C. Xu et al. / Tetrahedron Letters 49 (2008) 2454–2456
2455
oxide (Hammett constant H = +26.0)12 and has the lowest
O
O
solubility among the alkaline earth oxides (MgO, BaO,
CaO and SrO) and so can potentially be reused, as it will
not be lost due to leaching into the reaction mixture. Fur-
thermore, MgO is inexpensive and easily obtained so that it
could be applied to large scale manufacture. In view of our
success in using MgO as a basic catalyst in both Knoevena-
gel and especially Michael additions, it appeared that this
catalyst might be useful for triggering the ‘double’ Michael
addition of 1,3-dithianes to propargylic carbonyl systems 1
to give b-keto-1,3-dithianes 2. In this Letter, we report the
successful outcome of using this inexpensive and easily
obtained heterogeneous basic catalyst for this purpose
(Scheme 1).
The catalysts used in this study were obtained by calci-
nation of the precursors (450 °C, static air, 2 h) with a heat-
ing ramp of 10 °C/min (Table 1), except MgO-AD and
MgO-RIB. MgO-AD was the commercially available
MgO from Aldrich. MgO-RIB was obtained by igniting
Mg ribbon in air.
Initial trial reactions to test the various catalysts
involved the addition of the particular MgO sample
(2.5 equiv or 1.0 equiv) in one portion to a stirred solution
of 4-phenyl-3-butyn-2-one 1a (1 mmol) and propane-1,3-
dithiol (1 mmol) in THF (2 ml) cooled in ice-water. The
cooling bath was then removed and the mixture stirred at
ambient temperature and reaction progress followed by
withdrawing samples (about 0.5 ml), which were filtered
and the filtrate divided into two parts. One part was
directly analyzed by GC or GC–MS. The other part was
evaporated and the residue analyzed by 1H NMR and
13C NMR using CDCl3 as solvent. The conversions
observed are summarized in Table 1.
A series of substrates were then tested using MgO-CBC
as the catalyst, as this showed one of the highest surface
areas, clearly a key factor in the activity of this class of cat-
alyst. After the reaction, the solution was filtered through
silica gel to remove the catalyst and the solvent evaporated.
No further purification was usually necessary. The results
(Scheme 2) show that MgO-CBC is a very active catalyst
for the synthesis of b-keto-1,3-dithianes 2. Substrates
1a–g underwent clean additions with propane-1,3-dithiol
by this procedure to give the corresponding b-keto-1,3-
HS
SH
S
S
R1
R1
R2
MgO-CBC
THF, 0~20 ºC
R2
2
1
a) R1 = Me; R2 = Ph [1 h; 98%]
f) R1 = EtO; R2 = H [4 h; 95%]
g) R1 = MeO; R2 = H [2 h; 88%]
h) R1 = EtO; R2 = Me [24 h; 65%]
b) R1 = R2 = Ph [1 h; 90%]
c) R1 = H; R2 =
n-pent [3 h; 98%]
d) R1 = Me; R2 = SiMe3 [1 h; 95%] i) R1 = MeO; R2 = Ph [24 h; 30%]
e) R1 = Me; R2 = H; [1 h; 90%]
Scheme 2. The addition of 1,3-propanedithiol to ynones and ynoates 1
catalyzed by magnesium oxide.
dithianes 2. The experimental procedure is very simple
and the reactions are fast and clean.
Ley and co-workers proposed a mechanism6 for these
double conjugate additions of dithiols to propargylic car-
bonyl systems to generate b-keto-1,3-dithianes 2 when
using sodium methoxide as the trigger, which features
sequential deprotonation at sulfur, Michael addition and
enolate protonation to give both cis- and trans-isomers of
the a,b-unsaturated carbonyl intermediates 3. These are
themselves substrates for a second conjugate addition,
now an intramolecular process, which gives the desired b-
keto-1,3-dithianes 2. However, this second addition com-
petes with the reaction of substrate 3 with another molecule
of ynone to give dimeric side products 4, which are
observed in some cases (Scheme 3).
All five products 2a–e obtained from ynones and ynals
1a–e were essentially free of contamination by either inter-
mediates 3 or ‘dimers’ 4 and only traces of these were vis-
ible in the 1H NMR spectra of products 2f and 2g, derived
from propynoates 1f,g. This is clearly consistent with these
mechanistic proposals, as the propynoates would be
expected to be marginally poorer Michael acceptors rela-
tive to the ynones. The two final substrates, ethyl 2-butyno-
ate 1h and its phenyl homologue 1i, were understandably
significantly less reactive, the alkyne substituent doubtless
hindering the Michael addition steps significantly relative
to propynoates 1f,g (Scheme 2). This, coupled with the gen-
erally lower ester reactivity resulted in the requirement for
much longer reaction times and the inevitable accumula-
tion of by-products; both ‘dimer’ 4 and partly reacted
Table 1
O
S
1,3-Propanedithiol additions to 4-phenyl-3-butyn-2-one 1a in the presence
of various MgO catalysts (1.0 equiv) for 1 h in tetrahydrofuran
HS
SH
1
R1
R2 SH
MgO-CBC
THF, 0~20 ºC
Entry
Precursors
SBET (m2 gÀ1
)
Yield (%)
3
MgO-RIB
MgO-AD
MgO-CAD
MgO-COH
Mg ribbon
10
25
25
65
277
20
0
MgO (Aldrich)
MgO (Aldrich)
Mg(OH)2 (Fluka)
66
98
98
98
98
98
O
S
R2
O
MgO-ROH Mg(OH)2 (Fluka)a
+
2
R1
R2
S
R1
MgO-CBC
MgO-OX
MgO-CC
a
(MgCO3)4Mg(OH)2 (Fluka) 288
Mg[O(CO)2O]
MgCO3 (Acros Organics)
312
229
4
Scheme 3. Double Michael addition prior to final ring closure.
Calcined at 600 °C for 2 h, refluxed in H2O for 3 h, then oven-dried.