Thermal and flame retardant properties of transparent UV-curing epoxy acrylate coatings with POSS-based phosphonate acrylate

Article abstract of DOI:10.1039/c5ra11805k

A novel branched phosphonate acrylate monomer (BPA) containing phosphorus and polyhedral oligomeric silsesquioxanes (POSS) was successfully synthesized via thiol Michael addition of tri(acryloyloxyethyl) phosphate with octamercaptopropyl POSS, and then incorporated into epoxy acrylate (EA) resins in different ratios using ultraviolet curing technology. The structure of BPA was confirmed by Fourier transform infrared spectroscopy (FTIR) and nuclear magnetic resonance. The effects of BPA on the thermal stability, combustion performance, degradation process and transparency of EA were investigated. The modified EA (MEA) retained a high transparency and exhibited significantly enhanced flame retardancy, as evidenced by the increased limiting oxygen index values and greatly reduced peak heat release rate. The thermal properties of MEA indicated that the presence of BPA promoted the degradation of the EA matrix and produced additional char residues. The thermal degradation processes of the MEA were further investigated by real time FTIR. The char structure of MEA, investigated by Raman spectra, revealed that the addition of BPA increased the ratios of graphitized carbon in residual chars, thus enhancing the thermal stability of MEA.

Full text of DOI:10.1039/c5ra11805k

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Thermal and flame retardant properties of
transparent UV-curing epoxy acrylate coatings with
POSS-based phosphonate acrylate
Cite this: RSC Adv., 2015, 5, 75254
Bin Yu,ac Youji Tao,b Lu Liu,e Yongqian Shi,ac Hongyu Yang,ac Ganxin Jie,b
d
ac
a
ac
*
*
Siuming Lo, Qilong Tai,
Lei Song and Yuan Hu
A novel branched phosphonate acrylate monomer (BPA) containing phosphorus and polyhedral oligomeric
silsesquioxanes (POSS) was successfully synthesized via thiol Michael addition of tri(acryloyloxyethyl)
phosphate with octamercaptopropyl POSS, and then incorporated into epoxy acrylate (EA) resins in
different ratios using ultraviolet curing technology. The structure of BPA was confirmed by Fourier
transform infrared spectroscopy (FTIR) and nuclear magnetic resonance. The effects of BPA on the
thermal stability, combustion performance, degradation process and transparency of EA were
investigated. The modified EA (MEA) retained a high transparency and exhibited significantly enhanced
flame retardancy, as evidenced by the increased limiting oxygen index values and greatly reduced peak
heat release rate. The thermal properties of MEA indicated that the presence of BPA promoted the
degradation of the EA matrix and produced additional char residues. The thermal degradation processes
of the MEA were further investigated by real time FTIR. The char structure of MEA, investigated by
Raman spectra, revealed that the addition of BPA increased the ratios of graphitized carbon in residual
chars, thus enhancing the thermal stability of MEA.
Received 19th June 2015
Accepted 25th August 2015
DOI: 10.1039/c5ra11805k
www.rsc.org/advances
the range of applications, especially where high ame retard-
ancy is required. Thus, to widen the application of EAs,
1. Introduction
improving its ame retardancy and reducing the re hazards
are imperative.
Ultraviolet (UV)-curing technology is widely used in many elds,
ranging from polymeric coatings and composites to inks and
adhesives, as a means of rapidly and efficiently converting
liquid, multifunctional monomer resins into cross-linked
polymer networks.1–3 The advantages of photo-curable systems
have been well documented, including environmental friendli-
ness (without VOC emissions), a fast and controllable curing
system, lower energy consumption than thermal curing, and the
possibility of curing at room temperature.4,5 Among UV-cured
resins, much attention has focused on epoxy acrylates (EAs),
due to their excellent adhesion, good chemical and corrosion
resistance, and excellent electrical insulation. However, am-
mability is one of the principal disadvantages of EAs, restricting
Generally, incorporating phosphorus-,6 silicon-,7 nitrogen-8
and boron-containing compounds,9 nanoadditives,10 etc. into
EA matrix is an effective way of improving its ame retardance.
Of the different ame retardants (FRs) discussed in the litera-
ture, phosphorus-containing FRs, which promote char forma-
tion in the condensed phase through extracting water from the
pyrolysis substrate, have been demonstrated to be effective
ame retardants for most polymeric materials, such as poly-
propylene,11 polystyrene,12 and epoxy resins.13 However, the
incorporation of phosphorus-bearing FRs alone, especially
organic phosphorus compounds, will lead to the formation of
more unstable char residues, unable to resist the heat from
res. Therefore, synergistic combinations of phosphorus and
other inorganic additives or organic–inorganic hybrids have
been developed and reported to impart high thermal oxidation
resistance to the resulting char layers.14,15
aState Key Laboratory of Fire Science, University of Science and Technology of China,
96 Jinzhai Road, Hefei, Anhui 230026, P. R. China
bState Key Laboratory of Environmental Adaptability for Industrial Products, China
National Electric Apparatus Research Institute, Guangzhou, 510300, P. R. China
cUSTC-CityU Joint Advanced Research Centre, Suzhou Key Laboratory of Urban Public
Safety, Suzhou Institute for Advanced Study, University of Science and Technology of
China, 166 Ren'ai Road Suzhou, Jiangsu 215123, P. R. China. E-mail: yuanhu@
ustc.edu.cn; qltai@ustc.edu.cn; Fax: +86-551-3601664; Tel: +86-551-3601664
dDepartment of Building and Construction, City University of Hong Kong, Tat Chee
Avenue, Kowloon, Hong Kong, P. R. China
Polyhedral oligomeric silsesquioxanes (POSS), possessing a
cage-like structure with 1–3 nm particle size, are a class of
inorganic compounds. Their empirical formulas may be pre-
sented as (RSiO1.5)n, where R may be hydrogen or any alkyl,
alkylene, aryl, arylene or organo-functional derivative of alkyl,
alkylene, aryl or arylene groups.16–18 In prior work, POSS was
demonstrated to improve the char yield, the oxidation
eNano Science and Technology Institute, University of Science and Technology of
China, 166 Ren'ai Road, Suzhou, Jiangsu 215123, P. R. China
75254 | RSC Adv., 2015, 5, 75254–75262
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
resistance properties and ame retardancy of most poly- infrared spectrometry (TG-IR). The combustion properties
mers.14,19–21 However, the effect of POSS on the thermal behavior were evaluated by limiting oxygen index (LOI) and micro
and ame retardancy of the UV-cured epoxy acrylates have rarely combustion calorimeter (MCC). The transparency and dynamic
been reported because of poor compatibility between EA mechanical properties were investigated by UV-Vis spectra and
matrices and POSS. In addition, most POSS with varied struc- dynamic mechanical analysis (DMA). This work describes a
tures are additive, solid compounds and the presence of POSS good reactive ame retardant for UV-cured transparent resins.
may have a negative effect on the curing rate and degree of
curing, as well as the transparency of the resultant cured lms.
2. Experimental
In this work, ame retardant and transparent UV-cured EA
2.1 Raw materials
composites lled with different loadings of phosphorus- and
silicon-containing compound (BPA) were prepared using UV-
curing technology, where the compound was a liquid ame
retardant synthesized via thiol Michael addition. The structure
of BPA was characterized by Fourier transform infrared spec-
troscopy (FTIR) and nuclear magnetic resonance (NMR). The
thermal degradation of the as-prepared lms was investigated
by thermogravimetric analysis (TGA), real-time Fourier trans-
form infrared spectra (RT-IR) and thermogravimetric analysis/
Epoxy acrylate resin (EB600) was purchased from Cytec Indus-
tries Inc. Hydroxylethyl acrylate (HEA) was supplied by Beijing
Orient Chemical Co. Ltd. (3-Mercaptopropyl) trimethoxysilane
(MPTS) was obtained from Nanjing Shuguang Chemical Group
Co., Ltd. 2-Hydroxy-2-methyl-1-phenyl-1-propanone (Dar-
ocur1173), supplied by Ciba Specialty Chemicals, was used as a
photoinitiator. All the other chemicals were obtained from
China National Pharmaceutical Group (Shanghai, China).
Scheme 1 Synthetic routes of (a) TAEP, (b) OMP-POSS and (c) BPA.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 75254–75262 | 75255

View Article Online
RSC Advances
Paper
Phosphorus oxychloride (POCl3) and MPTS were distilled prior Co., China)). The formulations of the cured EA lms with
˚
to use. Triethylamine and HEA were distilled and dried over 4 A various BPA contents are shown in Table 1.
molecular sieves before use. Other reagents were used as-
received without further purication.
2.4 Characterizations
FTIR spectra (KBr) were collected on OMP-POSS and BPA using
a Nicolet 6700 FTIR spectrophotometer in the sample concen-
2.2 Synthesis of POSS-based branched phosphonate acrylate
(BPA)
tration range of 0.1–1.0 wt%.
1H NMR (400 MHz) spectra were measured with a Bruker
Tri(acryloyloxyethyl) phosphate (TAEP) was prepared via
substitution reaction between phosphorus oxychloride and
hydroxyethyl acrylate in our laboratory,6 and OMP-POSS was
AV400 NMR spectrometer, using chloroform-d6 as solvent.
LOI measurement was performed on a HC-2 LOI testing
synthesized by hydrolysis of 3-mercaptopropyl trimethoxysilane
device (Jiangsu, China) according to ASTM D2863 (the tested
specimens: 100 Â 6.5 Â 3 mm3).
under acid conditions according to the method described in
prior literature.22 BPA was obtained in an 8 : 1 molar ratio of
UL-94 vertical burning test was performed on a CZF-II hori-
TAEP to OMP-POSS through Michael addition, using triethyl-
zontal and vertical burning tester (Jiang Ning Analysis Instru-
amine as a catalyst. The synthetic route of (a) TAEP, (b) OMP-
POSS and (c) BPA were presented in Scheme 1. The typical
process for the preparation of BPA was as follows: TAEP (30.9 g,
78.7 mmol) and an appropriate amount of triethylamine (1 wt%
as total weight) were dissolved in 120 ml of dichloromethane
and then introduced into a 250 ml three-neck ask equipped
ment Company, China). The specimens used were 127 Â 12.7 Â
3 mm3 according to the standard ASTM D3801-2010.
TGA of samples was conducted with a Q5000 thermal
analyzer (TA Co., USA) from 50 to 700 ꢀC at a heating rate of
20 C minÀ1. The samples were run in triplicate: the tempera-
ꢀ
ture reproducibility of the instrument was 0.18 ꢀC, while the
mass reproducibility was 0.2%.
with
a dropping funnel, nitrogen inlet, condenser and
mechanical stirrer. Aerward, a solution of OMP-POSS (10 g,
9.84 mmol) in dichloromethane (30 ml) was added dropwise at
room temperature under magnetic stirring over a period of 1 h.
Aer completion of the addition, the reaction was conducted at
40 ꢀC under nitrogen for 24 h. Finally, the transparent and
viscous liquid was collected by evaporation under reduced press
to remove the solvent. (Yield: 98%).
DMA was measured using a DMA ꢀQ800 apparatus (TA
5 C minÀ1. The frequency of dynamic oscillatory loading was
Instruments Inc., USA) from À65 to 150 C at a heating rate of
ꢀ
10 Hz.
Small scale ammability tests were conducted on a pyrolysis-
combustion ow calorimeter (PCFC, Model MCC-2, Govmark,
Farmingdale, New York). The samples were tested according to
ASTM D7309-07.
TG-IR of the samples was performed using a TGA Q5000 IR
thermogravimetric analyzer that was interfaced to the Nicolet
6700 FTIR spectrophotometer. The sample was put in an
alumina crucible and heated from 30 to 600 ꢀC at a heating rate
of 20 ꢀC minÀ1 under nitrogen. The stainless steel transfer pipe
and gas cell were heated at 230 ꢀC to avoid the condensation of
volatile compounds.
OMP-POSS: 1H NMR (400 MHz, CDCl3, d): 0.65–0.85 (Si–CH2–),
1.33–1.47 (–SH), 1.72 (–CH2–), 2.50–2.60 (–CH2–S);
BPA: 1H NMR (400 MHz, CDCl3, d): 0.65–0.85 (Si–CH2–),
1.60–1.80 (–CH2–), 2.50–2.60 (–CH2–S), 5.8–6.5 (CH2]CH–),
4.0–4.55 (–OCH2CH2O–), 2.71–2.85 (–C(O)–CH2–), 2.60–2.71
(–C(O)CH2–CH2–);
2.3 Preparation of UV-cured EA/BPA lms
RT-IR was recorded using the Nicolet 6700 FT-IR spectro-
photometer, equipped with a ventilated oven with an additional
heating device. The cured samples were mixed with KBr powder,
and the mixture was pressed into a disc, which was then placed
UV-cured epoxy acrylate with various amounts of BPA was
designated as EAX, where X% represents the mass percentage of
BPA. Briey, the general procedure for preparing EA10 lm was
described below: BPA (2 g) was thoroughly dissolved in an
appropriate amount of dichloromethane under sonication, and
then introduced into a vial. Subsequently, EB600 (18 g) and
Darocur1173 (0.4 g, 2 wt%) were added into this solution. Aer
thorough stirring to form a homogeneous sꢀolution, the blended
was cast onto glass plates and dried at 60 C under vacuum to
remove the solvent. The cured lm was obtained when the
mixture was exposed to UV irradiation ((80 W cmÀ2, Lantian
into the oven. The tempÀer1ature of the oven was raised at a
ꢀ
heating rate of 10 C min
.
Raman spectra in the range of 500–2000 cmÀ1 were con-
ducted on a LabRAM-HR Confocal Raman Microprobe (Jobin
Yvon Instruments, France) using a 514.5 nm argon ion laser.
The transparency of the cured lms was evaluated with a
DUV-3700 UV-Vis spectrometer (Sahimadzu, Japan) in trans-
mission mode. The wavelengths studied were in the range of
400 to 800 nm.
Table 1 Formulations of the cured EA films with various BPA contents
Sample
EB600 (g)
BPA (g)
1173 (wt%)
3. Results and discussion
3.1 Structural characterization
EA0
100
90
80
—
10
20
30
2
2
2
2
EA10
EA20
EA30
The structures of OMP-POSS and BPA were characterized by
1
FTIR, and H NMR spectra. Fig. 1 presents the FTIR spectra of
70
OMP-POSS and BPA. In the prior literature, the absorption
75256 | RSC Adv., 2015, 5, 75254–75262
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
are assigned to the protons of thiol groups, reveals the
successful reaction of the thiol groups. Moreover, two new
signals in the range of 2.71–2.85 and 2.60–2.71, corresponding
to the protons of the methylene groups in –C(O)–CH2– and
–C(O)CH2–CH2–, respectively, occur in the 1H NMR spectrum of
the BPA. The above results clearly support that all thiol moieties
of OMP-POSS have participated in the thiol-acrylate reaction
and BPA was successful synthesized.
3.2 Thermal and ame retardant properties of composite
lms
Thermogravimetric analysis (TGA) is a widely used tool to
evaluate the thermal degradation behaviors of polymeric
materials. The TGA curves of the cured lms under nitrogen
and air atmosphere are shown in Fig. 3 and 4, respectively, and
the corresponding data are listed in Table 2. The initial degra-
dation temperature (T0.1) of the samples is the temperature
where 10% mass loss occurs, and Tmax represents the temper-
ature of the maximum mass loss rate. As can be seen from
Fig. 3, the control EA lm exhibits a one-step degradation
process in the temperature range of 350–500 ꢀC, with a Tmax of
424 ꢀC, which can be attributed to the degradation of the
principal EA chains. The initial degradation temperature and
char yields at 700 ꢀC of EA0 were 371 ꢀC and 12.9 wt%,
respectively. With the incorporation of BPA, all the composites
exhibit similar decomposition behaviors to that of pure EA, but
the T0.1 and Tmax are both lower than those of EA0. Moreover,
T0.1 gradually decreases with an increase in BPA content. A
Fig. 1 FTIR spectra of BPA and OMP-POSS.
bands may be observed in the FTIR spectra of OMP-POSS,
namely alkyl C–H vibration (2927, 2853 cmÀ1 for n C–H), S–H
stretching vibration (2560 cmÀ1), Si–C asymmetric stretching
vibration (1260 cmÀ1), Si–O–Si asymmetric stretching vibration
(1120 cmÀ1), Si–C stretching vibration (694 cmÀ1), and POSS
skeletal deformation vibrations (560 cmÀ1).22 For BPA, the
absorption peaks include some main peaks of TAEP and OMP-
POSS. The peak at 1270 cmÀ1 is attributed to P]O vibration; the
peaks which appear at 1038 and 980 cmÀ1 are attributable to
P–O–C vibration; the peak at 1726 cmÀ1 is assigned to C]O
vibration, while the peaks at 1635, 1410 and 810 cmÀ1 are
ascribed to C]C vibration.6 These results conrm the structure
of BPA. Moreover, no peak at 2560 cmÀ1 originating from the
S–H vibration in OMP-POSS is visible, indicating the occurrence
of thiol-acrylate Michael addition reaction.
The 1H NMR spectra of OMP-POSS and BPA are presented in
Fig. 2. The absence of the peaks around 1.33–1.47 ppm, which
Fig. 2 1H NMR spectra of OMP-POSS and BPA.
Fig. 3 TGA curves of the cured films under nitrogen atmosphere.
RSC Adv., 2015, 5, 75254–75262 | 75257
This journal is © The Royal Society of Chemistry 2015

View Article Online
RSC Advances
Paper
trend as the char yields do, due to the formation of increased
char and the barrier effect of char layer, which reduces the
release of combustible gases, thus reducing the heat release
rate.
The degradation behaviors of all the cured lms under air
(Fig. 4) are different from those under nitrogen. The control EA
lm exhibits a two-stage degradation process in the tempera-
ture ranges of 300–400 ꢀC and 600–700 ꢀC. The rst stage is
attributed to the degradation of the main EA chains and the
second is the oxidation of the unstable char residues. Aer the
incorporation of BPA, the T0.1 of the composite lms shis to a
lower temperature and is gradually reduced with an increase in
additives. Moreover, the temperature of maximum degradation
rate at the rst stage (Tmax1) displays a similar change trend to
the T0.1 of the cured lms. However, the incorporation of BPA
enhances the temperature of the second-stage ꢀ maximum
degradation rate (Tmax2) and the char yields at 700 C of cured
lms. For EA30 containing 30 wt% of BPA, the Tmax and char
yields are 652 ꢀC and 16.12 wt%, respectively, much higher than
those for EA0 (601 ꢀC, 0.46 wt%). These results indicate that the
degradation products of ame retardant in the condensed
phase improve the thermal oxidative resistance of residual
chars, thus enhancing the ame retardancy of the cured lms.
DMA was used to study the dynamic mechanical properties
of the UV-cured ame retardant coatings. Fig. 5 presents (a)
storage modulus and (b) tan d curves of the cured EA coatings.
The storage modulus of the cured ame retardant coatings
shows gradually increasing trend with increasing the concen-
tration of BPA. The storage modulus at 30 ꢀC for pure EA lm is
Fig. 4 TGA curves of the cured films under air atmosphere.
Table 2 TGA data for the cured films
T0.1 (ꢀC)
Tmax (ꢀC)
Char (700 ꢀC, wt%)
Sample
N2
Air
N2
Air
N2
Air
EA0
371
354
336
323
360
360
339
325
424
378
347
328
424, 601
379, 621
347, 651
329, 652
12.9
26.5
33.2
34.8
0.46
EA10
EA20
EA30
6.29
13.86
16.16
similar phenomenon has been observed in prior research and
this decrease in T0.1 may be explained by the early degradation
of unstable P–O–C bonds of BPA, along with the formation of
some acid derivatives.22,23 These acid derivatives will catalyze
polymer degradation to produce additional char residues.
Additionally, the aliphatic chains, which radiate out from the
POSS core and relatively poor thioether linkages lead to the low
thermal stability of MEA, while OMP-POSS degrades to form a
ceramic-like layer, which reinforces the forming char.24 In Table
2, the char formation in nitrogen of the cured lm gradually
increases as the BPA content increases from EA0 to EA30. For
example, the char yield of EA30 is as high as 34.8%, higher than
that of EA0, suggesting the improved thermal stability of EA30
at high temperature. The signicant increase in char yield is
due to the synergistic effect of phosphorous and silicon, where
phosphorous promotes char formation, while silicon reinforces
the char retarding further degradation of char residue. In the
DTG curves, the mass loss rate exhibits the same reduction
Fig. 5 (a) Storage modulus and (b) tan d curves of the cured EA films.
75258 | RSC Adv., 2015, 5, 75254–75262
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
around 1890 MPa. With adding 30 wt% BPA, the storage release rate (PHRR) is lower than that of pure EA lm. This
modulus of EA30 is improved by 56.6%, compared to that of result is caused by the early degradation of the relatively weak
pure EA lm, due to the nanoreinforcement of the POSS cages phosphate moieties, which promotes the formation of more
in the crosslinked polymer networks since the POSS cages were stable phosphorus–carbon–silicon complexes in the solid
covalently bonded with the crosslinking networks of the epoxy phase.23 Thus, the formation of more stable char layers helps
acrylate.25,26 Glass transition temperature (Tg) is determined decrease the release amounts and release rate of evolved
from the peak maximum of the tan d curve. The incorporation combustible gases, thus contributing to the low THR and
of BPA has little effect on the Tgs of the cured EA lms. The PHRR. The ame retardancy of cured lms is further investi-
reason for this phenomenon is that the exibility of organic gated by LOI and UL-94 tests, and the resulting values are listed
branches present in the POSS structure and the decrease in the in Table 3. Owing to the easy burning of virgin EA, a low LOI
crosslink density leads to the decrease of Tgs, whereas the value of 22.0% is obtained. As the BPA increases to 30 wt%, the
nanoreinforcement of POSS cages restricting the movement of LOI of the cured lms is gradually increased to a maximum of
the molecule chains in the matrix, thus increasing the Tgs.
29.0%, which is in agreement with the increased char residues
To estimate the ame retardancy of composite lms, MCC under nitrogen. However, no ratings (NRs) in the UL-94 tests are
was performed on a series of cured lms. Fig. 6 presents the observed for the FR EA systems. Consequently, the additive
heat release rate (HRR) curves of various cured lms versus loading level can not improve the vertical burning rating of the
temperature. The detailed data obtained from the MCC curves, system effectively.
including the peak heat release rate (PHRR), total heat release
(THR) and heat release capacity (HRC), are summarized in
Table 3. Neat EA lm is highly ammable with high PHRR and
THR values, which are 296 W gÀ1 and 20.5 kJ gÀ1, respectively.
Incorporating BPA into EA remarkably reduces the PHRR and
THR values. With the addition of 30 wt% BPA, the PHRR and
THR of the cured lm are decreased by 50.8% and 29.8%,
respectively, relative to those of pure EA. Moreover, it is worth
noting that the heat for the composite lms starts to release at a
relatively low temperature, and the temperature to peak heat
3.3 Gaseous and condensed phase analysis
TG-FTIR technique was used to investigate the evolved gas
products of ame retardant EA coatings during thermal
degradation process. The FTIR spectra of pyrolysis gaseous
products of EA0 and EA30 at maximum degradation rate are
presented in Fig. 7. The FTIR spectrum of EA30 is similar to that
of EA0, implying similar volatile compounds. The absorption of
gaseous products from the pyrolysis of BPA is not clearly
observed, probably due to the overlapping of the absorptions of
the degradation products from the EA matrix. The evolved
volatile products of EA0 and EA30 are identied by the char-
acteristic bands of water and/or phenol (3500–3750 cmÀ1),
various hydrocarbons (2950–2850 cmÀ1), CO2 (2250–
2400 cmÀ1), CO (2100–2200 cmÀ1), carbonyl-containing
compounds (1700–1760 cmÀ1
(1480–1650 cmÀ1).27
) and aromatic compounds
The absorbance of selected pyrolysis products vs. tempera-
ture for EA0 and EA30 is presented in Fig. 8. It can be observed
that all the pyrolysis products for EA30 begin to release earlier
than that of pure EA, due to the low thermal stability of BPA.
Fig. 6 HRR curves of various cured films versus temperature.
Table 3 MCC and UL-94 test results and LOI values of various cured
films
HRC
PHRR
THR
Sample
(J gÀ1 KÀ1
)
(W gÀ1
)
(kJ gÀ1
)
LOI (%)
UL-94
EA0
294
251
226
145
296.0
252.5
226.2
145.7
20.5
19.0
17.3
14.4
22.0
27.0
28.5
29.0
NR
NR
NR
NR
EA10
EA20
EA30
Fig. 7 FTIR spectra of pyrolysis gaseous products of EA0 and EA30 at
maximum degradation rate.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 75254–75262 | 75259

View Article Online
RSC Advances
Paper
Fig. 8 Absorbance of selected pyrolysis products vs. temperature for EA0 and EA30.
Moreover, BPA catalyzes the thermal degradation of the EA pyrolysis temperatures. It may be seen that the cured EA lm
matrix. The absorbance intensity of the degradation products presents some characteristic absorption peaks at approximately
for EA30 is much lower than that for EA0, especially for these 3450 cmÀ1 (O–H stretching vibration), 2970 cmÀ1 (stretching
ammable gases (hydrocarbons, aromatic and carbonyl vibration of C–H), 1730 cmÀ1 (C]O stretching vibration), 1610,
compounds). The reduction in the release of these organic 1510 cmÀ1 (C–C stretching vibration of aromatic ring), 1460,
volatiles is attributed to the formation of additional char resi- 1380 cmÀ1 (bending vibration of C–H), 1240 cmÀ1 (stretching
dues. The reduced ammable gases will lead to a reduction of vibration of CO–O–C), 1180 cmÀ1 (stretching vibration of C–O–
the heat release rate. Moreover, the absorbance intensity of C) and 830 cmÀ1 (deformation vibration of H atoms in
toxic gas (CO) is remarkably decreased. The most important re benzene).28,29 As the pyrolysis temperature increases, the peak at
hazards are classied as heat, smoke and toxic gases. Therefore, 3450 cmÀ1 shis to higher wavenumbers and nearly disappears
the addition of BPA effectively reduces the re hazards of EA at a temperature of 420 ꢀC, because of the release of water. The
coatings.
relative intensity of all the other bands nearly hardly alters
ꢀ
Real time FTIR was chosen to monitor the thermal oxidative before being heated to 300 C. Upon the pyrolysis temperature
degradation process of the cured lms in an air atmosphere. above 400 ꢀC, the relative intensity of the absorption peaks
Fig. 9 shows the RT-FTIR spectra of the sample EA0 at different dramatically decreases, implying that the primary decomposi-
tion of EA occurs. When the temperature rises to 560 ꢀC, no
obvious absorptions are observed in the condensed phase of
EA0, indicating complete decomposition of the EA matrix.
The FTIR spectra of the sample EA30 at different pyrolysis
temperatures are shown in Fig. 10. There is no distinct
difference of the absorption bands between EA0 and EA30,
except for a peak at 988 cmÀ1 (P–O–C), probably due to over-
lapping of the characteristic bands of BPA with those of EA0.
The peak at 988 cmÀ1 decreases quickly with increasing
temperature and disappears completely above 300 ꢀC due to
the degradation of P–O–C bonds, which is consistent with the
ꢀ
TGA results under air. Above 350 C, the rapid decrease in all
the peak intensities indicates that primary chain decomposi-
tion of EA occurs. Moreover, similar specic absorptions to
those of EA0 are observed in the condensed phase of EA30 at
560 ꢀC. The peak at 1290 cmÀ1 is attributed to the P]O
stretching vibration of the O]P–O–F structure, where F
represents a polynuclear aromatic structure.30,31 The presence
Fig. 9 RT-FTIR spectra of the sample EA0 at different pyrolysis
temperatures.
ꢀ
of an aromatic structure in char residues at 560 C is further
75260 | RSC Adv., 2015, 5, 75254–75262
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
residual char. The char for neat EA has a high ID/IG value of 2.62.
Adding 30 wt% BPA into EA reduces the ID/IG value of the char to
2.42, indicating the presence of more graphitized carbons in the
residual char. From the TGA and Raman discussions above, we
conclude that the phosphorus–silicon ame retardant not only
increases the char residue, but also enhances the thermal
stability of the residual char, suggesting signicant improve-
ment in the thermal properties of composite lms.
3.4 Transparency
Fig. 12 presents the UV-Vis transmittance spectra of the cured
EA lms over the visible wavelength range of 400–700 nm. The
pure EA exhibits high optical transparency with transmittance
of over 75% in the studied wavelength range. With the incor-
poration of ame retardant additives, the optical transparency
of the composite lm is slightly higher than that of neat EA lm.
For example, for 550 nm UV irradiation, the transmittances of
the lms containing 10 wt%, 20 wt% and 30 wt% of BPA FRs are
87.6%, 85.5% and 84.3%, respectively, as compared to 83.6%
for neat EA. These enhanced optical transmittances of the
composite lms imply a good dispersion of the additive in the
EA matrix. The attractive optical property can broaden the
application of UV-curable coatings in some areas where high
transparency is required.
Fig. 10 RT-FTIR spectra of the sample EA30 at different pyrolysis
temperatures.
veried by the peaks at 1610 and 750 cmÀ1, ascribed to the
stretching and deformation vibrations of the aromatic rings,
respectively. Moreover, the appearance of peaks at 1084 and
880 cmÀ1 resulting from symmetrical and asymmetrical
stretching vibrations of P–O–P structure conrms the presence
of pyrophosphate.30,32 Hence, a char layer composed of phos-
phorus–carbon complexes forms during combustion, which
acts as an effective barrier to protect the matrix beneath from
decomposing at high temperature.
Raman spectra were employed to investigate the structure of
char residues. The ame retardant efficiency of any ame
retardant is dependent on the quality of char layers. Fig. 11
shows the Raman spectra of residual chars of the cured lms
heated in a muffle furnace at 600 ꢀC for 10 minutes. These two
samples present two principal peaks around 1360 and
1600 cmÀ1. The former is called the D band, while the other is
called the G band and these are attributed to the L-point
photons of A1g symmetry and the rst order scattering of the E2g
phonon of sp2 C atoms, respectively.33 In general, the intensity
ratio of the D to G bands (ID/IG) is used as a measure of the
graphitization extent of the char. The lower the ID/IG values, the
higher the extent of graphitization and thermal resistance of the
Fig. 12 UV-Vis spectra of the cured films.
Fig. 11 Raman spectra of residual chars of the cured films heated in muffle furnace at 600 ꢀC for 10 minutes.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 75254–75262 | 75261

View Article Online
RSC Advances
Paper
10 Y. Guo, C. Bao, L. Song, X. Qian, B. Yuan and Y. Hu,
J. Nanosci. Nanotechnol., 2012, 12, 1776–1791.
4. Conclusions
11 B. Li and M. Xu, Polym. Degrad. Stab., 2006, 91, 1380–1386.
12 D. Price, L. K. Cunliffe, K. Bullett, T. R. Hull, G. J. Milnes,
J. R. Ebdon, B. J. Hunt and P. Joseph, Polym. Degrad. Stab.,
2007, 92, 1101–1114.
13 P. M. Hergenrother, C. M. Thompson, J. G. Smith,
J. W. Connell, J. A. Hinkley, R. E. Lyon and R. Moulton,
Polymer, 2005, 46, 5012–5024.
14 X. Wang, Y. Hu, L. Song, W. Xing and H. Lu, J. Polym. Sci.,
Part B: Polym. Phys., 2010, 48, 693–705.
15 M. Hussain, R. Varley, Z. Mathys, Y. Cheng and G. Simon,
J. Appl. Polym. Sci., 2004, 91, 1233–1253.
16 G. Li, L. Wang, H. Ni and C. U. Pittman Jr, J. Inorg.
Organomet. Polym., 2001, 11, 123–154.
17 R. H. Baney, M. Itoh, A. Sakakibara and T. Suzuki, Chem.
Rev., 1995, 95, 1409–1430.
18 S. H. Phillips, T. S. Haddad and S. J. Tomczak, Curr. Opin.
Solid State Mater. Sci., 2004, 8, 21–29.
19 M. Raimondo, S. Russo, L. Guadagno, P. Longo, S. Chirico,
A. Mariconda, L. Bonnaud, O. Murariu and P. Dubois, RSC
Adv., 2015, 5, 10974–10986.
A phosphorus- and silicon-containing branched compound
with active double bonds was rst synthesized by Michael
addition reaction and then the well-characterized co-monomer
was covalently incorporated into EA. The simultaneous pres-
ence of phosphorus and silicon elements signicantly improved
the ame retardancy and thermal properties of the EA. The
ame retardant EA composites exhibited signicantly improved
LOI values, as high as 29%, while the PHRR and THR values of
the composites were reduced by 50.8% and 29.8%, respectively.
TGA, TG-IR and RT-IR results revealed that the incorporation of
ame retardant accelerated the degradation of the EA matrix,
reduced the ammable and toxic gases release and promoted
char formation. Moreover, the formation of silicon dioxide
originating from the degradation of BPA on the surface of the
residual chars reinforced the char layer, thus enhancing
thermal oxidative resistance. The production of additional
graphitized carbon in char layers results in higher thermal
stability. The improved ame retardancy of EA was responsible
for a condensed phase ame retardant mechanism, including
the catalyzing charring effect of phosphorus ame retardants
and the char strengthening effect of silicon ame retardants,
thereby forming a more effective protective barrier.
´
20 E. Franchini, J. Galy, J.-F. Gerard, D. Tabuani and A. Medici,
Polym. Degrad. Stab., 2009, 94, 1728–1736.
21 A. Fina, H. C. Abbenhuis, D. Tabuani and G. Camino, Polym.
Degrad. Stab., 2006, 91, 2275–2281.
22 H. Lin, X. Wan, X. Jiang, Q. Wang and J. Yin, Adv. Funct.
Mater., 2011, 21, 2960–2967.
23 H. Liang, J. Ding and W. Shi, Polym. Degrad. Stab., 2004, 86,
217–223.
24 X. Wang, X. Wang, L. Song, W. Xing, G. Tang, W. Hu and
Y. Hu, Thermochim. Acta, 2013, 568, 130–139.
Acknowledgements
The work was nancially supported by the National Basic
Research Program of China (973 Program) (No. 2012CB719701),
the Science and Technology Program of Guangzhou, China (No.
2014J4100174) and the Fundamental Research Funds for the
Central Universities (No. WK2320000032).
25 S. Pellice, D. Fasce and R. Williams, J. Polym. Sci., Part B:
Polym. Phys., 2003, 41, 1451–1461.
References
26 Y. Ni, S. Zheng and K. Nie, Polymer, 2004, 45, 5557–5568.
27 X. Qian, L. Song, Y. Hu, R. K. Yuen, L. Chen, Y. Guo, N. Hong
and S. Jiang, Ind. Eng. Chem. Res., 2011, 50, 1881–1892.
28 J. Ding and W. Shi, Polym. Degrad. Stab., 2004, 84, 159–165.
29 X. Wang, Y. Hu, L. Song, W. Xing, H. Lu, P. Lv and G. Jie,
Polymer, 2010, 51, 2435–2445.
30 G.-A. Wang, W.-M. Cheng, Y.-L. Tu, C.-C. Wang and
C.-Y. Chen, Polym. Degrad. Stab., 2006, 91, 3344–3353.
31 G. H. Hsiue, S. J. Shiao, H. F. Wei, W. J. Kuo and Y. A. Sha, J.
Appl. Polym. Sci., 2001, 79, 342–349.
1 C. Decker, Macromol. Rapid Commun., 2002, 23, 1067–1093.
2 K. M. Schreck, D. Leung and C. N. Bowman, Macromolecules,
2011, 44, 7520–7529.
3 B. Yu, X. Wang, W. Y. Xing, H. Y. Yang, L. Song and Y. Hu,
Ind. Eng. Chem. Res., 2012, 51, 14629–14636.
4 W. Xing, Y. Hu, L. Song, X. Chen, P. Zhang and J. Ni, Polym.
Degrad. Stab., 2009, 94, 1176–1182.
5 Z. Huang and W. Shi, Polym. Degrad. Stab., 2007, 92, 1193–
1198.
6 H. Liang and W. Shi, Polym. Degrad. Stab., 2004, 84, 525–532.
7 X.-e. Cheng and W. Shi, Prog. Org. Coat., 2010, 69, 252–259.
8 H. Liang, A. Asif and W. Shi, Polym. Degrad. Stab., 2005, 87,
495–501.
32 M. Bugajny, S. Bourbigot, M. le Bras and R. Delobel, Polym.
Int., 1999, 48, 264–270.
33 W. Li, X.-Z. Tang, H.-B. Zhang, Z.-G. Jiang, Z.-Z. Yu, X.-S. Du
and Y.-W. Mai, Carbon, 2011, 49, 4724–4730.
¨
9 Y. Mulazim, M. V. Kahraman, N. K. Apohan, S. Kızılta¸s and
¨
¨
A. Gungor, J. Appl. Polym. Sci., 2011, 120, 2112–2121.
75262 | RSC Adv., 2015, 5, 75254–75262
This journal is © The Royal Society of Chemistry 2015