From 2H-phospholes to BIPNOR, a new efficient biphosphine for asymmetric catalysis

Article abstract of DOI:10.1016/S0022-328X(97)00739-0

For many years now, we have studied the 1H-/2H-phosphole equilibrium and its synthetic applications. On reaction with alkynes, 2H-phospholes yield the corresponding 1-phosphanorbornadienes. As ligands of rhodium(I), these phosphines show some potential in catalytic hydrogenation and hydroformylation of alkenes. Starting from 3,3′,4,4′-tetramethyl-1,1′-biphospholyl and tolan, we have similarly obtained the corresponding 2,2′-bis-(1-phosphanorbornadienyl) (BIPNOR) with two chiral, non-racemisable, phosphorus atoms at the bridgeheads. The pure enantiomers of BIPNOR appear to be efficient ligands in asymmetric hydrogenation of C=C and C=O double bonds.

Full text of DOI:10.1016/S0022-328X(97)00739-0

Journal of Organometallic Chemistry 577 (1998) 117–120
From 2H-phospholes to BIPNOR, a new efficient biphosphine for
asymmetric catalysis
Fran c¸ ois Mathey *, Fran c¸ ois Mercier, Fr e´ d e´ ric Robin, Louis Ricard
Laboratoire ‘H e´ t e´ ro e´ l e´ ments et Coordination’, URA 1499 CNRS, DCPH, Ecole Polytechnique, 91128, Palaiseau Cedex, France
Received 19 May 1997
Abstract
For many years now, we have studied the 1H-/2H-phosphole equilibrium and its synthetic applications. On reaction with
alkynes, 2H-phospholes yield the corresponding 1-phosphanorbornadienes. As ligands of rhodium(I), these phosphines show some
potential in catalytic hydrogenation and hydroformylation of alkenes. Starting from 3,3%,4,4%-tetramethyl-1,1%-biphospholyl and
tolan, we have similarly obtained the corresponding 2,2%-bis-(1-phosphanorbornadienyl) (BIPNOR) with two chiral, non-racemis-
able, phosphorus atoms at the bridgeheads. The pure enantiomers of BIPNOR appear to be efficient ligands in asymmetric
hydrogenation of CꢀC and CꢀO double bonds. © 1998 Elsevier Science S.A. All rights reserved.
Keywords: [1,5] sigmatropic shifts; [4+2] cycloadditions; enantioselective catalysis
1. Introduction
tribute to the reactivity of phospholes under normal
circumstances.
While studying systematically the Diels–Alder reac-
tivity of phospholes, we discovered in 1981 the equi-
librium between 1H- and 2H-phospholes [1] (Scheme
2
. Reactivity of 2H-phospholes: a brief overview
1
).
From a practical standpoint, the most interesting
reactions of these unstable 2H-phospholes are those
where they act as very reactive 1-phosphadienes giving
4+2] cycloadducts with alkenes and alkynes [1,5]
The origin of this equilibrium lies in the |,y overlap
between the P–R exocyclic bond and the dienic system.
The migration of R depends heavily on the nature of R.
Hydrogen migrates below 0°C [2] whereas sp-carbon
[
(
Eqs. (1) and (2))
2
substituents migrate around 100°C and sp -carbon sub-
stituents around 140°C [3]. Silicon, phosphorus and
sulphur substituents also migrate. Oxygen does not
result from the high strength of the P–O bond.
Bachrach has shown that 1H-, 2H- and 3H-phospholes
are close in energy but the activation barrier between
−
1
1
H and 2H is low (DH 16 kcal mol ) whereas it is
−
1
high between 2H and 3H (DH 27 kcal mol ) [4].
These data explain why 3H-phospholes do not con-
and those where they act as sources of aromatic phos-
5
pholide ions and p -phospholyl complexes. In the latter
case, the most recent applications concern the first
*
Corresponding author. Tel.: +33 1 69334079; fax: +33 1
6
9333960.
practical synthesis of h-functional phospholide ions [6]
0
022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(97)00739-0

1
18
F. Mathey et al. / Journal of Organometallic Chemistry 557 (1998) 117–120
Scheme 1. The 1H-l2H-phosphole equilibrium.
(
Eq. 3) and the straightforward preparation of a chelat-
More recently, Herrmann et al. [10] prepared a
sulphonated version of the phosphanorbornadiene of
Eq. 5. This water-soluble species shows exceptional
quality in the biphasic hydroformylation of propene. In
such a context, looking for possible applications of
homochiral 1-phosphanorbornadienes in asymmetric
catalysis was an obvious choice.
ing 2-(2%-pyridyl)phosphaferrocene [7] (Eq. 4).
Most of the homochiral bisphosphines currently in
use in enantioselective catalysis carry the chiral in-
formation on their carbon backbone (BINAP [11],
DuPHOS [12]...). With the only notable exception of
DIPAMP [13], the chiral information is seldom car-
ried by the phosphorus centre. Such a situation is
nevertheless intuitively attractive since in that case,
the transfer of chirality from the inductor to the
product is favoured by their close proximity within
the coordination sphere of the transition metal. The
main drawback is that P-homochiral phosphines
tend to racemise by a variety of routes including the
pyramidal inversion, the Berry pseudorotation and
the edge inversion [14]. Berry pseudorotation can
easily occur in protic media whereas edge inversion
does not necessitate the decoordination of phospho-
rus from the transition metal. In 1-phosphanorbor-
nadienes, the phosphorus atom is located at the
bridgehead position of a bicyclic system and none of
the racemisation pathways can operate. From that
standpoint, 1-phosphanorbornadienes are definitively
superior to any other classical phosphines. We first
investigated the synthesis of a monodentate ho-
mochiral species. The reaction of ethyl phenylpropio-
late with a 2H-phosphole afforded the corresponding
h- and i-functional phosphanorbornadienes whose
separation was performed on silica gel. The two h-
enantiomers were separated on a chiral phase as
their P-oxides [15] (Eq. 7).
However, the outcome of this chemistry in homoge-
neous catalysis is exclusively centered around the robust
and easily accessible 1-phosphanorbornadienes (Eq. 2).
3. From 1-phosphanorbornadienes to BIPNOR
These new 1-phosphanorbornadienes rapidly proved
to be excellent ligands of rhodium for the catalytic
hydrogenation of functional alkenes [8] (Eq. 5),
and the catalytic hydroformylation of terminal alkenes
[
9] (Eq. 6).

F. Mathey et al. / Journal of Organometallic Chemistry 557 (1998) 117–120
119
A 50% ee was observed in the rhodium-catalyzed
asymmetric hydrogenation of typical dehy-
a
droaminoacid. The next step was the discovery of a
simple access to a 2,2%-bis-(1-phosphanorbornadienyl)
(
BIPNOR)[16] (Eq. 8).
Scheme 2. Asymmetric hydrogenation of functional alkenes with
Rh(I)-BIPNOR catalysts.
The key step of this synthesis is a double [1,5] shift of
each phosphole around the other ring. The meso and rac
diastereomers were separated by chromatography on
silica gel as their PdCl complexes. The resolution of the
2
racemic mixture was carried out by coordination with an
optically active palladacycle and chromatographic sepa-
ration of the resulting diastereomers [17] (Eq. 9).
Scheme 3. Asymmetric hydrogenation of ketones with Ru(II)-BIP-
NOR catalysts.
As cationic Rh(I) complexes, these pure enan-
tiomers display high efficiency in the enantioselective
hydrogenation of functional alkenes (Scheme 2).
The observed ee’s are in the same range as those
obtained with BINAP- and DuPHOS-based catalysts.
The absolute configurations were established by X-ray
crystal structure analysis of one of the diastereomeric
palladium complexes. The free R,R and S,S-BIPNOR
were recovered from their complexes by treatment with
sodium cyanide (Eq. 10).

1
20
F. Mathey et al. / Journal of Organometallic Chemistry 557 (1998) 117–120
+
−
Fig. 1. View of the [Rh(−)-BIPNOR] moiety of [Rh(cod)(−)-BIPNOR] PF6 from the P RhP plane.
1
10
As Ru(II) complexes, the enantiomers of BIPNOR
appear to be less efficient than the BINAP-based cata-
lysts in the asymmetric hydrogenation of ketones
References
[
[
[
1] F. Mathey, F. Mercier, C. Charrier, J. Fischer, A. Mitschler, J.
Am. Chem. Soc. 103 (1981) 4595.
2] C. Charrier, H. Bonnard, G. de Lauzon, F. Mathey, J. Am.
Chem. Soc. 105 (1983) 6871.
3] F. Laporte, F. Mercier, L. Ricard, F. Mathey, Bull. Soc. Chim.
Fr. 130 (1993) 843.
(
Scheme 3), although the optimisation of the experi-
mental conditions has not been carried out yet in that
case.
It seems that the BIPNOR geometry is especially well
suited to the transfer of chirality in a square–planar
Rh(I) complex as shown in Fig. 1. The [Rh(−)-BIP-
NOR] moiety occupies two quadrants of the space
around Rh and leaves the two others free. This situa-
tion induces an optimum face-selection during the com-
plexation of the incoming prochiral alkene as discussed
by Knowles [13]. More work is currently being done in
order to define the scope of BIPNOR in enantioselec-
tive catalysis.
[
[
[
4] S.M. Bachrach, J. Org. Chem. 58 (1993) 5414.
5] P. Le Goff, F. Mathey, L. Ricard, J. Org. Chem. 54 (1989) 4754.
6] S. Holand, M. Jeanjean, F. Mathey, Angew. Chem. Int. Ed.
Engl. 36 (1997) 98.
[7] B. Deschamps, L. Ricard, F. Mathey, J. Organomet. Chem., in
press.
[
8] F. Mathey, D. Neibecker, A. Br e` que (SNPE), Fr. Pat. 25 88 197,
Oct. 3, 1985. Chem Abstr. 107 (1987) 219468v.
[
9] D. Neibecker, R. Reau, Angew. Chem. Int. Ed. Engl. 28 (1989)
5
00.
[10] W.A. Herrmann, C.W. Kohlpaintner, R.B. Manatsberger, H.
Bahrmann, H. Kottmann, J. Mol. Catal. A 97 (1995) 65.
[
[
[
[
11] R. Noyori, H. Takaya, Acc. Chem. Res. 23 (1990) 345.
12] M.J. Burk, J. Am. Chem. Soc. 113 (1991) 8518.
13] W.S. Knowles, Acc. Chem. Res. 16 (1983) 106.
14] A.J. Arduengo III, A.D. Dixon, D.C. Roe, J. Am. Chem. Soc.
108 (1986) 6821.
Acknowledgements
The authors thank Dr. Michel Spagnol (RP) for his
help in the evaluation of the performances of BIPNOR
in asymmetric catalysis. The work on BIPNOR was
supported by several contracts of Rh oˆ ne-Poulenc. The
previous work on the chemistry of 2H-phospholes was
supported by CNRS and Ecole Polytechnique.
[
[
[
15] A. Br e` que, J.-M. Alcaraz, L. Ricard, F. Mathey, New J. Chem.
1
2 (1989) 369.
16] M.-O. Bevierre, F. Mercier, L. Ricard, F. Mathey, Bull. Soc.
Chim. Fr. 129 (1992) 1.
17] F. Robin, F. Mercier, L. Ricard, F. Mathey, M. Spagnol, Chem.
Eur. J. 3 (1997) 1.
.