An unusual pathway to cyclobutane formation via desulfurative intramolecular photocycloaddition of an enone benzothiazoline pair

Article abstract of DOI:10.1021/ol900186y

Irradiation of the enone benzothiazoline 3 leads to the formation of cyclobutane 5. Preliminary mechanistic studies establish the intermediacy of an enecarbamate 14 in this photochemical transformation, which could be the result of sulfur extrusion from an episulfide intermediate. Photocycloaddition of the enecarbamate intermediate 14 leads to the formation of "crossed" photoadducts, i.e., 5, in excellent yield, with high levels of regio- and stereochemical control.

Full text of DOI:10.1021/ol900186y

ORGANIC
LETTERS
2009
Vol. 11, No. 8
1685-1687
An Unusual Pathway to Cyclobutane
Formation via Desulfurative
Intramolecular Photocycloaddition of an
Enone Benzothiazoline Pair
Hyunil Jo, Mark E. Fitzgerald,† and Jeffrey D. Winkler*
Department of Chemistry, UniVersity of PennsylVania,
Philadelphia, PennsylVania 19104
winkler@sas.upenn.edu
Received January 28, 2009
ABSTRACT
Irradiation of the enone benzothiazoline 3 leads to the formation of cyclobutane 5. Preliminary mechanistic studies establish the intermediacy
of an enecarbamate 14 in this photochemical transformation, which could be the result of sulfur extrusion from an episulfide intermediate.
Photocycloaddition of the enecarbamate intermediate 14 leads to the formation of “crossed” photoadducts, i.e., 5, in excellent yield, with high
levels of regio- and stereochemical control.
The selective synthesis of quaternary carbon centers remains
an important challenge in organic synthesis.1 We have re-
Scheme 1. Desulfurative Photocycloaddition
cently reported that intramolecular photocycloaddition of 1
leads to the efficient formation of 2, which contains an aryl-
substituted quaternary center.2 We describe herein the
synthesis and photoreaction of 3, the predominant product
of which is not the analogous thiolane product 4, but instead
cyclobutane 5, in which the sulfur is extruded in the course
of the photoreaction (Scheme 1).
The synthesis of racemic photosubstrate 3 is outlined in
Scheme 2. Addition of o-aminothiophenol to the known
ethynyl cyclohexenone 63 afforded 7, which on carbamate
formation and reaction with methoxide4 led to the formation
† Current Address: Broad Institute of MIT and Harvard, Cambridge, MA
02142.
(1) For a recent review, see: Trost, B. M.; Jiang, C. Synthesis 2006, 36,
9–396.
(2) Winkler, J. D.; Lee, E. C.-Y. J. Am. Chem. Soc. 2006, 128, 9040–
9041.
of the photosubstrate 3 (R ) COOMe). We found that
irradiation of 3 through a uranium filter in benzene gave the
(3) Larsen, D. S.; O’Shea, M. D. J. Chem. Soc., Perkin Trans. 1. 1995,
1019–1028.
10.1021/ol900186y CCC: $40.75
Published on Web 03/20/2009
2009 American Chemical Society

of enecarbamate 14 (R ) COOMe).6 Alternatively, Grob-
like fragmentation of 10/12 could lead to the direct formation
of 14 with loss of elemental sulfur.
Scheme 2. Formation of Cyclobutane Photoadduct 5
Triplet-mediated cycloaddition of 14 could then lead to
the formation of the crossed7 photoproduct 5 via 15. Careful
1H NMR analysis of the reaction mixture in the course of
the irradiation of 3 revealed the presence of 14. Isolation of
14 and its conversion to 5 in quantitative yield (uranium filter,
benzene, 0 °C, 1 h) established its competence as an
intermediate in the formation of 5 from 3. Failure to observe
the formation of 5 from 14 under nonphotochemical condi-
tion (benzene, 25 °C, 72 h; or benzene reflux, 1 h) suggested
that the formation of 5 was exclusively due to excited-state
reaction of 14.
While the intermediacy of enecarbamate 14 could be
directly observed on irradiation of 3, the presence of the
episulfide 13, which we propose as a possible precursor to
14, could not. In an effort to obtain evidence for the
intermediacy of the episulfide, we re-examined the photo-
reaction of 16, the sulfide corresponding to 1 (Scheme 4).
expected product 4 in only 7% yield, accompanied by the
formation of 5 as the major product (77% yield).5
Separate irradiation of 4 (uranium filter, benzene, 2 h) does
not lead to the formation of 5, establishing that the expected
thiolane product is not an intermediate in the formation of
the observed cyclobutane. A mechanistic proposal that is
consistent with the formation of both 4 and 5 from 3 is
outlined in Scheme 3.
Scheme 4. Episulfide Formation in the Photoreaction
Scheme 3. Mechanistic Proposal for the Formation of 4 and 5
Irradiation of 16 led to the formation of nearly equimolar
quantities of thiolane 18 and 3-phenyl-cyclohexenone 19.
The formation of the latter product is consistent with the
intermediacy of episulfide 20 in the reaction, although the
volatility of 20 would presumably preclude its isolation.
(4) Heindel, N. D.; Ko, C. C. H. J. Heterocycl. Chem. 1970, 7, 1007–
1011.
(5) The structure and stereochemistry of 4 were confirmed by X-ray
crystallographic analysis of the corresponding sulfoxide 26, which on
chemical reduction (PBr3, CH2Cl2, 25 °C, 100% yield) gave 4.The structure
and stereochemistry of 5 were confirmed by X-ray crystallographic analysis
of the corresponding N-acetylated product, which was obtained on irradiation
of 3 (R ) Ac).The N-acetylated photosubstrate 3 was prepared from amine
7 by acetylation and subsequent NaOMe-mediated cyclization.Experimental
details and spectral data are included in the Supporting Information.
(6) For an example of photochemically mediated extrusion of sulfur from
an episulfide, see: Puiatti, M.; Arguello, J.; Penonory, A. Eur. J. Org. Chem.
2006, 4528–4536.
We propose that the triplet of enone 3, shown in con-
formations 8a and 8b, can react at the ipso position of the
benzothiazoline to give 9 and 11, respectively. Rearomati-
zation with concomitant ejection of the thiyl radical would
give 10 and 12, respectively. Diradical 10, in which the sulfur
radical and the R-keto radical are cis on the reduced quinoline
ring, can undergo recombination to afford the expected
photoproduct 4, along with episulfide 13, which would result
from homolytic fragmentation of the indicated σ-bond in 10
(red). However, the trans stereochemical relationship be-
tween the radical centers in 12 precludes carbon-sulfur bond
formation but could still lead to the formation of episulfide
13. Extrusion of sulfur from 13 can then lead to the formation
(7) Exclusive formation of the bridged cyclobutane is an exception to
the empirical “rule of five” as described by: Srinivasan, R.; Carlough, K. H.
J. Am. Chem. Soc. 1967, 89, 4932–4936. For an example of the observation
of the crossed product as a minor component in a photochemical reaction,
see: Basler, B.; Schuster, O.; Bach, T. J. Org. Chem. 2005, 70, 9798–9808.
Le Blanc, S.; Pete, J.-P.; Piva, O. Tetrahedron Lett. 1993, 34, 635–638.
For a clever example of a conformationally constrained system that leads
to the selective formation of the crossed photoproduct, see: Crimmins, M.;
Hauser, E. Org. Lett. 2000, 2, 281–284.
1686
Org. Lett., Vol. 11, No. 8, 2009

However, irradiation of 218 led to the formation of 24, albeit
in modest yield. Furthermore, the presence of benzyl acrylate
was observed by H NMR spectroscopy during the course
Scheme 6. Retro-Mannich Reaction of the Photoproduct 5
1
of the irradiation of 21, a result that is consistent with the
extrusion of sulfur from the episulfide intermediate to gen-
erate alkene. We note, however, that this result does not
preclude the direct fragmentation pathway for the formation
of 14 from 10 and/or 12 (Scheme 3).
Based on the results that were obtained with 16 and with
1,2,9 we anticipated that irradiation of 25, the sulfoxide
corresponding to 3, would lead to the selective formation of
the thiolane S-oxide corresponding to 4, at the expense of
the episulfide/enecarbamate pathway. In the event, irradiation
of 25 (R ) COOMe), available by m-CPBA oxidation of 3
(generating a single diasteromer which is assumed to be the
anti compound), afforded 26 as a 1:1 mixture of diastereo-
meric sulfoxides in modest yield, although none of the
cyclobutane product 5 could be detected in the reaction
mixture (Scheme 5). Thus, the stability of the sulfur-centered
Scheme 7. Proposed Mechanism for the Formation of 28
Scheme 5. Photoreaction of S-Oxide 20
tautomerization and protonation could generate 30, which
on retro-vinylogous Mannich fragmentation and loss of the
methyl carbamate would give 28.
In conclusion, we have discovered a novel desulfurative
photocycloaddition and results that are consistent with in situ
generation of an episulfide that affords an enecarbamate
intermediate, which subsequently undergoes a highly regi-
oselective “crossed” intramolecular [2 + 2] photoreaction.
Further studies on the photochemistry of these enone-
enecarbamate systems are underway in our laboratory, and
our results will be reported in due course.
radical,10 corresponding to 10/12 in Scheme 3, appears to
be a crucial factor in the formation of the cyclobutane product
4.
The unexpected formation of cyclobutane 5 prompted us
to examine its retro-Mannich fragmentation, which could lead
to the formation of spirocyclic products. In the event,
exposure of 5 to pTSA in ethanol led to the formation of 27
in 62% yield, based on recovered starting material (Scheme
6). We observed a pronounced solvent effect in this reaction,
as reaction of 5 with pTSA in benzene led to the formation
of quinoline 28.
Acknowledgment. We dedicate this paper to our friend
and mentor Professor Deukjoon Kim in celebration of his
60th birthday. H.J. thanks the Department of Defense for a
Prostate Cancer Research Program Predoctoral Traineeship
Award (W81XWH-06-1-0092).
The formation of 28 could result from retro-Mannich
fragmentation of 5 to give 29 (Scheme 7), which on
Supporting Information Available: Experimental pro-
cedures and NMR spectra for 3, 4, 5, 7, 14, 21, and 24-28.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(8) The preparation of 21 is based on the related work of: Fu, X.; Zhang,
S.; Yin, J.; McAllister, T. L.; Jiang, S. A.; Tann, C.-H.; Thiruvengadam,
T. K.; Zhang, F. Tetrahedron Lett. 2002, 43, 573–576.
(9) Lee, E. C.-Y. Ph. D. Dissertation, University of Pennsylvania,
Philadelphia, PA, 2007
(10) Vos, B. W.; Jenks, W. S. J. Am. Chem. Soc. 2002, 124, 2544–
2547.
.
OL900186Y
Org. Lett., Vol. 11, No. 8, 2009
1687