Synthesis of versatile bicyclo[5.4.0]undecane systems from tetrachlorocyclopropene

Article abstract of DOI:10.1016/j.tetlet.2004.01.055

2,2,3,4-Tetrachloro-8-oxabicyclo[3.2.1]octa-2,6-diene, derived in a single step from the cyclocondensation of furan and tetrachlorocylopropene, serves as a key intermediate for the construction of bicyclo[5.4.0]undecane synthons. Conversion to a meso-1,3 diketone is followed by a high yielding Robinson annulation reaction. Studies on the reduction of the enone product reveals a powerful preference for formation of the cis-ring fusion.

Full text of DOI:10.1016/j.tetlet.2004.01.055

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 2093–2096
Synthesis of versatile bicyclo[5.4.0]undecane systems from
tetrachlorocyclopropene
William A. Batson, Khalil A. Abboud, Merle A. Battiste and Dennis L. Wright*
Department of Chemistry, University of Florida, Gainesville, FL 32611, USA
Received 15 December 2003; revised 14 January 2004; accepted 14 January 2004
Abstract—2,2,3,4-Tetrachloro-8-oxabicyclo[3.2.1]octa-2,6-diene, derived in a single step from the cyclocondensation of furan and
tetrachlorocylopropene, serves as a key intermediate for the construction of bicyclo[5.4.0]undecane synthons. Conversion to a meso-
1,3 diketone is followed by a high yielding Robinson annulation reaction. Studies on the reduction of the enone product reveals a
powerful preference for formation of the cis-ring fusion.
Ó 2004 Elsevier Ltd. All rights reserved.
The bicyclo[5.4.0]undecane system is an important
structural moiety found in a variety of terpenoid natural
products. We have been interested in several targets
containing this system (Fig. 1) because of their unique
and often potent biological activity.
furan are highly versatile synthons, which can be readily
functionalized.4 A direct strategy for the assembly of
bicyclo[5.4.0]undecane building blocks would be the
annulation of a six-membered ring on to these bicyclic
systems followed by opening of the bridging ether.
Herein, we report our efforts to apply a Robinson
annulation5 strategy to effect this conversion.
Erinacine C1 is an inducer of NGF, guanacastepene A2
is a potent antibiotic and frondosin D3 has shown
powerful anti-HIV properties. The prevalence of this
structure in many natural products has inspired a vari-
ety of efforts to develop general methods for the con-
struction of these systems.
The inspiration for the use of the Robinson annulation
was Tobey and LawÕs6 synthesis of the 1,3-diketone 3
(Scheme 1) in just two steps from commercially available
tetrachlorocyclopropene.
Recent efforts from these laboratories have shown that
the oxabicyclo[3.2.1]octadiene derivatives prepared
through the reaction of tetrahalocyclopropenes and
Cyclocondensation of 1 and furan proceeds through an
initial Diels–Alder reaction6;7 to produce 2 in high yield.
Direct hydrolysis of 2 under strongly acidic conditions
and elevated temperatures delivers the 1,3-dione 3.
Although the yield for this step is moderate (ꢀ50%), the
chloro diketone is relatively easy to produce in multi-
gram quantities. Intermediate 3 was attractive for the
synthesis of bicyclo[5.4.0]undecane systems because
such diketone moieties are highly effective in Robinson
annulation reactions.8 Moreover, since this hydrolysis
converts the racemic adduct into a meso-compound, it
may also be possible to execute the annulation with a
Figure 1. Natural products with a bicyclo[5.4.0]undecane ring system.
Keywords: Tetrachlorocyclopropene; Furan; Robinson annulation;
Bicyclo[5.4.0]undecane.
* Corresponding author. Tel.: +1-603-646-6481; fax: +1-603-448-9766/
646-6481; e-mail: dennis.l.wright@dartmouth.edu
Corresponding author concerning X-ray structural studies.
Scheme 1.
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.01.055

2094
W. A. Batson et al. / Tetrahedron Letters 45 (2004) 2093–2096
chiral catalyst (Hajos–Parrish reaction9) to produce
nonracemic adducts.
Initial studies involved the conversion of the known
chloro diketone 3 into a bicyclo[5.4.0]undecane system.
Our initial efforts to effect a Michael addition between 3
and methyl vinyl ketone were frustrated by the sensi-
tivity of 3 to the typical hydroxide or alkoxide bases
used to promote this reaction. Evidence was observed
for products arising from a ring-opening retro-Claisen
reaction. Fortunately, it was found that triethylamine
was an excellent base and produced a single Michael
adduct 4 in 75% isolated yield (Scheme 2).
An important issue in this reaction was the relative
configuration of the newly formed stereogenic center. It
was presumed that the conformation produced by the
bridging ether would bias attack of electrophiles to the
exo-face, syn to the oxabridge. An X-ray structure of
Michael adduct 4 verified this preference (Fig. 2).
Scheme 3.
The effectiveness of this process encouraged the inves-
tigation of more relevant compounds that have an
angular methyl group rather than the angular halogen
atom as in 6. Exchange of chlorine for a methyl group in
compound 3 was envisioned to give a more valuable
meso-diketone (Scheme 3).
Subjecting the trione 4 to an alkoxide base resulted in a
highly effective ring closure reaction. Interestingly, the
alkoxide bases that caused decomposition of dione 3
were compatible with the trione 4, perhaps because of
steric shielding of the endocyclic carbonyls from nucleo-
philic attack. The base catalyzed process led to the
formation of aldol product 5 along with minor amounts
of the annulation product 6. Treatment of the crude
mixture with tosic acid resulted in complete dehydration
to give the desired enone 6 in 70% isolated yield.
Initially, this transformation was envisioned as a one-
step operation whereby reduction of the chloro diketone
to the enolate would be followed by direct alkylation.
Unfortunately, it was not possible to produce the enol-
ate directly with zinc or magnesium, although reduction
to the parent diketone 7 under acidic conditions was
possible. However, the isolation of 7 was difficult owing
to high water solubility and ready deprotonation and
unfortunately, useful quantities of the unsubstituted
compound could not be obtained.
An alternative route involved initial alkylation with
methyl iodide to produce 8. Interestingly, the activated
proton in 3 is so acidic that triethylamine can be used as
a base for this alkylation. X-ray analysis (Fig. 3) once
again showed that the bias of the bridged bicyclic system
had forced the small electrophile exclusively to the exo-
face.
This compound was smoothly reduced by zinc and
hydrochloric acid to produce diketone 9 in good
yield. Conversion of this compound to the analogous
bicyclic building block proved relatively straightforward
(Scheme 4).
Scheme 2.
Figure 2. X-ray structure of 4.
Figure 3. X-ray structure of 8.

W. A. Batson et al. / Tetrahedron Letters 45 (2004) 2093–2096
2095
Figure 4. X-ray structure of ketone 12.
Scheme 4.
The X-ray structure clearly revealed that reduction had
occurred exclusively from the more hindered endo-face,
giving the cis-fused product 12 instead of the anticipated
13. In an attempt to reverse the selectivity and obtain the
trans-fused compound as well, the nonconjugated alkene
was selectively reduced with diimide. The conversion of
these sp2 centers to sp3 should increase the steric shield-
ing of the endo-face on the enone. However, reduction of
14 again delivered only the cis-fused isomer 12.
Amine promoted Michael addition proceeded unevent-
fully to produce tri-ketone 10 in excellent yield. A
variety of conditions were examined for the dehydrative
aldol cyclization and the use of pyrrolidine was found to
be far superior to the alkoxide bases employed previ-
ously. Under these conditions, the enone 11 was
obtained in a single step in 80% yield. This substructure
is common to both erinacine C and guanacastepene A
and is available in only six steps from furan.
In an attempt to explain this strong preference for the
formation of the cis-isomer, computational studies were
made on the two diketones. AM1 calculations11 revealed
that cis-isomer 12 is approximately 3.8 kcal/mol more
stable than trans-isomer 13, likely resulting from strain
on the seven-membered ring. This large difference in
thermodynamic stability may explain the strong prefer-
ence for formation of the cis-fused product.
A final aspect we wished to examine in these preliminary
studies was the stereoselectivity of the reduction of the
enone. Controlling the ring fusion on bicyclo-[5.4.0]un-
decanes is often difficult10 because there is very little
energy difference between the alternative isomers.
However, it seemed that the rigidity in the seven-mem-
bered ring imparted by the bridging ether may kineti-
cally control the formation of this new center.
Examination of an AM1 minimized structure11 (see
Scheme 4) suggested that approach to the endo-face of
the enone was hindered by the two-carbon bridge of the
oxabicyclo[3.2.1]octene ring as well as the endo-methyl
group at the angular position. Reduction from the exo-
face would deliver the trans-fused system common to the
erinacines (Scheme 5).
The adduct derived in a single step from furan and tet-
rachlorocyclopropene can be easily converted into var-
ious bicyclo[5.4.0]undecane building blocks through a
key Robinson annulation. We are currently examining
the asymmetric preparation of these compounds and
their application to the synthesis of terpenoid natural
products.
Reduction of the diene under an atmosphere of hydro-
gen and either a palladium or platinum catalyst yielded
a single product in very high yield. Determination of the
relative stereochemistry by NMR methods was difficult
and recourse was made to X-ray crystallography (Fig.
4).
Acknowledgements
The authors would like to thank the National Science
Foundation, the AlzheimerÕs Association and the Uni-
versity of Florida for Support of this work.
References and notes
1. Kawagishi, H.; Shimada, A.; Shirai, R.; Okamoto, K.;
Ojima, F.; Sakamoto, H.; Ishiguro, Y.; Furukawa, S.
Tetrahedron Lett. 1994, 35, 1569–1572.
2. (a) Brady, S. F.; Singh, M. P.; Janso, J. E.; Clardy, S. J. Am.
Chem. Soc. 2000, 122, 2116–2117; (b) Singh, M. P.; Janso,
J. E.; Luckman, S. W.; Brady, S. F.; Clardy, J.; Greenstein,
M.; Maiese, W. M. J. Antibiot. 2000, 53, 256–261.
3. Hallock, Y. F.; Cardellina, J. H.; Boyd, M. R. Nat. Prod.
Lett. 1998, 11, 153–160.
4. (a) Orugunty, R. S.; Wright, D. L.; Battiste, M. A.;
Abboud, K. A. Org. Lett. 2002, 4, 1997–2000; (b)
Orugunty, R. S.; Wright, D. L.; Battiste, M. A.; Helmich,
R. J.; Abboud, K. A. J. Org. Chem. 2003, 406.
Scheme 5.

2096
W. A. Batson et al. / Tetrahedron Letters 45 (2004) 2093–2096
5. For an example of a Robinson approach to related
systems, see: Ireland, R. E.; Aristoff, P. A.; Hoyng, C. F.
J. Org. Chem. 1979, 44, 4318–4322.
8. Jung, M. E. Tetrahedron 1976, 32, 3–31.
9. Hajos, Z. G.; Parrish, D. R. J. Org. Chem. 1974, 39, 1615–
1621.
6. Law, D. C. F.; Tobey, S. W. J. Am. Chem. Soc. 1968, 90,
2376–2386.
7. Orugunty, R. S.; Ghiviriga, I.; Abboud, K. A.;
Battiste, M. A.; Wright, D. L. J. Org. Chem. 2003,
570.
10. For a relevant example, see: Majetich, G.; Song, J. S.;
Ringold, C.; Nemeth, G. A.; Newton, M. G. J. Org. Chem.
1991, 56, 3973–3988.
11. AM1 calculations performed on a Macintosh G5 using the
Spartan modeling package from Wavefunction.