Halogen migration vs. hydrogen halogenide elimination in reactions of 1-chloro-2,2,2-trifluoroethansulfonyl chloride and 1,2,2,2-tetrafluoroethansulfonyl fluoride with amines: theoretical and experimental investigation

Article abstract of DOI:10.1016/j.jfluchem.2009.11.001

A simple and useful method has been proposed for preparing of 1-chloro-2,2,2-trifluoroethan sulfonylchloride. By aminolysis of 1-chloro-2,2,2-trifluoroethansulfonylchloride the chlorine migration proceeds forming the corresponding salts of 1,1-dichloro-2,2,2-trifluoroethansulfinic acid. This process as well as the alternative reaction, elimination of hydrogen halogenide, has been studied using quantum chemistry (DFT and MP2) methods. As the calculation data indicate, an intermediately formed anion undergo intramolecular chlorine migration via a three-membered cyclic transition state. The latter is characterized by the low activation energy (ΔE = 27.0 kcal/mol). The barrier of activation in the case of 1,2,2,2-tetrafluoroethansulfonylfluoride is considerably higher (ΔE = 41.6 kcal/mol). The structures of the 1,1-dichloro-2,2,2-trifluoroethan sulfinic acid and 1,1,2,2,2-pentafluoroethansulfinic acid anions can be considered as donor-acceptor complexes of perhalogenoalkyl anions with SO₂. Crown Copyright

Full text of DOI:10.1016/j.jfluchem.2009.11.001

Journal of Fluorine Chemistry 131 (2010) 254–260
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Halogen migration vs. hydrogen halogenide elimination in
reactions of 1-chloro-2,2,2-trifluoroethansulfonyl chloride and
1,2,2,2-tetrafluoroethansulfonyl fluoride with amines:
theoretical and experimental investigation
a
a
a
Yuriy M. Pustovit a, , Anatoliy N. Alekseenko , Nikolai D. Volkov , Mykhailo Yu. Fedorchuk ,
*
Alexander B. Rozhenko a,b,
*
a Institute of Organic Chemistry, National Academy of Sciences of Ukraine, 5 Murmanskaya Str., 02660, Kiev, Ukraine
b
¨
University of Bielefeld, Universitatstr. 25, 33615, Bielefeld, Germany
A R T I C L E I N F O
A B S T R A C T
simple and useful method has been proposed for preparing of 1-chloro-2,2,2-trifluoroethan
Article history:
A
Received 29 July 2009
sulfonylchloride. By aminolysis of 1-chloro-2,2,2-trifluoroethansulfonylchloride the chlorine migration
proceeds forming the corresponding salts of 1,1-dichloro-2,2,2-trifluoroethansulfinic acid. This process
as well as the alternative reaction, elimination of hydrogen halogenide, has been studied using quantum
chemistry (DFT and MP2) methods. As the calculation data indicate, an intermediately formed anion
undergo intramolecular chlorine migration via a three-membered cyclic transition state. The latter is
Received in revised form 31 October 2009
Accepted 4 November 2009
Available online 13 November 2009
Keywords:
characterized by the low activation energy (
DE = 27.0 kcal/mol). The barrier of activation in the case of
1-Chloro-2,2,2-
1,2,2,2-tetrafluoroethansulfonylfluoride is considerably higher (
DE = 41.6 kcal/mol). The structures of
trifluoroethansulfonylchloride
the 1,1-dichloro-2,2,2-trifluoroethan sulfinic acid and 1,1,2,2,2-pentafluoroethansulfinic acid anions can
be considered as donor–acceptor complexes of perhalogenoalkyl anions with SO2.
Crown Copyright ß 2009 Published by Elsevier B.V. All rights reserved.
1,1-Dichloro-2,2,2-trifluoroethansulfenic
acid
Halogen migration
Quantum chemistry
Calculations
DFT
MP2
Transition state
COSMO
1. Introduction
chloride from the accessible starting materials is an important
goal.
Since the 80s of the past century the 2,2,2-trifluoroethansul-
fonyl chloride (tresyl chloride) has been finding a wide application
in biochemistry and medicine as a reagent for the hydroxyl group
activation at the surface of various materials [1]. A selective
activation of the mGlu2 receptor using 2,2,2-trifluoroethansulfo-
nylamides as allosteric modulators was reported recently [2–4].
2. Results and discussion
We have elaborated a useful method of synthesis of the
previously unknown 1-chloro-2,2,2-trifluoroethansulfonyl chloride
(1a) (Scheme 1). The accessible sodium 1-chloro-2,2,2-trifluor-
oethansulfinate [5,6] has been treated with sulfuryl chloride
yielding the desired product with a high yield. In contrast to the
known 1-chloro-2,2,2-trifluoroethansulfonyl fluoride [7], 1a is a
stable compound and can be stored for a long time under laboratory
conditions. This demonstrates chloride 1a as an especially perspec-
tive reagent for the preparative organic synthesis.
Therefore, elaboration of
a simple and useful method for
preparing of new analogues of 2,2,2-trifluoroethansulfonyl
* Corresponding authors at: Institute of organic chemistry, National Academy of
Sciences of Ukraine, 5 Murmanskaya Str., 02660, Kiev, Ukraine.
Tel.: +380 44 492 4610; fax: +380 44 573 2643.
We have found that 1a reacts with pyridine to give a product
which has undergone a chlorine rearrangement, pyridinium salt of
1,1-dichloro-2,2,2-trifluoroethansulfinic acid (2a) (Scheme 2). A
E-mail addresses: ypus@ukr.net (Y.M. Pustovit), a_rozhenko@ukr.net
(A.B. Rozhenko).
0022-1139/$ – see front matter. Crown Copyright ß 2009 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2009.11.001

Y.M. Pustovit et al. / Journal of Fluorine Chemistry 131 (2010) 254–260
255
Scheme 1. Preparation of 1-chloro-2,2,2-trifluoroethanesulfonyl chloride (1a).
Fig. 1. MOLDEN plots of optimized (;C2(fc)/6–311++G**) structures 1a,b.
formation of the corresponding anion. The reaction energies for
such processes have been estimated using Eq. (1):
F3CꢀCHðHalÞꢀSO2Hal þ Py ! F3CꢀCðꢀÞHalꢀSO2ꢀHal
Scheme 2. Rearrangement of 1a in the presence of amines.
þ PyHðþÞ þ
D
E1
(1)
Structures of anions F3C–C(ꢀ)Cl–SO2Cl (4a) and F3C–C(ꢀ)F–SO2F
(4b), calculated as isolated molecules at the MP2(fc)/6–311++G**
approximation level, are shown in Fig. 2. The structural data
resulting from the DFT (B3LYP/6–311++G**) calculations differ
only slightly and, if nothing specially noted, will not be further
discussed here. More detailed data set (Cartesian coordinates for
the MP2 and DFT optimized structures, total energy values, ZPE
correction values, etc.) are included in the Supplementary Material
for this paper. Within the gas phase approximation, reaction (1) is
similar behaviour was reported for dichloromethanesulfonyl
chloride [8]. The formation of the 1,1-dichloro-2,2,2-trifluor-
oethansulfinic acid has been confirmed by comparing the 19F
and 13C NMR spectra of the pyridinium salt with those for the
known sodium salt [5].
In the reactions of 1a with more basic amines such as
morpholine or ammonia, the corresponding sulfonylamides 3a
and 3b, respectively have been isolated along with the migration
products, ammonium salts of 1,1-dichloro-2,2,2-trifluoroethan-
sulfinic acid (approximately 20%). Alternatively, the application of
a weaker base, such as, e.g. 3-(trifluoromethyl)aniline, only
sulfonylamide 3c arises as the reaction product. Probably, the
basicity of 3-(trifluoromethyl)aniline is not enough to provide
deprotonation of 1a at the first stage of the reaction (Scheme 3).
It was demonstrated previously [9] that 1,2,2,2-tetrafluor-
oethansulfonyl fluoride (1b) reacted with pyridine via alternative
pathway giving rise to a stable pyridine adduct of sulfene, as a
product of the hydrogen fluoride elimination. It should be noted
here that no such adducts have been identified in reaction of 1a
with amines, even as by-products.
strongly endothermic (D
E1 = 111.0 and 101.3 kcal/mol at the MP2
level of theory). Calculations for the solution in dichloromethane
using the COSMO procedure [10] (see Section 5) reduce the
energies for the C–H bond cleavage to 22.0 and 14.1 kcal/mol,
respectively. Taking into account the effect of counterion will
additionally stabilize the anions (see the next chapter).
In structures 4a,b the C–S bonds are rather short (1.672 and
˚
1.686 A, respectively), whereas the sulfur–halogen distances
˚
are significantly lengthened (2.292 and 1.741 A). For comparison,
the S–F bond length for SO2F2 determined using the electron
˚
diffraction method in the gas phase totals 1.552 A, whereas the
˚
In order to explain the difference in chemical behaviour of 1a
and 1b by treatment with pyridine and to elucidate the
thermodynamic and kinetic peculiarities of these two different
transformation pathways (elimination and migration), we have
decided to investigate these two reactions using quantum
chemistry methods. In the first part of the current paper we
analyze the results of calculations carried out for small model
structures using the gas phase approximation and empirically
corrected level for description of the solvent effects using the
polarizable continuum model (PCM).
S–Cl bond length for sulfuryl chloride SO2Cl2 is found to be 2.011 A
[11]. Thus, 4a,b might be represented as the adducts of sulfenes
F3C–C(Hal)55S(55O)2 (5a,b) with the corresponding halide anions. A
shorter C–S bond in 4b than in 4a and a larger sum of the bond
angles at the sulfur atom in the –S(55O)2 moiety (345.38 for 4a vs.
343.58 for 4b) indicate stronger S–Hal interactions in 4b than in 4a.
High negative NBO charges on the halogen atom attached to the
SO2 group calculated for 4a,b (ꢀ0.57 on fluorine in 4b and ꢀ0.47 on
chlorine in 4a) are in agreement with the representation as the
above mentioned donor–acceptor adducts.
The equilibrium structures for 1a and 1b are shown in Fig. 1.
The first stage for either the hydrogen halogenide elimination or
the halogen migration is deprotonation of sulfonyl halogenide and
Equilibrium structures of sulfenes 5a,b are shown in Fig. 3.
Formation of sulfenes can arise via halide anion elimination from
4a,b (Eq. (2)). These reactions are in the both cases endothermic
Scheme 3. Effect of amine basicity on the products of aminolysis of 1a.

256
Y.M. Pustovit et al. / Journal of Fluorine Chemistry 131 (2010) 254–260
Fig. 2. MOLDEN plots of optimized (;C2(fc)/6–311++G**) structures 4a,b.
Fig. 4. MOLDEN plots of optimized (;C2(fc)/6–311++G**) structures 2a,b.
We have also analyzed the intermolecular rearrangement for 4a
proceeding via the cyclic transition state (Scheme 4). The
corresponding transition state structures TS2 (C2 symmetry) and
TS3 (C1 symmetry, see Supplementary Material for more details)
correspond to the practically identical activation energy values,
which are definitely higher (DE DG
TS2 = 47.5 kcal/mol, TS2 = 54.1–
54.3 kcal/mol of 4a), than that calculated for the intramolecular
process. Thus, the bimolecular mechanism of the rearrangement
can be excluded from consideration.
Fig. 3. MOLDEN plots of optimized (;C2(fc)/6–311++G**) structures 5a,b.
The C–S bonds in anions 2a,b (Fig. 4) are significantly elongated
within the gas phase approximation (
D
E2 = 12.9 and 20.0 kcal/mol,
(1.962 and 2.123 A, respectively. Taking into account solvent
˚
respectively). This is not contrary to the experimental data
evidencing sulfenes as highly unstable species. They easily
undergo further transformations, for example, as substrates in
the Diels-Alder reactions [12,13] or in the reaction with amines
[9,14,15]. Structures and relative stabilities for series of sulfene
adducts with amines were analyzed recently in the theoretical
paper of Bucher [16].
effects (dichloromethane) imply some shortening of these bonds
˚
(1.921 and 2.011 A, respectively). Such bond lengths allow us to
consider 2a,b as the donor–acceptor adducts of perhalogenoalkyl
anions CF3–C(Hal)2(ꢀ) with the SO2 molecule (Scheme 1). A similar
significant bond elongation was also reported previously for the
trialkylammonium salts of trichloromethanesulfonic acid
R3NH(+)Cl3C–SO2
[17] and for salts 6a,b [14]) in solid. Such
(ꢀ)
In accordance with our experimental finding, anion 4a under-
goes a halotropic rearrangement of the halide anion towards
carbon giving an isomeric anion 2a as the product (reaction (2a)).
coordination character of the C–S bonds in 2a,b is a consequence of
(ꢀ)
the increased stabilities of the CF3–C(Hal)2
anions, due to the
charge delocalization on the electronegative halogen atoms. It is
easy to recognize in 6a,b donor–acceptor complexes of the
halocarbene with the amine as a donor and SO2 as an acceptor.
Such a consideration is in line with the dual nature of carbenes
[18]. Despite the rather high level of approximation (MP2(fc)/6–
311++G**), the calculated C–S bond lengths for 6a,b (2.026 and
The reaction is exothermic (
D
E
2a = ꢀ30.5 kcal/mol) and proceeds
via transition state 4a-TS (Scheme 1). The total energy for the
mentioned transition state structure is only 27.0 kcal/mol higher
(B3LYP/6–311++G**) than that calculated at the same level of
theory for 4a. In contrast to sulfene 5b, the formation of the
2a = ꢀ24.9 kcal/mol). How-
2.070 A, respectively) are longer than the experimental ones (1.904
˚
migration product 2b is favoured (
ever, a significantly higher activation energy (
D
E
TS = 41.6 kcal/
and 1.949 A) [14]. A possible explanation for the observed
˚
D
E
mol) has been predicted in this case for the corresponding
transition state structure, 4b-TS, compared to that calculated for
the chlorine migration via 4a-TS (27.0 kcal/mol):
discrepancies is probably the gas phase approximation used for
geometry optimization, which poorly describes the structures in
solid state. For comparison, the C–S bond length calculated for the
(ꢀ)
˚
Cl3C–SO2
anion (2.055 A) is also too long compared to the
F3CꢀCðꢀÞHalꢀSO2ꢀHal ! F3CꢀCHal2ꢀSO2ðꢀÞ þ
D
E2
(2)
experimental value for R3NH(+)Cl3C–SO2 (1.91 A) [17]. However,
(ꢀ)
˚
similarly to 4a,b, correction for the solvent effects using the
F3CꢀCðꢀÞHalꢀSO2ꢀHal ! F3CꢀCHal¼SO2 þ HalðꢀÞ þ
D
E2a
(2a)
˚
COSMO routine significantly shortens the C–S bond (to 1.962 A).
Therefore, the C–S bond lengthening should always be expected
for species R–SO2(ꢀ), where R can form a rather stable free anion. In
this connection, the calculated C–S distance in the series of
˚
haloalkyl sulfinic acids increases as follows: R = –CH3 (1.843 A) < –
˚
˚
˚
CF3 (1.922 A) < –CF2CF3 (2.026 A) < –CCl3 (2.055 A) < –CCl2EF3
˚ ˚
(2.070 A) < –CF(EF3)2 (2.184 A). The DE and DG values for the
Scheme 4. Monomolecular and bimolecular mechanisms of thansformation of 4a.

Y.M. Pustovit et al. / Journal of Fluorine Chemistry 131 (2010) 254–260
257
Table 1
Calculated reaction energy values for SO2 elimination from X–SO2
(ꢀ)
(DE and DG, kcal/mol).
X
B3LYP/6–311++G**
Gas phase
MP2(fc)/6–311++G**
Gas phase
CH2Cl2 solution
CH2Cl2 solution
D
E
D
G
D
E
D
G
D
E
D
G
D
E
DG
CH3
71.8
29.2
41.6
38.2
24.8
30.0
29.2
19.1
61.8
17.6
30.3
26.9
14.1
18.7
17.4
6.1
70.2
24.9
30.4
29.9
18.3
28.3
–
59.8
12.2
18.3
17.6
5.9
65.0
36.2
38.2
36.0
25.3
33.0
27.5
19.9
55.0
24.3
26.5
24.3
13.8
21.4
15.7
7.4
64.6
33.1
28.0
28.9
19.8
28.4
–
54.5
20.6
15.9
16.7
7.4
CCl3
CF3
CF3CF2
(CF3)2CF
CF3CCl2
Et3N(+)CHF
Et3N(+)CF2
16.0
–
16.0
–
–
–
–
–
Table 2
Calculated reaction energy values (
D
E and
D
G, kcal/mol) for reactions 1–6.
Reaction main product
Reaction
B3LYP/6–311++G**
Gas phase
MP2(fc)/6–311++G**
Gas phase
CH2Cl2 solution
CH2Cl2 solution
DE
DG
DE
DG
D
E
D
G
DE
DG
2c
2b
1
1
2
2
3
3
4
4
5
5
6
6
91.9
91.4
5.5
8.9
5.0
9.1
101.3
111.0
ꢀ24.9
ꢀ20.0
101.2
111.3
ꢀ25.1
ꢀ20.1
14.1
22.0
13.6
22.0
101.6
ꢀ25.3
ꢀ15.7
ꢀ7.7
101.6
ꢀ25.9
ꢀ16.0
1.5
4a
ꢀ30.3
ꢀ20.0
ꢀ30.5
ꢀ19.6
ꢀ30.5
ꢀ25.2
ꢀ30.3
ꢀ24.5
4b
7a
7b
ꢀ8.6
1.5
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
8a
25.7
25.3
8b
16.2
16.1
–
–
9a
ꢀ25.3
ꢀ29.6
ꢀ19.3
ꢀ23.7
ꢀ14.3
ꢀ19.2
ꢀ8.7
ꢀ12.7
ꢀ30.5
ꢀ28.8
ꢀ23.5
ꢀ28.3
ꢀ18.4
ꢀ16.4
ꢀ14.3
ꢀ17.1
9b
10a
10b
dissociation reaction, forming the free anion and SO2, are shown in
pyridinium salt, 7a,b. Within the gas phase approximation, such
coordination is exothermic for both halogenides, 1a and 1b
(Table 2). Taking entropy into account yields the free energy of the
reaction close to zero, obviously due to formation of an aggregate
from two separate molecules. However, based on the approximat-
ed quantum chemical considerations, it is usually difficult to
estimate the entropy contribution. First of all, the geometry
optimization is still carried out towards the total energy minimum,
not the Gibbs free energy, because the optimization algorithm
using the free energy as a criterion for optimization is hitherto
unknown. Secondly, calculations are usually performed for
isolated molecules. Since reactions are normally carried out in
solutions, molecules are not isolated, but exist in the form of
solvates. Desolvation is an endothermic process that decreases the
real exothermic effect of reaction. Taking all above-mentioned
conclusions into account, we can suggest that the formation of 7a,b
will probably be slightly exothermic and will not determine the
Table 1. The magnitude of
isolated (F3C)2CF–SO2
D
G for the C–S bond dissociation for the
(ꢀ)
anion (calculated at the MP2(fc)/6–
311++G** level of theory) amounts 19.8 kcal/mol. In the dichlor-
omethane solution, only 7.4 kcal/mol is necessary for this purpose.
Thus, in the presence of bases, the haloalkyl sulfinic acids will
probably easy loose SO2. The most pronounced disposition towards
SO2 loosing is demonstrated by salts 6a,b (Table 1). The
dissociation energy calculated for the isolated molecule 6b is only
ca. 6 kcal/mol.
Our consideration of the anionic conversions of 4a,b does not
take into account the counterion effect. Now we consider using the
DFT level of theory possible reaction stages and intermediates for
the 1a,b transformations in the presence of the base (pyridine)
(Scheme 5). The first stage for both the hydrogen halogenide
elimination and the halogen rearrangement is a coordination of
pyridine at the C–H hydrogen forming the corresponding
Scheme 5. Transformations of 1a,b in presence of pyridine.

258
Y.M. Pustovit et al. / Journal of Fluorine Chemistry 131 (2010) 254–260
Fig. 6. MOLDEN plots of optimized (;C2(fc)/6–311++G**) structures 10,b.
can result from the higher activation energy inherent to 1b in the
isomerization reaction than that calculated for 1a. Any case, the
formation of the sulfene adduct with pyridine in case of 1b can be
considered as a result of a subtle balance of different effects and
processes, which cannot simultaneously be taken into account
using the currently available quantum chemistry methods.
Nevertheless, such results do not exclude that 1b under
appropriate conditions might undergo isomerization, similarly
to 1a.
Fig. 5. MOLDEN plots of optimized (;C2(fc)/6–311++G**) structures 9,b.
total reaction rate. In both cases our gas phase calculations predict
a covalent character for the C–H bond where as the N–H bond
remains coordinative. Similarly to anions 4a,b, the S–Hal distances
3. Conclusions
˚
in 7a,b are noticeably elongated (2.112 and 1.627 A, respectively):
E3
(3)
A convenient way of synthesis of the previously unknown 1-
chloro-2,2,2-trifluoroethansulfonylchloride (1a) has been pro-
posed. In the presence of basic amines 1a undergoes easily
chlorine migration, where as in case of low-basic 3-(trifluoro-
methyl) aniline only the corresponding sulfonylamide is formed.
Two ways of transformation of 1a as well as the structurally similar
1,2,2,2-tetrafluoroethansulfonylfluoride (1b) with pyridine have
been investigated theoretically at the DFT and MP2(fc) levels of
approximation. It has been found that 1a preferably reacts with the
halogen migration from sulfur to carbon. It is different in the case
of 1,2,2,2-tetrafluoroethansulfonylfluoride that was previously
demonstrated to undergo hydrogen halogenide elimination
yielding in solution a stable adduct of the corresponding sulfene
with pyridine. Both reactions begin with deprotonation, the
corresponding anions indicate rather short C–S bonds and
elongated S–Hal bonds and can be better represented as adducts
of corresponding sulfenes with the halogenide anions. The further
conversion can proceed as the elimination or halogen migration
processes. The latter reaction proceeds intramolecularly with the
relatively low activation energy and is in both cases an exothermic
process. For the products, anions of the sulfinic acids, long C–SO2
distances have been predicted. One can better describe these
structures as adducts of the stable perhalogenated anions with the
SO2 molecule, and these should be disposed towards easy
dissociation, similarly to other anions of haloalkyl sufinic acids.
The elimination reaction is endothermic within the gas phase
approximation and becomes exothermic only if the unstable
sulfene is stabilized by complexation with pyridine. However, the
latter should be considered as the adduct of a perhalogenated
carbene, coordinated with an amine and SO2. For 1-chloro-2,2,2-
trifluoroethansulfonylchloride, the migration process is definitely
more favoured whereas 1,2,2,2-tetrafluoroethansulfonylfluoride
demonstrates similar MP2 reaction energies for both migration
and elimination. Thus, the conversion of the latter in the presence
of amines via the halogen migration cannot also be excluded.
CF3ꢀCHalðHÞꢀSO2Hal þ Py ! CF3ꢀCHalðH ꢁ PyÞꢀSO2Hal þ
D
A deprotonation, following the coordination of pyridine to 1a,b
(Eqn. (4)) is an endothermic process, however, the thermodynamic
reaction parameters are also effected by solvents and, probably, by
the nature of a counterion. Since the reaction has been carried out
in the relative low-polar dichloromethane, 4a,b will probably exist
in the form of close ion pairs. The structures 8a,b with the lowest
total energies have been located by geometry optimization at the
B3LYP/6–311++G** level of theory (see Supplementary Material for
details). In the both adducts, the pyridinium cation is coordinated
at oxygen of the SO2 group (Scheme 5). The predicted reaction
energies
D
E4 are endothermic (Table 2):
CF3ꢀCHalðH ꢁ PyÞꢀSO2Hal ! CF3ꢀCHalðꢀÞꢀSO2Hal ꢁ HPyðþÞ þ
D
E4
(4)
As the last stage either the halogen migration proceeds yielding
9a,b (Fig. 5), or the next pyridine molecule attacks nucleophilically
the C(Hal) carbon atom forming inner salts 10a,b (Fig. 6). As a
result, reactions (5) and (6) are exothermic:
CF3ꢀCðHalÞHꢀSO2Hal þ Py ! CF3ꢀCHal2ꢀSO2H ꢁ Py þ
D
E5
(5)
CF3ꢀCðHalÞHꢀSO2Hal þ 2Py
! CF3ꢀCðHalÞPyðþÞꢀSO2ðꢀÞ þ HalH ꢁ Py þ
D
E6
(6)
A comparison of the
D D
E5 and E6 values derived from the total
energies at the B3LYP/6–311++G** level of theory (Table 2)
indicates that reaction (5) is more preferable for the both
halogenides 1a,b. Geometry optimizations for 9a,b and 10a,b
carried out at the more superior level of approximation (MP2/6–
311++G**) make reaction (5) significantly more favoured for
chloride 1a (
D
E5 = ꢀ30.5,
D
E6 = ꢀ23.5 kcal/mol), whereas in case
4. Experimental
of 1b both reaction energy values are very close (
D
E5 = ꢀ28.8,
D
E6 = ꢀ28.3 kcal/mol). Changing to
D
G values makes reaction
4.1. General
(6) for 1b even slightly more favourable
(
D
G5 = ꢀ16.4,
D
G6 = ꢀ17.1 kcal/mol). Thus, the reaction pathway will strongly
1H NMR (500 MHz), 19F (470 MHz) and 13C NMR (125 MHz)
spectra were recorded on a Bruker Avance DRX 500 spectrometer.
depend upon the reaction conditions. An elimination product 10b

Y.M. Pustovit et al. / Journal of Fluorine Chemistry 131 (2010) 254–260
259
TMS (1H NMR and 13C NMR spectroscopy), and CFCl3
(
19F NMR
m.p. = 111–112 8C. 1H NMR (500 MHz, CDCl3):
d
= 5.06 (q,
spectroscopy) were used as internal standards. Infrared (IR)
spectra were recorded on a UR-20 spectrometer in KBr or neat
for 1-chloro-2,2,2-trifluoroethanesulfonyl chloride (1a). Melting
points were measured with a Buchi melting points apparatus and
are uncorrected. Boiling point for 1-chloro-2,2,2-trifluoroethane-
sulfonyl chloride (1a) is uncorrected.
Sodium 1,1-dichloro-2,2,2-trifluoroethanesulfinate and sodium
1-chloro-2,2,2-trifluoroethanesulfinate were synthesized as de-
scribed in literature [5,6]. All experiments were carried out under
inert atmosphere. Dichloromethane was distilled over phospho-
rous pentoxide under argon.
3JHF = 6.5 Hz, CH), 3.76 (m, 4H, CH2OCH2), 3.53 (m, 4H, CH2NCH2).
3
19F NMR (470 MHz, CDCl3):
d
= ꢀ68.94 (d, JFH = 6.5 Hz, 3F, CF3).
13C NMR (125 MHz, CDCl3):
d
= 121.0 (q, 1JCF = 281 Hz, CF3), 68.26
(q, JCF = 34 Hz, CHCl), 66.91 (s, CH2O), 47.47 (s, CH2N). IR (KBr),
(cmꢀ1): 2990–2920, 1690, 1480, 1340, 1290, 1190, 1180, 1170,
2
n
1060, 1040, 765,720. Anal. calcd. for C6H9ClF3NO3S: C 26.93; H
3.39; Cl 13.25; F 21.29; N 5.23; S 11.98. Found: C 27.07; H 3.60; Cl
13.38; F 21.01; N 5.24; S 12.04.
4.6. 1-Chloro-2,2,2-trifluoroethanesulfonamide (3b)
30 ml of dry dichloromethane was saturated at 0 8C with dry
ammonia. 4.34 g (20 mmol) of sulfonylchloride 1a was added
dropwise at 0 8C under stirring and continuous bubbling of dry
ammonia. Full conversion of 1a was controlled by the 19F NMR
spectroscopy. Formation up to 20% of 1,1-dichloro-2,2,2-trifluor-
oethanesulfinate ammonium salt was detected by NMR spectros-
4.2. 1-Chloro-2,2,2-trifluoroethanesulfonyl chloride (1a)
To 5 g (3 ml, 37 mmol) of sulfuryl chloride in 10 ml dichlor-
omethane 3.8 g (18.58 mmol) of sodium 1-chloro-2,2,2-trifluor-
oethanesulfinate was added in portions at 5 8C and the mixture
was vigorously stirred for 10 min. The reaction mixture was
allowed to warm to RT and then heated under stirring at 35 8C for
2 h. A precipitate was filtered off; distillation of the filtrate at
atmospheric pressure provided a colourless liquid (1a). Yield:
copy.
A
reaction mixture was allowed to warm to room
gaseous HCl. The
temperature and then saturated with
a
precipitate was filtered off; the solvent was evaporated under
vacuum. The residue (3b) was crystallized from hexane. Yield:
3.16 g (80%) as colourless hygroscopic crystals, m.p. = 86–87 8C. 1H
3.02 g (75%); b.p. = 135–136 8C. 1H NMR (500 MHz, CDCl3):
d
= 5.48
= ꢀ67.84
= 119.7 (dq,
3
3
(q, JHF = 5.5 Hz, 1H, CH). 19F NMR (470 MHz, CDCl3):
d
NMR (500 MHz, DMSO-d6):
d
= 6.29 (q, JHF = 6.8 Hz, CH), 8.11 (s,
(d, JFH = 5.5 Hz, 3F, CF3). 13C (125 MHz, CDCl3):
d
2H, NH2). 19F NMR (470 MHz, DMSO-d6):
3JFH = 6.8 Hz, 3F, CF3). 13C NMR (125 MHz, DMSO-d6):
(q, 1JCF = 280.4 Hz, CF3), 67.52 (dq, 1JCH = 163.5 Hz, 2JCF = 32.07 Hz,
CHCl). IR (KBr),
(cmꢀ1): 3310–3060, 2910, 1695, 1470, 1340,
d
= ꢀ67.34 (d,
= 121.52
3
1JCF = 283 Hz, JCH = 6.3 Hz, CF3), 76.3 (dq, 1JCH = 163.5 Hz,
d
2
2JCF = 35.2 Hz, CH). IR (neat),
n
(cmꢀ1): 1420, 1310, 1230, 1200,
1140, 900, 780, 700. Anal. calcd. for C2HCl2F3O2S: C 11.07; H 0.46;
Cl 32.68; F 26.27; S 14.78. Found: C 11.05; H 0.5; Cl 32.71; F 26.08;
S 14.74.
n
1290, 1200, 1160, 1150, 1040, 1010, 780, 720. Anal. calcd. for
C2H3ClF3NO2S: C 12.16; H 1.53; Cl 17.95; F 28.85; N 7.09; S 16.23.
Found: C 12.20; H 1.60; Cl 17.92; F 28.58; N 6.92; S 16.14.
4.3. Pyridinium 1,1-dichloro-2,2,2-trifluoroethanesulfinat (2a)
4.7. 1-Chloro-2,2,2-trifluoro-N-[3-
To a solution of 1.1 g (5 mmol) of 1-chloro-2,2,2-trifluoroetha-
nesulfonyl chloride (1a) in 20 ml of dry dichloromethane a solution
of 0.4 g, 0.41 ml (5 mmol) of dry pyridine in 10 ml of dry
dichloromethane was added dropwise at 5 8C and under vigorous
stirring. After 1 h stirring at this temperature hexane (10 ml) was
added. The mixture was filtrated and hexane was added to the
filtrate until crystallization begins. After 1 h 2a was filtered as the
colourless semisolid hydroscopic masse. Yield: 1.13 g (76.3%). 1H
(trifluoromethyl)phenyl]ethanesulfonamide (3c)
To a solution of 2.17 g (10 mmol) of 1-chloro-2,2,2-trifluor-
oethanesulfonyl chloride (1a) in 10 ml of dichloromethane at 5 8C
under stirring was added dropwise a solution 3.22 g (20 mmol) of
3-(trifluoromethyl)aniline in 10 ml of dichloromethane. A precipi-
tate of 3-(trifluoromethyl)aniline chlorohydrate was filtered off, a
filtrate was evaporated under reduced pressure, a residue was
crystallized from hexane. Yield of (3c) 2.74 g (80%) as colourless
NMR (500 MHz, CDCl3): d
= 7.94 (t, 3JHH = 6.8 Hz, 2H, 3HPy), 8.48 (t,
3JHH = 7.7 Hz, 1H, 4HPy), 8.95 (d, 3JHH = 5.5 Hz, 2H, 2HPy), 13.5(br s,
crystals with m.p. = 73–74 8C. 1H NMR (500 MHz, CDCl3):
d
= 5.05
1H, NH). 19F NMR (470 MHz, CDCl3):
NMR (125 MHz, CDCl3):
2JCF = 29 Hz, CCl2). IR (KBr),
d
= ꢀ72.53 (s, 3F, CF3). 13C
(q, JHF = 5.7 Hz, 1H, CH), 7.48 (br s, 1H, NH), 7.54–7.58 (m, 4H,
3
d
= 122.43 (q, 1JCF = 283 Hz, CF3), 96.49 (q,
n
H
apoM.). 19F NMR (470 MHz, CDCl3):
d
= ꢀ63.40 (s, 3F, CF3Ar),
(cmꢀ1): 3010–2650, 1620, 1470, 1210,
ꢀ68.12 (d, 3JFH = 5.7 Hz, 3F, CF3CHCl). 13C NMR (125 MHz, CDCl3):
2
1120, 1060, 910, 890, 760, 710.
d
= 134.44 (s, CAr), 132.63 (q, JCF = 34 Hz, CAr.), 130.72 (s, CAr),
125.75 (s, CAr), 124.0 (q, 3JCF = 3.8 Hz, CAr), 123.44 (q, 1JCF = 272 Hz,
1
3
4.4. Sodium 1,1-dichloro-2,2,2-trifluoroethanesulfinate
CF3Ar), 120.82 (q, JCF = 282 Hz, CF3CHCl), 119.39 (q, JCF = 6.0 Hz,
2
C
Ar), 67.25 (q, JCF = 34 Hz, CCHCl). IR (KBr), n
(cmꢀ1): 3010, 1695,
19F NMR (470 MHz, CD3COCD3):
d
= ꢀ71.53 (s, 3F, CF3). 13C NMR
1470, 1340, 1285, 1190, 1030, 785, 720, 665, 480. Anal. calcd. for
C9H6ClF6NO2S: C 31.64; H 1.77; Cl 10.38; F 33.36; N 4.10; S 9.38.
Found: C 31.73; H 1.87; Cl 10.26; F 33.18; N 4.10; S 9.16.
1
(125 MHz, CD3COCD3):
d
= 122.3 (q, JCF = 283 Hz, CF3), 95.5 (q,
2JCF = 30.2 Hz, CCl2). IR (KBr), (cmꢀ1): 1210, 1120, 1060, 910, 890,
n
760, 710.
5. Details of calculations
4.5. 4-[(1-Chloro-2,2,2-trifluoroethyl)sulfonyl]morpholine (3a)
The structures of the studied compounds were fully optimized
with the GAUSSIAN-03 set of programs [19] at the DFT (B3LYP [20])
and MP2(fc) level of theory (using frozen core approximation). In
the former case a three-parameter hybrid functional B3LYP,
including correlation fuctionals from Lee et al. [21] and VWN5
[22]. In both cases, geometry optimizations were carried out using
medium-size 6–311++G** basis sets. As default within the
GAUSSIAN packet the mentioned basis sets are defined as the
proper 6–311G Pople basis sets [23] for hydrogen and the second
period atoms (C, N, O, F) and the (12s,9p) McLean-Chandler basis
To a solution of 1.74 g (1.75 ml, 20 mmol) of morpholine in
15 ml of dichloromethane at 5 8C under stirring was added
dropwise 2.17 g (10 mmol) of 1-chloro-2,2,2-trifluoroethanesul-
fonyl chloride (1a) in 20 ml of dichloromethane. Formation up to
15% morpholinium salt of 1,1-dichloro-2,2,2-trifluoroethanesulfi-
nic acid was detected with the 19F NMR spectroscopy. A reaction
mixture was washed with water; the organic layer was separated,
dried with MgSO4 and evaporated under reduced pressure. A
residue (3a) was crystallized from hexane. Yield: 2.14 g (80%),

260
Y.M. Pustovit et al. / Journal of Fluorine Chemistry 131 (2010) 254–260
[4] E. Hu, P.C. Chua, L. Tehrani, J.Y. Nagasawa, A.B. Pinkerton, B.A. Rowe, J.-M.
Vernier, J.H. Hutchinson, N.D.P. Cosford, Bioorg. Med. Chem. Lett. 14 (2004)
5071–5074.
[5] H. Wei-Yuan, C. Jian-Long, Acta Chim. Sin. 45 (1987) 445–449.
[6] W. Dmowski, K. Piasecka-Maciejewska, J. Fluorine Chem. 126 (2005) 877–882.
[7] N.P. Aktaev, G.A. Sokol’skii, I.L. Knunyants, Izv. Akad. Nauk SSSR, Ser. Khim. 11
(1975) 2530–2536 (CA 84, 59345d).
[8] T. Kempe, T. Norin, Acta Chem. Scand. B 28 (1974) 609–612.
[9] L.I. Ragulin, P.P. Ropalo, G.A. Sokol’skii, I.L. Knunyants, Izv. Akad. Nauk SSSR, Ser.
Khim. 5 (1971) 1045–1049 (CA 75, 87784d).
[10] V. Barone, M. Cossi, J. Phys. Chem. A 102 (1998) 1995–2001.
[11] D.R. Lide (Ed.), Handbook of Chemistry and Physics, 84th ed., CRC Press, 2003–
2004, pp. 9–22.
[12] G. Opitz, M. Deissler, K. Rieth, R. Wegner, H. Irngartinger, B. Nuber, Liebigs Ann.
1995 (1995) 2151–2163.
[13] G. Opitz, M. Deissler, T. Ehlis, Thomas, K. Rieth, H. Irngartinger, M.L. Ziegler, B.
Nuber, Liebigs Ann. 1995 (1995) 2137–2149.
[14] H. Prizkow, K. Rall, S. Reimann-Andersen, W. Sundermeyer, Angew. Chem. 102
(1990) 80–81; Angew. Chem., Int. Ed. Engl. 29 (1990) 60–61.
[15] U. Hartwig, H. Pritzkow, W. Sundermeyer, Chem. Ber. 121 (1988) 1435–1439.
[16] G. Bucher, Eur. J. Org. Chem. (2003) 3868–3874.
[17] R. Allmann, W. Hanefeld, M. Krestel, B. Spangenberg, Angew. Chem. 99 (1987)
1175–1176; Angew. Chem. Int. Ed. Engl. 26 (1987) 1133–1134.
set [24] for sulfur and chlorine, expanded with the appropriate
polarization and diffuse Gaussian functions. Calculations of
vibration frequencies were performed at the both levels of
approximation for all optimized structures. The only exception
were structures 2a, 13a,b, 14 and 15a,b, of especially large size, for
which the vibration frequencies were calculated only using the
MP2/6–311G** approach. The first and second derivatives were
computed analytically. Zero and one imaginary frequency for
the structure indicated the local minimum and transition
state, respectively. GAUSSIAN-03 default setting (temperature
298.150 K, pressure
calculations. The calculated energy values were corrected by
addition of zero point energy (ZPE) values (for calculation of
magnitudes) or thermal correction to Gibbs free energy (TCGFE)
values (for calculation of G magnitudes). For some cases, the
1 atm) was used for thermochemistry
D
E
D
geometries were re-optimized using the empirical procedure
CPCM (COSMO) [10] and modelling the situation in dichloro-
methane. The atomic charges were calculated using the NBO
packet [25]. All structures within the paper are pictured using the
MOLDEN [26] visualization program set.
´
[18] S. Sole, H. Gornitzka, W.W. Schoeller, D. Bourissou, G. Bertrand, Science 292
(2001) 1901–1903.
[19] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery Jr., 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, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R.
Cammi, C. Pomelli, J.W. 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. Nanayak-
kara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez,
J.A. Pople, Gaussian 03, Revision B.03, Gaussian, Pittsburgh, 2003.
[20] A.D. Becke, J. Chem. Phys. 98 (1993) 5648–5652.
Acknowledgements
We thank Professor Dr. W.W. Schoeller and Professor Dr. U.
Manthe, University of Bielefeld (Germany) for the access to the
computer cluster and GAUSSIAN-03 set of program.
Appendix A. Supplementary data
[21] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jfluchem.2009.11.001.
[22] S.H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 58 (1980) 1200–1211.
[23] M.J. Frisch, J.A. Pople, J.S. Binkley, J. Chem. Phys. 80 (1984) 3265–3269.
[24] A.D. McLean, G.S. Chandler, J. Chem. Phys. 72 (1980) 5639–5648.
[25] NBO 5.0 written by E.D. Glendening, J.K. Badenhoop, A.E. Reed, J.E. Carpenter, J.A.
Bohmann, C.M. Morales, F. Weinhold, Theoretical Chemistry Institute, University
of Wisconsin, Madison, 2001.
References
[1] K. Nilsson, K. Mosbach, Meth. Enzymol. 104 (1984) 56–69.
[26] MOLDEN Program, Ver. 3.8 (2005) written by G. Schaftenaar, CAOS/CAMM Center
Nijmegen, Toernooiveld, Nijmegen, the Netherlands, 1991. See also: G. Schafte-
naar, J.H. Noordik, J. Comput.-Aided Mol. Des. 14 (2000) 123–134. For more
detailed information see: http://www.cmbi.ru.nl/molden/.
[2] M.P. Johnson, M. Baez, G.E. Jagdmann Jr., T.C. Britton, T.H. Large, D.O. Callagaro, J.P.
Tizzano, J.A. Monn, D.D. Schoepp, J. Med. Chem. 46 (2003) 3189–3192.
[3] D.A. Barda, Z.-Q. Wang, T.C. Britton, S.S. Henry, G.E. Jagdmann, D.S. Coleman, M.P.
Johnson, S.L. Andis, D.D. Schoepp, Bioorg. Med. Chem. Lett. 14 (2004) 3099–3102.