*3.3. TGA Analysis*

The determination of the modification range temperature, the e ffects of blending, and the comparative assessment of the mechanism of bonding relevant to pellet quality are among important instances where thermal analysis addresses specific needs. One of the tools proven to address these needs is a thermogravimetric analyzer (TGA). Thus, knowledge of the major compositional changes that take place as biomass undergoes thermal treatment can facilitate understanding of its behavior in pretreatment processes including in pelleting processes [48]. Figure 3 shows the TGA thermograms of the pure and blended pellet samples with the focus area clearly marked for less ambiguous interpretation.

**Figure 3.** Thermogravimetric analysis (TGA) thermograms of pure and blended samples of NSP and PSP.

Interpretation of the TGA plot in Figure 3 was limited to the focus area marked in the plot. Events beyond the focus area were not considered because, according to some studies [3,5,7,8,41,49,50], the optimum temperature range for pelleting biomass falls between 50 and 150 ◦C. Within this temperature range, flow and combustion characteristics relevant to bonding and pellet quality are usually formed [5,7,8]. However, it is important to bear in mind that thermal degradation of a sample under TGA proceeds differently in terms of the pattern, based on how the reduction in sample weight occurs as the sample undergoes combustion [51]. This means that the weight change of the sample (increase or decrease in weight) as temperature increases depends on a host of factors, including the type of sample, its stability, and analysis conditions [33,35,52]. Having said this, while the thermal behavior of NSP (100%) was controlled by its major organic components (cellulose, hemicellulose, and lignin), that of PSP (100%) was equally influenced by its primary organic macromolecular constituents (amylose and amylopectin), and the combustion behavior of the blend (NSP/PSP 50%/50%) was influenced by a combination of the contents of both materials (Norway spruce and pea starch).

Eye-tracking TGA curve information from Figure 3 shows that the weight change of all three pellet samples was completed at about 760 ◦C. The first noticeable feature, however, was the reduction in weight of the samples as temperature increased (indicated by the arrows), a condition which forms the core of the thermal performances of biomass materials [7,33]. Another observable feature is the differences in the degradation patterns of the curve, particularly around the focus area (focus point). This point represents the temperature range of chemical modification in which water molecules and other small molecular species relevant to bonding are liberated [7]. The release of moisture began at almost the same temperature for all samples (around 45 ◦C), but the temperature at which molecular species began to flow was slightly different between the pure and blended pellet samples, an indication of differences in thermal behavior attributed to compositional disparity. This temperature for NSP (100%) was about 55 ◦C, around 65 ◦C for PSP (100%), and ca. 60 ◦C for NSP/PSP (50%/50%). Shortly after moisture evaporation and the release of molecular species, critical evaluation of the plot showed that the degradation temperature for NSP (100%) relevant to pelleting began at ca. 85 ◦C, while that for the blend (NSP/PSP 50%/50%) started at about 75 ◦C, with that for PSP (100%) beginning at a moderately higher temperature of above 90 ◦C. This typically means that the release of the bonding components of PSP (100%) required more energy than those of NSP (100%) and NSP/PSP (50%/50%), a condition believed to be attributed to higher concentrations of polar functional groups, particularly the C–H group (from band 7 in the FT-IR spectra in Figure 2) in the structure of PSP (100%) with greater bond strength and bond energy as compared to NSP (100%) and NSP/PSP (50%/50%). The bond strength and bond energies of di fferent functional groups, as well as those of various attraction forces, can be found in References [53,54]. Atoms of polar functional groups are held together by a combination of strong dipole–dipole attraction and intermolecular forces that require more energy to be broken, resulting in the thermal behavior of PSP (100%) being di fferent than that of NSP (100%) and NSP/PSP (50%/50%), for which the energy required for bond breaking would be much less. The more concentrated these groups are in a material, the greater is the energy required to break the bonds holding atoms of these groups together, which can also be construed to mean that the stronger a bond is, the higher the energy required is to break the bond [7,50,54]. A review of the literature [53,54] also showed that the bond energies of di fferent functional groups and those of di fferent types of attraction forces vary.

For the quality of the pellets in terms of burning rate and e fficiency of combustion, this analysis (TGA) established that NSP (100%) may be a better quality pellet in comparison to PSP (100%) and NSP/PSP (50%/50%) because of its lower modification temperature, which was construed to mean faster burning velocity that translates into better combustion e fficiency due to greater contact between the pellet sample (NSP 100%) and the heating environment. Biomass combustion e fficiency is a function of good contact between the heating environment and the biomass [55]. According to Gil et al. [56], the burning rate of pure biomass pellets, which is determined by temperature, is usually higher than that of their blends. Furthermore, judging by the FT-IR data of the pellets presented in Figure 2, there were higher percentages of oxygen-containing polar functional groups for PSP (100%). This puts PSP (100%) at a grea<sup>t</sup> disadvantage as a feedstock in thermal energy conversion systems due to poor combustion-related issues that may significantly increase the operating and maintenance costs of the energy systems. This is because oxygen-containing polar functional groups like the –OH groups are hydrophilic in nature and have the ability to retain moisture, thereby increasing the viscosity of starch micelles (starch gelatinization) [57,58]. The issues of starch gelatinization are explained in greater detail in a correlative thermal analysis data presented in the section below. In the case of the blend (NSP/PSP 50%/50%), however, the oxygen-containing polar functional groups may have been compromised or compensated for by the cellulose, hemicellulose, and lignin contents of Norway spruce. In view of these results, therefore, the order of quality of the pure and blended samples of NSP and PSP in terms of burning rate and e fficiency of combustion was as follows: NSP (100%) > NSP/PSP (50%/50%) > PSP (100%).

It is vital to mention that there are no standard methods to determine combustion e fficiency; hence, data from thermal analysis of biomass pellets aimed at burning e fficiency cannot be compared without uncertainties [59].
