Conversion of N-Acyl-4-acyloxy-β-lactams into 1,3-oxazin-6-ones: Two consecutive pseudopericyclic processes

Article abstract of DOI:10.1021/ol0056168

(equation presented) N-Acyl-4-acyloxy-β-lactams are converted into 1,3-oxazin-6-ones under basic conditions. This transformation is believed to proceed via N-acylazetones, which rearrange to the final products by a sequence of two electrocyclic processes. The calculated (RHF and B3LYP) transrtion structures of both concerted reactions are shown to present characteristic pseudopericyclic orbital topologies.

Full text of DOI:10.1021/ol0056168

ORGANIC
LETTERS
2000
Vol. 2, No. 7
965-968
Conversion of
N-Acyl-4-acyloxy-â-lactams into
1,3-Oxazin-6-ones: Two Consecutive
Pseudopericyclic Processes
Mateo Alajar´ın,* Angel Vidal, Pilar Sa´nchez-Andrada, Fulgencio Tovar, and
Gine´s Ochoa
Departamento de Qu´ımica Orga´nica, Facultad de Qu´ımica, UniVersidad de Murcia,
Campus de Espinardo, 30100 Murcia, Spain
alajarin@fcu.um.es
Received February 3, 2000
ABSTRACT
N-Acyl-4-acyloxy-â-lactams are converted into 1,3-oxazin-6-ones under basic conditions. This transformation is believed to proceed via
N-acylazetones, which rearrange to the final products by a sequence of two electrocyclic processes. The calculated (RHF and B3LYP) transition
structures of both concerted reactions are shown to present characteristic pseudopericyclic orbital topologies.
In the course of our recent work on the intramolecular aza-
Wittig reaction of the â-lactam carbonyl group,1 we required
Scheme 1
a number of N-(2-azidobenzoyl)-â-lactams which were
prepared by acylation of N-unsubstituted â-lactams with
2-azidobenzoyl chloride in the presence of Et3N. In one of
such reactions, that in which the substrate was 4-acetoxy-
â-lactam (1), we noted the appearance of trace amounts of
an unidentified byproduct which seemed to derive from the
action of the base (Et3N) on the acylated lactam 2, as its
relative proportion in the final reaction mixture increased
by using more equivalents of base and extended reaction
times. Herein we disclose the structure of that byproduct,
2-(2-azidophenyl)-1,3-oxazin-6-one (3) (Scheme 1), describe
a new general method for the preparation of 1,3-oxazin-6-
ones by the action of organic bases on N-acyl-4-acyloxy-â-
lactams, and discuss a reasonable mechanistic explanation
of such transformations.
The structural elucidation of the 1,3-oxazin-6-one 3 was
straitghtforward from its analytical and spectral data.2 A
careful exploration of the reaction conditions (organic base,
solvent, temperature) for the conversion of 2 to 3 revealed
DBU (2 equiv) to be the base of choice and CH2Cl2, 25 °C,
and 2 h as the optimal conditions for obtaining the best yield
of 3 (78%). Moreover, we found that 1,3-oxazin-6-one 3
could be also efficiently obtained (41% yield) in one
experimental step when 1 was reacted with 2-azidobenzoyl
chloride (2 equiv) by using an excess of DBU (5 equiv) under
similar reaction conditions.
The validity of this one-step method for the preparation
of 1,3-oxazin-6-ones was checked for a number of combina-
tions of two different 4-acyloxy-â-lactams 4 with several
(1) Alajar´ın, M.; Molina, P.; Vidal, A.; Tovar, F. Synlett 1998, 1288.
(2) See Supporting Information.
10.1021/ol0056168 CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/15/2000

acyl chlorides (aroyl, heteroaroyl, alkenoyl and alkanoyl) and
resulted in compounds of general applicability (Scheme 2
and Table 1).
cyclic ring opening to the N-acylimidoylketene 8, which in
turn was transformed into the final 1,3-oxazin-6-one 5
through a six-centered electrocyclic ring closure (Scheme
3).
Scheme 2
Scheme 3
1,3-Oxazin-6-ones are versatile intermediates in hetero-
cyclic chemistry, and their synthesis and reactivity have been
reviewed.3 Compounds 5a, 5b, 5e, and 5f have been
previously synthesized by other methodologies.4 Their
reported analytical and spectral data were fully coincident
with those of the titled compounds here prepared. For most
of the entries in Table 1, the intermediate N-acyl-â-lactams
Table 1. One-Step Preparation of 2-Substituted
1,3-Oxazin-6-ones 5 from 4-Acyloxy-â-lactams 4
entry
R1
R2
product
(%) yielda
1
2
3
4
5
6
7
CH3
CH3
C6H5
4-Cl-C6H4
2-furyl
(E)-C6H5-CHdCH
(CH3)3C
C2H5
5a
5b
5c
5d
5e
5f
69
58
44
50
76
40
53
This mechanism is supported by the following: First, the
most general methods for the synthesis of 1,3-oxazin-6-ones
have been proposed to involve, as immediate precursors,
ring-opened valence tautomers such as acyliminoketene type
8,3 although they have never been isolated or detected by
spectroscopic means. The two most recent reports on this
theme are coincident in that proposal.4a,5
C6H5
C6H5
C6H5
C6H5
C6H5
4-Br-C6H4
5g
a Isolated yield.
Second, concerning the putative intermediate N-acylaze-
tone 7, it is worth pointing out that azetones have occasion-
ally been proposed in the literature as highly reactive
intermediates.6 Resonance of their carbonyl group leads to
formally antiaromatic structures. They have only been well-
characterized in their benzo-fused forms,7 whereas Wentrup
and co-workers claimed the isolation of an N-adamantylaze-
tone in an Ar matrix at 77 K together with its ring-opened
valence tautomer imidoylketene.8
could be isolated when the reactions were run using Et3N
instead of DBU and then cleanly converted into the corre-
sponding 5 by the action of this last base.
A reasonable mechanistic explanation for the conversion
of N-acyl-4-acyloxy-â-lactams 6 into 2-substituted 1,3-
oxazin-6-ones 5 by the action of DBU is the following: first,
the organic base promotes the â-elimination of 1 equiv of
carboxylic acid R1CO2H across the C3-C4 bond of the
â-lactam ring, giving rise to the highly strained N-acylazetone
7, which was not stable enough to survive in the reaction
conditions and rapidly experienced a four-centered electro-
(3) Steglich, W.; Jeschke, R.; Buschmann, E. Gazz. Chim. Ital. 1986,
116, 361.
(4) (a) McNab, H.; Withell, K. Tetrahedron 1996, 52, 3163. (b) Stajer,
G.; Szabo, A. E.; Fu¨lop, F.; Bernath, G. Synthesis 1984, 345.
(5) Millan, D. S.; Prager, R. H.J. Chem. Soc., Perkin Trans. 1 1998,
3245.
(6) Kato, H.; Wakao, K.; Yamada, A.; Mutoh, Y. J. Chem. Soc., Perkin
Trans. 1 1988, 189. Kappe, T.; Zadeh, R. K. Synthesis 1975, 247. Micetich,
R. G.; Maiti, S. N.; Tanaka, M.; Yamazaki, T.; Ogawa, K. J. Org. Chem.
1986, 51, 853.
(7) Olofson, R. A.; Van der Meer, R. K.; Hoskin, D. H.; Bernheim, M.
Y.; Stournas, S.; Morrison, D. S.J. Org. Chem. 1984, 49, 3367. Wentrup,
C.; Gross, G. Angew. Chem., Int. Ed. Engl. 1983, 22, 543.
(8) Kappe, C. O.; Kollenz, G.; Netsch, K.-P.; Leung-Toung, R.; Wentrup,
C. J. Chem. Soc., Chem. Commun. 1992, 488.
Ab initio calculations (HF/6-31G*) carried out by Nguyen
showed that ring opening of the parent azetone to the
unsubstituted imidoylketene is exothermic by 12 kcal/mol.9
The previous literature report more closely related to the
present work is due to Perica´s et al.10 These authors attempted
(9) Nguyen, M. T.; Ha, T.; O’Ferral, R. A. M. J. Org. Chem. 1990, 55,
3251.
(10) Perica´s, M. A.; Serratosa, F.; Valent´ı, E.; Font-Altaba, M.; Solans,
X. J. Chem. Soc., Perkin Trans. 2 1986, 961. In the context of the present
work, it is remarkable that the title of this article starts with the sentence
“Can N-acylazetones ever be obtained?”
966
Org. Lett., Vol. 2, No. 7, 2000

to prepare N-acylazetones 9 by [2 + 2] cycloaddition of
benzoyl isocyanate with di-tert-butoxyethyne, obtaining the
1,3-oxazin-6-one 11 as the main reaction product. They
rationalized the formation of 11 through the sequence
depicted below, involving the transient formation of azetone
9 and N-acylimidoylketene 10, a mechanism similar to the
one here proposed for the conversion of 7 into 5.
Scheme 4
carbonyl group more “ketone-like”. The absence of this effect
explains why more common N-alkyl- or -aryl-substituted
4-acyloxy-â-lactams, e.g., 13, do not undergo such elimina-
tion.
The conversions of 3-monosubstituted â-lactams cis-14
and trans-15 into 1,3-oxazin-6-one 16 were easily achieved,
in comparable yields (Scheme 4). The fact that the 3,4-
elimination is amenable irrespective of the relative cis or
trans relationship between the groups being eliminated, and
apparently at similar rate, is indicative, but not yet conclusive,
of an E1cB-like mechanism.
Following semiempirical MNDO calculations, the authors
of that work concluded that “... the conversion of N-
acylazetones into 1,3-oxazin-6-ones is predicted to take place
easily and completely, even at low temperatures, owing to
the important difference in thermodynamic stability between
the compounds.”
Finally, a brief comment on the â-elimination reaction of
N-acyl-4-acyloxy-â-lactams 6 presumably leading to N-
acylazetones 7. Although 4-acyloxyazetidinones are widely
employed synthetic materials,11 to our knowledge there are
no previous reports on elimination of carboxylic acids from
N-substituted substrates, whereas N-unsubstituted derivatives
are believed to undergo displacement of the 4-acyloxy group
by different O-, N-, and S-centered nucleophiles through the
intermediacy of the 1-azetin-4-one 12.11,12
At the present we are carrying out ab initio calculations,
at different levels of theory, on the full reaction path leading
from N-formylazetone to the simplest 1,3-oxazin-6-one (a
simplified model of the conversion 7 to 5), which will be
fully reported in due course. We here show the geometries
of the so far calculated transition states of the two exothermic
electrocyclic processes (ring opening and ring closure), as
they revealed orbital topologies different from the ones
expected for classical pericyclic reactions (Figure 1).
Whereas TS1q was located at both levels of theory studied,
q
RHF and B3LYP, TS2 could only be located at the RHF
level.13 The energy barriers for both reactions were notably
small (see Figure 1), particularly that of the ring closure
reaction. This value could explain why all attempts to locate
q
TS2 at the B3LYP level led directly to the ring-closed
oxazinone. Notably, both transition states TS1q and TS2q show
essentially planar geometries, proving to be nonrotatory, that
is, not to be of a classic, 4π conrotatory or 6π disrotatory
We carried out some additional experiences (Scheme 4)
in relation to this first step of the mechanism in Scheme 3.
N-Aryl-4-formyloxy-â-lactam 13 remained unchanged in the
presence of DBU (CH2Cl2, 25 °C, or toluene, 110 °C). We
believe that the suitability of the N-acylated derivatives 2
and 6 to undergo â-elimination along its C3-C4 bond is
due to the enhancing of the acidity of H-3 by the electron-
withdrawing N-acyl group, which makes the â-lactam
(13) The ring closure of a similar system, 5-oxo-2,4-pentadienal, was
calculated to have no barrier at the MP2/6-31G* level: Birney, D. M. J.
Org. Chem. 1996, 61, 243.
(14) Ross, J. A.; Seiders, R. P.; Lemal, D. M. J. Am. Chem. Soc. 1976,
98, 4325.
(15) Birney, D. M.; Xu, X.; Ham, S. Angew. Chem., Int. Ed. 1999, 38,
189. Birney, D. M.; Xu, X.; Ham, S.; Huang, X. J. Org. Chem. 1997, 62,
7114. Birney, D. M.; Ham, S.; Unruh, G. R. J. Am. Chem. Soc. 1997, 119,
4509. Ham, S.; Birney, D. M. J. Org. Chem. 1996, 61, 3962. Wagenseller,
P. E.; Birney, D. M.; Roy, D. J. Org. Chem. 1995, 60, 2853. Birney, D.
M.; Wagenseller, P. E. J. Am. Chem. Soc. 1994, 116, 6262.
(16) Fabian, W. M. F.; Kappe, C. O.; Bakulev, V. A. J. Org. Chem.
2000, 65, 47. Fabian, W. M. F.; Bakulev, V. A.; Kappe, C. O. J. Org.
Chem. 1998, 63, 5801. Liu, R. C.-Y.; Lusztyk, J.; McAllister, M. A.;
Tidwell, T. T.; Wagner, B. D. J. Am. Chem. Soc. 1998, 120, 6247. Luo, L.;
Bartberger, M. D.; Dolbier, W. R. J. Am. Chem. Soc. 1997, 119, 12366.
(11) Wild, H. In The Organic Chemistry of â-Lactam Antibiotics; Georg,
G. I., Ed.; VCH: New York, 1993; Chapter 2.
(12) Attrill, R. P.; Barret, A. G. M.; Quayle, P.; van der Westhuizen, J.;
Betts, M. J. J. Org. Chem. 1984, 49, 1679 and references therein.
Org. Lett., Vol. 2, No. 7, 2000
967

q
N-C σ-bond being broken in TS1 is giving rise to an sp2-
orbital on the nitrogen (not a p-orbital of the NdC π-bond)
and to a p-orbital of the carbonyl π-bond, rather than to that
q
of the cumulated CdC π-bond. Similarly, in TS2 the σ-bond
being formed is due to the interaction between an sp2-orbital
of the formyl oxygen atom and a p-orbital of the cumulated
carbonyl π-bond, both in the molecular plane, instead of
involving the two parallel p-orbitals of the reactant, on the
extremes of the cyclic array of six atoms. In both cases,
nonbonded pairs become bonded pairs as the new σ-bond is
broken or formed, thus giving rise to “disconnections” in
the cyclic array of overlapping orbitals because the atomic
orbitals which are switching functions are mutually orthogo-
nal.
q
All these characteristics of the transition structures TS1
q
and TS2 , corresponding to the two reactions shown in Figure
1, are consistent with Lemal’s definition of a pseudoperi-
cyclic process,14 a subset of pericyclic reactions that remained
almost ignored for about 20 years but which has been placed
on a solid foundation by the recent publications of Birney13,15
and others.16
Acknowledgment. We thank the Direccio´n General de
Ensen˜anza Superior (Project PB95-1019), the Fundacio´n
Se´neca-CARM (Project PB/2/FS/99), and Acedesa (a divi-
sion of Takasago) for financial support. F.T. and P.S.-A. also
thank the MEC and Cajamurcia for their respective fellow-
ships.
Figure 1. (a) Calculated (B3LYP/6-311++G**) transition state
TS1q for the concerted ring opening. (b) Calculated (RHF/6-31G*)
transition state TS2q for the concerted ring closure. The ∆E values
contain a ZPVE correction, computed at the same level of theory.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
electrocyclic nature, respectively. This means that the orbitals
interactions must be qualitatively different from those
commonly encountered in pericyclic reactions. Thus, the
OL0056168
968
Org. Lett., Vol. 2, No. 7, 2000