3.1. Microwave Torrefaction
Power levels from low to high were used in this work in order to determine the optimal and cost-effective power setting. Treatments with low to medium power usage (400 to 550 W) presented higher energy consumption due to prolonged time to reach desired temperature. At medium power levels (550 to 700 W), the time to reach the desired temperature was drastically reduced and the oat hull sample had a uniform torrefaction color. Under high power level (>750 W), the energy consumption slightly increased with no significant decrease in processing time, but it tended to damage the reactor vessel and components. Therefore, in order to avoid thermal shock and damage to the reactor, a reactor heating time of 8 to 12 min was set as a maximum recommended time to reach microwave torrefaction temperatures of 225 to 285 °C from room temperature (25 °C).
Table 1 shows the effect of microwave power on the process variable, where an increase in power level from 400 to 700 W decreased the total processing time, cost per tonne, and total kilowatt hour (kWh). A power level of 700 W may have been the most appropriate from an economic point of view. However, 650 W power level was set as the most appropriate based on safety of the reactor, processing time, and cost per tonne.
Another relevant result was the findings of constant mass yield and higher heating value (HHV) of the oat hull at different microwave power levels, in contrast to previous studies where it was mentioned that higher power levels would lead to a lower mass and an increase in the HHV. For example, Iroba et al. [
9] reported that municipal solid waste mass yield was mainly dependent on the microwave power level. In another study, Satpathy et al. [
12] mentioned that energy and mass yields decreased with an increase in microwave power and reaction time for wheat and barley straw. Lin [
31] stated that higher microwave power level decreased mass and energy yields and increased HHV of waste straw. Huang et al. [
13] described that the HHV of the rice straw and Pennisetum would increase with higher microwave power levels. Moreover, Wang et al. [
32] reported that high microwave power levels would result in a loss of the torrefied rice husk and sugar cane mass yield.
A trend in previous studies implied that the mass yield decreased with increase in microwave power for various biomass types. However, this is only a simplistic view and a better explanation could be offered by understanding how microwave power influences the biomass temperature level, which leads to an increase in mass loss and heating value. From
Figure 2a, it can be observed that a higher microwave power level would result in a reduced processing time to reach the torrefaction temperature (220 °C for this example), with coefficient of determination (R
2) of 0.99. This suggests that a second-order polynomial regression has a predictive capability on the time required for torrefaction as a function of microwave power operation condition. From
Figure 2b, the effect of increased microwave power produced a faster rise or increase in temperature level, with R
2 of 0.99, suggesting that there is a high positive linear association between variables.
The results showed that mass yield and HHV of torrefied biomass remained constant through different microwave power levels which is dissimilar to previous microwave torrefaction studies reported. An accurate control of the microwave torrefaction temperature by fiber optic temperature sensor would explain this dissimilarity. As mentioned by Basu [
6], reaction temperature is one of the most important factors in torrefaction. For example, the targeted microwave torrefaction was set to 255 °C and 6 min residence time. Once the biomass reached torrefaction temperature, the microwave power had to be manually regulated (decreased and increased) during the residence time, in order to hold the desired torrefaction temperature at 225 °C. Without microwave power regulation during residence time, the biomass temperature would continue to increase leading to a mass yield decrease and HHV increase. Based on temperature profiles of the different microwave power levels (
Figure 3), it was observed that higher power levels required less time to reach torrefaction temperature. When power levels were increased from 400 to 785 W, the time to reach torrefaction temperature decreased from 20 to 6.5 min, while the average heating rate simultaneously increased from 9.8 to 28.6 °C min
−1.
3.2. Physical Properties of the Samples
Table 2 presents the moisture content, particle size, and bulk density of torrefied and untorrefied oat hull, where increased microwave torrefaction temperatures decreased the geometric mean particle diameter (d
gw) and the standard deviation of particle diameter (S
gw). This could be explained by material shrinkage due to removal of moisture and volatiles through thermal treatment. For instance, Tumuluru et al. [
3] attributed the shrinkage to loss of volatiles leading to a reduction in weight, while Bergman and Kiel [
33] explained that biomass shrinkage was caused by a volumetric change due to torrefaction deep drying.
Figure 4 shows how the particle size distribution decreased with the increment in torrefaction level, where higher temperatures produced smaller particle sizes.
The ash content increased with an increasing torrefaction temperature, while moisture content of torrefied samples drastically decreased from 10.4 to its lowest 0.78% wb at high microwave torrefaction temperature. The bulk density did not present any significant changes at light microwave torrefaction temperature (225 °C). However, as the microwave torrefaction temperature increased to moderate (255 °C) and severe (285 °C), the bulk density decreased due to loss of mass specifically from hemicellulose. Rodrigues and Rousset [
34] reported similar results, as they indicated that low torrefaction temperatures (220 °C) did not cause any significant change in biomass density. The authors also determined that increased residence time did not cause any significant changes to bulk density or particle size between microwave torrefaction treatments.
Moisture absorption of microwave torrefied samples was found to be lower than that of untorrefied samples. After full saturation (72 h at 25 °C and 90% RH), the microwave torrefied oat hull samples at temperatures of 225, 255, and 285 °C had respective moisture contents of 15.37, 9.25, and 8.08% wt. vs. 17.96% wt. for the untreated. This difference of moisture absorption capability of oat hull could be explained by a depolymerization of polymers and removal of oxygen groups from the cell wall, replacing hydrophilic for hydrophobic groups resulting from torrefaction [
35]. Peng et al. [
20] attributed the decrease in moisture uptake of torrefied wood pellets to a decreased hydrophilic OH group and increased hydrophobic carbon content.
3.3. Torrefaction Performance
Figure 5 shows an increase in mass loss due to torrefaction temperature. The mass loss generally included part or all of the extractives of hemicellulose and cellulose contents depending on microwave torrefaction level (
Table 3). The maximum mass loss was recorded at 285 °C and 9 min residence time to be 41.3% wt. Moreover, increasing the torrefaction residence time did not significantly influence the mass loss as compared to increasing the temperature level. The average HHV of as-received oat hull was 16.8 MJ kg
−1, and the highest value of 22.70 MJ kg
−1 was obtained by severe microwave torrefaction and 9 min residence time (35.12% higher). The average energy density of light (97%), moderate (87%), and severe (79%) torrefaction gradually decreased by a reduction of the mass yield. Generally, a higher microwave power level contributed to the increase of HHV of torrefied oat hull. However, it was observed that similar to mass loss, HHV was not significantly increased as a result of a longer residence time. Almeida et al. [
36] and Huang et al. [
13] used mass loss and heating values as indicators of torrefaction severity. For the microwave torrefaction of oat hull, a severity factor (
SF) was used to integrate temperature and residence time [
37]. Equation (6) defines the
SF used in this study as:
where
t is the reaction time of microwave torrefaction in min;
TH is the reaction temperature in °C; and
TR is the reference temperature (100 °C).
Figure 5 presents a correlation between heating value and mass loss of oat hull after microwave torrefaction, where a decrease in the mass loss generally increased the HHV. Moreover, a severe decrease in the mass loss tended to reduce the energy yields by up to 58.63%. Mass loss and heating value presented a R
2 of 0.93 and 0.95, respectively, suggesting that the regression equations can fairly predict these variables with a high positive linear association between variables and SF.
As observed in
Table 3, high temperatures resulted in higher liquid and gas yields. During torrefaction, water and acids are produced from hemicellulose and cellulose degradation [
6,
38]. As described by Basu [
6], large structures of hemicellulose, cellulose, and lignin will degrade into smaller ones and can be classified as gas (CH
4, CO, H, CO
2, and steam), liquid (water, tar, phenols, acids, and carbonyls), and mass (char).
3.4. Energy Consumption during Grinding
For as-received untreated oat hull, the grinding process was shown to consume a lot of energy (121.54 KJ kg
−1). However, by reducing the sample moisture content, the energy consumption decreased by 44% (68.54 KJ kg
−1). Through the use of light (225 °C), moderate (255 °C), and severe (285 °C) torrefaction, the grinding energy reduced by 64, 80, and 86%, respectively, with corresponding energy consumptions of 43.70, 23.88, and 17.30 KJ kg
−1, respectively. The decrease in grinding energy of oat hull after microwave torrefaction was comparable with results obtained in other similar studies for biomass [
39,
40,
41]. For example, according to Repellin et al. [
41], dehydration decreased grinding energy by the breakdown of the fibrous structure of wood. Moreover, microwave torrefaction tended to degrade the lignocellulosic structure of the oat hull. Therefore, the higher torrefaction treatment level resulted in a greater lignocellulosic structure degradation with an improved grindability and a reduced energy consumption.
3.6. Organic Composition of Samples
This work also investigated the effect of microwave torrefaction level on lignocellulosic structure of oat hull. As observed in
Table 4, there was a significant degradation of hemicellulose by microwave torrefaction treatment from an initial content of 37.54 to 0.72% db through severe microwave torrefaction. Moreover, cellulose, and lignin contents increased, except for the severe microwave torrefaction treatment, where cellulose also decreased due to the higher thermal treatment. According to Nhuchhen et al. [
8], hemicellulose presents a lower degree of polymerization when compared to cellulose, resulting in higher thermal degradation, with severe mass yield reduction. Lignin increased from an initial content of 11.46 to 45.65% db through severe microwave torrefaction, suggesting that there was a much lower degradation when compared to hemicellulose and cellulose. This reduced degradation could be explained due to its large sterically hindered complex structure and lack of hydroxyl groups [
42]. The heating values increased with microwave torrefaction temperature level which can be explained by the removal of hemicelluloses. For example, hemicelluloses presented a much lower heating value (13.6 MJ kg
−1), when compared to lignin (27 MJ kg
−1) which was reported to increase [
40,
43]. According to Basu [
6] cellulose is composed of condensable vapors, while hemicellulose produces more non-condensable gases, less tar, and acetic acid.
Table 5 shows that microwave torrefaction resulted in significant elemental composition changes, making oat hull more suitable for fuel applications since percentage of carbon increased and oxygen contents reduced. For severe microwave torrefaction, the carbon content increased by 30.16%, while hydrogen and oxygen contents decreased by 8.10, and 19.40%, respectively thus lowering the C-H and C-O ratios. According to Phanphanich and Mani [
40] reduction of C-H and C-O bonds enhances the properties of a fuel by increasing heating values and reducing smoke and water vapors during combustion.
3.7. By-Products of Torrefaction
Component degradation of cellulose, hemicellulose, and lignin by pyrolysis of hardwood constitutes liquid formation, also referred as liquid smoke [
6,
23]. Commercial liquid smokes are widely used in the food industry to provide flavor, color, texture, and sometimes extend the product shelf life [
23]. Other advantages of liquid smoke over traditional direct smoke contact are higher homogeneity, reduced environmental pollution, better storability, lower processing time, and superior control of PAH groups [
44].
The main product of torrefaction is clearly the bio-char [
6]. However, another important factor that many previous studies did not take into account was the utilization of torrefaction by-products such as the liquid and gas yields [
9,
31,
40,
45,
46].
Figure 7 shows the liquid yield collected from different torrefaction temperatures. As observed, the higher the microwave torrefaction level, the darker the color of liquid yield (smoke) product. As microwave torrefaction temperature increased from 255 to 285 °C, the degree of cellulose polymerization decreased from 38.92 to 33.52% (
Table 4). According to Rama
Krishnan and Moeller [
23] and Budaraga et al. [
47], degradation of cellulose and hemicellulose leads to formation of carbonyl groups, which are responsible for the darker brown color of liquid smoke compositions. Therefore, a higher carbonyl concentration would lead to a darker color.
Table 6 presents the three microwave torrefaction levels and commercial liquid smoke properties. The results show that high microwave torrefaction levels decrease pH, while increasing total acids (acetic acid) and tar contents. It was reported that acetic acids were formed from thermal decomposition of cellulose and hemicellulose but mainly from degradation of the acetyl groups of hemicelluloses during torrefaction [
48]. The heat during torrefaction degraded hemicellulose almost completely when light (27.34%) to moderate (0.81%) microwave torrefaction was applied, explaining the drastic increase in total acids from 2.3% to 15% for the corresponding treatments. Furthermore, cellulose degradation produces water vapor, CO
2, CO, and condensable gases, while thermal degradation of lignin produces around 20% liquid yield and 15% tar [
6].
Liquid smoke is widely used in the food industry, having an average price of USD 10 per L and a minimum selling price of USD 2 per L [
14]. However, price may vary depending on the type of extraction, raw material, and flavour profile (e.g., sweet or woody flavor). These new types of liquid smoke could be of great value to the food industry, which despite having a low quality, can be used in low-end food products such as hot dogs and sausages which have a market size of USD 7.8 billion in 2019 [
49].