Preparation and Characterisation of Wood Polymer Composites Using Sustainable Raw Materials

In recent years, composites consisting of polymers and cellulosic materials have attracted increasing research attention. Polypropylene (PP) is among the most common polymer types found in excavated waste from landfills. Moreover, wood waste generated from wood products manufacturing such as sawdust (SD) offers a good potential for the fabrication of composite materials, and it is readily available in the environment. In this paper, wood polymer composites (WPC) consisting of recycled PP (rPP) and (SD) were prepared and characterised. A range of mechanical properties, including tensile strength, flexural properties, creep and hardness were studied, along with morphology, thermal properties, water degradation and contact angle. The results showed that the mechanical and thermal properties of rPP increased with an increase in 40 wt% of the SD content. Furthermore, the SD content significantly influenced the water uptake of the composites. Time–temperature superposition (TTS) was applied to predict the long-term mechanical performance from short-term accelerated creep tests at a range of elevated temperatures. The short-term creep test showed efficient homogeneity between the fillers and matrix with increasing temperature. The produced wood polymer composites displayed a comparable physical property to virgin polymer and wood and could potentially be used for various structural materials.


Morphology
The morphology of the cross-sectional surfaces of the rPPSD composites was analysed via scanning electron microscope (SEM, Hitachi 3030, Tokyo, Japan). The rPPSD composites were sputter-coated with gold. The observation was performed at a voltage of 20 kV perpendicular to the cross-sectional surface.

Thermogravimetric Analysis (TGA)
The thermal properties of the samples were analysed using a thermogravimetric analyser (TGA 4000, Perkin Elmer, Waltham, MA, USA). The thermographs were obtained at thermal scan temperatures from 30 to 800 °C at 30 °C/min under a nitrogen atmosphere.

Differential Scanning Calorimeter (DSC) Analysis
Differential scanning calorimeter (DSC 6000, Perkin Elmer, Waltham, MA, USA) was used to study the change in enthalpy of the rPPSD composites. First, the samples were heated from 30 to 190 °C at 15 °C/min. Next, the sample was held at 190 °C for 5 min to remove the thermal history, then cooled down to 30 °C at a rate of 15 °C/min to record the crystallisation rate. Subsequently, the samples were again heated from 30 to 190 °C at a rate of 15 °C/min to record the melting behaviour. To avoid oxidation in the samples, all heating and cooling runs were conducted under a nitrogen atmosphere. The percentage of crystallinity was calculated using Equation (1) [39,40] where ∆ is the melting enthalpy of rPP or rPPSD composites, is the wt% rPP in the composite and ∆ 0 is heat of fusion of 100% crystalline rPP, which is taken as 148 J/g [41].

Cleaning of Sawdust (SD)
Initially, the obtained SD was sieved as per ASTM E11 standard sieve to obtain 0.05 mm. Subsequently the SD was cleaned as per the methodology reported by Medupin et al. [36]. Briefly, 25 g of SD was mixed in 150 mL of 2 M NaOH solution and stirred for 45 min using overhead mechanically stirrer at 60 rpm. The mixture was then washed three times with deionised water then filtered and mixed in 10 M HCl to remove excess alkaline. Finally, the obtained SD was washed five times with deionised water and dried in a vacuum oven at 40 • C for 24 h [36].

Preparation of rPPSD Composites
The process of preparing the rPPSD composites is illustrated in Figure 1. Initially, the cleaned rPP and SD waste were pre-mixed in a zip lock bag according to the weight ratios ( Table 1). The materials were then fed into the hopper of an internal batch mixer (ZL-3011 Rubber Lab Banbury Kneader Mixer, Dongguan Zhongli Instrument Technology Co., Ltd., Dongguan, China). The spindle speed was kept at 8 rpm and the hopper temperature was maintained constant at 190 • C throughout the mixing process. The spindle rotation direction was changed every 2 min and mixing continued for 15 min to produce the composite. Subsequently, the obtained composite was crushed into small pieces using the plastic crusher. Finally, the obtained composite fragments were hot-pressed at 190 • C with a pressure of 7 N/mm 2 for 25 min to produce the dumbbell-shaped specimens for further characterisation. During the compression moulding process, the required thickness of the samples for further characterisation were prepared. Compression moulding shows great reproducibility and less cycle time [37,38].

Morphology
The morphology of the cross-sectional surfaces of the rPPSD composites was analysed via scanning electron microscope (SEM, Hitachi 3030, Tokyo, Japan). The rPPSD composites were sputter-coated with gold. The observation was performed at a voltage of 20 kV perpendicular to the cross-sectional surface.

Thermogravimetric Analysis (TGA)
The thermal properties of the samples were analysed using a thermogravimetric analyser (TGA 4000, Perkin Elmer, Waltham, MA, USA). The thermographs were obtained at thermal scan temperatures from 30 to 800 • C at 30 • C/min under a nitrogen atmosphere.

Differential Scanning Calorimeter (DSC) Analysis
Differential scanning calorimeter (DSC 6000, Perkin Elmer, Waltham, MA, USA) was used to study the change in enthalpy of the rPPSD composites. First, the samples were heated from 30 to 190 • C at 15 • C/min. Next, the sample was held at 190 • C for 5 min to remove the thermal history, then cooled down to 30 • C at a rate of 15 • C/min to record the crystallisation rate. Subsequently, the samples were again heated from 30 to 190 • C at a rate of 15 • C/min to record the melting behaviour. To avoid oxidation in the samples, all heating and cooling runs were conducted under a nitrogen atmosphere. The percentage of crystallinity was calculated using Equation (1) [39,40] where ∆H m is the melting enthalpy of rPP or rPPSD composites, X rPP is the wt% rPP in the composite and ∆H 0 m is heat of fusion of 100% crystalline rPP, which is taken as 148 J/g [41].

Tensile Testing
The tensile testing was conducted with a universal testing machine, Zhongli ZL-8001A (Dongguan Zhongli Instrument Technology Co., Ltd., Dongguan, China) at a crosshead speed of 3 mm/min with a load cell of 500 kN. The ASTM D638 Type 4-dimension was used for the tensile testing.

Flexural Testing
The flexural strength of the rPPSD composites was determined using a three-point flexural test on Instron 5890 (Norwood, MA, USA) universal testing machine according to ASTM standards (ASTM D790) [42].

Hardness
The hardness of the composites was measured using a Vickers hardness testing machine (DuraScan G5, Kuchl, Austria). Initially, a minor load of HV 0.2 was applied before the major load of HV 0.3. The load was then applied for 10 s, and the hardness value was recorded in HV units.

Creep Analysis
The creep compliance at lower temperature long time equalling creep compliance at higher temperature short time was used. The master curve of the creep compliance was measured in a three-point bending mode using dynamic mechanical analysis (DMA) (DMA 8000, Perkin Elmer, Waltham, MA, USA) at a frequency of 1 Hz. The dimensions of the sample were 30 mm × 10 mm with a thickness of 4 mm. The creep compliance is given by Equation (2) [43,44]; here, S is the creep compliance as a function of reference temperature (T ref ) and time (t), T elev is the elevated temperature and ∝ T is the shift factor. Creep and creep recovery cycles were conducted at isotherms between 20 and 70 • C at intervals of 5 • C. For each isotherm, 20% of the average flexural strength was applied for 1 h, followed by a 1 h recovery period.

Soil Burial Degradation
The rPPSD composites were buried at a depth of 2 cm in a mixture of 50% sand and 50% black humus soil at ambient temperature [45,46]. The relative humidity of the soil was about 50-60%. The degradation of the composites was assessed at predetermined intervals of 4, 8, 12, 16 and 20 days, respectively. Before weighing, the composites were removed from the soil and washed with distilled water. The weight loss of the composites buried in soil was obtained using Equation (3) [45,46]. Here, M 1 and M 2 represent the weight of the composites before and after soil burial, respectively.

Water Absorption
Samples with a dimension 30 × 30 × 3 mm 3 were used for the water absorption analysis. The samples were dried in an oven for 24 h at 80 • C to obtain constant weight. Subsequently, the initial weight (W 1 ) was recorded. The samples were then immersed in water for 2, 4, 6, 8, 10, 12, 14, 16, 18 and 20 days, respectively. The weight was recorded at regular time intervals to obtain the water absorption (%). The water absorption of the sample was calculated using Equation (4) [47].
where W 1 and W 2 are the sample weights before and after immersion, respectively.

Water Contact Angle
The wettability of the composites was measured using the Attension Theta Flux (Biolin Scientific, Västra Frölunda, Sweden). The contact angle measurement was conducted via sessile drop technique. The droplet was placed on the surface of the samples using a micrometre syringe, and the contact angle was measured by scanning the droplet profile for 15 s [48,49]. To avoid the effects of weight, the size of the water droplet was maintained at about 2 µL [50].

Statistical Analyses
GraphPad Prism 9.0 (GraphPad Software, Inc., San Diego, CA, USA) was used to evaluate statistical analyses using the ANOVA method. Three samples (n = 3) were investigated for each data set and presented as mean ± standard deviation (SD) unless otherwise stated. A significance level of the p-value of ≤0.05 was determined to be significant (*). Error bars in all figures represent the standard error of the mean [51][52][53].

Morphology
The properties of the rPPSD composites are highly dependent on the dispersion of SD in the matrix and the interaction between the SD and the polymer matrix. Therefore, the morphology of the SD, rPP and rPPSD composites was observed to provide further insight into the microstructures. Figure 2a shows the micrograph of the SD. As is evident from the image, SD was loose and rough. The roughness of the SD was further evident in the composites fabricated with rPP. Figure 2b shows the cross-sectional SEM image of the rPP, which was smooth and featureless [40]. The strength and interfacial interactions determined the composites failure mode and micromechanical deformation, as reported by Renner et al. [54]. Figure 2c,d shows the morphology of rPPSD4 composites at lower and higher magnification micrographs. The red circles highlighted the SD embedded in the polymer matrix, suggesting good interfacial adhesion between SD and polymer matrix, along with uniform dispersion of SD throughout the cross-section of the compressed samples [6,40].

Statistical Analyses
GraphPad Prism 9.0 (GraphPad Software, Inc., San Diego, CA, USA) was used to evaluate statistical analyses using the ANOVA method. Three samples (n = 3) were investigated for each data set and presented as mean ± standard deviation (SD) unless otherwise stated. A significance level of the p-value of ≤0.05 was determined to be significant (*). Error bars in all figures represent the standard error of the mean [51][52][53].

Morphology
The properties of the rPPSD composites are highly dependent on the dispersion of SD in the matrix and the interaction between the SD and the polymer matrix. Therefore, the morphology of the SD, rPP and rPPSD composites was observed to provide further insight into the microstructures. Figure 2a shows the micrograph of the SD. As is evident from the image, SD was loose and rough. The roughness of the SD was further evident in the composites fabricated with rPP. Figure 2b shows the cross-sectional SEM image of the rPP, which was smooth and featureless [40]. The strength and interfacial interactions determined the composites failure mode and micromechanical deformation, as reported by Renner et al. [54]. Figure 2c,d shows the morphology of rPPSD4 composites at lower and higher magnification micrographs. The red circles highlighted the SD embedded in the polymer matrix, suggesting good interfacial adhesion between SD and polymer matrix, along with uniform dispersion of SD throughout the cross-section of the compressed samples [6,40].

Thermogravimetric Analysis (TGA)
The TGA and DTG curves for the rPPSD composites are shown in Figures 3 and 4, while the thermal properties are shown in Table 2. The weight loss occurring between 70 and 110 • C was associated with the evaporation of absorbed moisture from the samples. In general, all the composites were thermally stable up to 200 • C. From the TGA curves, it is evident that hemicellulose started its degradation reactions followed by the more thermally stable cellulose domains at 250 • C [55][56][57]. Degradation of cellulose and hemicellulose involved complex reactions comprised in the temperature range of 250-370 • C. The degra-dation of lignin occurred in a wider temperature range from 250-480 • C. The addition of SD into the rPP matrix enhanced the thermal stability of composites. As shown in Table 2, the onset temperature progressively increased from 255 • C for rPP to 350 • C for rPPSD4. A similar trend can be observed in the end set temperatures of the rPPSD composites where the temperature increased from 452 • C (rPP) to 470 • C (rPPSD4). The contents at 700 • C are residual char or ashes from decomposition of saw dust. The increase in on set and end set temperature with increasing SD composition implied that SD enhanced thermal insulation behaviour of rPP matrix. According to Chun et al. [58] and Zander et al. [59], the rPP reinforced with cellulose fibres displayed improved thermal stability with an increase in fibre content. Similar observations were made with other types of WPCs [8,12,[60][61][62][63][64][65][66].
while the thermal properties are shown in Table 2. The weight loss occurring between 70 and 110 °C was associated with the evaporation of absorbed moisture from the samples. In general, all the composites were thermally stable up to 200 °C. From the TGA curves, it is evident that hemicellulose started its degradation reactions followed by the more thermally stable cellulose domains at 250 °C [55][56][57]. Degradation of cellulose and hemicellulose involved complex reactions comprised in the temperature range of 250-370 °C. The degradation of lignin occurred in a wider temperature range from 250-480 °C. The addition of SD into the rPP matrix enhanced the thermal stability of composites. As shown in Table 2, the onset temperature progressively increased from 255 °C for rPP to 350 °C for rPPSD4. A similar trend can be observed in the end set temperatures of the rPPSD composites where the temperature increased from 452 °C (rPP) to 470 °C (rPPSD4). The contents at 700 °C are residual char or ashes from decomposition of saw dust. The increase in on set and end set temperature with increasing SD composition implied that SD enhanced thermal insulation behaviour of rPP matrix. According to Chun et al. [58] and Zander et al. [59], the rPP reinforced with cellulose fibres displayed improved thermal stability with an increase in fibre content. Similar observations were made with other types of WPCs [8,12,[60][61][62][63][64][65][66].   while the thermal properties are shown in Table 2. The weight loss occurring between 70 and 110 °C was associated with the evaporation of absorbed moisture from the samples. In general, all the composites were thermally stable up to 200 °C. From the TGA curves, it is evident that hemicellulose started its degradation reactions followed by the more thermally stable cellulose domains at 250 °C [55][56][57]. Degradation of cellulose and hemicellulose involved complex reactions comprised in the temperature range of 250-370 °C. The degradation of lignin occurred in a wider temperature range from 250-480 °C. The addition of SD into the rPP matrix enhanced the thermal stability of composites. As shown in Table 2, the onset temperature progressively increased from 255 °C for rPP to 350 °C for rPPSD4. A similar trend can be observed in the end set temperatures of the rPPSD composites where the temperature increased from 452 °C (rPP) to 470 °C (rPPSD4). The contents at 700 °C are residual char or ashes from decomposition of saw dust. The increase in on set and end set temperature with increasing SD composition implied that SD enhanced thermal insulation behaviour of rPP matrix. According to Chun et al. [58] and Zander et al. [59], the rPP reinforced with cellulose fibres displayed improved thermal stability with an increase in fibre content. Similar observations were made with other types of WPCs [8,12,[60][61][62][63][64][65][66].    The DSC curves of rPPSD composites are shown in Figure 5. The values of T m (melting temperature), T c (crystallisation temperature), ∆H c (crystallisation enthalpy), ∆H m (melting enthalpy) and X c (crystallinity; %) are tabulated in Table 3. The T c increased from 117.1 • C to 120.9 • C as the weight of SD in rPP increased from 10 to 40 wt% [67]. The primary mechanism responsible for the crystallisation and melting behaviour of the matrix is heterogeneous nucleation on the SD surfaces [68,69]. Therefore, the increase in T c and T m values of the rPPSD composites was due to the presence of SD which acts as a nucleating agent [70][71][72].
(melting enthalpy) and Xc (crystallinity; %) are tabulated in Table 3. The Tc increased from 117.1 °C to 120.9 °C as the weight of SD in rPP increased from 10 to 40 wt% [67]. The primary mechanism responsible for the crystallisation and melting behaviour of the matrix is heterogeneous nucleation on the SD surfaces [68,69]. Therefore, the increase in Tc and Tm values of the rPPSD composites was due to the presence of SD which acts as a nucleating agent [70][71][72].
The values of ΔHc (crystallisation enthalpy) and ΔHm (melting enthalpy) for rPP were 95.0 and 93.7 J/g, respectively. The ΔHc and ΔHm were significantly increased with the increase in SD reinforcement into rPP. Ndiaye et al. [40] stated that the polymer in the wood polymer composite formed crystals much more easily due to the nucleation effect of the wood particles with the addition of lower content of wood particles. According to Lee et al. [73], more heat energy was absorbed by wood flour in melting the WPC. The Xc of the rPPSD composites increased with the increase in SD incorporation. The Xc for rPPSD1, rPPSD2, rPPSD3 and rPPSD4 were 56.5%, 57.7%, 58.5% and 59.8%, respectively. Beg et al. [74] reported a 3.3% increase of Xc with 40 wt% of wood flour in PP. Similar observations were made in the literature [73][74][75][76][77].    The values of ∆H c (crystallisation enthalpy) and ∆H m (melting enthalpy) for rPP were 95.0 and 93.7 J/g, respectively. The ∆H c and ∆H m were significantly increased with the increase in SD reinforcement into rPP. Ndiaye et al. [40] stated that the polymer in the wood polymer composite formed crystals much more easily due to the nucleation effect of the wood particles with the addition of lower content of wood particles. According to Lee et al. [73], more heat energy was absorbed by wood flour in melting the WPC. The X c of the rPPSD composites increased with the increase in SD incorporation. The X c for rPPSD1, rPPSD2, rPPSD3 and rPPSD4 were 56.5%, 57.7%, 58.5% and 59.8%, respectively. Beg et al. [74] reported a 3.3% increase of X c with 40 wt% of wood flour in PP. Similar observations were made in the literature [73][74][75][76][77].

Tensile Properties
The tensile performance of the rPPSD composites is shown in Figure 6. The rPP presented a tensile strength of 25.5 MPa and Young's modulus of 3.79 MPa. The incorporation of SD increased the ultimate tensile strength and Young's modulus with rPPSD4 showed the highest ultimate tensile strength and Young's modulus of 31 MPa and 6.95 MPa, as shown in Table 4. Ndiaye et al. [78] reported that the highest tensile strength was found in 70% PP with 30% wood composite. Najafi et al. [79] reported similar results with WPCs made of recycled plastics, where the stress concentration increased with an increase in the wood content. As shown in the SEM micrographs (Figure 2a-d), there was no clear gap between the SD and rPP matrix, showing a good interface bonding and indicating the stress transfer from the weaker matrix to the strong wood fibre.

Tensile Properties
The tensile performance of the rPPSD composites is shown in Figure 6. The rPP presented a tensile strength of 25.5 MPa and Young's modulus of 3.79 MPa. The incorporation of SD increased the ultimate tensile strength and Young's modulus with rPPSD4 showed the highest ultimate tensile strength and Young's modulus of 31 MPa and 6.95 MPa, as shown in Table 4. Ndiaye et al. [78] reported that the highest tensile strength was found in 70% PP with 30% wood composite. Najafi et al. [79] reported similar results with WPCs made of recycled plastics, where the stress concentration increased with an increase in the wood content. As shown in the SEM micrographs (Figure 2a-d), there was no clear gap between the SD and rPP matrix, showing a good interface bonding and indicating the stress transfer from the weaker matrix to the strong wood fibre.

Flexural Properties
The flexural behaviour of the rPPSD composites is shown in Figure 7. The flexural strength of the rPPSD composites increased linearly with the increase of SD content. The rPP showed the lowest flexural strength at 17 MPa, followed by rPPSD1, rPPSD2 and rPPSD3 with 19.5, 20.5 and 21.5 MPa, respectively. The rPPSD4 showed the highest flexural strength among all composites, at 23 MPa. The failure of specimens initiated with a crack on the tension side and grew until complete failure. Ratanawilai et al. [80] reported that when PP was mixed with wood flour at different concentrations, the highest flexural strength was found at 60% PP and 40% wood content. Furthermore, uniform dispersion and SD content play a significant role in determining the flexural properties of WPCs [78].

Flexural Properties
The flexural behaviour of the rPPSD composites is shown in Figure 7. The flexural strength of the rPPSD composites increased linearly with the increase of SD content. The rPP showed the lowest flexural strength at 17 MPa, followed by rPPSD1, rPPSD2 and rPPSD3 with 19.5, 20.5 and 21.5 MPa, respectively. The rPPSD4 showed the highest flexural strength among all composites, at 23 MPa. The failure of specimens initiated with a crack on the tension side and grew until complete failure. Ratanawilai et al. [80] reported that when PP was mixed with wood flour at different concentrations, the highest flexural strength was found at 60% PP and 40% wood content. Furthermore, uniform dispersion and SD content play a significant role in determining the flexural properties of WPCs [78].

Hardness
The optical micrographs of the microhardness indentation are shown in Figure 8a-e and hardness values of the composites are shown in Figure 8f. A continuous increase in hardness values was observed as the concentration of SD increased in the rPPSD composites [81]. This increase in hardness can be attributed to the fact that the addition of SD in

Hardness
The optical micrographs of the microhardness indentation are shown in Figure 8a-e and hardness values of the composites are shown in Figure 8f. A continuous increase in hardness values was observed as the concentration of SD increased in the rPPSD composites [81]. This increase in hardness can be attributed to the fact that the addition of SD in rPP imparted stiffness, thus making the composites rigid and hard and restricting the mobility of polymeric chains [82,83]. The rPP had a hardness of 1.5 HV, whereas the composites had a hardness of 3.7 HV (rPPSD1), 6.3 HV (rPPSD2), 7.8 HV (rPPSD3) and 9.8 HV (rPPSD4). The present results agreed with the prior literature [78,84,85].

Hardness
The optical micrographs of the microhardness indentation are shown in Figure 8a-e and hardness values of the composites are shown in Figure 8f. A continuous increase in hardness values was observed as the concentration of SD increased in the rPPSD composites [81]. This increase in hardness can be attributed to the fact that the addition of SD in rPP imparted stiffness, thus making the composites rigid and hard and restricting the mobility of polymeric chains [82,83]. The rPP had a hardness of 1.5 HV, whereas the composites had a hardness of 3.7 HV (rPPSD1), 6.3 HV (rPPSD2), 7.8 HV (rPPSD3) and 9.8 HV (rPPSD4). The present results agreed with the prior literature [78,84,85].

Creep Analysis
The creep analysis was performed using TTS to predict the long-term creep behaviour of the rPPSD composites from short-term accelerated creep tests at a range of elevated

Creep Analysis
The creep analysis was performed using TTS to predict the long-term creep behaviour of the rPPSD composites from short-term accelerated creep tests at a range of elevated temperatures. rPPSD4 was taken as the measuring sample. Figure 9 presents the unshifted short-term creep compliance and corresponding master curves of rPPSD4 at all tested temperatures, which were plotted against the test time on a log scale. With time and temperature increases, the creep compliance increased due to composite becoming less stable as the magnitude of creep strain increased over the same period of loading, which affects the viscoelastic region of the composite. For example, the creep compliance increased from 0.35 1/GPa at 30 • C to 0.44 1/GPa at 55 • C as the composite experienced greater deformation due to the constant applied stress. The shift factor was calculated using the modified William-Landel-Ferry (WLF) equation.
For the rPPSD4 composite, the modified WLF equation was employed as presented by Nielsen et al. [86] (see Equation (5) below) to calculate the shift factor when a temperature other than T g is chosen as the reference temperature.
Here, ∝ T is the horizontal shift factor for the corresponding elevated temperature, T ( • C); the reference temperature is T ref ( • C); and finally, C 1 and C 2 are the empirical constants determined from Equations (6) and (7).
Polymers 2022, 14, 3183 11 of 18 In the above, C 1g and C 2g are the empirical constants (C 1g = 17.44 and C 2g = 51.6 • C), while T g is the glass transition temperature ( • C).
By substituting the T g ( • C) of WPC materials into Equations (6) and (7), the values of C 1 and C 2 were calculated to be 10.5 and 85.70 • C, respectively. Finally, by substituting these values in Equation (5), the shift factor for the rPPSD4 composite was calculated to be 4.025 × 10 −3 for 45 • C [30,86,87].
Polymers 2022, 13, x FOR PEER REVIEW 11 of 18 temperatures. rPPSD4 was taken as the measuring sample. Figure 9 presents the unshifted short-term creep compliance and corresponding master curves of rPPSD4 at all tested temperatures, which were plotted against the test time on a log scale. With time and temperature increases, the creep compliance increased due to composite becoming less stable as the magnitude of creep strain increased over the same period of loading, which affects the viscoelastic region of the composite. For example, the creep compliance increased from 0.35 1/GPa at 30 °C to 0.44 1/GPa at 55 °C as the composite experienced greater deformation due to the constant applied stress. The shift factor was calculated using the modified William-Landel-Ferry (WLF) equation. For the rPPSD4 composite, the modified WLF equation was employed as presented by Nielsen et al. [86] (see Equation (5) below) to calculate the shift factor when a temperature other than Tg is chosen as the reference temperature.
Here, ∝ is the horizontal shift factor for the corresponding elevated temperature, T (°C); the reference temperature is Tref (°C); and finally, 1 and 2 are the empirical constants determined from Equations (6) and (7).
In the above, 1 and 2 are the empirical constants ( 1 = 17.44 and 2 = 51.6 °C), while is the glass transition temperature (°C). By substituting the (°C) of WPC materials into Equations (6) and (7), the values of 1 and 2 were calculated to be 10.5 and 85.70 °C, respectively. Finally, by substituting these values in Equation (5), the shift factor for the rPPSD4 composite was calculated to be 4.025 × 10 −3 for 45 °C [30,86,87].

Soil Burial Degradation
Soil burial degradation occurs due to moisture and enzymatic action in soil, leading to weight loss in the material [88,89]. Figure 10 shows the weight loss of each composite

Soil Burial Degradation
Soil burial degradation occurs due to moisture and enzymatic action in soil, leading to weight loss in the material [88,89]. Figure 10 shows the weight loss of each composite after being subjected to soil burial for 20 days. The weight loss percentage are 0.45% for rPP, 1.7% for rPPSD1, 2.3% for rPPSD2, 2.8% for rPPSD3 and 3.3% for rPPSD4. With the increase in wt% of SD, the weight loss percentage for soil burial degradation of rPPSD composites increased. The chemical contents such as cellulose, hemicellulose and lignin presented in SD reacted with soil and caused weight loss of the composite through soil burial [45,46]. The hydrolysis of the polymer backbone was the primary reason for the degradation of rPP. The increased water absorption (as depicted in Section 3.8) in composites with higher wt% of SD leads to more pronounced hydrolysis compared to rPP. Further, with the help of the moisture in the soil, the polymer chains were demolished by creating tiny fragments of rPPSD composite [46]. Finally, the microbial activities promoted the weight loss of the composite during soil burial degradation [89,90]. after being subjected to soil burial for 20 days. The weight loss percentage are 0.45% for rPP, 1.7% for rPPSD1, 2.3% for rPPSD2, 2.8% for rPPSD3 and 3.3% for rPPSD4. With the increase in wt% of SD, the weight loss percentage for soil burial degradation of rPPSD composites increased. The chemical contents such as cellulose, hemicellulose and lignin presented in SD reacted with soil and caused weight loss of the composite through soil burial [45,46]. The hydrolysis of the polymer backbone was the primary reason for the degradation of rPP. The increased water absorption (as depicted in Section 3.8) in composites with higher wt% of SD leads to more pronounced hydrolysis compared to rPP. Further, with the help of the moisture in the soil, the polymer chains were demolished by creating tiny fragments of rPPSD composite [46]. Finally, the microbial activities promoted the weight loss of the composite during soil burial degradation [89,90]. The study conducted by Yang et al. [46] on bamboo fibre-reinforced PP composites stated that with the increase in soil burial time and filler concentration, the weight loss of composites increased. The present study aligns with the study of Yang et al. [46].

Water Absorption
The water absorption behaviour of the rPPSD composites measured for 20 days with measurements taken at regular intervals is shown in Figure 11. The water absorption increased quickly in the first 3 days, and it slowed down as the immersion time prolonged until the specimen's water content was saturated [91,92]. When the samples were in an The study conducted by Yang et al. [46] on bamboo fibre-reinforced PP composites stated that with the increase in soil burial time and filler concentration, the weight loss of composites increased. The present study aligns with the study of Yang et al. [46].

Water Absorption
The water absorption behaviour of the rPPSD composites measured for 20 days with measurements taken at regular intervals is shown in Figure 11. The water absorption increased quickly in the first 3 days, and it slowed down as the immersion time prolonged until the specimen's water content was saturated [91,92]. When the samples were in an equilibrium state of water absorption (allowing the water absorption to change in time range and period of immersion), the percentages of water absorption of rPP, rPPSD1, rPPSD2, rPPSD3 and rPPSD4 were 1.15%, 4.95%, 9.25%, 12.82% and 17%, respectively. The rPP showed a low absorption percentage as it was hydrophobic and thus absorbed very little water [93]. The composites showed similar water absorption curves for the entire immersion period. Therefore, the water absorption percentages were considered to be consistent with Fickian diffusion [94][95][96]. Moreover, it shows that SD played a profound role on the water absorption behaviour of samples, that is, the water absorption increased with the increase of SD content, which is also consistent with the studies reported by other researchers [6,[97][98][99][100][101][102].

Water Contact Angle
The water contact angle was measured to investigate the hydrophilicity of the rPPSD composites, as shown in Figure 12. According to the literature, a contact angle below 90° indicates a good wetting surface by any liquid [53,[103][104][105]. The measurements of water contact were in accordance to those reported by Sdrobiş et al. [48] and Wang et al. [49]. Figure 12 shows the sessile drop images and the contact angle measurements of the composite surfaces. The rPP possessed the highest contact angle, with 66.1°, compared to the rPPSD composites as it was hydrophobic in nature, thus absorbing less water [106]. On the other hand, composites with the incorporation of SD showed a decrease in water contact angle, attributed to the hydrophilic nature of SD [107]. The average contact angles of rPPSD1, rPPSD2, rPPSD3 and rPPSD4 were 65.18°, 61.25°, 59.78° and 56.18°, respectively. Similarly, Lazrak et al. [108] studied the wetting behaviour of wood flour reinforced rHDPE composites, and it was stated that a decrease in contact angle can be attributed to an increase in wood flour, which is hydrophilic in nature. The present study aligns with the literature mentioned above.

Water Contact Angle
The water contact angle was measured to investigate the hydrophilicity of the rPPSD composites, as shown in Figure 12. According to the literature, a contact angle below 90 • indicates a good wetting surface by any liquid [53,[103][104][105]. The measurements of water contact were in accordance to those reported by Sdrobiş et al. [48] and Wang et al. [49]. Figure 12 shows the sessile drop images and the contact angle measurements of the composite surfaces. The rPP possessed the highest contact angle, with 66.1 • , compared to the rPPSD composites as it was hydrophobic in nature, thus absorbing less water [106]. On the other hand, composites with the incorporation of SD showed a decrease in water contact angle, attributed to the hydrophilic nature of SD [107]. The average contact angles of rPPSD1, rPPSD2, rPPSD3 and rPPSD4 were 65.18 • , 61.25 • , 59.78 • and 56.18 • , respectively. Similarly, Lazrak et al. [108] studied the wetting behaviour of wood flour reinforced rHDPE composites, and it was stated that a decrease in contact angle can be attributed to an increase in wood flour, which is hydrophilic in nature. The present study aligns with the literature mentioned above. the other hand, composites with the incorporation of SD showed a decrease in water contact angle, attributed to the hydrophilic nature of SD [107]. The average contact angles of rPPSD1, rPPSD2, rPPSD3 and rPPSD4 were 65.18°, 61.25°, 59.78° and 56.18°, respectively. Similarly, Lazrak et al. [108] studied the wetting behaviour of wood flour reinforced rHDPE composites, and it was stated that a decrease in contact angle can be attributed to an increase in wood flour, which is hydrophilic in nature. The present study aligns with the literature mentioned above.

Conclusions
This study successfully fabricated rPPSD composites using an internal batch mixer and compression moulding technique. The morphology, mechanical, thermal, soil burial degradation, water absorption and wettability properties of the produced rPPSD composites with varied SD and rPP contents were analysed. The morphological images showed that SD dispersed in the rPP matrix uniformly. The thermal properties showed that increased content of SD in the rPP matrix improved the thermal stability of rPPSD composites. Furthermore, the tensile and flexural strength increased from 25.5 MPa to 31 MPa and 18 MPa to 24 MPa for 40 wt% of SD in rPP, respectively. The creep compliance for rPPSD4 was increased from 0.35 1/GPa at 30 • C to 0.44 1/GPa at 50 • C as the composite experienced greater deformation under constant applied stress. In addition, the soil burial degradation showed a considerable weight loss, up to 3.3%, with the increase in SD content. Similarly, the water absorption increased with the increase in the SD concentration, while the wettability increased in rPPSD composite with higher concentrations of SD; for example, the water contact angle decreased and hydrophilicity increased. Hence, the results indicated that the properties and the performances of the rPPSD are similar or comparable to composites made of virgin wood and plastics. The rPPSD could potentially be used as an alternative material to replace non-sustainable composites.