Synthesis of fluorinated and hydrocarbon ester functionalized poly(p-phenylenes) and their solubility in supercritical fluids

Article abstract of DOI:10.1021/ma021615e

Monomers of the structure 2,5-dichloro-1,4-(RO₂C)₂C)₂benzene, where R = CH₂CF₂CF₃ (1a), -(CH₂)CH₃ (1b). -CH₂CH₂(CF₂)₅CF

Full text of DOI:10.1021/ma021615e

2242
Macromolecules 2003, 36, 2242-2247
Synthesis of Fluorinated and Hydrocarbon Ester Functionalized
Poly(p-phenylenes) and Their Solubility in Supercritical Fluids
Mich a el E. Wr igh t,*,† Kim ber ly M. Lott,‡ Ma r k A. McHu gh ,*,‡ a n d Zh ih a o Sh en ‡
Chemistry & Materials Division, NAWCWD, China Lake, California 93555-6100, and Departments of
Chemistry and Chemical Engineering, Virginia Commonwealth University, Richmond, Virginia 23284
Received October 23, 2002; Revised Manuscript Received J anuary 28, 2003
ABSTRACT: Monomers of the structure 2,5-dichloro-1,4-(RO2C)2benzene, where R ) CH2CF2CF3 (1a ),
-(CH2)2CH3 (1b), -CH2CH2(CF2)5CF3 (1c), and -(CH2)7CH3 (1d ), were prepared and characterized in
the study. Polymerization by the Ni-catalyzed/Zn-mediated homocoupling methodology afforded poly(p-
phenylenes) 2a -2d . The polymers obtained after precipitation into methanol showed molecular weights
(Mn, relative to polystyrene standards) in the range 7000-13 000. The regiochemistry in the polymer
was deduced to be approximately 70% head-to-tail with the remainder head-to-head. The polymers
displayed excellent solubility in common polar organic solvents such at THF and chloroform. The poly-
(p-phenylenes) 2a and 2b showed no solubility in supercritical ethane, propane, butane, and carbon dioxide
even at 190 °C and 35 000 psia. Polymer 2d was remarkably soluble in the supercritical hydrocarbon
solvents at low pressures (1000 psia). However, 2d does not dissolve in supercritical carbon dioxide. The
fluorinated poly(p-phenylene) 2c dissolved in the supercritical hydrocarbons at pressures below 6000
psia, and it dissolved in carbon dioxide at pressures as low as 1000 psia at 25 °C.
In tr od u ction
liquid organic solvents, which suggests that it may be
possible to dissolve and process these polymers in
supercritical fluid (SCF) solvents. Several studies have
been reported in the past decade or so on the thermo-
dynamics and solution behavior of functionalized, rod-
like polymers7,8 which provide a template to interpret
the polymer-SCF studies reported in the present work.
For example, Ballauff reported that polymer crystal-
lization and the possible formation of mesophases of
rodlike polymers can be inhibited by a substituent group
which acts as a tethered solvent that screens interchain
interactions and main chain-solvent interactions and
that increases the entropy of dissolution.7 Molecular
dynamics studies may provide further insight into the
role of side chain architecture on the characteristic ratio
and persistence length of PPP-type polymers.9 It is also
important to note that the regioregularity of the main
chain has been postulated to affect polymer backbone
flexibility that in turn could influence solubility.9 Hence,
the chemical architecture of the side chains and the
regioregularity of the main chain are both expected to
have a strong influence on the pressures and temper-
atures needed to dissolve substituted PPPs in SCF
solvents.
For the past several decades polymer chemists have
been interested in the unique properties of poly(p-
phenylene) (PPP)-based materials that are a conse-
quence of the “linear and rigid backbone” chain archi-
tecture of PPP.1 The outstanding thermal stability of
PPP motivated much of the early research efforts to
focus on the synthesis of processable derivatives of PPP.2
Within the past decade, there has been increased
attention directed at designing highly functionalized
and processable poly(phenylenes) due to the many high-
performance applications proposed for these type of
polymers.3 Functionalized poly(phenylenes) are now
readily synthesized in high yields and often as high
molecular weight materials by homo- or cross-coupling
reactions that are typically nickel- or palladium-
catalyzed.4 Functionalization of the PPP backbone is
necessary to ensure polymer solubility for synthesis as
well as for the later processing of these materials.
However, the backbone functional groups can be chemi-
cally modified or even completely removed at a later
stage. For example, Kaeriyama and co-workers gener-
ated an ester functionalized PPP from the nickel-
catalyzed/zinc-mediated homo-coupling of methyl 2,5-
dichlorobenzoate,5 isolated the PPP, hydrolyzed the
ester groups, and then decarboxylated the polymer to
produce unsubstituted PPP. Other ester-containing
PPPs have been reported in the patent literature6 which
suggests that PPP-ester-based materials will find
greater use in high-performance applications as creative
synthesis/solution processing schemes are developed.
From a fundamental viewpoint, substituted PPPs
represent the quintessential architecture to delineate
the influence of a substituent group on solubility since
unsubstituted PPPs are stiff chain, refractory polymers
that do not dissolve in organic liquid solvents. Func-
tionalized PPPs can be very soluble in conventional
The present study reports the synthesis of new,
fluorinated and nonfluorinated, propyl and octyl ester
functionalized PPPs. The role of the ester side chain on
the solubility of derivatized PPPs is investigated as a
function of the alkoxy tail length and its fluorine
content. The polarity of the ester group will remain
relatively constant in the examples studied. The super-
critical fluids of interest are CO2, ethane, propane, and
butane. In general, the level of polymer solubility in
supercritical CO2 is sensitive to polarity and backbone
flexibility.10,11 It is established that CO2 does not dis-
solve nonpolar polyolefins. Hence, fluorinating the
alkoxy tail is anticipated to impact solubility for the
poly(p-phenylenes) on the basis of the observations
made with other more flexible polymers.12,13 Supercriti-
cal alkanes are chosen in this study since these SCF
† NAWCWD.
‡ Virginia Commonwealth University.
10.1021/ma021615e CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/27/2003

Macromolecules, Vol. 36, No. 7, 2003
Synthesis of Poly(p-phenylenes) 2243
Sch em e 1
Sch em e 2
F igu r e 1. Thermal gravimetric analyses of polymer 2c were
carried out under a nitrogen atmosphere. The isothermal aging
study was performed at 250 °C with the x-axis in minutes,
and for the ramp TGA run (10 °C/min) the axis represents
temperature (°C).
1
two signals are observed in the H NMR spectrum that
can be assigned to end cap. One peak at δ 3.95 ppm is
assigned to the anticipated end cap, namely methyl
benzoate. The second verifiable end cap is assigned to
the octyl-ester resonance at δ 4.30 ppm. This latter end
cap must arise from a coupling reaction at the 5-chloro
position and then either leaving the 2-position unreacted
or possibly terminated by reduction (i.e., catalytic
replacement chloride by hydrogen). From the integra-
tion values we estimate that ∼50% of the end groups
result from addition of the 3-chlorobenzoate and the
reminder belong to reactions involving the monomer and
possibly phenyl transfer from the triphenylphosphine
ligand during the catalytic process.17
Polymers 2a -2d were subjected to thermal analysis
using both DSC and TGA instruments. Even though
reasonably sized samples (∼5 mg) of the polymers can
be loaded into DSC sample pans, consistent Tg events
were not observed. Although small thermal events were
observed, especially on the original scan, reproducible
(third and fourth scans) events of sufficient magnitude
were absent in the scans from -50 to 250 °C. The
thermal stability of the polymers was tested using TGA
operating in both the isothermal and ramp modes
(Figure 1). We find that at 250 °C under a nitrogen
atmosphere the polymers show less than ∼5% weight
loss over a period of 10 h. A decomposition breakpoint
of ∼350 °C was observed when a sample of 2c was
heated at 10 °C/min under a nitrogen atmosphere. It is
interesting to note that no “char residue” remains after
the temperature hits ∼570 °C.
solvents readily dissolve nonpolar and slightly polar
polyolefins.14 In this case, the length of the alkoxy tail
should again prove to be a controlling variable for
solubility since the polarity per molar volume of a
molecule decreases as this molar volume increases.15
Resu lts a n d Discu ssion
The alkyl and fluoroalkyl benzoate monomers were
synthesized through two different routes (Scheme 1).
The alkyl benzoate monomers 1b and 1d were synthe-
sized using 2,5-dichlorobenzoic acid and the correspond-
ing alcohol in a Fischer esterification reaction. The
fluoroalkyl benzoate monomers were not as easy to
synthesize. Low yields were obtained from the Fischer
esterification procedures. In addition, treatment of 2,5-
dichlorbenzoyl chloride with the fluoro alcohols resulted
in complicated product mixtures. Use of dicyclohexyl-
carbodiimide (DCC) also resulted in a mixture where
purification from the resulting urea byproduct proved
cumbersome. By far, we achieved our best results using
carbonyldiimidazole (CDI) as the coupling agent.16 The
gaseous and water-soluble products (i.e., imidazole) are
easily separated from the desired monomer. Isolated
yields of the fluoroalkyl monomers were generally above
70% using this procedure.
We have carried out a series of experiments where
the octyl 2,5-dichlorobenzoate was polymerized in the
presence of varying amounts of end cap. The experi-
ments follow the expected trend by affording a decrease
in the Mn (number-average molecular weight) as the mol
% of end cap added to the reaction mixture is increased
(Figure 2).
We have also carried out experiments to determine
the relationship of Mn as a function of reaction temper-
ature, and the results are depicted in Figure 3. The end
cap was maintained at 10 mol % for each temperature
studied. As the temperature increased from 70 to 80 °C,
there is a notable increase in the Mn even though each
reaction went to completion. At 90 °C a slight decrease
in Mn occurs. This is possibly due to catalyst decomposi-
tion or to an increase in the rate of competing side
reactions such as reduction of the aryl-chloride bond
or phenyl transfer from the phosphine ligand. Thus, 80
°C represents an optimized reaction temperature for this
monomer and catalyst combination.
The polymers used in this study were synthesized as
shown in Scheme 2 using the Ni-catalyzed homo-
coupling chemistry developed by Colon and Kesley.17 We
add a step where the reaction mixture was washed with
an aqueous sodium cyanide solution.18 This was imple-
mented to remove any possible remaining homogeneous
nickel complexes that might later become entrapped in
a precipitation process. The final polymeric materials
after precipitation and drying were fully characterized
by analytical and spectroscopic methods. The amount
of end cap added to the polymerization reaction was
used to help tune the polymer’s molecular weight,
leading to the four homologous polymers 2a -2d . Poly-
mer molecular weights were determined by SEC analy-
sis and are reported relative to polystyrene standards.19
In addition to SEC analysis, the polymer structure was
verified and analyzed on the basis of end-group analysis
by proton NMR spectroscopy.
1
The H NMR spectrum of polymer 2d indicates that
the head-to-tail (H-T) and head-to-head (H-H) repeat-
ing units are in an approximate ratio of 2. Furthermore,
The poly(p-phenylenes) with fluorinated and non-
fluorinated propyl chains (2a and 2b) do not dissolve

2244 Wright et al.
Macromolecules, Vol. 36, No. 7, 2003
F igu r e 2. Polymerization of octyl 2,5-dichlorobenzoate as in
Scheme 1 but with varying amounts of end cap added to adjust
polymer 2d molecular weight.
F igu r e 5. Comparison of the phase boundary curves for 1.3
wt % poly(p-phenylene) 2c in supercritical ethane, propane,
butane, and CO2. A single phase exists at conditions above
each of the curves and two phases exist below the curves. The
ethane, propane, and CO2 curves represent cloud points.
However, the entire butane curve represents bubble points
that are very close to the vapor pressure of pure butane.
∼16 000 psia in ethane, to 7000 psia in propane, to 2000
psia in butane. Dispersion interactions, which roughly
scale with polarizability, R,15 increase as the size of the
SCF solvent increases from ethane (R ) 4.4 Å3) to
propane (R ) 6.3 Å3) to butane (R ) 8.1 Å3), which
follows the decrease in cloud-point pressures observed
in Figure 4. Note that for the butane system the
transitions at 60 °C and lower are bubble points that
superimpose with the butane vapor pressure curve.
Hence, this poly(p-phenylene), 2d , readily dissolves in
butane at room conditions and pressures less than 100
psia. To precipitate the poly(p-phenylene) from solution,
it is necessary to increase the temperature above 60 °C
to decrease the density of butane which decreases its
solvent power. The phase behavior shown in Figure 4
demonstrates that the octyl side chain on the poly(p-
phenylene) has a significant effect on the solubility
behavior of this polymer.
F igu r e 3. Polymerization of octyl 2,5-dichlorobenzoate carried
out at various temperatures. Product 2d was analyzed by GPC
and NMR spectroscopy.
If the octyl side chain is fluorinated, it takes less
pressure to dissolve the resulting poly(p-phenylene) in
supercritical ethane, propane, and butane, and the
fluorinated polymer 2c readily dissolves in CO2. Figure
5 shows that the fluorooctyl poly(p-phenylene)-ethane
cloud-point curve, which is located at ∼5000-12 000
psia lower pressures compared to the nonfluorinated
octyl poly(p-phenylene) curve, increases in pressure as
the temperature is reduced below 60 °C. The fluorooctyl
poly(p-phenylene)-propane cloud-point curve exhibits
similar behavior, but the propane curve is located at
pressures as low as 800 psia. Note that the fluorooctyl
poly(p-phenylene)-butane cloud-point curve essentially
superposes onto the vapor pressure curve of pure butane
and that the transitions are bubble points rather than
fluid to liquid + liquid transitions.
It is possible to posit several plausible explanations
for the large differences in the solubility of the octyl and
fluorooctyl ester poly(p-phenylenes) in ethane, propane,
and butane. The large fluorooctyl ester side chains are
expected to inhibit chain-chain interactions since CF2
and CF3 groups exhibit very weak intermolecular
interactions. Hence, chain aggregation is expected to be
less in the fluorooctyl PPP-alkane mixtures compared
to the octyl PPP-alkane mixtures. However, it is not
readily apparent whether the fluorinated octyl ester side
chains adopt an orientation favoring interactions with
the aromatic backbone or with the alkane solvent. On
the basis of molecular mechanics calculations, Vaia et
al. suggested that flexibility of the PPP main chain is a
direct function of side group orientation and size.20 It
F igu r e 4. Comparison of the phase boundary curves for 1.3
wt % poly(p-phenylene) 2d in supercritical ethane, propane,
and butane. A single phase exists at conditions above each of
the curves and two phases exist below the curves. The ethane
and propane curves represent cloud points. However, the
butane curve represents cloud points (open circles) at temper-
atures greater than 60 °C and bubble points (filled circles) at
lower temperatures. The butane bubble-point pressures are
very close to the vapor pressure of pure butane.
in supercritical CO2, ethane, propane, or butane even
at 190 °C and 35 000 psia. Evidently, the alkoxy chains
are too short to screen interchain- and main chain-
solvent interactions and are not bulky enough to induce
chain flexibility as a consequence of steric consider-
ations.8 Even when the propyl chains are fluorinated,
the side chain-solvent interactions are not large enough
to induce dissolution of these rigid-rod polymers in SCF
solvents.
Figure 4 shows that poly(p-phenylene) 2d readily
dissolves in ethane, propane, and butane, but it does
not dissolve in CO2 even at 135 °C and 35 000 psia.
Since ethane, propane, and butane are nonpolar sol-
vents, dispersion interactions are likely to be the
dominant type of interaction15 that fixes the location of
the cloud-point curves shown in Figure 4. For example,
at 100 °C, the cloud-point pressures decrease from

Macromolecules, Vol. 36, No. 7, 2003
Synthesis of Poly(p-phenylenes) 2245
follows that the enhanced solubility of the fluorooctyl
ester PPP in these alkane solvents could be a result of
enhanced backbone distortion relative to the octyl ester
PPP-alkane systems. At present, it is not possible to
promote unequivocally one explanation over another for
the differences in observed phase behavior for these
octyl ester PPP-alkane mixtures, which certainly war-
rants further exploration.
Figure 5 also shows that fluorooctyl poly(p-phenylene)
2c is extremely soluble in CO2 as it only takes ∼1000
psia at 25 °C to dissolve this polymer. The fluorooctyl
poly(p-phenylene)-CO2 cloud-point curve exhibits a
positive slope and is located at pressures between 1000
and 5000 psia at temperatures from 20 to 120 °C,
respectively. At temperatures near 100 °C where dis-
persion interactions should be the dominate force of
attraction, the ethane and CO2 curves virtually overlap
even though the polarizability of ethane, 4.4 Å3, is much
larger than that of CO2, 2.7 Å3. However, Kazarian and
co-workers have shown that CO2 forms a CO2-ester
complex that is expected to increase in strength as the
temperature decreases.21 Hence, the CO2-ester complex
provides the necessary interaction that makes polymer
2c much more soluble in CO2 as compared to ethane at
temperatures below ∼115 °C and especially as the
temperature is reduced below 50 °C. It is important to
note that the impact of this CO2-ester complex is
magnified at low temperatures because the solution
density is very high in this region. The shape of the
fluorooctyl poly(p-phenylene)-propane cloud-point curve
in Figure 5 suggests that below 45 °C the pressures
needed to dissolve fluorooctyl poly(p-phenylene) (2c) in
propane increase sharply, which means that CO2 is now
a better solvent than propane. At low temperatures,
polymer 2c drops out of ethane and propane even as
the pressure is increased rapidly since ester-ester polar
interactions increase relative to ester-alkane solvent
interactions and since the alkane solvents do not
complex with the ester group. Figure 5 shows that
butane is a much better solvent than CO2 for polymer
2c probably due to the high polarizability of butane and
to the liquidlike density of butane at temperatures from
60 to 125 °C.
F igu r e 6. Comparison of the phase behavior of fluorooctyl
poly(p-phenylene) (2c) (filled squares) obtained in this study
with that of poly(1H,1H,2H,2H-tetrahydroperfluorodecyl meth-
acrylate, filled circles),13 poly(1H,1H,2H,2H-tetrahydroper-
fluorodecyl acrylate, open squares),22 and poly(1H,1H-dihydro-
perfluorooctyl acrylate, open circles) in supercritical CO2.23 The
polymer concentrations are between 1.0 and 4.0 wt %.
suggests that the side chain on the poly(p-phenylene)
acts as a tethered solvent that fixes the location of the
curve. At temperatures greater than ∼80 °C, the fluoro-
methacrylate curve also superposes with the other three
curves although below 80 °C the fluoromethacrylate
curve no longer superposes with the other curves, rather
it exhibits a zero slope at higher pressures. At “cold”
temperatures, in high-density CO2, the CO2-ester
complex should have a large effect on the location of
the cloud-point curve. The fluoroacrylate and fluoro-
methacrylate curves diverge from one another since
the methyl group on a methacrylate R-carbon more
effectively repels CO2 due to both steric hindrance and
repulsive interactions than does the hydrogen on an
acrylate R-carbon. Hence, CO2 has more restricted
access to the ester group in the methacrylate polymer
than in the acrylate polymer. In fact, Rindfleisch and
co-workers have shown that, at room temperature, poly-
(methyl acrylate) (PMA) dissolves in CO2 at a pressure
that is more than 24 000 psia greater than that needed
to dissolve poly(vinyl acetate) (PVAc) due to the easier
access CO2 has to the ester group in PVAc compared to
PMA.10 At temperatures greater than ∼80 °C, where
the strength of the CO2-ester complex is weaker, all of
the cloud-point curves superpose, suggesting that the
phase behavior is governed primarily by nonspecific
dispersion interactions. The phase behavior in Figure
6 shows that CO2 is a poor quality solvent that responds
to modest variations in the chemical architecture of
the polymer that changes the strength or type of inter-
molecular interactions.
As previously mentioned, chain-chain interactions
are expected to be reduced by the large fluorooctyl ester
groups; however, chain-chain interactions are now
expected to be even less in CO2 since CO2-fluoroalkane
attractive interactions are stronger than CO2-alkane
interactions as surmised from phase behavior studies.12
CO2 interacts with the ester group on the side chain,
and CO2-aromatic, quadrupole-quadrupole interac-
tions are favored at low temperatures.15 Hence, chain
aggregation is anticipated to be less with CO2 than with
the alkane solvents. It is still not possible to ascertain
whether the enhanced solubility of the fluorooctyl ester
PPP in CO2 is a result of increased chain flexibility
without performing complementary light scattering
studies.
Con clu sion s
From a synthetic point of view, a series of fluorinated
and nonfluorinated ester substituted PPPs were suc-
cessfully prepared and fully characterized. It is clear
from a comparison of NMR spectral and SEC data for
polymers 2a -2d that we can tailor molecular weight
through the addition of varying amounts of end cap;
however, there remains a substantial amount of poly-
mer terminus that is not due to the added end cap. Work
is continuing to optimize synthetic procedures in order
to attain greater control over end-group composition in
the new fluorinated and nonfluorinated PPPs.
The results from the present study also demonstrate
that the derivatized PPP polymers can be dissolved in
SCF solvents at modest operating conditions which
provides the opportunity for SCF-assisted processing
schemes for these novel materials. For example, CO2
Figure 6 presents a comparison of the phase behavior
of the fluorooctyl poly(p-phenylene) (2c) (filled squares)
obtained in this study with those previously reported
for poly(1H,1H,2H,2H-tetrahydroperfluorodecyl meth-
acrylate, filled circles),13 poly(1H,1H,2H,2H-tetrahydro-
perfluorodecyl acrylate, open squares),22 and poly(1H,1H-
dihydroperfluorooctyl acrylate, open circles)23 in super-
critical CO2. The two acrylate cloud-point curves super-
pose with the fluorooctyl poly(p-phenylene) curve which

2246 Wright et al.
Macromolecules, Vol. 36, No. 7, 2003
transitions at this concentration are expected to be close to
the maximum in the pressure-composition isotherms.26-28 The
transition is a bubble point if small gas bubbles appear in the
cell when the pressure is decreased. The gas phase is expected
to be essentially pure solvent, and the composition of the
predominant phase in the cell will equal the overall solution
composition since the mass in the bubble is negligible. The
system pressure is measured with Heise pressure gauges
accurate to within (10 psi for data to 10 000 psig and to within
(50 psi for data from 10 000 to 40 000 psig. Cloud points are
reproduced two to three times to within approximately (60
psi, and bubble points are also reproduced two to three times
to within approximately (20 psi.
can be used to process the fluoro-derivatized PPP
polymers at pressures as low as 1000 psia at 25 °C,
solubility conditions that are a strong function of the
length of the alkoxy chain on the fluoroester substituent
group. Interestingly, CO2 can distinguish fluorinated
poly(methacrylate)s from fluorinated poly(acrylate)s,
which demonstrates the sensitivity CO2 exhibits to
modest variations in the chemical architecture of the
polymer. The SCF solubility studies reported here follow
many of the trends previously reported on the solution
behavior of functionalized, rodlike polymers7,8 in liquid
organic solvents. Further work is in progress to ascer-
tain the physics underlying the enhanced solubility of
the fluorooctyl ester PPP relative to the octyl ester PPP
in SCF solvents.
Syn th esis of P en ta flu or op r op yl 2,5-Dich lor oben zoa te
(1a ). A reaction vessel was charged with CH2Cl2 (11 mL), 2,5-
dichlorobenzoic acid (1.10 g, 5.75 mmol), pentafluoropropanol
(0.86 g, 5.75 mmol, 0.6 mL), and 1,1-carbonyldiimidazole (0.93
g, 5.8 mmol). The solution was stirred at ambient temperature
for 24 h, diluted with diethyl ether (50 mL), washed with H2O
(2 × 50 mL), and dried with MgSO4, and the solvents were
removed under reduced pressure. The crude product was
purified by Kugelrohr distillation (∼0.5 Torr) to afford 1a as
a clear and colorless oil (1.26 g, 68%). 1H NMR (CDCl3): δ 7.84
(s, 1H), 7.43 (s, 2H), 4.78 (t, J ) 12.9 Hz, 2H). 13C NMR
(CDCl3): δ 162.2, 133.8, 133.3, 133.1, 132.8, 132.6, 131.9, 118.2
(CF2), 116.8 (CF3), 60.3. IR (neat): νCO 1755 cm-1. Anal. Calcd
for C10H5 Cl2F5O2: C, 37.18; H, 1.56. Found: C, 37.12; H, 1.51.
Syn th esis of Tr id eca flu or ooctyl 2,5-Dich lor oben zoa te
(1c). A reaction vessel was charged with CH2Cl2 (10 mL), 2,5-
dichlorobenzoate (2.41 g, 12.6 mmol), 1,1-carbonyldiimadazole
(2.00 g, 12.6 mmol), and tridecafluorooctanol (2.78 mL, 4.58
g, 12.6 mmol) and stirred at ambient temperature for 24 h.
The solution was diluted with ether (75 mL), washed with H2O
(2 × 100 mL), dried with MgSO4, filtered, and concentrated
to yield a yellow liquid. The crude product was purified by
Kugelrohr distillation (∼0.5 Torr) to afford 1c as a clear and
colorless oil (4.72 g, 76%). 1H NMR (CDCl3): δ 7.79 (s, 1H),
7.37 (s, 2H), 4.62 (t, J ) 4.8 Hz, 2 H), 2.70-2.50 (m, 2H). 13C
NMR (CDCl3): δ 164.0, 133.1, 132.9, 132.6, 132.5, 131.6, 130.5,
122-105 (multiple signals for CF2’s), 57.73 (OCH2), 30.58
(CF3). IR (neat): νCO 1740 cm-1. Anal. Calcd for C15H7
Cl2F13O2: C, 33.55; H, 1.32. Found: C, 34.22; H, 1.32.
P olym er iza tion of P en ta flu or op r op yl 2,5-Dich lor o-
ben zoa te (2a ). To a charged flask of NMP (∼14 mL), penta-
fluoropropyl 2,5-dichlorobenzoate (6.56 g, 20.3 mmol), methyl
3-chlorobenzoate (0.346 g, 2.03 mmol), and PPh3 (2.13 g, 8.12
mmol) were heated to 80 °C. After 5 min, NiBr2 (0.437 g, 2.03
mmol) and Zn (2.65 g, 40.6 mmol) were added simultaneously
to the mixture and heated at 80 °C for 2 h. The oil bath was
removed, and the mixture was allowed to cool to ambient
temperature and then diluted with CH2Cl2 (50 mL). The
mixture was filtered through celite, washed with H2O (2 ×
100 mL) and NaCN (5% aqueous solution, 100 mL), dried over
MgSO4, and then concentrated to a volume of ∼5 mL under
reduced pressure. The concentrated polymer solution was
precipitated into cold MeOH (75 mL). The product was
collected by filtration and dried under reduced pressure to
afford 2a as a white solid (3.41 g, 66%, Mn ) 15 400, PD )
1.98). 1H NMR (CDCl3): δ 8.31-8.14 (m), 8.01, 7.66, 7.54, 7.46,
4.79, 3.94, 3.37. 2.36. 13C NMR (CDCl3): δ 165.5, 132.9, 132.3,
131.5, 131.2, 129.3, 129.1, 110.0 (CF2), 59.5 (CH2), 51.2 (CF3).
IR (neat): νCO 1734 cm -1. UV-vis (THF): λmax ) 318 nm.
Anal. Calcd for (C10H5O2F5)n: C, 46.21; H, 2.01. Found: C,
46.13; H, 2.34.
Exp er im en ta l Section
Gen er a l Meth od s. All manipulations of compounds and
solvents were done under nitrogen using standard Schlenk line
techniques. Triethylamine (CaH2) and THF (Na°) were purified
from by distillation under nitrogen from the specified drying
agents. 1H NMR and 13C NMR measurements were performed
using one of the following instruments: Bruker AC 200 MHz,
Varian Mercury 300 MHz, or Varian Inova 400 MHz spec-
1
trometer. H NMR and 13C NMR chemical shifts are reported
vs the respective solvent residue peak (solvent, 1H, 13C
signals: CDCl3, δ 7.25 ppm, δ 76.9 ppm; DMSO-d6, δ 2.62 ppm,
δ 36.9 ppm, respectively). The octyl and propyl 2,5-dichloro-
benzoate esters were prepared by standard Fischer esterifi-
cation techniques. The 2,5-dichlorobenzoic acid, octanol, pro-
panol, pentafluoropropanol, triphenylphosphine, NMP (stored
over molecular sieves), and CDI were purchased from Aldrich
Chemical Co. and used as received. The zinc (powder, 97%)
was purchased for Fisher Scientific and used as received. The
nickel bromide (anhydrous, 97%) was purchased for Alfa
Chemical Co. and used as received. The CF3(CF2)5CH2CH2OH
was purchased from Apollo Scientific. DSC and TGA experi-
ments were carried out on Perkin-Elmer DSC-7 and TGA-7
system under a nitrogen atmosphere at a flow rate of 40 cm3
N2/min. Typical samples of 5-10 mg were run through several
heating cycles to identify reproducible and distinct Tg/Tmelt
events. SEC analyses were performed by dilution in THF (2
mg/mL) and then injection onto a Hewlett-Packard 1100 HPLC
(column: PL 300 × 7.5 mm, 5 µm particle size). Molecular
weights are calculated relative to polystyrene standards.
Elemental analyses were performed at Atlantic Microlab Inc.,
Norcross, GA.
Described elsewhere is the apparatus and techniques used
to obtain polymer-SCF phase behavior data.24,25 The main
component of the experimental apparatus is a high-pressure,
variable-volume cell (Nitronic 50, 7.0 cm o.d. × 1.6 cm i.d.,
∼30 cm3 working volume). The cell is first loaded with a
measured amount of polymer to within (0.002 g. To remove
entrapped air, the cell is degassed very slowly at pressures
less than 40 psi with the supercritical solvent of interest. Gas
is then transferred into the cell gravimetrically to within (0.02
g using a high-pressure bomb. The mixture in the cell is viewed
with a borescope (Olympus Corporation, model F100-024-000-
55) placed against a sapphire window secured at one end of
the cell. A stir bar activated by a magnet located below the
cell mixes the contents of the cell. The solution temperature
is held to within (0.3 °C, as measured with
a type K
thermocouple. A fixed polymer concentration of approximately
2 wt % is used for each constant-concentration, phase boundary
curve. The mixture in the cell is compressed to a single phase,
and the pressure is then slowly decreased until a second phase
appears. The transition is a cloud point if the solution becomes
so opaque that it is no longer possible to see the stir bar in
solution. These cloud points have been compared in our
laboratories to those obtained using a laser light setup where
the phase transition is the condition of 90% reduction in light
transmitted through the solution. Both methods gave identical
results within the reproducibility of the data. The cloud-point
P olym er iza tion of P r op yl 2,5-Dich lor oben zoa te (2b).
A NMP (∼6 mL) solution containing propyl 2,5-dichloro-
benzoate (2.41 g, 10.3 mmol), PPh3 (1.09 g, 4.14 mmol), and
methyl 3-chlorobenzoate (0.18 g, 1.03 mmol) was heated to 80
°C under a nitrogen atmosphere. After 5 min, NiBr2 (0.23 g,
1.03 mmol) and Zn (1.35 g, 20.66 mmol) were added simulta-
neously, and the solution was heated at 80 °C for 2 h with
stirring. The polymer was isolated as above to afford 2b as a
white powder (1.3 g, 78%, Mn ) 8400, PD ) 2.15). 1H NMR
(CDCl3): δ 8.40, 8.25, 8.05, 7.57, 7.48, 4.25, 4.11-4.10 (m),
3.93 (s, 3H), 2.54 (s), 0.78 (s). 13C NMR (CDCl3): δ 168.3, 133.2,

Macromolecules, Vol. 36, No. 7, 2003
Synthesis of Poly(p-phenylenes) 2247
(6) Marrocco, M. L.; Gagne, R. R.; Trimmer, M. S. U.S. Patent
5,756,581, 1998. Marrocco, M. L.; Gagne, R. R.; Trimmer, M.
S. U.S. Patent 5,565,543, 1996. Marrocco, M. L.; Gagne, R.
R.; Trimmer, M. S. U.S. Patent 5,227,457, 1993.
(7) Ballauff, M. Fluid Phase Equilib. 1993, 83, 349. Ballauff, M.
Angew. Chem., Int. Ed. Engl. 1989, 28, 253.
(8) Galda, P.; Kistner, D.; Martin, A.; Ballauff, M. Macromol-
ecules 1993, 26, 1595. Vaia, R. A.; Krishnamamoorti, R.;
Benner, C.; Trimer, M. J . Polym. Sci., Part B: Polym. Phys.
1998, 36, 2449.
(9) Park, K. C.; Dodd, L. R.; Levon, K.; Kwei, T. K. Macromol-
ecules 1996, 29, 7149. Vaia, R. A.; Dudis, D.; Henes, J .
Polymer 1998, 39, 6021.
(10) Rindfleisch, F.; DiNoia, T. P.; McHugh, M. A. J . Phys. Chem.
1996, 100, 15581.
131.7, 131.0, 130.8, 130.5, 129.0, 67.2, 21.9, 10.6. IR (neat):
νCO 1720 cm-1. UV-vis (THF): λmax ) 310 nm. Anal. Calcd
for (C10H10O2)n: C, 73.95; H, 6.19. Found: C, 73.95; H, 6.09.
P olym er iza tion of Tr id eca flu or ooctyl 2,5-Dich lor o-
ben zoa te (2c). A solution of NMP (∼6 mL) containing
tridecafluorooctyl 2,5-dichlorobenzoate (4.98, 9.27 mmol), PPh3
(0.97, 3.71 mmol), and methyl 3-chlorobenzoate (0.16 g, 0.93
mmol) was heated to 80 °C for 5 min. Then NiBr2 (0.20 g, 0.93
mmol) and Zn (1.21 g, 18.5 mmol) were added simultaneously
to the mixture and heated at 80 °C for 2 h. Work-up and
isolation as above afforded 2c as a white solid (2.52 g, 58%,
1
Mn ) 7200, PD ) 1.65). H NMR (CDCl3): br s at with some
fine structure at δ 8.35, 8.24-8.21, 8.06-8.04, 7.94-7.89, 7.51,
4.64, 4.42-4.35, 3.97, 3.91, 2.61, 2.29, 1.55. 13C NMR
(CDCl3): δ 167.4, 141.1, 133.0, 132.2, 130.7, 129.4, 128.9, 121.8,
120.5, 117.5, 113.5, 111.1, 105.0. IR (neat): νCO 1645 cm-1. λmax
) 305 nm. Anal. Calcd for (C15H7F13O2)n: C, 39.84; H, 1.66.
Found: C, 39.63; H, 1.69.
(11) Sarbu, T.; Styranec, T. J .; Beckman, E. J . Ind. Eng. Chem.
Res. 2000, 39, 4678.
(12) Kirby, C. F.; McHugh, M. A. Chem. Rev. 1999, 99, 565 and
references therein.
(13) McHugh, M. A.; Garach-Domech, A.; Park, I. H.; Barbu, E.;
Graham, P.; Tsibouklis, J . Macromolecules 2002, 35, 6479.
(14) McHugh, M. A.; Krukonis, V. Supercritical Fluid Extraction:
Principles and Practice, 2nd ed.; Butterworth-Heinemann:
Boston, 1994.
(15) Prausnitz, J . M.; Lichtenthaler, R. N.; de Azevedo, E. G.
Molecular Thermodynamics of Fluid-Phase Equilibria, 3rd
ed.; Prentice Hall: Englewood Cliffs, NJ , 1999.
(16) Gottikh, B. P.; Krayevesky, A. A.; Tarussova, N. B.; Purygin,
P. P.; Tsilevich, T. L. Tetrahedron 1970, 26, 4419.
(17) Colon, I.; Kelsey, D. R. J . Org. Chem. 1986, 51, 2627.
(18) Care should be taken when using sodium cyanide solutions,
and all work should be done in a hood with adequate air flow
and with the appropriate chemical lab ware.
(19) Because of the rigid nature of PPPs, GPC results (using
polystyrene standards) will lead to “inflated” molecular
weight values and degrees of polymerization: Holdcroft, S.
J . Polym. Sci., Part B: Polym. Phys. 1991, 29, 1585.
(20) In ref 9 the work of Vaia et al. suggests that backbone
distortion induced from side chain interaction is important
in describing polymer structure. Park and co-workers (ref 9)
suggest that the overall orientation of the side chains
(cylindrical vs “plain plate structure”) is primary to describing
polymer structure.
(21) Kazarian, S. G.; Vincent, M. F.; Bright, F. V.; Liotta, C. L.;
Eckert, C. A. J . Am. Chem. Soc. 1996, 118, 1729.
(22) Mawson, S.; J ohnston, K. P.; Combes, J . R.; DeSimone, J .
M. Macromolecules 1995, 28, 3182.
P olym er iza tion of Octyl 2,5-Dich lor oben zoa te (2d ). A
reaction vessel was charged with NMP (8 mL), octyl 2,5-
dichlorobenzoate (3.84 g, 12.7 mmol), triphenylphosphine (1.33
g, 5.1 mmol), and methyl 3-chlorobenzoate (0.21 g, 1.27 mmol)
was heated to 80 °C. After 5 min nickel(II) bromide (0.27 g,
1.27 mmol) and zinc (1.66 g, 25.4 mmol) were added to the
solution. The solution was heated with stirring for 2 h at 80
°C. Work-up and isolation as above afforded 2d as a white solid
(2.1 g, 71%, Mn ) 13 200, PD ) 2.1). 1H NMR (CDCl3): δ 7.94
(s, 1H), 7.55, 4.12 (s), 4.04 (s), 1.46 (s, 3H), 1.21 (s, 10H), 0.84
(s, 3H). 13C NMR (CDCl3): δ 168.3, 141.2, 131.7, 131.2, 130.7,
130.4, 129.1, 128.9, 65.7, 32.0, 29.4, 28.9, 28.5, 26.0, 22.8, 14.3.
IR (neat): 1717 cm-1. UV-vis (CH2Cl2): λmax ) 315 nm. Anal.
Calcd for (C15H20O2)n: C, 77.39; H, 8.62. Found: C, 77.14; H,
8.37.
Refer en ces a n d Notes
(1) Busch, H.; Weber, W.; Darboven, C.; Renner, W.; Hahn, H.;
Mathause, G.; Startz, F.; Zitzmann, K.; Engelhardt, H. J .
Prakt. Chem. 1936, 146, 1.
(2) Stille, J . K.; Gilliams, Y. Macromolecules 1971, 4, 515.
VanKerckhoven, H. F.; Gilliams, Y. K.; Stille, J . K. Macro-
molecules 1972, 5, 541 and references therein.
(3) Hamaguchi, M.; Sawada, H.; Kyokane, J .; Yoshino, K. Chem.
Lett. 1996, 527. Gin, D. L.; Conticello, V. P. Trends Polym.
Sci. 1996, 4, 217. Perec, V.; Hill, D. H. ACS Symp. Ser. 1996,
624, 2.
(4) Kovacic, P.; J ones, M. B. Chem. Rev. 1987, 87, 357 and
references therein. Percec, V.; Okita, S.; Weiss, R. Macro-
molecules 1992, 25, 1816-1823. Percec, V.; Zhao, M,; Bae,
J .-Y.; Hill, D. H. Macromolecules 1996, 29, 3727. Wang, Y.;
Quirk, R. P. Macromolecules 1995, 28, 3495. Goodson, F. E.;
Wallow, T. I.; Novak, B. M. Macromolecules 1998, 31, 2047.
Havelka-Rivard, P. A.; Nagai, K.; Freeman, B. D.; Sheares,
V. V. Macromolecules 1999, 32, 6418. Schlu¨ter, A. D. J .
Polym. Sci., Part A: Polym. Chem. 2001, 39, 1533.
(5) Chaturvedi, V.; Tanaka, S.; Kaeriyama, K. Macromolecules
1993, 26, 2607. Kaeriyama, K.; Mehta, M. A.; Masuda, H.
Synth. Met. 1995, 69, 507. Kaeriyama, K.; Mehta, M. A.;
Chaturvedi, V.; Masuda, H. Polymer 1995, 36, 3027.
(23) Luna-Barcenas, G.; Mawson, S.; Takishima, S.; DeSimone,
J . M.; Sanchez, I. C.; J ohnston, K. P. Fluid Phase Equilib.
1998, 146, 325.
(24) Meilchen, M. A.; Hasch, B. M.; McHugh, M. A. Macromol-
ecules 1991, 24, 4874.
(25) Mertdogan, C. A.; Byun, H.-S.; McHugh, M. A.; Tuminello,
W. H. Macromolecules 1996, 29, 6548.
(26) Allen, G.; Baker, C. H. Polymer 1965, 6, 181.
(27) Irani, C. A.; Cozewith, C. J . Appl. Polym. Sci. 1986, 31, 1879.
(28) Lee, S.-H.; LoStracco, M. A.; Hasch, B. M.; McHugh, M. A.
J . Phys. Chem. 1994, 98, 4055.
MA021615E