Enantioselective α-arylation of cyclohexanones with diaryl iodonium salts: Application to the synthesis of (-)-epibatidine

Article abstract of DOI:10.1002/anie.200501745

(Chemical Equation Presented) Short cut: Direct asymmetric α-arylation of prochiral ketones has been effected using chiral lithium amide bases and diaryl iodonium salts. The methodology has been employed in a short total synthesis of the alkaloid (-)-epibatidine (see scheme).

Full text of DOI:10.1002/anie.200501745

Communications
Synthetic Methods
(Scheme 2). Herein, we describe our success in achieving
these goals.
DOI: 10.1002/anie.200501745
Enantioselective a-Arylation of Cyclohexanones
with Diaryl Iodonium Salts: Application to the
Synthesis of (À)-Epibatidine**
Varinder K. Aggarwal* and Berit Olofsson
Scheme 2. Retrosynthesis of (À)-epibatidine.
The enantioselective introduction of electrophiles a to
carbonyl groups is a central topic in asymmetric synthesis.
Although asymmetric a-alkylations have been well devel-
oped, high enantioselectivity in a-arylation of ketones has
only been achieved in a very limited number of cases. These
involve arylation of trisubstituted enolates using binap–Pd
complexes which gives non-epimerizable, tetrasubstituted
ketones.[1,2] However, this strategy cannot be used in the
desymmetrization/arylation of cyclic ketones, since even if the
enolate was generated enantioselectively, the ketone product
would racemize the starting enolate (through proton transfer)
at the high temperatures required for arylation.
As possible electrophilic sources of aryl reagents, we
considered diaryl iodonium(iii) salts since scattered reports
existed of couplings with silyl enol ethers,[5] ketones,[6] and
malonates,[7] albeit with varying yields. The scope and
limitations of these couplings had not been investigated,
and no successful asymmetric reactions have been reported.[8]
Indeed, the dearth of examples of enolate arylations with
diaryl iodonium(iii) salts encouraged us to explore this area
further.
The coupling conditions were optimized with cyclohex-
anone (2a) and the commercially available diphenyl iodo-
nium salt 3 (Scheme 3).[9] Under optimized conditions aryla-
An alternative strategy involves a-arylation of b-ketoest-
ers with aryl lead reagents (Scheme 1).[3] This strategy can be
Scheme 1. Possible asymmetric routes toward 2-aryl cyclohexanones.
Scheme 3. a-Arylation of cyclohexanone (2a); HDMS=hexamethyldi-
silazide.
rendered asymmetric through 1) desymmetrization/b-
ketoester formation,[4] 2) arylation,[3] and 3) decarboxylation
and subsequent equilibration.
However, a direct asymmetric arylation of cyclohexa-
nones would clearly be superior. If this could be achieved, we
envisaged that this strategy could be applied to a short
asymmetric synthesis of the alkaloid (À)-epibatidine (1)
tion of cyclohexanone (2a) gave 4a in 83% yield (entry 1,
Table 1). It should be noted that 2 equivalents of base are
required; the second equivalent is used to deprotonate the
product ketone.[10] Despite the fact that part way through the
reaction enolates A and B are both present, no bis-arylated
product was observed under our optimized conditions
(Scheme 3).[11]
[*] Prof. V. K. Aggarwal, Dr. B. Olofsson
School of Chemistry
When applied to 4-substituted cyclohexanones 2b–d, the
coupling reaction afforded 2,4-disubstituted products 4b–d in
high yields and with high diastereoselectivities (Table 1).[12]
As expected, arylation of tert-butyl ketone 2b resulted in 4b
with high cis selectivity (entry 2), whilst ketone 2c yielded a
5:1 cis/trans mixture (entry 3). In contrast, reaction of the 4-
silyloxy ketone 2d led to 4d with high trans selectivity, the 4-
OTBDMS occupying the axial position (entry 4). It has been
suggested that an attractive electrostatic interaction and/or an
n,p* orbital overlap between the silyloxy group and the
carbonyl carbon favors the axial orientation of 4-alkoxy
substituents and so the trans isomer is likely to be the
University of Bristol
Cantock’s Close, Bristol, BS81TS (UK)
Fax: ( +44)117-929-8611
E-mail: v.aggarwal@bristol.ac.uk
Dr. B. Olofsson
Department of Organic Chemistry
Stockholm University (Sweden)
[**] We thank Merck and the Swedish Research Council for financial
support.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
5516
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5516 –5519

Angewandte
Chemie
Table 1: Couplings of ketones 2 and diphenyl iodonium salt 3.[a]
this occurred, lower enantioselectivity would have been
expected because the second equivalent of base would have
deprotonated the ketone at À458C instead of À1188C.
Evidently, the second equivalent of base acted to directly
deprotonate the arylated ketone.
Having established the asymmetric methodology for the
a-arylation of cyclic ketones, we sought to apply it to a short
enantioselective synthesis of epibatidine (1; Scheme 2). This
alkaloid, which was isolated from the Ecuadorian poison frog
Epipedobates tricolor in 1993, is 400 times more potent than
morphine as an analgesic, and is also an extremely potent
agonist of the nicotinic acetylcholine receptor.[21] Numerous
syntheses of epibatidine have since been published, most of
which are racemic.[22] One commonly employed strategy to
introduce the pyridyl group involves an organometallic
coupling with a 2-halosustituted cyclohexanone,[23,24] a reac-
tion that cannot be rendered asymmetric (since the starting
material is already chiral). As depicted in our retrosynthesis
(Scheme 2), the direct coupling of a 4-substituted cyclo-
hexanone with a pyridyl iodonium salt not only shortens the
synthetic pathway but also renders the synthesis asymmetric.
Our strategy required a pyridyl iodonium salt that had not
been previously prepared. After some experimentation,
pyridyl salt 7 was synthesized in good yield from compound
Entry
Ketone (R)
Base
Yield of 4 [%][b]
cis/trans
1
2
3
4
5
6
2a (H)
2b (tBu)
2c (Ph)
2d (OTBDMS)
2e (NBoc2)
2 f (NHBoc)
LHMDS
LDA
LDA
LDA
LHMDS
LDA
83
69
76
–
>20:1
5:1
1:15
>20:1
10:1
70
18[c]
31[d]
[a] Conditions: The ketone was treated with LHMDS or LDA at À788C in
THF. Salt 3 was added in DMF and the reaction was stirred at À458C for
4–5 h. LDA=lithium diisopropylamide, TBDMS=tert-butyldimethylsilyl,
Boc=tert-butoxycarbonyl. [b] Yields of isolated products. [c] The major
product was Boc-transfer product 5, see text. [d] 2,2-Diphenyl-4-NHBoc
ketone (38%) and 2 f (30%) were also isolated.
thermodynamically more stable product.[13,14] Application of
this methodology to 4-aminoketone 2e[15] was complicated by
an unexpected intramolecular Boc-transfer reaction, yielding
enol ester 5 as the major product (37%) (entry 5 and
Scheme 4).[16] The Boc-transfer could be avoided by employ-
8
and commercially available 2-chloro-5-bromopyridine
(Scheme 6).[25]
Scheme 6. Synthesis of pyridyl iodonium salt 7.
Scheme 4. Boc-transfer product 5 and chiral base 6 used in asymmet-
ric reactions.
Various cyclohexanones were treated with chiral base 6
and coupled with salt 7, leading to the corresponding arylated
ketones with high enantioselectivities and moderate to good
yields (Table 2).[22,26] The pyridyl group on the iodonium salt
had a negative impact on the yield of the coupling of ketones
2a and 2b. Whilst the diphenyl iodonium salt 3 gave 4a in
83% yield, the pyridyl iodonium salt 7 only gave a 51% yield
of coupling product 9a (entry 1, Tables 1 and 2 respectively).
Fortunately ketones 2d and 2e, both possible precursors to
epibatidine, proved to be better substrates in the coupling
reaction with the pyridyl iodonium salt 7. Ketone 2d has been
previously employed in asymmetric desymmetrization reac-
tions.[19,27] When this substrate was treated with base 6 and
coupled with iodonium salt 7, 2-pyridyl ketones 9d were
obtained in good yield and high enantioselectivity (entry 3).
Interestingly, whilst the 2-phenyl ketone 4d was formed with
high trans diastereoselectivity (Table 1, entry 4), the 2-pyridyl
ketone 9d was produced with modest cis selectivity.[28]
ing ketone 2 f, but the yield of 4 f was moderate due to
competitive deprotonation of the carbamate nitrogen.[17] This
entry shows that enolate B (Scheme 3) can indeed react to
give disubstituted product when enolate A is absent, as the
major product in this case was the corresponding diphenyl
compound (38%).
Employing Simpkinsꢀ (R,R)-base (6; Scheme 4) instead of
LDA or LHMDS, we found that tert-butylcyclohexanone (2b)
gave the arylated product (2R,4S)-4b in 84% yield and
90% enantiomeric excess (Scheme 5).[18] The enantioselec-
tivity was in the range typical for this class of desymmetriza-
tion,[19,20] which clearly indicates that proton transfer between
the product and the starting enolate was not occurring. Had
Although asymmetric desymmetrization of ketone 2e had
not been previously reported, we expected that the steric bulk
provided by the N(Boc)2 moiety would lock the conformation
of the cyclohexanone ring and thus lead to good asymmetric
induction. Indeed, enantioselective deprotonation with base 6
and subsequent coupling afforded pyridyl ketone 9e in 41%
Scheme 5. Asymmetric coupling with Simpkins’ base.
Angew. Chem. Int. Ed. 2005, 44, 5516 –5519
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5517

Communications
Table 2: Asymmetric couplings with pyridyl iodonium salt 7.[a]
been transformed into epibatidine in a further five steps,[23,29]
thus providing a short and high-yielding formal synthesis of
(+)-epibatidine. The optical rotation of 10 revealed that
although the 4-OTBDMS group in 2d prefers the axial
conformation, it must be the equatorial conformation that
becomes deprotonated by the chiral base.[30]
Entry
1
Ketone
Product[b]
Yield [%][c]
ee [%]
2a
51
–
Alternatively, transformation of ketone 2e to naturally
occurring (À)-epibatidine could be performed in five straight-
forward steps (Scheme 8). 2-Pyridyl ketone 9e was formed
2
3
4
2b
2d
2e
51
90
70[d]
90
41[e]
86 (94)[f]
[a] Conditions: The ketone was treated with base 6 at À1188C in THF
(entry 1: LHMDS at À788C). A solution of salt 7 in DMF was added and
the reaction was stirred at À458C for 4 h. [b] Pyr=5-(2-chloropyridyl).
[c] Yields of isolated products. [d] cis/trans 2:1, 13% recovered 2d.
[e] Compound 5 isolated in 14% yield. [f] After recrystallization.
yield and 86% ee with complete cis selectivity (entry 4).
Furthermore, a single recrystallization enhanced the enantio-
meric excess of 9e to 94%. This result indicates that salt 7 is
more reactive than the diphenyl salt 3, as much less of the
Boc-transfer product was observed in this reaction.
Scheme 8. Synthesis of (À)-epibatidine. Ms=methanesulfonyl, TFA=
trifluoroacetic acid.
Having established the asymmetric a-arylation reaction
we proceeded to complete the synthesis of epibatidine with
ketones 2d and 2e. Commercially available ketone 2d was
with high enantiomeric excess, as discussed above, and
reduced to alcohol 11 using sodium borohydride at low
temperature with high diastereoselectivity (d.r. 7:1 in crude).
Compound 11 was converted into (À)-epibatidine ([a]2D0 =
À6.2 (c = 0.21, CH2Cl2); literature value:[22b] [a]D25 = À6.7
(c = 0.21, CH2Cl2)) by sequential mesylation, deprotection
and ring-closure in 90% yield over the three steps.[31] The
total synthesis was thus accomplished in six steps and 31%
overall yield from commercially available 2 f with essentially
complete control of diastereoselectivity, thus providing the
shortest and most efficient asymmetric route to this important
compound to date.[32]
employed in
a
formal synthesis of (+)-epibatidine
(Scheme 7). As discussed above, the asymmetric iodonium
In summary, a direct asymmetric a-arylation of cyclo-
hexanones has been developed. The method employs Simp-
kinsꢀ base to desymmetrize 4-substituted cyclohexanones by
asymmetric enolization followed by coupling with diaryl
iodonium salts, leading to 2-aryl ketones in moderate to good
yields and high enantioselectivities. The method expands the
scope of the desymmetrization strategy pioneered by Simp-
kins and Koga, and highlights the utility of iodonium salts for
aryl transfer reactions in synthesis. This approach has been
applied in a short and efficient total synthesis of (À)-
epibatidine and a formal synthesis of (+)-epibatidine.
Scheme 7. Formal synthesis of (+)-epibatidine.
coupling afforded 2-pyridyl ketone 9d as a kinetically formed
2:1 cis/trans mixture, which was equilibrated with DBU (1,8-
diazabicyclo[5.4.0]undec-7-ene) to afford a separable 1:4
mixture with trans-9d as the major diastereomer. Ketone
trans-9d was subsequently reduced with sodium borohydride
to give the corresponding equatorial alcohol 10 in 84%
isolated yield (d.r. 10:1 in crude). Alcohol 10 has previously
Received: May 20, 2005
Published online: July 29, 2005
Keywords: cyclohexanone · desymmetrization · enolates ·
.
epibatidine · iodonium salts
5518
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5516 –5519

Angewandte
Chemie
Lett. 2001, 42, 3947; d) S. Aoyagi, R. Tanaka, M. Naruse, C.
[1] a) J. Šhman, J. P. Wolfe, M. V. Troutman, M. Palucki, S. L.
Buchwald, J. Am. Chem. Soc. 1998, 120, 1918; b) T. Hamada, A.
Chieffi, J. Šhman, S. L. Buchwald, J. Am. Chem. Soc. 2002, 124,
1261.
Kibayashi, J. Org. Chem. 1998, 63, 8397; e) H. Kosugi, M. Abe,
R. Hatsuda, H. Uda, M. Kato, Chem. Commun. 1997, 1857.
[23] M. T. Barros, C. D. Maycock, M. R. Ventura, J. Chem. Soc.
Perkin Trans. 1 2001, 166.
[2] For racemic or achiral examples of Pd-catalyzed arylation of
ketones, see a) M. Kawatsura, J. F. Hartwig, J. Am. Chem. Soc.
1999, 121, 1473; b) B. C. Hamann, J. F. Hartwig, J. Am. Chem.
Soc. 1997, 119, 12382; c) J. M. Fox, X. Huang, A. Chieffi, S. L.
Buchwald, J. Am. Chem. Soc. 2000, 122, 1360; d) M. Palucki,
S. L. Buchwald, J. Am. Chem. Soc. 1997, 119, 11108.
[3] G. I. Elliott, J. P. Konopelski, M. M. Olmstead, Org. Lett. 1999, 1,
1867.
[24] B. M. Trost, G. R. Cook, Tetrahedron Lett. 1996, 37, 7485.
[25] A modified procedure was developed from a) P. J. Stang, B.
Olenyuk, K. Chen, Synthesis 1995, 937; b) F. M. Beringer, R. A.
Nathan, J. Org. Chem. 1969, 34, 685, see Supporting information.
An unsymmetrical pyridyl-phenyl iodonium salt was also
prepared, but this gave consistently lower yields in the couplings
due to moderate selectivity of pyridyl-group transfer compared
to phenyl transfer.
[26] The 2-chloro-5-iodopyridine formed from 7 can be recovered
and used in the synthesis of 7.
[27] M. Majewski, N. M. Irvine, J. MacKinnon, Tetrahedron: Asym-
metry 1995, 6, 1837.
[4] J. Busch-Petersen, E. J. Corey, Tetrahedron Lett. 2000, 41, 6941.
[5] a) T. Iwama, V. B. Birman, S. A. Kozmin, V. H. Rawal, Org. Lett.
1999, 1, 673; b) K. Chen, G. F. Koser, J. Org. Chem. 1991, 56,
5764.
[28] The moderate diastereoselectivity could be explained by a
diminished preference for the axial position in the enolate
intermediate A (Scheme 1), which is quenched to give the
product mixture.
[29] a) E. Albertini, A. Barco, S. Benetti, C. De Risi, P. Pollini, R.
Romagnoli, V. Zanirato, Tetrahedron Lett. 1994, 35, 9297; b) A.
Barco, S. Benetti, C. De Risi, P. Marchetti, G. P. Pollini, V.
Zanirato, Eur. J. Org. Chem. 2001, 975.
[30] Analytical data were in agreement with literature. Optical
rotation differed in magnitude with the literature but had the
same sign. Value observed by us: [a]D = À7.2 (c = 0.32, CH2Cl2);
literature value:[23] [a]D = À10.6 (c = 0.32, CH2Cl2).
[31] The conversion of alcohol 11 to epibatidine follows the sequence
described for the corresponding NHBoc compound.[24] Analyt-
ical data of 1 (see Supporting information) were in agreement
with those reported in ref. [22].
[6] a) P. Gao, D. L. Larson, P. S. Portoghese, J. Med. Chem. 1998, 41,
3091; b) P. Gao, P. S. Portoghese, J. Org. Chem. 1995, 60, 2276;
c) J. H. Ryan, P. J. Stang, Tetrahedron Lett. 1997, 38, 5061.
[7] C. H. Oh, J. S. Kim, H. H. Jung, J. Org. Chem. 1999, 64, 1338.
[8] One report of a coupling between a chiral diaryl iodonium salt
and cyclic b-keto esters bearing chiral auxiliaries resulted in
modest eeꢀs, see M. Ochiai, Y. Kitagawa, N. Takayama, Y.
Takaoka, M. Shiro, J. Am. Chem. Soc. 1999, 121, 9233.
[9] The iodide salt was also tested and gave similar yields to the
triflate salt.
[10] Both LiHMDS and LDA could be successfully employed,
LiHMDS gave generally slightly higher yields.
[11] Significant double arylation was obtained at low temperature
(up to 31%) but above À458C only monoarylation was
observed.
[12] The diastereoselectivity of the initial coupling is irrevelant, as
the product is immediately deprotonated by the second equiv-
alent of base, see Scheme 3. The diastereoselectivity is deter-
mined during the protonation (in the quench) under kinetic
conditions.
[32] Compare ref. [22]; a) 9 steps, 21% yield; b) 13 steps, 13% yield;
d) 10 steps, 5.9% yield.
[13] a) Y. Nagao, M. Goto, M. Ochiai, M. Shiro, Chem. Lett. 1990,
1503; b) K. W. Baldry, M. H. Gordon, R. Hafter, M. J. T.
Robinson, Tetrahedron Lett. 1976, 2589.
[14] In the parent ketone, the C4 proton appears as a triplet of triplets
in the 1H NMR spectrum, J = 5.2 and 4.7 Hz, indicative of a
predominant axial conformation of the silyloxy substituent.
[15] 2e was obtained from commercially available NHBoc ketone 2 f
by Boc-protection in quantative yield, see the Supporting
information.
[16] When 2e was treated with LDA without addition of iodonium
salt, compound 5 was isolated in 75% yield. The intramolecular
nature of the reaction was proved by a cross-over experiment
with cyclohexanone. Only 5 and cyclohexanone were isolated;
tert-butyl 2-oxocyclohexanecarboxylate was not observed.
[17] Ketone 2 f could be recovered in 30%. Use of 3 equiv base did
not increase the yield of 4 f, instead more byproducts were
formed.
[18] The enantiomeric excess was measured by chiral HPLC, see
Supporting information. Absolute configurations were assigned
by analogy with previous desymmetrizations of cyclohexanones
with Simpkinsꢀ base.
[19] P. OꢀBrien, J. Chem. Soc. Perkin Trans. 1 1998, 1439.
[20] a) N. S. Simpkins, Pure Appl. Chem. 1996, 68, 691; b) P. J. Cox,
N. S. Simpkins, Tetrahedron: Asymmetry 1991, 2, 1.
[21] For a comprehensive review see: H. F. Olivo, M. S. Hemenway,
Org. Prep. Proced. Int. 2002, 34, 1.
[22] For some recent asymmetric syntheses see: a) Y. Hoashi, T.
Yabuta, Y. Takemoto, Tetrahedron Lett. 2004, 45, 9185; b) D. A.
Evans, K. A. Scheidt, C. W. Downey, Org. Lett. 2001, 3, 3009;
c) G. Pandey, S. K. Tiwari, R. S. Singh, R. S. Mali, Tetrahedron
Angew. Chem. Int. Ed. 2005, 44, 5516 –5519
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5519