Catalyzed reactions of α-amino diazoketones with allyl sulfides: A new synthetic protocol for α-amino homoallyl ketones en route to 3-piperidinol alkaloids

Article abstract of DOI:10.1016/S0040-4039(00)00282-3

Cu(acac)₂-catalyzed reactions of enantiopure α-amino diazoketones with allyl sulfides provide a facile synthetic route to α-amino homoallyl ketones via 2,3-sigmatropic rearrangements of the derived allyl sulfonium ylides. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00282-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 2965–2969
Catalyzed reactions of α-amino diazoketones with allyl sulfides:
a new synthetic protocol for α-amino homoallyl ketones en route
to 3-piperidinol alkaloids
Saumitra Sengupta ∗ and Somnath Mondal
Department of Chemistry, Jadavpur University, Calcutta 700 032, India
Received 30 July 1999; revised 8 February 2000; accepted 14 February 2000
Abstract
Cu(acac)2-catalyzed reactions of enantiopure α-amino diazoketones with allyl sulfides provide a facile synthetic
route to α-amino homoallyl ketones via 2,3-sigmatropic rearrangements of the derived allyl sulfonium ylides.
© 2000 Elsevier Science Ltd. All rights reserved.
Keywords: α-amino diazoketones; allyl sulfonium ylides; 2,3-sigmatropic rearrangements; α-amino ketones.
The Cassia and Prosopis species produce a number of 2,6-disubstituted-3-piperidinol alkaloids 1 (e.g.
azimic acid, spectaline, julifloridine, prosopinine, etc.)1 which have attracted much current synthetic
attention due to their broad spectrum of biological activities.2 In a unified synthetic approach towards
these alkaloids, we identified the enantiopure α-amino homoallyl ketones 2 as the key retrosynthetic
precursors and were in search of an efficient synthetic route towards the latter compounds (Scheme 1).
Enantiopure α-amino homoallyl ketones can be prepared via nucleophilic α-amino acylation reactions
of homoallyl lithium with α-amino Weinreb amides3 or, in the Rapoport-protocol, with NHTos α-amino
acids.4 However, due to the high costs of preparing Weinreb-amides and the need of a large excess of
homoallyl lithium in the Rapoport-protocol, these procedures were deemed unsuitable for our present
purpose. We, therefore, sought an organometallic-free approach for the preparation of enantiopure α-
amino homoallyl ketones and in this letter, we report a conceptually new synthesis of these compounds
Scheme 1.
∗
Corresponding author.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00282-3
tetl 16562

2966
Scheme 2. Reagents and conditions: (i) 10% Cu(acac)2, benzene, 80°C, 5 min; (ii) Zn, NH4Cl, ether, rt, 30 h
Table 1
Synthesis of α-amino homoallyl ketones (Scheme 2)

2967
via facile 2,3-sigmatropic rearrangements of allyl sulfonium ylides derived from catalyzed reactions of
enantiopure α-amino diazoketones 3 with allyl sulfides.
Catalyzed reactions of α-diazoketones with allyl sulfides leading to allyl sulfonium ylides and their
2,3-sigmatropic rearrangements are well documented in the literature.5 However, apart from a few
reactions of diazopenicillanates with allyl chalcogenides,6 there are no reports on the use of enantiopure
α-diazoketones in such sequences. This appeared to be a significant synthetic gap, especially since
catalyzed reactions of enantiopure α-diazoketones7,8 with allyl chalcogenides, via 2,3-sigmatropic
rearrangements of the derived ylides, promised a rapid and convergent assembly of functionalized
enantiopure building blocks under mild and neutral conditions. Such prospects provided additional
impetus for the present investigation.
Catalyzed reactions of allyl phenyl sulfide with enantiopure α-amino diazoketones 3a–c7 were
examined under a variety of conditions. The best results were obtained with Cu(acac)2 as the catalyst
(better than Rh2(OAc)4) in benzene at 80o using short contact times (≤5 min) which, reproducably, led to
the rapid formation of the α-amino-α0-thio homoallyl ketones 5a–c via 2,3-sigmatropic rearrangements
of the allyl sulfonium ylides 4 (Scheme 2). Reactions carried out in CH2Cl2 or THF (at reflux) gave
incomplete conversions, whereas prolonged heating of the reaction mixture or slow addition of 3 to a
heated mixture of the allyl sulfide and catalyst did not improve the yields, and in fact, led to increasing
amounts of the side products. The yields of these reactions were found to be moderate (Table 1), as
is often observed for intermolecular reactions of α-diazoketones with allyl sulfides, and this may be
attributed to dimerization problems and/or secondary reactions of the product thioketones with the
diazocarbonyl compounds.9 Nevertheless, the convergent nature of this synthesis, coupled with the
ready availability of starting materials, still made it an attractive synthetic sequence. The α-amino-α0-
thio homoallyl ketones 5a–c were all produced as variable mixtures of diastereomers which, without
separation, were subjected to desulfurization with Zn/NH4Cl10 in ether at rt to produce the desired α-
amino homoallyl ketones 6a–c in good to excellent yields (Scheme 2, Table 1).11 Raney-Ni could not
be used in this desulfurization step since it causes simultaneous saturation of the double bonds in the
products. That the enantiomeric purities of the starting α-diazoketones 3 were preserved during this two-
step sequence was shown by the following experiment: the α-amino homoallyl ketone 6b, upon catalytic
25
hydrogenation, produced the corresponding α-amino butyl ketone whose optical rotation value ([α]D
25
+35.2, c 2, CHCl3) was found to be in good agreement with the α-amino butyl ketone ([α]D +33.2,
c 2.6, CHCl3) independently synthesized via reaction of the N-CO2Et leucinyl Weinreb-amide with n-
BuLi (THF, −78°C). The Cu(acac)2-catalyzed reaction of the allyl sulfide with the L-threonine-derived
oxazolidinyl α-diazoketone 7 (entry 4, Table 1) similarly produced 8 (30%), which was subsequently
desulfurized with Zn/NH4Cl to give the oxazolidinyl homoallyl ketone 9 in 66% yield. N-Phthaloyl α-
amino diazoketones, e.g. 10, also reacted with allyl phenyl sulfide in the presence of 10% Cu(acac)2
to produce 11 (50%) (entry 5, Table 1), but desulfurization of the latter proved problematic due to
concomitant reduction of its phthaloyl moiety.
The N-CO2Et-L-proline-derived α-diazoketone 13, via the above two-step sequence, also produced
the homoallyl ketone 15 (Scheme 3) which promises to be a useful synthon towards synthesis of various
indolizidine alkaloids.
Scheme 3. Reagents and conditions: (i) (COCl)2, CH2Cl2, then excess CH2N2; (ii) 10% Cu(acac)2, benzene, 80°C, 5 min; (iii)
Zn, NH4Cl, ether, rt

2968
Scheme 4. Reagents and conditions: (i) Ref. 8; (ii) 10% Cu(acac)2, benzene, 80°C, 5 min; (iii) Zn, NH4Cl, ether, rt
The enantiopure α-acetoxy diazoketones 17a,b,8 readily prepared from the α-hydroxy acid chiral
pool, also participated in this two-step sequence to produce the α-acetoxy homoallyl ketones 19a,b in
somewhat better yields (Scheme 4). The latter promises to have important uses in the synthesis of deoxy-
C-glycosides.
In summary, we have shown that Cu-catalyzed reactions of enantiopure α-diazoketones with allyl
sulfides provide a facile synthetic route to enantiopure α-amino and α-hydroxy homoallyl ketones having
broad synthetic ramifications. Allyl selenides also reacted with these diazoketones in a similar fashion,12
but catalyzed reactions of 3 with benzyl sulfides/selenides have so far produced only complex product
mixtures. Syntheses of the 2,6-disubstituted-3-piperidinol alkaloids using the α-amino homoallyl ketones
prepared in this work are currently under investigation.
Acknowledgements
DST (SP/S1/G-14/97) and UGC (senior research fellowship to S.M.) are warmly thanked for financial
support.
References
1. Strunz, G. M.; Findlay, J. A. In The Alkaloids; Brossi, A., Ed.; Academic Press: New York, 1985; Vol. 26, Ch. 3.
2. For some recent syntheses, see: (a) Ojima, I.; Vidal, E. S. J. Org. Chem. 1998, 63, 7999; (b) Kiguchi, T.; Shirakawa, M.;
Honda, R.; Ninomiya, I.; Naito, T. Tetrahedron 1998, 54, 15 589; (c) Agami, C.; Couty, F.; Lam, H.; Mathieu, H. Tetrahedron
1998, 54, 8783; (d) Momose, T.; Toyooka, N.; Jin, M. J. Chem. Soc., Perkin Trans. 1 1997, 2005; (e) Oetting, J.; Holzkamp,
J.; Meyer, H. H.; Pahl, A. Tetrahedron: Asymmetry 1997, 8, 477; (f) Kadota, I.; Kawada, M.; Muramatsu, Y.; Yamamoto, Y.
Tetrahedron Lett. 1997, 38, 7469; (g) Luker, T.; Hiemstra, H.; Speckamp, W. N. J. Org. Chem. 1997, 62, 3592; (h) Hirai, Y.;
Watanabe, J.; Nozaki, T.; Yokoyama, H.; Yamaguchi, S. J. Org. Chem. 1997, 62, 776.
3. Sibi, M. P. Org. Prep. Proc. Int. 1993, 25, 15.
4. O’Neill, B. T. In Comprehensive Organic Synthesis; Trost, B. M.; Fleming, I., Eds.; Pergamon Press: Oxford, 1991; Vol. 1,
p. 397.
5. Padwa, A.; Weingarten, M. D. Chem. Rev. 1996, 96, 223.
6. (a) Giddings, P. J.; John, D. I.; Thomas, E. J.; Williams, D. J. J. Chem. Soc., Perkin Trans. 1 1982, 2757; (b) Giddings, P. J.;
John, D. I.; Thomas, E. J. Tetrahedron Lett. 1980, 21, 399; (c) Giddings, P. J.; John, D. I.; Thomas, E. J. Tetrahedron Lett.
1980, 21, 395.
7. (a) Doyle, M. P.; McKervey, M. A. J. Chem. Soc., Chem. Commun. 1997, 983; (b) Ye, T.; McKervey, M. A. Chem. Rev.
1994, 94, 1091.
8. Sengupta, S.; Sen Sarma, D.; Mondal, S. Synth. Commun. 1998, 28, 4409.
9. (a) Carter, D. S.; Van Vranken, D. L. Tetrahedron Lett. 1999, 40, 1617; (b) Aggarwal, V. K.; Ferrara, M.; Hainz, R.; Spey,
S. E. Tetrahedron Lett. 1999, 40, 8923, and references cited therein.
10. Holton, R. A.; Crouse, D. J.; Williams, A. D.; Kennedy, R. M. J. Org. Chem. 1987, 52, 2317.
11. All new compounds gave satisfactory spectral data. A typical procedure: (S)-6-Ethoxycarbonylamino-5-oxo-hept-1-ene
(6a): A solution of 3a (0.11g, 0.6 mmol), allyl phenyl sulfide (0.18 g, 1.2 mmol) and Cu(acac)2 (10 mol%) in benzene
(1.2 ml) was immersed in an oil bath preheated to 80°C. After 5 min, the benzene was removed at reduced pressure and
CH2Cl2 (5 ml) was added. It was then washed with water, dried and evaporated to give 5a (diastereomeric mixture) which
was purified by preparative TLC over silica gel (30% EtOAc in pet. ether). Activated Zn (0.40 g) and satd NH4Cl soln (2.5

2969
ml) were added to a solution of 5a (0.062 g, 0.2 mmol) in THF (3 ml) and the mixture stirred at rt for 30 h, then filtered
and the filtrate extracted with CH2Cl2 (3×5 ml). Removal of the solvent, followed by preparative TLC over silica gel (15%
EtOAc in pet. ether), gave 6a as a colorless oil (0.031g, 77%); [α]D28 +26.0 (c 0.9, CHCl3); IR: 3420, 2980, 1700 (br), 1495
cm−1; 1H NMR (300 MHz): δ 1.24 (t, 3H, J=7 Hz), 1.35 (d, 3H, J=7 Hz), 2.35 (q, 2H, J=7.2 Hz), 2.50–2.70 (m, 2H), 4.11
(q, 2H, J=7 Hz), 4.36 (m, 1H), 4.98–5.07 (m, 2H), 5.42 (br s, 1H), 5.72–5.86 (m, 1H); found: C, 60.51; H, 8.50; N, 7.33.
C10H17NO3 requires: C, 60.30; H, 8.54 and N, 7.03%.
12. Sengupta, S.; Das, D., unpublished results.