Kinetics and mechanism of the thermal gas-phase oxidation of perfluorobutene-2 in presence of trifluoromethyl hypofluorite

Article abstract of DOI:10.1002/kin.10138

The oxidation of perfluorobutene-2 (C₄F₈) initiated by trifluoromethyl hypofluorite (CF₃OF) in presence of O₂ has been studied at 323.1, 332.6, 342.5, and 352.0 K, using a conventional static system. The initial pressure of CF₃OF was varied between 4.8 and 23.6 Torr, that of C₄F₈ between 48.7 and 302.4 Torr, and that of O₂ between 51.5 and 270.4 Torr. Several runs were made in presence of 325.5-451.2 Torr of N₂. The main products were COF₂, CF₃C(O)F, and CF₃OC(O)F. Small amounts of compound containing -CF(CF₃)-O-C(O)CF₃ group were also formed, as detected by ¹³C NMR spectroscopy. The oxidation is a homogeneous short-chain reaction, attaining, at the pressure of O₂ used, the pseudo-zero-order condition with respect to O₂ as reactant. The reaction is independent of the total pressure. Its basic steps are as follows: the thermal generation of CF₃O radicals by the abstraction of fluorine atom of CF₃OF by C₄F₈, the addition of CF₃O to the alkene, the formation of perfluoroalkoxy radicals RO in presence of O₂, and the decomposition of these radicals via the C-C bond scission, giving products containing -C(O)F end group and reforming RO and CF₃O radicals. The mechanism consistent with experimental results is postulated.

Full text of DOI:10.1002/kin.10138

Kinetics and Mechanism
of the Thermal Gas-Phase
Oxidation of
Perfluorobutene-2 in
Presence of Trifluoromethyl
Hypofluorite
1
1,2
ROSANA M. ROMANO, CARLOS O. DELLA VEDOVA, JOANNA CZARNOWSKI3
´
1CEQUINOR (CONICET, UNLP) Departamento de Quı´mica, Facultad de Ciencias Exactas, Universidad de La Plata,
47 esq. 115, 1900 La Plata, Argentina
2Laboratorio de Servicios a la Industria y al Sistema Cientı´fico (UNLP-CIC-CONICET), Departamento de Quı´mica,
Facultad de Ciencias Exactas, Universidad de La Plata, 47 esq. 115, 1900 La Plata, Argentina
3Instituto de Investigaciones Fisicoquı´micas Teo´ricas y Aplicadas, Sucursal 4, Casilla de Correo 16, 1900 La Plata, Argentina
Received 16 September 2002; accepted 25 March 2003
DOI 10.1002/kin.10138
ABSTRACT: The oxidation of perfluorobutene-2 (C4F8) initiated by trifluoromethyl hypofluorite
(CF3OF) in presence of O2 has been studied at 323.1, 332.6, 342.5, and 352.0 K, using a conven-
tional static system. The initial pressure of CF3OF was varied between 4.8 and 23.6 Torr, that
of C4F8 between 48.7 and 302.4 Torr, and that of O2 between 51.5 and 270.4 Torr. Several runs
were made in presence of 325.5–451.2 Torr of N2. The main products were COF2, CF3C(O)F, and
CF3OC(O)F. Small amounts of compound containing CF(CF3) O C(O)CF3 group were also
formed, as detected by 13C NMR spectroscopy. The oxidation is a homogeneous short-chain
reaction, attaining, at the pressure of O2 used, the pseudo-zero-order condition with respect
to O2 as reactant. The reaction is independent of the total pressure. Its basic steps are as fol-
lows: the thermal generation of CF3O• radicals by the abstraction of fluorine atom of CF3OF by
Correspondence to: Joanna Czarnowski; e-mail: bdq782@
infovia.com.ar.
Contract grant sponsor: Alexander von Humboldt Stiftung.
Contract grant sponsor: The British Council.
Contract grant sponsor: Consejo Nacional de Investigaciones
Cient´ıficas y Te´cnicas (CONICET).
Contract grant sponsor: Deutsche Akademische Austauschdi-
enst, Germany (DAAD).
Contract grant number: PIP 4695.
Contract grant sponsor: ANCyT.
Contract grant number: PICT 6786.
Contract grant sponsor: Comisio´n de Investigaciones Cient´ıficas
de la Provincia de Buenos Aires (CIC), Repu´blica Argentina.
Contract grant sponsor: Facultad de Ciencias Exactas, Universi-
Contract grant sponsor: Fundacio´n Antorchas, Repu´blica
Argentina.
dad Nacional de La Plata, Repu´blica Argentina.
2003 Wiley Periodicals, Inc.
c
ꢀ

THERMAL GAS-PHASE OXIDATION OF PERFLUOROBUTENE-2
533
C4F8, the addition of CF3O• to the alkene, the formation of perfluoroalkoxy radicals RO• in pres-
ence of O2, and the decomposition of these radicals via the C C bond scission, giving products
containing C(O)F end group and reforming RO• and CF3O• radicals. The mechanism consis-
tent with experimental results is postulated. ꢀ 2003 Wiley Periodicals, Inc. Int J Chem Kinet 35:
C
532–541, 2003
The stable and easily handled trifluoromethyl hy-
pofluorite, CF3OF, containing a weak O F bond, is a
versatile and selective fluorinating agent [9]. The value
of 43.5 kcal mol−1 was determined for the dissociation
energy of O F bond [10–12]. The thermal addition of
CF3OF across the double bond of a variety of olefins
belongs to a very interesting class of reactions, where a
saturated compound with a relatively weak bond initi-
ates a series of reactions, important from the synthetic
and mechanistic point of view.
In this work the investigation of the thermal gas-
phase oxidation of C4F8 in the presence of CF3OF has
been undertaken to study the use of chemical initiators
in the oxidative processes of fluoroolefins, leading to
the formation of compounds containing the C O C
ether groups.
INTRODUCTION
In recent years the fluorine chemistry has been continu-
ously growing in areas such as new fluoropharmaceuti-
cals, nonsoluble, noncorrosivebiocompatiblematerials
for implants in medicine, agrochemicals, alternative re-
frigerants, paints, electronics, surface coatings for cor-
rosion and moisture protection, and high performance
lubricants.
The fluid perfluoropolyethers are used as lubricants
for the computer hard disks because of their good per-
formance as a protective layer between the hard disk
surface and the magnetic head. The major route utilized
for their commercial production is the photooxidation
[1–5] of perfluoroolefins (CF2CF2, CF3CFCF2) in liq-
uid phase at 233 K or lower, initiated by UV light.
These lubricants are mixtures of species with different
relative molecular masses, containing CF2
CF(CF3) O groups.
O and
EXPERIMENTAL
In the gas-phase photooxidation of perfluoro-
propene at 303 K [2], the main reaction products re-
ported were COF2, CF3C(O)F, CF3OC(O)F, and CF3
OCF2C(O)F and a mixture of two series of telomers,
CF3O(CF2O)nC(O)F and CF3O(CF2O)nCF2C(O)F,
where n = 1–12. The formation of CF3C(O)F and
CF3OC(O)F was observed in the mercury-photo-
sensitized oxidation of perfluorobutene, C4F8 [6].
It was found that in the photooxidation of liquid hex-
afluorobutadiene, C4F6, more than 50% of diolefin was
converted, through a kinetically complicated process,
into a high relative molecular mass liquid polyethers,
havingacomplexchainstructurewithC(O)Fendgroup
[2,7].
The use of chemical initiators, instead of UV light,
for synthesis of perfluoropolyethers has significant
practical and technological advantages. It is an interest-
ing alternative offering a better control over the struc-
ture and relative molecular mass of the products. The
possibility of synthesis of simpler molecules provides
useful tools for the studies of the unique properties
of perfluoropolyethers attributed to their different seg-
ments and end groups. The chemical initiator must be
able to generate free radicals that should be either ca-
pable of direct addition to the double bond or to do so
after their reaction with oxygen. Recently the fluorine-
initiated liquid-phase oxidation of C2F4 was reported
in the literature [7,8].
The reaction proceeded with pressure decrease. The
experiments were performed in a grease-free static sys-
tem allowing pressure measurements at constant vol-
ume and temperature. Two reaction vessels were used:
a spherical quartz bulb with a volume of 270 cm3
(surface-to-volume ratio, S/V = 0.75 cm−1) and an-
other one of similar dimension filled with small pieces
of quartz tubing (S/V = 4.7 cm−1). The pressure was
measured with a quartz spiral gauge and the tempera-
ture maintained within 0.1 K using a Lauda thermo-
stat. Infrared spectra were recorded on a Perkin-Elmer
325spectrometerandaPerkin-Elmer1600SeriesFTIR
spectrometer, using 10-cm cells provided with NaCl
and KBr windows, respectively. The resolution of the
reported wave numbers was 1 cm−1
.
All reactants were purchased commercially. CF3OF
(PCR, 97–98%) was washed with 0.1 mol dm−3 NaOH
solution and filtered at 80 K [13]. C4F8 (PCR, 97–
98%) was purified by repeated low-pressure trap-to-
trap distillations on a vacuum line, the middle fraction
being retained each time. The purified reactants were
spectroscopically pure. O2 (La Oxigena, 99.99%) and
N2 (La Oxigena, 99.9%) were bubbled through 98%
analytical-grade H2SO4, and passed slowly through
a Pyrex coil at 153 K and at liquid-air temperature,
respectively.
The reaction was followed, measuring the con-
sumption of C4F8 as a function of time. Twenty-nine

´
ROMANO, DELLA VEDOVA, AND CZARNOWSKI
534
experiments were carried out at 323.1, 332.6, 342.5,
and 352.0 K. The initial pressure of CF3OF was varied
between 4.8 and 23.6 Torr, that of C4F8 between 48.7
and 302.4 Torr, and that of O2 between 51.5 and 270.4
Torr. Several experiments were carried out in presence
of 325.5–451.2 Torr of N2. Two experiments were car-
the third fraction Fr3, volatile at 163 K, in addition to
the infrared absorption bands of CF3OC(O)F, a small
band at 1830 cm−1 was also observed. It can be at-
tributed to the C O group present in the CF(CF3)
OC(O)CF3 group of the product X. It was reported
that the absorption band corresponding to C O group
of the compound (CF3)2C(F)OOC(O)CF3 appeared at
1849 cm−1 [21]. The fourth fraction Fr4, remaining as
residue at 163 K, consisted of nonconsumed perfluo-
robutene.
The amounts of CF3OF consumed correspond clo-
sely to the calculated ones using the expression [CF3-
OF]consum = [CF3OF]i × {1 − exp(−k1ꢀt[C4F8]m)},
where ꢀt is the reaction time, [C4F8]m is the mean
pressure of perfluorobutene within the interval ꢀt,
and k1 = (5.34 0.3) × 106 exp(−(13.9 0.8) kcal
mol−1/RT ) dm3 mol−1 s−1 is the rate constant for the
gas-phase addition of CF3OF to the double bond of
C4F8 in absence of O2, obtained previously [22].
ried out in the reactor with S/V = 4.7 cm−1
.
The reaction is a homogeneous chain reaction, with
a very short chain length. Its course does not change
on increasing the surface-to-volume ratio. Under the
conditions of our work the reaction was independent
of the total pressure and attained the pseudo-zero-order
condition with respect to O2 as reactant.
In absence of CF3OF, no reaction was observed be-
tween C4F8 and O2 after several hours.
In presence of CF3OF, the following products
were detected by infrared spectroscopy: COF2 [14],
CF3C(O)F [15], and CF3OC(O)F [16–18]. Small
amounts of an unknown compound X, containing
CF(CF3)OC(O)CF3 group, were also formed, as de-
tected by 13C NMR spectroscopy.
In the 13C NMR spectrum of a sample containing the
unknown compound X (C(1)F3C(2)(O)OC(3)FC(4)
F3 ), obtained at room temperature and in a CDCl3
solution, using a Bruker AC 250 NMR spectrometer,
four signals were observed: δ = 114.49 ppm (c, C(1),
RESULTS
The analytical data of 23 selected experiments are sum-
marized in Table I.
The reaction mechanism proposed to explain the ex-
perimental results is presented in Table II. In this mech-
anism the considered telomerization steps, initiated by
the addition of radical CF3OC(O)•FCF3 to olefin and
giving CF3O(CF(CF3)O)nC(O)F, where n = 1,2 . . . ,
were not included. These telomerization steps should
occur only to a very small extent, not affecting substan-
tially the interpretation of the experimental results.
The corresponding global reaction for telomeriza-
tion can be written as
1 JCF = 284.67 Hz), δ = 161.79 ppm (c, C(2), 2 JCF
=
43.87 Hz), δ = 137.04 ppm (cd, C(3), 1 JCF = 296.59
2
Hz, JCF = 2.86 Hz), δ = 119.04 ppm (dc, C(4),
1 JCF = 269.65 Hz, 2 JCF = 10.96 Hz).
No absorption bands between 1450 and 1650 cm−1
were found in the infrared spectra obtained. It indi-
O
ꢀ
❝
cates that the perflurobutene epoxide, CF3CF CFCF3,
did not form in the reaction course. A strong ab-
sorption band in this frequency range is characteristic
of perfluorinated olefin epoxides, not being produced
by other perfluorocarbon compounds. Neither was the
very strong band of CF3OOCF3 at 1166 cm−1 [19,20]
observed.
CF3OC(O)•FCF3 + nCF3CFCFCF3 + nO2
→ CF3O(CF(CF3)O)nC(O)F + nCF3C(O)F + CF3
•
It was reported that the fluoroformate group of the
perfluoropolyethers containing C(CF3)F O C(O)F
end group decomposes thermally to give C(O)CF3
group and C(O)F2 [2]; then the formation of compound
X containing C(CF3)F O C(O) CF3 group can be
explained by the following global reaction:
To analyze the reaction mixture of each experiment
andtofollowtheconsumptionofthereactantsasafunc-
tion of time, the reaction vessel was rapidly cooled to
liquid-air temperature at selected intervals of reaction
time and the mixture separated by fractional conden-
sation. The nonconsumed O2 and any N2, if present,
were separated as volatile at liquid-air temperature as
the first fraction Fr1. The second fraction Fr2, volatile
at 123 K, consisted of CF3OF and COF2. To deter-
mine the concentration of COF2 in this fraction, an
infrared calibration curve was obtained using the pure
compound, thereby allowing conversion of the absorp-
tion intensity at 1942 cm−1 to the pressure of COF2.
The amount of CF3OF was obtained by difference. In
CF3O(CF(CF3)O)nC(O)F
→ CF3O(CF(CF3)O)n−1C(O)CF3 + COF2
where n = 2.
Applying the steady-state approximation to the pro-
posed mechanism, the following expressions were ob-
tained for the consumption of perfluorobutene-2 (C4F8)

THERMAL GAS-PHASE OXIDATION OF PERFLUOROBUTENE-2
535

´
ROMANO, DELLA VEDOVA, AND CZARNOWSKI
536
Table II The Mechanism for the Oxidation of C4F2-2 Initiated by CF3OF Proposed to Explain the Experimental Results
1. CF3OF + CF3CF CFCF3 → CF3CF2C•FCF3 + CF3O•
2. CF3CF2C•FCF3 + O2 + M → CF3CF2C(O2)•FCF3 + M
3. CF3CF2C(O2)•FCF3 + RO2 → CF3CF2C(O)•FCF3 + RO + O2
4. CF3CF2C(O)•FCF3 → CF3C•F2 + CF3C(O)F
•
5. CF3C•F2 + O2 + M → CF3CF2O2 + M
•
6. CF3CF2O2 + RO2 → CF3CF2O• + RO + O2
•
7. CF3CF2O• → CF3 + COF2
•
•
8. CF3 + O2 + M → CF3O2 + M
9. CF3O2 + RO2 → CF3O• + RO + O2
•
10. CF3O• + CF3CF CFCF3 → CF3OCF(CF3)C•FCF3
11. CF3OCF(CF3)C•FCF3 + O2 + M → CF3OCF(CF3)C(O2)•FCF3 + M
12. CF3OCF(CF3)C(O2)•FCF3 + RO2 → CF3OCF(CF3)C(O)•FCF3 + RO + O2
13. CF3OCF(CF3)C(O)•FCF3 → CF3OC•FCF3 + CF3C(O)F
14. CF3OCF(CF3)C(O)•FCF3 → CF3 + CF3OCF(CF3)C(O)F
•
15. CF3OC•FCF3 + O2 + M → CF3OC(O2)•FCF3 + M
16. CF3OC(O2)•FCF3 + RO2 → CF3OC(O)•FCF3 + RO + O2
17. CF3OC(O)•FCF3 → CF3 + CF3OC(O)F
•
18. CF3CF2O• + CF3CF CFCF3 → CF3CF2OCF(CF3)C•FCF3
19. CF3CF2OCF(CF3)C•FCF3 + O2 + M → CF3CF2OCF(CF3)C(O2)•FCF3 + M
20. CF3CF2OCF(CF3)C(O2)•FCF3 + RO2 → CF3CF2OCF(CF3)C(O)•FCF3 + RO + O2
21. CF3CF2OCF(CF3)C(O)•FCF3 → CF3CF2OC•FCF3 + CF3C(O)F
22. CF3CF2OC•FCF3 + O2 + M → CF3CF2OC(O2)•FCF3 + M
23. CF3CF2OC(O2)•FCF3 + RO2 → CF3CF2OC(O)•FCF3 + RO + O2
24. CF3CF2OC(O)•FCF3 → CF3 + CF3CF2OC(O)F
•
25. R• + CF3OF → RF + CF3O•
26. R• + R• → Recombination products
In this mechanism R• = CF3 , CF3C•F2, CF3CF2C•FCF3, CF3OC•F(CF3), CF3OCF(CF3)C•FCF3, CF3CF2OC•F(CF3), or CF3CF2OCF
•
(CF3)C•F(CF3).
and that of CF3OF:
−d[C4F8]/dt = −n d[CF3OF]/dt
where (ꢀ[C4F8]/ꢀt)exp is the experimental consump-
tion rate of C4F8.
The constants k at different temperatures were es-
timated plotting (ꢀ[C4F8]/ꢀt)exp/([CF3OF][C4F8]) as
a function of ([CF3OF]/[C4F8])0.5. Subsequently the
constant kꢁ equal to k25(k1k26)−0.5 was determined as a
slope, plotting (m − 1) vs. ([CF3OF]/[C4F8])0.5, where
m = (ꢀ[C4F8]/ꢀt)exp/k ([CF3OF][C4F8]). The plot is
illustrated in Fig. 1. As can be seen from this plot, kꢁ is
equal to 0.5 and does not depend on temperature.
In order to derive k25, the rate constant for the ab-
straction of fluorine atom from CF3OF by perfluo-
roalkyl radicals, from the value of kꢁ, k26 was assumed
to be equal to the collisional frequency factor, 1010 dm3
mol−1 s−1, and k1 was calculated using the expression
obtained previously in the study of the thermal gas-
phase addition of CF3OF to C4F8 in absence of O2
[22]:
= λk1[CF3OF][C4F8]
−d[CF3OF]/dt = k1[CF3OF][C4F8]
+ k25[R•][CF3OF]
(I)
(II)
where the kinetic chain length λ = nγ is the ratio be-
tween the consumption rate of perfluorobutene and the
rate of initiation k1[CF3OF][C4F8], and γ is the ratio of
the rate of consumption of CF3OF to that of initiation.
Taking into account that the termination rate k26[R]2
is equal to that of initiation, and that reaction (7) is more
rapid than reaction (18), the following expression was
obtained for λ:
λ = (1 + k10[CF3O•])/k1[CF3OF]
= n{1 + k25(k1k26)−0.5([CF3OF]/[C4F8])0.5
}
k1 = (5.34 0.3) × 106 exp(−(13.9 0.8)
kcal mol−1/RT ) dm3 mol−1 s−1 (V)
(III)
nk1 = k = (ꢀ[C4F8]/ꢀt)exp/{[CF3OF][C4F8]
× (1 + k25(k1k26)−0.5([CF3OF]/[C4F8])0.5)}
The k25 values of (2.32 0.7) × 103, (3.16
0.9) × 103, (4.27 1.2) × 103, and (5.63 1.5) × 103
(IV)

THERMAL GAS-PHASE OXIDATION OF PERFLUOROBUTENE-2
537
(3.0 2) × 109 and (3.7 0.5) × 109 dm3 mol−1 s−1
and the corresponding activation energies were 5.62
2.4 and 6.0 1.1 kcal mol−1, respectively.
,
As the termination rate (k26[R]2) and kꢁ are inde-
pendent of temperature, the value of 6.95 kcal mol−1
for the activation energy E25 was obtained using
ꢁ
the relation Ek = E25 − 0.5E1 − 0.5E26, where E1 =
13.9 kcal mol−1, E26 ≈ 0, and Ek ≈ 0.
ꢁ
The constants k at different temperatures were recal-
culated for each experiment, replacing k25(k1k26)−0.5 in
Eq. (IV) by the obtained value of kꢁ and using rate con-
stants k1 calculated by Eq. (V). They are summarized
in Table III.
The following mean values for k were ob-
tained: k(323.1 K) = (4.7 0.9) × 10−3 dm3 −m1 ol−1
,
s
−1, k(332.6 K) = (9.4 1.9) × 10−3 dm3 mol s−1
k(342.5 K) = (1.92 0.2) × 10−2 dm3 mol−1
s
−1, and
k(352.0 K) = (4.54 0.8) × 10−2 dm3 mol−1 s−1
.
Figure 1 Plot of (m − 1) vs. ([CF3OF]/[C4F8])0.5, where
m = (ꢀ[C4F8]/ꢀt)exp/k[CF3OF][C4F8]. The value of the
−0.5
.
slope is equal to kꢁ = k25(k1k26
)
DISCUSSION
dm3 mol−1 s−1 were obtained at 323.1, 332.6, 342.5,
and 352.0 K, respectively. The Arrhenius plot of k25
is illustrated in Fig. 2. The temperature dependence of
k25 can be expressed by:
The synthetic aspects of the thermal addition of CF3OF
to several haloalkenes have been reported by Johri
and DesMarteau [25] and Sekiya and Ueda [26]. In
their works the trifluoromethylhaloalkyl ethers were
the main products. The observed reactivity and the low
stereo- and regioselectivity indicated that these reac-
tions occurred via a free-radical mechanism.
To investigate the kinetic and mechanistic aspects
of the addition of CF3OF to alkenes in the gas phase,
the reactions of CF3OF with C3F6 [27], C4F8 [22],
CF2CCl2 [28], CHClCCl2 [23], and CCl2CCl2 [24] in
absence of O2 and with CF2CCl2 [29], CHClCCl2 [30],
CCl2CCl2 [31], and C3F6 [32] in presence of O2 have
been studied.
k25 = (1.14 0.5) × 108 exp(−(6.94 1)
kcal mol−1/RT ) dm3 mol−1 s−1 (VI)
The respective preexponential factors found for the
abstraction of fluorine atom from CF3OF by radicals
•
•
CF3OCHClCCl2 [23] and CF3OCCl2CCl2 [24] were
The primary path in these gas-phase reactions is the
homolytic cleavage of the O F bond in a bimolecular
reaction between CF3OF and alkene (A), giving rad-
icals CF3O• and F(A)•. The radicals CF3O• add rapidly
to the double bond, forming CF3O(A)•. The reaction
of CF3OF through the homolytic cleavage of its O F
bond was confirmed in the study of the addition of
CF3OF to a series of electron-poor haloalkenes (A) in
a nonpolar solvent [33]. The CF3O(A)• intermediate
radicals were observed by electron paramagnetic
resonance and electron-nuclear double resonance
techniques in the reaction systems CF3OF + per-
fluoroalkenes [33,34]. It indicates that when the
reactant–solvent interaction is weak, the gas-phase
reaction dynamics rules may be carried over to the
solution reactions [35], suggesting that, in this case, the
mechanism valid for the liquid-phase reaction does not
differ substantially from that for the gas-phase reaction.
Figure 2 Arrhenius plot of rate constant k25
.

´
ROMANO, DELLA VEDOVA, AND CZARNOWSKI
538
Table III Constants k and n = k/k1 at Different Temperatures
No. of Run T (K) ꢀt (min) CF3OFi (Torr) CF3OFf (Torr) C4F8i (Torr) C4F8f (Torr) O2i (Torr) k (105 Torr−1 min−1
)
n
26
27
28
2
3
4
22
23
24
6
7
8
323.1 303.9
323.1 237.0
323.1 449.8
19.8
12.0
11.5
23.6
14.6
13.0
19.5
13.4
8.5
14.2
11.7
6.5
20.0
19.7
4.8
16.4
10.4
10.5
22.4
6.4
9.0
16.5
8.8
5.3
8.6
5.5
2.0
16.8
15.1
3.1
7.0
14.9
8.7
3.4
16.2
3.3
7.0
16.0
97.8
89.3
58.8
88.8
85.5
54.1
92.8
183.6
92.7
93.6
289.5
191.3
84.6
184.3
186.2
37.9
88.0
83.3
86.4
84.7
38.9
89.0
86.3
185.5
85.8
86.3
81.3
81.1
51.5
1.44
1.39
1.37
2.75
2.60
2.71
2.66
2.89
2.70
5.33
5.44
5.39
5.47
5.48
5.30
2.25
2.17
2.14
2.39
2.26
2.36
2.31
2.51
2.35
2.60
2.65
2.63
2.67
2.67
2.59
3.50
3.47
3.73
3.64
3.58
3.50
3.70
3.58
332.6
27.0
94.8
270.4
154.8
99.1
102.0
153.6
150.6
118.0
149.8
150.6
77.5
106.9
100.6
111.6
113.6
69.2
101.7
111.6
159.7
262.6
109.0
332.6 358.6
332.6 226.2
332.6 123.3
332.6 124.0
332.6 191.0
342.5 188.0
342.5 181.3
342.5 298.5
342.5 187.0
342.5 130.0
342.5 239.4
352.0 134.8
204.8
100.3
100.5
302.4
198.9
100.6
202.5
200.6
48.7
102.2
88.0
102.4
100.9
51.2
9
10
11
12
13
14
15
16a
18
19
20b
11.0
19.1
11.3
6.5
20.0
7.1
12.1
352.0
70.5
12.0
12.9
12.6
12.4
12.1
12.8
12.4
352.0 170.8
352.0 195.4
102.1
101.3
200.7
100.8
101.5
352.0
58.3
352.0 116.0
352.0 125.0
10.4
19.9
352.0
59.8
Note: ꢀt is the reaction time. The subscripts i and f signify initial and final, respectively.
a 451.2 Torr of N2 was present.
b The surface-to-volume ratio of the reaction vessel used in this experiment was 6.3 times greater.
In absence of O2 the radicals CF3O(A)• abstract flu-
orine from CF3OF, regenerating CF3O• and initiating
the chain reaction with trifluoromethylhaloalkyl ethers,
CF3O(A)F, as the main products.
of CF3O• [36–41]. The values of the order of 1010 dm3
mol−1 s−1 were determined for the addition of CF3O•
to several alkenes [38–41].
•
The study of the equilibrium of R• + O2 → RO2
In presence of O2 the radicals CF3O(A)• react
rapidly with O2, leading to the formation of haloalkoxy
radicals CF3O(A)O•, which initiate a series of reactions
giving as the main products compounds containing the
C(O)X end group, where X = Cl or F.
[42] indicates that at the pressure of O2 used in our
work, the CF3O(C4F8)• radicals and other perfluo-
roalkyl radicals generated in this reaction system re-
act rapidly with O2. The high pressure limit values
of 1.5 × 109 and 1.26 × 109 dm3 mol−1 s−1 were re-
•
ported for the rate constants for the O2 addition to CF3
The products formed in the gas-phase oxidation of
C3F6, initiated by CF3OF [32], are COF2, CF3C(O)F,
CF3OC(O)F, and a perfluorodiether, CF3OCF2OCF2
C(O)F.
[43] and CF3CFHO• [44], respectively. The rate con-
stants for the reactions of haloalkyl radicals with O2
exceed those of fluorine atom abstraction from CF3OF
by several orders of magnitude [23,24,45,46]. Then the
contribution of reaction (25) to the disappearance of
CF3OF and to the generation of CF3O• radicals must
be very small, notwithstanding these few radicals con-
tributing to initiate new chains. The CF3OF is con-
sumed principally by reaction (1). The amounts of
CF3OF consumed correspond closely to the calculated
ones using k1, the rate constant obtained previously for
the gas-phase addition of CF3OF to the double bond of
C4F8 in absence of O2 [22].
At the temperature used in this work CF3OF and
perfluorobutene-2 are stable molecules and the reac-
tion is homogeneous; thus it can be concluded that
the primary process of this reaction is a bimolecular
reaction between CF3OF and C4F8, in which the ho-
molytic cleavage of a rather weak bond O F of CF3OF
occurs, generating F(C4F8)• and CF3O• radicals. The
CF3O• radicals add rapidly to the double bond of per-
fluorobutene, forming CF3O(C4F8)• radicals. The lack
offormationofCF3OOCF3 confirmswhatwasreported
earlier, that the addition of CF3O• to the double bond of
alkenes is considerably faster than any other reaction
The lack of formation of perfluorobutene epox-
O
ꢀ
❝
ide CF3CF CFCF3 indicates that the perfluoroalkyl

THERMAL GAS-PHASE OXIDATION OF PERFLUOROBUTENE-2
539
peroxyradicalsdonotaddtothedoublebondatthecon-
ditions used in our work, disappearing preferentially by
their self-reaction to give perfluoroalkoxy radicals. The
available data for the addition of methyl peroxy radicals
to alkenes [47], giving epoxy compounds as products,
indicate that the rate constants for these additions are
several orders of magnitude smaller than those for the
self-reactions of peroxy radicals.
It appears that the amounts of CF3CF2OC(O)F
formed through reaction sequence (18)–(24) were too
small to be detected, suggesting that in the temperature
range used in this work, the thermal decomposition of
•
CF3CF2O• radical to give COF2 and CF3 is more rapid
than its addition to the double bond of perfluorobutene.
It was found in the studies of oxidation of CF3CF2H
(HFC-125) [54–56], in the study of the self-reaction
•
The perfuoroalkoxy radicals decompose via the
C C bond cleavage [48,49], this decomposition be-
ing the atmospheric fate of these radicals. The perflu-
oroalkoxy radicals containing ether groups and a rel-
atively weak ꢀ-bond C C in the OCR2 C(O)FCF3
group, where R = F or CF3, decompose principally
by the ꢀ-scission, giving shorter perfluoroalkyl rad-
icals and CF3C(O)F as products [2,3]. Rates of ꢀ-
scission are determined by the stability of the rad-
icals produced, depending on the nature and degree
of substitution at the radical center [3,50]. It appears
that the oxygen atoms attached to the adjacent car-
bon atoms in the C C bond, as in the OC C(O)
group of the CF3OCF(CF3)C(O)•FCF3 radical, en-
hance the weakening of this bond, favoring ꢀ-scission
(reaction (13)). In the study of the thermal gas-phase
oxidation of C3F6, initiated by CF3OF [32], it was
found that the radical CF3OCF2C(O)•FCF3, analogous
to CF3OCF(CF3)C(O)•FCF3, also decomposes prin-
cipally by the ꢀ-scission of the C C bond of the
OC C(O) group, giving R• and CF3C(O)F as prod-
ucts. The decomposition rates of fluorinated alkoxy
radicals are very sensitive to change in molecular con-
figuration. The CF3CF2CFHO• radical decomposes 17
times more rapidly than the CF3CFHO• radical [51].
CF3CF2CFHO• can be viewed as CF3CFHO• with one
fluorine atom substituted by a CF3 group.
of CF3CF2O2 radicals [57], and that of the reaction of
•
CF3CF2O2 with NO [58], that the CF3CF2O• radical
was formed as a short-lived intermediate, whose fate
was decomposition via C C scission to give COF2 and
•
CF3 . The rate constant value of 5.3 × 106 s−1 was cal-
culated for the decomposition of CF3CF2O• at 300 K
[59]. It was also determined from the pulse radiolysis
experiments at 296 K that CF3CF2CF2O• radicals de-
compose with a rate greater than 1.5 × 105 s−1 [48].
The decomposition rates of perfluoroalkoxy radicals
increase with temperature.
The rate constant k25 was considered to be indepen-
dent of the perfluoroalkyl radical structure. This suppo-
sition is supported by the room temperature values of
the rate constants for the abstraction of a fluorine atom
•
•
from CF3OF by CF3 [45] and C2H5 [46], reported to
be 2 × 105 and >6 × 104 dm3 mol−1 s−1
.
Reaction(26)wasproposedaschaintermination. As
no products of this reaction were detected, it indicates
that in comparison with other products, only very small
amounts of these compounds were formed.
The CF3O• radicals are generated by reaction (1),
by the ꢁ C C scission of perfluoroalkoxy radicals in
presence of O2, and by the abstraction of the fluorine
atom from CF3OF by R•.
Reaction (14) was proposed as an alternative
route of decomposition of the radical CF3OCF
CONCLUSION
•
(CF3)C(O)•FCF3 by the ꢁ-scission, giving CF3 and
CF3OCF(CF3)C(O)F, but it is not expected to occur to
a great extent. No absorption bands that could be as-
cribed to CF3OCF(CF3)C(O)F were observed in the IR
spectra of the reaction products.
The cleavage of the O C bond in the C O C ether
group is not expected, as the activation energy for the
scission of O C bond in perfluorinated ethers is rather
high, near 100 kcal mol−1 [52]. Then the principal
route of decomposition of radicals CF3OC(O)•FCF3
(reaction (17)) and CF3CF2OC(O)•FCF3 (reaction
Our work demonstrates the possibility of activating the
complex mechanism of oxidation of fluorinated olefins
with molecular oxygen by employing such a simple
chemical initiator as CF3OF, thus offering an alterna-
tive to UV radiation. It also shows that this singular
oxidation mechanism has common characteristics as
compared with those for the oxidation of fluoroolefins
of different monomer structures. The efficiency of the
initiator to produce perfluoroalkylpolyethers with con-
trolled molecular mass depends on the competition be-
tween the rate of C C bond scission of the perflu-
oroalkoxy radicals formed and their rate of addition
to the olefin monomer. The lack of appreciable quan-
tity of telomeric products containing ether group in the
CF3OF + C4F8-2 + O2 system studied indicates that
within the temperature and pressure range used in our
(24)) is the ꢁ-scission, giving CF3OC(O)F and
•
CF3CF2OC(O)F, respectively, and radicals CF3 , which
reform the chain carriers CF3O•.
The extrusion of the fluorine atom adjacent to the
oxygen atom is unlikely because these processes are
highly endothermic [53].

´
ROMANO, DELLA VEDOVA, AND CZARNOWSKI
540
work, the principal route of disappearance of perflu-
oroalkoxy radicals CF3CF2O• and CF3OC(O)•FCF3 is
their decomposition giving CF3 + COF2 and CF3 +
CF3OC(O)F, respectively.
21. Minkwitz, R.; Reinemann, S. Z Anor Allg Chem 1998,
624, 1695–1698.
22. dos Santos Afonso, M.; Schumacher, H. J. Z Phys Chem
N F 1988, 158, 15–23.
23. Czarnowski, Z.; Czarnowski, J. J Chem Soc, Faraday
Trans 1993, 89, 451–455.
The authors thank Mr. Z. Czarnowski for helpful comments.
C.O.D.V. and R.M.R. appreciate the DAAD-Fundacio´n An-
torchas and Alexander von Humboldt Stiftung-Fundacio´n
Antorchas Awards to the German–Argentine cooperation
and the British Council-Fundacio´n Antorchas award to the
British–Argentinean cooperation.
24. Czarnowski, J. Int J Chem Kinet 1996, 28, 489–495.
25. Johri, K. K.; DesMarteau, D. D. J Org Chem 1983, 48,
242–250.
26. Sekiya, A.; Ueda, K. Chem Lett 1990, 609–612.
27. dos Santos Afonso, M.; Schumacher, H. J. Int J Chem
Kinet 1984, 16, 103–115.
28. dos Santos Afonso, M.; Czarnowski, J. Z Phys Chem N
F 1988, 158, 25–34.
29. Czarnowski, J. J Chem Soc, Faraday Trans 2 1989, 85,
1425–1437.
BIBLIOGRAPHY
1. Sianesi, D. ACS Polym Prepr 1971, 12, 411–419.
2. Sianesi, D.; Pasetti, A.; Fontanelli, R.; Bernardi, G.
C.; Caporiccio, G. La Chimica e l’Industria 1973, 55,
208–221.
3. Faucitano, A.; Buttafava, A.; Comincioli, V.;
Marchionni, G.; De Pascuale, R. J Phys Org Chem
1991, 4, 293–300.
4. Sianesi, D.; Marchionni, G.; De Pasquale, R. J. In Or-
ganic Fluorine Chemistry: Principles and Commercial
Applications; Banks, R. E.; Smart, B. E.; Tatlow, J. C.
(Eds.); Plenum: New York, 1994; pp. 431–461.
5. Bunyard, W. C.; Romack, T. J.; DeSimone, J. M. Macro-
molecules 1999, 32, 8224–8226.
30. Czarnowski, J. Z Phys Chem N F 1995, 191, 103–118.
31. Czarnowski, J. Z Phys Chem N F 1998, 203, 183–
197.
32. dos Santos Afonso, M.; Romano, R. M.; Della Vedova,
C. O.; Czarnowski, J. Phys Chem Chem Phys 2000, 2,
1393–1399.
33. Navarrini, W.; Russo, A.; Tortelli, V. J Org Chem 1995,
60, 6441–6443.
34. Conti, F.; Corvaja, C.; Cremonese, F.; Navarrini, W.;
Tortelli, V. J Chem Soc, Faraday Trans 1995, 91, 3813–
3820.
35. Li, Y. S.; Wilson, K. R. J Chem Phys 1990, 93, 8821–
8827.
6. Johnston, T.; Heicklen, J.; Stuckey, W. Can J Chem 1968,
46, 332–334.
7. Malavasi, M.; Sianesi, D. J Fluorine Chem 1999, 95,
19–25.
8. Sianesi, D.; Guarda, P. A.; Marchionni, G. Macro-
molecules 1999, 32, 7768–7780.
9. Rozen, S. Chem Rev 1996, 96, 1717–1736.
10. Czarnowski, J.; Castellano, E.; Schumacher, H. J. Z Phys
Chem N F 1969, 65, 225–237.
36. Czarnowski, J. Int J Chem Kinet 1986, 18, 1323–1331.
37. Czarnowski, J. J Chem Soc, Faraday Trans 2 1987, 83,
579–586.
38. Chen, J.; Zhu, T.; Young, V.; Niki, H. J Phys Chem 1993,
97, 7174–7177.
39. Kelly, C.; Treacy, J.; Sidebottom, H. W.; Nielsen, O. J.
Chem Phys Lett 1993, 207, 498–503.
40. Kelly, C.; Sidebottom, H. W.; Treacy, J.; Nielsen, O. J.
Chem Phys Lett 1994, 218, 29–33.
11. Czarnowski, J.; Schumacher, H. J. Z Phys Chem N F
1970, 73, 68–76.
41. Niki, H.; Chen, J.; Young, V. Res Chem Intermed 1994,
20, 277–301.
12. Kennedy, R. C.; Levy, J. B. J Phys Chem 1972, 76, 3480–
3488.
42. Knyazew, V.; Slagle, I. R. J Phys Chem A 1998, 102,
1770–1778.
13. Hohorts, F. A.; Shreeve, J. M. Inorg Synth 1967, 11,
143–146.
43. Breheny, C.; Hancock, G.; Morrell, C. Z Phys Chem N
F 2001, 215, 305–317.
14. Nielsen, A. H.; Burke, T. G.; Woltz, P. J. H.; Jones, E.
A. J Chem Phys 1952, 20, 596–604.
44. Maricq, M. M.; Szente, J. J.; Kaiser, E. W. Chem Phys
Lett 1992, 197, 149–156.
15. Loos, K. R.; Lord, R. C. Spectrochim Acta 1965, 21,
119–125.
45. Shoikhet, A. A.; Teitelboim, M. A.; Vedeneev, V. I.
Kinet Katal 1983, 24, 225–227.
16. Varetti, E.; Aymonino, P. J. J Mol Struct 1968, 1, 39–54.
17. Cutin, E. H.; Della Vedova, C. O.; Aymonino, P. J. An
Asoc Quim Argent 1985, 73, 171–180.
18. Hermann, A.; Trautner, F.; Gholivand, K.; von Ahsen,
S.; Varetti, E. L.; Della Vedova, C. O.; Willner, H.;
Oberhammer, H. Inorg Chem 2001, 40, 3979–3985.
19. Arvia, A. J.; Aymonino, P. J. Spectrochim Acta 1962,
18, 1299–1309.
46. Wang, N.; Rowland, F. S. J Chem Phys 1985, 89, 5154–
5155.
47. Osborne, D. A.; Waddington, D. J. J Chem Soc, Perkin
Trans II 1980, 925–930.
48. Giessing, A. M. B.; Feilberg, A.; Mogelberg, T. E.;
Sehested, J.; Bilde, M.; Wallington, T. J.; Nielsen, O. J.
J Phys Chem 1996, 100, 6572–6579.
49. Mogelberg, T. E.; Sehested, J.; Bilde, M.; Wallington,
T. J.; Nielsen, O. J. J Phys Chem 1996, 100, 8882–8889.
50. Zhang, X. J Org Chem 1998, 63, 3590–3594.
20. Durig, J. R.; Wertz, D. W. J Mol Spectrosc 1968, 25,
467–478.

THERMAL GAS-PHASE OXIDATION OF PERFLUOROBUTENE-2
541
51. Mogelberg, T. E.; Feilberg, A.; Giessing, A. M. B.;
179–199.
Sehested, J.; Bilde, M.; Wallington, T. J.; Nielsen, O.
J. J Phys Chem 1995, 99, 17386–17393.
56. Hasson, A. S.; Moore, C. M.; Smith, I. W. M. J Chem
Soc, Faraday Trans 1997, 93, 2693–2699.
52. Waltman, R. J. J Fluorine Chem 1998, 90, 9–16.
53. Rayez, J. C.; Rayez, M. T.; Halvick, P.; Duguay, B.;
Danneberg, J. J. Chem Phys 1987, 118, 265–272.
54. Edney, E. O.; Driscoll, D. J. Int J Chem Kinet 1992, 24,
1067–1081.
57. Sehested, J.; Ellermann, T.; Nielsen, O. J.; Wallington,
T. J.; Hurley, M. D. Int J Chem Kinet 1993, 25, 701–717.
58. Mogelberg, T. E.; Sehested, J.; Wallington, T. J.; Nielsen,
O. J. Int J Chem Kinet 1997, 29, 209–217.
59. Sommitz, H.; Zellner, R. Phys Chem Chem Phys 2001,
3, 2352–2364.
55. Tuazon, E. C.; Atkinson, R. J Atmos Chem 1993, 17,