Rapid syntheses of difluorinated polyols using a cleavable carbamate
Andrew S. Balnaves,a Michael J. Palmerb and Jonathan M. Percy*a
a School of Chemistry, University of Birmingham, Edgbaston, Birmingham, UK B15 2TT.
E-mail: jmpercy@chemistry.bham.ac.uk
b Pfizer Central Research, Sandwich, UK CT13 9NJ
Received (in Cambridge, UK) 6th September 1999, Accepted 28th September 1999
Trifluoroethyl
N-[2-(tert-butyldiphenylsilyloxy)ethyl]-N-
aldehyde electrophile. Stereoselectivities were low as expected
but the syn and anti diastereoisomers could be separated by
careful column chromatography.†
isopropylcarbamate undergoes smooth dehydrofluorina-
tion–metallation followed by BF3·OEt2 mediated addition to
aldehydes to afford a range of allylic alcohols; aldol reaction
with a second aldehyde, then reduction, affords products
which can be deprotected to afford difluorinated polyols.
To make further progress in the direction of polyol targets,
reduction of 4c was attempted under the stereoselective
conditions described by Kuroboshi and Ishihara,9 but very
sluggish reactions ensued (Scheme 3). We had hoped that the
reductions would proceed smoothly, and that the b-hydroxy
group would provide control over the stereochemical course of
the reduction, overriding any asymmetric induction exerted by
the adjacent (a) stereogenic centre. Under the syn-selective
conditions (DIBAL-H, ZnCl2, TMEDA), both the yield of 5a
and 5b (31% combined) and stereoselectivity (1.8:1 syn+anti)‡
were low from syn-4c and the conversion was poor (37%
recovered 4c after 6 h), a consequence presumably of the bulk
of the carbamoyloxy group. The anti-selective Meerwein–
Pondorf–Verley reduction [Al(OPri)3, PhH] conditions were
more successful affording a higher (66% combined) yield of 5a
and 5b. However, whereas the published procedure achieves the
reduction smoothly overnight at room temperature, we failed to
observe any reaction until the mixture had stirred for 4 days. We
then turned our attention to the less stereochemically complex
4b; both stereoselective reductions failed completely, so to take
compound through to the diol stage, we performed a simple
NaBH4 reduction and isolated a 2+1 mixture of syn- and anti-
diols 6a and 6b in 78% yield (Scheme 4).§ Attempts to cleave
the carbamate protection then ensued. We modified the
published conditions and found that dilute HF in aqueous
MeCN effected a satisfactory desilylation, allowing triols 7a
and 7b to be isolated in 84% yield. Exposure to KOH (1 equiv.)
in MeOH at room temperature allowed the isolation of 8a and
8b in a disappointing 20% yield. However, removing the MeOH
in vacuo then taking the residue up in EtOAc followed by an
extractive work up yielded separable triols 8a and 8b (78%),
We have been developing metallated difluoro-enol carbamates1
derived from trifluoroethanol as versatile building blocks2 for
the synthesis of selectively fluorinated polyfunctional mole-
cules. To date, we have concentrated on N,N-diethylcarbamates
which, while synthetically versatile3,4 and inexpensive to
synthesise, are chemically robust. The latter property represents
an apparent impediment to carbamate cleavage and hydroxy
group unmasking so recently, we made use of the cleavable
carbamate described by Derwing and Hoppe,5 synthesising 1 by
a standard route (Scheme 1). Dehydrofluorination–metallation
to 2a proceeded smoothly under our published conditions6 and
allylic alcohols 3a–c were synthesised in moderate to good
yield using BF3·OEt2 to attenuate the transacylation reaction.7
Acceptable yields of 3d and 3e (46 and 71% respectively) could
only be secured by transmetallating 2a to 2b using MgBr·OEt2
prepared freshly according to the method of Harwood et al.,8
and raising the reaction temperature to 230 °C before the
addition of the Lewis acid and electrophile. Next, we explored
the aldol chemistry and confirmed that our published conditions
transferred smoothly; adducts 4a–e were duly prepared
(Scheme 2) in moderate to good yields by treatment of the
allylic alcohols with BuLi at low temperature and allowing the
enolate solution to warm to 210 °C before the addition of the
Scheme 3 Reagents and conditions: i, DIBAL-H, ZnCl2•TMEDA, THF,
–78 °C; ii, Al(OPri)3, PhH, room temp., 4 days.
Scheme 1 Reagents and conditions: i, ButPh2SiCl (1.1 equiv.), DMAP (0.5
equiv.), CH2Cl2, reflux, 12 h; ii, diphosgene (0.5 equiv.), Et3N (1.5 equiv.),
CH2Cl2, 0 °C, then room temp., 16 h; iii, CF3CH2ONa, THF, reflux, 16 h;
iv, LDA, THF (2.0 equiv.), –78 °C; v, R1CHO; vi, BF3•OEt2, then warm to
0 °C; vii, NH4Cl (aq.); viii, MgBr2•OEt2.
Scheme 4 Reagents and conditions: i, NaBH4, MeOH, 0 °C; ii, 48% HF
(aq.)–MeCN (1+9), room temp.; iii, KOH (1.0 equiv.), MeOH, room
temp.
Scheme 2 Reagents and conditions: i, BuLi, THF, –78 °C; ii, warm to –10
°C; iii, R2CHO; iv, NH4Cl (aq.).
Chem. Commun., 1999, 2183–2184
This journal is © The Royal Society of Chemistry 1999
2183
Balnaves, Andrew S.
