Hybrid P-chiral diphosphines for asymmetric hydrogenation

Article abstract of DOI:10.1039/a808711c

A family of diphosphine ligands has been prepared by Michael addition of o-anisylphenyl phosphide to diethyl vinylphosphonate and elaboration to phospholanes based on hexane-2,5-diol or mannitol; some preliminary results of Rh-complex catalysed hydrogenations are reported.

Full text of DOI:10.1039/a808711c

Hybrid P-chiral diphosphines for asymmetric hydrogenation
Duncan Carmichael, Henri Doucet and John M. Brown*
Dyson Perrins Laboratory, South Parks Road, Oxford, UK OX1 3QY. E-mail: bjm@ermine.ox.ac.uk
Received (in Liverpool, UK) 6th November 1998, Accepted 8th December 1998
Me
Me
A family of diphosphine ligands has been prepared by
Ph
Ph
P
Michael addition of o-anisylphenyl phosphide to diethyl
vinylphosphonate and elaboration to phospholanes based on
hexane-2,5-diol or mannitol; some preliminary results of Rh-
complex catalysed hydrogenations are reported.
OMe
OMe
P
+
P
P
Me
Me
BH3
BH3
H3B
H3B
7a (R,R,RP)
8a (R,R,SP)
The scope of asymmetric hydrogenation of alkenes has been
gradually extended both in reactant structure and catalyst
efficiency over many years.1 For rhodium catalysis, the early
work of the Monsanto group using the P-chiral diphosphine
DIPAMP has provided an enduring standard.2 Only phospho-
lane ligands in the DUPHOS and BPE series have exhibited
superior enantioselectivity over a broad front.3
(iii)
Ph
P
OMe
R
H3B
(±)-3 R = H
(±)-4 R = (EtO)2P(O)CH2CH2
(±)-5 R = H2PCH2CH2
(i)
(ii)
(iv)
Me
RO
Me
RO
Me
Me
Ph
P
OMe
Me
P
Ph
P
Ph
P
OMe
OMe
RO
RO
P
P
P
P
+
Me
MeO
Ph
Me
BH3
BH3
H3B
Me
H3B
1
(R,R)-DIPAMP
2 (S,S)-MeBPE
11a-OH (S,S,S,S,RP)
10a-OH (S,S,S,S,SP)
[R = H]
[R = H]
It is a commonly held belief that C2 symmetric diphosphine
(or diol or diamine) ligands are endowed with superior
properties in catalysis, their attractiveness augmented by ease of
synthesis.4 An alternative view, stated most clearly by Achiwa
and co-workers,5 is that intermediates in a catalytic cycle lack
the intrinsic symmetry of the ligand and consequently the two
chelating atoms must fulfil different roles. The implication is
that lack of symmetry may be a positive advantage in an
appropriate case. In previous work we have endeavoured to
separate the functions of the two chelating phosphorus atoms in
asymmetric hydrogenation.6 The synthesis of a new class of
unsymmetrical ligands permits that approach to be extended
further.
Our basic idea was to combine the phosphorus moieties of
DIPAMP 1 and BPE 2 in a single ligand. The synthesis is based
on conjugate addition of racemic phosphineborane 37 to diethyl
vinylphosphonate (Scheme 1).8 Alane reduction of the product
4 gave the primary phosphineborane 5. Following deborona-
tion, stepwise double nucleophilic displacement on the cyclic
sulfate 69 via BuLi deprotonation permitted synthesis of the
target compounds 7a and 8a as a diastereomeric mixture in good
yield. These were separated by MPLC, with some difficulty
(EtOAc–pentane). The analogous compounds 10a-OH and
11a-OH, prepared from the mannitol derivative 9,10 proved
much more amenable to chromatographic separation, and
subsequently afforded the pure methyl ethers 10a-OMe and
11a-OMe.11 The absolute configuration of product boranes was
established by CD in comparison with that of the diborane from
(S,S)-DIPAMP, the phospholane part being essentially CD
transparent in the 240-400 nm region. This set of procedures
gives access to a family of unsymmetrical 1,2-phosphinoethane
ligands as their stable diborane complexes.12
Methylation :
10a-OH, 11a-OH
(v)
10a-OMe, 11a-OMe
(vi)
Deboronation :
7a, 8a,10a-OH, 10a-OMe,
7b, 8b, 10b-OH, 10b-OMe,
11a-OH, 11a-OMe
11b-OH, 11b-OMe
Me
Me
H
O
O
O
O
Me
Me
SO2
SO2
O
O
H
Me
Me
6
9
Scheme 1 Reagents: (i) CH2NCHP(O)(OEt)2, KOBut, THF, 95%; (ii)
DABCO, C7H8; AlH3, Et2O; H2O then CaH2; (iii) BuLi, THF, 278 °C then
6 then further BuLi; Me2S•BH3, 45% overall; (iv) BuLi, 278 °C then 9;
then repeat; Me2S•BH3, 30% isolated overall for (ii), (iv); (v) NaH, MeI,
THF, !80%; (vi) HBF4•OMe2 then NaHCO3.
match or mismatch of arylphosphine and phospholane chirality
have a significant effect on enantioselectivity? The results
recorded in Table 1 provide answers to these points but also
some surprises. A broad conclusion from these and parallel
results is that the ‘matched’ ligand14 gives significantly higher
ees than the ‘mismatched’ ligand, and also that the phospholane
configuration is dominant in defining the stereochemical course
of hydrogenation, but to an extent that depends on the substrate.
In the case of the bulkier pivalamide 13 and benzamide 14, the
configuration of the arylphosphine part plays a very minor role.
In comparison to the mannitol derived ligands 11, the simple
phospholanes 7b and 8b gave significantly inferior results in
these and related cases and were thus investigated in less
detail.
In most cases hydrogenation experiments were carried out by
in situ deboronation13 and reaction with (COD)2RhBF4 to
generate the catalyst. In initial studies of the hydrogenation of
simple dehydroamino acids and esters, two questions were
posed: Is the enantioselectivity governed predominantly by one
of the two phosphorus nuclei in the ligand? Does the alternative
Further results of interest came from a study of the itaconate
esters and half-esters recorded in Table 2. Here the mismatched
diastereomers of ligand 10b gave poor ees and are not included.
For the 1-substituted monoester 15, the hydroxy ligand 11b-OH
gives a superior ee to its methyl ether. The reverse is true for the
3-substituted monoester 16, where the methyl ether 11b-OMe
Chem. Commun., 1999, 261–262
261

Table 1
potential for rational structural alteration, and the impetus of
additional synthetic power and mechanistic information arising
from the distinct role of the two ligating atoms. The results
nicely complement those of Börner and co-workers on asym-
metric hydrogenation with mannitol-derived diphospho-
lanes.16
We thank EPSRC, DTI and Chiroscience (Dr Ulrich Berens),
Glaxo-Wellcome (Dr Andrew Payne), Robinson Bros. (Dr
Kelvyn Soars), SB (Dr Peter Sheldrake) and Zeneca FCMO (Dr
A. John Blacker) for support under the LINK Asymmetric
Synthesis Programme. We thank CNRS for support for leave of
absence (to D. C.). Johnson-Matthey kindly provided a loan of
RhCl3·3H2O and we thank Dr Ulrich Berens (Chirotech) for a
generous sample of the enantiomerically pure diol. We thank Dr
A. Boerner (Rostock) for a useful exchange of information (ref.
16).
Substrate
Catalyst precursor
Ee (%)
H
Ph
N
10b-OMe
11b-OMe
10b-OH
19 S
85 S
43 S
92 S
38 S
60 R
O
MeO
11b-OH
7b
Me
But
Ph
O
H
8b
H
12
Ph
58 S
67 S
82 S
88 S
10b-OMe
11b-OMe
10b-OH
11b-OH
7b
O
MeO
N
5
R
O
H
36 R
8b
H
13
Ph
O
77 S
72 S
90 Sa
11b-OMe
10b-OH
11b-OH
MeO
O
N
H
14
Conditions: substrate:catalyst 100:1, (COD)2RhBF4 as precursor
(HBF4·OMe2 deboronation in situ), 1.3 bar, MeOH, 1–3 h. a TfO2 instead
Notes and references
2
of BF4
.
1 H. Takaya, T. Ohta and R. Noyori, Catalytic Asymmetric Synthesis,
VCH, Weinheim, 1993, ch. 1; R. Noyori, Asymmetric Catalysis in
Organic Synthesis, Wiley, NY, 1994, ch. 2; A. Pfaltz, Houben-Weyl
Methods of Organic Chemistry, Vol. E21d 2.5.1.2, ed. G. Helmchen,
R. W. Hoffmann, J. Mulzer and E. Schaumann, Thieme, Stuttgart,
1995.
Me
Me
Ph
P
Ph
P
OMe
OMe
P
P
Me
Me
7b (R,R,RP)
8b (R,R,SP)
2 B. D. Vineyard, W. S. Knowles, M. J. Sabacky, G. L. Bachman and D. J.
Weinkauff, J. Am. Chem. Soc., 1977, 99, 5946.
RO
Me
RO
Me
P
Me
3 M. J. Burk, F. Bienewald, M. Harris and A. Zanotti-Gerosa, Angew.
Chem., Int. Ed., 1998, 37, 1931; M. J. Burk, C. S. Kalberg and A.
Pizzano, J. Am. Chem. Soc., 1998, 120, 4345, and earlier references.
4 J. K. Whitesell, Chem. Rev., 1989, 89, 1581; B. M. Trost and D. L. Van
Vranken, Angew. Chem., Int. Ed. Engl., 1992, 31, 228.
5 K. Inoguchi, S. Sakuraba and K. Achiwa, Synlett, 1992, 169; T.
Morimoto, M. Chiba and K. Achiwa, Chem. Pharm. Bull., 1993, 41,
1149; T. V. RajanBabu and A. L. Casalnuovo, J. Am. Chem. Soc., 1996,
118, 6325.
6 J. A. Ramsden, J. M. Brown, M. B. Hursthouse and A. I. Karalulov
Tetrahedron: Asymmetry, 1994, 5, 2033; J. A. Ramsden, T. Claridge and
J. M. Brown, J. Chem. Soc., Chem. Commun., 1995, 2469.
7 T. Imamoto, T. Oshiki, T. Onozawa, T. Kusumoto and K. Sato, J. Am.
Chem. Soc., 1990, 112, 5244; T. Imamoto, T. Yoshizawa, K. Hirose, Y.
Wada, H. Masuda, K. Yamaguchi and H. Seki, Heteroatom Chem.,
1995, 6, 99; M. Ohff, J. Holz, M. Quirmbach and A. Börner, Synthesis,
1998, 1391.
8 C.f. K. Bourumeau, A.-C. Gaumont and J.-M. Denis, Tetrahedron Lett.,
1997, 38, 1923.
9 M. J. Burk, J. E. Feaster, W. A. Nugent and R. L. Harlow J. Am. Chem.
Soc., 1993, 115, 10 125.
10 In 90% yield from the corresponding diol, L. F. Wiggins and D. J. C.
Wood, J. Chem. Soc., 1950, 1566; Y. Le Merrer, A Dureeault, C. Greck,
D. Micas-Languin, G. Gravier and J. C. Depezay, Heterocycles, 1983,
25, 541.
Ph
P
Ph
P
OMe
OMe
RO
RO
P
Me
10b-OH [R = H]
10b-OMe [R = Me]
(S,S,S,S,SP)
11b-OH [R = H]
11b-OMe [R = Me]
(S,S,S,S,RP)
provides the product of higher enantioselectivity. Changing the
solvent from MeOH to CH2Cl2 led to inferior rates and
selectivities in both these cases.
These preliminary results indicate that, contrary to expecta-
tion, the enantioselectivity is sensitive to remote oxygen
substituents in the phospholane ring. Inspection of molecular
models indicates that the MeO- or HO- groups are axial in the
5-membered ring of the phospholane, and in the vicinity of
substituents on the coordinated alkene. Hence the opportunity
exists for cooperative association through H-bonding between
ligand and coordinated reactant.15 The combination of good
enantioselectivities in simple unoptimised reactions make this
an attractive series of ligands for further investigation with the
Table 2
11 The absolute configuration of all product boranes was established by
CD in comparison with that of the diborane from (S,S)-DIPAMP, the
phospholane part being essentially CD transparent in the 240–400 nm
region. We warmly thank Dr Guiliano Siligardi, KCL, for this data.
12 Full details of the synthesis will be published separately; D. Carmichael
and J. M. Brown Tetrahedron: Asymmetry, in preparation.
13 L. McKinstry and T. Livinghouse, Tetrahedron Lett., 1994, 35, 9319.
14 (R,R)-DIPAMP (ref. 2) and the (S,S)-phospholane MeBPE (ref. 8) both
give S-amino acids on Rh asymmetric hydrogenation, hence 8b and 11b
constitute the configurationally matched ligands and 7b, 10b the
mismatched ligands.
Substrate
Catalyst precursor
Ee (%)
11b-OMe
11b-OH
85 R
95 R
HO2C
CO2Me
15
93 Ra
87 R
11b-OMe
11b-OH
MeO2C
MeO2C
CO2H
16
17
11b-OMe
11b-OH
85 R
15 M. Sawamura and Y. Ito, Chem. Rev., 1992, 92, 857; J. Holz, M.
Quirmbach and A. Borner, Synthesis, 1997, 983, and references cited
therein.
80 Ra
CO2Me
16 J. Holz, M. Quirmbach, U. Schmidt, D. Heller, R. Stürmer and A.
Börner, J. Org. Chem., 1998, 63, 8031.
Conditions: substrate:catalyst 100 :1, (COD)2RhBF4 as precursor,
(HBF4·OMe2 deboronation in situ), 1.3 bar, MeOH, 1–3 h. a 94% ee for 16
with TfO2 instead of BF4
.
2
Communication 8/08711C
262
Chem. Commun., 1999, 261–262