Very efficient phosphoramidite ligand for asymmetric iridium-catalyzed allylic alkylation

Article abstract of DOI:10.1021/ol048607y

(Chemical Equation Presented) Linear or branched allylic carbonates or acetates undergo enantioselective iridium-catalyzed allylic substitution with sodium malonate. The reaction is wide in scope and affords the branched product in high yield and with hig

Full text of DOI:10.1021/ol048607y

ORGANIC
LETTERS
2004
Vol. 6, No. 20
3529-3532
Very Efficient Phosphoramidite Ligand
for Asymmetric Iridium-Catalyzed Allylic
Alkylation
Alexandre Alexakis* and Damien Polet
Department of Organic Chemistry, UniVersity of GeneVa, 30, quai Ernest Ansermet,
CH-1211 GeneVa 4, Switzerland
alexandre.alexakis@chiorg.unige.ch
Received July 20, 2004
ABSTRACT
Linear or branched allylic carbonates or acetates undergo enantioselective iridium-catalyzed allylic substitution with sodium malonate. The
reaction is wide in scope and affords the branched product in high yield and with high regio- (up to >99:1) and enantioselectivity (up to 98%).
Ten aromatic or aliphatic substrates were successfully tested.
The asymmetric allylic substitution (eq 1) on an achiral
substrate is a potentially powerful method to create new
chiral centers from easily available starting material. Of
particular interest are the unsymmetrically substituted allyl
derivatives. They are particularly challenging substrates,
because, in addition to the requirement of enantiocontrol,
the problem of regioselectivity has to be solved.
Ir3,4 Mo,5 Ru,6 and W,7 which give regioselectivities in favor
of branched products. Nevertheless, there is still a need for
a general system that would allow the use of various types
(2) For exceptions, see: (a) Pre´toˆt, R.; Pfaltz, A. Angew. Chem., Int.
Ed. 1998, 37, 323. (b) Glorius, F.; Neuburger, M.; Pfaltz, A. HelV. Chim.
Acta 2001, 84, 3178. (c) You, S.-L.; Zhu, X.-Z.; Luo, Y.-M.; Hou, X.-L.;
Dai, L.-X. J. Am. Chem. Soc. 2001, 123, 7471 and refs cited herein.
(3) For a review on Ir, see: Takeuchi, R. Synlett 2002, 1954.
(4) (a) Takeuchi, R.; Kashio, M. Angew. Chem., Int. Ed. Engl. 1997,
36, 263. (b) Takeuchi, R.; Kashio, M. J. Am. Chem. Soc. 1998, 120, 8647.
(c) Takeuchi, R.; Tanabe, K. Angew. Chem., Int. Ed. 2000, 39, 1975. (d)
Janssen, J. P.; Helmchen, G. Tetrahedron Lett. 1997, 38, 8025. (e) Bartels,
B.; Helmchen, G. Chem. Commun. 1999, 741. (f) Bartels, B.; Garcia-Yebra,
C.; Rominger, F.; Helmchen, G. Eur. J. Inorg. Chem. 2002, 2569. (g)
Bartels, B.; Helmchen, G.; Garcia-Yebra, C. Eur. J. Org. Chem. 2003, 1097.
(h) Fuji, K.; Kinoshita, N.; Kawabata, T.; Tanaka, K. Chem. Commun. 1999,
2289. (i) Kanayama, T.; Yoshida, K.; Miyabe, H.; Kimachi, T.; Takemoto,
Y. J. Org. Chem. 2003, 68, 6197. (j) Takeuchi, R.; Ue, N.; Tanabe, K.;
Yamashita, K.; Shiga, N. J. Am. Chem. Soc. 2001, 123, 9525. (k) Kanayama,
T.; Yoshida, K.; Miyabe, H.; Kimachi, T.; Takemoto, Y. J. Org. Chem.
2003, 68, 6197. (l) Kanayama, T.; Yoshida, K.; Miyabe, H.; Takemoto, Y.
Angew. Chem., Int. Ed. 2003, 42, 2054.
Among the metals that are used for this reaction, palladium
is the most widely studied.1 However, despite the high
enantioselectivities obtained with many other substrates, this
metal gives preference to the linear instead of the branched
product (eq 1).2 The situation is different for complexes of
(5) (a) For a review on Mo, see: Belda, O.; Moberg, C. Acc. Chem.
Res. 2004, 37, 159 and refs cited therein. (b) See also: Trost, B. M.;
Hachiya, I. J. Am. Chem. Soc. 1998, 120, 1104.
(6) Trost, B. M.; Fraisse, P. L.; Ball, Z. T. Angew. Chem., Int. Ed. 2002,
41, 1059.
(1) (a) Trost, B. M.; Lee, C. In Catalytic Asymmetric Synthesis, 2nd ed.;
Ojima, I., Ed.; Wiley-VCH: New York, 2000; p 593. (b) Pfaltz, A.; Lautens,
M. In ComprehensiVe Asymmetric Catalysis I-III; Jacobsen, E. N., Pfaltz,
A., Yamamoto, H., Eds.; Springer, Berlin, 1999; p 833. (c) Guiry, P. J.;
Saunders, C. P. AdV. Synth. Catal. 2004, 346, 497.
(7) Lloyd-Jones, G. C.; Pfaltz, A. Angew. Chem., Int. Ed. 1995, 34, 462.
10.1021/ol048607y CCC: $27.50
© 2004 American Chemical Society
Published on Web 08/28/2004

Table 1. Enantioselective Ir-Catalyzed Allylic Alkylation of Carbonatesa
a Molar ratio: substrate/LiCl/NaHC(CO2Me)2/[Ir(COD)Cl]2/L ) 1/1/2/0.02/0.04. Scale: 0.5 mmol of substrate. b Isolated yield. c Determined by 400 MHz
1H NMR or GC-MS. d Determined by chiral GC or SFC. e Absolute configurations were determined by comparison with published data (ref 7); all the other
configurations were attributed by analogy. f Performed on a 1 mmol scale. g Performed on a 2 mmol scale with a molar ratio: substrate/LiCl/NaHC(CO2Me)2/
[Ir(COD)Cl]2/L ) 1/1/2/0.01/0.02.
of nucleophiles, as well as various substrates having different
types of substituents (aromatic or aliphatic ones), and would
be practical and cheap.
Strong π-accepting phosphoramidite monodentate ligands,
such as L1a-3a (Figure 1) have recently shown their
efficiency,8 in particular in the Ir-catalyzed allylic amination
reactions of allylic carbonates,9 as well as allylic substitution
of dienyl esters.10
In the allylic amination and etherification, using Ir and
L3a as the chiral source, Hartwig observed an induction
period; he isolated and characterized the active catalytic
species.11 Our group very recently described new phosphor-
amidite ligands (L1b-3b) that contain an ortho substituent
on the amine part of the ligand.12 These ligands, in particular
L3b, gave high enantio- and regioselectivities in the Cu-
(10) Lipowsky, G.; Helmchen, G. Chem. Commun. 2004, 116.
(11) Kiener, C. A.; Shu, C.; Incarvito, C.; Hartwig, J. F. J. Am. Chem.
Soc. 2003, 125, 14272.
(12) Alexakis, A.; Tissot-Croset, K.; Polet, D. Angew. Chem., Int. Ed.
2004, 43, 2426.
(8) Feringa, B. L. Acc. Chem. Res. 2000, 33, 346.
(9) (a) Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 15164.
(b) Lo´pez, F.; Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125,
3426.
3530
Org. Lett., Vol. 6, No. 20, 2004

Table 2. Ir-Catalyzed Allylic Alkylation of Acetates with
Ligand L3ba
Figure 1. Chiral ligands used in this study.
catalyzed alkylation and in the Ir-catalyzed amination of
cinnamyl carbonate. No induction period was observed,
which could lead to the hypothesis that L3b could work as
a bidentate P-O hemilabile ligand. This hypothesis was
supported by the fact that a para-methoxy group (instead of
ortho-methoxy) does not improve the result of L1a-3a.12
To further test the improvements brought by our new
ligands and to generalize our system, we wished to try other
nucleophiles, particularly malonates. It should be noticed that
Helmchen had already used ligands L1a-3a in this reaction
with 7-94% ee.4e-g
Our work began with the simple addition of sodium
malonate onto carbonate 1a under the classical experimental
conditions (4% Ir catalyst and 4% ligand).4g The reaction
was very slow, even at 60 °C, and led to low regio- and
enantioselectivities (Table 1, entry 1). The observation by
Helmchen4e and Fuji4h that lithium chloride gave a strong
salt effect, improving both the regio- and the enantioselec-
tivity, led us to add 1 equiv of lithium chloride early in the
reaction. An impressive acceleration of the reaction, together
with an improvement of the regio- (99:1) and enantioselec-
tivity (98%), was obtained (entry 2). Even a lower catalyst
loading (2% Ir catalyst) did not seem to be deleterious (entry
5). This result demonstrates again the efficiency of the new
ligand L3b, bearing the ortho-methoxy group, as compared
to L3a, which in the case of cinnamyl derivative gave 86%
ee and a 91:9 ratio.4g
a Molar ratio: substrate/LiCl/NaHC(CO2Me)2/[Ir(COD)Cl]2/L3b
)
1/1/2/0.02/0.04. Scale: 0.5 mmol of substrate. b Isolated yield. c Determined
by 400 MHz 1H NMR or GC-MS. d Determined by chiral GC or SFC.
e Performed on a 1 mmol scale.
We then successfully tested other allylic substrates,
showing that the reaction was general for every substrate,
whatever the electron demand on the aromatic substituents.5
The only exception was for the ortho-methoxy derivative
1h. This suggests that the O-coordinating ability of the
methoxy group on the substrate competes with the ortho-
methoxy of the ligand (entry 11). Of particular interest is
the substrate bearing a para-Cl substituent, 1f, which is
known to lead to the therapeutically useful GABAB receptor
agonist (R)-baclofen hydrochloride (Scheme 1) in high yield
Scheme 1. Synthesis of (R)-Baclofen According to Ref 14b
Comparison of ligands L3b (S,SS) and L2b (R,SS) clearly
shows a matched/mismatched situation, with L2b being the
mismatched one (only 89% ee). It should be noted that the
inversion of the binaphthol part of the ligand translates into
an inversion of stereochemistry on the resulting allylic
adduct. Interestingly, the simpler ligand L1b affords a
relatively high ee (93%). The absolute stereochemistry of
the adduct corresponds to the one obtained with the matched
ligand L3b. Thus, the biphenol part of the atropoisomerically
flexible ligand L1b seems to adopt the same atropoisomeric
configuration as the (S)-binaphthol of L3b. This behavior is
similar to that observed in Cu-catalyzed reactions, where the
induced atropoisomerism of the biphenol part has the same
configuration as the matched binaphthol configuration of the
binaphthyl-type ligand.13
and enantioselectivity.14 The reaction is equally effective with
substrates having a primary (entry 14) or a secondary (entry
13) alkyl substituent.
(13) Alexakis, A.; Rosset, S.; Allamand, J.; March, S.; Guillen, F.;
Benhaim, C. Synlett 2001, 1375.
(14) (a) Corey, E. J.; Zhang, F.-Y Org. Lett. 2000, 2, 4257. (b) Belda,
O.; Lundgren, S.; Moberg, C. Org. Lett. 2003, 5, 2275.
Org. Lett., Vol. 6, No. 20, 2004
3531

Allylic carbonates are usually the substrates of choice,
whereas allylic acetates are used less often due to their lower
reactivity and selectivity. In our case, the reaction on linear
allylic acetates did not seem to be different, except for a
slight decrease in rate (Table 2, entry 1). Thus, the substrate
4b needs to be heated to complete the reaction. The effect
of raising the temperature did not affect the enantioselectivity
but affected instead the regioselectivity (entries 2-5). Finally,
we also tested a branched substrate 5a. However, the
enantioselectivity dropped dramatically, showing that the
isomerization of the Ir π-complex is a very slow process, a
result also observed by Helmchen and Moberg.4e,14
to be the best for this purpose. This efficiency may probably
be ascribed to their bidentate P-O nature, due to the presence
of the ortho-methoxy group.
Acknowledgment. The authors thank Karine Tissot-
Croset, Ste´phane Rosset, and Xavier Rathgeb for help, the
BASF company for generous gift of chiral amines, and the
Swiss National Research Foundation No. 20-068095.02 and
COST action D24/0003/01 (OFES contract Nï. C02.0027)
for financial support.
Supporting Information Available: Experimental de-
tails, chromatograms, and spectral data for all compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
In conclusion, we have found that our new phosphor-
amidite ligands L1b-3b are excellent for the Ir-catalyzed
allylic substitution in terms of regio- and enantioselectivity
and in the wide scope of applicability. Ligand L3b appears
OL048607Y
3532
Org. Lett., Vol. 6, No. 20, 2004