Coupling between a fluorinated olefin and a perfluorinated iodide: A model study on the reaction mechanism of perfluorinated polymer cross-linking

Article abstract of DOI:10.1016/S0022-1139(03)00110-6

In a model study, ¹H, ¹⁹C, ¹³ C-¹H and ¹H-¹H correlated NMR techniques confirm a Markovnikov type reaction intermediate for the major coupling products between a short, low MW perfluorinated iodide C₂F₅I (I) and a short, low MW fluorinated olefin CF₃(CF₂)₇CH=CH₂ (II). The reaction is peroxide induced (di-t-butyl peroxide, DTBP) and is conducted at 140 °C for a 3 h reaction time in a sealed glass ampoule. Side reaction products due to the reaction of DTBP with radical reaction intermediates were also observed and identified. The study aimed to mimic as closely as possible the peroxide-initiated coupling reaction between an iodine terminated fluoropolymer (model compound I) and its fluorinated di-olefin coupling agent (model compound II). A mono-olefin was chosen to simplify the model reaction.

Full text of DOI:10.1016/S0022-1139(03)00110-6

Journal of Fluorine Chemistry 123 (2003) 119–126
Coupling between a fluorinated olefin and a perfluorinated iodide:
a model study on the reaction mechanism of perfluorinated
polymer cross-linking
I. Wlassics*, E. Barchiesi, M. Sala
Solvay Solexis, Via S. Pietro, 50-20021 Bollate (MI), Italy
Received 17 July 2002; received in revised form 31 March 2003; accepted 2 April 2003
Abstract
1
1
In a model study, H, 19C, 13C-1H and H-1H correlated NMR techniques confirm a Markovnikov type reaction intermediate for the
major coupling products between a short, low MW perfluorinated iodide C2F5I (I) and a short, low MW fluorinated olefin CF3(CF2)7CH¼CH2
(II). The reaction is peroxide induced (di-t-butyl peroxide, DTBP) and is conducted at 140 8C for a 3 h reaction time in a sealed glass
ampoule. Side reaction products due to the reaction of DTBP with radical reaction intermediates were also observed and identified.
The study aimed to mimic as closely as possible the peroxide-initiated coupling reaction between an iodine terminated fluoropolymer
(model compound I) and its fluorinated di-olefin coupling agent (model compound II). A mono-olefin was chosen to simplify the model
reaction.
# 2003 Elsevier B.V. All rights reserved.
Keywords: Polymer; Peroxide; Coupling; Model study
1. Introduction
reaction, proves to be of vital importance in the elucidation
of a radically induced reaction mechanism.
Peroxide-initiated coupling between an iodine terminated
fluoropolymer and a fluorinated di-olefin gives improved
thermal resistance to the resulting fluoropolymer [1,2] with
respect to the consolidated classical ionic [3] or radical
coupling mechanisms [4,5]. Previous model studies of the
coupling of elastomers performed by Logothetis [5],
Apotheker et al. [6], Krusic and Kochi [7] and Meakin
and Krusic [8], employing brominated substrates and a
peroxide [9,10] have unequivocally demonstrated, with
the aid of ESR, that the mechanism of peroxide curing of
fluoroelastomers is a radical process. Furthermore, it was
demonstrated that the most thermally stable cross-links were
obtained when triallyl isocyanurate (TAIC) was employed as
the coupling agent [6]. A reaction mechanism was proposed
[5,6] showing in part the fate of the radicals produced during
the initiation and propagation steps.
This model study aimed to synthesize, isolate and/or
identify, all of the reaction products generated in a radically
induced coupling reaction. As a result of this model study we
hope to have made progress in both confirming and com-
pleting the hypothesis of Apotheker and others [5–7].
2. Results and discussion
2.1. Model coupling reaction mechanism
A careful analysis of 1H, 19F, correlated 1H-13C NMR and
COSY 1H-1H NMR spectra of the reaction mixture, purified
of its starting materials I and II as well as of di-t-butyl
peroxide (DTBP) and its decomposition by-products acet-
one and t-butanol [11–14], reveals that the major event in the
model study is the Mꢀarkovnikov addition of the perfluori-
nated radical CF3CF2 to the terminal vinyl carbon of II,
C8F19CH¼CH2, thereby generating the more stable second-
ary radical at the C-9 position of III as shown in Fig. 1. The
other mechanism possible would be the anti-Markovnikov
The main intermediates and the reaction side products
were not isolated or identified. The fate of the halogen atoms
on the polymeric chain, in the course of the coupling
* Corresponding author. Tel.: þ39-02-38356367;
fax: þ39-02-38356355.
ꢀ
addition of the C2H5 radical to the C-9 position of com-
pound II generating the less stable primary radical at the
E-mail address: ivan.wlassics@solvay.com (I. Wlassics).
0022-1139/$ – see front matter # 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0022-1139(03)00110-6

120
I. Wlassics et al. / Journal of Fluorine Chemistry 123 (2003) 119–126
The discrimination between the Markovnikov product IV
shown in Fig. 6 and the anti-Markovnikov product VI shown
in Fig. 8 is demonstrated in the 1H-13C heteronuclear single
quantum correlation with gradient selection (g-HSQC)
NMR spectrum shown in Fig. 3. In fact, if compound VI
were present, we would expect a chemical shift of 0 ppm for
the primary iodide ÀCF2ÀCH2ÀI [21b,c] which was absent.
Furthermore, from both the shift and the multiplicity ana-
lysis of the peak at 5.2 ppm observed in Fig. 3 (doublet of
triplets, 1Jð C; HÞ ¼ 152 Hz and 2Jð C; 19FÞ ¼ 26 Hz,
coupling constant spectral data not shown), typical of the
secondary iodide moiety ÀCHðIÞÀCF2À [21b,c], one can
unequivocally assign the NMR spectrum shown in Fig. 3,
and consequently in Figs. 2, 4 and 5, to the Markovnikov
compounds IV and V.
The observation of the absence of the anti-Markovnikov
product VI confirms both the stability of the intermediate
radical III [20] and the hypothesis that the secondary radical
species III is the only major intermediate in the model
coupling reaction.
As was to be expected, multiple coupling events typical of
a radical coupling mechanism were observed. The principal
multiply coupled product observed arises from the coupling
of radical III formed as shown in Fig. 1 to the olefin II.
Fig. 1. Markovnikov reaction pathway leading to the radical intermediate
III.
C-10 position of compound III. This is in accord with the
studies by Tarrant and Tandon [15].
Olefin II is known to show moderate to good reactivity
under radical conditions [16–18] and in coupling reactions,
as shown by Arcella et al. [19].
13
1
13
Fig. 1 shows the expected mode of attack and follows
Markovnikov’s rule due mainly to the formation of the more
stable radical as postulated by Tedder and Walton [20]. The
existence of intermediate III depicted in Fig. 1 is demon-
1
1
strated by the 1H, 19F, H-1H COSY and H-13C correlated
NMR spectra shown respectively in Figs. 2–5. The NMR
spectra unequivocally assign the frequencies typical of the
secondary iodides IV and V shown respectively in Figs. 6
and 7. The secondary iodides IVand V can in fact only arise
from intermediate III. The NMR assignments are in accord
with those reported by Jeanneaux et al., Wiberg et al., Tonelli
et al. [21a,b,c] for similar compounds.
Fig. 2. 1H NMR spectrum of the isolated coupled products IV and V. The callouts correspond to the underlined portions of each structure:
.

I. Wlassics et al. / Journal of Fluorine Chemistry 123 (2003) 119–126
121
Fig. 3. 1H-13C heteronuclear single quantum correlation with gradient selection (g-HSQC) NMR spectrum of the isolated coupled products IV ( ) and V
(*). The callouts correspond to the underlined portions of each structure as in Fig. 2.
This coupling reaction again follows Markovnikov’s rule
[20] and a secondary iodide V with n ¼ 1 is the product
generated as shown in Fig. 7. Due to our experimental
conditions, we only observed V with n ¼ 1. It is likely that,
operating in a large excess of II, and employing a larger
quantity of fluorinated solvent (see Section 4.4), thereby
making the reaction mixture less viscous, one couꢀld perhaps
observe V with n > 1. The generation of CF3CF2 is shown
Fig. 4. 19F NMR spectrum of the isolated coupled products IV ( ) and V (*).

122
I. Wlassics et al. / Journal of Fluorine Chemistry 123 (2003) 119–126
Fig. 5. 1H-1H g-COSY correlated NMR spectrum of the coupled products IV ( ) and V (*). The callouts correspond to the underlined portions of each
structure as in Fig. 2.
propagation b and c. Although propagation sequence c may
occur, the most probable route to IV is propagation sequence
b, since the concentration of I is much higher than that of
CH3I (which arises from the decomposition of DTBP). The
same reasoning can be applied to the scheme shown in
Fig. 6. Isolated model coupled Markovnikov product IV.
further on in the propagation step of Fig. 9. The reactions
leading to products IV and V shown respectively in Figs. 9
and 10, can be viewed as a simplified version of the ‘‘pseudo-
living polymerization’’ ([22–24] and references therein).
The formation of the observed major product IV most
likely follows the generally accepted initiation–propagation
steps [25,26] shown in Fig. 9. Fig. 9 shows all of the steps
that occur which can explain the observed products but
does not take into consideration all of the theoretically
possible reactions. During the formation of IV, two propa-
gation sequences are proposed involving intermediate III,
Fig. 10; although propagation sequences c, d and e may all
occur, the most probable route to V is propagation sequence
c since the concentration of C2H5I is hꢀigher than the con-
centration of the methyl radical CH3 and the coupled
product IV is bulkier and therefore less mobile and is always
at a lower concentration than I.
It was observed that the selectivity of formation of product
IV with respect to product V was 60:40, respectively (see
Section 4.4). This, according to Apostolo et al. [22] and
Arcella and Apostolo [24], was expected; as the molecular
weight of the growing radical increases, its mobility
Fig. 7. Isolated model multiply coupled Markovnikov product V.
Fig. 9. Reaction pathway leading to the Markovnikov radical intermediate
Fig. 8. Anti-Markovnikov product VI.
III and the model coupled product IV.

I. Wlassics et al. / Journal of Fluorine Chemistry 123 (2003) 119–126
123
Fig. 10. Reaction pathway leading to the Markovnikov radical intermediate IIIa and the model multiply coupled product V.
decreases. The result is an almost exponential decrease in
yield of the multiply coupled products. This is the principle
of molecular weight polydispersity observed in polymer
chemistry. Our simple model therefore, fits well with the
reported literature [22–24].
The products shown in bold and in italics were the ones
actually observed and identified.
2.2. Coupling reaction side products
Careful GC–MS analysis of the worked-up reaction mix-
ture, shown in Fig. 11, confirmed the existence respectively
of the coupled and multiply coupled products IV and V.
The total isolated yield of the coupled products IV and V
was calculated to be 42 Æ 2% (mol%) based on the isolated
As can be seen from Fig. 12, along with the two major
coupling products IV (single coupling) and V (multiple
coupling) isolated and identified, three side products were
identified (but not isolated): (1) CF3CF2CH2COCH3, (2)
CF3CF2OC(CH3)3, (3) CH3I. All three side products were
1
1
weight and on the H NMR spectrum.
identified by H NMR spectroscopy in the crude reaction
The peaks appearing in the NMR spectrum shown in Fig. 2
not pertaining to compound IV and V are attributed to (a)
residual starting compound II (–CH¼CH2 resonance at 6.05
and 6.2 ppm); (b) residual t-butanol (–(CH3)3 at 1.1 ppm);
(c) residual ether 2 (–(CH3)3 at 1.3 ppm).
mixture after complete removal of any residual acetone and
t-butanol by repeated extractions of the crude mixture with
distilled H2O.
The total yield of the identified side products was calcu-
lated to be approximately 4 Æ 0:2% (mol%) with respect to
1
Fig. 12 schematically shows the complete coupling reac-
tion pathway along with all of the possible reaction products.
the limiting reagent (II) based on H NMR spectroscopic
analysis of the crude reaction mixture.
Fig. 11. GC–MS spectra of coupled products IV ( , C12F22H3I) and V (*, C22F39H6I).

124
I. Wlassics et al. / Journal of Fluorine Chemistry 123 (2003) 119–126
Fig. 12. Model of radical coupling showing the reaction mechanism, the coupled products obtained (bold) and the major side products observed (italics).
Aside from methyl iodide, whose presence is of no surprise
if DTBP is employed as a radical initiator in the presence of
an iodide [14,27,28], compounds 1 and 2 show that DTBP is a
competitor, along with the olefin II, for the iodide I. In a
separate experiment, it was demonstrated that as the [DTBP]
in the initial reaction mixture was increased, the yield of
side products 1 and 2 increased accordingly. Side product 1
shown above, arises from the coupling of the radical
ꢀCH2C(¼O)CH3 to the iodide I. It is reported [29] that
mechanism yielding secondary iodides. The secondary
iodides observed arise from both single and multiple cou-
pling reactions. Radical coupling evolves side products due
to the radical initiator employed, which adds mainly to the
perfluoroiodide.
Applying this model study to a real polymeric system
implies that one should expect, along with the cross-linked
products, polymers with ether or ketone end-groups.
ꢀ
proton abstraction by either CH3 or (CH3)3COꢀ [27,30] is
known to occur; indeed the authors observed the presence of
t-butanol in the crude reaction mixture (spectrum not shown).
It is therefore possible that the alkoxy radical (CH3)3COꢀ,
arising from the homꢀolytic cleavage of DTBP [11–13], or the
methyl radical CH3 (although the presence of CH4 was
not detected) abstracts a proton from acetone as is known
to occur [31,32]; the radical species thus generated either
4. Experimental details
4.1. Reagents
Both C2F5I (I) and C8F17I (Ia) were purchased from
Aldrich and were used without further purification. Di-t-
butyl peroxide was purchased from Carlo Erba and was used
without further purification.
ꢀ
couples with the perfluoro radical CF3CF2 or adds to the
corresponding iodide I. The species ICH2C(¼O)CH3 was not
detected via NMR spectroscopy. Side product 2 is the out-
come of the coupling of the alkoxy radical (CH3)3COꢀ to the
4.2. Analysis
1H, 19F, 13C, and correlated NMR spectra were recorded
on a Varian spectrometer operating at 400 MHz.
Gas chromatograms were recorded on a GC 8000Top
Carlo Erba gas chromatographer employing a capillary
0.54 mm, 25 m long silicone-packed column.
ꢀ
iodide I or to its corresponding radical species CF3CF2 .
Studies by Marchionni and Srinivasan [33] and by
Paciorek and Kratzer [34] have demonstrated that ketone
and ether terminals of the type shown in species 1 and 2
above are normally observed in peroxide-initiated polymer-
izations. Furthermore, both ketone and ether termini are
stable at 140 8C under radical conditions and therefore can
be detected and isolated in the crude reaction mixture at the
end of the polymerization reaction. This is in further support
of the side reaction products 1 and 2 observed.
Mass spectra were recorded on a Finnigan MAT SSQ700
chromatographer employing a methyl silicone CPSIL
(Varian-Chrompack) column (50 m, i:d: ¼ 0:32 mm; thick-
ness ¼ 5 m).
Boiling points of products IV and V were estimated
employing a micro-distillation apparatus.
A substantial presence of homocoupling of CH3ꢀ radicals
leading to C2H6 was not observed in contrast to Cope et al.
[25]. A probable explanation is that the concentration of
4.3. Synthesis 8-vinyl perfluoro-octane (II)
ꢀ
CH3 at any point in time during the course of the coupling
reaction was very low with respect to the iodide I and
therefore was quickly quenched.
The synthesis of (II) is already reported in the literature
[35–37]. We report here a slightly modified version.
Octyl perfluoro iodide, C8F17I (50 g, 91.6 mmol), was
placed in a 45 ml stainless steel autoclave equipped with a
magnetic stirrer and a digital pressure gauge. The autoclave
was mechanically sealed, evacuated and washed with N2.
The evacuated autoclave was placed in an oil bath and heated
in order to reach an internal temperature of 160 8C.
3. Conclusions
Peroxide-initiated radical coupling reactions of a per-
fluorinated alkyl iodide involve exclusively a Markovnikov

I. Wlassics et al. / Journal of Fluorine Chemistry 123 (2003) 119–126
125
Fifty atmospheres (35 mmol) of C2H4 were then loaded in
the autoclave. As the reaction proceeded, ethylene was
continuously added until no more ethylene was consumed.
The total reaction time was 8 h. The autoclave was then
cooled and the crude reaction product was analyzed by 19F
the sum of the three side products 1, 2 and 3 were, respec-
tively, 42 Æ 2 and 4 Æ 0:2% (mol%). The selectivity of
coupled products to side products was therefore approxi-
mately 9:1.
The crude reaction mixture was redistilled to eliminate
the low boiling side products 1 (bp ꢁ 70 8C), 2 (bp ꢁ 80 8C)
and 3 (bp ¼ 41 8C), the solvent (Galden1 D-100) and the
residual starting compound II (bp ¼ 135 8C). Starting com-
pound I was eliminated during work-up since it has a
bp ¼ 13 8C. A viscous orange residue was obtained which
was composed of 90% (mol%) of compounds IV and V and
10% (mol%) of residual starting material II.
1
and H NMR, FT-IR and found to be CF3(CF2)7CH2CH2I
(Ib) 51.6 g. Conversion ¼ 99%, yield ¼ 98:2%. The
crude product was used without further purification (1%
starting iodide and 0.5% telomeric product identified as
CF3(CF2)7(CH2CH2)2I were also obtained). Ib (50 g,
87.1 mmol) was dissolved in 50 ml of anhydrous ethanol
and stirred at 65 8C in a three necked round bottom flask.
When the reaction mixture reached 65 8C, 1.2 eq. of etha-
nolic KOH were slowly added by means of a graduated
dropping funnel. The reaction was moderately exothermic
and care was needed not to exceed 75 8C. The crude reaction
product was filtered to separate KI and the ethanol was
distilled under reduced pressure; 35 g, 78.5 mmol, of the
desired product II were obtained. Conversion ¼ 99%, yield
The yield of the singly coupled product IV, was 25% and
the yield of the multiply coupled product V was 16.7%, as
1
determined by H (Fig. 2) and 19F (Fig. 4) NMR spectro-
scopy. The selectivity of the singly coupled product IV and
of the multiply coupled product V is therefore, respectively,
60 and 40%.
2
a
b
4
5
6
NMR (IV): 1CF3 CF2 CH2 CHI3CF2 CF2 ðCF2Þ4 CF2
after distillation ¼ 88:5%. The product was identified by 19
F
7CF3.
1
and H NMR spectroscopy and by FT-IR.
19F: 1 (À86.2 ppm); 2 (À118 ppm, AB system); 3
(À99 and À110 ppm); 4 (À116.5 ppm AB system); 5
(À122 ppm); 6 (À126 ppm); 7 (À81.1 ppm).
1H: a (3.3 ppm); b (5 ppm).
19F NMR: À84 ppm (s, CF3–), À117 ppm (s, 8–CF2–),
À124,8 ppm (s, 7–CF2–), À125.2 ppm (s, 3,4–CF2–),
À126.1 ppm (s, 6–CF2–), À127 ppm (s, 3–CF2–),
À130 ppm (s, 2–CF2–).
13C: a (35 ppm); b (5.2 ppm); doublet of triplets,
1H NMR (CDCl3): 5.95 ppm (m, ÀCH¼CH2), 5.9 ppm
(m, ÀCH¼CH2).
1Jð C; 1HÞ ¼ 152 Hz and Jð C; 19FÞ ¼ 26 Hz.
13
2
13
bp (IV): 170–180 8C.
FT-IR (n, cmÀ1): 1651 (–CH¼CH2 st.), 1080–1200
(CF3– and –CF2– st.).
NMR (V): 1CF3 CF2 CH2 CH½eCH2 CHI3CF2 CF2
2
c
d
f
4
5
6
7
8
4
5
6
7
ðCF2Þ4 CF2 CF3ꢂ CF2 CF2 ðCF2Þ4 CF2 CF3.
GC (70 8C for 1 min, slope to 210 8C at 15 8C/min,
210 8C for 4 min): tret ¼ 3:92 min.
19F: 1 (À85.8 ppm); 2 (À118 ppm AB system); 3
(À97 and À112 ppm); 4 (À116.5 ppm AB system); 5
(À122 ppm); 6 (À126 ppm); 7 (À81.1 ppm).
1H: c (2.7–2.8 ppm); d (3.4 ppm); e (2.4 ppm); f (4.89
ppm).
4.4. Model coupling reaction
I (2.03 g, 5.87 mmol), II (1.745 g, 3.9 mmol), DTBP
(0.0427 g, 0.195 mmol, 0.05 eq.) and 2 ml of Galden1
D-100 (perfluorobutyltetrahydrofuran 80% and perfluoropro-
pyltetrahydropyran 20% with bp ¼ 100 8C), were placed in a
10 ml two necked, round bottom glass ampoule equipped
with a Teflon1 ‘‘sure seal’’ stopcock and a magnetic stir bar.
The ampoule was placed in a dry ice bath at À80 8C and the
air inside removed by flushing with N2 for 20 min. The
ampoule was then sealed and placed in an oil bath at
140 8C and magnetically stirred for 3 h equivalent to four
half lives of the DTBP. Due to the experimental conditions
chosen, all of the reagents were in the liquid phase during the
entire course of the reaction. The crude reaction product was
then washed in H2O. The organic layer was collected,
dissolved in CH2Cl2 and dried over MgSO4. The solvents
were carefully distilled and the crude mixture analyzed by
1H, 19F NMR and 13C NMR. Conversion (based on the
limiting reagent II) ¼ 46% (mol%).
13C: c (29 ppm); d (37.3 ppm); e (32.8 ppm); f (22.5
ppm).
bp (V): >240 8C (dec.).
1H NMR (side product 1—not isolated): –CH3 (2.1 ppm,
s; 3H); –CH2C(¼O)– (3.3 ppm; broad t; 2H).
19F NMR (side product 1—not isolated): –CF3 (À86 ppm;
shielded by IV in Fig. 2); –CF2– (À118 ppm; shielded by IV
in Fig. 2).
1H NMR (side product 2—not isolated): –CH3 (1.3 ppm;
s; 9H).
19F NMR (side product 2—not isolated): –CF3 (À86 ppm;
shielded by IV in Fig. 2); –CF2O– (À80 ppm).
1H NMR (side product 3—not isolated): –CH3 (2.2; s;
3H).
Acknowledgements
Two major coupling products IV and V and three minor
side reaction products 1–3 were observed by NMR spectro-
scopy in the crude, washed reaction mixture. The calculated
yields of the sum of the main coupled products IVand Vand
We wish to thank Mrs. Elisa Restelli and Dr. Cristina
Grossi for the NMR spectra, Dr. Patrizia Dardani for the
GC–MS spectra and Prof. Luciana Giovannetti for a critical
review of this work.

126
I. Wlassics et al. / Journal of Fluorine Chemistry 123 (2003) 119–126
[21] (a) F. Jeanneaux, M. Le Blanc, A. Cambon, J. Cambon, J. Fluorine
References
Chem. 4 (1974) 261;
(b) K.B. Wiberg, W.E. Pratt, W.F. Baily, J. Org. Chem. 45 (1980)
4936;
[1] I. Wlassics, V. Tortelli, W. Navarrini, M. Albano, US Patent
5,902,857 (1999); [CA 127 (1994) 6039].
(c) C. Tonelli, F.C. Schilling, R.E. Cais, Macromolecules 14 (1981)
560.
[2] M. Albano, C. Pizzi, I. Wlassics, US Patent 5,948,868 (1999); [CA
126 (1999) 344397].
[22] M. Apostolo, V. Arcella, G. Storti, M. Morbidelli, Macromolecules
35 (2002) 6154.
[3] D. Reyx, I. Campistron, C. Cailloud, Macromol. Chem. Phys. 196
(1995) 775.
[23] V. Arcella, M. Apostolo, Rubber World (2001) 27.
[24] V. Arcella, M. Apostolo, Fluoropolymers 2000, Presentation 26,
Savannah, GA, USA, 2000.
[4] W.W. Schmiegel, Die Angew. Macromol. Chem. 76/77 (1979) 39.
[5] A. Logothetis, Prog. Pol. Sci. 14 (1989) 251.
[6] A. Apotheker, J. Finley, P. Krusic, A. Logothetis, Rubber Chem.
Technol. 55 (1982) 1004.
[25] J. Cope, A.C. Grieg, J.M. Tedder, J. Chem. Faraday Trans. 1 71
(1975) 592.
[7] P. Krusic, J. Kochi, J. Am. Chem. Soc. 93 (1971) 846.
[8] P. Meakin, P. Krusic, J. Am. Chem. Soc. 95 (1973) 8185.
[9] B. Smart, P. Krusic, P. Meakin, R. Bingham, J. Am. Chem. Soc. 96
(1975) 7385.
[26] C. Walling, Free Radicals in Solution, Wiley, New York, 1957
(Chapter 3).
[27] J.P. Sloan, J.M. Tedder, J.C. Walton, J. Chem. Soc., Perkin Trans II
(1975) 1846.
[10] P. Krusic, R. Bingham, J. Am. Chem. Soc. 98 (1976) 230.
[11] G.R. McMillan, M.M. Wijnen, Can. J. Chem. 36 (1958) 1227.
[12] J.K. Allen, J.C. Bevington, Proc. R. Soc. A262 (1961) 271.
[13] F.F. Rust, F.H. Seubold, W.E. Vaughan, J. Am. Chem. Soc. 72 (1950)
228.
[28] J.M. Tedder, J.C. Walton, L.L.T. Vertommen, J. Chem. Soc., Perkin
Trans II (1978) 1040.
[29] M. Newcomb, R.M. Sanchez, J. Kaplan, J. Am. Chem. Soc. 109
(1987) 1195.
[30] H.Q. Xie, W.E. Baker, in: T.C. Chung (Ed.), New Advances in
Polyolefins, Plenum Press, New York, 1993, pp. 109–110.
[31] A.I.I. Taha, R.R. Kuntz, J. Phys. Chem. 73 (1969) 4406.
[32] H. Arica, N.L. Arthur, Aust. J. Chem. 36 (1983) 2185.
[33] G. Marchionni, P. Srinivasan, Fluoropolymers 1992, Paper 14,
UMIST—Manchester, GB, 1992.
[14] D.S. Ashton, D.S. Shand, J.M. Tedder, J. Am. Chem. Soc., Perkin
Trans II (1975) 320.
[15] P. Tarrant, J.P. Tandon, J. Org. Chem. 34 (1969) 864.
[16] F. Guida-Pietrasanta, A. Ratsimihety, G. Caporiccio, J. Fluorine
Chem. 68 (1994) 71.
[17] R.N Hazeldine, M.J. Newlands, B.R Iles, Br. Patent 1075101 (1967);
[CA 67 (1967) 100240].
[34] K.J.L. Paciorek, R.H. Kratzer, J. Fluorine Chem. 67 (1994) 169.
[35] B. Ameduri, B. Boutevin, M. Nouiri, M. Talbi, J. Fluorine Chem. 74
(1995) 191.
[18] Y.K. Kim, A.G. Smith, O.R. Pierce, J. Org. Chem. 38 (1973)
1615.
¨
¨
[36] Z. Szlavik, G. Tarkanyi, A. Gomory, J. Rabai, Org. Lett. 2 (2000) 2347.
[37] S. Kohjiya, M. Kozuyuki, S. Yamashita, Y. Shibata, J. Mater. Sci. 25
(1990) 3368.
[19] V. Arcella, G. Brinati, M. Albano, V. Tortelli, US Patent 5585449
(1996); [CA 123 (1996) 259509].
[20] J.M. Tedder, J.C. Walton, Acc. Chem. Res. 9 (1976) 183.