Behavior and Kinetic Mechanism Analysis of Dissolution of Iron Ore Particles in HIsmelt Process Based on High-temperature Confocal Microscopy

Abstract

In this study, high-temperature confocal microscopy (HTCM) was used to perform in situ observations of the dissolution of iron ore particles in slag at different temperatures. Moreover, the shrinking core model (SCM) is used to explain the kinetic mechanism of the dissolution of iron ore particles. The area of the undissolved fraction of iron ore particles was used to analyze the dissolution rate of iron ore particles. The results show that the kinetic mechanism of dissolution of iron ore particles can be well explained by the SCM. The dissolution of iron ore particles is controlled by the diffusion of iron ore fractions in the boundary layer. The dissolution temperature, the concentration difference of iron ore fraction at the two ends of the boundary layer, and the initial particle size of iron ore particles affect the dissolution rate of iron ore particles. The dissolution rate constant was fitted by introducing a dissolution mechanism. The activation energy of iron ore dissolved in the CaO–SiO₂–MgO–Al₂O₃–FeO slag system is 679.13 kJ/mol. The equation for the dissolution rate constant of iron ore in HIsmelt slag is summarized.

1. Introduction

Global warming is a severe global challenge that produces disasters such as climate change, sea level rise, starvation, and forest fires. The primary cause of global warming is rising CO₂ emissions from industrial production, with the steel industry accounting for over 5% of global CO₂ emissions.^1,2,3,4) Therefore, energy saving and emission reduction in the steel industry are imperative. In the conventional blast furnace route, the raw material handling process including pellet, sintering, and coking accounts for about 20% of the CO₂ emissions of the entire process. Researchers and engineers around the world have been trying to develop alternative processes to blast furnaces.^{5,6,7,8,9,10)} The HIsmelt ironmaking process is a new ironmaking process based on the principle of molten reduction. Iron ore fines and coal fines are blasted into the melt pool of the Smelting Reduction Vessel in this process, removing the raw material processing stage. As a result, HIsmelt reduces CO₂ emissions by 410 kg per ton of crude steel compared to the blast furnace route.^11,12)

Given the advantages of the HIsmelt process, scholars have conducted a series of studies on it. For example, Stephens¹³⁾ et al. performed kinetic simulations of the HIsmelt main reactor using computational fluid dynamics tools. Ma¹⁴⁾ et al. combined thermodynamic calculations and experimental analysis of the slag generation behavior of HIsmelt smelting vanadium-titanium magnetite. Theint¹⁵⁾ et al. used isothermal gravimetric analysis to investigate the gasification reactivity of carbonaceous materials in the molten reduction ironmaking process. Most of the previous studies on HIsmelt have focused on the simulation of its processes and the analysis of its material properties. However, studies on the dissolution behavior of the raw material after being blown into the melt pool have rarely been reported.

Since the dissolution of iron ore greatly affects the reaction kinetics of the HIsmelt smelting reduction. In addition, the study of the dissolution process has applications in many fields. In the study of steel materials, many studies have reported the dissolution behavior of inclusions or refractory materials in metallurgical slag for steel purity^{16,17,18,19,20)} or prevention of refractory erosion.^21,22,23) Different types of high-temperature confocal microscopes were used in these studies to make in situ observations of the dissolution mechanism of oxides.^{16,17,18,19,20,21,22,23,24,25,26)} And a dissolution model, the shrinkage core model (SCM) is widely used to analyze the kinetic mechanism of dissolution.^16,17,19,24) Inspired by the preceding, high-temperature confocal microscopy was used in this investigation to observe the dissolution process of iron ore particles in situ. The shrinkage core model was also utilized to investigate the dissolving kinetics of iron ore particles.

2. Materials and Methods

2.1. Preparation of Slag and Iron Ore Particles

Analytically pure SiO₂, CaO, Al₂O₃, MgO, and ferrous oxalate were mixed in a molybdenum crucible and heated to 1500°C using a tube furnace under argon protection and held for 1 hour to completely decompose the ferrous oxalate and completely melt the slag. The composition of the slag is shown in Table 1.

Table 1. Composition and content of slag.

CaO	SiO2	MgO	Al2O3	FeO
37.47	30.30	9.48	17.16	5.65

The iron ore particles used in the experiments were PB iron ore produced in Australia, which is widely used by steel companies. The composition is shown in Table 2. To verify that the iron ore did not melt due to the increase in temperature, the iron ore powder was heated to 1500°C using an argon-protected tube furnace and held for 1 hour. The samples before and after heating are shown in Fig. 1. The heated iron ore did not melt, but the color changed from red to black due to the decomposition reaction: 6 Fe₂O₃ → 4 Fe₃O₄ + O₂.

Table 2. Composition and content of iron ore.

Fe2O3	FeO	CaO	SiO2	MgO	Al2O3	MnO	V2O5	TiO2	P2O5
92.93	0.69	0.07	3.75	0.07	2.06	0.15	0.01	0.10	0.17

Fig. 1.

Comparison of iron ore before and after heating to 1500°C and holding for 1 hour (a) the sample before heating (b) the sample after heating.

2.2. In-situ Observation of Iron Ore Particles Dissolved in the Slag

In situ observation experiments of iron ore particles dissolved in slag were carried out using a high-temperature confocal microscope system. It consists of two main functional systems: the high-temperature stage (HTS) system and the microscopic imaging system, as illustrated in Fig. 2(a). The light from the halogen lamp is reflected by the gold-coated chamber onto the bottom surface of the platinum holder to heat the sample to the required temperature for the experiment. The warming and dissolving process of the iron ore particles was imaged by the microscope and recorded as a video by the monitor, as shown in Fig. 2(b).

Fig. 2.

Schematic view of (a) real product map of the high-temperature confocal microscope (b) Experimental principle of high-temperature confocal microscopy and (c) Experimental set-up for In-situ observation of dissolution experiments.

0.15 g broken slag was added into the Al₂O₃ crucible and heated to 1400°C under a high-temperature confocal microscope to fully melt. In the process of using a high-temperature confocal microscope, argon (flow rate of 200 cm³/min) was used as a protective gas to prevent slag oxidation. A Single iron ore particle was placed on the cooled pre-melted slag, and the Al₂O₃ crucible with iron ore particle and slag was placed on the holder of the HTCM as shown in Fig. 2(c). The experimental setup was heated to experimental temperature by a high-temperature confocal microscope in an argon atmosphere. To avoid the dissolution of the iron ore particles before heating to the target temperature, the sample was heated at a heating rate of 500°C/min. The heating rate was reduced to 300°C/min below the target temperature of 50°C to ensure that the samples were not overheated. After the samples reached the target temperature, the temperature was kept until the iron ore particles were completely dissolved. Taking the time when the sample was heated to the target temperature as the time zero. The temperature rise curve of the sample is shown in Fig. 3.

Fig. 3.

Temperature program settings for dissolution experiments using HTCM.

3. Results and Discussion

3.1. Analysis of the Behavior of Iron Ore Particles Dissolution

Figure 4 shows the dissolution process of iron ore particles in slag at 1350°C, 1400°C, and 1450°C respectively. Slag melted before iron ore particles change. There is no formation of product (ash) layer within the dissolving iron ore particles on its surface. The dissolution process was carried out for the same time, the higher the experimental temperature the more obvious the dissolution of iron ore particles. There are three phases to the breakdown of iron ore particles in slag. 1) slag melting, iron ore particles fell into the slag and were surrounded by slag; 2) iron ore particles began to split from the edge, and slag invaded the gap created by the split; 3) iron ore particles gradually dissolved, particle size decreased, and eventually fused with slag. The dissolving of iron ore particles is comprised of these three steps. Thermal expansion owing to temperature increase and the breakdown of Fe₂O₃ in iron ore particles cause the splitting of iron ore particles.

Fig. 4.

Images of the dissolution process of an iron ore particle at 1350°C, 1400°C and 1450°C during the HTCM experiments.

3.2. Modeling of the Kinetic Mechanism of Dissolution of Iron Ore Particles in Molten HIsmelt Slag

The shrinking core model (SCM) has been widely employed for the evaluation of the rate-controlling step of the dissolution process of a solid particle in a fluid. The SCM solution has two correlations between particle size and dissolution time with either linear dependence (reaction rate control) or parabolic dependence (boundary layer diffusion control).^16,19,27) Because there is no reaction between iron ore particles and slag components and no product layer formed during the dissolution process. The dissolving process can be divided into three steps: 1) The slag acts as a solvent, forming a boundary layer on the surface of the iron ore particles; 2) The soluble compounds in the iron ore particles dissolve into the boundary layer, forming solutes; and 3) The solute diffuses to the slag body in the boundary layer. The dissolution rate of component B in iron ore particles as shown in Eq. (1),   

$$\frac{-d{N}_{B(s)}}{dt}=\frac{d{N}_{(B)}}{dt}={\mathrm{\Omega }}_{l^{\prime }l}{J}_{B{l}^{\prime }}$$(1)

Where, N_B(s) is the mole number of solid B; N_(B) is the mole number of component B in slag; ${\mathrm{\Omega }}_{l^{\prime }l}$ is the interface area between the boundary layer and slag body, namely the outer surface area of the boundary layer; $J_{B{l}^{\prime }}$ is the absolute value of the diffusion rate of component B in the boundary layer.

At the unit time, the mole number of solute element B entering the slag through the unit boundary layer and slag interface area is as follows:   

$$J_{B{l}^{\prime }}=\left|J_{B{l}^{\prime }}\right|=|-D_{B{l}^{\prime }}\nabla c_{B{l}^{\prime }}|=D_{B{l}^{\prime }}\frac{\mathrm{\Delta }{c}_{B{l}^{\prime }}}{{\delta }_{l^{\prime }}}={D^{\prime }}_{B{l}^{\prime }}\mathrm{\Delta }{c}_{B{l}^{\prime }}$$(2)

Where ${D^{\prime }}_{B{l}^{\prime }}={\displaystyle \frac{D_{B{l}^{\prime }}}{{\delta }_{l^{\prime }}}}$; $\mathrm{\Delta }{c}_{B{l}^{\prime }}=c_{B{l}^{\prime }s}-c_{B{l}^{\prime }l}=c_{Bs}-c_{Bl}$; $D_{B{l}^{\prime }}$ is the diffusion coefficient of B in the boundary layer; $\nabla c_{B{l}^{\prime }}$ is the concentration gradient of component B in the boundary layer; ${\delta }_{l^{\prime }}$ is the boundary layer thickness. Assuming that the boundary layer area decreases but the thickness remains unchanged during the dissolution process; $c_{B{l}^{\prime }s}$ is the concentration of component B at the interface between solid and boundary layer, namely the concentration of component B in solid, so $c_{B{l}^{\prime }s}=c_{Bs}$. $c_{B{l}^{\prime }l}$ is the concentration of B in the boundary layer and slag interface, that is, the concentration of B in slag, so $c_{B{l}^{\prime }l}=c_{Bl}$. The mechanism of the SCM as shown in Fig. 5.

Fig. 5.

The SCM of iron ore particles dissolving in slag.

Thus, the dissolution rate of component B in iron ore particles as shown in Eq. (3). Assuming that the iron ore particles are approximately spherical during the dissolution process, the dissolution rate of component B in a spherical particle with radius r can be expressed as Eq. (4).   

$$\frac{-d{N}_{B(s)}}{dt}={\mathrm{\Omega }}_{l^{\prime }l}{J}_{B{l}^{\prime }}={\mathrm{\Omega }}_{l^{\prime }l}{D^{\prime }}_{B{l}^{\prime }}\mathrm{\Delta }{c}_{B{l}^{\prime }}$$(3)

  

$$\frac{-d{N}_{B(s)}}{dt}=4\pi r^2{D^{\prime }}_{B{l}^{\prime }}\mathrm{\Delta }{c}_{B{l}^{\prime }}$$(4)

For spherical iron ore particles with radius r, the molar number of component B can be shown as Eq. (5). Thus, the reduction rate of iron ore particle size r can be expressed as Eq. (6).   

$$N_{B(s)}=\frac{4}{3}\pi r^3\frac{{\rho }_B}{M_B}$$(5)

  

$$-\frac{dr}{dt}=\frac{M_B{D^{\prime }}_{B{l}^{\prime }}}{{\rho }_B}\mathrm{\Delta }{c}_{B{l}^{\prime }}$$(6)

Where, ρ_B is the density of component B in the particle; M_B is the molar mass of component B. The ratio of the amount of dissolved component B to the amount of initial component B is the conversion rate of component B.   

$$\begin{array}{l}{\alpha }_B=\frac{N_{B0}-N_{B(s)}}{N_{B0}}=\frac{\frac{4}{3}\pi r_0{}^3{\rho }_B/M_B-\frac{4}{3}\pi r^3{\rho }_B/M_B}{\frac{4}{3}\pi r_0{}^3{\rho }_B/M_B}\\ \hspace{1em}\hspace{0.5em}=1-{\left(\frac{r}{r_0}\right)}^3\end{array}$$(7)

The relationship between the conversion of component B and time can be obtained from Eqs. (6) and (7) as follows:   

$$1-{(1-{\alpha }_B)}^{\frac{1}{3}}=\frac{M_B{D^{\prime }}_{B{l}^{\prime }}\mathrm{\Delta }{c}_{B{l}^{\prime }}}{{\rho }_B{r}_0}t$$(8)

The particle mass is less than 0.1% of the slag mass. It can be considered that the dissolution of iron ore particles does not affect the slag composition. Thus, $\mathrm{\Delta }{c}_{B{l}^{\prime }}=c_{B{l}^{\prime }s}-c_{B{l}^{\prime }l}=c_{Bs}-c_{Bl}$ is a fixed value. The dissolution mechanism equation $G({\alpha }_B)=1-{(1-{\alpha }_B)}^{\frac{1}{3}}$ and the dissolution rate constant $k=\frac{M_B{D^{\prime }}_{B{l}^{\prime }}\mathrm{\Delta }{c}_{B{l}^{\prime }}}{{\rho }_B{r}_0}$ are introduced. The relationship between them is shown as follows.   

$$G({\alpha }_B)=kt$$(9)

Every other time, the video of dissolving particles was captured using the video processing software. The pixel number of the undissolved component of the particles in the image is treated as the cross-sectional area of the undissolved part, as illustrated in Fig. 6. The cross-sectional area of undissolved particles at time zero is expressed as S₀, and the cross-sectional area of undissolved particles at time t is expressed as S. Normalized dissolution patterns of iron ore particles in slag at 1350°C, 1400°C, and 1450°C are summarized in Fig. 7.

Fig. 6.

Images of the dissolution process of a particle at 1450°C during the HTCM experiments. (a) t = 6 s; (b) t = 10 s.

Fig. 7.

Normalized dissolution patterns of iron ore particles in slag at 1350°C, 1400°C, and 1450°C.

Thus, the conversion of component B and the mechanism function can be expressed by the cross-sectional area of particles.   

$${\alpha }_B=1-{\left(\frac{S}{S_0}\right)}^{\frac{3}{2}}$$(10)

  

$$G({\alpha }_B)=1-{\left(\frac{S}{S_0}\right)}^{\frac{1}{2}}=kt$$(11)

As shown in Fig. 8, there is a good linear relationship between G(α_B) and t, which proves that the SCM can be used to explain the kinetic mechanism of dissolution of iron ore particles in the HIsmelt slag. The dissolution rate constants k at 1350°C, 1400°C, and 1450°C were fitted to be 0.0025 s⁻¹, 0.0091 s⁻¹, and 0.0466 s⁻¹, respectively. So, the dissolution rate constant of particles is positively correlated with reaction temperature.

Fig. 8.

Relationship between mechanism function and time of dissolution of iron ore particles in slag at 1350°C, 1400°C and 1450°C.

According to the Arrhenius equation shown as follows,²⁸⁾ the dissolution rate constant k has a functional relationship with the activation energy and pre-exponential factor of the reaction.   

$$k=A\cdot exp(-Ea/RT)$$(12)

Where, k is the dissolution rate constant, R is the molar gas constant, T is the reaction temperature, Ea is the activation energy and A is the pre-exponential factor, respectively. Equation (12) was converted to be logarithmic on both sides to get Eq. (13) as follows. lnk and −1/T are linear, the slope of the line is Ea/R, and the intercept is lnA. Therefore, Ea and A can be calculated according to the k value measured at different temperatures.   

$$ln\mathrm{\hspace{0.17em} }k=-\frac{Ea}{RT}+ln\mathrm{\hspace{0.17em} }A$$(13)

The relationship between lnk and 1/T is shown in Fig. 9. The activation energy Ea of iron ore particles dissolved in HIsmelt slag obtained by fitting is 679.13 kJ/mol and pre-exponential factor A is 1.68 × 10¹⁹ s⁻¹.

Fig. 9.

The fitting of kinetic parameters of iron ore particle dissolution in slag.

The relationship between the dissolution rate constant of iron ore powder dissolved in HIsmelt slag and temperature is shown in Eq. (14).   

$$k=1.68\times {10}^{19}\times exp(-81\mathrm{\hspace{0.17em} }685/T)$$(14)

The S/S₀ changes with time are shown in Fig. 10(a). The dissolution rate constant and the complete dissolution time of iron ore particles at different temperatures are shown in Fig. 10(b). The dissolution rate constant increases with increasing temperature, and the increasing rate is temperature-dependent. The entire dissolution time of ore powder reduces with increasing temperature; however, the rate of decrease is negatively associated with temperature. At higher temperatures, increasing the temperature has no discernible influence on the total dissolution time.

Fig. 10.

Dissolution status, dissolution rate constant, and complete dissolution time of iron ore particles at different temperatures (a) S/S0 changes with time (b) Dissolution rate constant and complete dissolution time versus temperature.

As shown in Eq. (8), the dissolution rate constant k in iron ore particles is linearly related to the gradient diffusion coefficient ${D^{\prime }}_{B{l}^{\prime }}$, because temperature affects the dissolution rate of ore powder by changing the diffusion rate of components in the boundary layer. ${D^{\prime }}_{B{l}^{\prime }}$ is only related to temperature, as shown in Eq. (15). The initial radius of iron ore powder used in this experiment is 0.5 mm, ignoring the change of slag composition caused by the dissolution of iron ore particles, so $\mathrm{\Delta }{c}_{B{l}^{\prime }}=c_{Bs}=\frac{{\rho }_B}{M_B}$, ${D^{\prime }}_{B{l}^{\prime }}$ as shown in Eq. (16), unit m·s⁻¹.   

$${D^{\prime }}_{B{l}^{\prime }}=\frac{{\rho }_B{r}_0}{M_B\mathrm{\Delta }{c}_{B{l}^{\prime }}}\cdot k=\frac{{\rho }_B{r}_0}{M_B\mathrm{\Delta }{c}_{B{l}^{\prime }}}\cdot A\cdot exp(-Ea/RT)$$(15)

  

$${D^{\prime }}_{B{l}^{\prime }}=k\cdot r_0=8.4\times {10}^{15}\times exp(-81\mathrm{\hspace{0.17em} }685/T)$$(16)

The initial radius of ore powder is changed to $r_0^{\prime }$, and the ore powder is not assumed to be a pure substance. The concentration difference of component B in the boundary layer is set to $a\cdot \frac{{\rho }_B}{M_B}$, so the general equation of the dissolution rate constant of iron ore powder in HIsmelt slag is as follows:   

$$k=\frac{a{D^{\prime }}_{B{l}^{\prime }}}{r_0^{\prime }}=\frac{a}{r_0^{\prime }}\cdot 8.4\times {10}^{15}\times exp(-81\mathrm{\hspace{0.17em} }685/T)$$(17)

Equation (17) is compared with Eqs. (12) and (14). The pre-exponential factor is related to the initial particle size of iron ore particles and the concentration difference at both ends of the boundary layer. The particle size of ore powder was negatively correlated with the dissolution rate constant, and the concentration difference was positively correlated with the dissolution rate constant. The dissolution rate of iron ore powder can be improved by reducing the particle size of ore powder and the content of iron oxide in slag.

4. Conclusions

To clarify the kinetic mechanism of iron ore dissolution in the HIsmelt process, in situ observation experiments on the dissolution behavior of iron ore particles were conducted at 1350°C, 1400°C, and 1450°C using high-temperature confocal microscopy. The following conclusions were obtained.

(1) Iron ore does not melt below 1500°C in the absence of slag. The presence of slag is required for the molten iron ore reduction reaction to take place. Iron ore particles become loose after being heated in the slag. The iron ore particles are subsequently surrounded by slag, which dissolves them.

(2) The kinetic mechanism of iron ore dissolution, which is governed by the diffusion of ore components in the boundary layer, can be explained by the shrinking core model. The mechanistic function of dissolution is summarized as $G({\alpha }_B)=1-{(1-{\alpha }_B)}^{\frac{1}{3}}$. The dissolution rate constant is derived as $k=\frac{M_B{D^{\prime }}_{B{l}^{\prime }}\mathrm{\Delta }{c}_{B{l}^{\prime }}}{{\rho }_B{r}_0}$.

(3) According to the SCM, the activation energy of 1 mm diameter iron ore particles dissolved in slag is 679.13 kJ/mol, and the pre-exponential factor is 1.68 × 10¹⁹ s⁻¹. The dissolution rate of iron ore particles is proportional to the dissolution temperature. The equation for the dissolution rate constant of iron ore in HIsmelt slag is summarized as $k=\frac{a{D^{\prime }}_{B{l}^{\prime }}}{r_0^{\prime }}=\frac{a}{r_0^{\prime }}\cdot 8.4\times {10}^{15}\times exp(-81\mathrm{\hspace{0.17em} }685/T)$.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

CRediT Authorship Contribution Statement

Jing Pang: Conceptualization, Methodology, Software, Data curation, Writing-original draft. Zhenyang Wang: Conceptualization, Writing – review & editing. Jianliang Zhang: Supervision, Writing – review & editing. Shushi Zhang: Writing – review & editing. Peng Hu: Visualization, Investigation. Jiating Rao: Visualization, Investigation.

Acknowledgements

This work was supported by the China Postdoctoral Science Foundation (Project No.: BX20200045 and 2021M690370) and the State Key Laboratory of Vanadium and Titanium Resources Comprehensive Utilization (Project No.: 2020P4FZG06A).

References

• 1)   C. D.  Thomas,  A.  Cameron,  R. E.  Green,  M.  Bakkenes,  L. J.  Beaumont,  Y. C.  Collingham,  B. F. N.  Erasmus,  M. F.  de Siqueira,  A.  Grainger,  L.  Hannah,  L.  Hughes,  B.  Huntley,  A. S.  van Jaarsveld, et al.: Nature, 427 (2004), 145. https://doi.org/10.1038/nature02121
• 2)   O.  Hoegh-Guldberg,  P. J.  Mumby,  A. J.  Hooten,  R. S.  Steneck,  P.  Greenfield,  E.  Gomez,  C. D.  Harvell,  P. F.  Sale,  A. J.  Edwards,  K.  Caldeira,  N.  Knowlton,  C. M.  Eakin,  R.  Iglesias-Prieto,  N.  Muthiga, et al.: Science, 318 (2007), 1737. https://doi.org/10.1126/science.1152509
• 3)   M.  Meinshausen,  N.  Meinshausen,  W.  Hare,  S. C. B.  Raper,  K.  Frieler,  R.  Knutti,  D. J.  Frame and  M. R.  Allen: Nature, 458 (2009), 1158. https://doi.org/10.1038/nature08017
• 4)   M. T.  Ho,  A.  Bustamante and  D. E.  Wiley: Int. J. Greenh. Gas Control, 19 (2013), 145. https://doi.org/10.1016/j.ijggc.2013.08.003
• 5)   R.  Sarkar and  H. Y.  Sohn: Metall. Mater. Trans. B, 49 (2018), 1860. https://doi.org/10.1007/s11663-018-1300-1
• 6)   Q. F.  Hou,  M.  Samman,  J.  Li and  A. B.  Yu: ISIJ Int., 54 (2014), 1772. https://doi.org/10.2355/isijinternational.54.1772
• 7)   R. J.  Fruehan: Ironmaking Steelmaking, 32 (2005), 3. https://doi.org/10.1179/irs.2005.32.1.3
• 8)   B.  Anameric and  S. K.  Kawatra: Miner. Process. Extr. Metall. Rev., 28 (2007), 59. https://doi.org/10.1080/08827500600835576
• 9)   I. F.  Kurunov: Metallurgist, 54 (2010), 335. https://doi.org/10.1007/s11015-010-9300-7
• 10)   M. A.  Quader,  S.  Ahmed,  S. Z.  Dawal and  Y.  Nukman: Renew. Sustain. Energy Rev., 55 (2016), 537. https://doi.org/10.1016/j.rser.2015.10.101
• 11)   J. L.  Schenk: Particuology, 9 (2011), 14. https://doi.org/10.1016/j.partic.2010.08.011
• 12)   A.  Orth,  N.  Anastasijevic and  H.  Eichberger: Miner. Eng., 20 (2007), 854. https://doi.org/10.1016/j.mineng.2007.02.007
• 13)   D.  Stephens,  M.  Tabib,  M. P.  Schwarz and  M.  Davis: Prog. Comput. Fluid Dyn., 12 (2012), 196. https://doi.org/10.1504/pcfd.2012.047462
• 14)   H. B.  Ma,  K. X.  Jiao and  J. L.  Zhang: CrystEngComm, 22 (2020), 361. https://doi.org/10.1039/c9ce01695c
• 15)   T. T.  Htet,  Z. M.  Yan,  S.  Spooner,  V.  Degirmenci,  K.  Meijer and  Z. S.  Li: Fuel, 289 (2021), 119890. https://doi.org/10.1016/j.fuel.2020.119890
• 16)   S.  Michelic,  J.  Goriupp,  S.  Feichtinger,  Y. B.  Kang,  C.  Bernhard and  J.  Schenk: Steel Res. Int., 87 (2016), 57. https://doi.org/10.1002/srin.201500102
• 17)   S.  Feichtinger,  S. K.  Michelic,  Y. B.  Kang and  C.  Bernhard: J. Am. Ceram. Soc., 97 (2014), 316. https://doi.org/10.1111/jace.12665
• 18)   S. K.  Michelic and  C.  Bernhard: Scanning, 2017 (2017), 2326750. https://doi.org/10.1155/2017/2326750
• 19)   M.  Sharma,  W. Z.  Mu and  N.  Dogan: JOM, 70 (2018), 1220. https://doi.org/10.1007/s11837-018-2908-y
• 20)   C. J.  Xuan and  W. Z.  Mu: J. Am. Ceram. Soc., 102 (2019), 6480. https://doi.org/10.1111/jace.16509
• 21)   P. T.  Jones,  D.  Desmet,  M. X.  Guo,  D.  Durinck,  F.  Verhaeghe,  J.  Van Dyck,  J. H.  Liu,  B.  Blanpain and  P.  Wollants: J. Eur. Ceram. Soc., 27 (2007), 3497. https://doi.org/10.1016/j.jeurceramsoc.2007.01.022
• 22)   Y.  Hino,  H.  Matsunaga and  K.  Watanabe: ISIJ Int., 57 (2017), 697. https://doi.org/10.2355/isijinternational.ISIJINT-2016-583
• 23)   S. J.  Nam,  Y. B.  Kang,  S. M.  Jung and  Y.  Sasaki: ISIJ Int., 53 (2013), 1779. https://doi.org/10.2355/isijinternational.53.1779
• 24)   M.  Valdez,  K.  Prapakorn,  A. W.  Cramb and  S.  Sridhar: Ironmaking Steelmaking, 29 (2002), 47. https://doi.org/10.1179/030192302225001965
• 25)   Y.  Wang,  M.  Valdez and  S.  Sridhar: Metall. Mater. Trans. B, 33 (2002), 625. https://doi.org/10.1007/s11663-002-0042-1
• 26)   C. J.  Xuan and  W. Z.  Mu: J. Alloy. Compd., 834 (2020), 155168. https://doi.org/10.1016/j.jallcom.2020.155168
• 27)   Y.  Lee,  J. K.  Yang,  D. J.  Min and  J. H.  Park: Ceram. Int., 45 (2019), 20251. https://doi.org/10.1016/j.ceramint.2019.06.298
• 28)   K. J.  Laidler: J. Chem. Educ., 61 (1984), 494. https://doi.org/10.1021/ed061p494