Silica immobilized second generation Hoveyda-Grubbs: A convenient, recyclable and storageable heterogeneous solid catalyst

Article abstract of DOI:10.1002/adsc.200800211

The second generation Hoveyda-Grubbs metathesis catalyst was successfully immobilized on silica in pellet and powder form following a practical and fast synthesis procedure. The activity of the solid system is truly heterogeneous, efficient in various metathesis reaction types and stable for at least 4000 TON. Ru contamination of the products was very low (ppb levels). The successful use of the robust system has been demonstrated in a continuous reactor set-up.

Full text of DOI:10.1002/adsc.200800211

COMMUNICATIONS
DOI: 10.1002/adsc.200800211
Silica Immobilized Second Generation Hoveyda-Grubbs:
AConvenient, Recyclable and Storageable Heterogeneous
Solid Catalyst
Boris Van Berlo,a Kristof Houthoofd,a Bert F. Sels,a,* and Pierre A. Jacobsa,*
a
Centre for Surface Chemistry and Catalysis, K. U. Leuven, Kasteelpark Arenberg 23, 3001 Heverlee, Belgium
Fax : (+32)-16-32-1998; e-mail: bert.sels@biw.kuleuven.be or pierre.jacobs@biw.kuleuven.be
Received: April 5, 2008; Revised: June 23, 2008; Published online: August 13, 2008
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.200800211.
Abstract: The second generation Hoveyda–Grubbs
metathesis catalyst was successfully immobilized on
silica in pellet and powder form following a practi-
cal and fast synthesis procedure. The activity of the
solid system is truly heterogeneous, efficient in vari-
ous metathesis reaction types and stable for at least
4000 TON. Ru contamination of the products was
very low (ppb levels). The successful use of the
robust system has been demonstrated in a continu-
ous reactor set-up.
Keywords: heterogeneous catalysis; immobilization;
metathesis; N-heterocyclic carbenes; ring-opening
polymerization; ruthenium
Figure 1. Commercial ruthenium-metathesis catalysts 1–6;
Cy=cyclohexyl and Mes=2,4,6-trimethylphenyl.
ysis,[5] most reports on supported metathesis catalysts
describe a loss of activity as a result of immobiliza-
Olefin metathesis has since its discovery left a mark tion. Moreover, many catalysts barely survive a few
on organic synthesis, polymer chemistry and catalyst recycles, while only a handful of systems guarantee
development, making metathesis a key reaction of the fully recyclable activity. Unfortunately, despite ingeni-
last decade. The discovery of Grubbsꢀ ruthenium cata- ous organic syntheses, their preparation often consists
lyst 1 inspired the development of a new family of of multiple steps or requires additional reagents and
metathesis catalysts (2–6, Figure 1) with high activity modified Ru complexes, consequently making them
and remarkable tolerance towards functional groups, less accessible for general use. A practical, inexpen-
air and moisture.[1,2] Taking into account Ru cost and sive immobilization strategy for recyclable and reusa-
high end-product purity, straightforward catalyst sepa- ble catalysts, compatible with organic and combinato-
ration, recyclability and minimal complex leaching rial chemistry and with large-scale potential remains
become decisive criteria for widespread applications thus challenging.
of metathesis in organic synthesis.
In our search for immobilizing metathesis catalysts,
Various immobilization strategies for metathesis we were attracted by the recovery experiments of
catalysts have been recently reviewed.[3] Anchoring of Hoveyda et al.[6a] and later Fogg et al.[6b] They recy-
stable ruthenium alkylidenes 3 and 4a via phosphine cled catalyst 4a and related aryl oxide catalysts simply
or halogen exchange, via the N-heterocyclic carbene by silica gel column chromatography. While a chro-
ligand, or via alkylidene exchange (boomerang-type matographic recovery is time-, solvent- and material-
catalysts) has been performed most frequently on in- consuming, sometimes with metal losses of 5% Ru
soluble organic polymers and silica supports.[4] Where- per cycle and more,[3c] we envisioned that the silica-
as immobilization of homogeneous catalysts combines bound catalyst could potentially be used as a hetero-
advantages of homogeneous and heterogeneous catal- geneous catalyst by playing with solvent polarity.
Adv. Synth. Catal. 2008, 350, 1949 – 1953
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1949

Boris Van Berlo et al.
COMMUNICATIONS
However, it is not clear how the Ru catalyst was olefins (8 and 9) were polymerized with success (en-
anchored to the silica, whether the silica supported tries 8 and 9). The potential of 4c was also tested for
catalyst would retain its activity, and even if so, self-metathesis (entry 10–13). Remarkably, complete
whether its activity is truly heterogeneous.
conversion of 1-octene 10 is obtained within short
The silica immobilized metathesis catalyst in this time using a low Ru concentration (0.18 mol%). Even
report is prepared in an easy and straightforward the less reactive ether functionalized styrenes 11 and
manner, taking advantage of the high affinity of silica 12 are converted selectively and almost quantitatively
gel (Fluka) for the stable catalyst 4a. In a typical ex- into the corresponding stilbenes 17 and 18 in reflux
ample, 0.2 g silica gel[8] is contacted with 4a (5 mL, conditions. As expected from the bulky aromatic
0.5 mM in toluene), and the suspension is stirred at moiety, trans isomers are mainly formed. The catalyst
293 K for two hours to acquire the bright greenish 4c is also active in the self-metathesis of 13 to the
powder 4b (0.1 wt% Ru) after a single wash (hexane) 1,20-diester 19, a potential monomer for polyester
and dry step. Higher loadings (up to 0.06 mmol mÀ2 or and polyamide synthesis. Finally, catalyst pellets 4c
0.6 wt% Ru) are easily achieved upon increasing the also proved to be active in the ring-closing metathesis
initial Ru complex concentration in the preparation of silanes (14 and 15) (entries 14 and 15), albeit with
solution and by using high surface silica material of incomplete product yield for the bulky diphenylsilane
the MCM-41 and SBA-15 types. For potential combi- 21.
natorial usage, silica pellets (WR Grace) instead of
While the powdered silica-based catalyst 4b has to
silica powders were loaded identically with 4a result- be separated from the products by filtration, work-up
ing in catalyst 4c (Scheme 1). We found that the silica with pellets 4c simply involves decantation of the
source has no important impact on the catalyst perfor- product mixture. Its ease of handling and isolation
mance, as long as it is free of impurities such as Al or offers the pellet-based catalyst 4c potential use in fast
Ca. The use of impure silica results in inactive brown screening experiments and combinatorial applications.
solids.[9]
The recovered solid catalysts appear remarkably
As a first test, cyclooctene 7 was polymerized using stable. Even with filtration in air and usage of undis-
0.5 mol% of catalyst 4b (0.05 wt% Ru loading) at tilled reagent grade solvent in a batch recycling ex-
293 K in undistilled, reagent-grade hexane with quan- periment, no diminution of activity for ROMP of 7
titative conversion within 3 h (Table 1, entry 1). As (conditions of entry 4; stable initial TOF of 220 hÀ1)
expected, 7 is converted much faster upon using was observed after four cycles, each to full conver-
higher Ru loadings (0.1 and 0.2 wt% Ru, entries 2 sion, corresponding to 400 TONs (see Supporting In-
and 3). Interestingly, catalyst pellets 4c show a compa- formation, Table SI1).
rable activity with the powder, despite shaking the re-
To exclude any contribution of homogeneous activi-
action instead of stirring to avoid mechanical damage ty, we employed a more decisive experiment, the so-
(entry 4). As expected, higher ROMP activities were called split test. A standard ROMP reaction with 7
measured with a higher concentration in 7 (entry 6). (conditions of Table 1, entry 4) was set up, and after
Catalysts 4b and 4c were stored for days at 293 K in a 50% consumption of 7, 10 vol% of supernatant was
glass container (not shown) or for several months at transferred into another reactor. Both reactions were
248 K without loss of activity (entry 7).
continued under identical conditions, and sampled
The activity of catalyst 4c has been evaluated in after 1 hour. As derived from Figure 2, ROMP in the
various metathesis reactions. Next to 7, other cyclic presence of 4c went to completion, while the removed
Scheme 1. Schematic of the preparation of catalyst pellets 4c.
1950
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1949 – 1953

Silica Immobilized Second Generation Hoveyda–Grubbs
Table 1. Various metathesis reactions (ROMP, CM and RCM) utilizing immobilized catalysts 4b and 4c.[7]
COMMUNICATIONS
No.
Substrate [M]
Product[c]
Ru (mol%) t [min] (T [K]) Yield [%] (TOF [hÀ1])[d]
7
A
ACHTREUNG
E
ACHTREUNG
N
ACHTREUNG
ACHTREUNG
x=4
1[a]
2[a]
3[a]
4
0.5
1
1.5
1
1.1
0.22
1
1.5
1.5
180 (293)
80 (293)
50 (293)
80 (293)
120 (293)
25 (293)
80 (293)
180 (293)
300 (293)
98 (170)
95 (218)
97 (240)
98 (222)
98 (390)
98 (3610)
97 (220)
97 (30)
5[b]
6
7
8
9
8 [0.05]
9 [0.05]
x=8
x=3
96 (27)
10
10 [0.3]
16
0.18
240 (293)
95 (913)
11[e] 11 [0.15]
17
3
300 (333)
90 (36)
12[e] 12 [0.15]
18
19
3
1
300 (333)
300 (333)
67 (21)
13[e] 13 [0.05]
90 (190)
14
15
14 [0.05]
15 [0.05]
20 (Z=Me)
21 (Z=Ph)
1
2
240 (318)
600 (318)
80 (106)
44 (6)
[a]
[b]
[c]
[d]
[e]
Catalyst 4b.
Silica pre-dried at 7008C.
Determined by GC-MS and H NMR.
After 10 min.
Isolated yields (89% for entry 11, 65% for entry 12, 85% for entry 13)
1
supernatant was free of activity. These data nicely il- true heterogeneous metathesis. All reactions summar-
lustrate a true heterogeneous activity of immobilized ized in Table 1 have passed successfully the aforemen-
4a, when carried out in hexane. As a comparative ex- tioned leaching split test.
ample, when an identical test was carried out in dieth-
To explore the stability limits of 4c, we performed
yl ether (see Supporting Information, Figure SI2), long-run reactions by continuously feeding a cyclooc-
active Ru species were leached in the solution. Use of tene solution over a fixed bed. Again, no precautions
non-polar solvents is thus recommended to prevent were undertaken to avoid traces of air or to purify
Ru leaching. While Ru leaching free metathesis of the chemicals. Strikingly, continuous experiments
functionalized substrates such as 11, 12 and 13 is pos- appear rarely in Ru-based metathesis literature to
sible, the use of highly polar substrates such as alco- evaluate the robustness of the heterogeneous catalyst.
hols and acids should obviously be avoided to assure In one example, a more pronounced catalyst deactiva-
Adv. Synth. Catal. 2008, 350, 1949 – 1953
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1951

Boris Van Berlo et al.
COMMUNICATIONS
tion set-up, no coloration of this silica loading was ob-
served. As no Ru displacement occurs along the cata-
lyst bed and no detectable Ru is found in the end
product, the active Ru species stay firmly anchored to
the silica. To have an idea of the interaction of the
Ru complex with the silica support, we carried out
FT-IR and 29Si MAS NMR. As there is a clear de-
À
crease of the isolated O H vibration stretch at
3745 cmÀ1 (see Supporting Information, Figure SI5)
and of the Q3 and Q2 signals at À101 and À93 ppm
(Supporting Information, Figure SI6) upon contacting
the silica with the Ru complex 4a, we propose a
direct chemical interaction of the Ru center with the
silanols of the surface instead of a weak physisorp-
tion.
Figure 2. Example of split-test for entry 4, Table 1.
We conclude that the second generation Hoveyda–
tion of 4a and metal leaching was found compared to Grubbs catalyst 4a is successfully immobilized on
the corresponding batch experiments.[4e] Yet, stable commercial silica in pellet and powder form according
catalytic performance under continuous flow could to a practical one-step procedure. The activity is truly
offer a great leap forward to metathesis applications. heterogeneous, efficient in various metathesis reaction
The data of our continuous experiments are summar- types and stable for up to at least 4000 TON. The suc-
~
ized in Figure 3. In one example ( ), 7 (0.05M in cessful use of the robust system in a continuous reac-
tor has been demonstrated. While the anchoring
mechanism is under further investigation, we propose
a direct attachment of Ru to the surface silanols via
ligand exchange.
Experimental Section
All chemicals were commercially available and used as re-
ceived.
Typical Procedure for Metathesis Reaction
Batch reactions: The respective substrate, in 5 mLundistil-
led reagent-grade hexane, was added to a small reactor
vessel containing 0.2 g of catalyst 4b or 4c. Reactions with
Figure 3. Continuous flow reactions of cyclooctene 7 with
catalyst 4b were stirred magnetically at 1000 rpm, while
catalyst 4c.
those using catalyst 4c were shaken. Reactions were con-
ducted at room temperature, unless stated otherwise, and
hexane), dosed at 50 mLgcat À1 hÀ1, was fully convert-
sampled at regular time intervals. The crude mixture was an-
ed at a TOF of 90–100 hÀ1 during the first ten hours.
alyzed with GC and GC-MS using toluene as the internal
standard. When ethylene was formed during reaction, an
When more concentrated 7 (0.2M) was fed at
argon balloon was used to avoid major pressure build up in
150 mLgcat À1 hÀ1, more than 4000 TONs were ob-
the reactor.
&
tained ( ). To the best of our knowledge, this is one
Continuous reaction: A solution of hexane containing the
of the highest numbers reported for immobilized 4a
for metathesis carried out in unprotected and non-pu-
rified circumstances.[3] Along with the activity, metal
(complex) leaching was carefully monitored with ICP
during the course of the continuous reaction. For col-
lected samples of the continuous experiment, the Ru
content was below the detection limit of 7 ppb (see
Supporting Information). Interestingly, the Ru con-
tamination is substantially lower than many reported
values, typically around 10 ppm and more.[3]
substrate was fed continuously with a syringe pump over a
catalyst bed (0.2 g catalyst 4c pellets in a quartz tube),
which was placed between quartz wool. A picture of the re-
actor set-up is presented in the Supporting Information. The
outlet of the reactor was sampled at regular time intervals.
The crude mixture was analyzed with GC and GC-MS using
toluene as internal standard.
Typical Procedure for Split Test
The reaction has been carried out according to the afore-
mentioned standard procedure. Here, after Æ50% consump-
Moreover, when a bed of pure white silica was
placed behind the catalyst bed in the continuous reac- tion of substrate, 10 vol% supernatant (without solid cata-
1952
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1949 – 1953

Silica Immobilized Second Generation Hoveyda–Grubbs
COMMUNICATIONS
lyst) was transferred into another reactor. Removal of the
catalyst has been carried out at the corresponding reaction
temperature in order to avoid retrapping of the catalyst.
Both reactions were continued under identical conditions,
and both sampled at various times. In order to prove the
heterogeneous nature of the catalytic reaction, all reactions
in Table 1 were subjected to this heterogeneity test.
Kirschning, M. Mauduit, S. P. Nolan, Angew. Chem. Int.
Ed. 2007, 46, 6786–6801.
[4] a) Q. Yao, Y. Zhang, Angew. Chem. Int. Ed. 2003, 42,
3395–3398; b) S. T. Nguyen, R. H. Grubbs, J. Organo-
met. Chem. 1995, 497, 195–200; c) X. Elias, R. Pleixats,
M. W. Chi Man, J. J. E. Moreau, Adv. Synth. Catal. 2006,
348, 751–762; d) L. Li, J.-L. Shi, Adv. Synth. Catal. 2005,
347, 1745–1749; e) A. Michrowska, K. Mennecke, U.
Kunz, A. Kirschning, K. Grela, J. Am. Chem. Soc. 2006,
128, 13261–13267; f) J. O. Krause, O. Nuyken, K. Wurst,
M. R. Buchmeiser, Chem. Eur. J. 2004, 10, 777–784;
g) F. Michalek, W. Bannwarth, Helv. Chim. Acta 2006,
89, 1030–1037; h) T. S. Halbach, S. Mix, D. Fischer, S.
Maechling, J. O. Krause, C. Sievers, S. Blechert, O.
Nuyken, M. R. Buchmeiser, J. Org. Chem. 2005, 70,
4687–4694; i) J. Lim, S. S. Lee, S. N. Riduan, J. Y. Yinga,
Adv. Synth. Catal. 2007, 349, 1066–1076; j) X. Elias, R.
Pleixats, M. W. C. Man, J. J. E. Moreau, Adv. Synth.
Catal. 2006, 348, 751; k) F. Michalek, D. Mädge, J. Rühe,
W. Bannwarth, J. Organomet. Chem. 2006, 691, 5172.
[5] D. E. De Vos, M. Dams, B. F. Sels, P. A. Jacobs, Chem.
Rev. 2002, 102, 3615–3640.
[6] a) J. S. Kingsbury, J. P. A. Harrity, P. J. Bonitatebus,
A. H. Hoveyda, J. Am. Chem. Soc. 1999, 121, 791–799;
b) J. C. Conrad, H. H. Parnas, J. L. Snelgrove, D. E.
Fogg, J. Am. Chem. Soc. 2005, 127, 11882–11883.
[7] Active metal leaching and thus contribution of homoge-
neous catalysis is excluded for all entries, based on the
split-test.
[8] The silica support is pre-dried in an oven overnight at
1208C before contacting with 4 in toluene, except in
entry 5, where it was pre-dried at 7008C under vacuum.
[9] Immobilization attempts with silica-alumina Siral 1/Siral
40 (Sasol) and Redco CaO·SiO2·6H2O resulted in inac-
tive brown powders.
Supporting Information
Experimental details of the batch and continuous experi-
ments, split and Ru leaching tests and characterization data
are given in the Supporting Information.
Acknowledgements
Financial support in the IAP-PAI, IDECAT, GOA and
CECAT schemes are gratefully acknowledged.
References
[1] R. H. Grubbs, Handbook of Metathesis, Vol. 1, Wiley-
VCH, Weinheim, 2003, pp 61–67.
[2] For selected reviews: a) R. H. Grubbs, S. Chang, Tetra-
hedron 1998, 54, 4413–4450; b) T. M. Trnka, R. H.
Grubbs, Acc. Chem. Res. 2001, 34, 18–29; c) R. R.
Schrock, A. H. Hoveyda, Angew. Chem. Int. Ed. 2003,
42, 4592–4633.
[3] For selected reviews: a) C. CopØret, J.-M. Basset, Adv.
Synth. Catal. 2007, 349, 78–92; b) A. H. Hoveyda, D. G.
Gillingham, J. J. Van Veldhuizen, O. Kataoka, S. B.
Garber, J. S. Kingsbury, J. P. A. Harrity, Org. Biomol.
Chem. 2004, 2, 8–23; c) H. Clavier, K. Grela, A.
Adv. Synth. Catal. 2008, 350, 1949 – 1953
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1953