A novel fluoro-terminated hyperbranched poly(phenylene oxide) (FHPPO): Synthesis, characterization, and application in low-k epoxy materials

Article abstract of DOI:10.1039/c3ra40721g

A new AB₂ monomer 4-hydroxyl-4′,4′′- difluorotriphenylmethane was successfully synthesized via a Friedel-Crafts alkylation of phenol from 4,4′-difluorodiphenylmethanol. Based on the AB₂ monomer, novel fluoro-terminated hyperbranched poly(phenylene oxide)s (FHPPOs) were synthesized via the S_NAr reaction by self-condensation in one step. The FHPPOs were characterized by various techniques, including NMR, FT-IR, GPC, TGA and DSC. It was found that the molecular weight and polydispersity index of the FHPPOs increased with monomer concentration and reaction time. The degree of branching of the FHPPOs, determined by ¹³C NMR and ¹⁹F NMR with the aid of model compounds, decreased from 0.63 to 0.53 as the molecular weight increased. The glass transition temperature (T_g) of the FHPPOs increased with increasing molecular weight, up to 164°C when the M_n was over 6, 800. The FHPPOs showed excellent thermal stability up to a T_d5 temperature of 559°C. Because of the low polarity of the poly(phenyl oxide) (PPO) backbones, abundant fluoro-terminated groups, which have large molar free volume, low polarizability of C-F bonds, and inherent free volume or molecule-scale cavities in hyperbranched structures, the addition of FHPPO into diglycidyl ether of bisphenol A (DGEBA) could effectively lower the relative dielectric constant, the dissipation factor, and moisture absorption of the cured DGEBA/FHPPO composites. The free volume of the composites, which was quantified by positron annihilation lifetime spectroscopy (PALS), increased with increased FHPPO loading. The excellent dielectric and thermal properties make FHPPO a promising low-k modifier for epoxy resins. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3ra40721g

RSC Advances
View Article Online
View Journal | View Issue
PAPER
A novel fluoro-terminated hyperbranched
poly(phenylene oxide) (FHPPO): synthesis,
characterization, and application in low-k epoxy
materials
Cite this: RSC Advances, 2013, 3,
14509
Lijuan Luo,ab Teng Qiu,a Yan Meng,*b Longhai Guo,a Jing Yang,c Zhuoxin Li,c
Xingzhong Caoc and Xiaoyu Li*a
A new AB2 monomer 4-hydroxyl-49,499-difluorotriphenylmethane was successfully synthesized via a
Friedel–Crafts alkylation of phenol from 4,49-difluorodiphenylmethanol. Based on the AB2 monomer,
novel fluoro-terminated hyperbranched poly(phenylene oxide)s (FHPPOs) were synthesized via the SNAr
reaction by self-condensation in one step. The FHPPOs were characterized by various techniques, including
NMR, FT-IR, GPC, TGA and DSC. It was found that the molecular weight and polydispersity index of the
FHPPOs increased with monomer concentration and reaction time. The degree of branching of the
FHPPOs, determined by 13C NMR and 19F NMR with the aid of model compounds, decreased from 0.63 to
0.53 as the molecular weight increased. The glass transition temperature (Tg) of the FHPPOs increased with
increasing molecular weight, up to 164 uC when the Mn was over 6, 800. The FHPPOs showed excellent
thermal stability up to a Td5 temperature of 559 uC. Because of the low polarity of the poly(phenyl oxide)
(PPO) backbones, abundant fluoro-terminated groups, which have large molar free volume, low
polarizability of C–F bonds, and inherent free volume or molecule-scale cavities in hyperbranched
structures, the addition of FHPPO into diglycidyl ether of bisphenol A (DGEBA) could effectively lower the
relative dielectric constant, the dissipation factor, and moisture absorption of the cured DGEBA/FHPPO
composites. The free volume of the composites, which was quantified by positron annihilation lifetime
spectroscopy (PALS), increased with increased FHPPO loading. The excellent dielectric and thermal
properties make FHPPO a promising low-k modifier for epoxy resins.
Received 10th February 2013,
Accepted 13th May 2013
DOI: 10.1039/c3ra40721g
www.rsc.org/advances
Unlike dendrimers, hyperbranched polymers can be
synthesized through one-pot or pseudo-one step procedures
Introduction
Since the pioneering work of Kim and Webster, hyper-
branched polymers have attracted lots of attention and are
now at the forefront of polymer research.1–5 Unlike linear
polymers, hyperbranched polymers (HBPs) are highly
branched three-dimensional macromolecules and have inher-
ent internal cavities and abundant terminal groups, which
often lead to better processability, compatibility and solubi-
lity.6 Hyperbranched polymers have been used as rheology
modifiers, processing aids, and more recently, as functional
materials in coatings, catalysts, sensors, bio-materials, emit-
ting materials, and in nano-templating applications.6–12
but still share most of the properties of dendrimers, which
makes them easier and cheaper for commercialization.4
Recently, several hyperbranched poly(arylene ether)s have
been reported, including poly(ether sulfone),13–16 poly(ether
ketone),17–21 and poly(arylene ether phosphine oxide)s).22–26
The monomers used to prepare these hyperbranched polymers
contain electron-withdrawing groups, i.e.,
aState Key Laboratory of Organic and Inorganic Composites, Beijing University of
Chemical Technology, Beijing 100029, P. R. China. E-mail: lixy@mail.buct.edu.cn;
Tel: +8610-64419631
bKey Laboratory of Carbon Fiber and Functional Polymers, Ministry of Education,
Beijing University of Chemical Technology, Beijing 100029, P. R. China.
E-mail: mengyan@mail.buct.edu.cn; Fax: +8610-64452129; Tel: +8610-64419631
cKey Laboratory of Nuclear Analysis Techniques, Institute of High Energy Physics,
Chinese Academy of Science, Beijing 100049, P. R. China
in ortho and para positions of the leaving groups, which can
serve as the activating groups and are favourable for aromatic
nucleophilic substitution (SNAr) reactions. However, only a
handful of research has reported on the synthesis of
hyperbranched poly(phenyl oxide) (HPPO) via SNAr reactions
because the monomers used for preparing HPPO lack the
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 14509–14520 | 14509

View Article Online
Paper
RSC Advances
necessary activating groups as mentioned above, which greatly
limit the applications of HPPOs.
(Tg), and thermal stability of HPPO were investigated. By virtue
of the low polarity of the basic units of PPO, abundant
fluorinated terminal groups, and the inherent free volume (or
cavities) in the hyperbranched structures, the reported
hyperbranched macromolecules could have potential applica-
tions in low-k materials.
As an important engineering plastic, poly(phenyl oxide)
(PPO) has outstanding thermal, mechanical and dielectric
properties.27,28 However, its poor moldability, poor solubility
and inadequate adhesive strength limits its applications.
Compared with linear polymers, hyperbranched polymers
have lower viscosity and better solubility due to their abundant
functional groups and semi-global shape.4 As
a result,
Experimental
hyperbranched poly(phenyl oxide) (HPPO) could combine the
benefits of PPO with the merits of hyperbranched polymers.
Therefore, it is important to design new HPPOs with specific
structures and terminal groups to expand their applications.
Kim et al.18 reported a bromo-terminated HPPO, which was
derived from 3,5-dibromophenol by the Ullmann reaction.
However, applications of this HPPO have not been reported
until now, which is probably because of the poor ability of the
chemical modifications and the low functionality of the
terminal aryl bromide groups in the HPPO. A series of
hyperbranched poly(arylene ether)s based on 3,5-bis-(4-
fluoro-3-trifluoro-methylphenyl) phenol using the AB + AB2
and A2 + AB2 approaches have been reported by Ghosh and
coworkers.29,30 Using the A2 + B3 approach, Banerjee31 also
reported a hyperbranched poly(arylene ether), which shared
similar chemical structure with those of Ghosh’s. They
successfully introduced the trifluoromethyl groups in the
ortho positions of the leaving groups, which notably improved
the reactivity of the monomers. Those hyperbranched poly-
mers with abundant fluorinated substituents had high glass
transition temperatures (Tgs), good thermal stabilitiy and
hydrophobicity. However, further applications for those
hyperbranched polymers have not yet been reported, probably
due to the poor processability and the high cost of the raw
materials.
In our previous studies,32,33 HPPO was prepared from
4-bromo-49,499-dihydroxyltriphenylmethane by the Ullmann
reaction using CuCl as the catalyst. The phenolic terminated
HPPOs can be easily modified and used in epoxy resins,34
bismaleimide resins,35 and cyanate ester resins.36 In another
work of ours,37 the Ullmann reaction was replaced by the SNAr
reaction, and HPPO was synthesized using 4-fluoro-49,499-dihy-
droxyl triphenyl methane as an AB2 monomer. Subsequently,
the synthesized HPPO was used as a modifier for epoxy resins
and both the mechanical and thermal properties of the
modified epoxy systems were improved. In those cases, the
terminal groups of HPPO were phenolic groups. It is believed
that the terminal groups of hyperbranched polymers can
greatly affect the macromolecular properties, e.g., the glass
transition temperature (Tg), the solubility, dielectric proper-
ties, hydrophobicity, and thermal stability. Thus, in this work,
the terminal groups of the HPPO macromolecule are changed
into fluorinated substituents and those FHPPOs are synthe-
sized using 4-hydroxyl-49,499-difluorotriphenylmethane as an
AB2 monomer through the SNAr reaction. The effects of
monomer concentration and reaction time on the molecular
weight, degree of branching (DB), glass transition temperature
Materials
Phenol (98%) was purchased from Tianjin Chemical Reagents
Factory, China. 4,49-difluorodiphenylmethanol (99%) was
purchased from Changzhou Rising Chemical Co. Ltd.,
China. Fluorobenzene (99%) and 2-ethyl-4-methylimidazole
(2E4MZ, 99%) were obtained from Acros Organics. Other
solvents were purchased from Beijing Chemical Works, China.
Diglycidyl ether of bisphenol A (DGEBA) was purchased from
Yueyang Resin Factory, China (epoxy equivalent weight, EEW =
186.2 g/equiv.). Methyl nadic anhydride (MNA) (99.8%) was
supplied by POLYNT Chemical Co. Ltd., Italy. All the reagents
and solvents were used as received without further purifica-
tion.
Synthesis of AB2 monomer: 4-hydroxyl-49,499-difluorotri-
phenylmethane (compound 2)
A solution of phenol (42.0 g, 0.447 mol) and H2SO4 (20 ml) in
CH2Cl2 (250 ml) was cooled to 5 uC, then a solution of 4,
49-difluorodiphenylmethanol (87.6 g, 0.4 mol) in CH2Cl2 (250
ml) was added dropwise with mechanical stirring. After
maintaining at 5 uC for 6 h, the organic phase of the mixture
was treated with a saturated aq. NaHCO3 solution and
extracted with ethylacetate. The organic layer was dried with
anhydrous MgSO4 and concentrated under vacuum, and the
obtained crude product was recrystallized by 2 : 1 hexane/
toluene. Finally, the obtained compound 2 was dried under
vacuum at 80 uC thoroughly to give a yellow semi-solid (65.1 g,
55%).
In order to determine all the products, 2 g crude product
was purified by silica gel column chromatography using 1 : 1
CH2Cl2/petroleum ether as the eluent. After drying under
vacuum at 80 uC, desired AB2 momoner (denoted as 2 in
Schemes 1 to 3) (1.1 g, 60%) and by-product (denoted as 29 in
Scheme 2) (0.7 g, 39%) were obtained.
Compound 2. Purity (UPLC): 96.5%; MS m/z 295.3 (M2); H
1
NMR, dH (600 MHz; CDCl3; Me4Si) 4.67 (1H, s, OH), 5.44 (1H, s,
CH), 6.77 (2H, d, C6H4OH), 6.93 (2H, d, C6H4OH), 6.97 (4H, t,
C6H4F), 7.02 (4H, m, C6H4F); 13C NMR, dC (600 MHz; acetone-
d6; Me4Si) 54.41, 115.07, 115.25, 130.37, 130.66, 135.91,
139.71, 154.08, 160.61.
Compound 29. Purity (UPLC): 96.4%; MS m/z 497.7 (M2); 1H
NMR, dH (600 MHz; CDCl3; Me4Si) 5.33 (1H, s, OH), 5.67 (2H, s,
CH), 6.40 (1H, d, C6H3O), 6.76 (1H, d, C6H3O), 6.83 (1H, q,
C6H3O), 6.93 (2H, d, C6H3OH), 6.93 (8H, m, C6H4F), 6.96 (8H,
m, C6H4F); 13C NMR, dC (600 MHz; acetone-d6; Me4Si) 49.02,
54.34, 115.12, 115.87, 128.60, 130.47, 130.52, 131.37, 136.00,
138.01, 139.62, 151.69, 162.20.
14510 | RSC Adv., 2013, 3, 14509–14520
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
Scheme 1 Synthesis route of fluoro-terminated hyperbranched poly(phenylene oxide) (FHPPO).
C6H4); 13C NMR, dC (600 MHz; CDCl3; Me4Si) 54.57, 115.11,
115.31, 130.48, 130.67, 138.37, 155.66, 160.67.
Synthesis of fluoro-terminated hyperbranched poly (phenylene
oxide) (FHPPO)
AB2 (35.5 g, 0.12 mol) monomer (compound 2) and potassium
carbonate (21.5 g, 0.156 mol) were added into a mixture of
N-methylpyrrolidone (NMP) (600 ml) and toluene (150 ml).
After refluxing for 3 h, the water produced in the reaction was
removed by a Dean–Stark trap. Subsequently, the temperature
was elevated to 202 uC. After 2 days, the mixture was cooled to
room temperature and poured into 1500 ml of 10% acetic acid
solution. The precipitate was collected and reprecipitated from
THF into methanol. The resultant off-white solids were dried
under a vacuum at 60 uC (28.4 g, 80%). 1H NMR, dH (600 MHz;
CDCl3; Me4Si) 5.44 (s, CH(C6H5)3), 6.89 (br s, C6H4), 7.01 (br s,
Synthesis of model compound 3 (MC3)
AB2 monomer (4.72 g, 16 mmol), fluorobenzene (16.00 g, 160
mmol), and potassium carbonate (4.42 g, 32 mmol) were
added into 140 ml of NMP. The mixture was refluxed under
nitrogen and stirred for 72 h. It was then cooled to room
temperature, filtered, and slowly precipitated in 10 wt% acetic
acid solution. The crude precipitate was purified by silica gel
column chromatography using 1 : 2 CH2Cl2/petroleum ether
as the eluent. After drying under a vacuum at 80 uC, a white
solid was obtained (2.5 g, 42%). Purity (UPLC): 96.6%; 1H
NMR, dH (600 MHz; CDCl3; Me4Si) 5.52 (1H, s, CH), 6.97 (2H,
Scheme 2 Synthetic mechanisms of AB2 monomer.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 14509–14520 | 14511

View Article Online
Paper
RSC Advances
Scheme 3 Synthesis route of model compounds.
d, C6H5O), 7.01 (4H, m, C6H4F), 7.04 (2H, d, C6H4O), 7.06 (2H,
d, C6H4O), 7.09 (4H, m, C6H4F), 7.12 (1H, t, C6H5O), 7.35(2H, t,
C6H5O); 13C NMR dC (600 MHz; acetone-d6; Me4Si) 54.6, 115.3,
118.7, 119.0, 123.4, 129.8, 130.5, 130.8, 138.3, 139.6, 155.9,
157.1, 160.7.
mixtures were degassed in a vacuum oven at 80 uC and then
cured in silicone rubber molds. The curing schemes are as
follows: 120 uC for 2 h, 160 uC for 2 h, and 200 uC for 2 h.
Characterizations
The 1H and 13C NMR spectra were collected using a Bruker
Fourier Transform AVANCE 600 spectrometer. The 19F NMR
spectra were collected using a Bruker Fourier Transform
AVANCE 400 spectrometer. CDCl3 were used in all NMR
measurements as solvents. Acquisition parameters were
optimized for quantitative 13C NMR measurements, including
a 4.2-microsecond 45u pulse, inverse gated proton decoupling
and a 5 s delay time between pulses. A total of 2500 scans were
used for data averaging. For quantitative 19F NMR, a 7.5-
microsecond 45u pulse, inverse gated proton decoupling, an 8
s delay time between pulses, and a total of 64 scans were used.
Gel permeation chromatography (GPC) diagrams were
obtained with a Waters GPC 515-2410 system equipped with
2410 refractive index detector and Styragel column
(HT3_HT5_HT6E) using tetrahydrofuran as the eluent at 30
uC. The GPC columns were calibrated with linear narrow
polystyrene standards of known molecular weights in the
Synthesis of model compounds 4 (MC4) and 5 (MC5)
MC3 (1.70 g, 4.57 mmol), phenol (0.64 g, 6.86 mmol), and
potassium carbonate (1.23 g, 8.91 mmol) were added into a
mixture of NMP (40 ml) and toluene (30 ml). The mixture was
refluxed for 3 h with water collected in a Dean–Stark trap.
Subsequently, the temperature of the mixture was kept at 202
uC for 24 h. At the end of the reaction, the mixture was cooled
to room temperature and poured into 10 wt% acetic acid
solution. The crude precipitate was purified by silica gel
column chromatography using 1 : 2 CH2Cl2/petroleum ether
as the eluent. The resultant product was dried under a vacuum
at 80 uC. Finally MC4 (0.9 g, 88%) and MC5 (0.82 g, 69%) were
obtained.
1
MC4. Purity (UPLC): 97.4%; H NMR, dH (600 MHz; CDCl3;
Me4Si) 5.50 (1H, s, CH), 6.95 (4H, d, C6H5O), 7.00 (2H, t,
C6H4F), 7.04 (4H, d, C6H4O), 7.06 (4H, d, C6H4O), 7.12 (2H, d,
C6H4F), 7.10 (2H, t, C6H5O), 7.34 (4H, t, C6H5O); 13C NMR, dC
(600 MHz; acetone-d6; Me4Si) 54.7, 115.1, 118.6, 118.9, 123.3,
129.7, 130.5, 130.7, 138.6, 140.0, 155.8, 157.1, 160.6.
range of 160 to 85 000 000. The molecular weights (in g mol21
of the samples are rounded to the nearest hundred.
)
The infrared spectra of samples in the form of KBr pellets
were recorded on a Bruker Tensor 37 spectrometer at room
temperature. Mass spectra were obtained from a Waters
Quattro Premier XE with methanol as the solvents.
Purity of all synthesized compounds was determined by a
Waters Acquity Ultra Performance LC equipped with a TUV
detector. The analytical column was a Waters BEH C18, 50 6
2.1 mm (1.7 mm). The mobile phase consisted of methanol and
ultra-pure water. A linear gradient was used to ramp the
mobile phase composition from 30% (v/v) water to 1% (v/v)
water in 6 min. The runtime was 8 min per sample. The flow
rate was 0.25 mL min21 and the column oven temperature was
set at 40 uC. UV absorption was monitored at 235 nm.
1
MC5. Purity (UPLC): 98.2%; H NMR, dH (600 MHz; CDCl3;
Me4Si) 5.50 (1H, s, CH), 6.95 (6H, d, C6H5O), 7.03 (6H, d,
C6H4O), 7.08 (6H, d, C6H4O), 7.10 (3H, br s, C6H5O), 7.33 (6H,
t, C6H5O); 13C NMR, dC (600 MHz, acetone-d6, Me4Si): 54.8,
118.6, 118.9, 123.2, 129.7, 130.5, 138.8, 155.7, 157.2.
Preparation of DGEBA/FHPPO composites
Different amounts of FHPPO_2, i.e., 3, 5, 10, 15, and 20 wt% of
the total weights, were added into the stoichiometric mixtures
of DGEBA and MNA. The accelerator, 2E4MZ (1 wt% of weight
of the DGEBA) was then added under continuous stirring at 60
uC. In order to obtain a homogeneous mixture, a small amount
of tetrahydrofuran (THF) was used. The homogeneous
14512 | RSC Adv., 2013, 3, 14509–14520
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
The glass transition temperatures were determined by a
METTLER DSC1 differential scanning calorimetry on second
heating runs at a heating rate of 10 K min21 under nitrogen.
Thermal stability was measured using a Perkin-Elmer Pyris1
thermogravimetric analyzer (TGA) from 50 to 750 uC at a
heating rate of 20 K min21 under nitrogen.
Values of the dielectric constant and dissipation factor of
cured samples were characterized by an Anton Paar MCR 301
rheometer with a dielectro-rheological device (DRD) at 30 uC.
The test samples were round disks with a diameter of 25.0 mm
and a thickness of 2.0 mm.
Moisture absorption measurements were tested using the
following procedure: samples in the size of 60.0 mm 6 12.8
mm 6 3.2 mm were dried under vacuum at 100 uC to remove
all moisture. After cooling to room temperature, the speci-
mens were weighed. Then they were placed in boiling water for
2 h and weighed again. The moisture absorption (MA) was
calculated as
MA = (w/w0 2 1) 6 100%
(1)
where w is the sample weight after boiling water treatment,
and w0 is the sample weight before placing in water.
Positron annihilation lifetime spectroscopy (PALS) was used
to determine the free volumes or cavities of the cured DGEBA/
FHPPO composites. Detailed background information on the
theory and techniques of PALS can be found elsewhere.38,39
Generally speaking, the lifetimes and intensities of positrons,
which are generated from a 22Na source, are measured and
used to estimate the local free volumes inside polymers. Once
inside the polymers, the positron forms one of the two
possible types of positroniums, i.e., an ortho-positronium or a
para-positronium, which can pair with electrons abstracted
from the environment. After proper curve fitting, the lifetimes
of the various species and their intensity can be determined
from the decay spectra of positrons. In particular, the lifetime
of ortho-positronium (t3) and intensity (I3) are found to be
correlate well with the local free volume in polymers,40–42 in
which t3 corresponds to the size of individual free volume sites
and I3 corresponds to their number concentration.
Fig. 1 13C NMR spectra of compounds of 2 (a) and 29 (b).
assignments of each peak of AB2 monomer (2) and compound
29 are shown in Fig. 1.
The mechanism of the Friedel–Crafts alkylation is shown in
Scheme 2. Firstly, compound 1 (4,49-difluorodiphenylmetha-
nol) produces a stable carbocation 1+ under the protonation of
H2SO4. Then, a proton transfers from the oxygen to the carbon
atom, and the para and ortho positions are electron-rich.
Carbocation 1+ is captured by the para position of phenol to
form the main product 2; carbocations 1+ are captured by the
para- and ortho-positions of phenol to form the byproducts 29
and 299. Considering the steric hindrance of 1+ in the ortho-
substitution, compound 2 is the main product, and no
compound 299 is found in this reaction. In order to further
increase the yield of compound 2, low temperature (,5 uC)
and excess phenol could be used.
Results and discussions
Synthesis of AB2 monomer
In the AB2 monomer, the phenolic group and fluorine group
are represented by A and B, respectively. The synthesis route of
the AB2 monomer is shown in Scheme 1. The monomer can be
synthesized easily and cheaply with the commercially available
4,49-difluorodiphenylmethanol and phenol. The AB2 mono-
mer, 4-hydroxyl-49,499-difluorotriphenylmethane, was synthe-
sized successfully by Friedel–Crafts alkylation using H2SO4 as
the catalyst in CH2Cl2 solvent. The structure of compound AB2
was characterized by 1H NMR, 13C NMR, and MS. The by-
product compound 29, which was obtained by column
chromatography separation, was also determined by 1H
NMR, 13C NMR and MS. The spectra of 13C NMR and the
Synthesis of fluoro-terminated hyperbranched poly(phenyl
oxide)
Using the nucleophilic aromatic substitution (SNAr) reaction, a
series of fluoro-terminated hyperbranched poly(phenylene oxide)
with different molecular weights, (FHPPO)s, were prepared by one-
step polymerization of 4-hydroxyl-49,499-difluorotriphenylmethane
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 14509–14520 | 14513

View Article Online
Paper
RSC Advances
sized by the SNAr reaction using 4-hydroxyl-49,499-difluorotri-
phenylmethane as the AB2 monomer.
In the SNAr reaction, linkages are formed by the reaction
between a phenolic group and a leaving group in the presence of
a suitable base, such as potassium carbonate. Because 4-hydro-
xyl-49,499-difluorotriphenylmethane does not have any activat-
ing groups in the ortho or para positions of the leaving group,
polar solvents with high boiling points, N-methylpyrrolidone
(NMP), along with K2CO3, were used in order to produce the
phenolate ion. In order to increase the average molecular
weight, a mixture of NMP and toluene was used to remove the
water produced at the very beginning of the reaction.
In general, high monomer concentration and long reaction
time are beneficial for increasing the molecular weight. As
shown in Table 1, the average molecular weight number (Mn)
and the polydispersity index (PDI) of FHPPO increase with
reaction time and monomer concentration. In order to check
the reproducibility, FHPPO_2 was repeated three times under
the same reaction conditions. The Mns and PDIs from
different experiments are very close: for the first batch, Mn
y 2500, PDI y 2.2; for the second batch, Mn y 2800, PDI y
2.2; for third batch, Mn y 2500, PDI y 2.1. The solubilities of
the resulting FHPPO_2 and FHPPO_5 in different solvents are
summarized in Table 2. Both FHPPOs dissolve well in a variety
of common organic solvents, i.e., tetrahydrofuran (THF),
dichloromethane (CH2Cl2), chloroform (CHCl3), toluene, N,
N-dimethyl formamide (NMF), and N-methyl-2-pyrrolidinone
(NMP), but not in acetone or methanol. Ethanol, DMSO and
ethyl acetate are poor solvents for FHPPO_2, which has a lower
molecular weight, but the good solvents for FHPPO_5, which
has a higher molecular weight. One explanation may be that
the FHPPO with higher molecular weight has much more
terminal groups and thus is more soluble than that with lower
molecular weights. We note that the solubility of the FHPPO is
different from the HPPO synthesized in our previous
studies,32,33 indicating that the terminal groups play a big
role in the solubility of hyperbranched polymers.
Fig. 2 The 13C NMR spectrum of FHPPO.
(AB2 monomer ‘2’), as shown in Scheme 1. The chemical
structure of the FHPPO was determined by 13C NMR and FT-IR
measurements, as shown in Fig. 2 and 3, respectively. The
assignments of peaks of 13C NMR are shown in Fig. 2. In the FT-
IR spectrum, the peaks at 1599 and 1498 cm21 correspond to
the stretching vibrations of aromatic rings; the peaks at 1236
and 825 cm21 correspond to the Ph–O–Ph stretching and the
Ph–H distorting, respectively; the broad band at about 3400
cm21 corresponds to stretching vibration of phenolic hydroxyl
groups.
In our previous studies,32,33 HPPOs were prepared from
4-bromo-49, 499-dihydroxyltriphenylmethane by the Ullmann
reaction using CuCl as the catalyst. However, it was difficult to
remove all the copper salt in the reaction system. Thus, the
SNAr reaction is used to replace the Ullmann reaction in this
study. In order to take advantages of the desirable properties
of the fluorinated substituents, such as low polarizability of C–
F bonds,43 low moisture absorption, and good solubility,
HPPO with abundant fluoro-terminated groups was synthe-
Degree of branching of FHPPO
Aside from molecular weights and PDI, the degree of
branching (DB) is another important parameter for describing
hyperbranched molecules. Unlike dendritic molecules, which
are essentially perfectly branched and have a DB of 100%, the
DB of hyperbranched molecules is less than 100%. The DB can
´
be calculated using NMR spectroscopy. Hawker and Frechet
defined DB as44
DzT
DB~
(2)
DzLzT
where D is the number of dendritic units, T is the number of
terminal units, and L is the number of linear units.
Since the number of dendritic units is theoretically equal to
the number of terminal units plus one, Holter and Frey45
proposed that for higher molecular weights molecules, eqn (2)
is simplified to
2T
DB~
(3)
2TzL
Fig. 3 The FT-IR spectrum of FHPPO.
14514 | RSC Adv., 2013, 3, 14509–14520
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
Table 1 Properties of the fluoro-terminated hyperbranched poly(phenylene oxide)s (FHPPOs)
a
b
Time
(d)
Concentration
(M)
Mn
Tg
(uC)
Yield
(%)
Td5
Tmax
(uC)
Residue at
700 uC, in N2 (%)
Entry
(g mol21
)
PDI
(uC)
DBc
DBd
DBe
FHPPO_1
FHPPO_2
FHPPO_3
FHPPO_4
FHPPO_5
2
2
2
4
5
0.1
0.2
0.4
0.2
0.2
20 6 102
25 6 102
54 6 102
58 6 102
68 6 102
1.7
2.2
3.2
3.7
4.8
135
147
156
163
164
78
80
86
88
93
556
559
554
551
549
601
610
603
604
604
74.5
73.9
72.9
71.0
69.8
0.63
0.60
0.56
0.55
0.53
0.64
0.63
0.57
0.57
0.53
0.62
0.61
0.58
0.57
0.56
a
b
c
The temperature at 5 wt% decomposition. The temperature of the maximum decomposition rate. Degree of branching calculated
according to Hawker’s method using 13C NMR. Degree of branching calculated according to Holter’s method using 13C NMR. Degree of
branching calculated according to Holter’s method using 19F NMR.
d
e
each individual peak can be extracted. The DB values of the
FHPPO_1–5 based on eqn (2) and based on eqn (3) are
calculated and summarized in Table 1.
where T is the number of terminal units, and L is the number
of linear units.
The three types of structural units of FHPPO, i.e., the
terminal, linear, and dendritic units, are shown as following:
In addition to 13C NMR, for FHPPO, DB can also be
determined from 19F NMR. As shown in Fig. 5, the 19F NMR
spectra of MC3 and MC4 also show a distinct difference. The
two atoms of MC3 are equivalent and appear as a sharp singlet
at 2116.4 ppm whereas the single fluorine atom of MC4
appears at 2116.7 ppm. Thus, based on the different
resonance peaks of fluorine atoms between terminal and
linear structures, DB can also be determined from 19F NMR
spectra. Compared with the 19F NMR spectra of model
compounds (see Fig. 5), the resonance peaks of FHPPO at a
lower field (2116.4 ppm) can be assigned to the terminal
units, whereas the resonance at a higher field (2116.7) can be
assigned to the linear subunits. Using similar procedures like
13C NMR, the area of each individual peak can be extracted,
and the DB values of the FHPPO_1–5 based on eqn (3) were
calculated and are summarized in Table 1.
The NMR spectra contain all necessary information for
estimating DB. Therefore, model compounds of MC3, MC4,
and MC5 resembling the terminal, linear, and dendritic
structures, respectively, are synthesized to help determine
the DBs of the FHPPOs (see Scheme 3). The 13C NMR spectra
of MC3, MC4, MC5, and FHPPO (see Fig. 4) are used for the
determination of the DBs of the FHPPOs. The chemical shifts
at around 54.7 ppm, which correspond to the methylidyne
carbons, are the best choice for identifying the numbers of
terminal, linear, and dendritic units. Based on the character-
istic chemical shifts of methylidyne carbons in MC3, MC4, and
MC5, three distinct resonance peaks from lower to higher filed
in the spectra of the FHPPO are assigned to the dendritic,
linear, and terminal structures, respectively (see Fig. 4).
Assuming that the resonance peaks in the FHPPOs are
Gaussian and the peak positions are determined from the
second derivative of the spectra, the resonance peaks of the
FHPPO can be easily deconvoluted to extract the information.
For instance, Fig. 4d shows a typical curve-fitting result of the
FHPPO. The spectra of each individual component agree well
with their corresponding model compounds. Moreover, the
reconstructed spectrum, obtained by summing all of the three
deconvoluted peaks (shown as a dotted line) match well with
the measured spectrum (solid line), indicating the deconvolu-
tion process is valid. Following those procedures, the area of
As shown in Table 1, DBs calculated from different methods
(13C or 19F; eqn (2) or eqn (3)) all decrease with increasing
molecular weight. This trend could be attributed to the
difference in reactivity between the two B groups because of
the steric hindrance. That is to say, when one B group of the
AB2 monomer reacts with an A group, the other B group in the
same AB2 has less chance to react with the A group. As the
molecular weight increases, the hyperbranched molecule
becomes bigger, and the terminal B groups on the outside
surface become increasingly crowded. The crowding or the
steric hindrance effects inhibit the reaction between the
second B group and the A group, resulting in a more linear
architecture and thus a lower DB. As shown in Table 1, for the
same polymer, DBs defined by the Holter’s method are slightly
higher than those defined by the Hawker’s method. Similar
results were also reported by another group24 and could be
Table 2 Solution properties of the fluoro-terminated hyperbranched poly(phenylene oxide)s
Entry
CH2Cl2
CHCl3
DMSO
DMF
NMP
Toluene
Ethyl acetate
Acetone
Methanol
Ethanol
FHPPO_2
FHPPO_5
+
+
+
+
+
+
+
+
+
+
2
+
2
2
2
2
2
2
+
a
+, soluble; , partially soluble; 2, insoluble.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 14509–14520 | 14515

View Article Online
Paper
RSC Advances
Fig. 5 19F NMR spectra used to determine the degree of branching of FHPPO:
(a) model compound 3, (b) model compound 4. (c) The deconvoluted resonance
peaks of observed 19F NMR spectra of FHPPO (solid black line) into terminal (T)
and linear (L) structures. The reconstructed spectrum (dotted line) was obtained
by summing up the two deconvoluted peaks.
can inhibit the mobility of chain segments. It is also noted
from Table 1 that the Tg of FHPPOs tends to level off at high
molecular weights. This could be explained by the coupled
effects of two factors: the portions of the rigid triphenyl
structures from the AB2 monomers, the changes in DB and
molecular weights. As molecular weight increases, the propor-
tion of rigid triphenyl groups also increases, which increases
the rigidity of the hyperbranched structures and limits the
Fig. 4 13C NMR spectra used to determine the degree of branching of the
FHPPO: (a) model compound 3, (b) model compound 4, (c) model compound 5.
(d) The deconvoluted resonance peaks of the observed 13C NMR spectra of the
FHPPO (solid black line) into the dendritic (D), linear (L), and terminal (T)
structures. The reconstructed spectrum (dotted line) was obtained by summing
up the three deconvoluted peaks.
explained by the approximation used in the Holter method,
which is more valid for polymers with high molecular weights.
We note that the DB values defined by the Holter’s method
from 13C NMR and 19F NMR are not the same; however, the
difference is marginal. Thus, for FHPPO molecules, both 13C
NMR and 19F NMR can be used to estimate the DB.
Thermal properties of FHPPO
The DSC curves of FHPPOs with different molecular weights
are presented in Fig. 6. All samples show one clear Tg and no
endothermic peak was observed below 250 uC. Tg values of
FHPPO detected by DSC increase with increasing molecular
weight, and the results are summarized in Table 1. The Tg
values of FHPPO_1 to FHPPO_5 increase from 135 uC to 164
uC, which are all higher than the Tg of linear PPO (Tg # 95 uC)
with a high molecular weight.18 The increase in Tg may be
ascribed to the highly branched molecular architecture, which
Fig. 6 DSC thermograms of FHPPOs at a heating rate of 10 K min21
.
14516 | RSC Adv., 2013, 3, 14509–14520
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
Fig. 8 Dielectric properties of DGEBA/FHPPO composites as a function of FHPPO
loading.
Fig. 7 TGA thermograms of FHPPOs under N2 atmosphere at a heating rate of
20 K min21
.
fluoro-terminated groups, and inherent free volumes of
FHPPO can improve the dielectric properties of DGEBA/
FHPPO composites. The relative dielectric constant (k) and
dielectric dissipation factor (Df) of cured samples at 1 MHz are
shown in Fig. 8. Both the k and Df values of the DGEBA/FHPPO
composites are lower than those of the neat epoxy system. The
higher the FHPPO content in the composites, the lower the k
and Df values of the FHPPO modified epoxy resins are. At 20
wt% loading of FHPPO, the k value decreased from 3.00 to
2.77, and Df value decreased from 0.0244 to 0.0142. The
decrease of k and Df in DGEBA/FHPPO can be attributed to the
added FHPPO which has lower values of k and Df. Generally
speaking, low-k materials are materials with a dielectric
constant of 3.0 or less. When cured with amine, the majority
of the unmodified epoxy has a dielectric constant higher than
4.0; when cured with anhydrides, the dielectric constants of
the unmodified epoxy generally exceed 3.4. As mentioned
above, after addition of the FHPPO, the dielectric constant of
the modified epoxy is lowered to 2.77 (,3.0). Thus, the FHPPO
is a low-k modifier for epoxy resin.
The dielectric properties of epoxy materials depend on lots
of factors, including chemical structures, test conditions, and
moisture content. As a result, even for the same epoxy resin,
different curing agents and test conditions are often used,
making the direct comparison between different epoxy
systems difficult and misleading. Even so, a comparison is
made against some typical low-k modifiers for epoxy materials.
Wu46 used linear polyphenylene oxide (PPO) to modify a
DGEBA/cyanate ester curing system. The dielectric constant at
1 MHz decreased from 4.0 of the neat system to 3.5 at 20 wt%
PPO loading. In Wang’s work,47 a nanoporous additive,
polyhedral oilgomeric silisesquioxane (POSS) which contains
eight functional hexafluorine groups (OF), was added into a
UV-cured epoxy resin. The dielectric constant decreased from
3.71 of the neat epoxy to 2.80 at 15 wt% OF loading at 100 kHz.
In Yung’s work,48 hollow glass microspheres were added into a
dicyandiamide-cured epoxy system. The dielectric constant at
1 MHz decreased from 3.98 of the neat system to 3.42 at 20
wt% loading. As mentioned above, the dielectric constant
strongly depends on the chemical structure, i.e., the curing
agents. In our case, methyl nadic anhydride was chosen for the
mobility of the segments. However, the DB decreases with
increasing molecular weight, which means more linear
structures in FHPPOs. The increase in linear structures brings
in more ‘defects’ or flexible regions in the hyperbranched
structures, resulting in a decrease in the Tg. The change of the
Tg with molecular weight is different from that reported in our
previous stuides.32,33
The weight loss traces of FHPPOs with different molecular
weights in nitrogen as a function of temperature are shown in
Fig. 7. The values of 5% weight loss temperature (Td5),
maximum temperatures of decomposition (Tmax), and residue
at 700 uC are summarized in Table 1. The Td5s of FHPPOs
change very little firstly with molecular weight, and then
decrease slightly, which cover a range of 549 to 559 uC. The
Td5s of FHPPOs are notably higher than those of HPPOs with
phenolic terminal groups in our previous study, where the Td5
s
of the HPPOs were in the range of 258 to 295 uC. This suggests
that the terminal groups play an important in the thermal
stability of hyperbranched polymers. The Tmax values of the
FHPPOs are all about 600 uC, which are similar to those
phenolic-terminated HPPOs.32 As shown in Fig. 7, the
residuals at 700 uC decrease with increasing molecular weight
and are above 70%. To sum up, FHPPOs have excellent
thermal stability.
Dielectric performance of FHPPO/DGEBA composites
DGEBA/FHPPO composites were prepared by blending differ-
ent amounts of FHPPO_2 into DGEBA and cured with MNA.
Before being cured, the mixtures of FHPPO (¡20%) and epoxy/
MNA systems were all transparent at room temperature,
indicating that homogenous mixtures can be obtained using
current mixing procedures. Although FHPPO is a non-reactive
modifier for epoxy resins, the hydrogen bonding between
fluorine atoms of the FHPPO and hydrogen atoms of epoxy
resin can increase the compatibility and lead to homogenous
mixtures. DGEBA/FHPPO composites were prepared by curing
mixtures following the aforementioned curing procedures. It is
expected that the low polarity basic units of PPO, abundant
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 14509–14520 | 14517

View Article Online
Paper
RSC Advances
curing agent, which leads to a more densely cross-linked, less
polar network and more free volume from bicyclic structure
than other common curing agents (amine and cyanate ester).
Thus, the dielectric constant of the unmodified system is
already low (3.0) and further decrease in the dielectric constant
is more challenging. At 15 wt% FHPPO loading, the dielectric
constant decreased from 3.0 of the unmodified system to 2.8,
which has made a quite bit of progress for lower dielectric
constant although the improvement is not as big as the
examples mentioned above. In addition, the new modifier has
good solubility and can form homogeneous solutions upon
mixing, which will not complicate the fabrication process.
Moreover, after curing, the thermal stability and mechanical
properties are improved. (Mechanical and other properties will
be reported in more details in future work). Thus, FHPPO is
indeed a new promising low-k modifier for epoxy.
structures. Different from other linear fluorinated polymers,
free volume in hyperbranched molecules contributes more to
the lowering of dielectric properties, thus a quantitative
evaluation of the free volume in the DGEBA/FHPPO compo-
sites is very useful.
In order to quantify the free volume in the DGEBA/FHPPO
composites, PALS measurements were performed. The results
are summarized in Table 3. The lifetime of the ortho-
positronium (t3) corresponds to the size of the individual free
volume site, and the intensity (I3) corresponds to their number
concentration. Compared with neat DGEBA, the hyper-
branched polymers are found to have more free volume sites
both in size and number or concentration (Table 3). The free
volume fraction (Fv) can be expressed as
Fv = cVI3
(5)
Free volumes of FHPPO/DGEBA composites
where V is the total volume of a single site of free volume or
cavity, which can be calculated from t3, c is a constant
empirically determined from comparison with pressure-
volume-temperature data and is often found to be 0.018
0.002 nm23 for polymers, I3 is the lifetime of intensity. A
detailed background concerning the theory and techniques of
PALS can be found elsewhere.51,52 The results in Table 3 reveal
that the free volume fractions of DGEBA/FHPPO composites
increase monotonically with increasing FHPPO loading. At 20
wt% loading, the free volume fraction reaches the maximum
value of 4.35%, which is ca. 27% increase compared with that
of the neat DGEBA. The PALS results suggest that the DGEBA/
FHPPO composites have more free volume because of the
introduction of the FHPPO. The higher free volume of FHPPO
comes from two factors: the large molar free volume of fluoro-
terminal groups and the molecule-scale cavities from hyper-
branched structures.
According to the Clausius-Mossotti theory, the dielectric
constant can be estimated by the following formula:49,50
P
P
1z2ð Qi= ViÞ
P
P
k~
(4)
1{ Qi= Vi
where k is the dielectric constant, Qi is the molar polarization
of each building block, and Vi is the molar volume of each
building block. Obviously, decreasing the molar polarization
and/or increasing the molar volume of the building blocks can
lower the dielectric constant.
Poly(phenyl oxide) (PPO) has a low polarity structure and has
been used for electronic materials. C–F bond has low
polarizability as well as large molar volume.43 Therefore, the
incorporation of fluorinated substituents into the polymers
can effectively lower the dielectric constant by decreasing the
molar polarization and by increasing the molar volume. In
addition, the hyperbranched polymers contain inherent free
volumes or molecule-scale cavities. Thus, the dielectric
constant can be further reduced by increasing the molar
volume from hyperbranched polymer structures. In addition,
the dielectric dissipation factor is proportional to molar
polarity. Therefore, the low polarity PPO structure and
fluoro-terminal groups in FHPPO can reduce the dissipation
factor of the materials. In summary, the lower k and Df values
of DGEBA/FHPPO composites can be related to three factors,
i.e., the low polarity structure of PPO backbones, abundant
fluoro-terminated groups, which have a large molar free
volume and low polarizability of C–F bonds, and the inherent
free volume or molecule-scale cavities in the hyperbranched
Moisture resistance of DGEBA/FHPPO composites
The absorbed moisture can act as a plasticizer and impair the
mechanical and thermal properties of cured resins. Water has
a dielectric constant of 78; thus, even a small quantity of
absorbed water in the composites can have adverse effects on
the dielectric properties and shorten the service life of low-k
materials. Therefore, the moisture resistance is also an
important parameter for low-k materials. As shown in Fig. 9,
the moisture absorption (MA) of the DGEBA/FHPPO compo-
sites decreases with increasing FHPPO loading. At 20 wt%
loading of FHPPO, the MA decreases from 0.539% of the
unmodified system to 0.396%. The lower moisture absorption
Table 3 PALS data and free volume fractions of the DGEBA/FHPPO composites
Composition
FHPPO_2 content (wt%)
Lifetime of ortho-positronium t3 (ns)
Intensity I3 (%)
Free volume fraction, Fv (%)
Neat DGEBA
0
3
5
10
15
20
1.9880 0.0110
1.9372 0.0084
2.0180 0.0100
1.9890 0.0110
1.9610 0.0100
2.0048 0.0092
19.87 0.41
20.57 0.74
22.83 0.79
23.47 0.65
24.70 0.69
24.01 0.72
3.45
3.83
4.11
4.12
4.17
4.35
3 wt% FHPPO_2
5 wt% FHPPO_2
10 wt% FHPPO_2
15 wt% FHPPO_2
20 wt% FHPPO_2
14518 | RSC Adv., 2013, 3, 14509–14520
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
Acknowledgements
This work is financially supported by the National Natural
Science Foundation of China (No.2092023, No. 51173012).
Notes and references
1 Y. H. Kim and O. W. Webster, Macromolecules, 1992, 25,
5561–5572.
2 Y. H. Kim and O. W. Webster, J. Am. Chem. Soc., 1990, 112,
4592–4593.
3 Y. H. Kim, J. Polym. Sci., Part A: Polym. Chem., 1998, 36,
1685–1698.
4 B. I. Voit and A. Lederer, Chem. Rev., 2009, 109, 5924–5973.
5 C. Gao and D. Yan, Prog. Polym. Sci., 2004, 29, 183–275.
6 D. Yan, C. Gao and H. Frey, in Hyperbranched Polymers:
Synthesis, Properties and Applications, John Wiley& Sons,
Inc, New Jersey, 1st edn, 2011, ch.14, pp. 415–440.
7 D. Yan, Y. Zhou and J. Hou, Science, 2004, 303, 65–67.
8 D. Zhang, E. Liang, T. Li, S. Chen, J. Zhang, X. Cheng,
J. Zhou and A. Zhang, RSC Adv., 2013, 3, 3095–3102.
9 Q. Zhu, F. Qiu, B. Zhu and X. Zhu, RSC Adv., 2013, 3,
2071–2083.
10 D. Wang, H. Chen, Y. Su, F. Qiu, L. Zhu, X. Huan, B. Zhu,
D. Yan, F. Guo and X. Zhu, Polym. Chem., 2013, 4, 85–94.
11 H. Jin, W. Huang, X. Zhu, Y. Zhou and D. Yan, Chem. Soc.
Rev., 2012, 41, 5986–5997.
12 J. Liu, L. Yu, C. Zhong, R. He, W. Yang, H. Wu and Y. Cao,
RSC Adv., 2012, 2, 689–696.
Fig. 9 Moisture absorption of the DGEBA/FHPPO composites as a function of
FHPPO loading.
is attributed to the higher content of hydrophobic fluorinated
substituents and low-polarity chemical structure of PPO in the
skeleton of FHPPO.
The excellent thermal stability, low dielectric constant, low
dielectric dissipation factor, and low water absorption make
FHPPO a promising additive for epoxy resin in microelectronic
applications, such as in semiconductor devices, electromag-
netic components, printed circuit boards, and satellite com-
munication equipment.
13 M. Zou, J. Fang, J. Liu, C. Li and R. Guan, Solid State Ionics,
2012, 220, 23–31.
14 M. Kakimoto, S. J. Grunzinger and T. Hayakawa, Polym. J.,
2010, 42, 697–705.
Conclusions
15 P. Himmelberg and E. Fossum, J. Polym. Sci., Part A: Polym.
Chem., 2005, 43, 3178–3187.
16 Q. Lin, S. Unal, A. R. Fornof, I. Yilgor and T. E. Long,
Macromol. Chem. Phys., 2006, 207, 576–586.
17 Z. Yu, E. Fossum, D. H. Wang and L. Tan, Synth. Commun.,
2008, 38, 419–427.
18 I. In, H. Lee and S. Y. Kim, Macromol. Chem. Phys., 2003,
204, 1660–1664.
19 S. Kwak and D. U. Ahn, Macromolecules, 2000, 33,
7557–7563.
20 S. Kwak and H. Y. Lee, Macromolecules, 2000, 33,
5536–5543.
21 C. J. Hawker and F. Chu, Macromolecules, 1996, 29,
4370–4380.
22 H. Satpathi, A. Ghosh, H. Komber, S. Banerjee and B. Voit,
Eur. Polym. J., 2011, 47, 196–207.
23 L. Sennet, E. Fossum and L. Tan, Polymer, 2008, 49, 3731–3736.
24 D. P. Bernal, L. Bedrossian, K. Collins and E. Fossum,
Macromolecules, 2003, 36, 333–338.
Using the nucleophilic aromatic substitution (SNAr) reaction, a
fluoro-terminated hyperbranched poly(phenylene oxide)
(FHPPO) was readily prepared from a new AB2 monomer,
4-hydroxyl-49,499-difluorotriphenylmethane. The average mole-
cular weight and the PDI of the FHPPO increased with
increasing monomer concentration and reaction time. The
chemical structures of the FHPPOs were characterized by 13C
NMR and FT-IR, and the degree of branching (DB) of the
FHPPOs was estimated by 13C NMR and 19F NMR measure-
ments. The DBs of the FHPPOs decreased with molecular
weight. The FHPPOs showed excellent thermal stability and
high glass transition temperatures. Because of the low polarity
structure of the PPO backbones, abundant fluoro-terminated
groups which have large molar free volume, low polarizability
of C–F bonds, and the inherent free volume or molecule-scale
cavities in the hyperbranched structures, the relative dielectric
constant (k), the dielectric dissipation factor (Df), and moisture
absorption of the DGEBA/FHPPO composites apparently
decreased after the introduction of FHPPO_2. The increase
of free volume with addition of the FHPPO was confirmed by
PALS measurements. These excellent dielectric and thermal
properties make FHPPO a promising low-k modifier for
DGEBA epoxy.
25 M. Czupik and E. Fossum, J. Polym. Sci., Part A: Polym.
Chem., 2003, 41, 3871–3881.
26 D. P. Bernal, N. Bankey, R. C. Cockayne and E. Fossum, J.
Polym. Sci., Part A: Polym. Chem., 2002, 40, 1456–1467.
27 Y. Seike, Y. Okude, I. Iwakura, I. Chiba, T. Ikeno and
T. Yamada, Macromol. Chem. Phys., 2003, 204, 1876–1881.
28 C. Chiang and F. Chang, J. Polym. Sci., Part B: Polym. Phys.,
1998, 36, 1805–1819.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 14509–14520 | 14519

View Article Online
Paper
RSC Advances
29 A. Ghosh, S. Chatterjee, S. Banerjee, H. Komber and
B. Voit, J. Macromol. Sci., Part A: Pure Appl. Chem., 2011, 48,
509–517.
42 J. Kansy, G. Consolati and C. Dauwe, Radiat. Phys. Chem.,
2000, 58, 427–431.
43 D. W. Van Krevelen and K. T. Nijenhuis, in Properties of
Polymers, Elsevier, Amsterdam, 4th edn, 2009, ch. 11, pp.
319–354.
¨
30 A. Ghosh, S. Banerjee, H. Komber, A. Lederer, L. Haussler
and B. Voit, Macromolecules, 2010, 43, 2846–2854.
¨
31 S. Banerjee, H. Komber, L. Haussler and B. Voit, Macromol.
´
44 C. J. Hawker, R. Lee and J. M. J. Frechet, J. Am. Chem. Soc.,
Chem. Phys., 2009, 210, 1272–1282.
1991, 113, 4583–4588.
32 J. Zhang, H. Wang and X. Li, Polymer, 2006, 47, 1511–1518.
33 J. Zhang, H. Wang and X. Li, Chin. J. Polym. Sci., 2006, 24,
413–419.
34 P. Huang, A. Gu, G. Liang and L. Yuan, J. Appl. Polym. Sci.,
2012, 123, 2351–2359.
35 P. Huang, A. Gu, G. Liang and L. Yuan, J. Appl. Polym. Sci.,
2011, 120, 451–457.
36 P. Huang, A. Gu, G. Liang and L. Yuan, J. Appl. Polym. Sci.,
2011, 121, 2113–2122.
37 J. Lv, Y. Meng, L. He, T. Qiu, X. Li and H. Wang, J. Appl.
Polym. Sci., 2013, 128, 907–914.
38 J. Habsuda, G. P. Simon, Y. B. Cheng, D. G. Hewitt, D.
R. Diggins, H. Toh and F. Cser, Polymer, 2002, 43,
4627–4638.
39 P. G. Simon, Trends Polym. Sci., 1997, 5, 394–400.
40 D. M. Bigg, Polym. Eng. Sci., 1996, 36, 737–743.
41 R. A. Pethrick, Prog. Polym. Sci., 1997, 22, 1–47.
¨
45 D. Holter, A. Burgath and H. Frey, Acta Polym., 1997, 48,
30–35.
46 S. J. Wu, T. K. Lin and S. S. Shyu, J. Appl. Polym. Sci., 2000,
75, 26–34.
47 Y. Wang, W. Chen, C. Yang, C. Lin and F. Chang, J. Polym.
Sci., Part B: Polym. Phys., 2007, 45, 502–510.
48 K. C. Yung, B. L. Zhu, T. M. Yue and C. S. Xie, J. Appl. Polym.
Sci., 2010, 116, 225–527.
49 Y. Kornyushin, Sci. Sintering, 2009, 41, 225–245.
50 O. Ichiro, in Handbook of Low and High Dielectric Constant
Mateirals and Their Applications, Academic Press,
Burlington, 1999, vol. 1, ch 5, pp. 213–240.
51 D. Ratna and G. P. Simon, J. Appl. Polym. Sci., 2010, 117,
557–564.
52 Y. Y. Wang, H. Nakanishi, Y. C. Jean and T. C. Sandreczki,
J. Polym. Sci., Part B: Polym. Phys., 1990, 28, 1431–1441.
14520 | RSC Adv., 2013, 3, 14509–14520
This journal is ß The Royal Society of Chemistry 2013