Observations on the samarium diiodide-promoted C-C fragmentation/ring expansion chemistry of some aliphatic 1,4-diketones

Article abstract of DOI:10.1039/b008110h

The observations on the samarium diiodide-promoted C-C fragmentation/ring expansion chemistry of some aliphatic 1,4-diketones were presented. It was found that some photochemical electron transfer, as well as SmI₂-mediated transformations were carried out on non-aromatic substrates. It was shown that given the otherwise identical molecule, the ability of the SmI₂ HMPA complex to chelate the 1,4-dicarbonyl moiety in the one case and not in the other is the primary factor that determines the fragmentation reaction.

Full text of DOI:10.1039/b008110h

View Article Online / Journal Homepage / Table of Contents for this issue
Observations on the samarium diiodide-promoted C–C
fragmentation/ring expansion chemistry of some aliphatic
1,4-diketones
1
D. Bradley G. Williams,* Kevin Blann and Cedric W. Holzapfel
Department of Chemistry and Biochemistry, Rand Afrikaans University, PO Box 524,
Auckland Park, 2006, South Africa
Received (in Cambridge, UK) 9th October 2000, Accepted 7th December 2000
First published as an Advance Article on the web 10th January 2001
Various multicyclic and bridged 1,4-diketones were
subjected to the reductive conditions of the SmI2–HMPA
system, often affording surprising results and reactivity
upon fragmentation of the 2,3 C–C bond to provide the
corresponding ring enlarged product.
the SmI2–HMPA complex to chelate the 1,4-dicarbonyl moiety
in the one case (endo,endo) and not in the other (endo,exo) is the
primary factor that determines the success of the fragmentation
reaction. This was to be the first indication of the influence of
structure on the reactivity of 1,4-dicarbonyl systems.
In keeping with the theme of norbornene† derivatives, several
other 1,4-diketone Diels–Alder adducts were prepared accord-
ing to the literature,8c and subjected to reaction with SmI2–
HMPA in THF (addition of the substrate to the SmI2 solution
at room temperature).3 Diolefin 4 afforded the analogous
mono-olefin 5 (41%) and the fragmented macrocyclic product 6
(25%), upon reaction with 2.4 equivalents of SmI2 (Scheme 2).
Although some SmI2-promoted C–C fragmentation reactions
have been reported, these have been restricted to ring-strained
systems.1 Our previous endeavours in the field of SmI2-
mediated reactions2 have led us to investigate some fragmen-
tation reactions, and we have recently described the first
SmI2-mediated C–C fragmentation reactions of open chain
aromatic 1,4-diketones.3 As an extension of that work, we
investigated similar reactions making use of non-aromatic sub-
strates. Herein, we detail the often unpredicted reactivity of
some multicyclic/bridged 1,4-dicarbonyl substrates, and the
influence of reaction conditions and remote functionality on
the outcome of these reactions. In this respect, we noticed
that some photochemical electron transfer,4 as well as SmI2-
mediated,5 transformations have been carried out on similar
substrates, the latter produced results different from our own
(see conclusion).
Scheme 2
Subsequent to our work on aromatic 1,4-diketone sub-
strates,3 we turned our attention to more strained systems, in an
attempt to gauge the generality of this reductive transform-
ation, making use of the work of Ghosh6 as a benchmark. We
could readily repeat those results (Scheme 1), but found that
Making use of an excess (4.8 equivalents) of SmI2, the selectiv-
ity of the reaction could be reversed, affording the products 5
and 6 in yields of 17% and 41%, respectively. These results
indicate that the rate of the reduction of the conjugated double
bond is faster than the rate of fragmentation, and that diketone
5 is an intermediate in the transformation of 4 into 6. The latter
theory was confirmed by re-subjecting compound 5 to reaction
with SmI2–HMPA in THF, which afforded the anticipated
fragmented material 6.
The diolefin 4 was readily saturated under the action of H2 in
the presence of Pd/C, to afford the saturated product 7 in good
yield (Scheme 3). Surprisingly, this substrate afforded the prod-
uct of the fragmentation of the ‘external ’ bond (8, Scheme 3) in
Scheme 1
some starting material remained after completion of the reac-
tion (as identified by disappearance of the unmistakable colour
of the SmI2–HMPA complex), along with the expected product
of fragmentation 2. In addition to these compounds, a small
amount of the keto-alcohol 3 was isolated,7 the production of
which would also have consumed SmI2; this would account for
the unreacted starting material. During the synthesis of the
endo,endo substrate 1a,8 we also isolated some of the endo,exo
analogue 1b.9 We subjected this substrate to reaction with
SmI2–HMPA at room temperature, and found that the frag-
mented product 2 could be isolated in a yield of 18%, along
with an intractable mass. Heating to reflux, cooling to 0 ЊC, or
the addition of a proton source did not have any positive effect
on the outcome of the reaction. It can be reasonably argued
that, given the otherwise identical molecule, the ability or not of
Scheme 3
reactions in which 7 was added to a solution of SmI2–HMPA in
THF: no macrocyclic product of fragmentation of the ‘internal ’
C–C bond was detected from these reactions. This result,
together with the previous reaction of the mono-reduced
compound 5, indicated that the presence or absence of the
’norbornene double bond’ is determinative of which C–C bond
fragments. The change in selectivity is possibly due to subtle
DOI: 10.1039/b008110h
J. Chem. Soc., Perkin Trans. 1, 2001, 219–220
This journal is © The Royal Society of Chemistry 2001
219

View Article Online
energy differences in the ‘external ’ versus the ‘internal ’ bond,
these differences are due either to additional ring strain, or to
conformational considerations, induced by the double bond.
When added in a dropwise fashion to SmI2–HMPA in THF,
the adduct 9 yielded an altogether different product from any
other norbornene derivative that we had studied. While the
aromatic macrocyclic product of fragmentation was expected,
the reaction provided the quinone product 10 of a retro Diels–
Alder reaction (Scheme 4). The product was proved to be that
Scheme 6
fragmentation–pinacol coupling reaction to an intact cage unit
(based on our tentative structural assignment).7 The slow
addition of the SmI2–HMPA solution presumably afforded the
ketyl radicals (that had formed upon initial single electron
transfer to the ketone) sufficient time to combine in a pinacol-
type reaction, rather than undergo further reduction by excess
SmI2 to afford the fragmented monomeric material (12a)
previously isolated.
In summary, we have shown that certain strained compounds
undergo C–C fragmentation reactions, in an often unpredicted
manner. This methodology has allowed ready access to some
interesting macrocyclic and other structures, and has yielded
an unexpected retro Diels–Alder reaction. In contrast to previ-
ous work in which HMPA was not employed as cosolvent for
SmI2-promoted reactions of similar substrates,5 and in which
products of oxidation were obtained, our experiments consist-
ently afforded the products of reductive fragmentation. Once
again, the dramatic influence of HMPA on the outcome of
some SmI2-mediated reactions has been demonstrated.3
Scheme 4
of a reaction with SmI2 rather than with in situ generated
Sm(III) salt (which is known to be a Lewis acid)10 by a separate
reaction of substrate 9 with an ex situ generated SmI2OBn salt.
The reaction can reasonably be anticipated to have proceeded
via the disamarium(III) salt of hydronaphthoquinone; this
intermediate would have undergone facile auto-oxidation to
naphthoquinone upon contact with air. This observation might
represent an alternative means of protecting a naphthoquinone
moiety in synthesis.
In response to the interesting results obtained from the
norbornene derivatives, we wished to investigate the reactivity
of diketone cage compounds,11 which were readily prepared
from the corresponding Diels–Alder adducts by a [2ϩ2] photo-
induced cycloaddition. When the cage compounds 11 were
treated with SmI2–HMPA in THF at either the reflux temper-
ature of the reaction mixture or at any temperature above –50
ЊC, an intractable mixture of products was formed. However,
cooling of the reaction mixture to –78 ЊC allowed the isolation
of the fragmented products 12 (Scheme 5) after the SmI2–
Acknowledgement
We thank the Rand Afrikaans University for support of this
work.
Notes and references
† The IUPAC name for norbornene is bicyclo[2.2.1]hept-2-ene.
1 (a) R. A. Batey, J. D. Harling and W. B. Motherwell, Tetrahedron,
1996, 52, 11421; (b) G. L. Lange, L. Furlan and M. C. MacKinnon,
Tetrahedron Lett., 1998, 39, 5489.
2 (a) J. J. C. Grové, C. W. Holzapfel and D. B. G. Williams, Tetra-
hedron Lett., 1996, 37, 5817; (b) J. J. C. Grové, C. W. Holzapfel and
D. B. G. Williams, Tetrahedron Lett., 1996, 37, 1305.
3 D. B. G. Williams, K. Blann and C. W. Holzapfel, J. Org. Chem.,
2000, 65, 2834.
4 B. Pandey, A. T. Rao., P. V. Dalvi and P. Humar, Tetrahedron, 1994,
50, 3835.
Scheme 5
HMPA solution had been added in a rapid dropwise fashion.
Not unexpectedly, it was the more strained 2,3 C–C bond
that fragmented. The fact that the reaction had to be carried
out at such a low temperature demonstrated the remarkable
degree of ring strain, and hence the ease with which the C–C
bond is fragmented, in cage compounds. Similar fragmen-
tation reactions have been carried out on cage compounds
under reductive12 and under photochemical electron transfer4
conditions, with inconsistent results.
In a repeat reaction making use of cage compound 11a, the
THF solution of SmI2–HMPA was added to the substrate by
very slow dropwise addition. It was of interest to us to note
that, along with the C–C fragmented product 12a (15%), the
reaction yielded a dimeric material 13 (17%, Scheme 6), in
which one of the cage units had undergone a sequential
5 B. Pandey, K. Saravanan, A. T. Rao, D. Nagamani and P. Kumar,
Tetrahedron Lett., 1995, 36, 1145.
6 A. Haque and S. Ghosh, Chem. Commun., 1997, 2039.
7 All products provided satisfactory analytical data, which routinely
included IR, 1H and 13C NMR spectra, various 2D and other
NMR experiments (HETCOR, COSY or 1H–1H decoupling, NOE),
FAB-MS, and FAB-HRMS.
8 (a) O. Diels and K. Alder, Liebigs Ann. Chem., 1928, 98;
(b) S. Ghosh, S. S. Roy and A. Bhattacharya, Synth. Commun., 1989,
19, 3191; (c) J. Sauer, Angew. Chem., Int. Ed. Engl., 1966, 5, 211.
9 H. J. Wu, C. Y. Wu and C.-C. Lin, Chin. Chem. Lett., 1996, 7, 15.
10 D. A. Evans, A. R. Muci and R. Stürmer, J. Org. Chem., 1993, 58,
5307.
11 R. C. Cookson, E. Crundwell and J. Hudec, Chem. Ind. (London)
1958, 1003.
12 E. Wenkert and J. E. Yoder, J. Org. Chem., 1970, 35, 2986.
220
J. Chem. Soc., Perkin Trans. 1, 2001, 219–220