Facile synthesis of five 2D surface modifiers by highly selective photocyclic aromatization and efficient enhancement of oxygen permselectivities of three polymer membranes by surface modification using a small amount of the 2D surface modifiers

Article abstract of DOI:10.1016/j.polymer.2013.11.046

A facile synthesis of novel five 2D (planar) surface modifiers having a triphenylbenzene derivatives as a 2D structure has been achieved by the highly selective photocyclic aromatization reaction. Efficient enhancement of oxygen permselectivities through the three polymer membranes has been achieved by adding a small amount (<5.0 wt%) of the 2D surface modifiers. Among the five 2D surface modifiers, a modifier compound having oligoethylene oxide groups showed the best performance for the enhancement. These improvements were thought to be caused mainly by improvement of the solution selectivity on the membrane surface where the 2D surface modifiers were accumulated. In some of the surface-modified blend membranes, their plots in the PO₂-α graph were over or close to the upper boundary line by Robeson in 1991. Since all the membranes containing the 2D surface modifiers showed better permselectivities than the corresponding substrate membranes, it is very promising for the future.

Full text of DOI:10.1016/j.polymer.2013.11.046

Polymer 55 (2014) 1384e1396
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Facile synthesis of five 2D surface modifiers by highly selective
photocyclic aromatization and efficient enhancement of oxygen
permselectivities of three polymer membranes by surface
modification using a small amount of the 2D surface modifiers
,
a
a
Jianjun Wang a, Yu Zang b, Guanwu Yin a, Toshiki Aoki a , Hiroyuki Urita , Ken Taguwa ,
Lijia Liu c, Takeshi Namikoshi d, Masahiro Teraguchi a, Takashi Kaneko a, Liqun Ma b,
Hongge Jia b
*
a Department of Chemistry and Chemical Engineering, Graduate School of Science and Technology, Niigata University, Ikarashi 2-8050, Nishi-ku,
Niigata 950-2181, Japan
b Key Laboratory of Polymer Composition and Modification, College of Materials Science and Engineering, Qiqihar University, Wenhua street 42,
Qiqihar 161006, China
c Polymer Materials Research Center, College of Materials Science and Chemical Engineering, Harbin Engineering University, 145 Nantong Street,
Harbin 150001, China
d Materials Science and Engineering, Kitami Institute of Technology, 165 Koen-cho, Kitami, Hokkaido 090-8507, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
A facile synthesis of novel five 2D (planar) surface modifiers having a triphenylbenzene derivatives as a
2D structure has been achieved by the highly selective photocyclic aromatization reaction. Efficient
enhancement of oxygen permselectivities through the three polymer membranes has been achieved by
adding a small amount (<5.0 wt%) of the 2D surface modifiers. Among the five 2D surface modifiers, a
modifier compound having oligoethylene oxide groups showed the best performance for the enhance-
ment. These improvements were thought to be caused mainly by improvement of the solution selectivity
on the membrane surface where the 2D surface modifiers were accumulated. In some of the surface-
Received 24 September 2013
Received in revised form
26 November 2013
Accepted 29 November 2013
Available online 10 December 2013
Keywords:
modified blend membranes, their plots in the PO -a graph were over or close to the upper boundary
2
2D(planar) surface modifier
Highly selective photocyclic
aromatization(SCAT)
Oxygen permselectivity
line by Robeson in 1991. Since all the membranes containing the 2D surface modifiers showed better
permselectivities than the corresponding substrate membranes, it is very promising for the future.
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
had lower
a
values, and membranes having higher
showed a tradeoff relationship
[1e24]. Also membranes having higher PO values tended to be too
a values had
lower PO values, that is, PO and
a
2
2
Permselective membranes separating gases such as oxygen and
nitrogen have many practical uses to solve environmental and
energy problems. So, many studies have been reported on such
membranes in the form of not only organic polymers but also
inorganic compounds [1e24]. The requirements for oxygen perm-
2
flexible and those having higher
a values tended to be too brittle.
To solve the above problems, that is, to obtain materials that
simultaneously satisfy the above three requirements, we reported
surface modifications of conventional polymer membranes having
enough mechanical strength by solvent casting of a mixture of
small amounts of surface active polymers and the conventional
selective membranes are (1) a high permeability coefficient (PO
:
2
cm3(STP)$cm/cm2$s$cmHg), (2)
a
high separation factor
(
a
¼ PO =PN ), and (3) good membrane forming ability (high me-
base polymers [25e27]. It was an effective method to enhance a
2
2
chanical strength). However, it has proven difficult to realize all
with a small decrease in PO , while resulting in no change to the
2
three requirements at the same time in almost all membrane ma-
good membrane forming abilities of the base polymers. Although
similar reports have been made for nanofiltration of aqueous so-
lution [28e30], to the best of our knowledge, there have been no
reports except for our group [25e27] on such methods of surface
modifications of membranes for enhancing gas separation selec-
terials reported. For example, membranes having higher PO values
2
* Corresponding author.
tivities. However, the extent of enhancement of
a was not enough.
E-mail address: toshaoki@eng.niigata-u.ac.jp (T. Aoki).
0032-3861/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.polymer.2013.11.046

J. Wang et al. / Polymer 55 (2014) 1384e1396
1385
In this study, to improve the effectiveness of the surface modifiers,
five kinds of 2D(¼a planar structure) surface modifiers were
1H NMR (DMSO-d6, TMS):
d
¼ 8.16 (s, 1H, PhOH), 7.29(s, 2H, (PhH)2),
5.31(t, 2H, J ¼ 5.0 Hz, (CH2OH)2), 4.51(d, 4H, J ¼ 5.0 Hz, (CH2OH)2).
2.2.1.1.3. 2,6-Bis(acetoxymethyl)-4-bromo-1-phenyl acetate (2)
(Scheme 3). To a pyridine solution (50 mL) of 1 (10.0 g, 42.9 mmol),
acetic anhydride (30.0 mL, 317 mmol) was added dropwise at 0 ꢀC.
The solution was stirred for 1.5 h at room temperature and ethyl
acetate (100 mL) was added to the mixture. The mixture was
washed with a saturated aqueous solution of CuSO4$5H2O to
remove pyridine and extracted with CH2CL2. The organic layer was
dried over anhydrous MgSO4 for 1 h. After filtration and concen-
tration, the crude product was purified by silica-gel column chro-
matography to give 2 as a white solid. Yield: 78.4% (12.1 g).
Rf ¼ 0.30 (ethyl acetate/hexane ¼ 1/2 (v/v)). 1H NMR (DMSO-d6,
designed (Chart 1). They have a
p-conjugated planar (2D) part,
hydroxyl groups, and hydrophilic or hydrophobic groups.
We have reported recently a novel polymer reaction called SCAT
(highly selective photocyclic aromatization reaction) (Scheme 1)
[31]. It quantitatively gives 1,3,5-triphenylbenzene derivatives
having six hydroxyl groups and any kinds of three substituents by a
simple procedure (using only light irradiation) in high conversions
and selectivity. In addition the reaction has good tolerance for many
kinds of functional groups. These advantages of SCAT are very
useful for the synthesis of 2D surface modifiers having different
kinds of functional groups and therefore we selected SCAT as the
synthetic route in this study. In this article, a facile synthesis of the
TMS):
d
¼ 7.56 (s, 2H, ((PhH)2)), 4.97(s, 4H, Ph(CH2O)2), 2.34(s, 3H,
2D surface modifiers by the SCAT reaction and enhancement of
a
by
PhOCOCH3), 2.09(s, 6H, Ph(CH2OCOCH3)2).
surface modification using the 2D surface modifiers are reported.
2.2.1.1.4. 2,6-Bis(acetoxymethyl)-4-(trimethylsilylethynyl)-1-
phenyl acetate (3) (Scheme 3). A mixture of 2 (11.8 mg, 32.9 mmol),
triphenylphosphine (610 mg, 2.33 mmol), copper iodide (752 mg,
3.95 mmol) and trimethylsilylacetylene (6.50 mL, 45.7 mmol) in
triethylamine (130 mL) was refluxed for 24 h. After the mixture was
filtered, the solvent was removed by evaporation and the crude
product was purified by silica-gel column chromatography to give 3
as a brown liquid. Yield: 95.6% (11.8 g). Rf ¼ 0.30 (ethyl acetate/
2. Experimental
2.1. Material
All the solvents except for tetrahydrofuran(THF) used for
monomer synthesis and polymerization were distilled as usual. Dry
THF (99.5% purity) purchased from Kanto chemistry was used. The
polymerization initiator, [Rh(nbd)Cl]2 (nbd ¼ 2,5-norbornatiene)
purchased from Aldrich Chemistry was used as received. Poly(vinyl
alchohol)(PVA) purchased from Wako Pure Chemistry Industries,
Ltd. was used as received. 1-Phenyl-2-p-(trimethylsilyl)phenyl-
acetylene (DPA) [32] and p-(trimethylsilyl)phenylacetylene(SPA)
[33] were synthesized and polymerized according to the literature.
hexane ¼ 1/3 (v/v)). 1H NMR (CDCl3, TMS):
¼ 7.53 (s, 2H, (PhH)2),
d
4.98 (s, 4H, Ph(CH2OAc)2), 2.32(s, 3H, PhOCOCH3), 2.06(s, 6H,
Ph(CH2OCOCH3)2), 0.22(s, 9H, Si(CH3)3).
2.2.1.1.5. 2,6-Bis(acetoxymethyl)-4-ethynyl phenol (4) (Scheme 3).
To a mixture of lithium hydride (3.00 g, 79.1 mmol) and THF
(120 mL), a THF solution (15 mL) of 3 (14.9 g, 39.5 mmol) was added
dropwise at 0 ꢀC. After the mixture was stirred for 2 h at room
temperature, deionized water (140 mL) was added dropwise into
the reaction mixture at 0 ꢀC. The mixture was stirred for 12 h at
room temperature and was treated with a 2N HCl aq. solution to
precipitate aluminum salts. After the mixture was filtered, THF was
removed from the solution by evaporation. The product was dis-
solved in ethyl acetate and the solution was washed with water and
extract by CH2Cl2. The organic layer was dried over anhydrous
MgSO4 for 1 h. After filtration and concentration, the crude product
was purified by silica-gel column chromatography to give 4 as a
white solid. Yield: 70.3% (4.90 g). Rf ¼ 0.39 (ethyl acetate/
2.2. Synthesis of new 2D surface modifiers(T-R)(Chart 2)(Scheme 2)
2.2.1. Synthesis of the monomers (M-R) (Scheme 3)
2.2.1.1. Synthesis of M-EO, M-Do and M-S3 (Scheme 3)
2.2.1.1.1. Synthesis of 3,5-bis(hydroxymethyl)-4-{2-(1,4,7-trioxa
octyl)phenylmethyl}oxy-1-phenylacetylene (M-EO) (Scheme 3).
According to the synthetic route shown in Scheme 3, M-EO was
synthesized via compounds 1-7 in 8.1% as a total yield. All the
following reaction procedures were conducted under dry nitrogen.
2.2.1.1.2. 4-Bromo-2,6-bis(hydroxymethyl)-1-phenol (1) (Scheme 3)
[34]. According to the literature,1 was prepared. Yield: 56.8% (114 g).
hexane ¼ 2/3 (v/v)). 1H NMR (DMSO-d6, TMS):
d
¼ 8.98(s, 1H,
PhOH), 7.27(s, 2H, (PhH)2), 4.52(s, 4H, J ¼ 5.0 Hz, Ph(CH2OH)2),
3.91(s, 1H, HC^C), 3.31(t, 2H, J ¼ 5.0 Hz, (CH2OH)2).
2.2.1.1.6. 2-(2-Methoxyethoxy)ethyl 4-methylbenzenesulfonate
(5) (Scheme 3). To
87.6 mmol) and THF (50 mL), 4 (7.80 mL, 65.7 mmol) was added
dropwise at
ꢀC. Then p-toluenesulfonyl chloride (10.4 g,
a mixture of sodium hydride (3.50 g,
0
54.7 mmol) dissolved in THF (20 mL) was added dropwise at 0 ꢀC.
The solution was stirred for 9 h at room temperature and the re-
action mixture was treated with a 4N HCl aq. solution until the pH
read 4 and the aqueous layer was extracted with CH2Cl2. The
organic layer was dried over anhydrous MgSO4 for 1 h. After
filtration and concentration, the crude product was purified by
silica-gel column chromatography to give 5 as a colorless viscous
liquid. Yield: 64.3% (9.60 g). Rf ¼ 0.39 (ethyl acetate/hexane ¼ 1/1
(v/v)). 1H NMR (CDCl3, TMS):
d
¼ 7.80 (d, 2H, J ¼ 8.5 Hz, (PhH)2),
7.34(d, 2H, J ¼ 8.5 Hz, (PhH)2), 4.17(t, 2H, J ¼ 4.6 Hz, OCH2CH2),
3.69(t, 2H, J ¼ 4.6 Hz, OCH2CH2O), 3.57 (t, 2H, J ¼ 4.3 Hz, OCH2
-
CH2OCH3), 3.47(t, 2H, J ¼ 4.3 Hz, OCH2CH2OCH3), 3.35(s, 3H, OCH3),
2.45(s, 3H, PhCH3).
2.2.1.1.7. 2-{2-(2-Methoxyethoxy)ethoxy}-1-phenylmethynol (6)
(Scheme 3). To a mixture of 2-hydroxybenzyl alcohol (5.00 mg,
40.3 mmol), potassium carbonate (11.1 g, 80.6 mmol), and 18-
crown-6 (5.35 mg, 20.1 mmol) in dry acetone (200 mL), 5 (12.1 g,
44.3 mmol) dissolved in dry acetone(10 mL) was added dropwise at
Chart 1. Molecular design of the 2D surface modifier.

1386
J. Wang et al. / Polymer 55 (2014) 1384e1396
Chart 2. Chemical structures of the 2D surface modifiers.
0 ꢀC. This solution was refluxed for 48h and allowed to stand for 2 h
at 25 ꢀC. After the mixture was filtered, deionized water was added
to the resulting solution and the aqueous layer was extracted with
CH2Cl2. The combined organic layer was washed with a 2N HCl aq.
solution and dried over anhydrous MgSO4 for 1 h. After filtration
and concentration, the crude product was purified by silica-gel
column chromatography to give 6 as a colorless viscous liquid.
Yield: 66.2% (6.00 g). Rf ¼ 0.31 (ethyl acetate/hexane ¼ 1/1 (v/v)).
solution was stirred at 0 ꢀC for 1 h and a saturated NaCO3 aq. so-
lution was added to quench the reaction. The aqueous layer was
extracted with CH2Cl2, and the organic layer was washed with a
saturated NaCO3 aq. solution and dried over anhydrous MgSO4 for
1 h. After filtration and concentration, the crude product was pu-
rified by silica-gel column chromatography to give 7 as a colorless
viscous liquid. Yield: 86.5% (6.30 g). Rf ¼ 0.50 (ethyl acetate/
hexane ¼ 3/1 (v/v)). 1H NMR (CDCl3, TMS):
d
¼ 7.40(d, 1H, J ¼ 7.6 Hz,
1H NMR (CDCl3, TMS):
d
¼ 7.35(d, 1H, J ¼ 7.6 Hz,
), 7.24(dd,
), 7.28(dd, J ¼ 7.6 Hz, J ¼ 8.4 Hz, 1H,
), 7.01(dd, 1H,
J ¼ 7.6 Hz, J ¼ 8.4 Hz, 1H,
J ¼ 8.0 Hz,
), 7.02(dd, 1H, J ¼ 8.4 Hz,
J ¼ 8.4 Hz, J ¼ 8.0 Hz,
), 6.95(d, 1H, J ¼ 8.0 Hz,
),
4.57(s, 2H, PhCH2Br), 4.15(t, 2H, J ¼ 4.6 Hz, PhOCH2CH2), 3.86(t, 2H,
J ¼ 4.6 Hz, PhOCH2CH2O), 3.72(t, 2H, J ¼ 4.6 Hz, OCH2CH2OCH3),
3.53(t, 2H, J ¼ 4.6 Hz, OCH2CH2OCH3), 3.37(s, 3H, OCH3).
), 6.90(d, 1H, J ¼ 8.0 Hz,
),4.65(d, 2H,
J ¼ 5.0 Hz, PhCH2OH), 4.18(t, 2H, J ¼ 4.6 Hz, PhOCH2CH2), 3.83(t, 2H,
J ¼ 4.6 Hz, PhOCH2CH2O), 3.67(t, 2H, J ¼ 4.3 Hz, OCH2CH2OCH3),
3.54(t, 2H, J ¼ 4.3 Hz, OCH2CH2OCH3), 3.36(s, 3H, OCH3), 3.33(t, 1H,
J ¼ 5.0 Hz, CH2OH).
2.2.1.1.9. 3,5-Bis(hydroxymethyl)-4-{20-(1,4,7-trioxaoctyl)benzy-
loxy}-1-phenylacetylene (M-EO) (Scheme 3) [35]. To a mixture of 4
(1.10 g, 6.17 mmol), potassium carbonate (1.71 g,12.3 mmol), and 18-
crown-6 (1.79 g, 6.79 mmol) in dry acetone (33 mL), 7 (1.96 g,
6.79 mmol) was added dropwise at 0 ꢀC. This solution was refluxed
for 40 h and then allowed to stand for 2 h at 25 ꢀC. After the mixture
was filtered, deionized water was added to the resulting solution
and the aqueous layer was extracted with CH2Cl2. The organic layer
was washed with a 2N HCl aq. solution and dried over anhydrous
MgSO4 for 1 h. After filtration and concentration, the crude product
was purified by silica-gel column chromatography to give M-EO as a
white solid. Yield: 65.0% (1.50 g). Rf ¼ 0.17 (ethyl acetate/hexane ¼ 1/
2.2.1.1.8. 1-(1,4,7-Trioxaoctyl)-2-(bromomethyl)benzene
(7)
(Scheme 3). To a solution of 6 (5.71 g, 25.2 mmol) dissolved in
CH2Cl2 (200 mL), triphenylphosphine (10.0 g, 38.2 mmol) dissolved
in CBr4 (14.2 g, 42.9 mmol) were added dropwise at 0 ꢀC. This
1 (v/v)). 1H NMR (CDCl3, TMS):
d
¼ 7.53(s, 2H, C^C(PhH)2), 7.44(d,
1H, J ¼ 7.6 Hz, ), 7.34(dd, J ¼ 7.6 Hz, J ¼ 8.4 Hz,1H,
),
7.00(dd, 1H, J ¼ 8.4 Hz, J ¼ 8.0 Hz,
), 6.91(d, 1H, J ¼ 8.0 Hz,
), 5.01(s, 2H, PhOCH2Ph), 4.75(d, 4H,
J
¼
6.0 Hz,
Scheme 1. Highly selective photocyclic aromatization (SCAT).

J. Wang et al. / Polymer 55 (2014) 1384e1396
1387
Scheme 2. Synthesis of the 2D surface modifiers by a) helix-sense-selective polymerization (HSSP) and b) highly selective photocyclic aromatization (SCAT).
Ph(CH2OH)2), 4.19(t, 2H, J ¼ 4.4 Hz, PhOCH2CH2), 3.86(t, 2H, J ¼ 4.4
Hz, PhOCH2CH2O), 3.71(t, 2H, J ¼ 4.4 Hz, OCH2CH2OCH3), 3.58(t, 2H,
J ¼ 4.4 Hz, OCH2CH2OCH3), 3.36(s, 3H, OCH3), 3.03(s, 1H, HC^C),
2.82(t, 2H, J ¼ 6.0 Hz, Ph(CH2OH)2). 13C NMR (CDCl3, TMS):
83.30(HC^C), 76.91(HC^C), 72.57(Ph-CH2O-), 60.75(-CH2OH),
60.75(-OCH2CH2), 67.38(-OCH2CH2), 66.85(OCH2CH2OCH3),
68.97(OCH2CH2OCH3), 56.7(OCH2CH2OCH3). Anal. Cacld for
22H26O6: C, 68.38; H, 6.78; O, 24.84; Found: C, 68.41; H, 6.83; O,
24.76.
2.2.1.1.10. Synthesis of 3,5-bis(hydroxymethyl)-4-dodecyloxy-1-
C
d
¼
157.24(
), 155.94(
), 131.56( ), 130.54(
121.26( ), 118.41( ), 111.82(
), 135.06(
),
phenylacetylene (M-Do) (Scheme 3) [34]. M-Do was synthesized
according to our previous paper [30]. 1H NMR (CDCl3, TMS):
133.16(
(
), 125.14,
),
d
¼ 7.48(s, 2H, (PhH)2), 4.70(d, 4H, J ¼ 5.9 Hz, (CH2OH)2), 3.88(t, 2H,
J ¼ 6.6 Hz, OCH2CH2), 3.04(s, 1H, C^CH), 2.11(t, 2H, J ¼ 5.9 Hz,
(CH2OH)2), 1.79(quint, 2H, J ¼ 6.6 Hz, OCH2CH2CH2), 1.50-1.20(m,
)
Scheme 3. Synthesis of the monomers(M-R) as a material of the 2D surface modifiers.

1388
J. Wang et al. / Polymer 55 (2014) 1384e1396
18H, OC2H4(CH2)9CH3), 0.89(t, 3H, CH2CH3). IR (KBr): 3600e
3100(OH), 3308(HC^C), 3000e2800(CH), 2116(C^C), 1200e
1000(CO) cmꢁ1. Anal. Cacld for C22H34O3: C, 76.70; H, 9.36; O,13.94;
Found: C, 76.65; H, 9.76; O, 13.68.
C
28H29O3N: C, 78.66; H, 6.84; N, 3.28; Found: C, 77.68; H, 6.81; N,
3.28.
2.2.1.2.4. Synthesis of 3,5-bis(hydroxymethyl)-4-[40-{3-(trie-
thylsilyl)propyliminomethyl}benzyloxy]-1-phenyl-acetylene (M-TES)
(Scheme 3) [38]. 3-(Triethylsilyl)propylamine (9) (Scheme 3)
2.2.1.1.11. Synthesis of 3,5-bis(hydroxymethyl)-4-{40-(1-
heptamethyltrisiloxanyl)benzyloxy}-
(Scheme 3) [36]. M-S3 was synthesized according to our previous
paper [32]. 1H NMR (CDCl3, TMS):
PhOCH2(PhH)2), 7.50(s, 2H, C^C(PhH)2), 7.40(d, 2H, J ¼ 8.0 Hz,
Si(PhH)2), 4.96(s, 2H, PhOCH2), 4.67(d, 4H, J ¼ 6.0 Hz, Ph(CH2OH)2),
3.03(s, 1H, C^CH), 1.96(t, 2H, J ¼ 6.0 Hz, (PhCH2OH)2), 0.33(s, 6H,
PhSi(CH3)2), 0.06(s, 9H, Si(CH3)3), 0.03(s, 6H, OSi(CH3)2O). IR (KBr):
3600e3100(OH), 3312(HC^C), 2109(C^C), 1258(SiC) cmꢁ1. Anal.
Cacld for C24H36O5Si3: C, 58.97; H, 7.42; Found: C, 58.96; H, 7.49.
1-phenylacetylene
(M-S3)
H2PtCl6$6H2O (26 mg, 50 mmol) and toluene (15 mL) were added
to a flask and the mixture was stirred at 80 ꢀC until H2PtCl6$6H2O
was dissolved completely. Triethylsilane (8.00 mL, 50.2 mmol) and
allylamine (4.20 mL, 55.2 mmol) were added dropwise to the so-
lution at 40 ꢀC separately and this reaction solution was stirred at
85 ꢀC for 120 h. The solvent was removed by evaporation and the
crude product was purified by vacuum distillation at 54 ꢀC (250 Pa)
to give 9 as a colorless liquid. Yield: 67.7% (5.89 g). 1H NMR (CDCl3,
d
¼ 7.59 (d, 2H, J ¼ 8.0 Hz,
TMS):
d
¼ 2.65 (t, 2H, J ¼ 7.0 Hz, NH2CH2), 1.42 (m, 2H,
NH2CH2CH2CH2), 1.27(br, 2H, CH2NH2), 0.92 (t, 9H, J ¼ 8.0 Hz,
Si(CH2CH3)3), 0.59 (t, 2H, J ¼ 8.0 Hz, NCH2CH2CH2Si), 0.49 (q, 6H,
J ¼ 8.0 Hz, Si(CH2CH3)3).
2.2.1.2. Synthesis of M-TB and M-TES (Scheme 3). According to the
synthetic route as shown in Scheme 3, M-TB and M-TES were
synthesized via compounds 1-4, M-CHO, 8-9 in 12.0% as a total
yield. All the following reaction procedures were conducted under
dry nitrogen.
2.2.1.2.5. 3,5-Bis(hydroxymethyl)-4-[40-{3-(triethylsilyl)propyli-
minomethyl}benzyloxy]-1-phenylac -etylene (M-TES) (Scheme 3).
A mixture of M-CHO (500 mg, 1.68 mmol), 9 (586 mg, 3.36 mmol)
and Al2O3 (10.0 g) in dry THF (16 mL) was stirred for 3 days at room
temperature. After the mixture was filtered, the solvent was
removed by evaporation to yield a white solid. The crude product
was purified by recrystallization in chloroform/hexane(¼5/95 (v/
v)) to give M-TES as a white solid. Yield: 56.5% (428 mg). 1H NMR
2.2.1.2.1. Synthesis of 3,5-bis(hydroxymethyl)-4-{40-(20’-tert-
butyl-1-iminomethyl)benzyloxy}-1-phenylacetylene (M-TB) (Scheme
3) [37]. 4-Bromomethylbenzaldehyde(8) (Scheme 3)
4-bromomethylbenzonitrile (14.0 g, 71.4 mmol) was dissolved
in dry CH2Cl2 (260 mL) and the solution was stirred at 0 ꢀC for
20min.1.0 M diisobutylaluminium hydride hexane solution (DIBAL)
(70 mL, 70 mmol) was added dropwise to the solution at 0 ꢀC. The
solution was stirred for 15 min at 0 ꢀC and then DIBAL solution
(35 mL) was added dropwise again. The mixture was stirred for
15 min at 0 ꢀC and for 30 min at 25 ꢀC. This reaction mixture was
washed with 50% H2SO4 (150 mL) aq. solution to precipitate the
aluminum salt. After the mixture was filtered, deionized water was
added to the solution and the aqueous layer was extracted with
CH2Cl2. The organic layer was dried over MgSO4 for 1 h. This
mixture was filtered and concentrated to give 8 as a white solid.
(CDCl3, TMS):
d
¼ 8.27 (s, 1H, PhCH ¼ N), 7.76 (d, 2H, J ¼ 8 Hz,
), 7.50 (s, 2H, C^CPhH), 7.45 (d, 2H, J ¼ 8 Hz, ),
4.99 (s, 2H, PhOCH2Ph), 4.65 (d, 4H, J ¼ 6.0 Hz, Ph(CH2OH)2), 3.59 (t,
2H, J ¼ 7.0 Hz, NCH2CH2), 3.04 (s, 1H, HC^C), 1.90 (t, 2H, J ¼ 6.0 Hz,
Ph(CH2OH)2), 1.68 (m, 2H, NCH2CH2CH2), 0.91 (t, 9H, J ¼ 8.0 Hz,
Si(CH2CH3)3), 0.53 (t, 2H, J ¼ 8.0 Hz, NCH2CH2CH2Si), 0.49 (q, 6H,
J ¼ 8.0 Hz, Si(CH2CH3)3). IR (KBr): 3322 (OH), 3233 (HC^C), 1643
(C]N) cmꢁ1. Anal. Cacld for C27H37O3NSi: C, 71.80; H, 8.26; N, 3.10;
Found: C, 71.74; H, 8.17; N, 3.10.
Yield: 56.3% (8.00 g). 1H NMR (CDCl3, TMS):
7.85 (d, 2H, J ¼ 8.0 Hz, BrCH2(PhH)2), 7.54 (d, 2H, J ¼ 8.0 Hz,
(HPh)2CHO), 4.50 (s, 2H, BrCH2Ph).
d
¼ 10.0 (s, 1H, PhCHO),
2.2.2. Synthesis of the polymers (P-R) by helix-sense-selective
polymerization (HSSP) of M-R (Scheme 2a) [34]
2.2.1.2.2. 3,5-Bis(hydroxymethyl) -4-(40-formylbenzyloxy)phenyl-
acetylene (M-CHO) (Scheme 3). A mixture of 4 (3.79 g, 21.3 mmol), 8
(4.23 g, 21.3 mmol) and K2CO3 (8.80 g, 63.9 mmol) in DMF (107 mL)
was stirred for 50 h at 70 ꢀC. After the mixture was filtered, the
solvent was evaporated. The residue was washed with water and
extracted with ethyl acetate. The organic layer was dried over
MgSO4 for 1h. After filtration and concentration, the crude product
was purified by silica-gel column chromatography to give M-CHO
as a white solid. Rf ¼ 0.70 (CHCl3/MeOH ¼ 95/5 (v/v)). 1H NMR
2.2.2.1. Synthesis of P-EO by HSSP of M-EO (Scheme 2a). A solution
of [Rh(nbd)Cl]2 (1.37 mg, 2.98
amine (PEA) (153 L, 1.20 mmol) in CHCl3 (1.0 mL) was added to a
solution of M-EO (100 mg, 260 mol) in CHCl3 (1.6 mL). The reac-
mmol) and (S)- or (R)-phenylethyl-
m
m
tion solution was stirred at room temperature for 12 h. The crude
polymer was purified by reprecipitation of the CHCl3 solution into a
large amount of (hexane/ethyl acetate ¼ 2/3 (v/v)) and dried in
vacuo to give P-EO as a brownish red polymer in 55% yield.
Mw ¼ 2.6 ꢂ 105. 1H NMR (CDCl3/DMSO-d6 ¼ 3/2(v/v)):
d
¼ 7.28(s,
(CDCl3, TMS):
d
¼ 10.0 (s, 1H, PhCHO), 7.94 (d, 2H, J ¼ 8.0 Hz,
1H, HPhOCH2CH2O), 7.16(m, 1H, HPhOCH2CH2O), 6.81(m, 4H,
HC¼CPhH and HPhOCH2CH2O), 5.88(br, cis proton in the main
chain), 4.57e4.37 (m, 6H, PhOCH2Ph and Ph(CH2OH)2), 3.97e3.49
(m, 10H, OCH2CH2OCH2CH2OCH3 and Ph(CH2OH) 2), 3.36(s, 3H,
OCH3).
(HPh)2CHO), 7.63 (d, 2H, J ¼ 8.0 Hz, OCH2(PhH)2), 7.54 (s, 2H,
HC^C(PhH)2), 5.08 (s, 2H, PhOCH2Ph), 4.69 (d, 4H, J ¼ 6.0 Hz,
Ph(CH2OH)2), 3.07 (s, 1H, HC^C), 1.74 (t, 2H,
Ph(CH2OH)2).
J
¼
6.0 Hz,
2.2.1.2.3. 3,5-Bis(hydroxymethyl)-4-{40-(200-tert-butyl-1-
iminomethyl)benzyloxy}-1-phenyl-acetylene (M-TB) (Scheme 3).
A mixture of M-CHO (200 mg, 0.670 mmol), 2-tert-butylaniline
(0.21 mL, 1.4 mmol) and Al2O3 (700 mg) in dry THF (7 mL) was
stirred for 3 days at room temperature. After the mixture was
filtered, the solvent was removed by evaporation to yield a white
solid. The crude product was purified by recrystallization from
chloroform/hexane(¼10/90 (v/v)) to give M-TB as a white solid.
2.2.2.2. Synthesis of P-TB by HSSP of M-TB (Scheme 2a). A solution
of [Rh(nbd)Cl]2 (0.80 mg, 1.76
220 mol) in dry THF (0.44 mL) was added to a dry THF (0.44 mL)
solution of M-TB (37.0 mg, 88.0 mol). The reaction solution was
mmol) and (S)- or (R)-PEA (28.2 mL,
m
m
stirred at room temperature for 24 h. The crude polymer was pu-
rified by reprecipitation of the THF solution into a large amount of
methanol and the formed solid was dried in vacuo to give P-TB as a
red solid in 57% yield. Mw ¼ 7.9 ꢂ 107. 1H NMR (CCl4/DMSO-d6 ¼ 5/
Yield: 70.0% (200 mg). 1H NMR (CDCl3, TMS):
PhCH ¼ N), 7.90-6.81(m, 10H, PhH), 5.01(s, 2H, PhOCH2Ph), 4.62(d,
4H, Ph(CH2OH)2), 3.38(s, 1H, HC^C),1.43(s, 9H, PhC(CH3)3). IR
(KBr): 3326 (OH), 3232 (HC^C), 1646 (C]N) cmꢁ1. Anal. Cacld for
d
¼ 8.36(s, 1H,
1(v/v), TMS):
d
¼ 7.56e6.81(br, 10H, PhH), 5.93(br, cis proton in the
main chain), 4.79(br, 2H, PhOCH2Ph), 4.37(br, 4H, Ph(CH2OH)2),

J. Wang et al. / Polymer 55 (2014) 1384e1396
1389
1.51(br, 9H, C(CH3)3). IR (KBr): 3600e3100(OH), 3000e2800(CH),
1645 (C]N) cmꢁ1
purified by silica-gel column chromatography to give a brownish
red powder in 78.4% yield.
.
2.2.2.3. Synthesis of P-TES by HSSP of M-TES (Scheme 2a). The HSSP
of M-TES was carried out similarly to that of M-TB to give P-TES as
an orange solid in 98% yield. Mw ¼ 3.0 ꢂ 105. 1H NMR(CCl4/DMSO-
2.2.3.2. Synthesis of T-TB by SCAT of P-TB (Scheme 2b). SCAT of P-TB
(Mw ¼ 7.9 ꢂ 107) membrane gave a brown membrane in 100%
conversion and 65.8% selectivity. The crude product was purified by
Al2O3 chromatography to give a brown powder in 61.2% yield.
d6 ¼ 5/1(v/v)): ¼ 8.15 (br, 1H, PhCH ¼ N), 7.58e6.92 (br, 6H, PhH),
d
5.91 (br, cis proton in the main chain), 4.75 (br, 2H, PhOCH2Ph), 4.35
(br, 4H, Ph(CH2OH)2), 3.53 (br, 2H, NCH2CH2), 1.65 (br, 2H,
NCH2CH2CH2), 0.88 (br, 9H, Si(CH2CH3)3), 0.51 (br, 2H,
NCH2CH2CH2Si), 0.47 (br, 6H, Si(CH2CH3)3). IR (KBr): 3375 (OH),
2.2.3.3. Synthesis of T-TES by SCAT of P-TES (Scheme 2b). SCAT of P-
TES (Mw ¼ 3.0 ꢂ 105) membrane gave a faint orange membrane in
100% conversion and 79.5% selectivity. The crude product was pu-
rified by Al2O3 chromatography to give a slightly orange powder in
75.2% yield.
1645 (C]N) cmꢁ1
.
2.2.2.4. Synthesis of P-Do by HSSP of M-Do (Scheme 2a) [34].
P-Do was synthesized from M-Do according to our previous paper
in 36.5% yield. Mw ¼ 2.5 ꢂ 106.1H NMR ( CDCl3/DMSO-d6 ¼ 55/
2.2.3.4. Synthesis of T-Do by SCAT of P-Do (Scheme 2b). SCAT of P-
Do (Mw ¼ 2.5 ꢂ 106) membrane gave a brown membrane in 100%
conversion and 100% selectivity. The crude product was purified by
silica-gel column chromatography to give a brown powder in 91.1%
yield.
45(v/v), TMS):
d
¼ 6.77(s, 2H, (PhH)2), 5.76(br, cis proton in the
main chain), 4.61(s, 2H, PhOCH2), 4.32(s, 4H, Ph(CH2OH)2), 3.55(br,
2H, Ph(CH2OH)2), 1.29e1.64(m, 20H, OCH2CH2(CH2)9CH3), 0.89(b,
3H, CH2CH3). IR(KBr): 3700e3100(OH), 3000e2800(CH), 1200e
1000(CO) cmꢁ1
.
2.2.3.5. Synthesis of T-S3 by SCAT of P-S3 (Scheme 2b). SCAT of P-S3
(Mw ¼ 4.0 ꢂ 107) membrane gave a brownish red membrane in
100% conversion and 95% selectivity. The crude product was puri-
fied by silica-gel column chromatography to give a brownish red
powder in 90.7% yield.
2.2.2.5. Synthesis of P-S3 by HSSP of M-S3 (Scheme 2a) [35].
P-S3 was synthesized from M-S3 according to our previous paper in
89% yield. Mw ¼ 4.0 ꢂ 107. 1H NMR (DMSO-d6/CCl4 ¼ 1/5, TMS):
d
¼ 7.54e7.34(br, 6H, PhH), 5.89(br, cis proton in the main chain),
4.75(br, 2H, PhOCH2Ph), 4.38(br, 4H, (CH2OH)2), 0.34e0.04(br, 21H,
(eSi(CH3)3) and (eSiO(CH3)2e)2). IR(KBr): 3600e3100(OH),
2.3. Membrane preparation
1256(SiC), 1051 (SiO) cmꢁ1
.
The surface-modified membranes were prepared from the bi-
nary solution of the mixture consisting of small amounts (less than
5.0 wt%) of one of the 2D surface modifiers and one of the base
polymers (see Chart 2) by two kinds of solvent casting methods
(Method I and Method II) as shown in Scheme 4.
2.2.3. Synthesis of the 2D surface modifiers(T-R) by highly selective
photocyclic aromatization (SCAT) of P-R (Scheme 2b) [31]
The P-R membranes (thickness around 20 mm) were irradiated
under nitrogen at 25 ꢀC by visible light (400e500 nm, 2.54 mW/
cm2) for 2e4 weeks. The conversions and selectivities were
determined by GPC detected by UV. Visible light (400e500 nm)
irradiation was carried out by using a 300 W of Xe lamp (Asahi
Spectra, MAX-302 with vis mirror module) through a cutoff filter
2.3.1. PVA-based membranes by Method I
A dimethyl sulfoxide(DMSO) solution( 1.0 mL) of poly(vinyl
alcohol)(PVA) (40 mg) and 0.060e10.0 wt% of one of the 2D surface
modifiers was cast on a poly(tetrafluoroethylene) sheet. The
(Asahi Spectra, LUX400 (
l > 400 nm), XF541 (l < 510 nm), and
solvent(DMSO) was evaporated in
a
reduced pressure
(1.33 ꢂ 10ꢁ3 MPa) for 12 h and then by heating at 50 ꢀC at this
pressure for 24 h. The resulting membranes were allowed to stand
for 8 h at 25 ꢀC. Finally, the membranes were detached from the
poly(tetrafluoroethylene) sheet and were annealed in vacuo at
30 ꢀC for 24 h.
XF546 ( < 610 nm)).
l
2.2.3.1. Synthesis of T-EO by SCAT of P-EO (Scheme 2b). SCAT of P-
EO (Mw ¼ 2.6 ꢂ 105) membrane gave a brownish red membrane in
100% conversion and 83.2% selectivity. The crude product was
Scheme 4.

1390
J. Wang et al. / Polymer 55 (2014) 1384e1396
2.3.2. PSPA-based and PDPA-based membranes by two methods
(Methods I and II)
TES, TDo, and T-Si) groups. All the 2D surface modifiers were
synthesized by SCAT(Scheme 1) [27], which is our original method,
of the corresponding polyphenylacetylenes according to Scheme 2.
As a typical example, the 3D chemical structures of T-EO are shown
in Chart 3. It has 1,3,5-trisubstituted benzene as a hydrophobic 2D
part and three hydrophilic arms of an oligoethylene oxide which
can function as an anchor segment in hydrophilic base polymer
membranes such as poly(vinyl alcohol)(PVA). Chart 4 indicates an
example of an ideal 2D supramolecular structure of six T-EO mol-
ecules on a membrane surface. It has some molecular-size pores
which can recognize and separate gas molecules such as oxygen
and nitrogen.
In order to separate gas molecules more effectively, thinnest
membranes without any defects are desired. In addition, no dis-
tribution of the additives inside the base membrane is thought to
be better because such additives may change the original good
performance of the base membranes. Therefore, the minimum
amount of the 2D surface modifier (T-EO) needed for such a thin-
nest surface layer was calculated. As shown in Chart 5, the result of
calculation showed very little amount (¼6.0 ꢂ 10ꢁ3 wt%) of T-EO is
enough for making the ideal thin layer. Therefore we added small
amounts of the additives in this study.
Method I (conventional method): A toluene solution (1.0 mL) of
the base polymer (35 mg) and 0.060e5.0 wt% of the 2D surface
modifier was cast on a poly(tetrafluoroethylene) sheet. After the
solvent was evaporated for 24 h at room temperature, the mem-
brane was detached from the poly(tetrafluoroethylene) sheet and
dried in vacuo for 24 h at room temperature.
Method II (newly developed method):
A
toluene
solution(1.0 mL) of the base polymer (50 mg) and chloroform so-
lution( 0.50 mL) of the 2D surface modifier (0.175 mg) were first
blended together (In the case of T-EO, methanol was used instead of
chloroform.). And then the mixed solution was cast on a poly(-
tetrafluoroethylene) sheet. After the solvent was evaporated for
24 h at room temperature, the membrane was detached from the
sheet and dried in vacuo for 24 h at room temperature.
2.4. Measurement of oxygen permselectivities
Oxygen and nitrogen permeability coefficients (PO and PN : cm3
2
2
(STP)$cm/cm2$s$cmHg) of mixed gases of oxygen and nitrogen (O2/
N2 ¼ 20/80 (v/v)) were measured by gas chromatographic method
using YANACO GTR-10. The P, the oxygen separation factors (a), the
diffusion coefficients (D) and the solubility coefficients (S) were
calculated by the following equations:
3.2. New preparation method (Method II) of surface-modified
membranes
Qꢂl
P ¼
Aꢂ Pꢂt
D
The surface-modified membranes were prepared from the so-
a
¼ PO =PN
lution of the mixture consisting of small amounts (less than 5.0 wt
2
2
D ¼ l2=6T
S ¼ P=D
where Q, l, A, Dp, t, and T are the amount of the permeated gas, the
thickness of the membrane, the permeation area of the membrane,
the pressure difference across the membrane, the permeation time
and the time lag, respectively. The A and l of the membranes were
0.38e1.77 cm2 and 25e120
mm, respectively. Disc-type membranes
were used. The
Dp was 1 atm and the measurement temperature
was 25 ꢀC.
2.5. Characterization of membranes
ATR-FTIR(using ATR PR0450-S) and FT-IR spectra were recorded
on a JASCO FTIR-4200 spectrometer, contact angles of distilled
water droplets on the air-side surface of the membranes were
measured at 25 ꢀC with a DM301, Kyowa Interface Science Co., Ltd.
2.6. Apparatus
NMR spectra were recorded on a JEOL GSX 270 at 400 MHz for
1H NMR and 13C NMR. Average molecular weight (Mw) was esti-
mated by gel permeation chromatography (tetrahydrofuran as an
eluent, polystyrene calibration) using JASCO Liquid Chromatog-
raphy instruments with PU 2080, DG 2080 53, CO 2060, UV 2070,
CD 2095 and two polystyrene gel columns (Shodex KF 807L).
3. Results and discussion
3.1. Molecular design and synthesis of new 2D surface modifiers
In this study, five kinds of 2D surface modifiers (Chart 2) were
synthesized and used. All the compounds have
a 1,3,5-
trisubstituted benzene as a hydrophobic planar (2D) part, six hy-
droxyl groups, and three hydrophilic(T-EO) or hydrophobic(T-TB, T-
Chart 3. 3D molecular structure of T-EO.

J. Wang et al. / Polymer 55 (2014) 1384e1396
1391
hydrophilic base polymer and the solution was cast in the air, in the
case of this combination, i.e., a more hydrophobic additive and a
more hydrophilic base polymer, the resulting membrane prepared
by Method I (conventional method, see Scheme 4) forms more
hydrophobic surface. Therefore, the combination of T-EO and PVA,
i.e., a hydrophobic additive and a hydrophilic base polymer is
suitable for Method I. As a fact, T-EO was accumulated at the sur-
face of PVA judging from change in the contact angles (Table 1, nos.
1e6). However, in the case of the opposite combination, i.e., a more
hydrophilic additive and a more hydrophobic base polymer like T-
EO and poly(p-trimethylsilylphenylacetylene) (PSPA) were not
suitable for Method I (Table 1, nos. 12e14). To overcome the
problem, a new method, i.e., Method II (see Scheme 4) was
designed and carried out. Methanol and toluene were used as a
good solvent for T-EO and PSPA, respectively. As shown in Table 1,
nos. 15e17, T-EO was successfully accumulated on the surface of
PSPA membranes. Because the boiling point of methanol is lower
than that of toluene, T-EO may be concentrated and precipitated at
the surface first during the evaporation of the solvent after casting.
By this new method, membranes whose surfaces were covered by
more hydrophilic additives could be prepared easily.
3.3. Improvements of oxygen permselectivities by surface
modification using the 2D surface modifiers
3.3.1. Three kinds of membranes modified by the 2D surface
modifier having oligo(ethylene oxides) (T-EO)
Fig. 1 shows the results of oxygen permselectivity of T-EO con-
taining polymer membranes based on PVA, PDPA [28], or PSPA [29].
In all the blend membranes, oxygen permselectivities were higher
than that of pure membranes, PVA(B), PDPA(D), or PSPA(,)
membranes, respectively. In the case of PVA-based membranes
(Fig. 1,C) when 1.0 wt% of T-EO was added, the permselectivity
increased almost twice higher than that of the pure base polymer
membrane, i.e., the PVA membrane without any drop of perme-
ability. When 1.0 wt% of P-EO (the precursor polymer of T-EO, see
Chart 4. Possible 2D supramolecular structure of T-EO on the blend membrane
surface.
%) of one of the 2D surface modifiers and one of the base polymers
(see Chart 2) by two kinds of solvent casting methods as shown in
Scheme 4. In Method I (a conventional method), a solution of the
base polymer and 0.060e5.0 wt.-% of the 2D surface modifiers in a
common solvent was cast on a poly(tetrafluoroethylene) sheet. In
Method II (a new method), one solution of the base polymer in a
non-polar solvent having a higher boiling point and another solu-
tion of 0.060e5.0 wt.-% of the 2D surface modifiers in a polar sol-
vent having a lower boiling point were prepared and then the two
solutions were mixed and the binary solution was cast on a pol-
y(tetrafluoroethylene) sheet. Because hydrophobic compounds
tend to be concentrated at the surface when it was blended with a
Scheme 2) was added (Fig. 1,
), the permselectivity was not
;
improved and the permeability decreased. Therefore the small
amount of T-EO may form a certain surface structure which
enhanced the performance effectively. In particular, in the case of
PDPA-based membranes (Fig. 1, :), when a small amount (1.0 wt%)
of T-EO was added, both the permselectivity and permeability
increased simultaneously. In addition, the plots of T-EO/PDPA
membranes (:) are over the Robeson’s boundary line in 1991 [4]
indicating a relatively good performance. It may be because T-EO
may form an ideal surface structure. On the other hand, in the case
of PSPA-based membranes (Fig. 1,
-
), when 5.0 wt% of T-EO was
added, the permselectivity increased but permeability decreased a
Chart 5. Possible membrane structures of T-EO/PVA membranes having different amount of T-EO.

1392
J. Wang et al. / Polymer 55 (2014) 1384e1396
Table 1
Characterization of the surfaces of T-EO containning polymer membranes.
No.a
Additivesb
Content (wt%)
q
(ꢀ)c
1
2
3
4
5
6
7
8
None (PVA)
0.0
39.1
40.4
40.1
40.4
51.5
53.1
42.2
30.2
34.2
69.5
106
105
107
104
100
95
T-EO (Method I)d
0.06
0.50
1.0
5.0
10.0
1.0
P-EO (Method I)d
5.0
9
10.0
100(¼P-EO)
0.0
0.50
1.0
5.0
0.50
1.0
10
11
12
13
14
15
16
17
None (PSPA)
T-EO (Method I)d
T-EO (Method II)d
5.0
87
a
Nos. 2e9: Prepared by Method I using PVA as a base polymer; nos.12e14:
Prepared by Method I using PSPA as a base polymer; nos.15e17: Prepared by
Method II using PSPA as a base polymer.
b
For the code, see Chart 2.
Advancing contact angles for water droplets on the air surface of the blend
c
Fig. 2. Oxygen permselectivity of PDPA-based polymer membranes.
membranes.
d
Method I: Casting a solution of the base polymer and 2D surface modifier in a
common solvent; Method II: Casting a mixed solution of a 2D surface modifier in a
polar solvent and a base polymer in a nonpolar solvent.
3.3.3. Poly(p-trimethylsilylphenylacetylene) (PSPA)-based
membranes modified by the four kinds of 2D surface modifiers
Fig. 3 shows oxygen permselectivity of PSPA-based polymer
membranes. In all the blend membranes, oxygen permselectivities
were higher than that of the pure PSPA membrane(B) although the
permeability decreased. In particular, since the plot of a T-TES/PSPA
membrane(:) is closest to the Robeson’s boundary line among the
four additives, the performance is relatively good.
little. It may be because T-EO was present not only at the surface
but also inside the membranes (See Chart 5).
3.3.2. Poly(p-trimethylsilyldiphenylacetylene)(PDPA)-based
membranes modified by the four kinds of 2D surface modifiers
Fig. 2 shows oxygen permselectivity of PDPA-based polymer
membranes modified by the four kinds of 2D surface modifiers. In
all the blend membranes, oxygen permselectivities were higher
than that of the pure PDPA membrane(B) although the perme-
ability decreased except for T-EO(C). In particular, T-TB(
enhanced the permselectivity most effectively among the four ad-
ditives (T-EO(C), T-TB( ), T-Do( ), and T-S3(A)). Since the plots
-
)
-
;
of T-EO/PDPA membranes(C) are over the Robeson’s boundary line
in 1991 [4], they are relatively good.
Fig. 1. Oxygen permselectivity of T-EO containing polymer membranes.
Fig. 3. Oxygen permselectivity of PSPA-based polymer membranes.

J. Wang et al. / Polymer 55 (2014) 1384e1396
1393
Table 2
3.3.4. Summary of oxygen permselectivity thorough the surface-
Characterization of the surfaces of PDPA-based polymer membranes.a
modified membranes
No. Additivesb Content (wt%)
q
(ꢀ)c
ASd (ꢂ10ꢁ1
)
AB (ꢂ10ꢁ2
)
Rf
e
Fig. 4 shows the results of oxygen permselectivity of all the
blend membranes in this study together with other membranes
(D) having excellent performances reported by other researchers
[39e43]. All the oxygen permselectivities in this study were higher
than that of the corresponding pure membrane, i.e., PVA(B),
1
2
3
4
5
6
7
8
None
T-EO
0.0
1.0
3.0
5.0
1.0
3.0
5.0
1.0
3.0
5.0
1.0
3.0
5.0
103
93.5 7.03
e
e
e
7.10
8.64
9.73
9.22
9.35
16.0
6.71
7.83
12.8
8.80
9.91
11.8
9.90
8.77
7.94
3.28
4.72
3.49
7.45
6.56
4.21
3.65
4.04
3.47
92.2 7.58
89.6 7.73
96.1 3.03
93.2 4.42
92.5 5.59
90.7 5.00
85.1 5.14
83.9 5.40
93.1 3.22
89.9 4.01
86.8 4.10
T-TB
T-Do
T-S3
PDPA(D), or PSPA(,) membrane, respectively. In particular, T-EO/
PDPA membranes(:) and a T-TES/PSPA membrane(
-
) showed
relatively good performances whose plots are over or close to the
Robeson’s upper boundary line in 1991 [4]. Among the five addi-
tives (Chart 2), T-EO was the best to improve the performance of
the original membranes (C, :). In summary, only small amounts
(1.0e5.0 wt%) of the 2D modifiers were enough for the effective
improvement of oxygen permselectivities.
9
10
11
12
13
a
Prepared by Method II using PDPA as a base polymer (Method II: Casting a
mixed solution of a 2D surface modifier in a polar solvent and a base polymer in a
nonpolar solvent).
3.4. Reasons of the improvements of oxygen permselectivities
b
For the code, see Chart 2.
Advancing contact angles for water droplets on the air surface of the blend
c
3.4.1. Structures of the blend membranes
membranes.
d
To know structures of the membrane surface, contact angles of
water droplets and ATR-IR on the membrane surface were
measured. The results are listed in Tables 1e3. Table 1 shows the
values of contact angles of water droplets on the surface of T-EO
containing blend polymer membranes whose permselectivities
have been shown in Fig. 1. As shown in Table 1, nos. 1e6, judging
from the contact angles, T-EO was accumulated at the surface in the
T-EO/PVA membranes prepared by Method I whose performances
were improved efficiently. On the other hand, since the P-EO/PVA
membranes prepared by Method I whose performances were not
improved efficiently showed almost no change in the contact an-
gles, P-EO was not accumulated at the surface (Table 1, nos. 7e9).
In the case of the T-EO/PSPA membranes prepared by Method II
whose performances were improved efficiently, judging from the
contact angles, T-EO was found to be accumulated at the surface
(Table 1, nos. 15e17). On the other hand, since the T-EO/PSPA
membranes prepared by Method I whose performances were not
improved efficiently showed almost no change in the contact an-
gles, T-EO was not accumulated at the surface (Table 1, nos. 12e14).
Table 2 shows the values of contact angles of water droplets and
ATR-IR at the surface of the PDPA-based polymer blend membranes
prepared by Method II whose performances were improved effi-
AS ¼ A3450/A1250 of ATR-FTIR.
e
AB ¼ A3450/A1250 of FTIR.
f
R ¼ AS/AB.
Table 3
Characterization of the surfaces of PSPA-based polymer membranes.a
No.
Additivesb
Contents (wt%)
q
(ꢀ)c
1
2
3
4
5
6
7
None
T-EO
0.0
0.50
1.0
5.0
3.0
5.0
5.0
106
100
95.0
87.0
97.9
96.5
97.1
T-TB
T-TES
T-Do
a
Prepared by Method II using PSPA as a base polymer(Method II : Casting a mixed
solution of a 2D surface modifier in a polar solvent and a base polymer in a nonpolar
solvent).
b
For the code, see Chart 2.
Advancing contact angles for water droplets on the air surface of the blend
c
membranes.
ciently. The
q values show all the additives were present at the
surfaces because the surface became more hydrophilic. In addition,
Table 4
since the ratios (AS) of absorbance (A3450) for OH to that (A1250) for
Aging effects of the contact angles of the surfaces of PDPA-based polymer
membranes.a
No.
Additivesb
Content (wt%)
q
(ꢀ)c
Original
3 months
1
2
3
4
5
6
7
8
None
T-EO
0.0
1.0
3.0
5.0
1.0
3.0
5.0
1.0
3.0
5.0
1.0
3.0
5.0
103
96.3
94.6
94.0
93.2
96.4
93.8
92.6
91.7
85.7
84.1
93.7
90.4
89.5
93.5
92.2
89.6
96.1
93.2
92.5
90.7
85.1
83.9
93.1
89.9
86.8
T-TB
T-Do
T-S3
9
10
11
12
13
a
Prepared by Method II using PDPA as a base polymer (Method II : Casting a
mixed solution of a 2D surface modifier in a polar solvent and a base polymer in a
nonpolar solvent).
b
For the code, see Chart 2.
Advancing contact angles for water droplets on the air surface of the blend
c
Fig. 4. Oxygen permselectivity of all the surface-modified polymer membranes in this
study.
membranes just after the membrane fabrication and 3 months after the membrane
fabrication under air.

1394
J. Wang et al. / Polymer 55 (2014) 1384e1396
Table 5
Oxygen permselectivity of the T-EO containing polymer membranes.
e
No.a
Additivesb
Content (wt%)
PO (ꢂ10ꢁ2 barrer)c
DO d (ꢂ10ꢁ2
)
SO
PO =PN
DO =DN
SO =SN
2
2
2
2
2
2
2
2
2
1
2
3
4
5
6
7
8
9
10
None(PVA)
0.0
13.7
13.5
13.7
12.7
8.8
5.1
12.9
8.8
9.41
10.91
8.05
8.02
5.64
3.22
7.68
5.83
3.33
13.4
1.45
1.24
1.70
1.58
1.57
1.59
1.68
1.51
1.47
10.1
1.71
1.80
2.68
3.88
3.73
4.11
2.15
2.25
3.82
1.19
1.21
1.11
1.12
1.11
1.01
0.93
1.08
1.12
1.14
1.00
1.42
1.62
2.39
3.49
3.68
4.44
1.99
2.00
3.35
1.18
T-EO (Method I)f
0.06
0.50
1.0
5.0
10.0
1.0
5.0
10.0
100(P-EO)
P-EO (Method I)f
4.9
136
d
e
No.a
Additivesb
Content (wt%)
PO (barrer)c
DO
SO
PO =PN
DO =DN
SO =SN
2
2
2
2
2
2
2
2
2
11
12
13
14
15
16
17
None(PSPA)
0.0
0.50
1.0
5.0
0.50
1.0
171
150
136
120
142
128
111
3.35
2.95
2.60
2.32
3.25
3.10
2.98
51.1
50.8
52.4
51.8
43.7
41.3
37.2
2.36
2.45
2.55
2.63
2.58
2.88
3.22
0.95
0.98
1.04
1.02
0.97
1.02
1.06
2.48
2.51
2.45
2.58
2.66
2.82
3.16
T-EO (Method I)f
T-EO (Method II)f
5.0
a
Nos. 2e9: Prepared by Method I using PVA as a base polymer; no.10: a pure P-EO membrane; nos.12e14: Prepared by Method I using PSPA as a base polymer; nos.15e17:
Prepared by Method II using PSPA as a base polymer.
b
For the code, see Chart 2.
c
1 barrer ¼ 10ꢁ10cm3(STP)$cm cmꢁ2 sꢁ1 cmHgꢁ1
.
d
e
f
In 10ꢁ7 cm2 sꢁ1
.
In 10ꢁ3 cm3(STP) cmꢁ3 cmHgꢁ1
Method I: Casting a solution of the base polymer and 2D surface modifier in a common solvent; Method II: Casting a mixed solution of a 2D surface modifier in a polar
.
solvent and a base polymer in a nonpolar solvent.
SiCH3 of the surface in the infrared spectra were much higher than
those of the bulk (AB), in other words, AS values were much higher
than AB values (i.e., R ¼ AS/AB >1.0), the accumulation of the addi-
tives was confirmed. Table 3 shows the values of contact angles of
water droplets at the surface of the PSPA-based polymer blend
membranes prepared by Method II whose performances were
derivatives may have interaction with the hydrophobic substrate
polymers.
In order to estimate stability of the modified surface, aging ef-
fects of the contact angles were measured. Table 4 shows aging
effects of the contact angles on the PDPA-based polymer mem-
branes. Since almost no change in the contact angles was observed,
it was found that the surface layer was stable.
improved efficiently. The
q values show all the additives were
present at the surfaces because the surface became more hydro-
philic similarly to those of the PDPA-based membranes prepared by
Method II.
3.4.2. Reason of the improvements of oxygen permselectivities
As described in Sections 3.2 and 3.3.1, in all the blend membranes
containing the 2D surface modifiers, the oxygen permselectivities
were higher than the corresponding pure base polymer membranes
and the additives were accumulated at the surface. Therefore, the
surface layer must play an important role. To discuss the reason of
the enhancement, diffusion coefficients (D) were estimated by the
time lag methods and then solution coefficients (S) were calculated
from P ¼ D x S. The results are listed in Tables 5e7. Table 5 shows the
values of P, D, and S of oxygen and nitrogen and their ratios, i.e.,
In summary, it was found that all the additives were accumu-
lated at the surface in all the blend membranes containing the 2D
surface modifiers. The possible supramolecular structure of T-EO at
the surface of the T-EO/PVA membrane is shown in Chart 5. The
three hydrophilic oligoethylene oxide groups in T-EO can work as
anchors. In the case of the T-EO/PDPA and the T-EO/PSPA mem-
branes, the three oligoethylene oxide groups in T-EO can not work
as anchors. Instead, the hydrophobic part, i.e., the benzene
Table 6
Oxygen permselectivity of the PDPA-based polymer membranes.a
c
d
e
No.
Additivesb
Content (wt%)
PO (barrer)
DO
SO
PO =PN
DO =DN
SO =SN
2
2
2
2
2
2
2
2
2
1
2
3
4
5
6
7
8
None
T-EO
0.0
1.0
3.0
5.0
1.0
3.0
5.0
1.0
3.0
5.0
1.0
3.0
5.0
1520
3380
3190
3055
404
385
366
604
577
549
278
256
236
0.488
0.112
0.098
0.087
0.252
0.249
0.243
0.275
0.312
0.361
0.0375
0.0378
0.0371
311
3030
3230
3440
160
155
151
219
186
152
742
678
636
1.83
2.27
2.30
2.34
2.80
2.86
2.92
2.32
2.35
2.38
1.89
1.93
2.10
1.78
1.36
1.26
1.17
1.43
1.38
1.31
1.48
1.43
1.36
1.40
1.40
1.38
1.03
1.67
1.83
2.00
1.96
2.07
2.24
1.57
1.64
1.75
1.35
1.37
1.50
T-TB
T-Do
T-S3
9
10
11
12
13
a
Prepared by Method II using PDPA as a base polymer (Method II : Casting a mixed solution of a 2D surface modifier in a polar solvent and a base polymer in a nonpolar
solvent.).
b
c
For the code, see Chart 2.
1 barrer ¼ 10ꢁ10 cm3(STP)$cm cmꢁ2 sꢁ1 cmHgꢁ1
.
d
e
In 10ꢁ6 cmꢁ2 sꢁ1.
In 10ꢁ3 cm3(STP)$cmꢁ3 cmHgꢁ1
.

J. Wang et al. / Polymer 55 (2014) 1384e1396
1395
Table 7
Oxygen permselectivity of the PSPA-based polymer membranes.a
d
e
No.
Additivesb
Contents (wt%)
PO c (barrer)
DO
SO
PO =PN
DO =DN
SO =SN
2
2
2
2
2
2
2
2
2
1
2
3
4
5
6
7
None
T-EO
0.0
0.50
1.0
5.0
3.0
5.0
5.0
229
142
128
111
212
172
132
3.02
3.25
3.10
2.98
3.86
e
76.0
43.7
41.3
37.2
55.0
2.69
2.58
2.88
3.22
2.83
3.38
2.96
1.41
0.97
1.02
1.06
0.98
e
1.91
2.66
2.82
3.16
2.90
e
T-TB
T-TES
T-Do
e
0.981
135
1.17
2.54
a
Prepared by Method II using PSPA as a base polymer (Method II : Casting a mixed solution of a 2D surface modifier in a polar solvent and a base polymer in a nonpolar
solvent.).
b
c
For the code, see Chart 2.
1 barrer ¼ 10ꢁ10cm3(STP)$cm cmꢁ2 sꢁ1 cmHgꢁ1
.
d
e
In 10ꢁ7 cm2 sꢁ1
.
In 10ꢁ3 cm3(STP)$cmꢁ3 cmHgꢁ1
.
PO =PN (¼
a
), DO =DN , and SO =SN through the T-EO containing
(51103076), Grant-in-Aid for Young Scientists (B) (No. 25810068)
from the Japan Society for the Promotion of Science, and Grant in
Aid for Scientific Research (C) (No. 25390002) from the Japan So-
ciety for the Promotion of Science is acknowledged.
2
2
2
2
2
2
blend polymer membranes. As shown in Table 5, nos. 1e6, in the T-
EO/PVA membranes, the enhancements of values were found to be
mainly caused by those of SO =SN and the decreases of PO values
a
2
2
2
were found to be caused by those of DO . The extent of increase of
a
2
and SO =SN values in 1.0 or 5.0 wt% of T-EO/PVA membranes were
References
2
2
higher than those in 1.0 or 5.0 wt% of P-EO/PVA membranes (Table 5,
nos. 4,5 and 7,8). Therefore, T-EO was more effective than P-EO for
enhancing the performances because T-EO could show a better
surface layer. In the case of the T-EO/PSPA membranes prepared by
[1] Robeson LM. J Membr Sci 2008;320:390e400.
[2] Aoki T. Prog Polym Sci 1999;24:951e93.
[3] Stern SA. J Membr Sci 1994;94:1e65.
[4] Robeson LM. J Membr Sci 1991;62:165e85.
[5] Kim TH, Koros WJ, Husk GR. J Membr Sci 1988;37:45e62.
[6] Michaels AS, Bixler HJ. J Polym Sci 1961;50:413e39.
[7] Aoki T, Kaneko T, Teraguchi M. Polymer 2006;47:4867e92.
[8] Sato T, Yoshida N, Ishida A, Teraguchi M, Aoki T, Kaneko T. Polymer 2013;54:
2231e4.
[9] Hirayama Y, Yoshinaga T, Kusuki Y, Ninomiya K, Sakakibara T, Tamari T.
J Membr Sci 1996;111:169e82.
[10] Jung CH, Lee JE, Han SH, Park HB, Lee YM. J Membr Sci 2010;350:301e9.
[11] Sen SK, Banerjee S. J Membr Sci 2010;350:53e61.
[12] Ahn J, Chung WJ, Pinnau I, Song J, Du N, Robertson GP, et al. J Membr Sci
2010;346:280e7.
Method II (Table 5, nos.15e17), increase of
a values were found to be
caused by those of SO =SN similarly to those in T-EO/PVA. On the
2
2
other hand, in T-EO/PSPA prepared by Method I showed almost no
effects on enhancing SO =SN (Table 5, nos. 12e14). Therefore, the
2
2
surface thin layers produced by the 2D additives were effective for
enhancing SO =SN Tables 6and 7 show the values of P, D, and S of
.
2
2
oxygen and nitrogen and their ratios, i.e., PO =PN (¼
a), DO =DN , and
2
2
2
2
SO =SN through the PDPA- and PSPA-based blend polymer mem-
2
2
branes, respectively. All the enhancements of
a were found to be
[13] Thomas S, Pinnau I, Du N, Guiver MD. J Membr Sci 2009;333:125e31.
[14] Kaneko T, Yamamoto K, Asano M, Teraguchi M, Aoki T. J Membr Sci 2006;278:
365e72.
caused by those of SO =SN
.
2
2
[15] Barsema JN, Balster J, Jordan V, vander Vegt NFA, Wessling M. J Membr Sci
2003;219:47e57.
[16] Tanaka K, Kawai T, Kita H, Okamoto K, Ito Y. Macromolecules 2000;33:
5513e7.
[17] Centeno TA, Fuertes AB. J Membr Sci 1999;160:201e11.
[18] Freeman BD. Macromolecules 1999;32:375e80.
[19] Kusakabe K, Yamamoto M, Morooka S. J Memb Sci 1998;149:59e67.
[20] Robeson LM, Smith CD, Langsam M. J Membr Sci 1997;125:23e39.
[21] Kita H, Yoshino M, Tanaka K, Okamoto K. J Chem Soc Chem Commun 1997:
1051e2.
4. Conclusions
In this article, a facile synthesis of five 2D surface modifiers
including two new compounds are reported by the SCAT(highly
selective photocyclic aromatization) reaction and effective
enhancement of
a
( ¼ PO =PN ) through the three polymer mem-
2
2
branes are reported by adding a small amount of the 2D surface
[22] Ghosal K, Freeman BD. Polym Adv Technol 1993;5:673e97.
[23] Matsumoto K, Xu P. J Membr Sci 1993;81:23e30.
[24] Kim TH, Koros WJ, Husk GR, O’Brien KC. J Memb Sci 1988;37:45e62.
[25] Kawakami Y, Aoki T, Yamashita Y. Polym Bull 1987;17:293e7.
[26] Kawakami Y, Karasawa H, Kamiya H, Aoki T, Yamashita Y. Polym J 1986;18:
237e43.
[27] Kawakami Y, Aoki T, Yamashita Y, Hirose M, Ishitani A. Macromolecules
1985;18:580e2.
[28] Saenz de Jubera AM, Herbison JH, Komari Y, Plewa MJ, Moore JS, Cahill DG,
et al. Environ Sci Technol 2013;47:8642e9.
modifiers. Among the five 2D surface modifiers, T-EO showed the
best performance for the enhancement. The
a values in T-EO/PVA
increased with no decrease in PO and the and PO values in T-EO/
a
2
2
PDPA increased simultaneously. These improvements were thought
to be caused mainly by improvement of the solution selectivity on
the membrane surface where the 2D surface modifiers were
accumulated. In all the blend membranes, T-EO/PDPA, T-TB/PDPA,
and T-TES/PSPA membranes showed relatively good performance,
[29] Gao Y, Saenz de Jubera AM, Marinas BJ, Moore JS. Adv. Funct. Mater 2013;23:
598e607.
since the plots in the PO -a graph were close to the upper boundary
2
[30] Saenz de Jubera AM, Gao Y, Moore JS, Cahill DG, Marinas BJ. Environ Sci
Technol 2012;46:9592e9.
[31] Liu L, Namikoshi T, Zang Y, Aoki T, Hadano S, Abe Y, et al. J Am Chem Soc
2013;135:602e5.
line by Robeson in 1991 [4]. Since all the membranes showed better
permselectivities than the corresponding substrate membranes, it
is very promising for the future.
[32] Masuda T, Mishima K, Fujimori J, Nishida M, Muramatsu H, Higashimura J.
Macromolecules 1992;25:1401e4.
[33] Aoki T, Nakahara H, Hayakawa Y, Kokai M, Oikawa E. J Polym Sci Part A Polym
Chem 1994;32:849e58.
Acknowledgments
[34] Aoki T, Kaneko T, Maruyama N, Sumi A, Takahashi M, Sato T, et al. J Am Chem
Soc 2003;125:6346e7.
[35] Wang J, Aoki T, Liu L, Namikoshi T, Teraguchi M, Kaneko T. Chem Lett
2013;42:1090e2.
[36] Liu L, Zang Y, Hadano S, Aoki T, Teraguchi M, Kaneko T, et al. Macromolecules
2010;43:9268e76.
[37] Zang Y, Nakao K, Yotsuyanagi H, Aoki T, Namikoshi T, Tsutsuba T, et al.
Polymer 2013;54:1729e33.
Partial financial support through a Grant-in-Aid for Scientific
Research (B) (No. 19350054) from the Japan Society for the Pro-
motion of Science, the National Natural Science Foundation of
China (No. 21204014), and the Fundamental Research Funds for the
Central Universities, China (Nos. HEUCFR1227 and HEUCFT1009),
the National Natural Science Foundation of China (U1162123) and

1396
J. Wang et al. / Polymer 55 (2014) 1384e1396
[38] Zang Y, Aoki T, Liu L, Abe Y, Kakihana Y, Teraguchi M, et al. Chem Commun
2012;48:4761e3.
[39] Robeson LM. Curr Opin Solid State Mater Sci 1999;4:549e52.
[40] Sakaguchi T, Shiotsuki M, Masuda T. Macromolecules 2004;37:4104e8.
[41] Hu Y, Shiotsuki M, Sanda F, Masuda T. Chem Commun 2007:4269e70.
[42] Xua ZK, Dannenberg C, Springer J, Banerjee S, Maier G. J Membr Sci 2002;205:
23e31.
[43] Du N, Robertson GP, Pinnau I, Guiver MD. Macromolecules 2010;43:8580e7.