The use of methyl 2,2-difluoro-2-(fluorosulfonyl)acetate as the difluorocarbene source to generate an in situ source of difluoromethylene triphenylphosphonium ylide

Article abstract of DOI:10.1016/j.jfluchem.2013.02.026

Under moderate conditions in the presence of a demethylating reagent, such as iodide, methyl 2,2,-difluoro-2-(fluorosulfonyl)acetate (MDFA) releases difluorocarbene, which in the presence of triphenylphosphine forms difluoromethylene triphenylphosphonium ylide. When the process is carried out also in the presence of aldehydes or activated ketones, the ensuing in situ Wittig-type reaction of the ylide with the carbonyl reactants produces 1,1-difluoroalkenes in good yield. Density Functional Theory calculations were used to provide new estimates of the energies and structures of singlet and triplet states of CH₂:, CHF:, and CF₂: carbenes, as well as those of their respective triphenylphosphonium ylides.

Full text of DOI:10.1016/j.jfluchem.2013.02.026

Journal of Fluorine Chemistry 150 (2013) 53–59
Contents lists available at SciVerse ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
The use of methyl 2,2-difluoro-2-(fluorosulfonyl)acetate as the
difluorocarbene source to generate an in situ source of
difluoromethylene triphenylphosphonium ylide
*
Charles S. Thomoson, Henry Martinez, William R. Dolbier Jr.
Department of Chemistry, University of Florida, Gainesville, FL 32611-7200, United States
A R T I C L E I N F O
A B S T R A C T
Article history:
Under moderate conditions in the presence of a demethylating reagent, such as iodide, methyl 2,2,-
difluoro-2-(fluorosulfonyl)acetate (MDFA) releases difluorocarbene, which in the presence of
triphenylphosphine forms difluoromethylene triphenylphosphonium ylide. When the process is carried
out also in the presence of aldehydes or activated ketones, the ensuing in situ Wittig-type reaction of the
ylide with the carbonyl reactants produces 1,1-difluoroalkenes in good yield. Density Functional Theory
calculations were used to provide new estimates of the energies and structures of singlet and triplet
states of CH2:, CHF:, and CF2: carbenes, as well as those of their respective triphenylphosphonium ylides.
ß 2013 Elsevier B.V. All rights reserved.
Received 16 January 2013
Received in revised form 23 February 2013
Accepted 28 February 2013
Available online 13 March 2013
Keywords:
Difluorocarbene
Diphenylmethylene triphenylphosphonium
ylide
1,1-Difluoroalkenes
DFT calculations
1. Introduction
Fuqua and co-workers were the first to propose the intermedi-
acy of difluoromethylene triphenylphosphonium ylide and to
The introduction of fluorine atoms into organic compounds
alters their physical, chemical, and biological properties in such a
way that it often enhances the molecule’s biological activity.
Hence, there continues to be great interest in new methods that
allow the incorporation of fluorine into compounds of potential
pharmaceutical and agrochemical importance.
1,1-Difluoroalkenes have themselves generated interest as
potential enzyme inhibitors [1–4], but more commonly they have
been utilized as fluorinated building blocks. They can be reduced to
monofluoroalkenes [5], or hydrogenated to difluoromethyl groups
[6]. They can undergo addition of fluoride in the presence of a
proton source to give 2,2,2-trifluoroethyl groups [7,8], add
nucleophiles [9], and they are precursors of ring fluorinated
heterocycles via intramolecular cyclization reactions [10–12]. 1,1-
Difluoroalkenes are also precursors of esters and carboxylic acids
[13]. The versatility of their chemistry has led to various
approaches for their preparation, many of them involving a Wittig
reaction involving the presumed, but thus far undetected
difluoromethylene triphenylphosphonium ylide.
study the reaction of this presumed intermediate with aldehydes
to form 1,1-difluoroalkenes [14,15]. In 1964, they reported a
process involving a refluxing solution of sodium chlorodifluor-
oacetate in diglyme at 160 8C in the presence of Ph3P and aldehyde
substrate. Good yields were obtained, as shown in Scheme 1 below.
Aliphatic aldehydes were also decent substrates with heptanal
yielding 52% of 1,1-difluoro-1-octene.
Although the mechanism proposed by Fuqua involved initial
formation of difluorocarbene by thermal decomposition of
chlorodifluoroacetate, followed by trapping of the carbene by
triphenylphosphine to form the ylide, Burton has suggested that
alternative mechanisms, not involving the intermediacy of CF2:,
are more likely [16]. Their suggestion was based upon the fact that
the thermal decomposition of the sodium chlorodifluoroacetate in
diglyme was much faster in the presence of Ph3P, thus requiring
involvement of Ph3P in the rate determining step of the
mechanism, as well as on the fact that no cyclopropanation was
observed to occur when 2,3-dimethyl-2-butene was added to the
reaction mixture.
Herkes and Burton were successful in using essentially the
same process with activated ketone substrates, such as trifluor-
oacetophenone (68% yield) [16]. Burton subsequently introduced
another approach for preparing this ylide intermediate, via
reaction of CF2Br2 with triphenylphosphine at room temperature
* Corresponding author. Tel.: +1 352 392 5266.
E-mail address: wrd@chem.ufl.edu (W.R. Dolbier Jr.).
0022-1139/$ – see front matter ß 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jfluchem.2013.02.026

54
C.S. Thomoson et al. / Journal of Fluorine Chemistry 150 (2013) 53–59
Scheme 1. Fuqua procedure.
Scheme 2. Burton procedure.
(Scheme 2), which gave higher yields in reactions with activated
ketones, but lower with aldehydes [17,18]. Again, this process did
not require the intermediacy of difluorocarbene. In subsequent
papers, they found that using (Me2N)3P instead of Ph3P led to good
yields in reactions with non-activated ketones [19,20]. Another
example of a difluorocarbene/Ph3P procedure, using (CF3)2Hg as
the CF2: source, proved effective with non-activated ketones [21].
1,1-Difluoroalkenes can also be synthesized from aldehydes
and ketones via Horner–Wadsworth–Emmons/Horner–Wittig
[22,23], and Julia [24] or Julia–Kocienski protocols [25], which
require the preparation of fluorinated phosphonate, sulfinate or
sulfone precursors, the use of strong bases and low temperatures.
The latter reaction is exceptional in that it provides excellent yields
from non-activated ketone substrates.
the presence of aldehydes or activated ketones, this ylide reacts to
form 1,1-difluoroalkenes in good yields that are comparable and at
times better than those obtained when using either the Fuqua or
the Burton methodologies.
In terms of its use as a reagent, MDFA has previously been
primarily known as an excellent precursor of trifluoromethyl
copper (Scheme 3) [26]. Then recently we reported that MDFA
could be used as an effective source of difluorocarbene to
synthesize difluorocyclopropanes from alkenes (Scheme 3). In
the latter reaction, difluorocarbene formation was instigated by
demethylation of MDFA by iodide ion, with trimethylsilyl chloride
(TMSCl) being used to trap fluoride ion [27].
Initially it was hoped that Ph3P might be used both to
demethylate MDFA and combine with the generated CF2: to form
the ylide. Thus initial experiments involved simply adding Ph3P to
MDFA in THF at 100 8C in the presence of benzaldehyde (Scheme
2. Results and discussion
4). Although about 5% of desired
b,b-difluorostyrene was formed,
the major product of this reaction, quite unexpectedly, was
difluorotriphenylphosphorane [28,29] characterized unambigu-
In this paper, we report results where methyl 2,2-difluoro-2-
(fluorosulfonyl)acetate (MDFA) can be conveniently used as the
source of difluorocarbene, which then combines with Ph3P to form
difluoromethylene triphenylphosphonium ylide. When created in
ously by its 19F NMR signal, a doublet at
d
À41.1 (1JPF = 664 Hz).
This result indicated that the SN2 nucleophilic demethylation
Scheme 3. Some previous work with MDFA.

C.S. Thomoson et al. / Journal of Fluorine Chemistry 150 (2013) 53–59
55
Scheme 4. Initial results using Ph3P as both nucleophile and ylide component.
Table 1
Optimization of reaction.
Entry MeO2CCF2SO2F
equiv.
Solvent
Nucleophile Temp (8C) Yield (%)
1
2
3
4
5
6
7
8
9
2
THF
None
TBPB
KI
100
100
100
100
100
80
5
40
34
26
0
2
THF
2
THF
2
MeCN
TBPB
2
1,4 Dioxane TBPB
2
THF
TBPB
KI
50
64
26
56
2
MeCN
Toluene
MeCN
70
2
KI
70
2.5
KI
70
10
1.75
MeCN
KI
70
75
Fig. 1. Products from reaction with nitrobenzene.
reaction of Ph3P was not effectively competing with the alternative
mechanism that led to Ph3PF2 (most probably initiated by Ph3P
attack on the carbonyl oxygen of MDFA) [30].
amounts of the above-mentioned difluoromethyl phosphonium
co-product always accompanied formation of the desired 1,1-
difluoroalkenes.
This result led us to examine iodide and bromide ions as
alternative demethylating nucleophiles, the mechanism of which
we hoped would compete kinetically with the non-productive
reaction instigated by Ph3P alone. In examining various sources of
these nucleophiles, tetrabutyl phosphonium bromide (TBPB) and
potassium iodide (KI) were found to give the best results, and they
were thus chosen to be used in experiments designed to optimize
conditions of the reaction. Varying temperature and solvent, the
best yields were obtained using KI in acetonitrile at 70 8C (Entry 7,
Table 1). Finally, yields were improved when the amount of MDFA
was decreased to 1.75 equiv. (Entry 10, Table 1) (Scheme 5).
In the reaction with 4-nitrobenzaldehyde, desired product (2o)
was obtained in low yield. We considered that this might be due to
poor solubility of the starting material in CH3CN. An attempt to
increase the yield by changing solvents resulted in the formation of
4-nitro-(2,2,2-trifluoroethyl)benzene (3) (Fig. 1). Compound 3
presumably derives from nucleophilic attack by fluoride ion at the
terminal CF2 group of initially formed product 2o. Such reaction
has precedent and indeed has been demonstrated to be a
reasonable method for synthesizing 2,2,2-trifluoroethyl aromatics
from
b,b-difluorostyrenes [7,8].
The identities of known compounds 2a–2c, 2e–2h, 2j, 2l–2n,
and 3 were initially determined on the basis of their characteristic
fluorine NMR spectra, but were also confirmed by examination of
their proton spectra. New compounds 2d, 2i, and 2k were fully
characterized on the basis of their 1H, 13C, and 19F NMR spectra, and
by their exact masses as determined by HRMS.
In each case, formation of the desired
b,b-difluorostyrene was
accompanied by formation of variable amounts of what was
presumably difluoromethyl triphenylphosphonium iodide, which
was detected in the crude reaction mixture by 19F NMR. A doublet
3
2
of doublets at
ethyl acetate/acetonitrile mixture was consistent with the
reported 19F spectrum for Ph3P+CF2H BrÀ (dd,
d
À127.5 with JPF = 75.9 Hz and JFH = 45.5 Hz in
d
À126.2,
2.1. Mechanism and calculations
3JPF = 79.7 Hz, 2JFH = 46.6 Hz) in CDCl3 [31]. No attempt was made
to isolate this co-product. Presumably it was formed from the ylide
by protonation.
Although the involvement of difluoromethylene triphenylpho-
sphonium ylide, Ph3P55CF2, as the key intermediate in the Wittig-
type reactions reported by Fuqua, Burton and many others, as
discussed above, has never been questioned, the nature and
stability of this intermediate, in particular the strength of binding
between the Ph3P and CF2: entities has definitely been an issue
addressed in a number of papers. Our specific results appear to be
unambiguously derived from generation of difluorocarbene,
followed by its combination with Ph3P to form the ylide, Ph3P55CF2,
with subsequent Wittig reaction of this ylide with aldehydes and
ketones by the usual mechanism to form 1,1-difluoroalkenes.
Once best conditions were determined, a variety of aldehydes
and activated ketones were examined to determine the scope of
the reaction. The results of these reactions are found in Table 2.
These results indicate that MDFA gives the desired product in good
yield for substituted benzaldehydes, activated ketones, and
electron-rich heterocyclic aldehydes. Results were less satisfactory
for aliphatic aldehydes or highly electron deficient benzaldehydes,
or the highly electron deficient 2- or 3-pyridinecarboxaldehyde. 4-
(Dimethylamino)-benzaldehyde also failed as a substrate. Small
Scheme 5. Initial results when using iodide as nucleophile.

56
C.S. Thomoson et al. / Journal of Fluorine Chemistry 150 (2013) 53–59
Table 2
Yields of 1,1-difluoroalkenes from the respective aldehydes.
aYields determined by 19F NMR, using trifluorotolune as the internal standard.
bIsolated yields.
cUsing 2 equiv. of MDFA.
dUsing 2 equiv. of MDFA at 80 8C.
Therefore, we considered that it would be worthwhile to carry out
a computational reexamination of the binding and structure of this
ylide.
Difluorocarbene has, itself, been subject of several computa-
tional studies [32–34], with the large energy difference between its
singlet (S) and triplet (T) states (with the singlet being the ground
state) creating particular interest in this carbene. When compared
with CH2 carbene, each replacement of hydrogen with fluorine has
been shown to contribute significantly to the stabilization of the
singlet versus the triplet state.
These three carbenes can be considered to be ‘‘starting
materials’’, which when combined with triphenylphosphine,
form their respective triphenylphosphonium ylides. All three
of these ylides have been used as Wittig reagents, but in contrast
to methylene triphenylphosphonium ylide [35], neither the
CHF or the CF2 ylides are sufficiently stable to be isolated
[16,36].
In 1986, using ab initio molecular orbital theory (SCF level) [37],
Dixon and Smart examined the interaction between the CF2 (S) and
PH3. At this level of theory, it was found that the bond length
or discuss the discrepancy between their results and those of Dixon
and Smart.
Because of the apparent low interaction between the CF2(S) and
PH3 reported by Dixon and Smart, we considered that the use of
density functionals, and more important those that include
medium range attractive interactions such as M06-2X [39–41],
might provide greater insight in calculating the structures and
energies of these types of molecules.
Ground state calculations of the singlet and triplet state for the
CH2, CHF and CF2 carbenes were carried out and the results
compared with available experimental data. Both structures and
S–T energy gaps (Table 3) were found to be in good agreement with
available experimental and previous computational results [32–
34].
As expected, all singlet carbenes have greater bond lengths and
smaller X–C–Y angles than the triplet carbenes, consistent with the
presumed sp2 and sp hybridization respectively. Although energy
gaps have been calculated more precisely with different density
functionals and basis set [34], our calculated energy gaps between
the triplet and singlet carbenes using the M06-2X function are
close to the earlier reported values. Our structural calculations are
also consistent with the experimental data reported for CH2
[42,43].
With our calculated ground state structures and energies of the
singlet carbenes in excellent agreement with both previous theory
and experiment, the ground state structures of the CH2–PPh3, CHF–
PPh3 and CF2–PPh3 ylides were then calculated. These structures
are depicted in Fig. 2, with the data being provided in Table 4.
In comparing the structures of all three ylides, one can see that
the methylene carbon of CH2–PPh3 has the greatest sp2 character,
˚
between the CF2 carbon and the phosphorous was 3.54 A, which is
essentially two separate species with little interaction. In addition
the binding energy between the CF2 and the PH3 was calculated to
be only 1.2 kcal/mol, which was consistent with Burton’s evidence
for facile dissociation of the CF255PPh3 ylide [36]. However,
calculations by Allen and co-workers in 1988 at the HF/3-21G*
level reported that the C–P bond length of the same molecule was
˚
1.635 A [38], which could be interpreted as a double bond
considering that their calculations for the C–P bond length of
˚
CH2–PH3 was 1.646 A. Unfortunately, the authors did not mention

C.S. Thomoson et al. / Journal of Fluorine Chemistry 150 (2013) 53–59
57
Table 3
Geometrical data and energy gaps for various methyl carbenes at the M06-2X/6-311+G(2df,2p) level.a,b
˚
˚
C–H (A)
C–F (A)
X–C–Y (8)
DE (T–S) (kcal/mol)
CH2(S)
CH2(T)
1.106 (1.107)
1.076 (1.075)
–
–
102.0 (102.4)
134.3 (133.9)
À11.26 (À9.37)
9.59 (8.0–14.60)
54.25 (56.60)
CHF(S)
CHF(T)
1.118 (1.138)
1.083 (1.088)
1.299 (1.305)
1.307 (1.304)
102.3 (103.5)
121.9 (121.2)
CF2(S)
CF2(T)
–
–
1.293 (1.304)
1.307 (1.298)
104.5 (104.8)
118.9 (118.1)
a
In parenthesis are experimental values or difference-dedicated configuration interaction results, available from [34] and references therein.
X = H,F and Y = H,F.
b
Fig. 2. Depictions of the calculated structures of (a) CH25PPh3, (b) CHF5PPh3, and (c) CF25PPh3.
Table 4
Geometrical data for various phosphonium ylides at the M06-2X/6-311+G(2df,2p) level.a
˚
Distance (A)
Angle (8)
P–C
C–H
C–F
P–Ph(2, 3)
P–Ph1
X–C–Y
P–C–H
P–C–F
CH2–PPh3
CHF–PPh3
CF2–PPh3
1.674 (1.662)
1.707
1.082 (0.932)
–
1.821 (1.816)
1.814
1.847 (1.839)
1.842
116.0 (121.9)
111.8
116.2 (118.8)
–
1.085
–
1.391
1.378
115.6
–
115.3
109.2
1.815
1.808
1.82
105.1
a
Experimental values in parenthesis when available [35].
with this carbon taking on increasing sp3 character as each H is
replaced by F. These changes are reflected in the longer C–P bonds
and the greater pyramidalization as one proceeds from the CH2 to
the CHF to the CF2 ylide. Although our calculated bond length for
In order to quantify the stability of the CF2–PPh3 ylide, the
enthalpy and the free energy of the reaction between the singlet
carbene and triphenylphosphine to produce the ylide was
calculated (Scheme 6). Both gas phase and solvent were considered
and the data is shown in Table 5.
˚
the C–P bond in CF255PPh3 (1.815 A) is much shorter than the
˚
3.54 A calculated by Dixon and Smart for CF2–PH3, it is significantly
In the gas phase, formation of both the CH2 and the CHF ylide
are highly exothermic and exergonic, whereas for the CF2 ylide the
reaction is exothermic, but endergonic. However, since the dipole
moment of the CF2–PPh3 ylide is much larger than those of the CH2
and CHF ylides, this suggests that the CF2 ylide might be more
˚
longer than the 1.635 A calculated by Allen. The X–C–Y angle is
observed to decrease as the number of fluorine increases, a trend
consistent with the change from sp2 to sp3 hybridization at the
methylene carbon.
Scheme 6. Calculated reaction for formation of ylides from carbene and Ph3P.

58
C.S. Thomoson et al. / Journal of Fluorine Chemistry 150 (2013) 53–59
Table 5
135 mmol, 3 equiv.), potassium iodide (15.0 g, 90 mmol,
2 equiv.), and 4-bromobenzaldehyde (9.80 g, 45 mmol, 1 equiv.)
were added and let stir for 30 min. Methyl 2,2-difluoro-2-
(fluorosulfonyl)acetate (MDFA) (19.7 g, 78.75 mmol,1.75 equiv.)
was then added slowly over a period 1 h. The resulting mixture
Relative 298 K energies for CH2, CHF and CF2 triphenylphosphonium ylides.a
CH2–PPh3
CHF–PPh3
CF2–PPh3
CF2–PPh3-TS
D
D
D
D
H (gas)
À79.02
À67.15
À79.12
À69.15
2.96
À43.57
À30.85
À46.55
À35.73
4.42
À6.80
5.42
À3.21
8.38
G (gas)
H (solv)b
G (solv)b
À14.42
À4.10
5.78
À4.53
5.17
was stirred for three hours,
a nitrogen atmosphere being
maintained until the end of the reaction. Then the reaction
was quenched with water and extracted with ethyl acetate.
Trifluorotoluene (15 mmol) was added as internal standard and
the yield of the crude reaction was measured by 19F NMR (72%).
The ethyl acetate was then removed by rotary evaporation to
produce a black, semi-solid slurry. Product was isolated from the
slurry and separated from residual triphenylphosphine by
extraction 5 times with hexane (100 mL). The combined hexane
extracts were combined and concentrated. Additional impurities
were removed via column chromatography using a 95:5 mixture
of hexanes:methylene chloride to obtain pure product: (6.9 g
Dipolec
a
Energies in kcal/mol.
b
c
Solv: acetonitrile.
Dipole in Debye.
greatly favored than the others in polar solvents. When acetonitrile
was included as solvent in the calculation, formation of the CH2
and CHF ylides remained highly exothermic and exergonic, but the
values were not much different from those for the gas phase. On
the other hand, in the polar solvent, formation of the CF2 ylide
proved to be not only exothermic, but also exergonic to the extent
of À4.1 kcal/mol. These results suggest that formation of the CF2–
PPh3 ylide would be more favored in polar solvents, as the results
in Table 1 indicate, and it should thus be possible for difluor-
omethylene triphenylphosphonium ylide to participate in Wittig
reactions.
(70%), clear liquid); 1H NMR
J = 8.1 Hz, 2H), 5.25 (dd, J = 25 Hz, J = 3.1 Hz, 1H); 19F NMR
d 7.48 (d, J = 9.1 Hz, 2H), 7.21 (d,
d
À81.8 (dd, J = 30 Hz, J = 27 Hz, 1F), À83.7 (dd, J = 29 Hz,
J = 3.8 Hz, 1F). The above data is in accord with those reported
in the literature [25].
Since all but three of the products, 2a–2o were known
compounds, and their fluorine and vinylic proton NMR signals
were very characteristic for the 2,2-difluorovinyl group, most of
the reactions were run at much smaller scale (6 mmol) than the
example above with the reported yields for these small scale
reactions being obtained simply by fluorine NMR, using
trifluoromethylbenzene as internal standard. All new com-
pounds, 2d, i, and k, were of course isolated and fully
characterized.
In order to have a better understanding of the stability and
behavior of the CF2–PPh3 ylide, the transition state for the C–P
bond formation/dissociation was calculated (Scheme
6 and
Table 5), and the reaction was seen to have a relatively flat
potential energy surface at 298 K, with a calculated activation free
energy in acetonitrile of 5.17 kcal/mol, relative to the singlet CF2
carbene and PPh3 at infinite separation. Thus the dissociation of the
ylide to its carbene and phosphine components would only have to
overcome a 9.27 kcal/mol barrier. This is consistent with the kind
of reversibility of ylide formation that was observed by Burton
[36].
4.2.1. (2,2-Difluoroethenyl)benzene (2a)
1H NMR 5.40 (dd, 3JFH = 26.6 and 4.1 Hz, 1H), 7.55 (m, 5H); 19
d F
NMR
[15,44,45].
d
À82.9 (dd, J = 34 Hz, 27 Hz, 1F), À84.8 (d, J = 32 Hz, 1F)
3. Conclusions
In conclusion, it has been found that under moderate
conditions, in the presence of a demethylating reagent such as
iodide, methyl 2,2,-difluoro-2-(fluorosulfonyl)acetate releases
difluorocarbene, which in the presence of triphenylphosphine
forms difluoromethylene triphenylphosphonium ylide. When the
process is carried out in the presence of aldehydes or activated
ketones, an in situ Wittig-type reaction of the ylide with the
carbonyl reactants produces 1,1-difluoroalkenes in good yield.
Calculations indicated that the ylide intermediate was only weakly
bound, but was sufficiently stable to propose its involvement in the
kind of usual Wittig reaction expected of phosphonium ylides.
4.2.2. 4-Methyl-(2,2-difluoroethenyl)benzene (2b)
1H NMR
d
5.25 (3JFH = 26 and 4.2 Hz, 1 H), 7.16 (d, 3JHH = 8.4 Hz,
2H), 7.24 (d, JHH = 8.4 Hz, 2H); 19F NMR
28 Hz, 1F), À86.9 (d, J = 38 Hz, 1F) [44–46].
d
À84.9 (dd, J = 38 Hz,
3
4.2.3. 4-Methoxy-(2,2-difluoroethenyl)benzene (2c)
1H NMR 3.83 (s, 3H), 5.24 (dd, 3JFH = 26 and 4.5 Hz, 1H), 6.91
(d, 3JHH = 8.6 Hz, 2H), 7.29 (d, 3JHH = 8.6 Hz, 2H); 19F NMR
d
d
À87.3
(dd, J = 42 Hz, J = 27 Hz, 1F), À89.2 (d, J = 41 Hz, 1F) [44–46].
4.2.4. 4-Thiomethyl-(2,2-difluoroethenyl)benzene (2d)
1H NMR
d
2.48 (s, 3H), 5.23 (dd, 3JFH = 26 and 3.6 Hz, 1H), 7.23
4. Experimental
(m, 4H); 19F NMR
d
À85.3 (dd, J = 35 Hz, J = 26 Hz, 1F), À87.2 (d,
J = 35 Hz, 1F); GC–MS (exact mass, DART-TOF-MS): calcd C9H9F2S
(M)+ 187.0393, found 187.0399.
4.1. General experimental procedures
NMR spectra were obtained in CDCl3 using TMS and CFCl3 as the
4.2.5. 2-Bromo-(2,2-difluoroethenyl)benzene (2e)
3
internal standards for 1H/13
respectively.
C
NMR and 19F NMR spectra,
1H NMR
d
5.24 (dd, JFH = 26 Hz and 3.9 Hz, 1H), 7.04 (m, 2H),
7.29 (m, 2H); 19F NMR
d
À84.3 (dd, J = 26 Hz, J = 3 Hz, 1F), À85.3
(dd, J = 28 Hz, J = 2 Hz, 1F) [47].
4.2. Typical procedure for preparation of 1,1-difluoroalkenes
4.2.6. 4-Fluoro-(2,2-difluoroethenyl)benzene (2f)
3
A 250 mL, three necked round bottomed flask was equipped
with a stir bar. The vessel was then fitted with a condenser
topped by a T-tube with slow flow of N2, and then sealed with
septa. The vessel was flamed dried and then allowed to cool to
room temperature. Under the inert, nitrogen atmosphere,
acetonitrile (75 mL) was added to the vessel, and the tempera-
ture was increased to 70 8C. Then triphenylphosphine (34.5 g,
1H NMR
d
5.24 (dd, JFH = 26 Hz and 3.6 Hz, 1H), 7.02 (m, 2H),
7.25 (m, 2H); 19F NMR
d
À85.1 (dd, J = 37 Hz, J = 28 Hz, 1F), À87.2
(dt, J = 34 Hz, J = 3.7 Hz, 1F) [15,45].
4.2.7. 4-Trifluoromethyl-(2,2-difluoroethenyl)benzene (2g)
19F NMR
d
À77.9 (m,1F), À76.8 (qq, J = 10 Hz, J = 24 Hz,
J = 33 Hz, J = 47 Hz, 1F), À59.7 (dd, J = 10 Hz, J = 24 Hz, 3F) [48].

C.S. Thomoson et al. / Journal of Fluorine Chemistry 150 (2013) 53–59
59
References
4.2.8. 4-Benzyloxy-(2,2-difluoroethenyl)benzene (2i)
1H NMR
d
5.21 (dd, 3JFH = 26 and 3.6 Hz, 1H), 6.07 (s, 3H), 6.95
[1] I.A. McDonald, J.M. Lacoste, P. Bey, M.G. Palfreyman, M. Zreika, J. Med. Chem. 28
(1985) 186–193.
[2] W.R. Moore, G.L. Schatzman, E.T. Jarvi, R.S. Gross, J.R. McCarthy, J. Am. Chem. Soc.
114 (1992) 360–361.
[3] P.M. Weintraub, A.K. Holland, C.A. Gates, W.R. Moore, R.J. Resvick, P. Bey, N.P. Peet,
Bioorg. Med. Chem. 11 (2003) 427–431.
(m, 2H), 7.33–7.44 (m, 5H); 19F NMR
d
À87.1 (dd, JFF = 43 Hz,
2
2
3JHF = 26 Hz, 1F), À88.8 (d, JFF = 41 Hz, 1F); GC–MS (exact mass,
DART-TOF-MS): calcd C15H13F2O (M)+ 247.0934, found 247.0933.
4.2.9. 2,3,4,5,6-Pentafluoro-(2,2-difluoroethenyl)benzene (2j)
[4] J.-M. Altenburger, G.Y. Lassalle, M. Matrougui, D. Galtier, J.-C. Jetha, Z. Bocskei, C.N.
Berry, C. Lunven, J. Lorrain, J.-P. Herault, P. Schaeffer, S.E. O’Connor, J.-M. Herbert,
Bioorg. Med. Chem. 12 (2004) 1713–1730.
[5] S.-I. Hayashi, T. Nakai, N. Ishikawa, D.J. Burton, D.G. Naae, H.S. Kesling, Chem. Lett.
(1979) 983–986.
19F NMR
d
À77.3 (m, 1F), À81.1 (s, 1F), À141.3 (t, J) 18 Hz, 2F),
À157.4 (t, J = 21 Hz, 1F), À164.6 (m, 2F) [8].
[6] W.B. Motherwell, M.J. Tozer, B.C. Ross, J. Chem. Soc.: Chem. Commun. (1989)
1437–1439.
4.2.10. 2-(2,2-Difluoroethenyl) thiophene (2k)
1H NMR,
d 7.25 (t, J = 1 Hz, 1H), 7.24 (d, J = 1 Hz, 1H), 7.0 (d,
[7] L. Zhu, Y. Li, Y. Zhao, J. Hu, Tetrahedron Lett. 51 (2010) 6150–6152.
[8] B.V. Nguyen, D.J. Burton, J. Org. Chem. 62 (1997) 7758–7764.
[9] T. Yamazaki, H. Ueki, T. Kitazume, Chem. Commun. (2002) 2670–2671.
[10] J. Ichikawa, T. Wada, T. Okauchi, T. Minami, Chem. Commun. (1997) 1537–1538.
[11] M. Yokota, D. Fujita, J. Ichikawa, Org. Lett. 9 (2007) 4639–4642.
[12] J. Ichikawa, M. Yokota, T. Kudo, S. Umezaki, Angew. Chem. Int. Ed. 47 (2008) 4870–
4873.
[13] S.-I. Hayashi, T. Nakai, N. Ishikawa, Chem. Lett. (1980) 651–654.
[14] S.A. Fuqua, W.G. Duncan, R.M. Silverstein, Tetrahedron Lett. (1964) 1461–1463.
[15] S.A. Fuqua, W.G. Duncan, R.M. Silverstein, J. Org. Chem. 30 (1965) 1027–1029.
[16] F.E. Herkes, D.J. Burton, J. Org. Chem. 32 (1967) 1311–1318.
[17] D.G. Naae, D.J. Burton, J. Fluorine Chem. 1 (1971/1972) 123–126.
[18] D.J. Burton, Z.-Y. Yang, W. Qiu, Chem. Rev. 96 (1996) 1641–1716.
[19] D.G. Naae, D.J. Burton, Syn. Commun. 3 (1973) 197–198.
[20] P.J. Serafinowski, C.L. Barnes, Tetrahedron 52 (1996) 7929–7933.
[21] I. Nowak, M.J. Robins, Org. Lett. 7 (2005) 721–724.
3
3
J = 3 Hz, 1H), 5.55 (dd, JHF = 26 Hz, JHF = 2.1 Hz, 1H); 19F NMR,
d
2
3
2
À80.7 (dd, JFF = 28 Hz, JHF = 26 Hz, 1F), À87.9 (dd, JFF = 28 Hz,
3JHF = 1.7 Hz, 1F); 13C NMR:
d
156.0 (dd, J = 299 Hz and 291 Hz),
132.0 (br t, J = 6.7 Hz), 127.2 (s), 125.9 (dd, J = 6.8 Hz and 5.0 Hz),
124.9 (dd, J = 5.8 Hz and 3.6 Hz), 77.6 (dd, J = 33 Hz and 17 Hz);
GC–MS (exact mass, DART-TOF-MS): calcd C6H5F2S (M+H)+,
147.0080, found 147.0082.
4.2.11. 2-(2,2-Difluoroethenyl)furan (2l)
3
2
19F NMR:
d
À85.1 (dd, JHF = 32 Hz, JFF = 29 Hz, 1F), À86.7 (d,
2JFF = 29 Hz, 1F) [15].
[22] M. Obayashi, E. Ito, K. Matsui, K. Kondo, Tetrahedron Lett. 23 (1982) 2323–2326.
[23] M.L. Edwards, D.M. Stemerick, E.T. Jarvi, D.P. Mathews, J.R. McCarthy, Tetrahedron
Lett. 31 (1990) 5571–5574.
4.2.12. 1,1-Difluoro-1-heptene (2m)
19F NMR:
d
À85.1 (dd, J = 35 Hz, J = 28 Hz, 1F), À86.8 (d,
[24] J.S. Sabol, J.R. McCarthy, Tetrahedron Lett. 33 (1992) 3101–3104.
[25] Y. Zhao, W. Huang, L. Zhu, J. Hu, Org. Lett. 12 (2010) 1444–1447.
[26] X.-S. Fei, W.S. Tian, Q.-Y. Chen, J. Chem. Soc. Perkin Trans. 1 (1998) 1139–1142.
[27] S. Eusterwiemann, H. Martinez, W.R. Dolbier Jr., J. Org. Chem. 77(2012) 5461–5464.
[28] T. Mahmood, J.M. Shreeve, Inorg. Chem. 24 (1985) 1395–1398.
[29] K.M. Doxsee, E.M. Hanawalt, T.J.R. Weakley, Inorg. Chem. 31 (1992) 4420–4421.
[30] F. Ramirez, C.P. Smith, S. Meyerson, Tetrahedron Lett. (1966) 3651–3656.
[31] M.J. VanHamme, D.J.Burton,P.E.GreenlimbIII, Org.Magn.Reson.11(1978)275–280.
[32] C.W. Bauschlicher Jr., H.F. Schaefer III, P.S. Bagus, J. Am. Chem. Soc. 99 (1977)
7106–7110.
J = 32 Hz, 1F) [44].
4.2.13. 1,1-Difluoro-1-octene (2n)
19F NMR:
J = 27 Hz, 1F) [44].
d
À91.8 (d, J = 52 Hz, 1F), À94.5 (dd, J = 28 Hz,
4.2.14. 4-Nitro-(2,2-difluoroethenyl)benzene (2o)
19F NMR:
J = 18 Hz, 1F) [15,45].
d
À80.1 (dd, J = 26 Hz, J = 19 Hz 1F), À81.4 (d,
[33] E.A. Carter, W.A. Goddard III, J. Phys. Chem. 91 (1987) 4651–4652.
[34] D. Das, S.L. Whittenburg, J. Mol. Struct. (Theochem) 492 (1999) 175–186.
[35] J.C.J. Bart, J. Chem. Soc. B (1969) 350–365.
[36] D.J. Burton, D.G. Naae, R.M. Flynn, B.E. Smart, D.R. Brittelli, J. Org. Chem. 48 (1983)
3616–3618.
[37] D.A. Dixon, B.E. Smart, J. Am. Chem. Soc. 108 (1986) 7172–7177.
[38] M.M. Francl, R.C. Pellow, L.C. Allen, J. Am. Chem. Soc. 110 (1988) 3723–3728.
[39] Y. zhao, D.G. Truhlar, J. Chem. Phys. 124 (2006) 224105.
[40] Y. Zhao, D.G. Truhlar, Org. Lett. 9 (2007) 1967–1970.
4.2.15. 4-Nitro-(2,2,2-trifluoroethyl)benzene (3)
19F NMR:
d
À66.5 (t, J = 11 Hz, 3F) [15,49].
4.3. Computational method
[41] Y. Zhao, D.G. Truhlar, Acc. Chem. Res. 41 (2008) 157–167.
[42] E. Wasserman, V.J. Kuck, R.S. Hutton, W.A. Yager, J. Am. Chem. Soc. 92 (1970)
7491–7493.
[43] R.A. Bernheim, H.W. Bernard, P.S. Wang, L.S. Wood, P.S. Skell, J. Chem. Phys. 54
(1971) 3223–3224.
All quantum chemical calculations were performed using
Gaussian 09 Rev. A.02 [50] at the M06-2X/6-311+G(2df,2p) level
of theory [39–41], including frequency calculation to identify
the structures as either ground state or transition state. All
calculated transition states presented one and only one negative
frequency.
[44] D.R. Strobach, J. Org. Chem. 36 (1971) 1438–1440.
[45] W.F. Reynolds, V.G. Gibb, N. Plavac, Can. J. Chem. 58 (1980) 839–845.
[46] V.G. Nenajdenko, G.V. Varseev, V.N. Korotchenko, A.V. Shastin, E.S. Balenkova, J.
Fluorine Chem. 124 (2003) 115–118.
The influence of acetonitrile as a solvent was evaluated using
the SMD continuum solvation model [51]. Free energies in
solution were calculated by summing the gas phase thermal
contributions with the single point SMD/M06-2X energies.
Correction from the gas to the solution phase was made by
adding À1.9 kcal/mol (RT Ln([Sln]/[Gas])) to the free energy of
each molecule.
[47] P.S. Bhadury, P. Mechir, M. Sharma, S.K. Raza, D.K. Jaiswal, J. Fluorine Chem. 116
(2002) 75–80.
[48] P.S. Bhadury, P.P. Bhagwat, P. Mechir, D.K. Jaiswal, J. Fluorine Chem. 85 (1997)
115–116.
[49] M.M. Kremlev, W. Tyrra, A.I. Mushta, D. Naumann, Y.L. Yagupolskii, J. Fluorine
Chem. 131 (2010) 212–216.
[50] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, G.
Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H.P.
Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada, M. Ehara, K.
Toyota, R. Fukuda,J. Hasegawa, M. Ishida, T. Nakajima,Y. Honda, O. Kitao, H. Nakai, T.
Vreven, J. Montgomery, J.A.J.E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers,
K.N. Kudin, V.N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox,
J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev,
A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G.
Zakrzewski, G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, O.
Farkas, J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09, Revision A. 02,
Gaussian, Inc., Wallingford, CT, 2010.
Acknowledgments
HM wishes to acknowledge the University of Florida High
Performance Computer Center for the time that was provided to
allow our calculations.
[51] A.V. Marenich, C.J. Cramer, D.G. Truhlar, J. Phys. Chem. B 113 (2009) 6378–6396.