Rhodium-catalyzed allylic substitution with an acyl anion equivalent: Stereospecific construction of acyclic quaternary carbon stereogenic centers

Article abstract of DOI:10.1021/ja306602g

A highly regio- and stereospecific rhodium-catalyzed allylic alkylation of tertiary allylic alcohol derivatives with a cyanohydrin pronucleophile is described. This direct and operationally simple protocol provides a fundamentally novel approach toward the synthesis of α-quaternary substituted ketones and circumvents many of the inherent problems associated with conventional enolate alkylation reactions. The stereospecific variant of this reaction provides the enantiomerically enriched α-quaternary substituted allylic aryl ketone, which is a particularly challenging intermediate for more conventional enolate-based strategies.

Full text of DOI:10.1021/ja306602g

Communication
pubs.acs.org/JACS
Rhodium-Catalyzed Allylic Substitution with an Acyl Anion
Equivalent: Stereospecific Construction of Acyclic Quaternary Carbon
Stereogenic Centers
P. Andrew Evans,* Samuel Oliver, and Jungha Chae
Department of Chemistry, The University of Liverpool, Crown Street, Liverpool L69 7ZD, U.K.
S
* Supporting Information
Scheme 1. Comparison of Conventional Alkylation of a
Ketone Enolate with Allylic Substitution Using an Acyl
Anion
ABSTRACT: A highly regio- and stereospecific rhodium-
catalyzed allylic alkylation of tertiary allylic alcohol
derivatives with a cyanohydrin pronucleophile is described.
This direct and operationally simple protocol provides a
fundamentally novel approach toward the synthesis of α-
quaternary substituted ketones and circumvents many of
the inherent problems associated with conventional
enolate alkylation reactions. The stereospecific variant of
this reaction provides the enantiomerically enriched α-
quaternary substituted allylic aryl ketone, which is a
particularly challenging intermediate for more conven-
tional enolate-based strategies.
he stereoselective construction of all-carbon quaternary
T
stereogenic centers remains both an important and
challenging area of investigation.1 In this context, the synthesis
of this motif in acyclic systems remains arguably the greatest
challenge, despite the development of a number of elegant
approaches over the past decade.2 Among these methods,
enolate alkylation represents a particularly powerful synthetic
strategy and remains one of the most fundamentally important
carbon−carbon bond-forming reactions in synthetic organic
chemistry.3 This can be attributed to the ubiquity and versatility
of carbonyl compounds as synthetic intermediates, and their
presence in an array of pharmacologically important agents. For
example, chiral auxiliaries are frequently employed in this
manner, albeit this approach is now being superseded by
catalytic variants, which utilize organocatalysts and transition
metal complexes for the enantioselective alkylation of a wide
range of carbonyl compounds.4−6 Furthermore, these methods
have been extended to non-traditional electrophiles for the
arylation and vinylation of enolates.7−9 Nevertheless, the
stereocontrolled construction of acyclic quaternary carbon
centers via enolate alkylation has a number of inherent
limitations (Scheme 1A). For instance, the ability to
regioselectively form an enolate from an unsymmetrical ketone,
polyalkylation via enolate equilibration, and the necessity to
employ simple electrophiles adversely impact the chiral and
achiral versions of this venerable process. Additionally, this
process generally requires a specific enolate geometry, since the
origin of enantioselectivity is defined by a combination of
enolate geometry and π-facial selectivity in the alkylation.
We envisioned an alternative strategy for the construction of
enantiomerically enriched acyclic α-quaternary substituted
ketones (Scheme 1B), which involves the stereospecific
rhodium-catalyzed allylic substitution of chiral nonracemic
acyclic tertiary allylic carbonates with an acyl anion
equivalent.10−13 Hence, a critical component for the successful
implementation of this process is the ability to intercept the
putative rhodium-enyl intermediate in an analogous manner to
the corresponding secondary allylic carbonates.14,15 The
potential advantage to this strategy is the ability to employ an
array of diverse substituents (R1/R2), which obviates the
inherent stereoelectronic restrictions associated with conven-
tional electrophiles. In this context, we expect the inherent
stability of the nucleophile to be an important component,
since a stabilized nucleophile would provide retention of
configuration, whereas an unstabilized version would promote
direct attack on the metal center and afford the opposite
enantiomer.14 Herein, we now describe the successful
implementation of the first regioselective rhodium-catalyzed
allylic alkylation of tertiary allylic carbonates with an acyl anion
equivalent to facilitate the construction of acyclic α-quaternary
substituted ketones and the first stereospecific example using a
chiral nonracemic allylic carbonate (Scheme 1).
In accord with our hypothesis, the tert-butyldimethylsilyl-
protected cyanohydrin 2 was selected as the acyl anion
equivalent,16 which is readily prepared by the addition of tert-
butyldimethylsilyl cyanide to the corresponding aldehyde17 and
unmasked in the presence of fluoride. Additionally, we
envisioned that the anion derived from 2 would be sufficiently
Received: July 6, 2012
Published: November 16, 2012
© 2012 American Chemical Society
19314
dx.doi.org/10.1021/ja306602g | J. Am. Chem. Soc. 2012, 134, 19314−19317

Journal of the American Chemical Society
Communication
stabilized to permit the double inversion process outlined in
Scheme 1B.18 Treatment of the tertiary allylic carbonate 1a
with the lithium salt of 2a (Y = Ph) in the presence of
[RhCl(COD)]2, modified with triphenyl phosphite, followed by
the addition of tetra-n-butylammonium fluoride (TBAF)
furnished the α-quaternary substituted acyclic aryl ketone 4a
in 77% yield, albeit with modest regioselectivity (Table 1, entry
Table 2. Scope of the Rhodium-Catalyzed Allylic
Substitution Using the Cyanohydrin 2
a b c
, ,
Table 1. Optimization of the Rhodium-Catalyzed Allylic
a
Substitution Using the Cyanohydrin 2a (Y = Ph)
temp
(°C)
b/l for
yield of 4a
b
c
entry
phosphite L
L:M
3a
(%)
1
2
3
4
5
6
7
8
9
P(OPh)3
P(OPh)3
4:1
3:1
2:1
2:1
2:1
2:1
2:1
2:1
2:1
21
21
5:1
7:1
77
82
82
78
82
54
84
82
93
P(OPh)3
21
7:1
P(OMe)3
21
4:1
P(OCH2CF3)3
P(OTBS)3
P(O-2,4-ditBuC6H3)3
P(O-2,4-ditBuC6H3)3
P(O-2,4-ditBuC6H3)3
21
6:1
21
10:1
11:1
17:1
≥19:1
21
0
−10
a
All reactions were performed on a 0.5 mmol reaction scale using 2.5
mol % [Rh(COD)Cl]2, 1.3 equiv of 2a, and 1.8 equiv of LiHMDS in
THF for ca. 16 h, which was followed by the addition of 2.5 equiv of
b
TBAF at room temperature. Regioselectivity was determined by 500
MHz 1H NMR analysis of the crude reaction mixtures before
deprotection of the cyanohydrin adduct 3a, which is obtained as a 1:1
c
mixture of diastereoisomers.19 GC yields of 4a.
1).19 Although prior studies demonstrated that the ligand
stoichiometry was an important factor for improving selectivity,
a reduction in ligand stoichiometry provided only a modest
improvement in the formation of the branched isomer (entries
2 and 3).20 A particularly attractive feature of the rhodium-
catalyzed allylic substitution is the ability to readily modify the
stereoelectronics of the phosphite ligand, which allows ligands
to be readily interchanged without having to prepare a new
precatalyst. In this context, modifying the phosphite led to
further improvement in the branched selectivity (entries 4−7),
wherein the bulky phosphites provide optimal selectivity (entry
6 and 7). Although the bulky tris(tert-butyldimethylsilyl)-
phosphite and tris(2,4-di-tert-butylphenyl)phosphite ligands
provided similar levels of regioselectivity, the latter provides a
more efficient process. Finally, the selectivity could be further
improved by lowering the temperature to −10 °C (entries 8
and 9), which afforded the acyclic ketone 4a in 93% yield and
with ≥19:1 regioselectivity.
Table 2 summarizes the application of the optimized reaction
conditions (Table 1, entry 9) to a range of racemic tertiary
allylic carbonates and aryl cyanohydrin derivatives. The
alkylation is clearly tolerant of a range of electron-poor and
electron-rich aryl cyanohydrins (entries 1−5). Additionally, the
allylic carbonate provides access to an array of substituents,
some of which would prove extremely challenging for
conventional alkylation reactions. For instance, simple alkyl
and alkenyl groups afford the acyclic ketones 4 in good to
excellent yield with exquisite selectivity for the branched isomer
a
All reactions were performed on a 0.5 mmol reaction scale using 2.5
mol % [Rh(COD)Cl]2, 10 mol % P(O-2,4-di-tBuC6H3)3, 1.3 equiv of
2, and 1.8 equiv of LiHMDS in THF (5 mL) at −10 °C for ca. 16 h,
followed by the addition of 2.5 equiv of TBAF at room temperature.
b
1
Regioselectivity was determined by 500 MHz H NMR analysis of
c
d
the isolated aryl ketones 4. Isolated yields. P(OTBS)3 was used.
e
f
P(OPh)3 was used. Crude selectivity on the cyanohydrin adduct 3.
(entries 6−8). The ability to directly install gem-dimethyl
groups is particularly attractive, since this can often be quite
challenging for conventional alkylation reactions. Nevertheless,
the most impressive feature, as noted above, is the ability to
introduce branched and functionalized alkyl groups (entries 9−
15), which are notoriously challenging electrophiles in
conventional enolate alkylation reactions. Although these
substrates sometimes afford slightly reduced selectivity, the
ability to tailor the phosphite to improve reactivity and
selectivity makes this a particularly attractive process. Addi-
tionally, the introduction of a simple vinyl group provides an
extremely useful functional handle. Overall, this work provides a
versatile alternative to conventional enolate alkylation, which is
likely to provide a convenient disconnection for an array of synthetic
applications.
In order to showcase the synthetic utility of this trans-
formation, we elected to examine the stereospecific version
(Scheme 2). Although the alkylation of chiral nonracemic
secondary carbonates is well established for a range of stabilized
19315
dx.doi.org/10.1021/ja306602g | J. Am. Chem. Soc. 2012, 134, 19314−19317

Journal of the American Chemical Society
Communication
Scheme 2. Stereospecific Allylic Substitution through a
Double Inversion Process Using a Stabilized Cyanohydrin
Pronucleophile
ASSOCIATED CONTENT
* Supporting Information
Experimental procedures and spectral data for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
andrew.evans@liverpool.ac.uk
■
Notes
The authors declare no competing financial interest.
and unstabilized nucleophiles, the tertiary variants are
potentially more challenging.21 We envisioned that this process
would also proceed through an electronically biased configura-
tionally stable enyl intermediate in accord with our previous
studies, to facilitate the alkylation with overall retention of
configuration.15 Although enantiomerically pure tertiary allylic
alcohol derivatives were historically challenging intermediates, a
number of new methods have recently emerged for their
construction. For example, the enantiomerically enriched
tertiary allylic alcohol 5 was readily prepared using the method
reported by Aggarwal and co-workers.22
Interestingly, the conversion of the alcohol 5 to the
corresponding allylic carbonate resulted in significant rearrange-
ment to the achiral linear allylic carbonate. Although this
problem could be readily circumvented by in situ acylation and
subsequent allylic alkylation,23 the optimal reaction conditions
furnished the acyclic ketone 6 with poor stereospecificity, which
prompted the re-examination of the other phosphite ligands.
Gratifyingly, the electron-poor tris(2,2,2-trifluoroethyl)-
phosphite proved optimal in terms of both regio- and
stereospecificity, affording the acyclic ketone 6 in 87% yield
and with 91% conservation of enantiomeric excess (b/l≥ 19:1,
95:5 er). Although the chirality transfer is slightly lower than
that obtained with secondary derivatives, this reaction provides
proof-of-concept for this novel process. Additionally, the
absolute configuration of the major enantiomer was determined
by hydrogenation of the terminal alkene, to provide a
compound of known absolute configuration, thereby confirm-
ing that the process proceeds with overall retention.24
In conclusion, we have developed the first regio- and
stereospecific metal-catalyzed allylic alkylation of tertiary allylic
carbonates with a trialkylsilyl-protected cyanohydrin pronu-
cleophile. This process provides a direct and operationally
simple method for the construction of α-quaternary substituted
ketones, and thereby circumvents the inherent problems
associated with conventional enolate alkylation reactions.
Moreover, it provides broad substrate scope by facilitating the
introduction of branched and functionalized alkyl groups, which
are generally very challenging electrophiles for conventional
enolate alkylation reactions. The application of this method-
ology to the preparation of the enantiomerically enriched α-
quaternary substituted acyclic ketone is particularly significant
since it represents the first stereospecific rhodium-catalyzed
allylic alkylation of a chiral nonracemic tertiary allylic carbonate.
Future studies will focus on further development of this
important transformation, since we anticipate that this will be
applied to the preparation of complex bioactive biologically
important agents that contain quaternary carbon stereogenic
centers.
ACKNOWLEDGMENTS
■
We thank the National Institutes of Health (GM58877) for
generous financial support. We also thank the Royal Society for
a Wolfson Research Merit Award (PAE) and the EPSRC and
AstraZeneca (Alderley Park) for a Ph.D. studentship (S.O.).
We acknowledge Dr. Paul Kemmitt (AZ) for his support and
helpful discussions, and we are grateful to the EPSRC National
Mass Spectrometry Service Centre (Swansea, UK) for high-
resolution mass spectrometry.
REFERENCES
■
(1) For recent reviews on the catalytic enantioselective construction
of quaternary carbon stereogenic centers, see: (a) Corey, E. J.;
Guzman-Perez, A. Angew. Chem. Int. Ed. 1998, 37, 388. (b) Chris-
toffers, J.; Mann, A. Angew. Chem. Int. Ed. 2001, 40, 4591.
(c) Denissova, I.; Barriault, L. Tetrahedron 2003, 59, 10105.
(d) Trost, B. M.; Jiang, C. Synthesis 2006, 3, 369.
(2) For a recent review on the enantioselective construction of
acyclic quaternary carbon stereogenic centers, see: Das, J. P.; Marek, I.
Chem. Commun. 2011, 47, 4593.
(3) For a recent review on the enantioselective functionalization of
enolates, see: Stoltz, B. M.; Mohr, J. T. In Science of Synthesis; De Vries,
J. G., Molander, G. A., Evans, P. A., Eds.; Thieme: Stuttgart, Germany,
2010; Vol. 3, p 567.
(4) For chiral auxiliary-based methodologies, see: (a) Manthorpe, J.
M.; Gleason, J. L. Angew. Chem. Int. Ed. 2002, 41, 2338. (b) Abe, T.;
Suzuki, T.; Sekiguchi, K.; Hosokawa, S.; Kobayashi, S. Tetrahedron
Lett. 2003, 44, 9303. (c) Kummer, D. A.; Chain, W. J.; Morales, M. R.;
Quiroga, O.; Myers, A. G. J. Am. Chem. Soc. 2008, 130, 13231.
(5) For a recent example using organocatalysis, see: Brown, A. R.;
Kuo, W.-H.; Jacobsen, E. N. J. Am. Chem. Soc. 2010, 132, 9286.
(6) For recent examples of the asymmetric alkylation of acyclic
enolates, see: (a) Doyle, A. G.; Jacobsen, E. N. Angew. Chem. Int. Ed.
2007, 46, 3701. (b) Mukherjee, S.; List, B. J. Am. Chem. Soc. 2007, 129,
11336. (c) Trost, B. M.; Xu, J.; Schmidt, T. J. Am. Chem. Soc. 2009,
131, 18343.
(7) For a recent review on the transition metal-catalyzed enolate
arylation reactions, see: Johansson, C. C. C.; Colacot, T. J. Angew.
Chem. Int. Ed. 2010, 49, 676.
(8) For recent examples of transition metal-catalyzed enolate
arylation reactions, see: (a) Chen, G.; Kwong, F. Y.; Chan, H. O.;
Yu, W.-Y.; Chan, A. S. C. Chem. Commun. 2006, 1413. (b) Liao, X.;
Weng, Z.; Hartwig, J. F. J. Am. Chem. Soc. 2008, 130, 195. (c) Taylor,
A. M.; Altman, R. A.; Buchwald, S. L. J. Am. Chem. Soc. 2009, 131,
9900. (d) Ge, S.; Hartwig, J. F. J. Am. Chem. Soc. 2011, 133, 16330.
(9) For recent examples of transition metal-catalyzed enolate
vinylation reactions, see: (a) Chieffi, A.; Kamikawa, K.; Åhman, J.;
Fox, J. M.; Buchwald, S. L. Org. Lett. 2001, 3, 1897. (b) Huang, J.;
Bunel, E.; Faul, M. M. Org. Lett. 2007, 9, 4343.
(10) For an example of palladium-catalyzed allylic alkylation with a
cyanohydrin pronucleophile, see: Tsuji, J.; Shimizu, I.; Minami, I.;
Ohashi, Y.; Sugiura, T.; Takahashi, K. J. Org. Chem. 1985, 50, 1523.
(11) For related palladium-catalyzed allylic substitution reactions
with acyl silanes and stannanes, see: (a) Tsuji, Y.; Obora, Y.; Ogawa,
19316
dx.doi.org/10.1021/ja306602g | J. Am. Chem. Soc. 2012, 134, 19314−19317

Journal of the American Chemical Society
Communication
Y.; Imai, Y.; Kawamura, T. J. Am. Chem. Soc. 2001, 123, 10489.
(b) Tsuji, Y.; Obora, Y.; Nakanishi, M.; Tokunaga, M. J. Org. Chem.
2001, 67, 5835.
(12) For three-component palladium-catalyzed allylic substitution
reactions with carbon monoxide, see: (a) Tamaru, Y.; Yasui, K.;
Takanabe, H.; Tanaka, S.; Fugami, K. Angew. Chem. Int. Ed. 1992, 31,
645. (b) Sakurai, H.; Tanabe, K.; Narasaka, K. Chem. Lett. 1999, 309.
(13) For related metal-catalyzed allylic substitution reactions with
acyl anion equivalents, see: (a) Trost, B. M.; Dirat, O.; Dudash, J., Jr.;
Hembre, E. J. Angew. Chem. Int. Ed. 2001, 40, 3658. (b) Forster, S.;
̈
Tverskoy, O.; Helmchen, G. Synlett 2008, 2803. (c) Trost, B. M.;
Osipov, M.; Kaib, P. S. J.; Sorum, M. T. Org. Lett. 2011, 13, 3222.
(14) Evans, P. A.; Leahy, D. K. In Modern Rhodium-Catalyzed Organic
Reactions; Evans, P. A., Ed.; Wiley-VCH: Weinheim, Germany, 2005;
Ch. 10, pp 191−214.
(15) Evans, P. A.; Nelson, J. D. J. Am. Chem. Soc. 1998, 120, 5581.
(16) Mingo, P.; Zhang, S.; Liebeskind, L. S. J. Org. Chem. 1999, 64,
2145.
(17) Kurono, N.; Yamaguchi, M.; Suzuki, K.; Ohkuma, T. J. Org.
Chem. 2005, 70, 6530.
(18) Although the carbanion of the aryl-substituted cyanohydrin 2a
(Y = Ph) is stable under these reaction conditions, the carbanions of
the corresponding alkyl (Y = Pr) and alkenyl (Y = CHCH2)
derivatives are unstable, which leads to decomposition of the
nucleophile prior to alkylation.
(19) Regioselectivity was determined prior to desilylation due to the
reactivity of the linear acyclic aryl ketone under the reaction
conditions.
(20) Evans, P. A.; Nelson, J. D. Tetrahedron Lett. 1998, 39, 1725.
(21) For an example of a stereospecific iron-catalyzed allylic
substitution of a tertiary allylic carbonate, see: Jegelka, M.; Plietker,
B. Org. Lett. 2009, 11, 3462.
(22) (a) Stymeist, J. L.; Bagutski, V.; French, R. M.; Aggarwal, V. K.
Nature 2008, 456, 778. (b) Bagutski, V.; French, R. M.; Aggarwal, V.
K. Angew. Chem. Int. Ed. 2010, 49, 5142.
(23) Evans, P. A.; Leahy, D. K. J. Am. Chem. Soc. 2003, 125, 8974.
(24) (a) Kopecky, K. R.; Mojelsky, T. W.; Gillan, T.; Barry, J. A.;
Lopez Sastre, J. A. Can. J. Chem. 1977, 55, 1001. (b) Ros, A.; Aggarwal,
V. K. Angew. Chem. Int. Ed. 2009, 48, 6289.
19317
dx.doi.org/10.1021/ja306602g | J. Am. Chem. Soc. 2012, 134, 19314−19317