Impact of fluorine substituents on the rates of nucleophilic aliphatic substitution and β-elimination

Article abstract of DOI:10.1016/j.jfluchem.2011.10.008

A measure of the quantitative effect of proximate fluorine substituents on the rates of S_N2 and E2 reactions has been obtained through a study mainly of reactions of fluorinated n-alkyl bromides with weak base, strong nucleophile azide ion and strong base/nucleophile methoxide ion in the protic solvent methanol and the aprotic solvent, DMSO. The order of reactivity for S_N2 reactions of azide in methanol at 50 °C was found to be: n-alkyl-Br > n-alkyl-CHFBr > n-perfluoroalkyl-CH₂CH ₂Br n-perfluoroalkyl-CH₂Br > n-alkyl-CF₂Br. Approximate relative rates of reaction were: 1, 0.20, 0.12, 1 × 10 ^-4, <7.7 × 10^-5. The order of reactivity for E2 reactions was found to be: n-perfluoroalkyl-CH₂CH₂Br n-alkyl-CF₂Br > n-alkyl-CHFBr > n-alkyl-Br. The approximate relative rates for reaction of methoxide in methanol at 50 °C were: 1100, 4.4, 1.9, 1.

Full text of DOI:10.1016/j.jfluchem.2011.10.008

Journal of Fluorine Chemistry 135 (2012) 167–175
Contents lists available at SciVerse ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Impact of fluorine substituents on the rates of nucleophilic aliphatic substitution
and
b-elimination
*
Henry Martinez, Adele Rebeyrol, Taylor B. Nelms, William R. Dolbier Jr
Department of Chemistry, PO Box 117200, 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:
A measure of the quantitative effect of proximate fluorine substituents on the rates of SN2 and E2
reactions has been obtained through a study mainly of reactions of fluorinated n-alkyl bromides with
weak base, strong nucleophile azide ion and strong base/nucleophile methoxide ion in the protic solvent
methanol and the aprotic solvent, DMSO. The order of reactivity for SN2 reactions of azide in methanol at
50 8C was found to be: n-alkyl-Br > n-alkyl-CHFBr > n-perfluoroalkyl-CH2CH2Br ꢁ n-perfluoroalkyl-
Received 24 August 2011
Received in revised form 8 October 2011
Accepted 11 October 2011
Available online 18 October 2011
CH2Br > n-alkyl-CF2Br. Approximate relative rates of reaction were: 1, 0.20, 0.12, 1 ꢂ 10ꢃ4, <7.7 ꢂ 10ꢃ5
.
This paper is dedicated to the memory
of Prof. Merle A. Battiste, a great friend,
collaborator and true scholar, who loved
organic chemistry, and in particular
loved kinetics.
The order of reactivity for E2 reactions was found to be: n-perfluoroalkyl-CH2CH2Br ꢁ n-alkyl-
CF2Br > n-alkyl-CHFBr > n-alkyl-Br. The approximate relative rates for reaction of methoxide in
methanol at 50 8C were: 1100, 4.4, 1.9, 1.
ß 2011 Elsevier B.V. All rights reserved.
Keywords:
Nucleophilic aliphatic substitution
b-Elimination
Williamson synthesis
Effect of fluorine substitution
1. Introduction
nucleophilic substitutions on b-fluorinated substrates such as
CF3CH2X or RFCH2X requires rather harsh conditions, even with
relatively good nucleophiles, such as azide (Scheme 2) [3].
When a substrate’s fluorine substituents are two carbons or
more from the leaving group they are considered to have SN2
It is well recognized that nucleophilic aliphatic substitution
reactions of primary alkyl substrates proceed via the SN2
mechanism, and it is likewise recognized, perhaps not so
universally, that primary perfluoroalkyl substrates will not
participate in substitution reactions via the SN2 mechanism, but
rather are capable of undergoing nucleophilic substitution via the
reactivities ‘‘similar’’ to those of simple n-alkyl halides, whereas g-
fluorinated substrates also have significantly enhanced E2 reactiv-
ities. Such considerations are important to recognize when one is
designing syntheses of fluorous compounds using n-RFCH2CH2X
substrates.
S
RN1 mechanism when the nucleophile can undergo SET to initiate
the free radical chain process [1]. Indeed, simply the presence of
two -fluorines in substrates with the structure RCH2CF2X is
a
There are two early papers that provide some quantitative
understood to severely inhibit, if not prevent nucleophilic
substitution. To our knowledge there are no examples of successful
SN2 nucleophilic substitution reactions of such compounds. On the
other hand, Haas was able to observe apparent SN2 substitution
reactions with benzylic, arylCF2Br compounds (Scheme 1) [2].
It is also part of the ‘‘lore’’ of fluorine chemistry that fluorine
substituents proximate to the site of nucleophilic substitution will
inhibit such substitution, and the closer they are, the more impact
that they have. It is known, for example that carrying out
insight into the effects of fluorine substitution at the
positions of an alkyl halide on the rates of nucleophilic substitu-
tion. Hine examined the reaction of phenyl thiolate with
fluorinated ethyl iodide, [4] whereas McBee studied the reactions
b and g
b
-
of iodide ion with
[5].
b- and g-fluorinated alkyl bromides (Scheme 3)
Hine also provided the only data that is available on the effect of
-fluorine substitution with his study of the SN2 reactivity of
a
bromofluoromethane, which proved to be ꢀ350 times less reactive
than methylbromide in its reaction with iodide ion in acetone at
20 8C [6].
Kinetic data on
b-eliminations of fluorinated substrates is even
* Corresponding author. Fax: +352 846 1962.
E-mail address: wrd@chem.ufl.edu (W.R. Dolbier Jr).
more limited than that for substitution. In 1981, Koch reported
0022-1139/$ – see front matter ß 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2011.10.008

168
H. Martinez et al. / Journal of Fluorine Chemistry 135 (2012) 167–175
rates of elimination of HF for the series: PhCH2-R (R = CH2F, CHF2,
CF3, CF2Cl and CF2CF3) [7]. Their results were interpreted in terms
of variable degrees of E1cb character for the reactions, which
largely involved fluoride as the leaving group. While certainly
interesting and relevant, their data are not directly comparable to
data obtained in the present work.
The purpose of the present work is to provide comprehensive,
hopefully definitive kinetic data for substitution (SN2) and
Scheme 2. Literature examples of SN2 reactions of azide with
b-fluorinated
elimination (E2) reactions of
a-, b-, and g-fluorinated n-alkyl
substrates.
systems under variety of solvent, nucleophilic and basic
a
conditions, in order to provide synthetic organic chemists with
sufficient understanding of the chemistry of such systems to allow
them to plan successful syntheses when using partially fluorinated
aliphatic substrates.
the initial concentrations of both reactants, while A and B are the
concentrations at time t.
A + B ! C
Comparative kinetic and reaction outcome data will be
provided for the following types of partially fluorinated substrates:
n-RBr, n-RfCH2CH2Br, n-RfCH2Br, n-R-CHFBr, n-R-CF2Br, and
PhCF2Br in their reactions with two nucleophiles that have very
different character, the strongly basic nucleophile, methoxide, and
the weakly basic nucleophile, azide, the kinetics of their reactions
being examined in two solvents, the protic solvent, methanol and
the polar, aprotic solvent, DMSO.
dc
dt
dA
dt
¼ ꢃ
¼ k½Aꢄ½Bꢄ
(1)
(2)
1
AoB
Ln
¼ kt
Bo ꢃ Ao BoA
Computational results will also be presented in order to provide
a more complete understanding of the unusual kinetic effects of
fluorine substituents.
1
BoðAo ꢃ xÞ
Ln
Ao ꢃ Bo AoðBo ꢃ xÞ
¼ kt
(3)
Since experimentally it is usually easier to follow the reaction
by the decrease in the concentration (x) of one of the reactants,
Eq. (2) can be expressed in terms of x and the initial concentrations,
Ao and Bo (Eq. (3)). Given that the reaction is first order for each
reactant, the amount of x is the same in both cases. The plot
between [1/(Ao ꢃ Bo) ln ((Bo(Ao ꢃ x))/(Ao(Bo ꢃ x)) versus t gives the
rate constant directly from the slope. This rate expression has been
widely used for measuring the rate constants of second order
reactions [11].
Some of the reactions presented in this work undergo
competitive substitution and elimination reactions. Since each
reaction requires a 1:1 ratio between the substrate and the
nucleophile/base, the rate constant (kobs) derived from Eq. (3) will
2. Results and discussion
The specific substrates that were included in this kinetic study
are n-heptyl bromide (1), n-octyl iodide (2) and n-octyl tosylate (3),
1H,1H-perfluoro-n-butyl bromide (4), 1H,1H,2H,2H-perfluoro-n-
hexyl bromide (5), 1-bromo-1-fluorononane (6), 1-bromo-1,1-
difluorohexane (7), benzyl bromide (8), and bromodifluoromethyl-
benzene (9). Substrates 1, 2, 4, 5 and 8 were commercially
available. Tosylate 3 was obtained from the alcohol in the usual
manner [8]. 1-Bromo-1-fluorononane (6) was obtained essentially
by the method of Garcia Martinez (Scheme 4) [9]. 1-Bromo-1,1-
difluoro-hexane (7) was prepared as reported in our earlier study
of its radical chemistry [10].
be the sum of each of the individual rate constants, kS 2 and kE2. The
N
All of the kinetic experiments were run under second-order
reaction conditions, using two equivalents of nucleophile, and the
rates were measured using either proton or fluorine NMR to follow
the depletion of starting substrate and formation of substitution
and elimination products. In all cases, the reactions, as studied,
appeared to be quantitative, since no other products could be
detected by proton or fluorine NMR. In order to be able to compare
most of the reactions at the common temperature of 50 8C, it was
necessary to extrapolate the rate constants for very fast or very
slow reactions of some of the substrates to this temperature
through the use of the Eyring equation. Rates of such reactions at
two temperatures were used for this purpose.
ratio between the two products can then be used to obtain the
individual rate constants.
Because proton or fluorine NMR was being used to measure the
changes in concentration of substrate, it was necessary that the
particular starting material and product NMR signals be suffi-
ciently resolved to allow accurate individual integrations. This led
to some limitations with respect to reactions that could be studied.
For this reason, rates for the reactions of alkyl iodide with azide and
reactions of alkyl tosylate in DMSO could not be measured.
2.1. Kinetics
The second order character of bimolecular reactions such as SN2
and E2 has long been recognized, since the rates of such reactions
depend on the concentrations of both reactants (first order for each
one) (Eq. (1)). Integration of the rate expression results in a linear
relationship between ln (AoB/BoA) and time t (Eq. (2)). Ao and Bo are
Scheme 3. Early Hine and McBee kinetic studies.
Scheme 1. SN2 reaction of bromodifluoromethylbenzene.

H. Martinez et al. / Journal of Fluorine Chemistry 135 (2012) 167–175
169
of n-octyl iodide and n-heptyl bromide with methoxide ion
(pKa = 15.5), but interestingly no elimination was observed when
tosylate was the leaving group, either in methanol or DMSO.
Thus, based upon our admittedly limited data, tosylate, rather
than halide, would seem to be the preferred leaving group for
carrying out Williamson synthesis of ethers. Such a conclusion
does not appear to be part of the ‘‘lore’’ of the Williamson
synthesis [12]. Although reports of effective use of tosylates can be
found [13,14], including one paper involving solvent-free alkyla-
tions that clearly showed that alkyl tosylates gave much less
elimination in reactions with potassium tert-butoxide than alkyl
halides [15].
Scheme 4. Preparation of 1-bromo-1-fluorononane.
No trace of elimination products could be observed when using
the much less basic azide ion (pKa = 4.67). The kinetic data that
were obtained for these reactions of non-fluorinated n-alkyl
systems are given in Table 1.
2.2.1. Leaving group abilities, bromide versus iodide and tosylate
The relative leaving group abilities of bromide, iodide and
tosylate are compared in Table 2a. As expected, iodide is a
somewhat better leaving group than bromide in reactions with
methoxide. Such results are consistent with a reported study of the
reaction of hydroxide with methyl iodide versus methyl bromide,
where the iodide reacted slightly faster (ratio = 1.1) [16]. In studies
of SN1 reactions, the differential between iodide and bromide as
leaving groups is considerably greater, an example being the
solvolysis of phenylethyl iodide versus bromide in 80% ethanol at
75 8C where the rate ratio was 6.5 [17]. Tosylate was a better
leaving group than bromide in the methoxide/methanol reaction.
To our knowledge there is no literature comparison of tosylate
with bromide in an SN2 reaction. However, in the reaction of
hydroxide with methyl bromide and mesylate in water [16], the
mesylate was 3.6 faster. In the phenylethyl solvolysis study
mentioned just above, mesylate was found to be only slightly
Scheme 5. Relative amounts of substitution versus elimination in reactions of n-
alkyl iodide, bromide and tosylate with methoxide.
2.2. Hydrocarbon substrates
Kinetic data that directly compare leaving group abilities and
nucleophilicities for SN2 reactions of alkyl halides and tosylates in
protic and aprotic solvents can, of course, be found in the old
literature, but we believe that it is worthwhile to discuss these
factors as they have revealed themselves in this study. The
observed overall results for reactions of methoxide with n-alkyl
halides and tosylates are given in Scheme 5. As expected,
elimination was found to compete with substitution for reactions
Table 1
Second order rate constants for n-alkyl iodide, bromide and tosylate and for benzyl bromide.
Expt. No
Substrate
Solvent
Nucleophile
T (8C)
105k2 SN2
6.190.24
105k2 E2
1
2
n-C8H17Ia
CH3OH
DMSO
DMSO
DMSO
CH3OH
DMSO
DMSO
DMSO
CH3OH
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
CH3OH
CH3OH
CH3OH
CH3OH
CH3OH
DMSO
CH3OH
CH3OH
CH3OH
DMSO
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
50
23
33
50
50
18
23
50
50
18
23
50
25
30
50
50
50
15
25
50
15
15
25
50
15
0.4690.016
9.530.42
21.51.1
76.813
0.5660.032
3.480.075
5.290.009
40.41.1
noc
n-C8H17
n-C8H17
n-C8H17
I
I
I
35.71.4
79.33.7
3
4b
27641
5
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C7H15Br
n-C8H17OTs
n-C8H17OTs
Ph-CH2Br
Ph-CH2Br
Ph-CH2Br
Ph-CH2Br
Ph-CH2Br
Ph-CH2Br
Ph-CH2Br
Ph-CH2Br
3.950.24
9.560.20
14.60.064
6
7
8b
11615
6.890.064
1013.4
ꢃ
9
N3ꢃ
10
11
12b
13
14
15b
16d
17d
18
19
20b
21
22
23
24b
25f
N3ꢃ
N3ꢃ
N3
no
1623.5
no
1590300
no
Iꢃ
61.82.6
99.03.7
563140
no
Iꢃ
no
Iꢃ
no
CH3Oꢃ
12.70.70
23.81.0
13.50.59
30.51.8
no
ꢃ
N3
no
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
–
e
–
–
–
–
–
–
–
18819
102074
46.60.64
ꢃ
N3ꢃ
N3ꢃ
N3ꢃ
N3
1031.4
60620
7480250
a
The kinetics of reactions of n-C7H15I with azide could not be studied because of overlap of relevant peaks in the proton NMR.
b
These 2nd order rate constants were approximated for comparison purposes using the Eyring equation; the given error was obtained by propagating that of the two data
points.
c
no, not observed.
d
e
f
Reactions of n-C8H17OTs in DMSO could not be studied kinetically because of overlap of relevant peaks in the proton NMR.
– (n-dash) means product not possible.
Approximate, single point rate constant (56.8% of substrate consumed after 30 s).

170
H. Martinez et al. / Journal of Fluorine Chemistry 135 (2012) 167–175
Table 2a
Relative leaving group abilities in SN2 reactions at 50 8C.
Substrate
krel (CH3Oꢃ/CH3OH)
krel (CH3Oꢃ/DMSO)
krel (N3ꢃ/CH3OH)
krel (N3ꢃ/DMSO)
n-C7H15Br
1
1
1
1
–
–
n-C8H17
I
1.6
3.2
2.4
–
–
n-C8H17OTs
3.5
Table 2b
Relative nucleophilicities of azide versus methoxide in SN2 reactions.
Substrate
krel (CH3OH)
krel (DMSO)
n-C7H15Br (50 8C)
n-C8H17OTs (50 8C)
PhCH2Br (15 8C)
1.7
1.9
3.5
13.7
–
7.3
Scheme 6. Reactions of 1H,1H,2H,2H-perfluoro-n-hexyl bromide with methoxide
and azide.
Table 2c
Relative rate constants for reactions in DMSO versus methanol.
methanol (Table 2c), but rates for azide were affected more greatly
than those of methoxide. Thus, as expected from previous
literature reports [18,19], the polar aprotic solvent DMSO
facilitates SN2 reactions better than does the polar protic solvent,
methanol. Fuchs and Cole reported a rate ratio of 81 for the
reaction of azide with n-hexyl tosylate at 40 8C for DMSO versus
methanol [18]. The ratio that we observed for the analogous
reaction of azide with n-octyl tosylate at 50 8C was 54.
Also observed was a significant difference on the substitution/
elimination competition for the two solvents, with elimination
being significantly more competitive in DMSO (7–13% elimination
being observed in methanol as compared with 22–27% in DMSO).
The rate of elimination for the reactions of methoxide with n-
heptyl bromide and n-octyl iodide at 50 8C were 71 and 164 times
faster, respectively, in DMSO than in CH3OH.
One last insight can be gleaned from the data in Table 1, by
comparing the rates of benzyl bromide with those of n-heptyl
bromide. Benzyl bromide is considerably more reactive than the
primary alkyl bromide, 48 times faster in the reaction with
methoxide in methanol and 88 times faster with azide the
nucleophile in methanol, both at 50 8C. This can be compared with
a value of 78 given by Streitwieser for the reaction of iodide in
ethanol [20].
Although similar comparative rate data can be found scattered
throughout the early literature of nucleophilic substitution and
elimination, the above data on SN2 and E2 reactions of primary
alkyl halides, to our knowledge, are not replicated in the literature,
and they provide convenient access to quantitative information
related to leaving group ability, nucleophilicity and solvent effects
that can be useful for both teaching and research.
Substrate
n-C8H17
Nucleophile
Temp, 8C
krel
45
I
CH3Oꢃ
CH3Oꢃ
50
50
50
15
15
n-C7H15Br
n-C7H15Br
PhCH2Br
PhCH2Br
29
231
76
ꢃ
N3
CH3Oꢃ
ꢃ
N3
161
poorer than tosylate (ratio = 1.2) [17], so the methanol results in
Table 2a are more or less consistent with the literature.
2.2.2. Effect of the nucleophile, methoxide versus azide
Azide is a better nucleophile than methoxide in both protic and
aprotic solvents, but the difference is much greater in the aprotic
solvent (DMSO) (Table 2b). As expected, elimination competed
with substitution when methoxide was the nucleophile (Scheme
5), but no elimination product was observed when azide was the
nucleophile. Methoxide’s greater basicity in DMSO than in
methanol is reflected by the significantly greater amount of
elimination observed in DMSO. Iodide was examined as
a
nucleophile in order to provide a comparison of our results with
those of McBee [5]. It was a better nucleophile in DMSO than
methoxide (krel = 4.9), but not as good as azide (krel = 0.35). Azide
was not sufficiently soluble in acetone for rates to be measured.
Comparing McBee’s rate constants in acetone versus ours in DMSO,
acetone appears to be a slightly better solvent for SN2 reactions of
iodide ion.
2.2.3. Solvent effect, methanol versus DMSO
Lastly, it is to be noted that the SN2 reactions of both
nucleophiles were much faster in DMSO than they were in
Table 3
Second order rate constants for 1H,1H,2H,2H-perfluoro-n-hexyl bromide.
Expt. No.
Substrate
Solvent
Nuc
T (8C)
105k2 SN2
105k2 E2
5
26
27
28c
6
n-C7H15Br
CH3OH
CH3OH
CH3OH
CH3OH
DMSO
DMSO
CH3OH
CH3OH
DMSO
DMSO
DMSO
DMSO
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
50
10
20
50
18
18
50
50
50
18
23
50
3.950.24
noa
0.5660.032
45.31.0
93.52.2
63161
3.480.075
8.580b
noa
n-C4F9CH2CH2Br
n-C4F9CH2CH2Br
n-C4F9CH2CH2Br
n-C7H15Br
no
no
9.560.20
no
29
9
n-C4F9CH2CH2Br
n-C7H15Br
ꢃ
N3ꢃ
6.890.064
0.8230.056
1590
30
12c
31
32
33c
n-C4F9CH2CH2Br
n-C7H15Br
N3ꢃ
N3ꢃ
N3ꢃ
N3ꢃ
N3
no
no
n-C4F9CH2CH2Br
n-C4F9CH2CH2Br
n-C4F9CH2CH2Br
8.530.31
14.90.64
22771
no
no
no
a
no, not observed.
b
Calculated on the basis that 61% of starting material had reacted after 30 s of reaction.
c
These 2nd order rate constants were approximated for comparison purposes using the Eyring equation; the given error was obtained by propagating that of the two data
points.

H. Martinez et al. / Journal of Fluorine Chemistry 135 (2012) 167–175
171
Table 5
Second order rate constants for 1-bromo-1-fluorononane.
Expt. No. Substrate
Solvent Nuc
CH3OH CH3Oꢃ 50
T (8C) 105k2 SN2 105k2 E2
Scheme 7. Reactions of azide with 1H,1H-perfluoro-n-butyl bromide.
5
36
8b
n-C7H15Br
3.950.24
noa
0.5660.032
n-C8H17CHFBr CH3OH CH3Oꢃ 50
1.100.02
40.41.1
50.52.6
no
n-C7H15Br
DMSO
CH3Oꢃ 50
11615
no
2.3. Fluorinated substrates
37
9
n-C8H17CHFBr DMSO
n-C7H15Br
CH3Oꢃ 50
ꢃ
CH3OH N3ꢃ
50
50
50
30
40
50
6.890.064
1.420.04
1590300
63.41.0
1221.7
2265.3
38
12b
39
40
41b
n-C8H17CHFBr CH3OH N3ꢃ
no
2.3.1.
g-Fluorinated substrates
n-C7H15Br
DMSO
N3ꢃ
N3ꢃ
N3ꢃ
N3
no
In terms of chemoselectivity, the
g
-fluorine-substituted sub-
n-C8H17CHFBr DMSO
n-C8H17CHFBr DMSO
n-C8H17CHFBr DMSO
no
strate, n-C6F13CH2CH2I, exhibited completely different reactivity
with the two nucleophiles, methoxide and azide, as seen in Scheme
6. In its reactions with the strongly basic nucleophile, methoxide,
the RfCH2CH2Br substrate underwent exclusive elimination both in
methanol and in DMSO, whereas when using weakly basic
nucleophile, azide, as the nucleophile, only substitution was
observed. The rate data for this series of compounds are given
in Table 3.
This series of reactions allows us to make a direct comparison of
substitution and elimination rates of n-alkyl versus perfluoro-n-
alkylethyl bromides. The rate constants for elimination were much
larger for methoxide reacting with perfluoro-n-butylethyl bromide
than for its reaction with n-heptyl bromide (krel = 1115 in
methanol at 50 8C, and 2466 in DMSO at 18 8C). Comparing the
rates of substitution for n-heptyl bromide with those of n-
C4F9CH2CH2Br at 50 8C, in methanol the hydrocarbon bromide
underwent substitution 8.4 times faster than the perfluoroalk-
ylethyl bromide, whereas in DMSO the rate ratio was 7.0. This
compares with McBee’s ratio of 6.1 for the reaction of iodide in
acetone at 35 8C [5]. Thus, SN2 chemistry can be observed relatively
cleanly for perfluoroalkyethyl bromides when using a good
nucleophile that is a weak base, although the presence of the
perfluoroalkyl group has a significant damping effect upon the
rates of substitution. On the other hand, if the good nucleophile is
also a strong base, as in the case of methoxide, the rates of
elimination are increased to such an extent that there appears to be
little chance of observing any significant amount of substitution.
no
no
a
no, not observed.
b
These 2nd order rate constants were approximated for comparison purposes
using the Eyring equation; the given error was obtained by propagating that of the
two data points.
2.3.3.
As indicated by the data given in Table 5, a single fluorine
substituent at the -position inhibits substitution while acceler-
ating elimination, although both effects are modest. These effects
combine to give rise to exclusive elimination when methoxide is
used as nucleophile/base, and to exclusive substitution when azide
is used (Scheme 8).
In methanol, methoxide eliminated HBr from 1-bromo-1-
fluorononane 1.9 times faster than from n-heptyl bromide,
whereas in DMSO the ratio was 1.3. Elimination of the
compound in DMSO was itself 46 times faster than it was in
methanol.
Substitution of 1-bromo-1-fluorononane by azide ion, on the
other hand, was modestly slower than of 1-bromoheptane, the
non-fluorine-substituted compound undergoing reaction 4.9 times
a-Fluoro substrate
a
a-fluoro
faster in methanol and 7.0 times faster in DMSO than its
a-fluoro
analog. The azide substitution reaction of the 1-bromo-1-
fluoroalkane is 159 times faster in DMSO than it is in methanol.
No product deriving from replacement of fluoride was observed for
either the elimination or the substitution reactions.
2.3.2.
b-Fluorinated substrates
2.3.4.
a,a-Difluoro substrate
Substitution was the only process observed for reactions of
azide with 1H,1H-perfluoro-n-butyl bromide in either methanol or
DMSO (Scheme 7). However, the analogous reactions with
methoxide did not lead to any observable products of substitution.
No single product appeared to be formed in the methoxide
reactions, which although not yet characterized, may involve some
No substitution involving displacement of bromide from the
CF2Br group of 1-bromo-1,1-difluorohexane by either methoxide
or azide ion was observed in either solvent, and for the first time,
when the reaction was carried out in DMSO, azide was observed to
give rise to exclusive elimination (Scheme 9; Table 6).
Whereas substitution was strongly inhibited, elimination was
enhanced. The rate of
combination of
a- and b-elimination processes. There is precedent
b-elimination of n-pentyl-CF2Br was
for -elimination of fluoride in a related system [21]. The
b
enhanced relative to that of n-alkyl bromide, loss of HBr occurring
4.4 times faster in methanol and 59 times faster in DMSO at 50 8C.
With no substitution being observed for the reaction of azide in
methanol after 20 days at 50 8C, it was possible to calculate a
maximum rate for the SN2 process if one assumed that one could
detect a minimum of 0.5% of product in Expt. No. 46. Therefore the
SN2 reaction at a CF2Br site is calculated to be at least 13,000 times
slower than that at a CH2Br site of a primary alkyl bromide.
comparative rate data for the azide reactions can be found in
Table 4.
The rates for SN2 substitution of 1H,1H-perfluoro-n-butyl
bromide by azide were tremendously inhibited by the proximity
of the perfluoroalkyl group, undergoing reaction at 50 8C almost
10,000 times slower in methanol and more than 37,000 times
slower in DMSO.
Table 4
Second order rate constants for 1H,1H-perfluoro-n-butyl bromide.
Expt. No.
Substrate
Solvent
Nuc
T (8C)
105k2
ꢃ
9
34
n-C7H15Br
CH3OH
CH3OH
DMSO
DMSO
N3ꢃ
50
50
50
50
6.890.064
n-C3F7CH2Br
n-C7H15Br
N3ꢃ
N3ꢃ
N3
6.900.25 ꢂ 10ꢃ4
12a
35
1590300
4.230.32 ꢂ 10ꢃ2
n-C3F7CH2Br
a
This 2nd order rate constant was approximated for comparison purposes using
the Eyring equation; the given error was obtained by propagating that of the two
data points.
Scheme 8. Reactions of 1-bromo-1-fluorononane with methoxide and azide.

172
H. Martinez et al. / Journal of Fluorine Chemistry 135 (2012) 167–175
Scheme 10. Reactions of bromodifluoromethylbenzene with methoxide and azide.
Table 7
Second order rate constants for bromodifluoromethylbenzene.
Scheme 9. Reactions of 1-bromo-1,1-difluorohexane with methoxide and azide.
Expt. No.
Substrate
Solvent
Nuc
T (8C)
105k2 SN2
2.3.5. Fluorinated benzylic bromides
20a
48
PhCH2Br
PhCF2Br
PhCH2Br
PhCF2Br
PhCF2Br
PhCF2Br
PhCH2Br
PhCF2Br
PhCH2Br
PhCF2Br
PhCF2Br
PhCF2Br
CH3OH
CH3OH
DMSO
DMSO
DMSO
DMSO
CH3OH
CH3OH
DMSO
DMSO
DMSO
DMSO
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
CH3Oꢃ
50
50
15
60
50
15
50
50
15
60
50
15
18846
4.2930 ꢂ 10ꢃ3
Nucleophilic substitution at the CF2Br site of PhCF2Br has been
reported [2], and thus a study of this system where elimination
cannot compete should be able to provide a more exact measure of
the difference in SN2 reactivity of a CF2Br versus a CH2Br site. In
fact, both methoxide and azide were found to be effective
nucleophiles in their reactions with PhCF2Br (Scheme 10; Table 7).
From the data in Table 8, as well as in Table 6, one can see that
21
102074
49
1.250.045
50
0.5490.024
1.960.58 ꢂ 10ꢃ2
60620
51a
24a
52
ꢃ
N3ꢃ
N3ꢃ
N3ꢃ
N3ꢃ
N3ꢃ
N3
8.880.32 ꢂ 10ꢃ3
25b
53
7480250
2.340.11
the presence of two a-fluorines provides a formidable inhibition of
54
55a
1.060.058
4.351.8 ꢂ 10ꢃ2
SN2 chemistry, such that E2 chemistry will exclusively dominate,
except in those cases, such as in the benzyl system, where E2 is not
possible.
a
These 2nd order rate constants were approximated for comparison purposes
using the Eyring equation; the given error was obtained by propagating that of the
two data points.
b
2.4. Computational results
Approximate, single point rate constant (56.8% of substrate consumed after
30 s).
As can be witnessed by examination of any undergraduate
textbook, the bimolecular nucleophilic substitution reaction, SN2,
has been exhaustively studied, with the result that influences on
reactivity such as substrate structure, leaving group ability,
nucleophile effects, and solvent effects are considered to be well
understood. Each of these effects contributes individually, and
sometimes cooperatively, in such a way as to either increase or
decrease the reaction barrier. In the experimental part of this
paper, we have quantified the effect that near-by fluorines have on
raising the SN2 activation barrier, as well as their effect of lowering
the E2 energy.
electrostatic contribution. According to Liu’s work, the steric effect
is described by the Weizsacker kinetic energy (which is related to
the ordinary kinetic energy), the quantum effect corresponds
mainly to the exchange-correlation contribution, whereas the
electrostatic effect is defined by a combination of (a) nuclear–
electron attraction, (b) classical electron–electron Coulomb
repulsion and (c) nuclear–nuclear repulsion energies.
Consistent with general experimental observations, Lui’s
results show that the exchange of one hydrogen on the CH3Cl
by almost any other group leads to an increase in the energy
barrier due to an increase in steric interactions, but that this is
compensated by a decrease due to the quantum effect, with the
result that the observed net overall increase in barrier is due
largely to the impact of increasing electrostatic effect. Liu finds
that fluorine substituents do not fit this generalization. He does
find that the substitution of hydrogens by one or more fluorines
increases the energy barrier, but the steric effect was found to
decrease significantly for each hydrogen/fluorine replacement,
with this decrease being compensated now by an equivalent
increase in barrier from the quantum effect. Again, the net result
was that the overall increase in energy barrier derived largely
from the incremental increases in the electrostatic effect as one
replaced hydrogen with fluorine, the largest contribution to this
increase being the classical electron–electron Coulomb repulsive
interaction.
A recent computational paper by Liu et al. examined the effect
of substituents on the SN2 energy barrier [22]. Examining 54
different structures using CH3Cl as a reference and fluoride as a
nucleophile, Liu broke down the observed effects into three
contributions (Eq. (3)),
and Ee the
D
Es comprising the steric,
D
Eq the quantum,
D
D
E ¼
D
Es þ
DEq þ
D
Ee
(4)
Table 6
Second order rate constants for 1-bromo-1,1-difluorohexane.
Expt. No. Substrate
Solvent Nuc
T (8C) 105k2 SN2
105k2 E2
5
42
8b
n-C7H15Br
CH3OH CH3Oꢃ 50
3.950.24
noa
0.5660.032
2.500.016
40.4
n-C5H11CF2Br CH3OH CH3Oꢃ 50
n-C7H15Br
DMSO
CH3Oꢃ 50
116
43
44
45b
9
46c
12b
47
n-C5H11CF2Br DMSO
n-C5H11CF2Br DMSO
n-C5H11CF2Br DMSO
n-C7H15Br
CH3Oꢃ 15.5
CH3Oꢃ 20.5
no
2256.7
32811
no
CH3Oꢃ 50
no
2380660
no
ꢃ
CH3OH N3ꢃ
50
50
50
50
6.890.064
n-C5H11CF2Br CH3OH N3ꢃ
<5.2 ꢂ 10ꢃ4 no
Table 8
n-C7H15Br
DMSO
N3ꢃ
N3
1590300
no
Relative rates for SN2 reactions of PhCH2Br versus PhCF2Br.
n-C5H11CF2Br DMSO
no
1.550.037
a
Nucleophile
Solvent
T (8C)
krel (CH2Br/CF2Br)
no, not observed.
b
These 2nd order rate constants were approximated for comparison purposes
CH3Oꢃ
CH3Oꢃ
CH3OH
DMSO
CH3OH
DMSO
50
15
50
15
43,800
52,000
68,200
172,000
using the Eyring equation; the given error was obtained by propagating that of the
two data points.
ꢃ
N3ꢃ
c
No reaction was observed after 20 days at 50 8C; the maximum rate was
N3
approximated by assuming that 0.5% of product could have been detected.

H. Martinez et al. / Journal of Fluorine Chemistry 135 (2012) 167–175
173
As seen in Table 9, this repulsive effect due to the
a
-fluorines
appears to give rise to larger distances between the nucleophile
and the carbon-center at the TS for the -fluorinated molecules
a
than for the non-fluorinated standard: C2H5CF2I > C2H5CHFI > n-
C3H7I. At the same time the length of the bond to the carbon–iodine
leaving group is also larger than that for the non-fluorinated
substrate, which is consistent with the loss of the iodide being
assisted by the a-fluorines, stabilizing the partial positive charge at
the carbon-center. The net increase in barrier indicates that the
repulsive effect outweighs the stabilizing effect, slightly for the
mono-
substituents.
Evidence for the stabilization of the partial positive charge by
a-fluoro and greatly for the substrates with two a-fluorine
a
-
fluorines in the transition state is seen by comparison of the bond
lengths between the fluorine and the carbon center in the
transition states of the
a-fluoro substrates. They are shorter than
Scheme 11. Model structures used in the quantum chemical calculations.
˚
those in the respective ground states, i.e. C–F(A) changes from 1.42
to 1.36 for the reaction of C2H5CHFI and from 1.38 to 1.32 for the
reaction of C2H5CF2I.
Further insight can be obtained by looking at the geometry of
the transition state for the reaction of C2H5CHFI, where the OH
nucleophile is seen to approach the carbon-center from an angle
In our computational study of the impact of fluorine on SN2 and
E2 transition state energies, we examined a series of model
reactions involving hydroxide acting either as a nucleophile or a
base (Scheme 11).
The quantum chemical calculations were performed at MP2/6-
31+G(d,p)-LANL2DZ level of theory using DMSO, as included in the
Gaussian software, with the transition states being characterized
by one and only one negative frequency and the Intrinsic Reaction
Coordinate connecting both the starting material and the product.
All the calculations were performed using Gaussian 03 Software
closer to the
interpreted as the nucleophile avoiding the repulsive interaction
with the -fluorine.
Such avoidance would not be possible for the
compound, with the result that it would have a much higher
activation barrier, whereas the -monofluoro compound would
only be slightly less reactive than the fluorine-free compound.
Looking at the last column of Table 9, it is seen that all of the
substitution reactions are exothermic, as expected. Whereas the
a-hydrogen than to the fluorine (Fig. 1). This could be
a
a,a-difluoro
a
package [23]. Table 9 presents the energy barriers
D
Gz, some
relevant geometrical data for the SN2 transition state (TS), and the
enthalpy of reaction for the alkyl iodide substrates with hydroxide
as nucleophile.
enthalpies of the
greatly, those of the
b
,
g
, and non-fluorinated substrates do not differ
-fluorinated substrates are considerably
Experimentally we found that the magnitude of the activation
barrier for the SN2 reaction using azide as a nucleophile had the
following trend: n-C7H15Br < n-C4F9CH2CH2Br–n-C8H17CHFBr ꢅ
n-C3F7CH2Br < n-C5H11CF2Br. The calculated activation barriers
shown in Table 9 for hydroxide acting as a nucleophile follow the
same trend, which is consistent with the experimental observation
that the presence of fluorines near the reactive site increases the
energy barrier.
a
larger. This can be attributed to what has become known as
‘‘incremental geminal stabilization’’, wherein significant thermo-
dynamic molecular stabilization derives from increased fluorine
substitution on a carbon [24]. Similarly, it would be expected that
the presence of an electronegative oxygen substituent on the same
carbon as a fluorine or two fluorine substituents will provide
significant stabilization to the molecule.
During the experimental study, it was found that methoxide
acted exclusively as a base in reactions with all of the fluoro
compounds where E2 was possible. Such results undoubtedly
derive from a combination of two effects. First, the barriers for the
potentially competitive SN2 reactions are increased when fluorine
The incremental changes in the
D
Gz values found in the
calculations can be understood in two ways. First, the partial
positive charge that is generated at the carbon-center of an SN2
transition state should be inductively destabilized by the presence
of
b
or
g
fluorines, resulting in an increase of the energy barrier. In
-fluorines this partial positive charge might be
the case of
a
is present, and secondly, the acidities of those b-hydrogens that are
partially stabilized (by electron donation by fluorine’s lone pairs),
which brings us to the second factor. Taking into account Liu’s
work, the electrostatic repulsive interactions between the incom-
ing nucleophile and the electron clouds of the near-by fluorines
should increase the activation energy, with this effect being largest
for the gem-difluoro substrates.
Table 9
Relevant geometrical data and calculated energy barriers at the transition state for
the SN2 reaction between alkyl-iodide substrates and hydroxide.
a
b
Gzc
D
Hd
˚
˚
Substrate
O–C (A)
C–I (A)
D
n-C3H7I
2.22
2.17
2.13
2.24
2.31
2.54
2.52
2.50
2.58
2.70
19.24
19.65
25.92
19.74
29.17
ꢃ40.5
ꢃ43.0
ꢃ39.2
ꢃ54.4
ꢃ62.2
CF3C2H4I
C2F5CH2I
C2H5CHFI
C2H5CF2I
a
Fig. 1. Two perspectives of the calculated transition state of the SN2 reaction of
hydroxide with CH3CH2CHFI (colors seen in web version of paper are red = oxygen,
violet = iodine, and blue = fluorine). (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of the article.)
Oxygen–carbon-center bond length.
Carbon-center – iodide bond length.
b
c
Calculated energy barrier in kcal/mol
d
Calculated enthalpy of reaction in kcal/mol.

174
H. Martinez et al. / Journal of Fluorine Chemistry 135 (2012) 167–175
Table 10
stabilization of this transition state, with interaction of the partial
negative forming charge and the * orbital of the anti C–F bond. In
Relevant geometrical data, calculated energy barriers at the transition State and
overall enthalpy of reaction for the E2 reaction between alkyl-iodide substrates and
hydroxide as a base.
s
contrast, looking at the transition state for elimination from
C2H5CF2I, both fluorines are skewed at a dihedral angle of about 608
from the breaking C–H bond, which means that these fluorines can
only provide inductive stabilization of the developing negative
charge on this carbon.
a
b
c
d
Substrate
O–H (A)
H–C (A)
C–C (A)
C–I (A)
D
Gze
D
Hf
˚
˚
˚
˚
C3H7I
1.33
1.27
1.30
1.31
1.28
1.31
1.35
1.32
1.32
1.34
1.43
1.48
1.43
1.43
1.43
2.43
2.24
2.39
2.39
2.32
19.87 ꢃ36.4
3.86 ꢃ32.2
11.56 ꢃ33.3
11.67 ꢃ33.6
7.93 ꢃ28.4
CF3C2H4I
C2H5CHFI (E)
C2H5CHFI (Z)
C2H5CF2I
All of the elimination transition states were characterized
computationally as E2 processes. However, one can see that the
lowest
D
Gz values have the longest C–H bond lengths, and least C–
a
Oxygen–hydrogen bond length (new OH bond).
Hydrogen–carbon bond length (breaking bond).
Carbon–Carbon bond length (new double bond).
Carbon–iodide bond length (breaking bond).
Calculated energy barrier in kcal/mol.
C double bond character based on bond length of the forming
double bond, and the shortest C–I bond length. These results
suggest a highly unsymmetrical bimolecular elimination consis-
tent with E1cb character.
b
c
d
e
f
Calculated enthalpy of reaction in kcal/mol.
Finally, elimination of HI from the mono-fluoro, C2H5CHFI can
result in formation of either cis (Z) or trans (E) products. The
calculated energy barriers for these reactions were found to be very
similar, and this correlates with the almost 1:1 ratio observed
experimentally.
vicinal to fluorines are increased, such that the activation barriers
for their E2 processes are lowered. This combination of slowing
down the SN2 and enhancing the E2 reactions can explain the
experimental results.
The quantum chemical calculations, as represented in Table 10,
3. Conclusions
support these conclusions. Experimentally, the
a,a-difluoroalkyl
bromide has such a high energy barrier for the SN2 reaction, that
even azide would prefer to behave as a base, albeit at a slow rate.
The highest calculated energy barrier,
29.17 kcal/mol, which was for the 1,1-difluoro compound. Even
with such a relatively high barrier, it seems reasonable that this
reaction would have occurred were it not for the relative ease of
the elimination reaction. Comparing values from Tables 9 and 10
one can see that for each of the reactions of fluorinated compounds
In conclusion, an experimental measure of the quantitative
effect of proximate fluorine substituents on the rates of SN2 and E2
reactions has been obtained. The study was comprised mainly of
reactions of fluorinated n-alkyl bromides with the weakly basic
nucleophile, azide ion, and the strongly basic nucleophile,
methoxide ion, in the protic solvent methanol and the aprotic
solvent, DMSO. The order of reactivity for SN2 reactions of azide in
methanol at 50 8C was n-alkyl-Br > n-alkyl-CHFBr > n-perfluor-
oalkyl-CH2CH2Br ꢁ n-perfluoroalkyl-CH2Br > n-alkyl-CF2Br, with
D
Gz, for substitution was
where
b-elimination is possible, the energy barrier for the E2
approximate relative rates of reaction being: 1, 0.20, 0.12, 1 ꢂ 10ꢃ4
,
process is much lower than that for SN2.
7.7 ꢂ 10ꢃ5. The order of reactivity for E2 reactions was n-
perfluoroalkyl-CH2CH2Br ꢁ n-alkyl-CF2Br > n-alkyl-CHFBr > n-al-
kyl-Br, with the approximate relative rates for reaction of
methoxide in methanol at 50 8C being: 1100, 4.4, 1.9, 1. The data
for both elimination and substitution are presented graphically in
Fig. 2.
The calculated trend for the activation barrier in the E2 reaction
decreases as follows: n-C3H7I > C2H5CHFI > C2H5CF2I > CF3C2H4I,
which is consistent with the observed experimental trend. From
the data in Table 10 it is possible to glean other interesting insight.
Although both CF3C2H4I and C2H5CF2I have multiple fluorines in
the b-position with respect to the abstracted hydrogen, the barrier
to elimination of the CF2I compound is double that of CF3C2H4I. We
believe that this is due to the fact that in the transition state for the
elimination of CF3C2H4I, one of the fluorines from the CF3 group can
have a dihedral angle of almost 1808 with respect to the hydrogen
that is being removed. Thus, in addition to the inductive
stabilization that the CF3 group provides, there is hyperconjugative
It appears that the detrimental effect of fluorine substituents on
SN2 reactions derives from two main effects: their electron
withdrawing inductive effect, which destabilizes the partial
positive charge that develops at the carbon center undergoing
substitution, and the electrostatic repulsive influence of (particu-
larly
a-substituted) fluorine lone pairs on the approaching
Fig. 2. Graphic representation of the kinetic data.

H. Martinez et al. / Journal of Fluorine Chemistry 135 (2012) 167–175
175
nucleophile. The rate-enhancing effect of fluorine substituents on
E2 reactions appears to be largely due to their C–H acidifying
influence, which derives from two factors: their aforementioned
inductive effect, and when geometry allows, their powerful
to the NMR tube. The reaction was then followed by measurement
of the decrease of one particular NMR signal with respect to the
internal standard up to 30% of the reaction. For kinetic experiments
above room temperature, 0.22 mmol of substrate in 0.25 mL of
deuterated solvent were placed into an NMR tube, and then
0.75 mL of the corresponding 0.6 M stock solution and 5–10 mg of
internal standard were added. The NMR tube was warmed to the
desired temperature and the reaction followed by NMR up to 30%
of the reaction.
hyperconjugative stabilization of developing
b-negative charge.
4. Experimental
4.1. Synthesis of 1-bromo-1-fluorononane (6) [9]
In the following order 170 mL of anhydrous CH2Cl2, 5.1 mL
(28.1 mmol, 1 equiv., 95%) of nonanal and 3.9 mL (33.7 mmol,
1.2 equiv.) of 2,6-lutidine were placed into a 250 mL round
bottomed flask. The mixture was then cooled down to ꢃ5 8C
and 9.5 mL (56.2 mmol, 2 equiv.) of triflic anhydride in 20 mL of
anhydrous CH2Cl2 were added dropwise to the reaction mixture
over a period of 45 min. The reaction was stirred for 7 h at 0 8C and
was then placed in the refrigerator (0 8C) for an extra 17 h. The
relatively dilute nonanal solution, the presence of the base and the
slow addition of triflic anhydride were important in order to obtain
good yield in this reaction; otherwise a polymer is obtained as a
major product. The solvent was removed and the dark remaining
mixture was extracted with pentane (4 ꢂ 100 mL). The polymer is
insoluble in pentane. The organic layer was washed with aqueous
HCl 1.2 M (2 ꢂ 100 mL), followed by concentrated aqueous
NaHCO3 (2 ꢂ 100 mL) and then brine (1 ꢂ 100 mL). The solution
was dried over MgSO4, filtered, and the solvent removed to give a
dark yellow oil, 9.5 g, 80% yield of the geminal, bis-triflate (by
NMR). This material was used without further purification for the
following step. Once the solvent is removed, the product should be
kept in the refrigerator, since it not stable neat at room
temperature.
4.2.2. Computational methods
The quantum chemical calculations were performed with a
hybrid basis set at MP2/6-31+G(d,p)-LANL2DZ level of theory. The
effective core potential (ECP) of iodide was included in the
calculations in order to minimize the time in the optimization.
DMSO was used as a solvent in all calculations using the
polarizable continuum model (PCM). The transition states were
characterized by one and only one negative frequency and the
intrinsic reaction coordinate (IRC) connecting both the starting
material and the product. All the calculations were performed
using Gaussian 03 Rev. E01 Software package [23].
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jfluchem.2011.10.008.
References
[1] A.E. Feiring, J. Org. Chem. 48 (1983) 347–354.
[2] A. Haas, M. Spitzer, M. Lieb, Chem. Ber. 121 (1988) 1329–1340.
[3] Y.-Z. Wu, J. Deng, X. Fang, Q.-Y. Chen, J. Fluorine Chem. 125 (2004) 1415–1423.
[4] J. Hine, R.G. Ghirardelli, J. Org. Chem. 23 (1958) 1550–1552.
[5] E.T. McBee, R.D. Battershell, H.P. Braendlin, J. Am. Chem. Soc. 84 (1962)
3157–3160.
[6] J. Hine, C.H. Thomas, S.J. Ehrenson, J. Am. Chem. Soc. 77 (1955) 3886–3889.
[7] H.F. Koch, W. Tumas, R. Knoll, J. Am. Chem. Soc. 103 (1981) 5423–5429.
[8] G.W. Kabalka, M. Varma, R.S. Varma, J. Org. Chem. 51 (1986) 2386–2388.
[9] A. Garcia Martinez, J. Osio Barcina, A.Z. Rys, L.R. Subramanian, Syn. Lett. (1993)
587–588.
[10] M.D. Bartberger, W.R. Dolbier, J. Lusztyk, K.U. Ingold, Tetrahedron 53 (1997)
9857–9880.
[11] J.W. Moore, R.G. Pearson, Kinetics and Mechanism, John Wiley & Sons Inc., 1981.
[12] H. Feuer, J. Hooz, in: S. Patai (Ed.), The Chemistry of the Ether Linkage, Wiley, New
York, 1967, pp. 445–498.
[13] V.C. Sekera, C.S. Marvel, J. Am. Chem. Soc. 55 (1933) 345–349.
[14] D.A. Shirley, J.R. Zietz, W.H. Reedy, J. Org. Chem. 18 (1953) 378–381.
[15] J. Barry, G. Bram, G. Decodts, A. Loupy, P. Pigeon, J. Sansoulet, Tetrahedron 40
(1984) 2945–2950.
[16] S. Hartman, R.E. Robertson, Can. J. Chem. 38 (1960) 2033–2038.
[17] D.S. Noyce, J.A. Virgilio, J. Org. Chem. 37 (1972) 2643–2647.
[18] R.L. Fuchs, L.L. Cole, J. Am. Chem. Soc. 95 (1973) 3194–3197.
[19] R. Alexander, E.C.F. Ko, A.J. Parker, T.J. Broxton, J. Am. Chem. Soc. 90 (1968)
5049–5069.
The bis-triflate (9.5 g, 22.5 mmol, 1 equiv.) along with 150 mL
of anhydrous CH2Cl2 were placed in a 250 mL round bottomed
flask. The mixture was then cooled to 0 8C, and 22.5 mL of TBAF 1 M
solution in THF (22.5 mmol, 1 equiv.) were added and stirred for
12 h at 0 8C. Subsequently 14.5 g (45 mmol, 2 equiv.) of tetra-
butylammonium bromide were added and stirred at room
temperature for 16 h more.
The solvent was then removed and the remaining oily material
was filtered through silica gel using pentane as a solvent. The
solvent was removed and a second column in pentane was run to
obtain 68% of product. The result is a mixture of 88% 1-bromo-1-
fluorononane (6) and 12% 1,1-difluorononane, which was used for
the kinetic experiments as is.
6: 1H NMR (CDCl3)
d
6.36–6.53 (dt, 2JHF = 50.7 Hz, 3JHH = 5.3 Hz,
1H, CHF), 2.00–2.45 (m, 2H, CH2), 1.20–1.75 (m, 14H, CH2), 0.89 (t,
2
3H, CH3); 19F NMR (CDCl3)
d
ꢃ130.65 (ddd, JHF = 50.3 Hz,
3JHF = 20.5 Hz, JHF = 18.2 Hz, 1F); 13C NMR (CDCl3)
d 95.9 (d,
3
1JCF = 252.5 Hz, CHF), 40.9 (d, JCF = 18.7 Hz, CH2), 32.02 (CH2),
29.56 (CH2), 29.33 (CH2), 28.91 (CH2), 25.30 (CH2), 22.86 (CH2),
14.29 (CH3); HRMS: Calc. 223.0503 ((MꢃH)+). Found 223.0489).
[20] A. Streitwieser Jr., Chem. Rev. 56 (1956) 371–752.
2
[21] S.T. Patel, J.M. Percy, R.D. Wilkes, Tetrahedron 51 (1995) 9201–9216.
[22] S. Liu, H. Hu, L.G. Pedersen, J. Phys. Chem. A 114 (2010) 5913–5918.
[23] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery, T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J.
Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H.
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev,
A.J. Austin, R. Cammi, C. Pomelli, J. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth,
P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C.
Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V.
Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham,
C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen,
M.W. Wong, C. Gonzalez, J.A. Pople, Revision E. 01, Gaussian, Inc., Wallingford, CT,
2004.
4.2. 1-Octyl tosylate was prepared according to the literature [8]. It’s
NMR spectra were consistent with those in the literature [25]
4.2.1. General procedure for the kinetic experiments
Stock solutions of NaN3, NaOMe and NaI (0.6 M in MeOH-d4 and
DMSO-d6) were prepared prior to the kinetics experiments. For
kinetic experiments below room temperature, 0.75 mL of the
required 0.6 M stock solution, and 5–10 mg of internal standard
(toluene for 1H NMR and trifluorotoluene for 19F NMR) were placed
into an NMR tube and the tube cooled to the desired temperature
inside the NMR instrument. Then, 0.22 mmol of substrate
previously dissolved in 0.25 mL of deuterated solvent were added
[24] W.R. Dolbier Jr., K.S. Medinger, A. Greenberg, J.F. Liebman, Tetrahedron 38 (1982)
2415–2420.
[25] F. Kazemi, A.R. Massah, M. Javaherian, Tetrahedron 63 (2007) 5083–5087.