Novel interfacially-polymerized polyamide thin-film composite membranes: Studies on characterization, pervaporation, and positron annihilation spectroscopy

Article abstract of DOI:10.1016/j.polymer.2011.03.049

To improve the pervaporation performance of thin-film composite membranes, novel thin-film composite membranes were prepared via interfacial polymerization by reacting 5-nitrobenzene-1,3-dioyl dichloride (NTAC) or 5-tert-butylbenzene- 1,3-dioyl dichloride

Full text of DOI:10.1016/j.polymer.2011.03.049

Polymer 52 (2011) 2414e2421
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Novel interfacially-polymerized polyamide thin-film composite membranes:
Studies on characterization, pervaporation, and positron annihilation
spectroscopy
,
d
c
c
Wei-Chi Chao a, Shu-Hsien Huang b , Quanfu An , Der-Jang Liaw , Ying-Chi Huang ,
**
,
a
Kueir-Rarn Lee a , Juin-Yih Lai
*
a R&D Center for Membrane Technology, Department of Chemical Engineering, Chung Yuan University, Chung Li 32023, Taiwan
b Department of Chemical and Materials Engineering, National Ilan University, I-Lan 26047, Taiwan
c Department of Chemical Engineering, National Taiwan University of Science and Technology, Taipei 10607, Taiwan
d Department of Polymer Science and Engineering, Key Laboratory of Macromolecule Synthesis and Functionalization (Ministry of Education),
Zhejiang University, Hangzhou 310027, China
a r t i c l e i n f o
a b s t r a c t
Article history:
To improve the pervaporation performance of thin-film composite membranes, novel thin-film
composite membranes were prepared via interfacial polymerization by reacting 5-nitrobenzene-
1,3-dioyl dichloride (NTAC) or 5-tert-butylbenzene-1,3-dioyl dichloride (TBAC) with triethylenetetra-
amine (TETA) on the surface of a modified polyacrylonitrile (mPAN) membrane (TETA-NTAC/mPAN and
TETA-TBAC/mPAN). The effect of the acyl chloride monomers chemical structure on the pervaporation
separation of an aqueous ethanol solution was investigated. Attenuated total reflectance Fourier trans-
form infrared (ATR-FTIR) spectroscopy, X-ray photoelectron spectroscopy (XPS), atomic force microscopy
(AFM) and water contact angle measurements were applied to analyze the chemical structure, surface
chemical composition, surface roughness and hydrophilicity of the polyamide active layer of the
composite membrane. To correlate the variations in the free volume of the polyamide active layers with
the pervaporation performance, positron annihilation spectroscopy (PAS) experiments were performed
with a variable monoenergetic slow positron beam. From the results of the PAS and XPS experiments, the
S parameter, o-Ps annihilation lifetime s3 (corresponding to free volume size) and its intensity I3 (cor-
Received 7 November 2010
Received in revised form
20 March 2011
Accepted 24 March 2011
Available online 31 March 2011
Keywords:
Polyamide thin-film composite membrane
Positron annihilation spectroscopy
Pervaporation
responding to free volume concentration), the s3 and I3 of TETA-NTAC polyamide layer (positron incident
energy of 1e1.7 keV) were both higher than those of TETA-TBAC polyamide layer. The S parameter for
TETA-NTAC polyamide layer was also higher than that of the TETA-TBAC polyamide layer even though the
former was more crosslinking than that of the latter. In the aqueous ethanol solution dehydration
experiments, the TETA-NTAC/mPAN membrane produced both a higher permeation rate and water
concentration in the permeate than the TETA-TBAC/mPAN membrane.
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
generally used to concentrate ethanol aqueous solutions. However,
water/ethanol azeotropes in the ethanol aqueous solution restrict
Ethanol aqueous solution, a biofuel, is a product of sugar fer-
mentation in biomass resources, such as the stems of corn. It is a new
energy source being studied to address the concerns associated with
a lack of petroleum resources. Generally, the concentration of the
ethanol produced from fermentation is low. To obtain high pur-
ity ethanol, ethanol dehydration is performed using a separation
processes, such as distillation. Traditional distillation processes are
further concentration of the ethanol. To break the azeotrope, azeo-
tropic distillationwith an azeotrope-breaking component (entrainer)
or a hybrid system of multistage evaporation and distillation is used.
These two processes consume large amounts of energy and are not
economical. Pervaporation, an important membrane separation
process in chemical industries, can effectively separate azeotropic
mixtures and decrease the cost and energy consumption. Therefore,
the pervaporation separation process is suitable to be applied in the
dehydration of ethanol.
* Corresponding author. Tel.: þ886 3 2654190; fax: þ886 3 2654198.
** Corresponding author. Tel.: þ886 3 9357400; fax: þ886 3 9357025.
E-mail addresses: huangsh@niu.edu.tw (S.-H. Huang), krlee@cycu.edu.tw
(K.-R. Lee).
A composite membrane combines the advantages of a dense
membrane and an asymmetric membrane. It has the potential
to obtain a high permeation rate and a high selectivity during
0032-3861/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2011.03.049

W.-C. Chao et al. / Polymer 52 (2011) 2414e2421
2415
pervaporation. Interfacial polymerization is a useful technique to
Amine:
apply a thin selective top layer to a composite membrane. Poly-
amides are an excellent material for the dense selective layer
because of its high selectivity, high thermal stability, good
mechanical strength and high resistance to organic solvents. Inter-
facially-polymerized polyamide thin-film composite membranes
are usually studied for reverse osmosis [1e5] or nanofiltration
applications [6e10], but there have been few reports on pervapo-
ration [11e15]. In this study, polyamide thin-film composite
membranes were prepared via interfacial polymerization of the
amine monomer TETA and two novel acyl chlorides, NTAC and TBAC,
on the surface of modified polyacrylonitrile (mPAN) support
membranes. The effect of the monomer chemical structure on the
pervaporation separation performance of a 90 wt% aqueous ethanol
solution was investigated. The polyamide layer characteristics (i.e.,
surface roughness, hydrophilicity, degree of crosslinking and free
volume) were correlated with the pervaporation performance.
The free volume has a significant effect on the separation
performance of the membrane in the pervaporation operation.
However, the measurement of the free volume is difficult because
the free volume is a dynamic pore. Positron annihilation spec-
troscopy (PAS) with a variable monoenergetic slow positron beam
is a powerful technique to measure the free volume of a material.
The chemical and structural effects of the polymer on the free
volume have been reported [16e19], but there has not been much
work on the depth profile for the free volume variation in the thin-
film composite membranes [14]. The variations in the free volume
of these novel polyamide thin-film composite membranes were
measured by positron annihilation spectroscopy (PAS) with a vari-
able monoenergetic slow positron beam.
H
N
H2N
N
NH2
H
a
Acyl chloride:
ClOC
COCl
ClOC
COCl
NO2
H3C
CH3
CH3
b
c
Fig. 1. Chemical structures of the monomers used in preparing the polyamide active
layers. (a) TETA, (b) NTAC, (c) TBAC.
2.3. Synthesis of monomer TBAC acyl chloride
The diacyl chloride monomers of 5-tert-butylbenzene-1,3-dioyl
dichloride were synthesized by a mixture of 5-tert-butylbenzene-
1,3-dioic acid (3 g, 13.50 mmol), thionyl chloride (30 mL, 0.413 mol)
and 10 drops of DMF. The mixture was placed in a round-bottomed
flask, equipped with a reflux condenser. After refluxing for 24 h,
thionyl chloride was removed by distillation (120 ꢀC). The crude
product was purified by vacuum distillation (5 mmHg, 120 ꢀC)
twice to obtain white particles with 71% yield (2.48 g, 9.58 mmol).
The synthesis of TBAC was shown in Scheme 1(b). The spectro-
2. Experimental
2.1. Materials
scopic data for 1H NMR (500 MHz, CDCl3, ppm):
d8.4947e8.4919 (t,
1H, AreH,
J
¼
1.0000 Hz); d8.2741e8.2711 (d, 2H, AreH,
Polyacrylonitrile (PAN) polymer was supplied by the Tong-Hua
Synthesis Fiber Co. Ltd. in Taiwan. Reagent grade N-methyl-2-pyr-
rolidone (NMP) was used in preparing the casting solution of the
PAN polymer. Triethylenetetraamine (TETA) was purchased from
Merck Co. 5-Nitrobenzene-1,3-dioyl dichloride (NTAC) and 5-tert-
butylbenzene-1,3-dioyl dichloride (TBAC) were synthesized in our
laboratory. Amine (TETA) and acyl chloride monomers (NTAC and
TBAC) were used for the interfacial polymerization to form the
polyamide active layer. Distilled water was used in preparing the
aqueous amine solution, and reagent grade tetrahydrofuran (THF)
was used as the acyl chloride monomer solvent. The chemical
structures of the monomers used to prepare the polyamide active
layers via interfacial polymerization are shown in Fig. 1.
J ¼ 2.0553 Hz);
d
1.3879 (s, 9H, eCH3, J ¼ 9.7501 Hz). 13C NMR
(125 MHz, CDCl3, ppm):
34.70, 30.89.
d166.22, 151.82, 130.59, 130.15, 127.76,
2.4. Preparation of modified PAN (mPAN) porous membrane
supports
Flat PAN porous membrane supports were prepared by casting
PAN onto non-woven polyester fabrics. In a continuous preparation,
a polymer solution, containing 15 wt% of PAN, was cast onto the
non-woven polyester fabric by the use of a casting knife with
200 mm gap. The cast membrane was precipitated by immersion in
a water bath. The resulting PAN porous membrane supports were
washed in water several times for more than 1 day to remove the
remnants of the NMP solvent. In order to enhance the hydrophi-
licity of membrane support to benefit the sorption of the amine
2.2. Synthesis of monomer NTAC acyl chloride
The diacyl chloride monomers of 5-nitrobenzene-1,3-dioyl
dichloride were synthesized from a mixture of 5-nitrobenzene-
1,3-dioic acid (3 g, 14.21 mmol), thionyl chloride (30 mL, 0.413 mol)
and 1 mL of DMF. The mixturewas placed in a round-bottomed flask,
equipped with a reflux condenser. After refluxing for 24 h, thionyl
chloride was removed by distillation at 120 ꢀC. The crude product
was purified by vacuum distillation (5 mm Hg, 120 ꢀC) twice to
obtain white particles with 76% yield (2.67 g, 10.79 mmol). The
synthesis of NTAC was shown in Scheme 1(a). The spectroscopic
HOOC
COOH
ClOC
COCl
SOCl2
80oC, 24hr
Dried DMF
R
R
data for 1H NMR (500 MHz, CDCl3, ppm):
d9.2161e9.2128 (d, 2H,
R= (a) -NO2
or
AreH, J ¼ 1.9278 Hz); 9.1149e9.1083 (t, 1H, AreH, J ¼ 1.0000 Hz).
d
(b) -C(CH3)3
13C NMR (125 MHz, CDCl3, ppm):
130.84.
d165.60, 148.88, 137.38, 136.01,
Scheme 1. Synthesis of diacyl chloride monomer (a) NTAC or (b) TBAC.

2416
W.-C. Chao et al. / Polymer 52 (2011) 2414e2421
monomer in it, the PAN porous membrane support was hydrolyzed
by NaOH solution [20]. The modified PAN porous membrane was
prepared by immersing the PAN porous membrane support in a 2 M
NaOH solution at 50 ꢀC. The partial eCN groups of PAN can be
converted into eCOOH and eCONH2 groups after undergoing
hydrolysis in a NaOH solution. The mPAN porous membrane was
washed in a water bath for several hours to remove the remnants of
the NaOH solution.
2.7. Light transmission experiment
The reactivity between the aqueous amine solution and the
organic acyl chloride solutions was observed with the light trans-
mission experiments. In the light transmission experiment, the
beaker containing the aqueous TETA solution was placed on
a detector, and a light source was installed above the beaker
vertically. The light source was turned, letting the light transmit
through the aqueous TETA solution until the light transmission
value reached steady state. The organic diacryl chloride (NTAC or
TBAC) solution was poured into the beaker carefully and slowly
along its inside wall. When the organic diacryl chloride solution
contacted the light, the light transmission value was started to be
recorded by the detector. The light transmittance was calculated by
the following equation:
2.5. Preparation of the polyamide thin-film composite membrane
The polyamide thin-film composite membrane was prepared by
the interfacial polymerization of the amine monomer TETA and the
acyl chloride monomer (NTAC or TBAC). The mPAN membrane was
first immersed in a 1 wt% aqueous amine solution for 30 s. Then, the
excess aqueous amine solution on the surface of the mPAN
membrane was removed. The mPAN membrane, soaked with the
aqueous amine solution, was brought into contact with a 1 wt%
organic acyl chloride solution for 30 s to carry out the interfacial
polymerization. Finally, the resulting polyamide thin-film com-
posite membrane was washed in methanol and then dried at
atmospheric temperature. The schematic diagram for the chemical
reaction showing the structure of the synthesized polyamide as
a result of the interfacial polymerization between TETA and diacyl
chloride (NTAC or TBAC) on the surface of the mPAN membrane was
shown in Scheme 2.
Ei
E0
Light transmittance ¼
(1)
where E0 is the initial light transmission value (light transmission
value before pouring organic diacryl chloride solution into the
beaker), and Ei is the light transmission value after pouring organic
diacryl chloride solution into the beaker at each time interval.
When the organic diacryl chloride solution contacted the
aqueous TETA solution, a free-standing polyamide membrane was
formed at the interface between the solutions by means of inter-
facial polymerization. With the passage of time, the growth of the
free-standing polyamide membrane affected the variation in the
light transmittance, which was used to observe the reactivity
between the aqueous amine solution and the organic acyl chloride
solution.
2.6. Surface characterization
The surface roughness of the polyamide thin-film composite
membranes was observed with an atomic force microscope (AFM;
Digital Instruments, DI-NS3a, USA). Attenuated total reflectance
Fourier transform infrared spectroscopy (ATR-FTIR; PerkinElmer
Spectrum One) and X-ray photoelectron spectroscopy (XPS; Ther-
moFisher Scientific K-Alpha) were used to characterize the chem-
ical structure and the surface composition of the polyamide active
layer of the thin-film composite membrane, respectively. To
understand the surface hydrophilicity of the polyamide thin-film
composite membrane, the water contact angle was estimated with
an automatic interfacial tensiometer (FACE Mode 1 PD-VP).
2.8. Positron annihilation spectroscopy
PAS was used to estimate the free volume variations in the depth
profile of the polyamide thin-film composite membrane. A variable
monoenergetic slow positron beam as a function of the positron
incident energy (0e30 keV) at atmospheric temperature under
a vacuum of w10ꢁ8 torr. The radioisotope beam used 50 mCi of 22Na
as the positron source. Two positron annihilation spectrometers
were installed in this beamdDoppler broadening energy spec-
troscopy (DBES) and positron annihilation lifetime spectroscopy
Cross-linking
R
H
O
NH
N
O
N
H
HN
H
O
O
NH
N
H
N
N
R
CHN
N
O
O
NH
O
H
H2N
N
R
NH
N
H
NH2
R
O
HN
NH
O
-HCl
NH
HN
NH
on mPAN
HN
O
O
HN
O
NH
NH
Cl
Cl
+H3N
+H3N
R
O
Ionic Bond
R
CN
COOH
CONH2
R: -NO2 or -C(CH3)3
Scheme 2. Schematic diagram for the chemical reaction showing the structure of the synthesized polyamide as a result of the interfacial polymerization between TETA and diacyl
chloride (NTAC or TBAC) on the surface of the mPAN membrane.

W.-C. Chao et al. / Polymer 52 (2011) 2414e2421
2417
(PALS). The DBES was measured using an HP Ge detector at
a counting rate of approximately 2000 cps. The total number of
counts for each DBES was 1.0 million. The DBES is based on the
measured width of the annihilation gamma photon line, centered
at 511 keV. The Doppler-broadened linewidth can be described
with simple lineshape parameters, such as the shape (S) parameter.
The S parameter, defined as the ratio of the central part of
the annihilation spectrum to the total spectrum, reflects the posi-
tron annihilation with low momentum valence electrons. The S
parameter data from DBES was fitted by VEPFIT program. The PALS
experiment provides more quantitative information on the free
volume properties in polymeric systems. The PALS data were
obtained by taking the coincident events between the start signal
detected by a multichannel plate (MCP) from the secondary ele-
ctrons and the stop signal discerned by a BaF2 detector from the
Z
X
W =ZE
W =XE
aS
¼
(4)
where XW and XE are the weight fractions of water and ethanol in
the feed. ZW and ZE are the weight fractions of water and ethanol in
the membrane, respectively.
The water vapor ratio, which was absorbed in the membrane,
was calculated from the following equation:
ꢀ ꢁ
WS
WD
Water vapor sorption ratio
%
¼
(5)
where WD and WS are the weights of the dry membrane and the
water vapor absorbed in the membrane, respectively.
3. Results and discussion
annihilation photons at
a counting rate of approximately
200e300 cps. A PALS spectrum contains 2.0 million counts. From
the analysis of positron annihilation lifetime spectrum, positron
lifetimes (s1, s2 and s3) and relative intensities (I1, I2 and I3), which
3.1. Characterization of the polyamide thin-film composite
membranes
result from the positron and positronium annihilation in polymeric
membranes, can be determined. The lifetime s3 from pick-off
annihilation of ortho-positronium (o-Ps) and electron, which is on
the order of 1e5 ns in polymeric membranes, is related to the size
of the free volume [21]. s3 is usually used to calculate the mean
free-volume radius. I3 may directly indicate the concentration of
the free volume because of the amount of the formed Ps annihi-
lating with the electrons in microcavities may correspond to the
free volume. That is, an increase in I3 is approximately equivalent to
an increase in the concentration of the free volume. Therefore, s3
and I3 from PALS can be used to explore the depth profiles of the
size and concentration distributions of the free volume in poly-
meric thin-film composite membranes. All of the positron annihi-
lation lifetime spectra were analyzed by a finite term lifetime
analysis method using the PATFIT program [22].
The chemical structures of the polyamide thin-film composite
membranes were characterized by ATR-FTIR. Fig. 2 shows the
ATR-FTIR spectra of the mPAN and polyamide thin-film composite
membranes. Compared with the spectrum of mPAN (Fig. 2(a)), an
increase in the intensity of the peaks at wave number 1580 cmꢁ1
,
corresponding to the NeH (amide II) of the amide group, is shown
in the spectra for the polyamide thin-film composite membranes
(TETA-NTAC/mPAN and TETA-TBAC/mPAN) (Fig. 2(b) and (c)).
A shift of the amide I peak from 1631 cmꢁ1 (C]O of the acrylamide
group in mPAN) to 1664 cmꢁ1 (C]O of the amide group in the
polyamide) is shown in the spectra (Fig. 2). In addition, the peak at
1613 cmꢁ1, corresponding to the vibrational modes and the C]C
stretching of the aromatic ring, is shown in Fig. 2(b) and (c). The
results reveal that the active layers of the thin-film composite
membranes were composed of aromatic polyamides.
The surface chemical compositions of the polyamide thin-film
composite membranes were investigated by XPS analysis. Fig. 3
shows the C1s X-ray photoelectron spectra of the polyamide thin-
film composite membranes, and their relative bond assignments
from these spectra are shown in Table 1. There are two peaks that
correspond to the eNeC]O and eCOOH groups at 287.6 eV and
288.9 eV, respectively, in the C1s spectra of the TETA-NTAC/mPAN
and TETA-TBAC/mPAN membranes. The eNeC]O group was
formed by the reaction of an amine group with an acyl chloride
group during the interfacial polymerization process, and the
eCOOH group was generated by the hydrolysis of the unreacted
2.9. Pervaporation and sorption measurements
The pervaporation separation of an aqueous ethanol solution
through the polyamide thin-film composite membrane was per-
formed. The pervaporation apparatus has been described in an
earlier study [23]. The feed solution directly contacted the poly-
amide side of the composite membrane. The effective area of the
membrane for pervaporation was 10.24 cm2. The operating
temperature (feed solution temperature) was 25 ꢀC. The concen-
trations of the feed solution and the permeate were measured by
gas chromatography (GC; China Chromatography 8700 T). The
permeation rate (P) and the pervaporation separation factor of
water/ethanol (aP) were calculated from the following equations:
W
At
P ¼
(2)
(3)
Y
X
W =YE
W =XE
aP
¼
where W is the weight of the permeate, A is the effective membrane
area and t is the sampling time. XW and XE are the respective weight
fractions of water and ethanol in the feed. YW and YE are the
respective weight fractions of water and ethanol in the permeate.
The sorption of 90 wt% aqueous ethanol solution or water vapor by
the polyamide thin-film composite membrane was measured at
25 ꢀC. The sorption measurements and apparatus were described in
our previous study [24]. The solution separation factor of water/
ethanol (aS) was obtained from the following equation:
Fig. 2. ATR-FTIR spectra for (a) the pristine PAN membrane support and the polyamide
thin-film composite membranes: (b) TETA-NTAC/mPAN and (c) TETA-TBAC/mPAN.

2418
W.-C. Chao et al. / Polymer 52 (2011) 2414e2421
Fig. 3. C1s X-ray photoelectron spectra of the polyamide thin-film composite membranes: (a) TETA-NTAC/mPAN and (b) TETA-TBAC/mPAN.
acyl chloride group. A greater number of eCOOH groups on the
polymer chain of the interfacially-polymerized polyamide indicates
that many acyl chloride groups did not react with amine groups,
resulting in a shorter polymer chain and a lower polymer chain
entanglement. As shown in Table 1, the concentration of
the eNeC]O and eCOOH groups were 8.39% and 1.92% for the
TETA-NTAC/mPAN membrane and were 2.98% and 3.85% for the
TETA-TBAC/mPAN membrane, respectively. The relative ratio of
the eCOOH to eNeC]O groups was higher for the TETA-TBAC/
mPAN membrane than the TETA-NTAC/mPAN membrane. This
implies that the TETA-NTAC/mPAN membrane had a greater degree
of polymer chain entanglement and/or crosslinking than the TETA-
TBAC/mPAN membrane. Thus, a higher selectivity in the pervapo-
ration separation process was obtained, resulting from the
membrane with a higher polymer chain entanglement and a higher
polymer chain packing density.
Mean Depth ( m)
3.05
a
4.83
6.89
1.74
0.46
0
0.490
0.488
0.486
0.484
0.482
0.480
0.478
0.476
0.474
0.472
0.470
0.468
0.466
TETA- NTAC
TETA-TBAC
3.2. DBES of the polyamide thin-film composite membranes
To investigate the effects of the acyl chloride chemical structure
on the fine structure of the polyamide active layer, PAS experi-
ments, coupled with a variable monoenergetic slow positron beam,
were performed to obtain the DBES. The S parameter increases
concomitantly with increase in the relative contribution of low
momentum electrons to the positron annihilation in the polymeric
membranes [14,21,25e27]. Thereby, the DBES can be applied to
measure the variation in the fine structure of a composite
membrane, and the multilayer structure of a composite membrane
can be analyzed by the variation in the S parameter. Fig. 4 shows the
S parameter versus the positron incident energy (or mean depth)
for the polyamide thin-film composite membranes. The S value
close to the surface increases rapidly with an increase in the posi-
tron incident energy lower than 0.5 keV. This is due to the back-
diffusion of the positron annihilation. The S value shows a peak-like
region for the positron incident energies from 0.7 keV to 2.0 keV,
which penetrate deeper into the membrane. This region corre-
sponds to the polyamide active layer. And then the S value comes to
0
5
10
15
20
25
30
Positron Energy (keV)
Mean Depth (nm)
b
0
33
102
196
311
445
0.490
0.488
0.486
0.484
0.482
0.480
0.478
0.476
0.474
0.472
0.470
0.468
0.466
TETA-NTAC
TETA-TBAC
Table 1
Relative bond assignments from the C1s X-ray photoelectron spectra of the poly-
amide thin-film composite membranes.
0
1
2
3
4
5
Positron Energy (keV)
Membranes
CeC and
CeH
C]C and
CeH
CeN and
CeO
NeC]O
eCOOH
Fig. 4. S parameter versus the positron incident energy (mean depth) for the poly-
TETA-NTAC/mPAN
TETA-TBAC/mPAN
15.57%
22.20%
11.62%
9.32%
30.25%
32.80%
8.39%
2.98%
1.92%
3.85%
amide thin-film composite membranes: (C) TETA-NTAC/mPAN and ( ) TETA-TBAC/
;
mPAN.

W.-C. Chao et al. / Polymer 52 (2011) 2414e2421
2419
Table 2
a plateau-like region for the positron incident energies from
2.0 keV to 5.0 keV. This region corresponds to the transition layer
from polyamide layer to dense skin layer of mPAN support. For the
positron incident energies from 5.0 keV to 15 keV, the S value
increases to a maximum. This region is a transition layer from dense
skin to porous support layer in the mPAN. The S value decreases
when the positron incident is energy higher than 15 keV. This
region remains in the porous support of the mPAN. In general, the
effective separation layer of a composite membrane is the dense
top layer. Thus, we focused on the variation in the S parameter for
this region of the polyamide active layer. As shown in Fig. 4, the
S parameter was higher for the TETA-NTAC polyamide layer than
the TETA-TBAC polyamide layer. This indicates that the free volume
size of the TETA-NTAC polyamide layer was larger than that of the
TETA-TBAC polyamide layer.
The polymer chain packing density and light transmittance was
affected by the polymer formation rate during the interfacial
polymerization. To observe the light transmittance variation of the
free-standing polyamide membrane formation during the interfa-
cial polymerization process, light transmission experiments were
conducted (Fig. 5). For the TETA-NTAC and TETA-TBAC polyamide
layers, the light transmittance decreased during the initial 8 s of the
polymerization. This indicates that the polyamide membrane was
growing. After the polymerization time exceeded 8 s, there was no
significant change in the light transmittance, implying that the
polyamide layer had stopped growing. Compared with the TETA-
NTAC polyamide layer, the TETA-TBAC polyamide layer had the
lower light transmittance. This may have been due to the difference
in the polyamide formation rate between the diamine and diacryl
chloride monomers. Fig. 5 shows that the polyamide formation rate
of TETA-TBAC layer is higher than that of the TETA-NTAC layer.
Thus, a high density skin layer forms instantaneously when the
aqueous TETA solution and organic TBAC solution come in contact.
This results in a better arrangement of the polymer chain at the
TETA-TBAC polyamide layer. These results agree with Fig. 4. The
lower S parameter of the TETA-TBAC polyamide layer indicates that
the polymer chain packing density was higher in the TETA-TBAC
polyamide layer than the TETA-NTAC polyamide layer.
Solubility parameters and their differences for the amine and the acyl chloride
monomers.
a
Monomers
dt (MPa1/2
)
Ddt (MPa1/2
)
TETA
NTAC
TBAC
19.10
23.78
20.74
e
4.68
1.64
a
The solubility parameter difference of TETA and acyl chloride.
higher affinity between the amine and the acyl chloride monomers
increased the likelihood of collision and their subsequent reaction.
In addition, to illustrate the variation in the S parameter along
the depth of the composite membrane, the S parameter data as
a function of the positron incident energy were fitted by the VEPFIT
program analysis in a four-layer model and the results showed in
Fig. 6. As shown in Fig. 6, the polyamide thin-film composite
membrane was divided into four layers. These four layers indicated
are as follows: (I) interfacially-polymerized polyamide (PA) layer,
(II) transition layer from polyamide layer to dense skin layer of
mPAN support, (III) dense skin þ porous support layer of mPAN, and
(IV) porous support layer of mPAN. The effective selective layer of
the thin-film composite membrane is the polyamide active layer.
The thickness of the polyamide active layer also estimated from the
VEPFIT program analysis. The thickness of the TETA-NTAC and
TETA-TBAC polyamide layers were about 196 and 153 nm, respec-
tively. From the result of DBES analysis, the TETA-NTAC/mPAN
membrane had a higher S parameter and a thicker thickness for the
polyamide active layer than the TETA-TBAC/mPAN membrane. In
the pervaporation separation process, an increase in the free
volume and an increase in the thickness for a polymeric membrane
are favorable to an increase in the permeation rate and an increase
in the selectivity, respectively.
3.3. PALS of the polyamide thin-film composite membranes
In general, the S parameter from DBES experiment qualitatively
represents the fine structure of the polymeric membranes. In fact,
the fine structure of the polymer can be dominated by the size and
the concentration of the free volume quantitatively. To examine the
In addition, the solubility parameter difference between TETA
and TBAC was lower than that between TETA and NTAC (as shown
in Table 2), indicating that TETA and TBAC have higher affinity. The
0.486
1.0
0.8
0.6
(IV) Porous mPAN layer
(I)PA layer
0.484
(III) Dense mPAN layer + Porous mPAN layer
0.482
0.480
(a)
0.4
(II)Transition layer
TETA-NTAC
TETA-TBAC
(a) TETA-NTAC/mPAN
0.478
(b)
0.0
(b) TETA-TBAC/mPAN
0.2
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Mean depth ( m)
0.0
0
5
10
15
20
25
30
Fig. 6. Schematic diagram of four-layer depth structure obtained by using VEPFIT
program analysis of S parameter data from DBES for (a) TETA-NTAC/mPAN and (b)
TETA-TBAC/mPAN membranes. Description of four layers: (I) interfacially-polymerized
polyamide (PA) layer, (II) transition layer from polyamide layer to dense skin layer of
mPAN support, (III) dense skin þ porous support layer of mPAN, and (IV) porous
support layer of mPAN.
Time(sec)
Fig. 5. Light transmittance of the free-standing polyamide membrane formed between
the aqueous TETA and the organic acyl chloride solutions in the interfacial polymeri-
zation process (-) TETA-NTAC and (C) TETA-TBAC.

2420
W.-C. Chao et al. / Polymer 52 (2011) 2414e2421
Table 4
The surface roughness values of the polyamide thin-film composite membranes.
Membranes
Rmsa (nm)
Rab (nm)
Rmaxc (nm)
TETA-NTAC/mPAN
TETA-TBAC/mPAN
21.5 ꢂ 3.3
34.1 ꢂ 5.7
16.9 ꢂ 2.8
27.1 ꢂ 4.3
170.7 ꢂ 38.8
219.3 ꢂ 16.0
a
Rms: root mean square roughness.
Ra: average roughness.
b
c
Rmax: maximum roughness.
3.4. Surface hydrophilicity and roughness
The surface hydrophilicity of the polyamide thin-film composite
membrane was estimated by the surface water contact angle and
water vapor sorption measurements. The results are shown in
Table 3. Compared with the TETA-NTAC/mPAN membrane, the
TETA-TBAC/mPAN membrane had a lower water contact angle and
a higher water vapor sorption ratio. This indicates that the TETA-
TBAC/mPAN membrane had a higher hydrophilicity, resulting in
a higher water vapor sorption ratio. In addition to the innate char-
acteristic of the polymeric membrane, the surface roughness affects
the water contact angle. For a hydrophilic polymer membrane,
a rougher membrane surface produces a lower water contact angle,
indicating higher affinity between water molecular and polymer
membrane [28,29]. The surface roughness of the polyamide thin-
film composite membranes was measured by AFM, and the results
are shown in Table 4. The root mean square roughness (Rms), the
average roughness (Ra) and the maximum roughness (Rmax) were all
higher for the TETA-TBAC/mPAN membrane than for the TETA-
NTAC/mPAN membrane. This suggests an increase in the hydro-
philicity of the TETA-TBAC/mPAN membrane.
Fig. 7. o-Ps annihilation lifetime (s3) and intensity (I3) as a function of positron inci-
dent energy for TETA-NTAC/mPAN and TETA-TBAC/mPAN membranes.
free volume quantitatively, the PALS experiment was carried out to
obtain the o-Ps annihilation lifetime (s3) and intensity (I3) which
corresponded to the size and the concentration of the free volume,
respectively. The result of the PALS experiment is shown in Fig. 7. In
general, s3 represents the size of the free volume. An increase in s3
means an increase in free volume size. I3 may not directly indicate
the concentration of the free volume because the Ps formation also
depends on the decay rates and chemical environments, however,
the effects of chemical quenching and chemical inhibition were not
observed evident in this study even the eNO2 functional group still
survived in the TETA-NTAC polyamide layer on the mPAN (From
Fig. 2(b), the peak at 1052 cmꢁ1 is corresponding to the CeN
stretching of eNO2 functional groups). The amount of the formed
Ps annihilating with the electrons in microcavities may be in
response to the free volume only by assuming in the absence of
these effects for the current system. Therefore, an increase in I3 can
be only used very approximately for an increase in the concentra-
tion of free volume. The following observations are based on Fig. 7:
3.5. Pervaporation performance
The pervaporation performance of the polyamide thin-film
composite membranes with a 90 wt% ethanol aqueous solution at
25 ꢀC is shown in Table 5. The mPAN support membrane had a poor
pervaporation performance with a permeation rate of 2149 g/m2-hr
and a water concentration in permeate of 19.6 wt%. To improve the
pervaporation performance of the mPAN support membrane, the
TETA-NTAC/mPAN and the TETA-TBAC/mPAN composite membranes
were prepared via interfacial polymerization. The composite
membranes hadbetter pervaporationperformances than the pristine
mPAN support membrane. Compared with the TETA-TBAC/mPAN
membrane, the TETA-NTAC/mPAN membrane had both a higher
permeation rate and water concentration in the permeate. This
may be a result of the higher relative ratio of eCOOH to eNeC]O
groups for the TETA-TBAC/mPAN membrane (Table 1). This suggests
that the TETA-NTAC/mPAN membrane had greater polymer chain
entanglement and/or crosslinking than that of the TETA-TBAC/mPAN
membrane. Thus, a higher selectivity in the pervaporation separation
process was obtained, resulting from the membrane with a higher
polymer chain entanglement and higher polymer chain packing
density. In addition, from the result of the PAS experiment (Figs. 6
and 7), the s3 and I3 of the TETA-NTAC polyamide layer are both
(1) Gradually decrease in both the s3 and I3 intensity with increase
in the positron incident energy from 0.7 to 2 keV for TETA-NTAC/
mPAN membrane and 0.7 to 1.7 keV for TETA-TBAC/mPAN
membrane. This means that the free volume size decrease gradu-
ally in the region of the polyamide layer. (2) An increase in s3 and I3
after 2 keV for TETA-NTAC/mPAN membrane and 1.7 keV for TETA-
TBAC/mPAN. There is a transition layer from polyamide layer to
dense skin layer of mPAN support. As shown in Fig. 7, the s3 and I3 of
the TETA-NTAC polyamide layer are both higher than those of the
TETA-TBAC polyamide layer, that is, the size and concentration of
the free volume for the TETA-NTAC polyamide layer are both higher
than those for the TETA-TBAC polyamide layer. This result corre-
sponded with the DBES analysis using VEPFIT program in Fig. 6 very
well. The higher S parameter for the TETA-NTAC polyamide layer is
contributed to the larger free volume size and higher free volume
concentration.
Table 5
Effect of the acyl chloride monomer chemical structure on the pervaporation
performance of the polyamide thin-film composite membranes for 90 wt% aqueous
ethanol solutions at 25 ꢀC.
Table 3
The surface water contact angle and water vapor sorption values of the polyamide
thin-film composite membranes at 25 ꢀC.
Membranes
Permeation rate (g/m2-hr) Water conc. in permeate (%)
Membranes
TETA-NTAC/mPAN 45.5 ꢂ 1.2
TETA-TBAC/mPAN
38.5 ꢂ 1.0
Water contact angle (ꢀ) Water vapor sorption ratio (%)
mPAN
TETA-NTAC/mPAN
TETA-TBAC/mPAN
2149 ꢂ 290
537 ꢂ 71
452 ꢂ 52
19.6 ꢂ 1.7
98.2 ꢂ 0.8
97.1 ꢂ 1.2
8.60 ꢂ 3.58
11.48 ꢂ 4.08

W.-C. Chao et al. / Polymer 52 (2011) 2414e2421
2421
Table 6
TETA-TBAC/mPAN membrane. The solution effect primarily
controlled the pervaporation separation behavior of the polyamide
thin-film composite membrane in the dehydration of a 90 wt%
aqueous ethanol solution at 25 ꢀC.
Effect of the acyl chloride structure on the sorption and diffusion properties of the
polyamide thin-film composite membranes for a 90 wt% aqueous ethanol solution at
25 ꢀC.
Membranes
aP
aS
59 ꢂ 28
129 ꢂ 39
aD
TETA-NTAC/mPAN
TETA-TBAC/mPAN
558 ꢂ 261
326 ꢂ 159
6.7 ꢂ 2.1
2.7 ꢂ 0.9
Acknowledgements
aP ¼ pervaporation separation factor.
aS ¼ solution separation factor.
aD ¼ diffusion separation factor.
The authors wish to sincerely thank the Ministry of Economic
Affairs, the Ministry of Education Affairs and the National Science
and Technology Program-Energy from NSC of Taiwan for financially
supporting this work.
higher than those of the TETA-TBAC polyamide layer, that is, the size
and concentration of the free volume for the TETA-NTAC polyamide
layer are both higher than those for the TETA-TBAC polyamide layer,
resulting in a higher pervaporation performance of the TETA-NTAC/
mPAN membrane was obtained.
References
[1] Kwak SY, Jung SG, Yoon YS, Ihm DW. J Polym Sci Part B Polym Phys 1999;37:
1429e40.
The effect of the acyl chloride structure on the sorption and
diffusion of a 90 wt % aqueous ethanol solution through the poly-
amide thin-film composite membranes was investigated at 25 ꢀC,
and the results are shown in Table 6. The results indicate that aS was
higher than aD for all polyamide thin-film composite membranes,
which implies that the solution dominates the separation behavior
of the polyamide thin-film composite membranes. Compared with
the TETA-NTAC/mPAN composite membrane, the TETA-TBAC/mPAN
composite membrane had a higher aS and a lower aD. The higher aS
for the TETA-TBAC polyamide layer resulted from a lower surface
water contact angle and a higher surface roughness (Tables 3 and 4),
signifying a higher hydrophilicity and effective contact area. Thus,
the permeation rate of TETA-NTAC/mPAN membranes and the
water concentration of the resultant permeate were higher than
those of the TETA-TBAC/mPAN composite membranes.
[2] Kim CK, Kim JH, Roh IJ, Kim JJ. J Membr Sci 2000;165:189e99.
[3] Rao AP, Joshi SV, Trivedi JJ, Devmurari CV, Shah VJ. J Membr Sci 2003;211:
13e24.
[4] Louie JS, Pinnau I, Ciobanu I, Ishida KP, Ng A, Reinhard M. J Membr Sci 2006;
280:762e70.
[5] Li L, Zhang S, Zhang X, Zheng G. J Membr Sci 2007;289:258e67.
[6] Oh NW, Jegal J, Lee KH. J Appl Polym Sci 2001;80:2729e36.
[7] Chen SH, Chang DJ, Liou RM, Hsu CS, Lin SS. J Appl Polym Sci 2002;83:1112e8.
[8] Mohammad AW, Hilal N, Seman MNA. J Appl Polym Sci 2005;96:605e12.
[9] Verissimo S, Peinemann KV, Bordado J. J Membr Sci 2006;279:266e75.
[10] Yang F, Zhang S, Yang D, Jian X. J Membr Sci 2007;301:85e92.
[11] Huang SH, Li CL, Hu CC, Tsai HA, Lee KR, Lai JY. Desalination 2006;200:387e9.
[12] Li CL, Huang SH, Liaw DJ, Lee KR, Lai JY. Sep Purif Technol 2008;62:694e701.
[13] Huang SH, Hsu CJ, Liaw DJ, Hu CC, Lee KR, Lai JY. J Membr Sci 2008;307:73e81.
[14] Huang SH, Lin WL, Liaw DJ, Li CL, Kao ST, Wang DM, et al. J Membr Sci 2008;
322:139e45.
[15] Huang SH, Jiang GJ, Liaw DJ, Li CL, Hu CC, Lee KR, et al. J Appl Polym Sci 2009;
114:1511e22.
[16] Wang ZF, Wang B, Ding XM, Zhang M, Liu LM, Qi N, et al. J Membr Sci 2004;
241:355e61.
[17] Mondal S, Hu JL, Yong Z. J Membr Sci 2006;280:427e32.
[18] Wang ZF, Wang B, Qi N, Ding XM, Hu JL. Mater Chem Phys 2004;88:212e6.
[19] Honda Y, Shimada T, Tashiro M, Kimura N, Yoshida Y, Isoyama G, et al. Radiat
Phys Chem 2007;76:169e71.
[20] Oh NW, Jegal J, Lee KH. J Appl Polym Sci 2001;80:1854e62.
[21] Chen HM, Hung WS, Lo CH, Huang SH, Cheng ML, Liu G, et al. Macromolecules
2007;40:7542e57.
4. Conclusion
The polyamide thin-film composite membranes were success-
fully prepared via interfacial polymerization of TETA with NTAC or
TBAC and were applied to the dehydration of an aqueous ethanol
solution. The following conclusions were obtained:
[22] Kirkegaard P, Pederson NJ, Eldrup M. PATFIT package. Roskilde: Riso National
Laboratory; 1989.
As a result of the PAS experiments, the TETA-NTAC polyamide
layer of the composite membrane had a higher S parameter,
a higher s3 (free volume size), a higher I3 (free volume concentra-
tion), as well as a thicker polyamide layer thickness, compared with
the TETA-TBAC polyamide layer.
From the analyses of the XPS, AFM and water contact angle
measurements, the TETA-NTAC/mPAN membrane had higher
degree of crosslinking, lower surface roughness of the active layer
and lower hydrophilicity than the TETA-TBAC/mPAN membrane.
[23] Teng MY, Lee KR, Fan SC, Liaw DJ, Huang J, Lai JY. J Membr Sci 2000;164:
241e9.
[24] Lee KR, Lai JY. J Appl Polym Sci 1995;57:961e8.
[25] Jean YC, Mallon PE, Schrader DM. Principles and applications of positron &
positronium chemistry. World Scientific Publishing Co. Pvt. Ltd.; 2003
(chapter 11).
[26] Huang SH, Hung WS, Liaw DJ, Li CL, Kao ST, Wang DM, et al. Macromolecules
2008;41:6438e43.
[27] Hung WS, De Guzman M, Huang SH, Lee KR, Jean YC, Lai JY. Macromolecules
2010;43:6127e34.
[28] Luo N, Stewart MJ, Hirt DE, Husson SM, Schwark DW. J Appl Polym Sci 2004;
92:1688e94.
[29] Zhu LP, Zhang XX, Xu L, Du CH, Zhu BK, Xu YY. Colloids Surf B 2007;57:
189e97.
The TETA-NTAC/mPAN membrane showed both
a higher
permeation rate and water concentration in permeate than the