Iridium-catalyzed asymmetric allylic substitution with aryl zinc reagents

Article abstract of DOI:10.1021/ol0713842

Thanks to iridium catalysis, arylzinc reagents undergo regioselective allylic substitution with very high enantioselectivity, when associated with phosphoramidite ligands.

Full text of DOI:10.1021/ol0713842

ORGANIC
LETTERS
2007
Vol. 9, No. 17
3393-3395
Iridium-Catalyzed Asymmetric Allylic
Substitution with Aryl Zinc Reagents
Alexandre Alexakis,* Samir El Hajjaji, Damien Polet, and Xavier Rathgeb
UniVersite´ de Gene`Ve, De´partement de chimie organique, 30 Quai Ernest Ansermet,
1211 Gene`Ve 4, Switzerland
alexandre.alexakis@chiorg.unige.ch
Received June 12, 2007
ABSTRACT
Thanks to iridium catalysis, arylzinc reagents undergo regioselective allylic substitution with very high enantioselectivity, when associated
with phosphoramidite ligands.
The allylic substitution is a fundamental reaction in organic
chemistry.1 The reaction of carbon nucleophiles is usually
catalyzed by a transition metal, and chiral ligands around
this metal may allow the asymmetric version of this reaction.2
Copper is unique in this context, allowing nonstabilized
nucleophiles, such as alkyl groups, to be transferred.3 In
addition, the regiocontrol of the reaction, with unsymmetri-
cally substituted allylic substrates, is possible with this metal.4
We, and others, have developed efficient copper-catalyzed
procedures where a Grignard or a dialkylzinc reagent reacts,
enantioselectively, with an allylic substrate with high regio-
and stereocontrol (Scheme 1).5
nards or alkylzinc reagents, thanks to a chiral ligand to
copper. However, despite all efforts, the transfer of an aryl
group was not regioselective, resulting mainly in the achiral
R product.
Aryl metal reagents have scarcely been used with other
transition metals than copper.1 There are even fewer enan-
tioselective versions, most of them dealing with meso-type
π-allyl systems.6 For soft nucleophiles, most transition metals
show mainly R selectivity on the above type of allylic
(2) (a) Miyabe, H.; Takemoto, Y. Synlett 2005, 1641. (b) Trost, B. M.
J. Org. Chem. 2004, 69, 5813. (c) Trost, B. M.; Crawley, M. L. Chem.
ReV. 2003, 103, 2921.
(3) (a) Posner, G. H. Org. React. 1975, 22, 253. (b) Karlstro¨m, A. S. E.;
Ba¨ckvall, J. E. In Modern Organocopper Chemistry; Krause, N., Ed.; Wiley-
VCH: Weinheim, Germany, 2001; p 259.
(4) (a) Yamamoto, Y. Angew. Chem., Int. Ed. Engl. 1986, 25, 947. (b)
Karlstro¨m, A. S. E.; Ba¨ckvall, J. E. Chem. Eur. J. 2001, 7, 1981.
(5) For recent reviews in Cu-catalyzed asymmetric allylic substitution:
(a) Alexakis, A.; Malan, C.; Lea, L.; Tissot-Croset, K.; Polet, D.; Falciola,
C. Chimia 2006, 60, 124. (b) Yorimitsu, H.; Oshima, K. Angew. Chem.,
Int. Ed. 2005, 44, 4435. (c) Kar, A.; Argade, N. P. Synthesis 2005, 2995.
(6) (a) Cherest, M.; Felkin, H.; Umpleby, J. D. Chem. Commun. 1981,
681. (b) Hiyama, T.; Wakasa, N. Tetrahedron Lett. 1985, 26, 3259. (c)
Fiaud, J.-C.; Aribi-Zouioueche, L. J. Organomet. Chem. 1985, 295, 383.
(d) Consiglio, G.; Piccolo, O.; Roncetti, L.; Morandini, F. Tetrahedron 1986,
42, 2043. (e) Fotiadu, F.; Cros, P.; Faure, B.; Buono, G. Tetrahedron Lett.
1990, 31, 77. (f) Consiglio, G.; Indolese, A. Organometallics 1991, 10,
3425. (g) Nomura, N.; RajanBabu, T. V. Tetrahedron Lett. 1997, 38, 1713.
(h) Nagel, U.; Nedden, H. G. Inorg. Chim. Acta 1998, 269, 34. (i) Gomez-
Bengoa, E.; Heron, N. M. N. M.; Didiuk, M. T.; Luchaco, C. A.; Hoveyda,
A. H. J. Am. Chem. Soc. 1998, 120, 7649. (j) Chung, K.-G.; Miyake, Y.;
Uemura, S. J. Chem. Soc., Perkin Trans. I 2000, 2725. (k) Yasui, H.;
Mizutani, K.; Yoremitsu, H.; Oshima, K. Tetrahedron 2006, 62, 1410.
Scheme 1
The γ product could be obtained with >99:1 regioselec-
tivity and >98% enantioselectivity with many alkyl Grig-
(1) (a) Magid, R. M. Tetrahedron 1980, 36, 1901. (b) Trost, B. M.; van
Vranken, D. L. Chem. ReV. 1996, 96, 395. (c) Trost, B. M.; Lee, C. In
Catalytic Asymmetric Synthesis, 2nd ed.; Ojima, I., Ed.; Wiley, New York,
2000; p 593. (d) Pfaltz, A.; Lautens, M. In ComprehensiVe Asymmetric
Catalysis I-III; Jacobsen, E. N., Pfaltz A., Yamamoto, H., Eds.; Springer:
Berlin, Germany, 1999; p 833.
10.1021/ol0713842 CCC: $37.00
© 2007 American Chemical Society
Published on Web 07/19/2007

substrates.1,2 Iridium stands as the most notable exception:
γ selectivity, and high enantioselectivity, was found with
several heteroatomic nucleophiles as well as with stabilized
carbon nucleophiles such as malonates.7 Remarkably, Evans
has found that Rh catalysis allows a moderately regiocon-
trolled allylation with phenylzinc halide.8 We report in this
Letter that arylzinc reagents can be catalyzed by chiral
iridium complexes to afford, for the first time, the γ product
as the major regioisomer, along with good to excellent
enantiocontrol.
Since arylzinc halides are known to be compatible with
palladium catalysis,6 we felt that their association with a γ
selective metal, such as Ir, could provide the desired
regioselective substitution. In addition, our phosphoramidite
ligand L1 (Figure 1) affords excellent enantioselectivities
with this metal.9
transfer both phenyl groups, thus making the reaction more
atom economical.10 Phenyl Grignard itself reacts directly with
the carbonyl group of the carbonate. The addition of LiBr
has a strongly beneficial effect, as was the case for malo-
nates.7b,g Changing the leaving group (acetate, halide,
phosphate) gave lower selectivities.
This first screening identified the best experimental
conditions for increased γ selectivity. Further screening was
performed on many chiral and achiral phosphorus ligands
(Figure 1) to identify how the structural variation of the
ligand may affect both the regio- and enantioselectivity. L1
and L3 gave the best results (69/31, ee 74% and 47/53, ee
66%) in favor of the γ product. A strong match/mismatched
effect was observed with L2 and L4 (30/70, ee 84% and
13/87, ee 80%) where the ee increased while the regio-
selectivity strongly dropped. In both cases the absolute con-
figuration remained the same, meaning that the amine part
of the ligand imposes the stereochemistry of the product.
Finally, the first ligand we tried, L1, appeared to be the
best compromise between high γ selectivity and enantio-
selectivity.
We next turned our attention to the scope of the reaction
(Table). First, several substrates were tested with various
Figure 1. Phosphoramidite ligands used in this study.
Table 1. Reaction of Arylzinc Reagents with Various Allylic
Carbonates, with L1
entry substrate product convn (yield, %)
γ/R
ee, %
Carbonate 1 was selected as a typical allylic substrate for
our preliminary studies. The experimental conditions were
set identical with those used for the allylic alkylation with
malonates, that is THF, room temperature for 15-18 h
(overnight), 2% of [Ir(COD)Cl]2, and 4.4% of L1 (Scheme
2).7g It appeared that optimum γ selectivity can be attained
1
2
1
3
2
4
100 (72)
100
69/31 74 (S)
42/58 78 (S)a
49/51 90 (S)a
23/77 85 (S)b
33/67 91 (S)a
15/85 56 (S)b
55/45 99.2 (S)a
57/43 95 (R)
50/50 93 (R)
56/44 97 (S)a
53/47 92 (R)
73/27 79 (S)a
63/37 91 (S)
16/84 63 (R)a
45/55 82 (S)
40/60 80 (S)
32/68 86 (S)
3
4
5
5
6
6
100 (67)
100
5
6
7
7
8
8
100 (78)
100
7
8
9
9
10
12
14
16
18
20
18
10
14
16
22
100 (83)
100 (89)
100 (83)
100 (98)
100 (93)
100
100 (97)
98
100 (95)
100 (78)
100 (51)
11
13
15
17
19
21
21
21
21
11
10c,d
11
12
13
14
15
16
17
Scheme 2
with organozinc reagents prepared either from PhLi or, better,
from the corresponding Grignard reagent. Half an equivalent
of ZnBr2 is enough. The resulting diphenylzinc is able to
a Ent-L1 was used. b With Ent-L2. c The same experiment with the
Grignard prepared by Mg/I exchange11 gave 87% isolated yield, with
46/54 selectivity and 78% ee. d The same experiment with PhLi prepared
by Li/I exchange gave 53% isolated yield, with 44/56 selectivity and
87% ee.
(7) For an early review see: Takeuchi, R. Synlett 2002, 1954. For more
recent examples see: (a) Kanayama, T.; Yoshida, K.; Miyabe, H.; Takemoto,
Y. Angew. Chem., Int. Ed. 2003, 42, 2054. (b) Bartels, B.; Helmchen, G.;
Garcia-Yebra, C. Eur. J. Org. Chem. 2003, 1097. (c) Weihofen, R.; Dahnz,
A.; Tverskoy, O.; Helmchen, G. Chem. Commun. 2005, 3541. (d) Ohmura,
T.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 15164. (e) Shu, C.; Hartwig,
J. F. Angew. Chem., Int. Ed. 2004, 43, 4794. (f) Fischer, C.; Defieber, C.;
Suzuki, T.; Carreira, E. M. J. Am. Chem. Soc. 2004, 126, 1628. (g) Alexakis,
A.; Polet, D. Org. Lett. 2004, 6, 3529. (h) Polet, D.; Alexakis, A.; Tissot-
Croset, K.; Corminboeuf, C.; Ditrich, K. Chem. Eur. J. 2006, 12, 3596 and
references cited therein.
alkyl or aromatic groups. On the other hand, different aryl
groups were introduced via their Grignard reagent, or
sometimes via their organolithium reagent. The reaction with
phenylzinc reagent seems quite general with aryl-substituted
substrates. The enantioselectivities usually reach >90%, with
one example at 99.2% (entry 7)! The regioselectivities are
(8) See footnote 9 in: Evans, P. A.; Uraguchi, D. J. Am. Chem. Soc.
2003, 125, 7158.
(9) Tissot-Croset, K.; Polet, D.; Alexakis, A. Angew. Chem., Int. Ed.
2004, 43, 2426.
(10) Trost, B. M. Acc. Chem. Res. 2002, 35, 695.
Org. Lett., Vol. 9, No. 17, 2007
3394

usually slightly above 50%, but it should be noted that such
values are unprecedented. An ortho heteroatom on the
substrate, such as o-MeO-C6H4 3 (entry 2) or 2-furyl 19
(entry 12), lowers the ee values to 78% and 79%, while, for
an unexplained reason, a p-MeO-C6H4 substituent on sub-
strate 7 (entry 5) lowers the regioselectivity. The matched/
mismatched effect of L1 versus L2 could be seen again
with substrates 5 and 7 where both the regio- and enantio-
selectivities are lower (see entries 3-4 and 5-6). In general,
we do not see any marked effect of the electron withdrawing
or electron donating nature of the aromatic group (compare
entries 5 and 10). In addition, other aryl Grignard reagents
were investigated on cinnamyl carbonate 21. Although
commercial grade Grignard reagents were used, it is also
possible to use Grignard reagents made according to Knochel,
by magnesium-halogen exchange with i-PrMgCl‚LiCl.11 As
compared to the previous results with Ph2Zn, the enantiose-
lectivities are slightly lower (63-91%).
Scheme 4 . Synthetic Approach to Sertraline
was done by conjugate addition of a CuH species.18 A Pd-
catalyzed hydrogenation was avoided due to potential race-
mization of this sensitive substrate.19
Scheme 3
The stereochemical outcome merits some comments. The
selected face attack by the diarylzinc species is exactly the
same as the face selectivity of malonates and amines we
observed previously.7h Does that mean that the mechanism
of the substitution is the same? The lower regioselectivity
we observe in these reactions may point to a divergent
mechanism for the γ and the R products. The γ product could
arise from an anti attack of the intermediate π-allyliridium
complex, whereas the R product could be the result of a prior
transmetallation with the zinc nucleophile.11 Studies to clarify
this point are currently under investigation.
In conclusion, we have disclosed the first Ir-catalyzed
regioselective allylic substitution with aryl nucleophiles, with
high enantioselectivity. Although moderate, the observed γ
regioselectivity is the highest reported to date for such a
reaction, whatever the metal. The obtained diaryl-substituted
chiral alkenes are useful synthetic intermediates, as exempli-
fied by the formal synthesis of Sertraline.
The absolute configuration of the products of this reaction
was determined by comparison with known data on product,
2.12,13 In addition, product 14 was submitted to cross-
metathesis reaction with methyl acrylate14 and the Grubbs-
Hoveyda catalyst (Scheme 4).15 The resulting ester 23 was
reduced to 24, a well-known intermediate16 in the synthesis
of Sertraline (Zoloft), a major pharmaceutical agent for the
treatment of depression.17 The reduction of the double bond
Acknowledgment. We thank the Swiss National Science
Foundation (No. 200020-113332) for financial support.
Special thanks, also, to Ste´phane Rosset (University of
Geneva) for his help in the separation of enantiomers on
chiral GC or SFC.
(11) Knochel, P.; Dohle, W.; Gommermann, N.; Kneisel, F. F.; Kopp,
F.; Korn, T.; Sapountzis, I.; Vu, V. A. Angew. Chem., Int. Ed. 2003, 42,
4302.
(12) Breit, B.; Breuninger, D. Synthesis 2005, 147.
(13) Hayashi, T.; Iwamura, H.; Naito, M.; Matsumoto, Y.; Uozumi, Y.;
Miki, M.; Yanaki, K. J. Am. Chem. Soc. 1994, 116, 775.
(14) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34, 18.
(15) Garber, S. B.; Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H. J. Am.
Chem. Soc. 2000, 122, 8168.
Supporting Information Available: Experimental pro-
cedures and spectral analyses of all products. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL0713842
(16) Davies, H. M. L.; Stafford, D. G.; Hansen, T. Org. Lett. 1999, 1,
233.
(17) (a) Coe, B. K.; Weissman, A.; Welch, W. M.; Broune, R. G. J.
Pharmacol. Exp. Ther. 1983, 226, 686. (b) Welch, W. M.; Kraska, A. R.;
Sarges, R.; Coe, B. K. J. Med. Chem. 1984, 27, 1508. (c) Quallich, G. J.;
Woodall, T. M. Tetrahedron 1992, 48, 10239.
(18) Rainka, M. P.; Aye, Y.; Buchwald, S. L. Proc. Natl. Acad. Sci 2004,
101, 5821.
(19) Fehr, C.; Galindo, J.; Farris, I.; Cuenca, A. HelV. Chim. Acta 2004,
87, 1737.
Org. Lett., Vol. 9, No. 17, 2007
3395