Synthesis of Trifluorovinyl Ethers Incorporating Functionalized Hydrocarbon Ether Groups: Insight into the Mechanism of Trifluorovinyl Ether Formation from Trimethylsilyl 2-Alkoxy-2,3,3,3-tetrafluoropropionates

Article abstract of DOI:10.1021/jo971319d

Novel trifluorovinyl ethers (TFVEs, ROCF=CF2), where R is an oligoether, were synthesized from the corresponding sodium alkoxide and hexafluoropropene oxide.The sodium alkoxide ring opened hexafluoropropene oxide at the more highly substituted carbon (C2) to give 2-alkoxy-2,3,3,3-tetrafluoropropionic acid ester incorporating 2 equiv of the alcohol, ROCF(CF3)CO2R.Hydrolysis of the ester and reaction of the resulting sodium 2-alkoxy-2,3,3,3-tetrafluoropropionate with chlorotrimethylsilane gave the trimethylsilyl 2-alkoxy-2,3,3,3-tetrafluoropropionate, ROCF(CF3)CO2Si(CH3)3.Gas phase vacuum thermolysis of the trimethylsilyl ester at 140-150 deg C gave the corresponding TFVEs in 55-63percent yields.Thus, 1-<2-(ethoxyethoxy)ethoxy>-1,2,2-trifluoroethene and 1-<2-(2-tert-butoxyethoxy)ethoxy>-1,2,2-trifluoroethene were synthesized from 2-(2-ethoxyethoxy)ethanol and 2-(2-tert-butoxyethoxy)ethanol, respectively.Interestingly, thermolysis of sodium or potassium 2-alkoxy-2,3,3,3-tetrafluoropropionates resulted in negligible to low yields of TFVEs.¹ For example, thermolysis of sodium 2-<2-(ethoxyethoxy)ethoxy>-2,3,3,3-tetrafluoropropionate gave a trifluoroacetate ester, 2-(2-ethoxyethoxy)ethyl trifluoroacetate.Variable temperature 19F NMR spectroscopy of trimethylsilyl 2-<2-(ethoxyethoxy)ethoxy>-2,3,3,3-tetrafluoropropionate suggests that an equilibrium exists between two structural conformations of these trimethylsilyl esters: one in which there is an intramolecular "interaction" of silicon with fluorine and one in which there is no silicon-fluorine interaction.This interaction may affect the outcome of the trimethylsilyl ester thermolysis.

Full text of DOI:10.1021/jo971319d

7844
J . Org. Chem. 1997, 62, 7844-7849
Syn th esis of Tr iflu or ovin yl Eth er s In cor p or a tin g F u n ction a lized
Hyd r oca r bon Eth er Gr ou p s: In sigh t in to th e Mech a n ism of
Tr iflu or ovin yl Eth er F or m a tion fr om Tr im eth ylsilyl
2-Alk oxy-2,3,3,3-tetr a flu or op r op ion a tes
Robert D. Lousenberg† and Molly S. Shoichet*,†,†
Departments of Chemistry and of Chemical Engineering and Applied Chemistry, University of Toronto,
200 College Street, Toronto, Ontario M5S 3E5, Canada
Received J uly 18, 1997X
Novel trifluorovinyl ethers (TFVEs, ROCFdCF2), where R is an oligoether, were synthesized from
the corresponding sodium alkoxide and hexafluoropropene oxide. The sodium alkoxide ring opened
hexafluoropropene oxide at the more highly substituted carbon (C2) to give the 2-alkoxy-2,3,3,3-
tetrafluoropropionic acid ester incorporating 2 equiv of the alcohol, ROCF(CF3)CO2R. Hydrolysis
of the ester and reaction of the resulting sodium 2-alkoxy-2,3,3,3-tetrafluoropropionate with
chlorotrimethylsilane gave the trimethylsilyl 2-alkoxy-2,3,3,3-tetrafluoropropionate, ROCF(CF3)CO2-
Si(CH3)3. Gas phase vacuum thermolysis of the trimethylsilyl ester at 140-150 °C gave the
corresponding TFVEs in 55-63% yields. Thus, 1-[2-(ethoxyethoxy)ethoxy]-1,2,2-trifluoroethene and
1-[2-(2-tert-butoxyethoxy)ethoxy]-1,2,2-trifluoroethene were synthesized from 2-(2-ethoxyethoxy)-
ethanol and 2-(2-tert-butoxyethoxy)ethanol, respectively. Interestingly, thermolysis of sodium or
potassium 2-alkoxy-2,3,3,3-tetrafluoropropionates resulted in negligible to low yields of TFVEs.1
For example, thermolysis of sodium 2-[2-(ethoxyethoxy)ethoxy]-2,3,3,3-tetrafluoropropionate gave
a trifluoroacetate ester, 2-(2-ethoxyethoxy)ethyl trifluoroacetate. Variable temperature 19F NMR
spectroscopy of trimethylsilyl 2-[2-(ethoxyethoxy)ethoxy]-2,3,3,3-tetrafluoropropionate suggests that
an equilibrium exists between two structural conformations of these trimethylsilyl esters: one in
which there is an intramolecular “interaction” of silicon with fluorine and one in which there is no
silicon-fluorine interaction. This interaction may affect the outcome of the trimethylsilyl ester
thermolysis.
In tr od u ction
in the paint formulation industry.4 For biomaterial
applications, fluoropolymers have been found to be
relatively biologically inert yet still adsorb proteins. A
TFVE having an oligo-ethylene oxide pendant group was
synthesized to render the polymer less protein adsorp-
tive.5
TFVEs have been previously synthesized by two prin-
cipal synthetic routes which do not involve the use of
elemental fluorine, chlorine, or hydrogen fluoride. For
example, 1-methoxy-1,1,2-trifluoroethene was prepared6
by the reaction of sodium methoxide with tetrafluoro-
ethylene. This reaction was expanded to include ethox-
ide-, isopropoxide-, and tert-butoxide-substituted TFVEs.7
While straightforward in approach, this method required
high-pressure reaction equipment to achieve high tet-
rafluoroethylene pressures and long reaction times (and
in one instance an explosion was reported).7
Fluoropolymers, such as poly(tetrafluoroethylene) or
poly(tetrafluoroethylene-co-hexafluoropropylene), are dif-
ficult to process, insoluble in common organic solvents,
and chemically inert, requiring highly reactive species
for surface modification.2 Perfluorinated ether groups on
trifluorovinyl ethers (TFVEs) have been shown to im-
prove the processability of the resulting polymer.3 In-
corporating a hydrocarbon ether group into the fluo-
romonomer will likely further improve the processability
of the resulting polymers; however, no one has yet
synthesized (or polymerized) the hydrocarbon TFVEs
described herein.
Two new TFVEs were synthesized for copolymerization
with tetrafluoroethylene, promising polymers with im-
proved processability and solubility in common organic
solvents over existing fluoropolymers. To facilitate fur-
ther chemical modification, one of the TFVEs was de-
signed to have a terminal hydroxyl group (synthesized
protected as the tert-butylalkoxy). Unlike the existing
chemically inert fluoropolymers, the hydroxyl TFVE
provides a site for covalent bonding of numerous func-
tional groups. For example, an acrylate group may be
covalently attached to the hydroxyl group for applications
An alternate synthesis of TFVEs (i.e., (perfluoro-
alkoxy)-1,1,2-trifluoroethene) involved the reaction of a
perfluoro acid fluoride8,9 or perfluoro ketone8,10 with
hexafluoropropene oxide (HFPO) in the presence of an
alkali metal fluoride (e.g., potassium or cesium fluoride)
to yield the 2-(perfluoroalkoxy)-2,3,3,3-tetrafluoropropio-
nyl fluoride intermediate. Thermolysis of either this
(4) Reisch, M. S. Chem. Eng. News 1996, Oct. 14, 44-67.
(5) Prime, K. L.; Whitesides, G. M. J . Am. Chem. Soc. 1993, 115,
10714-21.
† Department of Chemistry.
† Department of Chemical Engineering and Applied Chemistry.
X Abstract published in Advance ACS Abstracts, October 1, 1997.
(1) Pellerite, M. J . J . Fluor. Chem. 1990, 49, 43-6.
(2) Shoichet, M. S.; McCarthy, T. J . Macromolecules 1991, 24, 982-
6.
(6) Dixon, S. U.S. Patent 2917548, 1959.
(7) Okuhara, K.; Baba, H.; Kojima, R. Bull. Chem. Soc. J pn. 1962,
35, 532-5.
(8) Fritz, C. G.; Moore, E. P., J r.; Selman, S. U.S. Patent 3114778,
1963.
(3) Feiring, A. E. In Organofluorine Chemistry: Principles and
Commercial Applications; (Banks, R. E., Tatlow, J . C., Smart, B. E.,
Eds.; Plenum Press: New York, 1994, Chapter 15.
(9) Moore, E. P., J r.; Milian, A. S., J r.; Eleuterio, H. S. U.S. Patent
3250808, 1966.
(10) Selman, S. U.S. Patent 3274239, 1966.
S0022-3263(97)01319-4 CCC: $14.00 © 1997 American Chemical Society

TFVEs with Functionalized Hydrocarbon Ether Groups
J . Org. Chem., Vol. 62, No. 22, 1997 7845
Ta ble 1. Yield s (%) for th e Ser ies of Rea ction s of a
Sod iu m Alk oxid e w ith HF P O To P r od u ce th e Desir ed
TF VEs 8a ,b
Sch em e 1
a
Reaction started at -78 °C.
intermediate in the vapor phase over a metal oxide or
the alkali metal salt of the corresponding carboxylic acid
yielded the TFVE.8,11 Thermolysis of the acid fluoride
intermediate was a “fair yielding step at best”;12 however,
the yield was improved by using 3-chloroperfluoropropene
oxide, instead of HFPO, producing a 2-(perfluoroalkoxy)-
3-chloro-2,3,3-trifluoropropionyl fluoride intermediate.13
The latter was reported to react with sodium carbonate
at temperatures between ambient and 80 °C to form a
TFVE in very high yields. 3-Chloroperfluoropropene
oxide is not commercially available.
Herein we describe the first reported syntheses of
alkoxy TFVEs by thermolysis of trimethylsilyl esters of
2-alkoxy-2,3,3,3-tetrafluoropropionates. Previous at-
tempts to synthesize hydrocarbon-substituted TFVEs by
thermolysis of 2-alkoxy-2,3,3,3-tetrafluoropropionate salts
gave unanticipated chemistry with negligible to low
yields of TFVE depending on the hydrocarbon substituent
and counterion.1 To date, fluoro/perfluorinated TFVEs
have been synthesized in high yields by thermolysis of
trimethylsilyl 2-fluoroalkoxy-2,3,3,3-tetrafluoropropi-
onates in the presence of potassium fluoride.12 Substi-
tuted acid fluorides have also been successfully converted
to the TFVEs;12,14,15 however, unlike the TFVEs described
herein, published examples were prepared from highly
fluorinated or perfluorinated acid fluorides or ketones.
Fluorine or trifluoromethyl substitution appears to sta-
bilize a fluoroalkoxide intermediate which then reacts
with HFPO.16
are significantly less reactive.18 2-(2-tert-Butoxyeth-
oxy)ethanol,1b, was first prepared by reacting diethylene
glycol with 1 equiv of isobutene in the presence of
Amberlyst 15 resin in methylene chloride.19 The mono-
ether was obtained in 70% yield (96% purity by gas
chromatography, GC) following vacuum distillation. A
negligible amount of the bis(2-tert-butoxyethyl) ether was
formed.
The reaction of the sodium alkoxide 2 with 0.59-0.75
equiv of HFPO in anhydrous ethylene glycol dimethyl
ether (DME) was exothermic, resulting in the formation
of an ester, 5, and an alkyl perfluoropropionate, 3. 5
contains two equivalents of the alcohol per equivalent of
HFPO (cf. Scheme 1). Ring opening of HFPO at the more
highly substituted carbon (C2) is favored by the electron-
withdrawing trifluoromethyl (CF3) group.18 The mech-
anism likely involves formation of an acid fluoride
intermediate, 4, which reacts with a second equivalent
of the alkoxide 2 to form 5.
Two possible reaction pathways lead to the formation
of the alkyl perfluoropropionate 3: (1) the alkoxide 2 may
attack at the least hindered carbon (C1) of HFPO with
elimination of a fluoride ion followed by rearrangement
to give 3, or more likely, (2) a fluoride ion may react with
HFPO at C2 to give the perfluoropropionyl fluoride which
reacts with the alkoxide to give 3. While HFPO has been
shown to be stable to fluoride ion attack in tert-butyl
alcohol, potassium tert-butoxide has been shown to react
with HFPO in tert-butyl alcohol at 20 °C to form tert-
butyl perfluoropropionate exclusively.20 This suggests
that nucleophilic attack occurs at C1 due to steric
hindrance at C2. However, HFPO has also been shown
to react with cesium fluoride to form perfluoropropionyl
fluoride in ethers.21 In the experiments described herein,
the relative amount of 5 vs 3 produced was controlled by
temperature. For example, a reaction started at -78 °C
yielded 40 mol % (or 22% yield) of 3a , whereas one started
at 0 °C yielded 18 mol % (or 12% yield) of 3a . At higher
temperatures, formation of 5 and thus alkoxide attack
at C2 were favored.
Resu lts
The overall synthetic route and reaction yields are
summarized in the Scheme 1 and Table 1, respectively.
Two equivalents of the sodium alkoxide react with HFPO
to yield the ester 5, which is converted to the sodium
carboxylate salt 6 and then to the trimethylsilyl ester17
7 prior to thermolysis to the desired TFVE 8. Conversion
of the alkoxy ester to the trimethylsilyl ester is critical
to the success of this synthesis as will be described.
The sodium alkoxide 2 was prepared (but not isolated)
by reaction of the appropriate alcohol 1 with sodium
hydride. The sodium alkoxide was required for reaction
with HFPO because, unlike lower alcohols (i.e., methanol,
ethanol) which react readily with HFPO, higher alcohols
The ester 5 was hydrolyzed to the sodium salt 6 (but
not isolated) in aqueous tetrahydrofuran (THF) with
approximately 1.1-1.2 equiv of sodium hydroxide. THF
and water (and the majority of byproduct alcohol 1) were
removed from the flask. The sodium carboxylate salt,
(11) Anderson, D. G.; Selman, S. U.S. Patent 3326984, 1967.
(12) Farnham, W. B. U.S. Patent 5391796, 1995.
(13) Ezzell, B. R.; Carl, W. P.; Mod, W. A. U.S. Patent 4515989, 1985.
(14) Pattison, D. B. U.S. Patent 3311658, 1967.
(15) Ezzell, B. R.; Carl, W. P.; Mod, W. A. U.S. Patent 4337211, 1982.
(16) Millauer, H.; Schwertfeger, W.; Siegemund, G. Angew. Chem.,
Int. Ed. Engl. 1985, 24, 161-79.
(18) Sianesi, D.; Pasetti, A.; Tarli, F. J . Org. Chem. 1966, 23, 2312-
6.
(19) Imaizumi, M.; Yasuda, M. German U.S. Patent 2450667, 1974;
Chem. Abstr. 83, P178305e.
(20) Bekker, R. A.; Asratyan, G. V.; Dyatkin, B. L.; Knunyants, I.
L. Tetrahedron 1974, 39, 3539-44.
(21) Moore, E. P., J r. U.S. Patent 3322826, 1967.
(17) Yur’ev, Y. K.; Belyakova, Z. V. Rus. Chem. Rev. 1960, 29 (7),
383-94.

7846 J . Org. Chem., Vol. 62, No. 22, 1997
Lousenberg and Shoichet
Sch em e 3
Sch em e 2
Sch em e 4
dissolved in anhydrous diethyl ether, reacted with tri-
methylsilyl chloride (TMSCl) to produce the trimethyl-
silyl ester 7. The sodium salt 6 is sensitive to acidic
conditions, while 7 is susceptible to hydrolysis. To
remove acidic species introduced from TMSCl and ensure
complete dryness, sodium hydride was added before the
addition of TMSCl to avoid conversion of 6 or 7 to the
corresponding acid.
2,3,3,3-tetrafluoropropionates decarboxylate to form an
intermediate carbanion which eliminates fluoride ion at
the â-position to form the desired olefin (Scheme 3).13 For
example, salts of 2-(fluoroalkoxy)-3-chloro-2,3,3-trifluo-
ropropionates decarboxylate with loss of chloride ion to
form the olefin at temperatures between ambient and 80
°C11 and thus require dramatically less energy (with
respect to what we observed) when a better leaving group
(i.e., chloride) replaces fluoride.
For purposes of characterization, pure sodium car-
boxylate salt 6a was obtained in near-quantitative yields
by reaction of 7a with an equivalent of sodium hydroxide
in aqueous THF. (Otherwise, 6a was prepared, but not
isolated, from the ester 5a prior to thermolysis.) Ther-
molysis of 6a , at 200 °C, resulted in the formation of a
complex mixture of products. The yield was approxi-
mately 27% by mass and consisted primarily of the
trifluoroacetate ester 11a as identified by GC analysis.
The trifluoroacetate ester 11a was isolated by vacuum
fractional distillation and further characterized by HRMS.
The trimethylsilyl ester 7 was injected at a constant
rate onto the evacuated preheated column as described
in the Experimental Section. Although the thermolysis
reaction is exothermic, the reaction column was heated
to between 140 and 150 °C to both maintain volatility of
and ensure efficient passage of the trimethylsilyl ester
through the column. GC analysis of the crude thermoly-
sis product indicated that it consisted primarily of TFVE,
fluorotrimethylsilane, and a small amount of the trifluo-
roacetate ester 11 (2-7% as determined by GC). The
fluorotrimethylsilane and residual carbon dioxide were
removed using an aspirator vacuum. TFVE 8 was
further purified by vacuum fractional distillation with
yields of 55-63% from the silyl ester.
The 19F NMR spectrum of 7 lends insight into the
thermolysis mechanism of trimethylsilyl esters to the
desired TFVEs. The 19F NMR spectrum of 7a in deuter-
ated chloroform, for example, was expected to have a
simple AX3 pattern. However, two sets of peaks were
observed as follows: (1) a doublet at -81.6 ppm and a
relatively broader singlet at -131.7 ppm with an integra-
tion ratio of 3:1, attributed to CF3 and CF, respectively,
and (2) a less intense doublet at -82.0 ppm and a broad,
considerably less intense, singlet at -131.2 ppm also with
an integration ratio of 3:1. Comparing this 19F NMR
spectrum to those of preceding intermediates, reaction
byproducts, and potential decomposition products of 7a
(i.e., the corresponding acid), it was determined that the
second set of peaks could also be attributed to the CF3
and CF of 7a . A similar 19F NMR spectrum was observed
for 7b.
In order to justify the assignment of the two sets of
peaks to the same fluorine atoms, an equilibrium be-
tween two different forms of 7 is proposed as follows: (1)
a conformation where there is no silicon-fluorine inter-
action and (2) a conformation with an intramolecular
interaction between silicon and fluorine (cf. Scheme 4).
The two possible interaction conformations that can be
rationalized include those of either CF or CF3 fluorine
with silicon through either a five- or six-membered ring,
respectively. The driving forces for the interaction
conformation may include the high silicon-fluorine bond
energy (43 kJ /mol) and the stability of the five- or six-
membered ring. The interaction conformation is disfa-
vored by the resultant steric crowding of the trimethyl-
silyl methyl groups in the interaction conformation which
may decrease entropy. The 19F NMR spectrum suggests
a CF fluorine-silicon interaction (or a five-membered
ring) because both peaks attributed to CF3 are doublets.
To support the proposed equilibrium, the effect of
temperature on relative NMR peak areas was studied.
19F NMR spectra were collected for 7a over a temperature
range of -30 °C (243 K) to 94 °C (367 K), the temperature
Discu ssion
Vacuum thermolysis of sodium 2-[2-(2-ethoxyethoxy)-
ethoxy]-2,3,3,3-tetrafluoropropionate, 6a , resulted in the
formation of a trifluoroacetate ester, 2-(2-ethoxyethoxy)-
ethyltrifluoroacetate, 11a . Sodium or potassium carbox-
ylate salts, similar to those described herein, have also
been shown to yield the trifluoroacetate esters as the
major product after thermolysis.1 A unimolecular reac-
tion pathway was proposed, similar to that shown in
Scheme 2: sodium fluoride, eliminated R to the carbonyl,
results in a zwitterionic intermediate which can exist in
either open-chain, 9, or cyclic, 10, forms. Loss of carbon
monoxide from 10 results in the observed trifluoroacetate
ester 11. The alkoxy substituent, which is relatively
electron-rich compared to a fluoroalkoxy substituent, may
also stabilize the zwitterion 9 through resonance with
the positive charge on the ether oxygen.
In contrast to the mechanism described for the forma-
tion of the undesirable trifluoroacetate ester 11 from 6,
it is generally accepted that salts of 2-fluoroalkoxy-

TFVEs with Functionalized Hydrocarbon Ether Groups
J . Org. Chem., Vol. 62, No. 22, 1997 7847
being limited only by the boiling point of the NMR solvent
(deuterated chloroform). Assuming that the less intense
set of peaks represents the interaction conformation, the
small to large peak area ratio22 may be expected to
increase with increasing temperature if the equilibrium
(as described in Scheme 4) is endothermic. As the
temperature increases the equilibrium constant (Keq)
shifts to the right.
Con clu sion
Two new trifluorovinyl ethers incorporating hydrocar-
bon oligoether groups were synthesized by a new method
from hexafluoropropene oxide and the appropriate so-
dium alkoxide. Conversion of the sodium (or potassium)
carboxylate to the trimethylsilyl ester prior to thermoly-
sis was essential for successful preparation of the desired
TFVE. Without conversion to the trimethylsilyl ester,
the undesired trifluoroacetate ester was formed. The
thermolysis mechanism was investigated in order to gain
insight into the importance of the silyl ester. 19F NMR
spectra indicate that either a five or six-membered ring
exists between silicon and either CF or CF3 fluorine. This
interaction may influence the thermolysis mechanism.
As shown in Figure 1, a van’t Hoff plot of ln Keq versus
1/T appears linear (R2 is 0.988); ln Keq decreases as 1/T
increases, or Keq increases with increasing temperature.
Assuming a similar equilibrium exists in the gas phase
as exists in the deuterated chloroform, Keq is estimated
at 0.10 for thermolysis temperatures. The standard
enthalpy for the equilibrium described in Scheme 4 was
calculated from the slope in Figure 1, ∆H° ) 3.0 ( 0.1
kJ mol-1, and the standard entropy from the intercept,
Exp er im en ta l Section
∆S° ) -12.1 ( 0.4 J mol-1 K-1
. The enthalpy and
entropy data indicate that entropy dominates Keq at
The reagents and solvents were purchased from Aldrich
Chemical Co. (Millwaukee, WI) and used as received unless
otherwise stated. Anhydrous diethyl ether and Celite 545 were
purchased from Fisher Scientific (Toronto, Canada). Dry
temperatures higher than 250 K.
sodium hydride was prepared prior to use from
a 60%
dispersion in mineral oil by dispersion in pentane followed by
filtration through a medium porosity glass frit funnel (four
times) and then dried under nitrogen. Hexafluoropropene
oxide (HFPO, 97%) was purchased from PCR Chemicals
(Gainesville, FL) and used as received.
Compounds were characterized by gas chromatography (HP
6890, Hewlett Packard, Ontario) using a Restek Rtx-5 column
(Chromatographic Specialties, Ontario; 0.530 mm × 15 m with
a 1.2 µm film thickness) with FID detector, helium carrier gas
(35 cm/s), and a split ratio of ∼25:1. A typical temperature
profile held the initial temperature at 80 °C for 1 min, then
ramped the temperature to 230 °C at 15 °C/min, and finally
held the temperature at 230 °C for 4 min.
Proton nuclear magnetic resonance (1H NMR) spectra were
taken on a 200 MHz Varian Gemini spectrometer using
tetramethylsilane as an external reference standard and
deuterated chloroform as the solvent. Fluorine (19F NMR)
spectra were taken on a 300 MHz Varian Gemini spectrometer
using fluorotrichloromethane as an external reference stan-
dard and deuterated chloroform as the solvent. Mass spectra
were obtained on a Micromass 70-250S (double-focusing) mass
spectrometer, arrayed with a HP 5890 gas chromatograph
(capillary column: J &W Scientific, DB-5ms, 30 m, 0.25 mm).
High-resolution data were obtained at 10 000 (10% valley)
resolution. 2σ ) 5.7 ppm based on 27 measurements of the
molecular ion of cholesterol.
The gas phase thermolysis apparatus consisted of a 1.9 cm
× 43 cm borosilicate glass tube with B24 female inlet and B24
male outlet ground glass joints. A 1.3 cm × 2.3 m 400 W
heating tape was wrapped along the middle 35 cm and
insulated with multiple layers of glass insulation tape. Sample
and nitrogen carrier gas were introduced via needles through
a septum. A B24/B14 reducer to vacuum side arm adapter
and a receiver flask were attached to the outlet. The tube was
packed with 40 g of approximately 3 mm diameter borosilicate
glass beads held in place by Vigreaux type indentations in the
tube, 1.3 cm above the outlet. Suspended on the glass beads
was approximately 2-3 g of potassium fluoride. The temper-
ature was maintained by a temperature controller connected
to a hose clamp thermocouple mounted on the outside of the
heating tape and under the insulation. The internal temper-
ature was monitored using a thermocouple mounted in the
center of the tube through a port located halfway along the
tube length.
1
F igu r e 1. Plot of ln Keq vs
/ for the equilibium defined in
T
Scheme 4 for trimethylsily ester 7a (R2 ) 0.988). From the
slope, ∆H° ) 3.0 ( 0.1 kJ mol-1. From the intercept, ∆S° )
-12.1 ( 0.4 J mol-1 K-1
.
The trimethylsilyl ester 7 is necessary for successful
thermolysis to the desired TFVE 8; however, it is not
clear that the proposed interaction conformation at
thermolysis temperatures contributes significantly to the
overall success. The proposed silicon-fluorine interac-
tion may serve to weaken the silicon-oxygen bond,
thereby facilitating decarboxylation in a similar way that
alkali metal 2-methoxy-2,3,3,3-tetrafluoropropionates are
more easily decarboxylated with weaker metal-oxygen
bonds.23 However, the concentration of the interaction
conformation is relatively low (estimated at 9% from
Figure 1) compared to the noninteraction conformation
at thermolysis temperatures. On contact with the potas-
sium fluoride catalyst, the trimethylsilyl ester is desilyl-
ated. Given that potassium carboxylates yield the un-
desired trifluoroacetate esters 11, it is likely that decar-
boxylation is well advanced before a potassium-oxygen
bond can form. The potassium fluoride catalyst is
necessary because, in its absence, most of the trimeth-
ylsilyl ester is recovered while a small amount (∼12 mol
%) of the undesirable trifluoroacetate ester is formed.
This observation further supports the proposed ther-
molysis mechanism.
2-(2-ter t-Bu toxyeth oxy)eth a n ol (1b). To a dry 500 mL,
3-neck round bottom flask that was fitted with a dry ice/
acetone condenser under a static nitrogen purge were added
30.0 g (283 mmol) of diethylene glycol, 250 mL of methylene
chloride, and 7.0 g of Amberlyst 15 resin. With magnetic
stirring, 19.0 g (337 mmol) of isobutene was slowly added and
the reaction mixture was maintained at 30 °C. After 7 h, GC
(22) The CF3 19F NMR signals were used for the van’t Hoff plot to
obtain the best signal to noise resolution. The TMS ester was distilled
twice to >99% purity as determined by GC.
(23) Rozov, L. A.; Ramig, K. Tetrahedron Lett. 1994, 35, 4501-4.

7848 J . Org. Chem., Vol. 62, No. 22, 1997
Lousenberg and Shoichet
analysis indicated that 90% of the diethylene glycol had been
converted to 2-(2-tert-butoxyethoxy)ethanol and that very little
(<1%) had been converted to bis(2-tert-butoxyethyl) ether. The
reaction was stopped by removal of the Amberlyst 15 resin by
gravity filtration. The solvent was removed by rotary evapo-
ration leaving a clear, oily liquid. 2-(2-tert-Butoxyethoxy)etha-
nol (32.1 g, 70% yield, >96% purity by GC) was isolated by
vacuum fractional distillation (∼0.05 mmHg, bp 36-37 °C).
1H NMR: δ 3.8-3.5 (m, 8H, CH2), 3.0 (br s, 1H, OH), 1.2 (s,
9H, C(CH3)3).
Tr im et h ylsilyl 2-[2-(2-E t h oxyet h oxy)et h oxy]-2,3,3,3-
t et r a flu or op r op ion a t e (7a ). To a 300 mL, round bottom
flask equipped with condenser and thermocouple side port
were added 26.9 g (68.2 mmol) of 2-(2-ethoxyethoxy)ethyl 2-[2-
(2-ethoxyethoxy)ethoxy]-2,3,3,3-tetrafluoropropionate and 175
mL of THF. Sodium hydroxide (3.4 g or 85 mmol) and 6.1 mL
(340 mmol) of deionized water were added to the flask with
magnetic stirring and maintained at 40 °C for 3 h during which
the translucent reaction mixture became slightly yellow. GC
analysis of a sample indicated that the ester was converted to
the 2-[2-(2-ethoxyethoxy)ethoxy]-2,3,3,3-tetrafluoropropionic
acid sodium salt. Most of the THF and water were removed
from the reaction mixture by rotary evaporation. Any remain-
ing water and most of the 2-(2-ethoxyethoxy)ethanol (∼80%)
were removed by short path vacuum distillation (∼0.05 mmHg,
stillpot temperature < 100 °C). After the reaction pot cooled
to ambient temperature, approximately 125 mL of diethyl
ether was added by cannula into the flask. The sodium salt
dissolved with magnetic stirring to give a translucent, light
amber solution. Sodium hydride (0.8 g or 30 mmol) was added
to the flask. With ice bath cooling, 15.0 g (138 mmol) of
trimethylsilyl chloride was slowly added to the flask during
which a slight exotherm and sodium chloride precipitation
were observed. After 3 h of stirring at room temperature, the
reaction mixture was filtered through 1 cm of Celite 545 on a
medium porosity glass frit funnel. Most of the diethyl ether
was removed by rotary evaporation. The crude product was
vacuum distilled using a 4 in. Vigreaux column from which
three fractions were isolated; the first fraction (bp 28-30 °C,
0.06 mmHg) was identified as 2-(2-ethoxyethoxy)ethanol, the
second fraction (bp 38-40 °C, 0.06 mmHg) was identified as
2-(2-ethoxyethoxy)ethyl trimethylsilyl ether, and the third
fraction (bp 56-57 °C, 0.04 mmHg) was identified as the
desired product, trimethylsilyl 2-[2-(2-ethoxyethoxy)ethoxy]-
2,3,3,3-tetrafluoropropionate (yield 18.6 g, 78%, >97% purity
by GC). 19F NMR (20 °C): δ -81.6 and -82.0 (d, CF3), -131.2
and -131.7 (br s, CF). 1H NMR: δ 3.9 (m, 2H, CFOCH2), 3.8-
3.45 (m, 6H, OCH2), 1.2 (t, 3H, CH3), 0.4 and 0.15 (s, 9H,
Si(CH3)3). HRMS: (M -H)+ calcd 349.1094, obsd 349.1086.
Tr im eth ylsilyl 2-[2-(2-ter t-Bu toxyeth oxy)eth oxy]-2,3,3,3-
tetr a flu or op r op ion a te (7b). To a 300 mL, round bottom
flask equipped with a condenser and thermocouple side port
were added 20.0 g (44.4 mmol) of 2-(2-tert-butoxyethoxy)ethyl
2-[2-(2-tert-butoxyethoxy)ethoxy]-2,3,3,3-tetrafluoropropi-
onate and 130 mL of THF. Sodium hydroxide (2.11 g or 52.8
mmol) and 4.0 mL (220 mmol) of deionized water were added
to the reaction flask with magnetic stirring and maintained
at 40 °C overnight during which a translucent, pale yellow
solution resulted. GC analysis of a sample indicated that the
ester was converted to the sodium salt. Most of the THF and
water were removed from the reaction mixture by rota-
ryevaporation. The remaining water and most of the 2-(2-tert-
butoxyethoxy)ethanol (∼70%) were removed by short path
vacuum distillation (∼0.05 mmHg, stillpot temperature < 100
°C). After the reaction pot cooled to ambient temperature,
approximately 125 mL of diethyl ether was added by cannula
into the flask. The sodium salt dissolved with magnetic
stirring to give a translucent, yellow solution; 0.5 g (21 mmol)
of sodium hydride was added to the reaction flask with
evolution of a small amount of hydrogen gas; 10.0 g (92.0
mmol) of trimethylsilyl chloride was slowly added to the flask
during which a slight exotherm and sodium chloride precipita-
tion were observed. After 3 h of stirring at room temperature,
the mixture was filtered through 1 cm of Celite 545 on a coarse
porosity glass frit funnel. Most of the diethyl ether was
removed by rotary evaporation. The crude product was
vacuum distilled using a 10 cm Vigreaux column from which
two fractions were isolated; the first fraction (bp 45-48 °C,
0.04 mmHg, yield 2.3 g) was identified as 2-(2-tert-butoxy-
ethoxy)ethyltrimethylsilyl ether, and the second fraction (bp
71-73 °C, 0.04 mmHg) was identified as the desired product,
trimethylsilyl 2-[2-(2-tert-butoxyethoxy)ethoxy]-2,3,3,3-tet-
rafluoropropionate (yield 10.8 g, 69%, >96% purity by GC).
19F NMR (ambient temperature): δ -81.6 and -82.0 (d, CF3),
-130.6 and -131.7 (br s, CF). 1H NMR: δ 3.9 (m, 2H,
CFOCH2), 3.75-3.45 (m, 6H, OCH2), 1.3 and 1.2 (s, 9H,
2-(2-Eth oxyeth oxy)eth yl 2-[2-(2-Eth oxyeth oxy)eth oxy]-
2,3,3,3-tetr a flu or op r op ion a te (5a ). To a dry 500 mL, 3-neck
round bottom flask that was fitted with a 50 mL pressure-
equalizing addition funnel, mechanical stirrer, and dry ice/
acetone condenser under a nitrogen purge was added 3.6 g
(150 mmol) of dry sodium hydride. Approximately 200 mL of
anhydrous ethylene glycol dimethyl ether (DME) was added
by cannula to the flask. The reaction flask was cooled to ∼0
°C, and the dispersion stirred while 20.0 g (149 mmol) of 2-(2-
ethoxyethoxy)ethanol was slowly added. In forming the alkox-
ide, hydrogen gas evolved and the resulting cloudy white
solution was stirred for 1 h. The addition funnel was replaced
with a septum through which 14.6 g (88.0 mmol) of HFPO was
slowly added using a transfer needle. Since the reaction was
very exothermic, care was taken to maintain the temperature
at or below 30 °C. A cloudy, pale yellow reaction mixture
resulted and was stirred for an additional 3 h at room
temperature. The reaction mixture was filtered through 1 cm
of Celite 545 on a coarse porosity glass frit funnel (to separate
sodium fluoride), and the supernatant was rotary evaporated,
leaving a yellow, oily liquid that was vacuum distilled using
a 10 cm Vigreaux column. Two fractions were isolated; the
first fraction (bp 43-45 °C, 0.05 mmHg) was identified as the
byproduct ester, 2-(2-ethoxyethoxy)ethyl perfluoropropionate
(yield 3.8 g, 13%), and the second fraction (bp 112-114 °C,
0.04 mmHg) was identified as the desired product, 2-(2-
ethoxyethoxy)ethyl 2-[2-(2-ethoxyethoxy)ethoxy]-2,3,3,3-tet-
rafluoropropionate (yield 22.8 g, 78%, >97% purity by GC).
19F NMR: δ -81.7 (d, CF3), -132.0 (br m, CF). 1H NMR: δ
4.5 (m, 2H, CO2CH2), 3.95 (m, 2H, CFOCH2), 3.7-3.45 (m,
16H, OCH2), 1.2 (t, 6H, CH3). HRMS: (MH)+ calcd 395.1693,
obsd 395.1705.
2-(2-ter t-Bu t oxyet h oxy)et h yl 2-[2-(2-ter t-Bu t oxyet h -
oxy)eth oxy]-2,3,3,3-tetr a flu or op r op ion a te (5b). To a dry
500 mL, 3-neck round bottom flask that was fitted with a 50
mL pressure-equalizing addition funnel, mechanical stirrer,
and dry ice/acetone condenser under a nitrogen purge was
added 3.3 g (140 mmol) of dry sodium hydride. Approximately
200 mL of anhydrous DME was added by cannula to the flask.
2-(2-tert-Butoxyethoxy)ethanol (22.0 g or 136 mmol) was slowly
added to the cooled reaction flask (∼0 °C) to form the alkoxide
with evolution of hydrogen gas. The resulting cloudy white
reaction mixture was stirred for 1 h to ensure complete
formation of the alkoxide. The addition funnel was replaced
with a septum through which 17.0 g (102 mmol) of HFPO was
slowly added using a transfer needle and during which the
reaction mixture was maintained below 30 °C. The slightly
cloudy, pale yellow reaction mixture that resulted was stirred
for an additional 3 h. The reaction mixture was filtered
through 1 cm of Celite 545 on a coarse porosity glass frit funnel
(to separate sodium fluoride), and the supernatant was
removed by rotary evaporation, leaving a yellow, oily liquid.
The crude product was vacuum distilled using a short path
distillation apparatus. Two fractions were isolated, the first
fraction (bp 53-56 °C, 0.05 mmHg, yield 4.3 g, 14%) was
identified by GC as a mixture of unreacted alcohol (0.3 g) and
byproduct ester 2-(2-tert-butoxyethoxy)ethyl perfluoropropi-
onate (yield 4 g, 13%), and the second fraction (bp 132-134
°C, 0.04 mmHg) was identified as the desired product, 2-(2-
tert-butoxyethoxy)ethyl 2-[2-(2-tert-butoxyethoxy)ethoxy]-2,3,3,3-
tetrafluoropropionate (yield 21.8 g, 71.5%, >99% purity by GC).
19F NMR: δ -81.7 (d, CF3), -132.0 (br m, CF). 1H NMR: δ
4.5 (m, 2H, CO2CH2), 3.95 (m, 2H, CFOCH2), 3.8-3.4 (m, 12H,
OCH2), 1.2 (s, 18H, C(CH3)3). HRMS: (M -CH3)+ calcd
435.2006, obsd 435.1988.

TFVEs with Functionalized Hydrocarbon Ether Groups
J . Org. Chem., Vol. 62, No. 22, 1997 7849
C(CH3)3), 0.4 and 0.15 (s, 9H, Si(CH3)3). HRMS: (M - CH3)+
calcd 363.1251, obsd 363.1237.
suspended small particulates were separated by centrifuga-
tion. Most of the THF and water were removed by rotary
evaporation, and final purification was done in a vacuum oven
at 40 °C for 2 h to give a very viscous, pale yellow liquid. The
yield of sodium 2-[2-(2-ethoxyethoxy)ethoxy]-2,3,3,3-tetrafluo-
ropropionate was 7.58 g (95%). 19F NMR: δ -81.9 (br s, CF3),
-131.8 (br s, CF). 1H NMR: δ 4.2-3.4 (br m, 10H, OCH2),
1.2 (m, 3H, CH3).
1-[2-(2-Eth oxyeth oxy)eth oxy]-1,2,2-tr iflu or oeth en e (8a).
Trimethylsilyl 2-[2-(2-ethoxyethoxy)ethoxy]-2,3,3,3-tetrafluo-
ropropionate (9.30 g or 26.5 mmol) was injected onto the
preheated thermolysis column (140 °C, ∼50 mL/min N2 flow,
∼0.2 mmHg) at a rate of ∼0.2 mL/min using a 10 mL Gastight
syringe and syringe pump. The thermolysis was exothermic,
and the column temperature increased to ∼160 °C during the
injection. The product was collected for an additional 1 h after
the injection was completed in a receiver flask cooled with
liquid nitrogen; 7.8 g of an oily, pale yellow liquid product was
collected. The crude product was fractionally distilled to give
3.58 g (63% yield, 95% purity by GC) of a product identified
as 1-[2-(2-ethoxyethoxy)ethoxy]-1,2,2-trifluoroethene (bp 22 °C,
0.15 mmHg). 19F NMR: δ -123.4 (dd, 1F, J ) 56, 104 Hz,
CF), -130.2 (dd, 1F, J ) 104, 108 Hz, CF), -135.1 (dd, 1F, J
) 56, 108 Hz, CF). 1H NMR: δ 4.15 (m, 2H, CFOCH2), 3.75
(t, 2H, OCH2), 3.7-3.45 (m, 6H, OCH2), 1.2 (t, 3H, CH3).
1-[2-(2-t er t -B u t o x y e t h o x y )e t h o x y ]-1,2,2-t r iflu o r o -
eth en e (8b). Trimethylsilyl 2-[2-(2-tert-butoxyethoxy)ethoxy]-
2,3,3,3-tetrafluoropropionate (11.8 g or 31.3 mmol) was injected
onto the preheated thermolysis column (150 °C, ∼50 mL/min
N2 flow, ∼0.2 mmHg) at a rate of ∼0.2 mL/min using a 10 mL
Gastight syringe and syringe pump. The thermolysis was
exothermic, and the column temperature increased to ∼170
°C during the injection. The product was collected in a receiver
flask cooled with liquid nitrogen for an additional 1 h after
the injection was completed. An oily, pale yellow liquid
product (8.72 g) was collected. The crude product was frac-
tionally distilled to give 4.2 g (55% yield, 95% purity by GC)
of a product identified as 1-[2-(2-tert-Butoxyethoxy)ethoxy]-
2-(2-Eth oxyeth oxy)eth yl Tr iflu or oa ceta te (11a ). To a
300 mL round bottom flask equipped with condenser and
thermocouple side port were added 30.4 g (77.1 mmol) of 2-(2-
ethoxyethoxy)ethyl 2-[2-(2-ethoxyethoxy)ethoxy]-2,3,3,3-tet-
rafluoropropionate and 175 mL of THF. Sodium hydroxide
(4.06 g or 102 mmol) and 7.13 mL (396 mmol) of deionized
water were added to the flask with magnetic stirring and
maintained at 40 °C for 3 h during which the reaction mixture
became slightly yellow. The reaction mixture was filtered
through 1 cm of Celite 545 on a coarse porosity glass frit
funnel. The majority of the THF and water were removed by
rotary evaporation. Most (92%) of the byproduct, 2-(2-ethoxy-
ethoxy)ethanol, was removed by short path vacuum distillation
(mantle temperature ∼ 120 °C). The temperature of the
heating mantle was increased to 200 °C, and the resulting
thermolysis product was collected in a receiver flask cooled
by liquid nitrogen. An oily, pale yellow liquid was obtained
(8.2 g or 27%, w/w, based on the starting ester). GC analysis
of a sample indicated that it was a complex mixture with one
predominant (∼50%) low-boiling component. By vacuum
fractional distillation (bp 25 °C, 0.07 mmHg, 87% purity by
GC), 3.3 g of a product was isolated and identified as 2-(2-
ethoxyethoxy)ethyl trifluoroacetate. 19F NMR: δ -75.3 (s,
CF3). 1H NMR: δ 4.5 (t, 2H, C(O)CH2), 3.8 (t, 2H, OCH2), 3.6-
3.45 (m, 6H, OCH2), 1.2 (t, 3H, CH3). HRMS: (MH)+ calcd
231.0844, obsd 231.0822.
1,2,2-trifluoroethene (bp 26 °C, 0.15 mmHg). 19F NMR:
δ
-123.5 (dd, 1F, J ) 56, 104 Hz, CF), -130.3 (dd, 1F, J ) 104,
108 Hz, CF), -135.1 (dd, 1F, J ) 56, 108 Hz, CF). 1H NMR:
δ ) 4.15 (m, 2H, CFOCH2), 3.75 (t, 2H, OCH2), 3.65-3.45 (m,
4H, OCH2), 1.2 (s, 9H, C(CH3)3), HRMS: (M - CH3)+ calcd
227.0895, obsd 227.0892.
Ack n ow led gm en t. The authors gratefully acknowl-
edge partial financial assistance from NSERC, the
University of Toronto, and DuPont Canada.
Sod iu m 2-[2-(2-Eth oxyeth oxy)eth oxy]-2,3,3,3-tetr a flu -
or op r op ion a te (6). To a 300 mL round bottom flask were
added 9.33 g (26.6 mmol) of trimethylsilyl 2-[2-(2-ethoxy-
ethoxy)ethoxy]-2,3,3,3-tetrafluoropropionate and 85 mL of
THF. With magnetic stirring, 1.22 g (30.5 mmol) of sodium
hydroxide followed by 2.52 mL (140 mmol) of deionized water
were added to the reaction flask. After 1 h of stirring at room
temperature, the reaction mixture was filtered through 1 cm
of Celite 545 on a medium porosity frit funnel. Residual
Su p p or tin g In for m a tion Ava ila ble: 1H and 19F NMR
spectra of both monomers 8a ,b (4 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
J O971319D