Impacts of multi-element flame retardants on flame retardancy, thermal stability, and pyrolysis behavior of epoxy resin

Impacts of multi-element flame retardants on flame retardancy, thermal stability, and pyrolysis behavior of epoxy resin

Polymer Degradation and Stability 167 (2019) 217e227 Contents lists available at ScienceDirect Polymer Degradation and Stability journal homepage: w...

4MB Sizes 0 Downloads 9 Views

Polymer Degradation and Stability 167 (2019) 217e227

Contents lists available at ScienceDirect

Polymer Degradation and Stability journal homepage:

Impacts of multi-element flame retardants on flame retardancy, thermal stability, and pyrolysis behavior of epoxy resin Bin Zhao*, Peng-Wei Liu, Kuan-Kuan Xiong, Hui-Hui Liu, Pei-Hua Zhao, Ya-Qing Liu** Research Center for Engineering Technology of Polymeric Composites of Shanxi Province, School of Materials Science and Engineering, North University of China, Taiyuan, 030051, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 April 2019 Received in revised form 28 June 2019 Accepted 7 July 2019 Available online 8 July 2019

N-substituted bis(diphenylphosphanyl)amine RN(PPh2)2 (PNP; R ¼ CH2CH2CH2Si(OEt)3) and its mononuclear nickel(II) ethanedithiolate complexe RN(PPh2)2Ni(SCH2CH2S) (PNS; R ¼ CH2CH2CH2Si(OEt)3) were prepared and used as multi-element flame retardants for epoxy resin (EP). Results of thermogravimetric analysis (TGA) revealed that the incorporation of PNP or PNS restrained the decomposition of EP and also improved the stability of char residues. Moreover, results of limited oxygen index (LOI), vertical burning test (UL-94), microscale combustion calorimeter (MCC), and cone calorimeter tests proved that both PNP and PNS endowed EP with good flame retardancy. Different chemical structures and action mechanisms of PNP and PNS resulted in different degrees of flame retardancy for EP. Although EP/PNP samples caused a little higher LOI values than those of EP/PNS ones, the incorporation of 7 wt% PNS endowed EP with a lower peak of heat release rate (PHRR) and total smoke production (TSP), thereby indicating better flame-retardant and smoke suppression effects. Furthermore, thermogravimetryFourier transform infrared spectroscopy (TG-FTIR) and pyrolysis gas chromatography/mass spectrometry (Py-GC/MS) were employed to analyze the pyrolysis behavior of EPs and flame retardants. The morphologies of char layers for EPs were examined by scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDX). In comparison to PNP, PNS promoted the formation of a stronger protective char barrier comprising by P/Si/S/Ni elements and also restrained the effects of flammable gases on EP. © 2019 Elsevier Ltd. All rights reserved.

Keywords: Flame retardancy Metal complexe P/N/Si-containing Epoxy resin Py-GC/MS

1. Introduction Epoxy resin (EP), as one of the important thermosetting resins, is extensively used in coatings, adhesives, electrical laminates, and composite matrices due to its easy curing and processing, good heat and chemical resistance, and excellent electrical and mechanical performances [1,2]. However, EP causes a disadvantage of higher flammability which must be seriously addressed [3,4]. The use of traditional halogenated flame retardants (bromine and chlorinebased) has been curbed due to ecological reasons [5,6]; hence, researches have paid considerable attention to develop halogen-free flame retardants for EP. Phosphorus and nitrogen are considered as excellent flameretardant elements because they are more environmentally

* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (B. Zhao), [email protected] (Y.-Q. Liu). 0141-3910/© 2019 Elsevier Ltd. All rights reserved.

benign than bromine and chlorine [7e9]; thus, in recent years, various effective phosphorus/nitrogen-containing flame retardants, such as 9, 10-dihydro-9-oxa-10-phosphaphenanthrene10-oxide (DOPO) and its derivatives [10e12], a variety of phosphazene compounds [13e15], modified ammonium polyphosphate (APP) [16], and imidazole derivatives [17], have been developed as additive and reactive fillers for EP. Silicon, is another important environmentally benign flame-retardant element and manifests good condensed-phase flame retardant mechanism for materials. Flame-retardant EPs can be also prepared from silicon-containing glycidyl monomers [12,18]. Silicon-containing components have been usually used as parts of several silicon/phosphorus/nitrogencontaining flame-retardant systems, such as polyhedral oligomeric silsesquioxanes (POSS)/DOPO [19], POSS/polyphosphazene [20], and some novel phosphorus-nitrogen-silicon molecules [21,22], have been also developed for EP. Moreover, sulfur has been proved to be an effective flame retardant element and plays an important role in the P/N-containing system for EP [23e25]. Metallic element-based flame retardants, such as aluminium


B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227

hydroxide, magnesium hydroxide, hypophosphite and phosphinates (Al3þ, Zn2þ, Ca2þ), and nano-fillers (Co/Ni-based layered double hydroxide and alpha-zirconium phosphate), are widely used in various polymers [26,27]. It is proved that metallic elements (Ni, Co, Cu, Zn, Zr) can effectively enhance the char yield of polymers and cause a better smoke suppression effect [28e30]. Wang et al. [31] showed that Co3O4-GO and SnO2-GO caused significant improvements in catalytic and smoke suppression effects of EP. Yu et al. [32] reported that Ni2þ played an important role in char formation and smoke suppression. Although metal-containing flame retardants possess many advantages, very few studies have reported the impacts of organic metallic flame-retardant molecules (such as phosphorus, nitrogen, silicon, and sulfur) on EP. In the present study, N-substituted bis(diphenylphosphanyl) amine RN(PPh2)2 (PNP; P/N/Si-containing, R ¼ CH2CH2CH2Si(OEt)3) and its mononuclear nickel(II) ethanedithiolate complexe RN(PPh2)2Ni(SCH2CH2S) (PNS; P/N/Si/S/Ni-containing R ¼ CH2CH2CH2Si(OEt)3) were prepared successfully and then used as flame retardants for EP. The thermal stability, the flame retardancy, and the pyrolysis behavior of different EPs were investigated in detail.

procedure depicted above. The detailed formulations of EP and its derivatives are presented in Table 1.

2.3. Characterization and measurements 2.3.1. Chemical structural characterization Fourier transform infrared (FTIR) analyses were directly conducted by attenuated total reflection (ATR) adjunct of a Nicolet IS50 spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). 1H NMR and 31P NMR spectra were detected from a BrukerAvanceIIITM 400 MHz spectrometer using Chloroform-d (CDCl3) and tetramethylsilane (TMS) as the solvent and the internal standard, respectively.

2.3.2. Thermogravimetric analysis (TGA) TGA was performed on a TA Instruments Q50 thermal gravimetric analyzer in the temperature range of 40  Ce700  C and at a heating rate of 10  C min1 under a nitrogen flow of 40 mL min1. The weights of the samples were 5 ± 0.5 mg, and the residue mass reproduction of each sample was within ±1%.

2. Experimental 2.1. Materials 3-triethoxysilylpropylamine (KH550), nickel chloride hexahydrate (NiCl2$6H2O), diphenylphosphonium chloride (Ph2PCl) and triethylamine were purchased from Shanghai Civi Chemical Technology Co. Ltd., Shanghai, China. 1, 2-ethanedithiol was purchased from J&K Scientific. Anhydrous ethanol and dichloromethane (CH2Cl2) were procured from Tianjin Guangfu Chemical Reagent Factory, Tianjin, China. 4, 4-diamino-diphenylmethane (DDM) was obtained from Aladdin Industrial Corporation, Shanghai, China. All reagents were analytical reagents (ARs) and used as received. Diglycidyl ether of bisphenol-A (DGEBA) (E51; NPEL128; epoxide equivalent weight ¼ 185 g/equiv) was obtained from Nan Ya Electronic Materials Co. Ltd., Kunshan, China. PNP and PNS were prepared according to the method described in previous works [33,34]. The detailed synthesis routes and characterization of PNP and PNS are presented in supporting information. The chemical structures for PNP and PNS are displayed in Scheme 1. 2.2. Preparation of EP thermosets Flame-retardant EP thermosets were prepared through the following process. PNP or PNS was first blended with DGEBA at 80  C to form a homogenous mixture. The stoichiometric DDM, after being melted at 110  C, was introduced into the mixture under vigorous stirring. Finally, the mixture of DGEBA/DDM/PNP or DGEBA/DDM/PNS was poured directly into a preheated Teflon mold after degassing and thermally cured it for 2 h each at 120  C and 150  C. In addition, the EP sample was also cured by the same

Scheme 1. Chemical structures of PNP and PNS.

2.3.3. Flame retardancy and fire behavior The limiting oxygen index (LOI) values of EP and flameretardant EPs were measured by a Motis oxygen index test instrument according to ASTM D2863-00 standard. The sheet dimension of each sample was 130.0 mm  6.5 mm  3.2 mm, and the deviation for each obtained data was ±0.2 vol%. Vertical burning tests were carried out in a CZF-5 instrument according to the ASTM D3801/UL-94V standard and the sheet dimension for each sample was 125.0 mm  13.0 mm  3.2 mm. Pyrolysis combustion flow calorimeter (PCFC) tests were performed on an FAA microscale combustion calorimeter. Cone calorimeter tests were executed in an FTT cone calorimeter device according to ISO 5660-1 standard, and all samples (100.0 mm  100.0 mm  3.0 mm) were irradiated under a heat flux of 35 kW m2. Each sample group was tested 2e3 times in order to maintain a deviation of 10% for the obtained results.

2.3.4. Gaseous analysis Thermogravimetry-Fourier transform infrared spectroscopy (TG-FTIR) was performed on a TA Q50 instrument, which was connected on an IS50 IR spectrophotometer through an insulated stainless-steel pipe. The pipe and the gas cell were both maintained at 250  C to avoid secondary reactions and the condensation of volatiles. Samples weighting 15 ± 0.5 mg were heated from 40  C to 700  C at a heating rate of 20  C/min under the nitrogen atmosphere. Pyrolysis-Gas chromatography/mass spectrometry (Py-GC/ MS) was carried out in a PerkinElmer Clarus 680 GC-SQ8 MS instrument, which was connected to a CDS 5200 cracking furnace (helium was used as the carrier gas). The injection temperature and the GC/MS interface temperature were maintained at 250  C and 280  C, respectively. The sample cracking temperature was set to 500  C due to the closed ignition temperature of epoxy resin [13].

2.3.5. Char residue analysis Scanning electron microscopy (SEM; HITACHI, SU8010, Japan) coupled with energy dispersive X-ray spectroscopy (EDX; IXRF Systems model 550i, USA) was performed to analyze the morphologies and the elemental compositions of char residues for EPs after cone calorimeter tests. All samples were sputtered with a thin layer of AuePt before performing observations.

B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227


Table 1 Formulations of EP, EP/PNP and EP/PNS. Sample

EP EP/3 wt%FR EP/5 wt%FR EP/7 wt%FR a b


80 80 80 80

DDM (g)

21.5 21.5 21.5 21.5

FR (PNP/PNS) (g)

/ 3.1 5.3 7.6

P contents (wt%)a

P contents (wt%)b





/ 0.32 0.52 0.74

/ 0.27 0.45 0.63

/ 0.35 ± 0.02 0.53 ± 0.02 0.80 ± 0.01

/ 0.31 ± 0.03 0.47 ± 0.03 0.71 ± 0.02

P contents in formulations were calculated based on molecular structures and relevant loadings of PNP/PNS. P contents were tested by ultraviolet spectrophotometry [35].

3. Results and discussion 3.1. Thermal stability The thermal stability of PNP and PNS was evaluated by TGA in

the nitrogen atmosphere. It is clear from Fig. 1 (A, B) that PNP underwent the main decomposition process during heating, whereas PNS manifested a lower initial decomposition temperature and typical multi-step TG and DTG curves. After heating, the amount of char residue for PNS (28.3 wt%) was found to be significantly higher

Fig. 1. TG and DTG curves of PNP, PNS and EPs in the N2 atmosphere.


B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227

than that of PNP (7.4 wt%). The noticeable difference in thermal stability of PNP and PNS can be ascribed to the SeNi chemical structure of PNS. The TGA and the DTG curves of EP and flame-retardant EPs under the N2 atmosphere are illustrated in Fig. 1 (C, D) and Fig. 1 (E, F), respectively, and the corresponding data, such as initial decomposition temperature (T5%), maximum mass loss rate (amax), temperature at amax (Tmax), and char residue values at 700  C, are summarized in Table 2. EP manifested a typical one-stage decomposition curve, and the values of T5% and Tmax were detected as 372.9  C and 392.1  C, respectively. However, EP/PNP composites caused a significant decrease in both T5% and amax; it can be attributed to the lower thermal stability of PNP than that of the EP matrix. The incorporation of PNP, due to its lower loading amount, had an imperceptible effect on the Tmax value of EP. It is noteworthy that the amount of char yield for EPs was enhanced from 17.2% to 24% with the increasing PNP filling from 3 wt% to 7 wt%. It is noticeable from Fig. 1 (E, F) that the incorporation of PNS, due to its lower thermal stability, endowed EPs with lower initial temperatures as compared to those of EP/PNP samples. The decrease in amax value of EP/PNS samples indicates that stable char layers formed during incunabular decomposition impeded the further decomposition of the samples; hence, EP/PNS composites caused enhanced char yields after heating. The experimental values of char yields for both EP/PNP and EP/PNS samples were found to be higher than the corresponding calculated ones. Results of TGA reveal the interactions among EP, PNP, PNS, and their decomposition products during heating, restrained the further decomposition of EP and led to an enhancement in char yield rate. PNS manifested better effects on the thermal stability of EP in the high-temperature zone; it happened mainly due to the degradation products of PNS (Ni could further catalyze the degradation of the EP matrix) [30].

Table 3 Results of LOI and UL 94 tests of EP, EP/PNP and EP/PNS. Samples

EP EP/3 wt%PNP EP/5 wt%PNP EP/7 wt%PNP EP/3 wt%PNS EP/5 wt%PNS EP/7 wt%PNS a

LOI (%)

23.3 32.4 33.5 34.0 29.1 30.5 32.2

UL-94 (3.2 mm) t1(s)a




>50 25 ± 3 8±2 25 ± 3 3±1 2±1 19 ± 3

/ 2±1 3±1 2±1 26 ± 3 22 ± 2 5±2

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

Yes No No No No No No

Average combustion duration after the first (t1) and the second ignitions (t2).

3.2. Flame retardancy The flame retardancy of EP, EP/PNP and EP/PNS was evaluated by LOI, the vertical burning and microscale combustion calorimeter (MCC) tests. The flame retardancy results are summarized in Table 3, and the digital photos of EPs captured after LOI and vertical burning tests are displayed in Fig. 2. The LOI value of EP was only 23.3%; thus during the vertical burning test, it completely burned out within 50 s after the first ignition. It is discernible from Fig. 2 that almost no char was formed from EP after the LOI test. On the contrary, PNP- and PNS-containing EPs form intumescent chars after burning; thus the LOI values EP/PNP samples increased from 23.3% to 32.4%, 33.5% and 34.0% with the increasing in PNP loading from 0 to 3 wt%, 5 wt% and 7 wt%, respectively. The EP/PNS samples yielded little lower LOI values than those of EP/PNP ones under the same loading of flame retardants; it can be attributed to the better flame inhibition of PNP than PNS in the gaseous phase [36]. Although only 3 wt% of both PNP and PNS endowed EP with the UL94 V-1 rating during the vertical burning test, EP with a higher

Fig. 2. Digital photos of EP and flame-retardant EPs after LOI and vertical burning tests.

loading of PNP or PNS could not pass the V-0 rating due to the presence of embers after the extinction of the flame. It is observable from Fig. 2 that very small damages occurred on the bottom surface of all flame-retardant EPs after the vertical burning test. EP/PNP and EP/PNS composites revealed different colors (transparency vs. black) due to different colors of PNP and PNS (white and black solids, respectively). MCC works on the principle of oxygen consumption calorimetry and can evaluate the flame retardancy of materials through milligram level samples [37]. Fig. 3 displays the heat release rate (HRR)

Table 2 Typical parameters obtained from TG analysis. Sample

T5% ( C)

Tmax ( C)

amax (%  C1)

Residue at 700  C (wt%) Test values

Calculated values

EP EP/3 wt%PNP EP/5 wt%PNP EP/7 wt%PNP EP/3 wt%PNS EP/5 wt%PNS EP/7 wt%PNS

372.9 368.9 363.4 358.9 354.1 354.1 333.8

392.1 390.7 392.1 391.2 390.3 374.9 387.4

2.00 1.67 1.46 1.31 1.21 0.97 0.87

16.2 19.0 19.4 20.1 21.6 22.9 23.8

/ 15.9 15.7 15.6 16.5 16.8 17.0

B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227


curves of EP and flame-retardant EPs, and the parameters obtained from the MCC test including peak heat release rate (pHRR), total heat released (THR), and temperature at pHRR (Tp) are presented in Table 4. The pHRR values of EP/3 wt%PNP, EP/5 wt%PNP, and EP/7 wt %PNP decreased from 727.3 W g1 (for EP) to 607.1 W g1, 538.4 W g1, and 531.0 W g1, respectively. With the addition of 3 wt %, 5 wt%, and 7 wt% PNS, the pHRR values of EPs were reduced by 45.7%, 51.4%, and 56.4%, respectively, thus suggesting better flame retardancy for EP/PNS samples. Moreover, the Tp values of flameretardant EPs gradually decreased with the increasing content of PNP or PNS. EP/PNS samples resulted in much lower pHRR, THR and Tp values than those of EP and EP/PNP ones due to the earlier decomposition and protective char forming of EP/PNS during heating and these results are in accordance with TGA observations. 3.3. Fire behavior Fig. 3. Heat release rate curves of EP and flame-retardant EPs.

Table 4 Combustion parameters of the EP and its composites obtained from MCC tests. Samples

pHRR (W g1)

THR (kJ g1)

Tp (ºC)

EP EP/3 wt%PNP EP/5 wt%PNP EP/7 wt%PNP EP/3 wt%PNS EP/5 wt%PNS EP/7 wt%PNS

727.3 601.7 538.4 531.0 393.4 353.2 317.0

36.4 33.6 33.5 33.5 29.5 29.4 28.8

396.6 390.5 387.0 388.3 385.7 379.6 374.1

Cone calorimeter test is an effective method to evaluate the fire behavior and the fire hazard of materials on a bench scale. The HRR, the THR, the total smoke production (TSP), and the smoke production rate (SPR) curves of EP, EP/7 wt%PNP, and EP/7 wt%PNS are displayed in Fig. 4, and the obtained data are summarized in Table 5 Cone calorimeter data for EP, EP/7 wt%PNP and EP/7 wt%PNS. Samples


EP/7 wt%PNP

EP/7 wt%PNS

TTI (s) THR (MJ m2) PHRR (kW m2) Av-EHC (MJ kg2) tp (s) FIGRA (kW m2 s1) TSP (m2) Residue (wt%)

72 95 1010 25 141 7.17 89 6.9

54 61 748 21 120 6.24 45 15.0

67 82 520 23 125 4.16 32 15.7

Fig. 4. Cone calorimeter test curves for EP, EP/7 wt%PNP and EP/7 wt%PNS. (A) HRR, (B) THR, (C) TSP, and (D) SPR.


B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227

Fig. 5. Digital photos of the char residues for (A) EP, (B) EP/7 wt%PNP, and (C) EP/7 wt%PNS.

Fig. 6. TGA-IR results for (a) EP, (b) EP/7 wt%PNP and (c) EP/7 wt%PNS.

Table 5. The time to ignition (TTI) values of EP/7 wt%PNP and EP/ 7 wt%PNS were found to be shorter than that of EP, and it can be ascribed to the early decomposition of flame retardants. In

comparison to EP, the peak of the heat release rate (PHRR) for EP/ 7 wt%PNP and EP/7 wt%PNS was reduced by 25.9% and 48.5%, respectively. Moreover, the HRR curve of EP/PNS manifested a

B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227


Fig. 7. Intensities of gaseous products of EP, EP/7 wt%PNP, and EP/7 wt%PNS against time: (a) Total; (b) Hydrocarbons; (c) Aromatic compounds; (d) Ether.

typical charring feature. With the addition of 7 wt% PNP and PNS, the THR values of EP/PNP and EP/PNS were reduced from 95.2 MJ m2 (for EP) to 61.3 MJ m2 and 82.3 MJ m2, respectively. The effective heat of combustion (EHC) was used to measure the combustion efficiency of EPs in the gaseous phase. The average EHC (Av-EHC) values of flame-retardant EPs were found to be lower than that of EP (Table 5), thus indicating their better flame inhibition abilities in the gaseous phase [38]. The main reason for lower Av-EHC values is that both PNP and PNS effectively inhibited the flammable gases of EP during combustion. It is worth noting that the Av-EHC value of EP/7 wt%PNP was lower than that of EP/7 wt% PNP; it suggests that PNP had a better flame inhibition ability than PNS and resulted in higher LOI values. Fire growth rate (FIGRA) is an important parameter to assess the fire hazard of material and can be obtained by dividing the PHRR by the time to PHRR (tp) [39]. It is observable from Table 5 that with the incorporation of 7 wt% PNP and PNS, the FIGRA values of EP/PNP and EP/PNS decreased from 7.17 kW m2 s1 (for EP) to 6.24 kW m2 s1 and 4.16 kW m2 s1, respectively, thus suggesting a delay in flame-flicker time and a reduction in fire hazard [40]. In addition, the TSP values of EP/7 wt% PNP and EP/7 wt%PNS decreased from 89 m2 (for EP) to 45 m2 (reduction of by 49%) and 32 m2 (reduction of by 64%), respectively, thus indicating outstanding smoke suppression performances by both PNP and PNS. Moreover, the smoke production rate (SPR) of EP/7 wt%PNS was found to be lower than those of EP and EP/7 wt% PNP ochar residue for EP/PNP and EP/PNS was enhanced from 6.9 wt% (for EP) to 15.0 wt% and 15.7 wt%, respectively. It is

discernible from Fig. 5 that flame-retardant EPs formed intumescent and unbroken chars, whereas EP led to thin, fragile, and damaged chars (Fig. 5B and C). Therefore, although both PNP and PNS reduced the fire hazard of EP during combustion, PNS endowed

Fig. 8. Gas chromatograms of the pyrolysis products for (a) PNP and (b) PNS at 500  C.


B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227

Table 6 Possible chemical structures of pyrolysis products for PNP and PNS at 500  C. No.

PNP Molecular mass(m/z)

No. Structure

PNS Molecular mass(m/z)













4, 6, 9










15, 19, 20




16, 17, 18, 21


EP with much lower heat and smoke release. 3.4. Gaseous analysis In order to analyze the thermal degradation mechanism, pyrolysis volatiles from EP, EP/7 wt%PNP, and EP/7 wt%PNS were examined by TG-FTIR, and the corresponding infrared spectra are displayed in Fig. 6. EP, EP/7 wt%PNP, and EP/7 wt%PNS exhibited almost the same signal characteristics peaks for water (40003400 cm1), hydroxyl (3649 cm1), hydrocarbon (2974 cm1), CO2 (2352 cm-1), carbonyl compound (1747 cm1), aromatic compound (1508e1512, 1605e1615, 827 and 752 cm1), and ether (1258 and 1171 cm1) [41]. The weak signals for P-containing volatiles as well as S and Si-containing compounds can be ascribed to low PNP/PNS loadings in EP, and relevant signals of these typical compounds got overlapped by the absorption peaks of aromatic and ether compounds. In order to further analyze the changes in thermally decomposed products, the absorption of total and typical products of EPs against time was examined (Fig. 7). The intensities of thermally decomposed products for both EP/PNP and EP/PNS were found to be lower than that of EP. Therefore, the introduction of PNP and PNS effectively suppressed the release of hydrocarbons, aromatic compound and ether, which were thought as flammable gases in previous report [42]. PNP and PNS manifested similar inhibiting effect for EP decomposition. In order to study the difference in pyrolysis behavior of PNP and PNS, Py-GC/MS was performed to identify different pyrolysis products of PNP and PNS. The recorded gaschromatograms of pyrolysis products are exhibited in Fig. 8, and the corresponding molecular formulae confirmed by mass spectrometry are summarized in Table 6. In addition, the pyrolysis pathways of PNP and PNS at 500  C are proposed in Scheme 2, and the structures of all detected chemicals and possible unstable intermediates (in dashed box) are illustrated. It is worth noting that both PNP and PNS yielded some same pyrolysis products including ethoxy silane (m/z ¼ 76), benzene (m/z ¼ 78), diphenylphosphine (m/z ¼ 186) and diphenylphosphoramide (m/z ¼ 217). For PNS,


besides diphenylphosphine and diphenylphosphoramide, phenylphosphine (m/z ¼ 110) can be detected during pyrolysis. For PNP, some other P-containing substances, such as diphenylphosphinamine (m/z ¼ 201) and triphenylphosphine (m/z ¼ 262), were detected in the gaseous phase. It is noticeable that P-containing fragments for both PNP and PNS were formed by the reactions of relevant intermediates, and these P-containing fragments played a critical flame inhibiting role in the gaseous phase during combustion. 3.5. Char residue analysis In order to study the effects of char residues on flameretardant EPs, the micro-morphologies of the char layers after combustion were examined by SEM. It is evident from Fig. 9 A and A0 that the outer and inner surfaces of the char layers of EP were very thin and prone to damage due to burning out of EP during combustion. However, EP/7 wt%PNP yielded much thicker char layers (Fig. 9B and B0 ). Furthermore, the addition of 7 wt% PNS endowed EP with complete, compact and continuous char layers after combustion (Fig. 9C and C’), thus resulting in effective flame retardancy for EP during combustion depending on good heat/ oxygen barrier effect. In order to further investigate the condensed flame-retardant action of PNS in EP, EDX tests were carried out to analyze the characteristics of char layers of EP/7 wt%PNS after the cone calorimeter test, and the obtained results are presented in Fig. 9C1, C2 and C3. In addition to C and O elements, different contents of P, Si, S and Ni existed in char layers of EP/7 wt%PNS after combustion. Moreover, the extremely high Ni content (34%) suggests Ni played an important role in condensed flame-retardant action. The content ratios of P/Ni and S/Ni in char layers were, respectively, found as 0.06 and 0.09, which are lower than those of PNS (1.05 and 1.08, respectively). Therefore, lower content ratios demonstrate that a considerable part of P and S elements was released into the gaseous phase during burning, and this finding complies with the results of Py-GC/MS analysis.

B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227


Scheme 2. Pyrolysis pathways of PNP and PNS at 500  C.

Therefore, it can be inferred that the introduction of PNP and PNS effectively suppressed the decomposition of EP and lowered the release of flammable volatiles such as hydrocarbons, aromatic compound and ether. Furthermore, both PNP and PNS revealed almost the same flame-retardant mechanism by pyrolyzing Pcontaining fragments in the gaseous phase. In the condensed phase, In comparison EP/PNP, EP/PNS yielded more compact and continuous char layers after combustion due to the presence of additional S and Ni elements. According to EDX results, in addition to C and O elements, P, Si, S, and Ni also participated in the formation of protective char layers; hence, PNS endowed EP with better flame retardancy. 4. Conclusions In the above-discussed experiment, two multi-element compounds PNP and PNS were used as flame retardants for EP. Results of TGA revealed that both PNP and PNS enhanced the thermal stability and the char yield rate of EP. Moreover, results of LOI,

vertical burning (UL-94), MCC, and cone calorimeter tests proved that the incorporation of PNP and PNS resulted in different degrees of flame retardancy for EP. PNP endowed EP with higher LOI values (32.4%e34.0%), whereas the values for EP/PNS samples ranged between 29.0% and 32.2%. Furthermore, the UL-94 V-1 rating was obtained when the content of PNP or PNS was between 3 wt% and 7 wt%. Results of MCC tests indicated that PNS, in comparison to PNP, endowed EP with better flame retardancy due to lower pHRR and THR values. Cone calorimeter tests expressed that the incorporation of 7 wt% PNS decreased the PHRR and the TSP values of EP by 48.5% and 64%, respectively. In order to further understand the reason for better flame retardancy of PNS, TG-FTIR, Py-GC/MS, and SEM-EDX were employed to analyze the gaseous products and morphologies of char layers. PNP and PNS both manifested a similar mechanism of flame inhibition in the gaseous phase; however, EP/PNS yielded better flame retardancy by forming a stronger protective multi-element char layer after combustion.


B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227

Fig. 9. SEM micrographs of EP composite char after the cone calorimeter test: the outer and inner surface of EP (A, A0 ), EP/7 wt%PNP (B, B0 ) and EP/7 wt%PNS (C, C0 ). EDX elemental mapping (C1, C2) and spectrum (C3) of char residues for EP/7 wt%PNS after the cone calorimeter test.

Acknowledgement Financial support by the National Natural Science Foundation of China (Grant 51403191) is sincerely acknowledged. Also, the authors wanted to acknowledge China Scholarship Council (CSC: 201808140038). Appendix A. Supplementary data Supplementary data to this article can be found online at References [1] F.L. Jin, X. Li, S.J. Park, Synthesis and application of epoxy resins: a review, J. Ind. Eng. Chem. 29 (2015) 1e11. [2] R. Auvergne, S. Caillol, G. David, B. Boutevin, J.P. Pascault, Biobased thermosetting epoxy: present and future, Chem. Rev. 114 (2014) 1082e1115. [3] E.D. Weil, S. Levchik, A review of current flame retardant systems for epoxy resins, J. Fire Sci. 22 (2004) 25e40. [4] C. Hamciuc, T. Vlad-Bubulac, D. Serbezeanu, I.D. Carja, E. Hamciuc, G. Lisa, rez, Environmentally friendly fire-resistant epoxy resins based on a V.F. Pe new oligophosphonate with high flame retardant efficiency, RSC Adv. 6 (2016) 22764e22776. [5] S.Y. Lu, I. Hamerton, Recent developments in the chemistry of halogen-free flame retardant polymers, Prog. Polym. Sci. 27 (2002) 1661e1712. [6] A.B. Morgan, The future of flame retardant polymers-Unmet needs and likely new approaches, Polym. Rev. 59 (2019) 25e54. [7] B. Schartel, Phosphorus-based flame retardancy mechanisms-old hat or a starting point for future development? Materials 3 (2010) 4710e4745. [8] H. Horacek, R. Grabner, Advantages of flame retardants based on nitrogen compounds, Polym. Degrad. Stabil. 54 (1996) 205e215.

[9] I. Van der Veen, J. de Boer, Phosphorus flame retardants: properties, production, environmental occurrence, toxicity and analysis, Chemosphere 88 (2012) 1119e1153. € ring, [10] B. Schartel, U. Braun, A.I. Balabanovich, J. Artner, M. Ciesielski, M. Do R.M. Perez, J.K. Sandler, V. Altst€ adt, Pyrolysis and fire behaviour of epoxy systems containing a novel 9, 10-dihydro-9-oxa-10-phosphaphenanthrene10-oxide-(DOPO)-based diamino hardener, Eur. Polym. J. 44 (2008) 704e715. [11] K.A. Salmeia, S. Gaan, An overview of some recent advances in DOPOderivatives: chemistry and flame retardant applications, Polym. Degrad. Stabil. 113 (2015) 119e134. [12] Y. Qiu, L.J. Qian, H. Feng, S. Jin, J. Hao, Toughening effect and flame-retardant behaviors of phosphaphenanthrene/phenylsiloxane bigroup macromolecules in epoxy thermoset, Macromolecules 51 (2018) 9992e10002. [13] L.J. Qian, L.J. Ye, Y. Qiu, S.R. Qu, Thermal degradation behavior of the compound containing phosphaphenanthrene and phosphazene groups and its flame retardant mechanism on epoxy resin, Polymer 52 (2011) 5486e5493. [14] G. Yang, W.H. Wu, Y.H. Wang, Y.H. Jiao, L.Y. Lu, H.Q. Qu, X.Y. Qin, Synthesis of a novel phosphazene-based flame retardant with active amine groups and its application in reducing the fire hazard of epoxy resin, J. Hazard Mater. 366 (2019) 78e87. [15] S. Qiu, Y. Zhou, X. Zhou, T. Zhang, C. Wang, R.K. Yuen, W. Hu, Y. Hu, Air-stable polyphosphazene-functionalized few-layer black phosphorene for flame retardancy of epoxy resins, Small 15 (2019) 1805175. [16] Y. Tan, Z.B. Shao, X.F. Chen, J.W. Long, L. Chen, Y.Z. Wang, Novel multifunctional organiceinorganic hybrid curing agent with high flame-retardant efficiency for epoxy resin, ACS Appl. Mater. Interfaces 7 (2015) 17919e17928. [17] Y.J. Xu, L. Chen, W.H. Rao, M. Qi, D.M. Guo, W. Liao, Y.Z. Wang, Latent curing epoxy system with excellent thermal stability, flame retardance and dielectric property, Chem. Eng. J. 347 (2018) 223e232. , J.A. Reina, Silicon-containing flame retardant epoxy [18] L.A. Mercado, M. Galia resins: synthesis, characterization and properties, Polym. Degrad. Stabil. 91 (2006) 2588e2594. [19] W.C. Zhang, X.M. Li, R.J. Yang, Study on flame retardancy of TGDDM epoxy resins loaded with DOPO-POSS compound and OPS/DOPO mixture, Polym. Degrad. Stabil. 99 (2014) 118e126. [20] Y.Q. Shi, B. Yu, Y.Y. Zheng, J. Guo, B.H. Chen, Z.M. Pan, Y. Hu, A combination of

B. Zhao et al. / Polymer Degradation and Stability 167 (2019) 217e227







[27] [28]

[29] [30]


POSS and polyphosphazene for reducing fire hazards of epoxy resin, Polym. Adv. Technol. 29 (2018) 1242e1254. P.J. Chao, Y.J. Li, X.Y. Gu, D.D. Han, X.Q. Jia, M.Q. Wang, T.F. Zhou, T. Wang, Novel phosphorus-nitrogen-silicon flame retardants and their application in cycloaliphatic epoxy systems, Polym. Chem. 6 (2015) 2977e2985. D. Yu, M. Kleemeier, G.M. Wu, B. Schartel, W.Q. Liu, A. Hartwig, Phosphorus and silicon containing low-melting organic-inorganic glasses improve flame retardancy of epoxy/clay composites, Macromol. Mater. Eng. 296 (2011) 952e964. R.K. Jian, P. Wang, W.S. Duan, J.S. Wang, X.L. Zheng, J.B. Weng, Synthesis of a novel P/N/S-containing flame retardant and its application in epoxy resin: thermal property, flame retardance, and pyrolysis behavior, Ind. Eng. Chem. Res. 55 (2016) 11520e11527. €ring, Bridged phosphorylated C. Klinkowski, S. Wagner, M. Ciesielski, M. Do diamines: synthesis, thermal stability and flame retarding properties in epoxy resins, Polym. Degrad. Stabil. 106 (2014) 122e128. M. Rakotomalala, M. Ciesielski, T. Zevaco, M. Doering, New phosphacyclic molecules and their application as flame retardants for epoxy resins, Phosphorus Sulfur 186 (2011) 989e996. H.D. Lu, L. Song, Y. Hu, A review on flame retardant technology in China. Part II: flame retardant polymeric nanocomposites and coatings, Polym. Adv. Technol. 22 (2011) 379e394. L. Chen, Y.Z. Wang, A review on flame retardant technology in China. Part I: development of flame retardants, Polym. Adv. Technol. 21 (2010) 1e26. Z.H. Cao, Y. Zhang, L.P. Zhao, M. Peng, Z.P. Fang, M. Klatt, Improving the flame retardancy and mechanical properties of high-density polyethylene-g-maleic anhydride with a novel organic metal phosphonate, J. Anal. Appl. Pyrolysis 102 (2013) 154e160. M. Lewin, M. Endo, Catalysis of intumescent flame retardancy of polypropylene by metallic compounds, Polym. Adv. Technol. 14 (2003) 3e11. T. Tang, X.C. Chen, H. Chen, X.Y. Meng, Z.W. Jiang, W.G. Bi, Catalyzing carbonization of polypropylene itself by supported nickel catalyst during combustion of polypropylene/clay nanocomposite for improving fire retardancy, Chem. Mater. 17 (2005) 2799e2802. X. Wang, W.Y. Xing, X.M. Feng, B. Yu, H.D. Lu, L. Song, Y. Hu, The effect of metal oxide decorated graphene hybrids on the improved thermal stability and the reduced smoke toxicity in epoxy resins, Chem. Eng. J. 250 (2014)


214e221. [32] Y.M. Yu, P.A. Song, C.D. Jin, S.Y. Fu, L.P. Zhao, Q. Wu, J.W. Ye, Catalytic effects of nickel (cobalt or zinc) acetates on thermal and flammability properties of polypropylene-modified lignin composites, Ind. Eng. Chem. Res. 51 (2012) 12367e12374. [33] P.H. Zhao, K.K. Xiong, W.T. Wang, Y.Q. Liu, Preparation of a halogen-free P/N/Si flame retardant monomer with reactive siloxy groups and its application in cotton fabrics, Chin. J. Chem. Eng. 25 (2017) 1322e1328. [34] X.F. Liu, X. Li, J. Yan, Synthetic and structural studies of the mononuclear nickel (II) ethanedithiolate complexes with chelating N-substituted bis (diphenylphosphanyl) amine, Polyhedron 85 (2015) 482e487. [35] T.R. Hurford, D.F. Boltz, Indirect ultraviolet spectrophotometric and atomic absorption spectrometric methods for determination of phosphorus and silicon by heteropoly chemistry of molybdate, Anal. Chem. 40 (1968) 379e382. [36] B. Zhao, L. Chen, J.W. Long, H.B. Chen, Y.Z. Wang, Aluminum hypophosphite versus alkyl-substituted phosphinate in polyamide 6: flame retardance, thermal degradation, and pyrolysis behavior, Ind. Eng. Chem. Res. 52 (2013) 2875e2886. [37] R.E. Lyon, R.N. Walters, A Microscale Combustion Calorimeter, US Department of Transport, Washington, 2002. Report Number: DOT/FAA/AR-01/117. [38] L.J. Russell, D.O.C. Marney, V.P. Dowling, in: C.A. Wilkie, G.L. Nelson (Eds.), Fire and Polymers IV Materials and Concepts for Hazard Prevention: in Search of Synergy Using Conventional Fire Tests, ACS, 2004, pp. 237e251. n, T.R. Hull, A.A. Stec, J.C. Ronda, M. Galia , V. C [39] M. Sacrista adiz, Cone calorimetry studies of fire retardant soybean-oil-based copolymers containing silicon or boron: comparison of additive and reactive approaches, Polym. Degrad. Stabil. 95 (2010) 1269e1274. lez, V.B. Heeralal, D.Y. Wang, Covalent assembly of MCM-41 [40] Z. Li, A.J. Gonza nanospheres on graphene oxide for improving fire retardancy and mechanical property of epoxy resin, Compos. B Eng. 138 (2018) 101e112. [41] X. Wang, Y. Hu, L. Song, W.Y. Xing, H.D. Lu, Thermal degradation mechanism of flame retarded epoxy resins with a DOPO-substitued organophosphorus oligomer by TG-FTIR and DP-MS, J. Anal. Appl. Pyrolysis 92 (2011) 164e170. [42] Y. Liu, Q.Q. Wang, Z.M. Jiang, C.J. Zhang, Z.F. Li, H.Q. Chen, P. Zhu, Effect of chitosan on the fire retardancy and thermal degradation properties of coated cotton fabrics with sodium phytate and APTES by LBL assembly, J. Anal. Appl. Pyrolysis 135 (2018) 289e298.