Highly regioselective terminal alkynes hydroformylation and Pauson-Khand reaction catalysed by mesoporous organised zirconium oxide based powders

Article abstract of DOI:10.1039/b512543j

Zirconia-silica mesoporous powders act as very efficient heterogeneous catalysts for both alkyne hydroformylation and Pauson-Khand reaction and yield regioselectivities opposite to those usually observed. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b512543j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Highly regioselective terminal alkynes hydroformylation and
Pauson–Khand reaction catalysed by mesoporous organised zirconium
oxide based powders{
a
Fr e´ d e´ ric Goettmann, Pascal Le Floch* and Cl e´ ment Sanchez*
b
a
Received (in Cambridge, UK) 6th September 2005, Accepted 18th October 2005
First published as an Advance Article on the web 15th November 2005
DOI: 10.1039/b512543j
Zirconia–silica mesoporous powders act as very efficient
heterogeneous catalysts for both alkyne hydroformylation
and Pauson–Khand reaction and yield regioselectivities oppo-
site to those usually observed.
Here we wish to report on the use of porous meso organised mixed
zirconia silica powders for terminal alkynes and CO activation.
In a recent paper we described the use of hybrid bidentate
ligands (HBL) for tandem 2,3-dimethylbut-2-ene hydroformyla-
1
5
tion. HBL are hybrid materials bearing organic ligands grafted
near an oxide surface so that they can bind metal centres both via
their organic functions and via hydroxyl moieties close to the
Hydroformylation of alkenes is, despite the high costs of the
rhodium or cobalt catalysts employed, the great success story of
homogeneous catalysis, with 7 million tons of aldehydes produced
1
6
surface. These systems are thus bridging the gap between surface
17
organometallic chemistry and standard immobilised homoge-
1,2
every year.
On the contrary, alkynes hydroformylation is
1
3
restricted to the field of academic synthesis despite the value of
3–6
the resulting enals.
neous complexes. For example, in the case of dimethylbutene
hydroformylation, two processes have to be successively carried
out: the isomerisation which yields 2,3-dimethyl-but-1-ene and the
hydroformylation of the external double bond that finally yields
The reaction conditions required can
compare with those of alkenes hydroformylation but the
7
regioselectivities are usually poor, except when a function like
an ester or a thioester on the substrate can coordinate the catalyst
4
and force the regioselectivity. This restriction explains why
3,4-dimethylpentanal. Both steps are usually catalysed by the same
18,19
homogeneous rhodium complex.
We found the way to
industrial development of this process is still lacking. The
8
Pauson–Khand reaction which is closely related to this process,
heterogenise the catalyst and separate these steps by using a
rhodium complex grafted on a zirconia-rich mesoporous material.
This material was a mesoporous mixed zirconia silica powder
as it also involves alkynes and CO (and an alkene for a [2 + 2 + 1]
cyclisation), has been thoroughly studied and constitutes one of the
easiest ways to produce functionalised cyclopentenones (Scheme 1).
This transformation, especially in an intramolecular version, was
4 4
obtained by spray drying an ethanolic sol of ZrCl and SiCl in a
molar ratio of 4/1 with cetyltrimethylammonium bromide as
surfactant. After consolidation and extraction of the organic
9–11
widely used in total synthesis.
In contrast with hydroformyla-
template this powder, labelled ZS20 , exhibited a specific surface
C
2
21
15
˚
tion, this process is usually regioselective and catalytic intermole-
area of 230 m g and an average pore diameter of 20 A. The
Lewis acidic sites of the support were capable of isomerising the
alkene, feeding the rhodium complex with a terminal olefin.
Indeed, a reference run showed that an ungrafted zirconia material
yielded on its own 30% of 2,3-dimethylbut-1-ene under hydro-
formylation conditions.{ Importantly, 3% of the hydroformylation
product was surprisingly detected attesting that our support could
promote the hydroformylation process. This observation
prompted us to investigate the catalytic activity of this support
under hydroformylation conditions and 2-octene and styrene were
tested.{ As could be expected from the results on dimethylbutene,
1
0
cular still remains an important challenge. Its mechanism is not
12
totally understood and there are still many unexplained results.
Development of a versatile, heterogeneous, inexpensive and
selective catalyst for both reactions would, of course, be a
breakthrough. In this perspective mesoporous materials, which
were sometimes considered as large pore zeolites and have widely
13,14
been used as catalysts,
were supposed to be good candidates.
C
ZS20 acted as a 2-octene isomerisation and hydroformylation
catalyst yielding 71% of aldehydes, from which 31% were nonanal.
Though this performance does not compete with homogeneous
catalysts usually employed to convert internal olefins into linear
aldehydes, it confirms that ZS20
C
is an active hydroformylation
Scheme 1 General representation of a Pauson–Khand reaction.
catalyst. In the case of styrene, the conversion rate was modest
(
8%) but surprisingly the regioselectivity of the reaction was poor
a
Laboratoire de Chimie de la Mati e` re Condens e´ e, Universit e´ Pierre et
20
and opposite to which is usually observed for styrene: 53% of
Marie Curie, 75005 Paris, France. E-mail: clems@ccr.jussieu.fr;
Fax: (+33) 1-44-27-47-69
Laboratoire H e´ t e´ ro e´ l e´ ments et Coordination, Ecole polytechnique,
linear aldehyde vs 47% of the branched one were obtained. This
result underlined the fact that the mechanism involved was
presumably not the standard one for hydroformylations catalysed
by noble metals and that it could lead, through reversed
regioselectivity, to interesting products.
b
9
1128 Palaiseau, France. E-mail: lefloch@poly.polytechnique.fr;
Fax: (+33) 1-69-33-39-90
Electronic supplementary information (ESI) available: reaction condi-
tions. See DOI: 10.1039/b512543j
{
1
80 | Chem. Commun., 2006, 180–182
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Therefore we decided to concentrate on alkynes, where there
were fewer possible pathways. Three alkynes were tested under the
same reaction conditions: 36 h reaction time at 110 uC in methanol
under 30 bars of CO/H
zirconia. The results are summarised in Table 1.
The first conclusion of these tests was that ZS20
2
mixture in a 1/1 ratio and 1 mol% of
C
is a real
hydroformylation catalyst for both CLC double bonds and CMC
triple bonds. The conversion rates were low except with
phenylacetylene, but this could be circumvented by using harsher
experimental conditions. However, it is important to keep in mind
that this catalyst is the first example of a hydroformylation catalyst
not featuring expensive transition metals like rhodium.{
Furthermore it is important to note that the regioselectivity was
found to be very good and in the case of phenylacetylene opposite
to that expected (Scheme 2). These results give some insight into
the reaction mechanism. Only terminal alkynes react and the more
acidic the proton the higher the conversion rate (for example
Scheme 3 A possible pathway for the first step of phenylacetylene
C
hydroformylation catalysed by ZS20 .
zirconium site can insert into the C–Zr bond to form a very
reactive -ynalate species that can be finally cleaved with H to
produce the corresponding aldehyde.§ (Scheme 3)
phenylacetylene has a pK
a
of 23.2 in cyclohexylamine whereas tert-
2
21
butylacetylene has a pK
a
of 25.5 in the same solvent ). One
possible pathway to rationalise both regioselectivity and acidity
dependency would be a mechanism involving a deprotonation
2
assisted by g -coordination of the alkyne onto the support
If this mechanistic proposal is true, we expected our material to
be also active for catalysis of Pausond–Khand reactions. Moreover
it would yield cyclopentenones in which the acetylene moiety
would be located in b of the carbonyl group (such products will
now on be referred to as b products). This could be very interesting
throught its p-system. Indeed, C–C triple bonds are known to have
15,22
a good affinity for zirconium(IV) species.
In a second step, one
2
may propose that this g -coordination mode promotes a
cooperative deprotonation of the alkyne associated with water
elimination. Very likely, this step is facilitated by the Brønsted
basic character of zirconia, this oxide having a point of zero charge
of approximately 9. This process is probably governed by an
equilibrium, strongly displaced towards the coordination of the
alkyne. However if even a small part of the alkyne is deprotonated,
then a carbonyl ligand, coordinated onto a neighbouring
12
as those substituents are usually placed in a position. To test
these two hypotheses, phenylacetylene, 1-octyne, 4-octyne and
methyl butyndioate were reacted with a symmetric alkene in order
to limit the potential products, cyclohexene, in presence of 4 mol %
of zirconium at 130 uC under 30 bars of CO for 48 h in toluene.
The results are summarised in Table 2.
Table 2 Pauson–Khand reaction of various alkynes with cyclohexene
catalysed by ZS20
a
C
a
Table 1 Hydroformylation of various alkynes with ZS20
C
Conversion
rate (%)
Rate of
b-product (%)
b
b
b
b,c
Entry
Substrate
Entry Substrate
Conversion rate (%)
Linearity (%)
1
2
3
4
a
Phenylacetylene
1-Octyne
4-Octyne
20
12
0
100
100
0
1
2
3
a
Phenylacetylene 23
83
100
0
1-Octyne
4-Octyne
8
0
Methyl butyndioate
0
0
Reaction conditions: 36 h, at 110 uC in methanol under 30 bars of
CO/H mixture in 1/1 ratio and mol of zirconium.
Conversion rates and linearities were determined by GC and
The reaction conditions are: 48 h reaction time at 130 uC in
2
a
1
%
b
toluene under 30 bars of CO and 4 mol % of zirconium. The
b
1
conversion rates and linearities were determined by H NMR.
1
confirmed by H NMR so that the detection limit sits below 0.5%.
Linearity is the rate of linear aldehydes on the amount of
aldehydes formed.
c
Scheme 4 Pauson–Khand reaction of terminal alkynes with cyclohexene
Scheme 2 Hydroformylation of phenylacetylene catalysed by ZS20
C
.
C
catalysed by ZS20 .
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 180–182 | 181

View Article Online
§
2
How the H activation proceeds remains unclear, even if it can be
Results obtained for terminal alkynes were quite good regarding
the relatively mild experimental conditions employed and the low
admitted that dihydrogene is able to interact through s-bonding with
zirconia surface. Therefore we do not speculate on the following steps of
the catalytic mechanism.
23
catalyst loading for intermolecular Pauson–Khand reactions.
Such reactions have already been catalysed using group IV metal
complexes but our system proves to be the first heterogeneous
1 B. Breit and W. Seiche, Synthesis, 2001, 1.
2 S. Scott, L’actualit e´ chimique canadienne, 1999, 7, 16.
10
catalyst able to promote this transformation. Interestingly, we
found that internal alkynes even strongly activated ones like
methyl butyndioate, do not react, thus substantiating our
mechanism. Moreover the regioselectivity is totally in favour of
the b-products (Scheme 4). This observation, which strongly
supports the mechanism proposed, is rather unusual for this
process.
3
B. G. Van den Hoven, B. El Ali and H. Alper, J. Org. Chem., 2000, 65,
131.
4
4 B. G. Van den Hoven and H. Alper, J. Org. Chem., 1999, 64, 9640.
5 B. G. Van den Hoven and H. Alper, J. Org. Chem., 1999, 64, 3964.
6
7
8
9
Y. Ishii, K. Miyashita, K. Kamita and M. Hidai, J. Am. Chem. Soc.,
997, 119, 6448.
C. F. Huo, Y. W. Li, M. Beller and H. J. Jiao, Chem.–Eur. J., 2005, 11,
1
889.
I. U. Khand, G. R. Knox, P. L. Pauson and W. E. Watts, J. Chem. Soc.
D, 1971, 36a.
I. Nakamura and Y. Yamamoto, Chem. Rev., 2004, 104, 2127.
The catalytic activity of a mesoporous zirconia/silica mixed
powder, both in hydroformylation and Pauson–Khand processes,
opens, even if the activities are still low, a wide field of
investigation. The mechanism still remains unclear and requires
additional investigation. On the other hand, the observed
regioselectivities are rather unusual and thus interesting. We
believe that catalytic application of these mesoporous materials
could emphasise alkyne hydroformylation as the industrial
outcomes are considerable.
10 S. E. Gibson, S. E. Lewis and N. Mainolfi, J. Organomet. Chem., 2004,
689, 3873.
1 J. Blanco-Urgoiti, L. Anorbe, L. Perez-Serrano, G. Dominguez and
J. Perez-Castells, Chem. Soc. Rev., 2004, 33, 32.
1
1
2 L. V. R. Bonaga and M. E. Krafft, Tetrahedron, 2004, 60, 9795.
13 A. Taguchi and F. Schuth, Microporous Mesoporous Mater., 2005, 77, 1.
14 A. Corma, Chem. Rev., 1997, 97, 2373.
15 F. Goettmann, C. Boissi e` re, D. Grosso, F. Mercier, P. Le Floch and
C. Sanchez, Chem.–Eur. J., 2005, DOI: 10.1002/chem.200500542.
6 F. Goettmann, D. Grosso, F. Mercier, F. Mathey and C. Sanchez,
1
Chemical Communications, 2004, 1240.
7 C. Cop e´ ret, M. Chabanas, R. Petroff Saint-Arroman and J.-M. Basset,
Notes and references
1
{
These results could naturally be assigned to traces of rhodium from
Angew. Chem., Int. Ed., 2003, 42, 156.
18 A. Moores, N. M e´ zailles, L. Ricard and P. Le Floch, Organometallics,
2005, 24, 508.
19 B. Breit, R. Winde, T. Mackewitz, R. Paciello and K. Harms, Chem.–
Eur. J., 2001, 7, 3106.
20 B. Breit, R. Winde and K. Harms, J. Chem. Soc., Perkin Trans. 1, 1997,
18, 2681.
21 R. O. Hutchins, B. E. Maryanoff and C. A. Milewski, J. Am. Chem.
Soc., 1971, 93, 1794.
previous catalyses in the autoclaves or to late transition metals leaching
from the walls. When it was not possible to fit the autoclaves with a glass
vessel (See Supplementary Information{), blank tests without zirconia
powder were undertaken that did not afford any detectable conversion of
the reactants. However it is not possible to exclude that the catalytic activity
relies on traces of metal included in the silica/zirconia powders. Those
pollutants could possibly be iron(III) species coming from the spraying
system used for the synthesis of the powders or other metals included in
ZrCl . Such chemicals could not be detected in our powders by X-ray
4
fluorescence. But to definitely exclude such a possibility, more sensitive
techniques should be used.
22 T. Takahashi, Z. F. Xi, A. Yamazaki, Y. H. Liu, K. Nakajima and
M. Kotora, J. Am. Chem. Soc., 1998, 120, 1672.
23 S. E. Gibson and A. Stevenazzi, Angew. Chem., Int. Ed., 2003, 42, 1800.
1
82 | Chem. Commun., 2006, 180–182
This journal is ß The Royal Society of Chemistry 2006