Intramolecular azide trapping of the Nazarov intermediate: Formation of peroxy-bridged indolizidinones via a deep-seated rearrangement and aerobic oxidation

Article abstract of DOI:10.1021/ol070053m

Cross-conjugated dienones with pendent azide side chains undergo interrupted Nazarov trapping, leading to peroxy-bridged indolizidinones in good yields. This process is proposed to involve skeletal rearrangement of the initial trapping product, with loss

Full text of DOI:10.1021/ol070053m

ORGANIC
LETTERS
2007
Vol. 9, No. 4
703-706
Intramolecular Azide Trapping of the
Nazarov Intermediate: Formation of
Peroxy-Bridged Indolizidinones via a
Deep-Seated Rearrangement and
Aerobic Oxidation
Ali Rostami, Yong Wang, Atta M. Arif,† Robert McDonald,‡ and F. G. West*
Department of Chemistry, UniVersity of Alberta, Edmonton, AB, Canada T6G 2G2
frederick.west@ualberta.ca
Received January 8, 2007
ABSTRACT
Cross-conjugated dienones with pendent azide side chains undergo interrupted Nazarov trapping, leading to peroxy-bridged indolizidinones
in good yields. This process is proposed to involve skeletal rearrangement of the initial trapping product, with loss of dinitrogen, to give an
intermediate 1,4-betaine, which then undergoes reaction with atmospheric oxygen. The endoperoxide products can be reduced under catalytic
hydrogenation conditions to furnish
r-hydroxylactams.
Domino processes1 involving the 2-oxidocyclopentenyl
intermediate formed during the Nazarov cyclization2 (i.e.,
the “interrupted Nazarov” reaction) furnish a wide range of
complex, polycyclic products.3 With the notable exception
of Dhoro and Tius’ recent report of interception by simple
amines,3c all examples of nucleophilic trapping have em-
ployed either carbon π nucleophiles or silyl hydride. We are
exploring intramolecular trapping by heteronucleophiles as
a rapid entry to polycyclic systems containing heterocyclic
rings. The initial focus has been cycloaddition processes
involving 1,3-dipoles, in analogy to previously reported
intramolecular [4 + 3] trapping by pendent 1,3-dienes.4
Given the expected stability of azides to the Lewis acidic
conditions employed to initiate the Nazarov cyclization, we
chose to test their suitability as traps.
Two possible reaction pathways were envisioned (Scheme
1): either nucleophilic capture of the allyl carbocation by
the internal nitrogen atom (path a) to give zwitterions 1 or
[3 + 3] cycloaddition (path b) to give dihydrotriazines 2.
Nucleophilic trapping of carbocations by pendent azides has
considerable precedent5 and generally involves a subsequent
bond migration with concomitant expulsion of dinitrogen.
Rearrangement of the zwitterionic intermediate 1 might be
expected to occur with migration of the enolate carbon to
form the Lewis acid complexed 1,4-dipole 3, which could
then suffer a 1,5-hydrogen shift to give a dihydropyridone.6
† X-ray Crystallographic Facility, Department of Chemistry, University
of Utah, 315 S. 1400 East, Salt Lake City, Utah 84112-0850.
‡ X-ray Crystallography Lab, Department of Chemistry, University of
Alberta.
(1) (a) Tietze, L. F. Chem. ReV. 1996, 96, 115-136. (b) Tietze, L. F.
Domino Reactions In Organic Synthesis; John Wiley: New York, 2006.
(2) Reviews: (a) Habermas, K. L.; Denmark, S. E.; Jones, T. K. Org.
React. (N.Y.) 1994, 45, 1-158. (b) Pellisier, H. Tetrahedron 2005, 61,
6479-6517. (c) Frontier, A. J.; Collison, C. Tetrahedron 2005, 61, 7577-
7606.
(3) Recent examples: (a) White, T. D.; West, F. G. Tetrahedron Lett.
2005, 46, 5629-5632. (b) Giese, S.; Mazzola, R. D., Jr.; Amann, C. M.;
Arif, A. M.; West, F. G. Angew. Chem., Int. Ed. 2005, 44, 6546-6549. (c)
Dhoro, F.; Tius, M. A. J. Am. Chem. Soc. 2005, 127, 12472-12473. (d)
Janka, M.; He, W.; Haedicke, I. E.; Fronczek, F. R.; Frontier, A. J.;
Eisenberg, R. J. Am. Chem. Soc. 2006, 128, 5312-5313. (e) Grant, T. N.;
West, F. G. J. Am. Chem. Soc. 2006, 128, 9348-9349. (f) Review: Tius,
M. A. Eur. J. Org. Chem. 2005, 2193-2206.
(4) Wang, Y.; Arif, A. M.; West, F. G. J. Am. Chem. Soc. 1999, 121,
876-877.
(5) For recent reviews concerning the reaction of azides with electro-
philes, see: (a) Bra¨se, S.; Gil, C.; Knepper, K.; Zimmermann, V. Angew.
Chem., Int. Ed. 2005, 44, 5188-5240. (b) Lang, S.; Murphy, J. A. Chem.
Soc. ReV. 2006, 35, 146-156.
10.1021/ol070053m CCC: $37.00
© 2007 American Chemical Society
Published on Web 01/26/2007

zonium intermediate 4 in analogy to Aube´’s result. Here we
describe the initial results of this study, including the
unexpected formation of a bridged endoperoxide via oxygen
trapping of the putative 1,4-dipole 3.
Scheme 1. Possible Azide Trapping Pathways
The substrates needed to test this process could be prepared
by a short reaction sequence (Scheme 2). Alkenyl bromide
5 was subjected to lithium-halogen exchange at low
temperature and then was allowed to react with unsaturated
aldehydes 6a-d. Protection of the dienol and removal of
the primary TBS ether gave alcohols 9a-d, which could be
converted to azides 10a-d. Finally, the doubly allylic acetate
was cleaved, and the resulting secondary alcohols were
subjected to oxidation with the Dess-Martin periodinane to
give dienones 11a-d.
Using 11a as a test case, the Lewis acid catalyzed
rearrangement was examined (Scheme 3). In the event, the
Scheme 3. Azide Trapping and Peroxide Formation
In the case of allylic cations, Pearson has observed both [3
+ 2] and [3 + 3] trapping.7 Schultz showed that a
2-oxidopentadienyl cation generated photochemically from
a cross-conjugated dienone underwent intramolecular [3 +
3] trapping,8 and more recently, Aube´ found that a simple
cyclopropanone-derived 2-oxidoallyl cation reacted with
benzyl azide in an apparent [3 + 3] process.9 Although the
[3 + 3] adduct 2 could be isolable,8 the Lewis acidic
conditions might be expected to promote either CN frag-
mentation to give zwitterion 1 or N-N cleavage to dia-
Scheme 2. Preparation of Dienone Substrates with Pendent
Azide Groups
starting dienone was cleanly consumed after treatment with
BF3‚OEt2 at low temperature, followed by warming to 0 °C.
(6) For other examples of formation of transient 1,4-dipoles and their
subsequent 1,5-hydrogen shifts, see: (a) Lenz, G. R. Synthesis 1978, 489-
518. (b) Ninomiya, I.; Naito, T. Heterocycles 1981, 15, 1433-1462. (c)
Potts, K. T.; Rochanapruk, T.; Padwa, A.; Coats, S. J.; Hadjiarapoglou, L.
J. Org. Chem. 1995, 60, 3795-3805. (d) Padwa, A.; Flick, A. C.; Lee, H.
I. Org. Lett. 2005, 7, 2925-2928.
(7) Pearson, W. H.; Fang, W.; Kampf, J. W. J. Org. Chem. 1994, 59,
2682-2684.
(8) Schultz, A. G.; Macielag, M.; Plummer, M. J. Org. Chem. 1988, 53,
391-395.
(9) Aube´, J.; Desai, P. Org. Lett. 2000, 2, 1657-1659.
704
Org. Lett., Vol. 9, No. 4, 2007

Careful analysis revealed three products, including a minor
amount of dihydropyridone 12a. The major components were
diastereomeric endoperoxides 13a and 14a. These com-
pounds were clearly stereoisomeric, and the indicated
structures were suggested by HRMS exact mass measure-
ments and 13C chemical shifts. Eventually, X-ray diffraction
analysis of 13a confirmed this assignment. We presume that
all three products arise from a common intermediate, Lewis
acid complexed betaine 3a, with 12a resulting from a
subsequent 1,5-hydrogen shift or from proton transfer.
Products 13a and 14a appear to arise from reaction with
adventitious oxygen. Although an example of singlet oxygen
trapping of a 1,4-dipole has been reported,10 efficient
scavenging of triplet oxygen appears to be without precedent.
When oxygen is rigorously excluded from the reaction, only
12a is isolated in 70% yield.11 However, we deemed the
additional functionality introduced via the final peroxide-
forming step to be advantageous, so the other examples were
run in the presence of oxygen. Substrates 11b,c furnished
the peroxy products in good yield, but 11d gave only minor
amounts of one of the desired products 13d or 14d, together
with several other inseparable components.12 Direct stereo-
chemical assignments for the products of the last two
examples were not possible; however, an X-ray crystal
structure of a derivative of 13c (vide infra) permitted
assignments as shown.
cation radicals or Lewis acids.17 Although this was initially
explained in terms of facilitation of intersystem crossing by
the catalyst, later work has implicated a chain process
involving diene cation radicals or their oxygen adducts.18
A
variant of this mechanism (Scheme 4) offers a reasonable
rationale for the formation of 13 and 14.
Scheme 4. Cation Radical Chain Mechanism for Peroxide
Formation
The mechanism by which intermediates 3 are efficiently
trapped by ambient oxygen merits some comment. Although
a direct electron-transfer process analogous to enolate
oxygenation is possible, this seems unlikely. Such reactions
are typically run under an atmosphere of oxygen,13 to permit
effective trapping of the enolate; in the present cases, an
unexpectedly long lifetime for the betaine intermediates
would be required to explain the efficient conversion to
peroxides. The lifetime of the putative 1,4-dipole intermediate
3 should be limited due to the availability of a 1,5-H shift
pathway. However, the rigidity of the bi- and tricyclic
betaines in this study may impede this process.14 It is notable
that apparently no oxygenation was seen for the structurally
related pyridinium enolates formed during Romo’s NCAL
reactions.15 Critical distinctions include the endocyclic
structure of dipoles 3 and the presence of BF3. There are
several examples of surprisingly facile oxidation of electron-
rich, cyclic dienes by 3O2.16 Barton has described the
formation of endoperoxides from steroidal cyclohexadienes
and triplet oxygen in the presence of catalytic ammoniumyl
Reduction of the peroxy bridge was also investigated,
using polycyclic endoperoxides 13c and 14c (Scheme 5).
Scheme 5. Peroxide Hydrogenolysis
After examining several conditions, we found that hydro-
genolysis furnished the reduced R-hydroxylactams 15c and
16c, along with minor amounts of unsaturated product 17c.
Product 15c was obtained as a crystalline solid, permitting
unambiguous structural assignment of both that compound
and its precursor, 13c. Although 16c was obtained as one
predominant diastereomer, rigorous assignment of the relative
configuration at the reduced bridgehead carbon was not
(10) Gotthardt, H.; Schenk, K.-H. Tetrahedron Lett. 1983, 24, 4669-
4670.
(11) See Supporting Information for a discussion of the relative config-
uration of the major isomer of 12a.
(12) We have previously observed that dienones incorporating a
cyclopentene ring undergo electrocyclization at a much slower rate,
presumably due to ring strain in the resulting bicyclo[3.3.0]octenyl cation:
Wang, Y.; Schill, B. D.; Arif, A. M.; West, F. G. Org. Lett. 2003, 5, 2747-
2750.
(13) Review: Chen, B.-C.; Zhou, P.; Davis, F. A.; Ciganek, E. Org.
React. (N.Y.) 2003, 62, 1-356.
(16) (a) Najjar, F.; Andre´-Barre`s, C.; Lauricella, R.; Gorrichos, L.; Tuccio,
B. Tetrahedron Lett. 2005, 46, 2117-2119. (b) Eichberg, M. J.; Dorta, R.
L.; Grotjahn, D. B.; Lamottke, K.; Schmidt, M.; Vollhardt, K. P. C. J. Am.
Chem. Soc. 2001, 123, 9324-9337.
(14) In preliminary studies, we have found that monocyclic betaines
analogous to 3 undergo a rapid 1,5-hydrogen shift with no detectable
peroxide formation: Song, D.; Rostami, A.; West, F. G. unpublished results.
(15) Henry-Riyad, H.; Lee, C.; Purohit, V. C.; Romo, D. Org. Lett. 2006,
8, 4363-4366.
(17) Barton, D. H. R.; Haynes, R. K.; Leclerc, G.; Magnus, P. D.;
Menzies, I. D. J. Chem. Soc., Perkin Trans. 1 1975, 2055-2065.
(18) (a) Tang, R.; Yue, H. J.; Wolf, J. F.; Mares, F. J. Am. Chem. Soc.
1978, 100, 5248-5249. (b) Nelsen, S. F.; Teasley, M. F.; Kapp, D. L. J.
Am. Chem. Soc. 1986, 108, 5503-5509.
Org. Lett., Vol. 9, No. 4, 2007
705

possible. However, delivery of hydrogen from the convex
face seems likely in this case.
Acknowledgment. We thank the NIH and NSERC for
support of this work and Dr. Patrick H. Dussault (University
of NebraskasLincoln) for helpful discussions.
This unexpected example of the interrupted Nazarov
reaction involves one of the first cases of trapping the
oxidocyclopentenyl cation intermediate with a nitrogen-based
moiety and entails a significant reorganization of the original
dienone framework. In the presence of air, additional
oxygenation to form peroxides is efficient, and these
intermediates are subject to stereoselective reduction. Further
applications of this chemistry are under investigation and
will be described in due course.
Supporting Information Available: Experimental pro-
cedures and spectral data for cyclization substrates, precur-
sors, and cyclization products, as well as X-ray data for 13a
and 15c (cif). This material is available free of charge via
the Internet at http://pubs.acs.org.
OL070053M
706
Org. Lett., Vol. 9, No. 4, 2007