Rapid syntheses of difluorinated dihydropyrans
Jonathan M. Percy* and Stéphane Pintat
School of Chemistry, University of Birmingham, Edgbaston, Birmingham, UK B15 2TT.
E-mail: jmpercy@chemistry.bham.ac.uk
Received (in Liverpool, UK) 4th January 2000, Accepted 15th February 2000
A very short reaction sequence opens with metal-mediated
addition of commercial bromodifluoropropene to aldehydes;
allylation under phase transfer catalysed conditions sets the
stage for a ring closing metathesis (RCM) in the presence of
commercial Grubbs’ catalyst to afford potentially useful
difluorinated dihydropyrans.
Difluorinated carbohydrate analogues are normally prepared by
DAST [(diethylamino)sulfur trifluoride] difluorination of
highly protected precursors in which a ketonic carbonyl group
has been isolated, chemistry which is highly vulnerable to
stereoelectronic effects as demonstrated many times by Cas-
tillo´n and coworkers.1 Even when such difficulties have been
overcome, the hazardous nature of the DAST reagent precludes
scale-up, while safer versions of the reagent sacrifice reactivity
in the cause of thermal stability.2 Building block chemistry then
appears most attractive and the Reformatsky reaction of ethyl
bromodifluoroacetate with aldehydes and other electrophiles is
a well-worked and productive theme.3 Our own approaches
have relied on metallated difluoroenol derivatives fashioned in
situ from trifluoroethanol but we were struck by the simplicity
of a possible RCM-based approach.4 Scheme 1 shows the
analysis; the route closely follows that published recently by
Carda et al.5 but forms the first instance where RCM technology
has been used for the synthesis of ring-fluorinated hetero-
cycles.
Scheme 2 Reagents and conditions: i, H2CNCHCF2Br, In, DMF, sonication,
room temp.; ii, H2CNCHCH2Br, NaOH, CH2Cl2, Bu4NHSO4, room temp.;
iii, 5 mol% Grubbs’ catalyst, CH2Cl2, rt, 24 h.
conditions described by Wilson and Guazzaroni (saturated
NH4Cl, THF).8 The zinc-mediated reaction9 was much more
effective in the case of benzaldehyde, affording the homoallyl
alcohol 2b in 84% yield, but less effective (37%) when the more
interesting and electrophilic benzyloxyacetaldehyde 1d was
present, despite our observation of an extremely clean 19F NMR
spectrum from the crude material. Pinacol coupling of the
aldehyde may be responsible for the low yield though we did not
isolate any products of this type.
Allylation10 was trouble-free as was the RCM reaction11 at
room temperature in dichloromethane, though reaction times of
24 h were required to consume the starting material entirely.
Initially, we were interested in the reproducibility of the
allylindium chemistry described by Momose and coworkers6 in
which bromodifluoropropene adds very smoothly to aldehydes.
With the homoallyl alcohols in hand, we would perform an
allylation and then expose the allyl homoallyl ether to an RCM
catalyst. Both alkenes are made less electron rich by electron-
withdrawing substitutents at the allyl position and we were
concerned that the overall cycle would be slow and that metallo-
[2 + 2] cycloaddition in the fluorinated allyl system could result
in b-fluoride elimination with concomitant formation of a
strong metal–fluorine bond destroying the catalyst.
In the event, our fears were groundless. The allylindium
chemistry was highly reproducible and homoallyl alcohols 2a–
d were synthesised in moderate to high (57–80%) yield
(Scheme 2).
The reactions worked best when the suspensions were shaken
or sonicated, and purification was facile.7 We also explored the
replacement of the indium metal with 325-mesh zinc under the
The methallyl ether 5 prepared by an analogous sequence
failed to cyclise at all under these conditions suggesting that the
additional steric hindrance imposed by a methyl group is
enough to prevent coordination and stop the cycle. Dihy-
dropyran 4d, formed in a particularly high isolated and purified
yield is of the type identified in the analysis, while 4c is
potentially interesting for the synthesis of higher sugars and
related species; we are continuing to explore the scope of the
reaction more fully. This short study clearly shows how
potentially important oxygen heterocycles can be assembled in
an extremely concise manner using off-the-shelf reagents.
We wish to thank the EPSRC (GR/K84882) and the
University of Birmingham for financial support and Dr Philip
Walker (Fluorochem Ltd.) for a generous gift of 1-bromo-
1,1-difluoropropene.
Notes and references
1 M. I. Barrena, M. I. Matheu and S. Castillón, J. Org. Chem., 1998, 63,
2184; R. Fernández and S. Castillón, Tetrahedron, 1999, 55, 8497.
2 G. S. Lal, G. P. Pez, R. J. Pesaresi, F. M. Prozonic and H. Cheng, J. Org.
Chem., 1999, 64, 7048.
3 For a full discussion of two-carbon fluorinated building blocks, see:
J. M. Percy, Top. Curr. Chem., 1997, 193, 131.
4 R. H. Grubbs and S. Chang, Tetrahedron, 1998, 54, 4413. For a recent
and most attractive approach to cyclitols, see: L. Hyldtoft, C. S. Poulsen
and R. Madsen, Chem. Commun., 1999, 2101.
Scheme 1
DOI: 10.1039/b000006j
Chem. Commun., 2000, 607–608
This journal is © The Royal Society of Chemistry 2000
607
Percy, Jonathan M.
