Batchwise and continuous nanofiltration of POSS-tagged Grubbs-Hoveyda-type olefin metathesis catalysts

Article abstract of DOI:10.1002/cssc.201200466

New molecular-weight-enlarged metathesis catalysts, which bear polyhedral oligomeric silsesquioxane (POSS) tags, were synthesized and characterized. The catalysts can be recovered from the reaction mixture by using nanofiltration techniques and can be reused. It was found that the membranes Starmem 228 and PuraMem 280 successfully separate the catalyst from the post-reaction mixtures to below 3 ppm. The application of these POSS-tagged catalysts in a continuous metathesis reaction was also investigated. Making metathesis plus nanofiltration POSS-ible: New molecular-weight-enlarged metathesis catalysts, which bear polyhedral oligomeric silsesquioxane (POSS) tags, have been synthesized and characterized. The catalysts can be recovered from the reaction mixture by using nanofiltration techniques and can be reused. The membranes Starmem 228 and PuraMem 280 can be used to successfully separate the catalyst from the post-reaction mixtures. Copyright

Full text of DOI:10.1002/cssc.201200466

CHEMSUSCHEM
FULL PAPERS
DOI: 10.1002/cssc.201200466
Batchwise and Continuous Nanofiltration of POSS-Tagged
Grubbs–Hoveyda-Type Olefin Metathesis Catalysts
[a]
[a]
[a]
[b]
´
Anna Kajetanowicz, Justyna Czaban, G. Rajesh Krishnan, Maura Malinska,
[b]
[c]
[c]
[c]
´
Krzysztof Wozniak, Humera Siddique, Ludmila G. Peeva, Andrew G. Livingston, and
Karol Grela*[a]
Dedicated to Professor Marek Chmielewski on the occasion of his 70th birthday
New molecular-weight-enlarged metathesis catalysts, which
bear polyhedral oligomeric silsesquioxane (POSS) tags, were
synthesized and characterized. The catalysts can be recovered
from the reaction mixture by using nanofiltration techniques
and can be reused. It was found that the membranes Star-
mem 228 and PuraMem 280 successfully separate the catalyst
from the post-reaction mixtures to below 3 ppm. The applica-
tion of these POSS-tagged catalysts in a continuous metathesis
reaction was also investigated.
Introduction
There has been an explosive growth in nanoscience and tech-
nology in recent years, primarily owing to the availability of
new strategies for the synthesis of nanomaterials and new
tools for their characterization and manipulation.[1] Nanotech-
nology influences almost all fields of basic and applied re-
search, and nanotechnology-based products are nowadays
making it from laboratory to market. The area of catalysis is
not an exception to this, and nanoparticle-based catalysts
show predominant activity compared to bulk-metal-based cat-
alysts.[2–5] There are two approaches to the use of nanoparticles
in catalysis. The first involves using nanoparticles themselves
as catalysts. In this approach, nanoparticles of metals, alloys,
metal oxides and non-metal oxides are exploited.[6] In the
second approach, nanoparticles are used as supports for cata-
lysts.[6–8] The advantages of using nanoparticles as supports for
homogeneous catalysts include the high surface area of the
supports, which allow the attachment of large amounts of cat-
alysts, and the tunable solubility of the nanoparticles, which
depends on the ligands used to stabilize them. Conventional
supports such as polymers or silica do not possess these ad-
vantages.
Homogeneous catalysts are frequently used in highly selec-
tive organic transformations and are especially significant in
asymmetric syntheses.[9] Nevertheless, in industry, many catalyt-
ic processes are still catalyzed by heterogeneous catalysts be-
cause of their facile separation from the product stream.[10]
In the field of homogeneous catalysis, separation of the cata-
lyst from the product mixture is rather complicated, which pre-
vents their use in large-scale industrial processes.[11]
An interesting and promising development in the area of
homogeneous catalyst recycling is the attachment of homoge-
neous catalysts to soluble organic supports and nanoparti-
cles.[6,11] In this way, macromolecular homogeneous catalysts
are created that can be recovered from the product stream by
ultra- or nanofiltration techniques and reused. Recycling cata-
lysts by nanofiltration represents an innovative route as it
could be employed as the final step in a complete discontinu-
ous production process or coupled to a batch reaction to lead
to a membrane reactor in a one-step continuous process.
To date, some examples of homogeneous-catalyst recycling by
using nanofiltration membranes (NF membranes) have been
reported,[11–22] but there are only few examples of its use in
olefin metathesis.[13,23,24]
[a] Dr. A. Kajetanowicz, J. Czaban, Dr. G. R. Krishnan, Prof. K. Grela
Institute of Organic Chemistry
Polish Academy of Sciences
Kasprzaka 44/52, 01-224 Warsaw (Poland)
Fax: (+48)22-632-66-81
E-mail: klgrela@gmail.com
This study involves the development of a Ru catalyst for
metathesis that is attached to polyhedral oligomeric silses-
quioxane (POSS) so that it can be effectively separated by
nanofiltration. Olefin metathesis seems to be the most impor-
tant field of research in organometallic chemistry over recent
decades.[25–28] It allows chemists to form C=C double bonds in
complicated natural products and is vital in polymer synthesis.
The introduction of well-defined metal-based catalysts
(Figure 1) resulted in enormous scientific progress and enabled
wide industrial applications.
Homepage: http://www.karolgrela.eu/
´
´
[b] M. Malinska, Prof. K. Wozniak
Department of Chemistry
Warsaw University
˙
Zwirki i Wigury Street 101, 02-089 Warsaw (Poland)
[c] H. Siddique, Dr. L. G. Peeva, Prof. A. G. Livingston
Department of Chemical Engineering
Imperial College London
Exhibition Road, London SW7 2AZ (UK)
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 182

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Results and Discussion
Ru catalysts are especially valued for their air and moisture
stability, which makes them easy to handle. A number of meta-
thesis reactions require high catalyst loadings, which results in
the presence of Ru in the final products.[29–35] In such cases, it
becomes necessary to develop extensive purification proce-
dures to remove the metal contamination, which makes meta-
thesis, from a commercial point of view, an expensive and time
consuming process overall. This can explain the urge to make
olefin metathesis a greener and cost-effective process.[36,37]
POSS derivatives comprise a family of molecularly precise,
near-isotropic molecules, which have diameters ranging from
1–3 nm, depending on the number of Si atoms in the central
cage and the nature of its peripheral substituent groups
(Figure 2). Unlike amorphous silica, the silsesquioxanes can be
Synthesis of ligands and catalysts
The synthesis of the POSS-supported Ru catalyst involves two
steps. In the first step, POSS was modified by using a styrene-
type ligand. The ligand was synthesized according to a multi-
step synthesis (Scheme 1).
Initially, commercially available 2-(4-hydroxyphenyl)acetic
acid (1) was esterified by using an excess of methanol in the
presence of a catalytic amount of H2SO4. The phenolic group
of methyl 2-(4-hydroxyphenyl)acetate (2) was converted into
the isopropyl ether 3 by treatment with isopropyl iodide in
DMF in the presence of K2CO3 and Cs2CO3. In the next step,
the acetic-acid-catalyzed bromi-
nation of 3 was carried out by
using molecular bromine in di-
chloromethane
(DCM).
The
bromo derivative 4 was convert-
ed into the styrene derivative 5
by applying Stille coupling with
n-butylvinylstannane with tetra-
kis(triphenylphosphine) as a cata-
lyst. The final step involves the
reduction of ester 5 to alcohol 6
by using LiAlH4. The styrene
Figure 1. Modern catalysts for olefin metathesis.
ligand 6 was obtained in 66% yield overall.
Ligand 6 was attached to POSS moiety through an urethane
linkage. For this, the hydroxyl group of the styrene molecule
was reacted with a POSS carrying an isocyanate group (7) in
the presence of a catalytic amount of dibutyltin dilaurate.
POSS ligand 8 was used in the preparation of the Grubbs–
Hoveyda-type catalysts (Scheme 2). The catalyst was prepared
from so-called second generation indenylidene catalysts that
possess a 1,3-bis-(2,4,6-trimethylphenyl)-4,5-dihydroimidazol-2-
ylidene (SIMes, 9a) or 1,3-bis-(2,4,6-trimethylphenyl)-imidazol-
2-ylidene (IMes, 9b) ligand by a ligand exchange method ac-
cording to an established procedure.[50] Catalysts 10a and 10b
Figure 2. Structure of the POSS moiety.
prepared with organic substituents (R) that can improve solu-
bility in protic and aprotic solvents.[38] Owing to their excep-
tional thermomechanical properties and oxidative stability,
a number of these compounds are used as ceramics fillers and
are incorporated into inorganic–organic hybrid co-polymers
that have diverse applications in electronics devices, as coating
films, and in lithography.[39–44] There are only few reports that
describe POSS as a support for homogeneous catalysts, and
POSS-supported catalysts for alkene polymerization, alkyne
metathesis, hydroformylation and oxidation were shown to be
highly efficient in both activity and recycling.[45–49]
1
were obtained as green powders. The H NMR spectra of these
materials showed singlet peaks at 16.51 and 16.67 ppm for
10a and 10b, respectively, which is a characteristic of the
Grubbs–Hoveyda catalyst.
This is the first time that the crystal structure of a Grubbs–
Hoveyda precatalyst with an attached POSS subunit has been
reported although crystal structures with POSS complexes with
Mo,[51] Pt,[52] Rh[53] and Zn[54] have been described. Single crys-
tals of 10a were obtained by crystallization from a DCM/MeOH
mixture.[55] Compound 10a crystallized in the monoclinic P21/
c space group with one molecule in the asymmetric unit.
The Ru atom in 10a is five coordinate, and the five-mem-
bered ring formed by Ru(1)C(22)C(23)C(24)O(3) is almost flat
(Figure 3). The Ru(1)C(22)C(23)C(24) torsion angle is À5.3(5)8.
The valence angle between the Ru atom and the two Cl atoms
is 160.91(4)8. The Ru(1)ÀC(1) and Ru(1)ÀC(22) bond lengths are
1.980(4) and 1.827(4) ꢁ, respectively. The geometry of the pre-
catalyst part of the complex remains nearly unchanged com-
pared to that reported.[56] The most significant discrepancy is
Here, we report the synthesis of new molecular-weight-en-
larged metathesis catalysts by attaching the Grubbs–Hoveyda
complex to a suitably functionalized POSS. The catalytic activi-
ty of this complex in various metathesis reactions was studied.
It appears to be comparable to the conventional, commercially
available homogenous metathesis catalysts. The catalyst can
be recovered from the final reaction mixture by using suitable
nanofiltration membranes and then it can be reused. The ap-
plication of the catalyst in a continuous metathesis reaction
was also investigated.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 183

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Scheme 1. Synthetic pathway for ligand precursors.
Scheme 2. Synthesis of 10a and 10b.
the conformation of the N-heterocyclic carbene (NHC) ligand;
the C(1)N(1)C(4) and C(1)N(2)C(13) valence angles are 123.8(4)8
and 129.2(4)8 in 10a, whereas in the Grubbs–Hoveyda struc-
ture they are 126.9(1)8 and 126.6(1)8, respectively. Furthermore,
the Ru(1)ÀO(3) bond is elongated to 2.244(4) ꢁ in 10a. The
POSS subunit is connected by an eight-atom chain with an
amide group. The POSS cage contains eight Si atoms with SiÀ
O bond lengths in the range from 1.616(3)–1.634(3) ꢁ. The
molecule forms a V-shaped structure connected to the neigh-
boring molecule by a very weak hydrogen bond between the
N(3)H(3) and Cl(1) atoms. The lengths of the contacts of
N(3)·Cl(1)x,Ày+1.5,z+0.5 and H(3)·Cl(1)x,Ày+1.5,z+0.5 are 3.288(4) and
2.49(4) ꢁ, respectively, whereas the N(3)ÀH(3) bond length and
N(3)ÀH(3)ÀCl(1) angle are 0.81(4) ꢁ and 167(4)8, respectively.
Molecules of 10a form layers in the crystal structure through
significant hydrophobic interactions between the POSS and
Figure 3. Atomic displacement parameters (ADPs) and the labeling of atoms
of 10a. Thermal ellipsoids are shown at the 50% probability level.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 184

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Figure 4. Packing of molecules in the crystal lattice of 10a viewed along the
y axis. Hydrogen atoms are omitted for clarity.
Figure 6. Time/conversion curve for the RCM reaction of diethyl diallylmalo-
nate (10b vs. Hov II’).
isobutyl substituents (Figure 4). The strongest interactions be-
tween the two neighboring layers are the CÀH···Cl
[C(29)Cl(2)1Àx,À0.5+y,0.5Àz 3.673(4) ꢁ] and CÀH···p interactions
[C(25)C(20)1Àx,À0.5+y,0.5Àz 3.726(6) ꢁ].
catalysts with the saturated NHC ligand SIMes and in Table 2
and Figure 6 for catalysts with the unsaturated NHC ligand
IMes.
As a starting point, we performed the RCM of diethyl diallyl-
malonate with 1 mol% of 10a (Table 1, entry I) and Hov II
(Table 1, entry V) in 0.1m concentration at room temperature
Activity studies and kinetic profiles
We selected the ring-closing metathesis (RCM) reaction of di-
ethyl diallylmalonate to study the activity of the synthesized
complexes (Scheme 3). All reactions were analyzed by means
Table 1. Detailed conditions for RCM reactions of 10a and Hov II.
Entry
Catalyst
Amount of
Solvent
Concentration
catalyst [mol%]
[m]
I
II
III
IV
V
10a
10a
10a
10a
Hov II
1
1
0.1
1
1
DCM
toluene
DCM
DCM
DCM
0.1
0.1
1
1
0.1
Scheme 3. Model RCM reaction of diethyl diallylmalonate.
of GC using internal standards to ensure that the correct prod-
uct was obtained. During the metathesis reactions promoted
by our new catalysts, product isomerization was not observed
(according to GC comparison with an authentic sample of the
product).
in DCM. The results obtained reveal that 10a initializes the re-
action slightly slower than the parent Hoveyda catalyst, but
after 4 h both achieve quantitative conversion. When the reac-
tion with 1 mol% of 10a was performed at higher concentra-
tion (c=1m, Table 1, entryIV), full conversion was achieved
after 30 min. Under the same conditions (RT, DCM, c=1m) but
with a smaller amount of 10a (0.1 mol%, Table 1, entry III), the
initiation was slightly slower, but the final result was better
than that of the reaction with 1 mol% of 10a and a concentra-
tion of 0.1m (Table 1, entry I). When the RCM of diethyl diallyl-
malonate was performed in toluene with 1 mol% of 10a, the
conversion rate was only 80% (Table 1, entry II), probably be-
cause of the fast decomposition of the catalyst.
The activity of the new POSS-tagged Hoveyda-type precata-
lysts 10a and 10b was compared with that of second-genera-
tion Hoveyda precatalysts with saturated and unsaturated li-
gands. The results are presented in Table 1 and Figure 5 for
Table 2. Detailed conditions for RCM reactions of 10b and Hov II’.
Entry
Catalyst
Amount of
Solvent
Concentration
catalyst [mol%]
[m]
I
II
III
IV
V
10b
10b
10b
10b
Hov II’
1
1
0.1
1
1
DCM
toluene
DCM
DCM
DCM
0.1
0.1
1
1
0.1
Figure 5. Time/conversion curve for the RCM reaction of diethyl diallylmalo-
nate (10a vs. Hov II).
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 185

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
A comparison of the activity
profiles of 10b and Hov II’ (with
the IMes ligand) in the RCM of
diethyl diallylmalonate was per-
formed. Under the standard con-
ditions (1 mol% of catalyst, c=
0.1m, RT, DCM), Hov II’ (Table 2,
entryV) initialized the reaction
faster than 10b (Table 2, entry I),
but achieved full conversion
after 5 h, whereas 10a achieved
it after 2 h. As expected, when
the reaction was performed with
the same amount of catalyst but
in a higher concentration (c=
1m, Table 2, entry IV), full con-
version was achieved after 1 h.
When the RCM of diethyl diallyl-
malonate was performed at the
same concentration but with
Table 3. Metathesis reactions catalyzed by 10a.[a]
Entry
Substrate
Product
t
[h]
Solvent
DCM
T
[8C]
Yield
[%]
1
2
3
0.5
25
99[b]
2
8
DCM
DCM
25
25
90[b]
96[b]
5
24
toluene
toluene
80
80
71
75
4
24
DCM
25
91
5
24
DCM
DCM
25
25
52
68
5
6
5
DCM
25
80
63[c]
a
lower
catalyst
loading
(0.1 mol%, Table 2, entry III),
almost quantitative conversion
was observed after 4 h. The reac-
tion was also performed in tolu-
ene (Table 2, entry II), which
gave a lower conversion, below
60%, probably because of the
low stability of 10b in this sol-
vent.
7
24
toluene
91[d]
[a] Conditions: c=0.1m, 1 mol% of 10a; Ts: p-toluenesulfonyl, TBDMS: tert-butyldimethylsily. [b] GC conver-
sion. [c] E/Z=85:15. [d] E/Z=90:10.
After initial studies involving
the RCM of diethyl diallylmalo-
nate, we found that both 10a
and 10b are almost as active as
the parent Hoveyda complexes.
As 10a gave rise to slightly
better results than its unsaturat-
ed analogue 10b, we decided
to present the preparative appli-
cations of 10a with a broader
range of substrates to explore
the catalytic activity of POSS-
tagged catalysts, and the results
are presented in Table 3.
Table 4. Membrane performance for catalyst separation with 10a.
Membrane
Conditions
Conversion[a] Catalyst
[%]
Physical appearance
rejection[b] [%]
Starmem 228
Starmem 240
toluene, RT, 5 h
toluene, RT, 5 h
91.89
91.89
100
72
98
smooth surface
creases at the edge as a result of swelling
smooth surface
PuraMem 280[c] toluene, 408C, 2.5 h 84
Duramem 500 DCM, RT, 3 h
96.40
71
creases at the edge as a result of swelling
[a] Based on GC analysis. [b] Based on UV analysis. [c] Results obtained during CSTR continuous experiment; re-
jection was calculated as: Rejection=(1Àcatalyst concentration in permeate/catalyst concentration in reten-
tate)ꢂ100.
The activity of 10a was tested in model reactions of ene–
yne (Table 3, entry 5)[57] and cross metathesis (Table 3, entries 6
and 7). Further examples of RCM are presented (Table 3, en-
tries 2–4), which include more demanding substrates with sub-
stituted double bonds. The yields of the products prove the
potential of the Ru complexes studied herein.
onate was performed with a catalyst loading of 0.02 mmol and
2 mmol of diethyl diallylmalonate per 20 mL of solvent. Upon
completion of the reaction, the post-reaction mixture was in-
troduced into the filtration cell, which was pressurized to 30ꢂ
105 Pa. Permeate (10 mL) was collected for rejection analysis of
the catalyst.
Table 4 indicates the rejection performance of three com-
mercial membranes that were initially tested in batch mode.
According to the supplier, Starmem 228 and Starmem 240 are
the most suitable for separations in toluene, which was there-
fore used as a reaction/separation media. The reaction/separa-
tion with Duramem 500 was performed in DCM. According to
the supplier, Duramem 500 is a cross-linked membrane stable
Nanofiltration
Membrane screening and batch studies
Initial screening of membranes was performed by separating
10a from the post-reaction mixture. RCM of diethyl diallylmal-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 186

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
in harsh solvents. The results in-
dicated that only Starmem 228
shows stable separation perfor-
mance with a catalyst rejection
of approximately 100% and
product rejection in the range of
10%. Starmem 240 and Dura-
mem 500 failed to separate the
catalyst properly and their stabil-
ity was also affected by the reac-
tion mixture (swelling and creas-
ing, which led to possible small
ruptures and resulted in low
rejection).
Figure 7. Schematic representation of the continuous metathesis reaction with catalyst recycling by using a mem-
brane (CSTR).
After the initial screening, ex-
periments were performed for
the recycling and reuse of the
catalyst by using Starmem 228 in batch mode. RCM was per-
formed in the reaction carousel with a catalyst loading of
0.04 mmol and diethyl diallylmalonate at 2 mmol per 20 mL of
solvent.
Initially, experiments were performed in the continuously
stirred tank reactor (CSTR) mode (Figure 7). As described in the
Experimental Section, the reaction was performed in 120 mL
toluene at 408C in a separate flask for 2.5 h to reach comple-
tion before being loaded in the CSTR. The catalyst seems to
show high reactivity, and after 2.5 h approximately 84% con-
version was achieved. The post-reaction mixture was loaded
into a CSTR and run at a very low dilution rate 0.2 mLminÀ1
(residence time 600 min). The commercial membrane Pura-
Mem 280[65] was used to retain the catalyst. The permeate was
The post-reaction mixture was introduced into the filtration
cell, the system was pressurized, and pure solvent was added
at the same rate at which the permeate was generated. Perme-
ate samples were taken at fixed intervals to assess the catalyst
rejection and product separation.
The rejection characteristics of Starmem 228 are presented
in Table 5. The catalyst rejection was again close to 100%, and
the Ru concentration in the permeate was below 3 ppm. Such
catalyst recovery is very high compared to other methods for
Ru-based-catalyst separation.[58–64] The product rejection was
again approximately 10%. The conversion dropped in the third
cycle owing to catalyst loss during the transfer from the filtra-
tion system to the reaction vessel and vice versa and possibly
catalyst deactivation on exposure to air. A small amount of Ru-
containing solid residue was also found deposited on the
membrane surface after the third cycle. This suggests that the
sudden conversion drop in the third cycle could be because of
partial catalyst decomposition.
Table 5. Study of catalyst recycling and reuse in batch mode with 10a
and toluene, using the Starmem 228 membrane the reaction was per-
formed at 608C overnight.
Cycle
Conversion[a]
[%]
Catalyst
rejection[b] [%]
Ru in permeate
[ppm][c]
1
2
3
99.9
90.2
78
100
100
100
2.10
2.32
1.02
[a] Based on GC analysis. [b] Based on UV analysis. [c] Based on ICP analy-
sis.
Metathesis reaction in continuous mode
With these encouraging results for a batch process in hand,
the metathesis reaction was performed in continuous mode
with catalyst recycling.
Figure 8. Continuous metathesis reaction with 10a: (a) % conversion in the
reaction with time and (b) rejection of substrate and product with time.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 187

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
colorless, which suggests a good retention of the catalyst. The
conversion remained constant for a few hours and then started
to decline (Figure 8a). Approximately 170 mL of permeate was
collected in which the Ru concentration varied from 1.86 ppm
in the first 50 mL, 1.54 ppm in the second 50 mL to 1.34 ppm
in the final 70 mL, which corresponds to approximately 2%
catalyst loss. It is unlikely that the conversion could be consid-
erably affected by this relatively small reduction in the catalyst
amount. Thus, the conversion drop was attributed to the possi-
ble deactivation of the catalyst and/or inhibition by ethylene,
which is a side product of the RCM reaction. A similar effect
was observed by Plenio et al.[13] The catalyst turnover number
was relatively low (ca. 144). The average product rejection was
approximately 10% (Figure 8b).
(PFR) during the continuous metathesis reaction performed by
using 0.1m diethyl diallylmalonate solution and 1 mol% of
10a with a residence time of approximately 133 min. The PFR
ran continuously for 32 h, and the reaction had a stable con-
version of approximately 100% throughout the run with a cata-
lyst turnover number of approximately 100. The permeate
stream contained approximately 0.34 mgLÀ1 Ru, which sug-
gests approximately 0.2% catalyst loss. The plug-flow configu-
ration seems promising, and further investigation should be
performed to evaluate its true potential.
Conclusions
In this study, organic solvent nanofiltration (OSN) was used to
separate a Grubbs–Hoveyda-type olefin metathesis catalyst
from the post-reaction mixture. The metathesis reaction of di-
ethyl diallylmalonate was used as a model reaction to test the
process feasibility. Among the
A plug-flow setup (Figure 9) was used to check the feasibility
for a continuous reaction. Figure 10 presents the conversion
over time at the outlet stream of the 40 mL plug-flow reactor
membranes
tested,
Star-
mem 228 and PuraMem 280
demonstrated the best catalyst
separation with rejections above
98%. In the batch mode, three
consecutive reaction–separation
cycles were successfully per-
formed. In the continuous
stirred tank reactor (CSTR) mode
for a coupled reaction–separa-
tion system with catalyst recy-
cling, catalyst stability is still
a challenge. A plug-flow reactor
(PFR) configuration could be
a viable alternative to perform
continuous ring-closing meta-
thesis (RCM) reactions coupled
with membrane separation of
the catalyst.
Figure 9. Schematic representation of the continuous metathesis reaction with catalyst separation by using
a membrane (PFR).
Experimental Section
Unless otherwise noted, all reac-
tions were conducted under an Ar
atmosphere. Flash column chro-
matography was carried out by using Merck silica gel 60 (230–400
mesh). NMR (1H and 13C) spectra were recorded by using Varian
Gemini 200 and 400 spectrometers; samples were dissolved in
CDCl3; chemical shifts (d) are reported relative to CDCl3. IR spectra
were recorded by using Perkin–Elmer Spectrum 2000 and 1170 FT-
IR spectrometers. ESI-MS were recorded by using an instrument
from Mariner Perseptive Biosystems, Inc. GC analyses were con-
ducted by using an HP 6890 chromatograph with an HP 5 column.
Elemental analyses were performed by the Institute of Organic
Chemistry, PAS, Warsaw.
Syntheses
Figure 10. Conversion over time at the outlet stream of the PFR during the
continuous metathesis reaction (the sudden drop in the conversion at 26.5 h
is because of a technical fault in the HPLC pump).
Methyl 2-(4-hydroxyphenyl)acetate (2): 2-(4-Hydroxyphenyl)acetic
acid (1) (29.0 mmol, 4.41 g) was dissolved in MeOH (120 mL).
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 188

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
To this solution was added H2SO4 (1.5 mL), and the reaction mix-
ture was heated to reflux overnight. Then the solvent was evapo-
rated, and the residue was dissolved in t-butyl methyl ether (TBME)
and washed with saturated NaHCO3, water, and brine (50 mL each).
The organic layer was dried over MgSO4, and the solvent was
evaporated under reduced pressure. The product obtained (yellow
oil, 28.4 mmol, 4.72 g, 98% yield) was directly used for the next re-
action without further purification. 1H NMR (200 MHz, CDCl3): d=
3.56 (s, 2H), 3.70 (s, 3H), 5.20 (bs, 1H), 6.76 (d, J=8.7 Hz, 2H),
7.09 ppm (d, J=8.7 Hz, 2H).[66]
(dd, J=1.6, 17.6 Hz, 1H), 6.83 (d, J=8.4 Hz, 1H), 7.03 (dd, J=11.2,
17.6 Hz, 1H), 7.12 (dd, J=2.4, 8.4 Hz, 1H), 7.38 ppm (d, J=2.4 Hz,
1H); 13C NMR (100 MHz, CDCl3): d=22.2, 40.4, 52.0, 70.9, 114.1,
114.2, 125.9, 127.4, 127.9, 129.4, 131.7, 154.3, 172.3 ppm; IR (CH2Cl2
film): n˜ =2978, 2952, 1740, 1625, 1492, 1435, 1384, 1373, 1336,
1308, 1248, 1154, 1138, 1119, 1015, 999, 957, 909, 855, 816, 709,
657, 422 cmÀ1
;
MS (ESI, MeOH): m/z=257.2 [M+Na]+, 491.2
[2M+Na]+; elemental analysis calcd (%) for C14H18O3 (234.29):
C 71.77, H 7.74; found: C 71.90, H 7.82.
2-(4-Isopropoxy-3-vinylphenyl)ethanol (6):
A
solution of
5
Methyl 2-(4-isopropoxyphenyl)acetate (3): 2 (28.0 mmol, 4.65 g),
K2CO3 (56.0 mmol, 7.74 g), and Cs2CO3 (5.60 mmol, 1.82 g) were dis-
solved in DMF (120 mL) and stirred for 1 h at RT. Isopropyl iodide
(56.0 mmol, 9.52 g, 5.60 mL) was added, and stirring was continued
for 16 h. Then the reaction mixture was poured onto 50 mL of
water, extracted with TBME (3ꢂ50 mL), washed with water (50 mL)
and brine (50 mL), dried over MgSO4, and the solvent was evapo-
rated. The residue was purified by using column chromatography
(30% of EtOAc in cyclohexane), which resulted in 3 as a colorless
(5.20 mmol, 1.22 g) in dry THF (52 mL) was cooled to 08C. To this
was added a solution of LiAlH4 (5.72 mmol, 0.217 g, 6.68 mmol) in
dry THF dropwise, and the mixture was stirred for 1 h at 08C. The
reaction mixture was warmed to RT and stirred for 1 h. The mixture
was diluted with diethyl ether (50 mL) and water (2 mL) and fil-
tered through a Celite plug. The filtrate was evaporated, and the
residue was purified by using column chromatography (10–20% of
EtOAc in cyclohexane), which resulted in a colorless oil (4.63 mmol,
1
0.96 g, 89%). H NMR (400 MHz, CDCl3): d=1.34 (d, J=6.0 Hz, 6H),
1
oil (25.2 mmol, 5.25 g, 90%). H NMR (400 MHz, CDCl3): d=1.33 (d,
2.81 (t, J=6.6 Hz, 2H), 3.83 (t, J=6.6 Hz, 2H), 4.50 (sept, J=6.0 Hz,
1H), 5.24 (dd, J=1.6, 11.2 Hz, 1H), 5.74 (dd, J=1.6, 18.0 Hz, 1H),
6.84 (d, J=8.4 Hz, 1H), 7.00–7.09 (m, 2H), 7.33 ppm (d, J=2.4 Hz,
6H); 13C NMR (100 MHz, CDCl3): d=22.2, 38.4, 63.7, 71.0, 114.0,
114.5, 127.0, 127.9, 129.2, 130.3, 131.8, 153.9 ppm; IR (CH2Cl2 film):
n˜ =3356, 2977, 2933, 2873, 1625, 1491, 1384, 1372, 1246, 1119,
1046, 957, 907, 814 cmÀ1; MS (ESI, MeOH): m/z=207.2 [M+H]+,
229.4 [M+Na]+; elemental analysis calcd (%) for C13H18O2 (206.28)
C 75.69, H 8.80; found: C 75.41, H 8.97.
J=6.0 Hz, 6H), 3.56 (s, 2H), 3.69 (s, 3H), 4.52 (sept, J=6.0 Hz, 1H),
6.81–6.88 (m, 2H), 7.14–7.20 ppm (m, 2H); 13C NMR (100 MHz,
CDCl3): d=22.0, 40.2, 51.9, 69.8, 115.9, 125.8, 130.2, 157.0,
172.4 ppm; IR (CH2Cl2 film): n˜ =2978, 1740, 1613, 1581, 1511, 1436,
1384, 1373, 1335, 1296, 1243, 1183, 1158, 1139, 1121, 1014, 955,
817, 786, 730, 690, 624, 553, 519, 478 cmÀ1; MS (ESI, MeOH): m/z=
231.2 [M+Na]+, 439.3 [2M+Na]+; elemental analysis calcd (%) for
C12H16O3 (208.25): C 69.21, H 7.74; found C 69.22, H 7.71.
Methyl 2-(3-bromo-4-isopropoxyphenyl)acetate (4): To a solution of
3 (24.7 mmol, 5.14 g) dissolved in DCM (80 mL) and acetic acid
(50 mL) was added bromine (24.7 mmol, 3.95 g, 1.27 mL) dropwise
at 08C, and the mixture was stirred for 3 h at RT. On completion of
the reaction (monitored by TLC), the mixture was quenched with
saturated aqueous Na2S2O5 (10 mL), poured into water (20 mL) and
extracted with diethyl ether (3ꢂ50 mL). The combined organic
layers were washed with water (50 mL), dried over anhydrous
MgSO4, and the solvent was evaporated to produce a colorless oil
Compound 8: 6 (1.10 mmol, 0.227 g) and 3-isocyanatopropylhep-
taisobutyl-POSS (7) (1.00 mmol, 0.901 g) were dissolved in dry DCM
(35 mL) under Ar. To this solution was added dibutyltin dilaurate
(ten drops), and the reaction mixture was stirred at RT for 18 h.
The solvent was evaporated, and the residue was purified by using
column chromatography (0–5% of EtOAc in cyclohexane), which
resulted in a white solid (0.978 mmol, 1.08 g, 97%). M.p. 97–988C
(DCM/CH3CN);1H NMR (400 MHz, CDCl3): d=0.58–0.64 (m, 14+2H),
0.95 (d, J=6.8 Hz, 42H), 1.34 (d, J=6.0 Hz, 6H), 1.54–1.64 (m, 2H),
1.79–1.86 (m, 7H), 2.86 (t, J=7.0 Hz, 2H), 3.12–3.20 (m, 2H), 4.24 (t,
J=7.0 Hz, 2H), 4.50 (sept, J=6.0 Hz, 1H), 4.68 (br s, 1H), 5.22 (dd,
J=1.6, 11.2 Hz, 1H), 5.72 (dd, J=1.6, 17.6 Hz, 1H), 6.81 (d, J=
8.4 Hz, 1H), 6.98–7.10 (m, 2H), 7.32 ppm (d, J=2.0 Hz, 1H);
13C NMR (100 MHz, CDCl3): d=9.0, 14.1, 22.2, 22.4, 22.5, 23.8, 23.9,
25.6, 25.7, 31.9, 34.8, 43.2, 65.4, 70.9, 113.9, 114.4, 126.9, 127.8,
129.0, 129.9, 131.9, 153.9, 156.4 ppm; IR (CH2Cl2 film): n˜ =3340,
2954, 1696, 1529, 1492, 1465, 1366, 1332, 1230, 1109, 957, 838,
742, 563, 482, 431 cmÀ1; MS (ESI, MeOH): m/z=1106.8 [M+H]+,
1128.8 [M+Na]+; elemental analysis calcd (%) for C45H87NO15Si8
(1106.85) C 48.83, H 7.92, N 1.27; found: C 48.58, H 7.88, N 1.14.
1
(24.2 mmol, 6.94 g, 97%). H NMR (400 MHz, CDCl3): d=1.37 (d, J=
6.0 Hz, 6H), 3.53 (s, 2H), 3.69 (s, 3H), 4.52 (sept, J=6.0 Hz, 1H),
6.84–6.88 (m, 1H), 7.11–7.16 (m, 1H), 7.44–7.48 ppm (m, 1H);
13C NMR (100 MHz, CDCl3): d=22.0, 39.8, 52.1, 72.2, 113.6, 115.6,
127.5, 129.1, 134.1, 153.7, 171.8 ppm; IR (CH2Cl2 film): n˜ =2979,
1740, 1605, 1567, 1493, 1435, 1385, 1374, 1337, 1281, 1255, 1217,
1154, 1110, 1047, 1015, 953, 903, 877, 850, 814, 728, 671, 578 cmÀ1
;
MS (ESI, MeOH): m/z=309.1 [M+Na]+, 596.9 [2M+Na]+; elemental
analysis calcd (%) for C12H15BrO3 (287.15): C 50.19, H 5.27, Br 27.83;
found: C 50.46, H 5.13, Br 27.74.
Methyl 2-(4-isopropoxy-3-vinylphenyl)acetate (5): In a flame-dried
Schlenck flask, 4 (6.00 mmol, 1.72 g) and tetrakis(triphenylphos-
phine)palladium(0) (0.300 mmol, 0.347 g) were dissolved in anhy-
drous toluene (35 mL) under Ar. To the reaction mixture was added
tributyl(vinyl)tin (7.20 mmol, 2.28 g, 2.10 mL), and the reaction mix-
ture was stirred at 1108C for 18 h. Then the reaction mixture was
cooled to RT and stirred with an aqueous solution of KF for 1 h.
The reaction mixture was filtered through a Celite plug, and the fil-
trate was extracted with DCM (3ꢂ20 mL). The combined organic
phases were washed with water (20 mL), dried over MgSO4, and
the solvent was evaporated. The residue was purified by using
column chromatography (0–5% of EtOAc in cyclohexane), which
resulted in a colorless oil (5.38 mmol. 1.26 g, 90%). 1H NMR
(400 MHz, CDCl3): d=1.34 (d, J=6.0 Hz, 6H), 3.56 (s, 2H), 3.69 (s,
3H), 4.52 (sept, J=6.0 Hz, 1H), 5.24 (dd, J=1.6, 11.2 Hz, 1H), 5.74
Catalysts 10a and 10b: Compound 8 (0.36 mmol, 398 mg), cop-
per(I) chloride (0.60 mmol, 59.4 mg) and 9a or 9b (0.30 mmol)
were placed in a Schlenk tube under Ar and dissolved in anhy-
drous toluene (10 mL). The mixture was heated at 808C for 40 min,
then cooled to RT, and the solvent was evaporated. The mixture
was then purified by using column chromatography (5–10% of
EtOAc in cyclohexane). The resulting product was concentrated in
vacuo, the green solid was dissolved in a minimal amount of DCM
and was washed with cold MeOH (10 mL). The precipitate was col-
lected by filtration, washed with MeOH (5 mL), and dried to afford
green crystals (44–91%).
Catalyst 10a: green solid (273 mmol, 428 mg, 91%); 1H NMR
(400 MHz, CDCl3): d=0.58–0.62 (m, 14+2H), 0.95 (d, J=6.4 Hz,
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 189

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
42H), 1.25 (d, J=6.0 Hz, 6H), 1.55–1.64 (m, 2H), 1.78–1.90 (m, 7H),
2.40 (s, 6H), 2.47 (s, 12H), 2.89 (t, J=6.8 Hz, 2H), 3.14–3.21 (m, 2H),
4.18 (s, 4H), 4.19 (t, J=6.8 Hz, 2H), 4.61–4.70 (m, 1H), 4.86 (sept,
J=6.0 Hz, 1H), 6.69–6.76 (m, 2H), 7.07 (s, 4H), 7.35 (dd, J=2.0,
8.4 Hz, 1H), 16.51 ppm (s, 1H); 13C NMR (100 MHz, CDCl3): d=9.2,
21.0, 21.1, 22.4, 22.4, 23.2, 23.8, 23.8, 25.6, 25.7, 34.1, 43.3, 51.5,
65.1, 74.9, 112.8, 120.1, 122.9, 129.3, 129.7, 131.7, 138.8, 145.3,
151.0, 156.3, 211.4, 297.1 ppm; IR (CH2Cl2 film): n˜ =2954, 2870,
1726, 1486, 1465, 1419, 1400, 1383, 1366, 1332, 1259, 1228, 1104,
1036, 837, 742, 578, 482, 432 cmÀ1; MS (FD): m/z=1571.3 [M+]; ele-
mental analysis calcd (%) for C65H111Cl2N3O15RuSi8 (157.25) C 49.69,
H 7.21, Cl 4.51, N 2.67; found: C 49.50, H 7.10, Cl 4.35, N 2.41.
Table 6. X-ray crystallographic details for 10a.
Crystal data
chemical formula
C65H111Cl2N3O15RuSi8
Mr
1571.26
crystal system, space group
monoclinic, P21/c
T [K]
100
a, b, c [ꢁ]
b [8]
37.462 (2), 10.808 (6), 19.693 (11)
92.769 (5)
7964.3 (8)
4
V [ꢁ3]
Z
radiation type
m [mmÀ1]
crystal size [mm]
MoKa
0.44
0.20ꢂ0.10ꢂ0.03
Catalyst 10b: green solid (132 mmol, 206 mg, 44%); 1H NMR
(400 MHz, CDCl3): d=0.58–0.64 (m, 14+2H), 0.95 (d, J=6.4 Hz,
42H), 1.34 (d, J=6.0 Hz, 6H), 1.56–1.64 (m, 2H), 1.78–1.90 (m, 7H),
2.27 (s, 12H), 2.46 (s, 6H), 2.92 (t, J=7.0 Hz, 2H), 3.14–3.22 (m, 2H),
4.21 (t, J=7.0 Hz, 2H), 4.65–4.70 (m, 1H), 4.89 (sept, J=6.0 Hz, 1H),
6.73 (d, J=8.4 Hz, 1H), 6.82 (d, J=2.0 Hz, 1H), 7.11–7.14 (m, 5H),
7.26–7.28 (m, 1H), 7.36 (dd, J=2.0, 8.4 Hz, 1H), 16.67 ppm (s, 1H);
13C NMR (100 MHz, CDCl3): d=9.2, 19.1, 21.0, 21.1, 22.3, 22.4, 23.3,
23.7, 23.8, 25.6, 34.2, 43.3, 65.1, 75.0, 112.8, 122.4, 124.7, 129.0,
131.8, 135.8, 138.0, 139.5, 145.5, 151.1, 156.3, 175.7, 293.0 ppm; IR
(CH2Cl2 film): n˜ =2954, 1725, 1488, 1465, 1401, 1383, 1366, 1319,
1228, 1105, 1037, 923, 837, 742, 695, 481, 431 cmÀ1. MS (FD): m/z=
1569.4 [M+]; elemental analysis calcd (%) for C65H109Cl2N3O15RuSi8
(1569.23) C 49.75, H 7.00, Cl 4.52, N 2.65; found: C 49.46, H 7.01,
Cl 4.56, N 2.58.
Data collection
diffractometer
absorption correction
KUMA4 CCD
multi-scan
0.917, 0.987
Tmin, Tmax
no. of measured, independent,
and observed [I>2s(I)] reflections
Rint
38413, 14530, 8715
0.088
Refinement
H-atom treatment
H atoms treated by a mixture of independent
and constrained refinement
no. of reflections
no. of parameters
no. of restraints
14530
872
0
0.055, 0.111, 1.00
0.68, À0.73
R[F2>2s(F2)], wR(F2), S
1max, 1min [eꢁÀ3
]
X-ray experiments
OSN set up and procedure for batch experiments
X-ray diffraction data were collected by using a Kuma KM4CCD
four-axis diffractometer with graphite-monochromated MoKa radia-
tion, equipped with an Oxford Cryosystems N2 gas-flow apparatus.
The crystal was positioned at 50 mm from the KM4CCD camera.
1203 frames were measured at 18 intervals with a counting time of
20 s. The data were corrected for Lorentz and polarization effects.
A multi-scan absorption correction was applied. Data reduction
and analysis were performed by using the Oxford Diffraction Ltd.
suit of programs.[67]
All batch nanofiltration experiments were performed in a Sepa ST
dead-end filtration cell (Figure 11, Osmonics USA) at 30ꢂ105 Pa
and 308C. A 14 cm2 disc of the membrane was fitted into a stainless
steel filtration cell with the active side of the membrane facing to-
wards the reaction mixture. A stainless steel sintered plate was
placed below the membrane to provide mechanical support to the
membrane during filtration. The filtration cell was closed by clamp-
ing the cell base against the body with a set of high pressure
clamps.[71]
The structure was solved by direct methods[68] and refined by
using SHELXL.[69] The refinement was based on F2 for all reflections
except those with very negative F2 values. Weighted R factors wR
and all goodness-of-fit S values are based on F2. Conventional R
factors are based on F with F set to zero for negative F2 values.
The reactions were completed by using a Carousel 12 (Radleys Dis-
covery Technologies) set up, and the post-reaction solution was in-
troduced in the dead-end cell for catalyst recovery. Pressure was
applied by connecting the filtration system to a pressurized N2 cyl-
inder. Continuous stirring was applied to avoid concentration po-
larization. After pressurizing the cell, fresh solvent was introduced
2
The F02 >2s[F0 ] criterion was used only to calculate R factors and
is not relevant to the choice of reflections for the refinement. The
R factors based on F2 are about twice as large as those based on F.
Scattering factors were taken from Tables 6.1.1.4 and 4.2.4.2 in
Ref. [70]. The crystallographic parameters are presented in Table 6.
Membranes used
Four commercial OSN membranes (Starmem 228, Starmem 240,
Duramem 500, and PuraMem 280) were purchased from Evonic
Membrane Extraction Technology, UK. The molecular weight cut
offs (MWCO, according to the supplier) are for Starmem 228 and
PuraMem 280 280 gmolÀ1, for Starmem 240 400 gmolÀ1and for Du-
ramem 500 500 gmolÀ1. Solvents used for MWCO determination
were toluene for Starmem 228, Starmem 240, and PuraMem 280,
DMF and acetone for Duramem 500.
Figure 11. Dead-end cell for catalyst recovery by nanofiltration.[45]
ChemSusChem 2013, 6, 182 – 192 190
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
into the filtration cell by using an HPLC pump at the same rate at
which the permeate was generated. The reaction mixture (20 mL)
was introduced to the filtration cell, and pure solvent (140 mL) was
used per wash (seven solvent wash volumes, sufficient to elute the
product through the membrane almost completely, estimated
from the product rejection data). After removal of the product, the
solution was transferred back into the carousel tube, and fresh
substrate was added for the next reaction cycle.
Acknowledgements
This work has been supported by the 7th framework programme
of the European Commission, Project CP-FP 214095–2 “HiCat”.The
authors would also like to thank Hybrid Catalysts for providing
the POSS compound.
Keywords: homogeneous catalysis · membranes · metathesis ·
microporous materials · ruthenium
Membrane pre-conditioning
[1] C. N. R. Rao, A. Muller, A. K. Cheetham, in The Chemistry of Nanomateri-
als: Synthesis Properties and Applications, Vol. 1 (Eds.: C. N. R. Rao, A.
Muller, A. K.Cheetham), Wiley-VCH, Weinheim, 2004, p. 1.
[2] K. J. Klabunde, in Nanoscale Materials in Chemistry (Ed.: K. J. Klabunde),
Wiley, New York, 2001.
All commercial membranes selected for this study were treated
with pure solvent prior to filtration of the reaction mixture to
remove any preserving agents from the membrane pores. Pure sol-
vent was passed through the membranes until steady-state flux
was achieved.
[3] G. Schmid in Nanoscale Materials in Chemistry (Ed.: K. J. Klabunde),
Wiley, New York, 2001.
[4] B. Hvolbæk, T. V. W. Janssens, B. S. Clausen, H. Falsig, C. H. Christensen,
J. K. Norskov, Nano Today 2007, 2, 14 and references therein.
[5] K. J. Klabunde, R. S. Mulukutla, in Nanoscale Materials in Chemistry (Ed.:
K. J. Klabunde), Wiley, New York, 2001.
[6] V. Polshettiwar, R. S. Varma, Green Chem. 2010, 12, 743.
[7] J. Fan, Y. Gao, J. Exp. Nanosci. 2006, 1, 457.
Continuous metathesis reaction with catalyst separation
through a nanofiltration membrane
[8] Y. Zheng, P. D. Stevens, Y. Gao, J. Org. Chem. 2006, 71, 537.
[9] R. Noyori, Asymmetric Catalysis in Organic Synthesis, Wiley, New York,
1994.
[10] K. P. de Jong, in Synthesis of Solid Catalysts (Ed.: K. P. de Jong), Wiley-
VCH, Weinheim, 2009, p. 3.
[11] H. P. Dijkstra, G. P. M. van Klink, G. van Koten, Acc. Chem. Res. 2002, 35,
798.
[12] I. F. J. Vankelecom, Chem. Rev. 2002, 102, 3779.
[13] D. Schoeps, K. Buhr, M. Dijkstra, K. Ebert, H. Plenio, Chem. Eur. J. 2009,
15, 2960.
[14] P. T. Witte, S. R. Chowdhury, J. E. ten Elshof, D. Sloboda-Rozner, R. Neu-
mann, P. L. Alsters, Chem. Commun. 2005, 1206.
[15] J. T. Scarpello, D. Nair, L. M. F. dos Santos, L. S. White, A. G. Livingston, J.
Membr. Sci. 2002, 203, 71.
[16] P. Vandezande, L. E. M. Gevers, I. F. J. Vankelecom, Chem. Soc. Rev. 2008,
37, 365.
[17] A. Datta, K. Ebert, H. Plenio, Organometallics 2003, 22, 4685.
[18] S. R. Chowdhury, P. T. Witte, D. H. A. Blank, P. L. Alsters, J. E. ten Elshof,
Chem. Eur. J. 2006, 12, 3061.
A flow-through nanofiltration cell[72] with a working volume of ap-
proximately 60 mL and a membrane area of 51 cm2 was set up and
used for continuous reaction–separation experiments.
Two modes of operation were investigated—CSTR with catalyst re-
cycling and PFR without catalyst recycling. For the CSTR experi-
ments, the reaction was performed in toluene (120 mL) at 408C
using 1 mol% of catalyst and 0.1m diethyl diallylmalonate. Initially,
the reaction was performed in a separate flask for 2.5 h to reach
completion before being loaded into the reaction vessel (Figure 7).
The reaction vessel was maintained at 408C with a heating/stirring
plate. Fresh 0.1m diethyl diallylmalonate feed was added continu-
ously to the reaction vessel (at 0.2 mLminÀ1 flow rate). The reac-
tion vessel content was continuously filtered through the mem-
brane in the flow-through nanofiltration cell using a filtration rate
equal to the feed supply. For the PFR experiments (Figure 9), the
catalyst and the feed were supplied as two separate streams
(0.15 mLminÀ1 each) at the entry point of the 40 mL loop. The
loop was maintained at 408C with a heating coil.
[19] R. Sablong, U. Schlotterbeck, D. Vogt, S. Mecking, Adv. Synth. Catal.
2003, 345, 333.
[20] A. Tsoukala, L. Peeva, A. G. Livingston, H. R. Bjørsvik, ChemSusChem
2012, 5, 188.
[21] P. G. N. Mertens, M. Bulut, L. E. M. Gevers, I. F. J. Vankelecom, P. A.
Jacobs, D. E. D. Vos, Catal. Lett. 2005, 102, 57.
[22] C. Mꢃller, M. G. Nijkamp, D. Vogt, Eur. J. Inorg. Chem. 2005, 4011–4021.
[23] A. Keraani, T. Renouard, C. Fischmeister, C. Bruneau, M. Rabiller-Baudry,
ChemSusChem 2008, 1, 927.
Analytical methods
UV spectroscopy: A Shimadzu UV-2101 PC, scanning between 200
and 350 nm, was used to calibrate and measure the concentration
of catalyst in the permeate and retentate samples.
[24] P. van der Gryp, A. Barnard, J.-P. Cronje, D. de Vlieger, S. Marx, H. C. M.
Vosloo, J. Membr. Sci. 2010, 353, 70.
[25] R. H. Grubbs, Handbook of Metathesis, Wiley-VCH, Weinheim, 2003.
[26] M. Michalak, Ł. Gułajski, K. Grela, Alkene Metathesis, Science of Synthesis,
Thieme, Stuttgart, 2010, 47a, 327.
[27] A. Fꢃrstner, Angew. Chem. 2000, 112, 3140; Angew. Chem. Int. Ed. 2000,
39, 3012.
[28] Y. Vidavsky, A. Anaby, G. N. Lemcoff, Dalton Trans. 2012, 41, 32.
[29] S. Maechling, M. Zaja, S. Blechert, Adv. Synth. Catal. 2005, 347, 1413.
[30] K. Vehlow, D. Wang, M. R. Buchmeiser, S. Blechert, Angew. Chem. 2008,
120, 2655; Angew. Chem. Int. Ed. 2008, 47, 2615.
[31] S. H. Hong, R. H. Grubbs, Org. Lett. 2007, 9, 1955.
[32] K. McEleney, D. P. Allen, A. E. Holliday, C. M. Crudden, Org. Lett. 2006, 8,
2663.
GC: An Agilent 6850 GC equipped with a flame ionization detector
and DB-FFAP capillary column 30 mꢂ0.25 mm was used to deter-
mine the substrate and product concentrations throughout the
metathesis reaction. The oven temperature was held at 1208C for
1 min then ramped to 2408C at 58CminÀ1 where it was maintained
for 5 min.
Inductively coupled plasma optical emission spectroscopy (ICP–
OES): A Perkin–Elmer 2000DV instrument was used to determine
the final elemental content of Ru in the permeate and retentate
samples. The samples were first evaporated at 308C to remove the
solvent, digested for 24 h with Aqua regia and finally diluted with
deionized water to a desired concentration range. The samples
were then analyzed for the respective metal contents.
[33] J. H. Cho, B. M. Kim, Org. Lett. 2003, 5, 531.
[34] T. Nicola, M. Brenner, K. Donsbach, P. Kreye, Org. Process Res. Dev. 2005,
9, 513.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 191

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
[35] M. T. Mwangi, M. B. Runge, N. B. Bowden, J. Am. Chem. Soc. 2006, 128,
14434.
[55] CCDC 880755 contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge from the Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[56] M. Barbasiewicz, M. Bieniek, A. Michrowska, A. Szadkowska, A. Makal, K.
Wozniak, K. Grela, Adv. Synth. Catal. 2007, 349, 193.
[36] H. Clavier, K. Grela, A. Kirschning, M. Mauduit, S. P. Nolan, Angew. Chem.
2007, 119, 6906; Angew. Chem. Int. Ed. 2007, 46, 6786.
[37] P. H. Deshmukh, S. Blechert, Dalton Trans. 2007, 2479.
[38] S. D. Hillson, E. Smith, M. Zeldin, C. A. Parish, J. Phys. Chem. A 2005, 109,
8371.
[57] M. Bieniek, A. Michrowska, D. L. Usanov, K. Grela, Chem. Eur. J. 2008, 14,
806.
[58] L. A. Paquette, J. D. Schloss, I. Efremov, F. Fabris, F. Gallou, J. Mendez-
Andino, J. Yang, Org. Lett. 2000, 2, 1259.
[59] Y. M. Ahn, K. Yang, G. I. Georg, Org. Lett. 2001, 3, 1411.
[60] H. Maynard, R. H. Grubbs, Tetrahedron Lett. 1999, 40, 4137.
[61] M. Westhus, E. Gonthier, D. Brohm, R. Breinbauer, Tetrahedron Lett.
2004, 45, 3141.
[39] J. Yu, Z. Qiu, ACS Appl. Mater. Interfaces 2011, 3, 890.
[40] S. W. Kuo, H. F. Lee, W. J. Huang, K. U. Jeong, F. C. Chang, Macromole-
cules 2009, 42, 1619.
[41] G. Lligadas, J. C. Ronda, M. Galia, V. Cadiz, Biomacromolecules 2006, 7,
3521.
[42] H.-J. Lee, J. Goo, S.-H. Kim, J.-G. Hong, H.-D. Lee, H.-K. Kang, S.-I. Lee,
M. Y. Lee, Jpn. J. Appl. Phys. Part 1 2000, 39, 3924.
[43] T. M. Chaudhry, L. T. Drzal, H. Ho, R. Laine, Proc. Annu. Meet. Adhes. Soc.
1996, 126.
[44] K.-I. Nakamatsu, K. Watanabe, K. Tone, T. Katase, W. Hattori, Y. Ochiai, T.
Matsuo, M. Sasago, H. Namatsu, M. Komuro, S. Matsui, Jpn. J. Appl. Phys.
Part 1 2004, 43, 4050.
[62] S. L. Balof, S. J. P’Pool, N. J. Berger, E. J. Valente, A. M. Shiller, H.-J.
Schanz, Dalton Trans. 2008, 42, 5791.
[63] C. C. Tzschucke, C. Markert, W. Bannwarth, S. Roller, A. Hebel, R. Haag,
Angew. Chem. 2002, 114, 4136; Angew. Chem. Int. Ed. 2002, 41, 3964.
[64] A. Michrowska, L. Gulajski, K. Grela, Chem. Commun. 2006, 841.
[65] Owing to the discontinuation of Starmem 228 production, Pura-
Mem 280 was offered as an alternative by Evonik Membrane Extraction
Technology.
[45] Y. Zhang, Z. Ye, Chem. Commun. 2008, 1178.
[46] F. A. He, L. M. Zhang, Nanotechnology 2006, 17, 5941.
[47] M. H. Cho, H. Weissman, S. R. Wilson, J. S. Moore, J. Am. Chem. Soc.
2006, 128, 14742.
[66] L. Dꢅez, P. Espinet, J. A. Miguel, M. P. Rodrꢅguez-Medina, J. Organomet.
Chem. 2005, 690, 261.
[67] Agilent Technologies, Data collection and data reduction, Version
171.35.15, 1995.
[48] M. Janssen, J. Wilting, C. Muller, D. Vogt, Angew. Chem. 2010, 122, 7904;
Angew. Chem. Int. Ed. 2010, 49, 7738.
[68] G. M. Sheldrick, Acta Crystallogr. Sect. A 1990, 46, 467.
[69] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112.
[70] International Tables for Crystallography (Ed.: A. J. C. Wilson), Kluwer, Dor-
drecht, 1992, Vol. C.
[71] Y. H. See Toh, X. X. Loh, K. Li, A. Bismarck, A. G. Livingston, J. Membr. Sci.
2007, 291, 120.
[49] M. T. Hay, S. J. Geib, D. A. Pettner, Polyhedron 2009, 28, 2183.
[50] H. Clavier, S. P. Nolan, M. Mauduit, Organometallics 2008, 27, 2287.
[51] F. Blanc, M. Chabanas, C. Copꢄret, B. Fenet, E. Herdweck, J. Organomet.
Chem. 2005, 690, 5014.
[52] J. I. van der Vlugt, J. Ackerstaff, T. Dijkstra, A. Mills, H. Kooijman, A Spek,
A. Meetsma, H. Abbenhuis, D. Vogt, Adv. Synth. Catal. 2004, 346, 399.
[53] B. Marciniec, I. Kownacki, A. Franczyk, M. Kubicki, Dalton Trans. 2011,
40, 5073.
[72] J. C.-T. Lin, A. G. Livingston, Chem. Eng. Sci. 2007, 62, 2728.
[54] C. Di Iulio, M. Jones, M. Mahon, D. Apperley, Inorg. Chem. 2010, 49,
10232.
Received: July 3, 2012
Published online on October 19, 2012
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 182 – 192 192