mol) and chloroacetaldehyde (50% aq. solution, 12.8 g, 0.231
mol) in pyridine (27 mL) was stirred and warmed at 50 ЊC
under nitrogen for 24 h. The cooled product was diluted with
ether (200 mL), washed with 10% HCl, water and brine. After
drying over Na2SO4 the solution was evaporated in vacuo, to
give an essentially pure ester 4 (8.0 g). ν (film) 1745 cmϪ1; δH
(CDCl3, 300 MHz) 1.46 (3H, s), 1.56 (3H, s), 2.24 (2H, q,
J = 7.2 Hz), 2.9 (2H, t, J = 7.2 Hz), 3.7 (3H, s), 5.05 (1H, t,
J = 7.2 Hz), 6.51 (1H, d, J = 1.5 Hz), 7.12 (1H, d, J = 1.5 Hz); δC
(CDCl3, 75 MHz) 17.2 (q), 25.4 (q), 26.4 (t), 26.5 (t), 50.8 (q),
110.5 (d), 112.8 (s), 122.7 (d), 132.3 (s), 140.0 (d), 162.4(s), 164.0
(s); HRMS (m/z) 208.1099 (208.1095 calcd. for C12H16O3).
The above ester 4 (8.0 g) was refluxed with sodium hydroxide
(4.34 g, 0.109 mol) in water (36 mL) for 2 h, cooled, washed
with ether (2 × 50 mL) and slowly acidified with 10% HCl. The
carboxylic acid 5 (5.48 g, 52%) was obtained by extractive
workup with dichloromethane. Mp 68–69 ЊC; ν (film) 3306–
2350, 1680 cmϪ1; δH 1.45 (3H, s), 1.56 (3H, s), 2.25 (2H, m), 2.93
(2H, t, J = 7.2 Hz), 5.03 (1H, t, J = 7.2 Hz), 6.56 (1H, d, J = 1.5
Hz), 7.12 (1H, d, J = 1.5 Hz), 12.44 (1H, br s); δC 17.4 (q), 25.6
(q), 26.6 (t), 27.9 (t), 111.0 (d), 113.0 (s), 122.8 (d), 133.0 (s),
140.5 (d), 164.0 (s), 170.0 (s); HRMS (m/z) 194.0935 (194.0939
calcd. for C11H14O3).
Decarboxylation of the acid 5 (9.69 g, 0.05 mol) was accom-
plished by heating with copper powder (0.64 g, 0.01 mol) and
quinoline (12.9 g, 0.1 mol) for 2 h. The cooled mixture was
filtered, diluted with dichloromethane and washed with 10%
HCl. Drying, concentration, silica gel chromatography, and
evaporative distillation (75 ЊC/3 Torr) led to the isolation of
3 (6.75 g, 90%). ν (film) 1598 cmϪ1; δH 1.40 (3H, s), 1.50 (3H, s),
2.14 (2H, q, J = 7.5 Hz), 2.45 (2H, t, J = 7.5 Hz), 4.98 (1H, t,
J = 7.5 Hz), 5.75 (1H, s), 6.03 (1H, s), 7.04 (1H, s); δC 17.7 (q),
25.6 (q), 26.6 (t), 28.2 (t), 104.7 (d), 110.0 (d), 123.5 (d), 132.4
(s), 140.7 (d), 156.1 (s).
Scheme 2
to obtain the amide alcohol 11a, we entered a cul-de-sac un-
wittingly because the subsequent reoxidation of the benzylic
alcohol to the aldehyde with MnO2 was attained by a compet-
ing lactonization to return a portion of the material to 10a.
This complication led us to hydrolyze the diester 8 selectively,
and convert the half-ester 12 to the amide ester 13 and thence to
11b. It was estimated that this molecule having larger alkyl
groups on the nitrogen atom, would be less prone to lactoniz-
ation (by virtue of changing the relative orientation of the
OH and C᎐O groups). We were glad to be able to obtain the
᎐
amide aldehyde 14 by Swern oxidation of 11b in 70% yield.
The low reaction temperature definitely contributed to this final
success.
Having in our possession the key intermediate 14, we investi-
gated the intramolecular ene reaction. After several unfruitful
attempts at using Lewis acids to catalyze the cyclization, we
conducted the thermal process in sealed vessels using toluene as
solvent. From a reaction at 200 ЊC in a sealed glass tube we
detected the formation of platyphyllide 1 and its diastereomer
15 in a ratio of 28:72. However, in a Teflon vessel (within a
resealable stainless steel bomb) only 15 was produced. The
difference between the two reactions may be attributed to the
favored transition state A in which the aldehyde and the amide
groups were oriented such as to minimize dipole–dipole inter-
action. On the other hand, the silicon atoms of the glass surface
may have provided a binding site for the oxygen atoms of both
groups, and reaction of molecules in such a bound state B
would lead to the platyphyllide precursor. The products result
from the subsequent lactonization.
We were rather pleased to observe the stereoselective form-
ation of one lactone in the specified reaction conditions, even
though an additional step or steps would be required to convert
15 to the target 1. To that end we tried a Mitsunobu reaction to
invert the configuration of the benzylic carbon. Relactonization
did not take place under standard conditions probably due to
steric hindrance in the formation of the phosphonium inter-
mediate. This forced us to attempt other options such as
mesylation after cleavage of the lactone. On acidification of the
NaOMe treated lactone, in an attempt to isolate the hydroxy
ester intermediate, platyphyllide 1 was detected. This finding led
us to investigate the process more thoroughly, which resulted in
establishing a procedure for the conversion of 15 to 1 in 84%
yield. Because the benzylic cation was readily generated and the
relactonization was under thermodynamic control (with 1,2-
asymmetric induction by the isopropenyl group) the synthesis
was rendered stereoselective.
Dimethyl 1-(4-methylpent-3-enyl)-7-oxabicyclo[2.2.1]hepta-2,5-
diene-2,3-dicarboxylate (6)
Furan 3 (6.75 g, 0.045 mol) was mixed with dimethyl acetylene-
dicarboxylate (6.39 g, 0.045 mol) and dichloromethane (30
mL), and was refluxed for 5 days, cooled, and the solvent
removed. The residue was chromatographed over silica gel
using 1:9 EtOAc–hexane as eluent to give 6 (11.89 g, 90.5%).
ν (film) 1712, 1639 cmϪ1; δH 1.54 (3H, s), 1.65 (3H, s), 2.07–2.15
(4H, m), 3.75 (3H, s), 3.81 (3H, s), 5.09 (1H, t, J = 4 Hz), 5.60
(1H, d, J = 2.1 Hz), 6.95 (1H, d, J = 5.4 Hz), 7.12 (1H, dd,
J = 5.4, 2.1 Hz); δC 17.3 (q), 23.2 (t), 25.5 (q), 28.7 (t), 51.7 (q),
83.0 (d), 97.1 (s), 123.2 (d), 132.0 (s), 144.1 (d), 144.9 (d), 151.1
(s), 155.8 (s), 164.7 (s), 164.7 (s); HRMS (m/z) 292.1302
(292.1305 calcd. for C16H20O5).
Dimethyl 3-(4-methylpent-3-enyl)phthalate (8)
The low-valent titanium reagent was prepared by sequential,
dropwise addition of titanium() chloride (3 mL, 0.0274 mol)
and then triethylamine (0.86 mL, 0.00617 mol) in THF (5 mL)
to an ice-cooled, stirred suspension of lithium aluminium
hydride (0.52 g, 0.0137 mol) in dry tetrahydrofuran (30 mL),
which was then heated at 65 ЊC for 0.5 h, recooled to room
temperature and treated with a solution of 6 (2.0 g, 0.00685
mol) in tetrahydrofuran (10 mL). After 24 h at room temper-
ature the reaction mixture was poured into water (150 mL), and
extracted with dichloromethane (3 × 50 mL). The crude prod-
uct was chromatographed over silica gel (eluent: 1:9 EtOAc–
hexane) to afford the phthalate 8 (2.1 g, 99%). ν (film) 1731,
1594 cmϪ1; δH 1.43 (3H, s), 1.57 (3H, s), 2.16 (2H, m), 2.51 (2H,
t, J = 9 Hz), 3.76 (3H, s), 3.82 (3H, s), 5.03 (1H, t, J = 4.8 Hz),
7.25–7.28 (2H, m), 7.72 (1H, dd, J = 6.75, 2.1 Hz); δC 17.3 (q),
25.5 (q), 29.7 (t), 33.3 (t), 52.1 (q), 122.9 (d), 127.4 (d), 127.6 (s),
128.6 (d), 132.3 (s), 133.6 (d), 134.9 (s), 139.6 (s), 165.8 (s), 169.2
(s); HRMS (m/z) 276.1361 (276.1356 calcd. for C16H20O4).
Experimental
2-(4-Methylpent-3-enyl)furan (3)
A mixture of methyl 3-oxo-7-methyloct-6-enoate (10.0 g, 0.154
1824
J. Chem. Soc., Perkin Trans. 1, 1999, 1823–1826