Synthesis of pentafluorosulfanyl trifluorovinyl ether and its facile rearrangement to difluoro(pentafluorosulfanyl)acetyl fluoride

Article abstract of DOI:10.1002/anie.200702425

A radical departure: A novel fluorinated vinyl ether 1 was synthesized and was found to undergo an unexpected facile rearrangement to give compound 2, which was confirmed to be a radical process by EPR spectroscopy. (Chemical Equation Presented).

Full text of DOI:10.1002/anie.200702425

Communications
DOI: 10.1002/anie.200702425
Fluorine Chemistry
Synthesis of Pentafluorosulfanyl Trifluorovinyl Ether and Its Facile
Rearrangement to Difluoro(pentafluorosulfanyl)acetyl Fluoride**
Libin Du, Bevan Elliott, Luis Echegoyen, and Darryl D. DesMarteau*
Amorphous fluoropolymers, such as teflon AF, hyflon AD,
and cytop, have attractive properties as optical materials for
use in a variety of applications.[1] However, their application
as soft polymer pellicles in 157-nm microlithography failed
because of photochemical degradation.[2,3] The degradation
was thought to be a result of b scission of the cyclic ether
structures in the polymer chains. In exploratory research on
potential alternative amorphous fluoropolymers, a polymer
=
based on F5SOCF CF2 (1) seemed promisingfor two main
reasons: 1) the bulkiness of the SF5 group might effectively
prevent crystallinity in the co-polymer with tetrafluoroethy-
lene (TFE), and 2) its linear structure might be less prone to
b scission under UV irradiation. Other perceived advantages
included high fluorine content and expected high thermal
stability. The polymerization of SF5-substituted vinyl ethers
had not been previously reported.
Scheme 1. Synthesis of pentafluorosulfanyl trifluorovinyl ether and its
rearrangement to difluoro(pentafluorosulfanyl)acetyl fluoride.
Four partially fluorinated SF5-substituted vinyl ethers
were previously prepared through dehydrohalogenation using
KOH.[4] However, an attempt to dehalogenate F5SOFClC
À
CF2Cl (2) to prepare 1 did not succeed.[5] Herein, we report
the first synthesis of 1 with reasonable yields through the
dechlorination of 2 usinghexaethyl phosphorus triamide
P(NEt2)3. Quite unexpectedly, 1 undergoes a facile rearrange-
ment at 228C to give difluoro(pentafluorosulfanyl)acetyl
=
fluoride F5SF2CC( O)F (3). Two intermediate free radicals
in the rearrangement were readily identified by EPR and a
plausible mechanism is proposed.
The preparation of 1 is summarized in Scheme 1. The
=
addition reaction of F5SOF to FClC CFCl led to 2 in high
yields (caution: this reaction can be explosive, and a small
scale reaction and careful control of the temperature are
essential). Repeated attempts to dehalogenate 2 usingzinc
dust failed, despite the fact that this method has been widely
used to prepare many other related compounds.[6] The
dehalogenation was subsequently carried out using P(NEt2)3
which had been used to prepare some unusual halogenated
olefins.[7] The 19F NMR spectrum of 1 with 95% purity is
shown in Figure 1. The only major impurity seen in the
Figure 1. 19F NMR spectrum of pentafluorosulfanyl trifluorovinyl ether,
recorded on a Bruker AC 200 spectrometer. The sample (in CDCl3 as
solvent and CFCl3 as reference) was sealed in a 4-mm NMR tube,
which was contained in a regular 5-mm NMR tube for the measure-
ment.
=
spectrum is F2C CFCl, which was assumed to be generated by
a loss of the F5SO unit rather than the Cl atom of 2. The
[*] Dr. L. Du, B. Elliott, Prof. L. Echegoyen, Prof. D. D. DesMarteau
Department of Chemistry
spectrum shows the strongly overlapped AB4 pattern for the
2
SF5 group and the JAB couplingconstant was not readily
Clemson University
Clemson, South Carolina 29634 (USA)
Fax: (+1)864-656-2545
determined. Often the AB4 patterns of SF5 groups are well
resolved.[4,5,8]
The vinyl ether 1 was quickly found to be unstable at 228C
and plans to use it as a co-monomer in polymerization with
TFE had to be abandoned. Its facile rearrangement to 3 at
room temperature was evident in both NMR and IR spectra.
A sample of 1 (3 mmol) sealed in an approximately 40-mL
E-mail: fluorin@clemson.edu
[**] We gratefully acknowledge the financial support of this study by
Solvay Solexis.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
6626
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 6626 –6628

Angewandte
Chemie
sample tube gave 2.7 mmol of pure 3 (90%) after 12 h at
228C. A sample of neat 1 (0.5 mmol) for EPR was completely
decomposed after 2 h at 228C.
A proposed radical mechanism for the rearrangement is
shown in Scheme 2. The process is comprised of a homolytic
Scheme 2. Proposed radical mechanism of the pentafluorosulfanyl
trifluorovinyl ether rearrangement.
À
cleavage of the S O bond of 1, followed by a rearrangement
Figure 2. EPR spectra of the rearrangement of 1; top: the experimental
of an oxygen-centered radical 5 to a more-stabilized carbon-
centered radical 6, and then a recombination of the radicals 4
and 6. An intermolecular mechanism was proposed to
account for the related thermal rearrangement (at 2608C)
spectrum recorded on a Bruker EMX spectrometer; bottom: simulated
spectrum (calculated by Bruker SimFonia). Neat liquid samples of 1
were placed in a 3-mm I.D. tubes, degassed, and sealed. X-band first
derivative absorption was then recorded at 228C. Signal intensity
decreased after 2 hours.
=
=
of F3CO(F)C CF2 into F3CF2CC( O)F, based on the analysis
of the side products.[9,10] In 1, the process is more likely
intramolecular in nature based on the high yields and the
EPR study. Comparison of the outwardly similar fluorinated
spectroscopic data,[4] although a possible mechanism of the
rearrangement was not proposed. However, we have pre-
[11]
=
=
vinyl ethers, F3COCF CF2 (very stable) and H3COFC CF2
(explosive) indicates the disparate properties of this com-
pound class.
=
pared F5SOHC CFCl (cis and trans) and the spectroscopic
data are totally different from those reported;[4] the pre-
The rearrangement shown in Scheme 2 is supported by
EPR. The experimental spectrum (top, Figure 2), was closely
matched by the simulated spectrum (bottom, Figure 2) and is
consistent with the proposed mechanism. Five spectroscopic
lines havingadditional second-order splittinsg were fitted
usingfour equivalent F nuclei with a hyperfine splittingof
144.0 G and a g value of 2.00290 (black in the simulated
spectrum). This spectrum is identical to that from an earlier
study of the SF5 radical, in which electron spin on four
equivalent equatorial F nuclei was observed with no resolv-
able couplingto the axial F nucleus. [12–14]
A second hyperfine pattern is also visible in the spectrum.
This pattern, a triplet of doublets with 1:2:1 relative line
intensities and a g value of 2.00318, can be simulated by
electron spin on two equivalent F nuclei (58.2 G) and a single
F nucleus (5.0 G) (blue in the simulated spectrum). This
spectrum, with hyperfine values well within the typical range
of known primary and secondary F hyperfine splittings,
corresponds well with the radical 6. This radical has not been
previously identified. No evidence for radical 5 or other
radical species was observed in the EPR, probably because of
its rapid intramolecular electron transfer and conversion into
6.
viously reported 19F NMR and IR spectroscopic data are
=
consistent with that expected for F5SCFHC( O)Cl and not
the alkene. The reported rearrangement was probably similar
to the rearrangement of 1. Amonga series of other SF
-
5
substituted vinyl ethers we have prepared, several also
undergo the same type of rearrangement, but with different
reaction rates.[15]
The EPR study suggests that the intermediate radicals 4
and 6 have reasonably longlifetimes, a property which can
potentially be utilized to initiate radical polymerizations.
Preliminary experiments demonstrated that 1 readily initiated
polymerization of TFE or vinylidene fluoride (VDF).
In summary, we have reported the first successful syn-
thesis of 1 usingthe atypical but effective dehalogenation
reagent P(NEt2)3. An EPR study identified two long-lived
radical intermediates in the facile but unexpected rearrange-
ment of 1. We believe the detection of both radicals in a
homolysis reaction is quite rare, as is their recombination to
form a rearranged product in high yield.
Experimental Section
=
=
2: FClC CFCl (20.2 mmol, 12% F2C CCl2) was vacuum transferred
into a stainless-steel bomb (150 mL) cooled to À1968C. Then SF5OF
(6.4 mmol) was condensed on the upper walls of the reactor, which
was allowed to warm slowly in a dewar from À1968C to 228C. The
=
A related interestingrearrangement of F 5SOClC CFH
(cis/trans) has been reported.[4] The rearranged product was
=
believed to be F5SOHC CFCl based on the NMR and IR
Angew. Chem. Int. Ed. 2007, 46, 6626 –6628
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 6627

Communications
reactor was then cooled to À1968C and further SF5OF (6.8 mmol)
[1] a) P. R. Resnick, W. H. Buck in Modern Fluoropolymers (Ed.: J.
Scheirs), Wiley, New York, 1997, pp. 397 – 419; b) N, Sugiyama
in Modern Fluoropolymers (Ed.: J. Scheirs), Wiley, New York,
1997, pp. 541 – 556; c) V. Arcella, A. Ghielmi, G. Tommasi, Ann.
N. Y. Acad. Sci. 2003, 984, 226 – 244.
[2] R. H. French, R. C. Wheland, W. Qiu, M. F. Lemon, E. Zhang, J.
Gordon, V. A. Petrov, V. F. Cherstkov, N. I. Delaygina, J.
Fluorine Chem. 2003, 122, 63 – 80.
[3] a) D. D. DesMarteau, N. Lu, P. Zimmerman, D. Miller, J.
Photopolym. Sci. Technol. 2004, 17, 651 – 654; b) I. Blakey,
G. A. George, D. J. T. Hill, H. Liu, F. Rasoul, A. K. Whittaker, P.
Zimmerman, Macromolecules 2005, 38, 4050 – 4053.
[4] R. D. Place, S. M. Williamson, J. Am. Chem. Soc. 1968, 90, 2550 –
2556.
added and the warmingrepeated. This cycle was repeated five times
to add a total of 32.2 mmol of SF5OF. Vacuum fractional condensation
through traps at À96 and À1968C, followed by reseparation of the
À968C trap through a À788C trap, gave 2 (14 mmol) contaminated
À
À
with traces of F5SOCl2C CF3 and F5SOF2C CFCl2. IR and NMR
spectroscopic data of 2 corresponded with those reported.[5]
1: Dry CH3CN (15 mL) was introduced into a predried flask
(250 mL). The startingmaterial 2 (3.2 mmol) was then condensed into
the flask through a vacuum line. The flask was immersed into an
ethanol/H2O bath at À158C. Compound 2 (1.1 mL, 3.2 mmol) was
slowly injected into the flask over 5 min. The resultingmixture
warmed to 08C over 30 min. A fractional vacuum condensation was
carried out using3 traps at À75, À88, and À1968C, respectively. The
compound 1 (2.7 mmol), contaminated with minor impurities, was
collected at À1968C. The unreacted startingmaterial 2 and CH3CN
were collected in the other 2 traps. 19F NMR (188.3 MHz, CDCl3,
258C, CFCl3): d = 61.6 (m, 5F; SF5), À133.9 (ddm, 3J(F, F) = 109, 3J(F,
F) = 74 Hz, 1F; CF2), À117.5 (ddm, 2J(F, F) = 72 Hz, 1F; CF2),
[5] C. J. Schack, R. D. Wilson, K. O. Christe, J. Fluorine Chem. 1989,
45, 283 – 291.
[6] a) W. S. Durrell, E. C. Stump, G. Westmoreland, C. D. Padgett, J.
Polym. Sci. Part A 1965, 12, 4065 – 4074; b) R. D. Chambers,
Fluorine in Organic Chemistry, CRC, Boca Raton, 2004, p. 171.
[7] V. Montanari, D. D. DesMarteau, J. Org. Chem. 1992, 57, 5018 –
5019.
À112.6 ppm (dd, 1F; CF). IR (gas phase, 6 torr): v = 1828 (s, vC C),
=
1343, 1292, 1175 (s, vCÀ ), 933, 890, 832 (s, v F), 722 (m), 607 (s), 577
À
S
F
(m), 538 cmÀ1 (w).
[8] a) D. D. DesMarteau, J. Am. Chem. Soc. 1972, 94, 8933 – 8934;
b) D. R. Eaton, W. A. Sheppard, J. Am. Chem. Soc. 1963, 85,
1310 – 1313; c) C. I. Merrill, G. H. Cady, J. Am. Chem. Soc. 1961,
83, 298 – 300.
[9] R. N. Haszeldine, A. E. Tipping, J. Chem. Soc. C 1968, 398 – 405.
[10] A. Zompatori, V. Tortelli, J. Fluorine Chem. 2004, 125, 199 – 204.
[11] a) D. C. England, C. G. Krespan, J. Org. Chem. 1968, 33, 816 –
819; b) A. W. Anderson, Chem. Eng. News 1976, 54, 5.
[12] a) J. R. Morton, K. F. Preston, Chem. Phys. Lett. 1973, 18, 98 –
101; b) J. R. Morton, K. F. Preston, J. Chem. Phys. 1973, 58,
2657 – 2659; c) J. C. Tait, J. A. Howard, Can. J. Chem. 1975, 53,
2361 – 2364.
3: The vinyl ether 1 (3.0 mmol) was vacuum transferred into a
glass tube (ca. 40 mL) at À1968C fitted with a glass-teflon valve. The
tube was then immersed in an ethanol bath at À308C and allowed to
warm slowly to 228C (over around 4 h). The tube was then held at
room temperature for 12 h. A mixture of the main product 3 along
with minor side products was obtained. GC separation yielded pure 3
(2.7 mmol, 90%). 19F NMR (188.3 MHz, CDCl3, 258C, CFCl3,) d =
2
62.1–65.3 (quintet of m, J(F, F) = 150.6 Hz, 1F; SF), 42.0–42.9 (dm,
4J(F, F) = 4.4 Hz, 4F; SF4), À91.9 (m, 3J(F, F) = 2.3, 3J(F, F) = 4.4 Hz,
2F; CF2), 21.4 ppm (septet, 1F; C(O)F). IR (gas phase, 15 torr): v =
1890 (s, vC O), 1285, 1232, 1132 (s, vCÀ ), 902, 869, 765 (s, v F), 680 (s),
=
À
S
F
614 (m), 576 (m), 523 cmÀ1 (w).
[13] a) A. R. Gregory, Chem. Phys. Lett. 1974, 28, 552 – 554; b) A. R.
Gregory, S. E. Karavelas, J. R. Morton, K. F. Preston, J. Am.
Chem. Soc. 1975, 97, 2206.R.D.
[14] A. Hasegawa, F. Williams, Chem. Phys. Lett. 1977, 45, 275 – 278.
[15] L. Du, PhD thesis, Clemson University, 2006.
Received: June 4, 2007
Published online: July 27, 2007
Keywords: EPR spectroscopy · fluorine · NMR spectroscopy ·
.
radical reactions · rearrangement
6628 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 6626 –6628