Communications
Table 1: Optimization of the palladium-catalyzed alkene–silyl enol ether
cyclization.[a]
Entry
R3Si
T
Conc.
Pd [mol%]
Ratio 8/9/6[b]
1
2
3
4
5
6
TMS
TMS
TIPS
TBS
TBS
TBS
RT
0.05m
0.2m
0.3m
0.1m
0.1m
0.05m
10
10
1
10
10
5[c]
40:50:trace
33:17:50
38:trace:60
80:20:trace
80:20:3
458C
458C
RT
458C
458C
>95:trace:trace
[a] Reactions were carried out with 0.5 mmol of the substrate 7 in DMSO
under oxygen (1 atm) for 24–48 h. [b] The product ratio was determined
1
by H NMR spectroscopy. [c] The catalyst was added slowly (over 4 h).
TBS=tert-butyldimethylsilyl, TIPS=triisopropylsilyl, TMS=trimethyl-
silyl.
ester 4[19] (Scheme 4). Thus, the formylation of 4 with isobutyl
formate, followed by the addition of allyl chloroformate, gave
an E/Z mixture of enol carbonates 13, which underwent
isomerization upon flash chromatography to form mostly the
Z isomer, the structure of which was assigned by NMR
spectroscopy, in 89% yield. This enol carbonate underwent a
rapid decarboxylative allylation[22,23] under Pd(OAc)2/PPh3
catalysis with the formation of a quaternary center to give
the 2-oxocyclohex-3-ene carbaldehyde 14 in high yield. In the
next step, the enone as well as the aldehyde functionality were
reduced with NaBH4 and CeCl3 in MeOH to give cyclo-
hexenone 15. Surprisingly, the vinylogous ester functionality
was also reduced under these conditions; the use of expensive
reducing agents, such as DIBAL-H, was thus avoided. The
crude hydroxyketone 15 was then subjected to pivaloylation
to afford pivalate 16 in 94% yield from aldehyde 14. Enone 16
was converted into silyl enol ether 17, which underwent
cyclization under our optimized conditions to give the 7-
Scheme 3. Synthesis and oxidative palladium-catalyzed transformation
of silyl enol ether 7: a) LDA (1.1 equiv), THF, ꢀ808C, 1 h, then allyl
bromide (1.3 equiv), ꢀ808C!RT, 12 h, 90%; b) LDA (1.1 equiv), THF,
ꢀ808C, 1 h, then methyl vinyl ketone (1 equiv), ꢀ808C, 0.5 h, 84%;
c) (CH2OH)2 (20 equiv), PPTS (0.4 equiv), benzene, reflux, 2–12 h;
d) DIBAL-H (1.5 equiv), THF, ꢀ80!ꢀ608C, 1 h; e) p-TsOH·H2O
(0.04 equiv), Et2O/H2O (100:1), 208C, 0.5 h, 80% from 5; f) LDA
(1.5 equiv), THF, ꢀ808C, 1 h, then R3SiCl (2 equiv), HMPA (1.5 equiv),
room temperature, 1.5 h, approximately 90%; g) Pd(OAc)2 (cat.),
DMSO, O2, 85% (see Table 1); h) p-TsOH·H2O (1 equiv), acetone,
208C, 40 min; i) KOtBu, THF, tBuOH (2:1), ꢀ808C, 2.5 h, 88% from
8; ketoenone 10 was used without purification. R3Si=Me3Si, iPr3Si, or
tBuMe2Si; DIBAL-H=diisobutylaluminum hydride, DMSO=dimethyl
sulfoxide, HMPA=hexamethylphosphoramide, LDA=lithium diisopro-
pylamide, PPTS=pyridinium p-toluenesulfonate, Ts=p-toluenesul-
fonyl.
bicyclic product into the decalin derivative 11 and then to
perform the key skeletal rearrangement on a tricyclic system
derived from 11. However, with our model system 6, which
was prepared by the Stork alkylation strategy[18] from 3-
isobutoxycyclohex-2-en-1-one[19] (4) under the conditions
reported by Toyota et al.,[17] the cyclization resulted in the
formation of a mixture of three products (Scheme 3, Table 1).
A change in the trialkylsilyl group of the silyl enol ether 7
from Me3Si to tBuMe2Si led to a decrease in the amount of the
desilylated cyclohexenone 6 formed but not to a decrease in
the formation of cyclohexadienone 9, which results from
oxidative desilylation.[20] Finally, a solution to this problem
was found in the slow addition of the catalyst to the reaction
mixture (Table 1, entry 6). Thus, the addition of the catalyst
solution over 4 h to a 0.05m solution of silyl enol ether 7 (R =
tBuMe2Si) in DMSO resulted in the clean formation of the
desired product 8 in 85% yield. Enone 8 was then converted
into the tricyclic diketone 11 in two steps (Scheme 3).
Unfortunately, the radical rearrangement of tricyclic deriva-
tives of 11 did not give the desired products.[21]
methylenebicyclo[3.2.1]oct-3-en-2-one 18. A subsequent
Scheme 4. Synthesis of the bicyclic ketoester 20: a) NaH, HCO2iBu,
24 h, 08C, then ClCO2allyl, KH (cat.), THF, 08C, 1 h; b) Pd(OAc)2
(1.3 mol%), PPh3, THF, 208C, 1 h, 92% from 4; c) NaBH4,
CeCl3·7H2O, MeOH, 08C, 0.5 h, then p-TsOH, H2O/Et2O, room temper-
ature, 0.5 h; d) PivCl (2 equiv), pyridine (4 equiv), DMAP (0.05 equiv),
CH2Cl2, room temperature, 48 h, 94% from 14; e) LDA (1.5 equiv),
TBSCl (2 equiv), HMPA (1 equiv), THF, ꢀ808C!RT, overnight, 88%;
After the optimization of this cycloalkenylation reaction
we focused on the synthesis of
a functionalized 7-
methylenebicyclo[3.2.1]octan-2-one, such as 18. This
approach would enable the formation of the cyclohexenone
ring of platencin after the rearrangement of the bicyclic
substructure. The racemic synthesis began with the vinylogous
=
f) O2, Pd(OAc)2 (0.058 equiv), DMSO, 85%; g) H2C C(OMe)OTBS
(19; 1.5 equiv), TiCl4 (1.2 equiv), CH2Cl2, ꢀ808C, 12 h, 88%.
DMAP=4-dimethylaminopyridine.
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 3685 –3688