Highly diastereoselective allylic azide formation and isomerization. Synthesis of 3(2′-amino)-β-lactams

Article abstract of DOI:10.1021/ol047815n

(Chemical Equation Presented) The stereoselective anti S _N2′ attack of NaN₃ to 3-alkenyl-3-bromo-azetidin-2- ones gave a mixture of diastereomeric azides in fast equilibrium. The [3,3]-sigmatropic rearrangement of allylic azides occu

Full text of DOI:10.1021/ol047815n

ORGANIC
LETTERS
2005
Vol. 7, No. 4
533-536
Highly Diastereoselective Allylic Azide
Formation and Isomerization. Synthesis
of 3(2′-Amino)-â-lactams
Giuliana Cardillo,* Serena Fabbroni, Luca Gentilucci, Rossana Perciaccante,
Fabio Piccinelli, and Alessandra Tolomelli
Department of Chemistry “G. Ciamician”, UniVersity of Bologna, Via Selmi 2,
40126 Bologna, Italy
giuliana.cardillo@unibo.it
Received October 25, 2004
ABSTRACT
The stereoselective anti SN2
The [3,3]-sigmatropic rearrangement of allylic azides occurred with complete stereocontrol, allowing the equilibrium to be directed preferentially
toward the (E)- or (Z)-isomer, useful precursors of 3(2 -amino)- -lactams.
′ attack of NaN3 to 3-alkenyl-3-bromo-azetidin-2-ones gave a mixture of diastereomeric azides in fast equilibrium.
′
â
One of the most useful reactions in organic chemistry is the
nucleophilic substitution of allylic halides. Many important
aspects of their behavior such as the stereo- and regiochem-
istry of the substitutions, have received considerable atten-
tion.1
is to be predicted for neutral nucleophiles, while anionic
nucleophiles approach from the anti direction; however, the
experimental evidence “is often contradictory”.1a This paper
concerns the nucleophilic attack of NaN3 to a particular
allylic moiety, the 3-alkenyl-3-bromo-azetidin-2-ones 1.3
This reaction shows interesting mechanistic aspects4 and
offers the opportunity to introduce in few steps and under
high regio- and stereocontrol the amino function in the side
The reaction of a nucleophile upon an allylic halide can
occur via an SN2 process with attack at CR or through an
SN2′ process with attack of the nucleophile at Cγ and
departure of the leaving group. While the stereochemistry
of the SN2 is defined, the SN2′ attack can occur syn or anti
with respect to the leaving group, depending on the nature
of the nucleophile.1a-c The theory2 suggests that a syn process
(2) (a) Fukui, K. Acc. Chem. Res. 1971, 4, 57-64. (b) Yates, R. L.;
Epiotis, N. D.; Bernardi, F. J. Am. Chem. Soc. 1975, 97, 6615-6621. (c)
Caramella, P.; Rondan, N. G.; Paddon-Row, M. N.; Houk, K. N. J. Am.
Chem. Soc. 1981, 103, 2438-2440. (d) Stohrer, W.-D. Angew. Chem. 1983,
95, 642-643. (e) Bach, R. D.; Wolber, G. J. J. Am. Chem. Soc. 1985, 103,
1352-1357.
(3) Benfatti, F.; Cardillo, G.; Fabbroni, S.; Gentilucci, L.; Perciaccante,
R. Piccinelli, F.; Tolomelli, A. Synthesis 2005, 61-70.
(4) (a) Panek, J. S.; Yang, M.; Muler, I. J. Org. Chem. 1992, 57, 4063-
4064. (b) Padwa, A.; Sa, M. M. Tetrahedron Lett. 1997, 38, 5087-5090.
(c) Padwa, A.; Sa`, M. M. J. Braz. Chem. Soc. 1999, 10, 231-236. (d)
Banert, K.; Hagedorn, M.; Liedtke, C.; Melzer, A.; Schoffler, C. Eur. J.
Org. Chem. 2000, 257-267.
(1) For a review, see: (a) Magid, R. M. Tetrahedron 1980, 36, 1901-
1930. (b) Toromanoff, E. Tetrahedron 1980, 36, 2809-2931. (c) Paquette,
L. A.; Stirling, C. J. M. Tetrahedron 1992, 48, 7383-7423. For recent
reviews on metal-catalyzed allylic substitution reactions, see: (d) Pfaltz,
A.; Lautens, M., In ComprehensiVe Asymmetric Catalysis II; Jacobsen, E.
N., Pfaltz, A., Yamamoto, H., Eds.; Springer-Verlag Telos: Berlin
Heidelberg, 1999; Chapter 24, pp 833-884. (e) Trost, B. M.; Crawley, M.
L. Chem. ReV. 2003, 103, 2921-2943.
10.1021/ol047815n CCC: $30.25
© 2005 American Chemical Society
Published on Web 01/21/2005

chain of the â-lactam5 (Figure 1). The propensity of allylic
azides to undergo [3,3]-sigmatropic rearrangement,6 thus
Table 1. Nucleophilic Attack of NaN3 to
3-Bromo-3-alkenyl-azetidin-2-ones (()-1a-d in DMF at 70 °C
yielda of
(E)-2 + (Z)-3
dr of Eb
dr of Zb
1
R
(%)
(%)
(%)
1a
1b
1c
1d
CH2Ph
C2H4CO2Et
CH2CHdCH2
(S)-phenylethyl-
69
78
70
68
92:8
>95:5
85:15
90:10
85:15
>95:5
88:12
90:10
a After purification by flash chromatography over silica gel. b Product
distribution was determined by 1H NMR integration at 300 MHz on the
crude mixture and confirmed by isolation of pure compounds.
azide 411 (Scheme 1). All the products were isolated by flash
chromatography on silica gel. The Z and E configuration of
Figure 1. Synthetic pathway to 3(2′-amino)-â-lactams.
Scheme 1. Nucleophilic Attack of NaN3 to
3-Bromo-3-alkenyl-azetidin-2-ones (()-1a-d
giving a mixture of isomers, is overcome in our case by the
complete stereocontrol at the equilibrium, which makes the
reaction of synthetic interest.7
In 1971, Bose and Manhas8 described the direct synthesis
of R-vinyl-â-lactams by the reaction of crotonyl chloride and
TEA with a Schiff base. We have applied the same reaction
conditions for the preparation of 3-alkenyl-3-bromo-azetidin-
2-ones,3 starting from R-bromo-â,γ-unsaturated ketenes9 and
a variety of Schiff bases. The reaction occurs smoothly in
moderate to good yields (50-60%), affording preferentially
the cis diastereomers, which were purified by flash chro-
matography or preparative HPLC and utilized as starting
materials for further transformations. The prospects of the
employment of R-alkenyl-R-bromo-azetidin-2-ones 1 as
precursors of new molecules, prompted us to verify the
feasibility of the substitution reaction. The presence of the
allylic bromide permitted easy SN2′ reaction. In fact, the
treatment of cis-(()-1a-c and (1′S,3R,4S)-1d with NaN3 in
DMF at 70 °C afforded 1:1 mixtures of (E)-2 and (Z)-3 azido
derivatives in good yield10 (Table 1) and 10-15% yield of
the double bond was attributed by comparison of the
chemical shift of vinyl proton H1. In accordance with
literature data,12 the vinyl proton resonating at 5.25 ppm was
attributed to the (Z)-3 isomer, while the one at 6.0 ppm was
attributed to the isomer (E)-2. Compounds 2 and 3 were
obtained, as oils, in high diastereomeric ratio, each (E)-2 or
(Z)-3 being accompanied by small amounts of the diastereo-
isomer with the C-N3 stereocenter in the opposite config-
uration (Table 1, columns 4 and 5). In accordance with the
theory on the SN2′ process for anionic nucleophiles, we
suggest for optically active (E)-2d major isomer the (2′R)
configuration and for the (Z)-3d major isomer the (2′S)
configuration. To confirm the correct attribution of the
stereochemistry, compound (()-1e was prepared starting
from benzyl-(4-nitro-benzylidene)-amine and R-bromo-hex-
enoyl chloride and obtained in 93% yield (Scheme 2). The
(5) For an example of azetidinone substituted at the side chain with an
amino group such as tabtoxin and its analogues, see: (a) Lee, D. L.;
Rapoport, H. J. Org. Chem. 1975, 40, 3491-3496. (b) Baldwin, J. E.;
Otsuka, M.; Wallace, P. M. Tetrahedron 1986, 42, 3097-3110. (c) Greenlee,
W. J.; Springer, J. P.; Patchett, A. A. J. Med. Chem. 1989, 32, 165-170.
(6) (a) Gagneux, A.; Winstein, S.; Young, W. G. J. Am. Chem. Soc.
1960, 82, 5956-5957. (b) Murahashi, S. I.; Taniguchi, Y.; Imada, Y.;
Tanigawa, Y. J. Org. Chem. 1989, 54, 3292-3303. (c) Safi, M.; Fahrang,
R.; Sinou, D. Tetrahedron Lett. 1990, 31, 527-530. (d) Chida, N.; Tobe,
T.; Murai, K.; Yamazaki, K.; Ogawa, S. Heterocycles 1994, 38, 2383-
2388.
(7) Trost, B. M:; Pulley, S. R. Tetrahedron Lett. 1995, 36, 8737-8740.
(8) (a) Bose, A. K.; Spiegelmann, G.; Manhas, M. S. Tetrahedron Lett.
1971, 3167-3170. (b) Zamboni, R.; Just, G. Can. J. Chem. 1979, 57, 1945-
1948. (c) Manhas, M. S.; Ghosh, M.; Bose, A. K. J. Org. Chem. 1990, 55,
575-580. (d) Manhas, M. S.; Banik, B. K.; Mathur, A.; Vincent, J. E.;
Bose, A. K. Tetrahedron 2000, 56, 5587-5601.
(10) Reaction of 1a carried out at rt afforded, in about 70 h, (E)-2a as
the major isomer (56%), (Z)-3a (29%), and 4a (15%).
(11) Trans configuration of compound 4 was determined on the basis of
the 1H NMR signal of H1′ resonating at 5.0 ppm; see Supporting
Information.
(12) (a) Anklam, S.; Liebscher, J. Tetrahedron 1998, 54, 6369-6384.
(b) Otto, H. H.; Bergmann, H. J.; Mayrhofer, R. Arch. Pharm. (Weinheim,
Ger.) 1986, 319, 203-216.
(9) (a) Cardillo, G.; De Simone, A.; Mingardi, A.; Tomasini, C. Synlett
1995, 11, 1131-1132. (b) Cardillo, G.; Fabbroni, S.; Gentilucci, L.;
Perciaccante, R.; Piccinelli, F.; Tolomelli, A. Tetrahedron 2004, 60, 5031-
5040. (c) Cardillo, G.; Fabbroni, S.; Gentilucci, L.; Perciaccante, R.;
Tolomelli, A. Tetrahedron: Asymmetry 2004, 15, 593-601.
534
Org. Lett., Vol. 7, No. 4, 2005

Scheme 2. Synthesis of 3-Bromo-3-alkenyl-azetidin-2-ones 1e
and Nucleophilic Attack of NaN3
Figure 2. X-ray-determined structure for the minor (2′S*-Z,4R*)-
3e diastereoisomer.
(Figure 3). The comparison of the spectra recorded at the
initial time and after 24 h showed that the ratio between the
products changed considerably, the amount of azide 4e (33%)
being decreased in favor of isomer (Z)-3e (63%) and (E)-2e
(4%). In a similar way, pure (E)-2e in CDCl3 converted in
72 h into a mixture of azide 4e (15%), (Z)-3e (30%), and
E-2e (55%).
presence of the nitro group on the aromatic ring generally
favors the crystallization of products. The SN2′ reaction,
performed on azetidin-2-one (()-1e, gave the products 2e/
3e/4e in 52/33/15 ratio and 88% total yield. The (Z)-isomer
was obtained as an 82:18 mixture of diastereoisomers with
opposite configuration at the C2′ stereocenter. Only the minor
(Z)-isomer could be crystallized from methanol/water solu-
tion, and the X-ray analysis13 showed the (2′S*-Z,4R*)
relative configuration (Figure 2), thus allowing the (2′R*-
Z,4R*) relative configuration to be attributed to the major
isomer (Z)-3e. On the basis of these results, the anti direction
of the nucleophile with respect to the bromide leaving group
was confirmed.2
1
Comparison of the H NMR chemical shift for (E)-2a-e
and (Z)-3a-e series revealed a complete regularity and
allowed us to determine the relative configuration. Further-
more, since compounds 2 and 3 are in equilibrium, we could
1
establish, on the basis of the H NMR analysis in CDCl3,
that this equilibrium involves the azide 4. The slow equili-
bration between the three isomers was observed even at low
temperature, changing the solvent (EtOAc) or in solvent-
free conditions.
In fact, a mixture of 4e (90%) and (Z)-3e (10%) was
dissolved in CDCl3 and analyzed by 1H NMR spectroscopy
Figure 3. Comparison of 1H NMR spectra of a mixture of 4e and
(Z)-3e at different times.
(13) Crystal data for the minor (Z)-isomer: C20H19N5O3 MW ) 377.40,
monoclinic, P21/n, a ) 15.8947(12) Å, b ) 7.1605(6) Å, c ) 17.1209(13)
Å, â ) 100.848(2)°, V ) 1913.8(3) Å3, Z ) 4, Dcalcd ) 1.310 Mg/m3, R1
) 0.0676, wR2 ) 0.1297 (all data), GOF on F2 ) 1.037.
The complete selectivity of the rearrangement was con-
firmed by a sequence of reactions carried out both on racemic
2a and on optically pure 2d. To this aim, pure (E)-2a was
Org. Lett., Vol. 7, No. 4, 2005
535

hydrogenated with Pd/C to the corresponding amino deriva-
tive, which was converted into the Cbz-derivative 5a (overall
yield ) 68%) (Scheme 3). The 3,4-cis stereochemistry of
isomer and separation by flash chromatography, was con-
verted exclusively into (2′R)-5d (yield 50%, [R]D22 ) +5.0,
Chiralcel OD:12 rt, 37.5 min), thus confirming the stereo-
chemical outcome of the equilibration.
The complete selectivity of the rearrangement was ratio-
nalized by the mechanism proposed in Figure 4, where a
Scheme 3. Conversion of (E)-2 into CBz-Derivative 5
5a was assigned on the basis of the coupling constants H3-
H4 (J ) 5.3 Hz). In a similar way, the hydrogenation and
22
protection of pure (2′R-E)-2d afforded 5d ([R]D ) +5.2,
Chiralcel OD:14 rt, 37.5 min) in 50% overall yield. On the
other hand, a solution of pure (Z)-3a in CHCl3 reached the
2:1 (Z)-3a/(E)-2a ratio after 24 h.
From this mixture (E)-2a was separated by flash chroma-
tography and reduced under the above-reported conditions.
The hydrogenation and the conversion to Cbz-derivative
afforded 5a as a single diastereoisomer, as confirmed by 1H
NMR and 13C NMR spectra (Scheme 4).
Figure 4. Mechanism proposed for the [3,3]-sigmatropic rear-
rangement.
Scheme 4. Conversion of (Z)-3 into Cbz-Derivative 6,
Equilibration to (E)-2, and Conversion to Cbz-Derivative 5
concerted [3,3]-sigmatropic rearrangement occurs via a cyclic
transition state with complete stereocontrol, in accordance
with the mechanisms suggested for the interconversion of
allylic azides.4
The (2′R-E)-2, generated by SN2′ attack on the re face of
the double bond in 1, is in equilibrium with trans-azide 4,
which converts to (2′S-Z)-3 via [3,3]-sigmatropic shift on
the si face of the double bond, with complete inversion of
the C2′ configuration. In a similar way, the sodium azide
SN2′ attack on the si face of the double bond in 1, affords
(2′S-Z)-3, which is in equilibrium with azide 4.
In conclusion, we showed that the SN2′ azide attack occurs
anti to the leaving bromide, in accordance with the theoretic
suggestions. The equilibrium of the â-lactam double-bond
occurs under complete stereocontrol via concerted [3,3]-
sigmatropic rearrangement of the azide. Finally, the equi-
librium (E)-2/(Z)-3 or (Z)-3/(E)-2 allows the use of this
reaction for synthetic purposes, giving, after hydrogenation,
stereodefined 3(2′-amino) derivatives.
Moreover, the hydrogenation of the pure (Z)-3a, followed
by protection with Cbz, afforded exclusively 6a (overall yield
50%), thus showing that the configuration of the newly
stereogenic center is opposite in 2a and 3a. The same reaction
sequence, performed on the optically active azides (2′S-Z)-
3d afforded exclusively (2′S)-6d (yield 50%, [R]D22 ) +52,
Chiralcel OD:12 rt, 23.1 min). In a similar way, compound
(2′R-E)-2d, obtained from equilibration of the (2′S-Z)-3d
Acknowledgment. We thank MIUR (PRIN 2002 and
FIRB 2001), CNR-ISOF, and University of Bologna (funds
for selected topics) for financial support.
Supporting Information Available: Experimental pro-
cedures, full characterization of new compounds, and crystal-
lographic data for the minor (Z)-isomer (CIF). This material
is available free of charge via the Internet at http://pubs.acs.org.
OL047815N
(14) Optical rotation values in CHCl3 (c 1.0); Chiralcel OD (Daicel
column), cellulose tris(3,5-dimethyl-phenyl)carbamate phase coated on 10
µm silica gel, n-hexane/2-propanol 90:10 solvent mixture, flow 0.5 mL/min.
536
Org. Lett., Vol. 7, No. 4, 2005