Scheme 1. Synthesis of 2 via Pathway aa
Scheme 2. Synthesis of ent-2 via Pathway ba
a Reagents and conditions: (i) (a) Et2BOTf (2.0 equiv), (iPr)2NEt
(2.2 equiv), CH2Cl2, -78 °C, (b) OHCCH2CHdCH2; (ii) RuCl2-
(CHPh)(PCy3)2, CH2Cl2; (iii) (a) Hg(OCOCF3)2, CH3CN, (b) satd
aq NaCl; (iv) Bu3SnH, AIBN; (v) LiOH, H2O2.
a Reagents and conditions: (i) (a) Et2BOTf (2.0 equiv), (iPr)2NEt
(2.2 equiv), CH2Cl2, -78 °C, (b) OHCCH2CH2CHdCHPh; (ii)
RuCl2(CHPh)(PCy3)2, CH2Cl2; (iii) (a) Hg(OCOCF3)2, CH3CN, (b)
satd aq NaCl; (iv) Bu3SnH, AIBN; (v) LiOH, H2O2.
Treatment of 11 with Hg(II)(OCOCF3)2 in acetonitrile14
gave 12 as a single regioisomer in quantitative yield. Reduc-
tive demercuration and hydrolysis then gave enantiomerically
pure ent-2.
With pathway b established as the preferred route, it was
next employed in the synthesis of 17, one of the enantiomers
of the “endo” isomer of 1 (Scheme 3). To prepare 17, the
this work, it was not clear what levels of regioselectivity
could be obtained in the proposed intramolecular oxymer-
curations involved in either of the pathways outlined in
Figure 1.7 It appeared plausible that there may be some
preference for closure onto C4 of cycloheptene 5 as this
would minimize any steric interaction between the nucleo-
philic alcohol and the adjacent acyl sultam. Thus, 5 was
prepared in good yield employing a sequence of syn-aldol8
and ring-closing metathesis (RCM) reactions.
Scheme 3. Synthesis of 17 via Pathway ba
Intramolecular oxymercuration of 5,9,10 with either
Hg(II)(OAc)2 or Hg(II)(OCOCF3)2, gave predominantly the
product from closure onto C4 (a 4:1 ratio of, respectively, 6
and 7, Scheme 1). Chloromercurial 6 could be recrystallized
pure in 60% yield (its structure is shown in Figure 2).
a Reagents and conditions: (i) (a) Et2BOTf (2.0 equiv), (iPr)2Net
(1.9 equiv), CH2Cl2, -78 °C, (b) OHCCH2CHdCH2; (ii) RuCl2-
(CHPh)(PCy3)2, CH2Cl2; (iii) (a) Hg(OCOCF3)2, CH2Cl2, (b) satd
aq NaCl; (iv) Bu3SnH, AIBN; (v) LiOH, H2O2.
C1-epimer of 11 (i.e., 15) was required. As shown in Scheme
3, this necessitated employing an anti-aldol addition in the
first step. Although a number of anti-aldol protocols are
Figure 2. X-ray crystal structure of chloromercurial 6.
(6) (a) Fallon, G. D.; Jones, E. D.; Perlmutter, P.; Selajarern, W.
Tetrahedron Lett. 1999, 40, 7435. (b) Perlmutter, P.; Selajarern, W. Aust.
J. Chem. 2000, 53, 349.
(7) There have been very few reports on intramolecular oxymercurations
which produce oxygen-bridged bicyclic products. See: Foehlisch, B.;
Joachimi, R. Chem. Ber. 1987, 120, 1951.
(8) “Xc” in Schemes 1 and 4 is used to represent the (-)-enantiomer of
Oppolzer’s camphor-derived sultam auxiliary. “Yc” in Schemes 2 and 3
represents the (+)-enantiomer. See ref 12 for more details on the preparation
and use of these auxiliaries.
(9) For reviews on electrophilic cyclisations, see: (a) Robin, S.; Rousseau,
G. Tetrahedron 1998, 54, 13681. (b) Rousseau, G.; Homsi, F. Chem. Soc.
ReV. 1997, 26, 453. (c) Harmange, J. C.; Figadere, B. Tetrahedron:
Asymmetry 1993, 4, 1711.
Reductive demercuration11 followed by hydrolysis12 then
provided enantiomerically pure 2.
Although the chemistry outlined in Scheme 1 provides a
practical route to 2, the approach was not completely
regioselective. Hence, pathway b was next explored in a
synthesis of the enantiomer of 2. The synthesis of the next
intramolecular oxymercuration candidate, the new cyclo-
heptene 11 (employing Crimmins’ method for preparing
3-butenal13), proceeded in good overall yield (Scheme 2).
894
Org. Lett., Vol. 6, No. 6, 2004