Ordered mesoporous materials as solid supports for rhodium-diphosphine catalysts with remarkable hydroformylation activity

Article abstract of DOI:10.1039/c0cc00924e

A rhodium-diphosphine complex based on Xantphos-type ligands was anchored on mesoporous silica SBA-15, and the material showed remarkably high catalytic activity in the hydroformylation of 1-octene, even exceeding its homogeneous analogue under certain conditions.

Full text of DOI:10.1039/c0cc00924e

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Ordered mesoporous materials as solid supports for rhodium–diphosphine
catalysts with remarkable hydroformylation activityw
Fabrizio Marras, Jia Wang, Marc-Olivier Coppens and Joost N. H. Reek*a
a
b
bc
Received 14th April 2010, Accepted 19th July 2010
DOI: 10.1039/c0cc00924e
A rhodium–diphosphine complex based on Xantphos-type
ligands was anchored on mesoporous silica SBA-15, and
the material showed remarkably high catalytic activity in the
hydroformylation of 1-octene, even exceeding its homogeneous
analogue under certain conditions.
been studied intensively and possess interesting properties for
6
several applications. It is well known from the literature on
2,7
supported catalysts that mass transfer limitation is one of the
crucial issues leading to deteriorated catalytic performance. The
combination of high surface area with accurate control over the
8
mesopore size and pore connectivity influences both the
Homogeneous transition metal catalysts generally display
high activity and selectivity in contrast to their heterogeneous
intrinsic catalytic properties and reduces transport limitations,
5
resulting in higher activities and better controlled selectivities.
1
counterparts. However, many industrial catalytic processes use
In addition, the controlled structure of mesoporous materials
9
could lead to isolated metal species on the support surface,
heterogeneous catalysts, because of the easy separation from
the reaction product. Several strategies for the separation of
homogeneous catalysts from the product mixture have been
10
possibly preventing deactivation and enhancing activity. This
had not yet been demonstrated for hydroformylation catalysts
1
11
reported so far. The covalent anchoring of homogeneous
based on Xantphos derivatives. Therefore, SBA-15, a meso-
porous silica with large surface area and narrow pore size
distribution, was explored as support for the covalent anchoring
of rhodium-based transition metal catalysts. The rhodium–
diphosphine catalytic system previously investigated by our
transition metal complexes on solid supports is particularly
interesting, as, after heterogenization, the catalyst can be applied
2
in reactors typically suited for heterogeneous catalysis. In spite
of the huge effort dedicated to immobilization, there is no
important industrial application, mainly because of catalyst
leaching and a loss of rate and selectivity upon immobilization.
In this communication we report a rhodium catalyst for hydro-
formylation reactions that upon immobilization to a SBA-15
support does not lose its rate compared to the homogeneous
analogue, and even displays higher rates in some experiments.
Our group previously studied the hydroformylation reaction
using rhodium complexes based on diphosphine ligands
3
group was covalently anchored on both silica gel and
SBA-15, and the effect of these different mesoporous supports
12
on the hydroformylation of 1-octene was studied, and com-
pared to the homogeneous catalyst.
SBA-15 with a pore diameter of about 10 nm was prepared
11b,13
according to literature procedures.
The immobilization of 1
(Scheme 1) was readily accomplished by grafting it to the support
in boiling toluene, and subsequent capping of the remaining free
silanols by refluxing in the presence of dimethoxydimethylsilane,
3
anchored to commercially available silica gel, resulting in
3
recoverable and recyclable rhodium-based catalysts for the
conversion of alkenes into aldehydes or alcohols. However, as
usual, these covalently supported catalytic systems showed
lower activity and selectivity than their homogeneous analogues.
In the early 1990s a new family of porous silica materials was
according to the previously developed procedures (see also
ESIw). The rhodium–diphosphine–SBA-15 material 2 was
analyzed after synthesis by means of elemental analysis, to
determine the ligand content (based on P), and by ICP-OES
spectroscopy, to measure the amount of rhodium present on the
4
À5
developed, which are obtained by using micelles as templates
silica mesoporous support, leading to 1.0 Â 10 mol Rh per
À4
during the preparation, leading to periodically ordered channels
5
of the same size in the mesoporous range (>2 nm). Their
gram and 1.5 Â 10 mol ligand per gram SBA-15, i.e., a ligand-
to-metal ratio of 15.
surface reactivity, both internal and external, is rather close to
that of silica gel, so that grafting or tethering organic func-
tionalities onto these silicates can be readily achieved. There-
fore, such surfactant templated mesoporous materials have
To evaluate the effect of the support in catalysis, the catalytic
performance of the silica-supported rhodium–diphosphine
complex 2 was investigated in the industrially important
a
Supramolecular and Homogeneous Catalysis Group, van’t Hoff
Institute for Molecular Sciences, University of Amsterdam,
Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands.
E-mail: j.n.h.reek@uva.nl; Fax: +31 20 5255604;
Tel: +31 20 5256459
DelftChemTech, Delft University of Technology, Julianalaan 136,
2628 BL Delft, The Netherlands
b
c
Department of Chemical and Biological Engineering,
Rensselaer Polytechnic Institute, 110 8th Street, Troy, NY 12180,
USA. E-mail: coppens@rpi.edu; Fax: +1 518 276 3089;
Tel: +1 518 276 2671
w Electronic supplementary information (ESI) available: General
experimental methods, synthetic procedures, material characterization,
catalysis procedures, full catalysis table. See DOI: 10.1039/c0cc00924e
Scheme 1 Diphosphine 1 can be easily anchored onto silica supports.
Chem. Commun., 2010, 46, 6587–6589 | 6587
This journal is ꢀc The Royal Society of Chemistry 2010

View Online
used under the same catalytic conditions. At 20 bars of syn-gas
pressure and at 80 1C, catalyst 2 supported on SBA-15 showed
activities (TOF of 306) that are twice as high as the homo-
geneous Rh complex (entries 3 and 12, respectively). Compared
to the 50 bar experiment the regioselectivity slightly increased,
but it still is lower than observed for the homogeneous phase
Scheme 2 Rhodium-catalyzed hydroformylation of 1-octene.
(
91%, 92% and 96% of linear product, respectively). In
a
Table 1 Rhodium-catalyzed hydroformylation of 1-octene
addition, the hydroformylation reaction using the SBA-15
supported catalyst 2 was performed in pure substrate resulting
in very high activities and very high TONs (entries 4 and 5).
The facile diffusion of the reactants through the porous solid
support of the mesoporous SBA-15 material is particularly
important under these conditions. Moreover, the supported
catalyst 2 could be recycled by simple filtration up to 14 times
retaining both activity and selectivity, and without significant
metal leaching (o3 ppb according to ICP-OES analysis).
We were particularly surprised by the high activity observed
for the SBA-15 supported catalyst 2 compared to the homo-
geneous catalyst at 20 bars and with a ligand-to-metal ratio of
15 (entries 3 and 12, respectively, Table 1). It is well established
that a high concentration of ligand in the reactor mixture
promotes the formation of bis-ligated metal species 4 (Fig. 1)
b
Conversion Linear
Entry/ Material
cycle (batch)
Cumul.
d e
l/b TOF TON
c
P/bar (%)
(%)
1
2
3
4
5
6
7
8
9
/1
/7
/8
/9
SBA-15
1)
SBA-15
1)
SBA-15
1)
SBA-15
1)
SBA-15
1)
SBA-15
1)
SBA-15
2)
SBA-15
2)
SBA-15
2)
0/— SiO
50
50
20
20
20
20
50
20
50
90.8
92.6
94.2
91.6
88.2
90.0
84.7
92.7
96.2
91
35 195 671
(
h
37 n.d. 3885
91
92
94
93
85
93
93
67
(
47 306
47 995
4597
(
f
14 748
(
f
h
39 n.d. 67 143
/14
/15
/1
(
28 226
27 101
40 192
67 716
664
(
(
/4
2717
4083
1
2
(
known to be inactive in the hydroformylation reaction.
/6
3
161
8
More importantly, high catalyst loading and low pressures
14
facilitate the formation of inactive dinuclear complex 5. To
(
g
1
1
1
2
50
50
20
—
—
—
86
93
95
19
—
—
—
study if the formation of these complexes could also explain
1/— Homog.
2/— Homog.
41 120
59 149
the lower activity of the homogeneous catalyst, a 15 mL
À5
a
À5
À1
solution of the homogeneous catalyst (10
mol Rh) at
[Rh] = 10 mol g , ligand/Rh = 15, substrate/Rh = 640, toluene =
3 mL, CO/H
b
1
lated after 23 hours. Samples were analyzed by means of GC analysis using
2
= 1/1, temperature = 80 1C. Conversions were calcu-
20 bars and with a ligand-to-metal ratio of 15 at 80 1C was
monitored with high pressure IR during the reaction (Fig. 2).
c
n-decane as an internal standard. l/b is linear/branched ratio of the
À1
Four carbonyl bands typical of the 3ee (n = 2034 cm and
d
products formed (Scheme 2). TOFs are calculated as (mol product)-
À1
À1
À1
1
969 cm ) and 3ea (n = 1990 cm and 1942 cm ) con-
formations indicate the presence of Rh species active in the
À1 À1
(mol rhodium) (hour) at 20–40% conversion. Variations in TOF stem
from differences in conversion and product distribution. Cumulative
e
À1
f
. Performed in neat
hydroformylation reaction.
TON calculated as (mol product)(mol rhodium)
1
determined.
4
-octene (substrate/Rh = 10 ). Data taken from ref. 2. n.d. = not
g
h
The IR spectra also evidenced a broad carbonyl band
À1
around 1720 cm , clearly indicating the presence of dinuclear
Rh complex 5, the formation of which is most probably
suppressed by site isolation in the supported SBA-15 complex
analogue. This explains the low activity of the homogeneous
Rh–diphosphine complex compared to the SBA-15 supported
hydroformylation of 1-octene (Scheme 2) in a batch reactor,
using 1 g of silica material (Table 1).
After each catalytic cycle the product mixture was separated
from the supported catalyst by simple filtration. Generally,
supported transition metal catalysts show lower activities and
1
–3
selectivities with respect to their homogeneous analogues.
3
Rhodium complex 2 supported on silica gel (entry 10, Table 1)
indeed shows activities (TOF of 8) that are 15 times lower than
its homogeneous analogue (TOF of 120, entry 11). Also the
regioselectivity is slightly lower (l/b of 19) when applied under
the same catalytic conditions. The rhodium catalyst supported
on mesoporous SBA-15 showed much better performance in
the hydroformylation reaction than the silica gel supported
analogue (entries 1–6 compared to 10). At 50 bars of syn-gas
pressure and at 80 1C the SBA-15 supported catalyst 2 showed
high activities with a TOF up to 195 and a l/b regioselectivity of
around 36 (entries 1 and 2), demonstrating that both activity
and selectivity are much higher than those of the catalyst
immobilized on amorphous silica support (entry 10).
Remarkably, the activity is comparable to, and in some
experiments higher than that of the homogeneous catalyst
Fig. 1 Active rhodium species 3ee and 3ea and inactive rhodium
species 4 and 5 that can be formed under hydroformylation conditions.
6
588 | Chem. Commun., 2010, 46, 6587–6589
This journal is ꢀc The Royal Society of Chemistry 2010

View Online
studies are required to determine the role of confinement of the
catalytic species in these materials. Such studies are currently
being carried out in our laboratories.
Financial support to MOC and JW by the Delft Research
Centre for Sustainable Industrial Processes, and to FM by the
Dutch National Science Foundation NWO (NWO/STW) is
gratefully acknowledged.
Notes and references
1
2
(a) D. J. Cole-Hamilton, Science, 2003, 299, 1702; (b) Catalyst
Separation Recovery and Recycling: Chemistry and Process Design,
ed. D. J. Cole-Hamilton and R. P. Tooze, Springer, Dordrecht,
The Netherlands, 2006; (c) A. F. Trinidade, P. M. P. Gois and
C. A. M. Afonso, Chem. Rev., 2009, 109, 418.
(a) For a thematic issue on catalyst immobilization see: Chem.
Rev., 2002, 102, 3215–3892; (b) J. N. H. Reek, P. W. N. M. van
Leeuwen, A. G. J. van der Ham and A. B. de Haan, Catalyst
Separation, Recovery and Recycling: Chemistry and Process Design,
ed. D. J. Cole-Hamilton and R. P. Tooze, Springer, Dordrecht,
The Netherlands, 2006, ch. 3.
Fig. 2 High-pressure FT-IR spectra of Rh–Nixantphos in Me-THF
at 20 bars of CO/H = 1/1 pressure and 80 1C.
2
catalyst under these conditions. We also studied the reproduci-
bility of the catalyst immobilization by making a new batch of
supported ligand, using the same batch of silica SBA-15 and
identical conditions. The SBA-15 metal complex 2 showed
slightly decreased performance when this second batch was
applied as hydroformylation reaction (entries 7–9, Table 1).
While selectivities remained comparable to the previous batch
of SBA-15 material (linear over branched ratios around 30 and
3 (a) A. J. Sandee, L. A. van der Veen, J. N. H. Reek, P. C. J. Kamer,
M. Lutz, A. L. Spek and P. W. N. M. van Leeuwen, Angew. Chem.,
Int. Ed., 1999, 38, 3231; (b) A. J. Sandee, J. N. H. Reek, P. C. J.
Kamer and P. W. N. M. van Leeuwen, J. Am. Chem. Soc., 2001,
123, 8468.
4
0, respectively, at 50 and 20 bars of syn-gas pressure),
activities almost halved, with TOFs going down to 101 and
92 for the process at 50 and 20 bars respectively. Decreased
4
C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli and
J. S. Beek, Nature, 1992, 359, 710.
(a) J. S. Beek, J. C. Vartuli, W. J. Roth, M. E. Leonowicz,
C. T. Kresge, K. D. Schmitt, C. T.-W. Chu, D. H. Olson,
E. W. Sheppard, S. B. McCullen, J. B. Higgins and
J. L. Schlenker, J. Am. Chem. Soc., 1992, 114, 10834;
1
5
catalytic performance is also observed in terms of catalyst
stability. While the first batch of supported metal complex 2
was recycled up to 15 times, leading to a cumulative TON of
(
2
b) M.-O. Coppens, J. Sun and T. Maschmeyer, Catal. Today,
001, 69, 331; (c) Catalysts for Fine Chemical Synthesis: Micro-
6
7 716, the second batch of supported catalyst already showed
deterioration after the 4th cycle (cumulative TON of 4083),
leading to increased 1-octene isomerization and decreased
regioselectivity. Differences in stability and performance
between the different batches might be due to the used immobili-
porous and Mesoporous Solid Catalysts, ed. E. G. Derouane, Wiley
Ltd., Chichester, England, 2006; (d) J. A. Bae, K.-C. Song,
J.-K. Jeon, Y. S. Ko, Y.-K. Park and J.-H. Yim, Microporous
Mesoporous Mater., 2009, 123, 289.
P. Selvam, S. K. Bhatia and C. G. Sonwane, Ind. Eng. Chem. Res.,
2001, 40, 3237.
(a) B. Pugin, H. Landert, F. Spindler and H. U. Blaser, Adv. Synth.
Catal., 2002, 344, 974; (b) R. Dorta, D. Broggini, R. Stoop,
H. Ruegger, F. Spindler and A. Togni, Chem.–Eur. J., 2004, 10,
267.
(a) X. Feng, G. E. Fryxell, L.-Q. Wang, A. Y. Kim, J. Liu and
K. M. Kemmer, Science, 1997, 276, 923; (b) J. Liu, X. Feng,
G. E. Frixell, L.-Q. Wang, A. Y. Kim and M. Gong, Adv. Mater.,
6
7
15
zation technique. It was shown that when rhodium complex 1
was supported by standard techniques, pre-condensation of
ligand 1 could form clusters of 1, grafted to the silica support.
This high local density of catalyst species on the pore surface,
which may vary between batches, could lead to local depletion
8
12,14
of substrates, resulting in lower performance.
In conclusion, we successfully applied complex 2, covalently
anchored to SBA-15, to the rhodium-catalyzed hydroformyla-
tion of 1-octene. High activities and selectivities were obtained
compared to its supported silica gel analogue, demonstrating
the effect of the application of structured supports. Up to 94%
of linear aldehyde was obtained, free of metal complex, by
simple filtration. Moreover, activities exceeding those of
homogeneous analogues were obtained in some experiments.
This could be explained by the formation of inactive dimers
under homogeneous conditions, as indicated by HP-IR experi-
ments, which may be suppressed by site isolation effects when
these catalysts are supported on SBA-15. While such dimer
formation could also be suppressed by immobilization on
commercially available silica, this effect is then counter-
balanced by the generally negative effects of immobilization.
The presented results using SBA-15 demonstrate the impor-
tance of the nature of the support, and also show that the
supported catalyst can have advantages beyond mere catalyst
recycling. The reproducibility of the catalyst immobilization
should be improved before industrial application, and detailed
1
998, 10, 161.
9 J. M. Thomas, R. Raja and D. W. Lewis, Angew. Chem., Int. Ed.,
005, 44, 6456.
2
1
0 (a) B. F. G. Johnson, S. A. Raynor, D. S. Shephard,
T. Mashmeyer, J. M. Thomas, G. Sankar, S. Bromley,
R. Oldroyd, L. Gladden and M. D. Mantle, Chem. Commun.,
1999, 1167; (b) C. M. Crudden, D. Allen, M. D. Mikoluk and
J. Sun, Chem. Commun., 2001, 1154; (c) C. Li, Catal. Rev., 2004,
4
1
6, 419; (d) T. J. Terry and T. D. P. Stack, J. Am. Chem. Soc., 2008,
30, 4945.
11 (a) D. Zhao, Q. Huo, J. Feng, B. F. Chmelka and G. D. Stucky,
J. Am. Chem. Soc., 1998, 120, 6024; (b) D. Zhao, J. Feng, Q. Huo,
N. Melosh, G. H. Fredrickson, B. F. Chmelka and G. D. Stucky,
Science, 1998, 279, 548.
2 P. W. N. M. van Leeuwen, C. P. Casey and G. T. Whiteker,
Rhodium Catalyzed Hydroformylation, ed. P. W. N. M. van
Leeuwen and C. Claver, Kluwer Academic Publishers, Dordrecht,
The Netherlands, 2001.
3 (a) A.-H. Lu, W.-C. Li, W. Schmidt, W. Kiefer and F. Schuth,
¨
Carbon, 2004, 42, 2939; (b) R. Pitchumani, W. Li and
M.-O. Coppens, Catal. Today, 2005, 105, 618.
4 P. W. N. M. van Leeuwen, Appl. Catal., A, 2001, 212, 61.
5 F. Marras, A. M. Kluwer, J. R. Siekierzycka, A. Vozza,
A. M. Brouwer and J. N. H. Reek, Angew. Chem., Int. Ed.,
2010, 49, 5480.
1
1
1
1
This journal is ꢀc The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 6587–6589 | 6589