Microwave Pyrolysis of Biomass: The Influence of Surface Area and Structure of a Layer

Abstract

The paper presents the results of experimental research into lab-scale microwave pyrolysis of wood biomass. The influence of the surface area and the structure of the biomass layer on the characteristics of pyrolysis during microwave heating are discussed. We have established that the biomass layer structure and surface area have a significant effect on the yield of pyrolysis gas. The approach of creating artificial deformation of the biomass layer was tested. The elements of artificial porosity made it possible to increase the CO yield by 18% and 32% compared to the pyrolysis of a biomass layer with artificial channels and a uniform layer, respectively. The concentration of H₂ was 33% higher compared to the layer without artificial pores and 3% lower compared to artificial channels. The yield of CO₂ increased by 25%, and the yield of CH₄ doubled. The experiments showed that the distribution of biomass on a half of the bottom of the crucible and the additional porosity of the biomass layer surface effectively increase the yield of the pyrolysis gas components. Recommendations for increasing the efficiency of microwave pyrolysis of biomass were formulated.

1. Introduction

1.1. The Prospects for the Use of Biomass as an Energy Source

The rapid growth of industry and technology requires more and more energy [1]. According to forecasts for the development of the world economy [2,3], annual energy consumption will be 1.000–1.500 EJ by 2050. Currently, at least 85% of the world’s energy consumption is provided by traditional fuels [4]. The limited reserves of non-renewable resources [5] as well as climate change [6], caused by greenhouse gas emissions, motivate the transition to bio-renewable raw materials [7]. In this regard, renewable energy sources such as solar energy, wind energy, tides and biomass are of growing interest all over the world [1]. Biomass has a wider application and, in addition to generating energy, can also be used for the production of chemicals and materials [8].

One of the promising methods of biomass recovery is thermochemical conversion [9,10,11,12,13,14,15,16,17,18,19,20], which includes pyrolysis [10,11,12,13], combustion [14,15], gasification [16,18] and advanced combined methods such as plasma gasification [19,20]. Pyrolysis has received much attention over the past decades, both on a laboratory and industrial scale [6,21,22]. Microwave pyrolysis has been considered a promising technology for waste recovery [23,24] that processes biomass, various agricultural wastes [6,21,25], sewage sludge [26,27], municipal solid waste and plastic [28,29], algae [30] and car tires [31].

Microwave heating has the following advantages compared to traditional pyrolysis [23]: uniform non-contact heating of raw materials; high energy efficiency and output; inexpensive and simple equipment; more flexible process control; high-quality final product (gas, bio-oil or semi-coke). The results of studies [23,24,25,26,27,28,29,30,31,32,33,34] substantiate these advantages. However, additional studies are still required to expand the current understanding of the characteristics of microwave pyrolysis.

1.2. Microwave Pyrolysis of Biomass and Waste: A Review of Modern Researches

The following factors and processes are often analyzed when studying microwave and thermochemical pyrolysis: fuel characteristics (particle size, sample distribution, humidity and density of a layer) [13,26,35,36,37,38]; the influence of metal catalysts (Al, Ni, Fe, Cu, etc.) [33,37,38] and absorbers [38,39]; heating characteristics (power and temperature) [37,38,40]; carrier gas properties (air, CO₂, H₂O, reaction gas) [26,40]; fuel pretreatment methods [13]. These factors and processes are considered in the context of their influence on the pyrolysis efficiency, yield and distribution of solid, liquid and gaseous end products.

When choosing and designing microwave reactors for research, it is necessary to take into account several important criteria that also affect the characteristics of pyrolysis [41,42,43]: (1) operating mode: periodic, semi-continuous or continuous; (2) type of reactor: fixed bed, fluidized bed or continuous flow; (3) orientation of the magnetron: side, top or bottom; (4) microwave frequency: 916 MHz or 2.45 GHz; (5) coating material: quartz or ceramic; (6) microwave resonator mode: single- or multi-mode; (7) the scale of the reactor: laboratory, pilot-scale or industrial.

The studies [26,32,36] present the results of numerical simulation of microwave pyrolysis. The data obtained [26,32,36] make it possible to evaluate various heating scenarios (for example, changing the structure and size of the object, observing the moving object, varying the characteristics of materials, etc.) and to compare the results with experimental data. In [26], Lin et al. studied spiral continuous microwave pyrolysis of sewage sludge. In comparison with periodic microwave pyrolysis, this technology provides instantaneous heating of the sludge and overcomes the challenge of producing large amounts of tar. In the experiment, the microwave power varied in the range of 1800–4200 W, the pyrolysis temperature varied from 300 to 600 °C, and fuel samples (moisture content 0–60%) were continuously fed into the reactor at a feed rate of 4 rpm. Nitrogen at a flow rate of 100 l/h was used to blow down the installation, and carbon dioxide was the medium for pyrolysis. The key factors for the analysis were pyrolysis temperature, fuel sample moisture content and CO₂ concentration. They were considered in terms of their influence on biogas yield, and H₂ and CO concentrations. It was found that an increase in temperature (400–600 °C) leads to an increase in the biogas yield (from 17.23 wt% to 53.09 wt%). At the same time, the combined use of H₂O and CO₂ had a positive effect on the biogas yield (40.65–55.67 wt%), which is significantly higher than the biogas yield obtained by conventional pyrolysis (18–20 wt%) or periodic microwave pyrolysis (25–42 wt%) under identical temperature conditions.

With an increase in humidity from 0 to 60 wt%, the biogas yield gradually decreased from 55.54 wt% to 40.72 wt%, which is associated with a lower quality of organic matter of the sewage sludge. However, water molecules reacted with light hydrocarbon compounds through steam reforming, increasing H₂ concentration (from 5.34 vol% to 31.51 vol%). CO concentration gradually increased with increasing temperature and CO₂ concentration. At a pyrolysis temperature of 600 °C and CO₂ concentration of more than 80 vol%, the CO concentration reached 30.92–37.47 vol%.

Various types and ratios of purge gases as well as catalysts have been used in many studies aimed at increasing the yield and quality of the generated gas. For example, in [6], the thermochemical catalytic pyrolysis of rice straw (particles less than 0.1 μm, the supplied fuel mass of 1.5 ± 0.02 g) was studied using various pyrolysis plants (one- and two-stage). CO/SiO₂ was used to improve the reaction kinetics. In [6], Jung et al. focused on the use of different N₂/CO₂ purge gas ratios (0/100, 25/75, 50/50, 75/25, and 100/0). They found that CO₂ accelerates homogeneous reactions in the gas phase between volatile organic compounds formed as a result of the rice straw pyrolysis [6]. The yield of CO started at temperatures above 480 °C in the presence of CO₂, while CO formation was not observed in the N₂ environment. In addition, the CO/SiO₂ catalyst additionally accelerated the kinetics of CO formation by a factor of 3 and H₂ by a factor of 6 relative to noncatalytic pyrolysis at 500 °C.

The synergistic effect of the CO/SiO₂ and CO₂ catalyst led to a significant increase in the concentrations of H₂ and CO at temperatures above 480 °C. Similar conclusions were made in [37]. Li et al. [37] analyzed the effect of microwave heating, reaction temperature, and types of catalysts (Ni/Fe), and also evaluated the removal efficiency of tar and nitrogen oxides during pyrolysis of plant residues in combination with catalytic reforming. The highest gas yield (77.47 wt%) and the lowest resin yield (7.40 wt%) were obtained at a temperature of 900 °C in the microwave mode. An increase in the moisture content of plant residues led to a significant decrease in the yield of a solid product and a simultaneous increase in gas production (the maximum gas yield was 82.60 wt% at a moisture content of 40%). With the combined effect of microwave radiation and a nickel-based catalyst, a maximum reduction of 74% was achieved in resin yield, and gas production also increased by 143%.

Different operation modes of the microwave reactor are studied to increase the yield of pyrolysis gas. In [34], the continuous microwave pyrolysis of cow manure with an initial mass of 50 g was performed using the CO₂ reforming technology. Luo et al. [34] studied the effect of CO₂ on the yield and quality of the gas produced, the pyrolysis mechanism and the energy balance of the process. It was found that continuous microwave pyrolysis combined with CO₂ reforming technology has advantages over other pyrolysis methods (batch microwave pyrolysis and traditional electric heating). The use of CO₂ increases the efficiency of biomass conversion by enhancing reactions with pyrolysis products such as tar, CH₄ and hydrocarbons. The resulting pyrolysis gas has a reduced content of H₂S, which was confirmed by many studies [44,45,46]. Luo et al. [34] compared the operating modes of a microwave reactor of periodic and continuous modes. They found that continuous microwave pyrolysis can increase the yield and calorific value of biofuels [34]. The maximum yield of biofuel during continuous microwave pyrolysis was 72.16%, and the maximum net calorific value of biogas and biooil was 19.50 MJ/Nm³ and 25.58 MJ/kg, respectively. Continuous microwave pyrolysis has low power consumption and high energy efficiency, reaching a maximum of 18.47%, which is 8.5% higher than for a batch microwave pyrolysis.

The authors of [13,47,48] discussed the influence of the biomass structure on the conditions and characteristics of microwave pyrolysis and gasification. Huang et al. [13] investigated the influence of the particle size and pretreatment of corn straw as well as the catalysis of Al₂O₃ on the efficiency of microwave pyrolysis, distribution and composition of end products. It was shown that the heating efficiency and the yield of the gaseous product increase as the particle size decreases. Huang et al. [13] also found that without any pretreatment, the effect of catalysis is insignificant. An increase in the microwave pyrolysis temperature was registered when the biomass was pretreated. The addition of Al₂O₃ increased the gas yield by 4–9 wt%. In comparison, the gas yield during catalytic microwave pyrolysis was about 40–60 wt% [35] without Al₂O₃ and 70–72 wt% with added Al₂O₃ [13].

The structure of the biomass and the dimensions of the feed sample are also important in the design and operation of microwave pyrolysis plants. In [47], microwave pyrolysis of pressed blocks of wheat and corn straw was studied (density of 80 kg/m³, block size of 1000 × 600 × 600 mm). The large size of the straw block contributed to the uneven distribution of temperature inside it. Zhao et al. [47] found that the content of H₂ reached the highest value of 35 vol%, and the content of synthesis gas (H₂ and CO) exceeded 50 vol%.

Research [48] focused on the microwave pyrolysis of oil palm biomass (fibers and shells) using char as a microwave absorber. Shells were used in the form of large particles (without grinding) with sizes ranging from 0.001 to 0.1 µm, while the fibers were crushed to a finer size (300–600 µm). For each experiment, the ratio of biomass and absorber was varied (1:0.25, 1:0.5, and 1:1). It was found that the addition of a microwave absorber to the biomass did not only increase the temperature, but also intensified the pyrolysis due to the formation of vapors. The maximum bed temperature at shell/absorber ratios of 1:0.25 and 1:0.5 was about 200 and 237 °C, respectively. The maximum gas yield was 30 wt% for shell/absorber ratios of 1:0.5 and 1:1. Salema and Ani [48] showed that microwave pyrolysis systems can use coarse particles, which reduces the cost of grinding and removing moisture. Salema and Afzal [49] obtained the heating characteristics of a biomass layer (fruit tree branches in the form of a uniform layer and granules) using numerical simulation of the fuel sample heating mode in a multimode microwave system. It has been established that the height of the biomass layer and its shape (layer or granules) affect the temperature profile. This indicates that there is an optimal loading height at which the maximum material temperature can be reached for a particular furnace size. Studies [49,50] confirm that the distribution of biomass has a direct effect on the yield of gases during microwave pyrolysis.

1.3. Motivation, Novelty, and the Main Objectives of the Study

A review of scientific achievements in the field of microwave pyrolysis illustrates its active development. At the same time, the pyrolysis process itself and its characteristics continue to require deeper insight in order to develop optimal technologies that maximize the production of useful end products. The purpose of this study was to determine the effect of the surface area and structure of the biomass layer, its distribution and dispersion on the characteristics of microwave pyrolysis. The following tasks were solved: (1) development of an experimental procedure and design of an experimental setup; (2) determination of the nomenclature and optimal concentrations of promising biofuel components; (3) formulation of generalizing criterion expressions and comparative analysis of biomass pyrolysis efficiency with varying surface area.

2. Materials

Wood biomass is a promising source of energy [1,8], as it is a fairly environmentally friendly material and a renewable energy source. Wood is considered to be CO₂-neutral [9,51] since when it is burned, the same amount of carbon dioxide is released that was required for the growth of a tree. In this work, sawdust is used as a typical waste of wood origin. Studies [52,53] show that sawdust has a high dielectric constant and, therefore, absorbs microwaves.

Table 1. Characteristics of sawdust.

Ultimate Analysis (%)					Proximate Analysis
C	H	O	N	S	Moisture, %	Volatiles, %	Fixed Carbon, %	Ash, %	Calorific Value, MJ/kg
54.3	5.2	40.0	0.4	–	3.5	80.1	15.1	1.1	18.67

presents the properties of the material used.

3. Experimental Setup and Procedures

3.1. Experimental Setup

A setup for the present study was designed on the basis of the review of experimental installations. The setup has the following advantages: simplicity of design and operation; flexible experimental procedure; real-time monitoring of gas concentrations; no need for additional equipment. The wavelength at a frequency of 2.45 GHz is 12.2 cm.

One of the main components of the microwave heating setup is a magnetron.

Table 2. Characteristics of magnetrons [54].

Output Power, kW	Frequency, MHz	Anode Voltage, kV	Efficiency, %	Cooling Method	Magnetic System
1.0	2375	3.6–4	62	Air	Permanent magnet
2.5	2375	3.6–4	60	Water	Electromagnet
5.0	2375	5.5	56–60	Water	Electromagnet
10.0	429	10	65	Water	Electromagnet
25.0	910	10	75–80	Water	Electromagnet

presents the characteristics of typical magnetrons in order to understand the possibilities of choosing setup elements for specific scientific and technological tasks.

Figure 1 presents a diagram of the experimental setup used in the present study. The main elements of the setup are a microwave oven for heating biomass and a gas analyzer (Test-1, Boner VT company, Novosibirsk, Russia) for taking a gas sample and determining the concentrations of the gas mixture components.

A ceramic crucible with a biomass sample was placed in a microwave oven. The oven was heated due to the energy produced by the magnetron. The oven control unit made it possible to adjust the heater power in the range from 200 to 800 W. In the study, we did not vary the power of the installation. The power was 800 W throughout the experiment. After the fuel was introduced into the furnace, it was hermetically sealed with a door into which a gas sampling hose was built. The gas analyzer used in the experiments is a laboratory system designed for long-term continuous measurement of the concentrations of gas mixture components formed during combustion and gasification of fuels. The device includes a standard set of elements: a modular probe, a condensate collector, a filtration system, a computing unit, a capillary and measuring electrochemical sensors (O₂, CO, SO₂, NO_x, NO₂, H₂S and HCl). Additionally, the gas analyzer was equipped with CO₂ and CH₄ optical sensors and a H₂ polarographic sensor. The characteristics of the measuring channels are presented in

Table 3. Specifications of the gas analyzer.

Component	Measurement Range	Accuracy	Response Time
O2	0–25%	±0.2 vol% (absolute)	≤15 s
H2	0–5%	±0.2 vol% (absolute)	≤35 s
CO2	0–30%	±2% (basic percentage error)	≤25 s
CH4	0–30%	±5% (of the indication)	≤25 s
CO	0–40000 ppm	±5% (of the indication)	≤35 s
NO	0–1000 ppm	±5% (of the indication)	≤35 s
NO2	0–500 ppm	±7% (of the indication)	≤45 s
SO2	0–1000 ppm	±5% (of the indication)	≤45 s
H2S	0–500 ppm	±5% (of the indication)	≤45 s
HCl	0–2000 ppm	±5% (of the indication)	≤45 s

.

The Test-1 gas analyzer was connected to a computer via an RS-232 interface. The software of the gas analyzer transmitted in real time the measured values of gas concentrations, and also archived, integrated and exported the resulting data in standard text and graphic formats. The study presents the results of direct measurement of the concentrations of individual components of the gas mixture. In the experiments, the gas mixture entering the gas analyzer was fed to several sensors at once. All sensors were only sensitive to certain components of the gas mixture. The analytical measurement range for each sensor was set by the manufacturer. During the experiments, the concentrations of individual gases did not exceed the measurement range of the sensors. Since the sensors were calibrated and metrologically verified, no further confirmation of the analytical range was required.

3.2. Experimental Technique for Measuring the Composition of Gases

The experimental procedure included the following stages:

A fuel sample was formed, which was weighed on an OHAUS Adventurer AX 324 analytical balance with an accuracy of 10⁻⁵ g.

The mass of the fuel sample in each experiment varied from 10 to 20 g, based on the goals of the experiment. Before the experiments, the sawdust was dried for 5 days under normal conditions, then soaked in a container with 200 mL of water. For the experiments with varying particle dispersion, the sawdust was preliminarily ground using a Pulverisette 14 rotary mill and sieved to separate the 140, 250 and 2000 µm fractions.

The fuel sample was placed in a ceramic heat-resistant crucible, which was used as a substrate. The wall thickness of the crucible was 0.5 cm, the height of the crucible was 9 cm and the diameter was 9.5 cm. Depending on the variable parameter (surface area, layer structure), the distribution of the fuel sample on the crucible bottom changed (Figure 1a). In other experiments, the fuel was located along the edge of the crucible from the side of the magnetron. The crucible was placed in a microwave oven, after which its door was hermetically closed (Figure 1b).

Oxygen was present in the microwave chamber initially before the start of the magnetron operation because the sample was loaded into the reactor through the door. Air entered the reactor through the same door due to natural convection. Forced air flow was not created. An inert atmosphere was not created in the reactor in the present study to match the most widely used air-type reactors.

Then, the microwave heating was turned on. After the concentrations of CO began to decrease, the heating was turned off. The release of pyrolysis gas took place when heat was no longer supplied from an external source. During the experiment, the probe sampled gaseous products. Then they fell into the moisture separator, where the liquid was separated. The purified gases were then sent to the sensors. Real-time parameters were displayed on the PC monitor. During the experiment, atmospheric air did not enter the oven. The duration of each experiment varied from 30 to 60 min.

After the experiment, the solid residue was weighed. Before starting the next measurement, the reactor and all gas channels were purged with compressed air. Within one series, from 2 to 3 experiments were carried out under identical conditions.

Next, data processing was carried out. Experimental results were averaged and random gross errors were eliminated. The results of experimental measurements were subjected to primary processing in order to eliminate random gross errors. A gross error is a result (time-averaged value of the gas concentration) that differs significantly from the expected one. Since the results of these experiments obeyed a normal distribution, it is advisable to use the 3-sigma rule to eliminate gross errors. However, this rule is reliable for a large number of measurements (over 20). Therefore, we relied on a modified approach (Chauvenet’s criterion). It is based on the calculation of the difference between the average and the test result of the measurement. This difference is then compared with (1.6 ÷ 2.0)·σ, where σ is the standard deviation. The numerical coefficient depends on the number of measurements in the experimental set. If the difference exceeds (1.6 ÷ 2.0)·σ, then the result can reasonably be considered a gross error and excluded from the sample. As a rule, there were no gross errors in the measurement results.

The average concentrations were calculated using the trapezoidal method, as in [55]. The area under the emission trend curve was approximated by rectangular trapezoids (the time step was 1 s). Next, the area of each trapezoid was determined. Then, we calculated the ratio of the sum of the areas of all trapezoids to the time during which the release of the gas component was recorded. After averaging the results of the experiments, the confidence intervals of the corresponding random errors were determined and gross errors were excluded. An example of calculating the average value of CO concentrations is shown in Figure 2. The release of the component started at 1 min of the experiment (Figure 2). The yield time of the component was about 1120 s. The area of the trapezium was 1719.82. By dividing the area by the emission time, an average value of 1.53 was calculated.

4. Results and Discussion

4.1. Distribution of Biomass on the Crucible Bottom

The influence of the distribution of biomass (sample mass of 15 g) on the crucible bottom on the pyrolysis gas yield characteristics was studied. Figure 3 illustrates the options for the distribution of biomass on the crucible bottom.

Figure 4 shows the trends in the concentrations of the gas mixture components released during the microwave pyrolysis of sawdust for different sample distributions in the crucible.

The main components of synthesis gas are H₂, CO, CO₂, CH₄, NO, H₂S. An intensive increase in the concentrations of the pyrolysis gas components began after 400 s (on average for three options for the biomass distribution in the crucible). This time period corresponded to the heating and drying of the biomass. The time of active release of gas components with the distribution of the biomass along the walls of the crucible and uniform distribution over the entire crucible bottom was 700 s and 800 s, respectively. When the biomass was distributed on half of the crucible bottom, this time was about 1500 s. Further, the release of volatile substances was intensified. Figure 4a shows that the concentrations of CO, H₂ and CH₄ increase, and then there is a slight decrease and again an increase (two peaks are recorded). This suggests that each fragment burns out separately from the others located separately along the edges of the crucible. The maximum values of CO, H₂ and CH₄ were 28.5, 3.9 and 10.5% (the first peak), 25, 3.7 and 6% (the second peak). For CO₂, there was a gradual increase and then a decrease in concentrations (one peak). The maximum CO₂ amounted to 18.7%. When the biomass was distributed on the crucible bottom in a uniform layer (Figure 4b), one peak of CO and CO₂ concentrations was observed (39% and 19% at 560 s and 700 s, respectively). CH₄ and H₂ still had two concentration peaks: 9.5% and 3.8% at 470 s and 460 s, respectively (first peak), and 8.3% and 3% at 670 s and 850 s, respectively (second peak). The nature of the trends presented in Figure 4c is generally similar to a gradual increase and decrease in gas concentrations with a uniform distribution of biomass over the entire surface of the crucible bottom (Figure 4b). In both cases, the biomass is distributed uniformly over the surface of the crucible bottom; only the area occupied by the fuel sample changes.

Figure 5 shows the average and maximum values of the concentrations of the gas mixture components.

Figure 5 illustrates that CO, CH₄ and NO trends have a similar behavior. The maximum values of average concentrations are achieved when the biomass is distributed on a half of the crucible bottom (Figure 5a). H₂S and SO₂ concentrations are maximum when the biomass is evenly distributed over the entire crucible bottom and minimum when the sample is located on a half of the crucible bottom. The concentrations of CO₂ and H₂ changed insignificantly with varying distribution of the biomass in the crucible. The difference was no more than 1%. Similar conclusions are valid for the maximum values of the concentrations of the gas components (Figure 5b). When the biomass was distributed on half of the crucible bottom, the average CO values were almost 2 times higher compared to the distribution along the crucible edges and 1.5 times higher compared to the distribution of the sample over the entire crucible bottom (Figure 5a). This result was caused by a different contact area of the biomass with the crucible surface. When the sample was located along the edges as separate fragments, the biomass was heated more intensively, receiving thermal energy from microwave radiation and the walls of the crucible. The fuel was pyrolyzed unevenly due to the lack of heat exchange between biomass particles. The minimum concentrations of CO, CO₂, NO and CH₄ were 8.6, 6.31, 2.77 and 1.8%, respectively. When the biomass was located over the entire crucible bottom, the maximum values of CO₂ and H₂ amounted to 18.74% and 3.8%, respectively (Figure 5b). With this distribution of the sample, it received thermal energy from microwave radiation and the walls of the crucible, and heat was better transferred between the biomass particles.

The analysis of the appearance of the solid residue (Figure 6) showed that the distribution of fuel on the crucible bottom and relative to the magnetron affects the uniformity of heating and the completeness of thermal decomposition. The distribution of fuel on a half the crucible bottom provided a more complete conversion. In this case, the mass of the solid residue relative to the initial mass of the sample was minimal (3.4%). When the fuel was distributed along the edges and over the entire crucible bottom, this indicator was 4.2% and 4.5%, respectively.

When the biomass is distributed along the crucible walls, heating is uneven since heat transfer between the particles is difficult. The areas located on the side of the magnetron begin to warm up first, and then the rest does. In the case of a uniform distribution over the surface, heat and mass transfer also proceeds more uniformly.

Table 4. The yield of the main gases.

Experiment	Gas	Average Concentrations (%)	Average Concentrations (ppm)	Average Concentrations (mg/m3)	Total Amount of Pyrolysis Gas (mg/m3)
Biomass is distributed along the edges of the crucible	CO	8.6	86,000	100,138.6	156,616,770
	CO2	6.31	63,100	115,443.82	146,382,763
	CH4	1.8	18,000	12,002.36	9,061,781
	H2	1.13	11,300	946.93	1,480,998
	H2S	0.0056	56.16	79.56	66,830
	SO2	0.00293	29.3	78.03	41,746
	NO	0.000277	2.77	3.46	3017
Biomass is distributed over the entire surface of the crucible bottom	CO	11.3	113,000	131,577	240,785,910
	CO2	7.02	70,200	128,433	160,412,817
	CH4	2.44	24,400	16,269.87	13,129,785
	H2	1.16	11,600	971.97	1,784,536.92
	H2S	0.0067659	67.66	95.86	83,494.06
	SO2	0.003446	34.46	91.77	53,410.14
	NO	0.00058	5.8	7.24	8514.24
Biomass is distributed over the surface of a half of the crucible bottom	CO	15.5	155,000	180,482.36	465,102,114
	CO2	7.01	70,100	128,250.58	138,382,375
	CH4	2.88	28,800	19,203.78	45,319,080
	H2	0.88	8800	737.36	22,739,200
	H2S	0.003759	37.59	53.26	50,104.47
	SO2	0.00165	16.5	43.94	18,298.5
	NO	0.00123	12.3	15.34	31,783.2

shows the gas flow rates obtained during the experiments. Calculations were made for normal conditions of 20 °C and 1 atm. The average values of the concentrations of the main gases were converted to mg/m³, after which the obtained values were multiplied by the time of gas yield.

The obtained data were compared with the results of numerical and experimental studies presented in [49,56,57]. It has been shown that the characteristics of sample heating during microwave pyrolysis strongly depend on the geometry and dimensions of the sample, electromagnetic and thermal parameters [56], and the distribution of the fuel inside the reactor [57]. In particular, Salema and Afzal [49] compared the experimental and numerical data obtained during the pyrolysis of biomass in the form of a layer and granules. The simulated surface temperature profile showed that most of the layer and biomass granules had a lower temperature compared to the edge parts of the fuel sample. The calculated data obtained agree with the experiment. As soon as a high temperature region (hot spot) is formed in the sample, it begins to expand, covering the surrounding mass of fuel. The biomass around the hotspot turns into a carbonaceous material, which absorbs large amounts of microwave energy and transfers heat to the environment. The temperature in the area of the hot spot is 165 °C higher than in the center of the biomass layer or granules. Such a temperature gradient may be due to the low thermal conductivity and diffusion capacity of the biomass. It is believed that the thermal conductivity of biomass materials limits heat and mass transfer. Usually, to achieve a very high rate of heat transfer, the biomass particles are crushed to small sizes.

Fragmented Distribution of Biomass

Figure 7 shows options for the distribution of the fuel sample in the crucible in the form of a combination of large and small fragments and a uniform distribution of fuel on a half of the crucible bottom.

Figure 8 shows the trends in the concentrations of the components of the gas mixture during the pyrolysis of sawdust. An intensive increase in the concentration of the main components of the gas mixture began with a decrease in O₂ concentration after 230 s (on average for the first two options of fuel distribution) and after 315 s when the fuel was distributed on a half of the crucible bottom. The time of intense release of gas components when the biomass was located in the form of large fragments, small fragments, and on half of the crucible bottom was 1220 s, 586 s and 1575 s, respectively. This suggests that the smaller fragments took less time to heat and dry up before the devolatilization began. The nature of emission trends (Figure 8a) is in good agreement with the data of Figure 4a. Two peaks are observed for the concentrations of CO, H₂, CH₄ and CO₂. The maximum values for CO and CO₂ were 22% and 11% (the first peak), 20% and 10.5% (the second peak).

The peak concentrations of CH₄ and H₂ were 7.7% and 4% (first peak), and 8.2% and 3.8% (second peak). When the biomass was distributed on the crucible bottom in the form of small fragments (Figure 8b), one concentration peak was observed for CO, CO₂ and CH₄, and two concentration peaks were for H₂. The nature of the trends in Figure 4c is generally similar to a gradual increase and decrease in gas concentrations when biomass was distributed in the form of small fragments (Figure 4b). The difference lies in the duration of gas yield, which was 2.5 times shorter for small fragments than when biomass was distributed in a uniform layer on a half of the crucible bottom. The maximum values of CO, CO₂ and CH₄ were 43%, 17% and 11%, respectively. The maximum H₂ concentration was 2.7% and 2.8% in the first and second peaks, respectively. Figure 9 shows the values of the average and maximum concentrations of the components of the gas mixture for different options of the fuel distribution in the crucible.

CO and CO₂ have the same distribution pattern of average concentrations, in which the average maximum concentrations were achieved when the biomass was distributed on a half of the crucible bottom, and the minimum when distributed in the form of large fragments (Figure 9a). When the fuel was evenly distributed, CO concentrations were 42% higher than when the biomass was distributed as large fragments and 33% higher than when distributed as small fragments. CH₄ and H₂ slightly depended on the method of fuel distribution. H₂S, SO₂ and NO had the same pattern of distribution of average concentrations (Figure 9a), in which the maximum values of gas concentrations were achieved when the biomass was distributed in the form of small fragments, and the minimum values were achieved when it was distributed in a uniform layer on a half of the crucible bottom. Similar conclusions are valid for the maximum values of H₂S, SO₂, NO (Figure 9b). The minimum values of CO and CO₂ (8.72% and 5.27%, respectively) were recorded in the experiment with the distribution of fuel in the form of large fragments because when the biomass was distributed as separate large fragments, the contact area of the entire sample of fuel with the bottom of the crucible decreased. Due to the remoteness of the biomass fragments from each other, the intensity of heat transfer between the particles decreased. Fragments remote from the magnetron warmed up more slowly than those closer to the magnetron.

Figure 10 shows the appearance of the solid residue of microwave pyrolysis of biomass with varying options for the distribution of fuel fragments on the crucible bottom. When the biomass was distributed into small fragments, the proportion of the solid residue relative to the initial mass was minimal. With the distribution of biomass along the crucible walls and on a half of the crucible bottom, this figure was 4.2% and 3.4%, respectively.

Figure 11 shows the emission trends for the main combustible components of the pyrolysis gas with varying distribution of fuel fragments on the crucible bottom.

Figure 11 shows that CO started to be released earlier when the fuel sample was distributed on half of the crucible bottom. CO₂ was released earlier (after about 240 s), when the biomass was distributed in the form of large and small fragments. When the biomass was distributed over a half of the crucible bottom, CO₂ emission began after 300 s. At the same time, for the last case of biomass distribution, the longest emission time of CO₂, CO and H₂ was observed: 1950 s, 2240 s and 1650 s, respectively. The shortest time for the emission of these gases was observed when the biomass was distributed in the form of small fragments (760 s, 750 s and 800 s, respectively). However, with the distribution of biomass in the form of small fragments, the highest concentrations of CO₂ and H₂ were recorded, equal to 18.7% and 4.2%, respectively (Figure 11b). The minimum values of CO and CO₂ were observed in the experiment with the distribution of biomass in the form of large fragments and amounted to 22.2% and 10.7%, respectively (Figure 11a,b). The minimum H₂ concentrations (2.6%) were recorded when the biomass was distributed in a uniform layer on a half of the crucible bottom (Figure 11c).

Table 5. The yield of the main gases.

Experiment	Gas	Average Concentrations (%)	Average Concentrations (ppm)	Average concentrations (mg/M3)	Total Amount of Pyrolysis Gas (mg/m3)
Biomass is distributed into large fragments	CO	8.72	87,200	101,535.88	203,070,000
	CO2	5.27	52,700	96,416.63	123,198,144
	CH4	3.2	32,000	21,337.54	25,455,041
	H2	1.47	14,700	2681.29	5,372,724
	H2S	0.004743	47.43	67.20	77,750
	SO2	0.00209	20.9	55.66	19,792
	NO	0.00182	18.2	22.71	45,510
Biomass is divided into small fragments	CO	10.52	105,200	122,495.12	153,853,720
	CO2	6.55	65,500	119,834.71	105,693,588
	CH4	3.67	36,700	24,471.49	28,117,179
	H2	1.08	10,800	904.94	1,237,576
	H2S	0.009053	90.53	128.26	172,288
	SO2	0.00285	28.5	75.90	30,056
	NO	0.002863	28.63	35.72	48,936
Biomass is distributed over a half of the crucible bottom	CO	15.5	155,000	180,482.36	465,102,114
	CO2	7.01	70,100	128,250.58	138,382,375
	CH4	2.88	28,800	19,203.78	45,319,080
	H2	0.88	8800	737.36	22,739,200
	H2S	0.003759	37.59	53.26	50,104.47
	SO2	0.00165	16.5	43.94	18,298.5
	NO	0.00123	12.3	15.34	31,783.2

shows the gas flow rates obtained during the experiments.

Thus, in the present study, the average concentrations of the components of the gas mixture had the following values: H₂—1.1 vol.%; CO—11.7 vol.%; CH₄—2.8 vol.%; CO₂—6.5 vol.% (Figure 5a and Figure 9a). Luo et al. [34] obtained the following concentrations of the components of the gas mixture formed during continuous microwave pyrolysis of cow manure: H₂—18.8 vol.%; CO—21.0 vol.%; CH₄—4.8 vol.%; and CO₂—48.8 vol.%. The pyrolysis temperature varied from 350 °C to 650 °C; the concentration of CO₂ used as the reaction medium varied from 0% to 100%. In [58], the concentrations of the components of the gas mixture during continuous microwave pyrolysis were the following: H₂—19.5 vol.%; CO—19.0 vol.%; CH₄—6.8 vol.%; and CO₂—16.9 vol.%. During periodic microwave pyrolysis, the concentrations were as follows: H₂—10.3 vol.%; CO—13.3 vol.%; CH₄—7.4 vol.%; and CO₂—12.0 vol.%. The differences in the obtained values can be explained, first of all, by the differences of the samples being pyrolyzed. Wood was used in the present experiments, and cow manure was used in [34,58]. The second reason is the different reaction mediums (air and carbon dioxide). CO₂ enhances the pyrolysis process and reacts with pyrolysis products. As a consequence, the pyrolysis gas has an increased concentration of components. Li et al. [52] performed microwave pyrolysis of spruce sawdust and recorded the following gas concentrations: H₂—38.1 vol.%; CO—39.5 vol.%; CH₄—8.13 vol.%; CO₂—11.8 vol.%. Catalysts were used in [52] to maximize the yield of hydrogen or pyrolysis gas. Zhao et al. [49] studied the pyrolysis of pine sawdust in a cylindrical furnace with electrical heating. The following concentrations of gases were obtained: H₂—7.0%; CO—8.3%; CH₄—14.8%; CO₂—40.0%. In study [52], a catalyst was used, which also enhanced the differences between the results of the present experiments and the data obtained earlier.

4.2. Surface Structure

This section details the effect of artificial modification of the biomass layer on the composition of the pyrolysis gas. Figure 12a,b shows the surface structure of the biomass. The artificial pores were small round voids in the surface layer of the biomass (Figure 12a). The channels were narrow, long, hollow spaces separating the biomass layer. Figure 13 shows the trends for the individual components of the gas mixture formed during the pyrolysis of sawdust with varying surface structure of the sample.

The release of CO, CH₄ and H₂ started faster when the biomass was distributed in an even layer. In addition, in this case, emission lasted the longest (Figure 13c). In the presence of channels and pores in the biomass structure, the start time and duration of gas release were almost identical (Figure 13a,b). During the pyrolysis of biomass with artificial channels in the structure, there were two peaks at which the maximum concentrations of CO, CO₂, CH₄ and H₂ gases were reached: 13%, 36%, 21% and 3.5% (first peak) and 13.8%, 34%, 17% and 2.8% (second peak). During the pyrolysis of a sample with artificial pores, one peak of CO and CO₂ concentrations was observed—52.5% and 18%, respectively. With the distribution of biomass in an even layer, we recorded a gradual increase and then a decrease in CO₂ concentrations with a maximum of 17%. The concentrations of CO, CH₄ and H₂ have two peaks: 25%, 3% and 2.5% (first peak), and 46%, 11% and 2.6% (second peak), respectively. Figure 14 shows the average and maximum concentrations of the gas mixture components.

The results showed that the creation of artificial pores on the surface of the biomass makes it possible to increase the CO yield by 18% and 32% compared to artificial channels and an even layer, respectively. The H₂ yield was 33% higher compared to an even layer and 3% lower than with artificial channels. The CO₂ yield increased by 23%, and the CH₄ yield was 2.2 times higher compared to the even layer. An increase in the concentrations of H₂, CO and CO₂ in the presence of artificial pores on the biomass surface shows that the pyrolysis proceeded more intensively. This conclusion is confirmed by the formation of a smaller ash residue (1.6%) in comparison with an even layer (3.5%). The creation of a large number of artificial channels on the surface of the biomass facilitates the diffusion of the pyrolyzing agent into the fuel layer, which intensifies the decomposition and the release of gaseous pyrolysis products. The creation of artificial pores on the surface of the biomass increased the CO yield by 6% compared to the pyrolysis of a uniformly distributed biomass. However, with artificial channels, CO yield decreased by 6% compared with a uniformly distributed biomass because of greater compaction of the layer between the channels. The material was compacted and, as a result, heat and mass transfer became more difficult compared to the case when the layer was uniformly distributed. Figure 15 shows the concentration trends of the main combustible components of the gas mixture.

Figure 15 shows that CO and H₂ emission started almost immediately for all three variants of the biomass surface structure. CO₂ began to be released later—after 125 s in the case of a structure with channels and pores, and after 370 s in the case of an even layer. In this case, for all three options for the structure of the biomass surface, the emission duration was the same. The maximum concentrations of CO and CO₂ were recorded for the structure with artificial pores (52.2% and 17.5%, respectively), and the minimum for the structure with artificial channels (35.5% and 13.5%, respectively). The values of H₂ concentrations changed insignificantly.

Table 6. The yield of the main gases.

Experiment	Gas	Average Concentrations (%)	Average Concentrations (ppm)	Average Concentrations (mg/m3)	Total Amount of Pyrolysis Gas (mg/m3)
Artificial channels	CO	14.63	146,300	170,352	168,818,832
	CO2	7.22	72,200	132,092	117,958,156
	CH4	6.02	60,200	40,141	40,462,128
	H2	1.33	13,300	1114	1,122,912
	H2S	0.007092	70.92	100	98,700
	SO2	0.00293	29.3	78	34,554
	NO	0.000012	0.12	0.15	4.8
Artificial pores	CO	16.47	164,700	191,777	197,146,756
	CO2	8.03	80,300	146,911	131,191,523
	CH4	6.45	64,500	430,008	144,948
	H2	1.21	12,100	1013	1,041,364
	H2S	0.010178	101.78	144	148,032
	SO2	0.003681	36.81	98.03	44,394
	NO	0.000424	4.24	5.29	2507.46
Uniform layer	CO	15.5	155,000	180,482.36	465,102,114
	CO2	7.01	70,100	128,250.58	138,382,375
	CH4	2.88	28,800	19,203.78	45,319,080
	H2	0.88	8800	737.36	22,739,200
	H2S	0.003759	37.59	53.26	50,104.47
	SO2	0.00165	16.5	43.94	18,298.5
	NO	0.00123	12.3	15.34	31,783.2

shows the gas flow rates obtained during the experiments.

In modern studies, insufficient attention is paid to the structure of the biomass layer. Biomass is mainly distributed either in the form of a layer over the entire crucible bottom or in the form of granules or bars [47,49,55]. Analytical studies [59] of microwave heating in rectangular, cylindrical and spherical coordinate systems show that the geometry and density of an object strongly affect the distribution of temperature or humidity. Brodie [59] showed that the highest temperatures in rectangular blocks usually occur in the corners, directly below the surface. For small-diameter cylinders with low dielectric losses, the resulting temperature distribution resembles a “dumbbell” with two temperature peaks along the longitudinal axis. However, as the loss factor or cylinder diameter increases, the temperature profile becomes a high-temperature ring under the top and bottom circular surfaces. Heat in small-diameter spheres is concentrated in the center, but, as in a cylinder, it is converted to subsurface heating as the diameter or material loss factor increases. The heating time also affects the temperature distribution, with the temperature peak shifting towards the core of most objects as the heating time increases. There is a transition to subsurface heating as the diameter or material loss factor increases.

4.3. Influence of Biomass Particle Size

This section presents the results on the effect of biomass particle fineness on the yield of pyrolysis gas. We used sawdust with an average particle size of 140, 250 and 2000 μm, which were pre-impregnated with moisture. The sample weight was 15 g. Figure 16 shows emission trends obtained during the pyrolysis of biomass with different particle sizes.

It has been established (Figure 16) that the H₂ release starts from the first seconds of heating during the pyrolysis of particles with a size of 2000 μm. In this case, the longest time for the release of pyrolysis gas was observed. The shortest gas emission time was recorded during the pyrolysis of particles with an average size of 140 μm. During the pyrolysis of particles of this size, one peak in the concentrations of CO and CO₂ (62.5% and 19%, respectively), as well as two peaks in the concentrations of CH₄, were observed (Figure 16a). Figure 17 shows the average and maximum concentrations of the gas mixture components.

Experiments have shown that pyrolysis of smaller particles (140 µm and 250 µm) gives higher concentrations of CH₄ and H₂ (Figure 17). The yield of CO during the pyrolysis of sawdust with a particle size of 140 µm increased by 12% and 65% compared to particles with a size of 250 µm and 2000 µm, respectively (Figure 17a). A similar comparison was made for other gases: the CO₂ yield increased by 1.5% and 15%, the CH₄ yield increased by 30% and 60%, and the H₂ yield increased by 1.5% and 43%. During the pyrolysis of biomass with a particle size of 140 µm, the maximum yield of CH₄ and H₂ increased by 3 and 1.5 times, respectively, compared with the particles of 2000 µm (Figure 17b).

Table 7. The yield of the main gases.

Experiment	Gas	Average Concentrations (%)	Average Concentrations (ppm)	Average Concentrations (mg/m3)	Total Amount of Pyrolysis Gas (mg/m3)
140 µm	CO	17.38	173,800	202,373	135,792,283
	CO2	6.91	69,100	126,421	86,851,227
	CH4	7.1	71,000	47,342	31,766,482
	H2	1.09	10,900	913	1,291,895
	H2S	0.0106	106	150	106,650
	SO2	0.0045049	45.05	120	32,160
	NO	0.00425	42.5	53	37,365
250 µm	CO	15.5	155,000	179,800	233,740,000
	CO2	7.01	700,100	1,281,183	1,076,193,720
	CH4	5.01	50,100	33,406	47,269,490
	H2	0.88	8800	913	1,291,895
	H2S	0.006342	63.4	89.85	125,935
	SO2	0.0018449	18.45	49.13	22,649
	NO	0.000008	0.08	0.10	0.8
2000 µm	CO	10.2	102,000	118,319.9	165,648,000
	CO2	8.07	807,000	1,476,810	2,658,258,000
	CH4	2.88	28,800	19,203.78	45,319,080
	H2	0.62	6200	523	1,113,000
	H2S	0.003759	37.59	53.26	50,104.47
	SO2	0.00165	16.5	43.94	18,298.5
	NO	0.00123	12.3	15.34	31,783.2

shows the gas flow rates obtained during the experiments.

The experimental results are in good agreement with the data obtained in [60]. Mohammed et al. [60] studied the effect of biomass particle size (0.3 to 1.0 mm) on the yield of gasification products such as gas, tar and char. The temperature in the reactor varied from 700 °C to 1000 °C. The results showed that smaller biomass particles produced more H₂, CO and CH₄ than larger particles. In particular, feedstock particles with a size of 0.3–0.5 mm provided a higher H₂ yield (33.93 vol.%), which is 1.5 times higher compared to particles with a size of 0.5–1.0 mm. The influence of the biomass particle size on the yield and composition of the pyrolysis gas was also studied in [61], where palm oil waste with particle sizes of 0.15–1.0, 1.0–2.0 and 2.0–5.0 mm was used as a fuel. The H₂ yield increased by 5% as the particle size decreased from 5 to 0.15 mm. In the present study, the minimum concentration of CO of 10.2% was recorded for particles with a size of 250 μm.

The established trends are due to the fact that raw materials of finer grinding have a larger total surface area of the particles. As a consequence, such fuel warms up better, and higher reaction rates can be achieved during pyrolysis. This contributes to the production of a larger amount of light gases and a smaller amount of coke and condensate [61,62]. With a larger particle size, the temperature gradient inside it increases. At the same moment of time, the temperature inside the particle is lower than on its surface, which leads to an increase in the yield of semi-coke and resins and, accordingly, the gas yield decreases [63]. Figure 18 shows the trends of changes in the individual components of the gas mixture formed during the pyrolysis of sawdust with varying particle sizes.

Experiments showed (Figure 18) that the release of CO started faster during the pyrolysis of biomass with a particle size of 250 µm. The release of CO₂ was the first to start during the pyrolysis of biomass with a particle size of 140 µm and 250 μm. The CO₂ concentration started to grow after 2 min. During the pyrolysis of sawdust with an average size of 2000 µm, the growth of CO₂ concentration occurred after 5 min. The shortest gas emission time was recorded for sawdust with a size of 140 µm; the longest gas emission time was recorded during the pyrolysis of sawdust with an average particle size of 2000 µm. The release of H₂ started immediately at the beginning of the experiment. In [64], microwave pyrolysis of wheat straw was studied. For a particle size of 0–0.09 mm and a microwave power of 800 W, the release of CO₂ began at the 2nd min of the experiment. The emission of CO and CH₄ began after 3 and 4 min of the experiment, respectively.

Based on the experimental data, the efficiency coefficient K_eff. was calculated using the formula: K_eff. = N_max/N_min, where N is the average gas concentration. The results of the calculations are shown in

Table 8. Coefficients of microwave pyrolysis efficiency.

Variable Parameter	Gas	Maximum Value	Minimum Value	Keff.
Distribution of biomass in the crucible	CO	15.5%	8.6%	1.8
	CO2	7.02%	6.31%	1.11
	CH4	2.88%	1.8%	1.6
	H2	1.16%	0.88%	1.3
	H2S	67.66 ppm	37.59 ppm	1.7
	SO2	34.46 ppm	16.5 ppm	2.08
	NO	12.31 ppm	2.77 ppm	4.44
Fragmented distribution of biomass	CO	15.5%	8.72%	1.7
	CO2	7.01%	5.27%	1.3
	CH4	3.67%	2.88%	1.2
	H2	1.47%	0.88%	1.6
	H2S	90.53 ppm	37.59 ppm	2.4
	SO2	28.5 ppm	16.5 ppm	1.7
	NO	28.63 ppm	12.31 ppm	2.32
Surface structure	CO	16.47%	14.63%	1.12
	CO2	8.03%	7.01%	1.14
	CH4	6.45%	2.88%	2.2
	H2	1.33%	0.88%	1.5
	H2S	101.78 ppm	37.59 ppm	2.7
	SO2	36.81 ppm	16.5 ppm	2.34
	NO	12.31 ppm	3.55 ppm	3.46
Average size of biomass particles	CO	17.38%	10.2%	1.7
	CO2	8.07%	6.91%	1.18
	CH4	7.1%	2.88%	2.4
	H2	1.09%	0.62%	1.75
	H2S	106 ppm	37.59 ppm	2.81
	SO2	45.05 ppm	16.5 ppm	2.73
	NO	42.57 ppm	12.31 ppm	3.4

. The analysis showed that the efficiency of microwave pyrolysis can vary in a fairly wide range when the process conditions change.

5. Conclusions

The shift of the biomass sample relative to the center to the crucible wall made it possible to double the CO yield as compared with its placement in the center and to increase it by 1.5 times as compared with the distribution of the biomass over the entire bottom of the crucible. The smallest solid residue (3.4% of the initial mass) was recorded during the pyrolysis of a biomass sample displaced from the center to the crucible wall. When the biomass was distributed along the crucible wall, the proportion of the solid residue was 4.2%, and when the sample was distributed over the entire bottom of the crucible, it was 4.5%.

The creation of artificial pores in the near-surface layer of biomass intensifies the diffusion of the pyrolyzing agent into the material. The processes of thermal decomposition and release of pyrolysis gas components were accelerated. Due to this approach, it was possible to increase the yield of carbon monoxide by 18% and 32% compared to biomass pyrolysis with artificial channels and without any artificial elements in a layer. The creation of artificial channels in a biomass layer did not lead to significant changes in the concentration of pyrolysis gas components.

Microwave pyrolysis of small biomass particles (140 and 250 µm) was more efficient than pyrolysis of 2000 µm particles. Smaller particle size increased the concentrations of the main combustible gas components by 42%. For example, the CO yield during the pyrolysis of sawdust with a particle size of 140 µm increased by 42% and 11% compared to the pyrolysis of sawdust with a particle size of 250 and 2000 µm, CO₂ decreased by 15% and 1.5%, and CH₄ yield increased by 30% and 60%. During the pyrolysis of small particles, the solid residue decreases and the gas yield increases.

The results obtained are of high importance for choosing the designs of reactors and platforms for loading biomass.