Reactivity and regiochemical behavior in the solvolysis reactions of (2,2-difluorocyclopyl)methyl tosylates

Article abstract of DOI:10.1016/S0022-1139(02)00248-8

The rate constants for acetolysis of (2,2-difluorocyclopropyl)methyl tosylate, and (2,2-difluoro-3-methylcyclopropyl)methyl tosylate at 92°C and of 1-(2,2-difluoropropyl)ethyl tosylate at 42°C are reported and the rectivities and regiochemistries of ring opening of these systems are discussed and compared with expectations based on computational results. The results are discussed in terms of the use of (2,2-difluorocyclopropyl)methyl systems as mechanistic probes.

Full text of DOI:10.1016/S0022-1139(02)00248-8

Journal of Fluorine Chemistry 119 (2003) 39±51
Reactivity and regiochemical behavior in the solvolysis reactions
of (2,2-di¯uorocyclopropyl)methyl tosylates
Merle A. Battiste, Feng Tian, John M. Baker, Olivia Bautista,
Janette Villalobos, William R. Dolbier Jr.*
Department of Chemistry, University of Florida, Gainesville, FL 32611-7200, USA
Received 3 July 2002; received in revised form 9 September 2002; accepted 9 September 2002
Dedicated to Professor Tatlow on the occasion of his 80th birthday.
Abstract
The rate constants for acetolysis of (2,2-di¯uorocyclopropyl)methyl tosylate, and (2,2-di¯uoro-3-methylcyclopropyl)methyl tosylate at
92 8C and of 1-(2,2-di¯uorocyclopropyl)ethyl tosylate at 42 8C are reported and the reactivities and regiochemistries of ring opening of these
systems are discussed and compared with expectations based on computational results. The results are discussed in terms of the use of (2,2-
di¯uorocyclopropyl)methyl systems as mechanistic probes.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: (2,2-Di¯uorocyclopropyl)methyl radical and cation; Solvolysis; Kinetics; DFT calculations
1. Introduction
peci®c cleavage of the proximal bond to form products
ostensibly derived from the 1,1-di¯uorobut-3-en-1-yl cation,
4. This intuitive prediction received support from prelimin-
ary DFT calculations that indicated a regiospeci®c conver-
sion of 3 to a homoallylic, stabilized version of 4, i.e. a
highly unsymmetrical homoallylic cation (4a), with vir-
tually no activation barrier [2].
The (2,2-di¯uorocyclopropyl)methyl radical, 1, under-
goes an extraordinarily fast, regiospeci®c unimolecular ring
opening distal to the geminal ¯uorine substituents to form
the 2,2-di¯uorobut-3-en-1-yl radical, 2 [1]. Its calculated
activation barrier of 1.9 kcal/mol and estimated rate constant
of 1:5 Â 1011
s
À1 at 25 8C quali®es this system as a ``hyper-
sensitive'' probe of reactions involving radicals as inter-
mediates [2]. As demonstrated by the extensive work of
Newcomb and Chestney [3], a radical probe is made much
more valuable if it can, by virtue of differences in regio-
chemistry, distinguish between a radical and a carbocation
intermediate.
In a preliminary communication, we reported experimen-
tal, solvolytic results that con®rmed this prediction for the
parent (2,2-di¯uorocyclopropyl)methyl system [6], but the
question remained regarding the degree to which alkyl
substituents at the third position would modify the regio-
selective proximal cleavage in favor of distal cleavage. To
the extent that distal cleavage prevailed, this would, of
course, diminish the usefulness of the (2,2-di¯uorocyclo-
propyl)methyl system as a probe capable of distinguishing
cationic from radical intermediates. A previous study of
tosylate 5 by Schlosser and coworkers [7] demonstrated that
the presence of two methyls at the third position induces
virtually exclusive distal cleavage. In this paper, the kinetic
and regiochemical impact of a single 3-methyl substituent
Because b-¯uorine substituents destabilize carbocations
[4], whereas a-¯uorines are stabilizing [5], it appeared likely
to us that reactions involving the (2,2-di¯uorocyclopropyl)-
methyl cation, 3, would undergo rearrangement via regios-
* Corresponding author.
E-mail addresses: battiste@chem.ufl.edu (M.A. Battiste),
wrd@chem.ufl.edu (W.R. Dolbier Jr.).
0022-1139/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 0 2 ) 0 0 2 4 8 - 8

40
M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
will be reported, along with additional pertinent experimen-
tal and computational results.
Fig. 2. Calculated structure of (2,2-difluoro-3-methylcyclopropyl)methyl
cation (7) (HF/6-31G(d) level of theory).
located and is 8.4 kcal/mol more stable than 4a. However,
because the bridged bicyclobutonium ion, which is the
intermediate required for the scramble, was not located, it
is unlikely that (C) would be formed from 4a.
2. Results and discussion
2.1. Computational results
Calculations were carried out attempting to locate the
(2,2-di¯uorocyclopropyl)methyl cation on the potential
energy surface at HF/6-31G(d), MP2/6-31G(d) and
B3LYP/6-31G(d) levels of theory. However, in all the cases,
the geometry optimizations starting from the (2,2-di¯uor-
ocyclopropyl)methyl cation (3) structure all resulted in the
optimized homoallylic, 1,1-di¯uorobut-3-en-1-yl cation
(4a). The B3LYP/6-31G(d) optimized geometry of homo-
allylic cation (4a) is depicted in Fig. 1. The planar geometry
The secondary cation (6) with a methyl group at the a-
position of the (2,2-di¯uorocyclopropyl)methyl system
behaved identically to (3) at both the HF/6-31G(d) and
B3LYP/6-32G(d) levels of theory. The structure of the
ring-opened cation has been depicted in Fig. 3.
‡
of the CF2 site of 4a, and its unusually short C±Fbonds
Ê
(1.272 A) are consistent with stabilization of the carbocation
Ê
by its ¯uorine substituents. The 1.340 A C±C bond indicated
a fully characterized double bond. The Mulliken charge of
the CF2 group is ‡0.55, and that of the CH2 at the original
carbinyl position is ‡0.21.
The attempt to locate the 2,2-di¯uorobut-3-en-1-yl cation
(A) also ended up as the 1,1-di¯uorobut-3-en-1-yl cation.
These results indicate that on the potential energy surface,
the (2,2-di¯uorocyclopropyl)methyl cation, at least in the
gas phase, is not a minimum, and also that there are no
saddle points connecting the (2,2-di¯uorocyclopropyl)-
methyl to the 1,1-di¯uorobut-3-en-1-yl cation; thus there
is no activation barrier for this transformation. Another
possible cation, the 3,3-di¯uorocyclobutyl cation (B), was
located, but it is 17.8 kcal/mol higher in energy than homo-
allylic cation 4a. The most stable cation of the formula
When the methyl substituent is at the third position on the
ring (7), preliminary calculations indicated that the regio-
chemistry of the ring opening might well be altered. At HF/
6-31G(d) level of theory, the geometry optimization starting
from the ring closed primary cation ended up with the cation
observed by the lengthening of C±C bond distal to the CF2
Ê
group (Fig. 2). The distance of C2±C3 is 2.03 A. This is not a
‡
C4H5F2 (C) that of (cyclopropyl)di¯uoromethyl cation was
Fig. 1. B3LYP/6-31G(d) optimized structure of the homoallylic, 1,1-
Ê
difluorobut-3-en-1-yl cation (4a). Distance is in A.
Fig. 3. Optimized structure for 1,1-difluoropent-3-en-1-yl cation (6)
UB3LYP/6-311‡G(2df, sp)//B3LYP/6-32G(d).

M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
41
well-de®ned ring-opened cation, more likely being a delo-
calized cation.
On the other hand, when this cation was optimized using
the DFT method at the UB3LYP/6-311‡G(2df, sp)//B3LYP/
6-32G(d) level of theory, the carbocation optimized to give
the CF2-based, 1,1-di¯uoro-2-methylbut-3-en-1-yl cation
(8) as depicted in Fig. 4. Without undertaking transition
state calculations for this system, the results must be con-
sidered ambiguous regarding kinetic control, but would
seem to predict that the CF2-based cation is more stable
for this system as well as the other two.
Fig. 4. Optimized structure for 1,1-difluoro-2-methylbut-3-en-1-yl cation
(8) UB3LYP/6-311‡G(2df, sp)//B3LYP/6-32G(d).
2.2. Syntheses of the solvolysis substrates
results compromise the ability to use this reagent in a
stereospeci®c di¯uorocarbene addition process. We have
no de®nitive rationale for these competing isomerizations,
although one possibility is that the SO2 co-product is indu-
cing the isomerization. In this case, the mixture of geome-
trical isomers was used in the solvolysis study.
Di¯uorocarbene was added to allyl benzoates, 9, 11, and
13 to prepare (2,2-di¯uorocyclopropyl)methyl derivatives
10, 12, and 14 in high yield, using the new di¯uorocarbene
reagent, trimethylsilyl 2,2-di¯uoro-2-(¯uorosulfonyl)ace-
tate (TFDA). Initially, it was planned to use directly the
p-nitrobenzoate, 10, as the solvolysis substrate, but it proved
too unreactive both in HOAc and in CF3CO2H. Thus, the
tosylate derivative, 15, was prepared in a straightforward
manner from 10, as described in Section 3.
Likewise, product 14, obtained by CF2: addition to alkene
substrate 13, is a near equivalent mixture of diastereoi-
somers. This mixture of diastereomers was used in the
subsequent solvolysis study.
Although starting from pure trans-11, the product, 12,
obtained from the TFDA reaction had a signi®cant cis-
component (trans/cis ˆ 9:1). We have generally observed
that the conditions of di¯uorocarbene±alkene addition reac-
tions using TFDA as the source of CF2: can give rise to cis±
trans isomerization of alkene substrate double bonds. Such
2.3. Solvolysis of (2,2-difluorocyclopropyl)methyl
tosylate, 15
Solvolysis of (2,2-di¯uorocyclopropyl)methyl tosylate,
15, in glacial acetic acid at 96.6 8C led to the formation

42
M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
of both ring-opened (16 and 17) and non-ring-opened pro-
ducts (18), with 16 and 17 re¯ecting exclusively the regio-
chemistry of ring opening (proximal) that had been
predicted computationally [2].
product 18, was assigned a Àp1seudo ®rst-order rate constant
.
In order to con®rm that competitive unimolecular and
bimolecular reactions were involved, and that all three
(kS) of 5.8 ꢀÆ0:1†  10À6 s
The assignment of structures to non-ring-opened product
(18) and ring-opened products 16 and 17 was straightfor-
ward, based upon their characteristic 19FNMR spectra: the
AB system of 18 at d À 129:8 (JF;F ˆ 161 Hz, JH;F ˆ 12 Hz)
and À144.0 (JF;F ˆ 161 Hz, JH;H ˆ 13 Hz), and the triplets
at d À67.9 (JF;F ˆ 13 Hz) and À71.0 (JF;F ˆ 12 Hz) derived
from 17 and 16, respectively (Fig. 5).
Assuming that the ring-opened and ring-closed products
derived from competing SN1 and SN2 processes, respec-
tively, the partial rate factors for the two processes were
calculated. Thus, the SN1 process, in which tosylate 15
undergoes heterolytic cleavage to (presumably) form ring-
opened cation 4, which then is trapped by either HOAc or
TsOÀ to give products 16 and 17, was assigned a ®rst-order
rate constant (kD) of 4.4 ꢀÆ0:1†  10À6 sÀ1, and the process
involving solvent participation, which led to ring-closed
products are not derived from the common intermediate
carbocation 4a, additional acetolysis experiments were car-
ried out in 0.200 and 0.400 M NaOAc/HOAc solutions. The
results are given in Table 1.
Correcting kS by subtracting the background kS
(HOAc) (which is the kS in pure HOAc), one can obtain
the pseudo ®rst-order rate constants kS (ÀOAc) that re¯ect
the rates of nucleophilic attack by acetate ion. The values
for kS (ÀOAc), 1.8 ꢀÆ0:1†  10À5 and 3.6 ꢀÆ0:4†Â
10À5 sÀ1 for 0.200 and 0.400 M [ÀOAc], respectively,
clearly demonstrate the dependence on [ÀOAc], hence
the SN2 nature of the process that leads to non-ring-
opened product 18. On the other hand, the values of kD for
the three reactions increase only slightly, consistent with
the in¯uence of a small salt effect on the ionization
process.
Fig. 5. 19FSpectrum of tosylate 15 HOAc solvolysis product mixture.
Table 1
Kinetic results for solvolyses of (2,2-difluorocyclopropyl)methyl tosylate (15) in NaOAc/HOAc solutions at 96.6 8C
Conditions
kobs (Â10À5 sÀ1
)
Mass balance (%)
Ratio 18:(16 ‡ 17)
kS (Â10À6 sÀ1
)
kD (Â10À6 sÀ1
)
Pure HOAc
1.0 Æ 0.1
2.87 Æ 0.07
4.7 Æ 0.3
92
92
90
57:43
84:16
90:10
5.8
24
4.4
4.6
4.7
0.200 M NaOAc
0.400 M NaOAc
42

M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
43
These results do not completely rule out the possibility
that some of 18 might be formed via the SN1 process.
However, when the solvolysis was carried out in tri¯uor-
oacetic acid, a solvent of high ionizing power but poor
nucleophilicity, only ring-opened products, 18 and 19, were
formed, under conditions where the potential ring-closed
product, (2,2-di¯uorocyclopropyl)methyl tri¯uoroacetate,
was demonstrated to be stable to the solvolytic conditions.
(such as 21 and 22) containing the RCF2OAc or RCF2OTs
functionalities, respectively, appear between À75 and
À85 ppm, and ®nally, product 23, with its CF2 group not
bound to oxygen, appears as an AB system centered at
ꢁ110 ppm. Thus, trans-24 was characterized by its AB
spectrum (d, d) at d À139.7 (JF;F ˆ 158 Hz, JF;H
ˆ
14:5 Hz) and À141 (JF;F ˆ 159 Hz, JF;H ˆ 12:2 Hz); 21
by its AB spectrum (d, d) at d 78.7 (JF;F ˆ 148 Hz,
JF;H ˆ 10 Hz) and 80.4 (JF;F ˆ 147 Hz, JF;H ˆ 14 Hz); 22
by its doublet at d À74.9 (JF;H ˆ 10 Hz) (with detectable AB
sidebands indicating a JF,F of 137 Hz); and product 23 by its
AB spectrum (d, t) at d 108.4 (JF;F ˆ 250:2 Hz, JF;H
ˆ
10 Hz) and 113.8 (JF;F ˆ 250:2 Hz, JF;H ˆ 12 Hz). All the
four products were isolated and characterized, with 22±24
being synthesized independently to con®rm structures. Pro-
ducts 21 and 22 were too unstable to isolate in a pure state.
The disappearance of 20 was observed to follow ®rst-order
kinetics with a rate constant of 1.49 ꢀÆ0:02†  10À4 sÀ1 at
96.6 (Æ0.1) 8C. Assuming competing kD and kS processes,
one can assign values of 6.40 ꢀÆ0:02†  10À5 sÀ1 and 8.50
ꢀÆ0:02†  10À5 to the ®rst-order rate constant, kD, and the
pseudo ®rst-order rate constant kS, respectively.
The above solvolytic results are completely consistent
with the ionization of 15 proceeding directly and regiospe-
ci®cally to the homoallylic carbocation 4a, as had been
predicted computationally [2].
2.4. Solvolysis of (3-methyl-2,2-difluorocyclopropyl)methyl
tosylate, 20
Solvolysis of tosylate 20 in glacial acetic acid at 96.3 8C
resulted in the formation of three, ring-opened products, 21±
23, as well as the non-ring-opened product 24.
In order to assess more fully the SN1 pathway for the
solvolysis of 20, its reaction in CF3CO2H was examined.
Only two signi®cant products were observed, both deriving
Ring-opened products 21 and 22 re¯ect proximal ring
cleavage, whereas the formation of 23 re¯ects distal ring
cleavage. Thus, it appears that there is kinetic competition
between proximal and distal ring cleavage in this solvolysis
reaction. Of course, the presence of 24 indicates that direct
displacement by solvent participation competes with the
SN1, ionization process that leads to the other three products.
Acetate product 21 is unstable under the acidic solvolysis
conditions, probably converting to the alcohol which ends up
as the acid ¯uoride.
from proximal bond cleavage. This must re¯ect thermody-
namic control of the CF3CO2H reaction to the extent that any
product resulting from distal bond cleavage must be rever-
sibly formed and end up as 22 or 25. Indeed, calculations
(UBLYP/6-311‡g(2df, 2p)//UB3LYP/6-31g(d)) of the
hypothetical reaction (Eq. (1)) indicate a signi®cantly
greater thermodynamic stability for structural type 22 (a-
alkoxy) in comparison to type 23 (b-alkoxy).
As was the case in the solvolysis of tosylate 15, differ-
entiating between ring-opened and non-ring-opened pro-
ducts was, on the basis of their 19FNMR spectra, a
simple task in analyzing products from the acetolysis of
20. The presence of any product retaining the gem-di¯uor-
ocyclopropyl entity (such as 24) would be unambiguously
recognized from its characteristic AB system, usually
between À130 and À150 ppm and with characteristic JF,F
coupling constants of ꢁ160 Hz. Any ring-opened product

44
M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
Thus, although, kinetically, nucleophilic attack on the inter-
mediate cation or cations leads competitively to products
derived from distal and proximal cleavage, the products
derived from proximal cleavage must be more stable than
those obtained from distal cleavage.
ꢀÆ0:02†  10À5
s
À1. After 36 h of reaction, all starting
material was gone, and the ratios of products had modi®ed
only slightly to 50:17:33, thus indicating that the products
were moderately stable to the reaction conditions and that
kinetic control of products had been obtained.
From these results it can be concluded that the presence of
a `single methyl substituent' at the third position of the (2,2-
di¯uorocyclopropyl)methyl system is suf®cient to induce
competitive solvolytic proximal and distal cleavage under
condition of kinetic control. The methyl substituent of 20
gives rise to approximately a 10-fold increase in the rate
constant kD as would be expected on the basis of its
carbocation-stabilizing ability.
In the only previous literature report that is relevant to
this result [8], Boger found that even with alkyl substitution
at the third position, the acid catalyzed ring opening of
spirocyclohexadienone, 26, led only to its proximal ring
cleavage.
The failure to observe acetate 30 at the higher temperature
(97 8C) is presumably due to an acid catalyzed rearrange-
ment of this more reactive secondary acetate to ring-opened
acetates 28. The observed ring-opened versus non-ring-
opened product distribution from tosylate 27 at the lower
temperature (42 8C), compared to those of primary tosylates
15 and 20 at 96.6 8C, suggests that kS was less competitive
with kD for the secondary substrate than for the primary
substrates.
2.6. Relative reactivity of (2,2-difluorocyclopropyl)methyl
cation systems
With tosylates 15, 20 and 27 undergoing ultra-fast ring
opening upon solvolysis, with well understood regiochem-
istries different from those exhibited by analogous radicals,
the (2,2-di¯uorocyclopropyl)methyl system certainly qua-
li®es as a hypersensitive probe capable of distinguishing
between cation and radical mechanisms.
In evaluating the potential ef®cacy of a new mechanistic
probe, in addition to its ability to differentiate between the
intermediacy of a radical and a carbocation, it should ideally
play simply the role of an ``observer'' of the reaction. That
is, it should not exert a signi®cant mechanistic in¯uence
upon the reaction system it is testing. If the diagnostic probe
itself has a steric or electronic bias so as to favor or disfavor
one of the possible mechanistic pathways, then the inter-
pretation of the results from use of such a probe may be
ambiguous. This is certainly one of the potential ¯aws of any
cyclopropylcarbinyl system that purports to distinguish
carbocation from radical intermediates, because although
the cyclopropyl group should exert little if any kinetic
in¯uence upon a radical-forming process, it is well recog-
nized to enhance greatly formation of carbocations. Cyclo-
propylmethyl tosylate, for example, undergoes acetolysis
with a rate constant more than 100,000 times that of the
model primary system, isobutyl tosylate (Table 2).
2.5. Solvolysis of 1-(2,2-difluorocyclopropyl)ethyl
tosylate, 27
Modi®cation of the (2,2-di¯uorocyclopropyl)methyl sys-
tem into a secondary tosylate, such as 27, was expected to
affect the reactivity of the system, but not the regiochemistry
of the ring opening process. Indeed, when tosylate 27 was
heated, in a preliminary experiment, in HOAc at 97 8C for
30 min (a length of time when little of the primary tosylate
15 would have reacted), all starting material disappeared,
and four ring-opened products derived from proximal ring
cleavage were observed to be formed. That these products
were ring opened was readily discerned from the 19FNMR
spectrum of the crude product mixture, which revealed two
pairs of triplets due to the cis- and trans-isomers of 28 and
29, in the range À74 to À82 ppm.
When the solvolysis was carried out at a lower tempera-
ture (42 8C), results similar to those observed for 15, includ-
ing formation of a signi®cant amount of non-ring opened
product (30), were obtained. The ratio of ring-opened to
non-ring-opened products was ꢁ2.3, and the rate constant
for disappearance of 27 at 42.1 8C was found to be 1.64
In evaluating the ef®cacy of this probe, the acetolysis rate
of the tosylate 15 which re¯ects the ease of cation formation,
was compared with those of isobutyl tosylate and cyclopro-
pylmethyl tosylate (which we extrapolated from their
respective activation parameters) (Tables 2 and 3) [9,10].

M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
45
standard (d ˆ 77:00). 19FNMR are reported in ppm using
Table 2
Relative rate (kD) of acetolysis of primary tosylates at 96.6 8C
CFCl3 as internal standard (d ˆ 0:00). All kinetic studies
were carried out using sealed NMR tubes submerged in an
oil bath with a Statim temperature controller.
Substrates
Relative rate Reference
Cyclopropylmethyl tosylate
(2,2-Difluorocyclopropyl)methyl tosylate (15)
(3-Methyl-2,2-difluorocyclopropyl)methyl
tosylate (20)
8.4 Â 104
[9]
All the reagents were obtained from Fisher except those
mentioned. THFwas dried through distillation from Na/
benzophenone. 2,2-Di¯uoro-2-(¯uorosulfonyl)acetic acid
was provided by Professor Qing-Yun Chen, from Shanghai
Institute of Organic Chemistry.
1
10
This work
This work
Isobutyl tosylate
0.56
[10]
Table 3
3.1. Computational methodology
Data for solvolysis kinetic run of 27
Density functional theory calculations were performed
using the Gaussian 98 program package [11]. Structures and
transition structures were optimized using restricted Becke's
hybrid three parameter functional (B3LYP) [12] and the 6-
31G(d) basis set [13]. Using the same level of theory,
vibrational frequency calculations were performed on all
stationary points to determine thermal energies. Thermal
energies, the sum of zero-point, vibrational, rotational, and
translational energies at 285.15 K, were obtained using
unscaled frequencies. Single-point energies were calculated
using UB3LYP level of theory using the 6-311‡G(2df, 2p)
basis set [14].
Time (s)
C0/Ct
ln(C0/Ct)
Time (s)
C0/Ct
ln(C0/Ct)
0
2,134
1
0
17,072
19,206
21,340
23,474
25,608
27,742
29,876
32,010
1.32169
1.37785
1.41654
1.46452
1.52605
1.56822
1.63016
1.6846
0.27891
0.32052
0.34821
0.38152
0.42268
0.44994
0.48868
0.52153
1.03299
1.07075
1.11002
1.15228
1.18538
1.22372
1.26287
0.03246
0.06836
0.10438
0.14175
0.17006
0.2019
0.2338
4,268
6,402
8,536
10,670
12,802
14,938
At 96.6 8C, the 8:4 Â 104 times decrease of acetolysis
rate of (2,2-di¯uorocyclopropyl)methyl tosylate (15) rela-
tive to that of cyclopropylmethyl tosylate, reveals a sub-
stantial inductive destabilizing in¯uence on the ionization
transition state of 15 due to the ¯uorine substituents on its
ring.
3.2. Preparative procedures
3.2.1. Trimethylsilyl 2,2-difluoro-2-(fluorosulfonyl)acetate
A dry 1 l three neck round bottom ¯ask, equipped with
magnetic stirrer and an additional funnel, was charged with
150 g (0.85 mol) 2,2-di¯uoro-2-(¯uorosulfonyl)acetic acid.
Under N2, at 0 8C, 329 g (3.05 mol, 3.6 eq.) of trimethylsilyl
chloride was added dropwise. The generated gas was passed
through 10% NaOH aqueous solution to remove HCl. Then,
the reaction mixture was stirred at room temperature over
night. Distillation at 27 mmHg, 62±63 8C gave 166 g
Since the rate constant (kD) of 15 is only 1.8 times greater
than that of isobutyl tosylate at 96.6 8C, the net result of the
¯uorine substituents is to essentially eliminate the rate
enhancing effect of the cyclopropylmethyl moiety (in 15)
relative to a typical primary substrate, such as isobutyl
tosylate. Thus, the (2,2-di¯uorocyclopropyl)methyl system
should exhibit no kinetic bias towards either the radical or
the carbocation mechanism. It is also not expected that (2,2-
di¯uorocyclopropyl)methyl systems, 15, 20 and 27 should
have any speci®c steric problems or stereoelectronic in¯u-
ences. Therefore, since they have been demonstrated to be
effectively electronically benign, we would conclude on the
basis of our kinetic, regiochemical and computational stu-
dies, that the (2,2-di¯uorocyclopropyl)methyl system
should be capable of acting as an effective, hypersensitive
probe of radical and/or carbocation mechanisms.
1
(0.66 mol, 78%). H NMR d: 0.40 (s); 13C NMR d: 155.1
(t, JꢀCÀF† ˆ 27:0 Hz), 112.2 (dt, JdꢀCÀF† ˆ 31:5 Hz, JtꢀCÀF†
ˆ
299:0 Hz), À1.1; 19FNMR d: 40.6 (1F, s), À103.7 (2F, s);
‡
HRMS (CI) C5H10O4S1Si1F3 (M ‡ 1) . Calcd.: 251.0212,
Found: 251.0015.
3.2.2. Allyl p-nitrobenzoate (9)
A 250 ml dry three necked round bottom ¯ask, equipped
with a magnetic stirrer, a condenser and an additional ¯ask,
was charged with 80 ml dry THFand 8.7 g (150 mmol,
1 eq., Aldrich) allyl alcohol and 11.9 g (150 mmol) pyridine.
At 0 8C, 100 ml dry THFsolution of 32.4 g of p-nitroben-
zoyl chloride (180 mmol, Acros) was added dropwise. After
stirring at room temperature for 8 h, the reaction mixture
was ®ltered, and solvent removed under reduced pressure.
Ethyl acetate (400 ml) was added, and the solution washed
with water, 5% NaHCO3, water and brine. After drying
(Na2SO4), solvent was removed under reduced pressure to
give 15.3 g of 9 (yellow liquid, 49%). 1H NMR d: 8.23 (4H,
aromatic AA0XX0), 6.02 (1H, ddt, JdꢀHÀH† ˆ 17:1, 10.5 Hz,
3. Experimental
All 1H NMR spectra were recorded at 300 MHz, 13C
NMR spectra at 75 MHz and 19FNMR spectra at 282 MHz
on Varian VXR-300, Mercury-300 and Gemini-300 NMR
spectrometers. The chemical shifts of 1H signals are
reported in ppm down ®eld relative to tetramethylsilane
(TMS) (d ˆ 0:00) in CDCl3. 13C signals are expressed in
ppm using the central peak of the CDCl3 signal as internal

46
M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
JtꢀHÀH† ˆ 5:9 Hz), 5.41 (1H, d and pseudo-q, JdꢀHÀH†
ˆ
was added. The reaction mixture was maintained at À5 8C
for 12 h, whereupon 3 ml of ice water was added, and the
solution was extracted with diethyl ether (2 ml  6). The
combined ether layers were washed with 6 M HCl/H2O,
water, and brine, then dried over Na2SO4 and Na2CO3. After
the solvent was removed under reduced pressure, 146 mg
(0.56 mmol, 53%) of pure product 15 (yellow oil) was
obtained. Tosylate 15 was kept in diethyl ether solution
in the presence of anhydrous Na2CO3. 1H NMR d: 7.56 (4H,
aromatic AA0XX0), 4.06 (2H, d, JdꢀHÀH† ˆ 7:8 Hz), 2.42
(3H, s), 1.91 (1H, m), 1.52 (1H, m), 1.18 (1H, m); 13C
NMR d: 145.1, 132.8, 129.9, 127.9, 112.3 (t,
JtꢀFÀC† ˆ 283:0 Hz), 66.9 (d, Jd ˆ 5:5 Hz), 21.7, 20.9 (dd,
17:1 Hz, JqꢀHÀH† ˆ 1:4 Hz), 5.31 (1H, d and pseudo-q,
JdꢀHÀH† ˆ 10:5 Hz, JqꢀHÀH† ˆ 1:4 Hz), 4.84 (2H, dt,
JdꢀHÀH† ˆ 5:9 Hz, JtꢀHÀH† ˆ 1:4 Hz); 13C NMR d:
164.278, 150.490, 135.454, 130.701, 123.490, 119.090,
‡
66.391; HRMS (CI) C10H10O4N (M ‡ 1) . Calcd.:
208.0610, Found: 208.0604.
3.2.3. (2,2-Difluorocyclopropyl)methyl
p-nitrobenzoate (10)
A 100 ml dry three necked round bottom ¯ask, equipped
with a magnetic stirrer, was charged with 12.25 g
(59.2 mmol) of 9, and a small amount of CsF. At 97 8C,
with stirring, 32.4 g (130 mmol) of TFDA was slowly added
using a syringe pump, at a rate of 1.67 ml/h. The reaction
was then cooled to room temperature and 0.125 g of product
was separated by ¯ash column chromatography (silica gel,
10% diethyl ether/hexanes) (Rf ˆ 0:28, 20% diethyl ether/
hexanes), from 0.28 g of the reaction mixture (total 16.7 g).
JdꢀCÀF† ˆ 12:1 Hz, JdꢀCÀF† ˆ 10:6 Hz), 15.3 (t, JtꢀCÀF†
ˆ
11:6 Hz); 19FNMR d: À129.7 (1F, dtdm, JdꢀFÀF† ˆ 162:3
Hz, JtꢀFÀH† ˆ 13:0 Hz, JdꢀFÀH† ˆ 4:0 Hz), À143.2 (1H, dddt,
JdꢀFÀF† ˆ 162:3 Hz, JdꢀHÀF† ˆ 13:3, 4.8 Hz, JtꢀHÀF† ˆ 1:7
‡
Hz; HRMS (CI) C11H13O3F2S (M ‡ 1) . Calcd.:
262.0554, Found; 263.0541 (3.7), 91.0362 (100).
1
This gave 10 in an isolated yield of 49% (74%, H NMR
yield). 1H NMR d: 8.24 (4H, aromatic AA0XX0), 4.52 (1H,
dddd, JdꢀHÀH† ˆ 12:0 Hz, Jd ˆ 7:5, 2.7, 1.2 Hz), 4.32 (1H,
ddd, Jd ˆ 12:0, 8.1, 1.8 Hz), 2.10 (1H, m), 1.60 (1H, m),
1.31 (1H, m); 13C NMR d: 164.5, 150.6, 135.1, 130.8, 123.6,
112.8 (t, JtꢀFÀC† ˆ 283:1 Hz), 62.6 (d, Jd ˆ 5:5 Hz), 20.9 (t,
JtꢀFÀC† ˆ 11:1 Hz), 15.1 (t, JtꢀFÀC† ˆ 11:6 Hz); 19FNMR
d: À129.8 (1F, dtt, JdꢀFÀF† ˆ 161:1 Hz, JtꢀFÀH† ˆ 12:1,
3.2.6. (2,2-Difluorocyclopropyl)methyl acetate (18)
A dry 5 ml round bottom ¯ask was charged with
100 mg (0.93 mmol) (2,2-di¯uorocyclopropyl)methanol,
1.2 ml pyridine. At 0 8C, 300 mg (2.94 mmol) of acetic
anhydride was added. The reaction mixture was maintained
at À5 8C for 12 h. Then, 3 ml of ice water was added, and the
solution was extracted with diethyl ether (2 ml  6). The
combined ether layer was washed with 6 M HCl/H2O, water,
and brine, and then dried over Na2SO4 and Na2CO3. After
the solvent was removed by distillation, 82 mg (0.54 mmol,
58%) of pure product 18 was obtained. 1H NMR d: 4.19 (1H,
dddd, JdꢀHÀH† ˆ 12:0 Hz, Jd ˆ 7:5, 2.7, 1.2 Hz), 4.03 (1H,
ddd, JdꢀHÀH† ˆ 12:0 Hz, Jd ˆ 8:1, 1.8 Hz), 2.07 (3H, s), 1.93
(1H, m), 1.50 (1H, m), 1.18 (1H, m); 13C NMR d: 170.9,
112.9 (t, JtꢀCÀF† ˆ 283:0 Hz), 61.2 (d, Jd ˆ 5:5 Hz), 21.0 (t,
JtꢀCÀF† ˆ 11:6 Hz), 20.8, 15.0 (t, JtꢀCÀF† ˆ 11:0 Hz); 19F
3.7 Hz), À143.7 (1F, ddd, JdꢀFÀF† ˆ 161:1 Hz, JdꢀFÀH†
ˆ
13:3 Hz, JdꢀFÀH† ˆ 4:5 Hz); HRMS (CI) C11H10O4F2N1
‡
(M ‡ 1) . Calcd.: 258.0578, Found: 258.0585.
3.2.4. (2,2-Difluorocyclopropyl)methanol
A 80 ml 10% of NaOH/H2O solution was added to the
reaction mixture mentioned above. At 80 8C, this mixture
was stirred for 2 h, and then cooled to room temperature.
Water (200 ml) was added, and the mixture was extracted
with diethyl ether (®ve times, 100 ml). The combined ether
layer was washed with 5% NaHCO3/H2O, H2O, and brine,
and then dried over Na2SO4. Ether was removed under
reduced pressure. Distillation of the residue gave 3.7 g
(34.3 mmol, two-step yield 58%) product (60 mmHg, 76±
80 8C). 1H NMR d: 3.75 (1H, dddd, JdꢀHÀH† ˆ 12:0 Hz,
Jd ˆ 6:6, 3.0, 1.2 Hz), 3.64 (1H, ddm, JdꢀHÀH† ˆ 12:0 Hz,
Jd ˆ 8:1 Hz), 1.87 (1H, m), 1.75 (1H broad), 1.44 (1H, m),
1.14 (1H, m); 13C NMR d: 113.7 (t, JtꢀFÀC† ˆ 282:0 Hz),
60.0 (d, Jd ˆ 5:6 Hz), 24.2 (t, JtꢀFÀC† ˆ 11:1 Hz), 14.5 (t,
NMR d: À129.8 (1F, dtdd, JdꢀFÀF† ˆ 160:9 Hz, JtꢀHÀF†
12:1 Hz, JdꢀHÀF† ˆ 3:7, 2.8 Hz), À144.0 (1F, dddt, JdꢀFÀF†
ˆ
ˆ
160:9 Hz, JdꢀHÀF† ˆ 13:3, 4.8 Hz, JtꢀHÀF† ˆ 1:5 Hz); HRMS
‡
(CI) C6H9O2F2 (M ‡ 1) . Calcd.: 151.0571, Found:
151.0643 (100).
3.2.7. 1,1-Difluorobut-3-en-1-yl tosylate (17)
A dry NMR tube with valve charged with 0.2 g (2,2-
sdi¯uorocyclopropyl)methyl tosylate, 0.2 ml tri¯uoroacetic
acid, was maintained at 53 8C for 12 h. Then, the reaction
mixture was cooled to room temperature. After vacuum
transfer, 1 ml diethyl ether was added into the residue.
The solution was treated with MgSO4and Na2CO3 at
0 8C. Filtration of the mixture, followed by vacuum removal
of the solvent, gave a mixture. Flash column chromatogra-
phy (hexanes/benzene ˆ 3:1) provided pure product 17
(Rf ˆ 0:12). 1H NMR d: 7.59 (4H, aromatic AA0XX0),
JtꢀFÀC† ˆ 11:0 Hz); 19FNMR d: À129.1 (1F, dt, JdꢀFÀF†
160:3 Hz, JdꢀFÀH† ˆ 12:7 Hz), À144.7 (1F, dd, JdꢀFÀF†
ˆ
ˆ
160:3 Hz, JdꢀFÀH† ˆ 12:7 Hz); HRMS (CI) C4H7O1F2
‡
(M ‡ 1) . Calcd.: 109.0465, Found: 109.0531 (0.23),
91.0363 (100).
3.2.5. (2,2-Difluorocyclopropyl)methyl tosylate (15)
A dry 5 ml round bottom ¯ask was charged with 115 mg
(1.06 mmol) (2,2-di¯uorocyclopropyl)methanol, 1.8 ml
pyridine. At 0 8C, 408 mg (2.14 mmol) of tosyl chloride
5.65 (1H, m), 5.24 (2H, m), 2.83 (2H, tdt, JtꢀFÀH†
12:3 Hz, JdꢀHÀH† ˆ 6:9 Hz, JtꢀHÀH† ˆ 1:2 Hz), 2.45 (3H,
s); 19FNMR d: À68.3 (2F, t, JtꢀHÀF† ˆ 12:7 Hz); 13C
ˆ

M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
47
NMR d: 145.7, 134.0, 129.8, 128.1, 126.3, 123.8 (t,
JtꢀCÀF† ˆ 276:0 Hz), 122.0, 40.6 (t, JtꢀCÀF† ˆ 27:4 Hz),
128.9, 125.2, 65.7, 17.9; HR-MS (FAB), C11H2O2 (M ‡ 1).
Calcd.: 177.0917, Found: 177.0917.
‡
21.8; HRMS (FAB) C11H13O3F2S (M ‡ 1) . Calcd.:
263.0553, Found: 263.0553 (25), 173.0157 (100), 91.0554
(36).
3.2.11. (2,2-Difluoro-3-methylcyclopropy)methyl
benzoate (12)
A dry 25 ml two necked round bottom ¯ask was charged
with 2-buten-1-yl benzoate 11 (7.3800 g, 41.88 mmol) and
sodium ¯uoride (0.100 g, catalyst). To this mixture TFDA
(16.00 ml, 83.76 mmol) was added via a glass syringe at a
rate of 0.35 ml/h while the mixture was heated to 120 8C
with re¯ux under nitrogen. The resulting material was
diluted with diethyl ether, washed with water and brine,
and dried over MgSO4. The ether was removed giving
impure 12, which was then puri®ed via column chromato-
graphy (Rf ˆ 0:3, 5% ethyl ether/hexanes) resulting in a
3.2.8. 1,1-Difluorobut-3-en-1-yl trifluoroacetate (19)
A dry NMR tube with valve charged with 0.2 g (2,2-
di¯uorocyclopropyl)methyl tosylate, 0.2 ml tri¯uoroacetic
acid, was maintained at 53 8C for 12 h. Then the reaction
mixture was cooled to room temperature. After vacuum
transfer, 1 ml CDCl3 was added into the volatile reaction
mixture. The solution was treated with MgSO4 and Na2CO3
at 0 8C. This gave a CDCl3 solution of 19. 1H NMR d: 5.73
(1H, ddt, JdꢀHÀH† ˆ 17:1, 10.2 Hz, JtꢀHÀH† ˆ 6:9 Hz), 5.33
1
(1H, dm, JdꢀHÀH† ˆ 10:2 Hz), 5.30 (1H, dm, JdꢀHÀH†
ˆ
ˆ
7.36 g mixture of both isomers (cis/trans ˆ 1:9, 76%). H
17:1 Hz), 3.04 (2H, dt, JdꢀHÀH† ˆ 6:9 Hz, JtꢀFÀH†
NMR d: 8.06 (2H, m), 7.58 (1H, dt), 7.45 (2H, m), 4.39 (2H,
m), 1.62 (2H, m), 1.22 (3H, m); 13C NMR d: 166.7, 133.4,
130.1, 129.9, 128.6, 115.5 (t, 1JCÀF ˆ 287:3 Hz), 61.9, 27.8
12:6 Hz); 19FNMR d: À71.6 (2F, t, JtꢀFÀH† ˆ 12:7 Hz),
‡
À76.0 (3F, s); HRMS (CI) C6H6O2F5 (M ‡ 1) . Calcd.:
2
2
205.0288, Found 205.0296 (1.12); C4H5F2 (M ‡ 1±
(t, JCÀF ˆ 10:8 Hz), 22.1 (t, JCÀF ˆ 10:3 Hz), 11.1; 19F
‡
2
CF3COOH) . Calcd.: 91.0359, Found: 91.0372 (100).
NMR cis isomer d: À126.0 (1F, dt, JFÀF ˆ 160:2 Hz,
3JHÀF ˆ 13:7 Hz), À153.7 (1F, d, 2JFÀF ˆ 160:2 Hz); trans
2
3
3.2.9. (2,2-Difluorocyclopropyl)methyl trifluoroacetate
A dry 5 ml round bottom ¯ask equipped with magnetic
stir bar was charged with 200 mg (1.85 mmol) of (2,2-
di¯uorocyclopropyl)methanol, and 218 mg of 2,6-lutidine
(2 mmol). The system was cooled to 0 8C. With stirring,
under the protection of N2, 390 mg of tri¯uoroacetic anhy-
dride was added, the temperature raised to room tempera-
ture, and 30 min later, the reaction mixture was distilled to
give 280 mg colorless liquid at 65 8C (1.37 mmol, 74%). 1H
NMR d: 4.40 (2H, AB pattern), 2.04 (1H, m), 1.62 (1H, m),
1.30 (1H, m); 13C NMR d: 157.4 (q, JqꢀCÀF† ˆ 42:8 Hz),
114.4 (q, JqꢀCÀF† ˆ 285:0 Hz), 112.2 (dd, JdꢀCÀF† ˆ 282:0,
284.0 Hz), 64.7 (d, Jd ˆ 6:0 Hz), 20.2 (dd, JdꢀCÀF† ˆ 12:6,
11.1 Hz), 15.2 (t, JtꢀCÀF† ˆ 11:6 Hz); 19FNMR d: À75.4 (3F,
s), À129.5 (1F, dt, JdꢀFÀF† ˆ 162:3 Hz, JtꢀFÀH† ˆ 10:5 Hz),
À143.3 (1F, ddd, JdꢀFÀF† ˆ 162:0 Hz, JdꢀFÀH† ˆ 12:7 Hz,
isomer d: À140.3 (2F, dd, JFÀF ˆ 160:0 Hz, JHÀF ˆ
14:9 Hz); HR-MS (FAB) C12H12O2F2 (M ‡ 1). Calcd.:
227.0884, Found: 227.0884.
3.2.12. (2,2-Difluoro-3-methylcyclopropyl)methanol
In a dry 150 ml one necked round bottom ¯ask equipped
with a condenser, 25 ml of 3 M NaOH and (2,2-di¯uoro-3-
methylcyclopropyl)methyl benzoate 12 (3.27 g, 14.5 mmol)
were combined and heated to 110 8C. After the reaction ran
overnight, the product was extracted with ether, brine and
dried over MgSO4. After ®ltration and removal of ether,
1.62 g of product was obtained (92%). 1H NMR d: 3.65 (2H,
m), 2.32 (1H, s), 1.38 (1H, m), 1.60 (3H, m); 13C NMR d:
1
2
115.9 (t, JCÀF ˆ 287:1 Hz), 59.9, 31.0 (t, JCÀF
ˆ
10:1 Hz), 21.5 (t, JCÀF ˆ 10:6 Hz), 11.1; 19FNMR cis
2
2
3
isomer d: À126.0 (1F, dt, JFÀF ˆ 158:7 Hz, JHÀF
ˆ
‡
2
JdꢀFÀH† ˆ 6:5 Hz); HRMS (CI) C6H6O2F5 (M ‡ 1) .
13:7 Hz), À154.5 (1F, d, JFÀF ˆ 158:7 Hz); trans isomer
d: À140.5 (2F, m); LR-MS (EI) C5H7OF2: 105 (46), 91 (38),
77 (54), 51 (100), 43 (64).
Calcd.: 205.0288, Found: 205.0283 (0.62), 91.0369
(100).
3.2.10. 2-Buten-1-yl benzoate (11)
3.2.13. (2,2-Difluoro-3-methylcyclopropyl)methyl
tosylate (20)
A dry 250 ml three necked round bottom ¯ask was
charged with dry diethyl ether (80 ml), crotyl alcohol
(5.00 g, 69.34 mmol), and triethylamine (14.16 g, 138.96
mmol). Under nitrogen and at 0 8C, benzoyl chloride
(9.94 g, 70.7 mmol) was added slowly via an addition funnel
overnight. The reaction mixture was diluted with 10 ml of
diethyl ether. The ether layer was washed with water, 0.5N
HCl, 5% NaHCO3, brine, and dried over MgSO4. After
®ltering and removing the solvent under reduced pressure,
10.33 g of pure ester 11 (yellow oil, yield 84%) were
obtained. 1H NMR d: 8.15 (2H, d), 7.65 (1H, m), 7.53
(2H, m), 5.79 (2H, m), 4.55 (2H, m), 1.55 (3H, m); 13C
NMR d: 166.5, 132.9, 131.4, 130.6, 130.4, 129.7, 128.9,
A mixture of (2,2-di¯uoro-3-methylcyclopropyl)metha-
nol (0.56 g, 4.62 mmol) and pyridine (1.85 g, 9.64 mmol)
was placed in a reaction vial. At 0 8C, tosyl chloride (1.35 g,
4.62 mmol) was added and the reaction mixture was main-
tained at À5 8C for 12 h. The resulting material was diluted
with diethyl ether and washed with water, 1 M HCl, and
brine and then dried over MgSO4 and Na2CO3. Filtration and
removal of ether resulted in a 0.63 g mixture of both isomers
(yellow oil, cis/trans ˆ 1:13, 46%). Tosylate 20 was stored
in diethyl ether solution over Na2CO3 (anhydrous). 1H NMR
d: 7.80 (2H, d), 7.36 (2H, d), 4.08 (2H, m), 2.45 (3H, s), 1.46
(2H, m), 1.15 (3H, m); 13C NMR d: 145.4, 133.0, 130.5,

48
M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
130.2, 128.0, 127.2, 114.6 (t, 1JCÀF ˆ 286:0 Hz), 67.4, 27.4
(t, 2JCÀF ˆ 11:0 Hz), 22.4 (t, 2JCÀF ˆ 10:0 Hz), 21.7, 10.8;
19FNMR cis isomer d: À125.6 (1F, dt, 2JFÀF ˆ 162:2 Hz,
3JHÀF ˆ 12:8 Hz), À153.4 (1F, d, 2JFÀF ˆ 160:1 Hz); trans
5% HCl was added and the mixture was stirred for 5 min.
Excess zinc was ®ltered off and the organic layer was
separated. The aqueous layer was extracted with ether.
The organic layer washed with saturated NaHCO3 and dried
over MgSO4. Filtration and removal of ether resulted in a
2
3
isomer d: À143.2 (2F, dd, JFÀF ˆ 162:3 Hz, JHÀF
ˆ
1
13:3 Hz); HR-MS (FAB) C12H14O3F2S (M ‡ 1). Calcd.:
0.39 g of 3,3-di¯uoropent-4-en-2-ol (yellow oil, 53%). H
NMR d: 5.98 (1H, m), 5.72 (1H, m), 5.56 (1H, d,
3Jcis HÀH ˆ 10:99 Hz), 3.96 (1H, m), 2.14 (1H, s), 1.26
277.0711, Found: 277.0749.
2
3.2.14. 1,1-Difluoro-2-methylbut-3-en-1-yl tosylate (22)
A dry reaction vessel was charged with (2,2-di¯uoro-3-
methylcyclopropyl)methyl tosylate (20) (0.2 g), tri¯uoroa-
cetic acid (0.2 ml) and maintained at 50 8C for 4 h. Then the
reaction mixture was cooled to room temperature. After
®ltration of the mixture and removal of the solvent, the
residue was puri®ed by ¯ash column chromatography
(Rf ˆ 0:38, 10% ethyl acetate/hexanes) which provided
(3H, d); 13C NMR d: 129.8 (t, JCÀF ˆ 25:5 Hz), 121.5
3
1
(t, JCÀF ˆ 9:5 Hz), 120.6 (t, JCÀF ˆ 240:9 Hz), 69.7 (t,
2JCÀF ˆ 29:5 Hz), 16.3 (t, JCÀF ˆ 2:3 Hz); 19FNMR d:
3
À112.2 (2F, dt, 2JFÀF ˆ 247:2 Hz, JHÀF ˆ 10:7 Hz); LR-
3
MS (EI) C5H7OF2: 105 (13), 103 (19), 83 (14), 77 (30), 59
(11), 51 (20), 43 (53), 45 (100).
3.2.17. 3,3-Difluoropent-4-en-2-yl acetate (23)
1
22 (yellow oil, 20%). H NMR d: 7.84 (2H, d), 7.35 (2H,
In a dry reaction vessel, 0.950 ml of dry diethyl ether,
triethyl amine (0.31 ml, 1.64 mmol), and 3,3-di¯uoropent-4-
en-2-ol (0.12 g, 0.98 mmol) were combined and cooled to
0 8C. Acetyl chloride (0.058 ml, 0.82 mmol) was added
slowly via a syringe and the reaction ran overnight. The
reaction mixture was diluted with ether and the product was
extracted with 0.6 M HCl, water, brine and dried over
MgSO4. Filtration and removal of ether resulted in a
3
3
d), 5.66 (1H, ddd, Jtrans HÀH ˆ 17:33 Hz, JHÀH ˆ 10:01,
10.25 Hz), 5.18 (2H, m), 2.30 (1H, m), 2.45 (3H, s), 1.12
(3H, d); 13C NMR d: 145.9, 134.4, 133.1 (t, 3JCÀF ˆ 2:0 Hz),
1
130.5, 128.6, 128.6, 128.3, 127.2 (t, JCÀF ˆ 134:4 Hz),
2
3
119.5, 44.3 (t, JCÀF ˆ 15:7 Hz), 21.9, 13.2 (t, JCÀF
ˆ
13:0 Hz); 19FNMR d: À74.6 (2F, d, JFÀF ˆ 9:6 Hz);
LRMS (EI) C12H14O3F2S: 155 (8), 104 (18), 91 (100), 77
(8), 65 (39).
2
1
0.080 g of 23 (yellow oil, yield 60%). H NMR d: 5.86
(1H, m), 5.66 (1H, m), 5.67 (1H, m), 5.09 (1H, m), 2.20 (3H,
2
3.2.15. (2,2-Difluoro-3-methylcyclopropyl)methyl
acetate (24)
s), 1.23 (3H, m); 19FNMR d: À111.1 (2F, dt, JFÀF
ˆ
3
250:2 Hz, JHÀF ˆ 9:2 Hz); LR-MS (EI) C7H9O2F2: 165
A dry reaction vessel was charged with (2,2-di¯uoro-3-
methylcyclopropyl)methanol (0.12 g, 0.98 mmol, pyridine
(0.13 ml, 1.97 mmol), and 0.95 ml of dry diethyl ether. After
cooling the mixture to 0 8C under nitrogen, acetic anhydride
(0.10 ml, 0.98) was added slowly via a syringe and the
reaction was allowed to run overnight. The resulting mixture
was diluted with ether and the product was extracted with
0.6 M HCl, 5% NaHCO3, and brine and dried over MgSO4.
Filtration and removal of ether resulted in a 0.13 g mixture
of both isomers of 24 (yellow oil, cis/trans ˆ 1:10, 98%). 1H
NMR d: 4.21 (1H, m), 4.01 (1H, m), 2.07 (3H, s), 1.46 (1H,
m), 1.19 (3H, m); 13C NMR d: 171.1, 115.2 (t,
(7), 145 (10), 77 (5), 51 (7), 43 (100).
3.2.18. 1,1-Difluoro-2-methylbut-en-1-yl
trifluoroacetate (25)
A dry reaction vessel was charged with (2,2-di¯uoro-3-
methylcyclopropyl)methyl tosylate (20) (0.2 g), tri¯uoroa-
cetic acid (0.2 ml) and maintained at 50 8C for 4 h. Then, the
reaction mixture was cooled to room temperature. After
vacuum transfer, 1 ml of CDCl3 was added to the volatile
reaction mixture and the solution was treated with MgSO4
and Na2CO3 at 0 8C which gave the CDCl3 solution of 1,1-
di¯uoro-2-methylbut-3-en-1-yl tri¯uoroacetate (25). 1H
2
3
3
1JCÀF ˆ 287:6 Hz), 61.4, 27.6 (t, JCÀF ˆ 10:6 Hz), 22.0
NMR d: 5.76 (1H, ddd, Jtrans HÀH ˆ 17:58 Hz, JHÀH ˆ
2
(t, JCÀF ˆ 10:1 Hz), 21.1, 11.1; 19FNMR cis isomer d:
9:52, 10.01 Hz), 5.28 (2H, m), 3.13 (1H, m), 1.26 (3H, d);
2
3
2
3
À126.4 (1F, dt, JFÀF ˆ 160:2 Hz, JHÀF ˆ 13:7 Hz),
19FNMR d: À79.5 (2F, dt, JFÀF ˆ 148:0 Hz, JHÀF
ˆ
2
À153.8 (1F, d, JFÀF ˆ 158:7 Hz); trans isomer d:
10:7 Hz), 76.0 (3F, s); LRMS (EI) C7H7O2F5: 83 (12), 55
(100).
2
3
À140.5 (2F, dd, JFÀF ˆ 174:0 Hz, JHÀF ˆ 15:3 Hz);
LRMS (EI) C7H10O2F2: 105 (100), 43 (100).
3.2.19. Buten-3-yl benzoate, 13
3.2.16. 3,3-Difluoropent-4-en-2-ol
To a mixture of 3-hydroxybutene (25.6 g, 0.355 mol),
triethylamine (48 g, 0.474 mol), and dry ethyl ether
(100 ml), a 32.0 g (0.228 mol) portion of benzoyl chloride
was added, dropwise. The reaction mixture was heated to
re¯ux and allowed to react overnight. A 150 ml portion of
ethyl acetate was added to the reaction mixture, which was
then washed with 0.5 M HCl and brine and dried over
MgSO4. The solvent was removed, and the crude product
was distilled (95 8C at 1 mmHg) giving 22.4 g of buten-3-yl
Acid washed zinc (0.40 g, 6.1 mmol) and acetaldehyde
(0.20 g, 4.5 mmol) in 3.0 ml of dry THFwere combined in a
dry reaction vessel under nitrogen. After the mixture was
cooled to 0 8C, a mixture of 1-bromo-1,1-di¯uoroprop-2-ene
(0.48 g, 3.1 mmol) (obtained from Oakwood Products), in
1.5 ml of dry THFwas added slowly to the zinc/aldehyde
mixture via a syringe. The reaction was cooled to room
temperature and stirred overnight. To the reaction mixture

M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
49
benzoate, 13 (57%). 1H NMR d: 8.07 (d, 2H), 7.5 (m, 3H),
3.3. Solvolysis procedures and kinetics
5.95 (m, 1H), 5.60 (m, 1H), 5.38 (d, 3J ˆ 17:2 Hz, 1H), 5.25
3
(d, J ˆ 10:5 Hz, 1H), 1.43 (d, 3H); 13C NMR d: 166.0,
3.3.1. Acetolysis of (2,2-difluoro-3-
methylcyclopropyl)methyl tosylate (20)
137.9, 133.1, 130.8, 129.8, 128.6, 116.1, 71.8, 20.3; HR-EI
C11H12O2. Calcd.: 176.0837, Found: 176.0873.
A capillary tube charged with (2,2-di¯uoro-3-methylcy-
clopropyl)methyl tosylate (20) (5.4 mg), HOAc (40 ml), and
Ac2O (7.9 ml) was maintained at 96.3 (Æ0.2) 8C in a Varian
VXR-500 NMR. The progression of the reaction was mon-
itored by19FNMRat282 MHzbyplacingthe sealed capillary
tube in a dry NMR tube, adding CDCl3, and obtaining a
spectrum every 15 min with 1.2 s of acquisition time. During
the course of the reaction, four major products were observed.
The doublet at À75.8 ppm was assigned as 1,1-di¯uoro-2-
methylbut-3-en-1-yl tosylate (22); the AB patterns centered at
À80.3, À110.1, and À140.5 ppm were assigned as 1,1-
di¯uoro-2-methylbut-3-en-1-yl acetate (21), 3,3-di¯uoro-
pent-4-en-2-yl acetate (23), (2,2-di¯uoro-3-methylcyclopro-
pyl)methyl acetate (24), respectively. The disappearance of
starting material followed ®rst-order kinetics (Figs. 6 and 7).
3.2.20. 1-(2,2-Difluorocyclopropyl)ethyl benzoate, 14
To 8.65 g (49.1 mmol) of the 13 and 13 mg (0.3 mmol) of
NaFat 120 8C, 2.7 equivalents of TFDA were added via a
pyrex syringe equipped with a Te¯on1 needle using a
syringe pump at a rate of 0.49 ml/h resulting in an equal
mixture of diastereomers of 14 (66%). 1H NMR d: 8.06 (m,
2H), 7.56 (m, 1H), 7.45 (m, 2H), 5.04 (m, 1/2H), 4.98 (m, 1/
2H), 1.82±1.21 (m, 6H); 13C NMR d: 166.2, 165.7, 133.3,
133.3, 130.5, 130.4, 129.9, 128.6, 128.6, 113.2 (t,
1
1JCÀF ˆ 284 Hz), 113.0 (t, JCÀF ˆ 283 Hz), 69.7, 69.4,
2
2
27.7 (t, JCÀF ˆ 10 Hz), 27.1 (t, JCÀF ˆ 11 Hz), 21.1,
2
2
19.6, 16.5 (t, JCÀF ˆ 10 Hz), 14.7 (t, JCÀF ˆ 12 Hz);
19FNMR d: À127.7 (dt, JFÀF ˆ 162 Hz, JHÀF ˆ 15 Hz,
2
3
2
3
1/2F), À129.0 (dt, JFÀF ˆ 162 Hz, JHÀF ˆ 13 Hz, 1/2F),
2
3
À142.4 (dd, JFÀF ˆ 162 Hz, JHÀF ˆ 13 Hz, 1/2F).
3.3.2. Trifluoroacetolysis of (2,2-difluoro-3-
methylcyclopropyl)methyl tosylate (20)
2
3
À144.3 (dd, JFÀF ˆ 162 Hz, JHÀF ˆ 13 Hz, 1/2F);
HRMS-EI (C12H12F2O2). Calcd.: 226.0805, Found:
226.0715.
A capillary tube (sealed) with (2,2-di¯uoro-3-methylcy-
clopropyl)methyl tosylate (6.2 mg) and tri¯uoroacetic acid
(50 ml) was maintained at 50.7 (Æ0.2) 8C in an oil bath. The
progression of the reaction was monitored by 19FNMR by
placing the sealed capillary tube in a dry NMR tube and
solvating with CDCl3. During the course of the reaction, two
major products were observed. The doublet of doublet signal
at À75.4 ppm in the 19FNMR was assigned to 1,1-di¯uoro-
2-methylbut-3-en-1-yl tosylate (22) and the doublet of
doublet signal at À81.0 ppm was assigned to 1,1-di¯uoro-
2-methylbut-3-en-1-yl tri¯uoroacetate (25). The ratio of
25:22 was found to be 1:2.
3.2.21. 1-(2,2-Difluorocyclopropyl)ethanol
1-(2,2-Di¯uorocyclopropyl)ethyl benzoate (14) (2.10 g,
9.3 mmol) in 20 ml 3 M aqueous NaOH was heated to re¯ux
overnight. Extraction with ethyl ether, washing with brine,
drying with anhydrous MgSO4, and removal of ethyl ether
1
resulted in 0.72 g of the alcohol (63% yield). H NMR d:
3.65 (m, 1/2H), 3.60 (m, 1/2H), 1.70 (bs, 1H), 1.65±1.08 (m,
1
6H); 13C NMR d: 114.1 (t, JCÀF ˆ 282 Hz), 113.5 (t,
2
1JCÀF ˆ 282 Hz), 67.1, 66.4, 29.9 (t, JCÀF ˆ 10 Hz),
2
2
29.7 (t, JCÀF ˆ 9:6 Hz), 23.7, 22.0, 15.3 (t, JCÀF
ˆ
10 Hz), 14.6 (t, JCÀF ˆ 9:6 Hz); 19FNMR d: À127.6
(dtd, 2JFÀF ˆ 162:8 Hz, 3JHÀF ˆ 12:8, 4.3 Hz, 1/2F), À128.6
3.3.3. Solvolysis of 1-(2,2-difluorocyclopropyl)ethyl
tosylate, 27
2
2
3
(dtd, JFÀF ˆ 162:8 Hz, JHÀF ˆ 12:8, 4.3 Hz, 1/2F),
In a 1.1±1.2 mm i:d: Â 10 mm pyrex capillary tube,
5.7 mg (0.021 mmol) of tosylate 27 was added to 5.9 mg
of acetic anhydride, 0.5 mg of a,a,a-tri¯uorotoluene, and
40 ml glacial acetic acid. The reaction tube was sealed using a
¯ame and placed in an NMR tube along with CDCl3. In a
Varian VXR 300 MHz NMR spectrometer, the sample was
heated to 40.0 8C where a 19FNMR spectrum was recorded
every 30 min to measure the disappearance of starting mate-
rial and the appearance of product. Four major ring-opened
products were observed in the 19FNMR. Based upon the
parent system, the triplets (J ˆ 13 Hz) at À67.9 and À68.3
ppm were assigned as cis- and trans-1,1-di¯uoropent-3-en-1-
yl tosylate, 29, respectively. The triplets (J ˆ 13 Hz) at
À71.1 and À71.6 ppm were assigned as cis- and trans-
1,1-di¯uoropent-3-en-1-yl acetate, 28, respectively.
2
3
À143.3 (dd, JFÀF ˆ 162:3 Hz, JHÀF ˆ 12:8 Hz, 1/2F),
2
3
À145.3 (dd, JFÀF ˆ 162:3 Hz, JHÀF ˆ 12:8 Hz, 1/2F);
MS-EI m/z (relative intensity): 105 (M À H2O, 100).
3.2.22. 1-(2,2-Difluorocyclopropyl)ethyl tosylate, 27
Under nitrogen, a 1.56 g (8.2 mmol) portion of tosyl
chloride was added to a mixture of 1-(2,2-di¯uorocyclopro-
pyl)ethanol (4.1 mmol) in pyridine (7.5 ml) at 0 8C for 5 h.
A 20 ml portion of ether was added to the reaction mixture
and washed with cold water and then 5% NaHCO3 followed
by 0.5 M HCl. The resulting organic mixture was dried over
sodium sulfate. The resulting tosylate 27 was obtained by
the removal of ether. 1H NMR d: 7.80 (d, 2H), 7.41 (d, 2H),
3.50 (m, 1H), 2.50 (s, 3H), 1.5±1.1 (m, 6H); 19FNMR d:
2
3
ꢀ
ꢁ
À127.5 (dt, JFÀF ˆ 162:8 Hz, JHÀF ˆ 12:8 Hz, 0.6F),
C0
Ct
2
3
ln
ˆ ꢀ1:64E À 05 Æ 2:3E À 07†t
À ꢀ2:08E À 03 Æ 4:3E À 03†
À129.7 (dt, JFÀF ˆ 162:8 Hz, JHÀF ˆ 12:8 Hz, 0.4F),
2
3
À141.9 (dd, JFÀF ˆ 162:3 Hz, JHÀF ˆ 12:8 Hz, 0.6H),
2
3
(1)
À143.8 (dd, JFÀF ˆ 162:3 Hz, JHÀF ˆ 12:8 Hz, 0.4H).

50
M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
Fig. 6. Acetolysis of (2,2-difluoro-3-methylcyclopropyl)methyl tosylate at (96.3 (Æ0.1) 8C).
Fig. 7. Plot of kinetic data for solvolysis of 27.
[3] M. Newcomb, D.L. Chestney, J. Am. Chem. Soc. 116 (1994) 9753±
9754.
Acknowledgements
[4] K.M. Koshy, T.T. Tidwell, J. Am. Chem. Soc. 102 (1980) 1216±
1218.
The support of this work in part by the National Science
Foundation is gratefully acknowledged.
[5] R.J. Blint, T.B. McMahon, J.L. Beauchamp, J. Am. Chem. Soc. 96
(1974) 1269±1278.
[6] F. Tian, M.A. Battiste, W.R. Dolbier Jr., Org. Lett. 1 (1999) 193±196.
[7] Y. Bessard, L. Kuhlmann, M. Schlosser, Tetrahedron 46 (1990)
5230±5236.
References
[8] D.L. Boger, T.J. Jenkins, J. Am. Chem. Soc. 118 (1996) 8860±8870.
[9] K.B. Wiberg, A.J. Ashe Jr., J. Am. Chem. Soc. 90 (1968) 63±74.
[10] I.L. Reich, A. Diaz, S. Winstein, J. Am. Chem. Soc. 91 (1969) 5635±
5637.
[1] F. Tian, W.R. Dolbier Jr., Org. Lett. 2 (2000) 835±837.
[2] F. Tian, M.D. Bartberger, W.R. Dolbier, J. Org. Chem. 64 (1999)
540±546.

M.A. Battiste et al. / Journal of Fluorine Chemistry 119 (2003) 39±51
51
[11] M.J.T. Frisch, G.W.H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R.
Cheeseman, V.G. Zakrzewski, J. Montgomery, J.A.R.E. Stratmann,
J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin,
M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B.
Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A.
Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D.
Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz,
A.G. Baboul, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-
Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,
P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L. Andres, C.
Gonzalez, M. Head-Gordon, E.S. Replogle, J.A. Pople, Gaussian 98,
Revision A.97, Gaussian Inc., Pittsburgh, PA, 1998.
[12] A.D. Becke, J. Chem. Phys. 98 (1993) 5648±5652.
[13] P.C. Hariharan, J.A. Pople, Theor. Chem. Acta 28 (1973) 213±
222.
[14] K. Raghavachari, G.W. Trucks, J. Chem. Phys. 91 (1989) 1062±
1065.