New perfluorovinylethers through the bis(fluoroxy)difluoromethane (BDM) chemistry

Article abstract of DOI:10.1016/j.jfluchem.2003.07.011

In the present work we give an overview of the CF₂(OF) ₂ radical reactivity and report the synthesis of new perfluorovinylethers. CF₂=CF-OCF₂OCF₂ CF₃ and CF₂=CF-OCF₂OCF₂ CF₂OCF₃ are prepared in a semi-continuous methodology starting from CF₂(OF)₂. These highly reactive vinylethers are characterized by the OCF₂O group directly bonded to the insaturation. For this reason they are excellent candidates for the preparation of very low T_g perfluororubbers.

Full text of DOI:10.1016/j.jfluchem.2003.07.011

Journal of Fluorine Chemistry 125 (2004) 189–197
New perfluorovinylethers through the bis(fluoroxy)difluoromethane
(BDM) chemistry
Walter Navarrinia,*, Sandra Cortib
aResearch and Development Center, Solvay Solexis, Viale Lombardia 20, 20021 Bollate Milan, Italy
bCiba Vision, 333 East Howard Avenue, Des Plaines, IL 60056, USA
Received 29 January 2003; received in revised form 14 July 2003; accepted 28 July 2003
Abstract
In the present work we give an overview of the CF2(OF)2 radical reactivity and report the synthesis of new perfluorovinylethers.
CF2¼CF–OCF2OCF2CF3 and CF2¼CF–OCF2OCF2CF2OCF3 are prepared in a semi-continuous methodology starting from CF2(OF)2. These
highly reactive vinylethers are characterized by the OCF2O group directly bonded to the insaturation. For this reason they are excellent
candidates for the preparation of very low Tg perfluororubbers.
# 2003 Elsevier B.V. All rights reserved.
Keywords: Synthesis; Hypofluorites; Bis(fluoroxy)difluoromethane; Perfluorovinylethers; Perfluororubbers
1. Introduction
fluoroolefins has been applied to the synthesis of perfluoro-
vinylethers [12] as well as to the peculiar hypofluorite-
initiated oxygen copolymerization with perfluoroolefins [13].
One of the most versatile hypofluorite is the bis(fluoroxy)-
difluoromethane (BDM).
We have recently utilized this hypofluorite as intermediate
for the preparation of sophisticated linear or cyclic viny-
lethers. These monomers are useful for the synthesis of high
[6] or low Tg [2] amorphous fluoropolymers. In addition, a
high content of vinylethers in the amorphous perfluoropo-
lymer gives good solubility in perfluorinated solvents [14].
During the last decade the hypofluorite chemistry has
attracted renewed interest from the industrial and academic
world. The recently discovered reactivity of this class of
compounds has in fact led to the identification of new
interesting perfluoromonomers and polymers [1,2].
One of the most attractive issues of the hypofluorite
chemistry is its great flexibility allowing the synthesis of
diversified perfluorovinylether monomers [1,3–5]. These
compounds represent the key building blocks for the pre-
paration of different and innovative polymers, like amor-
phous fluoropolymers with Tg ranging from þ300 8C [6,7] to
À70 8C [2,15], as well as sulphonic and carboxylic per-
fluoropolymers for fuel cell applications [8].
2. Results and discussion
The addition of BDM to the double bond, yields either the
related dioxolane and the fluorinated olefin product, as
described in Reaction 1, or the linear addition product, as
shown in Reaction 2, depending on the reaction conditions
adopted [5,7].
Due to the unique properties of the O–F bond in the
organic hypofluorite, this can act as a source of electrophilic
fluorine [9], electrophilic alkoxylium species [10] as well as
alkoxy radicals [11].
In our laboratories the chemistry of these compounds has
been extensively studied. Their free-radical reactivity with
These two reactions evolve through similar initiation,
propagation and termination steps as described in Scheme 1.
* Corresponding author. Tel.: þ39-02-38351; fax: þ39-02-38352129.
E-mail address: walter.navarrini@solvay.com (W. Navarrini).
0022-1139/$ – see front matter # 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2003.07.011

190
W. Navarrini, S. Corti / Journal of Fluorine Chemistry 125 (2004) 189–197
At the Propagation step 4 of Scheme 1, we have observed
two competing propagation reactions. One is a monomole-
cular internal homolytic substitution SHi giving right of the
dioxolane product, the second one is a bimolecular homo-
lytic substitution SH giving right of the linear addition
product as shown in the reaction steps 4 and 4b below:
The two pathways involve the same free-radical inter-
mediate. Competition between cyclization to dioxolane and
linear addition to bis-ethers has been observed. Considering
the two reactions 4 and 4b, it should be noted that ratio (lin/
diox) is proportional to the BDM concentration in the
reaction media, whereas it is independent from the concen-
tration of the radical intermediate FOCF2OCÀCꢀ ¼ ½int:Š as
described in Eq. (1) where K ¼ Klin=Kdiox
:
Initiation :
lin
Klin½CF2ðOFÞ2Š½int:Š
F
¼
¼ K½CF2ðOFÞ2Š
Eq. (1)
.
O
+
FO CF2
FO CF2 OF +
C C
.
C
C
1)
diox
Kdiox½int:Š
In fact by increasing the BDM concentration in the
reaction medium we observed an increment of the linear
addition as compared to the cyclic addition depending also
on the olefin, solvent and temperature chosen [5].
Under these experimental conditions, the olefin is added
to the solution of BDM. This procedure is exactly the
opposite of the methodology suggested in the literature in
order to obtain high yield of hypofluorite addition products,
where the addition order is reversed [11,12]. Hypofluorites
are indeed susceptible to thermo or chemical self-decom-
position that can cause very low addition yield or, in some
cases, extended decomposition.
In particular, BDM has been found to be susceptible to
ignition in gas or liquid phase, and this extended decom-
position may be due to traces of organic or inorganic
reactants. In order to safely handle BDM in gas phase a
1/3-dilution ratio with an inert gas like nitrogen or helium is
needed [1]. In the liquid medium its behavior is more
complex, since sudden exothermic reaction may locally
and easily change both liquid and gas BDM concentration
as well as the reaction course. This phenomenon has its
experimental evidence as micro-explosion within the liquid
phase. To avoid these hot spots, low temperature, low
volume and high thermal exchange surface are needed.
Propagation :
F
F
F
.
.
FO CF2 O
2)
3)
C
C
+
FO CF2 OF
+
C
C
F F
C
F
O
O
.
FO CF2
O
F
C C
+
.
C
C
F F
C
F F
C
O
O
O
O
.
F
+
4)
5)
.
SHi
C
C
C
C
F
.
F
C C
+
.
C
C
Termination :
F
F
F
.
.
C
6)
F
+
C
C
C
F
F
F
7)
.
C
C
2
C C C C
Scheme 1.

W. Navarrini, S. Corti / Journal of Fluorine Chemistry 125 (2004) 189–197
191
(a)
(a)
KlinOF
-110˚C
+
CF3CF2OCF2OF
CF2
CF2
CF3CF2OCF2OF
CF2
CF2
CF2
CF2
1) FOCF2OF
1) FOCF2OF
+
Klin
-80˚C
2) CF CF OCF OF
2) CF CF OCF OF CFCI
+
+
CF3CF2OCF2OCF2CF3
CF2 CF2
CF3CF3
CFCl
CF3CF2OCF2OCFCICF2CI
CF3CF2OCF2OCF CF2
3
2
2
3
2
2
Zn/DMF
CF3CF2OCF2OCFCICF2CI
3)
3)
CF
CF2
2
FOCF2OF
+
+
2
O
O
CF2
Scheme 2.
Scheme 4.
D.D. DesMarteau and G.H. Cady suggested the use of
very cold glass beads to be introduced in a millimoles scale
batch reactor or the use of fluorinated solvents. Another
approach is the use of a continuous system: the reactants are
overcooled to À114 8C and spread in a film form on a cold
vertical cylindrical reactor wall. In this way they can react at
a temperature close to À110 8C, while reaching the outlet of
the reactor, as described in Section 3.1. Following this latter
procedure, the addition yields were good and the decom-
position reactions were minimized even in the absence of
solvent.
The BDM/TFE and BDM/CF3OCF¼CF2 reactions are of
particular interest since the linear intermediate hypofluorites
(a) and (b) are utilized for the preparation of new perfluor-
ovinylethers through their addition to CFCl¼CFCl, as
described in Schemes 2 and 3.
directly obtained through the gas-chromatographic (GC)
areas of the uncorrected GC peaks. In the BDM/TFE range
ratio observed the dioxolane product evaluated by on-line
GC and by end products distillation was below 1% and
therefore considered constant in all the trials.
The experimental difficulty for the determination of the
correction factors for CF3CF2OCF2OF was due to the pre-
paration and handling of a pure sample. This intermediate
was simply characterized through the 19F NMR analysis and
through the relative addition product structure determina-
tion.
As can be see in Fig. 1 the yield on the converted BDM
(selectivity) increases with the increasing ratio BDM/TFE.
This set of experimental data can also be utilized to qualita-
tively determine the ratio between the observed kinetic con-
stants KlinOF and Klin of the consecutive reactions (1) and (2)
of Scheme 4.
The reactions (1) and (2) follow a scheme of consecutive
reactions with a order higher than one and can be described
by the differential Eq. (2) [16]:
In the BDM/TFE reaction, to reduce the competitive
dioxolane formation (Reaction 3 in Scheme 4), the ratio
BDM/TFE should be at least 0.6. At higher ratio also the
consecutive Reaction 2 can be minimized. In this way the
yield to linear addition on converted BDM also increases, as
can been seen by Fig. 1, where linOF ¼ CF3CF2OCF2OF
and lin ¼ CF3CF2OCF2OCF2CF3.
@½linOFŠ
@½BDMŠ
Applying this equation to a continuous reactor through the
Klin ½linOFŠ
KlinOF ½BDMŠ
¼ À1 þ
Eq. (2)
The experimental stationary points were obtained utiliz-
ing the apparatus and methodology described in Section 3.1.
By changing the ratio between BDM/TFE of the reacting
gases a set of stationary points was analyzed. The data were
material balance of the inlet and outlet at the stationary
(b)
-110˚C
+
1)
2)
CF3OCF2CF2OCF2OF
CF3OCFOCF2OF
CF3
FOCF2OF
CF2 CFOCF3
+
CFCI CFCI
+
+
CF3OCF2CF2OCF2OCFCICF2CI
CF3OCFOCF2OCFCICF2CI
CF3
CF3OCFOCF2OF
CF3
CF3OCF2CF2OCF2OF
Zn/DMF
+
CF3OCF2CF2OCF2OCF=CF2
CF3OCFOCF2OCF=CF2
CF3
+
3) CF3OCF2CF2OCF2OCFCICF2CI CF3OCFOCF2OCFCICF2CI
CF3
Scheme 3.

192
W. Navarrini, S. Corti / Journal of Fluorine Chemistry 125 (2004) 189–197
Fig. 1. Yield of CF3CF2OCF2OF (linOF) and CF3CF2OCF2OCF2CF3 (lin) on converted BDM.
points, we can substitute the differential with the relative
discrete concentration difference between the inlet and the
outlet at the stationary points. The concentrations can be
substituted with the millimoles of the related products formed
or reacted in the unit of time, since they are referred to the
same reactor volume, thus obtaining the following equation:
All the perfluorovinylethers obtained are characterized by
the unit OCF2O directly bonded to the insaturation of the
vinyl monomer. These methylene-oxy-vinylether monomers
(MOVE) homopolymerize and copolymerize with TFE or
VDF. The polymerization reactions are carried out in solu-
tion and are initiated by the perfluoropropionyl peroxide that
decomposes to carbon dioxide and perfluoroethyl radical
that initiates the polymerization. The end-groups of the
polymer thus obtained are completely perfluorinated and
do not introduce polarity in the polymer, that may cause
undesired aggregation and Tg misevaluations.
In this experimental conditions the only side reaction that
can introduce polar groups, such as carbonyl fluoride, at the
end of the polymer chains is a rearrangement of the growing
carbon radical bearing an oxyalky group in a position. Low
reaction temperature and a high reactivity of the comono-
mers can minimize carboxylic end-group formation as
described by Ameduri et al. [15].
ꢂ
ꢀ
ꢁꢃ
Klin
ðBDMÞ
DðlinOFÞ
DðBDMÞ
¼
1 þ
Eq. (3)
KlinOF ðlinOFÞ
Although the experimental values reported in Table 1 are
obtained by applying Eq. (3) and through a semi-quantitative
analysis, the average value of 0.41 for Klin/KlinOF indicates
that, in the experimental condition studied, BDM is around
2.4 times more reactive than linOF. Therefore, a single OF in
BDM is 1.2 times more reactive than OF in the LinOF, thus
appearing very similar.
The addition products to CFCl¼CFCl, CF3CF2OCF2-
OCFClCF2Cl and CF3OCF2CF2OCF2OCFClCF2Cl, can be
dehalogenated in good yields to the related perfluoroviny-
lether monomers, as shown in Sections 3.3 and 3.4.
The polymerization reactions are summarized in Table 2.
In the homopolymers Tg’s evaluation, where the only
reacting monomer is the vinylether, we experimentally
Table 1
Table 2
BDM/TFE ratio values and the related Klin/KlinOF ratios in the range BDM/
TFE 0.7–1.6
Polymerization reactions
TFE (%)
FVE (%)
Tg (8C)
BDM/TFE
Klin/KlinOF
Section 3.5
Section 3.6
Section 3.7
Section 3.8
Section 3.9
Section 3.10
–
–
100 (MOVE 1)
100 (MOVE 2)
24 (MOVE 1)
37 (MOVE 2)
23 (PVE)
À35.4
À52.6
À21.4
À44.5
þ15
0.70
0.78
0.87
0.97
1.08
1.22
1.59
0.34
0.38
0.40
0.38
0.47
0.44
0.45
76
63
77
77
23 (b-PDE)
À4.8
The values represent the molar monomer percentage contents incorporated
in the polymers.

W. Navarrini, S. Corti / Journal of Fluorine Chemistry 125 (2004) 189–197
193
observed slightly higher Tg’s than one would expect through
a linearization plotting of the Tg with the weight percent of
the vinylether monomer incorporated in the polymer. This
is probably due to the higher content of carbonyl fluoride
radicals located at the end of the homopolymer chains
compared to the copolymer with TFE. It is worth observing
that the Tg of the TFE copolymer with the MOVE 1
(CF2¼CFOCF2OCF2CF3) monomer in Section 3.7 is lower
as compared to the Tg of the copolymer obtained in the
Section 3.9 with CF2¼CFOCF2CF2CF3 (PVE), even at
equal modifying vinylether molar quantity incorporated in
the polymer. This is due to the higher O/C ratio in the MOVE
1 monomer than in the PVE monomer.
Surprisingly, the Tg of the copolymer of the TFE with
CF2¼CFOCF2CF2OCF3 (b-PDE) is higher than the one of
the copolymer of the TFE with MOVE 1 (CF2¼CFOCF2-
OCF2CF3), since both copolymers have the same molar
monomer content and the same O/C ratio. This can be
probably explained by the position of the oxygen atom in
the vinylether monomer side chain.
Indeed in the MOVE 1 monomer the unit OCF2O is
directly bonded to the insaturation, on the contrary in the
b-PDE the unit OCF2O is not present. In order to have a
better understanding of the low Tg variation in amorphous
TFE copolymers in relation to the oxygen atom position in
the vinylethers side chain, it is useful to compare the Tg’s of
the TFE/MOVE 2 (CF2¼CFOCF2OCF2CF2OCF3) copoly-
mers with the copolymers TFE/CF2¼CFOCF2CF2OCF2-
OCF3 at the same comonomer concentration, since also
in this case both monomers have the same O/C ratio and
only the MOVE 2 monomer contains the unit OF2O directly
bonded to the insaturation.
on contact with organic reagents safety precautions must be
taken in handling BDM [7].
3.1. General procedure, synthesis of
CF3CF2OCF2OCFClCF2Cl perfluoro-1,2-dichloro-3,
5-dioxaheptane
A 300 ml cylindrical flow reactor with i.d. of 3 cm, was
equipped with a magnetic dragging mechanical stirrer, a
turbine with recycle of the reacting gas placed at 20 cm from
the reactor top, an internal thermocouple, two internal
copper pipes (0.25 in o.d.) for the reactants feeding, ending
at about at 1 mm above the turbine, and a products outlet
placed at the bottom of the reactor. The apparatus was kept
vertically during the reaction.
In this reactor, inside of which the temperature was
maintained at À114 8C, 1.1 l/h of CF2(OF)2 diluted with
3.3 l/h of He were fed through one of the two inlet pipes; a
flow of 1.1 l/h of CF2¼CF2 diluted with 0.7 l/h of He was
introduced through the second inlet pipe. The feeding was
continued for 6.6 h.
After the reactants contacted each other on the fast
moving turbine (2000 rpm), an immediate and exothermic
reaction took place, and the liquid reactants quickly moved
on the perimetral wall and to the bottom of the reactor.
The reaction products were extracted from the bottom of
reactor through the product outlet (o.d. 0.25 in), by static
dropping. The reactions products were then evaporated
and continuously brought to room temperature, while the
gaseous mixture flow (unreacted starting material, reac-
tion product and the diluting helium) was monitored by
gas-chromatography.
Some data regarding TFE/CF2¼CFOCF2CF2(OCF2)n-
OCF3 (n ¼ 1–5) copolymers recently appeared in the patent
literature [17] further indicating that the monomers having
the unit OCF2O directly bonded to the insaturation have a
higher capability of decreasing the Tg’s of the TFE copo-
lymers.
The reaction mixture flow, consisting of unreacted BDM,
CF3CF2OCF2OF, CF3CF2OCF2OCF2CF3 and the diluting
He (TFE and the 1,3-perfluorodioxolane were completely
absent) was continuously fed, under mechanical stirring,
into a 250 ml reactor. This second reactor, maintained at the
temperature of À70 8C, was equipped with a thermocouple,
a dipping inlet for the incoming reaction mixture from the
first reactor, an outlet with head of inert gas, and contained
72.6 g of dichlorodifluoroethylene.
3. Experimental
At the end of the reagents addition (BDM, CF3CF2OC-
F2OF) into the second reactor, the reaction raw material was
distilled by a 90 plates column at atmospheric pressure,
collecting 41.5 g of the desired product (boiling point of
91 8C).
The yield of perfluoro-1,2-dichloro-3,5-dioxaheptene, cal-
culated with the respect of the moles of CF2(OF)2 utilized,
is 36%.
The remaining reaction products were mainly: CF3CF2-
OCF2OCF2CF3, CF2ClCF2Cl and 4,5-dicloroperfluoro-1,3-
diossolano. The 19F NMR of these products is consistent
with the 19F NMR known in the literature. Characterization
of perfluoro-1,2-dichloro-3,5-dioxaheptene is reported
below.
19F NMR spectra were recorded on a Varian 200 MHz
spectrometer with CFCl3 as internal standard. GLC analyses
were performed with a Carlo-Erba instrument equipped with
thermoconductivity detectors. IR spectra were recorded on a
Perkin-Elmer 1600 series FT-IR instruments. GLC/MS
spectra were performed on a Varian Mat CH7-A at 70 eV
in the electron impact mode. All boiling points are uncor-
rected.
Bis(fluoroxy)difluoromethane was prepared according to
standards methods described in the literature [18].
Olefins and other chemicals were obtained from com-
mercial sources and were purified appropriately as required.
Warning: Although it has been found that CF2(OF)2 has a
good thermal stability, because fire and explosion can occur
Boiling point at atmospheric pressure: 91 8C.

194
W. Navarrini, S. Corti / Journal of Fluorine Chemistry 125 (2004) 189–197
19F NMR d: À51.1/À52.1 (AB sys, 2F, –OCF2O–, J ¼
IR(cmÀ1) of the mixture A 79%, B 21%: 1388(w),
1288(vs), 1233(vs), 1151(w), 1104(vs), 1032(s), 846(m),
685(w).
78 Hz), À70.8/À71.4 (AB sys, 2F, C–CF2Cl, J ¼ 170 Hz),
À77.4 (1F, m, OCFCl–), À87.6 (s, 3F, CF3–C), À90.2/À90.8
(AB sys, 2F, C–CF2–O, J ¼ 158 Hz).
GC–MS m/z: 69 (CF3þ), 119 (C2F5þ), 151/153
(C2F3Cl2þ), 185 (C3F7Oþ) 100%.
IR (cmÀ1): 1407(w), 1235(vs), 1032(s), 929(w), 847(m).
3.3. Dehalogenation general procedure and synthesis of
the CF3CF2OCF2OCF¼CF2 perfluoro-3,5-dioxa-
1-heptene ‘‘MOVE 1 monomer’’
3.2. Synthesis of CF3OCF2CF2OCF2OCFClCF2Cl
perfluoro-1,2-dichloro-3,5,8-trioxanonane (isomer A) and
CF3OCF(CF3)OCF2OCFClCF2Cl perfluoro-1,
2-dichloro-3,5,7-trioxa-6-methyloctane (isomer B)
In a 250 ml three-necked flask, equipped with mechanical
stirrer, thermometer, dropping funnel, distillation column
equipped with water refrigerant and collecting trap main-
tained at À78 8C connected to a mechanical vacuum pump,
150 ml of DMF, 15 g of Zn dust, 0.5 g of K2CO3 and
100 mg of I2 were introduced. The internal temperature
was brought to 80 8C and 50 g of perfluoro-1,2-dichloro-
3,5-dioxaheptane were added drop by drop. When the
addition was over the mixture was allowed to react for
about 50 min. At the end the internal pressure was gradually
brought from 760 to 300 mmHg. After about 20 min the
collecting trap, containing 34.2 g of perfluoro-3,5-dioxa-1-
heptene (MOVE 1), was disconnected. The dehalogenation
yield was 85%.
As in Section 3.1, 1.55 l/h of CF2(OF)2 and 4.5 l/h of He
were fed through one of the two inlet pipes; through the
second inlet pipe 1.4 l/h of CF2¼CFOCF3 perfluoromethyl-
vinylether and 0.7 l/h of He were fed for about 4.5 h. The
reaction mixture extracted from the bottom of the reactor
was monitored by GC. The reaction mixture of the flowing
gases, consisting mainly of unreacted BDM, CF3OCF2-
CF2OCF2OF, CF3OCF2CF2O–CF2OCF2CF2OCF3 and the
diluting He (CF2¼CFOCF3 was completely absent) was
continuously fed into the second reactor containing 51 g
of dichlorodifluoroethylene at the temperature of À70 8C.
At the end of the addition (4.5 h) the reaction mixture
contained in the second reactor was distilled by a plate
column at the reduced pressure of 250 mmHg, yielding 50 g
of a mixture formed by two isomers, isomer A and isomer B,
respectively, perfluoro-1,2-dichloro-3,5,8-trioxanonane and
perfluoro-1,2-dichloro-3,5,7-trioxa-6-methyloctane in a
79:21 ratio. The mixture composition was determined by
GC and 19F NMR. The isomers were separated by prepara-
tive GC.
Characterization of perfluoro-3,5-dioxa-1-heptene.
Boiling point at atmospheric pressure: 42 8C.
19F NMR d: À57.0 (m, 2F, –OCF2O–), 87.9 (s, 3F, CF3–
C), À90.7 (t, 2F, C–CF2–O, J ¼ 11 Hz), À116.5 (dd, 1F, O–
C¼CF, J ¼ 87 Hz, J ¼ 67:5 Hz), À123.2 (ddt, 1F, OC¼CF,
J ¼ 112 Hz, J ¼ 87 Hz, J ¼ 5 Hz), À138.0 (ddt, 1F, O–
CF¼C).
GCþ–MS m/z: 69 (CF3þ), 81 (C2F3þ), 97(C2F3Oþ), 119
(C2F5
)
100%, 147 (C3F5Oþ), 185 (C3F7Oþ), 216
(C4F8Oþ), 282 Mþ.
IR (cmÀ1): 1839(m), 1407(w), 1307(vs), 1245(vs),
1117(vs), 907(m), 846(m).
The molar yield of isomer A þ isomerB with respect to
CF2(OF)2 was 38%. The remaining products were mainly:
CF2ClCF2Cl, 4,5-dicloroperfluoro-1,3-dioxolane and CF3-
OCF2CF2OCF2OCF2CF2OCF3. 19F NMR of these products
are consistent with the 19F NMR known in the literature.
The characterization of perfluoro-1,2-dichloro-3,5,8-trioxa-
nonane (isomer A) and perfluoro-1,2-dichloro-3,5,7-trioxa-
6-methyloctane (isomer B) is as follows.
3.4. Synthesis of the ‘‘MOVE 2 monomer’’
CF3OCF2CF2OCF2OCF¼CF2 perfluoro-3,5,8-trioxa-
1-nonene (isomer A) and the MOVE
CF3OCF(CF3)OCF2OCF¼CF2 (isomer B)
perfluoro-3,5,7-trioxa-6-methyl-1-octene
Mixture boiling point (A 79%, B 21%) at 250 mmHg:
82 8C.
As in Section 3.3, 110 ml of DMF, 10 g of Zn dust and
0.3 ml of Br2 were introduced in the 250 ml flask. The
internal temperature was brought to 80 8C and 30.5 g of
a binary mixture prepared in Section 3.2 (perfluoro-1,2-
dichloro-3,5,8-trioxanonane (79%) and perfluoro-1,2-
dichloro-3,5,7-trioxa –6-methyloctane (21%)) were added
drop to drop. When the addition was over, the mixture was
allowed to react for about 3 h. At the end the internal
pressure was reduced to 200 mmHg, after about 30 min
the collecting trap was disconnected. The corresponding
content was washed with water and dried over sodium
sulphate obtaining 24.0 g of a mixture consisting of isomer
A and isomer B in a 79:21 ratio. The overall dehalogenation
yield was 98%. The two isomers were separated by pre-
parative gas chromatography.
19F NMR d. Isomer A: À50.9/À51.8 (AB sys, 2F,–
OCF2O–, J ¼ 78), À70.6/À71.2 (AB sys, 2F, C–CF2Cl,
J ¼ 170 Hz), À77.2 (m, 1F, OCFCl–C), À56.0 (t, 3F,
CF3O, J ¼ 9 Hz), À90.0/À90.5 (AB sys, 2F, C–OC–
CF2OCO, J ¼ 144 Hz), À90.8 (q, 2F, O–CF2–COC,
J ¼ 9 Hz). Isomer B: À51.4 (AB sys, 2F –OCF2O–),
À70.8 (AB sys, 2F, C–CF2Cl), À77.2 (m, 1F, OCFCl–C),
À55.0 (dm, 3F, CF3OC), À86.2 (m, 3F, OC(CF3)CO),
À100.1 (m, 1F, OCF(C)O).
GC–MS m/z. Isomer A: 69 (CF3þ), 119 (C2F5þ) 100%,
151/153 (C2F3Cl2þ), 185 (C3F7Oþ), 251 (C4F9O2þ). Isomer
B: 69 (CF3þ), 97 (C2þF3Oþ), 135 (C2F5Oþ), 151/153
(C2F3Cl2þ), 185 (C3F7O ) 100%.

W. Navarrini, S. Corti / Journal of Fluorine Chemistry 125 (2004) 189–197
195
Characterization of the MOVE perfluoro-3,5,8-trioxa-1-
nonene (isomer A) and the MOVE perfluoro-3,5,7-trioxa-6-
methyl-1-octene (isomer B).
Boiling point range of the isomer mixture at atmospheric
pressure: 72.5–74.5 8C.
The 19F NMR carried out on the polymer dissolved in
perfluorobenzene is in accordance with the homopolymer
structure having a molecular weight of 50,000. This analysis
showed the absence of unreacted monomer. The DSC graph
did not show any melting endothermic curve, wherefore the
polymer was amorphous. The polymer Tg, evaluated by
DSC, is À35.4 8C. The TGA analysis showed a weight loss
of 2% at 332 8C and of 10% at 383 8C.
19F NMR d. Isomer A: CF3OCF2CF2OCF2OCF¼CF2:
À56.5 (t, 3F, CF3O–, J ¼ 9 Hz), À57.1 (m, 2F, –OCF2O–
), 90.8 (t, 2F, C–CF2–O, J ¼ 11 Hz), À91.2 (q, 2F, O–CF2–
C, J ¼ 9 Hz), À116.5 (dd, 1F, O–C¼CF, J ¼ 87 Hz,
J ¼ 66:5 Hz), À123.2 (ddt, 1F, OC¼CF, J ¼ 87 Hz, J ¼
112 Hz), À138 (ddt, 1F, O–CF¼C, J ¼ 112 Hz, J ¼
66:5 Hz).
3.6. Copolymer between ‘‘MOVE 2 monomers’’
perfluoro-3,5,8-trioxa-1-nonene (isomer A) and
perfluoro-3,5,7-trioxa-6-methyl-1-octene (isomer B)
Isomer B: CF3OCF(CF3)OCF2OCF¼CF2: À55.6 (dq, 3F,
CF3O–), À57.1 (m, 2F, –OCF2O–), 86.7 (m, 3F, CF3–C),
À100.3 (m, 1F, –CF–), À116.5 (dd,1F, O–C¼CF, J ¼
87 Hz, J ¼ 66:5 Hz), À123.2 (ddt, 1F, OC¼CF, J ¼ 112
Hz, J ¼ 87 Hz), À138 (ddt, 1F, O–CF¼C, J ¼ 112 Hz,
J ¼ 66:5 Hz).
In a glass reactor of the volume of 20 ml, equipped with
PTFE stopcock and magnetic stirring, the following com-
ponents were charged: 3.2 g of a MOVE mixture prepared
according to 3.4 consisting of perfluoro-3,5,8-trioxa-1-
nonene 83% and perfluoro-3,5,7-trioxa-6-methyl-1-octene
17%. Following the procedure described in Section 3.5,
150 ml of 3% (w/w) perfluoropropionyl peroxide solution
in CF2ClCFCl2 were added in three separated 50 ml portion.
After following the polymerization procedure, the reaction
raw materials appear as a slightly viscous, transparent,
colorless and homogeneous solution. After distillation of
the unreacted monomer, and subsequent stripping under
vacuum at 150 8C for 3 h, 380 mg of the homopolymers
were isolated. The IR analysis of the obtained polymer
showed the absence of the absorption bands corresponding
to the fluorinated monomer insaturation.
GC–MS m/z. Isomerþ A: 69 (CF3þ), 81 (C2F3þ), 97
(C2F3Oþ), 119 (C2F5
)
100%, 147 (C3F5Oþ), 185
(C3F7Oþ), 251 (C4F9O2þ), 282 (C5F10O2þ), Mþ.
Isomer B: 69 (CF3þ), 81 (C2F3þ), 97 (C2F3Oþ), 119
(C2F5
)
100%, 147 (C3F5Oþ), 185 (C3F7Oþ), 282
þ
(C5F10O2þ), Mþ.
IR (cmÀ1) of the AB mixture: 1839(m); 1343(s);
1248(vs); 1245(vs); 1145(vs); 918(m); 889(m).
3.5. Homopolymerization of ‘‘MOVE 1’’ monomer
perfluoro-3,5-dioxa-1-heptene
The 19F NMR carried out on the polymer dissolved in
perfluorobenzene is in accordance with the copolymer
structure having a molecular weight of 35,000 and the
incorporated monomer ratio equal to the ratio of the reacting
MOVE monomers. This analysis showed the absence of
unreacted monomer. The DSC graph did not show any
melting endothermic curve, therefore the polymer was
amorphous. The polymer Tg, evaluated by DSC, is
À52.6 8C. The TGA analysis showed a weight loss of 2%
at 280 8C and of 10% at 327 8C.
In a glass reactor of the volume of 20 ml, equipped with
PTFE stopcock and magnetic stirring, 20 ml of 3% (w/w)
perfluoropropionyl peroxide solution in CF2ClCFCl2, and
3 g of the perfluoro-3,5-dioxa-1-heptene ‘‘MOVE 1’’ pre-
pared in 3.3 were charged. The reactor was cooled to
À196 8C, evacuated, warmed to room temperature, cooled
again to À196 8C and evacuated. The freezing and thaw
procedure was repeated twice. At the end of the degassing
operation the reactor was thermostated at 30 8C while
stirring for 2 days. The reactor was then cooled to
À196 8C and a second portion of 20 ml of perfluoropro-
pionyl peroxide solution was charged under dry nitrogen
atmosphere. After three degassing operations as described
above, the reactor was thermostated at 30 8C while stirring
for 2 days. A third 20 ml portion of perfluoropropionyl
peroxide was finally added following the standard proce-
dure and the reactor was left under stirring at 30 8C for
2 days.
3.7. Amorphous copolymer between ‘‘MOVE 1 monomer’’
perfluoro-3,5-dioxa-1-heptene and TFE (general TFE
copolymerization procedure)
In a AISI-316 Parr microreactor of the volume of 40 ml,
equipped with magnetic stirring, pressure transducer and
an inlet for the reactant feeding and discharge, 250 ml of 3%
(w/w) perfluoropropionyl peroxide solution in CF2ClCFCl2,
9.8 mmol of perfluoro-3,5-dioxa-1-heptene ‘‘MOVE 1
monomer’’ and 18 mmol of tetrafluoroethylene were intro-
duced. The reactor was then cooled to À196 8C, evacuated,
warmed to À78 8C and cooled again to À196 8C and
evacuated. This procedure was repeated twice.
The reaction raw material appears as a viscous, transpar-
ent, colorless and homogeneous solution.
After distillation of the unreacted monomer, and subse-
quent stripping under vacuum at 150 8C for 3 h, 260 mg of
the homopolymers were isolated.
The IR analysis of the obtained polymer showed the
absence of the absorption bands corresponding to the fluori-
nated monomer insaturation.
At the end of the degassing operations the reactor was
thermostated at 30 8C while stirring for about 8 h. The
internal pressure decreased from 6.4 to 4.7 atm. The reaction

196
W. Navarrini, S. Corti / Journal of Fluorine Chemistry 125 (2004) 189–197
was quenched by cooling the reactor at À78 8C, while the
unreacted monomers were distilled under vacuum. The
copolymer was then stripped at 150 8C at 103 mmHg for
3 h, yielding 1100 mg of copolymer.
The molar percentage of TFE and perfluoro-3,5-dioxa-1-
heptene incorporated in the copolymer, determined by 19F
NMR analysis of the polymer dissolved in C6F6 at 50 8C
were respectively 76 and 24%.
The IR analysis of the obtained copolymer shows the
absence of the absorption bands corresponding to the mono-
mers insaturation and the presence of very small absorption
bands in the region of the carboxyl signals. The intensity of
these signals, compared with the analogous one obtained
from a film normalized at the identical thickness obtained
with a similar copolymer prepared in Section 3.9, is equal to
about 1/10 of the latter.
The DSC graph did not show any melting endothermic
curve, wherefore the polymer was amorphous. The polymer
Tg, evaluated by DSC, is À21.4 8C. The TGA analysis
showed a weight loss of 2% at 450 8C and of 10% at
477 8C. The polymer intrinsic viscosity measured at
30 8C in Fluorinert1 FC-75, is 35.5 ml/g.
intrinsic viscosity measured at 30 8C in Fluorinert1
FC-75, is 16.7 ml/g.
3.9. Amorphous copolymer between TFE and
perfluoropro-3-oxa-1-hexene (PVE)
In a polymerization reactor identical to that used in the
previous examples, 250 ml of 3% (w/w) perfluoropropionyl
peroxide solution in CF2ClCFCl2, 9.7 mmol of PVE per-
fluoropro-3-oxa-1-hexene and 18 mmol of TFE were in
sequence introduced.
The procedure described in Section 3.7 was followed until
thermosetting at 30 8C while stirring for 8 h.
ThereactionwasquenchedbycoolingthereactoratÀ78 8C
and the unreacted monomers were distilled under vacuum.
Thecopolymer was then stripped at150 8C at10À3 mmHg for
3 h and 540 mg of copolymer were recovered.
The molar percentage of TFE and PVE perfluoropro-3-
oxa-1-hexene incorporated in the copolymer, determined by
19F NMR analysis of the polymer dissolved in C6F6 at 50 8C
were respectively 77 and 23%.
The IR analysis of the obtained copolymer showed the
absence of the absorption bands corresponding to the mono-
mers insaturation and showed the presence of small absorp-
tion bands in the region of the carboxyl signals. The intensity
of these signals, compared with the analogous signals
obtained from a film normalized to the identical thickness,
obtained with a similar copolymer prepared in Section 3.7, is
10 times higher.
The DSC graph did not show any melting endothermic
curve, wherefore the polymer was amorphous. The polymer
Tg, evaluated by DSC, is 15 8C. The TGA analysis showed a
weight loss of 2% at 427 8C and of 10% at 463 8C. The
polymer intrinsic viscosity measured at 30 8C in Fluorinert1
FC-75, is 51 ml/g.
3.8. Amorphous copolymer between TFE and ‘‘MOVE 2
monomers’’: perfluoro-3,5,8-trioxa-1-nonene and
perfluoro-3,5,7-trioxa-6-methyl-1-octene
In a polymerization reactor identical to that used in
Section 3.7, 100 ml of 6% (w/w) perfluoropropionyl per-
oxide solution in CF2ClCFCl2, 9.7 mmol of ‘‘MOVE 2
monomers’’ mixture:
perfluoro-3,5,8-trioxa-1-nonene
(83%) and perfluoro-3,5,7-trioxa-6-methyl-1-octene (17%)
prepared in 3.4 and 10 mmol of TFE were introduced in
sequence. The general TFE copolymerization procedure
described in Section 3.7 was then followed until the thermo-
setting at 30 8C. The internal pressure decreased from 3.6
to 2.7 atm in about 8 h. The reaction was quenched by
cooling the reactor at À78 8C and the unreacted monomers
were distilled under vacuum. The copolymer was then strip-
ped at 150 8C at 10À3 mmHg for 3 h, yielding 652 mg of
copolymer.
The molar percentage of TFE and MOVE 2 monomers
incorporated in the copolymer, determined by 19F NMR
analysis of the polymer dissolved in C6F6 at 50 8C
were respectively 63 and 37%. The molar ratio of the
MOVE monomers, perfluoro-3,5,8-trioxa-1-nonene/perfluoro-
3,5,7-trioxa-6-methyl-1-octene incorporated in the polymer
is 83/17 and it is equal to the starting MOVE 2 monomers
feeding mixture.
3.10. Amorphous copolymer between TFE and
perfluoropro-3,6-dioxa-1-heptene CF3OCF2CF2OCF¼CF2
(b-PDE)
In a polymerization reactor identical to that used in the
previous examples, 250 ml of 3% (w/w) perfluoropropionyl
peroxide solution in CF2ClCFCl2, 10 mmol of perfluoropro-
3,5-dioxa-1-heptene and 18 mmol of TFE were in sequence
introduced.
The procedure already followed in Section 3.7 was
adopted until thermosetting at 30 8C under stirring.
The molar percentage of TFE and perfluoro-3,6-dioxa-1-
heptene incorporated in the copolymer, determined by 19F
NMR analysis of the polymer dissolved in C6F6 at 50 8C
were respectively 77 and 23%.
The IR analysis of the obtained copolymer showed the
absence of the absorption bands corresponding to the mono-
mers insaturation. The DSC graph did not show any melting
endothermic curve, therefore the polymer was amorphous.
The polymer Tg, evaluated by DSC, is À4.8 8C.
The IR analysis of the obtained copolymer showed the
absence of the absorption bands corresponding to the mono-
mers insaturation.
The DSC graph did not show any melting endothermic
curve, wherefore the polymer was amorphous. The polymer
Tg, evaluated by DSC, is À44.5 8C. The TGA analysis
showed a weight loss of 10% at 451 8C. The polymer

W. Navarrini, S. Corti / Journal of Fluorine Chemistry 125 (2004) 189–197
[2] W. Navarrini, Ausimont EP 1,148,041 (2001).
197
4. Conclusion
[3] A. Marracini, G. Perego, Ausimont US Patent 4,962,282 (1990).
[4] D.D. DesMarteau, J.D. Anderson, W. Navarrini EP 0,754,670
(1996).
The present work describes the preparation and charac-
terization of new perfluorovinylether monomers readily
available from standard reactants in fluorine chemistry. This
monomer family is characterized by a OCF2O unit directly
bonded to the insaturation. This structure gives to these
monomers a high capability of decreasing the Tg’s of TFE
and VDF copolymers. These intrinsic characteristics,
straight synthesis and high Tg decreasing capability, make
this monomer family a perfect candidate for the preparation
of low Tg perfluororubbers.
[5] W. Navarrini, V. Tortelli, A. Zedda, Ausimont EP 0,683,181
(1995).
[6] M.H. Hung, Macromolecules 26 (1993) 5829–5834.
[7] W. Navarrini, L. Bragante, S. Fontana, V. Tortelli, A. Zedda,
J. Fluorine Chem. 71 (1995) 111.
[8] B.V. Mislavsky, Fluorinated polymers with functional groups. in:
G. Hougham, et al. (Eds.), Fluoropolymers 1 Synthesis, Kluwer
Academic/Plenum Publishers, New York, 1999, pp. 91–108.
[9] S. Rozen, Chem. Rev. 96 (1996) 1717–1736.
[10] I. Ben-David, E. Mishani, S. Rozen, J. Org. Chem. 63 (1998) 4632–
4635.
[11] K.K. Johri, D.D. DesMarteau, J. Org. Chem. 48 (1983) 242–250.
[12] G. Guglielmo, G.P. Gambaretto, Ausimont US Patent 4,400,872
(1990).
Acknowledgements
[13] M. Malavasi, D. Sianesi, J. Fluorine Chem. 95 (1999) 19–25.
[14] V. Arcella, P. Colaianna, P. Maccone, A. Sanguineti, A. Giordano,
G. Clarizza, E. Orioli, J. Membr. Sci. 163 (1999) 203–209.
[15] B. Ameduri, B. Boutevin, G. Kostov, Prog. Polym. Sci. 26 (2001)
105–187.
We wish to thank Prof. Darryl D. DesMarteau and Prof.
D. Sianesi for their suggestions and helpful discussions.
[16] S.W. Benson, The Foundation of Chemical Kinetics, McGraw-Hill
Book Company, New York, 1960, pp. 42–46.
References
[17] S. Saito, H. TaTsu, Nippon Mektron, Li. US Patent 5,902,868
(1999).
[1] W. Navarrini, V. Tortelli, A. Russo, S. Corti, J. Fluorine Chem. 95
(1999) 27–39.
[18] M.J. Fifold, US Patent 4,499,024 (1985).