert-butyl-substituted aryloxy ligand only led to
a decomposition into phenol.
We stabilised this compound by addition of a
Lewis base, like quinuclidine or bipyridine. 1H NMR
of the stabilised complexes is in agreement with the
expected formulae A-1 and A-2. The elemental anal-
Et2O
Al OAr 2 Cl,Et OqAgReO
ArO 2 Al–O–ReO unstable, immediatelyused in metathesis
3 Ž .
Ž
.
Ž
.
2
4
quinuclidine
ArO 2 Al–O–ReO
ArO 2 Al–O–ReO quin
3 Ž .
Ž
.
Ž
.
3
OAr
2,6-Di-tert-butyl-4-methyl phenoxy
2,6-Di-tert-butylphenoxy
4-Bromo-2,6-di-tert-butyl phenoxy
A-1
A-2
A-3

(
)
134
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
The complex 2e derived from the 2-2X-methylene-
Table 1
Ž
.
Activity of complexes synthesized by the AgReO4 path type A
in 2-pentene metathesis
Ž
.
bis 6-tert-butyl-4-methyl-phenol reacted more
slowly. After 4 days, we could isolate the expected
complex A-4 as a white powder with a purity of
98%. Unlike monophenoxy-containing complexes, a
Lewis base was not necessary to stabilise the com-
plex, but an ether molecule remains coordinated,
most probably on aluminium. Actually, when a com-
plex containing an ether coordinated to rhenium is
dried, the ether is generally decoordinated under
vacuum. Thus, it appears that using a bidentate
Corresponding
stabilised
complex
Substituent in
para position
of the aryloxy
ligand
Activity k
Ž
y1
moly1 l2 hy1
g
cat
.
of the unstabilised
complex
A-1
A-2
A-3
A-4
Me
H
Br
0.4
0.04
0
0.007
Ž
.
bis aryloxy ligand on the aluminium increases the
stability of the corresponding perrhenate:
ing 50% at the equilibrium. Results are given in
Table 1.
These type A complexes have a poor catalytic
activity in 2-pentene metathesis. The most active
complex is A-1 with a para-methyl-substituted ary-
loxy ligand. When an electron-attracting group such
Ž
.
as bromine is in para position, the complex A-3
has no activity. The activity of the complex compris-
X
Ž
ing the 2-2 -methylene-bis 6-tert-butyl-4-methyl-
phenoxy ligand is very weak. However, we ignore
.
When the similar complex 2f, derived from the
how the presence of a coordinated ether influences
the catalytic activity.
We tried to improve the activity of these type A
complexes by adding a Lewis acid cocatalyst such as
X
Ž
.
2-2 -thia-bis 6-tert-butyl-4-methyl-phenol , was re-
acted with AgReO4 in diethylether or dioxane, a
1
purple solid is obtained. However, H NMR showed
that the crude product was constituted mainly by the
starting phenol and unidentified by-products. Stabili-
sation with quinuclidine did not allow to isolate any
rhenium complex.
Ž
.
Ž
.
AlEtCl2 , Al OAr R, Al OAr Cl,Et2O, GaBr3 and
2
2
AlCl3, in toluene solution. No effect on the catalytic
activity was observed. However, with AlCl3 in
chlorobenzene, the medium became black and activ-
ity was no more observed. Aluminium chloride is
well known to decompose aryloxyaluminium com-
3.3.1.2. Catalytic actiÕity in metathesis. The
Ž
.
metathesis of 2-pentene a cis–trans mixture to
2-butene and 3-hexene has been chosen as a test
reaction for the activity of the new complexes. These
complexes are used before their stabilisation with a
Lewis base and immediately after their synthesis,
owing to their unstability. The reaction was con-
ducted at 258C and started without the addition of
any cocatalyst. Detailed conditions are given in Sec-
tion 2.
w
x
plexes 31 . An alkylating cocatalyst such as tetram-
ethyltin did not improve the catalytic activity.
3.3.2. Oligomeric complexes synthesized by the
(
)
Re2 O7 path type B complexes
3.3.2.1. Synthesis and characterisation. The
Ž
.
reaction of an alkyl-bis aryloxy aluminium com-
plex with Re2O7 was performed in two diffe-
rent solvent: THF or dioxane. No reaction oc-
curred in toluene:
The activity of each complex is expressed by the
kinetic constant calculated according to a second-
w
x
order equilibrated reaction 18 , with conversion be-
Ž .
1 THF y358C . . . 208C
Al OAr 2 RqRe O
used in metathesis,
Ž
.
2
7
7
Ž .
2 extraction with pentane
Ž .
1 Dioxane 208C
Al OAr 2 RqRe O
used in metathesis.
Ž
.
2
Ž .
2 extraction with pentane

(
)
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
135
Using THF as a solvent, the aluminium complex
and Re2O7 are mixed at y788C. The medium is
allowed to warm slowly to room temperature. At
y358C, a black precipitate appears, due to the de-
composition of the alkyltrioxorhenium complex
RReO3 that is formed as a by-product. Indeed, isobu-
tane and isobutene were identified by gas chro-
matography and 1H NMR during the reaction be-
styrene resins at low temperature under argon but
this did not improve the purity of the product.
We stabilised these products by the addition of a
Ž
.
Lewis base bipyridine or quinuclidine like in the
case of type A complexes. After stabilisation, the
elemental analysis for C, H, N, Al, Re was per-
formed. The formulae that we could derive are more
complicated than those obtained in the AgReO4 path.
The complexes contained one or more Al–O–Al
linkages. The formulae are summarised in the Table
2, depending on the initial atomic ratio AlrRe used
in the synthesis and the nature of the solvent.
Ž
.
Ž
tween Al OAr iBu and Re2O7 ethane and ethylene
2
Ž
.
.
are found with Al OAr Et and Re2O7 . The black
colour was attributed to ReO3 usually red, but black
2
Ž
.
when finely divided .
When the reaction medium reached room temper-
ature, it was allowed to stir again for 1 h then THF
was removed. The time of stirring at room tempera-
ture mainly determines the final activity of the cata-
lyst. If the medium was stirred for too long a time
The product of the synthesis starting with com-
plexes 1h or 1j is different from those obtained with
Ž
.
the monodentate bis aryloxy aluminium complexes.
In THF, dioxane, Et2O, we obtained a purple solid.
1
Its H NMR spectrum showed the presence of phe-
Ž
.
more than 2 h or when the reaction was performed
nol and of an unknown product of which the chemi-
from the beginning at room temperature, THF poly-
cal shift of tert-butyls was close to the one of the
Ž
merised.
complex A-4 synthesized by the AgReO4 path ds
.
Once the solvent was removed, the residue was
extracted with pentane in order to separate the solu-
ble complex from the rhenium oxide precipitate.
After removal of the pentane, brown-red oils some-
times powders were obtained. If the oils were dis-
1.49 ppm . We were not able to isolate any complex.
However, it seemed that the presence of a bidentate
bis phenoxy ligand did not allow the synthesis of a
complex containing multiple Al–O–Al linkages.
Ž
.
Ž
.
solved in pentane, a brown-red precipitate appeared
after a few hours. Nonetheless, the solutions did not
seem to be colloidal because no sedimentation was
observed by centrifugation. On the other hand, the
solutions of the oils in toluene are much more stable.
They did not conduct to a precipitation and gave
reproducible results in catalysis.
In dioxane as a solvent, the reaction medium was
allowed to stir for 2 h at room temperature, then it
was treated as described above. Brown-red powders
were obtained. The black precipitate separated dur-
ing the pentane extraction was found to be
3.3.2.2. Catalytic actiÕity in the metathesis of 2-
pentene. The catalytic experiments were conducted
in the same conditions as for the type A complexes.
Again, complexes were used in metathesis before
their stabilisation with a base and immediately after
their synthesis. Table 3 summarises the results ac-
cording to the solvent of the synthesis. We noticed
Ž
that the atomic ratio AlrRe i.e. one, two or three
.
Al–O linkages had no influence on the final activity
of the complex.
Ž
. Ž
.
ReO3 C4 H8O2
confirming the decomposition of the alkyltrioxorhe-
satisfactory elemental analysis ,
The complexes of type B obtained through the
Re2O7 path are generally much more active in
metathesis catalysis than the type A complexes. As
Ž
.
nium into alkane, alkene and ReO3 solvent under
our conditions.
The characterisation of the products was difficult
because of the extreme unstability of these type B
complexes even under an inert atmosphere. 1H NMR
most often showed the presence of phenol mainly.
By analogy with the work of Kuhn et al. 33 on
alkylrhenium complexes, we tried purifications by
chromatography on grafted silica C18 or poly-
Ž
an example, 0.1 g catalyst 90 mmol with B-2 for-
.
mula without quinuclidine in 1 ml of toluene brought
about the metathesis of 10 ml 93 mmol of 2-pentene
Ž
.
y1
Ž
.
to equilibrium in 5 min TOF 100 min . Strong
differences are observed according to the nature of
the aryloxy groups. The catalyst synthesized by the
reaction of 1f and Re2O7 in dioxane gave the best
results.
w
x
¨
Ž
.

(
)
136
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
Table 2
Ž
.
Complexes synthesized by the Re2O7 path type B complexes

(
)
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
137
Table 4
3.3.2.3. Influence of the nature of the olefin
We compared the reactiÕity in metathesis of di-
fferent non-functional olefins with a type B-4
Ž
.
Activity of complexes synthesized by the Re2O7 path type B in
the metathesis of various olefins
Olefin
Corresponding
stabilised
complex
Activity k
Ž
catalyst before its stabilisation with
a base.
moly1 l2 hy1
of the unstabilised
g
cat
y1
.
Functional olefins were tested with a type B-5
Ž
.
catalyst before its stabilisation Table 4 .
complex
1-Hexene is much less reactive than 2-pentene.
Steric crowding around the double bond strongly
hinders metathesis. Indeed, 2-methyl-2-pentene and
2,4,4-trimethyl-2-pentene were not converted even
after 3 days of reaction. However, when 2-pentene
was added at that time, it was readily converted. The
lack of activity with bulky olefins is then due to an
impossibility of coordination rather than to an inhibi-
tion of the catalyst.
2-Pentene
1-Hexene
B4
B4
B4
B4
160
3
0
2-Methyl-2-pentene
2,4,4-Trimethyl-2-
pentene
Allyl acetate
Diallyl ether
Diallyl sulphide
0
B5
B5
B5
0
1400
1200
complexes are intermediates of type B complexes in
the Re2O7 path. With this aim, we performed the
following experiments.
AgReO4 path exhibit a much lower activity in
Table 3
Ž
Ø Complex A-1 was stirred with the ReO3 com-
Ž
.
Activity of complexes synthesized by the Re2O7 path type B in
.
ing from Re2O7 path black precipitate in dioxane
2-pentene metathesis
for 2 h and the medium was then extracted with
Corresponding
stabilised
complex
Substituent in
para position
of the aryloxy
ligand
Activity k
Ž
of the unstabilised
complex
1
pentane after removal of dioxane. H NMR showed
y1
moly1 l2 hy1
g
cat
.
that A-1 was not modified and there was no im-
provement in catalysis. Thus, ReO3 do not interact
with a type A complex.
Synthesis of the complex in THF
Ø Complex A-1 was allowed to stir with the
Non-isolated
Non-isolated
B2
Br
OMe
H
Me
tBu
2.4
1.9
75
125
35
Ž
.
aluminium complex 1a 1 Eq in dioxane for 2 h.
After removal of dioxane, the product looks like the
one observed at the end of the Re2O7 path. More-
over, its catalytic activity was considerably improved
B1
Non-isolated
y1
y1
l2
g
instead of 0.4 moly1 l2
Synthesis of the complex in dioxane
Ž
g
ks100 mol
cat
cat
y1
Non-isolated
Non-isolated
B3
OMe
H
Me
tBu
8
14
160
450
.
. When the same reaction was done in toluene,
there was no improvement in metathesis.
The complex resulting from the above reaction in
B6
Ž
dioxane was stabilised by quinuclidine complex

(
)
138
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
.
AB . Elemental analysis is close to the analysis of
the complex B-5 synthesized by Re2O7 path in
dioxane with an AlrRe ratio equal to 1. The Al
content in AB is found in the 4–5% by weight range
monomeric with a tricoordinated aluminium as is
w
x
known to be the methyl analog 34 , due to the bulky
ortho substituents. This steric bulk is likely to be
found also in the A and B type complexes. The
situation is more complicated because the syntheses
were done in an ether medium. The complexes iso-
lated from pentane extraction that are active in
metathesis comprise THF or dioxane molecules co-
Ž
.
2.79% in A-1 , a value which implies the presence
of Al–O–Al linkages. The following tentative for-
mula is proposed on the basis of the best fit for the
C, H, N, Al, Re analyses. The phenols could be
coordinated through the lone pairs of their oxygen to
the electron-deficient aluminium atoms. Then it
seems that there is a relation between the presence of
Al–O–Al linkages in the complex and a high activ-
ity in metathesis:
Ž
.
ordinated to aluminium and maybe to rhenium . On
the other hand, we have observed that an excess of
THF inhibits the catalysis. This suggests an equilib-
rium of coordination–decoordination of the ether.
Such equilibria have been described in the literature
w
x
35–37 .
The acidity of a bis aryloxy aluminium moiety
Ž
.
acting as a ligand for the perrhenate should then be
attributed to the tricoordination of the aluminium
center stabilised and protected by the bulky ortho
substituents of the aryloxy ligands. This conclusion
may explain the activity of these catalysts toward
functional olefins because the acidic function cannot
interact with the basic functional groups of the olefin
because of the steric bulk of aryloxy ligands.
3.5. Possible origin of the metathesis actiÕity
3.5.1. The respectiÕe roles of the aryloxy ligands and
aluminium
3.5.2. The difference of actiÕity between A complexes
and oligomeric B complexes
In our homogeneous catalysts, either of A or B
type, there is no correlation between the acidity of
The above conclusion cannot explain why A com-
plexes have a lower activity than B complexes. We
observed that the only way to activate A complexes
is to make them react with an alkyl-bis
Ž
the aryloxy ligand controlled by its para sub-
.
stituent and the activity in metathesis. The
Ž
.
isobutyl–aluminium complex 1a is most probably
aryloxy aluminium in an ether solvent.
Scheme 2. Inactive or low active complexes in metathesis catalysis.

(
)
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
139
Ž
.
2
tert-butyl-2,6-phenoxy ligand or ArO
is an
ortho-CH 2 –or –S-bridged di-tert-butyl-2,6-bis-
phenoxy ligand, and Y is an alkyl or chlorine ligand.
Ž
.
The reaction of bis aryloxy chloroaluminium
Ž
.
ArO Al–Cl,Et2O complexes with AgReO4 in ether
2
Ž
.
Ž
leads to new complexes ArO Al–OReO3 type A
2
.
complexes . These complexes were stabilised by co-
ordination with a Lewis base such as quinuclidine. In
the absence of this base, A complexes are poorly
active 2-pentene metathesis catalysts. Alkyl-
Scheme 3. Highly active complex in metathesis catalysis.
Ž
.
Ž
.
Ž
.
bis aryloxy aluminium ArO Al–R RsiBu, Et
2
complexes react with Re2O7 in THF or dioxane,
On the other hand, we tried to activate a perrhen-
leading to complexes that contain Al–O–Al
Ž
.
ate by the addition of a Lewis acid Scheme 2 .
w
Ž
.
oligomeric linkages such as in O3Re– Al OAr –
Tetrabutylammonium perrhenate, associated with a
x
Ž
.
O –ReO3 type B complexes . Some of these com-
2
Ž
.
Lewis acid AlCl3, AlEt2Cl, SnMe4 in chloroben-
zene, was inactive in metathesis. This result shows
that a simple enhancement of the acidity at the
plexes, unstable even under an argon atmosphere,
were identified after stabilisation with a Lewis base
such as quinuclidine or bipyridine. Before their sta-
Ž
rhenium assuming coordination of the Lewis acid to
bilisation, they show a fairly high activity in the
.
y1
an oxo ligand was not sufficient to obtain an active
catalyst. Even the trifluoroacetyl perrhenate contain-
Ž
.
.
metathesis of 2-pentene TOFs100 min
and
Ž
Ž
functional olefins diallylether, diallylsulphide . As
far as we know, they are the most active homoge-
neous rhenium-based catalysts for the metathesis of
internal olefins.
ing the trifluoroacetate electron-withdrawing group,
.
but unfortunately isolated with an acetonitrile ligand ,
in the presence of a few equivalents of Lewis acid,
was inactive in metathesis catalysis.
We have demonstrated that A type complexes
may be activated by reaction with alkylbis
As a conclusion, favourable conditions for a high
activity in metathesis seem to be met only when
Al–O–Al linkages are present in the rhenium envi-
ronment. Such linkages exhibit a strong Lewis acid-
ity related to the possibility of a double coordination
of two neighbouring aluminium atoms to a substrate,
Ž
.
aryloxy aluminium complexes in an ether solvent.
Al–O–Al linkages were then created, giving a B
complex-like structure. We propose that the possible
origin of the high catalytic activity is related to the
presence of these oligomeric Al–O–Al linkages. The
aluminium sites are protected by the steric hindrance
of the aryloxy ligands and they show a strong Lewis
acidity, eventually latent if the decoordination of a
solvent molecule should occur first. These results
exemplify again the determining role of the Lewis
acidity in such catalysts.
w
x
w x
as suggested by Hanawa et al. 38 , Ooi et al. 39
w
x
and Wuest 40 . In our case, according to Scheme 3,
an Al–O–Al linkage attached to the perrhenate may
induce a strong electrophilic activation of an oxo
ligand of the rhenium.
The competition of acidity between the alu-
minium and the trioxorhenium parts in the molecule
is then in favour of the aluminium, which at the end
extracts an oxo ligand out of the coordination sphere
of the rhenium, opening a vacant site where a car-
bene may be generated later from an incoming olefin.
Ž
.
The application of the bis aryloxy aluminium
complexes to the activation of the Re2O7rAl2O3
heterogeneous catalyst is the object of our current
investigations.
Acknowledgements
4. Conclusion
Ž
.
2
a di-
A
series of aluminium complexes ArO -
Pr. J.M. Basset, Drs. H. Olivier and L. Saussine
are gratefully acknowledged for helpful discussions.
Al–Y was synthesized, where ArO is

(
)
140
G. Doledec, D. CommereucrJournal of Molecular Catalysis A: Chemical 161 2000 125–140
w
w
w
x
Ž
.
.
20 M. Schmidt, H. Schmidbaur, Chem. Ber. 92 1959 2667.
x
References
Ž
21 M. Schmidt, H. Schmidbaur, Inorg. Synth. 4 1967 149.
x
22 A.P. Schreve, R. Mulhaupt, W. Fultz, J. Calabrese, W.
w x
1
Y. Chauvin, D. Commereuc, J. Chem. Soc., Chem. Commun.
Ž
.
Robbins, S.D. Ittel, Organometallics 7 1988 409.
Ž
.
1992 462.
w
w
x
x
Ž
.
23 S. Saito, K. Shimada, H. Yamamoto, Synlett 1996 720.
24 C. Floriani, F. Corazza, W. Lesueur, A. Chiesi-Villa, C.
w x
2
P. Amigues, Y. Chauvin, D. Commereuc, C.T. Hong, C.C.
Lai, Y.H. Liu, J. Mol. Catal. 65 1991 39.
Ž
.
.
Ž
.
Guastini, Angew. Chem., Int. Ed. Engl. 28 1989 66.
w x
Ž
3
w x
4
J.C. Mol, Catal. Today 51 1999 289.
F. Schekler-Nahama, O. Clause, D. Commereuc, J. Saussey,
Appl. Catal. 167 1998 237, 247.
w
w
w
x
25 F. Corazza, C. Floriani, A. Chiesi-Villa, C. Guastini, Inorg.
Ž
.
Chem. 30 1991 145.
Ž
.
x
26 J. Okuda, S. Fokken, H.C. Kang, W. Massa, Chem. Ber. 128
w x
5
A.M. Turek, I.E. Wachs, E. De Canio, J. Phys. Chem. 96
Ž
.
1995 221.
Ž
.
1992 5000.
x
27 S. Fokken, T.P. Spaniol, H.C. Kang, W. Massa, J. Okuda,
w x
6
X. Xiaoding, J.C. Mol, C. Boelhouwer, J. Chem. Soc.,
Faraday Trans. 1 82 1986 2707.
Ž
.
Organometallics 15 1996 5069.
Ž
.
w
w
x
x
Ž
.
28 S.D. Pastor, J.D. Spivack, J. Org. Chem. 49 1984 1297.
29 B.T. Ko, Y.L. Sun, C.H. Lin, C.C. Lin, 214th ACS National
w x
Ž
.
7
w x
8
W.A. Herrmann, J. Organomet. Chem. 500 1995 149.
C.C. Romao, F.E. Kuhn, W.A. Hermann, Chem. Rev. 97
˜ ¨
Meeting Las Vegas,1997, p. 229.
Ž
.
1997 3197.
w
w
w
w
w
w
w
w
w
w
w
x
30 C.H. Lin, L.F. Yan, F.C. Wang, Y.L. Sun, C.C. Lin, J.
w x
9
R.R. Schrock, R.T. Depue, J. Feldman, C.J. Schaverien, J.C.
Dewan, A.H. Liu, J. Am. Chem. Soc. 110 1988 1423.
Ž
.
Organomet. Chem. 587 1999 151.
Ž
.
x
31 M.D. Healy, J.W. Ziller, A.R. Barron, Organometallics 11
w
w
w
w
w
w
w
w
x
10 M.H. Schofield, R.R. Schrock, L.Y. Park, Organometallics
Ž
.
1992 3041.
Ž
.
10 1991 1844.
x
32 W.A. Herrmann, F.E. Kuhn, C.C. Romao, H.T. Huy, J.
¨
˜
x
11 I.A. Weinstock, R.R. Schrock, W.M. Davis, J. Am. Chem.
Ž
.
Organomet. Chem. 481 1994 227.
Ž
.
Soc. 113 1991 135.
x
Ž
.
33 F.E. Kuhn, J. Mink, W.A. Herrmann, Chem. Ber. 130 1997
¨
x
Ž
.
12 R. Toreki, R.R. Schrock, J. Am. Chem. Soc. 112 1990
295.
2448.
x
34 A.P. Shreve, R. Mulhaupt, W. Fultz, J. Calabrese, W. Rob-
x
13 R. Toreki, R.R. Schrock, W.M. Davis, J. Am. Chem. Soc.
Ž
.
bins, S.D. Ittel, Organometallics 7 1988 409.
Ž
.
114 1992 3367.
x
35 W.A. Herrmann, F.E. Kuhn, M.U. Rauch, J.D.G. Correia, G.
¨
x
14 R. Toreki, G.A. Vaughan, R.R. Schrock, W.M. Davis, J. Am.
Ž
.
Artus, Inorg. Chem. 34 1995 2914.
Ž
.
Chem. Soc. 115 1993 127.
x
36 J.T. Barry, M.H. Chisholm, J. Chem. Soc., Chem. Commun.
x
15 R. Toreki, R.R. Schrock, W.M. Davis, J. Organomet. Chem.
Ž
.
1995 1599.
Ž
.
520 1996 69.
x
37 R.L. Geerts, J.C. Huffman, K.G. Caulton, Inorg. Chem. 25
x
16 G.A. Vaughan, R. Toreki, R.R. Schrock, W.M. Davis, J. Am.
Ž
.
1986 1803.
Ž
.
Chem. Soc. 115 1993 2980.
x
38 H. Hanawa, N. Abe, K. Maruoka, Tetrahedron Lett. 40
x
Ž
.
17 A.M. LaPointe, R.R. Schrock, Organometallics 14 1995
Ž
.
1999 5365.
1875.
x
39 T. Ooi, T. Miura, K. Takaya, K. Maruaoka, Tetrahedron Lett.
w
w
x
18 G. Doledec, PhD Thesis, Paris 6 University, 1999.
Ž
.
40 1999 7695.
x
Ž
.
19 D. Commereuc, J. Chem. Soc., Chem. Commun. 1995 791.
x
Ž
.
40 J.D. Wuest, Acc. Chem. Res. 32 1999 81.