Preparation and flame retardancy of an intumescent flame-retardant epoxy resin system constructed by multiple flame-retardant compositions containing phosphorus and nitrogen heterocycle

Preparation and flame retardancy of an intumescent flame-retardant epoxy resin system constructed by multiple flame-retardant compositions containing phosphorus and nitrogen heterocycle

Polymer Degradation and Stability 119 (2015) 251e259 Contents lists available at ScienceDirect Polymer Degradation and Stability journal homepage: w...

2MB Sizes 16 Downloads 149 Views

Polymer Degradation and Stability 119 (2015) 251e259

Contents lists available at ScienceDirect

Polymer Degradation and Stability journal homepage: www.elsevier.com/locate/polydegstab

Preparation and flame retardancy of an intumescent flame-retardant epoxy resin system constructed by multiple flame-retardant compositions containing phosphorus and nitrogen heterocycle Shuang Yang*, Jun Wang, Siqi Huo, Liufeng Cheng, Mei Wang School of Materials Science and Engineering, Wuhan University of Technology, Wuhan 430070, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 March 2015 Received in revised form 14 May 2015 Accepted 31 May 2015 Available online 1 June 2015

A phosphorous/nitrogen-containing reactive phenolic derivative (DOPOeHPM) was synthesized via the addition reaction between 9,10-dihydro-9-oxa-10-phosphaphenanthrene-10-oxide (DOPO) and N-(4hydroxyphenyl) maleimide (HPM). The structure of DOPOeHPM was characterized by Fourier transform infrared spectroscopy (FTIR), 1H and 31P nuclear magnetic resonance (NMR) and elemental analysis (EA). The studied flame-retardant epoxy resin systems were prepared by copolymerizing diglycidyl ether of bisphenol-A (DGEBA) with DOPOeHPM, triglycidyl isocyanurate (TGIC) and 4,40 -diamino-diphenyl sulfone (DDS). Thermal and flame retardant properties of the cured epoxy resins were investigated by differential scanning calorimeter (DSC), thermogravimeric analysis (TGA), limited oxygen index (LOI) measurement, UL94 test and cone calorimeter. The DSC results indicated that the modified epoxy resins showed little fluctuation in glass transition temperatures (197e205  C). The results of combustion tests indicated that the modified epoxy resin systems exhibited excellent flame retardant properties. The P-1 and P-1.25 systems acquired LOI values of 37% and 38.5%, respectively, and achieved a UL94 V-0 rating. Compared with the P-0 system, the peak of heat release rate (pk-HRR), average of effective heat of combustion (av-EHC) and total heat release (THR) of P-1.25 system decreased by 61.4%, 23.4% and 34.9%, respectively. In addition, the total smoke production (TSP) of the modified epoxy resin systems decreased with the increasing content of flame retardants, indicating the smoke suppression effect of the flameretardant systems. Through visual observation, the char residues after cone calorimetry test exhibited intumescent structures with continuous and compact surfaces. The flame retardant mechanism was studied by FTIR, scanning electron microscope (SEM), cone calorimeter and pyrolysis-gas chromatography/mass spectrometry (Py-GC/MS). © 2015 Elsevier Ltd. All rights reserved.

Keywords: Epoxy resin DOPO TGIC Intumescent Flame retardant

1. Introduction Epoxy resins (EPs) are widely used as advanced matrix resin in the electronic and electrical industries due to their attractive characteristics of high tensile strength and modulus, high adhesion to substrates, good chemical and corrosion resistance, excellent dimensional stability and superior electrical properties [1e5]. However, conventional epoxy resins are flammable and can not satisfy high flame-resistance requirement of advanced materials [6,7]. So far, research works on improving the flame retardation of

* Corresponding author. School of Materials Science and Engineering, Wuhan University of Technology, 122 Luoshi Road, Hongshan District, Wuhan 430070, China. E-mail address: [email protected] (S. Yang). http://dx.doi.org/10.1016/j.polymdegradstab.2015.05.019 0141-3910/© 2015 Elsevier Ltd. All rights reserved.

epoxy resins are very attractive for advanced application. Traditionally, halogenated compounds have been widely used to endow epoxy resins with flame resistance. Currently, halogen-containing compounds are not preferred for environmental reasons [8e10]. Therefore, there is a trend to develop and apply halogen-free flame retardants. Phosphorus-containing flame retardants modified epoxy resins are considered to be more environmentally friendly and have received outstanding attention [11e16]. Among the phosphoruscontaining flame retardants, DOPO and its derivatives have received considerable attention due to their high reactivity with epoxy monomers, high thermal stability and flame retardant efficiency [17e21]. However, single flame retardant composition limits the further enhancing of flame retardancy of the modified epoxy resins [22e24]. Therefore, DOPO-based epoxy resin systems with

252

S. Yang et al. / Polymer Degradation and Stability 119 (2015) 251e259

Scheme 1. Synthesis route of DOPOeHPM.

multiple flame-retardant compositions have been prepared and the synergistic effect of multiple flame-retardant functional groups on flame retardancy of epoxy resins has been observed as reported in a few works [25e32]. In this work, a phosphorous/nitrogen-containing reactive phenolic derivative (DOPOeHPM) was synthesized via the addition reaction between DOPO and HPM. The structure of DOPOeHPM was characterized by Fourier transform infrared spectroscopy (FTIR), 1H and 31P nuclear magnetic resonance (NMR) and elemental analysis (EA). The investigated flame-retarded epoxy resin systems were prepared by copolymerizing DGEBA with DOPOeHPM, TGIC and DDS. Thermal and flame-retardant properties of the cured epoxy resins were investigated by differential scanning calorimeter (DSC), thermogravimeric analysis (TGA), limited oxygen index (LOI) measurement, UL94 test and cone calorimeter. The flame retardant mechanism was studied by FTIR, SEM, cone calorimeter and Py-GC/MS. 2. Experimental 2.1. Materials Diglycidyl ether of bisphenol-A (DGEBA) with an epoxide equivalent weight (EEW) of about 188 g/equiv was provided by Yueyang Baling Huaxing Petrochemical Co., Ltd. N-(4hydroxyphenyl) maleimide (HPM) was obtained from Puyang Willing Chemicals Co., Ltd. 9,10-dihydro-9-oxa-10-phosphaphenan threne-10-oxide (DOPO) was purchased from Huizhou Sunstar

Technology Co., Ltd. Triglycidyl isocyanurate (TGIC) was purchased from Jinan Zian Chemicals Co., Ltd. Triphenyl phosphine (TPP), 4,40 Diamino-diphenyl sulfone (DDS) and 1,4-dioxane were purchased from Sinopharm Chemical Reagent Co., Ltd. All the reagents were used as received. 2.2. Synthesis of DOPOeHPM DOPO (21.6 g, 0.1 mol), HPM (18.9 g, 0.1 mol) and 1,4-dioxane (150 ml) were introduced into a 250 mL, three-neck, round-bottom glass flask equipped with a mechanical stirrer, reflux condenser, thermometer and dry nitrogen inlet. The mixture was heated to 80  C under nitrogen atmosphere and stirred until DOPO and HPM dissolved completely. The mixture was further heated to reflux for 5 h and then distilled to remove 1,4-dioxane. The products were washed with ethyl acetate and vacuum-dried at 70  C for 24 h. The reaction formula is shown in Scheme 1. The yield was 90%. Element analysis: C: 64.99 (cal 65.2), N: 3.48 (cal 3.46), H: 3.97 (cal 3.95). IR (KBr, cm1): 3224 (eOH), 1777 and 1706 (C]O), 1393 (CeN), 1189 (P]O), 937 and 758 (PeO-Ph). 1H NMR (DMSO-d6, ppm): 9.7 and 9.8 (eOH, 1H), 6.7e8.3 (AreH, 12H), 3.9e4.2 (CH, 1H), 3.0e3.3 (CH2, 2H). 31P NMR (DMSO-d6, ppm): a single peak at 29.7. 2.3. Preparation of flame-retarded epoxy resins At first, DGEBA, TGIC, DOPOeHPM and TPP (0.4wt.%) were blended and prepolymerized at 135  C until a homogeneous

Scheme 2. Reactions between DOPOeHPM, TGIC and DGEBA.

S. Yang et al. / Polymer Degradation and Stability 119 (2015) 251e259

253

Table 1 Formulas of the cured epoxy resins. Sample code

DGEBA (g)

DDS (g)

DOPOeHPM (g)

TGIC (g)

N content (wt.%)

P content (wt.%)

P-0 P-0.25 P-0.5 P-0.75 P-1 P-1.25

100 100 100 100 100 100

33 34.4 36 38 40.3 43

0 4.62 10.02 16.29 23.65 32.57

0 3.38 7.35 11.94 17.35 23.88

2.8 3.17 3.35 3.93 4.31 4.69

0 0.25 0.5 0.75 1.0 1.25

solution was obtained (TPP was used as a reaction accelerator, TGIC and DOPOeHPM were equimolar). The reactions between DGEBA, TGIC and DOPOeHPM were illustrated in Scheme 2. After that, stoichiometric DDS (the sum of reactive hydrogen of DDS and DOPOeHPM was equal to that of epoxy groups of DGEBA and TGIC) was thoroughly blended at 135  C for 25 min. The mixture was then degassed under vacuum for 5 min to remove trapped air, and then poured directly into preheated mold and thermally cured in air convection oven at 160  C for 2 h and then at 185  C for 5 h. All the details of formula are listed in Table 1. 2.4. Measurements Fourier Transform Infrared (FTIR) spectra were obtained using a Nicolet 6700 infrared spectrometer. The powdered samples were thoroughly mixed with KBr and then pressed into pellets. 1 H and 31P NMR spectra were obtained on a Bruker AV400 NMR spectrometer using DMSO-d6 as solvent. Elemental analysis (EA) was performed on a Vario EL cube Elemental Analyzer. Differential scanning calorimetry (DSC) thermograms were recorded with PerkineElmer DSC 4000 at a heating rate of 10  C/ min under nitrogen atmosphere. Thermogravimetric analysis (TGA) was performed using NETZSCH STA449F3 at a heating rate of 10  C/min under nitrogen atmosphere from 50 to 800  C. The LOI values were measured at room temperature on a JF-3 oxygen index meter (Jiangning Analysis Instrument Company, China) according to ISO4589-1984 standard and dimensions of all samples were 130  6.5  3 mm3. Vertical burning (UL-94) tests

Fig. 1. FTIR spectra of DOPO, HPM and DOPOeHPM.

were carried out on the NK8017A instrument (Nklsky Instrument Co., Ltd, China) with the dimension of 130  13  3 mm3 according to UL-94 test standard. Cone calorimeter measurements were performed on an FTT cone calorimeter according to ISO 5660 under an external heat flux of 50 kW/m2. The dimension of samples was 100  100  3 mm3. The measurement for each specimen was repeated three times, and the error values of the typical cone calorimeter data were reproducible within ±5%. Morphological studies on the residual chars were conducted using a JSM-5610LV scanning electron microscope (SEM) at an acceleration voltage of 25 kV. Py-GC/MS analysis was carried out with Agilent 7890/5975 GC/ MS. The injector temperature was 250  C, 1 min at 50  C, temperature increase to 280  C at a rate of 8  C/min. The temperature of GC/MS interface was 280  C, and the cracker temperature was 500  C. 3. Results and discussion 3.1. Synthesis of DOPOeHPM As shown by FTIR spectra in Fig. 1, DOPOeHPM showed several characteristic absorption peaks: the peak at 3224 cm1 was assigned to the stretching vibration of OH; the peaks at 1777 and 1706 cm1 were assigned to the stretching vibration of C]O; the peak at 1393 cm1 was assigned to the stretching vibration of CeN; the peak at 1189 cm1 was assigned to the stretching vibration of P]O; the peaks at 937 and 758 cm1 were assigned to the stretching vibration of PeOePh. In addition, the stretching

Fig. 2. 1H NMR spectrum of DOPOeHPM.

254

S. Yang et al. / Polymer Degradation and Stability 119 (2015) 251e259

Fig. 4. TGA curves of the cured epoxy resins under N2 atmosphere.

Fig. 3.

maximum weight loss rate (Tmax) and char yields at different temperatures are listed in Table 2. The T5%, Tmax and Char yields at 400  C of the flame-retardant epoxy resin systems were lower than those of P-0 system, suggesting that the incorporation of flame retardants affected the decomposition process of EP matrix. In addition, the T5%, Tmax and Char yields at 400  C of the flameretarded thermosets decreased with the increasing content of DOPOeHPM and TGIC. The pyrolysis of phosphorus-containing groups (DOPO) induced the decomposition of the EP matrix in advance. The induced decomposition effect was also enhanced with increasing mass fraction of phosphorus, leading to the decreased thermal stability. It was worth noting that the char yields of the flame-retarded thermosets were higher than that of P-0 system at the temperature interval of 500e800  C, suggesting that the char residues of EP thermosets with multiple flame-retardant functional groups were more thermal stable than that of neat epoxy resin system. In condensed-phase, the DOPO groups decomposed in advance and promoted the charring of EP matrix; the thermal stable maleimide and triazine-trione groups retarded the decomposition of EP matrix at higher temperature region. The synergistic effect of multiple flame-retardant compositions resulted in highly

31

P NMR spectrum of DOPOeHPM.

vibration absorption peaks of C]C at 3110 cm1 and PeH at 2434 cm1 disappeared. The above discussion indicated that DOPO had reacted with HPM and DOPOeHPM was successfully synthesized. 1 H and 31P NMR spectra of DOPOeHPM were shown in Figs. 2 and 3, respectively. As shown in Fig. 2, DOPOeHPM showed the chemical shifts of OH at 9.8 and 9.7 ppm, aromatic hydrogen include benzene ring and phosphaphenanthrene group at 6.7e8.3 ppm, CH at 3.9e4.2 ppm and CH2 at 3.0e3.3 ppm. As shown in Fig. 3, the 31P NMR spectrum of DOPOeHPM showed a single peak at 29.7 ppm. These results further confirmed that DOPOeHPM was successfully prepared. 3.2. Thermal properties of the cured epoxy resins The glass transition temperatures (Tgs) of the investigated EP composites were measured by DSC and the results are summarized in Table 2. As shown in Table 2, the Tgs showed little fluctuation (197e205  C), which were consistent with that of P-0 system (200  C). DOPOeHPM with bulky rigid group increased the rotational barrier of the thermosets and TGIC with three active epoxy groups increased the crosslinking density of the thermosets, which compensated somewhat for the loss in crosslink density caused by the end-capping reaction between DOPOeHPM and DGEBA (Scheme 2), and therefore maintained the stability of Tgs. Thermal stability of the cured epoxy resins was assessed by TGA under nitrogen atmospheres. TGA curves of the EP thermosets are presented in Fig. 4. The characteristic thermal decomposition data, such as temperature at 5% weight loss (T5%), temperature at

Table 3 LOI and UL94 test results of the cured epoxy resins. Sample code

P content (%)

LOI (%)

UL-94 (3 mm)

P-0 P-0.25 P-0.5 P-0.75 P-1 P-1.25

0 0.25 0.5 0.75 1 1.25

22.5 31 33.5 36.5 37 38.5

NR NR V-1 V-1 V-0 V-0

Table 2 Thermal parameters of the cured epoxy resins. Sample code

P-0 P-0.25 P-0.5 P-0.75 P-1 P-1.25

Tg ( C)

200 197 201 199 202 205

T5% ( C)

384 356 342 332 323 317

Tmax ( C)

412 406 397 395 388 385

Char yields at different temperatures (%) 400 ( C)

500 ( C)

600 ( C)

700 ( C)

800 ( C)

83.6 71.2 52.5 52.1 46.5 46.3

23.3 24 27 26.6 29.2 31.2

19.9 21 24.3 23.9 26.2 27.7

18.6 19.8 23.3 22.6 25.1 26.4

17.6 18.9 22.4 21.7 24.1 25.4

S. Yang et al. / Polymer Degradation and Stability 119 (2015) 251e259

255

Table 4 Combustion parameters of the cured epoxy resins obtained from cone calorimeter test. Sample code

TTI (s)

TOF(s)

av-HRR (kW/m2)

pk-HRR (kW/m2)

av-EHC (MJ/kg)

THR (MJ/m2)

av-COY (kg/kg)

av-CO2Y (kg/kg)

TSP (m2)

P-0 P-0.25 P-0.5 P-0.75 P-1 P-1.25

47 40 43 41 41 34

430 393 384 378 319 337

177 153 135 140 129 114

1208 949 644 602 531 466

22.2 20.8 19.3 19 17 17

80.6 68.8 61.4 60 57.8 52.5

0.063 0.084 0.094 0.098 0.127 0.123

1.589 1.38 1.276 1.264 1.146 1.15

30 25.7 25 25.7 25.1 21

crosslinked and thermal stable char residues. 3.3. LOI and UL94 rating tests The flame-retarded properties of the EP thermosets were determined by LOI and UL94 vertical burning tests. The corresponding data are listed in Table 3. The LOI value of P-0 system was only 22.5%, whereas that of P-0.25 system with phosphorus content of only 0.25 wt.% reached 31%. However, P-0.25 system failed to pass the UL94 test. With the increasing phosphorus content, the LOI values of P-1 and P-1.25 systems further increased to 37% and 38.5%, respectively, and the samples achieved the UL94 V-0 rate. The above discussion disclosed that the studied flame-retarded epoxy resin systems exhibited higher LOI values and UL94 rating with lower phosphorus content compared with previously reported DOPO-based epoxy resin systems [21e24]. It was inferred that the excellent flame-retardant property was attributed to the synergistic effect of multiple flame-retardant compositions, which were composed of maleimide, phosphaphenanthrene and triazinetrione groups with certain flame retardant effect [33e36]. The flame-retardant groups reacted with EP matrix during combustion. As discussed in TGA, the char yields increased with the increasing content of DOPOeHPM and TGIC at high temperature region (500e800  C). Higher char yield means less release of pyrolysis gas and decreases the exothermicity during combustion, consequently to limit the flammability of epoxy resins. The flame retardant mechanism will be further discussed in the subsequent section.

Fig. 6. THR curves of the cured epoxy resins.

Cone calorimetry test was used to investigate the flameretarded behaviors of DOPOeHPM and TGIC on the EP thermosets. The characteristic parameters, such as the time to ignition

(TTI), time of flameout (TOF), average of heat release rate (av-HRR), peak of heat release rate (pk-HRR), average of effective heat of combustion (av-EHC), total heat release (THR), average CO yield (av-COY), average CO2 yield (av-CO2Y) and total smoke production (TSP) are summarized in Table 4. As shown in Table 4, the TTI of P-0 system was 47s whereas those of the flame-retarded thermosets were shortened to some extent. P-1.25 system with the highest phosphorus content had the minimum value of TTI. According to the above discussion on TGA, DOPO groups decomposed in advance and induced the degradation of EP matrix, which weakened the resistance to ignition. Therefore, the TTIs of the phosphorus-containing system decreased. Of course,

Fig. 5. HRR curves of the cured epoxy resins.

Fig. 7. Mass loss curves of the cured epoxy resins from cone calorimeter test.

3.4. Cone calorimeter analysis of the cured epoxy resins

256

S. Yang et al. / Polymer Degradation and Stability 119 (2015) 251e259

Fig. 8. The FTIR spectra of the char residues after UL94 test.

the degradation of phosphorus-containing thermosets ahead of time contributed to charring earlier during combustion. The char layer hindered the heat and oxygen from reaching the inner EP matrix, and thus promoted the flame-retarded properties. The curves of heat release rate (HRR) are shown in Fig. 5. As presented in Fig. 5, the neat EP thermoset burned rapidly after ignition and HRR reached a sharp peak with a pk-HRR of 1208 kW/ m2. In addition, P-0 system had the highest av-HRR of 177 kW/m2. With the addition of DOPOeHPM and TGIC, the av-HRR and pk-HRR of the modified EP composites decreased sharply. The av-HRR and pk-HRR of P-0.5 system decreased by 23.7% and 46.7%, respectively, compared with the neat epoxy resin system. The P-1.25 system with the most mass fraction of flame retardants had the lowest avHRR and pk-HRR, which were decreased by 35.6% and 61.4%

compared with the P-0 system. Moreover, the THR decreased with the increasing content of DOPOeHPM and TGIC as shown in Fig. 6 and Table 4. The THR of P-1 and P-1.25 systems decreased by 28.3% and 34.9%, respectively, compared with the P-0 system. The above discussion disclosed that the addition of DOPOeHPM and TGIC significantly reduced the HRR and THR, and thus enhanced the flame retardancy. This was in accordance with the TGA, LOI and UL94 results. The enhanced flame-retardant properties of the EP thermosets were further explained by average of effective heat of combustion (av-EHC), time of flameout (TOF), average CO yield (av-COY) and average CO2 yield (av-CO2Y), accompany with the mass loss curves of EP thermosets obtained from cone calorimeter test (Fig. 7). AvEHC, which is the ratio of average of heat release rate (av-HRR) to the average mass loss rate from the cone calorimetry test, discloses the burning rate of volatile gases in gaseous-phase flame during combustion. As shown in Table 4, P-0 system had the highest avEHC of 22.2 MJ/kg. The av-EHC gradually decreased with the increasing content of DOPOeHPM and TGIC. The less combustion heat generated from the burning of volatile gases in gaseous-phase was ascribed to the flame retardant quenching effect from the gaseous-phase pyrolysis products of DOPOeHPM and TGIC. This was further proved by the time of flameout (TOF). As presented in Table 4, P-0 system had the highest TOF of 430s. The TOFs were shortened to some extent and decreased with the increasing content of DOPOeHPM and TGIC. It was inferred that DOPOeHPM and TGIC decomposed to release incombustible gases and pyrolysis fragments with quenching effect during combustion. As shown in Fig. 7, the char yields of the EP thermosets increased with the increasing content of DOPOeHPM and TGIC after cone calorimeter test, indicating the flame retardant effect in the condensed-phase [37]. Table 4 also showed that the av-COY increased whereas the avCO2Y decreased with the increasing content of flame-retardant compositions, proving the flame retardant quenching effect in the gaseous-phase [37], since more av-COY and less av-CO2Y meant more incomplete combustion products (CO) and less complete combustion products (CO2). Moreover, the sum of av-COY and av-

Fig. 9. The SEM and digital images of the char residues after cone calorimeter test.

S. Yang et al. / Polymer Degradation and Stability 119 (2015) 251e259

CO2Y decreased with the increasing content of DOPOeHPM and TGIC, indicating that the flame retardants functioned in the condensed-phase and resulted in less release of combustible volatiles. Smoke in a real fire means more risk of suffocation, even more fatal than heat release. As shown in Table 4, it was evident that the total smoke production (TSP) was reduced with the addition of DOPOeHPM and TGIC. The TSP of P-1.25 system decreased by 30% compared with that of P-0 system, suggesting that the addition of DOPOeHPM and TGIC suppressed the formation of smoke. 3.5. FTIR study of the char residues after UL94 test The FTIR spectra of the char residues of the cured epoxy resins after UL94 test are shown in Fig. 8. The absorbance peaks of P0 system at 1590 and 1510 cm1 indicated the formation of polyaromatic carbons [36], which were observed in other spectra curves as well. For the flame-retardant systems, except the peaks at 1590 and 1510 cm1, new absorbance peaks at 1764, 1706, 1382, 756 and 717 cm1 appeared and became more and more obvious with the increasing content of DOPOeHPM and TGIC. The evident absorbance peaks of C]O at 1764, 1706 and 717 cm1 and CeN at 1382 cm1 indicated the relict structures of thermal stable maleimide (DOPOeHPM) and triazine-trione (TGIC) groups, which

257

retarded the decomposition of EP thermosets and were carbonized to provide char source during combustion. The new absorbance peak at 756 cm1 indicated the existence of PeOeC in the char residue [36], which further proved the flame-retarded effect of DOPO group in condensed-phase. 3.6. Morphological study of the char residues after cone calorimeter test Morphological study of the char residues were conducted by visual observation and SEM. The SEM and digital images of the char residues after cone calorimeter test are shown in Fig. 9. As shown in Fig. 9, the char of P-0 system showed a small amount of residual with a fragmentary structure which was unable to serve as a protective layer, whereas those of the modified epoxy resin systems exhibited higher char yields with crosslinked and rigid char layers. From the top views of the char layers, it was evident that the flameretardant systems exhibited a more continuous and compact surface compared with that of P-0 system. From the side views of the char residues, the expansion ratios of the flame-retardant samples were markedly elevated with the incorporation of DOPOeHPM and TGIC. The intumescent structures of the flame-retarded systems were clearly observed. The char residues were further investigated by SEM.

Fig. 10. The typical Py-GC/MS spectra of P-0 and P-1.25 systems.

258

S. Yang et al. / Polymer Degradation and Stability 119 (2015) 251e259

The interior and exterior char layers of the P-1 and P-1.25 systems were studied by SEM as shown in Fig. 9. As can be seen in Fig. 9, the exterior char layers of the P-1 and P-1.25 systems showed continuous, compact and sealed structures with obvious stretched track, indicating the highly crosslinked and tough char layers. The interior char layers of the P-1 and P-1.25 systems presented multihole and honeycomb-like structures with numerous bubbles separated by very thin layers. The continuous and compact surface of the char residue prevented the release of pyrolysis gases which were finally accommodated by the underlying char layers to form an intumescent structure. The sealed surface reduced the release of combustible volatiles, consequently to cut off the supply of fuel to weaken the combustion intensity. In addition, the intumescent char layer reduced the efficiency of heat- and oxygen-exchange. An intumescent char layer with a strong and sealed surface led to the promoted flammability property. 3.7. Analysis of Py-GC/MS For further disclosing the flame retardant mechanism, the P0 and P-1.25 thermosets were investigated by Py-GC/MS and some typical mass spectra are shown in Fig. 10. Fig. 10(a) and (b) showed two typical fragment flows of P-1.25 system, respectively. As shown in Fig. 10(a), the m/z value of 64 represented the typical PO2H fragment which was produced by PO combining with OH or H and O free radicals [29]; as shown in Fig. 10(b), the obvious m/z values of 168, 169 and 170 were considered as the fragments of o-phenylphenoxyl radical, o-phenylphenol and dibenzofuran which were derived from the decomposition of phosphaphenanthrene [29]. The above mentioned fragments functioned in gaseous-phase through quenching effect, which was responsible for the decrease in avEHC. Fig. 10(c) and (d) showed the mass spectra of P-0 and P-1.25 systems at the maximum release intensity of pyrolytic products, respectively. As clearly presented in Fig. 10(c) and (d), it was evident that the relative concentrations of larger chain segments (m/z >200) of P-1.25 system were significantly lower than those of P0 system and the fragments of P-1.25 system were mainly small molecules. This phenomenon further proved that the incorporation of flame retardants retarded the decomposition of EP matrix at higher temperature region (above 400  C). The larger chain segments remained in the condensed phase and formed the char layer. The char layer would swell through the release of small molecules, and finally form an intumescent structure. 4. Conclusions DOPOeHPM was successfully synthesized via the addition reaction between DOPO and HPM. The studied flame-retarded epoxy resin systems were prepared by copolymerizing DGEBA with DOPOeHPM, TGIC and DDS. The DSC results indicated that the modified epoxy resins showed little fluctuation in glass transition temperatures (197e205  C). The results of combustion tests indicated that the modified EP systems exhibited excellent flame retardant properties. Specifically, the P-0.25 thermoset had a LOI value of 31% when the phosphorus content was only 0.25wt.%. The P-1 and P-1.25 systems had LOI values of as high as 37% and 38.5%, respectively, and achieved a UL94 V-0 rating. The cone calorimeter test results indicated that the pk-HRR, av-HRR, av-EHC, THR and TSP were significantly decreased with the incorporation of DOPOeHPM and TGIC. The morphological study revealed that the char layer exhibited an intumescent structure with sealed surface. The flame retardant mechanism was explained as follows. In the condensed-phase, the phosphaphenanthrene group decomposed in advance and induced the degradation of EP matrix to form a char layer; in addition, the rigid and thermal stable maleimide and

triazine-trione groups retarded the decomposition of EP matrix, and finally a highly crosslinked char layer with high char yield was formed, which served as a protective layer for the underlying EP matrix. In the gaseous-phase, the flame-retardant systems released fragments with quenching effect during combustion. Moreover, the intumescent char layer with sealed surface reduced the release of pyrolysis gases and efficiency of heat- and oxygen-exchange. The multiple flame-retardant compositions functioned in the condensed-phase and gaseous-phase simultaneously led to the significantly promoted flame-retardant properties. References [1] Q. Lv, J. Huang, M.-J. Chen, J. Zhao, Y. Tan, L. Chen, et al., An effective flame retardant and smoke suppression oligomer for epoxy resin, Ind. Eng. Chem. Res. 52 (2013) 9397e9404. [2] D. Wang, K. Zhou, W. Yang, W. Xing, Y. Hu, X. Gong, Surface modification of graphene with layered molybdenum disulfide and their synergistic reinforcement on reducing fire hazards of epoxy resins, Ind. Eng. Chem. Res. 52 (2013) 17882e17890. [3] R.M. Perez, J.K.W. Sandler, V. Altst€ adt, T. Hoffmann, D. Pospiech, M. Ciesielski, et al., Novel phosphorus-modified polysulfone as a combined flame retardant and toughness modifier for epoxy resins, Polymer (Guildf.) 48 (2007) 778e790. [4] B. Schartel, A.I. Balabanovich, U. Braun, U. Knoll, J. Artner, M. Ciesielski, et al., Pyrolysis of epoxy resins and fire behavior of epoxy resin composites flameretarded with 9, 10-dihydro-9-oxa-10-phosphaphenanthrene-10-oxide additives, J. Appl. Polym. Sci. 104 (2007) 2260e2269. [5] B.K. Kandola, B. Biswas, D. Price, A.R. Horrocks, Studies on the effect of different levels of toughener and flame retardants on thermal stability of epoxy resin, Polym. Degrad. Stab. 95 (2010) 144e152. [6] M. Dogan, S. Murat Unlu, Flame retardant effect of boron compounds on red phosphorus containing epoxy resins, Polym. Degrad. Stab. 99 (2014) 12e17. €ß, M. Ciesielski, J. Diederichs, M. Do € ring, et al., [7] B. Perret, B. Schartel, K. Sto A new halogen-free flame retardant based on 9,10-dihydro-9-oxa-10phosphaphenanthrene-10-oxide for epoxy resins and their carbon fiber composites for the automotive and aviation industries, Macromol. Mater. Eng. 296 (2011) 14e30. €ring, Novel star-shaped and [8] L. Zang, S. Wagner, M. Ciesielski, P. Müller, M. Do hyperbranched phosphorus-containing flame retardants in epoxy resins, Polym. Adv. Technol. 22 (2011) 1182e1191. [9] H. Liu, K. Xu, H. Ai, L. Zhang, M. Chen, Preparation and characterization of phosphorus-containing Mannich-type bases as curing agents for epoxy resin, Polym. Adv. Technol. 20 (2009) 753e758. [10] X. Wang, Y. Hu, L. Song, W. Xing, H. Lu, Thermal degradation mechanism of flame retarded epoxy resins with a DOPO-substituted organophosphorus oligomer by TG-FTIR and DP-MS, J. Anal. Appl. Pyrolysis 92 (2011) 164e170. [11] H. Wang, Q. Wang, Z. Huang, W. Shi, Synthesis and thermal degradation behaviors of hyperbranched polyphosphate, Polym. Degrad. Stab. 92 (2007) 1788e1794. , G. Marosi, Intrinsically flame retardant [12] a. Toldy, P. Anna, I. Csontos, A. Szabo epoxy resinefire performance and backgroundepart I, Polym. Degrad. Stab. 92 (2007) 2223e2230. [13] B. Schartel, Phosphorus-based flame retardancy mechanismsdold hat or a starting point for future development? Materials (Basel) 3 (2010) 4710e4745. [14] X. Wang, Y. Hu, L. Song, W. Xing, H. Lu, P. Lv, et al., Flame retardancy and thermal degradation mechanism of epoxy resin composites based on a DOPO substituted organophosphorus oligomer, Polymer (Guildf.) 51 (2010) 2435e2445. [15] C. Wang, J. Shieh, Synthesis and properties of epoxy resins containing bis (3hydroxyphenyl ) phenyl phosphate, Eur. Polym. J. 36 (2000) 443e452. [16] H. Liu, K. Xu, H. Cai, J. Su, X. Liu, Z. Fu, et al., Thermal properties and flame retardancy of novel epoxy based on phosphorus-modified Schiff-base, Polym. Adv. Technol. 23 (2012) 114e121. € ring, et al., [17] B. Schartel, U. Braun, A.I. Balabanovich, J. Artner, M. Ciesielski, M. Do Pyrolysis and fire behaviour of epoxy systems containing a novel 9,10dihydro-9-oxa-10-phosphaphenanthrene-10-oxide-(DOPO)-based diamino hardener, Eur. Polym. J. 44 (2008) 704e715. [18] J. Hu, J. Shan, D. Wen, X. Liu, J. Zhao, Z. Tong, Flame retardant, mechanical properties and curing kinetics of DOPO-based epoxy resins, Polym. Degrad. Stab. 109 (2014) 218e225. [19] K.A. Salmeia, S. Gaan, An overview of some recent advances in DOPOderivatives: chemistry and flame retardant applications, Polym. Degrad. Stab. 113 (2015) 119e134. [20] S. Yang, J. Wang, S. Huo, L. Cheng, M. Wang, The synergistic effect of maleimide and phosphaphenanthrene groups on a reactive flame-retarded epoxy resin system, Polym. Degrad. Stab. 115 (2015) 63e69. [21] W. Zhang, X. Li, R. Yang, Blowing-out effect in epoxy composites flame retarded by DOPO-POSS and its correlation with amide curing agents, Polym. Degrad. Stab. 97 (2012) 1314e1324. [22] C.S. Wang, C.H. Lin, Synthesis and properties of phosphorus containing

S. Yang et al. / Polymer Degradation and Stability 119 (2015) 251e259 advanced epoxy resins, J. Appl. Polym. Sci. 75 (2000) 429e436. [23] X. Wang, Q. Zhang, Synthesis, characterization, and cure properties of phosphorus-containing epoxy resins for flame retardance, Eur. Polym. J. 40 (2004) 385e395. [24] C.S. Wang, C.H. Lin, Synthesis and properties of phosphorus-containing epoxy resins by novel method, J. Polym. Sci. Part A Polym. Chem. 37 (1999) 3903e3909. [25] L. Qian, Y. Qiu, J. Liu, F. Xin, Y. Chen, The flame retardant group-synergisticeffect of a phosphaphenanthrene and triazine double-group compound in epoxy resin, J. Appl. Polym. Sci. (2014) 1e8, 39709. [26] Y. Wang, J. Zhao, Y. Yuan, S. Liu, Z. Feng, Synthesis of maleimido-substituted aromatic s-triazine and its application in flame-retarded epoxy resins, Polym. Degrad. Stab. 99 (2014) 27e34. [27] Z. Li, R. Yang, Study of the synergistic effect of polyhedral oligomeric octadiphenylsulfonylsilsesquioxane and 9,10-dihydro-9-oxa-10phosphaphenanthrene-10-oxide on flame-retarded epoxy resins, Polym. Degrad. Stab. 109 (2014) 233e239. [28] X. Wang, Y. Hu, L. Song, W. Xing, H. Lu, P. Lv, et al., Effect of a triazine ringcontaining charring agent on fire retardancy and thermal degradation of intumescent flame retardant epoxy resins, Polym. Adv. Technol. 22 (2011) 2480e2487. [29] L. Qian, L. Ye, Y. Qiu, S. Qu, Thermal degradation behavior of the compound containing phosphaphenanthrene and phosphazene groups and its flame retardant mechanism on epoxy resin, Polymer (Guildf.) 52 (2011) 5486e5493. [30] L. Qian, Y. Qiu, N. Sun, M. Xu, G. Xu, F. Xin, et al., Pyrolysis route of a novel

[31]

[32]

[33]

[34] [35]

[36]

[37]

259

flame retardant constructed by phosphaphenanthrene and triazine-trione groups and its flame-retardant effect on epoxy resin, Polym. Degrad. Stab. 107 (2014) 98e105. W. Zhang, X. He, T. Song, Q. Jiao, R. Yang, The influence of the phosphorusbased flame retardant on the flame retardancy of the epoxy resins, Polym. Degrad. Stab. 109 (2014) 209e217. W. Zhang, X. Li, R. Yang, Study on flame retardancy of TGDDM epoxy resins loaded with DOPO-POSS compound and OPS/DOPO mixture, Polym. Degrad. Stab. 99 (2014) 118e126. X. Chen, A. Gu, G. Liang, L. Yuan, D. Zhuo, J. Hu, Novel low phosphorus-content bismaleimide resin system with outstanding flame retardancy and low dielectric loss, Polym. Degrad. Stab. 97 (2012) 698e706. C. Wu, Y. Liu, K. Hsu, Maleimide-epoxy resins: preparation, thermal properties, and flame retardance, Polymer (Guildf.) 44 (2003) 565e573. T.H. Chiang, C.Y. Liu, C.Y. Dai, A study of the thermal, dielectric, and flameretarding characteristics of various bismaleimide blended with halogen-free epoxy resin, J. Polym. Res. 20 (2013) 274. G. You, Z. Cheng, H. Peng, H. He, The synthesis and characterization of a novel phosphorus-nitrogen containing flame retardant and its application in epoxy resins, J. Appl. Polym. Sci. (2014) 1e8, 41079. S. Brehme, B. Schartel, J. Goebbels, O. Fischer, D. Pospiech, Y. Bykov, et al., Phosphorus polyester versus aluminium phosphinate in poly(butylene terephthalate) (PBT): flame retardancy performance and mechanisms, Polym. Degrad. Stab. 96 (2011) 875e884.