Perfluorocyclobutane (PFCB) polyaryl ethers: Versatile coatings materials

Article abstract of DOI:10.1016/S0022-1139(00)00233-5

The cyclopolymerization of aromatic trifluorovinyl ether (TFVE) monomers offers a versatile route to a unique class of linear and network fluoropolymers containing the perfluorocyclobutyl (PFCB) linkage. Polymerization proceeds by a thermal - radical medi

Full text of DOI:10.1016/S0022-1139(00)00233-5

Journal of Fluorine Chemistry 104 (2000) 109±117
Per¯uorocyclobutane (PFCB) polyaryl ethers:
versatile coatings materials
Dennis W. Smith Jr.a,*, David A. Babbb, Hiren V. Shaha,
Adrienne Hoeglunda, R. Traiphola, Dvora Perahiaa,
Harold W. Booneb, Charles Langhoffb, Mike Radlerb
aDepartment of Chemistry, Clemson University, Clemson, SC 29634, USA
bCorporate Research, The Dow Chemical Company, Midland, MI, USA
Abstract
The cyclopolymerization of aromatic tri¯uorovinyl ether (TFVE) monomers offers a versatile route to a unique class of linear and
network ¯uoropolymers containing the per¯uorocyclobutyl (PFCB) linkage. Polymerization proceeds by a thermal Ð radical mediated Ð
step-growth mechanism and provides well-de®ned polymers containing known ¯uoroole®n end groups. PFCB polymers combine the
engineering thermoplastic nature of polyaryl ethers with ¯uorocarbon segments and exhibit excellent processability, optical transparency,
high temperature performance, and low dielectric constants. An intermediate strategy utilizing Grignard and aryllithium reagents has been
developed which offers access to a wide variety of hybrid materials amenable to coatings applications. Liquid crystalline examples have
recently been achieved in addition to tailoring optical properties by co-polymerization. # 2000 Elsevier Science S.A. All rights reserved.
Keywords: Cyclopolymerization; Per¯urocyclobutane polymers; Tri¯uorovinyl ethers; Raman spectrocopy
1. Introduction
The cyclodimerization of TFVEs is activated thermally
and is favored for ¯uoroole®ns due to an increased double-
bond strain, a lower p-bond energy, and the strength of the
resulting ¯uorinated C±C single bond cycloadducts in con-
trast to hydrocarbon analogs [21]. As generalized in
Scheme 1, predominant head-to-head cycloaddition pro-
ceeds to form the more stable diradical intermediate fol-
lowed by rapid ring closure giving essentially an equal
mixture of cis- and trans-1,2-disubstituted PFCB linkages.
Both direct measurements on polymers [4] and model
studies [8] have consistently revealed an equal isomeric
distribution in PFCB polymers.
The PFCB approach has been to combine ¯exible, yet
thermally robust, aromatic ethers with ¯uorocarbon lin-
kages, thereby providing ¯uoropolymers which are easily
processed from solution or the melt. High molecular weight
PFCB thermoplastic polymers or solution advanced thermo-
setting resins can be spin coated and cured giving ®lms
which exhibit high glass transition temperature's (Tg), good
thermal stability, optical clarity, and isotropic dielectric
constants around 2.3 at 1 MHz. The step growth addition
type polymer chemistry which proceeds to form PFCB
polymers is thermally activated, does not require catalysts
or initiators, and does not produce condensation products
during polymerization or ®nal cure. Cycloaddition polymer-
Many practical coatings applications of highly crystalline
or high melt viscosity per¯uorinated materials, such as
Te¯onTM are often limited due to prohibitive processing
costs [1±3]. Incorporation of partially ¯uorinated segments
or substituents, however, yield polymers with low dielectric
constant, moisture resistance, low surface energy, thermal
and thermal-oxidative stability, and chemical resistance
without sacri®cing Ð and frequently improving Ð proces-
sability [1±3]. Our efforts have focused on the development
of a unique class of semi-¯uorinated polymers which utilize
the thermal cyclopolymerization of aryl tri¯uorovinyl ethers
(TFVEs) from which a variety of thermoplastic and thermo-
setting polymers containing the per¯uorocyclobutane
(PFCB) aromatic ether linkage have been obtained
(Scheme 1) [4±14]. The combination of processability
and performance provided by PFCB polymers makes them
natural candidates for applications, such as high perfor-
mance structural coatings, interlayer dielectrics, circuit
board laminates, dielectric wave guides, optical cladding
layers, and coatings for space applications [15±20].
* Corresponding author. Tel.: ‡1-8646565020.
E-mail address: dwsmith@clemson.edu (D.W. Smith Jr.)
0022-1139/00/$ ± see front matter # 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 0 0 ) 0 0 2 3 3 - 5

110
D.W. Smith Jr. et al. / Journal of Fluorine Chemistry 104 (2000) 109±117
Scheme 1. Cyclopolymerization of Trifluorovinyl Ether monomers.
ization of this type results in well de®ned linear or network
polymers containing known tri¯uorovinyl aromatic terminal
groups at any stage of polymerization [4].
with a DHˆ 50 kcal/mol due to this unique ¯uoroole®n
reactivity [22]. For TFVE monomers, exothermic polymer-
ization reaches a measurable rate (DSC at 108C/min) near
1408C and polymerizations are typically carried out at
temperatures between 150±2108C [5].
2. Results and discussion
The classical step growth kinetics by which PFCB poly-
mers are formed allows for easy control of parameters
important to coating technology in general. Monomer can
be solution advanced at 1508C in typical solvents to a
precisely controlled viscosity, molecular weight, and poly-
dispersity [5±6]. The pre-polymer solution formed can be
spin-coated giving defect free ®lms. Final cure can then be
achieved by post baking under nitrogen or air at tempera-
tures ranging from 235±3258C for several hours depending
on the application.
Polymerization kinetics for monomer 1 has recently been
studied in detail by Raman spectroscopy [5]. This technique
is well suited for ¯uoroole®n conversion due to the
enhanced signal of ¯uoroole®ns and the possibility of fast
and non-destructive measurement. Fig. 1 compares the
Raman spectra of pure monomer 1, an oligomeric ®lm spin
coated from mesitylene solution on silicon, and the fully
cured ®lm (3±6 mm in thickness).
2.1. Traditional PFCB coatings materials
Traditional TFVE monomers are prepared in two steps
from commercially available phenolic precursors via ¯uor-
oalkylation with BrCF2CF2Br followed by Zn mediated
elimination (Scheme 2) [10±11]. The trifunctional monomer
prepared from tris(hydroxyphenyl)ethane (1) has, thus, far
been the primary focus for PFCB technology targeted for
coatings applications. However, other monomers and poly-
mers containing bisphenyl¯uorene, [10] phosphine and
phosphine oxide [6,10], triphenyl benzene [11], siloxane
[4,12±13], and phosphazene [14] linkages have been
reported. Scheme 2 illustrates the structural versatility of
selected TFVE monomers prepared from phenolic precur-
sors.
The polymerization reaction of trifunctional monomer 1
forms a branched oligomer containing unreacted TFVE
terminal groups. As predicted for trifunctional monomers
undergoing step growth polymerization, the polydispersity
of the soluble oligomers grows steadily with molecular
weight prior to gelation near 50% conversion. Tri¯uorovinyl
aryl ether monomers undergo spontaneous exothermic
cyclopolymerization at relatively mild (>1308C) tempera-
tures due to a weak and strained C=C p-bond. The cyclo-
dimerization of tetra¯uoroethylene was reported to proceed
Clearly, the Raman signal representing the per¯uorinated
1
C=C stretching at 1831 cm can be used to monitor the
concentration of reactive groups remaining during initial
polymerization as well as ®nal cure on the substrate. The
1
1831 cm peak was normalized using the phenyl ring
1
vibration at 1605 cm as an internal standard. The spec-
trum of the fully cured material (3008C, 50 h) in Fig. 1
shows that the 1831 cm band will disappear after long
cure times at high temperature demonstrating that complete
1
Scheme 2. PFCB polymers prepared from bis-phenols.

D.W. Smith Jr. et al. / Journal of Fluorine Chemistry 104 (2000) 109±117
111
Table 1
Selected properties of PFCB thermoset polymer from monomer 1
Tensile strength (MPa)
Tensile modulus (MPa)
Flexural strength (MPa)
Flexural modulus (MPa)
% Elongation (break)
Tg (DMA)
66.0Æ1.4
2.270Æ79
74Æ12
2.320Æ13
4.1
3808C
2.35
Dielectric constant (10 kHz)
Dissipation factor (10 kHz)
% Water absorption (24 h)
Refractive index (800 nm)
0.0004
0.021
1.495
Fig. 1. Trifluorovinyl consumption in monomer 1 by Raman spectro-
scopy.
isothermal TGA analysis where the same polymer exhibits a
weight loss of <0.05%/h in nitrogen and 0.7%/h in air at
3508C. PFCB polymers, in general, thermally degrade by
homolytic cleavage at the cyclobutyl ether linkage forming
hexa¯uorocyclobutene [9]. Table 1 gives additional selected
properties for the thermoset polymer prepared from mono-
mer 1 [10].
The PFCB ®lms obtained exhibit exceptional planariza-
tion of topographical features, and have also demonstrated
the ability to smooth substrates, such as color pixels for ¯at
panel displays [20]. In addition, the pre-polymer has been
shown to planarize Canasite glass ceramic substrates used in
magnetic data storage devices [15]. PFCB polymers have
also been used as an electronic buffer layer in the prepara-
tion of multilayer Mach±Zender intensity modulator wave-
guiding devices [15±20].
conversions can be determined by the Raman technique.
This allows for the subtle effect of cure conversion on
performance to be revealed. For example, in thin ®lm
applications, Tg will change during processing depending
on the degree of cure, thereby affecting other critical inter-
related parameters, such as modulus, toughness, stress, and
adhesion.
The extent of cure versus time is plotted in Fig. 2 for cure
temperatures of 130, 150, 170, 190, and 2108C. The half-life
varies from ꢀ450 min to less than 10 min at these tempera-
tures. From this data, the Arrhenius relationship of ln (k)
versus 1/T gave an activation energy of 103 kJ/mol or
24.6 kcal/mol. This value agrees well with activation ener-
gies determined by other methods, such as DSC, GPC and
NMR.
2.2. The intermediate strategy
The thermal and thermal oxidative stability of PFCB
polymers, as well as the degradation mechanism and zeroth
order kinetic analysis has been reported in detail [9,6].
Using dynamic thermogravimetric (TGA) analysis at
108C/min in nitrogen, the fully cured polymer shown in
Fig. 2 retains 100% of its weight to >4708C after which
catastrophic degradation ensues (Tonsetˆ4758C). However, a
more practical measure of thermal stability is obtained by
The traditional synthesis of TFVE monomers relied on
¯uoroalkylation followed by zinc mediated elimination of
phenolic precursors (Scheme 2). Since cost and availability
are limiting factors to any technology, we undertook a study
to develop methods to deliver the tri¯uorovinyl aryl ether
group to a wide variety substrates directly from reasonably
accessable, and low cost, precursors. We chose simple
Fig. 2. Trifluorovinyl conversion vs. time for the neat polymerization of 1 at 1508C (reproduced from [5], with permission).

112
D.W. Smith Jr. et al. / Journal of Fluorine Chemistry 104 (2000) 109±117
Scheme 3. Monomers prepared by the Intermediate Strategy.
derivatives of phenol, such as 4-bromophenol in order to
access a nucleophillic organometalic reagent containing the
TFVE group intact (Scheme 3) [4,7,12±14].
¯uorovinyl ether stabilizes a p-carbanion by as much as
11 kcal/mol and destabilizes the protonated pyridinium
analog by 10 kcal/mol [7].
This strategy featured the application of the bromo
derivative (6) which could be easily converted to reactive
Grignard (7) and lithium (8) reagents and thereby access a
variety of new inorganic/organic hybrid ¯uorinated materi-
als amenable to coatings applications. Aryl bromide 6 is
easily prepared on the multi-pound per batch scale and is
isolated pure by vacuum distillation [4]. Grignard reagent 7
is generated by standard methods and is remarkably stable
in THF solution, even at elevated temperatures. The recently
discovered aryl lithium reagent 8, on the other hand, is much
more reactive and its formation is more speci®c [7] (Cau-
tion! When warmed above 208C, solutions of TFVE-Li
react very exothermically with subsequent decomposition).
However, stable solutions are easily prepared and used at
low temperature in diethyl ether.
The preparation and use of these nucleophilic organo-
metallic reagents in the presence of electrophilic ¯uorinated
ole®ns was a surprising result and prompted us to explore
theoretical support for their stability. Upon calculating (at
the ab initio level) deprotonation energies of tri¯uoroviny-
loxy substituted benzene and pyridine model compounds
versus the corresponding methyl ether, we found that per-
Grignard reagent 7 was ®rst used to prepare a novel
disiloxane-containing monomer (9, xˆ1) from which the
®rst elastomeric ¯uorosilicone PFCB polymers were pro-
duced [4]. Subsequently, a range of ¯uorosilicone PFCB
polymers have been prepared. Table 2 exhibits selected
properties for polymers prepared from (9) for xˆ1±3.
Alternatively, reaction with trichlorophosphine and sub-
sequent oxidation afforded the ®rst tris-tri¯uorovinyloxy-
phenyl phosphine oxide monomer (10) which polymerized
to a highly thermal and thermal oxidatively stable thermoset
[6]. In addition, a completely aromatic example (11) was
prepared by treatment of 7 with B(OMe)3 and subsequent Pd
catalyzed Suzuki coupling with 1,3,5-tribromobenzene
[11]. Polyphosphazene networks have also been obtained,
from 8, by either deprotonation/substitution of preformed
polyphosphazenes or thermolysis of TFVE functionalized
chlorophosphoranimines [14].
Organolithium intermediate 8 was further used to synthe-
size vinyl derivatives containing the TFVE group [7]. Vinyl
derivatives are reacted cleanly with a variety of commercial
poly(dimethylsiloxane) co-polymers containing Si±H lin-
kages by employing well known Pt(0) catalyzed hydrosila-
Table 2
Selected properties of PFCB fluorosilicones from monomer 9 in Scheme 3
a
a
x
GPC Mn
GPC Mw/Mn
DSC Tg (8C)b
TGA Tonset(8C)c
TGA total loss (%)c
1
2
3
20 000
12 000
12 400
3.4
1.6
1.7
18
19
34
434
432
430
100
86
86
a In THF versus polystyrene.
b 108C/min.
c 108C/min in N2 to 9008C.

D.W. Smith Jr. et al. / Journal of Fluorine Chemistry 104 (2000) 109±117
113
Fig. 3. Optical microscope images of Poly-4 with (left) and without (right) slight shear.
tion techniques [12±13]. Polymer compositions containing
varying amounts of TFVE were obtained based on the
amount of Si±H linkages present in the co-polymer. The
hydrosilated products ranged from highly viscous liquids to
tacky solids depending on the type of pendant attachment
and the amount present. Upon heating, pendant TFVE
groups cyclodimerize giving new silicone and phosphazene
thermosets.
by a simple heat/quench cycle. LC order was con®rmed in
poly-4 by both 2H NMR in the presence of cyclohexane-d12,
and polarized optical microscopy. Birefringent textures
observed by optical microscopy are consistent with a
nematic phase. Slight shear was essential in order to obtain
the patterns, which decay with time (Fig. 3). A complete
characterization of the homoetropic alignment and relaxa-
tion parameters versus surface energy (e.g. Fl incorporation)
is currently underway.
PFCB silicone networks are distinguished from typical
commercial ¯uorosilicones since catalysts and initiators are
not required and crosslinking occurs exclusively at the
¯uorocarbon site. Hybrid thermosets varied in transparency
and texture depending on the amount of crosslinking. Free
standing ®lms exhibited considerable swelling in organic
solvents and FTIR con®rmed the distinct presence of the
2.4. Optical fiber and waveguides
Organic polymer dielectric waveguides are increasingly
attractive alternatives to inorganic components in telecom-
munication devices [30±31]. Polymers offer ¯exibility, low
cost fabrication and connection, high transparency in the
visible and near-infrared spectra, and versatility in structure,
properties, and grades for task speci®c integration, such as
local-area-network applications. Fluoropolymers, in parti-
cular, represent viable alternatives to current optical mate-
rials due to their many complementary properties [1,32±33].
Halogenated polymers in general show negligible transmis-
sion losses in the range desired and ¯uoropolymers repre-
sent the lowest loss examples of organic polymers to date.
DuPont's Te¯onTM AF (Tgˆ160 or 2408C) [3] and Asahi's
CYTOPTM (Tgˆ1088C) [33] are two such amorphous
per¯uoroplastics for which much interest in low loss optics
exist. However, commercial per¯uoropolymers in general
do not exhibit the thermal and thermomechanical stability
required for commercial processes above 2508C. Partially
¯uorinated polymers, such as polyimides [34], and ¯uor-
oacrylate networks [35] have also been reported. Our goal
has been to exploit the well-de®ned polymerization
mechanism offered by PFCB chemistry and prepare co-
polymers with tunable thermal and optical properties [36].
We have found that polymers possessing variable refractive
indices, glass transition temperatures (ranging from 165±
3508C), and long-term 3508C thermal stability [10] have
been accessible by simple choice of comonomer composi-
tion (Table 3).
1
PFCB linkage at 963 cm [12±13].
2.3. Liquid crystalline (LC) PFCB coatings
Due to the well-de®ned nature and unique combination of
properties offered by PFCB chemistry, we have further
chosen to pursue the synthesis of LC ¯uoropolymers based
on a new TFVE monomer containing the mesogenic a-
methylstilbene linkage (4, Scheme 2) [23]. In particular, the
a-methylstilbene mesogen has been used extensively in a
variety of polymeric architectures where most recent atten-
tion has been devoted to LC epoxy thermosets [24±26].
Small molecule liquid crystals [27], as well as main and side
chain LC polymers [28±29], containing ¯uorine are also
known. The synthesis and LC phase ordering of the new
PFCB polymer was studied in an effort to probe the aniso-
tropy of the mesophase and surface orientation as a function
of controlled polymer dimensions. For example, semi¯uori-
nated PFCB liquid crystals may function well as alignment
layers in LC display applications.
Monomer synthesis utilized the traditional phenolic pre-
cursor route and polymerization was accomplished by
simply heating neat monomer 4 in an inert atmosphere at
temperatures above 1508C (Scheme 2). Exothermic cyclo-
dimerization of TFVE groups are typically detected just
above 1308C by DSC (108C/min) as was the case for 4.
Our initial approach to explore mesophase formation
relied on the ability of TFVE polymerization to provide
well-de®ned oligomers with intact reactive terminal groups
Possible cyclopolymerization rate differences for mono-
mers 1 and 2 (Scheme 2) were ®rst investigated by measur-
ing the ole®n consumption during solution polymerization
by 19F NMR spectroscopy. Bulk polymerization kinetics for

114
D.W. Smith Jr. et al. / Journal of Fluorine Chemistry 104 (2000) 109±117
Table 3
Selected properties of PFCB optical polymers
PFCB polymer from
Mn
Mw/Mn(GPC)
Tg (8C)
n (800 nm)e
e (10 kHz)
1
±
±
350c
165c
120
1.495
1.510
2.35
2.41
±
2
50 000a
60 000
8000a/3200b
2
5
1-co-2 (50:50) (pregel 24 h rxn)
1.9
5
1.3±1.4f
220d
1.505
±
a NMR end group analysis.
b GPC versus PS.
c DMS.
d Cured to 3508C by DSC.
e Abbe refractometer.
f Estimated by ellipsometry.
monomer 1 by Raman spectroscopy has also been reported
(vide supra) [5]. Fig. 4 shows the 19F NMR spectra for a pre-
gelled 50 wt.% co-polymer sample (1-co-2) after solution
polymerization in mesitylene (50 wt.%) for 24 h at 1508C.
Endgroup analysis by NMR has proven to be a reliable
asset in PFCB chemistry [4]. Although the three sets of non-
equivalent vinyl ¯uorine signals (ddd near 120, 127,
134 ppm) overlap surprisingly closely, it was possible to
determine from this data and GPC, that the relative rate of
disappearance for the two monomers is essentially the same.
Fig. 5 graphs the ¯uoroole®n conversion versus time for the
solution (50 wt.% mesitylene) homopolymerization of 1, 2,
and a 50 wt.% co-polymerization of the two (1-co-2) by 19F
NMR. From the data in Fig. 5, the rate of cyclopolymeriza-
tion for 1 and 2 monomer is essentially identical up to one
half life at 1508C.
Tgˆ2208C was measured for the co-polymer by DSC
(Table 3).
Polymer samples for optical analysis were prepared by
heating neat monomer or a melt mixed monomer mixture
(50 wt.%) in glass vials under nitrogen at 2008C for 12±14 h
after which near complete conversion of ole®n groups had
been achieved. The PFCB homopolymers and the co-poly-
mer exhibit low loss in the wavelength range suited for
optical waveguide applications (1535±1565 nm). Although
quantitative loss measurements have not been determined
for poly-2 and poly(1-co-2), an attenuation of 0.25 dB/cm at
1515±1565 nm has been reported for the poly-1 [17]. The
relative loss data presented here predicts that a similar value
may be obtained for the 50 wt.% co-polymer and a slightly
higher loss for the poly-2 and the co-polymers.
The dependence of refractive index on wavelength has
been determined for selected homopolymers and co-poly-
mers. As shown in Fig. 6, co-polymerization can be used to
tune exact optical properties by monomer composition. In
addition, we have recently developed a new TFVE mono-
mer prepared from 4,40-(hexa¯uoroisopropylidene)diphenol
(5 in Scheme 2) [37]. The increased degree of ¯uorination in
Integration of the multiple non-equivalent cyclobutane
¯uorines versus the clearly resolved vinyl ¯uorine, Fa or Fc,
gave a nˆ17 or Mnˆ8000 for the ideal linear structure
shown in Fig. 4. In contrast, Mnˆ3200 and Mw/Mnˆ5 was
measured by GPC (versus PS) which indicates substantial
branching has occurred. After heating to 3508C, a
Fig. 4. 19F NMR spectra of Poly(1-co-2) (50 wt.% in mesitylene after solution polymerization for 24 h at 1508C, calculated nˆ17, Mnˆ8000) and idealized
linear structure of the pre-gelled branched oligomer.

D.W. Smith Jr. et al. / Journal of Fluorine Chemistry 104 (2000) 109±117
115
Fig. 5. TFVE conversion vs. time for solution (50 wt.% mesitylene) cyclopolymerization at 1508C by 19F NMR.
a Bruker AC-200 spectrometer. Chloroform-d was used as
solvent, and chemical shifts reported are internally refer-
enced to tetramethylsilane (0 ppm), CDCl3 (77 ppm), and
1
CFCl3 (0 ppm) for H, 13C, and 19F nuclei, respectively.
Infrared analyses were performed on neat oils or free
standing ®lms using a Nicolet 550 Magna FTIR spectro-
photometer. Gas chromatography/mass spectrometry (GC/
MS) data were obtained from a Varian Saturn GC/MS. Gel
permeation chromatography (GPC) data were collected in
THF using a Waters 2690 Alliance System with refractive
index detection at 358C, and equipped with two consecutive
Polymer Labs PLGel 5 mm Mixed-D and Mixed-E col-
umns. Retention times were calibrated against Polymer
Labs Easical PS-2 polystyrene standards. DSC data were
obtained from a Mettler±Toledo 820 System under a nitro-
gen atmosphere at a scan-rate of 108C/min. Polarized
optical microscopy was performed at room temperature
using an Omega BX60 optical microscope equipped with
a Mettler-Toledo Hot Stage with a magni®cation of 50x.
Raman spectroscopy was performed on polymer samples
heated at 1508C in 5 mm NMR tubes [5]. Kinetic data were
collected using an ISA U-1000 double monochromator. The
532 nm line of a Coherent Radiation DPSS-200 YAG laser
was used as the excitation source. Raman Scattering was
collected using a 908 scattering geometry and collected with
a charged coupled device. Monomers and polymers were
prepared and characterized as generally described for 4 as
follows.
Fig. 6. Refractive index vs. wavelength for selected PFCB polymers.
poly-5 (Tgˆ1208C) causes a dramatic decrease in the
refractive index Ð well below that previously detected
for any PFCB polymer (Table 3). These results indicate
that a wide range of materials with tailorable optical and
thermal properties can be prepared by judicious choice of
comonomer.
3. Experimental
3.1. General information
3.2. 4,40-Bis(2-bromotetrafluoroethoxy)-a-methylstilbene
Bis-phenols were generously provided by the Dow Che-
mical. Other chemicals and reagents were purchased from
Aldrich or Fisher Scienti®c and used as received unless
otherwise stated. 1H NMR 300 MHz and proton-decoupled
13C NMR 75 MHz were obtained on a Bruker AC-300
spectrometer. 19F NMR 188 MHz data were obtained on
To a 3 l vessel equipped with a Dean±Stark azeotropic
distillation assembly were added 1.1 l solution of DMSO,
KOH (49.6 g, 0.88 mol), 4,40-dihydroxy-a-methylstilbene
(100 g, 0.44 mol) and 0.35 l of m-xylene. Water was
removed to less than 500 ppm by azeotropic vacuum

116
D.W. Smith Jr. et al. / Journal of Fluorine Chemistry 104 (2000) 109±117
1
distillation at 1008C over 6 h. The vessel was cooled to 208C
and dibromotetra¯uoroethane (252.5 g, 0.97 mol) was
added dropwise over 3 h and temperature was kept below
108C throughout the addition. The solution was stirred for
12 h, warmed to room temperature, extracted into 500 ml of
hexanes, washed with 3 mlÂ100 ml H2O, and dried over
MgSO4. The hexane solution was then ¯ashed over alumina
and evaporated giving 1 in 50% yield. 1H NMR (200 MHz,
CDCl3) d 2.26 (3H, s, trans Me), 6.79 (1H, s, trans Hv), 7.24
(4 H, m), 7.37 (2H, d, Jˆ8 Hz), 7.52 (2H, d, Jˆ8 Hz); 13C
NMR (75 MHz, CDCl3) 17.2 (trans Me), 108.7 (m, CF2),
112.9 (m, CF2), 115.9 (m, CF2), 116.4, (m, CF2), 121.4,
126.8, 130.3, 136.6, 137.0, 142.2, 147.3, 148.0; 19F NMR
temperature. H NMR (200 MHz, CDCl3) d 2.16 (s, cis
Me), 2.21 (s, trans Me), 6.42 (s), 6.73 (s), 6.86 (br, m), 7.13
(br, d), 7.21 (br, d), 7.29 (br, d), 7.45 (br, d), integration ratio
of Me to sum of aromatic and vinylic protons is 3±10; 13C
NMR (75 MHz, CDCl3) 17.4, 26.6, 105.9 (m, cyclobutyl-
F6), 109.3 (m, cyclobutyl-F6), 112.7 (m, cyclobutyl-F6),
115.4, 117.9, 118.2, 121.4, 125.9, 126.6, 127.2, 127.5,
128.4, 129.6, 134.9, 137.8, 140.6, 140.9, 151.0, 151.7;
19F NMR (188 MHz, CDCl3) d tri¯uorovinyl endgroups
at 119 (2F, dd, cis-CF=CF2, Fa), 126 (2F, dd, trans-
CF=CF2 Fb), (2F, dd, CF=CF2, Fc) (Jabˆ97 Hz, Jacˆ58 Hz,
Jbcˆ110); broad cyclobutyl-F6 at 127.2, 127.6, 128.3,
128.4, 129.0, 129.6, 130.2, 130.5, 130.8 (total
cyclobutyl-F); FTIR (neat): n 3021, 1608, 1515, 1319, 1217,
970 (cyclobutyl-F6), 775. 19F NMR endgroup analysis for
bulk polymerization (1508C for 3 h, 2008C for 12 h) gave
91% ole®n conversion and avg. DP (n)ˆ10, or Mnˆ4200.
GPC analysis found Mw>110,000 and Mw/Mn>20. Repro-
ducible glass transition temperatures are not observed
for soluble oligomers, yet heating to 3008C resulted in
an insoluble ®lm with a DSC (108C/min) measured
Tgˆ1108C.
‡
(188 MHz, CDCl3) d 66.24 (t), 86.60 (t); GC/MS (M
Calc. for C19H12Br2F8O2 584.31 g/mol), Obsv. M 584 with
distinctive dibromo isotopic mass ratio.
‡
3.3. 4,40-Bis(trifluorovinyloxy)-a-methylstilbene (4)
To a dry nitrogen purged 1 l vessel was added 1 (129.13 g,
0.22 mol) dropwise to 0.1 l of anhydrous acetonitrile and
30.0 g (0.46 mol) of Zn (30 mesh) at 808C. The solution was
re¯uxed for 18 h and the acetonitrile was removed from the
product/salt mixture in vacuo. The product was removed
from the by-product salts by extraction into hexanes and
¯ashed over alumina 3x giving 2 as a clear colorless liquid
which crystallized upon standing in 58% yield, mp (DSC)
178C; 1H NMR (200 MHz, CDCl3) d 2.17 (trace, s, cis Me),
2.23 (3H, s, trans Me), 6.78 (1H, s, trans Hv), 7.08 (2H, d,
Jˆ3.2 Hz), 7.09 (2H, d, Jˆ3.2 Hz), 7.32 (2H, d, Jˆ8.7 Hz),
7.48 (2H, d, 8.7 Hz); 13C NMR (75 MHz, CDCl3), 17.6
(trans Me), 26.8, 115.8, 126.4, 127.5, 130.7, 135.0 (ddd,
CF=CF2, 1Jˆ110 Hz, 2Jˆ156, 90 Hz), 135.5, 136.2, 147.0
Acknowledgements
We thank Clemson University, The Dow Chemical Com-
pany, 3M Corporation for supporting this work, and Mettler-
Toledo for donation of a DSC820 system to Clemson
University. Prof. R. Neilson, Dr. S. Narayan-Sarathy, and
Dr. J. Ji at Texas Christian University, and Dr. J. Laanne and
S.-N. Lee at Texas A & M University for collaboration in
part. We also acknowledge R.V. Snelgrove, Dr. K. Clement,
Dr. C.M. Cheatham D., and Dr. B. Ezzell at the Dow
Chemical Company for their pioneering efforts in PFCB
chemistry.
1
2
(ddd, CF=CF2, Jˆ62, 276 Hz, Jˆ273 Hz), 153.7, 154.5;
19F NMR (188 MHz, CDCl3) d 120 (2F, dd, cis-CF=CF2,
Fa), 127 (2F, dd, trans-CF=CF2, Fb), 134 (2F, dd,
CF=CF2, Fc) (Jabˆ97 Hz, Jacˆ58 Hz, Jbcˆ110); FTIR
(neat): n 3047, 1898 (w), 1830 (w, CF=CF2), 1609, 1514,
1302 (st, br), 1166 (st, br), 843 (w); GC/MS (Calc. for
References
‡
C19H12F6O2 386 g/mol), Obsv. m/z: 386 (M ), 191, 165,
152, 115, 89; Anal. Calc. for C19H12F6O2 (found): C, 59.07
(58.91); H, 3.14 (3.14).
[1] A.E. Feiring, in: R.E. Banks, B.E. Smart, J.C. Tatlow (Eds.),
Organofluorine Chemistry: Principles and Commercial Applications,
Plenum Press, New York, 1994, p. 339.
[2] M. Yamabe, in: R.E. Banks, B.E. Smart, J.C. Tatlow (Eds.),
Organofluorine Chemistry: Principles and Commercial Applications,
Plenum Press, New York, 1994, p. 397.
3.4. PFCB polymer from 4
To a 50 ml reactor equipped with mechanical stir and
nitrogen purge was added 2.0 g of 4. The neat monomer was
degassed via nitrogen sparge and heated at 1608C for 2 h
during which time the viscosity increased dramatically and
stirring was hindered. Alternatively, polymerization of
monomer 2 was accomplished by heating ca. 1.0 g of the
solid in an aluminum boat or glass tube. The THF soluble
polymer was obtained as a clear solid in essentially quanti-
[3] P.R. Resnick, W.H. Buck, in: J. Scheirs (Ed.), Modern Fluoropoly-
mers, Wiley, New York, 1997, p. 397.
[4] D.W. Smith Jr., D.A. Babb, Macromolecules 29 (1996) 852.
[5] C.M. Cheatham, S-N. Lee, J. Laane, D. Babb, D.W. Smith Jr.,
Polym. Int. 46 (1998) 320.
[6] D.A. Babb, H. Boone, D.W. Smith Jr., Rudolf J. Appl. Polym. Sci. 69
(1998) 2005.
[7] J. Ji, S. Narayan, R. Neilson, J. Oxley, D. Babb, N. Rondan, D.W.
Smith Jr., Organometallics 17 (1998) 783.
[8] D.A. Babb, B.R. Ezzell, K.S. Clement, W.F. Richey, A.P.J. Kennedy,
Polym. Sci.: Part A: Polym. Chem. 31 (1993) 3465.
[9] A.P. Kennedy, D.A. Babb, J.N. Bremmer, A.J.J. Pasztor Jr., Polym.
Sci.: Part A: Polym. Chem. 33 (1995) 1859.
2
tative yield. For H NMR and polarized microscopy sam-
ples, neat monomer was heated under nitrogen in a 10 mm
NMR tube at 1558C with periodic quenching to room

D.W. Smith Jr. et al. / Journal of Fluorine Chemistry 104 (2000) 109±117
117
[10] D. Babb, V. Snelgrove, D.W. Smith, Jr., S. Mudrich, in: J. Hedrick, J.
Labadie (Eds.), Step-Growth Polymers for High Performance
Materials, ACS Sym. Ser., No. 624, American Chemical Society,
Washington, DC, 1996.
[22] B. Atkinson, A.B. Trentwith, J. Chem. Soc. (1953) 2082.
[23] D.W. Smith Jr., H.W. Boone, R. Traiphol, H. Shah, D. Perahia,
Macromolecules 33 (4) (2000) 1126.
[24] C. Ortiz, R. Kim, E. Rodighiero, C.K. Ober, E.J. Kramer,
Macromolecules 31 (1998) 4074.
[11] H.W. Boone, D.W. Smith Jr., D.A. Babb, Polym. Prepr. Am. Chem.
Soc. Div. Polym. Chem. 39 (2) (1998) 812.
[25] B.C. Beniceicz, M.E. Smith, J.D. Earls, R.D. Priester, S.M.
Setz, R.S. Duran, E.P. Douglas, Macromolecules 31 (1998)
4730.
[12] S. Narayan, J. Ji, R.H. Neilson, D.W. Smith Jr., Polym. Prepr. (Am.
Chem. Soc. Div. Polym. Chem.) 39 (1) (1998) 530.
[13] S. Narayan, R.H. Neilson, D.W. Smith Jr., Polym. Prepr. (Am. Chem.
Soc. Div. Polym. Chem.) 39 (1) (1998) 609.
[26] J. Liu, C. Wang, G. Campbell, J. Earls, R. Priester, J. Polym. Sci.:
Polym. Chem. 36 (1998) 1457.
[14] R.H. Neilson, J. Ji, S. Narayan-Sarathy, D.W. Smith Jr., D.A. Babb,
Phosphorous, Sulfur and Silicon 144 (1999) 221.
[27] T. Inoi, in: R.E. Banks, B.E. Smart, J.C. Tatlow (Eds.), Organo-
fluorine Chemistry: Principles and Commercial Applications,
Plenum Press, New York, 1994, p. 263.
[15] D.J. Perettie, L.D. Bratton, J.N. Bremmer, D.A. Babb, in:
Proceedings of International Symposium on Electronic Imaging:
Science and Technology, IS&T/SPIE, 1993, pp.1911±03.
[16] C.W. Kim, C.O. Jeong, S.K. Rho, J.H. Lee, H.R. Nam, J.H. Choi,
M.P. Hong, S.Y. Kim, H.G. Yang, J.H. Souk, in: Proceedings of
Symposium on Society for Information Display (SID), Display
Manufacturing Technology Conference, January 1997.
[17] G. Fishbeck, R. Moosburger, C. Kostrzewa, A. Achen, K. Petermann,
Electron. Lett. 33 (6) (1997) 518.
[28] J.C. Wittmann, S. Meyer, P. Damman, M. Dosiere, H-W. Schmidt,
Polymer 39 (15) (1998) 3545.
[29] M. Sato, S. Ujiie, Y. Tada, Y. Goto, J. Polym. Sci. Part A: Polym.
Chem. 36 (1998) 1243.
[30] B.L. Booth, J. Lightwave Technol. 7 (10) (1989) 1445.
[31] A.C. Levy, C. Taylero, in: J. Kroschwitz (Ed.), Encycl. Polym. Sci.
and Eng, Vol. 7, Wiley, New York, 1987, pp. 1±15.
[32] T. Kaino, J. Polym. Sci. Part A 25 (1987) 37.
[33] N. Sugiyama, in: J. Scheirs (Ed.), Modern Fluoropolymers, Wiley,
New York, 1997, p. 541.
[18] D.J. Perettie, D.A. Babb, Q. Chen, J.H. Judy, in: Proceedings of
Symposium on IEEE Magnetics, MMM-Intermag Conference, 1994,
DQ-08.
[34] T. Matsuura, S. Ando, S. Sasaki, F. Yamamoto, Electron. Lett. 29 (3)
(1993) 269.
[19] T.A. Tumolillo Jr., A. Thomas, P.R. Ashley, Appl. Phys. Lett. 82 (24)
(1993) 3068.
[35] A. Hale, K. Quoi, D. DiGiovanni, Polym. Prepr. (Am. Chem. Soc.
Div. Polym. Chem.) 39 (2) (1998) 978.
[20] P. Townsend, E. Shaffer, M. Mills, J. Blackson, M. Radler, in: A.
Lagendijk, H. Treichel, K. Uram, A. Jones (Eds.), Low-Dielectric
Constant Materials II, Mat. Res. Soc. 443 (1996) 35.
[36] H. Shah, A. Hoeglund, M. Radler, C. Langhoff, D.W. Smith Jr.,
Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 40 (2) (1999)
1293.
[21] B.E. Smart, in: R.E. Banks, B.E. Smart, J.C. Tatlow, (Eds.),
Organofluorine Chemistry: Principles and Commercial Applications,
Plenum Press, New York, 1994, p. 73.
[37] D.W. Smith, Jr., H. Shah, B. Johnson, Polym. Prepr. (Am. Chem.
Soc. Div. Polym. Chem.) 41(1) (1998), in press.