NaCH(CO2Me)2 (2 equiv.)
[Ir(COD)Cl]2 (2 mol%)
P(OPh)3 (4 mol%)
OAc
CH(CO2Me)2
THF, RT, 66 h
83%
(R)-8
>99% ee
(R)-9
71% ee
Scheme 2
gioselectivity (ratio of 3:4) starting from achiral 2a was low
(3:4 = 62:38) and the enantioselectivity high [(R)-3a of 78%
ee], whereas with the regioisomeric racemic substrate 1a it was
just the opposite: the regioselectivity was high (3:4 = 95:5)
and the enantioselectivity low [(S)-3a of 8% ee] (entries 4–7).
These results demonstrate that properties of allyl–Ir inter-
mediates differ from those of allyl–Pd intermediates. As a rule,
in Pd-catalysed allylic substitutions regioisomeric starting
materials give rise to the same products; memory effects are
known, but are small for bidentate ligands.7
Fig. 1 Enantiomeric purity of (3) 1a and (+) 3a vs. conversion¶ for the
reaction of (±)-1a with dimethyl 2-sodiomalonate catalysed by [Ir-
(COD)Cl]2–(R)-6.
obtained in enantiomerically pure form beyond conversion of
ca. 80%.
For further clarification, the configurational stability of
intermediary allyl–Ir complexes and the configurative course of
the substitution were investigated. First, enantiomerically
enriched 1a and 1b were alkylated using P(OPh)3 as ligand.†
Starting from (R)-1a of 91% ee the product (R)-3a was obtained
with 51% ee with retention of configuration, i.e. 79% of the
substrate had reacted with retention of configuration (entry 9).
Similar results were achieved with the phenyl substituted
substrate 1b (entry 8).‡ Second, enantiomerically pure pent-
3-en-2-yl acetate† [(R)-8] was reacted with dimethyl 2-sodio-
malonate (Scheme 2); alkylation product (R)-9 with 71% ee was
formed.
As of now we cannot offer a conclusive explanation of our
results. As a working hypothesis we assume that the Ir-catalysed
reactions proceed, similar to the mechanism proposed for Rh-
catalysed reactions,4 by substitution of acetate to give s-allyl–Ir
complexes which further react with malonate again with
inversion; the s-complexes undergo slow racemisation (or
epimerisation) via s–p–s-rearrangement or sigmatropic 1,3-re-
arrangement (Scheme 3).
In conclusion, our results demonstrate that it is possible to
achieve high levels of enantioselectivity in Ir-catalysed alkyla-
tions of monoalkylallyl acetates. Presently it is necessary to use
racemic, branched substrates. Further progress will be achieved
on the basis of detailed mechanistic investigations. It appears of
particular importance to investigate the structure and dynamic
properties of allyl–Ir complexes.
This work was supported by the Deutsche Forschungsge-
meinschaft (SFB 247) and the Fonds der Chemischen Industrie.
We thank Degussa AG for iridium salts.
Notes and references
† Enantiomerically enriched 1a and 8 were prepared by enzyme (Novozym
435) catalysed esterification in vinyl acetate. (R)-1b was prepared from (R)-
1-phenylprop-2-en-1-ol which was purchased from Fluka.
‡ Change of the descriptors in the starting material and product is a
consequence of CIP priorities of substituents.
§ Isomerisation between 1a and 2a was not observed starting either from 1a
or from 2a (reaction conditions: 2 mol% [Ir(COD)Cl]2, 4 mol% P(OPh)3, 2
equiv. NaOAc, 18 h, room temperature). For enantiomerically enriched 1a
(83% ee) a low degree of racemisation was obtained when the same reaction
conditions were employed (70% ee).
The easily accessible phosphorus amidite 68 was used as a
monodentate chiral ligand (ratio Ir:6 = 1:1). For all substrates
this ligand is equivalent to P(OPh)3 with respect to catalytic
efficiency and regioselectivity (entries 11–14). Surprisingly,
and in contrast to the results obtained with the bidentate ligand
5, enantioselectivity induced by 6 was higher for the branched
substrate rac-1a than for the linear substrate 2a.
Reactions with added halide salts were investigated (entries
15-17) because a marked influence of halide ions on allylic
substitution has been reported.9 LiCl or LiBr indeed caused a
marked increase of regio- and enantioselectivity, although also
a small decrease of reactivity (entry 10). The effect of chloride
was further studied for alkylations of (R)- and (S)-1a (entries
18–23). The reaction with the substrate–ligand combination
(R)-1a–(R)-6 was faster and more enantioselective (matched
case) than with the combination (S)-1a–(R)-6 (mismatched),
yielding (R)-3a in both cases, i.e. control by the ligand is
stronger than that by the substrate. The influence of the ligand
is enhanced by addition of LiCl.
¶ The degree of conversion was determined by HPLC.
1 B. M. Trost and D. L. Van Vranken, Chem. Rev., 1996, 96, 395; T.
Hayashi, in Catalytic Asymmetric Synthesis, ed. I. Ojima, VCH, New
York, 1993, p. 325.
2 R. Prétôt and A. Pfaltz, Angew. Chem., Int. Ed., 1998, 37, 323; T.
Hayashi, K. Kishi, A. Yamamoto and Y. Ito, Tetrahedron Lett., 1990,
31, 1743; T. Hayashi, A. Ohno, S.-j. Lu, Y. Matsumoto, E. Fukuyo and
K. Yanagi, J. Am. Chem. Soc., 1994, 116, 4221; T. Hayashi, M.
Kawatsura and Y. Uozumi, Chem. Commun., 1997, 561; B. M. Trost
and F. D. Toste, J. Am. Chem. Soc., 1998, 120, 9074.
3 Mo: B. M. Trost and I. Hachiya, J. Am. Chem. Soc., 1998, 120, 1104; W:
G. C. Lloyd-Jones and A. Pfaltz, Angew. Chem., Int. Ed. Engl., 1995, 34,
462.
4 Fe: Y. Xu and B. Zhou, J. Org. Chem., 1987, 52, 974; Rh: P. A. Evans
and J. D. Nelson, J. Am. Chem. Soc., 1998, 120, 5581.
5 R. Takeuchi and M. Kashio, Angew. Chem., Int. Ed. Engl., 1997, 36,
263; R. Takeuchi and M. Kashio, J. Am. Chem. Soc., 1998, 120,
8647.
6 J. P. Janssen and G. Helmchen, Tetrahedron Lett., 1997, 38, 8025.
7 T. Hayashi, M. Kawatsura and Y. Uozumi, J. Am. Chem. Soc., 1998,
120, 1681; G. C. Lloyd-Jones and S. C. Stephen, Chem. Eur. J., 1998,
4, 2539; B. M. Trost and R. C. Bunt, J. Am. Chem. Soc., 1996, 118, 235;
B. M. Trost, M. G. Morgan and G. A. O’Doherty, J. Am. Chem. Soc.,
1995, 117, 9662.
8 B. L. Feringa, M. Pineschi, L. A. Arnold, R. Imbos and A. H. M. de
Vries, Angew. Chem., Int. Ed. Engl., 1997, 36, 2620.
9 U. Burckhardt, M. Baumann and A. Togni, Tetrahedron: Asymmetry,
1997, 8, 155; M. Kawatsura, Y. Uozumi and T. Hayashi, Chem.
Commun., 1998, 217.
As the acetates 1 and 2 do not isomerise under the usual
reaction conditions,§ it was of interest to explore a kinetic
resolution. The result for the reaction of rac-1a with NaHC-
(CO2CH3)2, catalysed by [Ir(COD)Cl]2–(R)-6, is shown in Fig.
1. According to this plot, (R)-1a is consumed ca. 12 times faster
than (S)-1a.10 Consequently, the slower reacting acetate is
[Ir]
R
R
[Ir]
10 The relative rate was estimated by comparison with a published plot of
conversion vs. ee of residual substrate for a kinetic resolution: V. S.
Martin, S. S. Woodard, T. Katsuki, Y. Yamada, M. Ikeda and K. B.
Sharpless, J. Am. Chem. Soc., 1981, 103, 6237.
[Ir]
[Ir]
[Ir]
R
R
R
Scheme 3
Communication 9/00864K
742
Chem. Commun., 1999, 741–742