1646
H. Dhimane et al. / Tetrahedron Letters 43 (2002) 1645–1648
the same experiment was conducted in the presence of
an excess of the diene, 5a was isolated as the major
product (68%), along with 6 (7%) and 1 (4%). These
results suggested an isomerisation of the kinetic
diastereoisomer 5b into the thermodynamic one 5a
either by successive retro-Mukaiyama/Mukaiyama
reactions (via the silylated enol of 6) or retro Diels–
Figure 1.
Alder/Diels–Alder reactions. The formation of
1
stemmed either from retro-Mukaiayama/retro-Michael
or from retro-Diels–Alder processes. Moreover, in
order to ascertain the relative stereochemistry of the
thermodynamic isomer 5a, the latter was transformed
in two steps into the corresponding crystalline benzoate
7a (Scheme 2), the configuration of which was secured
by X-ray analysis.12
(Cu(OTf)3),9 bismuth (BiCl3 and Bi(OTf)3)10 and ytter-
bium (Yb(OTf)3).11 Under these conditions, unactivated
dienes led to similar failures to those reported above. In
contrast, trimethylsilyloxybutadienes afforded, in most
cases, the corresponding cycloadducts (Table 1).
With the exception of BiCl3 which was totally ineffi-
cient (entry 3), reaction of 1 in the presence of these
Lewis acids with 1-trimethylsilyloxy-1,3-butadiene
afforded the expected cycloadducts 5a and/or 5b, along
with the monocyclic enal 6 (Scheme 2). The diastereose-
lectivity of 5 was highly dependant on the nature of the
Lewis acid employed (entries 1–5). The best chemical
yield in cycloadduct (5a and 5b) was obtained with
Yb(OTf)3, as the catalyst in refluxing CH2Cl2, for 24 h
(entry 5). Noteworthy in the latter case was the obser-
vation that a prolonged reaction time above 48 h
resulted in the obtention of only one diastereomer 5a
(entry 6).
2-Trimethylsilyloxy-1,3-butadiene was then reacted
with dienophile 1 in the presence of various Lewis
acids, among which only Sc(OTf)3, Cu(OTf)2 and
Yb(OTf)3 efficiently catalysed the expected cycloaddi-
tion (entries 7, 8 and 11). In all cases, azadecalone 8
was isolated (after acidic treatment) as a single
stereomer. Once again, the cis-configuration of this
compound was assigned based on X-ray data12 (Scheme
3).
As far as Danishefsy’s diene was concerned, the only
interesting results were obtained with BiCl3 and
Yb(OTf)3. In the former case (entry 14), three com-
pounds were identified in the crude mixture: desilylated
cycloadduct 2, enone 9 (resulting from 2 after elimina-
tion of methanol), along with a trace of phenol 10
(Scheme 3). Compound 2 was isolated as a mixture of
two diastereomers, in a ratio different from the one
obtained above. Attempts to isolate enone 9 by silica
gel chromatography mainly resulted in the obtention of
phenol 10, through retro-Michael and subsequent retro-
Formation of enal 6 is believed to result from the
protodesilylation of the Mukaiyama’s retroaldolisation
product of cycloadducts 5. Indeed, when submitted to
an overnight reflux in CH2Cl2 in the presence of
Yb(OTf)3, a purified diastereomeric mixture of 5a and
5b (30:70) mainly yielded enal 6 (38%) along with a
trace of 5a (7%) and substrate 1 (10%). Moreover, when
Table 1.
Entry
Diene
Lewis acid (equiv.)
Temp. (time)
Products (yield)a
1
2
3
4
5
6
Sc(OTf)3 (0.05)
Cu(TOf)3 (0.1)
BiCl3 (0.1)
Bi(OTf)3 (0.1)
Y(OTf)3 (0.05)
Y(OTf)3 (0.05)
−78 to 0°C (5 h)
rt (4 h)
rt
rt (5 h)
Reflux (24 h)
Reflux (48 h)
5a+5b (44%) [86:14]b 6 (34%)
5a+5b (35%) [57:43]b 6 (7%)
–
c
5a (46%) 6 (32%)
5a+5b (78%) [55:45]b 6 (7%)
5a (64%) 6 (16%)
7
8
9
10
11
Sc(OTf)3 (0.05)
Cu(TOf)3 (0.1)
BiCl3 (0.1)
Bi(OTf)3 (0.1)
Y(OTf)3 (0.05)
−78°C to rt (5 h)
rt (24 h)
rt
rt
8 (56%)
8 (56%)
c
–
–
c
Reflux (20 h)
8 (64%)
c
12
13
14
15
16
Sc(OTf)3 (0.05)
Cu(TOf)3 (0.1)
BiCl3 (0.1)
Bi(OTf)3 (0.1)
Y(OTf)3 (0.05)
rt
–
–
d
0°C (4 h)
rt (12 h)e
rt
2 (26%) 8 (50%) 10(15%)
c
–
rt (15 min)e
8 (56%) 10 (18%)
a Isolated yields.
b Diastereomeric ratio as estimated by 1H NMR and GC.
c No reaction was observed.
d Unreacted starting material was recovered. Only trace of 4-hydroxyacetophenone resulting from the dimerisation of the diene was isolated.
e Products isolated after submission of the crude reaction mixture to catalytic hydrogenation.