Immobilisation of an ionically tagged Hoveyda catalyst on a supported ionic liquid membrane: An innovative approach for metathesis reactions in a catalytic membrane reactor

Article abstract of DOI:10.1016/j.cattod.2010.04.024

This study aimed at developing an innovative strategy to recycle homogeneous olefin metathesis catalysts by the combination of complementary green processes, namely organic solvent nanofiltration coupled with ionic liquid advantages. The immobilisation of

Full text of DOI:10.1016/j.cattod.2010.04.024

Catalysis Today 156 (2010) 268–275
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Immobilisation of an ionically tagged Hoveyda catalyst on a
supported ionic liquid membrane: An innovative approach
for metathesis reactions in a catalytic membrane reactor
A. Keraania,c, M. Rabiller-Baudrya,c,∗, C. Fischmeisterb,c,∗∗, C. Bruneaub,c
a
Université Rennes 1, UMR « Sciences Chimiques de Rennes » CNRS - équipe Chimie et Ingénierie des Procédés, 263 avenue du Général Leclerc,
CS 74205, case 1011, 35042 Rennes cedex, France
b
CNRS, UMR « Sciences Chimiques de Rennes » - Université Rennes 1- équipe Catalyse et Organométalliques, 263 avenue du Général Leclerc,
CS 74205, case 1011, 35042 Rennes cedex, France
c
Université Européenne de Bretagne, France
a r t i c l e i n f o
a b s t r a c t
Article history:
This study aimed at developing an innovative strategy to recycle homogeneous olefin metathesis catalysts
by the combination of complementary green processes, namely organic solvent nanofiltration coupled
with ionic liquid advantages. The immobilisation of an ionically tagged Hoveyda catalyst in an ionic liquid
previously supported on a solvent-resistant polyimide membrane (Starmem 228) was prepared leading to
an original catalytic membrane successfully used in a membrane reactor for a model metathesis reaction.
Keywords:
Nanofiltration
Metathesis
©
2010 Elsevier B.V. All rights reserved.
Ionic liquid
Immobilisation
Catalytic membrane reactor
1
. Introduction
systems induce mass transport limitations due to IL high viscosity
and require large amounts of sometimes expensive ionic liquids
[17,18]. Moreover, the reaction products are generally removed
from the IL by solvent extraction, reducing the environmental
benefits claimed for ILs but with a remaining attractivity thanks
to the possible enhancement of the catalytic performances [19].
Recently, a new approach using the concept of supported ionic
liquid phase (SILP) has been developed as an alternative to pure
biphasic systems which consist in the immobilisation of a cata-
lyst into an ionic liquid supported on a solid phase [20–22]. This
approach reduces the amount of ionic liquid required and can over-
come mass transport limitation. SILP technology has been used in
various reactions such as hydroformylation [23,24], Friedel–Crafts
acylations [25], Suzuki cross-coupling [26,27], allylic substitu-
tion [28], Heck reaction [29] and hydrogenation [30] using a
broad range of solid phases such as silica [20,24,25,29,31], alu-
mina [23,27], MFI zeolite [26], chitosan [28] and carbon nanotubes
Ruthenium-catalysed olefin metathesis reactions represent a
powerful method for the formation of carbon–carbon double bonds
allowing synthesis of various intermediate compounds useful for
fine chemistry [1–8]. Among catalysts based on ruthenium, Hov-
eyda catalysts [9–11] are well known to be stable homogeneous
metathesis catalysts used at laboratory scale. These catalysts have
demonstrated their abilities to perform extremely selective organic
syntheses with high yields. Nevertheless, only few industrial pro-
cesses use metathesis reactions owing to the cost of homogeneous
ruthenium-catalysts together with difficulties of final separation
of the catalysts from the reaction media and their subsequent
recycling. To overcome this major problem limiting industrial
development, immobilisation of the Hoveyda catalysts in biphasic
systems using ionic liquids (IL) is an attractive approach for the sep-
aration and recycling [12–16]. However, biphasic IL-organic solvent
[
32].
Even though the necessity to recycle homogeneous
∗ Corresponding author at: Université Rennes 1, UMR « Sciences Chimiques de
organometallic catalysts is commonly accepted, the applied
separation technique has to be selected with an eco-design
objective. Indeed, the separation steps represent up to 70% of
the operating costs of a manufactured product and participate
as much as 45% of its global energy cost [33]. Organic Solvent
NanoFiltration (OSNF) is a recent process that allows separation
of organic mixtures at molecular level by applying a pressure
Rennes » CNRS - équipe Chimie et Ingénierie des Procédés, 263 avenue du Général
Leclerc, CS 74205, case 1011, 35042 Rennes cedex, France.
∗
∗ Corresponding author at: CNRS, UMR « Sciences Chimiques de Rennes » - Uni-
versité Rennes 1- équipe Catalyse et Organométalliques, 263 avenue du Général
Leclerc, CS 74205, case 1011, 35042 Rennes cedex, France.
E-mail addresses: murielle.rabiller-baudry@univ-rennes1.fr
M. Rabiller-Baudry), cedric.fischmeister@univ-rennes1.fr (C. Fischmeister).
(
0
920-5861/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2010.04.024

A. Keraani et al. / Catalysis Today 156 (2010) 268–275
269
gradient across a selective and dense membrane, offering a
2. Experimental
.1. Reagents and solvents
Diallyltosylamide (DATA) was synthesised by a standard organic
procedure using diallyltosylamine, allyl bromide and sodium
hydride in DMF. Grubbs 2nd generation catalyst (Benzylidene
sustainable and viable alternative for traditional separation tech-
niques. NF processes are widely developed at industrial scale in
water treatment and in food industry [34] while separations in
organic solvents (OSNF) have recently started to be exploited
2
[
34–36].
The integration of OSNF within the homogeneous catalytic pro-
[
(
1,3-bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro
tricyclohexylphosphine)ruthenium was purchased from Aldrich.
cess of olefin metathesis represents an innovative and soft route
since this separation process does not affect the reaction prod-
ucts that permeate through the membrane (no phase change).
OSNF was already used to recycle homogeneous catalysts at lab-
oratory scale [36–44] and the relevance of this technique has
yet been highlighted by the recycling of the Co-Jacobsen’s cata-
lyst (4 cycles) [45] or the NF-coupled Heck (5 cycles) and Suzuki
Toluene was freshly distilled over Na before use. CH Cl2 was
2
freshly distilled from CaH2.
2.2. Synthesis of the ionic liquid
(
10 cycles) reactions [46,47]. The OSNF of metathesis catalysts
has been reported only a few times: van Koten and co-workers
48] used a dendrimer-supported metathesis catalyst in a NF reac-
The ionic liquid 1-butyl-3-methylimidazolium hexafluorophos-
phate [bmim][PF ] (Fig. 1) was prepared according to a reported
6
[
procedure [56].
tor but a poor conversion (20%) was obtained. In 2006, Dow
Chemicals patented the principle of NF separation for metathesis
reactions [49] but only two questionable results were described.
Firstly, one of the complexes used for the demonstration was
not a metathesis catalyst and secondly, the proposed commercial
membrane materials is not the one reported by the manufac-
turer and should not be stable in the solvent used, nevertheless
the patent claimed for 87% rejection for the Grubbs’s 1st gen-
eration catalyst. Using a commercial polyimide (PI) membrane
This ionic liquid was selected due to its low miscibility with
toluene (solvent of RCM reaction) and the non coordinating PF6-
anion makes this ionic liquid fully compatible with ruthenium-
catalysed olefin metathesis transformations [12–16].
The ionic liquid structure was checked by H and 31P NMR spec-
troscopy.
1
1
◦
H NMR (200.12 MHz, CDCl , 25 C) ı ppm: 0.91 (t, J = 7.45 Hz,
3
3
3H); 1.39 (m, 2H); 1.89 (m, 2H); 4.15 (s, 3H); 4.31 (m, 2H); 7.65 (d,
J = 1.15, 1H); 7.72 (d, J = 1.15, 1H); 8.93 (s, 1H).
3
3
31
(
Starmem 228, MET) allowing insufficient rejection of Hoveyda cat-
P
NMR (81.019 MHz, (CD ) CO)
ı
ppm: −142.8 (sept,
3
2
−1
19
alyst (627 g mol , 64% recovery), we proposed in 2008 the coupling
of OSNF as green separation process and the use of tailor-made
homogeneous Hoveyda-type catalysts with enhanced NF recovery
JP–F = 705 Hz).
JF–P = 705 Hz).
F NMR (188 MHz, (CD ) CO) ı ppm: −72.6 (d,
3
2
(
up to 92%), both in toluene and dimethyl carbonate, respectively
2
.3. Synthesis of a new ionically tagged catalyst 2
known as a demanding solvent and an environmentally friendly
one and allowing at least 5 cycles [50]. More recently, Plenio
and co-workers also reported modified olefin metathesis cata-
lysts displaying very good properties in nanofiltration processes
A new tailor-made Hoveyda catalyst has been developed in
order to achieve an efficient immobilisation of this complex in ionic
liquids. Thus an ionically tagged ligand 1 derived from the original
Hoveyda ligand was prepared and the modified complex 2 prepared
in 75% yield according to reported procedure [11,15] (Fig. 2). In a
Schlenck tube maintained under an argon atmosphere, Grubbs 2nd
generation catalyst (50 mg, 0.06 mmol, 1.25 equiv.) and CuI (5.6 mg,
[
51].
If OSNF can be settled as a separate process, it can also
be coupled to a reaction batch to lead to a membrane reac-
tor or a catalytic membrane reactor (CMR) using an inert or an
active membrane, respectively [52]. The membrane appearing
clearly as a key point for further application, the preparation of
innovative membranes is a real breakthrough. An ideal inert or
active membrane must exhibit chemical, thermal and mechanical
stability together with good performances, e.g. high selectiv-
ity (rejection, both based on size and affinity of the solute
toward the membrane) and high permeability needed for pro-
ductivity purpose. Coupling advantages based on IL route and
NF appeared then as an interesting an innovative approach.
Immobilisation of ILs has been already achieved in pores of micro-
porous membranes commonly for microfiltration/ultrafiltration
0
.056 mmol, 1.2 equiv) were dissolved in 4 mL of dry and degassed
CH Cl . Ligand 1 (22 mg, 0.047 mmol, 1.0 equiv.) dissolved in 4 mL
of dry CH Cl2 was then added and the mixture was stirred at 40 C
for 12 h. The solvent was evaporated and the crude residue was dis-
solved in a mixture of pentane and CH Cl (1:1 (mL)). The unsoluble
2
2
◦
2
2
2
phosphine–copper adduct was filtered off. The filtrate was con-
centrated to dryness. 1 mL of toluene was added to dissolve the
excess of Grubbs catalyst which was removed using a filtrating
cannula. The new complex 2 was obtained as a greenish powder
32 mg, 75% yield). 2 was characterised by 1H, F, P NMR and
19
31
(
HRMS.
(
supported ionic liquid membrane, SILM) useful for gas separa-
tion [53], recovery of enzyme [54] and amino acid extraction
55].
This study presents our preliminary results in achieving, accord-
1
◦
H NMR (200.12 MHz, CDCl , 25 C) ı ppm: 1.25 (d, J = 6.10 Hz,
3
3
6
H, CH ), 2.21 (m, 2H, CH ), 2.46 (m, 21H, CH ), 3.51 (m, 5H,
3 2 3
[
CH3 and CH ), 4.21 (m, 5H, CH3 and CH ), 4.50 (s, 2H, CH ), 4.91
2
2
2
(sept, J = 6.15 Hz, 1H, CH), 6.90–7.05 (m, 4H, CH), 7.21 (bs, 4H,
3
ing to the best of our knowledge, the first CMR for olefin metathesis.
Inspired from SILM technology together with the well-known
approach for membrane surface modification based on polyelec-
trolyte adsorption (with or without cross-linking), we propose here
the modification of a PI dense membrane (Starmem 228®) by
31
CH), 7.62 (d, 2H, CH), 16.41 (s, 1H, CH). P NMR (81.019 MHz,
1
9
(
(
CD ) CO) ı ppm: −143.22 (sept, JP–F = 712 Hz). F NMR (188 MHz,
3
2
+
CD ) CO) ı ppm: −73.1 (d, JF–P = 712 Hz). MS (ESI): m/z: [M + H]
3
2
calculated for C40
H N O Cl Ru: 793.25891; found: 793.2582
53 4 2 2
(1 ppm).
IL ([bmim][PF ]) adsorption thus compatibilizing the membrane
6
surface with a tailor-made IL tagged Hoveyda catalyst hoping to
favour its anchorage by enhancement of polar and apolar interac-
tions between the two ILs’ parts. The resulting catalytic membrane
is subsequently used in a CMR running in a discontinuous mode
◦
in toluene at low temperature (35 C) with a model ring closing
metathesis reaction.
Fig. 1. Ionic liquid [bmim][PF6].

2
70
A. Keraani et al. / Catalysis Today 156 (2010) 268–275
Fig. 2. Hoveyda catalyst and synthesis of the new tailor-made catalyst 2 from grubbs 2nd generation catalyst.
2.4. Modification of the NF polyimide membrane
the ionic liquid. This membrane was further called “the catalytic
membrane”.
2
Three identical NF membrane coupons (Starmem 228®, 44 cm ,
−1
MWCO = 280 g mol
according to the manufacturer, polyimide,
2.5. Membrane material characterisations
MET, UK) were preconditioned by NF of toluene at 55 bar (maxi-
mum pressure according to the manufacturer) until a plateau value
of flux was reached. One membrane was kept as a reference and the
two other membranes were subjected to further modifications per-
formed in two steps. First the active layer side of the membranes
was impregnated by contact with a sufficient amount of IL in order
to cover the entire membrane surface during two days. The excess
of IL was removed by adsorption on a soft tissue. The obtained IL
Membrane material characterisations were achieved on
the supported ionic liquid membrane using attenuated total
reflectance Fourier transform infrared (ATR-FTIR), SEM-EDX micro-
analysis evidencing the presence of new element in order to check
the IL adsorption and SEM to see the impact of the adsorbed IL on
the membrane morphology. Adsorption of the ionically tagged cat-
alyst on the supported IL membrane was indirectly checked thanks
to the occurrence of the catalytic reaction (see below).
membranes were further called Starmem 228 [bmim][PF ]. One
6
membrane (no 1) was mounted in the filtration cell for measure-
ment of the evolution of its permeability to toluene owing to the
presence of IL (see below) and the second membrane (no 2) was
mounted in the filtration cell and poured with N2 gas in toluene
FTIR-ATR spectra were recorded on a PerkinElmer spectrome-
ter (Paragon 1000, spectrum for windows software) with an ATR
accessory equipped with a ZnSe crystal (incidence angle: 45 , 12
◦
reflections) totally covered by the membrane coupon after regis-
−
1
(
5 bar, t = 15 min) and then used for analysis purpose. Only the
tration of the background in the air (1 cm resolution, 20 scans).
The evidence of the IL adsorption was shown by comparison to
the virgin polyimide membrane, for which specific PI bands were
membrane no 1 was submitted to the second modification step
as follows. The new complex 2 (9.5 mg, 0.01 mmol) was suspended
in 9 mL of toluene and loaded into the filtration cell. A TMP of 3 bar
was applied during 30 min at room temperature in order to transfer
the catalyst from the toluene phase to the supported ionic liquid
membrane. It was expected that 2 was quantitatively dissolved into
−
1
−1
located at 1777 and 1715 cm for C O and 1362 cm for CN.
Pictures from SEM allowed investigating the morphological
evolution of the membrane surface after deposition of a thin
metal layer of Au–Pd on the dried membrane sample (JEOL-

A. Keraani et al. / Catalysis Today 156 (2010) 268–275
271
Scheme 1. RCM reaction of DATA.
Fig. 3. NF membrane reactor.
DATA conversion in c-DATA was determined by gas chro-
matography using an apolar stationary phase made of PDMS-
PDPS (95-5) in a semi-capillary column (Equity TM-5, length:
JSM 6400, Oxoford instruments). Moreover, occurrence of both P
and F element contained in the counter-anion of the ionic liq-
uid was checked using the EDX microanalysis coupled to the SEM
3
0 m × 0.53 mm × 1.5 ␮m film thickness, provided by Supelco).
(
SemQuant software).
◦
Injection at 250 C was followed by analysis at constant temper-
ature (230 C) and the detection was obtained by a FID detector
For all these characterisations, membranes were dried under
dynamic vacuum in a dessicator for at least one day.
◦
◦
(250 C). N2 (18 kPa) was used as the carrier gas. Relative Standard
Deviation on DATA and c-DATA concentrations was less than 1%
thus accuracy on conversion was better than 2%.
2.6. NF and membrane reactor
The stirred NF dead-end filtration cell of 300 mL volume was
used for permeability measurement (Fig. 3) in which a 44 cm2
membrane, either virgin or modified one, was inserted. The trans-
membrane pressure (TMP) was applied with pressurised nitrogen.
3. Results and discussions
3.1. Analyses of the supported ionic liquid membrane
◦
Standard conditions for NF were: TMP = 15 bar, rpm = 230, T = 35 C.
Permeability (Lp) of membranes to toluene, whatever their form,
was measured in the NF membrane reactor in standard conditions.
The same cell was used as membrane reactor, in which the
reagents were mixed under a nitrogen atmosphere in the pres-
ence of the catalytic membrane inserted in the cell before applying
pressure to recover the products in the permeate, according to a
discontinuous process that must be further optimised.
3.1.1. Membrane material analyses
The supported IL membranes were analysed by SEM-EDX. The
presence of the IL on the impregnated membrane was evidenced by
the detection of phosphorus and fluorine arising from the counter-
anion of the IL (not shown). The SEM pictures showing the different
morphologies between the virgin and the impregnated membrane
surfaces are presented in Fig. 4, evidencing a layer covering in an
irregular way the whole membrane surface.
2.7. Ring closing metathesis reaction
More sophisticated analysis was obtained from FTIR-ATR
(Figs. 5–7) allowing evidencing of both the cationic and anionic
The activity of the catalytic membrane was evaluated using
part of the IL. After IL adsorption new bands appeared located
at 3165 and 3118, 1574, 827 and 843, 805 and 783 emerging
from local minima on the virgin membrane spectrum. To evidence
the real position of the new bands, difference must be calcu-
lated between the two spectra, but as the penetration depth of
the IR beam in the PI membrane varied with the thickness of
the IL deposit, a correcting factor must be found as we already
explained in a previous article to evidence thin deposit on poly-
meric membranes [57]. Considering that bands located at 1714,
the ring closing metathesis reaction (RCM) of DATA (Scheme 1)
directly in the CMR described above. Standard conditions for the
RCM reaction used in the membrane reactor were: 100 mg of DATA
(
0.4 mmol) in 9 mL of toluene in the presence of 0.01 mmol (2.5
mol%) of catalyst assuming quantitative immobilisation of the cat-
◦
alyst in the IL. T = 35 C, rpm = 230, 1 h (non optimised reaction time)
in standard conditions. The cell was fully drained by applying TMP
after each RCM reaction before starting the next one, in order to
avoid the presence of c-DATA that could result in catalyst inhibition
−
1
1362 and 1089 cm only correspond to the virgin PI membrane,
it is possible to substrate the virgin membrane spectra from the
[
50].
Fig. 4. SEM pictures of the active layer of the virgin Starmem 228 (a) and Starmem after adsorption of [bmim][PF6](b).

2
72
A. Keraani et al. / Catalysis Today 156 (2010) 268–275
Fig. 5. FTIR-ATR raw spectra of virgin polyimide membrane and supported IL membrane region: 3400–2500 cm−1
.
supported IL one applying a 0.385 coefficient (supported IL– 0.385
virgin) for which these three bands are simultaneously cancelled.
3.1.2. Permeability to toluene
The permeability to toluene of the reference membrane and
−1
Thus a new band located at 834 cm
emerged from the large
the Starmem 228 [bmim][PF ] membrane (no 1) were measured at
6
−1
◦
peak located at 827 cm on the supported IL membrane (Fig. 8).
Assignation of the new bands could be as follows. 3165, 3118
35 C according to a standard procedure (Table 1). Applying 55 bar
pressure as a preconditioning procedure increased significantly the
membrane flux, probably thanks to both the removal of preserva-
tives and to the compaction of the active layer as it is quite classical
to observe in NF and UF in aqueous solution. This enhancement
(about 3 times appeared irreversible over many experiments not
shown here). In spite of the presence of the IL deposit on the mem-
−1
and 1574 cm
confirmed by comparison with a similar IL after exchange of the
counter-anion group, not shown here), the two first band corre-
bands correspond to the cationic part of the IL
(
−1
−
sponding to C–H vibrations. 834 cm is a specific band of the PF6
anion.
Fig. 6. FTIR-ATR raw spectra of virgin polyimide membrane and supported IL membrane region: 1847–1398 cm−1.

A. Keraani et al. / Catalysis Today 156 (2010) 268–275
273
Fig. 7. FTIR-ATR raw spectra of virgin polyimide membrane and supported IL membrane region: 1274–549 cm−1.
Fig. 8. FTIR-ATR difference spectra corresponding to supported Il membrane spectra −0.385 × virgin membrane spectra (zoom in the region 1064–647 cm−1 region).
brane top layer, the permeability increased significantly for the
supported ionic liquid membrane (about 2 times). Such behaviour
can be explained by a modification of the polar/apolar balance of
the membrane due to the IL presence, leading to a higher affin-
ity of toluene toward the modified membrane than the virgin one,
the consequence of which is a decrease in the whole membrane
resistance toward the solvent transfer.
Such behaviour was already observed in OSNF for membrane
coated with polymer owing to polyelectrolyte properties (poly-
cations and polyanions) or even with more simple ions issued
from salts as sodium chloride adsorbed on the PI membrane
[58]. The explanation based on the variation of the polar/apolar
balance of the membrane was already validated in ultrafiltra-
tion for water application for polyethersulfone (PES) membrane
after hydrophilic surfactant adsorption [59]. Of course in the
case of NF in solvent media, as dense membranes are used, the
transfer mechanisms are totally differents both for solvent and
solutes. Nevertheless, the hypothesis on flux variation due to
polar/apolar balance of membrane is consistent as convection
is very strongly lowered (down to zero, sometimes) with NF
dense membranes, increasing in a concomitant way the effect of
Table 1
Permeability of membranes to toluene.
−
−1
−1
bar )
Membrane
Permeability (L m 2
h
Starmem 228 (reference)
Starmem 228 [bmim][PF6] no 1
0.35
0.67

2
74
A. Keraani et al. / Catalysis Today 156 (2010) 268–275
Table 2
sequence, consisting in the dissolution of the catalyst in the ionic
liquid before the impregnation step, will also be envisaged.
RCM of DATA in the membrane reactor.
Cycle
Conversion (%)
Time (h)
1
2
3
99
98
52
1
1
1
2.5
Acknowledgement
The authors acknowledge the French Ministry of Research for a
Ph.D. grant given to A.K.
98
References
physico-chemical interactions in the membrane/solute/solvent
system.
[
[
1] Y. Chauvin, Angew. Chem. Int. Ed. 45 (2006) 3740.
2] R.R. Schrock, Angew. Chem. Int. Ed. 45 (2006) 3748.
3
.2. Metathesis reaction in catalytic membrane reactor
[3] R.H. Grubbs, Angew. Chem. Int. Ed. 45 (2006) 3760.
[
[
4] R.H. Grubbs (Ed.), Handbook of Metathesis, vol. 1–3, Wiley-VCH, Weinheim,
003.
5] S.J. Connon, S. Blechert, in: P.H. Dixneuf, C. Bruneau (Eds.), Ruthenium Catalysts
and Fine Chemistry, vol. 11, Springer, 2004, p. 93.
2
The ionically tagged Hoveyda catalyst immobilized in
2
[
bmim][PF ] dispersed on the Starmem membrane (catalytic mem-
6
◦
[6] A. Fürstner, Angew. Chem. Int. Ed. Engl. 39 (2000) 3012.
brane) was used in the catalytic reactor at 35 C for the RCM of DATA
[
[
7] A.H. Hoveyda, A.R. Zhugralin, Nature 450 (2007) 243.
8] P.H. Deshmuhk, S. Blechert, Dalton Trans. (2007) 2479.
(
Scheme 1). As depicted in Table 2, the first run proceeded quanti-
tatively in 1 h (not optimised reaction time). The reactor was then
drained and reloaded with the same amount of DATA resulting in
a similar result i.e. nearly quantitative conversion in 1 h. The pro-
cedure was repeated (cylcle 3) but the catalytic activity decreased
and 2.5 h was necessary to ensure a high conversion (Table 2).
The decrease in activity of the catalytic membrane (Table 2)
may have different origins ranging from the intrinsic stability of
the catalyst to that of the catalytic membrane itself over time.
Further experiments are then needed to determine the lifetime
of complex 2 under RCM conditions, conditioning of course the fea-
sibility of long term recycling. One must notice that such kinetic
studies are often lacking whatever the metathesis catalyst used.
On the other hand, the membrane stability must also be stud-
ied more extensively. Among possible problems are the transfer
of the catalyst in the permeate leading to a loss not linked to nat-
ural deactivation. This leaching could also be due to the transfer
mechanism itself as a too high affinity of the catalyst toward the
membrane surface can sometimes lead to an enhanced transfer
across the membrane in OSNF conditions not only based on size
exclusion but also on physico-chemical interactions. The other bot-
tleneck concerns the stability of the IL coating itself thanks to IL
weak interactions maintaining it on the PI membrane. Then as IL is
very slightly soluble in toluene it could progressively be dissolved
in the solvent and consequently few leaching can occur in permeate
when TMP is applied during OSNF.
[9] J.P.A. Harrity, D.S. La, D.R. Cefalo, M.S. Visser, A.H. Hoveyda, J. Am. Chem. Soc.
20 (1998) 2343.
10] J.S. Kingsbury, J.P.A. Harrity, P.J. Bonitatebus, A.H. Hoveyda, J. Am. Chem. Soc.
21 (1999) 791.
[11] S.B. Garber, J.S. Kingsbury, B.L. Gray, A.H. Hoveyda, J. Am. Chem. Soc. 122 (2000)
168.
12] N. Audic, H. Clavier, M. Mauduit, J.C. Guillemin, J. Am. Chem. Soc. 125 (2003)
248.
1
[
1
8
[
9
[13] H. Clavier, N. Audic, M. Mauduit, J.C. Guillemin, Chem. Commun. (2004) 2282.
[14] C. Thurier, C. Fischmeister, C. Bruneau, H. Olivier-Bourbigou, P.H. Dixneuf,
ChemSusChem 1 (2008) 118.
[
15] C. Thurier, C. Fischmeister, C. Bruneau, H. Olivier-Bourbigou, P.H. Dixneuf, J.
Mol. Catal. A: Chem. 268 (2007) 127.
[16] Q. Yao, Y. Zhang, Angew. Chem. Int. Ed. 42 (2003) 3395.
[
17] C. Fischmeister, Catalytic alkene metathesis in ionic liquids, in: Y. Imamoglu,
V. Dragutan, S. Karabulut (Eds.), Metathesis Chemistry, From Nanostructure
Design to Synthesis of Advanced Materials, Springer, Berlin, 2007, p. 483.
[18] P. Sledz, M. Mauduit, K. Grela, Chem. Soc. Rev. 37 (2008) 2433.
[
19] H.T. Wong, S. Han, A.G. Livingston, Chem. Eng. Sci. 61 (2006) 1338.
[
20] C.P. Mehnert, R.A. Cook, N.C. Dispenziere, M. Afeworki, J. Am. Chem. Soc. 124
(
2002) 12932.
[21] C.P. Mehnert, Chem. Eur. J. 11 (2005) 50.
[
[
[
[
22] A. Riisager, R. Fehrmann, M. Haumann, P. Wasserscheid, Eur. J. Inorg. Chem.
2006) 695.
23] A. Riisager, R. Fehrmann, M. Haumann, P. Wasserscheid, Top. Catal. 40 (2006)
91.
24] A. Riisager, R. Fehrmann, S. Flicker, R. Van Hal, M. Haumann, P. Wasserscheid,
Angew. Chem. Int. Ed. 44 (2005) 815.
25] M.H. Valkenberg, C. De Castro, W.F. Holderich, Green Chem. 4 (2002) 88.
(
[26] M.J. Jin, A. Taher, H.J. Kang, M. Choi, R. Ryoo, Green Chem. 11 (2009) 309.
[27] H. Hagiwara, K.H. Ko, T. Hoshi, T. Suzuki, Chem. Commun. (2007) 2838.
[
28] J. Baudoux, K. Perrigaud, P.J. Madec, A.C. Gaumont, I. Dez, Green Chem. 9 (2007)
346.
29] H. Hagiwara, Y. Sugawara, K. Isobe, T. Hoshi, T. Suzuki, Org. Lett. 6 (2004) 2325.
1
At first sight, few strategies could be followed in the future to
avoid such problems, if they are confirmed. On one hand graft-
ing of IL on the PI membrane may prevent from IL leaching. On
the other hand, a whole entrapping of both IL and catalyst in the
polymer structure, achieved during the membrane fabrication pro-
cedure might retain both in the membrane matrix. Nevertheless,
this last proposal would be probably limited thanks to the need in
accessibility of substrate to the catalyst during the metathesis and
probably applications would be restrained to small size substrates.
[
[30] A. Wolfson, I.F. Vankelekom, P.A. Jacobs, Tetrahedron Lett. 44 (2003) 1195.
[
[
31] H. Hagiwara, H. Sasaki, T. Hoshi, T. Suzuki, Synlett 4 (2009) 643.
32] L. Rodriguez-Pérez, E. Teuma, A. Falqui, M. Gómez, P. Serp, Chem. Commun.
(
2008) 4201.
[33] Department of Energy–FY. http://www.zyn.com/sbir/sbres/sbir/doe/doe09-
2.html 2009.
1
[
34] A.I. Schafer, A.G. Fane, T.D. Waite (Eds.), Nanofiltration-Principles and Applica-
tions, Elsevier, Amsterdam, The Netherlands, 2003.
[35] X. Li, P. Vandezande, I.F.J. Vankelecom, J. Membr. Sci. 320 (2008) 143.
[
[
36] P. Vandezande, L.E.M. Gevers, I.F.J. Vankelecom, Chem. Soc. Rev. 37 (2008) 365.
37] K. De Smet, S. Aerts, E. Ceulemans, I.F.J. Vankelecom, P.A. Jacobs, Chem. Com-
mun. (2001) 597.
[
[
38] J.T. Scarpello, D. Nair, L.M. Freitas dos Santos, L.S. White, A.G. Livingston, J.
Membr. Sci. 71 (2002) 203.
39] C. Müller, M.G. Nijkamp, D. Vogt, Eur. J. Inorg. Chem. (2005) 4011.
4
. Conclusions
This study reports the immobilisation of an ionically tagged
[40] J.T. Scarpello, D. Nair, L.M. Freitas dos Santos, L.S. White, A.G. Livingston, J.
Membr. Sci. 71 (2003) 203.
[
[
metathesis catalyst on an ionic liquid supported polyimide mem-
brane to prepare a catalytic membrane, the stability of which
must be carefully studied in the future according to several
proposal exposed above. Nevertheless, these preliminary experi-
ments proved, for the first time, the feasibility of the concept for
metathesis catalyst as the efficiency of the catalytic membrane was
demonstrated in a NF membrane reactor for a model metathesis
reaction in toluene at quite gentle temperature (35 C) in a dis-
continuous mode. Supplementary studies are ongoing to quantify
the exact amount of catalyst immobilised. Another immobilisation
41] H.P. Dijkstra, G.P.M. van Klink, G. van Koten, Acc. Chem. Res. 35 (2002) 798.
42] H.-T. Wong, Y.H. See-Toh, F. Castelo Ferreira, R. Crook, A.G. Livingston, Chem.
Commun. (2006) 2063.
[43] J.J.M. de Pater, B.-J. Deelman, C.J. Elsevier, G. van Koten, Adv. Synth. Catal. 348
2006) 1447.
[
(
44] S.A. Chavan, W. Maes, L.E.M. Gevers, J. Wahlen, I.F.J. Vankelekom, P.A. Jacobs,
N. Dehaen, D.E. de Vos, Chem. Eur. J. 11 (2005) 6754.
[45] S. Aerts, H. Weyten, A. Buekenhoudt, L.E.M. Gevers, I.F.J. Vankelecom, P.A.
Jacobs, Chem. Commun. (2004) 710.
[
◦
46] D. Nair, J.T. Scarpello, L.S. White, L.M. Freitas dos Santos, I.F.J. Vankelecom, A.G.
Livingston, Tetrahedron Lett. 42 (2001) 8219.
[47] A. Datta, K. Ebert, H. Plenio, Organometallics 22 (2003) 4685–4691.

A. Keraani et al. / Catalysis Today 156 (2010) 268–275
275
[
48] P. Wijkers, J.T.B.H. Jastzebski, P.A. van der Schaaf, R. Kolly, A. Hafner, G. van
[55] R. Fortunato, M.J. Gonzalez-Munoz, M. Kubasiewiez, S. Luque, J.R. Alvarez, C.A.
Alfonso, I.M. Coelhoso, J.G. Crespo, J. Membr. Sci 249 (2005) 153.
[56] J. Dupont, C. Consorti, P. Suarez, R. de Souza, Org. Synth. 10 (2004) 184.
[57] D. Delaunay, M. Rabiller-Baudry, J.M. Gozalvez-Zafrilla, B. Balannec, M. Frap-
part, L. Paugam, Chem. Eng. Process. 47 (2008) 1106–1117.
Koten, Org. Lett. 11 (2000) 1621.
[
49] K.A. Burdett, M.R. Collins, B.R. Maughon, M.L. Tulchinsky, WO 2006/047105 A2.
50] A. Keraani, T. Renouard, C. Fischmeister, C. Bruneau, M. Rabiller-Baudry, Chem-
SusChem 1 (2008) 927.
[
[
51] D. Schoeps, K. Buhr, M. Dijkstra, K. Ebert, H. Plenio, Chem. Eur. J. 15 (2009)
[58] I.J.F. Vankeleom, New materials and approaches for solvent resistant nanofil-
tration (SRNF), in: Keynote lecture at Euromembrane 2009 conference,
Proceeding, Euromembrane 2009, 6–10 September 2009, Montpellier, France,
2009, pp. 153–154.
[59] M. Rabiller-Baudry, L. Bégoin, D. Delaunay, L. Paugam, B. Chaufer, Chem. Eng.
Process. 47 (2008) 267.
2
960.
[
[
[
52] I. Vankelekom, Chem. Rev. 102 (2002) 3779.
53] Q. Gan, D. Rooney, M. Xue, G. Thompson, J. Membr. Sci. 280 (2006) 956.
54] M. Mori, R. Gomez Garcia, M.P. Belleville, D. Paolucci-Jeanjean, J. Sanchez, P.
Lozano, M. Vaultier, G. Rios, Catal. Today 104 (2005) 313.