Synthesis of 1,1'-bis(phosphetano)ferrocenes, a new class of chiral ligands for asymmetric catalysis

Article abstract of DOI:10.1055/s-1999-2968

The C₂-symmetric, chiral 1,1'-bis(phosphetano)ferrocenes 1 have been prepared from the cyclic sulfates of optically pure, 1,3-diols. They have been tested in the rhodium catalysed hydrogenation of unsaturated substrates.

Full text of DOI:10.1055/s-1999-2968

LETTER
1975
Synthesis of 1,1’-Bis(phosphetano)ferrocenes, a New Class of Chiral Ligands
for Asymmetric Catalysis
Angela Marinetti, Francis Labrue, Jean-Pierre Genêt
Laboratoire de Synthèse Sélective Organique et Produits Naturels. UMR CNRS 7573, ENSCP - 11, rue P. et M. Curie,
75231 Paris Cedex 05, France
Fax (33) 01 44 07 10 62; E-mail: marinet@ext.jussieu.fr
Received 8 September 1999
Abstract: The C2-symmetric, chiral 1,1'-bis(phosphetano)fer-
rocenes 1 have been prepared from the cyclic sulfates of optically
pure, 1,3-diols. They have been tested in the rhodium catalysed hy-
drogenation of unsaturated substrates.
Key words: chiral phosphetanes, ferrocene, rhodium, hydrogena-
tion
Nowadays, many successful chiral ligand designs, and es-
pecially design of chiral phosphines, take advantage from
the peculiar properties of the ferrocenyl moiety:1 the sig-
nificant sterical hindrance of the ferrocenyl group oper-
ates in planar-chiral species, while large bite angles and
restricted flexibility modulate the reactivity of the transi-
tion metal complexes of 1,1’-bis(phosphino)ferrocenes.
In C2 symmetric 1,1’-bis(phosphino)ferrocenes, chirality
is brought about by either disymmetric disubstitution of
both cyclopentadienyl rings (planar chirality)2 or by chiral
phosphorus substituents.3 The 1,1’-bis(phosphetano)fer-
rocenes 1 described hereafter, belong to this last class of
compounds, insofar as their chirality is afforded by opti-
cally pure, 2,4-disubstituted phosphetane units. They rep-
resent a new class of chiral phosphines and complement
our previous work on phosphetane based ligands. 4
Reagents and conditions: i. (a) n-BuLi (2.2 eq.), THF, -78 °C / 25 °C;
(b) cyclic sulfate 3 (2 eq.), -78 °C / 25 °C, 1h; (c) s-BuLi (2.2 eq.), -
78 °C / 25 °C, 1h. ii. BH3. SMe2, 25 °C, 10 min. iii. DABCO, 40 °C,
3h, benzene.
Scheme
Addition of 2.2 equivalents of s-BuLi led to deprotonation
of the remaining PH functions and final formation of the
ring-closure product 1a (or 1b). Complexation of the
phosphorus atoms with BH3 prevented air oxidation dur-
ing the purification steps and the protecting group was
easily removed. The phosphetane-borane complexes 28
were separated from the side products, including the 1-
(phosphetano)ferrocene-BH3 complexes, by chromatog-
raphy on alumina column, with a cyclohexane-dichlo-
romethane 1:1 mixture as eluent (Yield 50% on a 0.1 g
scale). After quantitative decomplexation of 1 by a phos-
phine–amine exchange reaction (excess DABCO, 40 °C,
3h, benzene), the excess amine was removed from the
crude reaction mixture by washing with aqueous HCl.
Phosphetanes 1a and 1b9 were crystallised from pentane.
The new chiral bis(phosphetanes) proved to be relatively
air stable, thus the above borane complexation could be
likely avoided. The experimental procedure is not fully
optimised to date.
Previous work from our group has shown that chiral 2,4-
disubstituted phosphetane rings are easily accessible from
optically pure anti-1,3-diol derivatives by reaction with
primary phosphines.4a,c The same approach has been ap-
plied here to the synthesis of the bis(phosphetanes) 1a and
1b from 1,1’-bis(phosphino)ferrocene. 5
1,1’-Bis(phosphino)ferrocene has been prepared from fer-
rocene, through a two-steps procedure, which requires the
synthesis and subsequent reduction of tetraethyl 1,1’-(fer-
rocenyl)bisphosphonite.5 In our hands this method afford-
ed 1,1’-bis(phosphino)ferrocene, contaminated with
small amounts of the starting ferrocene and of the mono-
phosphino substituted ferrocene. The phosphetane syn-
theses were performed on the crude mixture. An alterna-
tive procedure includes purification of the intermediate
tetraethyl 1,1’-(ferrocenyl)bisphosphonite by column
chromatography, after complexation of the phosphorus
atoms with BH3.7
Preparations of la and 1b are representative examples of a
general synthetic approach which could be extended to
many other bis(phosphetano)ferrocenes with different R
substituents, starting from the corresponding optically
Deprotonation of 1,1’-bis(phosphino)ferrocene with 2.2
equivalents of n-BuLi afforded the corresponding dianion
which was then treated with the cyclic sulfate of (R,R)-
2,4-pentanediol (or (R,R)-2,6-dimethyl-3,5-heptanediol).
Synlett 1999, No. 12, 1975–1977 ISSN 0936-5214 © Thieme Stuttgart · New York

1976
A. Marinetti et al.
LETTER
pure 1,3-diols. For instance, the pure cyclic sulfates 3 with there are no guidelines to predict the efficiency of these
R = Et, t-Bu, cyclohexyl, CH2Ph have already been pre- and analogous chiral ligands as a function of the R substit-
pared and applied to phosphetane syntheses.4c,d Further uent, a large number of tests is still required to establish
developments are in progress.
the potential of phosphetanes 1.
In order to get insight into the potential of ligands 1 in More detailed studies are in progress. Investigations will
asymmetric catalysis, we considered a few rhodium pro- be extended also to other fields of organometallic cataly-
moted hydrogenation reactions. Our first concern was in sis, where the specific properties of the ferrocene back-
fact to compare these phosphetane-based ligands with the bone should emerge.
analogous 1,1’-bis(phospholano)ferrocenes described in
ref. 5, which showed high catalytic activity and moderate
In summary, the first members of a new class of optically
pure phosphetanes have been prepared, characterised and
enantioselectivities in the hydrogenation of selected ole-
successfully tested in the enantioselective rhodium-catal-
fins and carbonyl groups. Thus, we performed the hydro-
ysed hydrogenations of model substrates.10 Studies are
genation of the same substrates under analogous
under way to optimise the synthetic approach and to find
conditions, by using a rhodium catalyst formed in situ
significant catalytic applications.
from (COD)2RhOTf and the chiral ligands 1a or 1b. The
results are given in Table 1.
References and Notes
(1) For recent reviews see: Ferrocenes. Togni, A. and Hayashi, T.
Eds, VCH, Weinheim 1995. Richards, C.J.; Locke, A.J.
Tetrahedron: Asymmetry 1998, 9, 2377.
(2) Selected examples: (a) Hayashi, T.; Yamamoto, A.; Hojo, M.;
Kishi, K.; Ito, Y., Nishioka, E.; Miura, H.; Yanagi, K. J.
Organomet. Chem. 1989, 370, 129. (b) Zhang, W.; Hirao, T.;
Ikeda, I. Tetrahedron Lett. 1996, 37, 4545. (c) Kang, J.; Lee,
J.H.; Ahn, S.H.; Choi, J.S. Tetrahedron Lett. 1998, 39, 5523.
(d) Schwink, L.; Knochel, P. Chem. Eur. J. 1998, 4, 950.;
(e) Perea, J.J.A.; Lotz, M.; Knochel, P. Tetrahedron:
Asymmetry 1999, 10, 375. (f) Reetz, M.T.; Beuttenmüller,
E.W.; Goddard, R.; Pasto, M. Tetrahedron Lett. 1999, 40,
4977.
(3) Selected examples : Maienza, F.; Wörle, M.; Steffanut, P.;
Mezzetti, A.; Spindler, F. Organometallics 1999, 18, 1041.
Nettekoven, U.; Kamer, P.C.J.; van Leeuwen, P.W.N.M.;
Widhalm, M.; Spek, A.L.; Lutz, M. J. Org. Chem. 1999, 64,
3996. Brunner, H.; Janura, M. Synthesis 1998, 45.
(4) (a) Marinetti, A. Kruger, V.; Buzin, F.-X. Tetrahedron Lett.
1997, 38, 2947. (b) Marinetti, A.; Kruger, V.; Buzin, F.-X.
Coord. Chem. Rev. 1998, 178, 755. (c) Marinetti, A.; Genêt,
J.-P.; Jus, S.; Blanc, D.; Ratovelomanana-Vidal, V. Chem.
Eur. J. 1999, 5, 1160. (d) Marinetti, A.; Jus, S.; Genêt, J.-P.;
Ricard, L. Tetrahedron, in press.
Concerning the catalytic activity of these phosphetane-
rhodium catalysts, we have not yet any reliable informa-
tion, because quite long reaction times have been applied
to ensure total conversion.
(5) Burk, M.J.; Gross, M.F. Tetrahedron Lett. 1994, 35, 9363.
(6) (R,R)-2,4-pentanediol and (R,R)-2,6-dimethyl-3,5-
heptanediol were conveniently prepared by ruthenium-
BINAP catalysed hydrogenations (see ref. 4c and references
therein).
(7) Tetraethyl 1,1’-(ferrocenyl)bisphosphonite bis-borane
complex:1H NMR (CDCl3): d 1.30 (t, JH-H = 7.1 Hz, Me), 3.9-
4.2 (m, CH2), 4.62 (CH), 4.63 (CH) ppm.
(8) The experimental procedure previously applied to the
preparation of other phosphetane-borane complexes4c,d has
been used here for the synthesis of 2. (S,S)-2a, 1H NMR
(CDCl3): d 1.00 (dd, JH-P = 15.8 Hz, JH-H = 7.4 Hz, Me), 1.46
(dd, JH-P = 18.8 Hz, JH-H = 7.4 Hz, Me), 2.2-2.4 (m, 4H), 2.6-
2.7 (m, 2H), 2.7-2.9 (m, 2H), 4.5 (m, 2H, Cp), 4.6 (m, 2H, Cp),
4.7 (m, 4H, Cp); 13C NMR (CDCl3): d 15.8 (Me), 15.9 (Me),
28.0 (1JC-P = 42.3 Hz, PCH), 29.2 (1JC-P = 41.0 Hz, PCH), 35.9
(2JC-P = 16.1 Hz, CH2), 67.7 (JC-P = 38.3 Hz, PC), 72.0 (JC-
P = 3.2 Hz, CH), 74.0 (JC-P = 5.5 Hz, CH), 74.7 (JC-P = 7.5 Hz,
CH), 75.8 (JC-P = 13.3 Hz, CH); 31P NMR (CDCl3): d 49 ppm.
MS (DCI/NH3) m/e 432 (M+NH4). [a]D = -35 (c = 0.5,
CH2Cl2). Anal. Calcd. For C20H34B2P2Fe: C, 58.04; H, 8.28.
Found: C, 58.02; H, 8.29. (S,S)-2b, 1H NMR (C6D6): d 0.51
(dd, JH-H = 6.6 Hz, JH-P = 0.7 Hz, Me), 0.72 (d, J = 6.5 Hz, Me),
0.90 (dd, JH-H = 6.6 Hz, JH-P = 0.5 Hz, Me), 1.1 (m, 2H), 1.32
As shown in Table 1, the enantioselectivity levels are
highly dependent on the nature of the R-substituent on the
phosphetane rings, however they are at least comparable
(entry 6) and sometimes higher (entries 1 and 4) than
those obtained with the phospholane analogues. 5 Thus, it
appears that in the series of 1,1'-disubstituted ferrocenes,
chiral phosphetane moieties can perform better than the
corresponding phospholanes in the enantioselective rhod-
ium-catalysed hydrogenations. On the whole, the prelim-
inary results reported in Table 1 are very promising, when
considering that they were obtained under non-optimised
conditions.
Moreover, it must be emphasized that many, variously
substituted phosphetanes 1 should be accessible. Thus, it
is likely that a “matching ligand” could be found for each
substrate, which would optimise both the catalytic activity
and stereoselectivity in the hydrogenation reactions. As
Synlett 1999, No. 12, 1975–1977 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Synthesis of 1,1’-Bis(phosphetano)ferrocenes, a New Class of Chiral Ligands
1977
(d, J = 6.4 Hz, Me), 1.9-2.1 (m, 10H), 4.3 (m, 2H, Cp), 4.5 (m,
phosphorus atoms): d 15.8 (t, JC-P = 1.5 Hz, Me), 19.3 (m, Me),
22.0 (dd, JC-P = 5.3 and 2.3 Hz, PCH), 22.8 (dd, JC-P = 9.2 and
3.0 Hz, PCH), 38.6 (CH2), 69.9 (t, JC-P = 2.3 Hz, CH), 70.1
(CH), 71.0 (t, JC-P = 3.8 Hz, CH), 74.4 (m, CP), 75.5 (m, CH);
31P NMR (C6D6): d 19 ppm. MS (DCI/NH3) m/e 387 (M+1).
(S,S)-1b, 1H NMR (C6D6): d 0.66 (d, J = 6.5 Hz, Me), 0.76 (d,
J = 6.4 Hz, Me), 1.08 (d, J = 6.4 Hz, Me), 1.30 (d, J = 6.4 Hz,
Me), 1.4 (m, 2H), 1.7-1.9 (m, 6H), 2.5-2.7 (m, 4H), 4.34 (m,
6H, Cp), 4.45 (m, 2H, Cp); 13C NMR (C6D6) d 15.5 (Me), 19.4
(Me), 21.1 (JC-P = 9.9 Hz, Me), 21.7 (JC-P = 13.6 Hz, Me), 31.4
(CH), 31.6 (JC-P = 19.3 CH), 35.1 (CH2), 35.9 (JC-P = 9.9 Hz,
CH), 37.3 (JC-P = 7.8 Hz, CH), 70.9, 71.6, 72.7 (CH), 75.6 (JC-
P = 32.5 Hz, PC), 78.0 (JC-P = 33.4 Hz, CH); 31P NMR (C6D6):
d 13 ppm.
2H, Cp), 4.7 (m, 2H, Cp), 4.8 (m, 2H, Cp); 13C NMR (C6D6):
d 19.8 (3JC-P = 12.2 Hz, Me), 21.0, 21.1, 21.2, 21.3 (2 Me),
23.0 (3JC-P = 4.7 Hz, Me), 29.5 (2JC-P = 2.7 Hz, CH), 30.7 (2JC-
P = 4.7 Hz, CH), 31.4 (2JC-P = 16.9 Hz, CH2), 40.4 (1JC-P = 38.5
Hz, PCH), 42.2 (1JC-P = 39.8 Hz, PCH), 68.9 (1JC-P = 40.5 Hz,
PC), 71.0 (CH), 74.4 (JC-P = 5.4 Hz, CH), 74.9 (JC-P = 7.4 Hz,
CH), 76.3 (JC-P = 14.2 Hz, CH); 31P NMR (C6D6): d 50 ppm.
Anal. Calcd. For C28H50B2P2Fe: C, 63.92; H, 9.58. Found: C,
62.96; H, 9.62.
(9) Decomplexation procedure: The phosphetane-borane
complex 2 (1mmol) was heated with DABCO (4 mmol) in
benzene (2 mL), under argon, at 40 °C for 3 h. 31P NMR
analysis of the crude reaction mixture showed quantitative
formation of 1. The benzene solution was washed twice with
degassed aqueous HCl (5%), dried over MgSO4, filtered and
concentrated. The residue was crystallised from pentane to
afford pure 1 in 80% yield. (S,S)-1a, 1H NMR (C6D6): d 0.83
(dd, J = 8.5 and 7.0 Hz, Me), 1.47 (dd, JH-P = 17.7 Hz, JH-
H = 7.2 Hz, Me), 2.2-2.3 (m, 6H), 2.4-2.5 (m, 2H), 4.2 (m, 6H,
Cp), 4.3 (m, 2H, Cp); 13C NMR (C6D6) (« m » indicates a
complex pattern due to the virtual coupling of the two
(10) Preliminary communication of this work has been made at the
10th IUPAC Symposium on Organometallic Chemistry
Directed Towards Organic Synthesis (OMCOS 10),
Versailles (France), 18-22 July 1999.
Article Identifier:
1437-2096,E;1999,0,12,1975,1977,ftx,en;G22499ST.pdf
Synlett 1999, No. 12, 1975–1977 ISSN 0936-5214 © Thieme Stuttgart · New York