The use of vinyl sulfonium salts in the stereo-controlled asymmetric synthesis of epoxide- and aziridine-fused heterocycles: Application to the synthesis of (-)-balanol

Article abstract of DOI:10.1002/anie.200602782

(Chemical Equation Presented) Lift your ylides: An asymmetric, epoxy-annulation reaction mediated by a vinyl sulfonium salt converts aminoaldehydes and -ketones into fused heterocyclic epoxides. The methodology was extended to an aziridine-annulation reac

Full text of DOI:10.1002/anie.200602782

Communications
Annulation
DOI: 10.1002/anie.200602782
The Use of Vinyl Sulfonium Salts in the Stereo-
controlled Asymmetric Synthesis of Epoxide- and
Aziridine-Fused Heterocycles: Application to the
Synthesis of (À)-Balanol**
Matthew G. Unthank, Nigel Hussain, and
Varinder K. Aggarwal*
The asymmetric synthesis of epoxides fused to carbo- or
heterocycles by epoxidation methods is inherently challeng-
ing because the two enantiotopic faces of the corresponding
cyclic (cis) alkene are poorly differentiated. The use of chiral
sulfur ylides offers a complementary method for preparing
epoxides, but this method has essentially only been employed
to make acyclic epoxides.[1] We considered the possibility of
extending sulfur ylide reactions to make fused bicyclic
epoxides, and in particular to the challenging substrates
described above. We were particularly attracted by the
possibility of designing a general asymmetric annulation
process for the synthesis of N-heterocycles through the
reaction of a bifunctional aminoaldehyde with a chiral vinyl
sulfonium salt (Scheme 1). This type of reaction has been
studied by Jimenez and co-workers in relation to the synthesis
of mitomycins[2] but has not been explored beyond that.
If we could achieve such an annulation process, we
envisaged that it could be employed in a short synthesis of the
protein kinase C inhibitor balanol[3] from readily available
Scheme 1. An annulation reaction mediated by a vinyl sulfonium salt
for the synthesis of fused bicyclic heterocycles.
[*] M. G. Unthank, Prof. Dr. V. K. Aggarwal
School of Chemistry
University of Bristol
Cantock’s Close, Bristol, BS81TS (UK)
Fax: (+44)117-929-8611
E-mail: v.aggarwal@bristol.ac.uk
N. Hussain
GlaxoSmithKline
Old Powder Mills, Tonbridge, Kent, TN119AN (UK)
[**] We thank GlaxoSmithKline and the EPSRC for financial support, and
Dr. J.P.H. Charmant for X-ray analyses.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
7066
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7066 –7069

Angewandte
Chemie
precursors (Scheme 2). We began our investigation with the
diphenyl vinyl sulfonium salt[2b] 1 and a-,[4] b-,[5] and g-[6]
aminoaldehydes. Through extensive screening of solvents,
bases and N-substituents, we found that the use of DBU in
We explored reactions with a-aminoketones, b-amino-
aldehydes, and b-aminoketones (a-aminoaldehydes were not
tested as they would give meso epoxides) (Table 1). Under
optimized conditions (method B), the a-aminoketones 8a–
Table 1: The reaction of chiral and achiral vinyl sulfonium salts with
aminoaldehydes and -ketones.
Entry Substrate
1
Product
Method Yield [%] ee [%]
A
B
90
76
–
97
A
B
90
65
–
99
Scheme 2. Proposed retrosynthesis of balanol.
2
3
4
THF with sulfonamides was optimum, and the reaction gave
the fused heterocyclic epoxides 2 and 3 in good yields
(Scheme 3). The g-aminoaldehyde substrate 4 proved to be a
A
B
96
80
–
92
A
B
67
67
–
98
A
B
50%
50
–
88
5
6
A
B
77
30
–
86
Method A: Diphenyl vinyl sulfonium salt 1 (1.2 equiv), DBU (2 equiv),
CH2Cl2 (0.09m), 2–4 h, 08C. Method B: Chiral vinyl sulfonium salt 6
(1.0 equiv), DBU (2 equiv), CH2Cl2 (0.015m), 2–5 days, À208C.
Scheme 3. The synthesis of five-, six-, and seven-membered heterocy-
clic epoxides through an annulation reaction mediated by 1.
DBU=1,8-diazabicyclo[5.4.0]undec-7-ene, Ts=toluene-p-sulfonyl,
Tf =triflate=trifluoromethanesulfonyl.
c[9,10] gave the desired epoxides 9a–c in good yield and with
very high enantioselectivity (Table 1, entries 1–3). Interest-
ingly, we achieved even higher yields in the annulation
process with the achiral, more robust diphenyl vinyl sulfo-
nium salt 1 (Table 1, method A, entries 1–3). The b-amino-
aldehydes[11] 10a[5] and 10b[12] were also effective and
furnished the corresponding epoxides 3 and 11b, again with
high enantioselectivity (Table 1, entries 4 and 5). The absolute
stereochemistry of epoxide 3 was determined by X-ray
analysis[13] and is consistent with our established model for
asymmetric induction; the remaining compounds are assigned
by analogy (Scheme 5).[1b,14] The b-aminoketone 10c[15] also
provided good enantioselectivity, but yields were low because
of a competing b elimination of the tosylamide[16] (Table 1,
method B, entry 6). This elimination was not observed in the
reaction with the achiral diphenyl vinyl sulfonium salt 1
(Table 1, method A, entry 6), which provided the epoxide 11c
in high yield. Interestingly, this methodology provides a class
of enantiomerically pure epoxides (for example, 3) that are
very difficult to make by the well-established epoxidation
methods, such as Jacobsen–Shi epoxidation of alkenes.[17]
Although other methods exist for the synthesis of some of
the other epoxides, for example, benzofused epoxides 11b,[18]
and trisubstituted epoxides 11c,[19] our method offers a
challenge, because: 1) the aldehyde exists predominantly in
the ring-closed, hemiaminal form, which leads to a low
concentration of the nucleophile, and 2) ring closure to form
the seven-membered ring 5 is much more difficult. Never-
theless, after optimization, even this substrate participated in
the annulation process, thus providing a new route to
functionalized azepines from simple and readily available
starting materials.
Having developed a new synthesis of fused heterocyclic
epoxides, we sought to render the process asymmetric using
the chiral vinyl sulfonium salt 6. This was prepared in two
steps from sulfide 7 (Scheme 4), which we had previously used
in both our catalytic[7] and stoichiometric[8] epoxidations
mediated by sulfur ylides.
Scheme 4. The synthesis of the chiral vinyl sulfonium salt 6.
Angew. Chem. Int. Ed. 2006, 45, 7066 –7069
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7067

Communications
Scheme 5. Model for asymmetric induction in the epoxy-annulation
reaction of b-aminoaldehyde 10a with chiral vinyl sulfonium salt 6.
a) Chiral vinyl sulfonium salt 6 (1.0 equiv), DBU (2 equiv), CH2Cl2
(0.015m), À208C, 3 days, 66% yield, 98% ee.
Scheme 7. Formal synthesis of (À)-balanol through an asymmetric
aziridine-annulation reaction mediated by vinyl sulfonium salt. a) Ti-
(OEt)4, tert-butylsulfinimine, CH2Cl2, 7 days, reflux, 95%; b) 1, NaH
(1.2 equiv), DMF (0.018m), 4 h, 08C, 68%; c) 1. HCl (3 equiv) in
dioxane, 20 min; 2. aq NH3, 42 h, 80%.
completely different disconnection and requires very differ-
ent starting materials, so is a complementary route.
Extension of this asymmetric process to the challenging
seven-membered azepine ring 5, the starting point of our
proposed asymmetric synthesis of balanol, was unsuccessful.
The hemiaminal was recovered, but the vinyl sulfonium salt 6
decomposed. Presumably the slower rate of addition of the
amide nucleophile to the vinyl sulfonium salt allowed
competing side reactions (decomposition pathways) of the
more sensitive dialkyl vinyl sulfonium salt to dominate.
However, even though the epoxide 5 has been used as an
intermediate in the synthesis of balanol,[20a] it is not ideal,
since the epoxide is opened regioselectively at the C4
position[19] (Scheme 6). Thus, the epoxide is first converted
to the corresponding aziridine 12, which undergoes selective
ring opening, again at C4, to furnish the required 3-amino-4-
hydroxyhexahydroazepine 13.
We therefore considered the direct formation of the
required aziridine, in which the identical annulation strategy
would be used. For this we would need an aminal, and in order
to induce asymmetry, we elected to use the Davis–Ellman tert-
butyl sulfinimine.[21] Aminal 14 was prepared by condensation
of the hemiaminal 4 with (R)-tert-butyl sulfinamide, and
reaction with diphenyl vinyl sulfonium salt 1 under the
optimized reaction conditions (NaH in DMF) furnished the
desired hexahydroazepine 15 in 68% yield as a 3:1 mixture of
diastereomers, which were separated by flash chromatogra-
phy (Scheme 7).[22] The stereochemistry of the minor diaste-
reomer was determined by X-ray analysis,[13] thus establishing
the stereochemistry of the major isomer.
Deprotection of aziridine 15 was achieved on treatment
with anhydrous HCl (3 equiv), which resulted in the unavoid-
able formation of the ring-opened product 16. Monitoring of
this reaction using LC-MS showed that the protected
aziridine was very susceptible to ring opening under these
conditions. In fact, it is likely that the protected aziridine 15
first underwent ring opening with HCl and the resulting 1,2-
chlorosulfinamide was then deprotected to yield the inter-
mediate 16. Upon workup with saturated aqueous ammonia
solution, the known chiral aziridine 17[19] was obtained, thus
completing a formal synthesis of balanol.
In conclusion, we have developed a new epoxy-annulation
method for converting a-, b-, and g-aminoaldehydes and
-ketones into racemic five-, six-, and seven-membered
epoxide-fused heterocycles. The reaction has been rendered
asymmetric in the synthesis of five- and six-membered rings,
using the chiral vinyl sulfonium salt 6, and practical levels of
asymmetric induction were achieved. The new process creates
À
À
one new C C bond and two C X bonds with control over
both relative and absolute stereochemistry. The methodology
has been extended to an asymmetric aziridine-annulation
process for the construction of the most challenging seven-
membered hexahydroazepine ring 15. This methodology now
Scheme 6. The regioselective ring opening of the hexahydroazepine core of balanol.
7068 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7066 –7069

Angewandte
Chemie
provides the shortest route (nine steps)[23] to the important
protein kinase C inhibitor, balanol.
proton shift and subsequent ring closure, the tosyl aziridine was
obtained in 45% yield.
Received: July 13, 2006
Published online: September 29, 2006
Keywords: annulation · asymmetric synthesis ·
.
natural products · ring opening · sulfur ylides
[1] a) V. K. Aggarwal, C. L. Winn, Acc. Chem. Res. 2004, 37, 611 –
620; b) V. K. Aggarwal, J. Richardson, Chem. Commun. 2003,
2644 – 2651.
[17] For recent reviews, see: a) E. N. Jacobsen, M. H. Wu in
Comprehensive Asymmetric Catalysis, Vol. 2 (Eds.: E. N. Jacob-
sen, A. Pfaltz, H. Yamamoto), Springer, Berlin, 1999, pp. 649 –
677; b) T. Katsuki in Comprehensive Organic Catalysis, Vol. 2
(Eds.: E. N. Jacobsen, A. Pfaltz, H. Yamamoto), Springer,
Berlin, 1999, pp. 621 – 648.
[18] a) P. G. M. Wuts, R. L. Gu, J. M. Northuis, T. A. Kwan, D. M.
Beck, M. J. White, Pure Appl. Chem. 2002, 74, 1359 – 1368; b) T.
Katsuki, Adv. Synth. Catal. 2002, 344, 131 – 147.
[2] a) K. Kim, L. S. Jimenez, Tetrahedron: Asymmetry 2001, 12,
999 – 1005, and references therein; b) Y. Wang, W. Zhang, J. V.
Colandrea, L. S. Jimenez, Tetrahedron 1999, 55, 10659 – 10672.
[3] a) P. Kulanthaivel, Y. F. Hallock, C. Boros, S. M. Hamilton, W. P.
Janzen, L. M. Ballas, C. R. Loomis, J. B. Jiang, B. Katz, J. R.
Steiner, J. Clardy, J. Am. Chem. Soc. 1993, 115, 6452 – 6453;
b) J. M. Defauw, M. M. Murphy, G. E. Jagdmann, Jr., H. Hu,
J. W. Lampe, S. P. Hollinshead, T. J. Mitchell, H. H. Crane, J. M.
Heerding, J. S. Mendoza, J. E. Davis, J. W. Darges, F. R. Hub-
bard, S. E. Hall, J. Med. Chem. 1996, 39, 5215 – 5227.
[4] a) K. A. Tehrani, T. N. Van, M. Karikomi, M. Rottiers, N.
De Kimpe, Tetrahedron 2002, 58, 7145 – 7152; b) S. Liu, R. P.
Hanzlik, J. Med. Chem. 1992, 35, 1067 – 1075.
[19] Y. Shi, Acc. Chem. Res. 2004, 37, 488 – 496.
[20] a) D. Tanner, A. Almario, T. Hogberg, Tetrahedron 1995, 51,
6061 – 6070; b) D. Tanner, L. Tedenborg, A. Almario, I. Petters-
son, I. Csoregh, N. M. Kelly, P. G. Andersson, T. Hogberg,
Tetrahedron 1997, 53, 4857 – 4868.
[5] H. Ohno, K. Miyamura, T. Tanaka, S. Oishi, A. Toda, Y.
Takemoto, N. Fujii, T. Ibuka, J. Org. Chem. 2002, 67, 1359 – 1367.
[6] a) S. A. Harkin, O. Singh, E. J. Thomas, J. Chem. Soc. Perkin
Trans. 1 1984, 1489 – 1499; b) J. Ahman, P. Somfai, Tetrahedron
1992, 48, 9537 – 9544.
[7] V. K. Aggarwal, E. Alonso, I. Bae, G. Hynd, K. M. Lydon, M. J.
Palmer, M. Patel, M. Porcelloni, J. Richardson, R. Stenson, J. R.
Studley, J. L. Vasse, C. L. Winn, J. Am. Chem. Soc. 2003, 125,
10926 – 10940.
[21] J. A. Ellman, T. D. Owen, T. P. Tang, Acc. Chem. Res. 2002, 35,
984 – 995.
[22] Attempts to improve the diastereoselectivity through the
reaction of aminal 14 with the chiral vinyl sulfonium salt 6 did
not result in the formation of the desired aziridine 15. Enoliza-
tion a to the imine 14 and subsequent conjugate addition of the
enolate to the chiral vinyl sulfonium salt 6 gave an ylide, which
underwent 1,3-proton shift and ring closure to form the a-
cyclopropylimine 18.
[8] a) V. K. Aggarwal, I. Bae, H.-Y. Lee, D. T. Williams, Angew.
Chem. 2003, 115, 3396 – 3400; Angew. Chem. Int. Ed. 2003, 42,
3274 – 3278.
[9] a) T. R. Kasturi, L. Mathew, J. A. Sattigeri, Indian J. Chem. Sect.
B 1990, 29, 1004 – 1006; b) W. R. McKay, G. R. Proctor,J. Chem.
Soc. Perkin Trans. 1 1981, 2435 – 2442.
[10] For reactions of a-aminoketones with vinyl phosphonium salts to
give dihydropyrroles, see: a) I. Burley, A. T. Hewson, Tetrahe-
dron Lett. 1994, 35, 7099 – 7102; b) C. M. Boynton, A. T.
Hewson, M. Douglas, J. Chem. Soc. Perkin Trans. 1 2000,
3599 – 3602.
[11] For reactions of b-aminoaldehydes with vinyl phosphonium salts,
see: S. Kumarn, D. M. Shaw, D. A. Longbottom, S. V. Ley, Org.
Lett. 2005, 7, 4189 – 4191.
[12] P. Y. Chong, S. Z. Janicki, P. A. Petillo, J. Org. Chem. 1998, 63,
8515 – 8521.
[13] CCDC 614698 and CCDC 614700 (3 and 15b) contain the
supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[14] a) V. K. Aggarwal, E. Alonso, G. Hynd, K. M. Lydon, M. J.
Palmer, M. Porcelloni, J. R. Studley, Angew. Chem. 2001, 113,
1479 – 1482; Angew. Chem. Int. Ed. 2001, 40, 1430 – 1433;
b) V. K. Aggarwal, J. Charmant, L. Dudin, M. Porcelloni, J.
Richardson, Proc. Natl. Acad. Sci. USA 2004, 101, 5467 – 5471.
[15] A. B. Reitz, E. Sonveaux, R. P. Rosenkranz, M. S. Verlander,
K. L. Melmon, B. B. Hoffman, Y. Akita, N. Castagnoli, M.
Goodman, J. Med. Chem. 1985, 28, 634 – 642.
[23] For the shortest previously reported route to balanol (ten linear
steps), see: a) A. Fꢀrstner, O. R. Thiel, J. Org. Chem. 2000, 65,
1738 – 1742; Leading references to other total syntheses of
balanol are contained therein.
[16] Enolate formation and subsequent elimination results in the
formation of N-tosylsulfonamide, which then underwent con-
jugate addition to the vinyl sulfonium salt. Following a 1,3-
Angew. Chem. Int. Ed. 2006, 45, 7066 –7069
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7069