Biochar/polypropylene composites: A study on the effect of pyrolysis temperature on crystallization kinetics, crystalline structure, and thermal stability

2.1 Materials

The polymer matrix used in this study for the preparation of the composite was an isotactic homopolymer polypropylene (LADENE PP570P) pellets obtained from the Saudi Basic Industries Corporation (SABIC, Saudi Arabia). This resin has a melt flow index (MFI) of 8 g/10 min (at 230 °C and 2.16 kg load) and a density of 0.905 g/cm³. The BC particles, used in this study, were prepared through the pyrolysis of date palm tree wastes. The pyrolysis process was carried out at different temperatures, and under inert atmosphere (N₂ gas flow rate of 100 ml/min), with heating rate of 10 °C/min and residence time of 4 h. The range of pyrolysis was kept in the range of 300 to 700 °C. The difference between the pyrolysis temperature was kept at 100 °C. To carry out the process an electrical muffle furnace equipped with a stainless-steel reactor, having length of 720 mm and 50 mm diameter, was used. All BC samples were milled in a home blender to reduce the particle size and sieved to <38 μm. The detailed analysis of these BC samples are given in our previous work (Elnour et al., 2019), while a summary of their particle size distributions and physicochemical characteristics are provided as supplementary information (Tables S1 and S2).

2.2 Composites preparation

The BC/PP composite samples were prepared through the melt compounding process. Before compounding, both BC and PP were dried in an oven at 110 °C for 24 h to eliminate the moisture content. Afterward, BC and PP were weighed and premixed as per the required BC loading percentages. The mixture was then introduced to a melt compounder (DSM Xplore micro-compounder, 15 cm³, Netherlands). The composites melt compounding process was performed in a speed-controlled mode at a temperature of 190 °C, screw speed of 100 rpm, and compounding time of 3 min. At the end of the compounding, the BC/PP composites were collected, stored, used for morphological, and thermal analysis to evaluate the microstructure-properties relationship and the crystallization behavior of the composite. The compositions of the samples and their treatment temperatures different formulations are presented in supplementary information (Table S3).

2.3 Composites characterization

The tensile-fractured surface of the BC/PP composites was investigated by using scanning electron microscopy (SEM), (JEOL Model: JSM-6360A, Japan). ASTM standard samples were subjected to a tensile load with an extension rate of 50 mm/min in order to obtain a fractured specimen. The small part of the specimen was placed on the SEM stub in such a manner that the broken surface is exposed on the top side for examination. Differential scanning calorimetric (DSC) measurements were performed, for the neat PP and the BC/PP composite samples, on a Shimadzu thermal analyzer (Model: TA-60WS), according to ASTM standard method (ASTM, 2015). The DSC instrument was calibrated with pure indium (In) metal before analysis. Each sample ∼ 6–7 mg was weighed carefully and placed in an aluminum pan, which was then covered with an aluminum lid and crimped prior to analysis. To eliminate the thermal history of the samples, the neat and BC/PP samples were first heated from (30 to 250 °C) at a heating rate of 10 °C/min and subsequently cooled to 30 °C at the same rate. The second heating–cooling cycle was performed right after the first cycle and on typical rate and conditions. To avoid any thermal degradation of the samples, the thermal analysis was performed under a nitrogen (N₂) atmosphere at a flow rate of 50 cm³/min. The melting temperature, crystallization temperature, and enthalpy of melting were analyzed using the TA-60WS software package based on the second heating–cooling cycle.

To determine the degree of crystallinity of the neat PP and BC/PP composites, the measured melting enthalpy ( $\mathrm{\Delta }{H}_m$ ) of each sample was compared to the value for 100% crystalline PP ( $\mathrm{\Delta }{H}_m^o$ ). Then, the crystallinity percentage of each sample was calculated by using the total enthalpy method based on the following equation.

The crystallization kinetics of The neat PP and its composites were analyzed by using the well-known Avrami model given by the following equation:

The relative crystallinity $\theta (t)$ can be defined as the area ratio for an exothermic peak at a specific time (t) divided by the total area of the exothermic peak. The relative crystallinity $\theta (t)$ at time t is given by the following equation:

Eq. (2) above can be transformed into a linear logarithmic in the following form.

For the crystallization kinetics, based on the experimental heating rate, the time t is related to the temperature T as follows:

If the crystallization kinetics is following Avrami’s theory well, a plot of $ln\left[-ln\left(1-\theta (t)\right)\right]$ versus $ln(t)$ should yield a straight line with a slope n and an intercept $ln(k)$ , from which k values could be obtained.

Also from equation (2), the half-time of crystallization (t_1/2) for crystallization kinetics, which is defined as the time at which the relative crystallinity of (50%) has been achieved (Li et al., 2006), can be expressed by the following equation:

To observe the effect of BC incorporation on the spherulite morphology of BC/PP composites, polarized light microscopy (PLM, BX51, Olympus, Japan) was used. A few fractions of samples are placed on a glass slide and heated to melt and allowed to form a film at 230 °C. The film was cooled at a rate of 40 °C/min to attain the temperature of 120 °C, where it was retained for the observation of its spherulitic evolution over time. Thermogravimetric analysis (TGA) was used to investigate the thermal stability of BC/PP composites. For determining the thermal decomposition temperature and the residual weight of the BC/PP composites. The TGA analysis was performed, according to the ASTM E1131 standard test method (ASTM, 2020), in a Shimadzu apparatus with a simultaneous DTA-TGA thermal analyzer (Model: DTG-60H). The aluminum pan was filled with ∼ 10 mg of the composite sample. Subsequently, the sample was heated from the ambient temperature up to 600 °C with a heating rate of 10 °C/min. Aluminum oxide was used as reference material. The analysis was done under air atmosphere with a flow rate of (50 cm³/min) and accordingly, the corresponding weight loss was recorded.

3.1 Morphology study

Fig. 1 illustrates the fracture morphology of the BC/PP composites. It is well known that the filler dispersion and the filler-polymer interfacial interaction are two significant determining factors that influence the properties of the resulted composites. The fractured surface micrographs of the composites at the maximum BC loading (i.e. 20 wt%) identify the different morphological characteristics of composites. The figure showed uniform distribution of BC into the PP matrix (some BC particles are highlighted with circles in Fig. 1). Furthermore; it can be observed from Fig. 1 that the interfacial bonding is highly dependent on the BC type, as evident from the degree of pull-out of BC particles (indicated with arrows in Fig. 1). The composites made with BC prepared at higher temperatures showed less pull-out effect than those made with BC prepared at lower temperatures. This indicates that the interface adhesion between BC particles and the PP matrix can be classified as (BC700 > BC600 > BC500 > BC400 > BC300). This later phenomenon of enhanced interface interactions can be correlated to the large surface area, porous structure, and surface chemistry of BC prepared at higher temperatures (Elnour et al., 2019). It is worth highlighting here that in this study no compatibilizer was used for improving the filler-polymer interfacial bonding, which confirms that this trend of enhanced interface quality is merely due to the inherent characteristics of BC type. Moreover; in our previous study we also illustrated the dependence of PP diffusion into BC pores on the type of BC and we showed that the molten PP is infiltrated into the BC pores, which resulted in better mechanical interlocking between the BC particles with the PP matrix, due to the enhanced filler-matrix interface (Tjong, 2012).

Close

Fig. 1

SEM morphologies of fractured-surface composites at the maximum BC loading (i.e. 20 wt%).

Export to PPT

3.2 Differential scanning calorimetry (DSC)

The crystalline structure as well as the degree of crystallinity of semi-crystalline polymers can affect the physical and mechanical properties of selected polymer significantly. As mentioned earlier, the crystallinity and crystallization kinetics of semi-crystalline polymers depend mainly on the processing conditions employed and the presence of carbonaceous fillers. During solidification from the melt, carbonaceous fillers can act as heterogeneous nucleation sites for the polymer crystallites. This reduces nucleation activation energy, leading to an acceleration of the crystallization and a decrease of the spherulite size (Elnour et al., 2019).

The thermal behaviors of the neat PP and the BC/PP composite systems are presented in Figs. 2 and 3 for the melting and crystallization peaks respectively. The numerical values of the DSC thermograms were extracted and the results are summarized in Table 1. The heat of melting (ΔH_m) and crystallization heats (ΔH_c) were determined from the areas of melting and crystallization peaks of the DSC thermograms, respectively.

Close

Fig. 2

Melting peaks of BC/PP composites, (A) 5%, (B) 10%, (C) 15% and (D) 20% loading.

Export to PPT

Close

Fig. 3

Crystallization peaks of BC/PP composites, (A) 5%, (B) 10%, (C) 15% and (D) 20% loading.

Export to PPT

From Fig. 2, it could be noticed that the neat PP showed a typical endothermic melting peak at ∼ 163 °C. When BC is incorporated into PP, the melting temperature (T_m) of BC/PP composites is increased compared to the neat PP. This increase in the T_m could be attributed to the heterogeneous nucleation effects exerted by the rigid fillers. Similarly, the crystallization temperatures (T_c) shown in Fig. 3 were also found to relatively increase toward higher temperatures (earlier onset crystallization temperature) for all the composites compared to the neat PP matrix. The BC particles act as points from where crystal growth is initiated. This type of early onset of crystallization temperature, as a result of carbonaceous filler addition, has been noticed and reported in the literature (Wang et al., 2018; Yu et al., 2018; Lin et al., 2018). It is also interesting to observe that crystallization temperature (T_c) increases proportionally with the amount of BC loading in all the composites. This might be related to the increased number of the nucleus that is available during the crystallization process as a result of the increased number of BC particles at higher loadings. Thus, the higher loading of BC particles aids in earlier crystallization of the PP matrix. Similarly, one can observe that at specific loading, BC prepared at higher temperatures induce more crystallization effect into PP, or in other words, the crystallization induction could be classified as (BC700 > BC600 > BC500 > BC400 > BC300). This behavior might be attributed to the higher surface areas and pore volumes of BC prepared at higher temperatures (Table S2). Higher surface area and pore volumes lead to more infiltration of the PP chain into the pores of BC particles and more BC-PP interlocking. A similar observation is also reported in the literature (Bhattacharyya et al., 2003). Another possible reason might be attributed to the BC hydrophobicity (Elnour et al., 2019), which increases with pyrolysis temperature and makes the BC particles more compatible with the hydrophobic PP matrix. Though the difference in the hydrophobicity of the BC prepared at different temperatures is not prominent, however, these factors (surface area, pore-volume, and hydrophobicity) could act synergistically for enhancement of the overall composite crystallization process.

The analysis of the melt crystallization process for different BC types by using Avrami’s equation can be very useful for understanding the crystallization kinetics.

It is well acknowledged that the Avrami model is ideally adapted for the prediction and quantification of crystal growth and nucleation under isothermal conditions, numerous studies have also been performed under non-isothermal conditions (Coburn et al., 2018; Kim Piew et al., 1999; Wang and Dou, 2007). This non-isothermal crystallization of semicrystalline polymers closely resembles the industrial melt polymer process.

Fig. 4 presents the relative crystallinity as a function of time for BC/PP composites at fixed BC loading. The BCs are prepared at different temperatures as discussed above. The gradual increase in relative crystallinity as a function of time for these composites can be seen in Fig. 4. Also, sigmoidal curves for all the composites were observed. The added BC type has shown a significant influence on the crystallization kinetics of the PP. From the data, it is deduced that BC addition enhanced the crystallization rate of resulted composites. This is obvious from the fact that the time required to reach a relative crystallinity of 100% became shorter. This effect is more pronounced in the case of composites fabricated with BC prepared at higher temperatures i.e. 500 °C onward. This indicates that the incorporation of BC particles into the PP matrix results in a decreased crystallization time of the PP chains. In other words, the induction time that is required for the overall crystallization process of BC/PP composites has decreased. Moreover, the increased BC concentration has also resulted in a reduced induction time, which suggests the hetero-nucleating action of BC particles. A similar type of effect is reported in the literature with the loading of carbon fillers (Lin et al., 2018; Bhattacharyya et al., 2003; Tian et al., 2017).

Close

Fig. 4

Relative crystallinity with time of neat PP and BC/PP composites, (A) 5%, (B) 10%, (C) 15%, and (D) 20% loading.

Export to PPT

Fig. 5 shows the typical Avrami plots of $ln\left[-ln\left(1-\theta (t)\right)\right]$ versus $ln(t)$ corresponding to the neat PP and BC/PP composites. It can be observed that both neat PP and BC/PP composites are following Avrami’s theory; since very good linearity between $ln\left[-ln\left(1-\theta (t)\right)\right]$ versus $ln(t)$ was obtained with an excellent correlation coefficient of (R ≥ 0.990) (Table 2). The yielded Avrami exponent n and crystallization rate constant k are given in Table 2. The value of n for the neat PP and it’s BC/PP composites was found to be between 2 and 3, indicating a two-dimensional crystal growth with a linear growth rate (Parija and Bhattacharyya, 2017). Surprisingly, the values of n for BC/PP composites prepared with BC300 and BC400 were higher than that of neat PP. However, the composites prepared with BC500, BC600, and BC700 showed lower n values. This variation in crystallization characteristics might be ascribed to the inherent difference between BC properties prepared at different temperatures. As n value is an indication of the nuclei growth dimensions, the values of 1 ≤ n ≤ 2 indicates 1-dimensional rod-like growth, whereas the values of 2 ≤ n ≤ 3 indicate 2-dimensional disc-like growth. Therefore, the relatively high porous structure and surface areas of the BC prepared at high-temperature (such as BC500, BC600, and BC700) points towards a 2-dimensional disc-like growth pattern rather than the rod-like pattern. This behavior could be correlated to the fact that PP chains can well penetrate the BC particle prepared at higher temperatures (BC500, BC600, and BC700) and diffuse through pores in BC/PP composites (Elnour et al., 2019).

Close

Fig. 5

Avrami plot of neat PP and BC/PP composites, (A) 5%, (B) 10%, (C) 15% and (D) 20% loading.

Export to PPT

The Avrami rate constant (k) could be a useful parameter to understand the nucleation and fastening effect of BC particles. The higher k value implies a higher extent of nucleation effect (Parija and Bhattacharyya, 2017). In all the composites, the Avrami rate constant (k) increases as compared to the neat PP, where neat PP showed a k value of $2.93\times {10}^{-5}$ , while the BC/PP composites exhibited a k value in the range between $2.00\times {10}^{-5}$ to $2.00\times {10}^{-4}$ . Besides, it can be observed that the BC prepared at higher temperatures (i.e. BC500, BC600, and BC700) showed the highest k values compared to lower temperature BC when considering the effect of the BC type and its loading percentage. This also justified faster crystallization of the BC/PP composites in comparison with the neat PP. Similarly, another useful parameter for understanding the nucleation and fastening effect of BC particles is the half crystallization time (t_1/2). This also showed a lower value for BC particles prepared at high temperature compared to the lower one.

In general, the addition of the BC particles into the PP matrix leads to an enhanced crystallization process due to their hetero-nucleating effect. This nucleating effect was found to be similar to other carbonaceous fillers reinforced PP composites such as carbon black (Li et al., 2006), single-walled carbon nanotubes (SWNTs) (Bhattacharyya et al., 2003), Multi-walled Carbon Nanotubes (MWCNTs) (Parija and Bhattacharyya, 2017), and carbon fibers (Tian et al., 2017).

3.3 Spherulite morphology

The analysis of spherulitic growth under optical microscopy was carried out at 120 °C. The selection of this temperature was merely based on the assumption that at a lower temperature, the spherulite growth occurs very rapidly. The morphological evolution of the neat PP sample is shown in Fig. 6I A-D, while that of composite made with 15% loading of BC500 is shown in Fig. 6II (A-D), for comparison. Here we have randomly selected the 15BC500/PP sample just for the sake of brevity, as we believe that in other composites the morphology will be just a duplication. From Fig. 6I(A-D), it can be seen, for neat PP, that the morphological evolution during crystallization at 120 °C, started with very small nuclei that are barely seen, shown as very tiny black dots in Fig. 6(IA), which required about 1 min to emerge. As the crystallization time progresses, at 2 min, it can be observed that the spherulites start to form (Fig. 6I(B)), at time 5 min or less, the neat PP spherulites grow continuously and become fully developed and well recognized at 10 min, which is an indication that the crystallization process is finished.

Close

Fig. 6

(I) POM images of neat PP at 120 °C, taken at times (A) < 1 min, (B) < 2 min, (C) 5 min and (D) > 10 min and (II) POM images of 15BC500/PP composites at 120 °C, images taken at (A) 0 min, (B) 2 min, (C) 5 min and (D) 5 min without cross polarizer.

Export to PPT

For the BC/PP composite under study (i.e: 15BC500), It can be seen that at time < 1 min, Fig. 6II (A), nuclei formation is detected, similar to neat PP sample. However, as the crystallization event proceeds further; at a time 2 min or higher (Fig. 6II(B–D) spherulites were observed. These were not similar to the one observed in the case of neat PP sample, rather a vast number of small-sized spherulites were formed. In other words, the number of spherulites developed for the model 15BC500 was significantly higher than the neat PP. This crystallization behavior of the composite could be attributed to the fact that heterogeneous nucleation caused by BC particles prevented homogeneous nucleation of PP, which resulted in the small-sized PP crystallites with fractural (imperfect) morphologies of PP spherulites. Similar behavior was observed for PP composites prepared with MWCNTs (Lin et al., 2018). Overall, this analysis of spherulitic morphology is in well accordance with the conclusions made from the DSC crystallization analysis that the nucleation rate of the BC/PP composites is evidently enhanced when BC is added to the PP matrix, and also the BC particles served as a nucleating agent for the PP matrix, which led to a faster crystallization rate of PP.

3.4 Thermogravimetric analysis (TGA)

Thermal stability of polymeric materials is an important parameter since thermal stability can be a limiting factor in polymer processing as well as its end-use applications (El Achaby et al., 2012). Thermal degradation of PP and BC/PP with different BC weight fractions as well as different BC types (prepared at different temperatures) was determined from the weight loss during heating. The weight loss (TGA) curves are presented in Fig. 7, and derivatives weight loss ((DTG)) are presented in supplementary information (Fig. S2), respectively. From these Figures, the temperature for 5 and 15% weight loss (T_5% and T1_5%) and the maximum decomposition temperature (T_max) were extracted and the results were summarized in Table 3. From Fig. 7, it could be noticed that the thermal degradation of the neat PP and its BC/PP composites occur as a single step process. The maximum decomposition temperature (T_max) for the neat PP was recorded at around 467 °C while the weight loss values of 5% (T_5%) and 15% (T_15%) were recorded at 384 and 422 °C, respectively. Also; it could be seen that all the BC/PP composites have shown a delayed degradation onset temperature compared to the neat PP matrix. This is an indication of the improved thermal stability of the composites due to the addition of the thermally stable BC particles. This improvement in the resistance to thermal degradation of the composite can be attributed to the hindered diffusion of volatile decomposition products within the composites due to the barrier effect of the added BC particles (Elnour et al., 2019). Moreover, this causes mass transfer limitations of produced volatile gases, and it is strongly dependent on the BC-PP interactions. The enhancement of thermal stability can also be attributed to the free radical trapping mechanism (Su et al., 2011). The (C–C) bond in the main chains of PP breaks at a high temperature to form carbon-based radicals, also; in the presence of oxygen O₂, the (C–H) bond can be attacked by O₂ to generate (HO•) and alkoxy free radicals (RO•) and subsequently leads to a series of polymer chain degradations through random scission. However; the addition of BC can delay the degradation of PP chains by trapping these generated free radicals from further propagation. Fig. S3(A) in supplementary information shows the residual weight percentage of BC/PP composites, it could be observed that the BC residue produced after each degradation cycle was proportional to the amount of BC added to the PP matrix. This effect was more pronounced in the composites having a higher amount of BC particles in them. Additionally, Fig. S3(B) presents the maximum degradation temperature (T_max) of BC/PP composites as a function of BC particles loading. It is clear from the figure that T_max increases gradually for all composite with loading, however, the trend of increase varies for the BC type. Furthermore, when comparing different types of BC at a particular loading, it is can be deduced that the thermal stability of BC/PP composites following the order of BC700 > BC600 > BC500 > BC400 > BC300. Thus, from the TGA experiments, it can be concluded that the addition of BC particles to PP enhances the thermal stability of PP by delaying the composites onset decomposition and its maximum temperatures along with producing a high amount of BC residues. Such important characteristics might widen the application spectrum of BC/PP composites; whenever thermal stability is a key factor. For instance, the application of composites where fire retardation property is a key consideration. Moreover, the as-prepared BC/PP composites exhibit good thermal stability in the range where polymer processing is performed. Thus BC can be incorporated into a PP matrix by melt blending without any noticeable degradation.

Close

Fig. 7

TGA curves of BC/PP composites, (A) 5%, (B) 10%, (C) 15% and (D) 20% loading.

Export to PPT