Radical addition of iodine monochloride to vinylidene fluoride

Article abstract of DOI:10.1016/S0022-1139(99)00308-5

The radical addition of iodine monochloride to 1,1-difluoroethylene or vinylidene fluoride (VDF) leading to a ClCF₂CH₂I (I)/ClCH₂CF₂I (II) mixture is described. Four different ways of initiation (thermal, photochemical, presence of radical initiators or redox catalysts) were used and all of them led to a high amount (≥98%) of isomer (I). The percentages of (I) and (II) isomers were determined by ¹⁹F NMR and they were also deduced from those of ClCF₂CH₃ and ClCH₂CF₂H obtained by selective reduction of the iodine atom of the product mixture, by tributylstannane. The reactivity of ICl to VDF and the high proportion of isomer (I) were interpreted by means of a thermodynamical approach from the enthalpy of formation of (I) and (II), respectively, determined by semi-empirical computations. In addition, heats of formation of both isomers and interactions between the SOMO of radicals and the HOMO of the fluoroolefin show that the mechanism of such a reaction occurs via the addition of I· to the less fluorinated side of VDF.

Full text of DOI:10.1016/S0022-1139(99)00308-5

Journal of Fluorine Chemistry 103 (2000) 145±153
Radical addition of iodine monochloride to vinylidene ¯uoride
*
M. Kharroubi, A. Manseri, B. Ameduri , B. Boutevin
Â
Ecole Nationale Superieure de Chimie de Montpellier, UMR 5076 (CNRS), Laboratory of Micromolecular Chemistry, 8 rue de l'Ecole Normale,
Â
Â
Â
F - 34296, Montpellier Cedex 5, France
Received 2 August 1999; accepted 11 November 1999
Abstract
The radical addition of iodine monochloride to 1,1-di¯uoroethylene or vinylidene ¯uoride (VDF) leading to a ClCF2CH2I (I)/ClCH2CF2I
(II) mixture is described. Four different ways of initiation (thermal, photochemical, presence of radical initiators or redox catalysts) were
used and all of them led to a high amount (ꢀ98%) of isomer (I). The percentages of (I) and (II) isomers were determined by 19F NMR and
they were also deduced from those of ClCF2CH3 and ClCH2CF2H obtained by selective reduction of the iodine atom of the product
mixture, by tributylstannane. The reactivity of ICl to VDF and the high proportion of isomer (I) were interpreted by means of a
thermodynamical approach from the enthalpy of formation of (I) and (II), respectively, determined by semi-empirical computations. In
addition, heats of formation of both isomers and interactions between the SOMO of radicals and the HOMO of the ¯uoroole®n show that
the mechanism of such a reaction occurs via the addition of Iꢁ to the less ¯uorinated side of VDF. # 2000 Elsevier Science S.A. All rights
reserved.
Keywords: Vinylidene ¯uoride; Iodine monochloride; Radical addition; Reduction; Semi-empirical computations; Nuclear magnetic resonance
1. Introduction
The literature quotes that iodine monochloride reacts with
VDF under ionic conditions leading to the formation of one
isomer ClCF2CH2I (I) only [45,51]. However no experiment
was conducted under radical conditions. The goal of this
paper concerns a deeper investigation of the addition of ICl
to VDF and involves a study of different initiators in order to
determine the best compromise between yield and selectiv-
ity (to isomer (I)). In addition, semi-empirical calculations
give an understanding of the experimental observations.
Vinylidene ¯uoride (VDF) or 1,1-di¯uoroethylene has
been extensively used in polymerization [1,2], copolymer-
ization [3,4], and telomerization [5]. In the case of telomer-
ization a wide range of telogens or transfer agents used may
possess various bonds, cleaved by the initiation step: C±H
[6±9], Si±H [10], P±H [11±13], S±H [14±17], S±S [18,19],
C±X (X ˆ Cl [8,9,20,21], Br [9,22±30], I [21,31±41]), O±Y
(Y ˆ F [42,43], Cl [43], Br [44]), H±Z (Z ˆ Br [45], I
[45,46]), S±X (X ˆ Cl [47±49], Br [50]) and I±Cl
[45,51]. Table 1 lists non-exhaustive researches involving
the transfer agents, different ways of initiation and the
telomerization degrees.
2. Experimental section
2.1. General comments
Among these transfer agents, iodine monochloride
seems to be an attractive reactant since the expected
product formed exhibits a CX2I (X ˆ F or H) end-group
which can be functionalized. Further, ICl is known to be a
very reactive transfer agent which has previously been
involved in radical additions to tri¯uoroethylene [52,53],
chlorotri¯uoroethylene [54±57] and hexa¯uoropropene
[55,58,59].
VDF was kindly supplied by Elf Atochem and Solvay.
Iodine monochloride (either pure or in a methylene chloride
solution) and tributylstannane were provided by Aldrich and
did not require puri®cation prior to use.
After reaction, the crude reaction mixture or the products
after work-up with alkali sodium bisul®te, were analyzed by
gas chromatography (GC) using a Delsi apparatus (model
330) equipped with a SE 30 column, 3m  1/8 in. (i.d). The
nitrogen pressure at the entrance to the column was main-
tained at 0.6 bar and the detector and injector temperatures
were 260 and 2558C, respectively. The temperature
* Corresponding author. Tel.: ‡33-4-67-14-4368; fax: ‡33-4-67-14-
7220.
E-mail address: ameduri@cit.enscm.fr (B. Ameduri).
Â
0022-1139/00/$ ± see front matter # 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 9 9 ) 0 0 3 0 8 - 5

146
M. Kharroubi et al. / Journal of Fluorine Chemistry 103 (2000) 145±153
Table 1
Telomerization of vinylidene fluoride with different telogensa
Telogens
Method of initiation
Structures of Telomers
References
[6,7]
HOCH2±H
DTBP, DBP or AIBN
DTBP
HOCH2(C2H2F2)nH
n > 10
Cl3C±H
Cl3C(C2H2F2)nH
n > 3
[8,9]
Cl3Si±H
H2P±H
UV
UV
Cl3SiCH2CF2H
H3 nP(C2H2F2)nH
n ˆ 1,2
[10]
[11]
(EtO)2(O)P±H
DTBP or perester
(EtO)2(O)P(C2H2F2)nH
n ˆ 1±5
[12,13]
HS±H
X ray
H2 nS(CH2CF2)nH
F3CSCH2CF2H
HOC2H4S(C2H2F2)nH
n ˆ 1,2 (or 1±6)
RS(C2H2F2)nSR
n ˆ 1±6
[14]
F3CS±H
X ray
[15]
HOC2H4S±H
AIBN, DBP (or DTBP)
[16,17]
RS±SR
UV
[18,19]
[8,9,20,21]
[22±25]
[26]
(R=CH3 or CF3)
RCCI2±Cl
Peroxide, FeCl3/Ni (or benzoin) or CuCl2
RCCl2(C2H2F2)nCl
n ˆ 1±4
(R=Cl, CO2CH3, CH2OH)
BrCF2±Br
Thermal (1908C), UV or g ray
BrCF2(C2H2F2) nBr
variable n
BrCF2CF2±Br
Br2CF±Br
Peroxide
UV
BrCF2CF2(C2H2F2)nBr
variable n
Br2CF(C2H2F2) nBr
n ˆ 1,2
[27]
Cl3C±Br
UV
Br3C(C2H2F2)nBr
low n
[9,23,24]
[28]
Br3C±Br
DTBP
Cl3C(C2H2F2)nBr
n ˆ 1,2
CF3CFBr±Br
CF3CBr2±Br
BrCF2CFCl±Br
g ray
CF3CFBr(C2H2F2)nBr
variable n
[25]
g ray
CF3CBr2(C2H2F2)nBr
variable n
[25]
Peroxides or salts
BrCF2CFCl(C2H2F2)nBr
variable n
[29,30]
CnF2n‡1±I
n ˆ 1±8
Thermal (180±2208C) UV
UV
CnF2n‡1(C2H2F2)pI
Variable p
[21,31±34]
[27,35]
[36]
AIBN/sc CO2
Peroxide
[27,37]
[34]
FeCl3/Ni
LTA
ICF2I
[38,39]
[40]
ICnF2n±I
n ˆ 2,4,6
ClCF2CClF±I
Thermal (180±2008C)
I(C2H2F2)pCnF2n(C2H2F2)qI
p,qˆ1,2,3
1818C/26h
ClCF2CClF(C2H2F2)nI
n ˆ 1±6
[22]
HCF2±I
UV
HCF2CH2CF2I
RF,WO(C2H2F2)nY
Variable n
[41]
R
F,WO±Y
Thermal ( 1458C to RT)
[42±44]
(W=H,Cl or Br and Y=F, Cl or Br)
Z±H (Z=Br, I)
Radical or ionic conditions
DBP
CH3CF2Z
[45,46]
[47±49]
FSO2±Cl
FSO2(C2H2F2)nCl
n ˆ 1±3
Cl3CSO2±Br
DBP
Cl3CSO2(C2H2F2)nBr
n ˆ 1,2
[50]
Cl±I
Ionic conditions
ClCF2CH2I
[45,51]
a AIBN, DBP, DTBP and LTA stand for azobisisobutyronitrile, dibenzoyl peroxide, ditert-butyl peroxide and lead tetraacetate, respectively.
programme started from 508C and reached 2508C at a heat-
ing rate of 158C min 1. The GC apparatus was connected to
a Hewlett Packard integrator (model 3390) which automati-
cally calculated the area of each peak on the chromatogram.
spectra were recorded on Bruker AC 200, 250 or WM 360
instruments, using deuterated chloroform and CFCl3 as
solvent and internal reference, for 19F NMR and tetra-
methylsilane as reference for H NMR. The letters d, t
and q designate doublet, triplet and quartet, respectively.
Coupling constants are given in Hertz (Hz).
1
1
The products were characterized by H and 19F NMR
1
spectroscopy at room temperature. The 19F and H NMR

M. Kharroubi et al. / Journal of Fluorine Chemistry 103 (2000) 145±153
147
2.2. Radical addition of ICl to vinylidene fluoride
For example in Experiment 15, VDF (12.3 g, 0.18 mol),
ICl (15.2 g, 0.09 mol), and 50 g of methylene chloride were
stirred at 1308C for 13 h. After reaction, 4.3 g (mainly
unreacted VDF) were released and after opening the auto-
clave, the purple product mixture was distilled. After
removal of the methylene choride, the monoadduct was
fractionated, bp ˆ 988C and 12.2 g (0.065 mol) of a pink
liquid were obtained (yield ˆ 60%).
2.2.1. Reaction in small Carius tubes
Radical additions of ICl to VDF were performed in thick
borosilicate Carius tubes in a batch process (length:
130 mm, internal diameter: 10 mm, thickness: 3.0 mm,
for a total volume of 8 cm3). After placing the initiator
or the catalyst and liquid ICl solution, the tube was con-
nected to a vacuum line and purged several times by
evacuating and ¯ushing with helium. After ®ve thaw-freeze
cycles at least to remove oxygen, the VDF was trapped
under vacuum, in the tube frozen in liquid nitrogen, after a
release from an intermediate metallic pressure calibrated
container. The required amount of VDF (0.500±1.000 g
with an error of 8 mg) introduced into the tube was assessed
by the relative drop of pressure in the release container,
initially fed by a cylinder of 300 g of VDF. The tube, under
vacuum and immersed in liquid nitrogen was sealed and
placed either into the cavity of a shaking oven at the required
temperature or on a clamp connected to a mechanical stirrer
(at the speed of 45 rpm) and irradiated by a UV lamp or
placed in the cavity of a thermoregulated oven and heated
while shaking.
Isomers (I) and (II) have not been separated.
Isomer
(I):
1-chloro-1,1-di¯uoro-2-iodo
ethane
ˆ
ClCF2CH2I 19F NMR (CDCl3) d: 50.63 (t, JFH
3
12.7 Hz, 2F). 1H NMR (CDCl3) d: 3.85 (t, 3JHF ˆ 12.4 Hz,
2H).
Isomer
(II):
1-iodo-1,1-di¯uoro-2-chloro-ethane
ICF2CH2Cl. 19F NMR (CDCl3) d: 36.8 (t, 3JFH ˆ 13.3Hz,
2F). 1H NMR (CDCl3) d: 5.90(t, 3JHF ˆ 13.1 Hz, 2H)
2.3. Reduction of (I)/(II) mixture into (I, red)/(II, red)
In a two necked round bottom ¯ask equipped with a
vigreaux column ®tted with a condenser 3.0 g (0.013 mol)
of IC2F2H2Cl were introduced under Argon. Then 4.0 g
(0.013 mol) of tributylstannane were added dropwise at
about 5±108C through a septum.
(I, red)/(II, red) mixture was trapped as soon as formed
in a round bottom ¯ask containing 1,1,2-tri¯uoro-1,2,2-
trichloroethane (F-113) cooled in a NaCl/ice bath.
Reduced derivative isomer (I, red): 1-chloro-1,1-di¯uor-
oethane CH3CF2Cl (colourless liquid). 19F NMR (CDCl3
1H,1H,2H,2H-iodoper¯uorodecane was used as the inter-
nal standard. The yield was calculated from the calibration
curve
A(ClC2F2H2I)/A(standard)
versus
weight-
(ClC2F2H2I)/weight(standard) determined by GC (A desig-
nated the area of the peak noted on the gas chromatogram,
assigned to the product).
After reaction, the tube was frozen in liquid nitrogen and
then opened and the total product mixture was analyzed by
NMR at room temperature (RT).
1
3
and F113) d: 45.50 (q, JFH ˆ 15.3 Hz, 2F). H NMR
3
(CDCl3 and F-113) d: 2.10 (t, JHF ˆ 14.8 Hz, 3H).
Isomer
(II,
red):
1,1-di¯uoro-2-chloroethane
HCF2CH2Cl (colourless liquid). 19F NMR (CDCl3and F-
1
2
3
113) d: 115.50 (dt, JHF ˆ 49.7 Hz, JHF ˆ 13.4 Hz). H
2.2.2. Reaction in medium Carius tubes
2
NMR (CDCl3 and F-113) d: 6.35 (tt, JHF ˆ 49.5 Hz,
For example, in Experiment 14, VDF (3.8 g, 46.3 mmol)
was condensed into a Carius tube (thickness: 2.5 mm;
length: 175 mm; internal diameter: 20 mm), containing
ICl (8.1 g, 50.0 mmol), benzophenone (0.169 g, 0.9 mmol)
and 10 g of methylene chloride. The tube was sealed and
then irradiated with UV source (mercury lamp Philips HPK
125 W) at room temperature.
3JHF ˆ 13.6 Hz, HCF2); 6.12 (td, 3JHF ˆ 12.2 Hz,
3JHH ˆ 6.8 Hz, CH2Cl).
3. Results and discussion
3.1. Addition of iodine monochloride to vinylidene
fluoride (VDF)
2.2.3. Reaction in autoclave
In contrast to Haszeldine and Osborne [45] and Park
et al. [51] who performed their investigations under
ionic initiations and showed that only one isomer was
produced, the radical addition of ICl to this ole®n led
to a mixture of two isomers with yet a major product, as
follows:
Other reactions were also performed in a HC 276 200 ml
Hastelloy autoclave. First, the vessel was charged with the
required amounts of ICl and methylene chloride and closed.
It was then cooled in an acetone/liquid nitrogen mixture to
condense VDF, after two cycles of degassing with nitrogen
under pressure followed by vacuum. By double weighting, a
suitable amount of gas was introduced under pressure and
trapped. The autoclave was then brought to the desired
temperature and the reaction medium was stirred with a
magnetic bar. After the reaction, the vessel was cooled down
to room temperature, then put into ice while the non-reacted
VDF was released.
ICl ‡ F2CˆˆCH2 ! ClCF2CH2I ‡ ICF2CH2Cl
ꢂI†
ꢂII†
The `normal' adduct (I) and the `reverse' adduct (II) were
observed, the ratio of which was determined by 19F and 1H
NMR spectroscopy. The addition of ICl to such an alkene

148
M. Kharroubi et al. / Journal of Fluorine Chemistry 103 (2000) 145±153
Table 2
Addition of iodine monochloride to vinylidene fluoride (VDF)a
Experiment. no.
Ro
Initiator or catalyst
co
T (8C)
t (h)
Yield (%)
I/II (%)
1
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.3
1.1
0.5
1.0
1.0
Thermal
Thermal
P-16
±
RT
54
21
21
21
21
21
21
20
20
20
20
20
20
20
18
21
18
55.6
49.6
49.7
51.1
48.9
53.3
57.3
62.2
55.2
8.2
98.9/1.1
99.2/0.8
99.5/0.5
98.7/1.3
97.8/2.2
99.7/0.3
99.5/0.5
99.7/0.3
99.3/0.7
100/0
2
±
3
0.02
±
54
4
Thermal
DBP
95
5
0.02
±
95
6
Thermal
UV
140
20
7
±
8
DTBP
CuCl
0.02
0.02
1.00
0.03
±
140
140
20
9
10
11
12
14
15
16
17
Tempo/UV
DPPH/UV
Thermal
UV/PhCOPh
Thermal*
Thermal*,
Thermal*,
20
15.0
58.6
70.1
60.3
54.8
61.5
100/0
54
99.4/0.6
100
0.02
±
RT
60
100
ꢁ
ꢁ
±
RT
60
99.1/0.9
99.5/0.5
±
a Reaction conditions (in methylene chloride or in acetonitrileꢁ) and methods of initiation: Ro and co mean initial [ICl]o/[VDF]o, [initiator or catalyst]o/
[VDF]o molar ratios; the yields were determined by gas chromatography using a calibration curve of 1H,1H,2H,2H-tetrahydroperfluorodecane as internal
standard ; the experiments were performed in borosilicate Carius tubes or in a Hastelloy autoclave*) (DBP, DTBP, P-16, PhCOPh, DPPH, and Tempo stand
for dibenzoyl peroxide, ditert-butyl peroxide, t-butylcyclohexyl peroxicarbonate, benzophenone, 2,2-diphenyl-1-picryl-hydrazyl and 2,2,6,6-tetramethyl-1-
piperidinyloxy, respectively).
was performed with different ways of initiation and under
various conditions (Table 2).
whereas the CH2Cl group underwent a slight low ®eld shift
from 5.90 ppm (isomer (II)) to 6.12 (isomer (II, red)). The
characteristic triplet (2JHF ˆ 49.5 Hz) of triplets
(3JHF ˆ 13.6 Hz) assigned to CF2H end-group was clearly
3.1.1. NMR characterization of isomers (I) and (II)
The 19F NMR spectrum exhibits two triplets assigned to
CF2 groups of both isomers. The more abundant one
corresponds to ClCF2CH2I and is represented by a triplet
(3JFH ˆ 12.7 Hz) centered at 50.63 ppm. The second
signal assigned to isomer (II) ICF2CH2Cl, is a triplet
(3JHF ˆ 13.3 Hz) centered at 36.8 ppm.
1
noted on the H NMR spectrum.
3.1.3. Influence of the initiation on the yield of the reaction
and on the (I)/(II) ratio
Table 2 lists the values of the (I)/(II) ratio obtained from
1H and 19F NMR of the products obtained from different
initiating systems.
1
The H NMR spectrum exhibits two groups of signals.
That centered at 3.85ppm, assigned to isomer (I) represents
triplet (3JHF ˆ 12.4 Hz) whereas the triplet
Several ways of initiation have been chosen: photoche-
mical, copper (I) catalyst or radical initiators, and ®nally
thermal. In all these cases, the reactions were performed
batch-wise and the conditions of the reaction were ®xed by
two initial molar ratios: Ro represents the initial molar ratio
of [ICl]o to [VDF]owhereas co stands for the initial [catalyst
or initiator]o/[VDF]o molar ratio. Usually, Ro was higher
than 1 since our objective is the monoadduct. A ratio of 1.0
led to satisfactory results when chlorotri¯uoroethylene [57]
or tri¯uoroethylene [53] were used as the ¯uorinated
a
(3JHF ˆ 13.1 Hz) centered at 5.90 ppm is ascribed to isomer
(II).
The ratio of both these above integrals slightly varied
according to the reaction conditions (Table 2).
3.1.2. Reduction of (I)/(II) mixture
The selective reduction of the iodine atoms of both
products was performed in the presence of tributylstannane
as follows:
2
alkenes. co was about 1 Â 10 2±4 Â 10 as used in pre-
vious works [53,57,60±62]. The reactions initiated photo-
chemically or with metallic catalysts or radical initiators
were performed in sealed Carius tubes, whereas those
investigated thermally took place either in Hastelloy auto-
claves or in Carius tubes.
ClCF2CH2I ‡ ClCH2CF2I Sn!Bu3H ClCF2CH3 ‡ ClCH2CF2H
ꢂI†
ꢂII†
ꢂI;red†
ꢂII;red†
The reaction was selective towards iodine and quantita-
tive and kept the same 98/2 isomer ratio. These derivatives
were characterized by both 1H and 19F NMR. The chemical
shift of the ClCF2 group underwent a slight low ®eld shift
(centered at 50.63 ppm for (I) to 45.50 ppm for (I, red)).
The 19F NMR spectrum of the minor isomer exhibits a
triplet (HCF2) with a smaller Dd and with a high ®eld shift
from ca. 36.8 ppm (ICF2) to 115.50 ppm (HCF2),
3.1.3.1. Photochemical initiation. In contrast to previous
radical additions of ICl performed photochemically, that to
VDF was mainly initiated thermally (Table 2). Our
investigations were also performed photochemically
under a high pressure mercury lamp irridiation with

M. Kharroubi et al. / Journal of Fluorine Chemistry 103 (2000) 145±153
149
benzophenone as photosensitizer (Table 2). Two different
ways of stirring were utilized but did not influence the yield
or the ratio of isomers: the Carius tube stood vertically, the
reactants being stirred by means of a magnet bar or the tube
was shaken mechanically describing a circle at a speed of
50 rpm. In these conditions, we have shown that the
photochemical initiation yielded the two isomers (I) and
(II) in 98/2 molar ratio.
Under similar conditions, the photochemical addition of
ICl to VDF was slower than that to chlorotri¯uoroethylene
[57] and to tri¯uoroethylene [53], since after 1.5 h-reaction
time, complete conversion of ICl was not achieved in the
®rst case, in contrast to the reaction involving CTFE.
However, the former reaction appears as regioselective
((I)/(II) ˆ 98/2) as that using TrFE (most of the results
have led to a ClCF2CFHI/ICF2CFHCl molar ratio of 97/3
[53]).
radical initiator [57], no propagation occurred when VDF
was used in the same conditions and even with a three fold
excess of VDF. This may be due to the reactive CFClI end-
group of Cl(CF2CFCl)xI (x ˆ 1, 2) telomers allowing them
to be more reactive and thus leading to Cl(CF2CFCl)x‡1I, as
observed in the telomerization of CTFE with per¯uoroalkyl
iodides [65]. But, the CH2I end-group of I(VDF)Cl is less
reactive than CFClI and thus does not form higher molecular
weight telomers. Moreover, the ClCF2 end-group is too
stable to allow a further reactivation. Hence, in the case
of VDF, ICl appears as a very good transfer agent thanks to
an easily cleavable bond, in contrast to per¯uoroalkyl
iodides [22,27,30±37], diethylphosphite (the transfer con-
stant cT is cT ˆ 0.34) [13], HCCl3 (cT ˆ 0.06) [9] or metha-
nol (0.008) [66].
According to the experimental results above, it was worth
attempting to understand the mechanism of the reaction.
First, using two solvents (acetonitrile and methylene chlor-
ide, Table 2) which have different dielectric constants
(eCH CN ˆ 37 and eCH Cl ˆ 9 at 298 K), it was noted that
3.1.3.2. Redox catalysis. Redox catalysts are usually used
in telomerization to favor the formation of low molecular
weight-adducts, particularly the monoadduct [5,60±62].
Copper (I) and copper (II) salts were utilized as catalysts
and did not show a great influence on the yield and on the
ratio of isomers (Table 2).
3
2
2
the results were unchanged. This polar effect may show that
an ionic mechanism is improbable. Second, the use of a
radical scavenger (e.g., 2,2-diphenyl-1-picryl-hydrazyl and
Tempo) (Table 2) led to a decrease in yield. Hence, both
these results are in favour of a radical mechanism, and to
study this radical mechanism, we used semi-empirical
computations.
3.1.3.3. Radical initiation. The radical additions of ICl to
VDF involving initiators were performed at a temperature at
which their half life is 1 h: tert-butylcyclohexyl
peroxidicarbonate (P-16) at 608C; dibenzoyl peroxide
(DBP) at 958C and ditert-butyl peroxide (DTBP) at 1408C.
As noted for the radical additions of ICl to CTFE [57] and
TrFE [53], no polyVDF was produced even with a three fold
excess of VDF to ICl. This can be explained by the assumed
high transfer constant of ICl at 808C, compared with per-
¯uoroalkyl iodide which led to higher molecular weight
telomers from the same Ro molar ratio [63].
3.2. Semi-empirical computations
We have attempted to explain the above results and to
propose a reaction mechanism by means of quantum chem-
istry by using a thermodynamic approach and then a mole-
cular orbital approach. Since experimental values of heats of
formation for our compounds are unavailable from the
literature, we have used semi-empirical methods: MNDO
[67] and AM1 [68] implemented in the MOPAC v.6.0
program [69]. Radicals, i.e. open shell systems, were
treated by the Unrestricted Hartree-Fock (UHF) method,
whereas the Restricted Hartree-Fock (RHF) method was
used for closed shell systems. The `precise option' was
employed to ensure very strict convergence criteria. The
geometries of the reactants, products and radical intermedi-
ates were ®rst fully optimized by the Davidon±Fletcher±
Powell (DFP) method [70,71]. In the same way, all sta-
tionary points were characterized as true minima by the
force method [72].
3.1.3.4. Thermal initiation. The thermal addition of ICl to
VDF was performed at room temperature, at 60, 95 and
1408C in Carius tubes and in a Hastelloy autoclave, and was
compared to the telomerization of this fluoroolefin with
perfluorobutyl iodide, C4F9CH2CF2I, (CF3)2CFI and
C5F11CF(CF3)I [64]. A previous study has shown that the
more electrophilic the telogenic radical, the higher the
amount of normal adduct and even at 1908C a telomeric
distribution was produced from Ro ˆ 1.
Interestingly, only the monoadduct was produced what-
ever the temperature and, at 1308C, the reaction was regio-
selective leading to isomer (I), only. It can be assumed that
the metals of the Hastelloy material (iron, chromium, nickel,
molybdenum, tungsten, . . .) may enhance the reaction,
since, at the same temperature, similar reactions performed
in a quartz Carius tube in the presence of copper (I) chloride
led to lower yields.
We have compared the reactivity of VDF with ICl
according to a thermodynamic approach taking into account
computations of heats of formation (DHf) of different
systems studied (reactants, intermediate radicals and pro-
ducts). Table 3 lists the results obtained from MNDO and
AM1 methods. Interestingly, both these methods predict
1
that isomer (I) (DHf ˆ 389.5 kJ mol
(AM1) and
(MNDO)) is more stable than isomer
1
Indeed, while di- and triadducts have been observed for
the telomerization of CTFE with ICl in the presence of a
390.4 kJ mol
(II) (DHf ˆ 370.7 kJ mol 1 (AM1) and 357.3 kJ mol
1

150
M. Kharroubi et al. / Journal of Fluorine Chemistry 103 (2000) 145±153
Table 3
Calculated heats of formation (DHf, kJ mol 1) in the gas-phase at room-temperature of the different systems studied
Products
AM1
MNDO
exp. [80]
Literature [81]
343.1
I±Cl
19.2
28.0
351.9
390.4
357.3
106.7
121.2
330.5
217.6
431.4
348.5
17.8
335.1
CH2=CF2
IH2C±CF2Cl
IF2C±CH2Cl
Iꢁ
352.7
389.5
370.7
106.7
121.2
322.6
292.0
423.4
327.6
106.7
121.2
Clꢁ
IH2C±CꢁF2
IF2C±CꢁH2
ClH2C±CꢁF2
ClF2C±CꢁH2
(MNDO)). Such observations are in good agreement with
experimental results since the most stable species are the
major products.
As shown above, the radical mechanism is the most
probable. Hence, two radical mechanisms were suggested
to produce the ICH2CF2Cl isomer in higher amount: either
that in which Clꢁ radical is added ®rst to the more hindered
side of the ¯uoroalkene or a second which enables Iꢁ to react
on the less hindered side of VDF.
In order to explain such results, we have considered the
radical reactivity which is often explained in terms of steric
[73,74] and polar [73,74] effects, or by the thermodynamical
[73±79] or orbital [76,77] controls.
From the experimental results (Table 2), the iodine atom
prefers to be added on the hydrogenated side of the VDF
which is in good agrement with the in¯uence of steric
effects [73,74].
The addition of Iꢁ to VDF (Table 4) shows that the
1
enthalpy of the reaction 5 (DHr3 is to 76.6 kJ mol
1
(AM1) and 85.3 kJ mol
hydrogenated side) whereas DHr4
(MNDO) (addition to the
1
ˆ
46.0 kJ mol
1
(AM1) and ‡27.6 kJ mol (MNDO) for the addition to
the halogenated side.
A similar behaviour was observed for the addition of Clꢁ
to VDF since the attack on the less ¯uorinated side of the
alkene is the most exothermic (DHr5 and DHr6).
To explain why the addition of Clꢁ towards the hydro-
genated sides of VDF was not experimentally observed, an
analysis of the nature of the initiation can supply an answer.
If one considers that the organic initiation, the following
steps can be written (A designates the initiator):
D
A2 ! 2Aꢁ
(1)
(2)
(3)
Aꢁ ‡ Cl I ! A Cl ‡ Iꢁ
Aꢁ ‡ I Cl ! A I ‡ Clꢁ
ꢂfavoured†
ꢂminimised†
Considering the thermodynamic control, we ®rst calcu-
lated the heat of the different reactions which may occur.
The heat of reaction (DHr) can be related to the sum of the
enthalpies of formation of the products (DHfp) and of the
reactants (DHfr) as follows:
Iodine is a good leaving atom and reacts with VDF
leading to two isomers: ICH2CꢁF2 and ICF2CꢁH2. Accord-
ing to semi-empirical calculations (Table 3), the former
radical is thermodynamically favoured, leading to a major
ICH2CF2Cl isomer.
X
X
DHr ˆ
ꢂDHfp†
ꢂDHfr†
The results are summarized in Table 4. It is shown that the
additions of Iꢁ and Clꢁ to the hydrogenated side are always
favoured thermodynamically even if the center is sometimes
the more hindered. This con®rms that both these reactions
are not controlled by steric effects.
On the other hand, for thermal or photochemical initia-
tion, the following equation is considered:
D
I Cl ! Iꢁ ‡ Clꢁ
or hn
According to Table 3, both radicals produced can be
added to VDF since they are in the same amount, in contrast
to the mechanism from organic initiation, in which the
concentration of iodine, [Iꢁ], is higher than that of [Clꢁ].
However, from Experiments 4 and 6 compared to 5 and 8
(Table 2), it is observed that the yields are almost the same
and above all the (I)/(II) ratios do not vary. Indeed, in the
photochemical or thermal initiations, the [Clꢁ] concentra-
tion decreased compared to [Iꢁ]. Hence, the mechanism
involving the addition of Iꢁ is the more favourable. To
con®rm this statement, further investigation is under pro-
gress (ab initio computations) and will be reported in a
forthcoming paper [82].
Table 4
Calculated gas-phase enthalpies (DHr, kJ mol 1) for the proposed
reactions
Reaction
no.
Reaction
AM1
MNDO
1
2
3
4
5
6
I±Cl ‡ CH2=CF2 ! IH2C±CF2Cl
I±Cl ‡ CF2=CH2 ! IF2C±CH2Cl
Iꢁ ‡ CH2=CF2 ! IH2C±CꢁF2
Iꢁ ‡ CF2=CH2 ! IF2C±CꢁH2
Clꢁ ‡ CH2=CF2 ! ClH2C±CꢁF2
Clꢁ ‡ CF2=CH2 ! ClF2C±CꢁH2
17.6
10.5
‡22.6
85.3
‡1.2
76.6
46.0
‡27.6
200.7
117.8
191.9
96.1

M. Kharroubi et al. / Journal of Fluorine Chemistry 103 (2000) 145±153
151
Table 5
These ®rst semi-empirical calculations are in good agree-
ment with the experimental results. Indeed, they led to the
conclusion that isomer (I) was the major product and is the
most stable thermodynamically. This is evidence that the
reactions were under thermodynamic control. The additions
of Iꢁ and Clꢁ radicals to both sites of VDF showed that the
addition of Clꢁ was the most exothermal but it reacts to the
hydrogenated side: this disagrees with experimental obser-
vation in contrast to the addition of Iꢁ. However, on the basis
of Frontier Molecular Orbitals, more explanations have
been given.
Frontier molecular orbital energies (eV) and atomic orbital coefficients of
both sites of addition of olefin (C1 and C2) in the HOMO and LUMO
SOMO
AM1
HOMO
MNDO AM1
LUMO
MNDO AM1
MNDO
Iꢁ
Clꢁ
C1H2=C2F2
2pz (C1)
2pz (C2)
10.50
12.19
10.68
12.93
10.09
0.72
0.52
10.46
0.71
0.57
0.15
0.67
0.69
0.62
0.68
0.69
These results can be con®rmed by an orbital analysis by
the means of the Frontier Molecular Orbital Theory [83,84].
The results are gathered in Table 5 where SOMO, HOMO
and LUMO represents the Single Occupied Molecular
Orbital for the radical systems, the Highest Occupied
Molecular Orbital and the Lowest Unoccupied Molecular
Orbital of ole®ns, respectively. This table shows that inter-
actions between SOMO of radicals and HOMO of the
alkenes are the strongest, since the difference of energy
between SOMO±LUMO is greater than that of SOMO±
HOMO.
Interestingly, the SOMO of radical Clꢁ [eSOMO(Clꢁ) ˆ
12.19 eV (AM1) and 12.93 eV (MNDO)] is very deep
compared with that of Iꢁ [eSOMO(Iꢁ) ˆ 10.50 eV (AM1)
and to 10.68 eV (MNDO)] and the latter is close to
HOMO of VDF [eHOMO is ca. 10 eV]. Consequently, Iꢁ
adds ®rst to the alkene.
A more detailed explanation can be found by the orbital
atomic coef®cients of both ethylenic carbon atoms of the P
orbital, according to the Frontier Molecular Orbital Theory
[83,84] and from the Canadell's rule [76,77]: `a radical
would be mainly added on the carbon atom having the
greater coef®cient in the HOMO'.
According to Table 5, it is shown that the hydrogenated
carbon atom exhibits the greatest coef®cients in P orbitals:
2pz(C1) ˆ 0.72 whereas 2pz(C2) ˆ 0.52. This may explain
the value of the molar (I)/(II) ratio of 99/1 in favour of an
electrophilic radical addition of Iꢁ to the hydrogenated site
of CH2=CF2 (C1).
4. Conclusion
Amongst four different initiations possible for the radical
addition of iodine monochloride to VDF, the thermal way
led to good yields and to regioselective formation of the
ClCF2CH2I isomer. This is the most convenient way since it
does not require any expensive initiator or any metallic
catalysts which may be regarded as toxic species. Interest-
ingly, this thermal reaction was selective in monoaddition in
contrast to that utilizing per¯uoroalkyl iodides which pro-
duced telomeric distributions and this indicates that ICl
exhibits a high transfer constant. In addition no poly(viny-
lidene ¯uoride) was observed, even for a large excess of
VDF.
Semi-empirical computations were in agreement with the
formation of (I) as the major isomer and indicated that the
mechanism of such a reaction is mainly radical, Iꢁ acting
®rst on the CH2 group of VDF. However, deeper theoretical
studies are still required and are in progress.
Compared to chlorotri¯uoroethylene and tri¯uoroethy-
lene, this ¯uoroalkene seems less reactive as far as the
initiation method is concerned. Hence, comparison on
reactivity, yield and regioselectivity of similar reactions
involving hexa¯uoropropene and vinylidene chloride [85]
need further study and are under investigation, and will be
presented in a forthcoming paper [82].
5. Supporting materials
In conclusion, the addition of ICl to VDF occurs through
a radical mechanism in which the iodine radical reacts ®rst
with the ¯uoroole®n. Hence, the following mechanism can
MOPAC output ®les of all optimized molecules may be
obtained from the authors.
be proposed (taking into account Eqs 1±3 above):
ꢀ
Acknowledgements
I Cl ! Iꢁ ‡ Clꢁ
Initiation
Iꢁ ‡ H2CˆˆCF2 ! ICH2CꢁF2
The authors would like to thank the Elf Atochem and
Solvay companies for the gifts of 1H,1H,2H,2H-tetrahy-
droper¯uorodecane
oethane, respectively. These companies also supplied
vinylidene ¯uoride which was acknowledged.
ICH2CꢁF2 ‡ H2CˆˆCF2
and
1,1,2-trichloro-1,2,2-tri¯uor-
! ICH2CF2CH2CꢁF2
Propagation ꢂif any†
ꢀ
ICH2CꢁF2 ‡ I Cl ! ICH2CF2Cl ‡ Iꢁ
ICH2CF2CH2CꢁF2 ‡ I Cl ! ICH2CF2CH2CF2Cl ‡ Iꢁ
Termination by Transfer of ICI
References
[1] S. Russo, K. Behari, S. Changji, M. Pianca, E. Barchiesi, G. Moggi,
Polymer 34 (1993) 4777.
Iꢁ ‡ Iꢁ ! I2
Termination by recombination

152
M. Kharroubi et al. / Journal of Fluorine Chemistry 103 (2000) 145±153
Â
Â
[2] S. Weinhold, M.H. Litt, J.B. Lando, J. Polym. Sci. Polym. Lett. Ed.
[40] A. Manseri, B. Ameduri, B. Boutevin, R.D. Chambers, G.
Caporiccio, A.P. Wright, J. Fluorine Chem. 74 (1995) 59.
[41] J.P. Sloan, J.M. Tedder, J.C. Walton, J. Chem. Soc. Perkin Trans. II
(1975) 1846
17 (1979) 585.
[3] C. Tournut, Thermoplastic copolymers from vinylidene fluoride, in:
J. Scheirs (Ed.), Modern Fluoropolymers, Chap. 31, pp. 577, Wiley,
New York, 1997.
[42] A. Sekiya, K. Ueda, Chem. Lett. (1990) 609
[43] B.B. Randolph, D.D. DesMarteau, J. Fluorine Chem. 64 (1993) 129.
[44] J.D.O. Anderson, D.D. DesMarteau, J. Fluorine Chem. 77 (1996)
147.
[4] A. Van Cleeff, Fluoroelastomers, in: J. Scheirs (Ed.), Modern
Fluoropolymers, Chap. 32, pp. 597, Wiley, New York, 1997.
Â
[5] B. Ameduri, B. Boutevin, Telomerisation reactions of fluorinated
alkenes, in: R.D. Chambers (Ed.), Topics in Current Chemistry, Vol.
192, pp. 165±233, Springer, Heidelberg, 1997.
[6] J. Oku, R.J.H. Chan, H.K. Hall, O.R. Hughes, Polym. Bull. 16 (1986)
481.
[45] R.N. Haszeldine, J.E. Osborne, J. Chem. Soc. (1956) 61
[46] C.S. Rondestvedt Jr., J. Org. Chem. 42 (1977) 1985.
[47] N.A. Grigor'ev, L.S. German, R.K. Freidlina, Bull. Acad. Sci. USSR
(1979) 863
Â
[7] M. Duc, B. Ameduri, B. Boutevin, M. Kharroubi, J.M. Sage,
[48] N.A. Grigor'ev, L.S. German, R.K. Freidlina, Izv. Akad. Nauk SSSR
Ser. Khim. (1979) 918 (Chem. Abstr. 91 (1979) 19833z)
[49] G.V.D. Tiers, US Patent, 2,846,472 (1958) (to 3 M)
[50] M.Y. Zhu, S.S.X. Sun, Y.H. Zhang, X.K. Jiang, Ch. Chem. Lett. 4
(1993) 583.
Macromol. Chem. Phys. 199 (1998) 1271.
[8] Y. Toyoda, N. Sakauchi, C. Nobuo, Japan Patent 77319 (to Kureha
Chem. Ind. Co. Ltd.), 02-12-1967
Â
[9] M. Duc, M. Kharroubi, B. Ameduri, B. Boutevin, J. Polym. Sci. Part.
A, Polym. Chem., submitted for publication.
[10] T.N. Bell, R.N. Haszeldine, M.J. Newlands, J.B. Plumb, J. Chem.
Soc. (1965) 2107
[51] J.D. Park, J. Abramo, M. Hein, D.N. Gray, J.R. Lacher, J. Org.
Chem. 23 (1958) 1661.
[52] E.R. Bissel, J. Org. Chem. 29 (1964) 252.
Â
[53] B. Ameduri, B. Boutevin, M. Kharroubi, G.K. Kostov, P. Petrova, J.
Fluorine Chem. 91 (1998) 41.
[11] G.M. Burch, H. Goldwithe, R.N. Haszeldine, J. Chem. Soc. (1963)
1083
[12] R.N. Haszeldine, D.L. Hobson, D.R. Taylor, J. Fluorine Chem. 8
(1976) 115.
[54] R.N. Haszeldine, J. Chem. Soc. (1955) 4291
[55] M. Hauptschein, M. Braid, A.H. Fainberg, J. Am. Chem. Soc. 83
(1961) 2495.
[13] M. Duc, B. Ameduri, B. Boutevin, Macromol. Chem. Phys.,
submitted for publication.
[56] R.E. Banks, R.N. Haszeldine, P. Mitra, T. Myerscough, S. Smith, J.
Macromol. Sci. Chem. A8 (1974) 1325.
[14] J.F. Harris, F.W. Stacey, J. Am. Chem. Soc. 85 (1963) 749.
[15] J.F. Harris, F.W. Stacey, J. Am. Chem. Soc. 83 (1961) 840.
[16] B. Boutevin, Y. Furet, Y. Hervaud, G. Rigal, J. Fluorine Chem. 69
(1994) 11.
Â
[57] B. Ameduri, B. Boutevin, G.K. Kostov, P. Petrova, J. Fluorine Chem.
74 (1995) 261.
[58] E. Meissner, J. Myszkowski, J. Szymanowski, Tenside Surf. Det. 32
(1995) 261.
 Â
[59] M. Kharroubi, A. Manseri, B. Ameduri, B. Boutevin, J. Fluorine
[17] B. Boutevin, Y. Furet, Y. Hervaud, G. Rigal, J. Fluorine Chem. 74
(1995) 37.
[18] D.W.A. Sharp, H.T. Miguel, Isr. J. Chem. 17 (1978) 144.
[19] G. Haran, D.W.A. Sharp, J. Chem. Soc. Perkin Trans. I (1972) 34.
[20] B. Boutevin, Y. Furet, L. Lemanach, F. Vial-Reveillon, J. Fluorine
Chem. 47 (1990) 95.
Chem., submitted for publication.
Â
[60] B. Ameduri, B. Boutevin, Telomerization, in: D. Bloor, R.J. Brook,
M.C. Flemings, S. Mahajan, R.W. Cahn (Eds), Encyclopedia of
Advanced Materials, Pergamon Press, Oxford, 1994, p 2766.
[61] A. Battais, B. Boutevin, Y. Pietrasanta, T. El Sarraf, Makromol.
Chem. 183 (1982) 2359.
[21] Q.Y. Chen, Z.Z. Ma, X.K. Jiang, Y.F. Zhang, S.M. Jia, Hua Hsueh
Hsueh Pao 38 (1980) 175, (Chem. Abstr. 94 (1981) 174184u).
[22] M. Hauptschein, M. Braid, F.E. Lawlor, J. Am. Chem. Soc. 80
(1958) 846.
Â
[62] B. Ameduri, B. Boutevin, C. Lecrom, L. Garnier, J. Polym. Sci, Part
A, Poly Chem. 30 (1992) 49.
Â
[63] B. Ameduri, B. Boutevin, unpublished results.
[23] J.M. Tedder, J.C. Walton, J. Chem. Soc. Faraday I 62 (1966) 1859.
[24] J.M. Tedder, J.C. Walton, J. Chem. Soc. Faraday I 70 (1974) 308.
[25] R.D. Chambers, L.D. Proctor, G. Caporiccio, J. Fluorine Chem. 70
(1995) 241.
 Â
[64] J. Balague, B. Ameduri, B. Boutevin, G. Caporiccio, J. Fluorine
Chem. 73 (1995) 237.
 Â
[65] J. Balague, B. Ameduri, B. Boutevin, G. Caporiccio, J. Fluorine
Á
[26] S. Modena, M. Pianca, M. Tato, G. Moggi, S. Russo, J. Fluorine
Chem., submitted for publication.
Â
[66] M. Duc, B. Ameduri, B. Boutevin, M. Kharroubi, Polym. Prepr. Am.
Chem. 43 (1989) 15.
[27] J.P. Sloan, J.M. Tedder, J.C. Walton, J. Chem. Soc. Faraday Trans. I
69 (1973) 1143.
Chem. Soc. Polym. Div. 39 (1998) 816.
[67] M.J.S. Dewar, G. Thiel, J. Am. Chem. Soc. 99 (1977) 4899.
[68] M.J.S. Dewar, E.G. Zoebisch, E.F. Haley, J.J.P. Stewart, J. Am.
Chem. Soc. 107 (1985) 3902.
[28] D.S. Ashton, D.J. Shand, J.M. Tedder, J.C. Walton, J. Chem. Soc.
Perkin Trans. II 4 (1975) 320.
[29] P. Tarrant, M.R. Lilyquist, J. Am. Chem. Soc. 77 (1955) 3640.
[30] B. Ameduri, B. Boutevin, G. Kostov, Macromol. Chem. Phys.,
submitted for publication.
[69] QCPE 504: Departement of Chemistry, Indiana University, Bloo-
mington IN, USA, 1990.
[70] R. Fletcher, M.J.D. Powell, Comput. J. 6 (1963) 163.
[71] W.C. Davidon, Comput. J. 10 (1968) 406.
[72] P. Pulay, Mol. Phys. 17 (1969) 197.
[31] M. Hauptschein, R.E. Oesterling, J. Am. Chem. Soc. 82 (1960) 2868.
[32] R.D. Chambers, J. Hutchinson, R.H. Mobbs, W.K.R. Musgrave,
Tetrahedron 20 (1964) 497.
[73] J.M. Tedder, Angew. Chem. Int. Ed. Engl. 21 (1982) 401.
[74] B. Giese, Angew. Chem. Int. Ed. Engl. 22 (1983) 753.
[75] R. Arnaud, New. J. Chem. 15 (1991) 615.
[33] G.C. Apsey, R.D. Chambers, M.J. Salisbury, G. Moggi, J. Fluorine
Chem. 40 (1988) 261.
 Â
[34] J. Balague, B. Ameduri, B. Boutevin, G. Caporiccio, J. Fluorine
[76] J.M. Poblet, E. Canadell, T. Sordo, Can. J. Chem. 61 (1983) 2068.
[77] E. Canadell, O. Eisenstein, G. Ohanessian, J.M. Poblet, J. Phys.
Chem. 89 (1985) 4856.
Chem. 70 (1995) 215.
[35] R.N. Haszeldine, B.R. Steele, J. Chem. Soc. (1954) 923
[36] J.R. Combes, Z. Guan, J.M. DeSimone, Macromolecules 24 (1994)
865.
à Â
[78] F. Delbecq, D. Ilavsky, Nguyen Trong Anh, J.M. Lefour, L'actualite
chimique, Janv.-Fev. 1986, p. 49.
[37] J.N. Cape, A.C. Greig, J.M. Tedder, J.C. Walton, J. Chem. Soc.
Faraday I (1975) 592
[79] D.V. Avila, K.U. Ingold, J. Lusztyk, W.R. Dolbier Jr., H.-Q. Pan, M.
Muir, J. Am. Chem. Soc. 116 (1994) 99.
[38] A.R. Li, Q.Y. Chen Synthesis (1997) 1481
[39] A.R. Li, Q.Y. Chen J. Fluorine Chem 82 (1997) 151
[80] D.R. Lide: Standard thermodynamic properties of chemical sub-
stances, in: D.R. Lide (Ed.), CRC Handbook of Chemistry and

M. Kharroubi et al. / Journal of Fluorine Chemistry 103 (2000) 145±153
153
È
Physics, 75th Edition (1913±1995), CRC Press, Boca Raton, FL,
1994 p. 5.
[83] K. Fukui, in: P.O. Lowdin, B. Pullman (Eds.), Molecular Orbitals in
Chemistry, Physics and Biology, Academic Press, New York, 1964.
[84] I. Fleming, Frontier Orbitals and Organic Chemical Reactions,
Wiley, London, 1976
[81] I.N. Rozhkov, Yu.A. Borisov, Izv. Akad. Nauk SSSR, Ser. Khim. 8
(1989) 1801.
 Â
[85] M. Kharroubi, A. Manseri, B. Ameduri, B. Boutevin, unpublished
[82] M. Kharroubi, P. De Sainte Claire, B. Ameduri, J. Phys. Chem.,
submitted for publication.
results