Fluorinated phenylethynyl-terminated imide oligomers with reduced melt viscosity and enhanced melt stability

Article abstract of DOI:10.1016/j.polymer.2010.11.007

Two novel fluorinated phenyethynyl-contained endcapping agents, 4-(3-trifluoromethyl-1-phenylethynyl)phthalic anhydride (3F-PEPA) and 4-(3,5-bistrifluoromethyl-1-phenylethynyl)phthalic anhydride (6F-PEPA) were synthesized, which were employed to synthesiz

Full text of DOI:10.1016/j.polymer.2010.11.007

Polymer 52 (2011) 138e148
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Fluorinated phenylethynyl-terminated imide oligomers with reduced melt
viscosity and enhanced melt stability
,
,
,
a,b,
Yang Yang a b, Lin Fan a, Ximing Qu a b, Mian Ji a , Shiyong Yang
*
*
a Laboratory of Advanced Polymer Materials, Institute of Chemistry, Chinese Academy of Sciences, Zhongguancun, Beijing 100190, PR China
b Graduate School of Chinese Academy of Sciences, Beijing 100049, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Two novel fluorinated phenyethynyl-contained endcapping agents, 4-(3-trifluoromethyl-1-phenyl-
ethynyl)phthalic anhydride (3F-PEPA) and 4-(3,5-bistrifluoromethyl-1-phenylethynyl)phthalic anhydride
(6F-PEPA) were synthesized, which were employed to synthesize two fluorinated model compounds, N-
phenyl-4(3-trifluoromethyl)-phenylethynylphthalimide (3F-M) and N-phenyl-4(3,5-bitrifluoromethyl)-
phenylethynyl phthalimide (6F-M). The thermal cure kinetics of 3F-M and 6F-M were analyzed using
DSC and compared to the unfluorinated derivative, N-phenyl-4-phenylethynylphthalimide (PEPA-M).
The thermal cure temperatures of 3F-M and 6F-M were 399 and 412 ꢀC, which were 22 and 35 ꢀC higher
than that of PEPA-M, respectively. The thermal cure kinetics of 3F-M and 6F-M best fit a first-order rate
law, although 3F-M and 6F-M reacted slower than PEPA-M. However, the exothermic enthalpy of 3F-M
and 6F-M were only half of PEPA-M. Based on the model compounds study, a series of fluorinated
phenylethynyl-terminated imide oligomers (F-PETIs) with different calculated molecular weights (Calc’d
Mn) were synthesized by thermal polycondensation of 2,3,30,40-biphenyltetracarboxylic acid dianhydride
(a-BPDA) and 3,40-oxydianiline (3,40-ODA) using 3F-PEPA or 6F-PEPA as the endcapping agent. The
substituent effects of the trifluoromethyl (ꢁCF3) groups on the thermal cure behavior and melt proc-
essability of F-PETIs were systematically investigated. Experimental results reveal that the melt proc-
essability of F-PETI was apparently improved by the reduced resin melt viscosities and the enhanced
melt stability due to the incorporation of the ꢁCF3 groups in the imide backbone. All of those F-PETIs
exhibit outstanding thermal and mechanical properties.
Received 25 March 2010
Received in revised form
20 October 2010
Accepted 7 November 2010
Available online 13 November 2010
Keywords:
Imide oligomer
Melt processability
Thermal-cured polyimides
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
The polyimide composites derived from PETI could be produced
either by autoclave [11e27] or by resin transfer molding (RTM)
In recent years, significant efforts have been attempted in
improving the melt processability of aromatic polyimides [1e8].
Introduction of phenylethynyl-endcapping groups into the low-
molecular-weight imide oligomers was confirmed to be an effective
approach [9,10]. Phenylethynyl-terminated imide oligomers
(PETIs), due to its good melt processability and the great combi-
nation of thermal and mechanical properties etc., have been
extensively employed as the matrices of carbon fiber-reinforced
polyimide composites for high temperature applications [11e33].
Compared with nadic- or maleic-endcapped resins, phenylethynyl-
endcapped imides could be thermally cured by much preferred
chain extension than crosslinking, affording thermoset resins with
better impact strength [34e36].
[28e33], resulting in composite materials with high strength and
modulus as well as excellent thermal stability. PETI-5, the repre-
sentative PETI with Calc’d Mn of 5000 g/mol could be melt pro-
cessed by autoclave at 360e370 ꢀC [16,18]. By reducing Calc’d Mn of
PETI-5 to 1000 g/mol, imide oligomers with much lower melt
viscosity and suitable for RTM have been successfully developed.
The representative RTM imide resins (such as PETI-298, PETI-330
and PETI-375) could be melt at 260e280 ꢀC and completed the
thermal curing at 360e370 ꢀC [29,30,32,33]. In the case of RTM
application, the melt processability of PETI was endorsed by its low
molecular weights as well as its flexible chemical backbone.
However, the very low molecular weights of PETI usually result in
the thermal-cured materials with poor impact toughness and the
flexible oligomer backbone leads to the low glass transition
temperatures (Tg) for the thermal-cured resins. Hence, to improve
the RTM processability of PETI by reducing the melt viscosity and
enhancing the melt stability is still of great concern to high
temperature polyimide composites. Moreover, many other efforts
* Corresponding authors. Laboratory of Advanced Polymer Materials, Institute of
Chemistry, Chinese Academy of Sciences, Zhongguancun, Beijing 100190, PR China.
Tel.: þ86 10 6256 4819; fax: þ86 10 6256 9562.
E-mail address: shiyang@iccas.ac.cn (S. Yang).
0032-3861/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2010.11.007

Y. Yang et al. / Polymer 52 (2011) 138e148
139
have been made to improve the RTM melt processability of PETI
such as (1) by introducing comonomers for copolycondensation;
(2) by mixing of low-viscous reactive diluents, or (3) by introducing
asymmetric aromatic structures into the resin chemical backbone
etc. [29e31,37e39].
and used as received. All other reagents were commercially
obtained from China National Medicines Corp., China and used as
received.
2.2. Measurements
Meanwhile, it was also found that by replacing the phenyl-
ethynyl-endcapping groups in the imide oligomers with naphthyl or
anthracenyl-endcapping groups could decrease the cure tempera-
ture by 30e80 ꢀC and accelerate the thermal cure reaction [40e43]. It
was reported that the imide oligomers endcapped with 4-(2-phe-
nylethynyl)-1,8-naphthalic anhydride (PENA), or 4-(1-naphethleth-
ynyl)-1,8-naphthalic anhydride (NENA) cure at faster rate at a
temperature nearly 30 ꢀC lower [44], attributed by the electron-
withdrawing effects of the multi-aromatic structures on the electron
density and distribution of the ethynyl group in the imide oligomers.
Researchers also noted that imide oligomers derived from 4-phe-
nylethynyl phthalic anhydride (4-PEPA) with different substituted
groups and 3-phenylethynyl phthalic anhydride (3-PEPA) exhibit
different thermal curing behaviors [21,45].
Usually the endcapping agent could affect the properties of the
imide oligomers and their final cured polyimides profoundly for the
extra-high density of endcapping groups in imide oligomers.
Although many researchers have designed novel endcapping agents
to develop new imide resins, few studies on the imide oligomers
derived from fluorinated phenyethynyl-contained endcapping
agents could be found in the literature. In efforts to enhance the melt
processability (including the melt viscosity and melt stability) of
imide oligomers with high molecular weights, it is reasoned that
introduction of the bulky trifluoromethyl in phenylethynyl-end-
capping group to lower the interchain interaction or reduce the
stiffness of polymer backbone, this should make it possible to
maintain higher-molecular weights for oligomers to enhance the
mechanical/thermal properties for their corresponding thermal-
cured resins. Meanwhile fluoro-containing substituents with out-
standing thermal stability could also affect the properties of the
cured polyimide.
Infrared spectra (IR) were measured on a PerkineElmer 782
Fourier transform infrared (FTIR) spectrometer. 1H and 13C NMR
spectra were obtained on a Bruker AVANCE 400 spectrometer at
frequencies of 400 MHz using chloroform-d or DMSO-d6 as
solvents. Differential scanning calorimetry (DSC) analysis was
performed on a TA-Q100 analyzer at a heating rate of 10 ꢀC/min in
nitrogen atmosphere. Thermogravimetric analysis (TGA) was
accomplished on a TA-Q50 series thermal analysis system at
a heating rate of 20 ꢀC/min in nitrogen atmosphere. Complex
viscosity was performed on a TA-AR 2000 rheometer using the test
specimen disks with diameter of 25 mm and thickness of 1.2 mm,
which were prepared by compression molding of the imide olig-
omer powders at room temperature. The test disks were loaded in
the rheometer fixture equipped with 25 mm diameter parallel
plates, in which the top plate was oscillated at a fixed strain of 5%
and a fixed angular frequency of 10 rad/s whereas the lower plate
was attached to a transducer for detecting the resultant torque. For
the tests of dynamic thermal melt complex viscosity, the test
specimens were scanned from 200 ꢀC at a heating rate of 4 ꢀC/min,
in which the melt viscosity (h*, complex viscosity as a function of
time) was determined. Gel permeation chromatography (GPC) was
performed on Waters system that was equipped with a model 1515
pump, a 2414 refractive index detector using NMP as the eluant at
a flowing rate of 1.0 ml/min. The sample concentration was
1 mmol/ml and monodisperse polystyrene was employed as the
standard sample. Matrix-assisted laser desorption/ionization time-
of-flight (MALDI-TOF) mass spectra were carried out with a Biflex
IIIMALDI-TOF mass spectrometer (Bruker, Billerica, Germany)
equipped with delayed extraction, a multisample probe, a time-of-
flight reflection analyzer, a nitrogen laser with a wavelength of
337 nm and a pulse width of 3 ns, and a linear flight path length of
100 cm, in which the flight tube was evacuated to 1027 Pa.
Mechanical properties were obtained with an Instron model 3365
universal tester at room temperature. The tensile strength,
modulus, and the elongation at breakage were measured in
accordance with GB/T 16421-1996 at a strain rate of 2 mm/min.
In the present research, two novel fluorinated phenyethynyl-
contained endcapping agents were synthesized. The work reported
herein was performed primarily to determine whether substituting
fluorinated groups (e.g. ꢁCF3) in the phenylethynyl group of PEPA
offered any distinct advantages in the physical and/or thermal cure
properties and also to determine the melt processability, the
thermal curing and mechanical properties of the fluorinated PETI
and their thermal-cured polyimides.
2.3. Synthesis of 3-trifluoromethyl-1-ethynylbenzene (3F-EB)
2. Experimental
1-bromo-3-(trifluoromethyl)benzene (22.50 g, 0.1 mol), 2-me-
thyl-3-butyn-2-ol (10.09 g, 0.12 mol), cuprous iodide (0.95 g,
5 mmol), triphenylphosphine (0.65 g, 2.5 mmol), bis (triphenyl-
phosphine)palladium dichloride (1.40 g, 2.0 mmol) and triethyl-
amine (200 ml) were added into a 500 ml three-necked flask
equipped with a water condenser, a mechanical stirrer and nitrogen
inlet/outlet. The reaction mixture was heated to reflux for 8 h, and
cooled to room temperature. The catalyst and produced salt were
separated by filtration to give a filtrate, which was then performed
on a rotary evaporator to remove the solvent and yield a white
solid. The crude product was washed twice with hot water and
transferred into a 250 ml flask equipped with DeaneStark trap
containing 100 ml cyclohexane and sodium hydroxide (0.50 g,
12.5 mmol). The resulting solution was heated to reflux for 2 h and
then hot filtrated. After cooled to ambient temperature, the filtrate
was vacuum distillated (5 mmHg) to give a colorless liquid with
boiling point of 87e88 ꢀC (11.74 g, yield: 69.0%). 1H NMR (chloro-
2.1. Materials
2,30,3,40-biphenyltetracarboxylic acid dianhydride (a-BPDA, mp:
196e197 ꢀC) and 4-phenylethynyl phthalic anhydride (4-PEPA, mp:
151e152 ꢀC) were synthesized in this laboratory according to the
reported methods [46,47]. 3,40-oxydianiline (3,40-ODA, mp:
84e85 ꢀC) was purchased from Shanghai Baicheng Chemicals Corp.,
China and recrystallized from ethanol/water (1:1) prior to use.
4-bromophthalic anhydride was purchased from Beijing Bomi
Chemicals Corp., China and recrystallized from acetic acid/acetic
anhydride prior to use. N-methyl-2-pyrrolidinone (NMP) was
purchased from Beijing Beihua Fine Chemicals, China and purified
by vacuum distillation over P2O5. Triethylamine was used after
vacuum distillation in the presence of calcium hydride. 1-bromo-
3-(trifluoromethyl)benzene and 1-bromo-3,5-bis(trifluoromethyl)
benzene were purchased from Liaoning Fuxin Sanbao Chemicals
Corp., China and used as received. 2-methyl-3-butyn-2ol was
purchased from Beijing Beihua Hengyuan Chemicals Corp., China
form-d, d ppm): 7.65 (s, 1H), 7.55e7.50 (t, 2H), 7.41e7.36 (t,1H), 2.57
(s, 1H). Anal. calcd for C9H5F3: C, 63.54%; H, 2.96%. Found: C, 63.57%;
H, 2.93%.

140
Y. Yang et al. / Polymer 52 (2011) 138e148
2.4. Synthesis of 3,5-bitrifluoromethyl-1-ethynylbenzene (6F-EB)
the following procedure: 4-(3-trifluoromethyl-1-phenylethynyl)
phthalic anhydride (31.60 g, 0.1 mol), aniline (9.31 g, 0.1 mol) and
150 ml DMAc were placed in a 500 ml flask equipped with a water
condenser and a mechanical stirrer. The reaction solution was
stirred at room temperature for 4 h. The solution was then chem-
ically cyclodehydrated with acetic anhydride and pyridine (2:1 v/v)
after stirring for 6 h. The resultant mixture was then poured into
300 ml of ethanol to yield precipitate. The solid was collected by
filtration, and then dried at 100 ꢀC in vacuum for 5 h to afford an
off-white solid (36.50 g, yield: 93.3%). mp: 240e241 ꢀC 1H NMR
This intermediate compound was obtained by the similar method
as described above except that 1-bromo-3-(trifluoromethyl)benzene
was replaced by 1-bromo-3,5-(bitrifluoromethyl)benzene. The
product was colorless liquid with boiling point of 61e62 ꢀC (15.24 g,
yield: 64.0%).1H NMR (chloroform-d,
d ppm): 7.91 (s, 2H), 7.84 (s,1H),
3.25 (s, 1H). Anal. calcd for C10H4F6: C, 50.44%; H, 1.69%. Found: C,
50.47%; H, 1.72%.
2.5. Synthesis of 4-(3-trifluoromethyl-1-phenylethynyl)phthalic
(chloroform-d, d ppm): 8.10 (s, 1H), 7.97e7.91 (m, 2H), 7.84 (s, 1H),
anhydride (3F-PEPA)
7.75e7.74 (d, 1H), 7.66e7.65 (d, 1H), 7.55e7.50 (m, 3H), 7.46e7.40
(m, 3H). FTIR (KBr cmꢁ1): 2205 (ꢁC^Cꢁ), 1843 (asym C]O str),
1811 (sym C]O str), 1383 (imide CꢁN str), 1141 (CꢁF). Anal. calcd
for C23H12F3NO2: C, 70.59%; H, 3.09%, N 3.58%. Found: C, 70.42%; H,
3.15%; N, 3.44%.
The model compound 6F-M was prepared according to above
method except that 3F-PEPA was replaced with 6F-PEPA to afford
off-white solid (43.51 g, yield: 94.7%). mp: 240e241 ꢀC by DSC. 1H
A mixture of 4-bromophthalic anhydride (22.70 g, 0.10 mol), 3F-
EB (18.70 g, 0.11 mol) and triethylamine (250 ml) were added into
a 500 ml flask equipped with a mechanical stirrer, a water
condenser and a nitrogen inlet. The reaction mixture was stirred at
room temperature until a homogeneous solution was obtained.
Cuprous iodide (0.95 g, 5.0 mmol), triphenylphosphine (0.66 g,
2.5 mmol), bis (triphenylphosphine)palladium dichloride (1.40 g,
2.0 mmol) were then added. After refluxed under nitrogen atmo-
sphere for 6 h, the reaction mixture was cooled to room tempera-
ture and then filtrated to remove the catalyst and salt generated in
the reaction. The filtrate was poured into 800 ml of water, and then
acidified to pH ¼ 2 with concentrated hydrochloric acid (35e38%).
The precipitate was collected and vacuum dried at 120 ꢀC for 5 h to
give a yellow powder. The obtained solid was dissolved in 200 ml of
toluene in a 500 ml flask equipped with mechanical stirrer, a water
condenser, a DeaneStark trap and a nitrogen inlet. The solution was
heated to reflux for 8 h under nitrogen atmosphere, in which the
water generated in the reaction was simultaneously removed by
azeotropic distillation. And the solution was then concentrated by
partly removal of toluene. The concentrated solution was cooled to
room temperature and yielded yellow solid. After filtrated and
rinsed with acetic acid, the crude product was purified by recrys-
tallization in acetic acid/acetic anhydride (2:1 v/v) to afford 23.10 g
of yellow crystal (yield: 73.0%). mp: 162e163 ꢀC [determined by
NMR (chloroform-d, d ppm): 8.12 (s, 1H), 8.01 (s, 2H), 8.00e7.93 (m,
2H), 7.89 (s, 1H), 7.54e7.50 (m, 2H), 7.45e7.41 (m, 3H). FTIR (KBr
cmꢁ1): 2117 (ꢁC^Cꢁ), 1853 (asym C]O str), 1820 (sym C]O str),
1392 (imide CꢁN str), 1147 (CꢁF). Anal. calcd for C24H11F6NO2: C,
62.75%; H, 2.41%; N 3.05%. Found: C, 62.77%; H, 2.42%; N, 3.05%.
The model compound PEPA-M as a comparative was prepared
according to the method as described above except that 3F-PEPA
was replaced with 4-PEPA to afford off-white solid (30.98 g, yield:
95.8%). mp: 203e204 ꢀC. 1H NMR (chloroform-d,
d ppm): 8.12 (s,
1H), 8.01 (s, 2H), 8.00e7.93 (m, 2H), 7.89 (s, 1H), 7.54e7.50 (m, 2H),
7.45e7.41 (m, 3H). FTIR (KBr cmꢁ1): 2117 (ꢁC^Cꢁ), 1853 (asym C]
O str), 1820 (sym C]O str), 1392 (imide CꢁN str), 1147 (CꢁF). Anal.
calcd for C24H11F6NO2: C, 81.72%; H, 4.05%; N 4.33%. Found: C,
81.84%; H, 4.13%; N, 4.21%.
2.8. Synthesis of the fluorinated imide oligomers (F-PETIs)
F-PETIs were prepared by the conventional two-stage polymer-
ization and imidization process. In a typical experiment, 3,40-ODA
(10.01g, 0.050 mol) and NMP (60 ml) were placed in a flask equipped
with a water condenser, a mechanical stirrer, a DeaneStark trap,
a thermometer and a nitrogen inlet/outlet. After stirred for 30 min at
room temperature to give a homogeneous solution, a slurry of a-
BPDA (7.35 g, 0.025 mol), 3F-PEPA (15.81 g, 0.050 mol) and NMP
(51 ml) was added. The reaction mixture, with 30% of solids (w/w),
was stirred at 80 ꢀC for 8 h and 150 ml toluene was then added. The
resultant solution was heated to reflux at 185 ꢀC for 10 h, in which
the water evolved in the thermal imidization was simultaneously
removed from the reaction system by azeotropic distillation. After
cooled down to about 120 ꢀC, the hot solution was poured slowly
into 1000 ml of water. The solid was collected by filtration and
washed with hot water three times, and then dried at 205 ꢀC in
vacuum for 10 h to give O1-1 as yellow powder (30.28 g, 96.6%).
Other F-PETIs including O1-2, O2-1, O2-2, O10-1, O10-2 as well
as the unfluorinated derivatives including O1-comp, O2-comp and
DSC]. 1H NMR (chloroform-d,
(s, 1H), 7.76e7.74 (d, 1H), 7.69e7.67 (d, 1H), 7.57e7.53 (t, 1H). 13C
NMR (chloroform-d, ppm): 162.9, 139.3, 136.0, 132.7, 131.4, 131.3,
131.1, 131.0, 129.1, 129.0, 128.5, 126.7, 126.6, 125.8, 123.6, 123.1121.3,
93.2, and 89.0. FTIR (KBr cmꢁ1): 2207 (ꢁC^Cꢁ), 1840 (asym C]O
str), 1782 (sym C]O str), 1146 (CꢁF). Anal. calcd for C18H6F6O3: C,
64.57%; H, 2.23%. Found: C, 64.36%; H, 2.08%.
d ppm): 8.14 (s, 1H), 8.01 (s, 2H), 7.84
d
2.6. Synthesis of 4-(3,5-bistrifluoromethyl-1-phenylethynyl)
phthalic anhydride (6F-PEPA)
6F-PEPA (yellow crystal, 26.88 g, yield: 70.0%) was obtained by
the similar method as described above except that 3F-EB was
replaced by 6F-EB. mp: 196e197 ꢀC 1H NMR (DMSO-d6,
d
ppm):
8.41 (s, 1H), 8.35 (s, 1H), 8.25 (s, 1H), 8.17 (s, 2H), 3.32 (s, 1H). 13C
NMR (chloroform-d, ppm): 162.9, 139.5, 133.0, 132.8, 132.7, 132.3,
d
132.0, 131.9, 130.4, 128.8, 127.8, 126.4, 125.3, 125.1, 123.5, 122.4,
119.5, 93.2, and 89.0. FTIR (KBr cmꢁ1): 2187 (ꢁC^Cꢁ), 1837 (asym
C]O str), 1795 (sym C]O str), 1136 (CꢁF). Anal. calcd for
C17H7F3O3: C, 56.27%; H, 1.57%. Found: C, 56.10%; H, 1.65%.
Table 1
Thermal data and kinetic parameters of the model compounds.
Sample Endothermal Exothermal
temperature temperature
D
H
E
n
A
2.7. Synthesis of model compounds
(J/g) (kJ/mol)
(ꢀC)
(ꢀC)
Model compounds including N-phenyl-4-(3-trifluoromethyl-
1-phenylethynyl)phthalimide (3F-M), N-phenyl-4-(3,5-bitrifluoro-
methyl-1-phenylethynyl)phthalimide (6F-M) and N-phenyl-4-(1-
phenylethynyl)phthalimide (PEPA-M) were prepared by chemical
cyclodehydration. In a typical experiment, 3F-M was prepared in
Peak
Onset Peak End
PEPA-M 202
320
343
350
377 434 516 152.9
399 432 270 168.9
0.94 8.1 ꢂ 1011
1.05 5.0 ꢂ 1012
1.09 8.2 ꢂ 1011
3F-M
6F-M
240
241
412
e
218 187.4

Y. Yang et al. / Polymer 52 (2011) 138e148
141
Table 2
Chemical compositions and molecular weights of F-PETIs and the unfluorinated
derivatives.
Sample
Chemical
Calc’d Calc’d
Pn
Mn
Mw
Mw/Mn
compositiona
Mn (g/mol)c (g/mol) (g/mol)
b
O1ꢁ1
O1e2
3,40-ODA/a-BPDA/
3F-PEPA ¼ 2/1/2
3,40-ODA/a-BPDA/
6F-PEPA ¼ 2/1/2
1
1254
1392
1118
1712
1850
1576
5376
5514
5238
2555
2854
2253
3410
3521
3248
4957 1.94
1
1
2
2
2
4881 1.71
5691 2.52
5564 1.63
8373 2.37
4661 1.49
O1ecomp 3,40-ODA/a-BPDA/
4-PEPA ¼ 2/1/2
O2e1
3,40-ODA/a-BPDA/
3F-PEPA ¼ 3/2/2
3,40-ODA/a-BPDA/
6F-PEPA ¼ 3/2/2
O2e2
O2-comp 3,40-ODA/a-BPDA/
4-PEPA ¼ 3/2/2
O10e1
3,40-ODA/a-BPDA/ 10
3F-PEPA ¼ 11/10/2
8197 16,172 1.97
10,826 20,943 1.93
10,027 22,947 2.29
O10e2
3,40-ODA/a-BPDA/ 10
6F-PEPA ¼ 11/10/2
Fig. 1. DSC curves of the model compounds at 20 ꢀC/min.
O10-comp 3,40-ODA/a-BPDA/ 10
4-PEPA ¼ 11/10/2
a
Molar ratios of the reactants.
Calculated polymerization degree.
Calculated molecular weight.
69.0% and 64.0%, respectively. The first step was the coupling reac-
tion of 2-methyl-3-butyn-2-ol with 1-bromo-3-(trifluoro methyl)
benzene or 1-bromo-3,5-(bitrifluoromethyl)benzene in the pres-
ence of palladium catalyst to yield the intermediates 1-(2-methyl-
3-butyn-2-ol)-3-(trifluoromethyl)benzene or 1-(2-methyl-3 -but-
yn-2-ol)-3,5-(bistrifluoromethyl)benzene, which was then treated
with NaOH to give the intermediate compounds. 3F-PEPA and 6F-
PEPA were prepared via the palladium-catalyzed coupling reaction
of 4-bromophthalic anhydride with 3F-EB or 6F-EB as showed in
Scheme 1. The resultant products were recrystallized in acetic acid/
acetic anhydride (2:1 v/v) in the total yield of 73.0% and 70.0%,
respectively. The molecular structures of 3F-PEPA and 6F-PEPA were
confirmed by means of FTIR,1H and 13C NMR and elemental analysis.
b
c
O10-comp were prepared according to the similar procedure, in
which the molar ratios of 3,40-ODA, a-BPDA and 3F-PEPA, 6F-PEPA,
or 4-PEPA were listed in Table 2.
2.9. Preparation of the thermal-cured polyimides
The imide oligomer resin powders were placed in a die, which
was then placed in a hot press preheated at 200 ꢀC. The die
temperature was increased gradually to 370 ꢀC at a rate of 4 ꢀC/min.
After it was kept there for 10e25 min, the die was applied with
a pressure of 1.0e3.5 MPa. After it was kept for 1h at 370 ꢀC, the die
was then cooled with the applied pressure to less than 200 ꢀC. The
thermally cured polyimide sheet was removed from the die at room
temperature and then cut to the desired sizes for thermal and
mechanical testing.
3.2. Thermal curing properties of model compounds
Prior to imide oligomers and the thermal-cured polyimide
studies, model compounds (3F-M, 6F-M and PEPA-M) were
synthesized by the chemical cyclodehydration in good yield and
characterized. The thermal behaviors of the model compounds
were investigated by DSC to determine the effect of the molecular
structures on their thermal curing behaviors. Fig. 1 shows the DSC
curves of the model compounds (3F-M and 6F-M) compared with
PEPA-M and Table 1 lists the DSC data. The sharp endothermic
peaks in the range of 202e241 ꢀC were assigned as the melt points
3. Results and discussion
3.1. Monomers
3F-EB and 6F-EB were successfully synthesized via a two-step
procedure of coupling and cleavage reaction in the total yield of
Scheme 1. Synthesis of the ꢁCF3 substituted phenylethynyl phthalic anhydrides.

142
Y. Yang et al. / Polymer 52 (2011) 138e148
Fig. 2. Curing kinetic analysis and calculation of the thermal curing activation energy
of the model compounds.
Fig. 3. Representative GPC curves of F-PETIs with different Calc’d Mn.
(Tms) and the wide exothermic peaks in the range of 377e412 ꢀC
were attributed to the thermal curing reactions of the phenyl-
ethynyl groups in the model compounds.
DSC method was employed to study the curing kinetics of the model
compounds and the resultant data are also tabulated in Table 1. In
the experiments, temperatures of the model compounds (3F-M, 6F-
M and PEPA-M) were scanned by DSC at different heating rates (2, 5,
10, 15 and 20 ꢀC/min, respectively). The exothermic peak tempera-
tures (Tp) in DSC curves were shifted into higher temperatures with
The ꢁCF3 groups had pronounced impacts on the Tms values and
the thermal curing reaction of the phenylethynyl groups. The
fluorinated model compounds (3F-M and 6F-M) showed the
identical melting points (240 and 241 ꢀC), 38e39 ꢀC higher than
PEPA-M (202 ꢀC). 3F-M and 6F-M had the onset temperatures (343
and 350 ꢀC, respectively), 23e30 ꢀC higher than PEPA-M (320 ꢀC).
Meanwhile, the exothermal peak temperatures of 3F-M and 6F-M
were 399 ꢀC and 412 ꢀC, respectively, 22e35 ꢀC higher than PEPA-M
(377 ꢀC), indicating that the fluorinated phenylethynyl groups
would undergo thermal curing at higher temperatures than the
unfluorinated analogue. In comparing the reaction exothermal
increasing of the heating rate (
fact that Tp varies with and assumes that the maximum reaction
rate (d /dt) occurs at peak temperatures [48]. Thus, the equation can
be expressed as follows (Equation (1)):
b). Kissinger’s method is based on the
b
a
ꢀ
ꢁ
ꢁln
b
=Tp2 ¼ lnðE=RÞ ꢁ lnðAnÞ ꢁ ðn ꢁ 1Þlnð1 ꢁ xÞpþE=RTp
(1)
data, 3F-M (
exothermic enthalpies than PEPA-M (
the fluorinated phenyethynyl group would undergo the thermal
curing with much lower exotherm.
D
H ¼ 270 J/g) and 6F-M ( H ¼ 218 J/g) had much lower
D
D
H ¼ 516 J/g), implying that
where E is the activation energy; R is the gas constant; A, n and x are
the pre-exponential factor, the order of the cure action, and the
extent of the cure action, respectively. Fig. 2 shows the scatter graphs
of ꢁInð
b
=Tp2Þ against 1/Tp. The calculated values of Es for PEPA-M,
In order to extensively understand the thermal curing reaction of
the phenylethynyl-endcapped imide oligomers, the nonisothermal
3F-M and 6F-M were 152.9, 168.9, and 187.4 kJ/mol, respectively.
Scheme 2. Synthesis of F-PETIs and the thermal-cured polyimides.

Y. Yang et al. / Polymer 52 (2011) 138e148
143
Fig. 4. MOLDI-TOF mass spectra of F-PETIs with different Calc’d Mn.
À
Á
Obviously, 3F-M and 6F-M, the fluorinated phenylethynyl-endcap-
ped imide compounds, have higher Es than PEPA-M. Meanwhile, 6F-
M, with two ꢁCF3 groups in the phenylethynyl group, has higher E
value (187.4 kJ/mol) than 3F-M (168.9 kJ/mol). This could be inter-
preted by the stronger electron-withdrawing effect and steric
hindrance of the ꢁCF3 in the phenylethynyl group. If the E value was
introduced into the Crane’s equation (2):
dðln
b
Þ=d 1=Tp z ꢁ E=nR
(2)
The scatter plot of ln(
b) as a function of 1/Tp can be obtained as
shown in Fig. 2, in which the n values were obtained as 0.94 for
PEPA-M, 1.05 for 3F-M, and 1.09 for 6F-M, respectively, demon-
strating that the thermal curing reaction of the model compounds
fit the first-order kinetics reaction rate law, similar to other electron-

144
Y. Yang et al. / Polymer 52 (2011) 138e148
Table 3
Molecular structures of the chemical species detected by MOLDI-TOF mass spectra for F-PETIs with different Calc’d Mn
Chemical structures of the resin species
n
Mass (m/z)
Intensity %
0
1
2
3
4
5
819.7
1277.4
1735.4
2193.4
2651.5
3110.7
1.5
100.0
84.7
62.2
25.5
6.1
O1-1
0
1
2
3
4
5
819.4
1277.4
1735.4
2193.4
2652.4
3110.6
3.6
87.8
100.0
92.9
29.6
2.0
F C
3
+
Na
O
O
O
O
N
O
N
N
O
N
CF
3
O
O
O
O1-2
O
n
double-endcapping species
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
1277.4
1735.4
2194.4
2652.4
3110.5
3569.0
4029.1
4487.6
4947.2
1021.5
1479.5
1937.4
2396.4
2854.5
3312.7
3771.1
16.3
69.4
100.0
84.7
54.1
33.7
20.4
8.2
F C
3
+
Na
O
N
O
O
O
O
O
N
N
O
N
CF
3
O
O
O
n
double-endcapping species
4.1
O1-10
32.7
62.2
86.7
61.2
31.6
15.3
7.1
O
Na +
O
O
N
O
N
+NaF
O
H2N
N
CF3
O
O
O
n'
mono-endcapping species
withdrawing arylethynyl-endcapped imide oligomers [40e43].
Moreover, the pre-exponential factors (A) in the range of 8.2 ꢂ 1011
to 5.0 ꢂ 1012 for the fluorinated model compounds could also be
obtained according to the Kissinger’s equation as shown in Table 1.
Besides the electronic effects, the trifluoromethyl group might
have some steric hindrance to retard the thermal curing reaction of
the phenylethynyl group, which resulted in the elevated thermal
curing temperatures.
3.3. Molecular structures of F-PETIs
A series of F-PETIs with various Calc’d Pn (calculated polymeri-
zation degree) and different endcapping agents including 3F-PEPA,
6F-PEPA or 4-PEPA were prepared by the synthetic pathway as
shown in Scheme 2 and their chemical compositions are summa-
rized in Table 2.
The stoichiometric compositions of the monomers including
aromatic dianhydride (i.e. a
-BPDA) and aromatic diamine (i.e. 3,40-
ODA) and the endcapping agents were changed systematically in
order to explore the influence of the oligomer molecular weights
Table 4
Thermal curing behaviors of F-PETIs with different Calc’d Mn determined by DSC.
Sample
Tg (ꢀC)
Tonset (ꢀC)
Texo (ꢀC)
DH (J/g)
O1-1
O1-2
O1-comp
O2-1
O2-2
O2-comp
O10-1
O10-2
O10-comp
136
143
137
165
170
168
229
217
218
312
325
293
323
334
315
388
398
379
406
421
397
413
424
392
435
441
422
192.6
174.7
240.1
139.1
109.6
178.5
7.81
5.67
9.64
Fig. 5. DSC curves of the representative F-PETIs with different Calc’d Pn.

Y. Yang et al. / Polymer 52 (2011) 138e148
145
and O2-1 had the major peaks in the relative sharp ranges, implying
that O10-1 was mainly consisted of the high-molecular-weight
resin species. In quantities, O10-1 has four peaks with different
relative intensities: i.e. 34% at 13.1 min, 100% at 15.1 min, 17.5% at
17.5 min, and 10% at 18.3 min respectively, compared with O2-1,
which also has four peaks in the elution time range of
12.7e19.1 min: 12% at 13.5 min, 99% at 16.8 min, 100% at 17.5 min
and 37% at 18.3 min respectively. Obviously, the major peak
intensity was intensified in the higher-molecular weight oligomer.
For the oligomers with higher Calc’d Mn (O10-1), the major peak at
15.1 min (100%) was new resin species which was not detected in
the lower Calc’d Mn oligomers (O1-1 and O2-1), implying that this
new species was coupled by the different lower-molecular molec-
ular resin species. It is obviously that, the molecular weights and
molecular weight distributions of F-PETIs were closely related to
their designed molecular weights.
Fig. 4 shows the MALDI-TOF mass spectra of the representative
imide oligomers with different Calc’d Pn (O1-1, O2-1 and O10-1) and
Table 3 summarizes the molecular structures of the resin species
detected. The imide oligomers are resin mixtures with different
molecular weights. O1-1 and O2-1 showed the same chemical
species with identical Pn (0e5) and molecular weights (m/z of
796.7e3087.7), but different 100% peak intensities (O1-1: Pn ¼ 1, O2-
1: Pn ¼ 2). For instance, O1-1, which has the lowest Calc’d Pn (¼1),
contains 6 resin species with molecular weights (Mn ¼ 796.7 to
3087.7) and Pn (0e5): 796.7 (Pn ¼ 0, m/z: 819.7e23.0 ¼ 796.7, 1.5%),
1254.4 (Pn ¼ 1, m/z: 1277.4e23.0 ¼ 1254.4,100%),1712.4 (Pn ¼ 2, m/z:
1735.4e23.0 ¼ 1712.4, 84.7%), 2170.4 (Pn ¼ 3, m/z: 2193.4e23.0
¼ 2170.4, 62.2%), 2628.5(Pn ¼ 4, m/z: 2651.5e23.0 ¼ 2628.5, 25.5%)
and 3087.7 (Pn ¼ 5, m/z: 3110.7e23.0 ¼ 3087.7, 6.1%), respectively, in
which the species with Pn ¼ 1 showed the 100% peak intensity. In
comparison, O10-1, which has the highest Calc’d Pn (¼10), was
composed with as many as 16 chemical species with molecular
weights of 965.5e4924.2, in which the 100% peak intensity was
detected in the species with Pn ¼ 3 (m/z ¼ 2194.4).
Fig. 6. Dynamic rheological behaviors of representative F-PETIs with different Calc’d
Mn.
and ꢁCF3 groups on the imide oligomer properties. The Calc’d Pn of
the imide oligomers were designed to be 1, 2 and 10, respectively,
by controlling the molar ratios of a-BPDA to 3,40-ODA along with
the endcapping agents including 3F-PEPA, 6F-PEPA or 4-PEPA
(Table 2), resulting in a series of F-PETIs with different Calc’d Mn in
the range of 1254e5514 g/mol.
Table 2 also shows the molecular weights of the synthesized
imide oligomers determined by GPC using NMP as solvent,
including the number average molecular weight (Mn), weight
average molecular weight (Mw), and the molecular weight distri-
bution (Mw/Mn). The measured molecular weights are much higher
than the calculated ones due to the polarity difference of the testing
samples compared to the GPC standard sample (polystyrene) [30].
For instance, O1-1 has the measured Mn of 2555 g/mol, almost
doubled the Calc’d Mn (1254). The molecular weight distributions
of F-PETIs were in the range of 1.63e2.37.
Fig. 3 depicts the representative GPC curves of F-PETIs with
different Calc’d Pn (1, 2 and 10), in which it can be seen that each
imide oligomer was composed of several chemical species with
different molecular weights in a broad elution time range. The
major molecular weight peaks of F-PETIs were increased with
increasing of the Calc’d Mn values. For instance, O10-1 (Calc’d
Mn ¼ 5376) showed a major elution time peak at 15.1 min,
compared with 16.8 min for O2-1 (Calc’d Mn ¼ 1712) and 18.2 min
for O1-1 (Calc’d Mn ¼ 1254), respectively. Moreover, the relative
peak intensities in GPC curves were also obviously changed with
Calc’d Mn. The higher the Calc’d Mn, the more intensified the peak
intensity. For instance, O10-1 showed an intense and broad peak
located at 18.8e11.6 min along with several trace peaks, while O1-1
Apparently, the major chemical species were polycondensed by
the aromatic dianhydride and aromatic diamine along with the
endcapping agents. It was also found that double-endcapped
chemical species were primarily produced for the lower Calc’d Mn
oligomers such as O1-1 and O2-1, and double- and mono-end-
capped species were simultaneously formed in the case of the
higher Calc’d Mn oligomer (O10-1). In comparing of the effect of
Calc’d Pn on the measured molecular weights, it can be concluded
that the higher-molecular-weight chemical species were produced
with increasing of Calc’d Pn.
3.4. Thermal curing of F-PETIs
Fig. 5 compares the DSC curves of F-PETIs with different Calc’d
Mn (O1-1, O2-1 and O10-1) and Table 4 summarizes the DSC data of
Table 5
Complex melt viscosity of F-PETIs and the unfluorinated derivatives.
Sample
Complex melt viscosity (Pa s) at different temperature (ꢀC)
Minimum melt
viscosity (Pa s/ꢀC)
Melt viscosity
variation at
280 ꢀC for 2 h
Melt viscosity
variation at
310 ꢀC for 2 h
250 ꢀC
280 ꢀC
310 ꢀC
340 ꢀC
370 ꢀC
O1e1
O1e2
O1ecomp
O2e1
O2e2
O2ecomp
O10e1
O10e2
O10ecomp
15.2
83.2
1.9
40.5
25.8
52.1
6952
5498
6803
0.4
0.9
0.9
4.5
4.8
0.2
0.5
0.8
1.6
2.4
0.1
0.4
0.8
0.8
2.2
0.2
0.7
10.7
3.6
31.9
32.3
65.2
21.6
83.8
0.1 at 341 ꢀC
0.4 at 338 ꢀC
0.8 at 320 ꢀC
0.8 at 347 ꢀC
2.3 at 342 ꢀC
2.9 at 324 ꢀC
64.8 at 369 ꢀC
22.9 at 367 ꢀC
41.7 at 355 ꢀC
0.46e0.51
0.65e0.93
1.27e2.73
3.47e7.25
4.95e6.44
6.19e7.37
e
0.29e85.7
0.50e9.75
0.88e452
1.71e134.1
1.51e87.14
2.93e176.5
e
6.2
3.0
3.1
1438
1213
4684
368.7
254.6
1204
120.2
55.94
75.6
e
e
e
e

146
Y. Yang et al. / Polymer 52 (2011) 138e148
the imide oligomers. The Tgs of the imide oligomers determined by
DSC were in the range of 136 ꢀCe229 ꢀC. The higher Calc’d Mn
oligomers showed higher Tg values. The strong curing exothermal
peaks were observed for the lower Calc’d Mn oligomers (O1-1 and
O2-1), in which the curing onset temperatures (Tonset) were
measured at 312e323 ꢀC, and the exothermic peaks (Texo) at
406e413 ꢀC, respectively. The normalized heat enthalpy (
DH) was
calculated at 139.1e192.6 J/g. However, the higher Calc’d Pn olig-
omer (O10-1) did not show apparently thermal curing exothermal
peak, probably due to the low endcapping group density in the long
imide chains.
The exothermic peaks of the thermal curing reactions for
F-PETIs appeared at 406e441 ꢀC, approximately 10e23 ꢀC higher
than the corresponding unfluorinated ones (392e422 ꢀC). This was
well accordance with the model compounds, implying that the
bulky ꢁCF3 substituent leaded to steric barrier and resulted in the
phenylethynyl group being much difficulty in thermal curing. In
Fig. 7. Isothermal melt viscosity at 280 ꢀC of the representative F-PETIs and the
unfluorinated derivative.
addition, the exothermic enthalpy (
D
H) decreased with increasing
H of 7.81 J/g, only 5.6% of
of the Calc’d Pn. For instance, O10-1 has
D
O2-1 (139.1 J/g) and 4.1% of O1-1 (192.6), respectively. This could be
interpreted by the contribution of the low endcapping group
density in the higher-molecular weight oligomers.
one (O1-1) as well as the unfluorinated phenylethynyl-endcapped
one (O1-comp). The isothermal standing time (Fig. 8) at which the
oligomer melt viscosity abruptly increased was 116 min for O1-2,
compared to 89 min for O1-1 and 70 min for O1-comp, indicating
that the fluorinated phenylethynyl groups showed obvious stabi-
lizing impacts on the oligomer melt viscosity, which is good
beneficial to the oligomer melt processability. Overall, the fluori-
nated phenylethynyl groups would not only contribute to reduce
the imide oligomer melt viscosity, but also be beneficial in
improving the stability of melt viscosity. These are due to the
introduction of the bulky trifluoromethyl in phenylethynyl end-
capping groups reduce the interchain interaction and the stiffness
of polymer backbone.
3.5. Melt processability of F-PETIs
The melt processability of F-PETIs was investigated by the
dynamic rheological method. Fig. 6 depicts the typical melt visco-
sityetemperature curves of the imide oligomers, in which it can be
seen that the melt viscosities were decreased gradually with
increasing of the scanned temperatures started at 200 ꢀC, then
down to a valley stage where the melt viscosity reached the
minimum point, and then went up with further increasing of the
temperature due to the thermal curing of the fluorinated phenyl-
ethynyl groups. Compared with the higher Calc’d Mn oligomers
(O2-1 and O10-1), O1-1 showed much broad low viscosity range
started from 260 ꢀC to 370 ꢀC, indicating that this oligomer has the
widest melt processing window.
3.6. Thermal properties of the thermal-cured polyimides
The complex melt viscosities at different scanned temperatures
and the minimum melt viscosities were summarized in Table 5.
Obviously, minimum melt viscosity temperatures were increased
with increasing of the molecular weights. For instance, the
minimum melt viscosity temperatures were increased from 341 ꢀC
(0.1 Pa s) for O1-1e347 ꢀC (0.8 Pa s) for O2-1 and 369 ꢀC (64.8 Pa s)
for O10-1. The minimum melt viscosities were also increased
gradually with increasing of the molecular weights. Hence,
reducing the resin molecular weights could result in sharp decrease
in the minimum melt viscosity and widen the processing temper-
ature window. In comparison, O2-1 (Calc’d Pn ¼ 2) has the melt
viscosity of 0.8 Pa s at 340 ꢀC, the same with O1-comp (Calc’d
Pn ¼ 1), indicating that the higher-molecular-weight F-PETI could
give the same low melt viscosity as the low-molecular-weight
unfluorinated derivative.
F-PETIs could be thermally cured at 370 ꢀC for 1 h into the
thermoset polyimides. Table 6 summarizes the thermal properties
of all cured polyimides determined by DSC and TGA. The Tgs of the
thermal-cured polyimides are closely related to the oligomer
molecular weights. The lower the un-cured oligomer molecular
weights, the higher the Tgs of the thermal-cured polyimides. For
instance, O1-1, with the lowest molecular weight, has the highest Tg
Figs. 7 and 8 show the melt viscosity stabilities of the imide
oligomers against the isothermal standing time at 280 ꢀC and
310 ꢀC, respectively. The melt viscosity variations were depended
on the resin molecular weights as well as their chemical structures.
O1-1 showed the lowest melt viscosity variation of 0.46e0.51 Pa s
after isothermal standing at 280 ꢀC for 2 h (Fig. 7), compared with
O1-2 (0.65e0.93 Pa s) and O1-comp (1.27e2.73 Pa s), implying that
the melt stability of the imide oligomers were closely related to
their molecular weights as well as the chemical structures derived
from the different endcapping agents. At higher isothermal
standing temperature (310 ꢀC), the bitrifluoromethyl phenyl-
ethynyl-endcapped oligomer (O1-2) showed the best melt viscosity
stability than the monotrifluoromethyl phenylethynyl-endcapped
Fig. 8. Isothermal melt viscosity at 310 ꢀC of the representative F-PETIs and the
unfluorinated derivative.

Y. Yang et al. / Polymer 52 (2011) 138e148
147
Table 6
Thermal properties of the thermal-cured polyimides cured at 370 ꢀC for 1 h
Table 7
Mechanical properties of the thermal-cured polyimides at 370 ꢀC/1 h.
Sample
Tg (ꢀC)
Tonset (ꢀC)
T5 (ꢀC)
T10 (ꢀC)
Sample
Tensile
Tensile
Elongation
strength (MPa)
modulus (GPa)
at break (%)
O1-1
O1-2
O1-comp
O2-1
O2-2
O2-comp
O10-1
O10-2
O10-comp
332
321
329
306
300
301
269
261
263
580
583
546
583
570
552
578
577
550
559
559
555
561
546
560
573
572
560
598
598
578
603
590
581
601
598
575
O1ꢁ1
41.7
39.0
47.5
112.2
101.8
121.5
1.2
1.4
1.3
1.1
1.2
1.3
6.2
5.8
6.0
18.6
21.2
23.5
O1e2
O1ecomp
O10e1
O10e2
O10ecomp
acceleration trend in the thermal curing reaction than O1-1. And
the monotrifluoromethyl phenylethynyl-endcapped oligomer (O1-
1) expressed similar thermal curing behavior with the unfluori-
nated derivative (O1-comp).
of 332 ꢀC, 26 ꢀC higher than O2-1 (306 ꢀC), and 63 ꢀC higher than
O10-1 (269 ꢀC). Additionally, the monotrifluoromethylphenyl-
ethynylendcapped oligomers (O1-1, O2-1 and O10-1) showed
a
little higher Tgs than the bitrifluoromethylphenyl-ethyny-
3.7. Mechanical properties of the thermal-cured polyimides
lendcapped ones (O1-2, O2-2 and O10-2). For instance, the
thermal-cured O2-1 has Tg of 306 ꢀC, 6 ꢀC higher than the thermal-
cured O2-2 (300 ꢀC).
Table 7 summarizes the mechanical properties of the thermal-
cured polyimides. After thermally cured at 370 ꢀC for 1 h, the
resulted thermoset polyimides with Calc’d Pn ¼ 1 (O1-1 and O1-2)
exhibited good combined mechanical properties, including tensile
strength of 39.0e41.7 MPa, tensile modulus of 1.2e1.4 GPa, elon-
gation at breakage of 5.8e6.2%. The tensile strength and elongation
at breakage of the thermal-cured polyimides were greatly
improved when the oligomer molecular weights were increased.
For instance, the thermal-cured polyimides with Calc’d Pn ¼ 10
(O10-1, O10-2) showed tensile strength of 101.8e112.2 MPa and
elongation at breakage of 18e21%, much better than O1-1 and O1-
2. In addition, F-PETIs could produce the thermal-cured polyimides
with mechanical properties comparable to the unfluorinated
derivatives, revealing that no obvious negative impacts on
mechanical properties of the thermal-cured polyimides were
observed due to the fluorination of the phenyethynyl groups.
The thermal-cured polyimides provided with the onset
decomposition temperatures (Tonset) in the range of 570e583 ꢀC, 5%
loss temperatures (T5) of 546e573 ꢀC and 10% loss temperatures
(T10) of 578e603 ꢀC (Table 6). No obvious difference in thermal
stability in nitrogen was observed for the thermal-cured poly-
imides. In addition, the thermal-cured fluorinated polyimides
exhibited little better thermal stability than the unfluorinated ones.
For instance, the thermal-cured O1-1 resin has Tonset of 580 ꢀC
(34 ꢀC higher than O1-comp (546 ꢀC)), and T10 of 598 ꢀC (20 ꢀC
higher than O1-comp (578 ꢀC).
The thermal curing kinetic analysis of the imide oligomers (O1-
1, O1-2 and O1-comp) were performed using DSC by monitoring of
the changes in Tg values as a function of the thermal curing
temperatures (360 ꢀC/1 h, 370 ꢀC/1 h, 380 ꢀC/1 h and 390 ꢀC/1 h,
respectively). The curing degree (x) was calculated by DiBenedetto
equation modified for highly crosslinked networks according to the
literatures [36,40e43,49]. Applying for the first-order data analysis
(ꢁlnð1 ꢁ xÞ ¼ Kt), the rate constant (K) at certain temperature was
obtained from the very good line fit. As shown in Fig. 9 in Arrhenius
relationship K ¼ AeꢁEa/RT between rate constant and temperature,
an activation energy (Ea) of 160.0 kJ/mol for O1-1 and 210.8 kJ/mol
for O1-2 were obtained, compared to 141.5 kJ/mol for O1-comp. Due
to the higher activation energy, O1-2 would have the higher rate
4. Conclusion
Novel fluorinated phenylethynyl-contained endcapping agents,
4-(3-trifluoromethyl-1-phenylethynyl)phthalic anhydride (3F-PEPA)
and 4-(3,5-bistrifluoromethyl-1-phenylethynyl)phthalic anhydride
(6F-PEPA) have been synthesized, which were employed to synthe-
size a series of F-PETIs. The experiments on the model compounds
indicated that the fluorinated model compounds exhibited 22e35 ꢀC
higherin thermalcure temperatures and 42e52% lowerin exotherms
than the unfluorinated derivative. F-PETIs also holded the thermal
curing temperatures of 20e30 ꢀC higher than the corresponding
unfluorinated analogues. The melt processability of F-PETIs was
apparently improved by the melt viscosities decreased and the melt
viscosity stability enhanced due to the ꢁCF3 groups in the imide
oligomers. F-PETIs could be thermally cured at 370 ꢀC to afford
the thermal-cured polyimides, which denoted thermal properties
with Tg of as high as 332 ꢀC, T5 of >570 ꢀC, and mechanical properties
with tensile strength of as high as 122 MPa, and elongation at
breakage of >18%.
References
[1] Serafini T, Delvig P, Lightsey G. J. Appl Polym Sci 1972;16:905e15.
[2] Landis AL, Bilow N, Boschan RH, Lawrence RE, Aponyi TJ. Polym Prepr
1974;15:537e41.
[3] Harris K, Sridhar K. Polym Prepr 1985;26:142e3.
[4] Chaudhari MA. 32nd Int SAMPE Symp 1987;32:597e601.
[5] Waters JF, Sutter JK, Meador MAB, Baldwin LJ, Meador MA. J Polym Sci Part A
Polym Chem 1991;29:1917e24.
[6] Baldwin LJ, Meador MAB, Meador MA. Polym Prepr 1988;29:263e4.
[7] Vaucraeynest W, Stille JK. Macromolecules 1980;13:1361e7.
Fig. 9. Arrhenius plots of the thermal curing kinetics of the representative F-PETIs and
the unfluorinated derivative.

148
[8] Bisshop KC. Chem Rev 1976;76:461e86.
[9] Connell JW, Smith Jr JG, Hergenrother PM. J Macromol Sci Rev Macromol
Chem Phys 2000;C40(2e3):207e30.
[10] Hergenrother PM. High Perform Polym 2003;15:3e45.
[11] Harris FW, Padaki SM, Vavaprath S. Polym Prepr 1980;21(1):3.
[12] Harris FW, Pamidimukkala A, Gupta R, Das S, Wu T, Mock G. Polym Prepr
1983;24(2):324e5.
Y. Yang et al. / Polymer 52 (2011) 138e148
[28] Smith Jr JG, Connell JW, Hergenrother PM. J Comp Matls 2000;34:614e28.
[29] Smith Jr JG, Connell JW, Hergenrother PM. J Comp Matls 2002;36:2255e65.
[30] Simone CD, Scola DA. High Perform Polym 2003;15:473e501.
[31] Zhou HW, Chen CH, Kanbara R, Sasaki T, Yokota R. High Perform Polym
2005;17:193e212.
[32] Connell JW, Smith Jr JG, Hergenrother PM. High Perform Polym 2003;15:
375e94.
[13] Harris FW, Sridhar K, Das S. Polym Prepr 1984;25(1):110e1.
[14] Harris FW, Pamidimukkala A, Gupta R, Das S, Wu T, Mock G. J Macromol Sci
Chem 1984;A21(8e9):1117e35.
[15] Meyer GW, Jayaraman S, McGrath JE. Polym Prepr 1993;34:540e1.
[16] Paul CW, Schultz R, Fennelli SP. In: Feger C, Khoyasteh MM, Htoo MS, editors.
Advances in polyimide science and technology. Lancaster, PA: Technomic;
1993. p. 220e354.
[33] Connell JW, Smith JG Jr., Hergenrother PM. Proceedings of 48th Int SAMPE
sym May; Long Beach CA (2003), CD Version.
[34] Cho D, Drzal LT. J Appl Polym Sci 2000;76:190e200.
[35] Fang X, Xie XQ, Simone CD, Stevens MP, Scola DA. Macromolecules
2000;33:1671e81.
[36] Fang X, Rogers DF, Scola DA, Stevens MP. J Polym Sci Part A Polym Chem
1998;36:461e70.
[17] Hergenrother PM, Bryant RG, Jensen BJ, Havens SJ. J Polym Sci Part A Polym
Chem 1994;32:3061e7.
[37] Knudsen RL, Jensen BJ. High Perform Polym 1996;8:57e66.
[38] Smith Jr JG, Connell JW. High Perform Polym 2000;12:213e23.
[39] Ding MX. Prog Polym Sci 2007;32:623e68.
[18] Meyer GW, Glass TE, Grubbs HJ, McGrath JE. Polym Prepr 1994;35:549e50.
[19] Takekoshi T, Terry JM. Polymer 1994;35:4874e80.
[20] Hergenrother PM, Smith Jr JG. Polymer 1994;35:4857e64.
[21] Johnston JA, Li FM, Harris FW. Polymer 1994;35:4865e73.
[22] Hinkley JA. J Adv Mater 1994;27:55e9.
[23] Preston CML, Jill DJT, Pomery PJ, Whittaker AK, Jensen BJ. High Perform Polym
1999;11:453e65.
[24] Connell JW, Smith Jr JG, Hergenrother PM. High Perform Polym 1998;
10:273e83.
[40] Wright ME, Schorzman DA. Macromolecules 1999;32:8693e4.
[41] Wright ME, Schorzman DA, Pence LE. Macromolecules 2000;33:8611e7.
[42] Wright ME, Schorzman DA. Macromolecules 2001;34:4768e73.
[43] Wright ME, Schorzman DA, Berman AM. Macromolecules 2002;35:6550e6.
[44] Thompson CM, Hergenrother PM. Macromolecules 2002;35:5835e9.
[45] Liu YF, Wang ZH, Yang HL. J Polym Sci Part A Polym Chem 2008;46:4227e35.
[46] Mengxian D, Wang Z, Yang Z, Zhang J. U.S. Patent, 5, 081, 281, 1992.
[47] Zhen W, Gao L, Mengxian Ding. Chinese Patent Z. L. 2003; 10115829.8.
[48] Kissinger HE. Anal Chem 1957;29:1702e6.
[25] Hou T, Jensen BJ, Hergenrother PM. J Comp Matls 1996;30(1):109e22.
[26] Hergenrother PM. Sampe J. 2000; 36(1): 30ꢁ41.
[27] Hergenrother PM, Conell JW, Smith Jr JG. Polymer 2000;41:5073e81.
[49] Ayambem SJ, Mecham SJ, Sun Y, Glass TE, McGrath JE. Polymer 2000;41:
5109e24.