An unusual samarium diiodide mediated reductive ring contraction of a tricyclic oxazine to a highly-functionalized cyclopentane and cyclobutane.

Article abstract of DOI:10.1021/ol000057q

[reaction--see text] Samarium diiodide mediated reductive ring contraction of a substituted tricyclo[2.2.2]oxazine at 25 degrees C leads to a mixture of cyclopentane and cyclobutane rearrangement products with complete diastereoselectivity in each case. A

Full text of DOI:10.1021/ol000057q

ORGANIC
LETTERS
2000
Vol. 2, No. 10
1457-1459
An Unusual Samarium Diiodide
Mediated Reductive Ring Contraction of
a Tricyclic Oxazine to a
Highly-Functionalized Cyclopentane and
Cyclobutane
Barry J. McAuley, Mark Nieuwenhuyzen, and Gary N. Sheldrake*
School of Chemistry, DaVid Keir Building, The Queen’s UniVersity of Belfast,
Belfast BT9 5AG, U.K.
g.sheldrake@qub.ac.uk
Received March 14, 2000
ABSTRACT
Samarium diiodide mediated reductive ring contraction of a substituted tricyclo[2.2.2]oxazine at 25 °C leads to a mixture of cyclopentane and
cyclobutane rearrangement products with complete diastereoselectivity in each case. At −78 °C, the anticipated amidocyclohexanol reduction
product is obtained exclusively, while the cyclopentane is the sole product at reflux in THF.
Samarium diiodide is a versatile single-electron reducing
reagent which has been used widely in the past two decades
for a variety of reductions,1 pinacol coupling reactions,2
Barbier-type reactions,3 and many other rearrangements
involving radical intermediates. The utility of samarium
diiodide for the selective reduction of N-O bonds in
oxazines was demonstrated by Keck and co-workers for
the preparation of amido alcohol 2 from the corresponding
bicyclic oxazine 1 (Scheme 1).
In the course of our synthetic studies with enzymatically
produced arene cis-dihydrodiols,5 we obtained cis-dihydro-
diol 3 as a single enantiomer from toluene via enzymatic
cis-dihydroxylation using a whole-cell culture of the mutant
microorganism Pseudomonas putida UV4. This organism
contains a dioxygenase enzyme, which catalyses the arene
cis-dihydroxylation, but lacks the diol dehydrogenase en-
zyme, which catalyses the subsequent step in the normal
metabolic pathway.5 Protection of the cis-dihydrodiol as a
1,3-dioxalane and then Diels-Alder cycloaddition with
nitrosocarbonyl dienophile 4, generated in situ by periodate
4
Scheme 1. Samarium Diiodide Reduction of a Bicyclic
(1) Soderquist, J. A. Aldrichimica Acta 1991, 24, 15-23. Molander, G.
A.; Chem. ReV. 1992, 92, 29-68. Molander, G. A.; Harris, C. R. Chem.
ReV. 1996, 96, 307-338.
Oxazine
(2) Nomura, R.; Matsuno, T.; Endo, T. J. Am. Chem. Soc. 1996, 118,
11666-11667.
(3) Girard, P.; Namy, J. L.; Kagan, H. B. J. Am. Chem. Soc. 1980, 102,
2693-2698. Curran, D. P.; Gu, X.; Zhang, W.; Dowd, P. Tetrahedron 1997,
53, 9023-9042.
(4) Keck, G. E.; McHardy, S. F.; Wager, T. T. Tetrahedron Lett. 1995,
36, 7419-7422.
(5) Boyd, D. R.; Sheldrake, G. N. Nat. Prod. Rep. 1998, 15, 309-324.
10.1021/ol000057q CCC: $19.00 © 2000 American Chemical Society
Published on Web 04/21/2000

oxidation of benzohydroxamic acid, gave tricyclic oxazine
5 with complete regio- and diastereoselectivity (Scheme 2).
absolute configuration of 8 was assigned from the relative
stereochemistry determined in the crystal structure on the
assumption that the configurations at carbons C-2 and C-3
had remained unchanged: it is difficult to imagine a
mechanism that would result in the simultaneous inversion
of configuration at both of these positions.
Scheme 2. Chemoenzymatic Preparation of Substituted
Tricyclic Oxazine 5
Careful examination of the crude product mixture by NMR
spectroscopy revealed that 7 and 8 were formed as single
diastereoisomers indicating that the two ring contractions had
occurred with complete stereochemical control. The relative
1
configuration of cyclopentane 7 was assigned by H NMR
spectroscopic analysis.6 The key assignments are the trans
relationship between the methine hydrogens H-1 and H-5 (J
) 0 Hz, consistent with a 90° dihedral angle) and the cis
relationship between H-1 and H-2 (J ) 5.9 Hz). Again,
assuming that the configurations at C-2 and C-3 remain
unchanged during the ring contraction, this also establishes
the absolute configurations at C-1 and C-5.
Attempts to optimize the formation of either or both of
the rearrangement products by variation of the reaction
conditions led to some interesting observations. While a large
excess of commercial samarium diiodide solution at first
appeared to be necessary to obtain complete conversion of
the starting material, rigorous exclusion of air and moisture
and preparation of a fresh SmI2 solution 7 immediately prior
to reaction established that 2 equiv of the reductant was
required. Performance of the reaction at -78 °C led to
exclusive formation of amido alcohol 6, while reaction in
refluxing THF (67 °C) led to formation only of cyclopentane
product 7. So far we have been unable to improve the ratio
of products 8:7 to better than 1:4 (0.5 h at 0 °C, at which
temperature small amounts of amido alcohol 6 were also
observed).
Samarium diiodide was chosen as the reagent for the mild
and selective reduction of 5 to give amido alcohol 6.
When the reductive cleavage was carried out at 25 °C,
with an excess of samarium diiodide, two products (7 and
8, Scheme 3) were isolated (ratio 9:1 by 1H NMR spectros-
Scheme 3. Samarium Diiodide Reduction of Oxazine 5 at
Different Temperatures
Investigation of the mechanism(s) of these unusual rear-
rangements is ongoing, but at this early stage a possible
mechanistic pathway may be conjectured. To obtain the
methyl ketone functionality it would seem likely that the
initial transfer of an electron from the Sm2+ ion to the N-O
bond results in an oxygen-centered radical and an anion on
the amide nitrogen (Scheme 4).
Under “normal” N-O reduction conditions a second
electron is transferred from another Sm2+ ion to give an oxy-
anion which is then protonated either immediately or during
workup. This process is observed when the reaction is carried
out at -78 °C. At higher temperatures it appears that the
oxygen-centered radical fragments to give an oxygen-
stabilized, carbon-centered dioxalanyl radical. All attempts
to effect this rearrangement starting from amido alcohol 6,
under acidic or basic conditions, have failed to produce either
of the ring-contraction products 7 or 8. One possible
explanation is that the radical is transferred from oxygen to
C-1 in a concerted process involving formation of the C-O
double bond and homolytic fission of the C-1/C-6 bond.
Stereoselective addition of this radical to the top face of the
alkene double bond, possibly with coordination assistance
by Sm2+ or Sm3+, would lead to the observed products via
copy), with a combined yield of 85%. Neither of the products
had spectroscopic data compatible with the expected amido
alcohol product 6. The products were separated using
preparative thin-layer chromatography, and mass spectro-
scopic analysis indicated that they were isomeric. IR and
NMR spectroscopy revealed that both products were methyl
ketones and that both retained the dioxalane-protected cis-
diol functionality.
Unequivocal confirmation of the cyclobutane structure for
8 was obtained by X-ray crystallographic analysis.6 The
(7) Namy, J. L.; Girard, P.; Kagan, H. B.; Caro, P. E. NouV. J. Chim.
1981, 5, 479-484.
(6) See Supporting Information.
1458
Org. Lett., Vol. 2, No. 10, 2000

Scheme 4. Suggested Mechanism for the Ring Contractions to
the Cyclopentane 7 and Cyclobutane 8
Scheme 5. Reduction of Desmethyl Oxazine 9 with Samarium
Diiodide To Give Amido Alcohol 10 without Competing Ring
Contractions
cannot be persuaded to undergo rearrangement. It is possible
that the presence of the extra methyl group in 5 sterically
inhibits approach of further Sm2+ ions to the tertiary oxygen
radical long enough for fragmentation to be preferred over
a second electron-transfer process.
This unusual rearrangement process has provided a new
method for the preparation of highly functionalized cyclo-
pentane and cyclobutane ring systems with a high degree of
stereochemical control. These enantiopure products have four
asymmetric centers and are readily prepared in synthetically
useful quantities from toluene in four steps. Utilizing the
enantiocomplementary strategies for obtaining both enanti-
9
omers of the arene cis-dihydrodiols developed in these
a 5-endo-trig or 4-exo-trig process. Ring closure to the five-
membered ring would be favored from ring-strain consid-
erations, but addition of the electron-rich nucleophilic radical
to the alkene double bond would also be facilitated by the
conjugation of this alkene to the ketone. This could account
for the unusual ring closure to the cyclobutane. Reduction
of the rearranged radical with a second Sm2+ ion and then
protonation during workup would complete the process.
While we currently have no direct evidence to support this
mechanism, there have been a few reported examples of this
type of rearrangement process involving dioxalanyl radicals.
For example, Murphy and co-workers have demonstrated that
dioxalanyl radicals generated from nitrite esters undergo
stereoselective intramolecular additions to alkenes.8
laboratories, there is also ready access to both antipodes of
the rearrangement products. Synthetic applications for these
new and versatile additions to the chiral pool are currently
under investigation.
Acknowledgment. We thank DENI for financial support
through a Distinction Award (B.M.), Mr. D. Clarke
(QUESTOR Centre, The Queen’s University of Belfast) for
carrying out the biotransformations, and Avecia Life Science
Molecules, Blackley, Manchester, for permission to use the
microorganism Pseudomonas putida UV4.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for compounds 5-8 and
X-ray crystallographic data for compound 8. This information
is available free of charge via the Internet at http://pubs.acs.org.
One observation which is not fully explained by our
mechanistic interpretation is that the des-methyl Diels-Alder
adduct 9 gives only amido alcohol reduction product 10 under
any of the reaction conditions described (Scheme 5) and
OL000057Q
(9) Allen, C. C. R.; Boyd, D. R.; Dalton, H.; Sharma, N. D.; Brannigan,
I.; Kerley, N. A.; Sheldrake, G. N.; Taylor, S. C. Chem. Commun. 1995,
117-118.
(8) Batsanov, A. S.; Begley, M. J.; Fletcher, R. J.; Murphy, J. A.;
Sherburn, M. S. J. Chem. Soc., Perkin Trans. 1 1995, 1281-1294.
Org. Lett., Vol. 2, No. 10, 2000
1459