*3.4. DSC Analysis*

According to Back [39], su fficient bonding areas are created when wood components are plasticized above their glass transition temperature (Tg). In addition, since combustion is the final destination of most biomass pellets, an ultimate method for testing the quality of biomass pellets is thermal analysis, because good-quality pellets are expected to burn easily and generate less complex degradation patterns [60,61]. Figure 4 presents the DSC plot of pure and blended samples of NSP and PSP.

**Figure 4.** Differential scanning calorimetry (DSC) thermograms of pure and blended samples of NSP and PSP.

A step change in the heat flow of the DSC thermogram presented in Figure 4 constituted the first observable feature of the plot and represented the glass transition (Tg) range temperature for all samples. Tg conditions facilitate the deformation of particles, reduce viscosity, and increase the movement of natural binding components [41]. According to the DSC plot in Figure 4, these temperature conditions fell between 45 and 85 ◦C for NSP (100%), between 45 and 80 ◦C for PSP (100%), and between 45 and 75 ◦C for NSP/PSP (50%/50%). Moisture is one of the most useful agents that can act as a lubricant during pelleting of biomass materials because it strengthens and promotes bonding through a combination of attraction forces by increasing the particle contact area [3,41,62]. Just like the previous thermal analysis data, transition at 45 ◦C in the DSC plot typically indicated the moisture evaporation from all samples. The Tg range of the samples under DSC analysis were consistent with the previous modification temperature range presented in Section 3.3, except that, in this analysis (DSC analysis), the maximum Tg for PSP (100%) occurred at a much lower temperature (80 ◦C) than its previous maximum modification temperature of around 90 ◦C in the previous section. The complexity of the thermal behavior of starch and differing analysis conditions may be responsible for the difference in the maximum modification and transition temperatures between this thermal analysis and the previous for 100% PSP. The behavior of starch under thermal conditions is much more complex than other thermoplastic materials like wood because the physicochemical changes that occur during combustion of starch or products containing starch involve the motion of water, gelatinization, glass transition, and melting, as well as the change of crystal structure and molecular degradation, which are thermal events that are dependent on material moisture content, and the water contained in starch is often not stable during combustion [63].

The DSC plot in Figure 4 also indicated that 100% NSP displayed two endothermic peaks, with its first in the Tg range, and its second way beyond the Tg range (between 350 and 370 ◦C). However, further interpretation of the DSC plot beyond the Tg range for all samples was less of a focus because the study aimed to establish differences in the mechanism of bonding between the pellet samples relevant to quality, and the Tg is considered the temperature range at which natural binders relevant to bonding are activated [6,41]. Therefore, inter-diffusion of the natural binders in 100% NSP increased between adjacent particles to facilitate bonding and the formation of solid bridges at its Tg. The polymeric constituents of the blend (NSP/PSP 50%/50%) were released faster than those of the pure pellet samples (NSP 100% and PSP 100%) because of lower Tg and the presence of water vapor created by moisture. Differences in Tg between the pure and blended pellet samples can be traced back to

their content of polar functional groups (Figure 2) including the oxygen-containing groups whose concentrations were higher with stronger bond energies in PSP (100%) than NSP (100%) and the blend (NSP/PSP 50%/50%).

For the quality of the pellets in terms of combustion e fficiency, which was considered in this analysis (DSC analysis) from the viewpoint of burning and heat flow rates, NSP (100%) seemed to dominate because of the two endothermic peaks formed in its DSC plot (Figure 4). The formation of the two endothermic peaks and the modification temperature for NSP (100%) from the previous thermal analysis data (55 ◦C as compared to 60 ◦C and 65 ◦C for NSP/PSP 50%/50% and PSP 100%, respectively) were evidence of increased burning and heat flow rates that were facilitated by auto-oxidation reactions because of the presence of oxygen-containing functional groups in the structure of NSP (100%). This means that the burning and heat production rates of NSP (100%) were significantly higher in comparison to those of PSP (100%) and NSP/PSP (50%/50%). Although, PSP (100%) may have greater concentrations of oxygen-containing polar functional groups, there is the possibility of the formation of hydrogen bridges at elevated temperatures because of increased probabilities of the promotion of retrogradation reactions that may result in the formation of thickened paste, which reduces combustion efficiency. However, this thickened paste can facilitate particle-to-particle bonding to form pellets with increased strength due to the presence of rigid irreversible gel [64]. The concentrations of the oxygen-containing polar functional groups in the blend were compromised by mixing the two materials at a 50/50 ratio. However, because of the balance between the ratio of the two materials in the blend (NSP/PSP 50%/50%), and perhaps the higher percentage of carbon in NSP (100%) in comparison to PSP (100%), combustion issues would be minimized when the blended pellet is used as feedstock in thermal conversion systems. Therefore, the order of quality of the pellets in terms of combustion efficiency was as follows: NSP (100%) > NSP/PSP (50%/50%) > PSP (100%).
