28% yield under Ar suggested that oxidants are essential for
catalytic turnover (entry 2). Encouraged by these results, we
next screened Cu(II) salts as possible catalysts. Among the
“monodentate-type” carboxylate salts (entries 1-5), Cu(O-
COCF3)2 gave the best yield (entry 3). Although the benzoate
salts resulted in a moderate yield (entry 5), the bidentate-
type salicylate salts (5 and 6)6 provided 3i in higher yields.
In place of air as an oxidant, NaIO4 could be used and
OXONE was found to accelerate considerably the catalytic
reaction to provide 3i in 2 h in high yield (entry 9). Even
using 2 mol % of the catalyst, 3i was obtained in good yield
(entry 10). From these results, the rate determining step
would be reoxidation of Cu(I) to Cu(II) in this catalytic
system. Finally, the Wittig reagent 2b having 2,6-dimeth-
ylphenylthioester was found to provide 3i in excellent yield
using only 2 mol % of the catalyst, probably because the
sterically more congested Cu(II)SAr complex (see Figure 3)
might be more easily converted into disulfide and Cu(I).
To establish the generality of this catalytic lactonization,
various kinds of acyloins were reacted with 2b using 6 as a
catalyst in toluene.7 As shown in Figure 2, several cyclic and
gested acyloin (1j), the key intermediates (1k, 1l) for the
syntheses of the bioactive natural products, sundiversifolide8
and heritonin,9 respectively, also afforded the corresponding
lactones. Compared to the low yield (1.4%) by the intramo-
lecular Horner-Emmons-lactonization,9a our method gave the
product in much higher yield.
A plausible mechanism for this catalytic reaction is shown
in Figure 3. The thiol ester would be activated by Cu(II)
Figure 3. Proposed mechanism.
catalyst as Masamune proposed.2a,b In this catalytic system,
the phosphorus ylide on the R position would be an essential
substituent, because simple thiol esters without a phosphorus
ylide were not converted into the corresponding O-ester by
this catalyst. Therefore, we estimated that the ketene-like
intermediate, as illustrated in 7, would be generated by a
push-pull mechanism, which involves the double activation
of the C-S bond by both the soft Lewis acidic Cu (II) (pull)
and the electron-donating phosphorus ylide (push). The
ketene intermediate 7 would undergo rapid acylation, fol-
lowed by the intramolecular Wittig reaction, to provide the
lactone 3. A redox reaction of the thiolate complex 8 would
afford a thiyl radical 10, which would be converted into the
disulfide, and the Cu(I) salt 9, which would be oxidized to
regenerate the catalyst 6.
In conclusion, we have found the catalytic acylation of
alcohols with a thiol ester present in Wittig reagents under
neutral conditions catalyzed by the Cu(II) salt by a push-pull
mechanism. Furthermore, we have developed a new meth-
odology for the one-pot lactonization of acyloins by a copper
catalyst. Further investigations to elucidate the detailed
mechanism of this reaction are now in progress.
Figure 2. Tandem acylation-Wittig lactonization of acyloins.
Acknowledgment. We thank Dr. Y. Sunada at Kyushu
University for the X-ray crystal structure analysis. This
research was partially supported by a Grant-in-Aid for
Scientific Research (B) (19390007, 22390002) and the
Program for Promotion of Basic and Applied Research for
acyclic acyloins were transformed into the desired lactones in
high yields. The R-hydroxyl lactone (1h), the sterically con-
(5) (a) Uemura, S. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon Press: Oxford, 1991; Vol. 7, pp 757-786. (b)
Witt, D. Synthesis 2008, 16, 2491–2509.
(6) For X-ray crystal structure analysis of 6, see Supporting Informa-
tion.
(8) Ohno, S.; Tomita-Yokotani, K.; Kosemura, S.; Node, M.; Suzuki,
T.; Amano, M.; Yasui, K.; Goto, T.; Yamamura, S.; Hasegawa, K.
Phytochemistry 2001, 56, 577.
(7) Toluene is safer than THF under oxidative conditions. Furthermore,
the catalyst turnover number in toluene was fairly better than in THF.
5348
Org. Lett., Vol. 12, No. 22, 2010