Generation of aminoacyl radicals from 1-carbamoyl-1-methylcyclohexa-2,5-dienes: A new tin-free homolytic route to β- and γ-lactams

Article abstract of DOI:10.1039/b007454n

Radical induced homolyses of 1-carbamoyl-1-methylcyclo-hexa-2,5-dienes took place cleanly to yield aminoacyl radicals, with no competition from the alternative dissociation to methyl radicals: β- and γ-lactams were obtained from ring closures of suitably

Full text of DOI:10.1039/b007454n

Generation of aminoacyl radicals from 1-carbamoyl-1-methylcyclohexa-2,5-
dienes: a new tin-free homolytic route to b- and g-lactams
Leon V. Jackson and John C. Walton*
University of St. Andrews, School of Chemistry, St. Andrews, Fife, UK KY16 9ST. E-mail: jcw@st-and.ac.uk
Received (in Cambridge, UK) 13th September 2000, Accepted 3rd October 2000
First published as an Advance Article on the web 9th November 2000
Radical induced homolyses of 1-carbamoyl-1-methylcyclo-
hexa-2,5-dienes took place cleanly to yield aminoacyl
radicals, with no competition from the alternative dissocia-
tion to methyl radicals: b- and g-lactams were obtained from
ring closures of suitably unsaturated model compounds.
afforded g-lactam 12 as the main product (53%) together with
smaller amounts of formamide 10 (37%) from direct reduction.
Similar results were obtained simply by heating 7 with
dibenzoyl peroxide (1.5 equiv.) in benzene for 24 h. Acyl
radicals often cyclise in the endo mode6 and, in the case of 9,
this would have produced 1-benzylpiperidin-2-one containing a
6-membered ring. However, spectral evidence was unequivocal
in support of structure 12 and none of the piperidine derivative
was perceptible under our conditions.7 Most significantly, none
of the aromatic amide of type 4 was detectable, even by GC-MS,
and hence the adverse dissociation of delocalised radical 8 to
Me· was negligible. This implied that amides of type 1 had high
potential as clean aminoacyl radical sources, with promise of
considerable generality for syntheses of a variety of lactams.
Carbapenems and nocardicins are important monocyclic
antibiotic classes containing smaller, b-lactam rings that might,
therefore, be accessible starting from appropriate amidocyclo-
hexadienes. Radical cyclisations to afford 4-membered rings via
4-exo-trig ring closures are not generally favoured, but
instances leading to b-lactams have been reported.8–11 Suitably
unsaturated aminoacyl radicals were generated from amides 13
and 17a,b. EPR spectroscopic observations with amide 13
followed the same pattern as with amide 7 i.e. on photolysis
with DTBP the spectrum showed the cyclohexadienyl radical at
lower temperatures and aminoacyl radical 14 at higher tem-
peratures ( > ca. 40 °C). In preparative scale experiments at
60 °C, carbapenem derivative 15 was isolated as the major
product (34%) along with formamide 16 (31%). Analogous
aminoacyl radicals containing propargyl and cyanomethyl
chains were generated from amides 17a and 17b. However, the
formamides 18a,b were the major products isolated. Neither of
these radicals underwent efficient 4-exo-ring closure, pre-
The quest for ‘cleaner’ free-radical precursors, independent of
tin and other toxic metal involvement, and hence suitable for
pharmaceutical syntheses,1 was aided by the discovery that
esters of 1-methylcyclohexa-2,5-diene-1-carboxylic acid and of
2,5-dihydrofuran-2-carboxylic acid selectively furnished alkyl
radicals on induced homolysis.2 These reagents were employed
with moderate success in benign chain alkylations of olefins,
and in cyclisations, affording product yields in the range
35–65%.3 The main limitation to the scope of their deployment
was unwanted competition from an alternative dissociation of
intermediate 1-methyl-1-carboxylatocyclohexadienyl radicals,
that generated methyl radicals and benzoate esters as by-
products. These findings triggered the idea that analogous
amides 1 might function as sources of aminoacyl radicals 3. It
was anticipated that the greater stability of aminoacyl radicals,
in comparison to alkoxyacyls, would favour the desired
dissociation of the delocalised radical 2 to aminoacyl 3, over the
alternative dissociation to Me· and amide 4 (Scheme 1).
Moreover, it was expected that aminoacyls would not de-
carbonylate at moderate temperatures and hence could be
incorporated in free-radical chain cyclisations.
To test this possibility, amide 7 was prepared as illustrated in
Scheme 2. Benzylimine 5 was reduced to N-but-3-enylbenzyl-
amine 6 with sodium borohydride and hence, by reaction with
1-methylcyclohexa-2,5-diene-1-carbonyl chloride,3 to amide 7.
Preliminary observations were carried out using EPR spectros-
copy to monitor radical intermediates generated on photolysis
of a solution of 7 in di-tert-butyl peroxide (DTBP) as initiator
(In). Below about 30 °C the EPR spectrum showed a single
radical with hyperfine splittings (hfs) and g-factor entirely as
expected for cyclohexadienyl radical 8; and similar to para-
meters previously reported for related radicals.3,4 Above this
temperature the spectrum of radical 8 weakened and by about
60 °C was entirely replaced by a new spectrum consisting of a
simple nitrogen triplet (g = 2.0018, a(N) = 2.21 mT, DHpp
=
0.24 mT). These EPR parameters are very similar to those of
archetype aminoacyls e.g. EtNHC·(O) (trans-radical: g =
2.0018, a(N) = 2.24 mT)5 and we attribute the spectrum to
radical 9. Clear-cut spectroscopic evidence for the ring closed
radical 11 was not forthcoming; partly because of sample
boiling and weak spectra at higher temperatures.
Photolysis of a solution of amide 7 in DTBP with unfiltered
light from a 400 W medium pressure Hg lamp at 50 °C for 8 h
Scheme 1
Scheme 2
DOI: 10.1039/b007454n
Chem. Commun., 2000, 2327–2328
This journal is © The Royal Society of Chemistry 2000
2327

View Article Online
Notes and references
1 P. A. Baguley and J. C. Walton, Angew. Chem., Int. Ed., 1998, 37,
3072.
2 G. Binmore, J. C. Walton and L. Cardellini, J. Chem. Soc., Chem.
Commun., 1995, 27.
3 G. Binmore, L. Cardellini and J. C. Walton, J. Chem. Soc., Perkin Trans.
2, 1997, 757.
4 L. Lunazzi, G. Placucci and L. Grossi, J. Chem. Soc., Perkin Trans. 2,
1982, 875; R. H. Schuler, G. P. Laroff and R.W. Fessenden, J. Chem.
Phys., 1973, 77, 456.
5 R. Sutcliffe and K. U. Ingold, J. Am. Chem. Soc., 1981, 103, 7868; H.
Hefter and H. Fischer, Ber. Bunsen-Ges. Phys. Chem., 1970, 74, 493.
6 C. Chatgilialoglu, D. Crich, M. Komatsu and I. Ryu, Chem. Rev., 1999,
99, 1991.
7 For example: carbamoylcyclohexadiene 7 (0.5 g, 1.8 mmol) in DTBP
(1.3 cm3) was photolysed in a quartz tube at 60 °C for 6 h with light from
a 400 W medium pressure Hg lamp. Remaining DTBP was evaporated
and the residue chromatographed to give g-lactam 12 (0.18 g, 53%); dH
1.25 (3H, d, J = 7 Hz), 1.55–1.65 (2H, m), 2.52 (1H, sextet, J = 7 Hz),
3.20 (2H, m), 4.48 (2H, AB), 7.2–7.4 (5H, m); dC 16.4 (CH3), 27.1
(CH2), 36.8 (CH), 44.6 (CH2), 46.8 (CH2), 127.5 (CH), 128.1 (CH),
128.6 (CH), 136.7 (C), 177.4 (CNO). Found: M+, 189.1158. C12H15NO
requires M, 189.1154. Formamide derivative 10 (0.12 g, 37%) was also
isolated.
8 G. B. Gill, G. Pattenden and S. J. Reynolds, Tetrahedron Lett., 1989, 30,
3229; G. B. Gill, G. Pattenden and S. J. Reynolds, J. Chem. Soc., Perkin
Trans. 1, 1994, 369.
9 H. Ishibashi, C. Kameoka, H. Iriyama, K. Kodama, T. Sato and M.
Ikeda, J. Org. Chem., 1995, 60, 1276; H. Ishibashi, C. Kameoka, K.
Kodama, T. Sato and M. Ikeda, Tetrahedron, 1996, 52, 489.
10 A. J. Clark and J. L. Peacock, Tetrahedron Lett., 1998, 39, 1265.
11 L. Boiteau, J. Boivin, B. Quiclet-Sire and S. Z. Zard, Tetrahedron, 1998,
54, 2087.
Scheme 3
sumably because of the extra strain involved in forming
4-membered rings containing exocyclic double bonds.
In summary, therefore, radical induced homolyses of cyclo-
hexadienyl amides of type 2 took place cleanly to yield
aminoacyl radicals, with no competition from the alternative
dissociation to methyl radicals. Model b- and g-lactams were
obtained from ring closures of alkenylaminoacyl radicals.
Consequently, these amides, which are easily prepared from
unsaturated amines and 1-methylcyclohexa-2,5-diene-1-car-
boxylic acid, furnish mild, tin-free routes to small, and probably
medium ring lactams, eminently suitable for conversion to
useful biologically active compounds.
We thank the EPSRC (grant GR/L49185) for financial
support.
2328
Chem. Commun., 2000, 2327–2328