Comparison of olefin metathesis by rhenium-containing γ-alumina or silica-aluminas and by some mesoporous analogues

Article abstract of DOI:10.1016/j.jcat.2008.05.026

Oxide materials with rhenium(V) oxochloro complexes highly dispersed on the surface were prepared by grafting Re(V) alkoxochloro complexes onto γ-Al₂O₃, chlorinated alumina, 25 and 75 wt% SiO₂ silica-aluminas, mesoporous alumina with high specific surface area (685 m² g^-1) and mesoporous silica-alumina (950 m² g^-1). The activities of these catalysts for 1-octene metathesis at room temperature are compared in the presence of a promoter: Bu₄Sn or Ph₂SiH₂. Systems consisting of pure alumina and chlorinated alumina supports with rhenium species and Bu₄Sn lead to selective formation of 7-tetradecene and ethylene at 90-100% conversion. There is a dramatic change of selectivities with Bu₄Sn and silica-alumina or with Ph₂SiH₂ and aluminas. The difference is attributed to the formation of hydrido and carbene rhenium surface species which promote at the same time isomerization, self- and cross-metathesis. Mesostructured oxides do not show any clear advantage over commercial supports. Interestingly, modifying γ-alumina with hydrogen chloride increases the catalytic activity even at low rhenium loading (ca. 0.7 wt%).

Full text of DOI:10.1016/j.jcat.2008.05.026

Journal of Catalysis 258 (2008) 61–70
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Comparison of olefin metathesis by rhenium-containing γ -alumina or
silica–aluminas and by some mesoporous analogues
Paul Celestin Bakala a, Emmanuel Briot a,∗, Yannick Millot a, Jean-Yves Piquemal b, Jean-Marie Brégeault a
a
Systèmes Interfaciaux à l’Echelle Nanométrique, Université Pierre et Marie Curie-Paris 6/CNRS, UMR 7142, Case 196, 4 place Jussieu, F-75252 Paris Cedex 05, France
ITODYS, Université Denis Diderot-Paris7/CNRS, Case 7090, 2 place Jussieu, F-75251 Paris Cedex 05, France
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Oxide materials with rhenium(V) oxochloro complexes highly dispersed on the surface were prepared by
Received 21 February 2008
Revised 21 May 2008
Accepted 27 May 2008
Available online 3 July 2008
grafting Re(V) alkoxochloro complexes onto γ -Al2O3, chlorinated alumina, 25 and 75 wt% SiO2 silica–
aluminas, mesoporous alumina with high specific surface area (685 m2 −1) and mesoporous silica–
g
alumina (950 m2 g−1). The activities of these catalysts for 1-octene metathesis at room temperature
are compared in the presence of a promoter: Bu4Sn or Ph2SiH2. Systems consisting of pure alumina
and chlorinated alumina supports with rhenium species and Bu4Sn lead to selective formation of 7-
tetradecene and ethylene at 90–100% conversion. There is a dramatic change of selectivities with Bu4Sn
and silica–alumina or with Ph2SiH2 and aluminas. The difference is attributed to the formation of
hydrido and carbene rhenium surface species which promote at the same time isomerization, self- and
cross-metathesis. Mesostructured oxides do not show any clear advantage over commercial supports.
Interestingly, modifying γ -alumina with hydrogen chloride increases the catalytic activity even at low
rhenium loading (ca. 0.7 wt%).
Keywords:
Silica
Alumina
1-Octene metathesis
Mesoporous materials
MAS NMR
Chlorinated alumina
Rhenium
© 2008 Elsevier Inc. All rights reserved.
Grafting
Diphenylsilane
1. Introduction
lated to the presence of Al–O–Al linkages which should induce a
strong bidentate Lewis acidity [9–11]. The best heterogeneous cat-
alysts involve supported Mo(VI), W(VI) or Re(VII) on aluminas or
Since the pioneering work of many groups [1–4], alkene
metathesis is now widely used in large-scale industrial processes
for the synthesis of linear olefins, polymeric materials and phar-
maceutical intermediates. It is now extensively used in various
kinds of reactions, such as ring-opening metathesis polymeriza-
tion (ROMP), ring-closing metathesis (RCM), acyclic diene metathe-
sis polymerization (ADMET), and cross-metathesis (CM or XMET)
[1–4]. Many homogeneous catalysts or precursors, containing tran-
sition metal complexes, mainly with Mo, W, Re and Ru [1], have
served in olefin disproportionation. Among rhenium complexes,
methyltrioxorhenium(VII), [CH3ReO3], denoted MTO [5,6], requires
activation by a co-catalyst, a Lewis acid and an alkylating promoter
(AlRnX3−n for homogeneous systems), unless it is grafted onto
an acidic support. Alkyl-alkylidene-alkylidyne rhenium complexes
such as [Re(C-t-Bu)(CH-t-Bu)(CH2-t-Bu)(CH3CN)(OTf)] metathesize
cis-2-pentene, without the need for a co-catalyst, but they are
not long-lived [7,8]. Interestingly, also without the help of any
co-catalyst, [{O3ReO–Al(OAr)}2O] species and parent analogues, ho-
mogeneous models of rhenium oxide-on-alumina catalysts, show
◦
silicas: MoO3/Al2O3 or MoO3/SiO2 from 100 to 200 C, WO3/Al2O3
◦
◦
at 150 C; WO3/SiO2 at 400 C and Re2O7/Al2O3 which exhibits
◦
high activity and selectivity at temperatures below 100 C [1].
These rhenium-based catalysts were thoroughly investigated be-
cause of their high performance even near room temperature,
either in heterogeneous or in homogeneous systems [12]. The
improvement in the catalytic activity of rhenium-containing ma-
terials has been attributed to Brönsted and/or Lewis acid sites;
aluminas and silica–aluminas which contain these two types of
acid sites are then better supports than pure siliceous materials
with rather weak Brönsted centers. Thus, ≡Si–O–ReO3 moieties
may be regarded as completely inactive or weakly active rhenium
species at room temperature in olefin metathesis (on standard
amorphous silica [13] or mesoporous materials [14]). The surface
complex ≡Si–O–[Re(C-t-Bu)(CH-t-Bu)(CH2-t-Bu)] supported on sil-
ica is an active metathesis catalyst or precursor [15], even for
functionalized olefins [16]; this is certainly due to its well defined
preformed complex with the key alkylidene ligand (vide supra and
[7,8]), and not to the acid properties of the silica aerosil. Thus,
even if several studies conclude that the acidity of the support
is one of the key parameters in promoting the activity, prefer-
ential formation and site isolation of =Al–O–ReO3 linkages with
high rhenium dispersion are certainly important. Estimation of the
◦
fairly high activity also with 2-pentene at 25 C, tentatively re-
Corresponding author.
E-mail address: emmanuel.briot@upmc.fr (E. Briot).
*
0021-9517/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2008.05.026

62
P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
plexes to exist. Thus, comparison of Ph2SiH2 and Bu4Sn will be
considered in conjunction with the role of the support (aluminas
and silica–aluminas, and some mesoporous analogues).
The solid samples were characterized by elemental analysis,
X-ray powder diffraction, transmission electron microscopy (TEM)
and EDX analysis, dinitrogen adsorption–desorption at 77 K, and
solid state 27Al and 1H MAS-NMR spectroscopies. The activities of
these catalysts for 1-octene metathesis at room temperature are
compared.
Scheme 1. Grafting of alkoxochlororhenium complexes on oxides.
number of active sites present in conventional ReOx/γ -Al2O3 (pre-
+
−
−
+
pared with NH4 , ReO4 or H , ReO4 ) catalysts by energy dispersive
X-ray analysis (EDX) suggested that a very low percentage (less
than 1%) of the rhenium was really involved in the metathesis
reaction [17]. This proposal led to the design of novel heteroge-
neous catalysts obtained by grafting alkoxochlororhenium species,
[ReClx(OEt)y]n, onto γ -alumina [17]. The surface reaction is rep-
resented in Scheme 1 where Matrix–(OH) denotes bare Brönsted
sites, preferentially =Al–O–H.
2. Experimental
2.1. Reagents
After thermal treatment, the activity of the resulting pre-
catalyst with ca. 1.5 wt% Re [18] exceeds that of known ReOx/Al2O3
prepared by the conventional method (capillary or dry impregna-
tion with NH4ReO4 or perrhenic acid involving 7.5 wt% Re). These
results suggest that carbene rhenium surface species represent
only a small part of the rhenium deposited by “dry impregna-
tion” of γ -alumina. An improvement in the catalytic activities was
already claimed (vide infra) when active phases were supported
on mesostructured materials, by comparison with conventional γ -
Al2O3- or SiO2-based catalysts. However, are mesoporous materials
containing aluminum oxide really more promising than classical
industrial supports, to develop rhenium alkylidene intermediates
involved in the propagation of the catalysis cycle?
Since the early 1990s and the advent of mesoporous silica [19,
20], several transition metal species, particularly with molybde-
num [14,21,22], tungsten [14,23,24], rhenium [14,25–31] and ruthe-
nium [32–35], have been grafted onto or inserted into mesoporous
materials. Great efforts have been devoted to improving the in-
corporation of Al(III) into mesoporous silica in relation with: (i)
the synthesis of classical silica–alumina analogues; (ii) the sup-
posed easy substitution of Si by Al, and (iii) the creation of Lewis
acid sites and an increase in Brönsted acidity and ion exchange
capacity (vide supra). Few rhenium-containing mesoporous mate-
rials with aluminum oxide have been reported, probably because
mesoporous alumina or well defined aluminum-containing meso-
porous silica is more difficult to synthesize than the correspond-
ing pure siliceous materials. Rhenium oxide supported on ordered
mesoporous aluminas has been claimed as more active and selec-
tive than Re2O7/γ -Al2O3 for alkene metathesis, but high optimum
loadings of 6–15% are considered [25,27,29–31] and isomerization
occurring on Brönsted acid sites can also occur [25,29,30]. The cat-
alysts were prepared with ammonium perrhenate, perrhenic acid
(incipient wetness impregnation, thermal spreading, etc.) or graft-
ing of MTO [26,28,36–38]. A second question can be raised: are the
low rhenium contents in the range of those proposed for the graft-
ing of alkoxochlororhenium species onto γ -alumina [17] likely to
work with alumina-based mesoporous materials?
Keeping in mind these two questions, the present paper con-
siders the objective of preparing low-loaded precursors (0.5–2 wt%
Re) via a route involving the grafting (Scheme 1) of [ReClx(OR)y],
previously synthesized from [Re2Cl10]/anhydrous ethanol [17], γ -
alumina, silica–aluminas (25 and 75 wt% SiO2), and mesoporous
alumina or silica–alumina. We observed that promotion by chlori-
nation is particularly effective in alkene metathesis, this leads us
to use chlorinated alumina as a support of chlororhenium species.
How an activating agent or promoter forms active sites is still
a subject of interest. One of the first model complexes for the
activation of silanes, [Ph2SiH2Re2(CO)8], probably has an agostic
Si–H···Re interaction [39]; this led us to use Ph2SiH2 to modify
the surface of rhenium deposits on aluminum oxide and to gener-
ate hydrido- and/or even carbene rhenium species [18]. It is well
known that Si–H bonds represent a particularly interesting case
because the interaction can be strong enough for η2-SiH2R2 com-
1-Propanol (Rectapur), sodium silicate solution (27 wt% SiO2,
8 wt% Na2O, Rectapur), C6H5Cl (PhCl, 99.6%) and 4 Å molec-
ular sieves were purchased from Prolabo. Aluminum tri-sec-
butoxide (97%) [Al(OCH(CH3)(C2H5))3], aluminum isopropoxide
(98%) [Al(OCH(CH3)2)3], dry hexane (99%), lauric acid (99%) CH3-
(CH2)10COOH and [Re2Cl10] were obtained from Aldrich. Bu4Sn
(98%), 1-octene (99%) (without peroxides, stored under dry dinitro-
gen), and tetradecyltrimethylammonium bromide (C14 TMABr, 98%)
were purchased from Fluka, Ph2SiH2 [(C6H5)2SiH2] from Merck,
and H2SO4 and C2H5OH (anhydrous, 99.9%) from Carlo Erba. Per-
rhenic acid [HReO4] in aqueous solution was prepared by slow
addition of finely powdered samples of metallic rhenium to hy-
drogen peroxide (30 wt% aqueous H2O2, stabilizer: stannate); it
was neutralized by aqueous ammonia and concentrated to get
white crystalline ammonium perrhenate, NH4ReO4 [40]. CAUTION:
self-heating can rapidly increase the decomposition rate of desta-
bilized hydrogen peroxide. Carry out the reaction on
a small
scale. Tetrabutylstannate (Bu4Sn) is a highly toxic flammable liq-
uid. Appropriate safety precautions and procedures should be
employed during all stages of the handling and disposal of this
compound. γ -Alumina and silica–aluminas with 25 or 75 wt%
SiO2 were commercial products (Rhodia) with surface areas of
224, 280 and 330 m2 g−1, and pore volumes of 0.53, 0.80 and
0.80 cm3 g−1, respectively. Elemental analysis did not reveal the
presence of any impurity. Chlorinated alumina (γ -Al2O3–Cl) was
obtained by interaction of HCl(g)/H2 (2/1) with γ -alumina (sur-
face area: 206 m2 g−1, pore volume: 0.46 cm3 g−1) [41,42] (see
Section 2.3.4).
2.2. Characterizations
X-ray diffraction patterns were recorded on a Bruker D8 diffrac-
◦
tometer using Cu K radiation (1.54 Å) with 0.02 (2θ) step size
α
◦
and 6 s step time in the range 2 < 2θ < 10 . Adsorption and des-
orption isotherms for nitrogen were obtained at 77 K using a Mi-
cromeritics ASAP 2010. The samples were outgassed at 393 K and
0.3 Pa for 4 h before measurements. Elemental analyses were car-
ried out at the Service Central d’Analyse (CNRS-Lyon) by inductive
coupling plasma-atomic emission spectroscopy (ICP-AES) after al-
kaline fusion with Li2B4O7. TEM and EDX analyses were performed
using a JEOL JEM 2010 transmission electron microscope operating
at 200 kV, equipped with a PGT Imix-PC system. The X-ray peaks
(Si K at 1.75 keV, Al K at 1.487 keV, Cl K at 2.62 keV, Re L
α
α
α
α
at 8.652 keV, Re L at 10.009 keV, Re L
at 10.273 keV and Re
β1
β2
L
at 11.684 keV) emitted from a part of the bulk and/or from the
γ 1
surface (spot beam analysis) were recorded in the 0–15 keV range.
Data do not include oxygen composition because the instrument is
not sensitive to oxygen. The samples were dispersed in cyclohex-
ane by sonication and placed on carbon-covered copper grids and
dried in dry air. Room-temperature solid-state magic-angle spin-
ning nuclear magnetic resonance (MAS NMR) experiments were
performed on a Bruker AVANCE400 spectrometer at 9.4 T in 4 mm

P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
63
zirconia rotors. 27Al one-pulse experiments (aqueous solution of
Al(NO3)3 as reference) were performed by spinning at 12 kHz with
a selective pulse (<π/6) duration of 1 μs, recycle time 1 s and
2048 acquisitions. The triple quantum MAS (3QMAS) spectra were
acquired with Z-filtering [43] and hypercomplex (States) phase de-
tection. The pulse durations were set empirically to 4 and 1.5 μs
for the first and the second pulses, respectively (ΩI /2π = 90 kHz),
and to 8.5 μs for the third selective pulse (ΩI /2π = 10 kHz). These
values maximized the observed signal. 32 t1 increments of 31 μs
were acquired in the first dimension. A total of 2000 scans was
accumulated with a recycle time of 1 s. Data processing including
shearing and scaling according to the Cz convention [44] was per-
formed using the xfshear “au” NMR program. Room-temperature
single-pulse 1H experiments at 400 MHz (TMS as reference) were
Scheme 2. Synthesis of chlorinated alumina: initiation steps.
for the replacement of some surface hydroxyl groups by chlorine
and the formation of water and =Al–Cl surface species. Detailed
analyses of the dehydroxylation process cannot determine the co-
ordination of the last OH groups after thermoevacuation.
◦
performed with a 90 pulse duration of 2 μs and a recycle delay of
The material, shaped in ball form, was a γ -alumina from Akzo
40 s (8 scans). The MAS equipment for rotation (14 kHz) was care-
fully cleaned to avoid spurious proton signals. The probe signal was
subtracted from the total FID. For the experiments with Ph2SiH2
and mesoporous alumina, the samples were treated in vacuum (be-
low 1 Pa) at temperatures between 300 and 550 K. Ph2SiH2 (10 μL)
was introduced with a micro-syringe under reduced pressure.
(Ketjen CK-300) with BET area 206 m2 g−1, and pore volume
0.46 cm3 g−1. This solid (1.5 g) was first dehydrated at 350 C
◦
(12 h) in a silica glass reactor freed of grease. Chlorination was
◦
then carried out at atmospheric pressure and 580 C, with a gas
flow (0.5 cm3 s−1) containing HCl/H2 (2/1) for 4 h. The pre-catalyst
◦
thus obtained was quickly cooled to 60 C in a reduced flow of the
◦
chlorination gas and dry dinitrogen 99.99% to 20 C. The chlorine
2.3. Catalyst preparation
content was in the range 1.90–2.10 wt%.
2.3.1. Mesoporous alumina (Al2O3/meso)
2.3.5. Rhenium grafting onto γ -Al2O3, silica–alumina and onto
mesoporous supports
Mesoporous alumina was synthesized according to a slightly
ˇ
modified procedure described by Cejka et al. [45], with anionic
It is essential that the matrix material used for grafting be free
of adsorbed water, as this would merely react with and destroy the
rhenium(V) complex (it is known that [Re2Cl10] hydrolysis gives
Re(IV) and Re(VII) species). 1 g of the support was evacuated at
surfactant. Typically, to a solution of 3.4 g (0.015 mol) of lauric
acid in 100 mL (1.2 mol) of 1-propanol and 3.1 mL (0.172 mol) of
H2O, stirred for 30 min at room temperature, 13.7 g (0.05 mol)
of aluminum tri-sec-butoxide was added and the reaction mixture
stirred for 20 min. The gel was aged in a static Teflon-coated stain-
◦
◦
200 C (35 C h−1) for 2 h. The solid was then added to a solu-
−4
◦
tion of [Re2Cl10] (40 mg, 1.1 × 10
mol) and anhydrous C2H5OH
less steel autoclave and held at 100 C for 50 h. The molar compo-
(20 mL) under anaerobic conditions (dry dinitrogen). The mixture
sition of the synthesis gel was: 1 Al/3.44 H2O/0.3 lauric acid/24
C3H7OH. After cooling to room temperature, the resulting mixture
was suction-filtered, the recovered solid treated with C2H5OH and
dried overnight over P4O10. Before grafting with the solution of
Re(V) chloroalkoxides, solids were calcined for 2 h in dinitrogen
◦
was refluxed for 6 h in N2 at 65 C and the solid dried overnight
over P4O10 after removing the solvent. The solid was calcined for
◦
◦
2 h in dry N2 (1 cm3 s−1) at 350 C (120 C h−1). It should be
noted that the entire rhenium grafting procedure must be done in
a dry atmosphere. As the solid thus obtained is highly moisture-
sensitive, it is kept in a sealed container in dry dinitrogen.
◦
◦
−1
99.99% (2.5 cm3 s−1) at 410 C (30 C h−1) and in air (2.5 cm3 s
)
◦
◦
at 420 C for 9 h (30 C h−1).
2.3.6. Rhenium oxide (ReOx) on γ -Al2O3 prepared by dry or incipient
wetness impregnation, denoted Re–Al2O3 (reforming)
2.3.2. MCM-41 mesoporous silica (SiO2/meso) [46]
9.9 g (0.0408 mol) of sodium silicate solution were diluted with
30 mL (1.66 mol) of water and slowly added to a mixture of 7.5 g
(0.022 mol) of C14TMABr in 80 mL (4.4 mol) of H2O. After stirring
for 30 min at room temperature, the pH of the gel was adjusted
to 10 with 2 M H2SO4, the solution was stirred for 30 min more
Extruded samples with Cl content lower than 200 ppm (1.5 g)
were submitted to constant and gentle shaking, and an aqueous
solution of a rhenium salt such as NH4ReO4 or perrhenic acid
was slowly added. The solid and the solution were placed in ther-
◦
◦
moregulated (50 C) Erlenmeyer flasks and graduated burettes, re-
and aged in a stainless steel autoclave and held at 100 C for two
spectively. For the rhenium content of 7.0 wt%, three dry impregna-
tion steps with NH4ReO4 were necessary. Each step was followed
days. The solid was suction-filtered and dried over P4O10, and then
◦
calcined for 6 h in air (2.5 cm3 s−1) at 550 C.
◦
◦
by drying at 110 C and calcination at 550 C for 6 h in flowing air
(1.7 cm3 s−1).
2.3.3. Aluminum grafting onto MCM-41 silica (SiO2–Al2O3/meso)
[47,48]
2.4. Catalysis tests
A solution of 1.4 g (0.006 mol) of aluminum isopropoxide in dry
hexane (150 mL) was added to a suspension of 2.0 g of all-silica
MCM-41, synthesized according to Section 2.3.2, in 50 mL of dry
hexane. The mixture was stirred for 10 min and aged for 30 h at
room temperature, the rest of the synthesis being identical with
that for mesoporous silica (see Section 2.3.2).
◦
Catalysts were tested in 1-octene metathesis at 20 C with a
substrate/Re molar ratio of about 100. In a typical experiment,
with stirring at room temperature and atmospheric pressure in
a very slow flow of dry N2, ca. 0.4 g of the catalyst was placed
in a micro-reactor without solvent or with 2 mL of dry PhCl and
10 μL of tetrabutyltin (Bu4Sn) or of diphenylsilane (Ph2SiH2), active
sites are generated quickly; then 0.7 mL of 1-octene (4 mmol) was
added. Ethylene was continuously released from the reactor. The
progress of the reaction was monitored by GC with 0.2 μL of su-
pernatant at given intervals (Delsi 30 gas chromatograph equipped
with 0.25-mm-diameter 50 m OV-1701 capillary column and a
2.3.4. γ -Alumina chlorinated with hydrogen chloride (Al2O3–Cl)
Hydrogen chloride is known to react with the hydroxyl groups
◦
of γ -alumina from 280 to 730 C. The reaction is thought to occur
via two dissociative pathways (Scheme 2), the latter playing a mi-
nor role. The main step generates initially Cl–Al(OH) sites; some of
◦
them are dehydroxylated at 580 C. Infrared data provide evidence

64
P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
Table 1
Elemental analysis and textural properties of initial supports and rhenium-containing catalysts
b
c
Entry
Pre-catalyst
Re
(wt%)
Cl
(wt%)
Cl/Re
(mol/mol)
Si/Al
(mol/mol)
Al/Re
(mol/mol)
SBET
V p
a0
(Å)
dp
wd
(Å)
(m2
g
)
(cm3
g
(Å)
−1
−1
)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
SiO2–Al2O3 (75/25)
Re–SiO2–Al2O3 (75/25)
SiO2–Al2O3 (25/75)
Re–SiO2–Al2O3 (25/75)
Al2O3
Re–Al2O3
Al2O3-Cl
Re–Al2O3–Cl
Re–Al2O3 (reforming)
SiO2/meso
–
0.42
–
1.11
–
2.18
–
0.67
6.98
–
–
1.50
–
–
0.40
–
0.80
–
0.90
1.90
2.27
<200 ppm
–
–
1.52
–
2.54e
2.54
0.28f
0.28
–
–
–
–
–
–
330
329
280
279
224
214
198
199
155
1285
949
874
686
668
0.80
0.78
0.80
0.75
0.53
0.50
0.46
0.46
0.43
1.18
0.79
0.74
0.50
0.35
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
9
14
14
43
55
5.0
3.8
2.2
2.9
217 (45)a
–
246 (134)a
–
164 (178)a
–
541 (178)a
45
–
–
–
–
–
46
47
48
72
76
37
33
34
29
21
SiO2–Al2O3/meso
Re–SiO2–Al2O3/meso
Al2O3/meso
3.40
3.40
–
5.3
5.9
44 (36)a
–
Re–Al2O3/meso
2.87
3.23
–
124 (178)a
a
Expected Al/Re molar ratio from rhenium grafting on the supports with [Re2Cl10 ]/EtOH solution.
b
c
d
e
f
The mesopore volume V p is determined at P/P0 ≈ 0.97.
The mean pore diameter dp = 4V p/SBET
.
The wall thickness w = a0 − dp
EDX analysis Si/Al = 2.56.
EDX analysis Si/Al = 0.29.
.
flame ionization detector linked to a Delsi Enica 10 electronic inte-
grator). GC–MS analyses were performed on a Trace GC 2000 series
(ThermoQuest) coupled to a Trace MS.
ing NH4ReO4 (see Section 2.3.6) have very inhomogeneous distri-
butions with ReOx or Re2O7 clusters and uncovered zones (entry 9,
Table 1).
3. Results and discussion
3.2. X-ray analysis
3.1. Rhenium and chlorine contents of pre-catalysts
The X-ray diffraction pattern of calcined SiO2/meso before alu-
minum and rhenium grafting (Fig. 1a) displays three well resolved
broad peaks at low angle, corresponding to the (100), (110) and
(200) reflexions. Another peak, attributed to the (210) reflexion, is
Two classes of materials can be distinguished in the grafting
of Re(V) complexes (Table 1): silica–aluminas incorporate less Re
than expected (see entries 2, 4, 12), whereas with pure alumina
supports, there is apparently an increase in surface reactivity (see
entries 6, 14). The corresponding Al/Re molar ratios (Table 1) are
easily understood by the formation of soluble =Al–O–[Re] species
commonly observed for other anchoring processes on silica or
alumina [22–24,49]. This is related to the poor stability of the
mesostructured alumina (entry 14). The analytical data suggest
that some catalyst supports are chlorinated during the grafting
process (entries 2, 4, 8, 12, 14). Thus, it was interesting to consider
also chlorinated alumina. The pre-catalyst Re–Al2O3–Cl (entry 8),
with 0.67 wt% Re has, with SiO2–Al2O3 (75/25), the lowest num-
ber of Al sites to accommodate rhenium surface species. These
results are interpreted by the stronger affinity of Re(V) species
for residual Al(III)–OH than Si(IV)–OH groups (see Section 2.3.4);
for Re–Al2O3–Cl, the low Re content can be explained by a defi-
ciency of –OH groups, due to the presence of major =Al–Cl surface
species (see Section 2.3.4 and Scheme 2). We observed that pure
alumina supports (entries 6, 14) are better than silica–alumina
(75/25) and, to some extent, silica–alumina (25/75) for rhenium
grafting, since the entire Re(V) used in the synthesis is preferen-
tially grafted onto the surface aluminum centers.
◦
present at 2θ = 5.8 , indicating that the hydrothermal treatment
used in the synthesis (see Section 2.3.2) enhances the matura-
tion of the solid, leading to organized mesoporous material. The
three first reflexion peaks are also observed for SiO2–Al2O3/meso
(not shown) and Re–SiO2–Al2O3/meso (Fig. 1b) and their slight
shift indicates that post-grafting of Al(III) and subsequently Re(V)
onto mesoporous silica does not drastically modify its mesoporous
structure in the bulk, the unit cell parameter a0 being approxi-
mately the same after these chemical modifications of the surface
(see Table 1).
Low-angle powder X-ray diffraction patterns of calcined Al2O3/
meso and Re–Al2O3/meso, exhibit only one very broad peak at 2θ
◦
around 1.5 (Fig. 2), indicating their relatively poor mesoporous
organization. It is well known that mesoporous aluminas are more
difficult to synthesize than siliceous analogues, and the diffraction
patterns usually published, are very similar to those obtained in
Fig. 2 [45]. Well-ordered mesoporous alumina materials have also
been prepared [50] but the BET specific surface area is significantly
lower.
The Cl signals (K line of EDX spectra) reveal that the chlo-
α
3.3. Dinitrogen sorption at 77 K
rine distribution in the sample Al2O3–Cl is quasi-uniform over
the alumina surface, with mean Irel. Cl/Irel. Al ca. 49 × 10−2. Des-
orption of HCl and/or Cl2 was controlled conductimetrically [17];
3.3.1. Conventional supports: aluminas and silica–aluminas
Physisorption measurements of reference samples (bare sup-
ports) (Fig. 3) all exhibit type IV isotherms, according to IUPAC
nomenclature [51], being characteristic of materials with wide
pores and low specific surface areas (150–330 m2 g−1, see Table 1).
It should be noted that the grafting of these supports with alkox-
ochloro complexes of Re(V) does not drastically change their sur-
face properties, since V p, SBET and the shape of the isotherms are
approximately constant.
◦
◦
they desorb over the temperature range 70–600 C (120 C h−1, N2:
0.85 cm3 s−1). The detection limit of the method is 0.1% of the total
chloride content of the samples. Analyzing the chlorine contents of
the solids after the catalysis test showed that Cl was not released
at room temperature.
EDX multiple-spot analysis revealed that rhenium distributions
in the grafted solids were nearly homogeneous. However, samples
prepared according to the conventional impregnation method us-

P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
65
Fig. 1. X-ray powder diffraction patterns of calcined (a) SiO2/meso and (b) Re–SiO2–Al2O3/meso.
Fig. 2. X-ray powder diffraction patterns of calcined (a) Al2O3/meso and (b) Re–Al2O3/meso.
3.3.2. Mesoporous materials
All the mesoporous supports (Fig. 4) display nearly hysteresis-
free type IV isotherms, with a capillary condensation near P/P0
0.3, except for Al2O3/meso and Re–Al2O3/meso for which the ad-
sorbed N2 does not increase very much in the entire range of
relative pressure.
The physisorption of dinitrogen onto mesoporous alumina can
be related to a broad distribution of mesopores with an average
diameter of 20–30 Å (Table 1, entries 13–14), resulting from a
poor organization of the pore network and a high specific sur-
face area (686 m2 g−1) after calcination at moderate temperature
(see Section 2.3.1). The modification of the surface properties of
these mesoporous supports after Al and/or Re grafting depends
on whether alumina or silica–alumina is involved. Pore volume
and SBET of mesoporous silica are clearly more affected by alu-
minum grafting than by subsequent rhenium grafting. The slight
decrease in SBET and the large change in V p after rhenium graft-
ing for Al2O3/meso may be due to some pore blocking and to the
formation of soluble =Al–O–[Re] species (see Section 3.1).
3.4. Solid state 27Al MAS-NMR
One-pulse 27Al MAS-NMR spectra of as-synthesized Al2O3/meso
materials (Fig. 5a) give a broad peak centered at about 6 ppm
and another centered at 67 ppm, attributed to hexa- and tetra-
coordinated aluminum, respectively; for SiO2–Al2O3/meso (not
shown), there is an additional weak peak at 35 ppm.
The lack of crystallinity in typical silica–aluminas precludes
the occurrence of relatively well separated 29Si MAS-NMR peaks
found for zeolites with different n values in Si(OAl)n(OSi)4−n struc-
tural units. Thus, the deconvoluted 29Si MAS contributions near
−90, −100 and −110 ppm (with respect to TMS), attributed to
Si(OSi)2(OAl)2, Si(OSi)3(OAl) and Si(OSi)4, respectively, also corre-
spond to the Q2, Q3 and Q4 sites of pure silicas [21]. Similarly,
27Al spectra have not provided additional structural information,
although further studies may differentiate Al2O3 from SiO2–Al2O3.
After calcination (Fig. 5b), the distribution of the two coordinating
sites of the aluminum center is modified and a new peak appears
at 35 ppm. The 2D MQMAS spectra presented in Fig. 6 (see also δiso
=

66
P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
Fig. 4. Dinitrogen adsorption–desorption isotherms at 77 K of calcined: (a) SiO2/
meso, (b) SiO2–Al2O3/meso, (c) Re–SiO2–Al2O3/meso, (d) Al2O3/meso and (e) Re–
Al2O3/meso.
Fig. 3. Dinitrogen adsorption–desorption isotherms at 77 K. Solid lines, isotherms
of the bare supports: (a) SiO2–Al2O3 (75/25), (b) SiO2–Al2O3 (25/75), (c) Al2O3 and
(d) Al2O3–Cl. Dashed lines, isotherms of the grafted supports after thermal treat-
ments: (a) Re–SiO2–Al2O3 (75/25), (b) Re–SiO2–Al2O3 (25/75), (c) Re–Al2O3, (d) Re–
Al2O3–Cl and (e) Re–Al2O3 (reforming).
3.5. Solid state 1H MAS-NMR
and CQ η values) allow us to attribute this intermediate peak un-
ambiguously to penta-coordinated Al. Calcined SiO2–Al2O3 displays
approximately the same spectrum (not shown) as mesoporous ma-
terials but the distribution of aluminum surface species is slightly
different (very little penta-coordinated Al, denoted Al(5)), although
the peaks at 67, 35 and 6 ppm are poorly resolved. For γ -Al2O3,
the spectrum (not shown) of the calcined sample is also poorly re-
solved but there is no peak corresponding to penta-coordinated Al.
Under our conditions, with the lack of an internal standard and
the very real possibility that a significant fraction of the Al is not
detected once the materials are calcined, the 27Al NMR of amor-
phous oxides at intermediate field strengths is notoriously non-
quantitative. The grafting of materials by Re(V) species (Figs. 5b, 5c,
6a, and 6b) does not modify the 27Al NMR spectra, meaning that
the overall structure is unchanged by grafting and thermal treat-
ment. We must keep in mind that 27Al MQ-MAS NMR spectrom-
etry is not sensitive enough to differentiate surface modifications
from bulk properties. The signals are asymmetric with a tailing
high-field edge and a steep low-field edge (Fig. 5), characteristic of
a distorted environment. The MQ-MAS signals obtained for the Al
species (Fig. 6) are very spread out along the QIS axis (Quadrupolar
induced shift) indicating a large distribution of quadrupolar con-
stants. This confirms the disorder [52].
It appears that calcined alumina and silica–alumina meso-
porous materials cannot be easily distinguished by 27Al MAS-NMR
(vide supra), whereas 1H NMR spectra (Fig. 7) clearly indicate
that both mesoporous silica–alumina and alumina possess Brön-
sted acid sites, in different amounts and strengths. The peaks at
1.0–1.7 ppm (Fig. 7a) and at 0.9–2.0 ppm (Fig. 7b) are attributed to
isolated (non-interacting) terminal Al–OH and silanol groups [53–
55]. As shown previously [54], microdomains with Y zeolite struc-
ture can exist in “non-equilibrated” partially dehydrated samples.
They give weak signals and chemical shift values near 4 ppm (3.7
and 3.6 ppm, Fig. 7, a and b). Adventitious water is difficult to ex-
clude from mesoporous materials and even from conventional sup-
ports that are “practically dry” [54]. The peak at 5.4 ppm (Fig. 7b)
is attributed to hydroxyl and residual water [56–58]. With “in
equilibrium amorphous materials,” all protons in silica-rich silica–
aluminas are associated with isolated silanol sites and OH groups
bonded to aluminum atoms with a low symmetry environment
[36,54,58,59]. However, we are not able to estimate the relative
strength of these Brönsted acid sites in silica–alumina (3.7 ppm)
and in alumina (3.6 and 5.4 ppm), although there are more pro-
tons in alumina than in silica–alumina. The lack of a signal near
7 ppm, characteristic of strong Brönsted acid sites [56], indicates
that they are absent from SiO2–Al2O3/meso sample [57].
The 1H NMR spectrum (Fig. 7c) of Ph2SiH2 adsorbed on
Al2O3/meso shows well resolved maxima. EDX data support a

P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
67
(a)
Fig. 5. One-pulse 27Al MAS-NMR spectra at 9.4 T of: (a) Al2O3/meso (as-synthe-
sized), (b) calcined Al2O3/meso and (c) calcined Re–Al2O3/meso. ∗: spinning side-
bands.
homogeneous distribution of the Si of the promoter in these ex-
periments. The peaks near 7 ppm and 4.5 ppm belong to aromatic
and Si–H protons, respectively. They are at 7.58, 7.36, 7.33 and
4.93 ppm in the liquid. The rest of the spectrum, being identi-
cal to that of Al2O3/meso, indicates that Ph2SiH2 does not interact
strongly with the alumina surface, even if the change of chemical
shift from liquid to surface species may suggest a certain proxim-
ity. However, the increase in the peak widths (Fig. 7d) indicates
that rhenium grafting onto Al2O3/meso may induce: (i) a stronger
Ph2SiH2/Re surface species interaction, (ii) a paramagnetic system
due to the proximity of rhenium and protons. We propose that
Fig. 7d is the fingerprint of a strong promoter (Ph2SiH2)/Re surface
complex interacting with an agostic Si–H···Re bonding as found
for [Ph2SiH2Re2(CO)8] [39].
(b)
Fig. 6. Sheared 27Al (I = 5/2) 3QMAS-NMR spectra at 9.4
T
of calcined:
iso
(a) Al2O3/meso, hexa-coordinated Al, denoted Al(6) (δ = 10.91 ppm, CQ η
=
CS
iso
CS
4.66 MHz), penta-coordinated Al, denoted Al(5) (δ
= 38.19 ppm, CQ η
=
iso
5.32 MHz), tetra-coordinated Al, denoted Al(4) (δ = 73.82 ppm, CQ η = 5.26 MHz)
CS
and (b) Re–Al2O3/meso, hexa-coordinated Al (δiso = 11.73 ppm, CQ η = 4.60 MHz),
CS
iso
penta-coordinated Al (δ = 38.36 ppm, CQ η = 5.40 MHz), tetra-coordinated Al
CS
iso
CS
(δ = 73.04 ppm, CQ η = 5.17 MHz).
metal center. Because ethylene was released from the reaction
mixture, the equilibrium was shifted towards the products and
conversions as high as 99–100% were achieved.
3.6. Catalysis tests
3.6.1. Bu4Sn as promoter
With Bu4Sn as promoter, silica–aluminas do not behave like
alumina supports: the first group shows high conversions and low
selectivities for 7-tetradecene (cis and trans), denoted C14 (runs 1,
2 and 6) and the second class of supports displays fair conver-
sions (56-90%) and high selectivities for C14 (runs 3, 4, 5 and 7).
Two competitive reactions are involved for the first group (runs
1, 2, 6): metathesis and isomerization. This observation is often
described as the result of the existence of Brönsted acid sites on
Conversion in 1-octene metathesis increases with the catalyst
activation temperature, but for better performance and compara-
tive study, the systems were activated with Bu4Sn or Ph2SiH2. In
this paper, a preliminary comparative study is based on 1-octene
metathesis (Table 2). 1-Octene is a moderately demanding olefin,
much less reactive than 2-pentene; it is one of the most appro-
priate substrates for semi-quantitative studies, other bulky olefins,
such as 2-methyl-2-pentene being unable to coordinate with the

68
P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
Table 2
◦
1-Octene metathesis over Re-containing catalysts, after 24 h reaction at 20 C in PhCl or without solvent
Run
Pre-catalyst
Promoter
Conv.
(%)
Sel. C14
(%)
Other major products
selectivity (%)
TOF
(h
−1
)
1
2
3
4
5
6
7
8
9
Re–SiO2–Al2O3 (75/25)
Re–SiO2–Al2O3 (25/75)
Re–Al2O3
Bu4Sn
Bu4Sn
Bu4Sn
Bu4Sn
Bu4Sn
Bu4Sn
Bu4Sn
Ph2SiH2
Ph2SiH2
Ph2SiH2
Ph2SiH2
93
89
56
90
78
99
80
84
78
61
80
34
63
100
96
100
30
99
42
89
20
74
28 (C13 ), 18 (C12
)
17
6
2
10
0.8
5
2
3
0.8
7
25 (C13 ), 8 (C12
–
)
Re–Al2O3–Cl
<4 (C13
)
Re–Al2O3 (reforming)
Re–SiO2–Al2O3/meso
Re–Al2O3/meso
31 (C13 ), 23 (C12 ), 13 (C11
–
)
a
Re–Al2O3
35 (C13 ), 6 (C12
4 (C15 ), 4 (C13
31 (C13 ), 10 (C12 ), 24 (C9), 10 (C7)
20 (C13
)
Re–Al2O3 (reforming)a
Re–Al2O3-Cla
)
10
11
Re–Al2O3/mesoa
)
2
a
No solvent.
display results intermediate between alumina (run 3) and silica-
rich materials (run 1).
With alumina supports, conversion and selectivity are related
to the physical and/or chemical treatments performed before the
catalysis tests. The yield with mesoporous alumina (run 7) is
slightly better than with alumina (run 3) and Re–Al2O3 (reforming)
(run 5) corresponding to the pre-catalyst obtained by capillary im-
pregnation with NH4ReO4 (see Section 2.3.6). Chlorinated alumina
(run 4), although it contains one of the lowest Re wt%, displays the
highest activity of all the catalysts (high conversion, selectivity and
best TOF of selective catalysts). As for silica–alumina, the use of
mesoporous supports does not markedly improve the catalytic ac-
tivity of Re-alumina, whatever the physical and/or chemical treat-
ments before the catalysis test. Oikawa et al. [25] have also ob-
served approximately the same conversion for 1-octene metathesis
with 6 wt% Re (prepared with NH4ReO4) on γ -Al2O3 or meso-
porous Al2O3 (TOF = 2 h−1). However, Hamtil et al. [30], with
catalysts prepared by thermal spreading of ammonium perrhen-
ate on alumina, have claimed that mesoporous supports are better
than γ -Al2O3 with the same loading (83% conversion after 10 h at
◦
40 C, TOF = 34 h−1, 7 wt% Re).
3.6.2. Ph2SiH2 as promoter
Several experiments with alumina supports have been per-
formed using Ph2SiH2 as promoter without solvent, in order
to create in situ “Re–O–Al–O–Si” species with hydrido entities
such as Si–H···Re keeping in mind the crystal structure [39] of
[Ph2SiH2Re2(CO)8] (runs 8 to 11). For Re–Al2O3, the conversion
is improved (run 8), but with others, it remains of the same or-
der (runs 9 and 11) or is below the previous values (run 10)
with Bu4Sn as promoter. In all cases, the selectivities obtained
with Ph2SiH2 as co-catalyst are below the values obtained with
Bu4Sn. It should be emphasized that there is no synergy between
Ph2SiH2 and the alumina pre-catalysts, but there is an interac-
tion of Ph2SiH2 with the Re centers, as evidenced by the 1H NMR
spectra (vide supra, Fig. 7d), and it leads to isomerization and
metathesis.
Fig. 7. 1H MAS NMR spectra at 9.4 T of calcined: (a) SiO2–Al2O3/meso, (b) Al2O3/
meso, (c) Ph2SiH2/Al2O3/meso and (d) Ph2SiH2/Re–Al2O3/meso.
3.6.3. Isomerization properties of rhenium complexes at room
temperature
silica–alumina, which are believed to enhance the rate of isomer-
ization. However, since most of the samples display approximately
the same acidity (Lewis and Brönsted acid sites in both samples,
of comparable strengths), acidity cannot be the only parameter re-
sponsible for the observed differences at room temperature.
Among silica–alumina supports (runs 1, 2 and 6), silica-rich ma-
terials (runs 1 and 6) behave in the same manner, indicating that
mesoporous supports do not improve the catalytic activity, even at
higher Cl and Re contents (run 6, with Table 1, entry 12). Moreover,
the turnover frequency (TOF) is three times less with the meso-
porous supports (runs 1 and 6). Aluminum-rich materials (run 2)
According to previous investigations, allylic alcohols undergo
apparent double-bond isomerization at room temperature when
catalyzed by [ReClx(OR)y] complexes or species grafted on alu-
mina [17]. This was illustrated by the isomerization of 2-methyl-
3-buten-2-ol (A) to 3-methyl-2-buten-1-ol (B), in the absence of
◦
air and water (equilibrium is obtained in 3 min at 20 C), with
subsequent formation of ethers C and D (Scheme 3). The product
B is indicative of the formation of alkoxo complex intermediates
and of a cyclic rhenium-centered transition state [17,60,61]. Diallyl
ethers C and D are acid and/or rhenium-catalyzed (see Supplemen-
tary material, 1).

P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
69
Scheme 5. Proposed formation of [Re]–H complex by protonation [4].
reacting with the Al centers of the alumina. This forces the alkox-
ochlororhenium complexes to interact subsequently with residual
surface hydroxyl groups, leading to a better distribution of the
=Al–O–Re active sites. Rigorous proof of heterogeneity was ob-
tained by: (i) filtering the catalyst at the reaction temperature
before completion of the reaction and testing the filtrate and
(ii) analyzing the rhenium content of the filtrate (leaching of ac-
tive species does not occur in the presence of weak nucleophilic
reagents such as 1-octene and chlorobenzene). A correlation be-
tween metathesis activity and ZnCl2-modified aluminas has been
noted, but no structures for the grafted sites were proposed [26].
−
Zinc halides and halocomplexes with Cl ions may also serve as
chlorinating agents of the support, leading also to a better disper-
sion of MTO.
4. Conclusion
Estimation of the number of active sites in conventional ReOx/
γ -Al2O3 catalysts (7 wt%) prepared with NH4ReO4 (or HReO4),
suggests that less than 1% of the rhenium deposit is really ac-
tive for 1-octene metathesis. In contrast, the intimate grafting of
rhenium alkoxochloro species appears to be a reliable alternative
for preparing room-temperature active catalysts with low rhenium
contents (e.g. 0.4–0.7 wt%).
Scheme 3. Isomerization of allylic alcohols and subsequent formation of allylic
ethers.
Mesoporous solids (Al2O3/meso, SiO2–Al2O3/meso) were pre-
pared and compared to standard amorphous supports (γ -alumina,
silica–aluminas with 25 and 75 wt% SiO2). The synthesis of these
mesoporous materials with high specific surface areas was not,
at present, the trump card for metathesis under very mild condi-
tions. Variable amounts of penta-coordinated aluminum (Al(5)) are
evidenced by NMR studies, and they probably do not create pre-
ferred sites for anchoring rhenium centers. Conventional supports
such as alumina, reacted with rhenium alkoxochloro complexes,
with subsequent thermal treatments, show a clear increase in sur-
face reactivity. From the molar ratio Cl/Re, it appears that there
is a partial spreading and a promoting effect of chlorine, which
is confirmed with chlorinated alumina and grafted chloro rhenium
species. The precursors thus obtained display interesting catalytic
activity towards terminal olefin metathesis. The ease and low cost
of their synthesis without sophisticated ligands, as compared to
mesoporous materials, make them good candidates for alkene re-
actions at room temperature. The use of low-nuclearity rhenium
complexes should be considered for the preparation of fully re-
cyclable catalysts with site isolation of surface species, low metal
content and high dispersion.
We think that =Al–Cl groups have a protecting role in prevent-
ing deactivation and dimerization of rhenium surface species and
thus a stabilizing effect. Systems with pure alumina and chlori-
nated alumina supports, rhenium species and Bu4Sn lead to the
formation of 7-tetradecene and ethylene at 90–100% 1-octene con-
version. Coupling silica–alumina based catalysts with Bu4Sn or alu-
mina with a silane (here Ph2SiH2) leads to isomerization and self-
and cross-metathesis by dual catalysis due to hydrido and/or car-
bene rhenium species. 1H NMR studies show a strong interaction
between rhenium center and Ph2SiH2. Work is continuing on the
search for highly efficient low rhenium-content catalysts.
Scheme 4. Proposed formation of [Re]–H species for the isomerization of alkenes
via a hydride addition-elimination mechanism [4].
These results were extended to three rhenium complexes:
[ReO3(OSiR3)] (R = Me, Ph) [62] and MTO [63,64]. The mechanism,
with a 1–3 oxygen shift, does not correspond to classical isomer-
ization of alkene, which is a more demanding process, without a
transition metal complex. Thus, at room temperature, 1-octene is
completely isomerized in the presence of a catalyst with an acidity
function −H0 = 12 (H2SO4, 100%) with the formation of tertiary
carbocations [65]. To the best of our knowledge, the highly acidic
solids (γ -alumina/H2O(g), sulfated zirconia, chlorinated alumina,
etc.) are only active for the isomerization of butenes at higher
◦
temperature (250 C for γ -Al2O3). Control experiments with the
bare thermally treated supports, including chlorinated alumina
(without grafted rhenium species), confirm the absence of iso-
merization at room temperature with 1-octene. As an example,
sulfated γ -alumina showed 50% selectivity towards isobutene at
◦
approximately 50% conversion of butenes only at 450 C and mass
−1
hour space velocity, GHSV of 0.6 g cm3 h
[66]. Interestingly, even
at room temperature, isomerization of 1-octene or product olefins
is very high (runs 1, 2, 6, 8, 10 and 11). This can only be ex-
plained by a mechanism involving hydrido rhenium [Re]–H species
(see Supplementary material, 2), which are more easily gener-
ated with silanes such as Ph2SiH2 (Scheme 4), a hydride reagent
[39] or by protonation of the rhenium centers by surface Brön-
sted species (silica–alumina). Many inorganic complexes behave as
Lewis bases due to electron-rich metal centers and protonation is
+
a common reaction (see Scheme 5 for Re–H formation by H ad-
Mesoporous materials have no decisive advantage over indus-
trial extruded oxides. The challenge is to better control the as-
sembly, hydrolysis and condensation mechanisms in order to ob-
tain narrow pore size distributions and excellent thermal stability.
Some of the catalysts prepared now may have applications for
dition and [4]).
3.6.4. Chlorinated alumina systems
It appears that the activity of Re–Al2O3 can be increased by
treating the oxide first with an acid (HCl/H2 mixture) capable of

70
P.C. Bakala et al. / Journal of Catalysis 258 (2008) 61–70
the conversion of alkenes and even of alkanes by the combina-
tion of dehydrogenation/hydrogenation reactions under study in
our group. We think that promising synthetic perspectives may
also arise if it is possible to switch selectively from metathesis to
isomerization and back again at room temperature. There are many
examples of the development of useful synthetic methods where
the “occasional occurrence” of olefin isomerization reactions com-
bined with another reaction opens up useful synthetic methods
[4,67]. Tailor-made catalysts can be pure metathesis or “assisted
Tandem” catalysts.
[26] T. Oikawa, Y. Masui, T. Tanaka, Y. Chujo, M. Onaka, J. Organomet. Chem. 692
(2007) 554.
[27] J. Aguado, J.M. Escola, M.C. Castro, B. Paredes, Appl. Catal. A: Gen. 284 (2005)
47.
[28] C.D. Nunes, M. Pillinger, A.A. Valente, I.S. Gonçalves, J. Rocha, P. Ferreira, F.E.
Kühn, Eur. J. Inorg. Chem. (2002) 1100.
ˇ
[29] H. Balcar, R. Hamtil, N. Žilková, J. Cejka, Catal. Lett. 97 (2004) 25.
ˇ
[30] R. Hamtil, N. Žilková, H. Balcar, J. Cejka, Appl. Catal. A: Gen. 302 (2006) 193.
ˇ
[31] H. Balcar, R. Hamtil, N. Žilková, Z. Zhang, T.J. Pinnavaia, J. Cejka, Appl. Catal. A:
Gen. 320 (2007) 56.
[32] S. Krompiec, N. Kuz´nik, R. Penczek, J. Rzepa, J. Mrowiec-Bialon´ , J. Mol. Catal. A:
Chem. 219 (2004) 29.
[33] L. Li, J.L. Shi, Adv. Synth. Catal. 347 (2005) 1745.
[34] K. McEleney, D.P. Allen, A.E. Holliday, C.M. Crudden, Org. Lett. 8 (2006) 2663.
[35] A. Sakthivel, F.M. Pedro, A.S.T. Chiang, F.E. Kühn, Dalton Trans. (2007) 320.
[36] A.W. Moses, C. Raab, R.C. Nelson, H.D. Leifeste, N.A. Ramsahye, S. Chattopad-
hyay, J. Eckert, B.F. Chmelka, S.L. Scott, J. Am. Chem. Soc. 129 (2007) 8912.
[37] A.W. Moses, H.D. Leifeste, N.A. Ramsahye, J. Eckert, S.L. Scott, in: S.R. Schmitt
(Ed.), Catalysis of Organic Reactions, CRC Press–Taylor and Francis, Boca Raton,
2007, p. 13.
Acknowledgments
We thank Dr. Patricia Beaunier for EDX measurements and Dr.
John Lomas for constructive discussions and for correcting the
manuscript.
[38] A. Salameh, J. Joubert, A. Baudoin, W. Lukens, F. Delbecq, P. Sautet, J.-M. Basset,
C. Copéret, Angew. Chem., Int. Ed. Engl. 46 (2007) 3870.
Supplementary material
[39] J.K. Hoyano, M. Elder, W.A.G. Graham, J. Am. Chem. Soc. 91 (1969) 4568.
[40] L. Schriver, M. Jungfleish, S. Tribalat, in: G. Pannetier, A. Pacault (Eds.), Complé-
ments au nouveau traité de chimie minérale – technétium – rhénium, Masson,
Paris, 1978, p. 135.
The online version of this article contains additional supple-
mentary material.
Please visit DOI: 10.1016/j.jcat.2008.05.026.
[41] J.W. Myers, Ind. Eng. Chem. Prod. Res. Dev. 10 (1971) 200.
[42] G. Clet, J.M. Goupil, G. Szabo, D. Cornet, Appl. Catal. A: Gen. 202 (2000) 37.
[43] J.-P. Amoureux, C. Fernandez, S. Steuernagel, J. Magn. Reson., Ser. A 123 (1996)
116.
References
[1] K.J. Ivin, J.C. Mol, Olefin Metathesis and Metathesis Polymerization, Academic
Press, New York, 1997.
[2] R.H. Grubbs (Ed.), Handbook of Metathesis, Wiley–VCH, Weinheim, 2003, vols.
1–3: Catalyst Development (vol. 1); Applications in Organic Synthesis (vol. 2);
Applications in Polymer Synthesis (vol. 3).
[3] T.J. Katz, Angew. Chem., Int. Ed. Engl. 44 (2005) 3010.
[4] J.-M. Brégeault, Catalyse homogène par les complexes des métaux de transition,
Masson, Paris, 1992, pp. 128–157.
[5] W.A. Hermann, J. Organomet. Chem. 500 (1995) 149.
[6] C.C. Romão, F.E. Kühn, W.A. Hermann, Chem. Rev. 97 (1997) 3197.
[7] R. Toreki, R.R. Schrock, J. Am. Chem. Soc. 112 (1990) 2448.
[8] A.M. LaPointe, R.R. Schrock, Organometallics 14 (1995) 1875.
[9] D. Commereuc, J. Chem. Soc., Chem. Commun. (1995) 791.
[10] G. Doledec, D. Commereuc, J. Mol. Catal. A: Chem. 161 (2000) 125.
[11] D. Commereuc, H. Olivier-Bourbigou, V. Kruger-Tissot, L. Saussine, J. Mol. Catal.
A: Chem. 186 (2002) 215.
[44] J.-P. Amoureux, C. Fernandez, Solid State NMR 10 (1998) 211.
ˇ
[45] J. Cejka, N. Žilková, J. Rathouský, A. Zukal, Phys. Chem. Chem. Phys. 3 (2001)
5076.
[46] C.D. Nunes, A.A. Valente, M. Pillinger, A.C. Fernandes, C.C. Romão, J. Rocha, I.S.
Gonçalves, J. Mater. Chem. 12 (2002) 1735.
[47] R. Mokaya, Chem. Commun. (2000) 1891.
[48] R. Mokaya, W. Jones, Chem. Commun. (1997) 2185.
[49] F. Schekler-Nahama, O. Clause, D. Commereuc, J. Saussey, Appl. Catal. A:
Gen. 167 (1998) 247.
[50] K. Niesz, P. Yang, G.A. Somorjai, Chem. Commun. (2005) 1986.
[51] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity, second ed., Aca-
demic Press, London, 1995.
[52] A.A. Tsyganenko, P.P. Mardilovich, J. Chem. Soc., Faraday Trans. 92 (1996) 4843.
[53] C.E. Bronnimann, I.-S. Chuang, B.L. Hawkins, G.E. Maciel, J. Am. Chem. Soc. 109
(1987) 1562.
[54] C. Dorémieux-Morin, C. Martin, J.-M. Brégeault, J. Fraissard, Appl. Catal. 77
(1991) 149, and references therein.
[55] C. Dorémieux-Morin, P. Batamack, C. Martin, J.-M. Brégeault, J. Fraissard, Catal.
Lett. 9 (1991) 403.
[56] E.C. DeCanio, J.C. Edwards, J.W. Bruno, J. Catal. 148 (1994) 76.
[57] M. Hunger, D. Freude, H. Pfeifer, H. Bremer, M. Jank, K.-P. Wendlandt, Chem.
Phys. Lett. 100 (1983) 29.
[58] A. Omegna, J.A. van Bokhoven, R. Prins, J. Phys. Chem. B 107 (2003) 8854.
[59] G. Crépeau, V. Montouillout, A. Vimout, L. Mariey, T. Cseri, F. Mangé, J. Phys.
Chem. B 110 (2006) 15172.
[60] F. Derdar, J. Martin, C. Martin, J.-M. Brégeault, J. Mercier, J. Organomet. Chem.
338 (1988) C21.
[61] J.-M. Brégeault, C. Lepetit, F. Ziani-Derdar, O. Mohammedi, L. Salles, A. Deloffre,
Stud. Surf. Sci. Catal. 110 (1997) 545.
[62] S. Bellemin-Laponnaz, H. Gisie, J.-P. Le Ny, J.A. Osborn, Angew. Chem., Int. Ed.
Engl. 36 (1997) 976.
[63] J. Jacob, J.H. Espenson, J.H. Jensen, M.S. Gordon, Organometallics 17 (1998)
1835.
[64] A. Thome, M. Roeper, H.J. Kneuper, Patent DE-422 8887, August 29, 1992 (BASF
AG).
[65] S. Yolou, J.-M. Brégeault, Bull. Soc. Chim. Fr. (1979) 485.
[66] V. Macho, M. Králik, L. Jurecek, E. Jurecekova, J. Balazova, Appl. Catal. A:
Gen. 203 (2000) 5.
[12] J.C. Mol, Catal. Today 51 (1999) 289.
[13] L.G. Duguette, R.C. Cielinski, C.W. Jung, P.E. Garrou, J. Catal. 90 (1984) 362.
[14] J.-Y. Piquemal, E. Briot, M. Vennat, J.-M. Brégeault, G. Chottard, J.-M. Manoli,
Chem. Commun. (1999) 1195.
[15] M. Chabanas, A. Baudoin, C. Copéret, J.-M. Basset, J. Am. Chem. Soc. 123 (2001)
2062.
[16] M. Chabanas, C. Copéret, J.-M. Basset, Chem. Eur. J. 9 (2003) 971.
[17] J.-M. Brégeault, B. El Ali, J. Martin, C. Martin, F. Derdar, G. Bugli, M. Delamar,
J. Mol. Catal. 46 (1988) 37.
[18] F. Derdar, Ph.D. thesis, Pierre et Marie Curie University, 1988.
[19] T. Yanagisawa, T. Shimizu, K. Kuroda, C. Kato, Bull. Chem. Soc. Jpn. 63 (1990)
988 and 1535.
[20] J.S. Beck, J.C. Vartuli, J.W. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.W.
Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, J. Am.
Chem. Soc. 114 (1992) 10834.
[21] J.-Y. Piquemal, E. Briot, G. Chottard, P. Tougne, J.-M. Manoli, J.-M. Brégeault,
Microporous Mesoporous Mater. 58 (2003) 279, and references therein.
[22] P.C. Bakala, E. Briot, L. Salles, J.-M. Brégeault, Appl. Catal. A: Gen. 300 (2006)
91, and references therein.
[23] E. Briot, J.-Y. Piquemal, M. Vennat, J.-M. Brégeault, G. Chottard, J.-M. Manoli,
J. Mater. Chem. 10 (2000) 953, and references therein.
[24] J.-M. Brégeault, J.-Y. Piquemal, E. Briot, E. Duprey, F. Launay, L. Salles, M. Vennat,
A.P. Legrand, Microporous Mesoporous Mater. 44–45 (2001) 409, and refer-
ences therein.
[67] G.W. Parshall, S.D. Ittel, Homogeneous Catalysis, second ed., John Wiley and
Sons, New York, 1992.
[25] T. Oikawa, T. Ookoshi, T. Tanaka, T. Yamamoto, M. Onaka, Microporous Meso-
porous Mater. 74 (2004) 93.