At this point, treatment with titanium tetrachloride served
to form compound 4 in essentially quantitative yield.
From compound 4, three different cyclization substrates
were prepared. For the simple, â-unsubstituted compound
1, treatment with excess lithium aluminum hydride, followed
by acidic hydrolysis of the intermediate diol, afforded the
desired compound 1 in 52% yield. The dimethyl substrate 5
was prepared in similar fashion by adding an excess of
methyllithium, followed by acidic hydrolysis. Finally, the
monomethyl substrate 6 was prepared by addition first of 1
equiv of DIBAL-H, then 1 equiv of methyllithium, and
finally acidic hydrolysis. This sequence afforded compound
6 in a reasonable 55% yield overall.
trans, since it is not readily cleaved to afford the starting
materials 5 and 6 by treatment with sodium periodate.
Since nonelectrochemical conditions are also known to
promote the reductive cyclization reaction, we chose to
investigate one of these alternatives and determine what
effect it had on the cyclization and product outcome of these
reactions.7 Our choice was the use of samarium diiodide.
Treatment of substrate 1 with 3 equiv of samarium diiodide
at room temperature in THF under argon afforded the
anticipated cyclization product 7 in 25% yield. Again, the
product was isolated as a single isomer, the same as under
the electrochemical conditions. Interestingly, treatment of 5
and 6 under these samarium diiodide conditions also afforded
the reductive cyclization products 10 and 11 as single
isomers, with none of the pinacol product being detected.
The stereochemistry of product 11 is assigned as cis/anti/cis
on the basis of the stereochemistry of product 7 and the fact
that product 11 is symmetrical (only eight signals observed
in the 13C NMR spectrum and only one methyl signal
With these three compounds in hand, the reductive
cyclization was first explored under electrochemical condi-
tions (Scheme 2). Subjection of compound 1 to a constant
Scheme 2. Electrochemical Cyclizations
1
observed in the H NMR spectrum). Further evidence for
the cis ring fusion comes from the observation of an NOE
between the ring fusion methyl and the ring fusion proton.
The cis/anti/cis arrangement of product 10 is assigned by
analogy with the outcome of the other two cyclizations.
While these results clearly indicated that reductive cy-
clizations can be achieved on these shorter tether compounds,
they also raised the question as to why the electrochemical
conditions were leading to pinacol and the samarium diiodide
conditions to reductive cyclization. This question is even
more perplexing, since no similar difference was observed
in the cyclization of a compound related to 5 but with two
carbons in the tether between the two enones.8 In examining
these systems, one possible hypothesis is that chelation and
steric effects play the determining factor. Assuming that the
reductive cyclization of these compounds is proceeding via
a one-electron reduction of each enone to a radical anion,
followed by radical/radical coupling (the most commonly
postulated mechanism for such reactions), then the situation
outlined in Scheme 4 could help to explain the observed
results.1
current of 200 mA for 1 h using a sacrificial tin anode and
a platinum cathode in aqueous acetonitrile with tetraethyl-
ammonium chloride as the supporting electrolyte afforded
the anticipated cyclization product 7 in 73% yield as a single
isomer.4 The stereochemistry of this product was confirmed
to be cis/anti/cis by double Wolff-Kishner reduction to
afford the known perhydrofluorene compound.5
When there is no hindrance at the â-position, then the
reaction proceeds via the normal reductive cyclization
Interestingly, the reductive cyclizations of substrates 5 and
6 did not follow the same pattern. Both of these compounds
afforded only the pinacol-type products 8 and 9 under
electrochemical conditions. Although somewhat surprising,
this is not unprecedented, since steric hindrance at the
â-position is known to favor the pinacol product in the
intermolecular hydrodimerization of cyclic enones.6 The
stereochemistry of the pinacol products is assigned as being
Scheme 3. Samarium Diiodide Cyclizations
(4) Conditions were adapted from those reported by Kise and co-workers.
Kise, N.; Iitaka, S.; Iwasaki, K.; Ueda, N. J. Org. Chem. 2002, 67, 8305-
8315.
(5) Meusinger, R.; Rohloff, J.; Schumann, F.; Herzschuh, R. J. Prakt.
Chem./Chem.-Ztg. 1997, 339, 128-134.
(6) In addition to ref 1, for a specific example, see: Tissot, P.; Surbeck,
J.-P.; Guelacar, F. O.; Margaretha, P. HelV. Chim. Acta 1977, 60, 1472-
1477.
1554
Org. Lett., Vol. 7, No. 8, 2005