Synergistic Effects of Flame Retardants on the Flammability and Foamability of PS Foams Prepared by Supercritical Carbon Dioxide Foaming.

Halogen-free flame-retardant polystyrene (PS) foams prepared by supercritical carbon dioxide (SC-CO2) foaming have been achieved. The flame-retardants include expandable graphite (EG) and melamine phosphate (MP), and their influence on the foamability, decomposition behavior, fire performance, and mechanical properties of PS foams were investigated. It has been shown that flame retardants can generate inert gases and catalyze the char formation from PS, and the formed thick char layer with a notable barrier property can greatly decrease the heat release of PS foams. The addition of triphenyl phosphate (TPP) or hexaphenoxycyclotriphosphazene (HPCTP), which acts as a flame-retardant plasticizer, can obviously improve the foamability and fire performance of the foams. TPP or HPCTP can generate active phosphorous species and phenoxyl radicals to enhance the gas phase flame-retardant effect; therefore, the flame-retarded PS foams (with 25 wt % MP/EG) achieve HF1 and V-0 ratings, with limiting oxygen index (LOI) values of 30.1 or 29.6%, respectively. The numerical assessment of synergistic effects of TPP and HPCTP on further enhancing flame retardancy of PS foams has been provided by the microcalorimeter (MCC) test. Further X-ray photoelectron spectroscopy (XPS) investigation on char residues of PS foams demonstrates the formation of the P-O-C and other stable structures.

Introduction

Polystyrene (PS) foam has many advantages, such as lightweight, heat insulation, shock absorption, noise reduction, and easy processing.[1−3] Therefore, PS foam is widely used in various fields of national economy, for instance, building, packaging and automobile.[4−7] However, due to the chemical composition and structural features, the PS foam is extremely flammable. Consequently, its limiting oxygen index (LOI) is very low, only 18.0%.[8] Meanwhile, the combustion process is very rapid and will release a lot of heat and toxic gases. A large number of serious building fire accidents were caused by an external insulation of PS foams, resulting in huge loss of people’s life and properties.[9] It has been reported that the direct losses from fire was accounted for 0.05–0.22% of GDP among industrialized nations.[10] Usually, the halogen flame-retardant is with high efficiency in reducing combustibility of PS foams, but it will generate corrosive and toxic gases during the burning process.[11,12] Government has gradually banned the use of halogen flame retardants; therefore, it is of very important practical significance to develop halogen-free flame-retardant PS foams with satisfactory properties. Intumescent flame retardant, comprising an acid source, a gas source, and a carbon source, is one of the important halogen-free candidates to improve the flame-retardant property of PS foams, and best known example of this is ammonium polyphosphate (APP) in combination with a carbon source.[11,13] During the combustion process, intumescent flame retardants will form a foam-like carbonized layer to reduce heat, oxygen, and fuel transfers.

Nowadays, the PS foam is mainly prepared by the physical foaming methods using hydrocarbons or chlorofluorocarbons as the blowing agents. However, due to the environment and safety concerns,[14] the replacement of conventional foaming agents is a fundamental technological innovation. Carbon dioxide (CO2) is an ideal physical blowing agent in the production of polymeric foams because it is nonflammable, low cost, nontoxic, and environmentally friendly.[15−17] In addition, the critical state (critical temperature: 31.1 °C; critical pressure: 7.38 MPa) of CO2 is mild and easy to achieve.[18] To obtain polymeric foams, the substrates absorb supercritical carbon dioxide (SC-CO2) to saturation state and then are followed by fast depressurization at a constant temperature.[19] Some research has focused on the utilization of SC-CO2 to prepare flame-retardant polymeric foams, such as poly(lactic acid) (PLA),[20] polypropylene (PP),[21] and poly(styrene-co-acrylonitrile) (SAN).[22,23] It has been reported by Urbanczyk et al.[22] that the SAN foams modified by (organo)clays/melamine polyphosphate (MPP) were prepared by SC-CO2 foaming, and a significant decrease in the heat release rate was observed owing to the synergistic effect of the flame retardants. However, there is a rare report dealing with the preparation of halogen-free flame-retardant PS foams by SC-CO2 foaming.

In our previous work,[24] we have proved the synergistic effect between melamine phosphate (MP) and expandable graphite (EG) on the flame-retardant PS. Based on this, a novel method is adopted in the present work to prepare flame-retardant PS foams with MP and EG by SC-CO2 foaming. The flame-retardant granulates are obtained by melt extrusion, and the SC-CO2 saturated granulates are used to prepare flame-retardant foams by hot-press molding. To achieve a high expansion ratio and improve flame-retardant property, we adopt triphenyl phosphate (TPP) and hexaphenoxycyclotriphosphazene (HPCTP) to act as a flame-retardant plasticizer,[25] which can decrease the viscosity of composite melt and increase the chain segments movement ability. Through this method, the commercial PS resin or waste PS foams can be used to produce flame retardant PS foams, and we provide a new simple method to realize the recycling or functionalization of waste PS foams. The impact of MP/EG on the foamability of PS composites was evaluated by a rheology test, differential scanning calorimetry (DSC), and cell morphology analysis. The fire performance, heat release rate, and thermal oxidation stability of the foams were systematically revealed by multiple test methods. The formation of C–O–P in the char residue was confirmed through an X-ray photoelectron spectroscopy (XPS) test. According to the results, the flame retardant mechanism combining of gas phase and condensed phase is proposed.

Results and Discussion

Effect of Flame Retardants on Foaming Performance

The foaming performance of PS composites is related to the viscosity and glass transition temperature (Tg). Figure shows the rheology and DSC curves of PS composites. As the increasing of EG/MP content, which can obstruct the movement of PS molecular chains, the viscosity of PS composites shows a dramatic increase. The neat PS and flame-retardant PS composites are with obvious shear thinning behavior. With the increase of shear rate, the viscosity decreases gradually due to the orientation of molecular chains and graphite flakes during the test.[26] Adding MP/EG leads to the increase of viscosity, but has little influence on the glass transition behavior, which happens in the range of 100–120 °C. TPP and HPCTP can act as the flame-retardant plasticizer, which can weaken the forces between polymer chain segments and make the segments movement easier. When adding 3 wt % TPP or HPCTP, the Tg of 75PS composite decrease from 105 °C to 96 or 100 °C, respectively. In addition, the viscosity decreases clearly, especially for TPP3 composite, the viscosity of which is much lower than the neat PS. For amorphous polymers, the cell growth occurs above Tg, and the melt is subjected to elongational deformation during the cell growth step, which is impacted by the melt viscosity. The melt viscosity will have an influence on the cell size, and the low viscosity will promote cell growth, which will lead to the formation of larger cells.[27,28] As for the cell stabilization, the morphology must be stabilized and the cell growth has to cease, and the main factor is the increase of melt viscosity that is caused by a reduction of the polymer’s temperature. So, the Tg is a key parameter for the cell stabilization because the polymer will behave like a solid and the cell growth will stop by hardening of cell walls when the temperature is below Tg. The decrease of the viscosity and Tg will influence the cell growth and cell stabilization steps of the foaming process and will result in the decline of the foaming temperature.

Figure 1

(a) Rheology and (b) DSC curves of different flame-retardant PS composites.

The SEM images, expansion ratio, and average cell size of flame-retardant PS foams are presented in Figures and 3. The 100PS foam with a uniform cell size is obtained by SC-CO2 foaming, and its average cell size, cell density, foam density, and expansion ratio are 78.95 μm, 6.62 × 107 cells/cm3, 26.91 kg/m3, and 38.79 times, respectively. MP/EG can reduce the absorption of CO2 and will lead to the heterogeneous nucleation, so it will influence the foaming performance. The microvoids between EG/MP and PS-melt can act as nucleation points, and gas will diffuse to the microvoids,[29] so the cell will grow around EG/MP. The sheet filler EG leads to the collapse or collision of cells in the PS matrix because the sharp edges and harsh surface of EG will cause a cell rupture and hinder the cell growth.[30] Therefore, the sheet filler EG contributes to the merging of cells, which results in forming of the large size cells. The uneven dispersion and agglomeration of the flame retardants happen with the increase of MP/EG content, which will further deteriorate the foamability. The 25 wt % MP/EG addition results in the increase of cell density and foam density and the decrease of expansion ratio and average cell size.

Figure 2

SEM images of (a) 100PS, (b) 75PS, (c) TPP3, (d) HPCTP3 foams, and (e) MP/EG.

Figure 3

Average cell size and expansion ratio of different flame-retardant PS foams.

Although, it has been reported that SC-CO2 has a plasticizing effect to decrease the polymer melt viscosity,[31] the effect is still not enough under a high flame-retardant amount. Adding TPP or HPCTP decreases the viscosity and Tg, which is beneficial to the cell growth and will cause the merging of cells. This is because a relative low melt viscosity and the enhancement of viscous behaviors of TPP3 and HPCTP3 (Figure S1) are in favor of the diffusion of gas through polymer melt, which will reduce the resistance of cell growth.[32] In consequence, the cell density decreases slightly, but the average cell size and expansion ratio increase clearly. Compared to that of 75PS foam, the foam density decreases from 61.89 kg/m3 to 52.86 or 44.73 kg/m3, and the average cell size increases from 39.73 μm to 53.31 or 46.55 μm, by the addition of 3 wt % TPP or HPCTP. As a result, TPP and HPCTP can act as a plasticizer to improve the foamability.

Thermal Oxidation Stability of Flame-Retardant PS Foams

TGA as a valid approach to study the degradation behaviors of materials was adopted to evaluate the thermal oxidation stability of PS foams. Figure presents the TGA and derivative thermos-gravimetric (DTG) curves of flame-retardant PS foams and the detailed results are recorded in Table . The neat PS foam decomposes rapidly at 300–400 °C according to the radical chain mechanism, and its initial decomposition temperature (temperature at 5% weight loss, T5 wt %) and maximum decomposition rate are 315 °C and 23.14 wt %/min. The thermal oxidation of PS generates styrene monomer, oligomers (dimer, trimer, and tetramer), and products of oxidation (benzaldehyde, acetophenone, phenol, styrene oxide, etc.),[33−35] which can fuel the combustion process.

Figure 4

(a) TGA and (b) DTG curves of different flame-retardant PS foams at a heating rate of 10 °C/min under an air atmosphere.

Table 1

Data of TGA and DTG Thermograms of Different Flame Retardant PS Foams at a Heating Rate of 10 °C/min under an Air Atmosphere

sample	T5% (°C)	residue at 700 °C (wt %)	Tmax (°C)	mass loss rate at Tmax (wt %/min)
100PS	315	0.27	399	23.14
90PS	326	6.04	383	20.20
80PS	321	11.51	364	17.42
75PS	315	14.80	360	15.99
70PS	316	18.92	363	14.85
TPP3	296	15.88	362	14.65
HPCTP3	317	18.55	364	14.70

The addition of MP/EG can increase the initial decomposition temperature of PS foams, which illustrate that the thermal oxidation stability of foams is enhanced. However, the increments gradually decrease with the increase of MP/EG amount, this is due to the lower initial decomposition temperature of EG (190 °C, Figure S2) and MP (285 °C). The redox reaction between H2SO4 and the graphite of EG, which releases abundant blowing gases (CO2, SO2, and H2O), leads to the formation of the expanded graphite with a vermicular structure and large volume.[36,37] Meanwhile, high viscosity products (such as polyphosphate compounds) are generated by the decomposition of MP (Figure S2), which are helpful to strengthen the bond between the expanded graphite to form a thick barrier. At the early stage, the barrier can suppress the escape of the volatiles, which leads to the weight loss delay and initial decomposition temperature increase. Adding MP/EG enhances the char yield of the foams at 700 °C, and the maximum weight loss rate of the foams decrease as well. The escaping time of the volatile products are prolonged by a char layer, which will cause the increase of free radicals life, resulting in accelerating the thermal oxidation of PS. Compared with 100PS foam, the corresponding temperature of maximum mass loss rate (Tmax) of flame-retardant foams are shifted to the lower temperature. As for 75PS foam, the char residue at 700 °C and maximum weight loss rate are 14.80 wt % and 14.85 wt %/min, respectively.

The char yield at 700 °C of TPP3 and HPCTP3 foams are 15.88 and 18.55 wt %. The addition of a small amount of HPCTP results in the char residue increase, which will promote the formation of the dense barrier to strengthen the condensed phase flame-retardant effect. Compared with that of 100PS foam, the maximum mass loss rate of TPP3 and HPCTP3 foams are 14.65 and 14.70 wt %/min, which have a decline of 36.69 and 36.47%, respectively. As a result, TPP and HPCTP can act as the synergistic flame retardants to enhance the flame retardancy of PS foams.

To further estimate whether there is an interaction between PS and flame retardants, the experimental and calculated TGA curves were plotted in Figure . The initial decomposition temperature of 75PS, TPP3, and HPCTP foams are shifted from 41, 48, and 76 °C to the higher temperature than the calculated one, respectively. Meanwhile, the experimental char yield is significantly higher than the calculated one. The char yield at 700 °C of 75PS, TPP3, and HPCTP foams is 14.80, 15.88, and 18.55 wt %, respectively. It is worth noting that 13.43 wt % (53.72 wt % × 25/100), 12.75 wt % (46.90 wt % × 28/103), and 13.80 wt % (50.77 wt % × 28/103) residues are from the flame retardants, and the rest (1.37, 3.13, and 4.75 wt %) are from PS. Therefore, the true residues from PS are 1.83, 4.30, and 6.52 wt %. Compared with that of 100PS foam (0.27 wt %), the true residues from PS of 75PS, TPP3, and HPCTP foams have about 6.78, 15.93, and 24.15 times increase. The results indicate that adding flame retardants can promote PS to participate in carbonization reaction and improve the thermal oxidation stability of PS foams.

Figure 5

Experimental and calculated TGA curves of 75PS, TPP3, and HPCTP3 at a heating rate of 10 °C/min under an air atmosphere.

MCC and Cone Calorimetric Analysis

Microcalorimeter (MCC) was utilized to assess the fire performance improvement of different flame-retardant PS foams, and the results are presented in (Figure ) and summarized in Table . The 100PS foam shows a very rapid heat release, with a high peak heat release rate (Peak-HRR) and total heat release (THR) values of 929 W/g and 37.8 kJ/g, respectively. As for flame-retardant foams, the Peak-HRR and THR values decrease clearly with the increasing of flame-retardants amount. Adding a small amount of TPP or HPCTP can further reduce Peak-HRR and THR values. Compared with 100PS foam, the Peak-HRR and THR values of HPCTP3 foams decrease to 496 W/g and 27.9 kJ/g , which have a reduction by 46.6 and 26.2%, respectively. Heat release capacity (HRC) is an important assessment parameter to the fire safety of polymer foams,[38] the HRC value of HPCTP3 foam reduces from 962 to 511 J/g·k, with a decrease by 46.9%.

Figure 6

HRR curves of different flame-retardant PS foams from the MCC test.

Table 2

Data of Different Flame-Retardant PS Foams from the MCC Test

sample	Peak-HRR (W/g)	Tmax (°C)	THR (kJ/g)	HRC (J/g·k)	EFFa	SEa
100PS	929	422	37.8	962
90PS	726	435	33.4	748	20.28
80PS	647	430	31.6	664	14.10
75PS	561	419	30.4	591	14.73
70PS	520	423	27.5	542	13.64
TPP3	513	419	28.3	534	17.15	1.16
HPCTP3	496	418	27.9	511	17.82	1.21

Flame retardant effectivity (EFF) and synergistic effectivity (SE) were calculated by MCC data as follows: EFF = (Peak-HRRpolymer – Peak-HRRcomposite)/Flame-retardant content; SE = EFFFlame-retardant + synergists/EFFFlame-retardant.

To provide a numerical assessment of synergistic effects of TPP and HPCTP, the flame retardant effectivity (EFF) and synergistic effectivity (SE), as shown in Table , of flame-retarded PS foams were calculated. As defined by Lewin,[39−41] the EFF values and SE values of PS foams are calculated by the reduction of Peak-HRR, which are obtained from MCC results. With the increase of the flame-retardants amount, the EFF value decreases gradually, which illustrates that the contribution of a unit mass of flame retardants to the flame retardant efficiency decreases gently. Adding 3 wt % TPP or HPCTP, the EFF values of foam increase (17.15 and 17.82, respectively). The SE values of TPP3 and HPCTP3 foams are 1.16 and 1.21, respectively, indicating that TPP and HPCTP can act as the synergists to further enhance the flame retardancy of PS foams.

The cone calorimeter test, which is able to effectively assess burning behaviors of polymeric materials in actual fire accidents,[42,43] was adopted to evaluate the fire safety of PS foams as well. Figure presents the HRR, THR, fire growth rate index (FIGRA), weight loss, and smoke production rate (SPR) curves of 100PS and flame-retarded PS foams, and the detailed data are shown in Table . The ignition times of flame-retarded PS foams are shorter than that of 100PS foam. During the heating process, the rapid melting of 100PS foam results in the increase of the distance between the heated surface and heater, so that the heating efficiency is reduced. Meanwhile, adding flame retardants increases both of the viscosity and thermal conductivity of flame-retarded foams, which will cause the fast rise of surface temperature to decomposition temperature.[44] As a result, flame-retarded PS foams are ignited in a shorter time. After ignition, 100PS foam releases abundant heat within 100 s. For 75PS and TPP3 foams, the slightly higher THR values is due to their higher foam density. All flame-retarded PS foams show typical HRR curves of the char-forming substance and are with lower HRR values. It is noteworthy that the obvious reduction of Average-HRR and Peak-HRR values are considered as the most significant element in reducing fire loss. The Peak-HRR, Average-HRR, and THR values of HPCTP3 foam are 169 kW/m2, 74 kW/m2, and 18.63 MJ/m2, which decrease by 56.92, 43.51, and 23.24%, respectively, compared to those of 100PS foam. Adding a small amount of HPCTP, which presents a good synergistic effect both on enhancing foaming and fire retardant properties, can further decrease the heat release of the foams. The prolonged burning time and reduced HRR values are induced by the thick char layer, which can suppress the escape of the combustible volatiles. The FIGRA, which is calculated by the ratio of HRR on time (HRRi/ti), is very useful to evaluate the contribution of a material to fire.[45] The FIGRA curves of flame-retarded PS foams are lower than 100PS foam, which shows that adding flame retardants can decrease the dedication of PS foams to a fire. Compared to that of 100PS foam, the maximum FIGRA value of HPCTP3 foam decreases from 5.63 to 4.12 kW/m2/s. Meanwhile, the decrease of the MAHRE (maximum average rate of heat emission) value also indicates the fire performance improvement of the flame-retarded foams.

Figure 7

Cone calorimetric curves of 100PS and flame-retarded PS foams. (a) HRR, (b) FIGRA, (c) THR, (d) mass retention, (e) SPR, (f) CO production rate, and (g) CO2 production rate.

Table 3

Data of 100PS and HPCTP3 Foams from the Cone Calorimetry

sample	100PS	75PS	TPP3	HPCTP3
TTI (s)	39	18	17	14
Peak-HRR (kW/m2)	392	193	178	169
Average-HRR (kW/m2)	131	77	77	74
MAHRE (kW/m2)	165	125	120	115
THR (MJ/m2)	24.27	27.45	23.85	18.63
Peak-MLR (g/s)	0.18	0.13	0.08	0.07
Average-MLR (g/s)	0.09	0.03	0.03	0.02
Peak-SPR (m2/s)	0.16	0.09	0.09	0.08
TSP (m2/kg)	9.53	9.07	8.43	6.61
SEA (m2/kg)	1247	1013	1073	1118
CO yield (kg/kg)	0.07	0.11	0.15	0.13
CO2 yield (kg/kg)	2.29	2.92	3.34	3.08

In Figure d, the weight loss rate of flame-retarded PS foams are much lower than 100PS foam, and flame-retarded PS foams are of higher mass retention. The peak mass loss rate (Peak-MLR) and Average-MLR of HPCTP3 foam reduce from 0.18 and 0.09 g/s to 0.07 and 0.02 g/s, with a reduction by 61.11 and 77.78%, respectively. This result indicates that the formed thick char layer can cut off the fuel supply to the combustion process, which will cause the flame to self-extinguish. During the combustion process, PS foam will generate a large volume of toxic smoke, resulting in the death by suffocation or inhalation.[46] The Peak-SPR and total smoke production (TSP) of HPCTP3 foam are 0.08 m2/s and 6.61 m2/kg, which decrease by 50.00 and 30.64% in comparison to 100PS foam, respectively. The CO and CO2 production rates of flame-retarded PS foams are lower than that of 100PS foam. The slowly release of toxic smoke creates the conditions for the safety evacuation of personnels. It is worth noting that adding flame retardants increase the CO yield (defined as the mass of CO formed from the unit mass of burning materials). It is known that the CO production results from incomplete combustion. The active phosphorous species generated by the decomposition of the flame retardants can act as radical scavengers and cause the incomplete combustion, thus correspondingly more CO is produced under forced flaming conditions. In other words, the flame retardants are with a flame retardant effect in the gas phase.[47]

Figure presents the residues of 100PS and HPCTP3 foams after cone calorimetric test. The 100PS foam has almost no residue left after the test, but HPCTP3 foam has about 21 wt % char residue retention. In the heating process, EG forms the worm-like expanded graphite, and MP decomposes to form high-viscosity phosphate-containing compounds, which can strengthen the bond between the expanded graphite. The formed thick char residue possesses an obvious barrier effect, which can suppress the exchange of heat, air, and volatile products. The flame retardancy of PS foams is greatly improved due to the condensed phase flame retardant effect.

Figure 8

Residues of (a) 100PS and (b) HPCTP3 foams after the cone calorimetric test.

Flame Retardant Mechanism of PS Foams

As shown in Figure , the chemical composition of the char residue for 75PS, TPP3, and HPCTP3 foams were characterized by XPS analysis. The results show that the C content is very high indicating that the expanded graphite is enriched on the surface of the residues and can play a role as barrier. Take HPCTP3 foam as an example to illustrate the detailed information of C1s, O1s, P2p and N1s spectra. The peaks at 284.650, 286. 087, and 288. 650 eV are assigned to C–H and C–C in aliphatic or aromatic species, C–O–P in phosphate compounds, and C=O in carbonyl compounds, respectively. The peaks of O1s spectra with a binding energy around 531.819 and 533.104 eV are the contributions of =O in phosphate or carbonyl groups and −O– in C–O–P or C–O–C groups.[48−50] The peak at 134.306 eV corresponds to P–O–C or PO3– groups in phosphate species.[51] For N1s spectra, the peak at 399.740 is attributed to nitrogen in pyrrole- or pyridine-type structures.[48,49] The binding energy around 401.530 eV is corresponding to the quaternary nitrogen and the formation of oxidized nitrogen compounds.[52] The formation of the stable structures (such as C–O–P) is beneficial to the retention of more char residues at a high temperature.

Figure 9

XPS spectra of the heat-treated (a) 75PS, (b) TPP3, and (c) HPCTP3 systems.

Figure shows the flame retardant mechanism of PS foams. Under the heating condition, EG will come into forming a worm-like expanded graphite by the redox reaction (>190 °C) between H2SO4 and graphite. MP will undergo a series of complex reactions (including condensation, chain scission, and crosslinking) during the decomposition process, and will decompose into various intermediates, such as melamine pyrophosphate, melamine polyphosphate, melam ultraphosphate, water, ammonia, and melamine.[53] The high-viscosity phosphate-containing products cover the surface of worm-like expanded graphite, which can prevent the oxidation of expanded graphite. Therefore, the thick char layer with excellent barrier effect is formed. The inert gases (SO2, CO2, H2O, NH3, etc.) generated by EG and MP can dilute the concentration of the fuels and oxygen. Furthermore, TPP and HPCTP can act as the synergists to improve the flame retardancy. The decomposition of TPP and HPCTP will generate active phosphorous species (such as PO·, P·, and PO2·) and phenoxyl radicals,[54,55] which can quench down the flammable free radicals (such as H· and OH·) and inhibit the combustion process. Due to both of gas phase and condensed phase flame retardant effect, the flame retardancy of PS foams is improved obviously.

Figure 10

Flame retardant mechanism of PS foam.

Flammability of Flame-Retardant PS Foams

The LOI test, vertical burning test, and horizontal burning test were adopted to evaluate the flammability of PS foams, as presented in Table . The 100PS foam quickly burns up to the holding clamp and produces an abundant black smoke, with a low LOI value of 18.4%. Increasing the MP/EG content, the flame retardancy of PS foams improves gradually by the formation of the thick char layer. Adding 25 wt % MP/EG, the LOI value of 75PS foam increases to 26.3% but can only reach HBF and V-1 ratings. Further increase of the MP/EG content will enhance the flame retardancy of PS foams but will deteriorate the foaming performance. TPP and HPCTP, which can act as the flame-retardant plasticizer, were added to enhance both of the foamability and fire performance of PS foams. The LOI values of TPP3 and HPCTP3 foams are 30.1 and 29.6%, respectively, and both can achieve HF1 and V-0 ratings.

Table 4

LOI and UL-94 Testing Results of Different Flame-Retardant PS Foams

sample	foam density (kg/m3)	LOI	horizontal burning	vertical burning
100PS	26.91	18.4	NR	NR
90PS	30.81	20.8	NR	NR
80PS	40.99	24.1	NR	NR
75PS	61.89	26.3	HBF	V-1
70PS	72.76	27.9	HF1	V-0
TPP1	57.53	26.4	HF1	V-1
TPP2	59.00	28.4	HF1	V-0
TPP3	52.86	30.1	HF1	V-0
TPP4	70.98	29.7	HF1	V-0
HPCTP1	54.40	28.1	HF1	V-1
HPCTP2	47.24	28.8	HF1	V-0
HPCTP3	44.73	29.6	HF1	V-0
HPCTP4	43.21	29.9	HF1	V-0

Thermal Conductivity and Mechanical Properties of PS Foams

Figure presents thermal conductivity of different flame-retardant PS foams. The thermal conductivity of flame-retarded PS foams increase with the increase of MP/EG amount. There are two main reasons for this phenomenon: first, the foam density increases significantly with the MP/EG content increasing; second, part of the EG particles, which are of higher thermal conductivity than PS, form the heat conduction path. The thermal conductivity of 75PS foam is 0.0435 W/(m·k). After the introduction of TPP or HPCTP into PS composites, the improvement of the foaming performance leads to the decrease of the foam density. Therefore, the thermal conductivity of TPP3 and HPCTP3 foams decreases to 0.0372 and 0.0363 W/(m·k).

Figure 11

Thermal conductivity of different flame-retardant PS foams.

The bending strength and compressive strength of 100PS, 75PS, TPP3, and HPCTP3 foams were investigated, as shown in Figure . The bending strength and compressive strength of 75PS foam are higher than that of 100PS foam because of the increase of the foam density and the decrease of the cell size. However, adding TPP or HPCTP causes the reduction in the foam density of TPP3 and HPCTP3 foams. The continuity and stability of the cell structure deteriorate as well, due to the formation of the large cells induced by EG. Compared with 75PS foam, the bending strength of TPP3 and HPCTP3 foams reduce to 0.660 and 0.560 MPa and the compressive strength of those decrease to 0.180 and 0.178 MPa, respectively.

Figure 12

Bending strength and compressive strength of 100PS, 75PS, TPP3, and HPCTP3.

Conclusions

Synergistic effects of flame retardants on the foamability and fire performance of PS foams prepared by supercritical carbon dioxide foaming are observed. The results show that the MP/EG content will increase the melt viscosity of PS composites and deteriorate the foaming performance. The addition of TPP or HPCTP decreases the melt viscosity and glass transition temperature of PS composites and improves the foaming performance. Meanwhile, adding another 3 wt % TPP or HPCTP further enhances the flame retardant performance of 75PS foam (with 25 wt % MP/EG addition), achieving HF1 and V-0 ratings, with LOI values of 30.1 or 29.6%, respectively. The formation of the stable structures (such as C–O–P) are beneficial to retain more char residues at a high temperature, which can play a role in suppressing the exchange of heat, air, and volatile products. The inert gases (H2O, NH3, etc.), active phosphorous species (such as PO·, P·, and PO2·), and phenoxyl radicals generated by the decomposition of the flame retardants are helpful to suppress the combustion process. Due to both of gas phase and condensed phase flame-retardant effect, the heat release rate, total heat release, and smoke production rate of the foams decrease significantly, which create the conditions for the safety evacuation of personnels. The thermal conductivity, bending strength, and compressive strength of HPCTP3 foam are 0.0363 W/(m·k), 0.560 MPa, and 0.178 MPa, respectively.

Experimental Section

Materials

PS (GPPS-500N) was purchased from CNPC Dushanzi Petrochemical Co. (Xinjiang, China). MP and EG (EG-E300, 80 mesh) were supplied by Sichuan Institute of Fine Chemical Industry Research and Design (Sichuan, China) and Qingdao Yanhai Carbon Materials CO., Ltd. (Shandong, China). TPP and HPCTP were provided by Chengdu Kelong Chemical Co., Ltd. (Sichuan, China) and Otsuka Chemical Co., Ltd. (Japan) (Scheme ).

Scheme 1

Chemical Structure of (a) TPP and (b) HPCTP

Preparation of Flame Retardant PS Foams

Figure reveals the preparation process of flame retardant PS foams, and the formulation of different foams are summarized in Table . PS and flame retardants were premixed and then extruded by a TSSJ-25/33 corotating twin-screw extruder (Φ = 25 mm, L/D = 33, Chenguang Research Institute of Chemical Industry, China) with an extrusion temperature of 180 °C and a rotation speed of 150 rpm. The dried extruded pellets were saturated with SC-CO2 in a stainless steel autoclave (Beijing Century Senlang Experimental Apparatus Co., Ltd., China) at a temperature of 45 °C and a pressure of 12 MPa for 4 h. The saturated PS granulates were used to prepare foams by hot-press molding (YJ63 × 2 Plate Vulcanizing Press Machine, Chengdu Rich Chuanghong Technology Co., Ltd., China) at corresponding foaming temperature under 20 MPa. After 15 min, the flame-retardant foams were obtained by rapid pressure drop.

Figure 13

Preparation process of flame-retardant PS foams by SC-CO2 foaming.

Table 5

Formulation of Different Flame-Retardant PS Foams

sample	PS content (wt %)	MP/EG(1:2) content (wt %)	synergista content (wt %)	foaming temperature (°C)	foam density (kg/m3)
100PS	100			130	26.91
90PS	90	10		130	30.81
80PS	80	20		130	40.99
75PS	75	25		126	61.89
70PS	70	30		126	72.76
TPP1	75	25	1	126	57.53
TPP2	75	25	2	123	59.00
TPP3	75	25	3	123	52.86
TPP4	75	25	4	120	70.98
HPCTP1	75	25	1	126	54.40
HPCTP2	75	25	2	126	47.24
HPCTP3	75	25	3	126	44.73
HPCTP4	75	25	4	126	43.21

synergist: TPP or HPCTP.

Characterization

The rheology test was carried out on a 25 mm parallel-plate rotational rheometer (AR2000ex, TA instruments, USA) using a 1.8 mm gap. The dynamic frequency sweeps of different flame retardant PS composites were carried out in a frequency range from 100 to 0.01 Hz using 5% strain at 180 °C.

The DSC test was performed on a Q20 DSC analyzer (TA instruments, USA) under a nitrogen environment in an aluminum pan. The flow rate is 50 mL/min, and the heating rate is 10 °C /min. The glass transition temperature was calculated from 40 to 180 °C.

The morphology of PS foams was observed by scanning electron microscopy (SEM) (FEI Instrument Co., Ltd., Netherlands) after gold sputter-coating.

The horizontal and vertical burning tests were performed on an HK-HVR horizontal and vertical flame tester (Zhuhai Huake Testing Equipment Co., Ltd., China) according to UL-94, ASTM D4986, and ASTM D3801, with sample dimensions of 150 × 50 × 13 mm3 and 127 × 12.7 × 13 mm3.

The LOI value was measured by a JF-3 oxygen index instrument (Nanjing Jiangning Analysis Instrument Company, China) with 130 × 10 × 10 mm3 samples according to ASTM D2863.

The TGA test was carried on a TA-Q50 instrument (TA instruments, USA) from room temperature to 700 °C under an air atmosphere. The heating rate is 10 °C/min, and the gas flow rate is 60 mL/min.

The MCC test was carried out at a FAA-PCFC microcalorimeter (Fire Testing Technology Limited, UK). The sample (about 4 mg) was heated to 750 °C at a heating rate of 1 °C/s under a nitrogen atmosphere. The combustor was set to 900 °C with an oxygen/nitrogen flow rate of 20/80 mL/mL.

The combustion behaviors were investigated by cone calorimeter (Fire Testing Technology Limited, UK). The samples with 100 × 100 × 15 mm3 were exposed to a radiant heat flux of 35 kW/m2 according to ISO 5660.

XPS spectra of the char residue obtained at 700 °C were recorded by Shimadzu/Kratos AXIS Ultra DLD Multifunctional X-ray Photoelectron Spectrometer (Manchester, UK).

Bending strength and compressive strength of different PS foams were tested by INSTRON 5567 tensile and compression tester (Instron Corporation, USA). The bending strength was completed according to ISO 1209-1:2004 with a testing speed of 10 mm/min with a sample size of 120 × 25 × 20 mm3. For the compressive strength test, samples were measured at a crosshead speed of 2 mm/min with dimensions of 50 × 50 × 30 mm3 according to ISO 844:2007. At least five samples were used to calculate the average values. Thermal conductivity of the foams was obtained by a hot disk thermal analyzer (Hot Disk AB, Uppsala, Sweden).