Soluble high-Tg polyetherimides with good flame retardancy based on an asymmetric phosphinated etherdiamine

Article abstract of DOI:10.1002/pola.24552

This study focuses on preparing soluble, high-T_g polyetherimides (PEIs) with good flame retardancy. To achieve this goal, a phosphinated aminophenol, 1-(4-hydroxyphenyl)-1-(4-aminophenyl)-1-(6-oxido-6H-dibenz(c,e)(1, 2)oxaphosphorin-6-yl)ethane (1), was prepared. Then, an asymmetric etherdiamine, 1-(4-(4-aminophenoxy)phenyl)-1-(4-aminophenyl)-1-(6-oxido-6H-dibenz(c,e)(1,2) oxaphosphorin-6-yl)ethane (3), was prepared from the nucleophilic substitution of (1) with 4-chloronitrobenzene, followed by catalytic hydrogenation. Light colored, tough, and foldable PEIs (5) with good thermal stability, organo-solubility, and good flame retardancy were synthesized from the condensation of (3) with various aromatic dianhydrides in N,N-dimethylacetamide, followed by thermal imidization. Properties of the resulting PEIs were evaluated and compared with those of symmetric PEIs (5′) and polyimides (5′). The structure-property relationships were discussed.

Full text of DOI:10.1002/pola.24552

Soluble High-Tg Polyetherimides with Good Flame Retardancy Based on
an Asymmetric Phosphinated Etherdiamine
CHING HSUAN LIN, SHENG LUNG CHANG, PO WEN CHENG
Department of Chemical Engineering, National Chung Hsing University, Taichung, Taiwan
Received 22 September 2010; accepted 7 December 2010
DOI: 10.1002/pola.24552
Published online 20 January 2011 in Wiley Online Library (wileyonlinelibrary.com).
ABSTRACT: This study focuses on preparing soluble, high-Tg
polyetherimides (PEIs) with good flame retardancy. To achieve
this goal, a phosphinated aminophenol, 1-(4-hydroxyphenyl)-1-
(4-aminophenyl)-1-(6-oxido-6H-dibenzhc,eih1,2ioxaphosphorin-
6-yl)ethane (1), was prepared. Then, an asymmetric
etherdiamine, 1-(4-(4-aminophenoxy)phenyl)-1-(4-aminophenyl)-
sized from the condensation of (3) with various aromatic
dianhydrides in N,N-dimethylacetamide, followed by thermal
imidization. Properties of the resulting PEIs were evaluated and
compared with those of symmetric PEIs (50) and polyimides (500).
C
V
The structure–property relationships were discussed.
2011
Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 49:
1331–1340, 2011
1-(6-oxido-6H-dibenzhc,eih1,2ioxaphosphorin-6-yl)ethane
(3),
was prepared from the nucleophilic substitution of (1) with 4-
chloronitrobenzene, followed by catalytic hydrogenation. Light
colored, tough, and foldable PEIs (5) with good thermal stabil-
ity, organo-solubility, and good flame retardancy were synthe-
KEYWORDS: asymmetric; phosphinate; polyimides; solubility;
structure–property relationship
INTRODUCTION Polyimides (PIs) are useful high-perform-
ance electronic materials because of their high glass transi-
tion temperature, high thermal stability, and good mechani-
cal properties.1–8 However, PIs are difficult to process due to
their poor solubility in organic solvents. It has been reported
that incorporating bulky pendants9–13 and flexible ether link-
ages14–16 can impart good solubility.
catalytic hydrogenation. Based on diamines (3), a series of
PEIs (5) were prepared. The solubility, crystallinity, glass
transition, dimensional stability, and thermal stability of
the PEIs were investigated and compared with those of
symmetric PEIs (50)34 and PIs (500)41 (Fig. 1). As shown in
Figure 1, PEIs (50) contain two phenoxy linkages, PEIs (5)
contain one phenoxy linkage, but PI (500) does not contain
any phenoxy linkage.
It has also been reported that phosphorus-functionalized
polymers, whether in the phosphine oxide form,17–21 in the
phosphonate22,23 or in the phosphinate form,24–33 can
improve organo-solubility, adhesion to metal, flame retard-
ancy, and atomic oxygen resistance.
EXPERIMENTAL
Materials
9,10-Dihydro-oxa-10-phosphaphenanthrene-10-oxide (DOPO,
from TCI), 4-hydroxyacetophenone (from Acros), aniline
(from Acros), p-toluenesulfonic acid monohydrate (p-TSA,
from SHOWA), potassium carbonate, 10% palladium on car-
bon (Pd/C, from Acros), 4-chloronitrobenzene (from Acros),
and calcium hydride (from Acros) were used as received.
Pyromellitic dianhydride (PMDA, from Acros) was dried at
In our previous work, we developed polyetherimides (PEIs)
based on a dietheramine, 1,1-bis(4-(4-aminophenoxy)phenyl)-1-
(6-oxido-6H-dibenzhc,eih1,2ioxaphosphorin-6-yl)ethane.34
Although the organo-solubility of the resulting PEIs was
improved, their solvents were limited to high-boiling-point
ones. Studies have shown that incorporating asymmetric struc-
ture into polymer chains can impart better organo-solubility
because of the disturbed sequence distribution.35–40 Herein,
continuing our work in organo-soluble phosphinated PIs,
an asymmetric etherdiamine, 1-(4-(4-aminophenoxy)phenyl)-1-
(4-aminophenyl)-1-(6-oxido-6H-dibenzhc,eih1,2ioxaphosphorin-
6-yl)ethane (3), was prepared from the nucleophilic substitu-
tion of a new phosphinated aminophenol, 1-(4-hydroxyphenyl)-
1-(4-aminophenyl)-1-(6-oxido-6H-dibenzhc,eih1,2ioxaphosphorin-6-
yl)ethane (1), with 4-chloronitrobenzene, followed by
ꢀ
170 C overnight before use. 3,30,4,40-Benzophenonetetracar-
boxylic dianhydride (BTDA, from Acros), 4,40-oxydiphthalic
anhydride (ODPA, from Chriskev), and 3,30,4,40-biphenyltetra-
carboxylic dianhydride (BPDA, from Chriskev) were recrys-
tallized from acetic anhydride. N,N-Dimethylacetamide
(DMAc, from TEDIA) was purified by distillation under
reduced pressure over calcium hydride (from Acros) and
stored over molecular sieves. The other solvents used are
commercial products (HPLC grade) and were used without
any further purification.
Additional Supporting Information may be found in the online version of this article. Correspondence to: C. H. Lin (E-mail: linch@nchu.edu.tw)
C
V
Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 49, 1331–1340 (2011) 2011 Wiley Periodicals, Inc.
SOLUBLE HIGH-Tg PEIs WITH GOOD FLAME RETARDANCY, LIN, CHANG, AND CHENG
1331

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
was in the range of 10–30 s without any dripping, the poly-
mer was considered to be a VTM-1 material.
Synthesis of (1)
DOPO 25.94 g (0.12 mol), 4-hydroxyacetophenone 16.34 g
(0.12 mol), aniline 44.70 g (0.48 mol), and p-TSA 1.038 g (4
wt % of DOPO) were introduced into a 250-mL round-bot-
tom glass flask equipped with a nitrogen inlet and a mag-
netic stirrer. The reaction mixture was stirred at 130 ꢀC for
24 h. After that, the white precipitate was filtered and
recrystallized from methanol, and then dried in a vacuum
ꢀ
oven at 150 C for 8 h. Light white crystal (55% yield) with
ꢀ
a melting point of 294 C (DSC) was obtained. HR-MS(EI) m/
z: calcd. for C26H22O3N1P 427.1337; found 427.1332. 1H
NMR (DMSO-d6), d ¼ 1.57 (3H, H18), 4.99 (2H, NH2), 6.41
(2H, H2), 6.57(2H, H21), 7.07 (2H, H3), 7.08 (2H, H20),7.10
(1H,H15), 7.12(1H, H9), 7.17(1H, H13), 7.34 (1H, H14), 7.36
(1H, H8), 7.66 (1H, H7), 7.98 (1H, H12), 8.81 (1H,
H6),9.36(1H,OH) 13C NMR (DMSO-d6), d ¼ 24.17 (C18), 51.62
(C17), 113.34 (C21), 114.37(C2), 118.98 (C9), 120.88 (C11),
122.31 (C5), 123.34 (C10), 123.83(C13), 125.31 (C12), 127.78
(C8), 128.93 (C3), 128.97 (C21), 129.54(C4), 130.54 (C14),
131.82 (C15), 133.20 (C7), 135.98 (C19), 147.03 (C1), 150.73
(C22), 156.18 (C16). 31P NMR (DMSO-d6), d ¼ 42.30.
FIGURE 1 Structures of PEIs (5), PEIs (50),34 and PIs (500).41
Characterization
NMR measurements were performed using a Varian Inova
600 NMR in DMSO-d6, and the chemical shift was calibrated
by setting the chemical shift of DMSO-d6 as 2.49 ppm. The
assignment of the individual peak of (1–3) was assisted by
the correlations shown in the 1H-1H COSY and 1H-13C HET-
COR NMR spectra. IR Spectra were measured by Perkin-
Elmer RX1 infrared spectrophotometer. High-resolution mass
measurements were performed by a Finnigan/Thermo Quest
MAT 95XL mass spectrometer by fast atom bombardment.
Elemental analysis was performed on an Elementar Vario EL
III. Differential scanning calorimeter (DSC) scans were
obtained by a Perkin-Elmer DSC 7 in a nitrogen atmosphere
Synthesis of (2)
To a round-bottom 250-mL glass flask equipped with a nitro-
gen inlet, a condenser, and a magnetic stirrer, (1) 10.0g
(23.3 mmol), 4-chloronitrobenzene 3.87 g (24.5 mmol), po-
tassium carbonate 3.39 g (23 mmol), and DMAC 46 mL were
introduced. The reaction mixture was heated to 120 ꢀC and
maintained at that temperature for 12 h. Then, the reaction
mixture was filtered, and the filtrate was poured into 450
mL of enthanol/water (1/2, V/V) solution under stirring.
ꢀ
at a heating rate of 20 C/min. Tg was taken as the midpoint
of the heat capacity transition. Dynamic mechanical analysis
(DMA) was performed byꢀ a Perkin-Elmer Pyris Diamond
DMA at a heating rate of 5 C/min. Thermo mechanical anal-
ysis (TMA) was performed by a SII TMA/SS6100 at a heating
rate of 5 ꢀC/min. The onset of transition in dimension was
defined as Tg, and coefficient of thermal expansion (CTE)
before Tg was defined by calculating the TMA data from 50
to Tg ꢁ20 ꢀC. Thermal gravimetric analysis (TGA) wꢀas per-
formed by a Seiko Exstar 600 at a heating rate of 20 C/min
in a nitrogen or air atmosphere. Flame retardancy of PIs was
performed by a UL-94VTM vertical thin test. In that test, an
800 ꢂ 200 sample was wrapped around a 1/200 mandrel, and
then taped on one end. The mandrel was removed, leaving a
cone-shaped sample that was relatively rigid. The two flame
applications were 3 s. After the first ignition, the flame was
removed and the time for the polymer to self-extinguish (t1)
was recorded. Cotton ignition was noted if polymer dripping
occurred during the test. After cooling, the second ignition
was performed on the same sample and the self-extinguish-
ing time (t2) and dripping characteristics were recorded. If
t1 plus t2 was less than 10 s without any dripping, the poly-
mer was considered to be a VTM-0 material. If t1 plus t2
The yellow precipitate was filtered and recrystallized from
ethyl acetate/DMAC, and then dried in a vacuum oven at
110 ꢀC for 6 h. Light yellow crystals (66% yield) with a
melting point of 180 ꢀC (DSC) were obtained.HR-MS(EI) m/
z: calcd. for C32H25O5N2P 548.1501; found 548.1507. 1H
NMR (DMSO-d6), d ¼ 1.73 (3H, H18), 5.05 (2H, NH2), 6.45
(2H, H2), 6.93 (2H, H21), 7.03 (2H,H24), 7.12 (2H, H3), 7.21
(2H,H13), 7.29 (1H, H20), 7.32 (1H, H15), 7.34 (1H, H9), 7.36
(1H, H14), 7.42 (1H, H8), 7.70 (1H, H7), 8.01 (1H, H12), 8.12
(1H, H6), 8.29 (1H, H25).13C NMR (DMSO-d6), d ¼ 24.31
(C18), 52.43 (C17), 113.48 (C2), 117.30 (C24), 119.26 (C21),
120.61 (C11), 121.99 (C5), 123.38 (C6), 123.83 (C13), 125.31
(C12), 126.19 (C25), 127.78 (C8), 128.93 (C3), 130.54 (C14),
131.58 (C20), 131.82 (C15), 131.94 (C9), 133.20 (C7), 135.98
(C19), 136.21 (C4), 137.33 (C22), 147.03 (C26), 147.60 (C1),
1332
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA

ARTICLE
SCHEME 1 Synthesis of (1).
150.73 (C22), 156.18 (C16), 162.10 (C23). 31P NMR (DMSO-
d6), d ¼ 40.09.
(1H, H9), 7.32 (1H, H14), 7.37 (1H, H8), 7.68 (1H, H7), 7.94
(1H, H12), 8.06 (1H, H6). 13C NMR (ppm, DMSO-d6), d ¼
30.51 (C18), 52.09 (C17), 113.41 (C2), 114.81 (C25), 115.43
(C21), 119.43 (C15), 122.22 (C24), 122.86 (C5), 123.21 (C6),
123.84 (C13), 125.26 (C12), 128.04 (C8), 129.11 (C20), 130.44
(C14), 130.85 (C3), 131.80 (C9), 132.66 (C10), 133.41 (C11),
134.22 (C7), 136.78 (C4), 137.59 (C19), 145.61 (C26), 146.23
(C1), 147.87 (C23), 151.22 (C16), 156.36 (C22). 31P NMR
(DMSO-d6), d ¼ 40.48.
Synthesis of (3)
(2) 10 g, 10% Pd/C 0.15 g, and DMF 50 mL were introduced
into a Parr autoclave at room temperature. The reaction sys-
tem was purged with nitrogen for 30 min before introducing
hydrogen. The pressure of the autoclave was maintained at
about 140 psi under hydrogen atmosphere for 24 h. Then,
the reaction mixture was filtered to remove Pd/C, and the
filtrate was poured into 500 mL of water under stirring. The
precipitate was filtered and then dried in a vacuum oven at
70 ꢀC. Off-white powders (80% yield) were obtained. Mass
spectra (FAB) (m/e, relative intensity): 507 (M þ 1þ 78),
398 (M ꢁ 108þ, 30), 154 (biphenyl, 100). HR-MS(EI) m/z:
calcd. for C32H27O3N2P 518.1759; found 518.1763. 1H NMR
(ppm, DMSO-d6), d ¼ 1.63 (3H, H18), 5.02 (4H, NH2), 6.40
(2H, H2), 6.58 (2H, H25), 6.61 (2H, H21), 6.68 (2H, H24), 7.04
(1H, H15), 7.06 (2H, H3),7.08 (2H, H20), 7.12 (1H, H13), 7.23
Preparation of PEIs
Phosphinated PEIs were prepared by reacting (3) with equal
molar dianhydrides (a–d). PEIs syntheses were exemplified
by the specific synthesis of (5a) from the condensation of
(3) and PMDA. To a 100-mL three-neck round-bottom flask
equipped with a nitrogen inlet and magnetic stirrer, (3)
1.037 g (2 mmol) and CaH2-dried DMAc 5.89 g were added.
After the monomer had dissolved completely, the reactor
was placed into an ice-bath cooling system. After the
FIGURE 2 1H NMR spectrum of (1).
SOLUBLE HIGH-Tg PEIs WITH GOOD FLAME RETARDANCY, LIN, CHANG, AND CHENG
1333

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
SCHEME 2 Proposed mechanism for the regio-selective electrophilic substitution. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
reaction mixture reached 3–5 ꢀC, PMDA 0.436 g (2 mmol)
was added quickly and reacted at that temperature for 2 h.
Then, the viscous poly(amic acid) (4a) was diluted with
DMAc 2.45 g (solid content 15 wt %) and then casted on
glass by an automatic film applicator. The film was dried at
80 ꢀC overnighꢀt, and then imidized at 100 ꢀC (1 h), 200 ꢀC
(1 h), and 300 C (1 h), respectively.
nism for the synthesis of (1) was proposed (Supporting In-
formation Scheme S1). In that mechanism, DOPO initially
attacks the carbonyl group of 4-hydroxyacetophenone via a
nucleophilic addition, forming an intermediate with a terti-
ary hydroxyl group. The tertiary hydroxyl group is then pro-
tonated by acid. Spontaneous dissociation of the protonated
hydroxyl group occurs to yield a tertiary carbocation. Then,
aniline reacts with the resulting carbocation via an electro-
philic substitution, yielding (1). Figure 2 shows the 1H NMR
spectrum of (1). An amino peak at 5.0 ppm and a phenol
peak at 9.4 ppm were observed. Signals of the methyl group
at 1.5 ppm were split into two peaks with a coupling con-
RESULTS AND DISCUSSION
Monomer Synthesis
Compound (1) was prepared by the reaction of DOPO, 4-
hydroxyacetophenone in excess aniline using p-TSA as cata-
lyst via a one-pot procedure (Scheme 1). The excess aniline
plays the roles of reactant and reaction medium. A nucleo-
philic addition, followed by electrophilic substitution mecha-
3
stant of 17.4 Hz because of a JPAH coupling. Signals from H2
and H21 (ortho to the amino and ether group, respectively)
show the smallest chemical shift at 6.4–6.6 ppm because of
SCHEME 3 Synthesis of (2) and (3).
1334
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA

ARTICLE
FIGURE 3 1H NMR spectrum of (2).
FIGURE 4 1H NMR spectrum of (3).
SOLUBLE HIGH-Tg PEIs WITH GOOD FLAME RETARDANCY, LIN, CHANG, AND CHENG
1335

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
SCHEME 4 Synthesis of PEIs (5).
the shielding effect of the electron-donating amino and ether
groups. The other spectroscopic data such as 13C NMR,
1H-1H COSY, and 1H-13C HETCOR spectra (Supporting Infor-
mation Figs. S1–S3) support the structure and purity of (1).
Compound (2) was easily synthesized from the nucleophilic
substitution of (1) and 4-chloronitrobenzene at 120 C using
ꢀ
potassium carbonate as catalyst (Scheme 3). In our previous
work,42 we reported that PAC (aliphatic carbon) is stable in
an alkaline condition. That claim can be further supported in
this case. Figure 3 shows the 1H NMR spectrum of (2). Sig-
nals from H24 and H25 indicate that the nucleophilic substi-
tution was successful. Because of the inductive and aniso-
tropic deshielding effect of the nitro group, the signal from
H25 shows the largest chemical shift at around 8.3 ppm. In
contrast, the signal from H2 (ortho to the amino group)
shows the smallest chemical shift at 6.4 ppm because of the
shielding effect of the electron-donating amino group. The
other spectroscopic data such as 13C NMR, 1H-1H COSY, and
1H-13C HETCOR spectra (Supporting information Figs. S4–S6)
also support the structure and purity of (2).
Because the phosphorus and aliphatic carbon in (1) are both
chiral centers, four stereoisomers should result. Theoreti-
cally, each stereocenter can be either R or S configuration,
and hence, the possible combinations are RR, RS, SR, and SS
configurations. The four stereoisomers can be grouped into
two pairs of enantiomers, resulting in two pairs of diaster-
eomers: RR þ SS and RS þ SR. The RR stereoisomer is the
enantiomer of the SS stereoisomer and is diastereomerically
related to the RS and SR stereoisomers. However, according
to the spectroscopic data of (1), only racemic mixture with
equal RR and SS configurations or equal RS and SR configu-
rations was obtained. It is speculated that the electrophilic
substitution of aniline on the 3ꢀ carbocation (Scheme 2) is
limited to one direction, leading to regioselective reaction.
However, a more detailed research is required to explain this
phenomenon.
Compound (3) was readily obtained by catalytic reduction of
(2). Figure 4 shows the 1H NMR spectrum of (3). An amino
peak corresponding to four hydrogen’s at 5.02 ppm was
observed, indicating that the two amino groups have almost
the same chemical shift. Because of the shielding effect of
FIGURE 5 X-ray single crystal dif-
fractogram of (2).
1336
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA

ARTICLE
TABLE 1 Inherent Viscosity and Solubility Data of PEIs (5)
Solvent for Solubility Testb
ginh
Polymer
(dL gꢁ1 a
)
NMP
DMAc
DMF
DMSO
m-cresol
CH2Cl2
CHCl3
THF
5a
5b
5c
5d
5c0
a
0.51
0.61
0.58
0.73
0.64
þhc
þh
þh
þh
þh
þh
þh
þh
þh
6
þꢁ
þh
þh
6
ꢁ
þh
þh
6
ꢁ
þ
6
6
6
ꢁ
þ
ꢁ
þ
6
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
6
þh
þh
ꢁ
þ
6
6
c
Measured at a polymer concentration of 0.5 g/dL in NMP at 30 ꢀC.
þ, soluble at room temperature; þh, soluble on heating; 6, partially
b
Solubility was tested with a 5 mg sample in 0.5 g of solvent at room
soluble on heating; ꢁ, insoluble on heating.
temperature.
the electron-donating amino groups, signals of H25 upshifted
from 8.3 ppm in (2) to 6.6 ppm in (3), demonstrating the
successful reduction of the nitro group. The other spectro-
which leads to the formation of configuration isomers of the
repeat unit and resulted in the disruption of the packing
sequence, should be responsible for the improved solubility.
In particular, PEI (5b), derived from flexible ODPA, is soluble
in low-boiling solvents, such as CH2Cl2 and CHCl3.
scopic data such as 13C NMR, 31P, 1H-1H COSY, and 1H-13
C
HETCOR spectra (Supporting Information Figs. S7–S9) sup-
port the structure and purity of (3).
Film Quality and Thermal Properties
Supporting Information Figure S11 shows a picture of (5a),
in which PEI (5a) displays light color and is foldable. The
other PEIs also show light-color and foldable properties.
Because all PEIs (5) are tough and foldable, DMA and TMA
were applied to evaluate their thermal mechanical properties
and dimensional stability. Figure 6 shows the DMA curves of
PEIs (5), and the DMA data are listed in Table 2. The tem-
peratures corresponding to 1 GPa were as high as 328, 288,
290, and 303 ꢀC for (5a–5d), respectively, and the Tg’s
obtained from the peak temperatures of tan d are 336, 291,
300, and 311 ꢀC for (5a–5d), respectively. Among the PEIs,
(5a) shows the highest Tg due to the rigid pyromellitic moi-
eties. In contrast, (5b) displays the lowest Tg because of the
flexible ether linkage. Figure 7 shows the TMA curves of
PEIs (5), and the TMA data are listed in Table 2. Tg’s meas-
ured by TMA are in the range of 290–327 ꢀC, which is
slightly lower than those measured by DMA. The CTEs
before Tg are in the range of 44–52 ppm/ꢀC, displaying mod-
erate dimensional stability. Figure 8 shows the structure-Tg
Polyimide Synthesis
PEIs (5) were prepared by the reaction of (3) with various
commercially available dianhydrides (a–d) in DMAc to form
poly(amic acid)s (4), followed by thermal imidization
(Scheme 4). According to the principle of NMR, a smaller
chemical shift of an amino proton means a higher electron
density of nitrogen, corresponding to a higher reactivity. The
chemical shift of both amino groups of (3) is at 5.02 ppm.
This suggests that the reactivity of the amine (next to C1) is
not affected by the electron-withdrawing P¼¼O group. Al-
though the single crystal data of (3) is not successfully pre-
pared, the configuration of (3) might be approximately de-
picted by the single crystal data of (2), as shown in Figure 5.
The bulky phosphinate pendant is oriented to the 4-(4-ami-
nophenoxy)phenyl linkage. Therefore, little steric hindrance
on the left 4-aminophenyl linkage is expected. As a result,
PEIs with moderate-to-high inherent viscosity, ranging from
0.51 to 0.73 dL/g (Table 1), can be prepared. Supporting In-
formation Figure S10 shows the IR spectra of (5), the char-
acteristic imide absorption bands at 1773 cmꢁ1 (C¼¼O asym-
metric stretch), 1740 cmꢁ1 (C¼¼O symmetric stretch), 1370
cmꢁ1 (CAN stretch), 1116 and 717 cmꢁ1 (imide ring defor-
mation) indicate the formation of the imide ring. Further, the
absorptions of PAPh absorption at 1594 cmꢁ1, P¼¼O absorp-
tion at 1205 cmꢁ1, and PAOAPh absorption at 922 cmꢁ1
indicate no cleavage of the biphenylene phosphinate pendant
after thermal imidization.
Properties of the PEIs
Organosolubility
The solubility properties of PEIs (5) are listed in Table 1.
When compared with the symmetric PEIs (50) with two flexi-
ble ether linkages, asymmetric PEIs (5), even with only one
ether linkage, display much better solubility. For example,
PEI (5c) is soluble in NMP, DMAc, DMF, and DMSO, but PEI
(5c0) is soluble only in NMP. The asymmetric structure,
FIGURE 6 DMA curves of PEIs (5).
SOLUBLE HIGH-Tg PEIs WITH GOOD FLAME RETARDANCY, LIN, CHANG, AND CHENG
1337

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
TABLE 2 Thermal Properties of PEIs (5)
Td 5% (ꢀC)e
Char Yieldf
Film
E0
(GPa)a
Tan d
(ꢀC)
Tg (ꢀC)
(DSC)b
Tg (ꢀC)
(TMA)c
CTE
(ppm/ꢀC)d
Sample
Quality
N2
Air
N2
Air
5a
5b
5c
5d
a
Foldable
Foldable
Foldable
Foldable
2.51
4.58
4.60
4.57
336
291
300
311
329
281
290
296
327
277
290
299
44
52
49
44
452
450
447
463
447
455
444
463
56
56
55
59
49
39
47
50
d
Measured by DMA at a heating rate of 5 ꢀC/min; storage modulus (E0)
Coefficient of thermal expansion are recorded from 50 ꢀC to Tg ꢁ20 ꢀC.
e
are recorded at 50 ꢀC.
Temperature corresponding to 5% weight loss by thermogravimetry at
b
From the second DSC heating scan at a heating rate of 20 ꢀC/min.
Measured by TMA at a heating rate of 5 ꢀC/min.
a heating rate of 20 ꢀC/min.
c
f
Residual wt % at 800 ꢀC.
relationship of PEIs (5), PEIs (50), and PIs (500). Obviously, Tg
is strongly related with the number of phenoxy linkages. For
example, Tg reduced from 392 ꢀC for (5a00) to 336 ꢀC for
(5a), and to 314 ꢀC for (5a0). PEIs (50), with two phenoxy
linkages, show the lowest Tg. In contrast, PIs (500), with no
phenoxy linage, show the highest Tg. A 20–30 ꢀC difference
in Tg for a phenoxy linkage was observed for (b–d) series.
Figure 9 shows the structure-CTE relationship of three series
of phosphinated PIs. PEIs (50) show the highest CTE, and PIs
(500) show the lowest CTE. As expected, the trend of CTE is
inversely related with that of Tg.
The reason for the relatively high thermal stability of PEIs
(5) may be attributed to fact that the O¼¼PAO group is pro-
tected by the two phenylene groups, and therefore, the cyclic
O¼¼PAO chain was more thermally stable than the open
O¼¼PAO chain. In addition, on heating, the decomposition of
the weak PAO bond decreases the molecular weight rapidly
for main-chain type phosphinates. In contrast, the cleavage
of the phosphorus pendant has little effect on the integrity
of the polymer main chains, so they are thermally more
The thermal stability of PEIs (5) was evaluated by TGA (Ta-
ble 2). The 5 wt % degradation temperatures of PEIs (5) in
a nitrogen atmosphere range from 453 to 467 ꢀC, and the
char yields are in the range of 56–58 wt %. In an air atmos-
phere, the 5 wt % (Td 5%) degradation temperatures range
from 447 to 463 ꢀC, and the char yields are in the range of
39–50 wt %. The degradation temperature is relatively high
compared with those of main-chain type phosphinates.26,27
FIGURE 8 Structure-Tg relationship of PEIs (5), PEIs (50), and
PIs (500). DMA data (top) and TMA data (bottom).
FIGURE 7 TMA curves PEIs (5).
1338
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA

ARTICLE
UL-94
TABLE 3 Data of UL-94 Test for PEIs (5) Series and Kapton
Film
First
Second
Burning
Sample Thickness Burning
Code
(lm)
Time (s) Time (s) Dripping Grade
5a
31
33
31
35
23
0.6
1.2
0.9
0.7
4.6
0.1
0.3
0.3
0.2
1.8
No
No
No
No
No
V-0
V-0
V-0
V-0
V-0
5b
5c
5d
Kapton
bon adjacent to the phosphorus is critical to the thermal sta-
bility of the resulting PIs. It is speculated that PEIs (50), with
two electron-donating ether groups, increased the electron
density of the carbon next to phosphorus, so PEIs (50) dis-
play the highest Td. In contrast, PIs (500), with no electron-
donating ether group to increase the electron density of the
carbon, show the lowest Td. As to the flame retardancy,
according to the UL-94VTM vertical thin test (Table 3), the t1
þ t2 is less than 2 s for PEIs (5), so they belong to the VTM-
0 grade. In contrast, the t1 þ t2 is higher than 6 s for Kap-
ton. Although wholly aromatic PIs usually possess VTM-0
grade without halogen and phosphorus components, they
are usually organo-insoluble. The combination of organo-sol-
ubility and flame retardancy makes PEIs (5) attractive for
industrial application.
FIGURE 9 Structure-CTE relationship of PEIs (5), PEIs (50), and
PI (500).
stable than the main-chain type phosphinates. In addition,
according to bonding energy, the PAC chain has a lower one,
so the initial degradation of polymer occurred at that chain.
In our previous studies,43,44 we reported that the thermal
stability of phosphorus-containing thermosets is related to
the electron density of the carbon adjacent to the phospho-
rus. The lower is the electron density of the carbon, the
lower is its thermal stability. In this case, the high thermal
stability of PEIs (5) can also be explained by the electron
density of the aliphatic carbon adjacent to the phosphorus.
The P¼¼O group, an electron withdrawing group, reduced the
electron density of the carbon, destabilizing the PAC bond.
However, the methyl group, an electron donating group,
increased the electron density of the carbon, increasing the
thermal stability of the PAC bond.
CONCLUSIONS
We report a facile procedure to prepare a phosphinated ami-
nophenol (1). A ‘‘nucleophilic addition and then electrophilic
substitution’’ mechanism was proposed for the preparation
of (1). Then, an asymmetric etherdiamine (3) was then pre-
pared from a K2CO3-catalyzed nucleophilic aromatic substitu-
tion of (1) and 4-chloronitrobenzene, followed by catalytic
hydrogenation. PEIs (5) based on (3) are light colored,
tough, and foldable. In the structure-Tg, and structure-CTE
relationships, PIs (500) show the highest Tg and lowest CTE,
whereas PEIs (500) display the lowest Tg and highest CTE.
The values of Tg and CTE are strongly related with the num-
ber of phenoxy linkage in the repeating unit polymer back-
bone. In the structure-Td relationships, PEIs (50) show the
highest thermal stability, whereas PIs (500) display the lowest
one. It is speculated that electron-donating ether groups
increased the electron density of the carbon next to phos-
phorus, and thus enhanced the PAC bond. As for the struc-
ture-solubility relationship, PEIs (5) show the best organo-
solubility. In particular, PEI (5b), derived from flexible ODPA,
is soluble in low-boiling solvents, such as CH2Cl2 and CHCl3.
The asymmetric structure, which leads to the formation of
configuration isomers of the repeat unit and results in the
disruption of the packing sequence, might be responsible for
the improved solubility. According to the UL-94VTM vertical
test, PEIs (5) belong to the VTM-0 grade. The combination of
organo-solubility and flame retardancy makes PEIs (5)
attractive for industrial application.
Figure 10 shows the structure-Td relationship of the three se-
ries of phosphinated PIs. PEIs (50) show the highest Td, and
PIs (500) show the lowest Td. This trend is inversely related
to the order in Tg. Generally, polymers with high Tg usually
exhibit high thermal stability, so the trend is unexpected.
According to the previous discussion, the density of the car-
FIGURE 10 Structure-Td relationship of PEIs (5), PEIs (50), and
PI (500).
SOLUBLE HIGH-Tg PEIs WITH GOOD FLAME RETARDANCY, LIN, CHANG, AND CHENG
1339

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
22 Cheng, S. Z.; Chen, L.; Wang, Y. Z. J Polym Sci Part A:
REFERENCES AND NOTES
Polym Chem 2008, 46, 5752–5759.
1 Yang, C. P.; Su, Y. Y. J Polym Sci Part A: Polym Chem 2006,
23 Canadell, J.; Manteco´ n, A.; Ca´diz, V. J Polym Sci Part A:
44, 3092–3102.
Polym Chem 2007, 45, 1980–1992.
2 Li, H. S.; Liu, J. G.; Rui, J. M.; Fan, L.; Yang, S. Y. J Polym
24 Canadell, J.; Manteco´ n, A.; Ca´diz, V. J Polym Sci Part A:
Sci Part A: Polym Chem 2006, 44, 2665–2674.
Polym Chem 2007, 45, 1920–1930.
3 Mathews, A. S.; Kim, D.; Kim, Y.; Kim, I.; Ha, C. S. J Polym
25 Banerjee, S.; Palit, S. K.; Maiti, S. J Polym Sci Part A: Polym
Sci Part A: Polym Chem 2008, 46, 8117–8130.
Chem 1994, 32, 219–227.
4 Liu, Y.; Xing, Y.; Zhang, Y.; Guan, S.; Zhang, H.; Wang, Y.;
Wang, Y.; Jiang, Z. J Polym Sci Part A: Polym Chem 2010, 48,
3281–3289.
26 Liu, Y. L.; Hsiu, G. H.; Lee, R. H.; Chiu, Y. S. J Appl Polym
Sci 1997, 63, 895–901.
5 Sakayori, K.; Shibasaki, Y.; Ueda, M. J Polym Sci Part A:
27 Liou, G. S. Hsiao, S. H. J Polym Sci Part A: Polym Chem
Polym Chem 2006, 44, 6385–6393.
2001, 39, 1786–1799.
6 Chung, C. L.; Lee, W. F.; Lin, C. H.; Hsiao, S. H. J Polym Sci
28 Liu, Y. L.; Hsu, C. Y.; Wu, C. S. J Appl Polym Sci 2003, 89,
Part A: Polym Chem 2009, 47, 1756–1770.
791–796.
7 Chern, Y. T.; Tsai, J. Y.; Wang, J. J. J Polym Sci Part A:
29 Lin, C. H.; Lin, C. H. J Polym Sci Part A: Polym Chem 2007,
Polym Chem 2009, 47, 2443–2452.
45, 2897–2912.
8 Wang, S.; Zhou, H.; Dang, G.; Chen, C. J Polym Sci Part A:
30 Lin, C. H.; Wang, C. S. Polymer 2001, 42, 1869–1878.
Polym Chem 2009, 47, 2024–2031.
31 Liu, Y. L.
J Polym Sci Part A: Polym Chem 2002, 40,
9 Li, W.; Li, S.; Zhang, Q.; Zhang, S. Macromolecules 2008, 40,
359–368.
8205–8211.
32 Liu, Y. L.; Tsai, S. H. Polymer 2002, 43, 5757–5762.
10 Wang, C. Y.; Li, G.; Zhao, X. Y.; Jiang, J. M. J Polym Sci
33 Shieh, J. Y.; Wang, C. S. J Polym Sci Part A: Polym Chem
Part A: Polym Chem 2009, 47, 3309–3317.
2002, 40, 369–378.
11 Kulkarni, M.; Potrekar, R.; Kulkarni, R. A.; Vernekar, S. P. J
34 Lin, C. H.; Chang, S. L.; Cheng, P. W., under revision.
Polym Sci Part A: Polym Chem 2008, 46, 5776–5793.
35 Kim, W. G.; Hay, A. S. Macromolecules 1993, 26,
12 Chern, Y. T.; Tsai, J. Y. Macromolecules 2008, 41, 9556–9564.
5275–5280.
´
13 Calle, M.; Lozano, A. E.; de La Campa, J. G.; de Abajo, J.
36 Chung, I. S.; Kim, S. Y. Macromolecules 2000, 33,
Macromolecules 2010, 43, 2268–2275.
3190–3193.
14 Hsiao, S. H.; Guo, W. J.; Chung, C. L.; Chen, W. T. Eur
37 Choi, H.; Chung, I. S.; Hong, K.; Park, C. E.; Kim, S. Y. Poly-
Polym J 2010, 46, 1878–1890.
mer 2008, 49, 2644–2649.
15 Yang, C. P.; Su, Y. Y. J Polym Sci Part A: Polym Chem 2006,
38 Hsiao, S. H.; Lin, K. H. J Polym Sci Part A: Polym Chem
44, 3140–3152.
2005, 43, 331–341.
16 Liaw, D. J.; Chang, F. C. J Polym Sci Part A: Polym Chem
39 Chern, Y. T.; Ju, M. H. Macromolecules 2009, 42, 169–179.
2004, 42, 5766–5774.
40 Wang, C. Y.; Li, G.; Jiang, J. M. Polymer 2009, 50,
17 Watson, K. A.; Palmieri, F. L.; Connell, J. W. Macromole-
1709–1716.
cules 2002, 35, 4968–4974.
41 Chang, C. W.; Lin, C. H.; Cheng, P. W.; Hwang, H. J.; Dai, S.
18 Kwak, S. M.; Yeon, J. H.; Yoon, T. H. J Polym Sci Part A:
A. J Polym Sci Part A: Polym Chem 2009, 47, 2486–2499.
Polym Chem 2006, 44, 2567–2578.
42 Lin, C. H.; Chang, S. L.; Wei, T. P.; Ding, S. H.; Su, W. C.
19 Liu, Y. L.; Hsiue, G. H.; Lan, C. W.; Kuo, J. K.; Jeng, R. J.;
Polym Degrad Stab 2010, 95, 1167–1176.
Chiu, Y. S. J Appl Polym Sci 1997, 63, 875–882.
43 Wang, C. S.; Lin, C. H.
J Appl Polym Sci 2000, 78,
20 Jeong, K. U.; Kim, J. J.; Yoon, T. H. Polymer 2001, 42, 6019–6030.
228–235.
21 Sponto´ n, M.; Ronda, J. C.; Galia´, M.; Ca´diz, V. J Polym Sci
44 Wang, C. S.; Lin, C. H. J Appl Polym Sci 2000, 75, 429–436.
Part A: Polym Chem 2007, 45, 2142–2151.
1340
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA