ω-Ethylenic allylic substrates as alternatives to cyclic substrates in copper- and iridium-catalyzed asymmetric allylic alkylation

Article abstract of DOI:10.1021/ol100162y

"Chemical Equation Presented" A new strategy to access highly enantioenriched cyclic compounds (up to 98%) Is proposed using ω-ethylenic allylic substrates through a one-pot asymmetric allylic alkylation and ring-closing metathesis. Such starting compounds can be seen as synthetic equivalents of cyclic allylic substrates.

Full text of DOI:10.1021/ol100162y

ORGANIC
LETTERS
2010
Vol. 12, No. 6
1156-1159
ω-Ethylenic Allylic Substrates as
Alternatives to Cyclic Substrates in
Copper- and Iridium-Catalyzed
Asymmetric Allylic Alkylation
Francesca Giacomina, David Riat, and Alexandre Alexakis*
Department of Organic Chemistry, UniVersity of GeneVa, 30, quai Ernest Ansermet
CH-1211, GeneVa 4, Switzerland
alexandre.alexakis@unige.ch
Received December 8, 2009
ABSTRACT
A new strategy to access highly enantioenriched cyclic compounds (up to 98%) is proposed using ω-ethylenic allylic substrates through a
one-pot asymmetric allylic alkylation and ring-closing metathesis. Such starting compounds can be seen as synthetic equivalents of cyclic
allylic substrates.
Asymmetric allylic alkylation (AAA) is one of the funda-
mental C-C bond-forming transformations used as an
important tool in organic synthesis to achieve high regio-
and enantioselectivity using different nucleophiles and dif-
ferent allylic electrophiles.1 Many metals can catalyze this
reaction and, among them, palladium has been extensively
studied and successfully applied in this reaction using
stabilized nucleophiles, such as malonates or amines.2
The low regioselectivity obtained in palladium-catalyzed
allylic alkylation with nonsymmetrical (or monosubstituted)
allylic substrates has hampered the general use of this
reaction. In addition, only stabilized nucleophiles provided
high selectivities, forcing many authors to explore other
alternatives.
the use of nonstabilized organometallic reagents for the direct
introduction of alkyl groups into prochiral allylic substrates,
and more interestingly, it proceeds with high SN2′ regiose-
lectivity.3
In our laboratory, we were interested in cyclic allylic
substrates as class of electrophiles because of the numerous
biologically active compounds bearing a cyclic moiety in
the scaffold, which could be obtained starting from an
asymmetric allylic alkylation reaction.4
Many examples for the formation of enantioenriched cyclic
compounds can be found in the literature using palladium
as catalyst, as high enantioselectivity is reached using an
appropriate chiral ligand which allows deracemization of the
meso π-allyl intermediate formed upon displacement of the
allylic leaving group.2
In recent years, more attention has turned to copper.
Copper-catalyzed allylic alkylation could be considered
complementary to palladium. In fact, copper catalysis allows
(3) For reviews, see: (a) Alexakis, A.; Malan, C.; Lea, L.; Tissot-Croset,
K.; Polet, D.; Falciola, C. Chimia 2006, 60, 124. (b) Falciola, C. A.;
Alexakis, A. Eur. J. Org. Chem. 2008, 3765. (c) Alexakis, A.; Ba¨ckvall,
J. E.; Krause, N.; Pamies, O.; Dieguez, M. Chem. ReV. 2008, 108, 2796.
(d) Harutyunyan, S. R.; den Hartog, T.; Geurts, K.; Minnaard, A. J.; Feringa,
B. L. Chem. ReV. 2008, 108, 2824.
(1) (a) Trost, B. M.; Lee, C. In Catalytic Asymmetric Synthesis, 2nd
ed.; Oijima, I., Ed.; Wiley: New York, 2000; 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.
(2) Trost, B. M.; Crawley, M. L. Chem. ReV. 2003, 103, 2921.
(4) Trost, B. M. J. Org. Chem. 2004, 69, 5813.
10.1021/ol100162y  2010 American Chemical Society
Published on Web 02/18/2010

In contrast, there are some limitations to using copper as
catalyst with cyclic compounds because of the different
mechanism involved in this reaction, compared to the one
of palladium.5 Exclusively SN2′ (or γ) syn or anti adducts
can be obtained by choosing a suitable chelating leaving
group on a chiral allylic substrate (Scheme 1).6 However,
groups.8d On the other hand, aryl Grignard reagents are not
very γ regioselective.3 One key feature of our strategy is
the possibility to perform the reaction in a one-pot procedure
as we have already demonstrated the tolerance of Grubbs’
catalyst toward the allylic alkylation’s conditions.8
Initially, we considered three allylic chlorides, differing
from each other for the length of the carbon chain between
the allylic double bond and the ethylenic bond to have access
to enantioenriched five-, six-, and seven-membered rings.
Copper-catalyzed asymmetric allylic alkylation was per-
formed using 3 mol % of copper thiophene carboxylate (CuTC),
3.3 mol % of a chiral phosphoramidite-type ligand (Figure 1),
Scheme 1. Copper-Mediated Allylic Alkylation
starting from a racemic allylic substrate the enantiodiscrimi-
nating step, namely the oxidative addition of copper, would
generate both possible σ-allyl intermediate, producing after
the reductive elimation step a mixture of enantiomers.7
For this reason we looked for an alternative way to obtain
this important chiral synthons.
Here we propose the use of an ω-ethylenic allylic substrate
as synthetic equivalent, which can undergo copper-catalyzed
asymmetric allylic alkylation using Grignard reagents and
bears a convenient functionality to be cyclized through a ring
closing metathesis reaction (Scheme 2, pathway a). This
Figure 1. Chiral ligands used in this work.
Scheme 2
.
Alternative Approach for the Synthesis of
Enantioenriched Cyclic Systems
and 1.3 equiv of Grignard reagent in dichloromethane at -78
°C. Without quenching the reaction, Grubbs’ catalyst9 was added
and the flask was warmed to room temperature. The results
reported in the schemes are the best obtained after a screening
of different phosphoramidite-type ligands.
As shown in Table 1, regioselectivities were excellent in
all cases and subsequent ring-closing metathesis, using 5 mol
% of Grubbs’ catalyst (first generation), allowed the deter-
mination of the enantiomeric excess when it was not possible
after the alkylation step. Entries 1 and 2 of Table 1 show
effectively that no loss of enantioselectivity takes place
during the ring-closure step. Although the isolated yields over
the two steps are good to moderate, it should be mentioned
that the metathesis step is the most delicate, needing recently
bought (or prepared) Grubbs’ catalyst.
By looking more in detail at the results, bulky phosphora-
midite ligand L210 gave the best enantiomeric excess of 93%
for the formation of a six-membered ring with a (4-tert-
butoxy)butyl substitution (Table 1, entry 6). L2 was also
effective for the formation of a phenethyl-substituted six-
membered ring product in 86% ee, and the same level of
enantioselectivity was obtained with the standard phosphora-
midite ligand L111 for obtaining a phenethyl-substituted five-
functionality can be also introduced through the allylic
alkylation step from the nucleophile, as it has already been
reported from our group (Scheme 2, pathway b).8 Depending
on the R group to be introduced, either pathway a or b can
be chosen. For example, linear R alkyl groups on the allylic
substrate usually afford lower enantioselectivities than aryl
(5) Yamanaka, M.; Kato, S.; Nakamura, E. J. Am. Chem. Soc. 2004,
126, 6287.
(6) For syn selectivity, see: (a) Gallina, C.; Ciattini, P. G. J. Am. Chem.
Soc. 1979, 101, 1035. (b) Breit, B.; Demel, P. AdV. Synth. Catal. 2001,
343, 429.
(7) For anti selectivity, see: (a) Persson, E. S. M.; van Klaveren, M.;
Grove, D. M.; Ba¨ckvall, J.-E.; van Koten, G. Chem.sEur. J. 1995, 1, 351.
(8) (a) Alexakis, A.; Croset, K. Org. Lett. 2002, 4, 4147. (b) Tissot-
Croset, K.; Polet, D.; Alexakis, A. Angew. Chem., Int. Ed. 2004, 43, 2426.
(c) Tissot-Croset, K.; Polet, D.; Gille, S.; Hawner, C.; Alexakis, A. Synthesis
2004, 43, 2586. (d) Tissot-Croset, K. Ph.D. thesis (no. 3634), University
of Geneva: Geneva, Switzerland, 2005.
(9) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34, 18.
(10) (a) Alexakis, A.; Albrow, V.; Biswas, K.; d’Augustin, M.; Prieto,
O.; Woodward, S. Chem. Commun. 2005, 2843. (b) Li, K.; Alexakis, A.t
Chem.sEur. J. 2007, 13, 3765.
Org. Lett., Vol. 12, No. 6, 2010
1157

sponding allylic substrate.13 To the best of our knowledge,
no examples are reported in the literature about this
transformation catalyzed either by copper or palladium. In
fact, as already discussed for simple cyclic substrates, copper-
catalyzed allylic alkylation would afford a mixture of
enantiomers because of the formation of the two possible
diastereomeric species. In addition, the presence of a
substituent in the γ-position would afford a mixture of
regioisomers because of the hindered position.
Table 1. Cu-AAA/RCM of Allylic Chlorides 1a-ca
Furthermore, concerning palladium chemistry, it is well-
known that this metal is generally regioselective for the least
hindered position but the presence of a substituent in the γ
position would lead to the formation of a nonmeso π-allyl
intermediate, preventing deracemization of the substrates and
thus resulting in the formation of a mixture of enantiomers.1
For these reasons, we wished to apply our strategy to
synthesize these challenging compounds starting from an
ω-ethylenic allylic substrate and performing the asymmetric
allylic alkylation-ring-closing metathesis in a one-pot pro-
cedure.
ee of product
ee of
yieldd (%) 4c (%)
entry substrate
R
L
γ/Rb 3c (%)
4
1
1a
1a
1a
1b
1b
1b
1c
1c
2a L1 92/8
2b L3 >99/1
2c L4 >99/1
2a L2 >99/1
2b L1 94/6
2c L2 >99/1
2a L1 >99/1
2b L3 >99/1
86
82
4a
4b
4c
4d
4e
4f
62
74
52
66
68
68
62f
55f
86
82
73
86
74
93
85
80
2
3
4
5
6
7e
4g
4h
Table 2 summarizes the results obtained for the copper-
catalyzed asymmetric allylic alkylation for allylic chlorides
5a and 5b.
8e
a General conditions: 3 mol % of CuTC, 3.3 mol % of L*, 1.3 equiv of
RMgX, in CH2Cl2, at -78 °C. b Determined by 1H NMR or GC/MS.
c Determined by chiral GC or SFC analysis. d Yield of isolated product
after a flash column chromatography on SiO2. e Reaction performed in a
two-pot procedure. f Overall yield.
Table 2. Cu-AAA/RCM of Allylic Chlorides 5a,b
membered ring (Table 1, entry 4 and 1). An 85% ee was
also obtained with this ligand for the corresponding seven-
membered ring product (Table 1, entry 7). The introduction
of a secondary alkyl group was also explored: a good 82%
and 80% ee was achieved with the use of chiral ligand L312
to obtain five- and seven-membered ring products (Table 1,
entry 2 and 8), while only 74% ee was obtained for the six-
membered ring with ligand L1 (Table 1, entry 5). Finally, it
is worth mentioning the positive effect the ω double bond
has on the enantioselecivity. The saturated analogue of 1a
gave, under the same conditions as entry 2, only 38% ee.
We then turned our attention to a challenging cyclic
substrate for the asymmetric allylic alkylation reaction having
a substitution on the γ-position (Scheme 3). In fact, this kind
product
ee of
entry substrate
R
L
γ/Ra
7
yieldb (%) 7c (%)
1
5a
5a
5b
5b
5b
2b L3 >99/1
2c L1 >99/1
2a L4 92/8
2b L3 >99/1
2c L1 >99/1
7a
7b
7c
7d
7e
55
74
68
67
64
53
2
60
3d
4
62e
70
5d
60e
Scheme 3. Allylic Alkylation on γ-Substituted Cyclic Systems
a Determined by 1H NMR or GC/MS. b Yield of isolated product after
a flash column chromatography on SiO2. c Determined by chiral GC or SFC
analysis. d Reaction performed in a two-pot procedure. e Overall yield.
All the reactions proceeded in highly regioselective
manner. Biphenol ligand L3 was reasonably efficient for the
cyclohexyl alkylation of both substrates with ee’s up to 74%
(Table 2, entry 1), while phosphoramidite ligand L1 afforded
(11) Feringa, B. L. Acc. Chem. Res. 2000, 33, 346.
(12) Alexakis, A.; Rosset, S.; Allamand, J.; March, S.; Guillen, F.;
Benhaim, C. Synlett 2001, 1375.
of allylic system is not accessible in a highly selective way
from a direct asymmetric allylic alkylation on the corre-
(13) Falciola, C. A.; Tissot-Croset, K.; Reyneri, H.; Alexakis, A. AdV.
Synth. Catal. 2008, 350, 1090.
1158
Org. Lett., Vol. 12, No. 6, 2010

68% ee for the formation of a five-membered ring with a
(4-tert-butoxy)butyl substitution but only 53% for the six-
membered ring (Table 2, entries 2 and 5).
emerged as highly regioselective, as reported by our group
and other authors.14
After optimization of the reaction conditions, high regio-
and enantioselectivities with both substrates were obtained.
It should be pointed out that because of the different
reaction’s conditions (i.e., different reaction’s solvent) the
whole process was performed in a two step procedure. Ligand
L48b,c afforded up to 95% γ-selectivity and up to 98% ee.
In conclusion, we have described a new strategy to access
highly enantioenriched cyclic systems using ω-ethylenic
allylic substrates as synthetic equivalent of cyclic allylic
compounds. This is a complementary strategy to our previ-
ously described one,8 as it allows an easy variation of the
organometallic reagent instead of the substrate. It was applied
in the copper-catalyzed enantioselective allylic alkylation by
Grignard reagents and iridium-catalyzed enantioselective
allylic alkylation with dimethyl malonate followed by a ring-
closing metathesis reaction. Enantiomeric excesses up to 93%
and 98% were obtained, respectively.
By comparing the results obtained on the allylic substrate
without having a substituent on the terminal double bond we
observed a decrease of the enantiomeric excess, especially in
the case of the addition of a secondary Grignard reagent. We
could assume that the ethylenic double bond could play a role
in the enantiodetermining step and in particular during the
coordination of copper to the substrate. Indeed the metal can
coordinate to this terminal double bond before the formation
of the σ-allyl complex so that the presence of the methyl group
can modify by steric hindrance the chiral pocket.
Finally, we were interested in investigating the asymmetric
allylic alkylation reaction using stabilized nucleophiles.
Therefore, we performed the iridium-catalyzed asymmetric
allylic alkylation on allylic carbonate 8a and 8b with
dimethyl malonate (Scheme 4). The choice of using this
Acknowledgment. We thank the Swiss National Research
Foundation (grant no. 200020-126663) and COST action D40
(SER contract no. C07.0097) for financial support, as well
as BASF for the generous gift of chiral amines.
Scheme 4. Ir-AAA/RCM of Allylic Carbonates 8a,b
Supporting Information Available: Experimental pro-
cedures, 1H and 13C spectra, and chiral separations for
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL100162Y
(14) (a) Polet, D.; Alexakis, A.; Tissot-Croset, K.; Corminboeuf, C.;
Ditrich, K. Chem.sEur. J. 2006, 12, 3569. (b) Ohmura, T.; Hartwig, J. F.
J. Am. Chem. Soc. 2002, 124, 15164. (c) Helmchen, G.; Dahnz, A.; Du¨bon,
P.; Schelwies, M.; Weihofen, R. Chem. Commun. 2007, 675, and references
cited therein.
metal as catalyst instead of palladium is due to the fact that
iridium-mediated allylic alkylation reaction has recently
Org. Lett., Vol. 12, No. 6, 2010
1159