First elaboration of an olefin metathesis catalytic membrane by grafting a Hoveyda–Grubbs precatalyst on zirconia membranes

Article abstract of DOI:10.1016/j.crci.2017.04.003

This study is aimed at developing an innovative concept for the preparation of an olefin metathesis catalytic membrane. This goal was achieved by grafting a Grubbs II precatalyst on a zirconia membrane generating a tailor-made Hoveyda–Grubbs precatalyst o

Full text of DOI:10.1016/j.crci.2017.04.003

C. R. Chimie xxx (2017) 1e15
Contents lists available at ScienceDirect
Comptes Rendus Chimie
www.sciencedirect.com
ꢀ
Full paper/Memoire
First elaboration of an olefin metathesis catalytic membrane
by grafting a HoveydaeGrubbs precatalyst on zirconia
membranes
,
,
*
c, **
Adel Keraani a b, Murielle Rabiller-Baudry a , Cedric Fischmeister
,
ꢀ
David Delaunay a 1, Amandine Baudry a, Christian Bruneau c,
,
Thierry Renouard a
a
ꢀ
ꢀ
ꢀ ꢀ
Universite de Rennes-1, CNRS, Universite Bretagne Loire, Institut des sciences chimiques de Rennes, 263, avenue du General-Leclerc, CS
74205, case 1011, 35042 Rennes cedex, France
b
ꢀ
ꢀ
ꢀ
ꢀ
Laboratoire de chimie analytique et electrochimie, Departement de chimie, Faculte des sciences de Tunis, Universite de Tunis El Manar,
campus universitaire de Tunis, El Manar, 2092 Tunis El Manar, Tunisia
c
ꢀ
ꢀ
ꢀ ꢀ
CNRS, Universite de Rennes-1, Universite Bretagne Loire, Institut des sciences chimiques de Rennes, 263, avenue du General-Leclerc, CS
74205, 35042 Rennes cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
This study is aimed at developing an innovative concept for the preparation of an olefin
metathesis catalytic membrane. This goal was achieved by grafting a Grubbs II precatalyst
on a zirconia membrane generating a tailor-made HoveydaeGrubbs precatalyst on the
surface of an ultrafiltration inorganic membrane. The obtained membranes were charac-
terized by different techniques evidencing the grafting. The catalytic activity was also
evaluated in a model ring-closing metathesis reaction. According to our knowledge, this
manuscript reports the first ever so sophisticated catalytic membrane based on an
organometallic complex of ruthenium. The proof of concept was achieved, although the
membrane grafting must be improved for better catalytic activity.
Received 30 January 2017
Accepted 25 April 2017
Available online xxxx
Keywords:
Catalytic membrane
HoveydaeGrubbs II precatalyst
Olefin metathesis
Membrane modification
Supported catalysis
OSN in toluene
© 2017 Académie des sciences. Published by Elsevier Masson SAS. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
metathesis development and achievements shared the
Nobel Prize in Chemistry in 2005 [3e5]. For examples of
Olefin metathesis reactions represent a powerful cata-
lytic tool for the formation of carbonecarbon double bonds
allowing the preparation of various chemical intermediates
useful for the synthesis of fine chemicals, pharmaceuticals,
and polymers [1,2]. The three major contributors to olefin
industrial applications of olefin metathesis technology, the
reader is directed to a review by Fogg et al. [6]. The suc-
cessful progress in olefin metathesis stems from the
transformation of unsaturated renewable resources such as
fat and oils [7], terpenes [8], and Z-selective trans-
formations [9]. Such reactions are catalyzed by organo-
metallic soluble catalysts of ruthenium, tungsten or
molybdenum, which are all quite expensive compounds
[10,11]. Among olefin metathesis catalysts, commercially
available ruthenium-based Grubbs and HoveydaeGrubbs
precatalysts (Fig. 1) have received a lot of attention at a
laboratory scale as they combine high stability and activity
with an excellent tolerance toward polar functional groups,
* Corresponding author.
** Corresponding author.
E-mail addresses: murielle.rabiller-baudry@univ-rennes1.fr (M. Rabiller-
Baudry), cedric.fischmeister@univ-rennes1.fr (C. Fischmeister).
1
ꢀ
ꢀ
Present address: Universite de Caen Basse-Normandie, Unite de
ꢀ
ꢀ
recherche Aliments Bioprocedes Toxicologie Environnements (UR ABTE
EA No. 4651), IFR 146 ICORE, France.
http://dx.doi.org/10.1016/j.crci.2017.04.003
1631-0748/© 2017 Académie des sciences. Published by Elsevier Masson SAS. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

2
A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
hence allowing reactions with various substrates/reactants.
It is important to note that the actual catalytic species is
generated from precursors such as Grubbs and Hoveyda
eGrubbs complexes (herein called precatalysts) by in situ
activation in the presence of the substrate. Despite the
broad range of applications accessible by olefin metathesis,
only few industrial processes use metathesis reactions
because of the cost of homogeneous catalysts together with
difficulties for the final separation of the reaction products
from the various metal-containing species at the end of the
reaction. Depending on the target application the residual
amount of metal must be very low in the final product.
Therefore, several procedures have been developed to
reduce this metal content of which precatalyst immobili-
zation has to be considered as a preferred strategy if the
productivity of the reaction can be ensured [12e14].
Furthermore, precatalyst immobilization is also a very
valuable strategy to increase the catalyst productivity
quantified through its turnover number (TON¼moles of
reactant converted per mole of catalyst) obtained for a
less fluidized bed, avoiding the subsequent separation step
by filtration. However, such process suffers from pressure
drop, which must be carefully controlled and managed.
Another possible strategy is based on the mastering of
the separation itself without any modification of the pre-
catalyst. Following this idea, in the recent years, membrane
technology has appeared as a promising and environmen-
tally friendly alternative approach for process intensifica-
tion in organic (fine) chemistry as it allows, in principle, the
separation and reuse of homogeneous catalysts [28,29].
This idea has been developed especially with organic sol-
vent nanofiltration (OSN) after the development of suitable
polymeric membranes such as Koch (PDMS), Starmem (PI),
and DuraMem [30]. Many research groups have investi-
gated OSN for separation and reuse of homogeneous cata-
lysts using such “inert” membranes [31e33]. In the case of
olefin metathesis reactions, previous works from our group
[34e36] and others [37e41] implemented OSN for the re-
covery of HoveydaeGrubbs precatalysts as the starting
point to increase the apparent turnover number in selected
model reactions. Nevertheless, to the best of our knowledge
the transfer from laboratory to industrial scale has not been
achieved as technical improvements are still required. We
previously proposed a comparison of several systems
applied to olefin metathesis based either on simple sepa-
rations by OSN or on the use of a synthesis membrane
reactor running in continuous or discontinuous mode [35].
What about the coupling of the two strategies described
just above? In others words, what about the immobiliza-
tion of the catalyst on a porous membrane support, inte-
grating synthesis, and separation in a single operation?
Such a strategy would increase the process intensification
by proposing a real continuous process in a catalytic
membrane reactor reducing problems related to pressure
drop management encountered with more or less packed
bed.
single load of the precatalyst and
a sufficient load
(continuous or discontinuous) of the substrate. Precatalyst
recovery and subsequent reuse, a process quoted as “cata-
lyst recycling” has been addressed via heterogenization of
the HoveydaeGrubbs precatalysts by immobilization in
nonconventional solvents [15e18] or on an insoluble
porous solid material, generally silica, the particles of
which are being easily separated by classical filtration of
the reaction medium. This immobilization can be achieved
either by a covalent grafting [19e23], by adsorption owing
to electrostatic attractive interactions leading to an ionic
bond [24,25] or by simple adsorption [26,27].
Among these different immobilization routes, grafting
through covalent bonding seems advantageous for the
design of stable and recoverable immobilized Hoveyda
eGrubbs precatalysts. This route should prevent leaching
problems allowing for the possible adaptation to scalable
industrial processes. Grafting through covalent bonding is
generally carried out by a reaction of hydroxyl groups
(ꢀOH) of porous oxide particles with an organosilane linker
of the general formula (R0O)3SiR.
In this respect, modified polymeric membranes have
been used as catalytic membranes [28,42]. The hetero-
genization of a homogeneous catalyst in polymeric mem-
branes can be achieved by occlusion, entrapment or
encapsulation of the homogeneous catalyst inside or on the
membrane surface [28,43].
Using such approach, the global catalytic process can be
run in a semicontinuous mode with oxide particles in
suspension in a batch reactor and sequential filtration to
separate beads from the solution. It can also be achieved in
a continuous mode by packing the particles in a more or
For instance, Vankelecom et al. [44] studied the
encapsulation of the Ru-MeDuPhos (MeDuPhos ¼ (ꢀ)-1,2-
Bis[(2R,5R)-2,5-dimethylphospholano]benzene)
catalyst
on a PDMS-based membrane (PDMS ¼ PolyDiMethyl
Fig. 1. Example of Grubbs II and HoveydaeGrubbs precatalysts (Cy ¼ cyclohexyl).
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
3
Siloxane) for the hydrogenation of methyl acet-
amidoacrylate. Mac Leod et al. [45] studied the encapsu-
lation of the Jacobsen catalyst in a PDMS-based membrane
and used it for oxidation of alkanes and alkenes. Our group
studied the elaboration of a metathesis catalytic membrane
by immobilization of an ionically tagged HoveydaeGrubbs
precatalyst on a polyimide membrane previously modified
by adsorption of an ionic liquid [46]. Drioli et al. [47]
embedded Salen-Co(II) complexes into polymeric mem-
branes and used them for the cyclopropanation reaction in
a catalytic reactor.
the development of metathesis catalytic membrane re-
actors, we studied three fields of research: catalysis,
modification of ceramic membrane, and reactor engineer-
ing. This work reports the results of the chemical grafting of
tailor-made HoveydaeGrubbs precatalysts onto the surface
of a zirconia UF membrane with the aim of elaborating
olefin metathesis catalytic membrane reactors.
2. Strategy for membrane grafting
2.1. General approach
However, such catalytic membranes suffer from catalyst
leaching issues resulting from the noncovalent linkage
between the organometallic catalyst and the polymer
membrane. To overcome the leaching problem and design a
stable catalytic membrane, a similar approach as the
grafting on silica particles [48], that is, by a covalent link
between an organometallic catalyst and a ceramic mem-
brane, seems advantageous. Furthermore, higher chemical
and mechanical stabilities will be brought by these inor-
ganic membranes as compared to organic ones.
In the literature, heterogeneous catalysts (metal, cata-
lyst particles) have been widely used to elaborate catalytic
membranes [49,50]. Generally, immobilization of hetero-
geneous catalysts in the active layer of a ceramic membrane
was realized by an impregnation process or by incorpora-
tion of catalyst particles during the preparation of the
active layer. The presence of hydroxyl groups on the surface
of inorganic membranes allows modification and compa-
tibilization of the membrane for a wide variety of applica-
tions in diverse fields such as OSN [51,52], separation of gas
[53], and membrane distillation [54].
Among the different techniques for the modification of
inorganic membrane surfaces, chemical grafting with
organosilanes is the most frequently used method [52]. The
chemical grafting of zirconia membranes has been reported
by many researchers for different applications [55]. For
instance, in 2004 Picard et al. [56,57] reduced the hydro-
philicity of the surface of zirconia ultrafiltration (UF)
membranes by grafting of perfluoroalkylsilanes.
In 1999, our group [58] patented the modification of a
zirconia membrane by grafting of organotitanate-bearing
pyrophosphate, alkyl-substituted pyrophosphate, or
amine groups to improve protein rejection and prevent
fouling in UF and also to prepare membranes for the che-
lation of metal cations. Using the same modification
methodology, our group studied the modification of a zir-
conia membrane aiming at recovering the commercial
HoveydaeGrubbs precatalyst by nanofiltration [59].
To the best of our knowledge only one study has re-
ported the elaboration of a catalytic membrane by grafting
a homogeneous catalyst on the surface of a ceramic
membrane [60]. In this article, Liu et al. reported the
grafting of Salen-Mn(III) onto an APTES-modified ceramic
membrane (APTES ¼ aminopropyltriethoxysilane) and its
performances in epoxidation of styrene.
To graft a HoveydaeGrubbs precatalyst onto the surface
of a zirconia UF membrane, we have adapted the procedure
that we have already used to graft silica and zirconia
powders [61]. This procedure consisted in the direct
grafting of the silylated styrene linker 3 onto a zirconia
powder (Fig. 2a) allowing the creation of a covalent bond
between the linker 3 and zirconia through a SieOeZr
bond. Here, we performed the grafting directly on the zir-
conia active layer of a membrane (Fig. 2b) such as the M5-
Carbosep membrane (Fig. 2c). Subsequent grafting of the
organometallic species was achieved by a reaction of the
grafted linker 3 and Grubbs II precatalyst (Fig. 3a) leading
to the final catalytic membrane (Fig. 3b).
2.2. Two approaches to graft linker 3
From a practical point of view, the linker 3 was syn-
thesized from 1 and the silane 2 as depicted in Fig. 4.
Consequently, two different procedures can be proposed
for the membrane preparation depending on the way
chosen to graft the linker 3, either in one or two steps.
In the one-step procedure, the direct grafting of linker 3
is achieved on the membrane according to the same pro-
tocol as those already used to graft silica or zirconia powder
[61].
In the two-step procedure, grafting of linker 3 was
obtained by first grafting the silane 2 as shown in Fig. 5. The
modified surface was then treated with 1 in the presence of
sodium hydride (NaH) (Fig. 6).
3. Experimental section
3.1. Reactants
All reactions were carried out under a dry argon atmo-
sphere using standard Schlenk tube techniques. Toluene
was freshly distilled over Na and THF over Na/benzophe-
none before use. DMF and dichloromethane were freshly
distilled over CaH2 before use. Grubbs second-generation
catalyst (further denoted Grubbs II, Fig. 1) was purchased
from SigmaeAldrich and stored under argon. 2,5-
Dihydroxybenzaldehyde was purchased from SigmaeAl-
drich and was used as received. (3-Iodopropyl)trimethox-
ysilane 2 was purchased from SigmaeAldrich and was used
as received. All other reagents were purchased from com-
mercial sources and were used as received. Dia-
The main objective of the present study was to develop
an innovative strategy bridging the concepts of (1) heter-
ogenization of homogeneous catalysts by grafting on a
porous material and (2) chemical grafting modification of
ceramic membranes to elaborate catalytic membranes. For
llyltosylamide (DATA) was prepared according to
reported procedure [62].
a
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

4
A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
Fig. 2. (a) Grafting procedure of zirconia surface, (b) grafting on the active layer of a tubular membrane, and (c) picture of a zirconia M5-Carbosep pristine
membrane.
Fig. 3. (a) Grafting procedure of the Grubbs II catalyst on the silylated zirconia surface and (b) obtained catalytic tubular membrane.
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
5
Fig. 4. Preparation of silylated styrene linker 3.
3.3. Zirconia powder and membranes
3.3.1. Zirconia powder
ZrO2 powder (P354, ca. 1
mm average size) was prepared
from the same reactants as those used for the preparation
of the active layer of the tubular membrane. This mem-
brane, which is no longer produced, was kindly provided by
Tech-Sep (now CTI Salindres, Saint-Maurice-de-Beynost,
France) and used as received. The zirconia crystallographic
lattice was cubic owing to stabilization provided by addi-
tives (confidential provider's data). The specific surface
Fig. 5. Grafting procedure of silane 2 on the zirconia membrane.
area of this powder was 64 m2 gꢀ1
.
Nevertheless, the specific surface area of the active layer
of the UF membrane was certainly a little lower than that of
the powder because of some differences in their respective
thermal treatments (confidential provider's data).
3.2. Synthesis of intermediate compounds
3.2.1. Synthesis of 1
The synthesis of 1 was achieved according to the
scheme depicted in Fig. 7. Compound 1 was obtained in
four steps starting from commercially available 2,5-
dihydroxybenzaldehyde. Compound 1 was isolated in 85%
yield and was easily purified as determined from 1H NMR
analysis (data not shown).
3.3.2. Tubular inorganic membranes
Tubular zirconia membranes, M5-Carbosep, were kindly
provided by Tech-Sep (now CTI) but are no longer
commercialized. These UF membranes have an active layer
made of zirconia on a carbon support and an Molecular
Weight Cut-Off (MWCO) of 10 kg molꢀ1 corresponding to a
pore diameter of ca. 6 nm (Fig. 2c). The external and in-
ternal diameters of this tubular membrane were 10 and
6 mm, respectively. The membranes used for the catalytic
membrane preparation according to the dynamic protocol
(vide infra) were of 200 mm length, whereas those for the
immersion procedure were close to 25 mm length.
3.2.2. Synthesis of linker 3
The synthesis of the silylated styrene linker 3 from
Compound 1 was achieved as already reported in Ref. [61]
(Fig. 7). Linker 3 was obtained in a 53% yield and a purity of
90% as determined by 1H NMR analysis. 1H NMR
(200.12 MHz, CDCl3, 25 ꢁC, TMS,
d (ppm)): 0.85 (t, 8.1 Hz,
2H, CH2); 1.31 (d, 6.0 Hz, 6H, CH3); 1.92 (m, 2H, CH2); 3.65
(s, 9H, CH3); 3.90 (m, 2H, CH2); 4.40 (sept, 6.1 Hz, 1H, CH);
5.20 (d, 10.1 Hz, 1H, CH); 5.71(d, 17.3 Hz, 1H, CH); 6.71e6.85
(m, 2H, CH); 6.90e7.15 (m, 2H, CH).
3.4. Membrane grafting
The main impurity was identified as remaining silane 2.
Because of purification difficulty, the grafting solution was
used as obtained and did not correspond to pure linker 3
even if the presence of silane 2 is not mentioned in the
following for the preparation of the CM1 membrane (vide
infra). This difficulty in the purification of linker 3 was one
of the main reasons for developing an alternative two-step
grafting procedure.
Grafting was performed according to two different
procedures depending on the linker 3 grafting, either in
one or two steps. In both cases, the final step consisted in
the reaction of the Grubbs II precatalyst with immobilized
linker 3 regardless of its immobilization procedure.
Preliminary tests aimed at checking that linker 3
grafting onto zirconia powder and the tubular membrane
was obtained, regardless of its grafting route.
Fig. 6. Grafting of 1 onto the silylated grafted membrane.
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

6
A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
Fig. 7. Synthesis of 1 and linker 3 according to Ref. [61].
The direct grafting on zirconia powder was achieved by
successively with dichloromethane, diethylether, and
pentane. A portion of P2 powder was used for analyses and
the other part was used for the next modification step. The
final modification step was achieved in a Schlenk tube
maintained under an argon atmosphere by a reaction of
20.9 mg Grubbs II precatalyst with 293 mg of P2-modified
powder in refluxing CH2Cl2 (15 mL) for 12 h. After washing
and drying, a final pale green powder was obtained, further
denoted as CP3. Such color indicated the formation of the
desired HoveydaeGrubbs supported precatalyst.
immersion in a solution of linker 3 followed by Grubbs II
immobilization. Such protocol was checked step by step,
and this material was used as reference (vide infra).
A second set of preliminary tests was performed to
check that silane grafting (either linker 3 for the one-step
protocol or silane 2 for the two-step protocol) can occur
on the active layer of the tubular membrane in the same
manner as for ZrO2 powder. The main difference between
these two materials was attributed to the carbon support of
the tubular membrane that could play the role of an
adsorbent toward all organic compounds involved in the
synthesis and could consequently lower the grafting ratio.
This was checked by immersion of small membranes of
25 mm length in silane solutions followed by reaction of
the immobilized silane either by 1 in the case of silane 2
grafting or by Grubbs II precatalysts in the case of linker 3
direct grafting.
3.4.2. Grafting of linker 3 in two steps on 25 mm membranes
by immersion followed by Grubbs II immobilization
Four similar samples of about 25 mm length of the zir-
conia virgin membranes were used. The grafting of linker 3
on these membrane samples was obtained in two steps by
grafting first silane 2 (Fig. 5) followed by reaction with 1
(Fig. 6).
A second procedure implemented under dynamic con-
ditions was performed in a similar manner to filtration
conditions. This procedure is aimed at proposing a scalable
procedure for further possible industrial applications.
A preliminary activation of the hydroxyl groups (ꢀOH)
of the zirconia surface was achieved with an alkaline so-
lution. This alkaline pretreatment creates
a strongly
nucleophilic Oꢀ, which can then react with the silane 2.
Alkaline pretreatment was performed by immersing the
zirconia membranes in a NaOH solution (0.1 mol Lꢀ1) for
16 h. The four membrane samples were rinsed with water
and dried in an oven at 120 ꢁC for 8 h. One membrane
(further considered as the pristine membrane) was kept as
a reference, and the three other membranes were subjected
to further modification steps (F1, F2, and CMF3).
3.4.1. Direct grafting of linker 3 on zirconia powder by
immersion
Grafting on zirconia powder was performed to have a
reference powder material. Therefore, only one grafting
procedure was implemented in that case. The pristine zir-
conia powder was dried for 24 h at 70 ꢁC under vacuum.
The modification was then carried out via direct grafting of
linker 3 in a Schlenk tube maintained under an argon at-
mosphere. A total of 351 mg of dry zirconia powder was
added at room temperature to a solution of 260 mg of
linker 3 in 15 mL of dry toluene and refluxed for 24 h. The
resulting white powder, further called P2, was then washed
The first step consisted in the reaction between the
activated hydroxyl groups of the pristine zirconia and the
methoxy groups of silane 2. The three membranes F1, F2,
and CMF3 were soaked separately in 10 mL of dry toluene
with 95.15
then washed successively with dichloromethane and
m
L of silane 2 at 120 ꢁC overnight. They were
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
7
diethylether. The F1 membrane was kept for analyses to
check the grafting efficiency and only F2 and CMF3 mem-
branes were subjected to the second modification step.
The second step consisted in an etherification reaction
of the silylated grafted sample membrane with 1 in the
presence of NaH. The F2 and CMF3 membranes were
soaked separately in 1.5 mL of dry THF with 41 mg of 1 and
5.8 mg of NaH at room temperature for 16 h and were then
washed successively with dichloromethane and dieth-
ylether. The F2 membrane was kept for analyses to check
the second reaction occurrence.
membrane surface. Alternatively, the membrane was dried
in an oven at 80 ꢁC or 100 ꢁC for one night and cooled down
to room temperature under vacuum in a desiccator.
The membrane was then mounted in the inox carter and
installed on the pilot (Fig. 8). Dried toluene was then
circulated in the membrane under a nitrogen atmosphere.
During this membrane conditioning, nitrogen at 1 bar was
continuously bubbled in toluene, and the transmembrane
pressure (TMP) was close to 1 bar to have permeate and to
carefully rinse the whole membrane structure.
The two routes for linker 3 grafting were implemented.
The final Grubbs II immobilization was achieved on the
CMF3 membrane in a Schlenk tube maintained under an
argon atmosphere by soaking CMF3 in 20 mg of the Grubbs
II precatalyst in CH2Cl2 (15 mL) at 40 ꢁC for 12 h. The
membrane was then washed with dichloromethane and
diethylether and finally dried.
The direct grafting of linker 3 was not achieved on
25 mm membranes and was directly performed according
to the dynamic procedure (see subsequently, CM1
membrane).
3.4.3.2. Dynamic direct grafting of linker 3. After the first
conditioning step exposed previously, toluene was replaced
by 100 mL of a 5.5 g Lꢀ1 toluene solution of silylated styrene
linker 3. The grafting solution was circulated in the
membrane tube at a low cross-flow velocity and a TMP of
1 bar for 13 h at 120 ꢁC. The solution was allowed to
naturally cool down to 30e40 ꢁC while maintaining the
circulation and the nitrogen atmosphere at 1 bar. The
membrane maintained under a nitrogen pressure was
successively rinsed with toluene, dichloromethane, and
diethylether to remove the excess of linker 3 that could be
adsorbed (and not grafted) on the membrane surface.
The final rinsing solvent (diethylether) was then
replaced by dry dichloromethane that was circulated in the
membrane under 1 bar nitrogen atmosphere for condi-
tioning. The solvent was then replaced by 100 mL of the
Grubbs II precatalyst dissolved in dry CH2Cl2 (400 mg Lꢀ1).
The solution was circulated in the membrane tube for 10 h
at 40 ꢁC with a low cross-flow velocity and a TMP of 1 bar.
The membrane maintained under nitrogen pressure was
successively rinsed with CH2Cl2 and diethylether to remove
the ungrafted precatalyst.
3.4.3. Grafting of tubular membranes in dynamic conditions
This dynamic procedure was implemented as a scalable
protocol for further possible industrial purposes. Accord-
ingly, we have considered that grafting in conditions close
to cross-flow filtration could be achieved leading to a
possible homogeneous modification of the active layer,
thanks to the circulation of the grafting solution. According
to such protocols grafting could occur on the membrane
surface but also in the pores.
An original setup was developed by our laboratory for
this dynamic grafting (Fig. 8). Two grafting procedures,
detailed subsequently, were used for the elaboration of the
tubular catalytic membranes.
The catalytic membrane obtained according to this
procedure was further called CM1. This membrane was
stored in dry toluene before catalytic tests. After the olefin
metathesis test, the membrane active layer was scrapped
for analytical purposes (see subsequently).
3.4.3.1. Membrane conditioning common to the two dynamic
grafting procedures. The first step of membrane condition-
ing was the same regardless of the following grafted ligand,
either linker 3 or silane 2.
The 200 mm tubular membrane was first dried for 24 h
in a desiccator under dynamic vacuum at room tempera-
ture to remove the molecular water adsorbed at the
3.4.3.3. Dynamic grafting of linker 3 in two steps. After the
first conditioning step exposed previously, toluene was
replaced by 100 mL of the grafting solution made of silane
Fig. 8. Experimental setup for dynamic grafting of the tubular membrane: (a) scheme; (b) picture.
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

8
A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
2 at 5.5 g Lꢀ1 in 100 mL toluene. The grafting solution, in
which nitrogen bubbling was maintained, was circulated in
the membrane tube at a low cross-flow velocity and a TMP
of 1 bar for 9 h at 114 ꢁC. The solution was allowed to
naturally cool down to 30e40 ꢁC while maintaining the
circulation and the nitrogen atmosphere at 1 bar.
The membrane was rinsed with toluene to remove the
excess of silane 2 that could be adsorbed (and not grafted)
on the membrane surface. The membrane was maintained
in toluene overnight, then toluene was replaced by 200 mL
of the grafting solution of 1. This solution was previously
freshly prepared by pouring 535 mg of 1 dissolved in the
minimum volume of diethylether in a 20 mL DMF sus-
pension of 0.0781 g of NaH at 0 ꢁC for 1 h before mixing
with 180 mL of toluene. The solution was circulated in the
membrane tube at a low cross-flow velocity and a TMP of
1 bar for 5 h at 120 ꢁC. The membrane was then carefully
rinsed with DMF.
Conversion ð%Þ ¼ ½ðinitial substrate concentration
ꢀ final substrate concentrationÞ
ꢂ =initial substrate concentrationꢃ
ꢂ 100
(1)
3.6. Analysis
3.6.1. Attenuated Total ReflectioneFourier transform infrared
Infrared (Fourier transform infrared [FTIR]) analyses
allow us to evidence functional groups made of elements
belonging to polar/polarizable bounds. FTIR analyses were
performed using the attenuated total reflection (ATR) mode
directly on powder obtained by scrapping of the active
layer of the inorganic membrane, either initial or grafted
one.
DMF was then replaced by 180 mL of a solution made of
150 mg of the Grubbs II precatalyst in dry CH2Cl2 in the
presence of 36 mg CuCl as phosphine scavenger. The so-
lution was circulated in the membrane tube for 15 h at
40 ꢁC with a low cross-flow velocity and a TMP of 1 bar. The
membrane was then rinsed with CH2Cl2 to remove the
ungrafted precatalyst.
The spectra were recorded between 4000 and 600 cmꢀ1
with a spectrometer provided by PerkineElmer (Spectrum
100, spectrum for windows software) equipped with a
diamond crystal with an incidence angle of 45ꢁ allowing
one reflection. The background spectrum was recorded in
the air. Each spectrum was the result of 20 scans with a
2 cmꢀ1 resolution.
The Grubbs II grafting step in DMF was repeated ac-
cording to the same protocol to increase, if necessary, the
grafting ratio. The catalytic membrane obtained according
to this procedure was further called CM2. This membrane
was then rinsed with toluene and stored in dried toluene
before catalytic tests.
3.6.2. Diffuse reflectance infrared Fourier transform
Diffuse reflectance infrared Fourier transform (DRIFT) is
another infrared analysis allowing similar characterization
as ATR-FTIR with a diamond crystal. Obtained spectra were
close to those obtained with ATR-FTIR techniques, except
some absorbance intensities more dependent on grinding
with DRIFT than with ATR. Slight variations in the wave-
number location can also be observed between the two
techniques.
DRIFT spectra were recorded directly on powders (that
must be carefully grinded) slightly dispersed in KBr.
The spectra were recorded between 4000 and 600 cmꢀ1
with a spectrometer provided by PerkineElmer (Spectrum
1000, spectrum for windows software) equipped with a
DRIFT accessory. The background spectrum was recorded in
the air with KBr powder. Each spectrum was the result of 20
scans with a 2 cmꢀ1 resolution.
3.5. Olefin metathesis test
A model reaction of ring-closing metathesis of N,N-
ꢁ
diallyltosylamide in toluene at 30 C was used to evaluate
the catalytic activities of the catalytic membranes and for
sake of comparison of the free catalyst in solution (Fig. 9)
[62].
From a practical point of view, catalytic tests were
performed in the cross-flow loop of the same pilot as that
used for the membrane grafting after a careful cleaning of
the pilot and a first conditioning of the membrane in
toluene. During the metathesis reactions the TMP was
1 bar and both retentate and permeate were collected over
time.
3.6.3. Liquid-state NMR
Liquid 1H NMR data were recorded on a Bruker DPX 200
spectrometer. Chemical shifts (d) are given in ppm related
The activity of the catalytic membranes was monitored
according to two criteria: the final conversion of the sub-
strate (Eq. 1) and the time to reach this conversion.
to TMS and calibrated with residual nondeuterated solvent.
Coupling constants are reported in Hertz (Hz); multiplicity
of signals is indicated by using the following abbreviations:
s ¼ singlet, d ¼ doublet, t ¼ triplet, q ¼ quartet, and
b ¼ broad.
3.6.4. Solid-state NMR
Solid-state NMR was used to evidence the grafting on a
zirconia active layer of modified membranes. The modified
zirconia powder was obtained by scrapping the internal
part of the tubular (catalytic) membranes.
To evidence the grafting of organic compounds on zir-
conia, solid-state NMR spectra were registered with two
different techniques. The magic angle spinning (MAS) was
Fig. 9. Model ring-closing metathesis reaction of diallyltosylamide (DATA).
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
9
used for 1H solid-state NMR. MAS was coupled with cross
polarization (CP) for 13C, 29Si solid-state NMR.
In both case, MAS-NMR and CP-MAS NMR spectra were
recorded on a Bruker 300 spectrometer using acquisition
parameters provided in Table 1.
nitrogen at 18 kPa. Quantification by the internal standard
method was obtained with a precision of 1%. Consequently
conversion ratio during metathesis was 2%.
4. Results and discussion
3.6.5. Inductively coupled plasmaeoptical emission
spectrometry
4.1. Zirconia powder modification
The content of ruthenium in grafted powder was
determined at the UMR-CNRS “Institut des sciences chi-
miques de Rennes” of the University of Rennes-1 by
inductively coupled plasmaeoptical emission spectrometry
(ICP-OES) analysis.
Zirconia powder was only modified according to the
first procedure starting by direct linker 3 grafting and
followed by Grubbs II immobilization.
4.1.1. Direct linker 3 grafting
About 10 mg of modified zirconia powder was obtained
by scrapping of the internal part of the tubular catalytic
membranes. The scrapped powder was dissolved in 2%
nitric acid. The Ru content was measured with a precision
better than 0.02%.
To evidence the effective grafting of linker 3, modified
zirconia powder (P2) was analyzed by SEM-EDX and DRIFT.
The presence of the linker 3 on this modified powder was
confirmed by the detection of both C and Si. Confirmation
of grafting was also provided by IR measurements. Indeed,
comparison of DRIFT spectra of pristine and P2-modified
zirconia powder highlighted the appearance of new bands
located in three regions: 2900e3100, 1350e1750, and 800
e1700 cmꢀ1 (Fig. 10). Assignments of some of the new
bands were proposed as follows: wavenumber (cmꢀ1):
3.6.6. Scanning electron microscopyeenergy dispersive X-ray
spectrometery
Scanning electron microscopyeenergy dispersive X-ray
spectrometery (SEM-EDX; microanalysis coupled with
SEM) was used to evidence the different chemical elements
present on the zirconia active layer of modified mem-
branes, especially looking for Si and Ru each one being
characteristics of a particular grafting step contrary to C
that can be either because of 1, silane 2, linker 3, or Grubbs
II, whereas Zr only imaged the zirconia. The modified zir-
conia material was obtained by scrapping the internal part
of the tubular (catalytic) membranes.
Measurements were performed on a scanning electron
microscope (JEOL JSM 6400, Japan) equipped with an ul-
trathin window energy dispersive X-ray spectrometer and
a microanalysis system (LINK INCA Oxford Instruments,
UK) allowing quantitative determination of elements. Re-
sults are expressed as atomic percentages with a precision
better than 5 atom %.
2974:
metric, 1652:
(CH3), 1312:
n
(CH3, CH2) symmetric, 2937:
(C]C) styrene, 1452: (C]C) aromatic, 1387:
(CH2). Finally, the broad signal observed
n(CH3, CH2) asym-
n
d
n
d
between ꢀ45 and ꢀ55 ppm in the solid-state 29Si CP-MAS
NMR spectra also evidenced the grafting of linker 3
(Fig. 11). This broad signal results from the contribution of
three different microstructures noted as T1, T2, and T3 cor-
responding to [eOSiR(OR0)2], [eO2Si(OR0)], and [O3SiR],
respectively, with R ¼ linker and R0 ¼ OMe, OH [63].
Fig. 12 shows the CP-MAS NMR 13C spectra on which the
following assignments were proposed. 13C CP-MAS NMR:
d
(ppm): 7 (CH2), 20 (CH3), 49 (OeCH3), 73 (CH, CH2), 110
e119 (3CH aromatic, CH2 styrene), 126 (C aromatic, CH
styrene), 135 (C aromatic), 137 (C aromatic).
4.1.2. Grubbs II precatalyst immobilization
3.6.7. Gas chromatography
To evidence the effective immobilization of Grubbs II
precatalyst on grafted linker 3, CP3-modified zirconia
powder was analyzed.
SEM-EDX confirmed the presence of C and Si evidencing
that a silane remained grafted to zirconia. However, it was
not possible to detect ruthenium on CP3 powder probably
because of the very small amount of Ru grafted and the
detection limit close to 5 atom % for each element. How-
ever, ICP-OES analysis proved the effective presence of
ruthenium on CP3-modified zirconia powder (Table 2).
DATA and c-DATA concentrations were determined
from gas chromatography (GC) quantification with a Shi-
madzu GC gas chromatograph (GC-2014). The semicapillary
column used was purchased from Supelco (Equity-5,
30 m ꢂ 0.53 mm ꢂ 1.5
mm film thickness). This column is
mainly apolar, thanks to the stationary phase made of
PDMS-PDPS (95-5). Injection achieved at 250 ꢁC was fol-
lowed by an analysis at a constant temperature set at
230 ꢁC. The detection was obtained by an Flame Ionization
Detector (FID) detector set at 250 ꢁC. The carrier gas was
4.1.3. Catalytic test: ring-closing metathesis of DATA
Table 1
The CP3 powder was evaluated in the catalytic test for
sake of comparison with the catalytic membranes. A total of
106 mg of CP3 was used with a first single load of 80 mg
DATA corresponding to 0.7 mol % of Ru to substrate ratio. A
maximum conversion ratio of DATA into c-DATA was
reached after 211 min at 30 ꢁC (Table 3). A second cycle was
achieved with a second load of the same amount of DATA
highlighting a little decrease in the activity that was
confirmed by a third cycle performed with a third load of
the same amount of DATA.
Conditions for solid-state NMR according to the element.
Nucleus
1H
13C
29Si
Technique
MAS
CP-MAS
CP-MAS
NMR frequency
Spinning rate (kHz)
Rotor size (mm)
Contact time (s)
300
14
4
75
8
4
59.64
12
4
e
2
3.5
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

10
A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
Fig. 10. DRIFT spectra of pristine and linker 3emodified (P2) zirconia powders.
zirconia active layer of the membrane and that the porous
carbon support was not a limitation (for instance massive
adsorption of reactants not free to diffuse toward the active
layer).
After alkaline activation, one pristine membrane was
kept as a reference and the other membranes were sub-
jected to further modification steps.
4.2.1. First step: grafting of silane 2
Three membranes (F1, F2, and CMF3) were grafted with
silane 2 according to the experimental protocol explained
previously. Among them, one membrane was kept for in-
termediate analyses and the two others were used for the
next modification step.
Fig. 11. CP-MAS 29Si NMR of P2.
Fig. 13 compares the ATR-FTIR spectra of the unmodified
and the silane 2-modified (F1) active layer scrapped on
membranes. The ATR-FTIR spectra evidenced the presence
of silane 2 onto a zirconia sample membrane through the
alkyl groups bore by silicon as the modified membrane
exhibited new bands located in two regions: 2800e2950
4.2. Membrane modified by immersion: grafting of linker 3 in
two steps followed by Grubbs II immobilization
This protocol was established by using four membranes
of 25 mm length and was validated step by step to check
that chemical reactions occurred as expected on the
and 1200e1550 cmꢀ1
Two new adsorption bands around 2844 and 2913 cmꢀ1
were assigned to (CH3, CH2) asymmetric and (CH3, CH2)
.
n
n
symmetric of the alkyl groups of the silane, respectively.
An enlargement of the 1200e1550 cmꢀ1 region showed
the presence of four new absorption bands at 1203, 1373,
1411, and 1539 cmꢀ1 corresponding to
the silane alkyl groups, respectively.
d(CH2) and d(CH3) of
Table 2
Ruthenium content as determined by ICP-OES.
Material
Ru (wt %)
Ru (mmol gꢀ1
)
CP3 (powder)
CM1 (membrane)
0.209
0.005
2.07 ꢂ 10ꢀ2
Fig. 12. Solid-state 13C CP-MAS NMR spectra of the zirconia powder before
(P2) and after (CP3) reaction with Grubbs IIe*new signals.
4.9 ꢂ 10ꢀ4
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
11
Table 3
Catalytic activity and multiple-use of modified zirconia in powder (CP3) and membrane form (CM1 and CM2).
Cycle
Material
CP3
CM1
CM2
Time (min)
Conversion (%)
Time (min)
92
Conversion (%)
5
Time (min)
Conversion (%)
1
2
3
211
240
392
97
88
54
90
210
12
15
4.2.2. Second step: grafting of 1
Difference spectrum 2 ¼ raw spectrum of F2 membrane
ꢀ raw spectrum of F1 membrane
The two remaining membrane samples (F2 and CMF3)
were subjected to another modification step by grafting 1
aiming at linker 3 final grafting.
Fig. 14 shows the ATR-FTIR spectra of the active layer of
the F2 membrane successively modified by silane 2 and
then 1. The 2800e2950 cmꢀ1 region showed a modified
absorption profile when compared to the F1 membrane
(3)
Fig. 15 shows the difference spectrum 1 determined
from Eq. 2 allowing to evidence the presence of alkyl
groups on the membrane.
Fig. 16 shows the difference spectrum 2 determined
from Eq. 3 allowing to evidence that alkyl groups on the F2
membrane were different from those grafted on the F1
membrane.
that could be assigned to new n(CH3, CH2) asymmetric and
n
(CH3, CH2) symmetric. Moreover, modification of the ab-
sorption in the 1200e1550 cmꢀ1 region showed the pres-
ence of three new absorption bands at 1238, 1477, and
1430 cmꢀ1 corresponding to
n(CeOeC),
n(C]C) styrene,
4.2.3. Final step: grafting of Grubbs II precatalyst
and
n(C]C) aromatic, respectively.
The remaining F2 membrane sample was subjected to
the final modification step by grafting Grubbs II precatalyst
leading to the expected catalytic membrane (CMF3).
ATR-FTIR failed at evidencing the grafting of the pre-
catalyst probably because of the very small amount of the
Grubbs II precatalyst grafted on the membrane. Neverthe-
less, the presence of ruthenium was detected by ICP-OES
analysis of a scrapped active layer of the CMF3 membrane
(quantification was not possible with sufficient accuracy).
As such modification on the ATR-FTIR spectra before
(F1) and after (F2) modification by 1 are quite subtle, the
effective grafting of 1 (leading finally to linker 3 grafting)
on the zirconia was more clearly evidenced by differences
between raw spectra calculated according to Eqs. 2 and 3:
Difference spectrum 1 ¼ raw spectrum of F2 membrane
ꢀ raw spectrum of pristine active layer
(2)
Fig. 13. ATR-FTIR spectra of the active layers scrapped from the pristine zirconia membrane and the silane 2egrafted membrane (F1 membrane).
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

12
A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
Fig. 14. ATR-FTIR spectra of the active layers scrapped from the (silane 2 þ 1)-grafted membrane (F2 membrane).
4.3. Membranes modified according to the dynamic protocol
4.3.1. Direct linker 3 grafting followed by Grubbs II precatalyst
immobilization
Two different grafting procedures were used to immo-
bilize linker 3 onto the active layer of two 200 mm length
zirconia tubular UF membranes either in one or two steps
before the final immobilization of the Grubbs II precatalyst
leading to a final catalytic membrane bearing a tailor-made
prototype HoveydaeGrubbs precatalyst of second
generation.
Few analyses were achieved to prove the grafting on the
tubular CM1 membrane as these analyses are destructive
except the measurement of the catalytic activity.
Ruthenium quantification by ICP-OES proved the effec-
tive presence of the GrubbseHoveyda precatalyst and
consequently the efficiency of the whole grafting scheme
(Table 2), whereas SEM-EDX failed at showing the presence
Fig. 15. Difference ATR-FTIR spectrum 1 between spectra of the active layers scrapped from the F2-grafted membrane (silane 2 þ 1) and from the pristine zirconia
active layer.
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
13
Fig. 16. Difference ATR-FTIR spectrum 2 between spectra of the active layers scrapped from the F2-grafted membrane (silane 2 þ 1) and the F1-grafted
membrane (silane 2).
of Si (linker 3 grafting) and Ru (Grubbs II immobilization)
because of the effective low grafting ratio. For the same
reason, the CP-MAS 29Si solid-state NMR of the active layer
of CM1 did not allow us to detect the signal of the covalent
ZreOeSi, whereas the 1H NMR solid-state CP-MAS of active
layer of CM1 could not really draw unambiguous
information.
Despite the evidence for low grafting, the catalytic ac-
tivity of the membrane was tested on the Ring Closing
Metathesis (RCM) of DATA with an entire 200 mm mem-
brane. Five percent conversion of DATA was reached in
92 min (Table 3). This result proves that a catalytic mem-
brane was prepared even if performances were disap-
pointing but in agreement with the low catalyst grafting
suggested by analysis of the catalytic membrane.
the first time, that such a sophisticated catalytic membrane
can be prepared for olefin metathesis catalysis although the
grafting ratio needs to be improved in the future.
4.4.1. Discussion
The CM2 membrane appeared better than the CM1 one
but it is not possible to definitively draw that the two-step
procedure applied for linker 3 grafting was better than the
one-step procedure. Up to now, it appeared very difficult to
identify the origin of such differences.
Performances of the catalytic membranes were less than
that of the similarly modified powder certainly because of a
low grafting ratio of the precatalyst on the membrane. This
could be because of a greater specific area of the P354
pristine powder compared to that of the active layer of the
UF M5-Carbosep membrane. Consequently, further modifi-
cation improvements could be done on a nanofiltration
membrane to increase this rate as its active layer might have
a greater specific area because of thermal treatments at
lower temperature during the active layer synthesis.
Moreover, the UF membrane flux to toluene was not
significantly modified by the grafting when compared to
the pristine zirconia membrane (and consequently prob-
ably the pore diameter remained constant), and no reten-
tion of DATA and c-DATA was observed and thus no
selectivity between the substrate and the product of the
olefin metathesis reaction. Using a nanofiltration mem-
brane would allow us to induce a better selectivity between
4.4. Grafting of linker 3 in two steps followed by Grubbs II
catalyst immobilization
Because of lengthy synthesis and membrane consump-
tion by analysis procedures, we considered that the con-
version of DATA in the ring-closing metathesis reaction was
sufficient to prove the grafting of a metathesis catalyst. The
catalytic activity of the CM2 membrane was thus tested on
the entire 200 mm membrane and a load of 500 mg of
DATA in 110 mL toluene. Under these conditions, 12% of
DATA conversion was reached in 90 min (Table 3). The
membrane was rinsed with toluene and a second cycle was
initiated with a new load of 500 mg of DATA. A similar
conversion of 15% was reached after prolonged reaction
time, hence proving that the catalytic membrane was not
deactivated after the first cycle. These results showed, for
some substrate and associated products allowing
a
continuous extraction of the product from the reaction
medium and not only a use of the catalytic membrane as a
membrane reactive contactor.
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

14
A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
Advanced Materials. NATO Science Series, II: Mathematics, Physics
5. Conclusion
and Chemistry, Springer, Dordrecht, Netherlands, 2007, pp. 483
e501.
The design and the grafting of the HoveydaeGrubbs II
precatalyst was developed according to two different
grafting procedures depending on the silylated ligand
grafted during the first modification step but leading,
theoretically, to the same overall modified membrane. To
the best of our knowledge, the catalytic membranes re-
ported in this manuscript are the first examples of a Hov-
eydaeGrubbs II precatalyst type grafted onto the surface of
a zirconia membrane. The catalytic activity of the “best”
membrane was tested in two consecutive cycles of a model
olefin metathesis reaction with no significant decrease in
the substrate conversion. Although the overall protocol
including the catalytic activity and the grafting ratio as well
as the process setup and efficiency must be improved, the
results obtained in the present study are innovative and
promising as the concept of a catalytic membrane based on
organometallic complexes was proved.
[16] C. Fischmeister, C. Bruneau, Top. Organomet. Chem. 51 (2015) 287
e306.
[17] P. Sledz, M. Mauduit, K. Grela, Chem. Soc. Rev. 37 (2008) 2433
e2442.
[18] Q. Yao, Y. Zhang, J. Am. Chem. Soc. 126 (2004) 74e75.
[19] M.R. Buchmeiser, New J. Chem. 28 (2004) 549e557.
[20] K. Melis, D. De Vos, P. Jacobs, F. Verpoort, J. Mol. Catal. A: Chem. 169
(2001) 47e56.
[21] D. Fischer, S. Blechert, Adv. Synth. Catal. 347 (2005) 1329e1332.
[22] D.P. Allen, M.M. Van Wingerden, R.H. Grubbs, Org. Lett. 11 (2009)
1261e1264.
[23] X. Elias, R. Pleixats, M.W. Chi Man, J.J.E. Moreau, Adv. Synth. Catal.
11 (2007) 1701e1713.
[24] A. Michrowsak, K. Menneke, U. Kuntz, A. Kirschning, K. Grela, J. Am.
Chem. Soc. 128 (2006) 13261e13267.
ꢀ
[25] N. Clousier, A. Filippi, E. Borr, E. Guibal, C. Crevisy, F. Caijo,
M. Mauduit, I. Dez, A.-C. Gaumont, ChemSusChem 7 (2014) 1040
e1043.
[26] B. Van Berlo, K. Houthoofd, B.F. Sels, P.A. Jacobs, Adv. Synth. Catal.
350 (2008) 1949e1953.
[27] H. Yang, Z. Ma, Y. Wang, Y. Wang, L. Fang, Chem. Commun. 46
(2010) 8659e8661.
[28] I.F.J. Vankelecom, Chem. Rev. 102 (2002) 3779e3810.
[29] P. Marchetti, M.F. Jimenez, G. Szekey, A.G. Livingston, Chem. Rev.
114 (2014) 10735e10806.
Acknowledgments
[30] www.kochmembrane.com/and www.duramem.evonik.com/.
[31] M. Janssen, C. Müller, D. Vogt, Green Chem. 13 (2011) 2247e2257.
[32] A.V. Volkov, G.A. Korneeva, G.F. Tereshchenko, Russ. Chem. Rev. 77
(2008) 983e993.
This study was financially supported by the French
Ministry for Research (A.K., PhD), the ANR (ANR-09-CP2D-
[33] P. Vandezande, L.E.M. Gevers, I.F.J. Vankelecom, Chem. Soc. Rev. 37
(2008) 365e405.
[34] A. Keraani, T. Renouard, C. Fischmeister, C. Bruneau, M. Rabiller-
Baudry, ChemSusChem 1 (2008) 927e933.
[35] M. Rabiller-Baudry, G. Nasser, T. Renouard, D. Delaunay, M. Camus,
Sep. Purif. Technol. 116 (2013) 46e60.
[36] G. Nasser, T. Renouard, S. Shahane, C. Fischmeister, C. Bruneau,
M. Rabiller-Baudry, ChemPlusChem 78 (2013) 728e736.
[37] K. A. Burdett, M. R. Collins, B. R. Maughon, M. L. Tulchinsky, Mem-
brane separation of a metathesis reaction mixture, Patent, WO2006/
047105 A2.
[38] D. Schoeps, K. Buhr, M. Dijkstra, K. Ebert, H. Plenio, Chem. Eur. J. 15
(2009) 2960e2965.
[39] P. Van Der Gryp, A. Barnard, J.P. Cronje, D. De Vlieger, S. Marx,
H.C.M. Vosloo, J. Membr. Sci. 353 (2010) 70e77.
ꢀ
11-01, program NanoRemCat2), and “Rennes Metropole”
(project UB453. 10ST422-01).
The authors acknowledge Francis Gouttefangeas from
CMEBA of University Rennes-1 (http://www.cmeba.univ-
rennes1.fr) for SEM-EDX analysis and Yann Le Gall and
Maxence Moine for ICP-OES measurements.
Solid-state NMR spectra were acquired at the CRMPO
(Centre regional de mesure physique de l'Ouest de l'uni-
ꢀ
ꢀ
versite de Rennes-1, http://www.crmpo.univ-rennes1.fr),
and the authors thank Dr. Marie Le Floch for her contri-
bution and helpful discussions about solid-state NMR.
The authors thank Tech Sep/CTI for providing zirconia
tubular membrane and zirconia powder. The authors also
[40] A. Kajetanowicz, J. Czaban, G.R. Krishnan, M. Malinska, K. Wozniak,
H. Siddique, L.G. Peeva, A.G. Livingston, K. Grela, ChemSusChem 6
(2013) 182e192.
[41] D. Ormerod, B. Bongers, W. Porto-Carrero, S. Giegas, G. Vijt,
N. Lefevre, D. Lauwers, W. Brustena, A. Buekenhoudt, RSC Adv. 3
(2013) 21501e21510.
ꢀ
thank Frederic Bon (IUT RenneseUniversite de Rennes-1)
for discussions.
[42] S.S. Ozdemir, M.G. Buonomenna, S.H. Choi, E. Drioli, Appl. Catal., A
307 (2006) 167e183.
References
[43] M.G. Buonomenna, S.H. Choi, E. Drioli, J. Chem. Eng. 5 (2009) 26e34.
[44] I.F.J. Vankelecom, A. Wolfson, S. Geresh, M. Landau, M. Gottlieb,
M. Hershkovitz, Chem. Commun. (1999) 2407e2408.
[45] D.E.C. Guedes, T.C.O. Mac Leod, M.C.A.F. Gotardo, M.A. Schiavon,
I.V.P. Yoshida, K.J. Ciuffi, M.D. Assis, Appl. Catal., A 296 (2005) 120
e127.
[46] A. Keraani, M. Rabiller-Baudry, C. Fischmeister, C. Bruneau, Catal.
Today 156 (2010) 268e275.
[47] A. Caselli, M.G. Buonomenna, F. de Baldironi, L. Laera, S. Fantauzzi,
F. Ragaini, E. Gallo, G. Golemme, S. Cenini, E. Drioli, J. Mol. Catal. A;
Chem. 317 (2010) 72e80.
[1] A. Fürstner, Chem. Commun. 47 (2011) 6505e6511.
[2] H. Mutlu, L. Montero De Espinosa, M.A.R. Meier, Chem. Soc. Rev. 40
(2011) 1404e1445.
[3] Y. Chauvin, Angew. Chem., Int. Ed. 45 (2006) 3740e3747.
[4] R.R. Schrock, Angew. Chem., Int. Ed. 45 (2006) 3748e3759.
[5] R.H. Grubbs, Angew. Chem., Int. Ed. 45 (2006) 3760e3769.
[6] C.H. Higman, J.A.M. Lumiss, D.E. Fogg, Angew. Chem., Int. Ed. 55
(2016) 3552e3565.
[7] A. Rybak, P.A. Fokou, M.A.R. Meier, Eur. J. Lipid. Sci. Technol. 110
(2008) 797e804.
[8] H. Bilel, N. Hamdi, F. Zagrouba, C. Fischmeister, C. Bruneau, Green
Chem. 13 (2011) 1448e1452.
[48] K. Li, Ceramic Membranes for Separation and Reaction, John Wiley&
Sons, New York, USA, 2007.
[9] S. Shahane, C. Bruneau, C. Fischmeister, ChemCatChem 5 (2013)
3436e3459.
[10] G.C. Vougioukalakis, R.H. Grubbs, Chem. Rev. 110 (2010) 1746
e1787.
[11] R.R. Schrock, A.H. Hoveyda, Angew. Chem., Int. Ed. 42 (2003) 4592
e4633.
[49] A. Juble, D. Farrusseng, C. Guizard, J. Membr. Sci. 181 (2001) 3e20.
[50] J.G. Sanchez-Marcano, Th T. Tsotsis, Catalytic Membranes and
Membranes Reactors, Wiley-VCH Verlag GmbH & Co. KGaA, Wein-
heim, Germany, 2002.
€
[51] M. Amirilargani, M. Sadrzadeh, E.J.R. Sudholter, L.C.P.M. de Smet,
Chem. Eng. J. 289 (2016) 562e582.
[12] G.C. Vougioukalakis, Chem. Eur. J. 18 (2012) 8868e8880.
[13] H. Clavier, K. Grela, A. Kirschning, M. Mauduit, S.P. Nolan, Green
Chem. 46 (2007) 6786e6801.
[14] M.R. Buchmeiser, Chem. Rev. 109 (2009) 303e321.
[15] C. Fischmeister, Catalytic Alkene Metathesis in Ionic Liquids in
Metathesis Chemistry: From Nanostructure Design to Synthesis of
[52] V. Meynen, A. Buekenhoudt, in: M.G. Buonomenna, G. Golemme
(Eds.), Hybrid Organic-Inorganic Membranes for Solvent Filtration
in Advanced Materials for Membrane Preparation, Bentham Pub-
lishers, Sharjah, U.A.E, 2009.
[53] C. Leger, H.D. Lira, R. Paterson, J. Membr. Sci. 120 (1996) 187e195.
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003

A. Keraani et al. / C. R. Chimie xxx (2017) 1e15
15
[54] J. Kujawa, W.M. Kujawski, S. Koter, K. Jarzynka, A. Rozicka, K. Baja,
S. Cerneaux, M. Persina, A. Larbot, Desalination Water Treat. 51
(2013) 1352e1361.
[55] S.P. Nunes, K.V. Peinemann, Surface Modification of Membranes,
Membrane Technology in the Chemical Industry, Second, Revised
[59] M. Rabiller-Baudry, G. Nasser, D. Delaunay, A. Keraani, T. Renouard,
D. Roizard, H. Ben Soltane, C. Fischmeister, J.-L. Couturier, D. Dhaler,
Interests and Challenges of Organic Solvent Nanofiltration for Sus-
tainable Chemistry: The Case of Homogeneous Catalysis of
Metathesis, WIT Transaction on Ecology and the Environment, vol.
154, WIT Press, UK, 2011, pp. 141e152, http://dx.doi.org/10.2495/
CHEM110141.
and Extended Edition, Wiley-VCH Verlag GmbH
& Co. KGaA,
Weinheim, Germany, 2006, pp. 39e43.
[56] C. Picard, A. Larbot, E. Tronel-Peyroz, R. Berjoan, Solid State Sci. 6
(2004) 605e612.
[57] S.R. Krajewski, W. Kujawski, F. Dijoux, C. Picard, A. Larbot, Colloids
Surf., A Physicochem. Eng. Aspects 243 (2004) 43e47.
[58] B. Chaufer, J. K. Liou, A. Bouguen, M. Rabiller-Baudry, L. Millesime,
[60] M. Liu, Z.P. Zhao, K.C. Chen, W.F. Liu, Catal. Commun. 64 (2015) 70e74.
[61] A. Keraani, C. Fischmeister, T. Renouard, M. Le Floch, A. Baudry,
C. Bruneau, M. Rabiller-Baudry, J. Mol. Catal. A: Chem. 357 (2012) 73
e80.
[62] C. Lo, R. Cariou, C. Fischmeister, P.H. Dixneuf, Adv. Synth. Catal. 349
(2007) 546e550.
ꢁ
ꢀ
ꢀ
Matiere inorganique de filtration modifiee par greffage d'organo-
ꢀ
ꢀ
ꢀ
mineraux et son procede de preparation, French patent,
WO9929402, 1999.
[63] J. Kujawa, W. Kujawski, ACS Appl. Mater. Interfaces 8 (2016) 7509
e7521.
Please cite this article in press as: A. Keraani, et al., First elaboration of an olefin metathesis catalytic membrane by grafting a
HoveydaeGrubbs precatalyst on zirconia membranes, Comptes Rendus Chimie (2017), http://dx.doi.org/10.1016/
j.crci.2017.04.003