Low Temperature Reduction Mechanism of Carbon-Iron Ore Composite Using Woody Biomass

Abstract

Utilization of biomass, which is regarded as a carbon neutral fuel, has been discussed to decrease in the carbon dioxide emission from the ironmaking processes. Previous reports insisted the volatile matters in the uncarbonized biomass contributed to the reduction of iron ore at low temperature. However, its mechanism has not been explained by the pyrolysis reaction of biomass components. In this study, the low temperature reduction behavior of the carbon-iron ore composite using uncarbonized biomass was compared with that using its components such as cellulose and lignin. Furthermore, the effect of volatile matters in the biomass on the reduction of iron ores was examined.

The composite using uncarbonized biomass was started to be reduced at 400–450°C by CO produced by pyrolysis of cellulose. As the temperature was further increased, iron ore was reduced to metallic iron at approximately 800°C by hydrogen produced by the interaction between cellulose and lignin. The metallic iron may contribute to the gasification of biomass char and fast reduction reaction as the catalyst.

1. Introduction

In the ironmaking process, an innovative technology to drastically decrease carbon dioxide emissions has been required. However, it is difficult since the ironmaking process requires a large amount of energy in principle and the current blast furnace process is already highly efficient. Utilizations of green hydrogen and carbon-neutral carbonaceous materials are the substantial ways to decrease the carbon dioxide emission, although it will be difficult to use sufficient amount of green hydrogen in the immediate future, especially in the East Asian area.

There are also some issues on the utilization of biomass such as its low energy density and inefficient collection systems. In Brazil, a part of molten iron is produced by small size blast furnaces using biomass char.¹⁾ Further, a trial was made that woody biomass was utilized in the coke making process.^2,3,4,5) Montiano et al.³⁾ investigated the effect of adding sawdust to coal blends on the quality of the cokes. They noted that the 5 mass% addition of sawdust impairs the quality of the cokes because the bulk density of coal blends decreases. Ueki et al.⁴⁾ evaluated the possibility of adding biomass to coke. They indicate volatile matters of biomass inhibit the connectivity between coal particles in the coke. Florentino-Madiedo et al.,⁵⁾ however, pointed that lignin which is a component of biomass promotes the strength of cokes because agglomerated lignin during heating integrates within the coke matrix. The obtained char was also injected to the blast furnace through tuyeres.^6,7) The heat of 120–148 J/g⁸⁾ is required for the carbonization depending on the type of biomass, its water content and carbonization conditions. Some of heats will not be necessary if biomass can directly use for the reduction of iron ores.

The authors have proposed a rapid carbonized process using a rotary kiln-type furnace applying heat storage material balls, namely HSM-PC balls which promote heat transfer and pulverization of produced char.⁹⁾ Further, iron ores fed into this process were reduced to metallic iron by CO and H₂ generated during carbonizing of biomass.¹⁰⁾ Ueki et al. reported on the metallic iron formation observed for the composites using ceder powder and hematite reagent heated at 1000–1300°C for 0.3 ks.¹¹⁾ Wei et al. examined the reduction experiment of the composite using pine sawdust at 800–1250°C. They reported iron oxides were rapidly reduced by two-step reduction reaction, namely reduction by volatile carbon and that by carbonized char.¹²⁾ The metallic iron formation was also observed in the composite using Japanese ceder wood and pisolite ore heated up to 800°C.¹⁰⁾ Each report points to the rapid reduction reaction of iron oxides by volatile matters in the woody biomass. The detail mechanism, however, has been not clarified yet.

The main components of woody biomass are cellulose, lignin and hemicellulose. Orfão et al. investigated the changes in the ratio of pyrolysis residue of woody biomass and its main components under inert atmosphere with time. They validated a model describing the pyrolysis behavior of woody biomass by independent those of its main component.¹³⁾ This indicates the possibility that the reduction mechanism of carbon-iron ore composite using woody biomass will be described by the pyrolysis of its components.

In this study, the relation between the reduction reaction of iron ore and the pyrolysis of woody biomass and its component reagents was investigated. Further, low temperature reduction mechanism of the carbon-iron ore composite using woody biomass was clarified by examining the effect of volatile matters in the composite on the reduction behavior.

2. Experimental

Australian pisolite ore (T. Fe: 57.2 mass%, SiO₂: 5.5 mass%, Al₂O₃: 2.5 mass%, LOI: 10.1 mass%) and uncarbonized biomass (Chamaecyparis obtusa) were pulverized and adjusted −250 μm using the sieve. Alumina powder (ca. 1 μm) was also prepared. Ultimate and proximate analyses data of the biomass measured on air dried basis are listed in Table 1. The value of water is not included in that of volatile matters. The content ratio of cellulose, lignin and hemicellulose in the biomass were 45.0%, 36.3% and 17.0%, respectively. The reagents of these major components were also used. The particle sizes of reagents of cellulose and lignin were under 200 μm and 50 μm, respectively. Hemicellulose is a general term of polysaccharides, excluding cellulose. Therefore, a reagent of xylan, the main component of hemicellulose, with a particle size of under 200 μm was used.¹⁴⁾ Model biomass was obtained to mix the reagents at same ratio of used biomass component. Torrefied biomass samples were also prepared by dry distillation of the uncarbonized biomass heated up to 400°C and 700°C, respectively. The particle sizes were adjusted −150 μm. The ratios of weight loss and volatile content of samples torrefied at 400°C and 700°C were 60.5%, 68.1% and 77.3%, 44.0%, respectively. Torrefied model biomass heated up to 700°C was prepared. Specific surface areas of the biomass and model biomass torrefied at 700°C were measured by BET method.

Table 1. Values of ultimate and proximate analyses of biomass sample (mass%).

Ultimate analysis					Proximate analysis
C	H	N	O	S	F. C	V. M.	Ash	Water
46.85	6.27	0.20	46.34	0.08	12.48	87.26	0.26	10.78

The appearances of the carbonaceous materials are shown in Fig. 1. Fibrous structure is observed in the pulverized biomass and the cellulose reagent while torrefied biomass shows grain structure. This is because grindability of biomass is improved due to dry distillation.⁶⁾ The reagents of lignin and xylan also show grain structure with several tens of μm.

Fig. 1. SEM images of carbonaceous materials (a) woody biomass, (b) model biomass, torrefied biomass heated up to (c) 400°C and (d) 700°C.

The cylindrical composites with 10 mm in diameter were prepared by press-shaping the mixture of 1.0 g of ore and carbonaceous materials under a pressure of 98 MPa. Table 2 shows the mass ratio of carbonaceous materials to ore or alumina in each composite sample. In order to evaluate the gasification behavior of woody biomass only, W_A was prepared using alumina powder instead of ore. The values in parentheses show the ratio calculated by the weight of biomass before dry distillation treatment. C/O ratio, the molar ratio of total carbon in the biomass to oxygen in iron oxides, was set as to 1.0 for the sample W. The sample R was prepared to mix the model biomass with ore. The samples RC and RL were made by only cellulose or lignin reagent as a carbonaceous material, respectively. The C/O ratio of R, RC and RL were 1.0, 0.43 and 0.41, respectively. The biomass torrefied at 400°C and 700°C were used for the sample W₄₀₀ and W₇₀₀, respectively. These were prepared to be same amount of biomass before the dry distillation as that of the sample W.

Table 2. Mixing ratios of carbonaceous materials to iron ore or alumina powders (mass%).

Sample	Biomass	Cellulose	Lignin	Xylan	Mixed with
W	43.5	–	–	–	Ore
WA	43.5	–	–	–	Alumina
R	–	20.0	16.1	7.4	Ore
RC	–	20.0	–	–	Ore
RL	–	–	16.1	–	Ore
W400	17.0 (43.5)	–	–	–	Ore
W700	10.0 (43.5)	–	–	–	Ore

The composite sample was set in the experimental apparatus as shown Fig. 2.¹⁰⁾ Alumina balls with 3 mm in diameter were filled in the bottom of the sample holder as a heater for supplied gas from lower part of the furnace. After evacuating air in the chamber, Ar-5%N₂ gas was introduced at the rate of 8.33×10⁻⁶ Nm³/s under atmospheric pressure. Then the sample was heated up to 1200°C at a heating rate of 0.17°C/s using an infrared image furnace. The temperature at 1 mm upper the surface of the sample was measured using an R-type thermocouple. The concentrations of N₂, CH₄, CO, CO₂, C₂H₄, C₂H₆ and H₂O of the outlet gas were measured during the experiment at 90 s intervals by a gas chromatography. N₂ gas was used as a tracer to estimate the amount of gas generated from the sample. The sample was cooled to room temperature by turning off the furnace after reaching certain temperature.

Fig. 2. Schematic diagram of an experimental apparatus. (Online version in color.)

Shrinkage ratio of the sample was calculated by Eq. (1).

  

$$\mathrm{S}\mathrm{h}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{k}\mathrm{a}\mathrm{g}\mathrm{e}\hspace{0.17em}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}(\%)=\left(1-\frac{{\mathrm{d}}_{\mathrm{a}}^2{\mathrm{h}}_{\mathrm{a}}}{{\mathrm{d}}_{\mathrm{b}}^2{\mathrm{h}}_{\mathrm{b}}}\right)\times 100$$(1)

d and h are diameter and height of the samples measured using a caliper, respectively. The subscripts of a and b means before and after the experiment, respectively. Remained carbon in the reduced composite was removed by the previous method after pulverizing the sample in liquid nitrogen.¹⁵⁾ The compounds existed in the sample were identified by XRD (RIGAKU SmartLab, 40 kV, 30 mA, Fe Ka (0.19373 nm)). Further, oxygen content in the sample was measured by inert gas fusion – infrared absorption method using ONH836 combustion analyzer (LECO). Reduction degree was defined by Eq. (2).

  

$$\text{Reduction}\mathrm{\hspace{0.17em} }\text{degree}\left(\text{\%}\right)=\left(1-\frac{{\text{O}}_{\text{m}}-{\text{O}}_{\text{total}}+{\text{O}}_{\text{Fe}}}{{\text{O}}_{\text{Fe}}}\right)\times 100$$(2)

O_Fe is the molar amount of oxygen in the iron oxide in the sample. O_m is the molar amount of oxygen in the reduced sample after removing remained carbon. O_total is the molar amount of oxygen in the ore sample.

3. Results and Discussion

3.1. Reduction Behavior of Carbon-Iron Ore Composite Using Woody Biomass

Appearances of sample W heated up to 400°C to 1200°C is shown in Fig. 3. The sample heated up to 1200°C maintains cylindrical shape. Furthermore, alumina balls filled in the sample holder adheres to the bottom of the sample. Those heated up to 800°C and 1000°C, on the other hand, are partially collapsed during sample removal from the equipment. This indicates that molten slag binds among particles in the sample and between the sample and alumina balls.

Fig. 3. Appearances of the composites prepared using woody biomass heated up to different temperatures. (Online version in color.)

The changes in shrinkage ratio and reduction degree with temperature of sample W are shown in Fig. 4. The shrinkage ratio of the sample increases as the temperature increases. Especially, it increases significantly above 800°C. This is because reduced iron particles are sintered each other. The reduction reaction started to proceed between 400°C and 600°C. That heated up to 800°C shows the relatively lower value than that to 600°C. The reason is explained in the next paragraph. The value of reduction degree is 98% at 1200°C.

Fig. 4. Changes in shrinkage ratio and reduction degree of the composite prepared using woody biomass with temperature. (Online version in color.)

XRD profiles of the sample W heated up to 400°C, 450°C and 800°C are shown in Fig. 5 in order to investigate the reduction behavior at low temperature. Only Fe₂O₃ phase is detected as iron oxide in the sample heated up to 400°C. That heated up to 450°C, on the other hand, contains Fe₃O₄. This indicates the reduction of iron oxide is proceeded above 400°C. The intense of FeO peak is relatively low against those of Fe and Fe₃O₄ in the sample heated up to 800°C. This is considered to be because reduced iron particles are re-oxidized during sample removal to air atmosphere after heating experiment, and consistent with the result that calculated reduction degree shown in Fig. 4 decreases at 800°C.

Fig. 5. XRD profiles of the composites prepared using woody biomass heated up to 400°C, 450°C and 800°C. (Online version in color.)

3.2. Effects of Biomass Components on Reduction Behavior

The changes in CO gas ratio (P_CO / (P_CO + P_CO2)) obtained for the sample W, R, RC and RL with temperature, plotted on the phase diagram of Fe–C–O system is shown in Fig. 6. The equilibrium line of the Boudouard reaction (C + CO₂ = 2CO) is represented as dashed line in the diagram. The temperature at which the CO gas ratio of W reaches the equilibrium lines of Fe₃O₄–FeO and FeO–Fe are 680°C and 810°C, respectively. The changes in the CO gas ratio of R and RC are almost consistent with that of W up to around 800°C. That of RL, on the other hand, shifts to about 150°C higher than that of the others. This indicates the pyrolysis of cellulose have large effects on the reduction of the sample W up to 800°C. Changes in the generation rates of CO for each sample with temperature are shown in Fig. 7. The rate in the case of W is decreased compared to that of W_A after showing the peak around 400°C. This is because CO is consumed for the reduction of the iron ore. Then, the rate is increased again around 650°C. Such behavior of CO generation is also observed in R and RC. In RL, however, CO generation is not detected below 800°C. Cellulose has a linear structure made by glycosidic bond with C₆H₁₀O₅ as one unit. As the pyrolysis of cellulose proceeds, levoglucosan is produced by cleaving the linear structure of cellulose.¹⁶⁾ Its boiling point is 385°C.¹⁷⁾ It is reported that volatilized levoglucosan is decomposed to CO.¹⁸⁾ Thus, the generation rate of CO of W, R and RC shows a peak around 400°C. The rates of the samples containing cellulose start to increase again at approximately 700°C. In the case of RC, the rate is decreased above 840°C due to its lower carbon content than the others. Drastic increase of the rate is only observed in W.

Fig. 6. Changes in P_CO/(P_CO+P_CO2) observed for the composites prepared using woody biomass and reagents with temperature. (Online version in color.)

Fig. 7. Changes in the generation rates of CO gas from the composites prepared using woody biomass, reagents and alumina powders with temperature. (Online version in color.)

Figure 8 shows the phase diagram of Fe–H–O system with the changes in H₂ gas ratio (P_H2 / (P_H2 + P_H2O)) obtained for the sample W, R, RC and RL with temperature. The temperature at which the H₂ gas ratio of W reaches the equilibrium lines of Fe₃O₄–FeO and FeO–Fe are 730°C and 820°C, respectively. They are slightly higher than the case of CO gas ratio. The H₂ gas ratio of the samples made of reagents reach the equilibrium line of FeO–Fe at higher temperature than that of W. This means that factors other than the pyrolysis of the biomass components may have effects on the hydrogen reduction of FeO.

Fig. 8. Changes in P_H2/(P_H2+P_H2O) observed for the composites prepared using woody biomass and reagents with temperature. (Online version in color.)

Changes in the generation rates of H₂ for each sample with temperature are shown in Fig. 9. The rates increase in W, R and W_A around 600°C. This is considered to be because the production of H₂ increases due to interaction of the pyrolysis of lignin and cellulose.^19,20) Furthermore, drastic increase of H₂ generation rate is also observed around 750°C in W.

Fig. 9. Changes in the generation rates of H2 gas from the composites prepared using woody biomass, reagents and alumina powders with temperature. (Online version in color.)

In order to investigate the difference of generation behaviors of CO and H₂ around 750°C between W and R, SEM images of biomass and model biomass after dry distillation at 700°C are shown in Fig. 10, respectively. Several numbers of pores can be observed in the torrefied woody biomass. The specific surface areas of woody and model biomass after torrefied were 371.6 m²/g and 0.7 m²/g, respectively. Lignin melts over 200°C and has adhesion properties.²¹⁾ Giudicianni et al.²²⁾ also torrefied the mixture of the reagents of cellulose, lignin and xylan. They reported molten lignin matrix incorporates the porous cells organized structure of cellulose. In the case of R, the contact between the reducing gas and iron ores may be obstructed by spreading molten lignin on the iron ores. Lignin in biomass, however, is considered to be difficult to form such matrix because it coils around cellulose with several nano meters in thickness.²³⁾ These results indicate the morphology of the biomass have effects on reducing properties for iron ores.

Fig. 10. SEM images of woody and model biomass after dry distillation at 700°C.

3.3. Effects of CO and H₂ from Volatile Matters on Reduction Behavior

XRD profiles of W, W₄₀₀ and W₇₀₀ heated up to 800°C for 3.0 ks are shown in Fig. 11. W has significant peaks of Fe. The main phases of W₄₀₀ and W₇₀₀, on the other hand, are FeO and Fe₃O₄, respectively. Fe₂SiO₄ is detected from W and W₄₀₀. This is because the formed wustite reacts with SiO₂ in ash.

Fig. 11. XRD profiles of the composites prepared using woody and torrefied biomass heated at 800°C for 3.0 ks. (Online version in color.)

Figure 12 shows changes in equilibrium oxygen partial pressures, $P_{{\text{O}}_2}^{\text{CO}}$ and $P_{{\text{O}}_2}^{{\text{H}}_2}$, of each sample on the Fe–O phase diagram. $P_{{\text{O}}_2}^{\text{CO}}$ and $P_{{\text{O}}_2}^{{\text{H}}_2}$ are calculated assuming that the reactions shown as Eqs. (3) and (4) are in equilibrium, respectively.

  

$$2\text{CO}+{\text{O}}_2=2{\text{CO}}_2$$(3)

  

$$2{\text{H}}_2+{\text{O}}_2=2{\text{H}}_2\text{O}$$(4)

Fig. 12. Changes in oxygen partial pressures observed for the composites prepared using woody and torrefied biomass with temperature. (Online version in color.)

W₇₀₀ has higher oxygen partial pressures than the other sample, while W₄₀₀ shows similar tendency of oxygen partial pressure changes to W. This result is consistent with the indication that reducing gases derived from volatile matters generated above 400°C contributed to the reduction of iron ores.

In this temperature region, there is a possibility that the water-gas shift reaction proceeds due to presence of H₂, CO, CO₂ and H₂O gases derived from volatile matters. If the water-gas shift reaction has to be taken into account in the analysis, it becomes difficult to distinguish the reactions: reduction by H₂ or by CO. To determine the temperature at which the water-gas shift reaction reaches equilibrium, the equilibrium constant from Eq. (5), K defined as Eq. (6),²⁴⁾ and the experimentally obtained value, Q, defined by Eq. (7) were compared.

  

$$\text{CO}+{\text{H}}_2\text{O}={\text{CO}}_2+{\text{H}}_2$$(5)

  

$$\text{K}=exp\left(\frac{36\mathrm{\hspace{0.17em} }045-31.07\text{T}}{\text{RT}}\right)$$(6)

  

$$\text{Q}=\frac{P_{{\text{CO}}_2}\cdot P_{{\text{H}}_2}}{P_{\text{CO}}\cdot P_{{\text{H}}_2\text{O}}}$$(7)

P_i is the partial pressure of gas i. The changes in the ratio of the measured gas ratio to the equilibrium constant (Q/K) obtained from W and W₄₀₀ are shown in Fig. 13. The water-gas shift reaction reaches equilibrium when Q/K is 1. Above 400°C, the value of Q/K is below 1. Especially, in the temperature region of 800–1100°C, the value is below 0.5. Furthermore, the generation rate of CO is largest during this temperature range. The reaction should proceed to generate hydrogen, and it leads to increase Q value if this rate reaction is fast. These indicate that water-gas shift reaction has less effect on the composition of generated gas from the composite especially in the region of 800–1100°C compared with that of the carbon gasification. Therefore, the water-gas shift reaction is not considered in the analysis of this experiment above 800°C.²⁵⁾

Fig. 13. Changes in the ratios of the measured gas ratio (Q) to the equilibrium constant of water-gas shift reaction obtained from the composites prepared using woody and torrefied biomass with temperature. (Online version in color.)

$T_{\text{Fe}}^{\text{CO}}$ and $T_{\text{Fe}}^{{\text{H}}_2}$ are defined as the temperature at which $P_{{\text{O}}_2}^{\text{CO}}$ and $P_{{\text{O}}_2}^{{\text{H}}_2}$ reach the Fe stable phase shown in Fig. 12, respectively. Their changes with volatile matter content ratio of biomass are shown in Fig. 14. $T_{\text{Fe}}^{{\text{H}}_2}$ is higher than $T_{\text{Fe}}^{\text{CO}}$, and both of the temperature decreases with increasing volatile matter ratio. Reduced metallic iron by CO can be re-oxidized by H₂O since $P_{{\text{O}}_2}^{{\text{H}}_2}$ is dominated by the equilibrium of FeO reduction in the temperature range between $T_{\text{Fe}}^{\text{CO}}$ and $T_{\text{Fe}}^{{\text{H}}_2}$. In other words, metallic iron is considered to be stable above $T_{\text{Fe}}^{{\text{H}}_2}$. Changes in the generation rates of CO for W and W₄₀₀ with temperature is shown in Fig. 15. $T_{\text{Fe}}^{{\text{H}}_2}$ of W and W₄₀₀ are also shown. The rates start to increase drastically around $T_{\text{Fe}}^{{\text{H}}_2}$. This is considered to be because metallic iron has catalytic effect on the gasification of carbon.²⁶⁾ Once the oxygen partial pressure of the composite reaches the Fe stable condition, metallic iron can exist in the composite. The CO gas drastically generated by the catalytic effect of initially reduced metallic iron may contribute to further reduction reaction. These results reveal that woody biomass has more reducing potential than torrefied one.

Fig. 14. Effect of volatile matter ratio of biomass on the stable temperature of metallic iron for the oxygen partial pressures corresponding to CO and H₂ gases, respectively.

Fig. 15. Changes in the generation rates of CO gas from the composites prepared using woody and torrefied biomass with temperature. (Online version in color.)

4. Conclusions

In this study, the objective is elucidating of low temperature reduction mechanism of the carbon-iron ore composite using woody biomass. The relation between the reduction reaction of iron ore and the pyrolysis of woody biomass and its component reagents was investigated. Gasification and reduction behaviors of the carbon-iron ore composites using torrefied biomass at different treatment temperatures were examined. The following results were obtained.

• The carbon-iron ore composite using woody biomass starts to be reduced around 400°C, and can be reduced to metallic iron around 800°C by CO and H₂. The CO is mainly derived from the pyrolysis of cellulose and the H₂ was produced by interaction between cellulose and lignin. The factors of low temperature reduction of the composite using woody biomass are not only volatile matters but also its morphology.

• Reduced metallic iron existing stably at low temperature promotes CO generation since acting as catalyst for gasification of carbon. Drastic increase in the generation rate of CO contributes the rapid reduction of the carbon-iron ore composite using biomass.

• In the case of using torrefied biomass, the rapid reduction to metallic iron shifts to higher temperature than that using woody biomass due to reoxidation of metallic iron by H₂O.

Acknowledgment

This work was supported by JSPS Grant-in-Aid for JSPS Fellows Grant Number 22KJ0282, Steel Foundation for Environmental Protection Technology and JFE 21st Century Foundation.

References

• 1)   R.  Taniguchi and  Y.  Serizawa: Tetsu-to-Hagané, 68 (1982), 2086 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.68.15_2086
• 2)   M.  Ikemori,  Y.  Nunome,  Y.  Ueki,  R.  Yoshiie,  I.  Naruse,  Y.  Nishibata and  S.  Aizawa: Proc. Coal Sci. Conf., 49 (2012), 58 (in Japanese). https://doi.org/10.20550/jiesekitanronbun.49.0_58
• 3)   M. G.  Montiano,  E.  Díaz-Faes,  C.  Barriocanal and  R.  Alvarez: Fuel, 116 (2014), 175. https://doi.org/10.1016/j.fuel.2013.07.070
• 4)   Y.  Ueki,  Y.  Nunome,  R.  Yoshiie,  I.  Naruse,  Y.  Nishibata and  S.  Aizawa: ISIJ Int., 54 (2014), 2454. https://doi.org/10.2355/isijinternational.54.2454
• 5)   L.  Florentino-Madiedo,  E.  Díaz-Faes and  C.  Barriocanal: Renew. Energy, 146 (2020), 1717. https://doi.org/10.1016/j.renene.2019.07.139
• 6)   S.  Ueda,  R.  Inoue and  T.  Ariyama: Tetsu-to-Hagané, 94 (2008), 468 (in Japanese). https://doi.org/10.2355/tetsutohagane.94.468
• 7)   H.  Tsuruta,  M.  Asanuma,  A.  Fujibayashi and  A.  Murao: Proc. Biomass Sci. Conf., 8 (2012), 10 (in Japanese). https://doi.org/10.20550/jiebiomassronbun.8.0_10
• 8)   M. V.  Velden,  J.  Baeyens,  A.  Brems,  B.  Janssens and  R.  Dewil: Renewable Energy, 35 (2010), 232. https://doi.org/10.1016/j.renene.2009.04.019
• 9)   D.  Maruoka,  T.  Nakamura,  T.  Murakami and  E.  Kasai: Tetsu-to-Hagané, 102 (2016), 730 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2016-025
• 10)   T.  Murakami,  R.  Higashi,  D.  Maruoka and  E.  Kasai: ISIJ Int., 61 (2021), 2971. https://doi.org/10.2355/isijinternational.ISIJINT-2021-342
• 11)   Y.  Ueki,  R.  Yoshiie,  I.  Naruse,  K.  Ohno,  T.  Maeda,  K.  Nishioka and  M.  Shimizu: Fuel, 104 (2013), 58. https://doi.org/10.1016/j.fuel.2010.09.019
• 12)   R.  Wei,  D.  Cang,  Y.  Bai,  D.  Huang and  X.  Liu: Ironmak. Steelmak., 43 (2016), 144. https://doi.org/10.1179/1743281215Y.0000000061
• 13)   J. J. M.  Orfão,  F. J. A.  Antunes and  J. L.  Figueiredo: Fuel, 78 (1999), 349. https://doi.org/10.1016/S0016-2361(98)00156-2
• 14)   J.  Wang,  E.  Minami and  H.  Kawamoto: J. Wood Sci., 66 (2020), Article number: 41. https://doi.org/10.1186/s10086-020-01888-x
• 15)   R.  Higashi,  T.  Murakami,  D.  Maruoka and  E.  Kasai: ISIJ Int., 61 (2021), 1808. https://doi.org/10.2355/isijinternational.ISIJINT-2020-746
• 16)   H.  Kawamoto,  M.  Murayama and  S.  Saka: J. Wood Sci., 49 (2003), 469. https://doi.org/10.1007/s10086-002-0487-5
• 17)   T.  Shoji,  H.  Kawamoto and  S.  Saka: J. Anal. Appl. Pyrolysis, 109 (2014), 185. https://doi.org/10.1016/j.jaap.2014.06.014
• 18)   T.  Hosoya,  H.  Kawamoto and  S.  Saka: J. Anal. Appl. Pyrolysis, 83 (2008), 64. https://doi.org/10.1016/j.jaap.2008.06.008
• 19)   T.  Hosoya,  H.  Kawamoto and  S.  Saka: J. Anal. Appl. Pyrolysis, 85 (2009), 237. https://doi.org/10.1016/j.jaap.2008.11.028
• 20)   H.  Kawamoto: J. Wood Sci., 63 (2017), 63. https://doi.org/10.1007/s10086-016-1606-z
• 21)   L.  Xiao,  Z.  Lin,  W.  Peng,  T.  Yuan,  F.  Xu,  N.  Li,  Q.  Tao,  H.  Xiang and  R.  Sun: Sustain. Chem. Process., 2 (2014), Article number: 9. https://doi.org/10.1186/2043-7129-2-9
• 22)   P.  Giudicianni,  G.  Cardone,  G.  Sorrentino and  R.  Ragucci: J. Anal. Appl. Pyrolysis, 110 (2014), 138. https://doi.org/10.1016/j.jaap.2014.08.014
• 23)   P. E.  Marriott, L. D, Gómez and S. J. McQueen-Mason: New Phytol., 209 (2016), 1366. https://doi.org/10.1111/nph.13684
• 24)   H. D.  Baehr,  O.  Knacke,  O.  Kubaschewski and  K.  Hasselmann: Thermochemical properties of inorganic substances. Forsch Ing-Wes., 58 (1992), 103. https://doi.org/10.1007/BF02561491
• 25)   Y.  Takyu,  T  Murakami,  S. H.  Son and  E.  Kasai: ISIJ Int., 55 (2015), 1188. https://doi.org/10.2355/isijinternational.55.1188
• 26)   Y.  Kashiwaya,  S.  Nakaya and  K.  Ishii: Tetsu-to-Hagané, 77 (1991), 759 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.77.6_759