An unusual synthesis of N-unsubstituted benzazepinones

Article abstract of DOI:10.1021/ol3026044

A short route to novel bicyclic N-unprotected benzazepinones is described starting from N-acetoxyanilides involving radical addition and cyclization with concomitant homolytic rupture of the N-O bond.

Full text of DOI:10.1021/ol3026044

ORGANIC
LETTERS
2012
Vol. 14, No. 21
5514–5517
An Unusual Synthesis of N‑Unsubstituted
Benzazepinones
ꢀ
Beatrice Quiclet-Sire, Ngoc Diem My Tran, and Samir Z. Zard*
ꢁ
Laboratoire de Synthese Organique, CNRS UMR 7652 Ecole Polytechnique,
91128 Palaiseau Cedex, France
zard@poly.polytechnique.fr
Received September 21, 2012
ABSTRACT
A short route to novel bicyclic N-unprotected benzazepinones is described starting from N-acetoxyanilides involving radical addition and
cyclization with concomitant homolytic rupture of the NꢀO bond.
We describe here an unexpected route to N-unsubsti-
tuted benzazepinone we discovered while examining a
radical-based approach to cyclic hydroxamic acids. The
strong chelating ability of hydroxamic acids toward
metal ions such as zinc has made them popular targets
for medicinal chemists. Numerous members of this family
have indeed been reported to be potent inhibitors of his-
tone deacetylase and matrix metalloproteinases.1 Other
interesting biological properties include antibiotic, anti-
fungal, anti-inflammatory and anticancer activities.1,2
While most hydroxamic acids described are linear deriva-
tives, such as the antibiotic fosmidomycin 1, a few cases
where the hydroxamate motif is part of a ring appear to
be particularly interesting (Figure 1). Examples include
cobactin T, a siderophore growth promoter isolated
from mycobacteria,3 and PF-04859989, 3, an irreversible
kynurenine aminotransferase II inhibitor developed by
Pfizer for treating schizophrenia.4
Figure 1. Examples of biologically active hydroxamic acids.
As part of our ongoing work on xanthates,5 we found
that lactams fused to aromatic or heteroaromatic rings
could be readily constructed by direct radical cyclization
(1) (a) Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J. Chem.
Rev. 1999, 99, 2735. (b) Rao, B. G. Curr. Pharm. Des. 2005, 11, 295.
(c) Flipo, M.; Charton, J.; Hocine, A.; Dassonneville, S.; Deprez, B.;
Deprez-Poulain, R. J. Med. Chem. 2009, 52, 6790.
(2) (a) Montgomery, J. I.; Brown, M. F.; Reilly, U.; Price, L. M.;
Abramite, J. A.; Arcari, J.; Barham, R.; Che, Y.; Chen, J. M.; Chung,
S. W.; Collantes, E. M.; Desbonnet, C.; Doroski, M.; Doty, J.; Engtrakul,
J.; Harris, T. M.; Huband, M.; Knafels, J. D.; Leach, K. L.; Liu, S.;
Marfat, A.; McAllister, L.; McElroy, E.; Menard, C. A.; Mitton-Fry, M.;
Mullins, L.; Noe, N. C.; O’Donnell, J.; Oliver, R.; Penzien, J.; Plummer,
M.; Shanmugasundaram, V.; Thoma, C.; Tomaras, A. P.; Uccello, D. P.;
Vaz, A.; Wishka, D. J. J. Med. Chem. 2012, 55, 1662. (b) Tardibono, L. P.,
Jr.; Miller, M. J. Org. Lett. 2009, 11, 1575.
(4) Dounay, A. B.; Anderson, M.; Bechle, B. M.; Campbell, B. M.;
Claffey, M. M.; Evdokimov, A.; Evrard, E.; Fonseca, K. R.; Gan, X.;
Ghosh, S.; Hayward, M. M.; Horner, W.; Kim, J. ꢀY.; McAllister,
L. A.; Pandit, J.; Paradis, V.; Parikh, V. D.; Reese, M. R.; Rong, S. B.;
Salafia, M. A.; Schuyten, K.; Strick, C. A.; Tuttle, J. B.; Valentine, J.;
Wang, H.; Zawadzke, L. E.; Verhoest, P. R. ACS Med. Chem. Lett.
2012, 3, 187.
(5) For reviews on the xanthate radical transfer reaction, see: (a)
Zard, S. Z. Angew. Chem., Int. Ed. 1997, 36, 672. (b) Quiclet-Sire, B.;
Zard, S. Z. Chem.;Eur. J. 2006, 12, 6002. (c) Quiclet-Sire, B.; Zard, S. Z.
Top. Curr. Chem. 2006, 264, 201. (d) Zard, S. Z. Aust. J. Chem. 2006, 59,
663. (e) Quiclet-Sire, B.; Zard, S. Z. Pure Appl. Chem. 2011, 83, 519.
(3) Yang, S. ꢀM.; Lag, B.; Wilson, L. J. J. Org. Chem. 2007, 72, 8123.
r
10.1021/ol3026044
Published on Web 10/18/2012
2012 American Chemical Society

onto the aromatic nucleus.6 It seemed therefore from the
outset a simple matter to extend this approach to the
synthesis of aryl-fused cyclic hydroxamic acids 6 as shown
in Scheme 1. An appealing aspect of this route is the ready
availability of the requisite precursors 5 from arylhydrox-
ylamines 4, which in turn can be easily prepared by partial
reduction of nitroarenes.7
is apparently significantly slower as compared with the
corresponding anilide radicals 11, which ring-close effi-
ciently under similar conditions.6,9
Scheme 3
Scheme 1
The formation of five-membered rings by radical cycli-
zation is usually the most efficient, so we first attempted
accessing N-hydroxyoxindoles (6, n = 0). The required
xanthate precursor 7a was prepared from p-methyl-phe-
nylhydroxylamine 4a by chloroacetylation of the nitrogen,
displacement of the chlorine with potassium O-ethyl
xanthate and acetylation (Scheme 2).
Scheme 2
Unfortunately, whenarefluxingsolution of thismaterial
in ethyl acetate was treated with lauroyl peroxide, added
portion-wise in stoichiometric amounts, none of the de-
sired N-acetoxy-oxindole 8a was obtained; only the trivial
reduced product 9awasisolatedin 95% yield.8 For reasons
that are not yet clear, but that may have to do with bond
angles, the cyclization of radicals 10 onto the aromatic ring
In contrast, we found that the desired cyclization took
place to a certain extent in the homologous series (5, 6;
n = 1) but not in the manner we had anticipated. Two
major products were formed when compound 14a, itself
prepared by radical addition of cyanomethyl xanthate 13a
to butenyl hydroxamate 12a, was subjected to the action of
lauroyl peroxide in refluxing ethyl acetate (Scheme 3). The
first product, isolated in 40% yield, proved to be open
chain hydroxamate 17a, arising through a radical Smiles
rearrangement.10 This ultimately leads to N-acetoxyami-
dyl radical 16, which then abstracts a hydrogen atom
from the solvent. The second product turned out to be
(6) For a review, see: El Qacemi, M.; Petit, L.; Quiclet-Sire, B.; Zard,
S. Z. Org. Biomol. Chem. 2012, 10, 5707.
(7) Oxley, P. W.; Adger, B. M.; Sasse, M. J.; Forth, M. A. Org. Synth.
1988, 67, 187.
(8) In Schemes 3ꢀ5, placing lauroyl peroxide in parentheses indicates
that it is used as the initiator in substoichiometric amounts. No
parentheses means that the peroxide is used both as initiator and as a
stoichiometric oxidant.
(9) (a) Axon, J.; Boiteau, L.; Boivin, J.; Forbes, J. E.; Zard, S. Z.
ꢀ
(10) (a) Studer, A. In Radicals in Organic Synthesis; Renaud, P., Sibi,
M. P., Eds; Wiley-VCH: Weinheim, 2001; Vol. 2, 44. (b) Studer, A.;
Bossart, M. Tetrahedon 2001, 57, 9649. (c) E. Bacque, E.; El Qacemi,
Tetrahedron Lett. 1994, 35, 1719. (b) Bacque, E.; El Qacemi, M.; Zard,
ꢀ
ꢀ
S. Z. Org. Lett. 2004, 6, 3671. (c) Bacque, E.; El Qacemi, M.; Zard, S. Z.
Heterocycles 2012, 84, 291.
Org. Lett., Vol. 14, No. 21, 2012
M.; Zard, S. Z. Org. Lett. 2005, 7, 3817.
5515

dihydroquinolone 20, obtained in 35% yield, where the
acetoxy group of the hydroxamate moiety had been lost.
As depicted in Scheme 3, the cyclization of radical 15 gives
cyclohexadienyl radical 18, which then undergoes β-scission
of the weak NꢀO bond.11 The acetyloxy radical thus pro-
duced evolves through rapid extrusion of carbon dioxide
into a methyl radical, which can transfer a xanthate
group from the starting material 14a to give radical 15 and
S-methyl xanthate 21a.
Scheme 4
Scheme 5
Evidence for this mechanism was adduced by repeating
theoperation with steroidhydroxamate14b. Asoutlinedin
Figure 2. Examples of benzazepinones.
(11) Acyloxy radicals have been generated by treatment of N-acylox-
yphthalimides with stannyl radicals. See: Barton, D. H. R.; Blundell, P.;
Jaszberenyi, J. Cs. Tetrahedron Lett. 1989, 30, 2341.
Scheme 4, the reaction furnished in the yields shown the
expected open chain hydroxamate 17b and dihydroquinolone
5516
Org. Lett., Vol. 14, No. 21, 2012

20, in addition to steroid xanthate 21b. Decarboxylation of
the bile acid moiety had indeed taken place in accord with the
mechanism displayed in Scheme 3. The fact that cyclization
did proceed in the case of xanthates 14a and 14b, albeit in
modest yield, encouraged us to examine the behavior of the
even higher homologue 5, n = 2. The cyclization in this case
would lead to a benzazepinone 6, n = 2. We had found in the
past that it was possible in some cases to construct seven
membered rings by direct radical cyclization onto aromatic or
heteroaromatic rings. Furthermore, the observation that
radicals 10 (Scheme 2) did not readily cyclize opened the
possibility of their intermolecular capture by an alkene 22,
thus providing a direct and flexible access to xanthate pre-
cursors 23 needed for the formation of benzazepinones 24, as
indicated in Scheme 5.9b,c,12
The intermolecular addition to unhindered terminal
alkenes indeed proved successful, as did the cyclization
step. The examples assembled in Figure 2 give an idea of
the scope and functional group tolerance of the process.
No complications arising from a radical Smiles rearrange-
ment were observed, in contrast to the lower analogous
series. In this case, the Smiles rearrangement would have
had to proceed through a temporary six-membered ring,
which kinetically is a much slower process.
Thus, alkenes 22aꢀi bearing a range of useful func-
tional groups underwent the desired radical addition with
xanthates 7aꢀd to give new adduct xanthates 23aꢀi, and
these in turn were converted into benzazepinones 24aꢀi in
synthetically useful yields. Particularly noteworthy is the
compatibility of the method with the presence of an
aromatic bromide and, especially, iodide, as this allows
further elaboration through the myriad transition metal
catalyzed coupling reactions. The ring-closure in the pre-
sence of a meta-iodo substituent is only moderatly regio-
selective (2:1 in favor of a distal cyclization). In the case of
24f/24f0, the two regioisomers could not be separated by
chromatography, but a pure sample of the major isomer
could be obtained by crystallization. Another interesting
aspect is the ease of introducing a boronate (24i) oracomplex
carbohydrate motif (two separable epimers 24j and 24j0).
Direct access to N-unsubstituted benzazepinones can-
not be accomplished by cyclization of secondary amide
xanthates 25 (Scheme 5), presumably because of the re-
latively high rotation barrier and the predominance of the
rotamer with a geometry unfavorable for ring-closure.12,13
The best approach we have recently developed relies on
methanesulfonamides of type 26 (Scheme 5), which give N-
unsubstituted benzazepinones by extrusion of a methane-
sulfonyl radical, but high temperatures must be used in the
cyclization step.14
(12) (a) Kaoudi, T.; Quiclet-Sire, B.; Seguin, S.; Zard, S. Z. Angew.
Chem., Int. Ed. 2000, 39, 731. (b) Petit, L.; Botez, I.; Tizot, A.; Zard, S. Z.
Tetrahedron Lett. 2012, 53, 3220.
(13) Yamasaki, R.; Tanatani, A.; Azumaya, I; Saito, S.; Yamaguchi,
K.; Kagechika, H. Org. Lett. 2003, 5, 1265 and references cited therein.
(14) Charrier, N.; Liu, Z.; Zard, S. Z. Org. Lett. 2012, 14, 2018.
(15) See for example: A. Zhang, A.; Neumeyer, J. L.; Baldessarini,
R. J. Chem. Rev. 2007, 107, 274.
(16) S.Z.Z. would like to put on record his debt to Prof. Ireland for
being indirectly responsible for the discovery of a mild source of
hydroxyl radicals: In the spring of 1989, I happened to be a visiting
scholar at Texas A&M University when Bob Ireland was invited to give a
seminar. During our brief meeting, he said he was interested in using the
Barton decarboxylation reaction in one of his syntheses and asked if I
could provide him with a reliable procedure for obtaining N-hydro-
xypyridine-2-thione from the commercial aqueous solution of its sodium
salt. In those pre-internet days, I had to wait until I returned to France a
couple of weeks later to retrieve the procedure from the thesis where it
was detailed. I dutifully copied the procedure and decided to enclose a
couple of grams of N-hydroxypyridine-2-thione, so that Bob Ireland’s
student would not have to wait to test the decarboxylation on his
substrate. However, as I was about to place the sample in a small
The present convergent approach allies a flexibility in
the choice of reacting partners with the simplicity and
mildness of the experimental procedure. It constitutes a
concise and cheap route to a class of highly privileged
structures in medicinal chemistry.15 Furthermore, the sub-
stitution pattern accessible by xanthate technology is not
easily attained by more conventional synthetic methods.
Acknowledgment. We dedicate this paper with respect
to the memory of Professor Robert E. Ireland (University
of Virginia).16 N.D.M.T. thanks Ecole Polytechnique for a
scholarship.
ꢀ
glassine envelope, Prof. Marcel Fetizon came by and we chatted at the
door of my office. It was a grim and overcast day but, as we were
chatting, the sun just came out for a few minutes and the sunlight struck
the vial containing the N-hydroxypyridine-2-thione. When I eventually
returned to my desk, I was dismayed to see that the nice off-white solid
had turned into a brown goo on the side of the vial pointing towards the
window. The initial irritation at the photochemical destruction of our
main supply of the material soon gave way to curiosity, and we
ultimately found that mere irradiation with a tungsten filament lamp
of N-hydroxypyridine-2-thione resulted in the clean generation of
Supporting Information Available. Experimental pro-
cedures, full spectroscopic data and copies of 1H and 13
C
NMR spectra for all new compounds. This material is
available free of charge via the Internet at http://pubs.acs.
org.
ꢀ
hydroxyl radicals ( Boivin, J.; Crepon, E.; Zard, S. Z. Tetrahedron Lett.
1990, 31, 6869). This convenient source of hydroxyl radicals was later
used by various research groups to cleave DNA strands.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 21, 2012
5517