1
348
M. E. B. Smith et al. / Tetrahedron Letters 42 (2001) 1347–1350
Scheme 1. (i) SOCl , MeOH, 0°C, 2 h, quantitative; (ii) Boc O, TEA, DCM, 0°C to rt, 2 h, 97%; (iii) DBU, DCM, rt, 3 days,
2
2
®
9
2%; (iv) Oxone , Na EDTA, Bu NHSO , Me CO, KH PO (aq., pH 7), 5 h, 38%; (v) DBU, DCM, 16 h, quantitative; (vi)
Oxone , H O, pH 6, 7 h, 33%; (vii) Boc O, DMAP, DCM, 0°C to rt, 5 h, 67%; (viii) NaOMe, MeOH, 0°C then rt, 16 h,
2 4 4 2 2 4
®
2
2
quantitative.
this white solid was suspended in DCM and treated
The absolute stereochemistry of 10a, 10b, 11a, 11b, 12a
14
firstly with TEA then Boc O to give N-Boc ester 4 as a
and 12b is assumed to be related to that of lactam 3.
2
The relative stereochemistry of products 10a and 10b
was assigned via the synthesis of pure methyl ester 10b
for which the assignment is unambiguous (Scheme 3).
The relative stereochemistry of 11a and 11b was deter-
mined via NOE experiments performed on the related
O-acetylated derivatives. The relative stereochemistry
of 12a and 12b was determined via previously reported
physical data for 12b and NOE experiments per-
formed on the related O-acetylated derivatives.
Diastereomeric ratios were determined using gas chro-
matography using a Chirasil Dex CB column.
waxy, white solid in 97% yield. Again without purifica-
tion, 4 was treated with 1 equivalent of DBU in DCM
to generate crude alkene 5. Recrystallisation from ethyl
acetate and heptane yielded pure alkene 5 in 92% yield.
®
Alternatively, treatment of N-Boc ester 4 with Oxone
gave the crude cis-epoxide 6 which was crystallised to
purity from tert-butyl methyl ether (MTBE) to give the
desired product, albeit in only 38% yield. Facile rear-
rangement of epoxide 6 with DBU in DCM gave the
desired allylic alcohol 7 as a white solid in quantitative
15
®
yield. Treatment of lactam 3 with Oxone gave crude
trans-epoxide 8 as a 10:1 mixture of diastereomers. The
mixture was crystallised to purity from MTBE and
Hydrogenation of substrate 5 to product 10a was
achieved in high yield and excellent selectivity (92%
d.e.) using [(R,R)MeBPE]Rh (A). The catalysts [(R,R)-
MeDuPHOS]Rh (C) and [(S,S)MeDuPHOS]Rh (D)
were also seen to give reasonable selectivities. However,
selectivity using catalyst [(S,S)MeBPE]Rh (B), which is
the opposite enantiomer of A, was poor presumably
due to a mismatching effect. Interestingly, both Crab-
tree’s catalyst (F) and [1,4-dppb]Rh (G) gave 10b as the
predominant product highlighting that carbamate-
directed reaction had not occurred in these cases,
although the latter catalyst had been successfully
employed in the carbamate-directed hydrogenation of
treated with Boc O in the presence of DMAP to give
2
epoxide 8 in 33% yield (overall yield for two steps). A
methanolic solution of 8 was treated with catalytic
sodium methoxide. The initial formation of epoxide
methyl ester was exothermic and required cooling. The
subsequent rearrangement to allylic alcohol 9 was
slower requiring the reaction temperature to be elevated
to room temperature to give complete reaction. Allylic
alcohol 9 was isolated as a white solid in quantitative
yield. Both allylic alcohols 7 and 9 could be recrys-
tallised from a range of solvents but in our experience
purification at this stage was not necessary.
1
acyclic substrates.
For substrate 7, where both carbamate and hydroxyl
groups can potentially direct hydrogenation to the same
alkene face, conversion to 11a was achieved in excel-
lent yield and selectivity by catalysts A, B and C. Cata-
lyst C gave an exceptionally selective transformation
yielding crude product of 98% d.e. A directed hydro-
genation was also obtained with catalyst E but with
poor selectivity. Catalyst G gave a predominance of
Substrates 5, 7 and 9 were then subjected to a hydro-
genation screen using various chiral and achiral hydro-
genation catalysts as illustrated in Scheme 2 with the
results displayed in Table 1. All reactions were per-
formed according to the example experimental proce-
1
3
dure provided,
and were seen to proceed to
completion with crude products isolated in quantitative
yields unless specified otherwise. All diastereomeric
excesses quoted were measured on crude reaction
products.
1
1b and catalyst F gave no reaction at all. We found
that the synthesis of 11b was best achieved by using
palladium on carbon as catalyst, giving the desired