ing the intrinsic barrier of RCF2Br toward SN2 substitutions.7
In principle, a synthetically attractive approach to the
nucleophilically substituted difluoroallene 2 is the copper-
mediated SN2′displacement of bromide from 1sthe only
readily available difluoropropargyl electrophile.8 Although
this type of reaction has been utilized in the synthesis of
allenes,3b its application to gem-difluoro systems is severely
impaired because of the facile loss of fluorine through
R-elimination, which may lead to carbenoid intermediates
and complex mixtures.9 We are now pleased to report a
highly regioselective synthesis of difluoroallene 2 from
difluoropropargyl 1, employing an excess amount of a
Grignard reagent and Cu(I) salt.
This difference in charge density is caused by the strong
electron withdrawing effect of two fluorine atoms on the C1-
carbon. If 2b was synthesized from 1b by using an SN2′
displacement, its CF2 terminus would undergo a competitive
nucleophilic attacksdriven by an energetically favorable
addition-elimination process (â-elimination of fluoride
ion)11syielding complicated mixtures of products (see also
Scheme 2).
Table 1 summarizes the results of our SN2′ optimization
study in the reaction between a Grignard nucleophile
(EtMgBr) and difluoropropargyl bromides 1c or 1d. Loss
of fluorine was observed in the absence of a copper source,
most likely via a bromine-magnesium exchange, leading
to a carbenoid intermediate that undergoes two succesive
R-eliminations of F- (Table 1, entry 1). The combination of
Ab initio calculations of NBO (natural bond order)10
charge densities in propargyl bromide 5 vis-a`-vis its fluori-
nated analogue 1 (Figure 1) demonstrate why SN2′ substitu-
Table 1. Optimization of the Synthesis of Difluoroallene 2
temp,
equiv of
EtMgBr CuX (equiv) (°C, min) (°C, h)
incubationa time
2
(%)b
entry
1
1
2
3
4
5
6
7
8
9
1c
1c
1c
1c
1c
1c
1c
1d
1d
2.2
2.2
2.2
2.2
2.2
2.2
2.2
2.2
1.0
none
-40, 1
-40, 1
-78, 1
-15, 5
+20, 5
-15, 5
-15, 5
-15, 5
-60, 30
-40, 3
-40, 3
-78, 1 NR
-78, 1 81 (70)c
-78, 1 14
CuCl (1.1)
CuCl (4.0)
CuCl (4.0)
CuCl (4.0)
CuBr (4.0)
CuCN (4.0)
CuCl (4.0)
CuBr (2.0)
LiBr (2.0)
CuBr (2.0)
LiBr (2.0)
CuBr (2.0)
LiBr (2.0)
CuBr (2.0)
S(Me)2 (2.0)
CuBr (2.0)
S(Me)2 (2.0)
8
-78, 1 38
-78, 1 NR
-78, 1 51d
-60, 2 52
10
11
12
13
1d
1d
1d
1c
1.5
1.7
1.7
1.7
-60, 30
-60, 30
-60, 30
-60, 30
-60, 2 82
Figure 1. Ab initio calculation of gem-difluoropropargyl bromides
and their nonfluorinated counterparts. Numbers refer to NBO
(natural bond order) charge densities.
-60, 1 89
-60, 1 94 (91)c
-60, 1 80
tion is problematic in the fluorinated model. Electron
densities on the C1- and C3-carbons of 5a and 5b indicate
that the C3-carbon in both species is significantly more
electrophilic than the C1-carbon and therefore prone to
undergo an SN2′ attack. Conversely, the charge densities on
the C1- and C3-carbons in gem-difluopropargyl bromides
1a or 1b reveal that both C1- and C3-carbons are electro-
philic. To complicate matters further, the C1-carbon of
difluoroallene 2b has a positive charge density, whereas its
nonfluorinated counterpart 6 does not (+0.337 vs -0.875).
a Reactions were conducted in 1 mmol scale. A rbf was charged with
the copper salt and THF and cooled to the temperature indicated, after which
the Grignard reagent was introduced and the solution was stirred for an
indicated period of time. Substrate was introduced at the end of the
incubation time. b Yield was determined by 19F NMR, using R,R,R-
trifluorotoluene as an internal standard. c Isolated yield. d Major byproduct
was monofluoro allene 7a:
(7) Chambers, R. D. Fluorine in Organic Chemistry; Blackwell Publish-
ing Ltd./CRC Press: Boca Raton, FL, 2004; p 123.
(8) Xu, B.; Mae, M.; Hong, J. A.; Li, Y.; Hammond, G. B. Synthesis. In
press.
(9) Pohmakotr, M.; Ieawsuwan, W.; Tuchinda, P.; Kongsaeree, P.;
Prabpai, S.; Reutrakul, V. Org. Lett. 2004, 6, 4547-4550.
(10) Computational analysis was carried out with Gaussian 03, Revision
C.02 at the B3LYP/6-311+g(3d) level of theory. Frisch, M. J., et al., see
Supporting Information.
CuCl and EtMgBr yields the desired product 2c in low yield
(8%, Table 1, entry 2). A purple color in the solution of the
Grignard reagent and the copper salt signaled the formation
(11) For synthetic applications that take advantage of fluoride elimination
see: Ichikawa, J.; Wada, Y.; Fujiwara, M.; Sakoda, K. Synthesis 2002,
1917-1936.
480
Org. Lett., Vol. 8, No. 3, 2006