Rhodium-catalyzed asymmetric allylic substitution with boronic acid nucleophiles

Article abstract of DOI:10.1021/ol061777l

An enantio-, regio-, and diastereoselective rhodium(I)-catalyzed desymmetrization of a meso-cyclic allylic dicarbonate with organoboronic acid nucleophiles is described. The rhodium(I) catalyst formed in situ from [Rh(cod)OH]₂ and Xyl-P-PHOS al

Full text of DOI:10.1021/ol061777l

ORGANIC
LETTERS
2006
Vol. 8, No. 20
4569-4572
Rhodium-Catalyzed Asymmetric Allylic
Substitution with Boronic Acid
Nucleophiles
Frederic Menard, Timothy M. Chapman, Chris Dockendorff, and Mark Lautens*
DaVenport Laboratories, Department of Chemistry, UniVersity of Toronto, Toronto,
Ontario, Canada M5H 3H6
mlautens@chem.utoronto.ca
Received July 19, 2006
ABSTRACT
An enantio-, regio-, and diastereoselective rhodium(I)-catalyzed desymmetrization of a meso-cyclic allylic dicarbonate with organoboronic acid
nucleophiles is described. The rhodium(I) catalyst formed in situ from [Rh(cod)OH]2 and Xyl-P-PHOS allowed the SN2 allylic substitution
product to be obtained with a range of arylboronic acids in enantiomeric excesses of up to 92% with regioselectivities of up to >20:1.
′
The utility of the asymmetric allylic substitution reaction1
has been the subject of intense study, as its application to
the synthesis of a range of natural products demonstrates.2
Usually, the enantiodetermining step is either the ionization
of an enantiotopic leaving group or nucleophilic addition to
a chiral π-allylmetal complex.1b Although asymmetric induc-
tion with palladium catalysts and soft nucleophiles, especially
malonates, is well documented, there are very few examples
of the use of unstabilized, hard carbon nucleophiles.3
Recent advances have been made in the asymmetric allylic
substitution reaction for the introduction of hard nucleophiles
through copper-catalyzed reactions with Grignard reagents
or organozincs, with both acyclic4 and cyclic5 substrates. As
these methods use mainly alkyl nucleophiles, we report
herein complementary work that makes use of air and
moisture tolerant organoboronic acids to achieve the regio-
and enantioselective allylic substitution of meso-cyclic di-
carbonates with a Rh(I)-biarylphospine catalytic system.
As an extension to our recently reported asymmetric
rhodium-catalyzed ring-opening of oxabicyclic alkenes with
boronic acids (eq 1),6 we attempted to use similar conditions
for the desymmetrization of meso-cyclic allylic diol deriva-
tives (Scheme 1). The ready availability, low toxicity, and
high functional group tolerance of boronic acids makes this
reaction especially appealing.
Precedents by Evans in the case of acyclic chiral allylic
alcohol derivatives have shown the stereospecific nature of
the rhodium-catalyzed asymmetric allylic alkylation with
both soft7 and hard8 nucleophiles. Also, Hayashi reported a
(1) (a) Trost, B. M.; Van Vranken, D. L. Chem. ReV. 1996, 96, 6, 395.
(b) Tsuji, J. Palladium Reagents and Catalysts: InnoVations in Organic
Synthesis; Wiley: New York, 1996. (c) Pfaltz, A.; Lautens, M. In
ComprehensiVe Asymmetric Catalysis; Jacobsen, E. N., Pfaltz, A., Yama-
moto, H., Eds.; Springer: Berlin, 1999; Vol. 2, p 833. (d) Paquin, J.-F.;
Lautens, M. In ComprehensiVe Asymmetric Catalysis; Jacobsen, E. N.,
Pfaltz, A., Yamamoto, H., Eds.; Springer: Berlin, 2004; Vol. 2, p 73.
(2) For reviews, see: (a) Trost, B. M. J. Org. Chem. 2004, 69, 5813. (b)
Trost, B. M.; Crawley, M. L. Chem. ReV. 2003, 103, 2921. (c) Graening,
T.; Schmalz, H.-G. Angew. Chem., Int. Ed. 2003, 42, 2580. (d) Trost, B.
M. Chem. Pharm. Bull. 2002, 50, 1.
(4) (a) Tissot-Croset, K.; Polet, D.; Alexakis, A. Angew. Chem., Int. Ed.
2004, 43, 2426. (b) Hoveyda, A. H.; Kacprzynski, M. A. J. Am. Chem.
Soc. 2004, 126, 10676. (c) Lopez, F.; van Zijl, A. W.; Minnaard, A. J.;
Feringa, B. L. Chem. Commun. 2006, 409.
(5) (a) Piarulli, U.; Claverie, C.; Daubos, P.; Gennari, C.; Minnaard, A.
J.; Feringa, B. L. Org. Lett. 2003, 5, 4493. (b) Piarulli, U.; Daubos, P.;
Claverie, C.; Roux, M.; Gennari, C. Angew. Chem., Int. Ed. 2003, 42, 234.
(6) (a) Lautens, M.; Dockendorff, C.; Fagnou, K.; Malicki, A. Org. Lett.
2002, 4, 1311. (b) For a review of selected ring-openings prior to 2003,
see: Lautens, M.; Fagnou, K.; Hiebert, S. Acc. Chem. Res. 2003, 36, 48.
(7) (a) Evans, P. A.; Nelson, J. D. J. Am. Chem. Soc. 1998, 120, 5581.
(a) Evans, P. A.; Lawler, M. J. J. Am. Chem. Soc. 2004, 126, 8642. (b)
Evans, P. A.; Leahy, D. K. J. Am. Chem. Soc. 2003, 125, 8974.
(3) (a) Indolese, A. F.; Consiglio, G. Organometallics 1994, 13, 2230.
(b) Chung, K.-G.; Miyake, Y.; Uemura, S. J. Chem. Soc., Perkin Trans. 1
2000, 15.
10.1021/ol061777l CCC: $33.50
© 2006 American Chemical Society
Published on Web 09/01/2006

Scheme 1
Scheme 3
highly enantioselective alkylation with malonate nucleophile
using a chiral rhodium catalyst.9 But there have been no
reports of the use of chiral rhodium catalysts with nonsta-
bilized carbon nucleophiles for the asymmetric allylic
substitution reaction, despite its efficient use in conjugate
addition reactions.10
Meso-cyclic allylic diol derivatives are challenging sub-
strates for this reaction due to the possibility of competiting
reaction pathways. Indeed, the allylic displacement reaction
may take place via an SN2- or SN2′-type substitution, with
overall inversion or retention of stereochemistry in both cases
(Scheme 2).
temperature. The resulting regioisomeric alcohols were easily
separated by chromatography and their enantiomeric excesses
were determined by chiral HPLC (Chiralcel OD-H column,
76% ee for 4, 72% ee for 5).
Optimization of the reaction conditions was then under-
taken in an effort to improve the conversion, regioselectivity,
and enantioselectivity. Among the different parameters
studied (leaving group,14 base,15 solvent,16 temperature,
concentration, and chiral ligand), variation of the ligand had
the most significant effect. Biarylbisphosphine ligands proved
to be the most effective class for this reaction (Table 1).17
Scheme 2
Table 1. Desymmetrization with Selected Biaryl Ligands
We found that cis-1,4-cyclopentene-diethyl carbonate 111
could be desymmetrized in a highly diastereo- and enantio-
selective manner. Initial studies used a [Rh(cod)OH]2
catalyst12 and phenylboronic acid with (S)-BINAP as the
ligand in the presence of Cs2CO3 in THF at 50 °C (Scheme
3). The 1,2-trans-substituted product 2 was obtained as the
major product, in a 2:1 ratio over the 1,4-trans-substituted
product 3. Importantly, none of the diastereomeric cis isomers
were observed. Although the conversion of starting material
was low, the reaction profile was clean: only 2, 3, and
unreacted starting material were present in the mixture. While
the carbonate products could not be separated at this stage,
they were readily hydrolyzed to yield the corresponding
known alcohols 45b and 513 by simple methanolysis at room
entry ligand conversiona (%)
2/3a
2:1
2:1
1:1
>20:1
13:1
>20:1
% ee (2)b % ee (3)b
1
2
3
4
5
6
6ac
6bc
7c
8ad
8bc
8bd
45
59
60
74
65
70
76
82
90
82
88
90
72
60
>98
>98
1
a Determined by H NMR spectroscopy. b Determined by chiral HPLC
(Daicel Chiralcel OD-H column) on the deprotected products. c Conditions:
substrate (1.0 equiv), PhB(OH)2 (2.0 equiv), Rh (5 mol %), ligand (10
mol %), Cs2CO3 (1.0 equiv), THF, 50 °C. d Conditions as for footnote c
except ligand (12 mol %).
(8) Evans, P. A.; Uraguchi, D. J. Am. Chem. Soc. 2003, 125, 7158.
(9) Hayashi, T.; Okada, A.; Suzuka, T.; Kawatsura, M. Org. Lett. 2003,
5, 1713.
Xyl-BINAP 6b gave an increased conversion versus BINAP
6a, and it showed a small improvement in ee for 2, but the
(10) Hayashi, T.; Yamasaki, K. Chem. ReV. 2003, 103, 2829.
(11) Arnau, N.; Corte´s, J.; Moreno-Man˜as, M.; Pleixats, R.; Villarroya,
M. J. Heterocycl. Chem. 1997, 34, 233.
(12) The use of [Rh(cod)OH]2 was vital to the success of the reaction;
[Rh(cod)Cl]2 with aqueous base gave no desired product. See the Supporting
Information for catalyst screening details.
(13) (a) Kobayashi, Y.; Murugesh, M. G.; Nakano, M.; Takahisa, E.;
Usmani, S. B.; Ainai, T. J. Org. Chem. 2002, 67, 7110. (b) Tueting, D. R.;
Echavarren, A. M.; Stille, J. K. Tetrahedron 1989, 45, 979. (c) Marino, J.
P.; Ferna´ndez de la Pradilla, R.; Laborde, E. J. Org. Chem. 1987, 52, 4898.
(14) Other leaving groups giving poorer conversion, regioselectivity, and/
or enantioselectivity were -OC(O)C6H5, -OC(O)C6H4NO2, -OC(O)CH3,
-OC(O)OCH(CF3)2, and -OP(O)(OEt)2.
(15) Other bases giving poorer conversion, regioselectivity, and/or
enantioselectivity were Na2CO3, KOH, Et3N, DIPEA, and KF.
(16) Other solvents giving poorer conversion, regioselectivity, and/or
enantioselectivity were toluene, dioxane, acetonitrile, DMF, methanol, and
2-propanol.
(17) See the Supporting Information for complete ligands screening.
4570
Org. Lett., Vol. 8, No. 20, 2006

regioselectivity remained poor (entry 2). Excellent enantio-
selectivity was obtained with Cl-OMe-BIPHEP 7 (entry 3),
although no regioselectivity was achieved. Xyl-P-PHOS 8b18
gave good regioselectivity and high ee with acceptable
conversion (entry 4). With a slight increase in ligand loading,
the conversion was increased to 70%, the regioselectivity
for 2:3 was >20:1, and an enantiomeric excess of 90% for
2 was achieved (entry 5). P-PHOS 8a gave good conversion
and regioselectivity but poorer enantioselectivity than Xyl-
P-PHOS (entry 6).
The reaction is either complete or stops within the first
hour. All attempts to lower the rhodium loading have so far
led to significantly decreased yields. The relatively high
catalyst loading arises from the rapid loss of activity,
particularly with electron-rich arylboronic acids. Slow ad-
dition of PhB(OH)2 actually led to a decrease in conversion.
Also, premixing and heating of the rhodium catalyst and
PhB(OH)2 for 30 min at 50 °C before the substrate was added
did not impede the reaction, but lower conversion was
observed.
enantioselectivity (entries 2-4). 4-Halosubstituted boronic
acids gave good regioselectivity and enantioselectivities
(entries 5-7), albeit with lower conversions than for phe-
nylboronic acid. o-Tolylboronic acid gave low conversion
and decreased regioselectivity but high ee for the 1,2-product
(entry 16). The m- and p-tolylboronic acids gave comparable
results to phenyl (entries 8 and 13), but with lower ee.
Electron-rich arylboronic acids gave good regioselectivity
and ee but poor conversion (entry 9-11), whereas 4-dim-
ethylaminophenylboronic acid gave no reaction (not shown).
The 3-substituted arylboronic acids all gave good conversion
and ee (entries 11-13). The substrates that did not react (4-
dimethylaminophenylboronic acid, 4-pyridylboronic acid) all
contain a nitrogen atom that may interfere with the reaction
by coordination to the rhodium complex, preventing catalyst
turnover. 2-Naphthylboronic acid gave excellent results (entry
14), whereas 1-naphthylboronic acid suffered from a lack
of regioselectivity and gave the 1,2- product in lower ee
(entry 15).
The mechanism of the reaction is still unclear, as at least
two major catalytic cycles can be considered. The first
involves an enantioselective ionization of the leaving group
by rhodium to generate a σ-enyl cation intermediate C, as
suggested by Evans (Scheme 4).7,8 The formation of the
We next investigated the scope of the reaction with respect
to the boronic acid (Table 2). Arylboronic acids substituted
Table 2. Scope of Desymmetrization with Arylboronic Acids
Scheme 4. Proposed Catalytic Cycle Proceeding via a σ-Enyl
Intermediate
entry product
R
yieldsa (%)
2/3a
% ee (2)b,c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
2a
2b
2c
2d
2e
2f
2g
2h
2i
2j
2k
2l
2m
2n
2o
2p
H
87
95
94
86
46
53
35
70
32
49
63
87
78
78
50
25
18:1
>20:1
>20:1
13:1
7:1
13:1
10:1
20:1
>20:1
>20:1
>20:1
10:1
92
90
88
88
86
90
86
84
84
89
92
90
92
90
70
92
4-CO2Me
4-Ac
4-CF3
4-F
4-Cl
4-Br
4-Me
4-NHBoc
4-OMe
3-OMe
3-Cl
3-Me
20:1
>20:1
1:1
2-naphthyl
1-naphthyl
2-Me
6:1
rhodium(III) complex C can arise from a transmetalation
occurring first to generate the Rh(I)-Ar species A, followed
by ionization to generate the σ-enyl intermediate C. On the
other hand, the ionization by the presumed Rh-OH species
could give the Rh(III)-OH intermediate B, followed by a
transmetalation step with the boronic acid. Once the 1,2-σ-
enyl complex C is formed, it could reductively eliminate to
yield the 1,2-substituted product 2 or equilibrate to the
isomeric 1,4-σ-enyl complex D, from which a reductive
elimination would lead to the 1,4-substituted product 3.
Accordingly, the rate of isomerization versus the rate of
reductive elimination would be pivotal in determining the
a Isolated yields of carbonates, average of at least two runs. b Determined
by chiral HPLC (Chiralcell OD-H column) on the deprotected products.
c Absolute stereochemistry of products was determined by the MTPA
Mosher ester method.
with strongly electron-withdrawing groups in the 4-position
gave excellent conversion and high regioselectivity and
(18) Wu, J.; Kwok, W. H.; Lam, K. H.; Zhou, Z. Y.; Yeung, C. H.;
Chan, A. S. C. Tetrahedron Lett. 2002, 43, 1539. Xyl-P-PHOS 8b and
P-PHOS 8a are available from Strem Chemicals, Inc., in both (R)- and
(S)-enantiomeric forms.
Org. Lett., Vol. 8, No. 20, 2006
4571

regioselectivity of the reaction. At this point, we cannot
determine if both ionization mechanisms are operative.
A second option involves a carborhodation of the alkene
as the key step (Scheme 5). Transmetalation of the aryl
need to occur on the face of the olefin anti to the leaving
groups. The leaving groups might prevent metalation from
the syn face of the olefin and would not coordinate to the
metal. However, this mechanism cannot explain the forma-
tion of the competing 1,4-adducts. If this mechanism is
operational, the observed 1,4-product would imply that there
is at least a partial contribution from the proposed σ-enyl-
type mechanism.
Scheme 5. Proposed Catalytic Cycle Proceeding by a
Carborhodation Mechanism
In summary, a highly enantio-, regio-, and diastereose-
lective meso-desymmetrization with a rhodium(I) catalyst and
readily available organoboronic acid nucleophiles has been
developed. The enantioselectivities in this reaction are
consistently good to excellent. We are working to improve
the problematic cases by examining alternative nucleophiles
and catalysts. Mechanistic investigations, as well as the use
of nonaromatic nucleophiles and the extension to acyclic
alkenes substrates, are the focus of ongoing efforts.
Acknowledgment. We thank NSERC, Merck Frosst
Canada, and the University of Toronto for support of our
programs. T.M.C. thanks The Royal Society, U.K., for a
postdoctoral fellowship. C.D. would like to thank the W. C.
Sumner foundation for a graduate fellowship. We thank
Solvias AG and Digital Speciality Chemicals for providing
some of the ligands used in this study.
moiety from boron to rhodium would form A, followed by
a stereo- and enantioselective carborhodation which would
generate the key intermediate E. Subsequent â-alkoxy
elimination would then yield the 1,2-product. This mecha-
nism is based on our previous work for the Rh-catalyzed
ring-opening reaction in more strained systems.6 Interestingly,
the desymmetrization of the cis-cyclopentene-1,4-diols yields
exclusively the trans products, compared to the exclusive
cis products observed with the ring-opening chemistry. To
explain the stereochemistry, the carborhodation step would
Supporting Information Available: Experimental de-
tails, additional data tables, and characterization data (includ-
ing 1H NMR, HR-MS, and HPLC conditions). This material
is available free of charge via the Internet at http://pubs.acs.org.
OL061777L
4572
Org. Lett., Vol. 8, No. 20, 2006