A new approach to improving flame retardancy, smoke suppression and anti-dripping of PET: Via arylene-ether units rearrangement reactions at high temperature

Article abstract of DOI:10.1016/j.polymer.2015.09.016

In order to avoid the serious melt-dripping problems caused by conventional phosphorus flame retardants, a new arylene-ether containing monomer named 2,2′-(4,4′-(1,4-phenylenebis(oxy))bis(4,1-phenylene))bis(oxy) diethanol (PBPBD) was synthesized and intro

Full text of DOI:10.1016/j.polymer.2015.09.016

Polymer 77 (2015) 21e31
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
A new approach to improving flame retardancy, smoke suppression
and anti-dripping of PET: Via arylene-ether units rearrangement
reactions at high temperature
,
,
, **
De-Ming Guo a b, Teng Fu a, Chao Ruan a, Xiu-Li Wang a *, Li Chen a, Yu-Zhong Wang a
a Center for Degradable and Flame-Retardant Polymeric Materials, College of Chemistry, State Key Laboratory of Polymer Materials Engineering, National
Engineering Laboratory of Eco-Friendly Polymeric Materials (Sichuan), Sichuan University, Chengdu, 610064, China
b College of Light Industry, Textile and Food Engineering, Sichuan University, Chengdu, 610064, China
a r t i c l e i n f o
a b s t r a c t
Article history:
In order to avoid the serious melt-dripping problems caused by conventional phosphorus flame re-
tardants, a new arylene-ether containing monomer named 2,20-(4,40-(1,4-phenylenebis(oxy))bis(4,1-
phenylene))bis(oxy) diethanol (PBPBD) was synthesized and introduced into the PET main-chain via
condensation polymerization. With the incorporation of arylene-ether units, the thermal stability and
the char residue from combustion of flame-retardant-element-free copolyesters containing PBPBD
(BDxPETs) increased dramatically. However, the crystallinity and melting temperature of these co-
polymers were lower than the same quantities for PET. Thermal degradation kinetic was investigated
using the Ozawa-Flynn-Wall method, which illustrated that the apparent activation energy for decom-
position of the copolyesters was enhanced with increasing the conversion and the content of PBPBD. Py-
GC/MS results showed that the PBBPD structural units undergo rearrangement reactions at high tem-
perature, and ultimately form conjugated heteroaromatic structures, which lead to the formation of
stable char residues with unique “stick-shaped” micro-morphology. Due to these rearrangement re-
actions, BDxPETs showed an expected flame retardant performance and a reduction in smoke generation
during combustion. For example, the LOI value for BD10PET was 28.4, and the total smoke production
(TSP) was 10.4 m2, much smaller than that for PET, 18.8 m2. Moreover, melt-dripping was restricted.
© 2015 Elsevier Ltd. All rights reserved.
Received 25 June 2015
Received in revised form
30 August 2015
Accepted 7 September 2015
Available online 9 September 2015
Keywords:
2,20-(4,40-(1,4-phenylenebis(oxy))bis(4,1-
phenylene))bis(oxy) diethanol
PET copolyester
Arylene-ether units
Anti melt-dripping
Rearrangement
Flame-retardant-element-free
1. Introduction
burning surface of the material. Melt dripping is a real danger in a
fire, which leads to secondary damage and immediate empyrosis
Many different methods have been utilized to endow poly(-
ethylene terephthalate) (PET) with flame retardancy. Incorporating
phosphorus entities into the main-chain of PET is considered to be
one of the most efficient methods [1e4]. However, flame retarda-
tion of most phosphorus-containing copolyesters is achieved
through melt dripping, specifically by removing heat from the
for the human body. Recently, thermo-cross-linkable monomers or
ionic monomer have been introduced into the molecular chain of
PET [5e7]. Through chemical cross-linking or ionic aggregation, the
carbonization reactions during thermal degradation of the copo-
lyester were promoted, leading to an enhanced formation of char
residue. This reduces the rate of release of volatile products, and
inhibits melt dripping [8,9].
Besides the functional groups above-mentioned, some wholly
aromatic liquid-crystalline polyesters [10], all-aromatic poly(ether
imide)s (PEIs) [11] and poly(ether ether ketone) (PEEK) [12,13],
present excellent thermal properties and undergo thermal degra-
dation with formation of char as a consequence of the presence of
aromatic entities in the main-chain. Good thermal stability and
char forming capability are beneficial for providing outstanding
flame retardance and anti-melt-dripping property during a fire.
Other similar polymers, such as those containing arylene-ether
* Corresponding author. Center for Degradable and Flame-Retardant Polymeric
Materials, College of Chemistry, State Key Laboratory of Polymer Materials Engi-
neering, National Engineering Laboratory of Eco-Friendly Polymeric Materials
(Sichuan), 29 Wangjiang Road, Sichuan University, Chengdu, 610064, China.
** Corresponding author. Center for Degradable and Flame-Retardant Polymeric
Materials, College of Chemistry, State Key Laboratory of Polymer Materials Engi-
neering, National Engineering Laboratory of Eco-Friendly Polymeric Materials
(Sichuan), 29 Wangjiang Road, Sichuan University, Chengdu, 610064, China.
E-mail addresses: xiuliwang1@163.com (X.-L. Wang), yzwang@scu.edu.cn
(Y.-Z. Wang).
http://dx.doi.org/10.1016/j.polymer.2015.09.016
0032-3861/© 2015 Elsevier Ltd. All rights reserved.

22
D.-M. Guo et al. / Polymer 77 (2015) 21e31
units, i.e., “AreOeAr” structural units or “AreOeAr-OeAr” struc-
tural units may also display good flame retardance. These structural
units may play a critical and positive role in enhancing char for-
mation during thermal degradation thus improving the flame
retardant performance of these materials.
with water. After that, the brown product was collected, washed
with water, and dried at reduced pressure. The PBBPD was obtained
as brown powders. Yields: 85.3%; 1H-NMR (400 MHz, CDCl3,
d,
ppm): 2.60 (s, 6H), 6.98e7.09 (m, 4H), 7.12 (s, 4H), 7.86e8.10 (m,
4H).
In
this
work,
2,20-(4,40-(1,4-Phenylenebis(oxy))bis(4,1-
phenylene))bis(oxy) diethanol (PBPBD) containing arylene-ether
units has been synthesized and then introduced into the main-
chain of PET via melt polycondensation. The chemical structures
of these novel flame-retardant-element-free copolyesters
(BDxPETs) were characterized using 1H-NMR, 13C-NMR. Melt-
crystallization behavior, thermal stability, thermal degradation ki-
netic, pyrolysis behavior, flame retardant performance, and char
residue morphology were well investigated using DSC, TGA, Py-GC/
MS, LOI, cone calorimeter, and SEM.
2.1.2. Synthesis of 4,40-(1,4-phenylenebis(oxy))bis(4,1-phenylene)
diacetate (PBPD) [11,14]
PBBPD (34.6 g, 0.1 mol), 300 mL CHCl3 and m-CPBA (27.6 g,
0.2 mol) were added into a 500 mL round bottom flask. The mixture
was heated to reflux with stirring for 5 h, and then washed with
saturated NaHSO3 solution (250 mL), saturated NaHCO3 solution
(2 ꢀ 250 mL), and water (250 mL). The solvent was removed by
distillation, and the crude product was recrystallized from meth-
anol. The PBPD was acquired as golden crystals. The synthesis
process is shown in Scheme 1. Yields: 80.2%; 1H-NMR (400 MHz,
2. Experimental section
CDCl3, d, ppm): 2.32 (s, 6H), 6.84e7.19 (m, 12H).
2.1. Materials
2.1.3. Synthesis of 4,40-(1,4-phenylenebis(oxy)) diphenol (PD)
[11,14]
Hydroquinone (AR grade), dimethylacetamide (DMAc, AR
grade), chloroform (AR grade), methanol (AR grade), potassium
carbonate (AR grade), potassium hydroxide (AR grade) and potas-
sium iodide were received from Kelong Chemical Reagent Factory
(Chengdu, China). Ethylene carbonate (AR grade) and 4'-fluo-
roacetophenone (AR grade) were purchased from Aladdin-Reagent.
meta-Chloroperoxybenzoic acid (m-CPBA, AR grade) was received
from Hainachuan Technology Development Co., Ltd (Tianjin,
China). Dimethyl terephthalate (DMT, AR grade), ethylene glycol
(EG, AR grade), phenol, tetrachloroethane, and zinc acetate were all
supplied by Chengdu Chemical Industries Co. (Chengdu, China).
Tetrabutyl titanate was purchased from Kelong Chemical Reagent
Factory, and dissolved in anhydrous toluene to prepare a 0.2 g/mL
solution before use. Other materials were commercially available
and used as received.
The PBPD (37.8 g, 0.1 mol) was dissolved in 50 mL methanol and
treated with a 0.5 M KOH/methanol solution (30 mL). The reaction
mixture was heated to reflux for 3 h. The solvent was removed by
distillation, and the crude product was treated with a 1 M HCl so-
lution. The product was collected by filtration, washed with water,
and dried. The PD was gained as dark brown powders. The syn-
thesis process is also presented in Scheme 1. Yields: 90.2%; 1H-NMR
(400 MHz, DMSO, d, ppm): 6.73e6.78 (m, 4H), 6.82e6.87 (m, 4H),
6.88 (d, 4H), 9.30 (s, 2H).
2.1.4. Synthesis of 2,20-(4,40-(1,4-phenylenebis(oxy))bis(4,1-
phenylene))bis(oxy) diethanol (PBPBD)
2,20-(4,4'-(1,4-phenylenebis(oxy))bis(4,1-phenylene))bis(oxy)
diethanol (PBPBD) was prepared as previously described
[11,14e16], and its synthesis process is also given in Scheme 1. The
detailed synthesis procedure is: PD (29.4 g, 0.1 mol), ethylene
carbonate (17.6 g, 0.2 mol) and KI (0.2 g) were added into a 100 mL
two-necked round bottom flask. With the protection of N2 and a
magnetic stirrer, the mixture was heated to 175 ꢁC for 8 h, and then
washed with a 2 M NaOH solution and water. After filtration and
drying, the final product PBPBD was obtained as grey powders.
2.1.1. Synthesis of 1,10-(4,40-(1,4-phenylenebis(oxy))bis(4,1-
phenylene))diethanone (PBBPD) [11,14]
The chemical structure of PBBPD is shown in Scheme 1, and its
detailed synthesis process is: A 500 mL three-necked round bottom
flask equipped with a magnetic stirrer, reflux condenser and ni-
trogen inlet was charged with hydroquinone (11.0 g, 0.1 mol), 40-
fluoroacetophenone (27.6 g, 0.2 mol), K2CO3 (27.6 g, 0.2 mol), and
250 mL DMAc. The mixture was stirred and refluxed for 12 h. The
dark reaction product was cooled to room temperature and diluted
Yields: 92.0%; 1H NMR (400 MHz, DMSO,
12H), 3.68e3.73 (t, 4H), 3.92e3.98 (t, 4H), 4.76 (s, 2H); 13C NMR
(400 MHz, DMSO, , ppm): 155, 153, 150, 120, 119, 116, 70, and 60.
d, ppm): 6.84e7.12 (m,
d
Scheme 1. Synthesis processes for PBBPD, PBPD, PD, and PBPBD.

D.-M. Guo et al. / Polymer 77 (2015) 21e31
23
2.2. Preparation of BDxPETs
capillary column (DN-1701 FAST 10 m 0.10 mm 0.10 mm) of GC is as
following: 2 min at 45 ꢁC, after that the temperature was increased
to 280 ꢁC at a rate of 15 ꢁC/min then kept at 280 ꢁC for 5 min. The
injector temperature was 280 ꢁC. The MS indicator was operated in
the electron impact mode at electron energy of 70 eV, and the ion
source temperature was kept at 180 ꢁC. The detection of mass
spectra was carried out using a NIST library.
BDxPETs copolyesters were synthesized using DMT, EG, PBPBD
as raw materials via transesterification and polycondensation. Zinc
acetate and tetrabutyl titanate were used as the catalysts for the
transesterification and polymerization, respectively. The prepara-
tion procedure is shown in Scheme 2.
The route for preparation of BDxPETs containing 5 mol% PBPBD
is presented here as a representative procedure: DMT (30.6 g,
0.1576 mol), EG (14.6 g, 0.2364 mol), PBPBD (2.95 g, 0.0078 mol),
zinc acetate (0.08 g, 4.35 ꢀ 10ꢂ4 mol), and tetrabutyl titanate so-
lution (0.05 mL, 4.35 ꢀ 10ꢂ4 mol) were added into a 100 mL
polymerization bottle equipped with mechanic stirring and nitro-
gen inlet. Then the system was heated to 210e220 ꢁC for 2 h under
the protection of N2, and the transesterification was carried out.
After that, the mixture was heated to 270 ꢁC, and the pressure of the
system was reduced to lower than 50 Pa. The polycondensation was
maintained for 3 h. Finally, the reactor was cooled to room tem-
perature at reduced pressure. Other BDxPET copolyesters were
prepared by similar process.
Cone calorimeter tests were carried out using a FTT cone calo-
rimeter according to ISO 5660-1 standard. The specimens were
prepared with a size of 100 ꢀ 100 ꢀ 3 mm3 and a weight of 35 g,
tested under a heat flux of 50 kW/m2.
The LOI measurements were performed on the Oxygen Index
Flammability Gauge (HC-2C) according to ASTM D2863-97 stan-
dard. The samples were compression molded, and cut into a stan-
dard size of 120 ꢀ 6.5 ꢀ 3.2 mm3.
The microstructures of the residual chars collected after cone
calorimeter tests were observed by scanning electron microscopy
(JEOL JSM 5900LV) with an acceleration voltage of 10 kV. A thin
layer of gold was sprayed on the surface before SEM observation.
3. Results and disscussion
2.3. Characterization
3.1. Structure characterization of BDxPETs
Intrinsic viscosities [h] of PET and BDxPETs were determined
with an Ubbelohde Viscometer at 25 ꢁC in phenol/1,1,2,2-
tetrachloroethane solution (v/v, 1/1).
The chemical structures of the resulting copolymers were
characterized using 1H-NMR and 13C NMR spectroscopy, which are
shown in Fig. 1 and Fig. 2, respectively. The chemical shifts of 4.86
and 8.17 ppm in the 1H-NMR spectrum are assigned to the ethylene
glycol and terephthalic acid units, respectively. And their corre-
sponding carbon chemical shifts are found at 57 (-CH2-CH2- in
ethylene glycol units), 128, 130 (Ar-C in terephthalic acid units), and
169 ppm (C]O in terephthalic acid units). Besides, the resonance
signals occurring at 4.50, 4.80 and 6.93e7.07 ppm in the 1H-NMR
spectrum are ascribed to PBPBD units, and associated with the
peaks at 61, 58 ppm (-CH2-CH2- in PBPBD units), 147, 146, 145, 113
and 113 ppm (Ar-C in PBPBD units) in the 13C-NMR spectrum. The
above results demonstrated that BDxPETs were successfully ob-
tained through the melt polycondensation.
The experimental content of PBPBD was calculated by the in-
tegral area ratio of Ifþgþh and Ia, and the detailed results are listed in
Table 1. It is clear that over 80% PBPBD monomer is incorporated
into the backbone of BDxPET. The loss of PBPBD may be ascribed to
its volatility during the polycondensation process at reduced
pressure. And the intrinsic viscosities [h] of BDxPETs are all more
than 0.70 dL/g (PET), which illustrates that comparatively high
molecular weight copolyesters are positively prepared.
The NMR spectra (1H, 400 MHz; 13C, 400 MHz) for BDxPETs were
obtained with Bruker AVANCE AV II-400 NMR instrument, in which
CF3COOD was used as the solvent, and tetramethylsilane was the
reference.
Differential scanning calorimetry (DSC) curves were recorded
with a TA Q200 DSC apparatus, and calibrated with pure indium
standards. The measurements were performed in aluminum pans
with weighted samples of 5.0 0.5 mg under nitrogen with a purge
flow of 50 mL/min. The test procedure is given as follows: in order
to eliminate the influence of thermal history and the effect of heat
treatment on the crystalline structure of materials, the specimens
were firstly heated at 280 ꢁC and maintained for 5 min. Subse-
quently they were cooled down to 0 ꢁC (i.e. 1st cooling), and then
reheated to 280 ꢁC (i.e. 2nd heating). All the scanning rates for this
measurement were 10 ꢁC/min.
The thermal stability and thermal degradation kinetic of the
specimens in nitrogen atmosphere were performed using
a
NETZSCH TG 209 F1 apparatus. The samples (5.0 0.5 mg) were
heated from 40 to 700 ꢁC at 10 ꢁC/min. The Ozawa-Flynn-Wall
method was used to analyze the thermal degradation kinetic pa-
rameters, and the samples (5.0 0.5 mg) were heated from 40 to
700 ꢁC at various rates: 5, 10, 20 and 40 ꢁC/min.
3.2. Melting and crystallization behavior of BDxPETs
Py-GC/MS tests were performed in a CDS5200 pyrolyzer. The
pyrolysis chamber was under helium atmosphere, and the samples
The thermal transition behavior of BDxPETs and PET were
investigated using DSC. The glass transition temperature (Tg), the
melting temperature (Tm), the crystallization temperature (Tc), and
(300 m
g) were heated from ambient temperature to 500 ꢁC at a rate
of 1000 ꢁC/min and kept for 20 s. The pyrolyzer was coupled with
DANI MASTER GC-TOF-MS System, and helium was used as the
carrier gas. For the operation, the temperature program of the
the corresponding enthalpies (DHm and DHc) are listed in Table 2
and Fig. 3 shows the second heating (a) and the first cooling
Scheme 2. Preparation process for BDxPET, x denotes the molar content of PBPBD in the copolyester.

24
D.-M. Guo et al. / Polymer 77 (2015) 21e31
Fig. 1. 1H-NMR spectrum of BD20PET.
Fig. 2. 13C-NMR spectra of BD20PET.
curves (b) of the samples. Considering the same thermal history of
the prepared samples, the results of the second heating and the first
cooling scans are discussed. It can be seen that, with increasing the
content of PBPBD, the crystallization peak of BDxPETs becomes
decreased crystallization. And the Tg values for BDxPETs are slightly
higher than the Tg of neat PET, which can be attributed to the fact
that the chain rigidity of the copolyester is increased due to the
aromatic PBPBD structural units. Moreover, the third monomer
destroys the regularity of the macromolecular chains, and leads to
duller and broader, while the Tc and
D
Hc decreases, indicating the
Table 1
Table 2
Basic characteristics of samples.
Thermal behavior of PET and BDxPETs obtained from DSC.
Samples
PBPBD (mol%)
Theoretical
[
h] (dL/g)
Samples
Second heating scans
First cooling scans
Experimentala
Tg (ºC)
Tm (ºC)
D
Hm (J/g)
Tc (ºC)
DHc (J/g)
PET
e
e
0.70
0.84
0.74
0.82
PET
78
83
81
81
243
238
220
210a
31.7
32.1
15.2
e
172
155
145
e
40.0
27.0
4.7
e
BD5PET
BD10PET
BD20PET
5
10
20
4.3
8.3
18.2
BD5PET
BD10PET
BD20PET
a
a
Calculated from 1H-NMR.
Measured using the Melting Point Apparatus.

D.-M. Guo et al. / Polymer 77 (2015) 21e31
25
Fig. 3. DSC thermograms of PET and BDxPETs for second heating scans (a) and first cooling scans (b) at 10 ºC/min in N2.
the crystalline property decrease [17e19]. Therefore, the Tm of
BD10PET decreases from 238 ꢁC (BD5PET) to 220 ꢁC. The Tm and Tc of
BD20PET have not been detected from DSC, which indicates that
BD20PET cannot crystallize during DSC testing.
the char forming capability of PET-based copolyesters significantly.
3.4. Thermal degradation kinetic
In order to illustrate the thermal degradation behavior of PET
and BDxPETs, Ozawa-Flynn-Wall method (Eq. (1)) is used [20e22].
3.3. Thermal stability of BDxPETs
AE
R
E
RT
The thermal stability of the samples was evaluated by ther-
mogravimetric analysis (TGA), and their TG as well as DTG curves
under nitrogen atmosphere are shown in Fig. 4. The detailed data
such as the initial decomposition temperature (T5%), the maximum
rate decomposition temperature (Tmax), and the residue at 700 ꢁC of
all samples are collected in Table 3. From Fig. 4, it is clear that all
samples exhibit a single decomposition process. As seen in Table 3,
T5%, Tmax and the residue at 700 ꢁC of copolyesters increase with
increasing PBPBD content. Introducing only 5 mol% PBPBD into PET,
the char residue increases sharply from 9.9% (PET) to 17.2%
(BD5PET). With further increasing the content of PBPBD, the char
residue of BD20PET reaches 22.5%, which achieves more than twice
as that of PET. These results demonstrate that introducing PBPBD
into the main-chain of PET can improve the thermal stability and
log10b ¼ log10 ꢂ log10FðaÞ ꢂ 2:315 ꢂ 0:4567
(1)
Where A is the pre-exponential factor, E is the activation energy,
Table 3
TGA data of samples in N2 atmosphere.
Samples
T5% (ºC)a
Tmax (ºC)b
Residues at 700 ꢁC (wt%)
PET
389.9
394.0
396.6
402.0
428.7
430.8
434.4
438.3
9.9
17.2
18.2
22.5
BD5PET
BD10PET
BD20PET
a
Temperature of 5 wt% weight loss.
Temperature at the max rate weight loss.
b
Fig. 4. (a) TG and (b) DTG curves of samples in N2 atmosphere.

26
D.-M. Guo et al. / Polymer 77 (2015) 21e31
b
is the heating rate, T is the temperature, and F (
a
) is the integral
BD20PET are very similar, which suggests that their degradation
function of the conversion. Via this method the apparent activation
energy for decomposition of a reaction can be determined without
specifically knowing its reaction mechanism. From Eq. (1), E can be
calculated from the slope of a plot of logb versus 1/T at a fixed
weight loss. Fig. S1 shows the thermal degradation curves of PET
and BDxPETs at different heating rates. The thermal stability of the
samples increases with increasing heating rates, as expected, given
the time lag for thermal conduction and volatilization of the
decomposition compounds [23]. Fig. S2 shows the plots of logb
processes partly overlapped. The assigned structures of pyrolysis
products of PET listed in Table 5 are: vinyl benzoate (peak 1),
benzoic acid (peak 2), divinyl terephthalate (peak 3), dimethyl
terephthalate (peak 4), methyl vinyl terephthalate (peak 5), [1,10-
biphenyl]-4-carbaldehyde (peak 6), [1,10-biphenyl]-4-carboxylic
acid (peak 7), ethane-1,2-diyl dibenzoate (peak 8), and butane-
1,4-diyl dibenzoate (peak 9). Based on our results and those pre-
viously reported [24e27], the proposed pyrolysis process of PET is
given in Fig. 6. It can be sure that PET breaks down at the ester links
according to the random chain scission mechanism, and the prin-
versus 1/T at 5, 10, 20, 30, 40, 50, and 60% conversion (a%), and the
fitted lines are nearly parallel as well as the correlation coefficients
are almost above 0.99 (Table 4), which indicates that this method is
valid. The calculated activation energy for decomposition of the
samples with different conversion is also listed in Table 4.
It could be found that, as the conversion and the content of
PBPBD monomer increasing, the activation energy for decomposi-
tion of the samples increases obviously. Higher activation energy
means more energy is needed to break down the chemical bonds of
the samples and their char formations become more and more
stable. The above results indicate that the incorporation of PBPBD
significantly improves the thermal stability of PET-based copo-
lyesters, which is consistent with the TGA results.
cipal point of weakness in PET is the b-methylene group.
For pyrolysis products of BD20PET, it is clear that one major
difference from those of PET is the absence of methyl vinyl tere-
phthalate (peak 5), [1,10-biphenyl]-4-carbaldehyde (peak 6), and
[1,10-biphenyl]-4-carboxylic acid (peak 7), which come from the
degradation of benzoic acid (peak 2; 20) or divinyl terephthalate
(peak 3; 30) (Fig. 6). This phenomenon illustrates that, compared to
neat PET, BD20PET will release less volatile products at high tem-
perature. The other major difference is the typical product of
BD20PET (peak 70). From its assigned structure in Table 5, it can be
deduced that this product may come from the rearrangement re-
actions of PBPBD structural units. During the decomposition pro-
cess, some PBPBD units are separated from the main-chain of the
copolyester, and then chain scissions occur at the “Ar-OCH2CH2O”
links. With the electrical effects of the oxygen atoms linking the
benzene rings, the adjacent carbon atoms become more active, and
then link with the free radicals of ethy and ethoxyl [13,28]. After
that, they undergo the dehydrogenation, ultimately forming the
3.5. Pyrolysis behavior
In order to further investigate the thermal degradation process
of BDxPET copolyesters, the pyrolysis GC/MS study was performed.
The pyrograms of BD20PET and PET are presented in Fig. 5. Major
corresponding peaks in Fig. 5 are numbered and their assignments
are listed in Table 5.
(E)-(14-methylchromeno[2,3-b]xanthen-7(14H)-ylidene)
meth-
anol (peak 70, assigned mass spectrum is shown in Fig. S3). Note-
worthily, product of peak 70 is a stable and conjugated structure
As shown in Fig. 5 and Table 5, pyrolysis products of PET and
Table 4
Decomposition Activation energy for samples determined by Ozawa-Flynn-Wall method.
a
(%)
Activation energy (kJ/mol)
Correlation coefficient R2
PET
BD5PET
BD10PET
BD20PET
PET
BD5PET
BD10PET
BD20PET
5
179
186
189
194
197
200
202
187
194
200
203
205
211
212
189
200
206
208
211
214
218
199
209
211
213
222
224
228
0.9999
0.9998
0.9992
0.9995
0.9998
0.9998
0.9996
0.9895
0.9982
0.9985
0.9990
0.9989
0.9989
0.9985
0.9920
0.9994
0.9985
0.9987
0.9971
0.9954
0.9884
0.9973
0.9992
0.9994
0.9998
0.9999
0.9998
0.9998
10
20
30
40
50
60
Fig. 5. The pyrograms of PET and BD20PET.

D.-M. Guo et al. / Polymer 77 (2015) 21e31
27
with heteroaromatic rings, and this specific proposed pyrolysis
3.6. Burning behavior of BDxPETs
process is given in Fig. 7. Combining the results of TGA and thermal
degradation kinetic, i. e. considering the high thermal stability
(5 wt% mass loss observed at 394e402 ꢁC), the high char residue
yield (17.2e22.5 wt%) and the comparatively high apparent
decomposition activation energy for BDxPETs, the formation of
these conjugated heteroaromatic structures (peak 70) will play a
flame-retardant role in condensed phase [1], which will be illus-
trated in the following parts.
The combustion performance of BDxPET copolyesters and neat
PET, including heat release rate (HRR), total heat release (THR),
smoke production rate (SPR) and total smoke production (TSP), are
plotted in Fig. 8. And the detailed data are summarized in Table 6.
The HRR, especially the peak value for HRR (PHRR), is considered as
the most important parameter in evaluating the fire safety of ma-
terials, which determines the rate of fuel-feeding and flame spread
in the combustion. A lower PHRR value denotes slower flame
Table 5
Compounds identified in the pyrograms of PET and BD20PET.
Peak NO.
1; 10
Retention time (min)
6.02
M
Assigned structure
148
2; 20
3; 30
6.98
9.89
122
218
4; 40
10.62
11.18
194
206
5
6
7
12.24
12.36
182
199
8; 50
9; 60
70
14.00
16.41
17.13
270
298
328

28
D.-M. Guo et al. / Polymer 77 (2015) 21e31
Fig. 6. Proposed pyrolysis process for PET.
spread and minor fire hazard, suggesting better flame retardance
[7,29]. In this study, the introduction of PBPBD decreases the heat
release of the copolyesters. The PHRR value for BD10PET is
734.2 kW/m2, which is 73% of the value for neat PET; the PHRR
value for BD20PET further decreases to 541.9 kW/m2, which is
nearly half of that for neat PET. Based on the HRR curves in Fig. 8 (a),
fire growth rate (FIGRA) has been calculated through PHRR/tPHRR to
assess flame spread in developing fire, the scale of a fire, and the fire
hazards of samples [30]. As well as the PHRR, the FIGRA of
copolyester is also decreased with the increase of the mole percent
PBPBD. The FIGRA value for neat PET is 9.6, and the FIGRA values for
BD5PET, BD10PET and BD20PET decrease to 9.5, 7.7 and 6.4, respec-
tively. Low FIGRA values indicate delayed times to flashover, which
allows enough time to evacuate and/or for fire extinguishers to
arrive [31]. And the THR values for BDxPET copolyesters range from
53.4 to 57.3 MJ/m2, which are close to the THR value for neat PET
(55.5 MJ/m2), retaining an acceptable level.
Smoke produced in the fire is even the most important factor
Fig. 7. Proposed pyrolysis process for PBPBD structural units in BD20PET.

D.-M. Guo et al. / Polymer 77 (2015) 21e31
29
Fig. 8. Cone calorimetric results of PET and BDxPETs: (a) HRR; (b)THR; (c)SPR; (d)TSP.
which directly puts people to death by suffocation and/or inhala-
tion of the toxic gases [32]. Thus, smoke suppression is really
important for the flame-retardant polymers. As shown in Fig. 8 (d)
and Table 6, the TSP values for BD5PET, BD10PET and BD20PET are
10.2 m2, 10.4 m2, and 13.0 m2, respectively. These values are much
lower than the TSP value for neat PET (18.8 m2), which illustrates
that the incorporation of PBPBD endows BDxPETs with good smoke
inhibition property. One reasonable interpretation should be that
PBPBD structural units can ultimately form conjugated hetero-
aromatic structures, and release less volatile products compared to
PET at high temperature, which suppresses the smoke generation
effectively.
The limiting oxygen index (LOI) of the specimens was measured
to estimate their flame retardance. The detailed LOI values are lis-
ted in Table 6. It can be seen that neat PET is a flammable material
with a low LOI value of 22.0, and with the incorporation of PBPBD,
the LOI value for BD5PET increases to 26.1. Further increasing the
content of PBPBD, the LOI value for BD20PET reaches 32.0, meaning
the copolyester obtains expected flame retardance. Fig. 9 shows the
digital images of specimens after LOI test, serious melt-dripping
and few char residue from combustion can be observed for neat
PET. In contrast, apparently intumescent char formation and no
melt-dripping phenomenon are observed for all the three samples
of BDxPETs.
heteroaromatic structures through the rearrangement reactions of
PBPBD structural units that promoted plenty of char formation
during combustion, which can isolate the oxygen and the heat,
suppress the smoke production simultaneously, as well as inhibit
the dripping effectively.
3.7. Char residue morphology
To investigate the morphology of the char residues of BDxPETs,
SEM analysis was carried out. Inner surface images of char residues
of PET and BD20PET after cone calorimeter measurements are given
in Fig. 10. We can observe clearly that the char residue morphology
of neat PET is smooth and with scattered ashes (Fig. 10 (a)).
Compared with neat PET, in Fig. 10 (b), BD20PET presents rougher
surface, and many grass-shape objects can be observed. After
Table 6
Cone calorimetry and LOI results for PET and BDxPETs.
Samples
PET
BD5PET
BD10PET
BD20PET
PHRR (kW/m2)
Av-HRRa (kW/m2)
THR (MJ/m2)
TSP (m2)
1004.5
104.7
55.5
18.8
105
950.1
105.0
56.1
10.2
100
734.2
93.7
53.4
10.4
95
541.9
89.5
57.3
13.0
85
tPHRRb (s)
All the results above-mentioned demonstrate that the incor-
poration of PBPBD can lower the HRR, FIGRA and TSP values while
raise the LOI values for the copolyesters, providing outstanding
flame retardant performance. Combining the pyrolysis GC/MS re-
sults, we can deduce that it is the formation of conjugated
FIGRAc
9.6
9.5
7.7
6.4
LOI (vol%)
22.0
26.1
28.4
32.0
a
Av-HRR, the average value of HRR.
tPHRR means Time to PHRR.
Fire Growth Rate, calculated by dividing the PHRR by the tPHRR
b
c
.

30
D.-M. Guo et al. / Polymer 77 (2015) 21e31
Fig. 9. Digital images of BDxPETs and PET after LOI test.
Fig. 10. SEM images (500 ꢀ ; zoom in part, 10000 ꢀ ) of the inner surfaces of residues: (a) PET; (b) BD20PET.
magnification, it is clear that these grass-shape objects are actually
“stick-shaped” char formations with a size of nearly 5 m long.
dripping but obvious intumescent char formation was observed
after LOI test. Despite the absence of any flame retardant element,
BDxPETs copolyesters exhibited outstanding flame retardant per-
formance and char forming capability. This modification might
provide a new solution to flame retarded polymers, which is more
friendly to environment and human body than traditional methods.
m
These unique “stick-shaped” char residues should be ascribed to
the formation of conjugated heteroaromatic structures via the
rearrangement reactions of PBPBD structural units during com-
bustion. And these “stick-shaped” char formations exhibit good
heat insulation and smoke inhibition, providing outstanding flame-
retardant effects in a fire. Regrettably, like neat PET, BD20PET con-
tains only carbon, hydrogen and oxygen, thus it is difficult to further
investigate the composition of its char residue from combustion.
Acknowledgments
This work was financially supported by the National Natural
Science Foundation of China (50933005, 51421061 and J1103315),
Program for Changjiang Scholars and Innovative Research Team in
University of Ministry of Education of China (IRT1026).
4. Conclusion
In this study,
a monomer containing arylene-ether units
(PBPBD) and a series of novel flame-retardant-element-free PET
based copolyesters using PBPBD as a comonomer (BDxPETs) were
synthesized successfully. Their chemical structures were confirmed
using 1H-NMR, 13C-NMR spectroscopy. With increasing the content
of PBPBD, the thermal stability and the char residue from com-
bustion of BDxPETs increased dramatically, whereas the crystal-
linity and melting temperature of these copolymers decreased. The
results of thermal degradation kinetic illustrated that the apparent
decomposition activation energy for the copolyesters was
enhanced with increasing the conversion and the content of PBPBD.
The pyrolysis investigation showed that PBBPD structural units
undergo rearrangement reactions at high temperature, ultimately
forming conjugated heteroaromatic structures. And this formation
made BDxPETs have plenty of residues with special “stick-shaped”
char morphology. Thanks to these efficient rearrangement and
carbonization reactions promoted by PBPBD units, the flame
retardance and smoke suppression of BDxPETs were simulta-
neously enhanced significantly. Meanwhile, for BDxPETs, no melt-
Appendix A. Supplementary data
Supplementary data related to this article can be found at 10.
1016/j.polymer.2015.09.016.
References
[1] L. Chen, Y.Z. Wang, Materials 3 (10) (2010) 4746e4760.
[2] Y. Deng, Y.Z. Wang, Z.J. Zong, X.H. Liu, X.H. Du, Chin. J. Polym. Sci. 26 (01)
(2008) 111e116.
[3] S.C. Yang, J.P. Kim, J. Appl. Polym. Sci. 106 (5) (2007) 2870e2874.
[4] X.K. Jing, X.G. Ge, X. Xiang, C. Wang, Z. Sun, L. Chen, Y.Z. Wang, Polym. Int. 58
(10) (2009) 1202e1208.
[5] X.K. Jing, X.S. Wang, D.M. Guo, Y. Zhang, F.Y. Zhai, X.L. Wang, L. Chen,
Y.Z. Wang, J. Mater Chem. 1 (32) (2013) 9264e9272.
[6] Y. Zhang, L. Chen, J.J. Zhao, H.B. Chen, M.X. He, Y.P. Ni, J.Q. Zhai, X.L. Wang,
Y.Z. Wang, Polym. Chem. 5 (2014) 1982e1991.
[7] H.B. Zhao, L. Chen, J.C. Yang, X.G. Ge, Y.Z. Wang, J. Mater Chem. 22 (37) (2012)
19849e19857.
[8] H.B. Zhao, B.W. Liu, X.L. Wang, L. Chen, X.L. Wang, Y.Z. Wang, Polymer 55 (10)
(2014) 2394e2403.

D.-M. Guo et al. / Polymer 77 (2015) 21e31
31
[9] H.B. Zhao, X.L. Wang, Y. Guan, X.L. Wang, L. Chen, Y.Z. Wang, Polymer 70
(2015) 68e76.
[21] H. Zhao, Y.Z. Wang, D.Y. Wang, B. Wu, D.Q. Chen, X.L. Wang, K.K. Yang, Polym.
Degrad. Stab. 80 (1) (2003) 135e140.
[10] H. Han, P.K. Bhowmik, Prog. Polym. Sci. 22 (7) (1997) 1431e1502.
[11] T.J. Dingemans, E. Mendes, J.J. Hinkley, E.S. Weiser, T.L. StClair, Macromole-
cules 41 (7) (2008) 2474e2483.
[12] P. Patel, T.R. Hull, R.E. Lyon, S.I. Stoliarov, R.N. Walters, S. Crowley,
N. Safronava, Polym. Degrad. Stab. 96 (1) (2011) 12e22.
[13] P. Patel, T.R. Hull, R.W. McCabe, D. Flath, J. Grasmeder, M. Percy, Polym.
Degrad. Stab. 95 (5) (2010) 709e718.
[14] G.W. Yeager, D.N. Schissel, Synthesis 1 (1991) 63e68.
[15] Y.M. Sun, C.S. Wang, J. Polym. Sci. Pol. Chem. 34 (9) (1996) 1783e1792.
[16] N. Lotti, M. Colonna, M. Fiorini, L. Finelli, C. Berti, Polymer 52 (4) (2011)
904e911.
[17] H.B. Chen, L. Chen, Y. Zhang, J.J. Zhang, Y.Z. Wang, Phys. Chem. Chem. Phys. 13
(23) (2011) 11067e11075.
[22] B. Wu, Y.Z. Wang, X.L. Wang, K.K. Yang, Y.D. Jin, H. Zhao, Polym. Degrad. Stab.
76 (3) (2002) 401e409.
[23] H.B. Chen, X. Dong, D.A. Schiraldi, L. Chen, D.Y. Wang, Y.Z. Wang, J. Anal. Appl.
Pyrol. 99 (2013) 40e48.
[24] L. Buxbaum, Angew. Chem. Int. Ed. 7 (3) (1968) 182e190.
[25] H. Zimmerman, N.T. Kim, Polym. Eng. Sci. 20 (10) (1980) 680e683.
[26] H.A. Pohl, J. Am. Chem. Soc. 73 (12) (1951) 5660e5661.
[27] H.B. Chen, Q. Zhou, X. Dong, Y. Zhang, L. Chen, Y.Z. Wang, Polym. Degrad. Stab.
97 (6) (2012) 905e913.
[28] C. Tsai, L. Perng, Y. Ling, Rapid Commun. Mass Sp. 11 (18) (1997) 1987e1995.
[29] B. Schartel, M. Bartholmai, U. Knoll, Polym. Degrad. Stab. 88 (3) (2005)
540e547.
[30] H. Breulet, T. Steenhuizen, Polym. Degrad. Stab. 88 (1) (2005) 150e158.
[31] Y.W. Yan, L. Chen, R.K. Jian, S. Kong, Y.Z. Wang, Polym. Degrad. Stab. 97 (8)
(2012) 1423e1431.
[18] H.B. Chen, Y. Zhang, L. Chen, W. Wang, B. Zhao, Y.Z. Wang, Polym. Advan
Technol. 23 (9) (2012) 1276e1282.
[19] H.B. Chen, J.B. Zeng, X. Dong, L. Chen, Y.Z. Wang, CrystEngComm 15 (14)
(2013) 2688.
[32] T.R. Hull, B.K. Kandola, Fire Retardancy of Polymers: New Strategies and
Mechanisms, Royal Society of Chemistry, Cambridge, 2009.
[20] T.J. Ozawa, Therm. Anal. Calorim. 2 (3) (1970) 301e324.