Mesoporous silica-supported catalysts for metathesis: Application to a circulating flow reactor

Article abstract of DOI:10.1039/b917986k

Using click chemistry for linkage, a ruthenium-based metathesis catalyst was efficiently immobilized on nanoporous silica. The heterogenized catalyst exhibited good activity and recyclability for various substrates. An interesting application was demonstr

Full text of DOI:10.1039/b917986k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Mesoporous silica-supported catalysts for metathesis: application
to a circulating flow reactorw
Jaehong Lim, Su Seong Lee* and Jackie Y. Ying*
Received (in Cambridge, UK) 2nd September 2009, Accepted 9th November 2009
First published as an Advance Article on the web 30th November 2009
DOI: 10.1039/b917986k
Using click chemistry for linkage, a ruthenium-based metathesis
catalyst was efficiently immobilized on nanoporous silica. The
heterogenized catalyst exhibited good activity and recyclability
for various substrates. An interesting application was demon-
strated for a continuous process using a circulating flow reactor.
Ring-closing metathesis (RCM) is of great interest in the
chemical and pharmaceutical industries.1 Unfortunately,
conventional ruthenium catalysts face challenges such as high
cost and metal leaching.2 Of the various approaches for
circumventing these problems, catalyst immobilization has a
clear advantage for industrial applications.3 Another key issue
for large-scale production lies in the type of reactors
employed. Flow reactors have numerous benefits including
more consistent product stream, less catalyst attrition, less
solvent waste, and less maintenance cost.4 Nevertheless, only a
few studies have been reported for a continuous metathesis
process,5 and their efficiencies need to be raised substantially.
Scheme 1 Synthesis of the immobilized catalyst 6.
We have developed a novel reactor system that made use of
immobilized catalysts6 in a continuous process by circulating
the reaction mixture to facilitate the removal of in situ
generated by-products.
MCF with excellent overall yield (see ESIw). Elemental
analysis revealed that the ruthenium loading in 6 was
0.16 mmol gÀ1, while the ligand density of 5 was 0.19 mmol gÀ1
.
This implied that 480% of the immobilized ligand was loaded
with ruthenium.
Taking advantage of the robust and orthogonal nature of
click chemistry,7 the triazole 3 was readily formed from
cycloaddition of the alkyne 1 and the azide 2 catalyzed by
Cu(I) (Scheme 1). Introducing a sub-stoichiometric amount of
hexamethyldisilazane (HMDS) to siliceous mesocelluar foam
(MCF) microparticles8 produced a partially trimethysilane
(TMS)-precapped MCF (4).9 A smooth grafting of 3 at
elevated temperature, followed by further treatment with
HMDS to block the unreacted silanol groups, led to the
immobilized ligand 5. A green powder 6 was then generated
by treatment with the second-generation Grubbs’ catalyst in
the presence of Cu(I) chloride.10 MCF is a stable mesoporous
silica with ultralarge and cell-like pores of 29 nm, inter-
connected by windows of 17 nm.8 This novel support material
has a large BET surface area (561 m2 gÀ1) and pore volume
(2.1 cm3 gÀ1), and is well-suited for fixating bulky complexes
and facilitates substrate diffusion.
In catalytic studies, diene 7 was smoothly and completely
cyclized in 30 min using 5 mol% of 6 in dichloromethane
(DCM) at ambient temperature without forming noticeable
by-products. Our previous study uncovered that the reaction
rates were influenced by the character of linker in 6.9 Kinetic
studies showed that the n-propyl linker gave the best catalytic
activity, while a more rigid (benzyl) or a more flexible
(n-pentyl) counterpart exhibited a lower conversion rate.
However, the difference was not very significant (see ESIw).
Table 1 summarizes the RCM of several dienes catalyzed by
6. In most cases, good catalytic activities were maintained for
5–10 runs at 50 1C in DCM. While the benchmark substrate 7
and an oxygen-containing diene 11 produced remarkable
results, the cyclization to a trisubstituted olefin 10 gave
relatively poor recycling efficiency. Good efficiency was
achieved in the production of a seven-membered ring 14 and
nitrogen-containing rings such as 16 and 18.
Cross-polarization magic angle spinning nuclear magnetic
resonance (CP-MAS NMR) 13C and 29Si spectra indicated the
high dispersion of Hoveyda-type ligands on the surface of
Based on the good catalyst recyclability, a continuous
process was investigated. A solution of diene 7 was eluted
through a small reactor packed with 6 (225 mg, reactor size =
4.6 mm  50 mm) at 50 1C at 0.5 ml minÀ1, which was the
highest flow rate for complete conversion at the outlet. As
depicted in Fig. 1, the conversion decreased sharply both in
toluene and DCM, presumably due to the deactivation and
leaching of 6. The deactivation might be due to inhibition by
Institute of Bioengineering and Nanotechnology, 31 Biopolis Way,
138669, The Nanos, Singapore. E-mail: sslee@ibn.a-star.edu.sg,
jyying@ibn.a-star.edu.sg; Fax: (+65) 6478-9020
w Electronic supplementary information (ESI) available: Detailed
experimental procedures, catalytic studies of 6 and its derivatives,
leaching and recycling studies, NMR spectra of the novel compounds
in Scheme 1. See DOI: 10.1039/b917986k
ꢀc
This journal is The Royal Society of Chemistry 2010
806 | Chem. Commun., 2010, 46, 806–808

View Article Online
Table 1 RCM of several dienes
Fig. 1 RCM of 7 catalyzed by 6 in a continuous flow reactor
(conditions: catalyst = 36 mmol, substrate concentration = 0.15 M,
flow rate = 0.5 ml minÀ1, T = 50 1C).
a Conditions: 5 mol% of 6, 50 1C, 0.05 M in DCM. b Determined by
gas chromatography (GC) or liquid chromatography (LC).
the increasing amount of products. In particular, ethylene
generated in a packed column could adversely impact the
active ruthenium species.11 An interesting report revealed that
ethylene pretreatment of a Grubbs’ type of ruthenium catalyst
prior to metathesis in a continuous flow reactor produced a
significant effect on catalyst stability.12 The authors pointed
out that efficient removal of products in the continuous flow
reactor was important to retain the catalytic activity. The
decomposition studies by Grubbs and coworkers also
supported our observation.13 Interestingly, the conversion
rates decreased more steeply in toluene than in DCM, losing
475% of the catalytic activity in 2 h (Fig. 1). This relatively
large solvent effect might be attributed to the higher solubility
of ethylene in toluene.14 Higher ethylene content in toluene
presented more interference in a packed column, and therefore
led to more catalyst deactivation.
Fig. 2 Schematic of the circulating flow reactor.
bed reactor by removing the by-product generated
(i.e. nitrogen gas).15 The circulating flow reactor successfully
demonstrated increased turnover frequency (TOF) and
reduced reaction time, while retaining the catalytic activity
and selectivity.15 In this study, 495% conversion of diene 7
was repeatedly attained in DCM for 12 runs (see Fig. 3).
With a high reactant concentration and a high flow rate, the
circulating flow reactor produced a good turnover frequency
of 1.5/min, which was 5 times higher than that of a batch
reactor. Again, toluene was found to be a less efficient solvent,
with a decreased conversion of only 63% at run #7. This lower
efficiency could be attributed to the higher solubility of
ethylene in toluene,14 leading to slower ethylene removal from
the reaction system.
To overcome the catalyst deactivation problem, we devised
a circulating flow reactor to facilitate the removal of in situ
generated ethylene from the packed bed reactor (see Fig. 2).
Instead of a one-time continuous flow process, our modified
system eluted the reactant solution through the packed bed
reactor (which was locally heated at 50 1C) repeatedly at a
relatively high flow rate (5 ml minÀ1) until the targeted
conversion had been achieved. During the course of circulation,
gaseous ethylene could be expelled simultaneously by degasser
and open reservoir so as to minimize the deactivation of
catalyst 6. A similar approach was previously reported
to facilitate the cyclopropanation of styrene in a packed
Continuous production of an oxygen-containing six-
membered ring 12 was efficient, transforming diene 11 at a
high conversion of 490% for 8 runs. Despite the higher
substrate concentration and substrate-to-catalyst ratio in the
1
circulating flow system, H-NMR spectra of the concentrated
crude product solution showed no trace of by-products. Thus,
the circulating flow reactor was superior in performance as
compared to the continuous flow reactor.
Inductively coupled plasma-mass spectroscopy (ICP-MS)
revealed that leaching of ruthenium species occurred mostly
at the early stage in both types of reactors. In the RCM of 7 in
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 806–808 | 807

View Article Online
deactivation, and ethylene solubility in solvents adversely
impacted the catalytic conversion. We are examining catalysts
with lower loading and bulky pre-capping groups to further
reduce catalyst leaching and improve catalyst recyclability.
These advances will contribute towards establishing more
environmentally benign industrial processes.16
This work is supported by the Institute of Bioengineering
and Nanotechnology (Biomedical Research Council, Agency
for Science, Technology and Research, Singapore).
Notes and references
1 P. A. Evans, J. Cui, S. J. Gharpure, A. Polosukhin and
H.-R. Zhang, J. Am. Chem. Soc., 2003, 125, 14702; A. Deiters
and S. F. Martin, Chem. Rev., 2004, 104, 2199; T. Nicola,
M. Brenner, K. Donsbach and P. Kreye, Org. Process Res. Dev.,
2005, 9, 513.
2 Ruthenium residue levels must be less than 5 ppm and 0.5 ppm for
oral and parental drug products according to the guidelines of
European Agency for the Evaluation of Medicinal Products. For
details, see: A. Thayer, Chem. Eng. News, 2005, 83, 55.
3 A representative example: Q. Yao, Angew. Chem., Int. Ed., 2000,
39, 3896.
4 G. Jas and A. Kirschning, Chem.–Eur. J., 2003, 9, 5708.
5 F. Sinner and M. R. Buchmeiser, Macromolecules, 2000, 33, 5777;
D. Schoeps, K. Buhr, M. Dijkstra, K. Ebert and H. Plenio,
Chem.–Eur. J., 2009, 15, 2960.
6 A recent example: J. Lim, S. S. Lee and J. Y. Ying, Chem.
Commun., 2008, 4312.
7 A recent review: J.-F. Lutz, Angew. Chem., Int. Ed., 2007, 46, 1018.
8 P. Schmidt-Winkel, W. W. Lukens, Jr., D. Zhao, P. Yang,
B. F. Chmelka and G. D. Stucky, J. Am. Chem. Soc., 1999, 121,
254; P. Schmidt-Winkel, W. W. Lukens, Jr., P. Yang,
D. I. Margolese, J. S. Lettow, J. Y. Ying and G. D. Stucky, Chem.
Mater., 2000, 12, 686; J. S. Lettow, Y. J. Han, P. Schmidt-Winkel,
P. Yang, D. Zhao, G. D. Stucky and J. Y. Ying, Langmuir, 2000,
16, 8291; J. S. Lettow, T. M. Lancaster, C. J. Glinka and
J. Y. Ying, Langmuir, 2005, 21, 5738; Y. Han and J. Y. Ying,
Angew. Chem., Int. Ed., 2005, 44, 288; T. M. Lancaster, S. S. Lee
and J. Y. Ying, Chem. Commun., 2005, 3577; S. S. Lee and
J. Y. Ying, J. Mol. Catal. A: Chem., 2006, 256, 219.
9 A recent example: J. Lim, S. S. Lee, S. N. Riduan and J. Y. Ying,
Adv. Synth. Catal., 2007, 349, 1066.
10 J. S. Kingsbury, J. P. A. Harrity, P. J. Bonitatebus, Jr. and
A. H. Hoveyda, J. Am. Chem. Soc., 1999, 121, 791.
11 M. Ulman and R. H. Grubbs, J. Org. Chem., 1999, 64, 7202.
12 Z. Lysenko, B. R. Maughon, T. Mokhtar-Zadeh and
M. L. Tulchinsky, J. Organomet. Chem., 2006, 691, 5197.
13 S. H. Hong, M. W. Day and R. H. Grubbs, J. Am. Chem. Soc.,
2004, 126, 7414; S. H. Hong, A. G. Wenzel, T. T. Salguero,
M. W. Day and R. H. Grubbs, J. Am. Chem. Soc., 2007, 129, 7961.
14 N. Intaragamjon, T. Shiono, B. Jongsomjit and P. Praserthdam,
Catal. Commun., 2006, 7, 721.
15 J. Lim, S. N. Riduan, S. S. Lee and J. Y. Ying, Adv. Synth. Catal.,
2008, 350, 1295.
16 P. T. Anastras and J. C. Warner, Green Chemistry Theory and
Practice, Oxford University Press, 2000; R. A. Sheldon, Green
Chem., 2000, 2, G1.
Fig. 3 RCM of 7 and 11 catalyzed by 6 in a circulating flow reactor
(conditions: catalyst = 36 mmol, substrate = 2.2 mmol/run, time =
40 min/run, substrate concentration = 0.15 M, flow rate =
5.0 ml minÀ1, T = 50 1C).
DCM using the circulating flow reactor (Fig. 3), the average
concentration of ruthenium leached out to the reaction
mixture was 8.4 ppm for the first 4 runs, and 2.9 ppm for
the subsequent 8 runs. A similar trend was also found in DCM
in the case of the continuous flow reactor. Ruthenium levels
were detected in the reaction mixture at 11.3 ppm for the first
60 min, and at 1.6 ppm at 180–240 min (Fig. 1). Overall, more
ruthenium leached out in the continuous flow reactor in the
conversion of a given amount of diene 7. 13C CP-MAS NMR
spectra also indicated that less active catalysts were lost in the
circulating flow reactor (see ESIw). These results suggested
that less ruthenium leaching and higher recycling efficiency
could be attained by reducing the catalyst loading. Tuning the
catalyst microenvironment by pre-capping the MCF with a
bulkier group to suppress the intermolecular decomposition
pathways of immobilized catalysts should also help to improve
recyclability.6 Although the amounts of ruthenium leaching
were not significantly different for the two reactor systems, the
lower recyclability in the continuous flow reactor indicated
that longer residence time of ethylene in the packed bed
reactor could lead to other deactivation pathways. Overall,
the circulating flow reactor was more beneficial, providing
good conversion rates over multiple runs, while the continuous
flow reactor was more vulnerable to catalyst deactivation.
In summary, a highly active and recyclable RCM catalyst
has been effectively achieved via the facile immobilization of
ligands onto MCF with click chemistry. A circulating flow
reactor has been successfully developed for a continuous
catalytic process. Studies with packed bed reactors showed
that in situ generated ethylene played a role in catalyst
ꢀc
This journal is The Royal Society of Chemistry 2010
808 | Chem. Commun., 2010, 46, 806–808