17
water added. Use of water as a cosolvent was deleterious to the
reaction.
20, presumably due to amide hydrolysis. Treatment of 21 with
methoxide gave a very slow protodesilylation unfortunately
accompanied by epimerisation. The fact that no TBAF was
required in this particular reaction suggests that protodesilyl-
ation is assisted by the β-amido group in much the same way
Mechanistically it is known that hydroxide and alkoxide
11
anions can form pentavalent anions with silicon, even more
14
so if at least one substituent is a phenyl group (Scheme 4).
18
as analogous β-hydroxy examples.
In summary we have demonstrated two complementary
and mild methods for the protodesilylation of C-2 trialkylsilyl
terminal alkenes. The hydroboration–Peterson elimination
protocol should be applicable to a wide range of trialkylsilyl
groups where other functional groups are resistant to hydro-
boration and the one pot t-BuOK–18-C-6–TBAF protocol
serves as a valuable alternative for the phenyldimethylsilyl
congeners. These protocols are useful in our own research, in
providing stereochemically pure amino acid building blocks
from the aza-[2,3]-Wittig sigmatropic rearrangement, and we
hope will also be for others engaged in synthesis.
Scheme 4
Experimental
General details
Collapse to give the neutral silanol 11 would occur via expul-
sion of phenyl anion, probably acquiring a proton from the
solvent as it separates. There is no experimental evidence to
suggest that hydroxide or alkoxide anion can displace the vinyl
1a
Our general experimental details have been reported.
Representative protodesilylations are given for the hydro-
boration–Peterson elimination protocol 3 to 7 and the one pot
t-BuOK–18-C-6–TBAF 15 to 17.
11,15
moiety.
The formation of protodesilylated material 13,
in the absence of TBAF, can conceivably arise from alkoxide
attack on the silanol 11 or unimolecular decomposition from
the silanoate anion 12, the latter mechanism having some pref-
5-Phenylpent-1-ene (7)
16
erence in the literature. We do not invoke the expulsion of
vinyl anion, but as is probable in the phenyl case, it acquires a
proton from the reaction mixture as it leaves. We have no evi-
dence for the formation of silyl ethers derived from alkoxide
substitution of phenyl groups or the formation of silanols 14.
Neat alkene 3 (90 mg, 0.413 mmol) was treated with a solution
of 9-BBN in THF (2.48 mL of a 0.5 M solution, 1.24 mmol, 3.0
equiv.) at rt and then warmed to 70 ЊC until 3 had disappeared
by TLC analysis. After 5 h the reaction was cooled to 0 ЊC,
treated with EtOH (1 mL), NaOH (1 mL of a 2 M aq. solution),
16
The former may not be stable under the reaction conditions.
H O2 (30%, 1 mL) and after 5 min warmed to 50 ЊC for
2
The one pot procedure proved useful in the protodesilylation
of rearrangement products 15 and 16 to give 17 and 18 in 82%
and 71% yield respectively (Scheme 5). The slightly lower yield
3
0 min. The reaction was then diluted with Et O, separated and
2
extracted with Et O. The combined ethereal layers were washed
2
with brine, dried (MgSO ) and concentrated in vacuo to give the
4
crude β-hydroxysilane (169 mg).
The crude product in THF (0.4 mL ϩ 0.4 mL wash) was
added to a suspension of KH (276 mg of a 30% dispersion in
mineral oil, prewashed with dry hexane, 2.07 mmol, 5.0 equiv.)
in THF (0.4 mL) at 0 ЊC, warmed to rt and stirred for 2.5 h. The
reaction mixture was cooled to 0 ЊC, quenched by the addition
of saturated aq. NH Cl (0.5 mL), and extracted with Et O. The
4
2
combined ethereals were washed with brine, dried (MgSO ),
4
filtered and concentrated in vacuo to give a crude oil which was
purified by flash column chromatography (hexane) to give 7
(
85%) as a clear oil which possessed identical analytical data to
5
that prepared in the literature.
(
1S*,2R*)-N-tert-Butoxycarbonyl-2-methyl-1-phenylbut-3-
enylamine (17)
To a stirred solution of 15 (83 mg, 0.21 mmol) in DMSO (2 mL)
was added TBAF (1.05 mL of a 1 M THF solution, 5 equiv.)
dropwise. The resulting mixture was cooled to 0 ЊC and treated
with potassium tert-butoxide (27 mg, 0.24 mmol, 1.1 equiv.),
1
8-crown-6 (10 mg, 0.038 mmol, 0.2 equiv.). After 1 h the
mixture was warmed to rt and stirred for 36 h after which
time MeOH (0.02 mL, 2 equiv.) was added. The mixture was
washed with H O (2 × 10 mL), dried (MgSO ) and concen-
2
4
trated in vacuo. Purification by flash-column chromatography
5% EtOAc–light petroleum) gave 17 (45 mg, 82%) as a white
solid which possessed identical analytical data to the sample
(
1a
prepared previously.
Scheme 5
Acknowledgements
1
7
of 18 could be due to some amide hydrolysis. It is noteworthy
that no epimerisation of 18 was detected by H NMR. Proto-
desilylation of 19, where the TIPS protecting group precludes
the use of the one pot procedure containing TBAF, was
achieved by stirring with t-BuOK and 18-crown-6 in THF for
We would like to thank the EPSRC and Merck Sharp and
Dohme for financial support.
1
Notes and references
3
0 minutes to give desired 20 in 67% isolated yield along with
† Stirring with TBAF alone for 5 days at rt gave only recovered starting
material (90%).
silanol 21 (21%). Longer reaction times led to lower yields of
3
026 J. Chem. Soc., Perkin Trans. 1, 2000, 3025–3027