Improvement of the flame retardancy of wood-fibre/polypropylene composites with ideal mechanical properties by a novel intumescent flame retardant system

Abstract

To improve the flame retardancy and maintain the ideal mechanical properties of the widely used wood fibre (WF) reinforced polypropylene (PP) composite (WPC), a novel intumescent flame retardant (IFR) system consisting of poly[N⁴-bis(ethylenediamino)-phenyl phosphonic-N²,N⁶-bis(ethylenediamino)-1,3,5-triazine-N-phenyl phosphonate] (PTPA) and ammonium polyphosphate (APP) was developed. Addition of APP, a high performance flame retardant, improves flame retardancy but has a negative effect on flexural properties and the Notched Izod impact strength. Combining self-designed PTPA with WPC has a positive effect on the mechanical properties whereas it just slightly increases the flame retardancy. The flame retarded WPC system with both PTPA and APP, i.e. WPC/PTPA/APP, combines bilateral advantages. The selected formula has a flexural modulus of 4.4 GPa, flexural strength of 37.2 MPa and impact strength of 1.8 kJ m⁻², only the flexural strength is slightly lower than that of neat WPC. And the formula has a high limiting oxygen index (LOI) of 31.5%, which passes the UL-94 V-0 rating for a vertical burning test. The results of cone calorimetry show that the heat release rate (HRR) of the WPC/PTPA/APP system is significantly reduced compared to the neat WPC. The pyrolysis and char residue of the WPC/PTPA/APP system were investigated by thermogravimetric analysis (TGA), Fourier transform infrared (FTIR) and X-ray photoelectron spectroscopy (XPS). These results reveal the flame-retardant mechanism that PTPA/APP can synergistically catalyze the etherification, dehydration and char formation of WPC.

---

Introduction

Chemical products that are based on renewable resources are an important issue to sustainable development.¹ Polypropylene (PP), which is widely used in cars, cables, electronics, and constructive materials, contributes a large percentage to waste and polluting the environment^2,3 due to its non-degradability. The utilization of renewable resources such as wood fibres (WF) into the PP matrix, which results in a WF/PP composite or WPC, is being actively pursued due to improvement in material properties. Furthermore, this natural fibre composite has attracted intensive attention from both academic and industrial societies due to its renewability, cost-effectiveness, long-term performance, shape flexibility and anti-corrosion etc. They are widely used in building, transportation, packaging, and consumer products.^4–8 As organic materials, polymers and wood fiber are sensitive to flame. Especially PP which is a highly combustible material. In combustion, PP releases large amounts of heat and easily melt flow, which promotes the spreading of flames. To safely apply this material in public places, it is hence necessary to flame retardant this composite. Unfortunately, the incorporation of the flame retardants always significantly deteriorates the mechanical properties of composites.^9–14 Therefore, how to flame retard wood/plastic composites without sacrificing mechanical properties is one of the most critical issues for their wide applications.

Suppakarn et al.¹⁵ added magnesium hydroxide (Mg(OH)₂) and zinc borate into sisal/PP composites. Owing to the results of horizontal burning rate and mechanical properties, the addition of Mg(OH)₂ and zinc borate improved flame retardancy of sisal/PP composites without sacrificing their mechanical properties. Abu Bakar et al.¹⁶ used three types of flame retardants (FRs) based on ammonium polyphosphate to improve the FR properties of wood flour-filled polypropylene composites. The presence of 30 wt% FRs in the PP/WF composites improved the LOI value to 24%, UL-94 achieved V-0. The mechanical property study revealed that, with the incorporation of FRs, the tensile strength and flexural strength were decreased, but the tensile and flexural moduli were increased in all cases. Schartel et al.¹⁷ added an ammonium polyphosphate based flame retardant (APS) and expandable graphite (EG) in polypropylene/flax biocomposites. In terms of peak of the heat release, PP/flax/EG25 has the best fire retardancy. In terms of fire hazards such as CO and smoke production, PP/flax/APS25 shows an increase whereas PP/flax/EG15 and PP/flax/EG25 shows an increasingly improvement with increasing amount of expandable graphite.

Sain et al.¹⁸ studied the effect of adding boric acid or zinc borate combining with magnesium hydroxide sawdust/rice husk filled polypropylene composites. It was found that magnesium hydroxide can effectively reduce the flammability (almost 50%) of natural fibre filled polypropylene composites. Inorganic particles reinforce PP composites by their inherent strength and restricting chain movement of the polymers. However, addition of inorganic particles always requires a high content to fulfil the flame retardant requirements for their comparatively low effect. Furthermore, effective synergetic effect was less observed in inorganic flame retardants reinforced organic composites in contrast to organic ones. Therefore, organic flame retardants combining other merits, environment-friendly for instance, are more desirable in polyolefin composites. In our previous work,¹⁹ a single flame retardant, ethanolamine-APP or ETA-APP, was used to prepare flame-retardant wood/PP composite. The flame-retardant efficiency is improved greatly: the limiting oxygen index is 43.0%, and the vertical burning test can pass UL-94 V-0 rating. The flexural properties of WPC/30 wt% ETA-APP increase a lot in comparison with WPC/30 wt% APP. But the mechanical properties of WPC/30 wt% ETA-APP decreased in comparison with WPC. Based on the reported results in present literature, it seems inevitably to sacrifice some mechanical strength to gain lower burning capacity with a high loading of flame retardants. To improve the flame retarding efficiency, environment-friendly property and to uphold the mechanical properties at the same time, the novel highly-efficient halogen-free flame retardants are desirable. We designed compounds based on triazine and its derivatives with high thermal stability and high flame retarding efficiency.^20,22 These studies indicated that phosphorus oxynitride structure was helpful for the formation of a stable char layer.²²

In the present work, we adopted the newly designed polymeric flame retardant, poly[N⁴-bis(ethylenediamino)-phenyl phosphonic-N²,N⁶-bis(ethylenediamino)-1,3,5-triazine-N-phenyl phosphonate] or PTPA,²² combining with rich acid source (APP) to prepare a wood fibre/PP composite with enhanced mechanical properties and flame retardancy. The mechanical properties and flame retardancy of the novel intumescent flame retardants modified WPC were investigated, and corresponding mechanisms were studied.

Results and discussion

Mechanical properties and flammability of WPC/APP or WPC/PTPA

Mechanical properties of WPC with a single FR. As WPC is widely used as constructive materials, the mechanical properties of WPC containing FR are very important, and were first investigated (Fig. 1). Fig. 1a shows that the flexural modulus of WPC is ca. 2 GPa when WF content is 20 wt%. The increase of WF content to 30 wt% or 40 wt% only results in a flexural modulus of ca. 2.5 GPa. The flexural modulus increases sharply when WF content is 60 wt%, which is ca. 4.5 GPa. However, the flexural strength increases from ca. 44 MPa to ca. 47 MPa when WF content changes from 20 wt% to 30 wt%, and then decrease to ca. 37 MPa when the content of WF is further increased to 40 wt% or 50 wt%. The flexural strength of WPC with a WF content of 60 wt% moderately increases again to ca. 40 MPa. It knows that higher filler content always results in greater modulus.²¹ To maintain a PP-based composite nature, the WPC with a WF content of 60 wt% (W₆₀PC), which has the highest comprehensive flexural properties, was chosen as the framework to be flame retarded in this work.

Fig. 1 (a) Flexural strength and flexural modulus of WPC with different content of wood fibre; (b) flexural strength and flexural modulus of WPC with 60 wt% of wood fibre (W₆₀PC), W₆₀PC with 30 wt% APP((W₆₀PC)₇₀/APP₃₀) and W₆₀PC with 30 wt% PTPA((W₆₀PC)₇₀/PTPA₃₀); (c) impact strength of W₆₀PC, (W₆₀PC)₇₀/APP₃₀ and (W₆₀PC)₇₀/PTPA₃₀.

Fig. 1b shows the flexural properties of W₆₀PC with a single flame retardant of PTPA or APP. The addition of flame retardants shows a generally negative effect on the flexural properties of the composites. Especially, the flexural strength of (W₆₀PC)₇₀/APP₃₀ system decreases by about 30% compared to that of W₆₀PC. Additionally, during the notched impact tests of W₆₀PC with a single component FR (Fig. 1c), addition of PTPA makes the notched impact strength of WPC at the same level or even higher, but addition of 30 wt% APP decreases the corresponding values. Poorer compatibility of APP than PTPA in WPC may cause this phenomenon. The reason could be attributed to the lignin in the wood fibre. Lignin is the secondary component in wood fibres, which is a complex polymer of aromatic alcohols. Therefore, the WPC matrix has a better interfacial adhesion and stronger interfacial regions between PTPA than those of APP. During the impact test, cracks travel through the composite along the weaker interfacial regions. The weak interfacial regions cannot resist crack propagation as effectively as the well adhesion ones, hence reducing the mechanical strengths.¹⁸ Incorporation of FRs is also thought to hinder the movement of the polymer chains, thereby lowering the ability of the system to absorb energy during fracture propagation.

In order to better understand the differences in mechanical properties for different composite systems, we further investigated the composites by SEM. Fig. 2 shows the morphological structures of APP, PTPA and the cryogenic fracture surface of W₆₀PC, (W₆₀PC)₇₀/PTPA₃₀ and (W₆₀PC)₇₀/APP₃₀, respectively. The surfaces of APP (Fig. 2a) are clean in contrast to rough ones for PTPA (Fig. 2b). Besides, PTPA particles stick to each other. From Fig. 2c, it is easy to observe that WF are well embedded in the PP matrix. Addition of coupling agent (PP-g-MHA) was not fully cured the voids and cavities in the W₆₀PC composite, which implies the interfacial adhesion between the WF and PP is not strong, for the reason that WF is a polar material comparing to a nonpolar polyolefin. For its polar nature, WF is prone to agglomerate, which result in poor interfacial adhesion inside the composite. However, as shown in Fig. 1a, the moduli increase with higher content of WF, which is resulted by the incorporation of rigid fibre reinforcements into the polyolefin. It also shows the ability of WF to impart greater stiffness to the composite. This result is in agreement with previous studies of lignocellulosic-filled thermoplastics.^16,21 Some studies have shown that if adhesion was not good, there would be voids around the WF and pull-out of WF could be observed.^23,25 Comparing with Fig. 2d and e, Fig. 2d shows that there is no aggregation of PTPA particles in the composite, PTPA disperses more uniformly and even plays a role as glue in the matrix apparently. There is no obvious gap at the contact zone between WF and the matrix. The surface of (W₆₀PC)₇₀/APP₃₀ (Fig. 2e) shows randomly scattered APP particles with clean surface, and some of the particles even agglomerated in the fracture surface of WPC (circles in Fig. 2e) and some voids around the WF. This phenomenon indicates that APP particles can be pulled out easily from the WPC matrix by breaking the interface because of the poor adhesion between APP and the WPC matrix.^24,25 These results supported the analysis for the differences of the mechanical properties.

Fig. 2 SEM micrographs of (a) APP, (b) PTPA and cryogenic fracture surface of (c) W₆₀PC, (d) (W₆₀PC)₇₀/PTPA₃₀ and (e) (W₆₀PC)₇₀/APP₃₀, respectively. WF are marked by arrows and APP are marked by circles respectively for clarity.

Flammability of WPC with a single FR. In order to better understand the mechanism of flame retardancy, the flammability of WPC with a single component FR were conducted first. Table 1 lists the limiting oxygen index (LOI) and vertical burning test (UL-94) results of W₆₀PC, (W₆₀PC)₇₀/APP₃₀ and (W₆₀PC)₇₀/PTPA₃₀. The W₆₀PC is an easy flammable material with a LOI value of 19.0%. In fact, WF has flame-retarded PP by playing a role as promoter for char formation which delayed the degradation of PP (see below).¹⁶ The resultant char reduces the burning rate of WPC by preventing the oxygen from reaching the pyrolysis products. However, the char was not sufficient. In the absence of FRs, W₆₀PC burned easily with accompanying melt dripping, which indicates WPC is sensitive to flame. As the entire samples were consumed during the burning, W₆₀PC could be classified as no rating according to UL-94 vertical burning standard specifications.

Table 1 Flame retardant properties of WPC/FR composites

Formulation	LOI (%)	UL-94
		Dripping	Rating
W60PC	19.0	Little	No rating
(W60PC)70/PTPA30	25.5	No	V-1
(W60PC)70/APP30	29.0	No	V-0

---

For (W₆₀PC)₇₀/PTPA₃₀, the LOI value is 25.5% and it is V-1 according to UL-94 standard specifications. For (W₆₀PC)₇₀/APP₃₀, the LOI value is 29.0% and it passes V-0 rank for UL-94. No melt dripping was observed from the burning samples of these composites. The presence of 30 wt% PTPA or APP in the W₆₀PC improved the flame retardant properties of the composites. Char formation on the composite surfaces can be observed for both W₆₀PC/single FR systems. Charring would effectively lead to a self-extinguishment during combustion.

The cone calorimeter is also used to evaluate the flame retardant performances of W₆₀PCs. The heat release rate (HRR) is used to evaluate the flammability of W₆₀PC, and the relevant results are shown in Fig. 3. Advances in fire research have emphasized the heat release rate as the primary fire hazard indicator for a material.²⁶ The HRR curve of W₆₀PC shows one peak in the beginning and sustain above 250 kW m⁻² for about 130 s. After addition of FR, HRR is decreased in comparison with W₆₀PC. Wood fibres are constituted of cellulose and lignin and highly oxygen-containing. When WPC/APP composite are exposed to fire or heat, the FR decomposes, commonly into poly(phosphoric acid) and ammonia. Ammonia forms the blowing source to dilute the flammable volatiles and promotes intumescent layers. The poly(phosphoric acid) reacts with hydroxyl or other groups of a synergist to form a unstable phosphate ester. The dehydration of the phosphate ester follows next. A carbon foam is built up on the surface against the heat source. The char acts as an insulation layer and prevents volatile, combustible gases or further decomposition of the material.²⁷ PTPA is a novel polymeric flame retardant containing triazine ring, benzene ring and phosphorus oxynitride structure which may endow a material better thermal stability and charring ability.²² During combustion, PTPA released nonflammable gases to dilute the flammable volatiles of decomposition of WPC. It also promoted the charring process. Comparing with APP, the minor reduction of HRR may due to the lack of acid source in the molecular structure in PTPA.

Fig. 3 (a) Heat release rate (HRR) and (b) total heat release (THR) curves of W₆₀PC, (W₆₀PC)₇₀/PTPA₃₀ and (W₆₀PC)₇₀/APP₃₀.

Mechanical properties and flammability of W₆₀PC/PTPA/APP

Mechanical properties of W₆₀PC/PTPA/APP. According to the aforementioned study, we can find out that the addition of PTPA maintains the mechanical properties of WPC and addition of APP imparts the material good flame-retardant properties. Therefore, it is promising to obtain a flame-retardant wood fibre reinforced PP composite with good mechanical properties by combining both PTPA and APP. Therefore, WPC/IFR composites were prepared with a total content of FRs of 30 wt%. To achieve a better flame retardancy, the minimum content of APP was limited to 15 wt%. The ratio of PTPA to APP was systematically adjusted with 1 : 1, 1 : 2 or 1 : 3. Fig. 4a shows flexural strength and flexural modulus of W₆₀PC combined with both FRs. Comparing with the results presented in Fig. 1b and c, gradual increasing PTPA content in WPC/APP composite strengthened the composites. This result implies that PTPA promoted the integration of APP with WPC. Especially, when the ratio of PTPA/APP was 1 : 2, i.e. the sample (W₆₀PC)₇₀/PTPA₁₀/APP₂₀, it gives the relatively satisfied mechanical properties among all of the WPC/IFR systems, the flexural modulus is 4.4 GPa and the impact strength is 1.8 kJ m⁻², which are the same with W₆₀PC without addition of FR. Only the flexural strength slightly decreased from 40.2 MPa of W₆₀PC to 37.2 MPa.

Fig. 4 (a) Flexural strength and flexural modulus of (W₆₀PC)₇₀/PTPA/APP systems, (b) impact strength of (W₆₀PC)₇₀/PTPA/APP systems.

Flammability of WPC/PTPA/APP. To investigate the flammability of the intumescent flame-retardant composites, the limiting oxygen index (LOI) and UL-94 combustion tests were made. The data are listed in Table 2. The LOI value of (W₆₀PC)₇₀/PTPA₁₅/APP₁₅ significantly increases from 19.0% of W₆₀PC to 28.0% and pass V-0 class of UL-94 rating system. The LOI value of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ is further increased to as high as 31.5% and the UL-94 rating is V-0, showing an excellent synergy between APP and PTPA when comparing with single FR composites. The LOI values were not further improved when APP content increased from 2 : 1 to 3 : 1 comparing with PTPA. As aforementioned results, further increase APP content will deteriorate the mechanical properties of the composites. The results above show that there exists a good synergistic effect between PTPA and APP, which endows the formula of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ good mechanical properties and flame retardancy. Therefore, the composites with PTPA/APP ratio of 1 : 2 were chosen to do the analysis for the combustion process and the flame retardant mechanism.

Table 2 Formulation and flame retardant properties of WPC with double flame retardants

Formulation	LOI (%)	UL-94
		Dripping	Rating
(W60PC)70/PTPA15/APP15	28.0	—	V-0
(W60PC)70/PTPA10/APP20	31.5	—	V-0
(W60PC)70/PTPA7/APP23	31.5	—	V-0

---

Fig. 5 and Table 3 show the results of cone calorimeter of the flame-retardant WPCs at an incident heat flux of 50 kW m⁻². Fig. 5 gives the HHR and THR curves of W₆₀PC and W₆₀PC/PTPA₁₀/APP₂₀. The HRR of W₆₀PC/PTPA₁₀/APP₂₀ decreases greatly compared to W₆₀PC with only one peak. After the HRR increases to the maximum, it decreases slowly in a longer duration. The peak of HRR of W₆₀PC/PTPA₁₀/APP₂₀ significantly decreases from 300 kW m⁻² of W₆₀PC to 223 kW m⁻². The total heat release (THR) of W₆₀PC decreases from 53 MJ m⁻² to 41 MJ m⁻² when PTPA₁₀/APP₂₀ was incorporated. Moreover, the addition of PTPA₁₀/APP₂₀ system enhanced the residue from 23% of W₆₀PC to 45%. Although addition of mono-component APP decreased the heat release most significantly, as we have mentioned, it deteriorated the mechanical properties seriously. In the end, using a composite flame retardant system at a chosen ratio could well balance fire resistance and mechanical strength.

Fig. 5 (a) Heat release rate (HRR) and (b) total heat release (THR) curves of W₆₀PC and (W₆₀PC)₇₀/PTPA₁₀/APP₂₀.

Table 3 Cone calorimetry tests of WPC/FR systems

Sample	pHRR^1 (kW m^−2)	pHRR^2 (kW m^−2)	THR (MJ m^−2)	t (ignition) (s)	Residue (%)
WPC	296	—	53	12	23.2
(W60PC)70/PTPA30	266	—	51	16	28.6
(W60PC)70/APP30	202	171	39	17	45.4
(W60PC)70/PTPA10/APP20	223	—	41	19	40.5

---

Flame-retardant mechanism of WPC/PTPA/APP

Thermal degradation of composite materials. Thermogravimetric analysis (TGA) is a widely used technique for rapid evaluation of the thermal stability of different materials, and it can also indicate the decomposition behavior of materials at various temperatures. Fig. 6 shows the TG and DTG curves of PP, WF, W₆₀PC, PTPA₁₀/APP₂₀ and (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ and resulted key data were listed in Table 4. Based on 5% mass loss, pure PP showed the highest thermal stability with a very high onset temperature (T_onset) of 381 °C. The TGA curve of pure PP showed a single mass-loss step with a maximum degradation rate, i.e. the first peak degradation rate (T_max1), at 445 °C. At 450 °C, the residue content of PP was 26.9 wt%, with nearly no residue content at 480 °C. Wood fibre has a low thermal stability with a T_onset of 261 °C. It degrades to carbon backbones and some volatiles when temperature is higher than 200 °C. When temperature further increases, some carbon free radicals form. These free radicals may attack hydrogen on PP chain to form new long chain free radicals, which can accelerate the reactions of the thermal degradation.²⁸ This process is responsible for the T_onset decrease of W₆₀PC. After the initial loss, W₆₀PC exhibited two decomposition steps at 367 °C (T_max1) and 482 °C (T_max2), which were related to the decomposition of WF and PP, respectively.^29–31 The presence of WF increased the thermal decomposition temperature of the PP component from 445 to 482 °C. This suggests that WF promoted char formation and thereby delayed the degradation of PP. WF has a component called lignin that acts as a char former, which could reduce the thermal degradation of WPC. Lignin is an amorphous polyphenolic plant constituent, and it represents 20–30 wt% of the wood.³² Lignin can form char during thermal degradation. Char formation is a basic aspect of FR additives because the char can protect the matrix from heat and further decomposition.

Fig. 6 (a) TGA curves of PP, WF, W₆₀PC, PTPA₁₀/APP₂₀ (PTPA : APP (w/w) = 1 : 2) and (W₆₀PC)₇₀/PTPA₁₀/APP₂₀. (b) Corresponding DTG curves. All curves were done under a nitrogen flow.

Table 4 Thermal degradation data for different WPC/FR composites in N₂

Sample	T onset ^a (°C)	T max1 ^b/Rmax2^c (°C/% min^−1)	T max2 ^b/Rmax2^c (°C/% min^−1)	T max3 ^b/Rmax3^c (°C/% min^−1)	Residue (%) at 700 °C
a T onset, the temperature at which mass loss is 5 wt%. b T max, the temperature at which the mass loss rate is maximum. c R max, the maximum mass loss rate value.
PP	381	445/−22.8	—	—	0.18
WF	261	372/−19.8	—	—	17.6
W60PC	281	367/−12.4	482/−17.9	—	12.1
PTPA10/APP20	285	297/−1.5	405/−3.9	514/−1.0	47.1
(W60PC)70/PTPA10/APP20	246	286/−5.4	498/−17.2	—	28.6

---

It can be seen that there are three degradation steps for PTPA/APP (1 : 2, w/w) at 285–310 °C, 310–450 °C, and 450–700 °C, respectively. The thermal decomposition process has been carefully studied in our previous work.²² There is a reaction between the decomposition products of PTPA and APP which not only promotes their degradation and volatilization of their products, but also promotes the formation of more thermal stable char residue at lower temperature. Therefore, the mass loss rate of IFR is small at temperature above 450 °C. Briefly, the first step of decomposition is polymerization or aromatization of carbon–carbon double bonds formed from the degradation of PTPA promoted stable char residue. The second step involves the broken of P–O–P bonds in the polyphosphate chain and forming of stable cross-linking structures. An alternative route for polyphosphate chain scission is due to the reaction between the decomposition products of PTPA and APP which involves the formation of non-flammable vaporing compounds and some stable compounds containing P–N bond.

The thermal degradation of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ undergoes a two-step decomposition. The initial decomposition temperature is lowered to 246 °C by combining PTPA₁₀/APP₂₀ with W₆₀PC. It is because PTPA₁₀/APP₂₀ accelerates the first step of decomposition, and promotes the cross-linking and char forming. The char formed can protect the matrix from heat and further decomposition. The residue of W₆₀PC with PTPA₁₀/APP₂₀ at 700 °C is 28.6 wt%, which is much higher than that of W₆₀PC. During the process of burning, (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ undergoes an expansion process. A large-volume, high-carbon protective layer is developed, which can effectively prevent the underlying matrix from the attack of the heat and oxygen, resulting improved flame retardancy of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀. It can be seen that W₆₀PC begins to degrade at 281 °C and the maximum decomposition temperature is at 482 °C with a small amount of char residue of 12.1% at 700 °C. However, the onset decomposition temperature of WPC/PTPA/APP is 246 °C which is related with the decomposition and cross-linking reactions, forming the char which can protect the matrix from heat and decomposition. When the temperature is increased, PP with IFR is further decomposed, further cross-linked and char is formed. The char residue of flame-retardant WPC is much more than that of WPC at 700 °C. The maximum weight loss rate is smaller and appears at higher temperature, which also means that the matrix is protected during burning.

FTIR analysis of the condensed products of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀. Fig. 7 shows the FTIR spectra of condensed products of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ at different temperatures in N₂. At 25 °C, P=O (1261 cm⁻¹), P–O–P (1018 cm⁻¹, 894 cm⁻¹) are attributed to APP. The typical absorptions of PTPA are evident at 1632 cm⁻¹ (C=C of benzene ring) and 3000–2800 cm⁻¹ (–CH₂–). With the increase of temperature, the absorptions at 3000–2800, 3270, and 3116 cm⁻¹ which are attributed to –CH₂–, –NH₂ and aromatic hydrocarbons, respectively, are significantly weakened. The exact association of these absorptions to specific bonds (P–N or P–O) is not easy by FTIR because of broad and overlapping peaks.³³ The reactions among these products and wood fibre lead to a cross-linked residue structure. When the temperature is above 246 °C, there is a new absorption appearing at 2135 cm⁻¹ (–CN). This peak disappeared at higher temperature, which indicates the decomposition of this new nitrile compound. Cross-linkage of the residue species is the main reason for the increase of residue observed in TGA. Furthermore, the cross-linked residue is stable, showing the rise in stability during thermo-oxidative decomposition.³⁴

Fig. 7 FT-IR spectra of condensed products of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ at different temperature in N₂.

XPS analysis of the condensed products of W₆₀PC/PTPA₁₀/APP₂₀. XPS is used to study the atomic concentration of the condensed products of flame retardant WPC. Table 5 gives the concentration of carbon (C), oxygen (O), nitrogen (N) and phosphorus (P) of the solid decomposition products of WPC/IFR under N₂ atmosphere at different temperatures. The C content decreases significantly from 80.38 wt% at 286 °C to 42.49 wt% at 485 °C, which is because of the incomplete degradation of wood fibre and PTPA, and the char formation. As temperature increases from 485 to 600 °C, the slight decrease of C means that part of the char residue decomposes, which means the char formed is stable. The content of O increases from 14.61 wt% to 46.42 wt% when the temperature is increased from 286 to 600 °C, which might be because of the formation of the aldehydes, ether, alcohol, ketone, etc. during the degradation of WPC/IFR. Compared with the change of O and C contents, the change of N content is not obvious. At first it decreases slightly because of the release of a little NH₃, then increases with the formation of some stable structures containing N. The content of P keeps increasing from 300 to 600 °C, which should be caused by the formation of some stable compounds containing P during the thermally decomposing. The results of XPS are consistent with that of FTIR test of condensed products.

Table 5 XPS data of condensed products of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ under N₂ atmosphere at different temperatures

Sample	T (°C)	C (wt%)	N (wt%)	O (wt%)	P (wt%)
(W60PC)70/PTPA10/APP20	286	80.4	2.2	14.6	2.8
	485	42.5	2.1	45.9	9.5
	600	41.2	2.8	46.4	9.7

---

Morphology of char residues. Fig. 8 shows the digital photographs of residual char after cone calorimeter test for different systems respectively. There nearly is rare residue for W₆₀PC after cone calorimeter test in Fig. 8a. Comparing with nearly no residue for pure PP (Table 3), the char residue was increased by composition with wood fibre for the amorphous polyphenolic structure. The residue increased when the flame retardants were added in as shown in Fig. 8b–d. The loosen surfaces and cracks are found on the char layer as shown in Fig. 8b and c. In contrast to those results, smoother, thicker and more compact char layer was formed when both PTPA and APP were applied as shown in Fig. 8d. This firm charred layer prevents the heat transferring, oxygen penetration and the transportation of decomposed products between melting polymer and surface, thus the HRR and THR are reduced.³⁴

Fig. 8 The digital photographs of residual char after cone calorimeter test for (a) W₆₀PC, (b) (W₆₀PC)₇₀/PTPA₃₀, (c) (W₆₀PC)₇₀/APP₃₀ and (d) (W₆₀PC)₇₀/PTPA₁₀/APP₂₀.

To further clarify the morphology of the char formed during the process of burning, the outer surface of char residues were studied by SEM. The SEM micrographs of the char residue after cone calorimeter test for different systems are shown in Fig. 9. The char residue of WPC/PTPA (Fig. 9a-1) is coralloid network structure with lots of holes. That is because PTPA is rich in carbon (charring agent) and nitrogen (blowing agent), but poor in phosphorus (acid source). So the char layer becomes fragmented when a plenty of decomposed gases easily permeates through the char layer. The char residue of (W₆₀PC)₇₀/APP₃₀ (Fig. 9b-1 and b-2) is continuous with some holes and flocculent structure in the surface because the cross-linked polyphosphoric acid formed on heating can provide a stable softened glassy coating on the surface of the polymer, but it is not firm enough. It can be permeated through by decomposed gases. The intumescent char of (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ shown in Fig. 9c-1 and c-2 is more continuous and compact, but it is still not firm enough when comparing with PTPA/APP flame-retarded PP.²² The reason can be attributed to the different decomposition processes of WF and PP in WPC and the poor interface compatibility between them.

Fig. 9 The SEM micrographs of residual char after cone calorimetry tests a(1 and 2) (W₆₀PC)₇₀/PTPA₃₀, b(1 and 2) (W₆₀PC)₇₀/APP₃₀, c(1 and 2) (W₆₀PC)₇₀/PTPA₁₀/APP₂₀.

Experimental

Materials

PP (T30S) was purchased from Lanzhou Petroleum Chemical Co., China. The coupling agent maleic anhydride grafted polypropylene (PP-g-MAH, CMG5001) was supplied by Shanghai Rizhishen New Technology Co., Ltd. Wood fibre (Pine 80 mesh) was supplied by Qingdao Fumulin Wood Plastic Composite Co., Ltd (Qiangdao, China). APP was supplied by Taifeng New-Type Flame Retardants, Co., Ltd (Shifang, China). PTPA was synthesized and well characterized in previous work,²² which was used as received. The chemical structure of PTPA is shown in Scheme 1.

Scheme 1 Chemical structrue of PTPA.

Sample preparation

All samples in this work were prepared in a Banbury mixer (SX300) at 180 °C and 50 rpm for about 7 min, and then pressed on a curing machine at 190 °C for 3 min to form sheets. The wood fibre was dried at 100 °C in an air-circulating oven for 24 h, PTPA and APP were dried in the oven at 80 °C for 8 h. Based on usual practice, a unit of wt% was used to replace weight percentage wt% in this work. First, WF was added into PP matrix with a weight percentage of, for example, 60 wt%, which resulted in a composite of W₆₀PC. This formula was fixed because of mechanical properties (see below). In addition, 5 wt% maleic anhydride grafted polyethylene (PP-g-MAH) were used to replace a minor part of PP, which is widely used for better compatibility and mechanical properties of WPC.¹⁰ Then this composite was flame retarded by PTPA or/and APP with a total weight percentage of 30% (30 wt%). All the composites were marked by its weight ratio as subscript. For example, sample (W₆₀PC)₇₀/PTPA₁₀/APP₂₀ means that 10 wt% PTPA and 20 wt% APP was combined with 70 wt% WPC, and as mentioned before, the WPC composite is with 60 wt% WF content (W₆₀PC).

Measurement

The flexural property was measured by a CMT4104 tester according to GB/T 9341-2008 at the speed of 20 mm min⁻¹ at room temperature. The size of the samples is 150 mm × 10 mm × 4 mm. The Notched Izod impact property was measured by Notched Izod impact instrument (ZBC1400-2) according to GB/T 1843–2008 at room temperature. The size of the samples is 80 mm × 10 mm × 4 mm. 5 specimens per sample were tested for mechanical properties. Scanning electronic micrograph (SEM) observed on a Philips XL-30S was used to investigate the cryogenic fracture surfaces of the IFR-WPC system.

The LOI value was tested on an HC-2C oxygen index meter (Jiangning, China) according to ASTM D2863-97 and the size of all samples is 130 mm × 6.5 mm × 3.2 mm. The UL-94 vertical burning level was tested on a CZF-2 instrument (Jiangning, China) according to ASTM D3801 and dimension of all samples is 130 mm × 13 mm × 3.2 mm. The flammability of the sample was measured with a cone calorimeter (Fire Testing Technology). The dimension of all samples is 100 mm × 100 mm × 3 mm. The samples were exposed to a radiant cone at an external heat flux of 50 kW m⁻².

Thermogravimetric analysis (TGA) was carried on a TG 209 F1 (NETZSCH, Germany) thermogravimetric analyzer from 40 to 700 °C at a heating rate of 10 °C min⁻¹ under nitrogen atmosphere with a flowing rate of 50 mL min⁻¹. 3–5 mg of the samples were taken and measured in TGA.

Fourier transform infrared spectra (FTIR) were obtained using a Nicolet FTIR 170 SX spectrometer. The samples were heated to the corresponding temperature in TG 209 F1 (NETZSCH, Germany) at a heating rate of 40 °C min⁻¹ under a nitrogen atmosphere and kept for 10 min.

X-ray photoelectron spectroscopy (XPS) of the condensed products of the IFR-WPC system at different temperatures was recorded by a XSAM80 (Kratos Co, UK), using Al Kα excitation radiation (hν – 1486.6 eV). The samples were heated to the corresponding temperature in TG 209 F1 (NETZSCH, Germany) at a heating rate of 40 °C min⁻¹ under a nitrogen atmosphere and kept for 10 min.

Scanning electronic micrograph (SEM) observed on a Philips XL-30S was used to investigate the char residues of the WPC/IFR systems, which were obtained after combustion by cone calorimeter tests. SEM graphs were recorded after gold coating surface treatment.

Conclusions

Wood fibre/polypropylene composites were prepared by melted blend method. A novel intumescent flame retardant system was incorporated to the WF/PP system to prepare the flame retarded composites. The flexural properties and Izod impact strength of the composites are well maintained once incorporation of the PTPA into WF/PP composites, comparing to the fire retardancy improved by APP. To combine two advantages, both FRs were composited with WPCs. The fire performances of flame-retardant WPCs are improved significantly with the addition of both PTPA and APP. W₆₀PC containing PTPA₁₀/APP₂₀ not only has a high LOI value of 31.5% and achieve UL-94 V-0 rating, but also has a decrease in heat release rate compared with W₆₀PC. PTPA/APP system can change the thermal degradation behavior of WPC, induce the earlier degradation of WPC and promote charring. PTPA itself shows a good ability of char formation: the residue of W₆₀PC/PTPA₃₀ reaches 21.4 wt% at 700 °C based upon TGA test. Combining PTPA with APP, the char formation can be further promoted because of the synergism of PTPA and APP. The char residue is increased to 28.6% at 700 °C.

Acknowledgements

We gratefully acknowledge financial supports from the Natural Science Foundation of China (Grant No. 50933005 and 51421061) and the Program for Changjiang Scholars and Innovative Research Teams in University of China (IRT 1026).

References

1. L. C. Hollaway, Construct. Build. Mater., 2010, 24, 2419–2445.
2. L. Soccalingame, A. Bourmaud, D. Perrin, J.-C. Bénézet and A. Bergeret, Polym. Degrad. Stab., 2015, 113, 72–85.
3. N. Ayrilmis, S. Jarusombuti, V. Fueangvivat, P. Bauchongkol and R. H. White, Fibers Polym., 2011, 12, 919–926.
4. S.-L. Du, X.-B. Lin, R.-K. Jian, C. Deng and Y.-Z. Wang, Chin. J. Polym. Sci., 2015, 33, 84–94.
5. J. Summerscales, N. P. J. Dissanayake, A. S. Virk and W. Hall, Composites, Part A, 2010, 41, 1329–1335.
6. R. R. Devi and T. K. Maji, Ind. Eng. Chem. Res., 2012, 51, 3870–3880.
7. M. Pracella, D. Chionna, I. Anguillesi, Z. Kulinski and E. Piorkowska, Compos. Sci. Technol., 2006, 66, 2218–2230.
8. R. Malkapuram, V. Kumar and Y. S. Negi, J. Reinf. Plast. Compos., 2009, 28, 1169–1189.
9. G. Bai, C. Guo and L. Li, Construct. Build. Mater., 2014, 50, 148–153.
10. R.-M. Li, C. Deng, C.-L. Deng, L.-P. Dong, H.-W. Di and Y.-Z. Wang, RSC Adv., 2015, 5, 16328–16339.
11. B. Li and J. M. He, Polym. Degrad. Stab., 2004, 83, 241–246.
12. Z. X. Zhang, J. Zhang, B.-X. Lu, Z. X. Xin, C. K. Kang and J. K. Kim, Composites, Part B, 2012, 43, 150–158.
13. J. Gong, R. Niu, J. Liu, X. Chen, X. Wen, E. Mijowska, Z. Sun and T. Tang, RSC Adv., 2014, 4, 33776–33784.
14. G. Malucelli, F. Bosco, J. Alongi, F. Carosio, A. Di Blasio, C. Mollea, F. Cuttica and A. Casale, RSC Adv., 2014, 4, 46024–46039.
15. N. Suppakarn and K. Jarukumjorn, Composites, Part B, 2009, 40, 613–618.
16. M. B. Abu Bakar, Z. A. M. Ishak, R. M. Taib, H. D. Rozman and S. M. Jani, J. Appl. Polym. Sci., 2010, 116, 2714–2722.
17. B. Schartel, U. Braun, U. Schwarz and S. Reinemann, Polymer, 2003, 44, 6241–6250.
18. M. Sain, S. H. Park, F. Suhara and S. Law, Polym. Degrad. Stab., 2004, 83, 363–367.
19. Y.-H. Guan, J.-Q. Huang, J.-C. Yang, Z.-B. Shao and Y.-Z. Wang, Ind. Eng. Chem. Res., 2015, 54, 3524–3531.
20. X. P. Hu, W. Y. Li and Y. Z. Wang, J. Appl. Polym. Sci., 2004, 94, 1556–1561.
21. M. J. Zaini, M. Y. A. Fuad, Z. Ismail, M. S. Mansor and J. Mustafah, Polym. Int., 1996, 40, 51–55.
22. Z.-Z. Xu, J.-Q. Huang, M.-J. Chen, Y. Tan and Y.-Z. Wang, Polym. Degrad. Stab., 2013, 98, 2011–2020.
23. L. Gao, Z. Wang and W. J. Guo, J. Beijing For. Univ., 2010, 32, 247.
24. L. Gao, Z. Wang and W. J. Guo, J. Beijing For. Univ., 2010, 32, 247–250.
25. K. Oksman and C. Clemons, J. Appl. Polym. Sci., 1998, 67, 1503–1513.
26. B. Schartel and T. R. Hull, Fire Mater., 2007, 31, 327–354.
27. P. Lv, Z. Z. Wang, K. L. Hu and W. C. Fan, Polym. Degrad. Stab., 2005, 90, 523–534.
28. C. Amen-Chen, H. Pakdel and C. Roy, Bioresour. Technol., 2001, 79, 277–299.
29. H. Yang, R. Yan, H. Chen, D. H. Lee and C. Zheng, Fuel, 2007, 86, 1781–1788.
30. N. P. G. Suardana, M. S. Ku and J. K. Lim, Mater. Des., 2011, 32, 1990–1999.
31. R. C. Pettersen, Adv. Chem. Ser., 1984, 57–126.
32. A. de Chirico, M. Armanini, P. Chini, G. Cioccolo, F. Provasoli and G. Audisio, Polym. Degrad. Stab., 2003, 79, 139–145.
33. M. le Bras, S. Duquesne, M. Fois, M. Grisel and F. Poutch, Polym. Degrad. Stab., 2005, 88, 80–84.
34. H. Seefeldt and U. Braun, Macromol. Mater. Eng., 2012, 297, 814–820.

---