Gas transport in polymers prepared via metathesis copolymerization of exo-N-phenyl-7-oxanorbornene-5,6-dicarboximide and norbornene

Article abstract of DOI:10.1021/ma030285a

This work reports the synthesis of exo-N-phenyl-7-oxanorbornene-5,6-dicarboximide and its ring-opening metathesis copolymerization with norbornene to yield poly(exo-N-phenyl-7-oxanorbornene-5,6-dicarboximide-co-norbornene), with molar ratio 50/50. The glass transition temperature of the copolymer is 125°C. Permeation and sorption processes of different gases (hydrogen, nitrogen, oxygen, carbon monoxide, carbon dioxide, methane, ethylene, and ethane) were measured in membranes prepared by casting from solutions of the copolymer in chloroform. The Langmuir capacity of the gases is relatively small due to the nearness of the glass transition temperature of the polymer to the working temperature. The solution of the most condensable gases in the continuous phase of the membrane is apparently described by the Flory-Huggins theory of polymer-diluent mixtures. In general, the membranes exhibit a reasonably high separation coefficient of hydrogen with respect to ethane, ethylene, nitrogen and methane. The value of α(O₂/N₂) at room temperature lies in the vicinity of 5.

Full text of DOI:10.1021/ma030285a

Macromolecules 2003, 36, 8483-8488
8483
Gas Transport in Polymers Prepared via Metathesis Copolymerization
of exo-N-Phenyl-7-oxanorbornene-5,6-dicarboximide and Norbornene
Mik h a il A. Tlen k op a tch ev a n d J oel Va r ga s
Instituto de Investigaciones en Materiales, Universidad Aut o´ noma de M e´ xico,
Apartado Postal 70-360 CU, Coyoac a´ n, M e´ xico DF 04510, M e´ xico
Ma r ı´ a d el Ma r L o´ p ez-Gon z a´ lez a n d Eva r isto Ria n d e*
Instituto de Ciencia y Tecnolog ´ı a de Pol ´ı meros (CSIC), 28006 Madrid, Spain
Received May 22, 2003; Revised Manuscript Received August 12, 2003
ABSTRACT: This work reports the synthesis of exo-N-phenyl-7-oxanorbornene-5,6-dicarboximide and
its ring-opening metathesis copolymerization with norbornene to yield poly(exo-N-phenyl-7-oxanorbornene-
5
,6-dicarboximide-co-norbornene), with molar ratio 50/50. The glass transition temperature of the
copolymer is 125 °C. Permeation and sorption processes of different gases (hydrogen, nitrogen, oxygen,
carbon monoxide, carbon dioxide, methane, ethylene, and ethane) were measured in membranes prepared
by casting from solutions of the copolymer in chloroform. The Langmuir capacity of the gases is relatively
small due to the nearness of the glass transition temperature of the polymer to the working temperature.
The solution of the most condensable gases in the continuous phase of the membrane is apparently
described by the Flory-Huggins theory of polymer-diluent mixtures. In general, the membranes exhibit
a reasonably high separation coefficient of hydrogen with respect to ethane, ethylene, nitrogen, and
2 2
methane. The value of R(O /N ) at room temperature lies in the vicinity of 5.
In tr od u ction
bornene-5,6-dicarboximide, and N-phenyl-exo-norbornene-
,6-dicarboximide as well as the copolymers N-(1-
5
The ever increasing use of membranes as an alterna-
tive to the cryogenic industries for producing nitrogen-
enriched air, as well as the use of membrane technology
in the petrochemical industry, has encouraged studies
of the relationship between structure and transport
properties of different classes of polymers.1,2 Polynor-
bornene and derivatives have been widely used in the
past decade for permeation and sorption studies. Nor-
bornene monomers have been the subject of interest due
to facile preparation and functionalization.3 Function-
alized norbornenes easily participated in ring-opening
metathesis polymerization (ROMP) with approximate
complete conversion to high molecular weight polymers
with good mechanical properties.6 Some of the polynor-
bornenes showed very attractive gas transport proper-
ties. Thus, fluorine-containing glassy polynorbornene as
well as organosilicon-substituted polynorbornene have
adamantyl)-exo-norbornene-5,6-dicarboximide/norbor-
nene, with molar compositions 70/30, 50/50, and 30/70.
The permselectivity of these membranes was found to
be dependent on the type of gases to be separated. For
example, the permselectivity of oxygen with respect to
nitrogen, R(O2/N2), is ca. 5.50 for membranes obtained
from N-(1-adamantyl)-exo-norbornene-5,6-dicarboximide
(50/50) copolymers. For the separation of H2 from C2H6,
the membranes prepared from N-(1-adamantyl)-exo-
norbornene-5,6-dicarboximide exhibit very good perfor-
mance. In general, this study showed that the presence
of imide groups enhances the permselectivity of polynor-
bornenes.
-5
,7
It is important to study how slight modifications in
the structure of polymers may affect the gas transport
properties of membranes prepared from them. In this
work, we extended the preceding studies to the synthe-
sis of exo-N-phenyl-7-oxanorbornene-5,6-dicarboximide
(PhONDI), and its further copolymerization with nor-
bornene (NB). Then we studied the transport charac-
teristics of the membranes prepared from poly-
(exo-N-phenyl-7-oxanorbornene-5,6-dicarboximide-co-
norbornene) (50/50) to different gases: hydrogen, argon,
oxygen, nitrogen, carbon monoxide, carbon dioxide,
methane, ethane, and ethylene. From permeation re-
sults, the permselectivity coefficients of the membranes
for different pairs of gases were obtained and compared
with the pertinent results reported for other polynor-
bornenes containing imide groups in the structure. Only
sorption experiments of the most soluble gases were
performed, specifically, carbon dioxide, ethane, ethylene,
and methane, because the relatively small amount of
copolymer available precluded the possibility of obtain-
ing with enough precision the solubility coefficient of
the less soluble gases.
been synthesized,8
-11
observing in both cases an in-
crease in glass transition temperature and gas perme-
1
2
ability. More recently, Steinh a¨ usler and Koros re-
ported that a high order tacticity has a positive effect
on both permeation and separation processes. It has also
been reported that introduction of tosylate group into
the five-membered ring of the polynorbornene main
chain resulted in increase of glass transition tempera-
ture and significant improvement in selectivity for O2/
1
3
N2.
In most cases, however, substituted polynor-
bornenes8 display lower permselectivity and higher
permeability than polyimides, one of the materials that
exhibit better gas separation performance.11 Recently
the need to investigate how the presence of imide groups
in polynorbornenes could affect the permselectivity of
,14
membranes prepared from these materials has become
evident. For this purpose we proceeded7
,15
with the
synthesis and polymerization of N-(1-adamantyl)-exo-
norbornene-5,6-dicarboximide, N-cyclohexyl-exo-nor-
1
0.1021/ma030285a CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/03/2003

8
484 Tlenkopatchev et al.
Macromolecules, Vol. 36, No. 22, 2003
Exp er im en ta l Section
Ma ter ia ls. exo-7-Oxanorbornene-5,6-dicarboxylic anhydride
(exo-ONDA, 1), norbornene, aniline, and other chemicals were
purchased from Aldrich Chemical Co. Norbornene was distilled
from sodium metal. 1,2-Dichloroethane and toluene were dried
over anhydrous calcium chloride and distilled under nitrogen
over CaH
hydroimidazol-2-ylidene)(PCy
2
. Catalyst I, 1,3-bis(2,4,6-trimethylphenyl)-4,5-di-
1
6
3
2
)CI RudCHPh, was purchased
from Stream Chemical Co. and used as received.
Mea su r em en ts. H NMR and 13C NMR spectra were
1
recorded with a Varian spectrometer at 300 and 75.5 MHz
frequencies, respectively, in CDCl
TMS) as internal standard.
The glass transition temperature was measured under
3
with tetramethylsilane
(
nitrogen with a Du Pont 2100 instrument, at a heating rate
of 10 °C/min. The samples were encapsulated in standard
aluminum DSC pans in duplicate. Each pan was run twice on
the temperature range 30-200 °C. The glass transition
temperature was taken as the temperature at which the
baseline in the glassy region intercepts the straight line drawn
through the middle point of the endotherm. FTIR spectra were
F igu r e 1. Synthesis route of exo-N-phenyl-7-oxanorbornene-
5
,6-dicarboximide (3).
sphere terminated the polymerization, and the solution was
poured into an excess of methanol. The polymer was purified
by solubilization in chloroform containing a few drops of 1 N
HCl and precipitation into methanol, and it was further dried
in a vacuum oven at 40 °C to constant weight.
obtained on a Nicolet 510p spectrometer.
Copolymer compositions were determined by 1H NMR
integration of the olefinic peak of exo-N-phenyl-7-oxanor-
bornene-5,6-dicarboximide (PhONDI) (6.15-5.85 ppm) units
relative to the olefinic protons of norbornene units (5.34 ppm).
Molecular weights and molecular weight distributions with
reference to polystyrene standards were determined with a
Varian 9012 GPC instrument at 30 °C in chloroform (universal
Syn th esis of Cop oly(exo-N-ph en yl-7-oxa n or bor n en e-
5
,6-d ica r boxim id e/n or bor n en e) (4). One gram (4.15 mmol)
of 3, 0. 40 g (4.15 mmol) of NB, and 0.0071 g (0.0084 mmol) of
catalyst I were stirred in 8.4 mL of 1,2-dichloroethane at room
temperature for 1 h. The polymer obtained (Figure 2) was
-
1
soluble in chloroform and dichloromethane.
column and a flow rate of 1 mL min ).
1
H NMR (300 MHz, CDCl ): δ (ppm) ) 7.47-7.25 (5H, m),
3
Membranes were prepared by casting from chloroform
solutions. Gas permeation was measured at 30 °C using a
thermostated experimental device described in detail else-
6
(
.15 (2H, s, trans), 5.88 (2H, m, cis), 5.64 (2H, m, trans), 5.34
2H, m, cis), 4.46 (2H, m), 3.35 (2H, m), 2.58 (2H, m), 2.44
(2H, s), 2.06-1.72 (2H, m), 1.60-1.0 (4H, m).
1
7
where. Sorption experiments were performed in a thermo-
stated experimental device made up of a reservoir chamber
connected by a valve to the sorption chamber that contained
1
3
C NMR (75 MHz, CDCl ): δ (ppm) ) 174.6, 133.3, 132.5,
3
1
1
31.4, 129.1, 128.7, 126.3, 82.5, 52.6, 43.1, 40.6, 32.1.
1
8
FT-IR: 2997, 2940, 2854, 1775, 1716, 1555, 1498, 1376,
the polymer in film form. The reservoir and the sorption
chambers were equipped, respectively, with Gometrics (0-35
bar) and MKS-722 (0-33 atm) pressure transducers. Gas at
a given pressure was introduced into the reservoir chamber,
and once it reached the temperature of interest, the valve
separating the reservoir from the sorption chamber was
suddenly opened and closed. The variation of the pressure of
the gas by effect of the sorption process was recorded as a
function of time in the sorption chamber. The leaks and the
adsorption of gas in the walls of the latter chamber were
previously measured in a blank experiment. The gas concen-
-
1
308, 1178, 740, 691 cm
The values of the number-average molecular weight, M
heterodispersity index, M /M , and glass transition temper-
.
n
,
w
n
ature of poly(exo-N-phenyl-7-oxanorbornene-5,6-dicarboximide-
5
co-norbornene) were, respectively, 1.8 × 10 , 1.4, and 125 °C.
Resu lts
P er m ea tion Resu lts. An illustrative curve depicting
the variation of the pressure of nitrogen with time in
the downstream chamber is shown in Figure 3. As
usual, a transitory appears at short times followed by
a steady region at which the pressure is a linear
function of time. The dependence of the pressure on
time, calculated from the integration of Fick’s second
3
3
tration in cm of gas(STP)/cm of polymer at the pressures and
temperature of interest was obtained by means of the gas
equation using the suitable compressibility coefficients.
Mon om er Syn th esis. exo-N-Phenyl-7-oxanorbornene-5,6-
dicarboximide (PhONDI, 3) was prepared according to the
1
9
20
literature. exo-ONDA 1 (5 g, 30 mmol) was dissolved in 17
mL of toluene. To the stirred solution of exo-ONDA 2.8 g of
aniline (30 mmol) in 10 mL of toluene was added dropwise.
The reaction was maintained at 40 °C for 2 h and then cooled
to room temperature. A precipitate was filtered and dried to
give 7.6 g (29 mmol) of amic acid. Obtained amic acid 2 (7.6 g,
law assuming the diffusion coefficient constant, is
shown in Figure 3 for nitrogen. The good fitting between
computed and experimental results suggests that the
assumption D ) constant holds. The same occurs for
the other gases. The extrapolation of the steady-state
2
1
region intercepts the abscissa axis at the “time lag” θ.
2
9 mmol), anhydrous sodium acetate (1.12 g, 14 mmol), and
2
The diffusion coefficient was obtained from D ) L /6θ,
where L is the membrane thickness, whereas the
permeability coefficient, P, was calculated from the
permeation results in steady-state conditions.
acetic anhydride (23 g, 224 mmol) were heated at 70 °C for 3
h and then cooled. The solid crystallized on cooling was filtered,
washed several times with water, and dried in a vacuum oven
at 50 °C overnight. Pure monomer 3 (Figure 1) was obtained
after twice recrystallization from methanol: yield ) 89%, mp
Values at 30 °C of the permeability and diffusion
coefficients for different gases are given in the second
and third columns of Table 1, respectively. The results
show that P(H2) > P(CO2) > P(O2) > P(Ar) > P(CO) >
P(C2H4) = P(CH4) > P(N2) > P(C2H6), whereas the
values of the diffusion coefficient follow the trend D(H2)
. D(O2) > D(Ar) > D(CO2) = D(N2) > D(CO) > D(CH4)
)
164-165 °C.
1
H NMR (300 MHz, CDCl ): δ (ppm) ) 7.49-7.26 (5H, m),
3
6
1
1
.56 (2H, s), 5.39 (2H, s), 3.0 (2H, s).
C NMR (75 MHz, CDCl
28.7, 126.5, 81.3, 47.5.
1
3
3
): δ (ppm) ) 175.3, 136.6, 129.1,
FT-IR: 3064, 3021, 3003, 1775, 1715, 1595, 1496, 1380,
287, 1186, 874, 713 cm-1
.
>
D(C2H4) > D(C2H6). The results for the apparent
Mon om er Meta th esis P olym er iza tion . It was carried out
in glass vials under a dry nitrogen atmosphere at room
temperature. Adding benzaldehyde under a nitrogen atmo-
solubility coefficient, obtained from the ratio P/D, are
shown in the fourth column of Table 1. It can be seen

Macromolecules, Vol. 36, No. 22, 2003
Copolymerization of PhONDI and Norbornene 8485
F igu r e 2. Scheme showing synthesis of poly(exo-N-phenyl-7-oxanorbornene-5,6-dicarboximide-co-norbornene) (50:50) via ROMP.
F igu r e 4. Isotherms at 30 °C representing the sorption of
F igu r e 3. Dependence of the pressure of nitrogen in the
2 2 4 2 6 4
CO , C H , C H , and CH as a function of pressure. The
dashed isotherm for CO represents the concentration of this
2
gas in the membrane calculated assuming ideal gas behavior.
downstream chamber on time, at 30 °C. Continuous line
represents the isotherm calculated by integration of Fick’s
second law assuming the diffusion coefficient independent of
concentration.
Ta ble 2. Va lu es of Hen r y’s La w Solu bility Con sta n t a n d
La n gm u ir P a r a m eter s for Ca r bon Dioxid e, Meth a n e,
Eth ylen e, a n d Eth a n e a t 30 °C a n d 1 a tm P r essu r e
Ta ble 1. P er m ea tion a n d Sor p tion Resu lts for Differ en t
Ga ses a t 30 °C a n d 1 a tm P r essu r e
3
3
3
3
kD × 10 , cm (STP)/
C′H, cm (STP)/
b × 10 ,
8
3
D × 10 ,
S × 10 ,
(cm3 cmHg)
3
-1
gas
cm
(cmHg)
2
3 a,b
from sorptionb
50.2
gas
P, barrers
cm /s
S × 10
CO2
9.4
0.8
3.4
3.5
7.9
1.8
8.8
8.1
8.5
5.4
4.4
3.6
CO2
O2
N2
Ar
CO
H2
CH4
C2H4
C2H6
6.14
0.99
0.21
0.57
0.38
9.21
0.27
0.32
0.22
1.50
3.94
1.42
2.59
1.03
129
41.0
2.5
1.5
2.2
3.7
CH4
C2H4
C2H6
results obtained by both techniques though they follow
the same trend, that is, S(CO2) > S(C2H4) > S(C2H6) >
S(CH4).
0.7
0.62
0.16
0.06
4.3
19.4
37.8
7.9
32.5
26.4
As usual, the isotherms display concavity with respect
to the abscissa axis observed in the sorption curves of
most glassy systems. The dual-mode model, which
assumes the glassy state as formed by a continuous
phase where microcavities accounting for the excess
volume are dispersed, has traditionally been used to
interpret gas sorption in glassy polymers. The solubility
in the continuous phase obeys Henry’s law, while the
microcavities act like Langmuir sites in which adsorp-
a
S ) P/D. b Units of S: cm of gas(STP)/(cm of polymer cmHg).
3
3
that S(CO2) > S(C2H4)> S(C2H6) > S(CH4) > S(CO) >
S(O2) > S(Ar) > S(N2) > S(H2). The small solubility
coefficient of hydrogen in the membranes is responsible
for the rather low flow rate of this gas across the
membrane, only somewhat larger than that of carbon
dioxide.
Sor p tion Resu lts. Sorption experiments of carbon
dioxide, methane, ethylene, and ethane were conducted
by monitoring the decrease of pressure with time in a
chamber containing the polymer in film form. Gas
concentrations are represented as a function of pressure
in Figure 4. The values of the solubility coefficients at
2
2,23
tion processes take place.
As shown in Figure 4, the
sorption results fit rather well those predicted by the
dual-mode model using the values given in Table 2 for
the absorption Henry’s constant, kD, the Langmuir
capacity, C′H, and the Langmuir affinity parameter, b.
An inspection of the values of these parameters indi-
cates that the gas concentration in Langmuir sites
follows the trend: C2H4 = CO2 = C2H6 > CH4. In any
case, the value of C′H for CO2 is significantly lower than
that reported for polynorbornenes and fluorine-contain-
1
atm of CO2, CH4, C2H4, and C2H6, obtained at 30 °C
from sorption measurements, are given in the fifth
column of Table 1, while those obtained from perme-
ation experiments appear in the fourth column. As often
is the case, there are small differences between the
9
ing polynorbornenes.

8
486 Tlenkopatchev et al.
Macromolecules, Vol. 36, No. 22, 2003
F igu r e 7. Arrhenius plots for the parameters of the dual-
D H
mode model of carbon dioxide: (b) k , (0) b and (2) C′ .
F igu r e 5. Isotherms showing the dependence of the solubility
coefficient of CO on pressure at different temperatures.
2
7
-oxanorbornene-5,6-dicarboximide/norbornene) to the
working temperature.
By assuming that the continuous phase in the dual-
mode model behaves like a liquid, the solubility Henry’s
constant can be obtained from the variation of the
chemical potential of the gas in the liquid state dissolved
in the polymer. The Flory-Huggins theory,27 in con-
junction with the Clausius-Clapeyron equation that
allows for the determination of the vapor pressure of
the gas in the liquid form at the temperature of interest,
leads to the following expression for Henry’s con-
2
8,29
stant
2
7
2414
6Vh
λ
RTb
T
Tb
kD )
exp -(1 + ø) -
1 -
(1)
[
(
)
]
where Tb is the boiling temperature of the gas at 1 bar,
λ is the latent heat of vaporization at Tb, ø is an
enthalpic parameter that accounts for the polymer-gas
interaction and Vh is the partial molar volume of the gas
2
F igu r e 6. Arrhenius plots for the solubility coefficient of CO .
The results for the solubility coefficient of CO2 at
different temperatures, plotted as a function of pressure
in Figure 5, show that S undergoes a sharp decrease at
low pressures, remaining nearly constant at pressures
higher than 10 atm.. Moreover, the decrease of the
solubility coefficient is lower, the higher the tempera-
ture is. As shown in Figure 6, the temperature depen-
dence of the solubility coefficient follows Arrhenius
behavior and the value of the sorption heat obtained
from the Arrhenius plot is -5.2 kcal mol- at 1 atm.
Therefore, the solution of CO2 in the membranes is an
exothermic process, as occurs with the solubility of most
gases in glassy polymers.24 Arrhenius plots for the
Henry’s constant, kD, the gas concentration at the
Langmuir sites, C′H, and the affinity parameter, b, are
shown in Figure 7. The activation energies amount to
3
in cm /mol. According to eq 1, the higher Tb is and/or
the lower ø is, the higher the value of the Henry’s
constant is. As shown in Table 3, a good concordance
between the experimental and calculated values of kD
is obtained using the reasonable values of the interac-
tion parameter given in the fifth column of this table.
This is a remarkable fact considering that in the
development of eq 1 not only the Flory-Huggins theory
was used that, in principle, is only valid for rubbery
polymers, but also a rather wide extrapolation interval
of temperature was used to obtain the vapor pressure
of the gas by means of the Clausius-Clapeyron equa-
tion. It should be pointed out that the molar partial
volume used in the calculations was that corresponding
to Tb at 1 bar. To use the true value of Vh at 30 °C only
would slightly alter the results obtained for ø.
Occasional fluctuations in the glassy state give rise
to the formation of channels through which the diffusant
molecule may slip to a nearby cavity. The possibility of
slippage requires not only a suitable velocity of the
molecule when the fluctuation takes place but also the
radius of the channel to be larger than the radius of
the diffusant. The size of the molecule of a gas can be
estimated from the Lennard-J ones collision diameter σc,
determined on the basis of the molecular interactions
1
-
1
-
6.6, -4.1, and -1.2 kcal mol for kD, b, and C′H,
respectively.
Discu ssion
The dual-mode model suggests that the Langmuir
2
5,26
capacity of glassy membranes depends on Tg - T.
According to this approach, a plot of vgC′H/(Rg - Rl) vs
Tg - T for glassy membranes should give a straight line
passing through the origin. Notice that Rg and Rl are,
respectively, the thermal expansion coefficient of the
glass and the liquid at temperature T and vg is the
specific volume. Therefore, the relatively low value of
C′H found for the membranes used in this study may
be due to the closeness of the Tg of copoly(exo-N-phenyl-
of a gas, and the kinetic diameter which is close to the
molecular sieving dimension of the molecule.3
0,31
The
collision and kinetic diameters are, respectively, widely
accepted correlation parameters for diffusivity in the
rubbery and glassy states. The values of the natural

Macromolecules, Vol. 36, No. 22, 2003
Copolymerization of PhONDI and Norbornene 8487
Ta ble 3. Va lu es of th e En th a lp ic P a r a m eter Th a t Accou n ts for th e In ter a ction P olym er Ga s, a t 30 °C
3
3
3
3
1
0 × kD,calcd, cm (STP)/
10 × kD,expt, cm (STP)/
λ, kcal mol-
1
Vh , cm /mol
a
3
ø
(cm3 cmHg)
(cm3 cmHg)
gas
Tb, °C
CO2
-78.5
-159
-88.6
-103.7
6.06
2.21
3.51
46
38
55
49.4
0
2
2.5
2.5
8.6
0.9
3.8
2.6
9.4
0.8
3.5
3.4
CH4
C2H6
C2H4
b
b
3.23
a
Molar partial volume at Tb. b Reference 32.
F igu r e 9. Experimental separation factor of hydrogen with
respect to different gases as a function of the permeability
coefficients of the latter.
F igu r e 8. Plot of natural logarithm of the diffusion coefficient
against the square of the kinetic diameter of different gases
indicated in the figure.
logarithm of the diffusion coefficients of H2, N2, CO, Ar,
O2, CH4, C2H4, and C2H6, represented as a function of
the square of the kinetic diameter in Figure 8, fit rather
well to a straight line. It is worth noting that the
diffusion coefficient of carbon dioxide falls well below
the straight line, suggesting that some kind of polymer-
gas interaction hindering the diffusive process occurs.
The high solubility of CO2 and the anomalous low
diffusion coefficient of this gas may be related to the
polarity of the imide groups. CO2 is a nonpolar molecule,
but it still has a strong quadrupole moment that makes
possible dipole-dipole interaction with the imide groups.
These interactions may enhance solubility, but severely
hinder diffusant motions.
the permselectivity of the membranes prepared from
1
2
these polymers. Also, introduction of fluorine atoms
in the main chain or in side groups as well as the
attachment of organosilicon moieties to polynorbornene
seems to increase the permeability without detriment
to the selectivity.
It is worth noting that R(H2/CO2) = 1.5 for poly(exo-
N-phenyl-7-oxanorbornene-5,6-dicarboximide-co-nor-
bornene) membranes in spite of the fact that D(H2)/
D(CO2) ) 86. In this case, however, the low solubility
of hydrogen in comparison with that of carbon dioxide
is responsible for the poor discriminative properties of
the membrane for the separation of hydrogen from
carbon dioxide.
The rather low permeability coefficient of the mem-
branes, despite the relatively bulky side groups of the
monomer containing imide groups, could be caused by
the stereochemical compositions of polynorbornene
The performance of the membranes described in this
work for separation of hydrogen from N2, O2, Ar, CO,
CO2, CH4, C2H4, and C2H6 is depicted in Figure 9. The
3
2,33
values of R(H /B) represented in terms of increasing
chains. It is well-known
that materials obtained by
2
values of the permeability coefficient of B gives a curve
ROMP may give rise to the formation of blocks of cis
and/or trans configurations having different specific
chain lengths. The incompatibility between blocks could
result in microphase separation that increases the
tortuosity of the diffusive path, thus decreasing the
diffusion coefficient.
that shows a strong decrease of the separation coef-
ficient with increasing values of P(B). The values of R-
(
H2/N2), R(H2/CH4), and R(H2/C2H6) are significantly
higher than those reported for poly(vinyltrimethylsilyl-
norbornene) and fluorine-containing ring-opened polynor-
bornenes. However, the permselectivity factors depict-
9
The separation or selectivity factor for two gases A
and B can be written as
ing the separation of hydrogen from the other gases are
on average similar to those reported for membranes of
poly[(1-adamantyl)-exo-norbornene-5,6-dicarboximide]
and poly(N-cyclohexyl-exo-norbornene-5,6-dicarboxim-
ide) as well as for copolymers of N-(1-adamantyl)-exo-
norbornene-5,6-dicarboximide/norbornene, with molar
P(A) D(A) S(A)
R(A/B) )
)
(2)
P(B) D(B) S(B)
Equation 2 indicates that the discriminative effects to
gas transport in membranes may be governed by the
diffusive step, the sorption process, or both. In general,
polynorbornene is characterized for having rather low
permeability accompanied by low permselectivity. For
1
5
compositions 70/30, 50/50, and 30/70.
The membranes exhibit a separation factor R(O2/N2)
) 4.7, somewhat higher than that reported for the
membranes prepared from polynorbornenes containing
imide functions in their structure, with the exception
of the copolymer (1-adamantyl)-exo-norbornene-5,6-di-
carboximide/norbornene (50/50), for which R(O2/N2) )
9
example, the value of R(O2/N2) for polynorbornene is
only 1.9, significantly lower than the value of 5 or larger
1
reported for polyimides, polysulfones, etc. An increase
1
6
in the tacticity of polynorbornene has proved to enhance
5.5.

8
488 Tlenkopatchev et al.
Macromolecules, Vol. 36, No. 22, 2003
Con clu sion s
(8) Finkelshtein, E. Sh.; Gringolts, M. L.; Ushakow, N. V.;
Lakhtin, V. G.; Sloviev, S. A.; Yampol’skii, Yu. P. Polymer
The synthesis of poly(exo-N-phenyl-7-oxanorbornene-
2
003, 44, 2483.
5
,6-dicarboximide-co- norbornene) is reported. The Lang-
(9) Yampol’skii, Yu. P.; Bespalova, N. B.; Finkelshtein, E. Sh.;
Bondar, V. I.; Popov, A. V. Macromolecules 1994, 27, 2872.
10) Finkelshtein, E. Sh.; Makovetskii, K. L.; Yampol’skii, Yu. P.;
Portnykh, E. B.; Ostrovskaja, I. Ya.; Kaliuzhnyi, N. E.;
Pritula, N. A.; Gol’berg, A. I.; Yatsenko, M. S.; Plate, N. A.
Makromol. Chem. 1991, 192, 1.
(11) Bondar, V. I.; Kukharskii, Yu. M.; Yampol’skii, Yu. P.;
Finkelshtein, E. Sh.; Makovetskii, K. L. J . Polym. Sci., Part
B: Polym. Phys. 1993, 31, 1273.
muir capacity of the most condensable gases in mem-
branes of this polymer prepared by casting is relatively
low, as one would expect from the relatively low glass
transition temperature of the copolymer. The mem-
branes exhibit pretty good properties to separate hy-
drogen from other gases such as nitrogen, methane,
ethylene, and ethane, in comparison with other mem-
branes prepared from norbornenes containing fluorine
atoms or organosilicon moieties.
A shortcoming of these membranes is their low
permeability presumably caused by either the good
packing of the chains or by the increase in the tortuosity
of the diffusive path arising from the possible existence
of segregate stereoregular blocks of cis and trans
configurations. It would be important to increase the
permeability of the membranes either by using polym-
erization methods that enhance the stereoregularity of
the chains or by fluorinating or introducing organosili-
con moieties in norbornenes containing imide groups in
the structure.
(
(12) Steinh a¨ usler, T.; Koros, W. J . J . Polym. Sci., Part B: Polym.
Phys. 1997, 35, 91.
(13) Contreras, A. P.; Tlenkopatchev, M. A.; Ogawa, T.; Naka-
gawa, T. Polym. J . 2002, 34, 49.
(14) Yampol’skii, Yu. P.; Shishatskii, S.; Alentiev, A.; Loza, K. J .
Membr. Sci. 1998, 149, 203.
(
15) Contreras, A. P.; Tlenkopatchev, M. A.; L o´ pez-Gonz a´ lez, M.
M.; Riande, E. Macromolecules 2002, 35, 4677.
16) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999,
1, 953.
(
(17) Laguna, M. F.; Saiz, E.; Guzm a´ n, J .; Riande, E. Macromol-
ecules 1998, 31, 7488.
18) Saiz, E.; L o´ pez-Gonz a´ lez, M. M.; Riande, E.; Guzm a´ n, J .;
Compa n˜ , V. Phys. Chem. Chem. Phys. 2003, 5, 2862.
19) Anderson, W. K.; Milowsky, A. S. J . Org. Chem. 1985, 50,
5423.
(
(
(
20) Crank, J . The Mathematics of Diffusion; Oxford University
Press: New York, 1975.
Ack n ow led gm en t. We thank the CONACyT for
(
(
(
(
(
21) Barrer, R. M. Trans. Faraday Soc. 1939, 35, 628.
22) Vieth, W. R.; Sladek, K. J . J . Colloid Sci. 1965, 20, 1014.
23) Reference 2, p 60.
24) van der Vegt, N. F. A. J . Membr. Sci. 2002, 205, 125.
25) Koros, W. J .; Paul, D. R. J . Polym. Sci., Polym. Phys. Ed.
1978, 16, 1947.
generous support of this research with Contract No. NC-
2
04. The authors thank Miguel Angel Canseco, J uan
Manuel Garcia Leon, and Gerardo Cedillo for their
assistance in thermal, GPC, and NMR analyses. Sup-
port of this work by the CAM (Spain) thorugh Grant
Bio-009-2000 is also gratefully acknowledged.
(26) Muruganandam, N.; Koros, W. J .; Paul, D. R. J . Polym. Sci.,
Part B: Polym. Phys. 1987, 25, 1999.
(
27) Flory, P. J . Principles of Polymer Chemistry; Cornell Univer-
Refer en ces a n d Notes
sity Press: Ithaca, NY, 1953; Chapter 21.
(
(
(
(
(
1) Kesting, R. E.; Fritzsche, A. K. Polymeric Gas Separation
Membranes; Wiley-Interscience: New York, 1993.
2) Ohya, H.; Kudryavtsev, V. V.; Semenova, S. I. Polyimide
Membranes; Gordon and Breach Publishers: Tokyo, 1996.
3) Tlenkopatchev, M. A.; Fomine, S.; Miranda, E.; Fomina, L.;
Ogawa, T. Polym. J . 1995, 27, 1173.
4) Tlenkopatchev, M. A.; Fomine, S.; Fomina, L.; Gavi n˜ o, R.;
Ogawa, T. Polym. J . 1997, 29, 622.
5) Maya, V. G.; Contreras, A. P.; Canseco, M. A.; Tlenkopatchev,
M. A. React. Funct. Polym. 2001, 49, 145.
(28) Petropoulos, J . H. Pure Appl. Chem. 1993, 65, 219.
(29) Compa n˜ , V.; L o´ pez-Gonz a´ lez, M. M.; Riande, E. Macromol-
ecules, in press.
(30) Breck, D. W. Zeolite Molecular Sieves; Wiley-Interscience:
New York, 1974; p 636.
(31) Costello, L. M.; Koros, W. J . Polym. Sci., Part B: Polym. Phys.
1994, 32, 701.
(32) Ivin, K. J .; Mol, J . C. Olefin Metathesis and Metathesis
Polymerization; Academic Press: San Diego, CA, 1997.
(
33) Reid, R. C.; Prausnitz, J . M.; Poling, B. E. The Properties of
(
6) Asrar, J . Macromolecules 1992, 25, 5150.
7) Contreras, A. P.; Cerda, A. M.; Tlenkopatchev, M. A. Mac-
romol. Chem. Phys. 2002, 203, 1811.
Gases and Liquids, 4th ed.; McGraw-Hill: New York, 1987.
(
MA030285A