Synthesis and gas permeability of ester substituted poly(p-phenylene)s

Article abstract of DOI:10.1016/j.polymer.2011.03.036

Polymerizations of various ester substituted 2,5-dichlorobenzoates [substituent: linear alkyl groups (1a-f), branched alkyl groups (1g-l), cyclohexyl groups (1m-o), phenyl groups (1p-r), and oxyethylene units (1s-v)] were investigated with Ni-catalyzed/Zn-mediated system in 1-methyl-2-pyrrolidone (NMP) at 80 °C. Most of monomers bearing linear and branched alkyl groups successfully polymerized to give relatively high-molecular-weight polymers (M_n = 10,000-20,800). However, the molecular weight of the polymer having eicocyl groups was low because of steric hindrance of long alkyl chain. The polymerizations of cyclohexyl 2,5-dichlorobenzoate and phenyl 2,5-dichlorobenzoate produced low-molecular-weight polymers, while the polymerizations of monomers with alkyl cyclohexyl and alkyl phenyl groups proceeded to afford polymers with relatively high-molecular-weights. The polymers possessing oxyethylene units were obtained, but the molecular weights were low when the oxyethylene chains were long. The gas permeability of membranes of poly(p-phenylene)s with alkyl chains increased as increasing the length of alkyl chain. The membranes of poly(p-phenylene)s with phenyl groups and oxyethylene units exhibited high densities and relatively low gas permeability. However, the CO₂/N₂ separation factor of membrane of poly(p-phenylene) having oxyethylene units was as large as 73.6.

Full text of DOI:10.1016/j.polymer.2011.03.036

Polymer 52 (2011) 2163e2169
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Synthesis and gas permeability of ester substituted poly(p-phenylene)s
*
Toshikazu Sakaguchi , Shinichi Tominaga, Tamotsu Hashimoto
Department of Materials Science and Engineering, Graduate School of Engineering, University of Fukui, Bunkyo 3-9-1, Fukui 910-8507, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Polymerizations of various ester substituted 2,5-dichlorobenzoates [substituent: linear alkyl groups (1aef),
branched alkyl groups (1gel), cyclohexyl groups (1meo), phenyl groups (1per), and oxyethylene units (1sev)]
were investigated with Ni-catalyzed/Zn-mediated system in 1-methyl-2-pyrrolidone (NMP) at 80 ꢀC. Most of
monomers bearing linear and branched alkyl groups successfully polymerized to give relatively high-molec-
ular-weight polymers (Mn ¼ 10,000e20,800). However, the molecular weight of the polymer having eicocyl
groups was low because of steric hindrance of long alkyl chain. The polymerizations of cyclohexyl 2,5-
dichlorobenzoate and phenyl 2,5-dichlorobenzoate produced low-molecular-weight polymers, while the
polymerizations of monomers with alkyl cyclohexyl and alkyl phenyl groups proceeded to afford polymers
with relatively high-molecular-weights. The polymers possessing oxyethylene units were obtained, but the
molecular weights were low when the oxyethylene chains were long. The gas permeability of membranes of
poly(p-phenylene)s with alkyl chains increased as increasing the length of alkyl chain. The membranes of poly
(p-phenylene)s with phenyl groups and oxyethylene units exhibited high densities and relatively low gas
permeability. However, the CO2/N2 separation factor of membrane of poly(p-phenylene) having oxyethylene
units was as large as 73.6.
Received 16 December 2010
Received in revised form
11 March 2011
Accepted 19 March 2011
Available online 12 April 2011
Keywords:
Poly(p-phenylene)
Membrane
Gas permeability
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
use, and the design and synthesis of novel polymers are essential
from the viewpoint of industrial issue.
Gas permeable polymers have been extensively studied because
of their promising application in gas separation membrane [1e5]. A
large number of polymers such as substituted polyacetylenes [6e8]
and polyimides [9e11] have been studied as materials for gas
separation membrane. The main chains of these polymers are
composed of double bonds, aromatic rings, etc., and they are rigid
and restricted in rotation. When such polymers possess spherical
bulky substituents, their membranes usually showed large free
volume and high gas permeability because of the steric repulsion of
bulky substituents. For instance, poly(1-trimethylsilyl-1-propyne)
exhibits the highest gas permeability among all the polymers
known [12], but it has some problems toward the practical use; e.g.,
low permselectivity [13], unstability toward heat and atmospheric
oxidation [14e17], decrease of permeability for aging [18e21], and
so on. Recently, new types of polymer membranes such as carbon
molecular sieve membranes and polymers of intrinsic micropo-
rosity have been developed and they are reported to show high
performance in gas separation [22e25]. However, there are a few
polymers which have sufficient performance to apply to practical
Poly(p-phenylene) is the simplest aromatic polymer, which has
linear and rigid backbone. Thus, poly(p-phenylene)s possibly
exhibit high gas permeability, and they are interesting as materials
for gas separation membrane. However, gas permeation properties
of membranes of poly(p-phenylene) derivatives have never been
known so far. Unsubstituted poly(p-phenylene) is insoluble in any
organic solvents and infusible, but the incorporation of appropriate
substituents can make poly(p-phenylene) soluble and fusible [26].
Soluble poly(p-phenylene)s with substituents are synthesized by
homo- or cross-coupling reactions that are typically nickel- or
palladium-catalyzed [27]. Among them, Yamamoto coupling reac-
tion using Ni/Zn catalyst afforded relatively high-molecular-weight
ester substituted poly(p-phenylene)s [28e30]. Ester substituted
poly(p-phenylene)s obtained by the polymerization with Ni/Zn
catalyst would form a free-standing membrane due to their high-
molecular-weights.
In the present study, various ester substituted poly(p-phenyl-
ene)s were synthesized by Ni/Zn catalyzed coupling reaction to
develop materials for gas separation (Scheme 1). The gas perme-
ability of their free-standing membranes was examined, and it was
found that poly(p-phenylene) bearing diethylene glycols exhibited
pretty high CO2 permselectivity. This is the first report on gas
permeability of membranes of poly(p-phenylene) derivatives.
* Corresponding author. Tel.: þ81 776 27 9779; fax: þ81 776 27 8904.
E-mail address: sakaguchi@matse.u-fukui.ac.jp (T. Sakaguchi).
0032-3861/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2011.03.036

2164
T. Sakaguchi et al. / Polymer 52 (2011) 2163e2169
Scheme 1. Synthesis of ester substituted poly(p-phenylene)s by Ni-catalyzed polymerization.
2. Experimental
dichlorobenzoic acid (50 g, 0.262 mol) and DMF (40 mL) were added
into the flask, and the mixture was stirred at 50 ꢀC for 30 min. Thionyl
chloride (62 g, 0.525 mol) was slowly added to the flask through
dropping funnel at 50 ꢀC, and the mixture was stirred for 24 h. 2,5-
Dichlorobenzoyl chloride was obtained by distillation under reduced
pressure as colorless liquid (48 g, 0.229 mol, yield 88%).
2.1. Measurements
The molecular weight distributions (MWDs) of polymers were
measured by gel permeation chromatography (GPC) in chloroform (at
a 1.0 mL/min flow rate) at 40 ꢀC on a Shimadzu LC-10AD chromato-
graph equipped with four polystyrene gel columns (Shodex K-805L ꢁ 1
and K-804L ꢁ 3) and a Shimadzu RID-6A refractive index detector. The
number-average molecular weight (Mn) and polydispersity ratio
[weight-average molecular weight/number-average molecular weight
(Mw/Mn)] were calculated from chromatograms based on a polystyrene
calibration. 1H (500 MHz) and 13C (125 MHz) NMR spectra were
recorded on Jeol LA-500 instrument in CDCl3 at room temperature.
Thermogravimetric analyses (TGA) were conducted with Rigaku TG-
DTA 8078G1 under N2 at a 10 ꢀC/min heating rate. Thermal decom-
position temperature was defined as the temperature of 5% weight loss
of the sample. Gas permeability coefficients were measured with
a Rikaseiki K-315-N gas permeability apparatus at 25 ꢀC under 1 atm
upstream pressure. The permeabilitycoefficient (P) expressed in barrer
unit (1 barrer ¼ 10ꢂ10 cm3(STP) cm cmꢂ2 sꢂ1 cm Hgꢂ1) was calculated
from the slope of the steady-state line. The D value was determined by
the time lag method using the following equation:
To a 500 mL, three-necked round-bottom flask equipped with
a magnetic stirring bar and dropping funnel, were added 1-eicosanol
(13 g, 0.044 mol), triethylamine (7.3 g, 0.072 mol), and 40 mL of
CH2Cl2. A solution of 2,5-dichlorobenzoyl chloride (8.4 g, 0.040 mol)
in CH2Cl2 (20 mL) was slowly added to the flask through dropping
funnel at 0 ꢀC, and the mixture was stirred for 24 h at room temper-
ature. The reaction mixture was washed with water three times. The
solution was dried over anhydrous sodium sulfate, and then
concentrated at reduced pressure. The crude product was purified by
silica gel column chromatography (eluent: hexane/CHCl3:1/1) to give
1f (16 g, 0.033 mol, 83%) as white solid.1H NMR (CDCl3, ppm): 7.79 (s,
1H, Ar), 7.38 (s, 2H, Ar), 4.33 (t, J ¼ 6.7 Hz, 2H, OCH2), 1.76 (quint, 2H,
J ¼ 7.0 Hz, OCH2CH2), 1.31e1.23 (m, 34H, OCH2CH2(CH2)17), 0.88 (t,
J ¼ 6.8 Hz, 3H, OCH2CH2(CH2)17CH3). 13C NMR (CDCl3, ppm): 164.6,
132.6, 132.3, 132.2, 132.0, 131.7, 131.2, 66.1, 31.9, 29.69, 29.67, 29.64,
29.62, 29.61, 29.5, 29.4, 29.3, 29.2, 28.5, 25.9, 22.7. Anal. Calcd. for
C27H44O2Cl2: C, 68.77; H, 9.41. Found: C, 68.66; H, 9.63.
D ¼ l2=6
q
2.3.2. 2-Butyl 2,5-dichlorobenzoate (1g)
here, l is the membrane thickness and q is the time lag, which is given
by the intercept of the asymptotic line of time-pressure curve to the
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and 2-butanol. H NMR (CDCl3, ppm):
1
time axis. The S value was calculated by using equation S ¼ P/D.
7.75 (s,1H, Ar), 7.39 (s, 2H, Ar), 5.12 (sext, J ¼ 6.4 Hz,1H, OCH),1.79e1.65
(m, 2H, OCHCH2),1.35 (d, J ¼ 6.2 Hz, 3H, OCHCH3), 0.99 (t, J ¼ 7.4 Hz, 3H,
OCHCH2CH3). 13C NMR (CDCl3, ppm): 164.3, 132.6, 132.3, 132.1, 131.8,
131.0,130.9, 74.4, 28.8,19.4, 9.7. Anal. Calcd. for C11H12O2Cl2: C, 53.46; H,
4.89. Found: C, 53.21; H, 4.78.
2.2. Materials
2,5-Dichlorobenzoic acid, thionyl chloride, dichloro bis(triphe-
nylphosphine)nickel (II), sodium iodide, zinc powder, and common
organic solvents were commercially obtained from Wako Pure
Chemicals, Ind., Ltd. and used without further purification. Alcohols
used for esterification were purchased from Tokyo Kasei Kogyo,
Ind., Ltd. and used without further purification. Monomers (1aev)
were synthesized by chlorination and esterification as shown in
Scheme 2. Synthesis of monomers 1aee, 1j, and 1p were reported
previously [30e32], and monomers 1fei, 1keo, 1q, and 1rev were
newly synthesized in the present study. The details of the proce-
dures and analytical data are described below.
2.3.3. 3-Pentyl 2,5-dichlorobenzoate (1h)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and 3-pentanol. 1H NMR (CDCl3,
ppm): 7.76 (s, 1H, Ar), 7.37 (s, 2H, Ar), 5.04 (quint, J ¼ 6.2 Hz, 1H,
OCH), 1.72 (quint, J ¼ 7.4 Hz, 4H, OCHCH2), 0.97 (t, J ¼ 7.4 Hz, 6H,
OCHCH2CH3). 13C NMR (CDCl3, ppm): 164.5, 132.6, 132.3, 132.1,
132.0, 131.8, 130.9, 79.0, 26.4, 9.7. Anal. Calcd. for C12H14O2Cl2: C,
55.19; H, 5.40. Found: C, 54.07; H, 5.22.
2.3. Monomer synthesis
2.3.4. 4-Heptyl 2,5-dichlorobenzoate (1i)
This monomer was prepared by the same method as for 1f
using 2,5-dichlorobenzoyl chloride and 4-heptanol. 1H NMR
(CDCl3, ppm): 7.75 (s, 1H, Ar), 7.37 (s, 2H, Ar), 5.18 (quint,
J ¼ 6.2 Hz, 1H, OCH), 1.72e1.58 (m, 4H, OCHCH2), 1.47e1.35 (m,
4H, OCHCH2CH2), 0.97 (t, J ¼ 7.4 Hz, 6H, OCHCH2CH2CH3). 13C
2.3.1. Eicosyl 2,5-dicholorobenzoate (1f)
A 500-mL three-necked flask was equipped with a reflux
condenser, a dropping funnel, a three-way stopcock, and a magnetic
stirring bar. After the flask was flushed with nitrogen, 2,5-

T. Sakaguchi et al. / Polymer 52 (2011) 2163e2169
2165
Scheme 2. Synthesis of monomers 1aev.
NMR (CDCl3, ppm): 164.4, 132.5, 132.2, 132.1, 132.0, 131.8, 130.9,
76.2, 36.2, 18.6, 14.0. Anal. Calcd. for C14H18O2Cl2: C, 58.14; H, 6.27.
Found: C, 57.64; H, 6.22.
ppm): 7.82 (s, 1H, Ar), 7.46e7.34 (m, 7H, Ar), 5.37 (s, 2H, OCH2). 13
C
NMR (CDCl3, ppm): 164.2, 135.2, 132.6, 132.5, 132.3, 132.2, 131.3,
131.1, 128.6, 128.5, 128.4, 67.6. Anal. Calcd. for C14H10O2Cl2: C, 59.81;
H, 3.59. Found: C, 59.07; H, 3.73.
2.3.5. 4-Decyl 2,5-dichlorobenzoate (1k)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and 4-decanol. 1H NMR (CDCl3, ppm):
7.74 (s, 1H, Ar), 7.37 (s, 2H, Ar), 5.16 (quint, J ¼ 6.2 Hz, 1H, OCH),
1.73e1.58 (m, 4H, OCH(CH2CH2CH3)CH2), 1.46e1.22 (m, 10H, OCH
(CH2CH2CH3)CH2(CH2)4CH3), 0.95 (t, J ¼ 7.4 Hz, 3H, OCHCH2CH2CH3),
0.88 (t, J ¼ 7.0 Hz, 3H, OCH(CH2)5CH3). 13C NMR (CDCl3, ppm): 164.4,
132.5, 132.3, 132.1, 132.0, 131.8, 130.9, 76.4, 36.2, 34.1, 31.7, 29.2, 25.3,
22.6, 18.7, 14.0, 13.9. Anal. Calcd. for C17H24O2Cl2: C, 61.63; H, 7.30.
Found: C, 61.34; H, 7.53.
2.3.11. 2-Phenylethyl 2,5-dichlorobenzoate (1r)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and 2-phenylethanol. 1H NMR (CDCl3,
ppm): 7.71 (s, 1H, Ar), 7.38e7.23 (m, 7H, Ar), 4.55 (t, J ¼ 7.2 Hz, 2H,
OCH2), 3.08 (t, J ¼ 7.2 Hz, 2H, OCH2CH2). 13C NMR (CDCl3, ppm):
164.3, 137.4, 132.5, 132.4, 132.2, 132.1, 131.2, 131.1, 128.9, 128.6, 126.7,
66.3, 35.0. Anal. Calcd. for C15H12O2Cl2: C, 61.04; H, 4.10. Found: C,
61.16; H, 4.23.
2.3.12. 2-Methoxyethyl 2,5-dichlorobenzoate (1s)
2.3.6. 2-Decyl 2,5-dichlorobenzoate (1l)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and ethylene glycol monomethyl ether.
1H NMR (CDCl3, ppm): 7.83 (s, 1H, Ar), 7.39 (s,1H, Ar), 7.38 (s,1H, Ar),
4.49 (t, J ¼ 4.8 Hz, 2H, C(]O)OCH2), 3.73 (t, J ¼ 4.8 Hz, 2H, C(]O)
OCH2CH2), 3.42 (s, 3H, OCH3). 13C NMR (CDCl3, ppm): 164.3, 132.6,
132.5, 132.2, 132.1, 131.3, 131.1, 70.2, 64.7, 59.0. Anal. Calcd. for
C10H10O3Cl2: C, 48.22; H, 4.05. Found: C, 48.35; H, 3.89.
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and 2-decanol. 1H NMR (CDCl3, ppm):
7.75 (s, 1H, Ar), 7.37 (s, 2H, Ar), 5.17 (sext, J ¼ 7.0 Hz, 1H, OCH),
1.75e1.68 (m, 2H, OCHCH2),1.64e1.55 (m, 2H, OCHCH2CH2),1.47e1.22
(m, 13H, OCH(CH3)CH2CH2(CH2)5), 0.88 (t, J ¼ 6.8 Hz, 3H, OCH
(CH2)6CH3). 13C NMR (CDCl3, ppm): 164.2, 132.5, 132.3, 132.1, 132.0,
131.8, 130.9, 73.3, 35.9, 31.8, 29.44, 29.41, 29.2, 25.4, 22.6, 19.9, 14.1.
Anal. Calcd. for C17H24O2Cl2: C, 61.63; H, 7.30. Found: C, 61.74; H, 7.25.
2.3.13. 2-(2-Methoxyethoxy)ethyl 2,5-dichlorobenzoate (1t)
This monomer was prepared by the same method as for 1f
using 2,5-dichlorobenzoyl chloride and diethylene glycol mono-
methyl ether. 1H NMR (CDCl3, ppm): 7.84 (s, 1H, Ar), 7.39 (s, 1H,
Ar), 7.38 (s, 1H, Ar), 4.50 (t, J ¼ 4.8 Hz, 2H, C(]O)OCH2), 3.84 (t,
J ¼ 4.8 Hz, 2H, C(]O)OCH2CH2), 3.69 (t, J ¼ 4.8 Hz, 2H, OCH2CH2),
3.57 (t, J ¼ 4.8 Hz, 2H, OCH2CH2), 3.39 (s, 3H, OCH3). 13C NMR
(CDCl3, ppm): 164.2, 132.5, 132.4, 132.2, 132.1, 131.3, 131.1, 71.8,
70.5, 68.8, 64.7, 59.0. Anal. Calcd. for C12H14O4Cl2: C, 49.17; H, 4.81.
Found: C, 49.33; H, 4.58.
2.3.7. Cyclohexyl 2,5-dichlorobenzoate (1m)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and cyclohexanol. 1H NMR (CDCl3,
ppm): 7.77 (s,1H, Ar), 7.37 (s, 2H, Ar), 5.05 (quint, J ¼ 5.0 Hz,1H, OCH),
1.98e1.93 (m, 2H, OCH(CH2)CH2), 1.84e1.76 (m, 2H, OCH(CH2)CH2),
1.64e1.31 (m, 6H, OCHCH2CH2CH2CH2CH2). 13C NMR (CDCl3, ppm):
164.0,132.5,132.2,132.1,132.0,131.8,130.9, 74.6, 31.5, 25.3, 23.6. Anal.
Calcd. for C13H14O2Cl2: C, 57.16; H, 5.17. Found: C, 56.91; H, 5.14.
2.3.8. Cyclohexylmethyl 2,5-dichlorobenzoate (1n)
2.3.14. 2-[2-(2-Methoxyethoxy)ethoxy]ethyl 2,5-dichlorobenzoate
(1u)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and cyclohexylmethanol. 1H NMR
(CDCl3, ppm): 7.80 (s, 1H, Ar), 7.38 (s, 2H, Ar), 4.15 (d, J ¼ 6.3 Hz, 2H,
OCH2), 1.85e1.66 (m, 5H, OCH2CH(CH2)CH2), 1.33e1.04 (m, 6H,
OCH2CHCH2CH2CH2CH2CH2) 13C NMR (CDCl3, ppm): 164.6, 132.6,
132.3,132.2,132.0, 131.7, 131.2, 71.1, 37.1, 29.7, 26.3, 25.6. Anal. Calcd.
for C14H16O2Cl2: C, 58.55; H, 5.62. Found: C, 57.88; H, 5.55.
This monomer was prepared by the same method as for 1f
using 2,5-dichlorobenzoyl chloride and triethylene glycol mono-
methyl ether. 1H NMR (CDCl3, ppm): 7.83 (s, 1H, Ar), 7.39 (s, 1H,
Ar), 7.38 (s, 1H, Ar), 4.49 (t, J ¼ 4.7 Hz, 2H, C(]O)OCH2), 3.83 (t,
J ¼ 4.7 Hz, 2H, C(]O)OCH2CH2), 3.71e3.62 (m, 6H, OCH2
-
CH2OCH2CH2), 3.54 (t, J ¼ 4.9 Hz, 2H, OCH2CH2OCH3), 3.37 (s, 3H,
OCH3). 13C NMR (CDCl3, ppm): 164.3, 132.6, 132.5, 132.2, 132.1,
131.3, 131.2, 71.9, 70.7, 70.6, 70.5, 68.9, 64.8, 59.0. Anal. Calcd. for
C14H18O5Cl2: C, 49.87; H, 5.38. Found: C, 50.09; H, 5.41.
2.3.9. Cyclohexylethyl 2,5-dichlorobenzoate (1o)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and 2-cyclohexylethanol. 1H NMR
(CDCl3, ppm): 7.78 (s, 1H, Ar), 7.38 (s, 2H, Ar), 4.37 (t, J ¼ 7.0 Hz, 2H,
OCH2), 1.78e1.62 (m, 8H, OCH2CH2CH(CH2CH2)CH2CH2CH2), 1.45
(m, 1H, OCH2CH2CH), 1.28e0.92 (m, 4H, OCH2CH2CH(CH2CH2)
CH2CH2) 13C NMR (CDCl3, ppm): 164.6, 132.6, 132.3, 132.2, 132.0,
131.7, 131.2, 64.3, 35.8, 34.5, 33.4, 26.4, 26.2. Anal. Calcd. for
C15H18O2Cl2: C, 59.81; H, 6.02. Found: C, 59.62; H, 6.19.
2.3.15. Heptaethylene glycolyl 2,5-dichlorobenzoate (1v)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and heptaethylene glycol monomethyl
ether. 1H NMR (CDCl3, ppm): 7.83 (s, 1H, Ar), 7.39 (s, 1H, Ar), 7.38 (s,
1H, Ar), 4.49 (t, J ¼ 4.8 Hz, 2H, C(]O)OCH2), 3.83 (t, J ¼ 4.8 Hz, 2H, C
(]O)OCH2CH2), 3.68e3.60 (m, 22H, O(CH2CH2O)5CH2CH2OCH3),
3.54 (t, J ¼ 4.8 Hz, 2H, CH2CH2OCH3), 3.38 (s, 3H, OCH3). 13C NMR
(CDCl3, ppm): 164.3, 132.5, 132.4, 132.2, 132.1, 131.3, 131.2, 71.9, 70.6,
70.56, 70.52, 70.45, 68.8, 64.7, 58.9. Anal. Calcd. for C22H34O9Cl2: C,
52.38; H, 6.88. Found: C, 51.14; H, 6.23.
2.3.10. Phenylmethyl 2,5-dichlorobenzoate (1q)
This monomer was prepared by the same method as for 1f using
2,5-dichlorobenzoyl chloride and benzyl alchoal. 1H NMR (CDCl3,

2166
T. Sakaguchi et al. / Polymer 52 (2011) 2163e2169
Table 1
À
Áꢀ
À
Áꢀ
Results of polymerization of 1aer.a
FFV ¼ vsp ꢂ v0 vspz vsp ꢂ 1:3vw vsp
Yieldb
Mn
Mw/Mn
Pn
c
c
d
Monomer
where vsp is the polymer specific volume, and v0 is the occupied
volume of the polymer. The occupied volume is typically estimated
as 1.3 times the van der Waals volume (vw), which is calculated
using the group contribution method [33].
1a
1b
1c
1d
1e
1f
1g
1h
1i
89
72
78
73
91
87
69
67
49
89
72
85
65
64
78
54
57
54
51
35
66e
30e
13,200
13,300
13,200
20,800
11,800
6270
11,900
14,700
10,000
17,100
2020
14,300
3130
12,100
17,300
2010
2.72
3.30
4.05
3.64
3.03
2.00
2.31
2.84
2.36
3.63
1.35
3.52
1.70
3.47
4.06
2.75
2.66
3.39
3.11
2.10
2.04
2.02
75
65
57
80
41
15
68
77
46
74
8
55
15
56
76
10
21
80
86
116
18
13
3. Results and discussion
3.1. Polymerization
1j
1k
1l
The polymerizations of monomers 1aev were carried out using
Ni/Zn catalyst in 1-methyl-2-pyrrolidone (NMP) at 80 ꢀC, whose
results are summarized in Table 1. The polymerization of 1a with
a n-butyl group as a side chain provided a polymer (2a) in 89%
yield, whose number-average molecular weight (Mn) was 13,200.
The polymerizations of monomers 1be1e with a linear alkyl chain
(n-hexyl, n-octyl, n-decyl, and n-dodecyl) also proceeded
smoothly, and polymers 2be2e with Mn ¼ 11,800e20,800 were
obtained in good yields. It is reported that the polymerizations of
linear alkyl 2,5-dichlorobenzoates (n-hexyl, n-octyl, and n-
dodecyl) by Ni-catalyzed coupling to give polymers with
Mn ¼ 7000e13,000 [30,34]. Both the previous and the present
results indicate that the length of alkyl side chain does not much
affect the polymerizability when the alkyl group is shorter than n-
dodecyl group. However, monomer 1f having an eicosanyl group,
which is the longest side group in the present study, polymerized
1m
1n
1o
1p
1q
1r
1s
1t
4460
17,900
15,300
25,800
4770
1u
1v
5970
a
In NMP at 80 ꢀC for 96 h; [M]0 ¼ 1.04 M, [Ni] ¼ 70.0 mM, [NaI] ¼ 70.0 mM,
[Ph3P] ¼ 0.486 mM.
b
Methanol-insoluble product.
Measured by GPC (polystyrene calibration).
Degree of polymerization calculated from GPC.
Ether-insoluble product.
c
d
e
2.4. Polymerization
Polymerization was carried out in a glass flask equipped with
a three-way stopcock under dry nitrogen at 80 ꢀC. A detailed
procedure of polymerization is as follows: Dichloro bis(triphenyl-
phosphine)nickel (II) (0.732 g, 1.12 mmol), sodium iodide (0.168 g,
1.12 mmol), zinc powder (2.18 g, 33.3 mmol), triphenylphosphine
(2.04 g, 7.78 mmol), and 1-methyl-2-pyrrolidone (NMP) (8 mL)
were placed in a 100-mL glass flask. The mixture was heated at
80 ꢀC, and stirring for 4 days. Polymerization was quenched by
adding a small amount of methanol. The formed polymer was
isolated by precipitation into a large excess of methanol. Its yield
was determined by gravimetry.
2.5. Membrane preparation
Polymer membranes (thickness ca. 70e100 mm) were prepared
by casting toluene solutions of the polymers (concn. 1.0 wt%) onto
a glass plate. The plate was covered with a glass vessel to slow
down solvent evaporation (ca. 3 days). After drying, the membrane
was peeled off, and it was further dried at room temperature under
reduced pressure for 24 h.
2.6. Density and fractional free volume (FFV) of polymer membrane
The densities of membranes were determined by hydrostatic
weighing using a Mettler Toledo balance and a density determina-
tion kit. In this method, a liquid with known density (r0) is needed,
and the membrane density (r) is given by the following equation:
r
¼
r0 ꢁ MA=ðMA ꢂ MLÞ
where MA is membrane weight in air and ML is membrane weight in
the auxiliary liquid. An aqueous sodium nitrate was used as the
auxiliary liquid. FFV is calculated by the following equation:
Fig. 1. 1H NMR spectra of monomer 1a (top) and polymer 1b (bottom) in CDCl3.

T. Sakaguchi et al. / Polymer 52 (2011) 2163e2169
2167
Fig. 2. 1H NMR spectra of polymers 2g, 2o, 2r, and 2t in CDCl3.
to afford a relatively low-molecular-weight polymers (Mn ¼ 6270).
The low polymerizability of 1f may be accounted for by the steric
hindrance of long alkyl side chain to the coupling reaction.
Monomers 1ge1l bearing a branched alkyl side group also poly-
merized to give corresponding poly(p-phenylene)s. The Mn’s of
polymers 2gej and 2l were over 10,000, and the branched
structure does not affect the polymerizability. The molecular
weight of polymer 2k was one order of magnitude smaller than
those of the polymers bearing linear and branched alkyl side
groups. However, the reason for the low polymerizability of 1k is
not cleared. Introduction of bulky substituents such as cyclohexyl
and phenyl groups decreased the polymerizability (1m and 1p),
and the resultant polymers had relatively low-molecular-weights.
The steric hindrance of bulky substituents would results in the
decrease of degree of polymerization. The polymerizations of
monomers having an alkyl spacer between bulky substituents and
ester oxygen smoothly proceeded to give relatively high-molec-
ular-weight polymers. This is because the flexible spacer reduces
the steric hindrance of bulky substituents. Monomers 1s and 1t
possessing methoxyethyl and methoxyethoxyethyl groups,
respectively, polymerized to produce relatively high-molecular-
weight polymers, while the polymerizations of monomers 1u and
1v having a longer oxyethylene side chain afforded the polymers
with low-molecular-weights. This may be due to both the steric
hindrance of long side groups and the interaction between metal
catalyst and polar ether oxygen.
Table 2
Solubility of the polymers.a
hexane toluene CHCl3 THF CH2Cl2 acetone DMF DMSO
2a
2b
2c
2d
2e
2f
2g
2h
2i
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
þ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
ꢂ
ꢂ
ꢃ
ꢃ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
ꢃ
ꢃ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
ꢃ
þ
þ
þ
þ
þ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
þ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
þ
þ
þ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢃ
þ
þ
þ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢃ
þ
þ
þ
2j
2k
2l
2m
2n
2o
2p
2q
2r
2s
2t
2u
2v
a
Symbols: þ soluble, ꢃ partially soluble, ꢂ insoluble.

2168
T. Sakaguchi et al. / Polymer 52 (2011) 2163e2169
Table 3
Density, gas permeability coefficients (P), and permselectivity of membranes.
Density (g/cm3)
FFV
PN2 (barrer)a
PO2 (barrer)a
PCO2 (barrer)a
DCO2 ꢁ 10ꢂ6b
SCO2 ꢁ 10ꢂ3c
PO2/PN2
PCO2/PN2
2b
2c
2d
2e
2j
2l
2r
2t
1.11
1.07
1.04
1.02
1.07
1.04
1.22
1.23
0.153
0.160
0.165
0.170
0.157
0.167
0.132
0.148
1.55
4.78
10.8
10.8
1.47
4.84
0.120
0.390
4.74
14.4
28.0
29.1
5.07
14.6
0.280
1.88
28.6
82.2
149
0.237
0.944
1.72
12.1
8.71
8.66
8.64
10.3
8.54
6.29
13.5
3.06
3.00
2.59
2.70
3.46
3.00
2.32
4.82
18.5
17.2
13.8
14.2
18.8
16.1
10.1
73.6
153
1.77
27.6
78.1
1.22
28.7
0.268
0.915
0.0194
0.213
a
1 barrer ¼ 1 ꢁ 10ꢂ10 cm3(STP) cm/(cm2 s cm Hg).
b
c
In the units of cm2/s.
In the units of cm3(STP)/(cm3 cm Hg).
Fig. 1 shows 1H NMR spectra of monomer 1a and polymer 2a
the permeability measurement were too small. The nitrogen
permeability coefficient (PN2) of membrane of 2b was 1.55 barrers,
whose value is near those of conventional glassy polymers such as
low-density-polyethylene and polystyrene [35]. The permeability
was similar to those of alkyl-substituted poly(p-phenyleneethyn-
ylene)s, which have linear and rigid backbone as same as poly(p-
phenylene) [36].The PN2 values of membranes of 2c and 2d are 4.78
and 10.8 barrers, respectively. The nitrogenpermeability increased as
increasing the length of alkyl side chain. However, membrane of 2e
exhibited the same PN2 valueasmembraneof 2d. Whenthe alkyl side
groupisenoughlong, thelengthofalkylsidegroup doesnotaffectthe
gas permeability. Densities of membranes of 2bee decreased as
increasing the length of alkyl side groups. The FFVs of membranes of
2bee increased as increasing the length of alkyl side groups.
Increasing the FFV would result in the increase of gas permeability of
poly(p-phenylene)s. Another possible reason for the increasing gas
permeability is that the increment of content of flexible linear alkyl
chain enhances the diffusivity of gas molecules in polymer matrix.
These DCO2 values, in fact, increased as increasing the length of alkyl
chains although the SCO2 values did not significantly depends on
alkyl side groups.
in CDCl3. The 1H NMR spectrum of polymer 2a shows a complex
multiplet in the aromatic region. The peaks at 4.2 ppm and
1.6e1.1 ppm in the aliphatic region were also complex, while
these peaks were clearly observed in the spectrum of the
monomer. The complexity of the peaks in polymer is probably
due to the presence of various constitutional isomeric structural
units. Generally, the 5-chloride substituent in 2,5-dichlor-
obenzoate is consumed to a greater degree than the 2-chloride
substituent [31]. Therefore, the coupling is occurred between two
5-chloride substituents at the first stage of polymerization, and
then between 5-chloride and 2-chloride substituents. At the last
stage of polymerization, the reaction is occurred only between
two 2-chloride substituents. In this way, the various constitu-
tional isomeric structures are generated. The spectra of all the
other polymers in this study exhibited complex maltiplet, indi-
cating that the substituted poly(p-phenylene)s have constitu-
tional isomeric structures. The spectra of typical polymers were
shown in Fig. 2.
3.2. Solubility property
The density of membrane of 2j which has 2-ethylhexyl groups
was 1.07 g/cm3, and it is the same as that of 2c which has n-octyl
groups. Similarly, the density of membrane of 2l with 2-octyl
groups was the same as that of 2d with n-decyl groups. The FFV
values of 2j and 2l were also close to those of 2c and 2d, respec-
tively. However, the gas permeability of membranes of 2j and 2l
The solubility properties of polymers 2aev were studied, whose
results are summarized in Table 2. Polymers 2aef with linear alkyl
chains showed the same solubility and totally dissolved in toluene,
CHCl3, THF, and CH2Cl2. Polymers 2gel having branched alkyl
chains were also soluble in such organic solvents. Polymer 2k was
soluble even in hexane and acetone because its molecular weight
was very low. Polymers 2meo bearing cyclohexyl groups dissolved
in toluene, CHCl3, THF, and CH2Cl2, and they showed the same
solvent-solubility as the polymers with linear and branched alkyl
groups. On the other hand, polymers 2per having phenyl groups
showed poor solubility. They completely dissolved only in CHCl3,
and they were partially soluble and insoluble in other organic
solvents. Polymer 2s having methoxyethyl groups exhibited similar
solubility to the polymers having alkyl groups, while polymers
2tev with longer oxyethylene side groups dissolved in toluene,
CHCl3, THF, CH2Cl2, acetone, DMF, and DMSO.
3.3. Gas permeation properties
Free-standing membranes of polymers 2bee, 2j, 2l, 2r, and 2t
could be prepared by solution-casting, and they were enough tough
to measure gas permeability coefficients. However, the membranes
of the other polymers were too brittle to measure the membrane
properties. Table 3 lists densities, FFV, permeability coefficients,
diffusivity coefficients for CO2, solubility coefficients for CO2, and
ratios of the coefficients for the O2/N2 and CO2/N2 in 2bee, 2j, 2l, 2r,
and 2t at 25 ꢀC. Unfortunately, the diffusivity and solubility coeffi-
cients for O2 and N2 could not be determined because the time lags in
Fig. 3. Plots of PCO2/PN2 versus PCO2 for the present poly(p-phenylene)s together with
the data for other polymers: (C) poly(ether imide) segmented copolymers; (:) poly
(2,5-dialkyl-p-phenyleneethynylene)s; (-) poly(diphenylacetylene)s bearing poly-
ethylene glycols.

T. Sakaguchi et al. / Polymer 52 (2011) 2163e2169
2169
bearing branched alkyl groups was lower than those of 2c and 2d,
respectively. The lower gas permeability of membranes of 2j and 2l
can be accounted for by the fact that branched alkyl chains have
lower mobility than linear alkyl chains, which results in lower gas
permeability was revealed for the first time. The gas permeability of
membranes of ester substituted poly(p-phenylene)s was found to
be increased as increasing alkyl side chains. The PCO2/PN2 value of
membrane of poly(p-phenylene) bearing diethylene glycol moieties
was as large as 73.6.
diffusion. The DCO2 values in 2j and 2l were 0.268 ꢁ 10ꢂ6
,
0.915 ꢁ 10ꢂ6 cm2/s, respectively, which were much smaller than
that of 2d.
The PN2 value of 2r bearing 2-phenylethyl groups was 0.120
barrers, which is much smaller than those of membranes in this
study. Phenyl groups probably make the membrane dense by p-p
References
[1] Pinnau I, Freeman BD. Advanced materials for membrane separation. In:
ACS symposium series 876. Washington DC: American Chemical Society;
2004.
stacking. This is supported by the data of FFV for membrane of 2r
(FFV ¼ 0.132). The DCO2 value of 2r was one or two orders of
magnitude smaller than those of the other polymers in the present
study. Both the small FFV and less local mobility of bulky phenyl
side groups may cause such low diffusivity in membrane of 2r. The
density of membrane of 2t was 1.23 g/cm3, which is the largest
among all the membranes in this study. Polymer 2t also made
a dense membrane because of strong interaction of polar side
chains, and the FFV value was as small as 0.148. However, the gas
permeability of 2t was one order of magnitude higher than that of
2r. This is probably because the flexible oxyethylene chains
promote gas diffusion in the polymer matrix. It is noteworthy that
the CO2/N2 separation factor (PCO2/PN2) of membrane of 2t was as
large as 73.6. The polymer membranes with ether oxygen moieties
such as polyethylene glycol (PEG) are considered to be favorable for
CO2 separation membrane [37], and it has been reported that
crosslinked PEG membranes exhibit high CO2 permselectivity
[38,39]. Polymer 2t owes the high PCO2/PN2 to high affinity of CO2
molecules to oxyethylene side groups. Okamoto et al. reported
excellent CO2/N2 separation performance of poly(ether imide)
segmented copolymers [40]. Their PCO2 and PCO2/PN2 were
24e140 barrers and 65e70, respectively, whose data is one of the
best performance among all the polymers. The PCO2 and PCO2/PN2
of 2t were 28.7 barrer and 73.6, respectively. It is therefore said that
poly(p-phenylene) bearing diethylene glycols exhibits good
performance for CO2/N2 separation.
[2] Tsujita Y. Prog Polym Sci 2003;28:1377.
[3] Nunes SP, Peinemann KV. Membrane technology in the chemical industry.
New York: Wiley; 2001.
[4] Pandey P, Chauhan RS. Prog Polym Sci 2001;26:853.
[5] Baker RW. Ind Eng Chem Res 2002;41:1393.
[6] Nagai K, Masuda T, Nakagawa T, Freeman BD, Pinnau I. Prog Polym Sci 2001;
26:721.
[7] Aoki T. Prog Polym Sci 1999;24:951.
[8] Sakaguchi T, Yumoto K, Shida Y, Shiotsuki M, Sanda F, Masuda T. J Polym Sci
Pol Chem 2006;44:5028.
[9] Wind JD, Paul DR, Koros WJ. J Membr Sci 2004;228:227.
[10] Heuchel M, Hofmann D, Pullumbi P. Macromolecules 2004;37:201.
[11] Burns RL, Koros WJ. Macromolecules 2003;36:2374.
[12] Masuda T, Isobe E, Higashimura T, Takada K. J Am Chem Soc 1983;105:
7473.
[13] Masuda T, Isobe E, Higashimura T. Macromolecules 1985;18:841.
[14] Nagai K, Nakagawa T. J Appl Polym Sci 1994;54:1651.
[15] Tasaka S, Inagaki N, Igawa M. J Polym Sci Pol Phys 1991;29:691.
[16] Tanaka A, Nitta K, Maekawa R, Masuda T, Higashimura T. Polym J 1992;24:
1173.
[17] Masuda T, Tang BZ, Higashimura T, Yamaoka H. Macromolecules 1985;18:
2369.
[18] Hill AJ, Pas SJ, Bastow TJ, Burgar MI, Nagai K, Toy LG, et al. J Membr Sci 2004;
243:37.
[19] Nagai K, Freeman BD, Hill AJ. J Polym Sci Pol Phys 2000;38:1222.
[20] Korkenoo KD, Pfromm PH. Macromolecules 2000;33:3747.
[21] Nagai K, Nakagawa T. J Membr Sci 1995;105:261.
[22] Ismail AF, David LIB. J Membr Sci 2001;193:1.
[23] Kim YK, Park HB, Lee YM. J Membr Sci 2005;255:265.
[24] Budd PM, Msayib KJ, Tattershall CE, Ghanem BS, Reynolds KJ, McKeown NB,
et al. J Membr Sci 2005;251:263.
[25] Budd PM, Ghanem BS, Makhseed S, McKeown NB, Msayib KJ. Chem Commum;
2004:230.
Fig. 3 depicts the CO2 permeabilitycoefficients (PCO2) and CO2/N2
separation factor (PCO2/PN2) of the present polymers. The data of
poly(ether imide) segmented copolymers reported by Okamoto
[40], poly(p-phenyleneethynylene)s having alkyl groups [36], poly
(diphenylacetylene)s having polyethylene glycols [41], and Robe-
son’s upper bound [42] are also plotted in Fig. 3. It is clearly observed
from Fig. 3 that poly(p-phenylene)s with alkyl side chains (2bel)
exhibited higher CO2 permselectivity than the poly(p-phenyl-
eneethynylene)s with alkyl side chains. The CO2 permeability of poly
(p-phenylene)s is also higher than poly(p-phenyleneethynylene)s
when they have the same alkyl side groups. Poly(diphenylacetylene)
s bearing ethylene glycols show high CO2 permeability, and poly
(ether imide) segmented copolymers show high CO2 permse-
lectivity. Compared to such polymers, poly(p-phenylene) with
diethylene glycols exhibited low CO2 permeability but high selec-
tivity, and its data gets close to Robeson’s upper bound.
[26] Kaeriyama K. Synthesis and properties of poly(p-phenylene) and its deriva-
tives. In: Wise DL, Knek GE, Trantolo DJ, Cooper TM, Gresser JD, editors.
Photonic polymer systems. New York: Marcel Dekker; 1998. p. 33.
[27] Kovacic P, Jones MB. Chem Rev 1987;87:357.
[28] Yamamoto T. Bull Chem Soc Jpn 2010;83:431.
[29] Yamamoto T, Osakada K, Wakabayashi T, Yamamoto A. Makromol Chem
Rapid Commun 1985;6:671.
[30] Kaeriyama K, Kouyama S, Sekita M, Nakayama T, Tsukahara Y. Macromol
Rapid Commun 1999;20:50.
[31] Percec V, Bae JY, Zhao M, Hill DH. Macromolecules 1995;28:6726.
[32] Shen Z, McHugh MA, Lott KM, Wright ME. Fluid Phase Equilib 2004;216:1.
[33] van Krevelen KW. Properties of polymer: their correlation with chemical
structure; their numerical estimation and prediction from additive group
contributions. 3rd ed. Amsterdam: Elsevier Science; 1990. pp. 71e107.
[34] Wright ME, Lott KM, McHugh MA, Shen Z. Macromolecules 2003;36:2242.
[35] Pauly S. Permeability and diffusion data. In: Brandrup J, Immergut EH,
Grulke EA, editors. Polymer handbook. New York: Willy; 1998. p.
VI-545e62.
[36] Sakaguchi T, Nojiri D, Hashimoto T. Polym Bull 2008;60:271.
[37] Lin H, Freeman BD. J Mol Struct 2005;739:57.
[38] Lin H, Wagner EV, Swinnea JS, Freeman BD, Pas SJ, Hill AJ, et al. J Membr Sci
2006;276:145.
[39] Lin H, Freeman BD. J Membr Sci 2004;239:105.
[40] Okamoto K, Fujii M, Okamyo S, Suzuki H, Tanaka K, Kita H. Macromolecules
1995;28:6950.
[41] Sakaguchi T, Kameoka K, Hashimoto T. J Appl Polym Sci 2009;113:3504.
[42] Robeson LM. J Membr Sci 2008;320:390.
4. Conclusions
A series of ester substituted poly(p-phenylene)s (2aev) were
synthesized by Ni/Zn-catalyzed coupling reaction, and their gas