Synthesis, characterization, and evaluation of novel polyhydantoins as gas separation membranes

Article abstract of DOI:10.1002/pola.26806

Novel polyhydantoins (PHYs) were synthesized from original aromatic diisocyanates and bisiminoacetates by a two-step polycondensation procedure, which involved the cyclization of polyurea intermediates promoted by acid catalysis. The physical properties of the novel PHYs were evaluated by comparing them with a classical PHY derived from 4,4′-methylenediphenyl diisocyanate. All PHYs were soluble and could be processed into dense films, which showed good mechanical properties (tensile strength up to 110 MPa) and thermal stability of >400 °C. High glass transition temperatures (T _gs), ranging from 260 to 410 °C, were observed. Fractional free volume (FFV) was strongly dependent on the chemical structure, and a linear correlation between gas permeability and FFV of PHYs could be found. The gas separation properties were comparable to those of the commercial polyimide Matrimid, with the exception of one of the PHYs which exhibited very promising properties as its gas productivity was comparable to the gas separation performance of well-established experimental polyimides.

Full text of DOI:10.1002/pola.26806

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Synthesis, Characterization, and Evaluation of Novel Polyhydantoins as
Gas Separation Membranes
ꢀ
ꢀ
ꢀ
Ruben Tejero, Angel E. Lozano, Cristina Alvarez, Javier de Abajo
ꢀ
ꢀ
ꢀ
Departamento de Quımica Macromolecular Aplicada, Instituto de Ciencia y Tecnologıa de Polımeros, Consejo Superior de
ꢀ
Investigaciones Cientıficas (ICTP-CSIC), Madrid 28006, Spain
Correspondence to: J. de Abajo (E-mail: deabajo@ictp.csic.es)
Received 1 April 2013; accepted 25 May 2013; published online 23 July 2013
DOI: 10.1002/pola.26806
ABSTRACT: Novel polyhydantoins (PHYs) were synthesized
from original aromatic diisocyanates and bisiminoacetates by
a two-step polycondensation procedure, which involved the
cyclization of polyurea intermediates promoted by acid cataly-
sis. The physical properties of the novel PHYs were evaluated
by comparing them with a classical PHY derived from 4,40-
methylenediphenyl diisocyanate. All PHYs were soluble and
could be processed into dense films, which showed good
mechanical properties (tensile strength up to 110 MPa) and
thermal stability of >400 ꢀC. High glass transition tempera-
tures (Tgs), ranging from 260 to 410 ꢀC, were observed. Frac-
tional free volume (FFV) was strongly dependent on the
permeability and FFV of PHYs could be found. The gas sepa-
ration properties were comparable to those of the commercial
R
polyimide MatrimidV, with the exception of one of the PHYs
which exhibited very promising properties as its gas produc-
tivity was comparable to the gas separation performance of
C
V
well-established experimental polyimides.
2013 Wiley Peri-
odicals, Inc. J. Polym. Sci., Part A: Polym. Chem. 2013, 51,
4052–4060
KEYWORDS: diisocyanate; fractional free volume; heteroatom-
containing polymers; high performance polymers; membrane
gas separation; polyhydantoin; structure-property relations;
thermal properties
chemical structure, and
a linear correlation between gas
structural characteristics to be considered as film-forming
polymer materials for gas separation and purification. Never-
theless, it seems that no significant research effort has been
devoted to this application so far. The absence of research on
this field in recent years has moved us to initiate a study to
assess the suitability of PHYs for gas separation applications.
In this regard, an advantageous characteristic of these poly-
mers relies on their high chemical versatility as both mono-
mers, diisocyanates and bisiminoacetates, can be produced
from a widespread number of commercial diamines.
INTRODUCTION Heterocylic polymers such as polyimides,
polybenzimidazoles and polybenzoxazoles have generated a
significant level of interest in both industry and academia in
recent decades.1 Particularly, aromatic polyimides have found
application in aeronautics and electrical–electronic industries,
thanks to their outstanding mechanical and thermal proper-
ties. In recent years, they have received special attention as
membrane-forming polymers for gas separation operations,
which are directly related to environmental protection and
energy-saving challenges.2–4 In fact, there are many examples
of polyimides with high fractional free volume (FFV) and
which are consequently highly permeable to vapor and
gases.5–8 Aromatic polyhydantoins (PHYs) are a class of poly-
heterocycles which underwent short but important technical
developments in the 1970s, but they have not received special
interest in recent years.9–12 PHYs, which are commonly syn-
thesized by the polycondensation of bisiminoacetates and dii-
socyanates, have been used successfully as copper-wire
enamel insulators and high-temperature adhesives.13–17 These
polymers are mainly notable owing to their mechanical prop-
erties, chemical resistance and their exceptional thermal sta-
bility. As with other polyheterocycles, PHYs possess attractive
As has been well established, for a polymer to exhibit prom-
ising properties for gas separation, it should show a rela-
tively high free volume, to permit the easy transportation of
gas molecules. Additionally, it should possess the capability
to efficiently separate gases of different molecular sizes,
which is mainly determined by chain rigidity, that is molecu-
lar stiffness.
Herein, PHYs have been synthesized from newly designed
monomers, which contain the structural elements which pre-
dictably would work in terms of both good selectivity and
Additional Supporting Information may be found in the online version of this article.
C
V
2013 Wiley Periodicals, Inc.
4052
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
high permeability. Thus, monomers were prepared contain-
ing fluorene units, which provide favorable chain stiffness
and improve gas selectivity. Furthermore, bulky side groups
were incorporated which can inhibit the dense packing of
polymer chains and improve the permeability.18 To investi-
gate the effect of these “cardo” bridging units and bulky
groups on the chain stiffness and free volume, a classical
PHY from methylene diphenyl diisocyanate (MDI) was also
prepared and evaluated as a reference.
4.21 (c, J 5 7.2 Hz, 4H; 2OCH2), 3.84 (s, 4H; 2CH2), 1.27
(t, J 5 7.2 Hz, 6H; 2CH3); 13C NMR (75 MHz, CDCl3, d): 171.6
(2C@O), 148.6 (2C), 130.6 (2C), 104.1 (2C), 61.7 (2C), 46.3
(2C), 14.7 (2C); HRMS (ESI, m/z): [M 1 H]1 calcd for
C
C
14H20N2O4, 281.1423; found, 281.1501. Anal. calcd for
14H20N2O4: C 60.40, H 7.14, N 10.00; found: C 60.00, H 7.21,
N 9.54.
Dimethyl 2,20-[((9H-fluorene-9,9-diyl)bis(4,1-
phenylene))bis(azanediyl)] diacetate
Particular attention was paid to characterize and conveniently
evaluate the set of synthesized polymers, and a study has been
accomplished to find out structure–property relationships to
figure out the real capabilities of these PHYs as special poly-
mer materials, and in particular as polymer membranes for
gas separation This approach allowed for the critical evalua-
tion of the prepared PHYs in terms of special polymer materi-
als, specifically for this application. For this assessment, their
separation performances were compared to that of a commer-
Over a stirred mixture of 9,9-bis(4-aminophenyl)fluorene
(10.7 g, 0.030 mol) and 2,6-lutidine (8.4 mL, 0.072 mol) dis-
solved in acetonitrile (500 mL) methyl bromoacetate (7.3 mL,
0.066 mol) was added dropwise and the reaction mixture was
refluxed for 8 h. The final reaction mixture was partitioned
with ethyl acetate/water, the combined organic layers were
dried over magnesium sulfate, and the solvent was removed
under vacuum. The residue was purified by column chroma-
tography over silica gel (hexane–ethyl acetate, 2:1 v/v), to
yield 8.0 g (53%) of dimethyl 2,20-[((9H-fluorene-9,9-diyl)-
bis(4,1-phenylene))bis(azanediyl)] diacetate (BIS-2) as a white
solid. mp: 132–133 ꢀC.
R
cial polyimide, MatrimidV,19 which is a polymer widely used in
gas separation. Furthermore, the present PHYs were also com-
pared with the experimental polyimide 6FDA-6FpDA,20 which
is a highly fluorinated polyimide considered as a reference for
experimental, high-performance gas polymer membranes. The
modulation of the composition of these PHYs allowed for the
study of the effects of the chemical structure on general prop-
erties such as solubility, thermal resistance, crystallization
ability, and mechanical strength.
1H NMR (300 MHz, CDCl3, d): 7.65 (dd, J 5 7.6 Hz, 2H; Ar),
7.32 (d, J 5 7.3 Hz, 2H; Ar), 7.29 (dd, J 5 1.3, 7.6 Hz, 2H; Ar),
7.16 (dd, J 5 1.3, 7.3 Hz, 2H; Ar), 6.97 (d, J 5 8.6 Hz, 4H; Ar),
6.36 (d, J 5 8.6 Hz, 4H; Ar), 3.97 (s, 2H; 2NH), 3.74 (s, 4H;
2CH2), 3.64 (s, 6H; 2OCH3); 13C NMR (75 MHz, CDCl3, d):
171.5 (2C@O), 152.1 (2C), 145.4 (2C), 139.8 (2C), 135.6 (2C),
128.9 (4C), 127.4 (2C), 126.9 (2C), 125.9 (2C), 119.8 (2C),
112.5 (4C), 63.9, 52.0 (2C), 45.5 (2C); HRMS (ESI, m/z):
[M 1 H]1 calcd for C31H28N2O4, 493.2049; found, 493.2131.
Anal. calcd for C31H28N2O4: C 75.59, H 5.73, N 5.69; found: C
75.37, H 5.46, N 5.76.
EXPERIMENTAL
Materials
Methyl bromoacetate (97%), ethyl bromoacetate (98%), m-
phenylenediamine (MPD, 99%), triethylamine (TEA, 99%), tri-
phosgene (98%), and anhydrous 1-methyl-2-pyrrolidinone
(NMP, 99.5%) were purchased from Aldrich and were used
without further purification. 9,9-Bis(4-aminophenyl) fluorene
(98%) was purchased from TCI Europe and 1,4-bis(2-trifluor-
Synthesis of Aromatic Diisocyanates
9,9-Bis(4-isocyanatophenyl)29H-fluorene
Over a stirred mixture of triphosgene (2.08 g, 0.007 mol) in
ꢀ
anhydrous toluene (50 mL) at 0 C under a nitrogen atmos-
omethyl-4-aminophenoxy)22,5-bis-tert-butylbenzene
was
phere, a solution of 9,9-bis(4-aminophenyl)fluorene (5.0 g,
0.014 mol) in anhydrous toluene (50 mL) was dropped for 1
h. Then, the mixture was refluxed for 10 h. After cooling, the
solvent was evaporated under vacuum and the residue was
extracted several times with anhydrous hexane. The combined
organic layers were dried over magnesium sulfate. Removal of
the solvent under vacuum yielded a residue which was subli-
mated under vacuum at 180 ꢀC to yield 2.3 g (83%) of 9,9-
bis(4-isocyanatophenyl)29H-fluorene (DII-1) as a white solid.
mp: 140–141 ꢀC; Attenuated total reflectance-Fourier trans-
form infrared (ATR-FTIR): 3046 (@CH), 2258 (N@C@O) cm21
purchased from Aurora Fine Chemicals and were used as
received; m-cresol (99%) was purchased from Aldrich and
was distilled before use. 4,4’-Methylene-bis(phenylisocyanate)
(4,40-MDI) was purchased from Aldrich and was sublimated
before use. Other solvents and reactants were all purchased
from Aldrich and were used without further purification.
Synthesis of Aromatic Bisiminoacetates
Diethyl 2,20-[(1,3-phenylenebis(azanediyl))] diacetate
A mixture of MPD (5.0 g, 0.046 mol) and ethyl bromoacetate
(11.7 mL, 0.106 mol) was refluxed in TEA (60 mL) for 30
min under a nitrogen atmosphere. After cooling, the solvent
was evaporated under vacuum and the residue was purified
by column chromatography over silica gel (hexane–ethyl ace-
tate, 1:1 v/v) to yield 7.5 g (58%) of diethyl 2,20-[(1,3-phe-
nylenebis(azanediyl))] diacetate (BIS-1) as a white solid. mp:
1H NMR (300 MHz, CDCl3, d): 7.74 (d, J 5 8.5 Hz, 2H; Ar),
7.35 (m, J 5 1.5, 6.9 Hz, 2H; Ar), 7.29-7.25 (m, J 5 1.5, 6.9,
8.5 Hz, 4H; Ar), 7.09 (d, J 5 8.6 Hz, 4H; Ar), 6.92 (d, J 5 8.6
Hz, 4H; Ar); 13C NMR (75 MHz, CDCl3, d): 149.9 (2NCO),
142.8 (2C), 139.6 (2C), 131.6 (2C), 128.7 (6C), 127.5 (2C),
127.4 (2C), 125.3 (2C), 124.2 (4C), 119.9 (2C), 64.1. Anal.
calcd for C27H16N2O2: C 80.99, H 4.03, N 7.00; found: C
81.03, H 4.25, N 6.98.
ꢀ
70–71 C.
1H NMR (300 MHz, CDCl3, d): 6.98 (dd, J 5 8.1 Hz, 1H; Ar),
6.03 (dd, J 5 2.2, 8.1 Hz, 2H; Ar), 5.84 (dd, J 5 2.3 Hz, 1H; Ar),
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060
4053

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
4,40-[(2,5-Di-tert-butyl-1,4-phenylene)bis(oxy)]bis(1-
isocyanato-3-trifluoromethylbenzene)
Microanalyses were made with a Carlo Erba EA1108 elemen-
tal analyzer. ATR-FTIR spectra were recorded on a Perkin
Elmer RX-1 instrument with Universal ATR Sampling Acces-
sory module. Inherent viscosities were measured at 25 ꢀC
with an Ubbelohde viscometer using NMP as solvent at a
concentration of 0.5 gꢁdL21. Differential scanning calorimetry
(DSC)ꢀdata were recorded on a TA Q-2000 device in nitrogen
at 20 Cꢁmin21. Thermogravimetric analysis (TGA) data were
recorded on a TA Q-500 analyzer using Hi-ResTM option. Hi-
Res scans were run at 50 Cꢁmin with resolution and sen-
sitivity parameters of 4.0 and 1.0, respectively, in the tem-
perature range of 25–850 ꢀC. The purge gas was nitrogen
(60 mLꢁmin21) and the sample mass was about 5 mg. Ten-
sile strength, elongation at break, and Young’s module of the
membranes were measured with a MTS Synergie 200 appa-
ratus fitted with a 100 N load cell at room temperature. The
samples of 5 mm width and 5 cm length were clamped at
both ends with an initial gauge length of 10 mm at an elon-
gation rate of 5 mm min21. At least eight samples were
tested for each film. Wide-angle X-ray scattering (WAXS) pat-
terns were recorded in the reflection mode at room tempera-
ture, using a Bruker D8 Advance diffractometer provided
with a Goebel Mirror and a PSDVantec detector. CuKa (wave-
length, k 5 1.54 Å) radiation was used. A step-scanning
mode was used for the detector, with a 2h step of 0.024ꢀ
and 0.5 s per step.
In
a
similar way, 4,40-[(2,5-di-tert-butyl-1,4-phenylene)
bis(oxy)]bis(1-isocyanato-3-trifluoromethylbenzene) (DII-2)
was synthesized from 1,4-bis(2-trifluoromethyl-4-amino-
phenoxy)22,5-bis-tertbutylbenzene and triphosgene. It was
isolated in 72% of yield after purification by sublimation
under vacuum at 220 ꢀC. mp: 166–167 ꢀC.
21
ꢀ
ATR-FTIR: 3025 (@CH), 2990 (ACH), 2980 (ACH), 2270
(N@C@O) cm21 1H NMR (300 MHz, CDCl3, d): 7.43 (d,
.
J 5 2.6 Hz, 2H; Ar), 7.18 (dd, J 5 2.6, 8.9 Hz, 2H; Ar), 6.85–
6.75 (m, 4H; Ar), 1.27 (s, 18H; 6CH3); 13C NMR (75 MHz,
CDCl3, d): 150.2 (4C), 137.9 (2C), 129.3 (2C), 129.0 (2C),
128.9 (q, 1JCAF 5 271.4 Hz, 2C; CF3), 128.2 (4C), 120.1 (q,
2JCAF 5 33.1 Hz, 2C; CF3), 118.8 (4C), C30H26F6N2O4: C 60.81,
H 4.42, N 4.73; found: C 60.53, H 4.32, N 4.47.
Synthesis of PHYs
As an example of the general synthetic route to PHYs,21–24
the preparation of the homopolymer PHY-3 is described
here in detail: A mixture of BIS-2 (2.46 g, 0.005 mol) and
DII-1 (2.00 g, 0.005 mol) in 5 mL of freshly distilled m-cre-
ꢀ
sol was heated to 50 C and stirred under a nitrogen atmos-
phere for 4 h. As the reaction progressed, the solution
became very viscous. The second stage of the reaction
(chemical cyclization) was carried out by the addition of 0.5
mL of concentrated HCl and 5 mL of m-cresol. Then, the
reaction was heated at 50 ꢀC for 3 h. The viscous solution
obtained was poured into 400 mL of ethanol (EtOH), the
precipitated polymer was filtered, and washed several times
with EtOH, and it was finally dried to constant weight in a
Fractional Free Volume
FFV was calculated from the density (q) of polymer films
using the following relation:
V21:3Vw
FFV 5
(1)
ꢀ
vacuum oven at 200 C for 12 h. The yield was quantitative.
V
ATR-FTIR: 3056 (@CH), 1776 (C@O), 1718 (C@O) cm21 1H
.
where V 5 1/q is the polymer-specific volume (cm3ꢁmol21),
Vw is the van der Waals volume, which was estimated by
using the Hyperchem 7.0 computer program.25 This program
considers the dependence of van der Waals volume on the
chemical environment and it can estimate Vw for the struc-
tural unit of a polymer chain.
NMR (300 MHz, CDCl3, d): 7.95 (m, J 5 8.4 Hz, 2H; Ar), 7.55 (d,
J 5 7.8 Hz, 4H; Ar), 7.46–7.28 (m, J 5 7.8 Hz, 8H; Ar), 7.14 (d,
J 5 8.4 Hz, 2H; Ar), 4.55 (s, 2H; CH2); 13C NMR (75 MHz,
CDCl3, d): 168.5, 153.5, 150.8, 150.3, 145.6, 141.2, 139.9,
136.9 (2C), 130.9 (2C), 128.6 (4C), 128.4 (4C), 128.2, 127.4,
126.6, 126.4, 121.1, 118.9 (2C), 64.9, 50.3; Anal. calcd for
(C28H18N2O2)n: C 81.14, H 4.38, N 6.76; found: C 78.93, H 4.36,
N 6.62.
The density was determined experimentally using a top-
loading electronic XS105 Dual-Range Mettler Toledo balance
coupled with a density kit based on Archimide’s principle.
The samples were weighed in air and into a known-density
liquid (high-purity isooctane). The measurement was per-
formed at room temperature and the density was calculated
as follows:
Film Preparation
PHYs dense films were prepared as follows: a 10% solution
of polymer in NMP (w/v) was filtered through a 3.1-mm
R
fiberglass SymtaV syringe filter and, after degassing, the
polymer solution was poured into a glass ring placed on a
leveled glass plate, and the solvent was slowly evaporated at
60 C for 8 h, and then at 80 C for 3 h. Polymer films were
peeled off from the glass and the residual solvent was evapo-
rated under vacuum for 12 h at 250 ꢀC. Thicknesses of the
films ranged from 50 to 80 mm.
ꢀ
ꢁ
x
air 2xliquid
q5qliquid
(2)
ꢀ
ꢀ
xair
where qliquid is the density of isooctane, xair is the weight of
the sample in air, and xliquid is its weight when is submerged
in isooctane.
Characterization
1H and 13C NMR spectra were recorded on a Bruker 300
spectrometer operating at 300 and 75 MHz. Melting points
Permeability Measurements
The gas permeation properties were determined for pure
€
were determined using
a
Buchi 504392-S instrument.
gas feeds using a constant volume/variable pressure
4054
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
RESULTS AND DISCUSSION
Monomers and Polymers
Bisiminoacetates BIS-1 and BIS-2 were prepared from the
corresponding diamines and alkyl bromoacetates by the gen-
eral method reported previously.26,27 Unreported diisocya-
nates DII-1 and DII-2 were prepared by the method
reported elsewhere from diamines and using triphosgene as
the acylation agent.28 These electrophilic monomers were
obtained in high yield and could be purified to monomer
grade by sublimation. DII-1 and DII-2 were identified by
spectroscopic methods. Data of IR and NMR were consistent
with the proposed structures as shown in Figure 1, where
the spectra of DII-1 are shown as examples. The appearance
of the characteristic band of the isocyanate group in the
ATR-FTIR spectra, around 2200 cm21 (N@C@O), confirmed
the chemical structure of the monomer; in the same way, all
peaks of the NMR spectra could be easily assigned to the
protons of the monomer. Additional data on the spectro-
scopic characteristics of all the monomers have been sum-
marized in Supporting Information material.
The polycondensations of diisocyanates and bisiminoacetates
were carried out using the classical method in solution in m-
cresol as shown in Scheme 1. The precursor polyureas,
which were formed in the first step of the reaction, were
converted into the final PHYs by an acid-catalyzed cycliza-
ꢀ
tion at moderate temperature (approx., 50 C). The complete
cyclization of precursors to PHYs was confirmed by the
appearance in the ATR-FTIR spectra of two characteristics
bands of the carbonyls in the hydantoin ring at 1770 and
1710 cm21 (C@O), and by the absence of bands around
3300 (NAH) and 1690 (C@O) cm21 of the urea groups. As
FIGURE 1 ATR-FTIR (top) and 1H NMR (bottom, CDCl3) spectra
of DII-1.
1
an example, Figure 2 shows the ATR-FTIR and H-NMR spec-
tra of PHY-3. ATR-FTIR, 1H and 13C NMR spectra, and ele-
mental analyses confirmed that PHYs were successfully
synthesized in high yield (91–98%) and with full conversion
via the polyurea chemical cyclization method (Supporting
Information).
apparatus at 30 ꢀC. The downstream pressure was kept
below 1022 mbar, whereas the upstream pressure was main-
tained at 3 bar.
Permeability coefficients (P) were determined from the slope
of downstream pressure versus time, plotted once steady
state had been achieved, according to the equation:
The polymer’s chemical composition and the results of the
polycondensation reactions have been listed in Table 1, along
with inherent viscosities measured for the set of polymers
and their solubility in organic solvents. All polymers showed
good solubility in highly polar organic solvents such as m-
cresol, NMP and DMAc. Polymers PHY-3 and PHY-4 were
soluble even in THF and CHCl3. Inherent viscosities in the
range of 0.40–0.70 dLꢁg21 should correspond to medium to
high molecular weights.
273 Vl dpðtÞ
P5
(3)
76 ATp0 dt
where A and l are, respectively, the effective area and the
thickness of the film, T is the measurement temperature in
K, and p0 is the pressure of the feed gas in the upstream
chamber. P is usually expressed in barrers [1 barrer 5 10210
(cm3 (STP)ꢁcm)ꢁ(cm22ꢁs21ꢁcmHg21)]. To evaluate the possi-
ble existence of micropores (pinholes) in the membranes,
they were subjected to a preliminary test to measure the
permeability to Helium at different pressures, from 1 to 5
bar. To ensure reproducibility of results, each membrane was
measured three times. The ideal selectivities of the O2/N2
and CO2/CH4 gas pairs were calculated from the ratio of
pure gas permeabilities (aPA/PB).
Consistent with this assumption, films of good quality with
high mechanical resistance could be attained by the casting
of polymer solutions in all cases (Table 2). Polymer films
showed elastic moduli close to 2.0 GPa and tensile strength
values of between 50 and 110 MPa, indicating that these
polymers have excellent mechanical properties, comparable
to other polyheterocycles and engineering thermoplastics.29
These values allow these polymers to be considered as
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060
4055

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 1 Synthesis of PHYs.
suitable candidates for the preparation of dense membranes
able to withstand high gas pressures for long periods of
time.
hydantoin rings are present. The trend observed in the Tgs
was consistent with the chain stiffness that was predicted.
Thus, PHY-2 and PHY-3 have Tg values of 345 and 405 ꢀC,
respectively, as corresponds to the presence of bulky fluo-
rene moieties. On the other hand, PHY-1 and PHY-4 do not
reach such high values of Tg because both polymer have
ACH2A and AOA linkages between the phenylene rings,
which work for a higher conformational freedom and lower
Tg. Endothermic peaks, which could be associated with
enthalpy of fusion, were not observed in any case (Fig. 4),
therefore these polymers should be considered as amor-
phous in nature.
TGA was used to investigate the thermal resistance (Fig. 3).
Data summarized in Table 2 reveal that these materials pos-
sess high temperatures of initial decomposition, Td, with two
of them, PHY-2 and PHY-3, showing Td higher than 440 ꢀC.
ꢀ
All of them left char residues close to 50% at 800 C.
DSC was used to determine the glass transition temperature,
Tg, of these polymers. The samples were subjected to a heat-
ing–cooling–heating cycle, the first heating until 300 ꢀC to
ensure the removal of water and traces of solvent. Tgs are
listed in Table 2 where it can be seen that all of the PHYs
To confirm this, the molecular order of the current PHYs
was also investigated by WAXS. As expected, the X-ray pat-
terns showed the typical halo seen in amorphous polymers
(Fig. 5). Moreover, PHYs containing fluorene units exhibited
a more or less defined maximum, close to 14ꢀ, which corre-
sponds to an intersegmental distance of 6.3 Å, indicating
that these bulky units cause chain separation and hinder effi-
cient chain packing.
ꢀ
have Tgs higher than 250 C. This agrees well with the chem-
ical composition of these polymers, where only aromatic and
Evaluation of PHYs as Gas Separation Membranes
On applying eq (1), values of FFV were obtained through the
experimental density of the polymer films (Table 3). PHY-1
presented the lowest value of FFV (0.081), whereas PHY-4
showed the highest FFV (0.194). The values of FFV followed
a trend just opposite to that of densities, according to a cor-
rect application of eq (1). PHY-1 was the only polymer with
no bulky fluorene moiety in its repeating unit and it also did
not have side substituents, which made it in terms of effi-
cient molecular packing and high density. PHY-2 and PHY-3
showed similar values of FFV although their chemical com-
positions were very different. The presence of diphenylme-
thane units in PHY-2 should not give as high FFV as the
fluorene unit does in PHY-3, but the molecular heterogeneity
provided by the different nature of the starting diisocyanate
and bis-iminoacetate of PHY-2 can explain the reason why
both polymers exhibit similar values of density and FFV. The
high FFV of PHY-4 can be attributed to the presence of bulky
groups such as tert-butyl and trifluoromethyl along with the
fluorene units. One could presume that the molecular mobil-
ity that generally impart the AOA linkages to aromatic poly-
mers would allow for good molecular packing and a higher
FIGURE 2 ATR-FTIR (top) and 1H NMR (bottom, DMSO-d6)
spectra of PHY-3.
4056
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
TABLE 1 Yield, Inherent Viscosity (ginh) and Solubility of PHYs
Ar-R-Ar0
Solvent
NMP
Yield
(%)
ginh
(dL g21
Polymer
PHY-1
Ar (Bisiminoacetate)
Ar0 (Diisocyanate)
)
m-Cresol
THF
2
CHCl3
2
98
96
0.69
0.72
1
1
1
1
PHY-2
2
2
PHY-3
PHY-4
91
94
0.42
0.52
1
1
1
1
1
1
1
1
TABLE 2 Thermal and Mechanical Properties of PHYs Films
Polymer
Tensile Strength (MPa)
Elongation at Break (%)
Young’s Module (GPa)
Tg (ꢀC)
Char Yieldb (%)
a
Td (ꢀC)
PHY-1
PHY-2
PHY-3
PHY-4
110 6 10
100 6 10
50 6 10
60 6 10
10 6 1
14 6 4
2.0 6 0.2
1.8 6 0.2
2.3 6 0.7
2.0 6 0.3
265
345
405
310
430
450
490
440
52
48
55
49
2.5 6 0.4
3.5 6 0.3
a
Onset temperature at which decomposition begins.
Char yield at 800 ꢀC.
b
density for PHY-4, but it is not so because the presence of
tert-butyl groups placed ortho to the AOA linkage greatly
restricts the rotation around the PhAOAPh bonds, as it has
been reported previously.20
PHY-4 presented the best permeability, even higher than that
of 6FDA-6FpDA, as could be expected of a polymer that has a
very high FFV. However, the selectivity CO2/CH4 was signifi-
cantly lower than that of the reference polymer. Their proper-
ties are actually also fairly comparable with those of some
recently reported polyimides.30,31 As it has been reported, it
is possible to establish a correlation between log P and
1/FFV for a wide variety of macromolecular structures.32–34
The gas permeabilities and selectivities of PHYs are listed in
R
Table 4. PHY-2 showed values similar to MatrimidV, whereas
those of PHY-1 were considerably lower in agreement with its
lower FFV. PHY-3 exhibited higher permeability than
R
MatrimidV although CO2/CH4 and O2/N2 selectivities were
Park and Paul35 reported that this correlation is accurate
when a specific family of polymers is studied. Also, they
lower.
FIGURE 3 TGA thermograms of PHYs.
FIGURE 4 DSC thermograms of PHYs.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060
4057

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
R
TABLE 4 Permeabilities and Selectivities of PHYs, MatrimidV,19
and 6FDA-6FpDA20
Permeability (barrer)
Selectivity
aPCO2/ aPO2/
Polymer
PHe PN2 PO2 PCH4 PCO2 PCH4
PN2
PHY-1
PHY-2
PHY-3
PHY-4
4.6 0.025 0.17 0.015 0.85
57
31
18
21
36
35
6.8
5.8
5.6
4.8
7.0
5.0
16
29
0.33
0.75
5.2
1.9
4.2
25
0.32
0.77
5.2
10
14
120
22
110
8.7
70
R
MatrimidV
0.27
3.6
1.9
18
0.24
2.0
6FDA-6FpDA 151
On plotting selectivity (a) against permeability (P) in a chart,
which includes the upper bound limits of Robeson (1991 and
2008)36,37 for a given pair of gases, a promising behavior of
these polymers in this kind of separation could be observed.
FIGURE 5 WAXS patterns of PHYs.
found a much better correlation when FFV was calculated
considering the occupied volume, which is specific for each
gas, irrespective of the polymer studied. To find out whether
PHYs fit the classical correlation between gas permeability
and FFV, a graph is shown in Figure 6, where the values of
The productivity of gas separation for this family of PHYs is
shown in Figure 7, which also includes the values of
R
MatrimidV and 6FDA-6FpDA. PHY-1 showed a moderate per-
meability/selectivity balance for the gas pairs CO2/CH4
R
and O2/N2. PHY-2 had similar values to MatrimidV for the
R
MatrimidV and 6FDA-6FpDA are also included. A fairly good
correlation for the PHYs and reference polyimides could be
observed, both for CO2 and for O2.
This relationship could be employed for further research on
these polymers which should be accomplished in the near
future.
R
TABLE 3 Densities and FFV of PHYs, MatrimidV,19 and 6FDA-
6FpDA20
Polymer
q (g cm23
)
FFV
1.345
0.081
1.256
1.249
0.102
0.112
1.217
0.194
1.266
1.466
0.110
0.208
FIGURE 6 Log P 2 1/FFV diagram for CO2 (top) and O2
(bottom).
4058
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
bulky side groups, such as tert-butyl and trifluoromethyl,
gave comparatively high FFV as well.
The presence of fluorene units in the backbone of PHYs sig-
nificantly improves the gas permeabilities, with acceptable
losses in selectivity, in accordance with the highest FFV dis-
posable for gas transport. Nonetheless, it was necessary to
introduce additional bulky side groups in the monomers to
get gas separation properties comparable to those of the ref-
erence 6FDA-6FpDA polyimide.
Data reported here indicate that the convenient modification
of the starting monomers can lead to PHYs with outstanding
physical properties and with gas permeation capabilities
superior to those of many polyimides. This study should be
considered as an original approach to reach valuable infor-
mation to design new materials, with improved gas separa-
tion properties for selected applications, mainly related to
CO2 capture and air component separation.
ACKNOWLEDGMENTS
The authors are indebted to the Ministry of Economy and Com-
petitiveness of Spain (MINECO) for the economic support of
this work (MAT2010–20668/MAT and MAT2011–25513/
MAT). The authors also acknowledge financial support from
the program Consolider Ingenio 2010 (project CSD-0050-MUL-
TICAT). The help provided by Sara Rodriguez and Alberto Tena
in measuring gas permeability is greatly appreciated. R. Tejero
thanks CSIC for a predoctoral JAE fellowship. The Authors
acknowledge Maria Nash for the revision of this manuscript.
REFERENCES AND NOTES
FIGURE 7 Selectivity–permeability (a-P) diagram for CO2/CH4
pair (top) and O2/N2 pair (bottom).
1 T.-S. Chung, J. Polym. Sci. Part C: Polym. Rev. 1997, 37, 277–
301.
CO2/CH4 separation, O2/N2 pair. PHY-3 had lower selectivity
2 B. Sedlacek, J. Kahovec, J. Polym. Sci. Part C: Polym. Lett.
1987, 26, 242.
for CO2/CH4 although its permeability was better than that
R
of MatrimidV. The best results, with a high permeability to
3 M. Mulder, Basic Principles of Membrane Technology;
Kluwer Academic Publishers: London, 1991.
O2 and CO2 along with a good selectivity, were by far for
PHY-4. Furthermore, it showed better performance in terms
of permeability/selectivity than 6FDA-6FpDA for the O2/N2
pair as it lays closer to the Robeson limit. However, for the
CO2/CH4 pair, PHY-4 presented lower productivity than
6FDA-6FpDA. Nonetheless, this PHY was better for gas sepa-
ration than many glassy polymers reported so far.38,39
4 K. Scott, Handbook of Industrial Membranes; Elsevier:
Oxford, 1995.
5 C. E. Sroog, Prog. Polym. Sci. 1991, 16, 561–694.
6 R. Wang, C. Cao, T.-S. Chung, J. Membr. Sci. 2002, 198, 259–
271.
7 J. de Abajo, J. G. De la Campa, A. E. Lozano, J. Espeso, C.
ꢀ
Garcıa, Macromol. Symp. 2003, 199, 293–305.
8 Y. Xiao, B. T. Low, S. S. Hosseini, T.-S. Chung, D. R. Paul,
Prog. Polym. Sci. 2009, 34, 561–580.
CONCLUSIONS
€
9 R. Merten, W. Dunwald, K.-H. Mielke, E. Reese (Bayer AG).
It has been demonstrated that it is possible to prepare new
PHYs in high yield based on an optimized synthetic
approach by using new monomers derived from commercial
aromatic diamines. These PHYs can be processed into dense
films with good mechanical properties. Furthermore, these
PHYs are amorphous materials, soluble in many organic sol-
vents, which have high thermal stability.
U.S. 3,397,253, 1965.
10 R. Merten, W. Zecher (Farbenfabriken Bayer AG). U.S.
3,448,170, 1969.
11 J. A. Preston, F. Wayne (Essex International). U.S. 3,681,282,
1972.
12 K. B. Wagener, B. W. Spivey, J. M. Chapman (Akzona Incor-
porated). U.S. 4,053,441, 1977.
This research has also proved that PHYs containing fluorene
bridging groups have high Tgs. The additional presence of
13 E. Reese, J. Wank, R. Binsack, (Bayer Aktiengesellschaft).
U.S. 4,206,110, 1980.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060
4059

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
€
14 W. Zecher, R. Merten, W. Dunwald (Bayer Aktiengesell-
27 K. Yasuda (Kokai Tokkyo Koho), JP 133483, 2006.
schaft). U.S. 4,517,353, 1985.
28 W. Xiang-Hui, H. Yong-Ning, Appl. Chem. Ind. 2008, 37,
15 G. Heywang, F. Jonas, R. Merten, W. Schmidtberg (Bayer
Aktiengesellschaft). U.S. 4,849,492, 1989.
1019.
29 J. M. Margolis, Ed. Engineering Thermoplastics; Dekker:
New York, 1985.
16 F. Jonas, R. Merten (Bayer Aktiengesellschaft). U.S.
5,006,641, 1991.
30 Y. Yampolskii, B. D. Freeman, I. Pinnau, Materials Science
of Membranes for Gas and Vapor Separation; John Wiley &
Sons Ltd: New York, 2006.
17 C.-I. Tseng (Lonza). U.S. 5,785,963, 1998.
18 Y. Yampolskii, Macromolecules 2012, 45, 3298–3311.
31 H. Lin, B.D. Freeman, J. Mol. Struct. 2005, 739, 57–74.
19 M. D. Guiver, G. P. Robertson, Y. Dai, F. Bilodeau, Y. S.
Kang, K. J. Lee, J. Y. Jho, J. Polym. Sci. Part A: Polym. Chem.
2002, 40, 4193–4204.
32 C. L. Aitken, W. J. Koros, D. R. Paul, Macromolecules 1992,
25, 3651.
20 M. Calle, A. E. Lozano, J. de Abajo, J. G. de la Campa, C.
Alvarez, J. Membr. Sci. 2010, 365, 145–153.
33 T. H. Kim, W. J. Koros, G. R. Husk, K. C. O’Brien, J. Membr.
Sci. 1988, 37, 45.
ꢀ
ꢀ
21 R. H. Salle, B. J. Sillion, G. P. Gaudemaris, J. Polym. Sci.
34 D. H. Weinkauf, D. R. Paul, J. Polym. Sci. Part B: Polym.
Phys. 1992, 30, 837.
Part A: Polym. Chem. 1966, 4, 2903–2906.
22 Y. Imai, J. Polym. Sci. Part A: Polym. Chem. 1967, 5, 2289–
2296.
35 J. Y. Park, D. R. Paul, J. Membr. Sci. 1997, 125, 23–39.
36 L. M. Robeson, J. Membr. Sci. 1991, 62, 165–185.
37 L. M. Robeson, J. Membr. Sci. 2008, 320, 390–400.
23 M. M. Koton, Vysokomol. Soyed. A18, 1976, 9, 2092.
24 R. Merten, Angew. Chem. Int. Ed. 1971, 10, 294–301.
38 M. M. Dal-Chin, A. A. Kumar, A. Layton, J. Membr. Sci.
2008, 323, 299–308.
25 Hyperchem, Computational Chemistry, version 7.0, Hyper-
cube Inc., Gainesville, Florida, 2002.
39 L. M. Robeson, B. D. Freeman, D. R. Paul, B. W. Rowe, J.
Membr. Sci. 2009, 341, 178–185.
26 H. Kwon, Bull. Korean Chem. Soc. 1997, 18, 476–480.
4060
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4052–4060