The polymer matrix used in this study was the mixture of the ready-to-eat pPP packages obtained from Seven-Eleven Mini Mart at Srinakharinwirot University, Ongkharak branch. It was a mixture of 50 wt% thermoformed red and 50 wt% injection molded white PP trays as shown in Fig. 1a. The collected pPP packages were then cleaned using washing liquid and tap water and let it dried naturally indoors. The dried samples were then ground into small pieces suitable for the extrusion process (Fig. 1b). HGM (S60) was kindly donated from 3M Co., Ltd. (Thailand), and employed as an additive for thermal conductivity and density reduction purposes. Its average particle size, thermal conductivity and true density were 30 μm, 0.208 W/(m·K) and 0.604 g/cm³, respectively. APP was added as a fire-retardant which mainly liberated NH₃ and yielded metaphosphoric char to prevent the combustion reaction of oxygen. A virgin PP (vPP) homopolymer (POLIMAXX 1100NK, IRPC Plc, Thailand) was used as a standard material for testing and characterizations.

The production process of pPP composites started with grinding pPP trays by using a blade cutting mill. The ground pPP, with a top-cut size of around 1 cm (Fig. 1b), was then dry-mixed with HGM particles and APP following the formulation shown in Table 1. Each formulation was then melt-blended using a co-rotating twin-screw extruder (Charoen Tut, CTE-D16L32, Thailand) and pelletized using a blade pelletizer (Fig. 1c). The extrusion process was operated at barrel temperatures of 175/185/195/210/195°C with a screw speed of 80–100 rpm. The extrudates were then processed using an injection molding machine (Charoen Tut, INJ101T, Thailand) for some mechanical thermal and mechanical tests as illustrated in Fig. 2. The barrel temperature of the injection process was set from 195°C, 190°C and 200°C for the back, middle and front ends, respectively. The mold temperature and the injection pressure were 45°C and 100 MPa, respectively. The total injection molding time was 60 s. Rectangular and dumbbell-shaped samples were prepared for mechanical, thermal, and morphological tests.

For the thermal conductivity test, each formulation was fabricated using a compression molding machine (Charoen Tut, PR2D-W300L350-PM-WCL-HMI, Thailand). One hundred and twenty grams of the sheet were put into a 20 × 20 × 0.3 cm³ mold and articulated with steel sheets and pre-heated for 5 min. Compression molding was processed at 190°C. A pressure of 30 bars was applied for 2 min, followed by a pressure of 100 bars for 5 min. The cooling cycle was 3 min. The final compression molded products are clearly shown in Fig. 3.

The tensile properties of vPP and pPP composites were measured in accordance with the ASTM D638 standard, using a Universal Testing Machine (Instron, 5966, USA). A gauge length of 2.54 cm and a speed of 50 mm/min were set for this test. The flexural strength was evaluated in accordance with ASTM D790 under a three-point bending mode using a Universal Testing Machine (INSTRON, 5966, USA). A support span and crosshead speed were 48 mm and 1.2 mm/min, respectively. Both flexural and tensile tests were carried out at room temperature. Five replicated samples were averaged for the tensile and flexural tests. The density of the vPP, pPP and its composites was measured by using a densimeter following the Archimedes principle.

The thermal transition behavior of vPP, pPP and pPP composites was observed using a Differential Scanning Calorimetry (DSC) technique (Perkin Elmer, DSC6000) under nitrogen condition. The test was conducted with a 10 mg sample which was firstly heated from room temperature to 220°C at a ramp rate of 10°C/min. They were then isothermally kept for 5 min and cooled down to room temperature at the identical ramp rate and held again for 5 min. Finally, they were re-heated to 220°C at 10°C/min. Data of the thermal profiles, glass transition temperature (T_g), melting temperature (T_m) and crystallization temperature (T_c) were monitored and collected. The degree of crystallinity (X_c) value was further calculated from the area under the melting peak of the second heating (Eq. (1)).

(1) XC(%)=100ΔHmΔHf×XPP

where ΔH_m is the enthalpy of melting obtained from the area under the melting peak. X_PP is the weight fraction of PP. ∆H_f  is the specific heat of melting of completely crystalline PP, which is equal to 207 J/g [27].

Thermogravimetric analysis (TGA) of each formulation was measured using a Perkin Elmer Pyris 6 TGA under air atmosphere. The test was operated at a temperature range of 30–800°C at a heating rate of 10°C/min. Data of the thermal decomposition curve was reported together with the residual weight, 5% weight loss temperature (T_D5) and 50% weight loss temperature (T_D50).

The fracture morphology of vPP, pPP and pPP composites was observed using a field emission scanning electron microscope (FE-SEM) (JEOL, JSM7800F, Japan) equipped with PCSEM software. Each specimen was fixed to an aluminum stub and coated with a conductive thin layer of gold. Each sample was tested at an acceleration voltage of 1 kV.

A vertical burning test (VB) was performed using a handmade burning chamber with the sample size of 127 × 12.7 × 3.2 mm³ (Fig. 4). The test was operated in accordance with the UL94 standard. Data of burning time at the first (t₁) and second ignition (t₂) were recorded together with the clamp fire propagation, fire dipping and cotton ignition observation.

The conductive heat transfer coefficient (k) value was estimated using a heat flow meter (NETZSCH HFM 436) which was tested using a method described in ASTMC518. Further calculation of the k value, in W/(m·K), was done by Eqs. (2) to (4):

(2) q=k(T1−T2)Δx

(3) R=(T1−T2)q

(4) k=ΔxR

where q is the heat flux, in W/m², T₁ and T₂ are the temperature difference, Δx is the thickness of the test sample (m). R is the thermal resistance in m²·K/W.

Table 2 shows the tensile and flexural properties of vPP, pPP and pPP composites. With 5 wt% APP (A5), the tensile strength decreased slightly compared with that of pPP. It continued to decrease with the further addition of APP and HGM. The same behavior was also found in flexural strength. This value reduced from 49.9 MPa, in pPP, to 30.1 MPa when HGM and APP were simultaneously added at 5 and 20 wt% respectively. These are because of the poor interfacial bonding between pPP, HGM and APP [28,29]. The pPP was less tough, which was reflected in the decrease in tensile elongation at break with HGM and APP loading. This is mainly due to the restriction of pPP chain entanglement from HGM and APP additions. Another reason was the PP chain scission during reprocessing which lowered its chain length and average molecular weight [30]. The tensile strength and elongation at break of poly(butylene succinate) (PBS) composites were found to be reduced with the HGM contents due to the poor interfacial adhesion and the fracture of HGM particles [29]. With the APP, both tensile strength and elongation at break were remarkably reduced due to the poor interfacial interaction. The authors suggested that the addition of silane coupling agent was found to greatly improve the dispersibility and interfacial compatibility between the APP and PP matrix [31].

Based on the mechanical test data, there are four possible mechanisms for pPP, HGM and APP debonding. The applied load was first applied to the pPP matrix, which resulted in matrix deformation. This load was then transferred to the rigid APP and HGM particles. These particles, however, were much more rigid than pPP. Thus, the deformation of APP and HGM particles was much lower than pPP. At this point, the pPP phase began to separate from the APP and HGM particles, resulting in debonding at the interfacial points [32].

Microscopic observation of vPP, pPP and pPP composites was clearly shown in Fig. 5. HGM showed a spherical shape while APP illustrated an asymmetry shape. In Figs. 5a and 5b, vPP showed a rougher surface compared with pPP due to its virginity. With 5 wt% APP, the pulled-out particles and the interfacial boundaries were found. This was believed to restrict the filler-matrix stress transfer, which lowered the tensile and flexural strength [3]. Some pulled-out of HGM and APP particles were found in Figs. 5d–5g. This indicated the poor interfacial bonding between pPP, APP and HGM.

Fig. 6 illustrates the DSC thermograms of vPP, pPP and pPP composites obtained from nonisothermal DSC scanning. Their corresponding T_c, T_m, ΔH_c, ΔH_m and X_c are clearly listed in Table 3. In Fig. 6a, all formulations showed identical crystallization peak characteristics of a single T_c peak. The pPP/APP and pPP/APP/HGM composites exhibited lower T_c and ΔH_c. vPP and pPP showed T_c values of 120°C and 122°C, respectively. It slightly decreased with the addition of HGM and APP by around 2°C–4°C together with the decrease of ΔH_c. This indicated the hindrance of the HGM and APP on the spherulite formation and growth under the cooling process [33].

In Fig. 6b, vPP showed only a single endothermic peak observed at 161°C, which corresponded to the melting of the α-crystalline phase [4,13]. For pPP, pPP/APP and pPP/APP/HGM composites, small endothermic jumps (at around 130°C) were found. For this case, it was further explained that the wasted PP was more structurally heterogeneous due to their different industrial processing and origins. Under industrial production, these products were quenched under the cooling process to reduce the production cycle time and increase productivity which induced the different crystal structures of the recycled products [34]. The shift of the T_m peak by 1–2°C was found for all formulations compared with that of pPP and vPP. The ΔH_m and X_c were found to reduce with increasing APP concentration. This indicated that both HGM and APP obstructed pPP crystallization. X_c was found to reduce from 40.75, in pPP, to 30.48% in pPP with 20 wt% APP and 5 wt% HGM. The reduction of the X_c was previously reported in recycled PP composites with fly ash [2], wood/glass fiber mixture [9] and construction debris [10].

The thermal stability of vPP, pPP and pPP composites obtained from TGA is graphically presented in Fig. 7. Their corresponding onset temperature (T_onset), end set temperature (T_{end set}), 5% weight loss (T_D5) and 50% weight loss (T_D50) were clearly indicated in Table 4. The decomposition curve characteristic of vPP took place over a single step. It started to decompose at around 397°C and it continued decomposing with T_D5 and T_D50 of 407 and 452°C, respectively. The T_{end set} and residual mass of 481°C and 0.18% were observed.

For pPP, the initial mass loss started earlier at a T_onset of 341°C and took place over two steps due to the decomposition of the shorter PP chain. This was the consequence of the pPP chain scission that occurred under the recycling process [35]. The second mass loss at approximately 400°C was the decomposition of the main pPP molecules. The residual mass of 0.13% was recorded at 481°C. The same decomposition behavior was observed for pPP/APP and pPP/APP/HGM composites. The T_D50 increased slightly with increasing APP concentrations. This is mainly due to the endothermic decomposition of APP which required more heat for NH₃ gas and phosphoric acid liberations. The residual component increased with APP loading due to the residue of HGM and metaphosphoric acid resulting from APP decomposition [36] as shown in Fig. 8.

Table 5 shows the data from the UL94 vertical burning test. vPP and pPP showed the non-flame-retardance property due to its high burning time together with fire dripping, flaming to the clamp and cotton ignition. With the addition of APP (A5), the burning time t₂ was significantly reduced compared with that of vPP and pPP, but this value was still longer than the limitation of the UL 94 standard. A synergistic flame retardancy behavior was found when combining HGM and APP in the H5A5 formulation. This resulted in the reduction of t₂ down to 49 s. Further addition of APP to 15 and 20 wt% was found to have faster self-extinguish with V-0 classification. The achieved fire-retardance property was mainly due to the liberation of NH₃ and H₂O together with the metaphosphoric formed during combustion. This was believed to block the oxygen to contact with pPP composites [37]. A reaction of the NH₃ liberation and phosphoric acid formation was illustrated in Fig. 8a. Further decomposition of phosphoric acid released one molecule of water and metaphosphoric acid residue as shown in Fig. 8b. To improve the performance of APP, many papers recommended adding the fire retardance facilitators such as pentaerythritol derivatives, carbohydrates or spumific agents. This was previously reported to significantly improve the flame-retardancy of APP with higher char formation [31,36,38].

Table 6 shows the effect of HGM and APP on the k value of pPP. The k value of vPP and pPP were 0.259 and 0.262 W/(m·K), respectively, while the k value of HGM was 0.208 W/(m·K). Adding APP to pPP (A5) was found to slightly increase the k value by around 3%, which was expected based on the previous works [39,40] and the high thermal conductivity of APP. The k value was significantly reduced due to the presence of low conductivity HGM. The minimum value of 0.203 W/(m·K) was observed in pPP with 5wt% HGM and 5 wt% APP. However, there was an upturn of the k value with the increase of the APP amount. Previous reports indicated the important role of HGM in reducing the thermal conductivity of polymer composites. In low-density polyethylene (LDPE), the k value was found to dramatically decrease with the HGM concentrations from 0.209 W/(m·K) to 0.171 W/(m·K) for the composites with 20 wt% HGM. Further investigation on the effects of HGM diameter indicated that larger HGM was more effective in reducing the k value than that of smaller ones. This is due to the longer heat transfer path of the larger HGM diameter, which delayed the heat transfer rate [41].

Based on the above result, HGM played an important role in k value reduction. The possible heat transfer mechanism was proposed in Fig. 9. By Assuming that all formulations were free of interfacial void or the heat transfer path was too small compared with that of HGM. The conductive heat transfer in pPP composites with HGM and APP was a combination of three possible ways. The first one was the thermal conduction by solid pPP and APP. Due to the higher k of APP, this may have increased the rate of heat transfer (q). The second was the radiation heat transfer on the surface between neighboring HGM particles. This slightly delayed the heat transfer by thermal dissipation. The last mechanism occurred through the HGM particles. In this case, the heat was transferred through the HGM wall by conduction followed by the convection of gas in the HGM particles [42]. This was believed to be the main reason for the k value reduction. To maximize the HGM performance, Li et al. recommended the use of lower density HGM. They indicated that the thermal conductivity through the HGM particles was found to be significantly reduced with a reduction in HGM’s true density [43].

Further observations on the density are clearly listed in Table 6. The density was found to increase with the addition of APP, but it tended to reduce with HGM addition. However, they were all higher than that of vPP and pPP. This tended to follow the rule of the mixture when the higher density APP was added. In epoxy, the density was found to reduce with the increasing volume fraction of HGM. Further comparison of the HGM wall thickness indicated that the thinner wall GHM exhibited more efficiency than that of thicker ones [44].