Effect of CO₂ Content in Quicklime on Dissolution Rate of Quicklime in Steelmaking Slags

Abstract

The dissolution rate of lime in the molten slag is important for the efficient of steelmaking reactions. The dissolution rates of quicklime were conventionally measured by a rotating cylinder method, and they were quite lower compared with the estimated rates from actual steelmaking operations. Previously, the authors reported that the quicklime used in the actual operation had a much faster dissolution rate than completely calcined lime. During the dissolution of quicklime used in the actual operation, quicklime emits CO₂ gas twice, and the second gas formation effectively enhances the dissolution rate. Though the dissolution rates of quicklime with a CO₂ content of 0, 2, 4, and 9 mass% had been analyzed, the dissolution rates were scattered. The reason for this scattering of the data was that the CO₂ content of individual quicklime samples varied significantly within the same grade of quicklime, because the samples used in the previous study were produced by a rotary kiln process. Consequently, the dissolution rates were inconclusive, and the effect of the CO₂ content in quicklime on the dissolution rate of quicklime could not be fully clarified. In this study, the CO₂ content was controlled through the laboratory-based preparation of spherical quicklime samples and thus, the effect of the CO₂ content on the dissolution rate of quicklime in the molten slag could be precisely analyzed. Eventually, this approach allowed to propose the dissolution rate of quicklime with gas formation due to the thermal decomposition of the CaCO₃ core existing in the center of quicklime samples.

1. Introduction

In steelmaking processes, a CaO-based slag is used to remove impurities from steel. Various solid oxides, such as quicklime and dolomite, are used for producing the slag. The dissolution rates of these oxides in the molten slag play an important role for increasing the reaction rate between the steel and slag, decreasing the amount of flux, decreasing the energy consumptions, and reducing slag emission. Though the rate of the lime dissolution affects the rate for removing phosphorus from hot metal,^{1,2,3,4,5,6,7,8,9)} a part of the lime added as a flux still remains without dissolution after the refining. Therefore, it is required to increase the dissolution rate of quicklime.

Many researchers have measured the dissolution rates of various solid oxides.^{10,11,12,13,14,15,16,17)} Matsushima et al. determined the dissolution rate of lime in the molten slag by recording the decrease in diameter of an immersed and rotating sintered lime rod.¹⁰⁾ They reported that the dissolution rates increased with the increases in each, the relative velocity between lime and the molten slag, the difference of the CaO content between the slag and the interfacial phase, and in the experimental temperature. These experiments, however, were carried out using densely sintered rods, whereas oxides materials used as industrial fluxes were usually porous, fragile, and non-uniform in structure. In consequence, the rates measured in their study was slower compared to the rates estimated from the actual operation.

As reported in previous studies, the authors have developed an experimental technique for measuring the dissolution rate of porous or fragile oxides in molten slag.^{18,19,20,21,22,23)} In this technique, the dissolution rate is determined by the compositional change of the molten slag after the addition of flux particles. This technique was applied to CaO dissolution.¹⁸⁾ The results showed that the dissolution rates of dense and porous lime prepared in the laboratory in a CaO–SiO₂–FeO system slag were comparable when a dicalcium silicate (C₂S) layer was formed around the lime particles, though the dissolution rate of porous materials was generally higher than that of dense materials. Interestingly, the dissolution rate of porous lime was faster than that of the dense lime in the case of dissolution in a CaO-B₂O₃-FeO-system slag. In this case, the no-intermediate layer formed between lime and the slag. Consequently, the C₂S layer inhibited the dissolution of lime in the molten slag. Guo et al. also reported that the dissolution rate of lime in a CaO–Al₂O₃–SiO₂ system slag was limited when an interfacial layer was formed between lime and the slag, which they could observe in-situ with a confocal scanning laser microscope.^24,25,26) Notably, the dissolution rate of quicklime used in the actual steelmaking operation is much higher than that of porous and dense lime prepared in a laboratory.^19,27) An in-situ observation revealed that the quicklime used in the actual operation dissolved in the molten slag with gas formation.^19,28,29,30) This gas formation is caused by the thermal decomposition of the CaCO₃ core that exists in the center of most quicklime. The reason is that quicklime is produced from limestone (CaCO₃) particles by the calcination and mainly consists of CaO. The calcination reaction starts from the surface of a limestone particle and takes place at the definite boundary between the CaO and CaCO₃ phases.³¹⁾ Therefore, when the calcination reaction is insufficient, a CaCO₃ core remains in the central part of the quicklime particle. Subsequently, when such quicklime particle is put into the molten slag and heated, the CaCO₃ core emits CO₂ gas due to the thermal decomposition.

According to the results of the author’s earlier experiments,^27,28,29,30) the dissolution rate is enhanced during the gas emission. This result indicates that the existence of a CaCO₃ core is important for the rapid dissolution of quicklime. The authors measured the CO₂ content in each quicklime particle; and it was revealed that the CO₂ content of each particle were distributed widely due to the fact that the quicklime used in this study was prepared industrially. Therefore, the actual CO₂ content of individual quicklime particles used in each dissolution test was unclear, though the average CO₂ content was given. Consequently, the purpose of the current study is to clarify the effect of the CO₂ content on the dissolution rate of quicklime by using lab-produced quicklime particles with a well-defined CO₂ content.

For the experiments, limestone particle is shaped in sphere first, and then it is calcined in a controlled way to produce samples with a particular amount of CO₂. Subsequently, the dissolution rate and foaming phenomena of quicklime samples with well-defined CO₂ contents were investigated.

2. Experimental Procedure

The method employed in this study to measure the dissolution rate of quicklime in the molten slag is similar to that used in previous studies.^{18,19,20,21,22,27,30,32)} Figure 1 shows a schematic diagram of the experimental apparatus for measuring the dissolution rate and observing the dissolution behavior of spherical quicklime particles. A mullite tube (Φ50 mm O.D. × Φ42 mm I.D. × 1000 mm) furnace with SiC spiral heating element was used. The dissolution rate of spherical quicklime particles was measured according to the following procedure: Synthesized slag (100 g) was melted in an iron crucible with an inner diameter of 35 mm at 1673 K. After the slag had been melted, Ar gas was introduced into the slag for stirring with the flowrate of 50 mL/min. One of the spherical quicklime sample particles (kept at room temperature) was dropped into the molten slag from the top of the furnace. The slag was sampled using a molybdenum sampling tool at pre-defined time intervals and quenched. The composition of the slag samples was analyzed by X-ray fluorescence (XRF) analysis using the 1:10 dilution glass beads method after fine grinding of the samples. Dissolution rates were determined based on the variation of the CaO content in the molten slag. During the dissolution experiments, the dissolution behavior was recorded as a movie by a digital camera through the mirror mounted on the top of the furnace. The recorded movie was used to analyze the gas formation behavior of the spherical quicklime particles.

Fig. 1.

Schematic diagram of the experimental setup.

The same system slag was used as in the previous study,³⁰⁾ in which the initial slag composition was set to 20%CaO-40%FeO-40%SiO₂. The slag was prepared by melting a mixture of reagent grade oxide powders in an iron crucible at 1673 K under Ar atmosphere for one hour.

The CO₂ content of individual spherical quicklime particles was determined by the following method. First, a spherical limestone particle with a diameter of about 13 mm was made from a natural limestone lump by grinding. The weight of spherical limestone particles were about 3.0 g. Next, the spherical limestone sample was calcined at 1173 K in air atmosphere for a specified time. Figure 2 shows a photograph of a prepared spherical limestone and a spherical quicklime sample. The color of the sample changed from gray to white after calcination. The changes of the sample diameter due to the calcination was negligibly small, i.e. the density of sample decreased by the calcination. The weight of sample decreased because of the thermal decomposition of CaCO₃. The ratio of the weight loss could be controlled by changing the duration of the calcination. The CO₂ content in a spherical quicklime particle was determined by measuring the weights before and after the calcination, because CaO in natural limestone exists in the form of CaCO₃. The CO₂ content in the quicklime samples prepared in this study was in the range from 0.1 to 25.0 mass% to investigate the effect of the CO₂ content on the dissolution rate of quicklime in the molten slag. Because the limestone used in this study was a naturally occurring material, some impurities might be contained in the limestone, and it should be noted that the samples with 0.1 and 0.2 mass% CO₂ corresponded to completely calcined quicklime particles (i.e. CO₂=0.0 mass%).

Fig. 2.

Photographs of a spherical limestone (a) and a spherical quicklime sample (b).

The radius of the CaCO₃ core and the thickness of the decomposed CaO layer in the quicklime particles was estimated based on the assumption that the thermal decomposition reaction proceeds uniformly from the surface to the inside of the limestone sample, as illustrated in Fig. 3. The weight of the spherical quicklime particles ranged from 1.66 to 2.36 g, depending on their CO₂ contents. When the spherical quicklime particles were completely dissolved, the CaO content in the slag stoichiometry increased by 1.7 mass%.

Fig. 3.

Radiuses of the undecomposed CaCO₃ core and thickness of the decomposed CaO layer in spherical quicklime samples with a varied CO₂ content.

3. Experimental Results

3.1. Observation of the Quicklime Dissolution Behavior

The dissolution behavior of spherical quicklime samples in the molten slag was observed. The dissolution of spherical quicklime samples showed similar behavior to those occurring during the dissolution of quicklime used in the actual operation reported in the previous studies.^28,29,30) During the initial period, the gas injection tube, edge of the iron crucible, and injected Ar bubbles on the slag surface were observed. After a spherical quicklime sample was immersed, bubble formation around the quicklime particle started immediately and slag foams could be observed, referred to as “first foaming”. Several seconds later, the first foaming finished and only Ar bubbles were observed as same as before the immersion of any quicklime. Several more seconds later, gas formation set in again and slag foams were observed for a second time, referred to as “second foaming”. The second foaming is much stronger than the first foaming; and the edge of the iron crucible disappeared from the view area because the slag overflowed from the iron crucible due to the generation of a large number of bubbles in the slag phase. Again, several seconds later, the second foaming finished and the edge of the iron crucible was observed again as well as injected Ar bubbles on the slag surface. These observations revealed that the dissolution of spherical quicklime samples proceeds with two occurrences of gas emission and slag foams due to the gas formation from the immersed quicklime samples.

The start and end times of the first and second foaming were analyzed from the recorded movie. Figure 4 shows the effect of the CO₂ content in the spherical quicklime samples on the start and end times of the first and second foaming. In all samples, the first foaming starts immediately when the quicklime is immersed into the slag. The first foaming stops from 11 to 27 s after it set in. There seemed to be no relationship between the CO₂ content of the spherical quicklime samples and the end time of the first foaming. The start time of the second foaming is shortened with increasing CO₂ content of the spherical quicklime particles. Further, the end time of the second foaming is delayer with increasing CO₂ content. In cases where the CO₂ content was lower than 1.48 mass%, the second foaming was not observed.

Fig. 4.

The effect of the CO₂ content in spherical quicklime samples on the start and end times of the first and second foaming.

3.2. Variation of the CaO Content in Slag

Figure 5 shows the variation of the CaO content in the molten slag with time after immersing a spherical quicklime sample in the molten slag. The CaO content increases with time and reaches about 22% after a lapse of 3 min in the cases that CO₂ content is more than 2.3 mass%. In these cases, almost all the quicklime might become part of the slag. The CaO content in some sampled slag was specifically large compared with values of the previously and next sampled slag. In the electron probe micro analysis (EPMA) of these samples, CaO-rich phases, such as dicalcium silicate or calico-wustite, were observed, though a uniform phase is observed in ordinary sampled slag.³⁰⁾ It is considered that these CaO-rich phases were formed when undissolved quicklime came into contact with the sampled slag and thus, the slag with the CaO-rich phases was collected incidentally. In the following analysis, the dissolution rates were evaluated by neglecting the sampled slag with a specifically large CaO content. The initial CaO content in the slag showed variations from 19.6 to 20.5 mass% in each experiment, though the initial CaO content was set to be 20.0 mass%. For a better understanding of the dissolution behavior of quicklime in steelmaking slags, the increment of the CaO content in the slag, which is defined as the difference in CaO content between the initial and the observed state, was analyzed.   

$$\mathrm{\Delta }{\left(\%\text{CaO}\right)}_t={\left(\%\text{CaO}\right)}_t-{\left(\%\text{CaO}\right)}_{t=0}$$(1)

where, (%CaO) is the CaO content in the slag [mass%] and the subscript t denotes the time [s]. Figure 6 shows the variations of the increment of the CaO content in the slag. The CaO increment increases with time and in turn, the rate of the CaO increment is enhanced with increasing CO₂ content. By comparing the duration of second foaming as shown in Fig. 4, it can be seen that the increasing rate of CaO content in slag is clearly enhanced during second foaming.

Fig. 5.

Variation of the CaO content in the molten slag as a function of elapsed time after the immersion of a spherical quicklime particle in the molten slag.

Fig. 6.

Variations of increment of the CaO content in the slag.

4. Discussion

4.1. Duration of the First Foaming and Second Foaming

Figure 7 shows the relation between the CO₂ content in spherical quicklime samples and the duration of the first and second foaming, which is the difference between the start and end times of either foaming. Though the duration of the first foaming decreased slightly with increasing CO₂ content, the effect of the CO₂ content on the duration of the first foaming was hardly pronounced. The first foaming is caused by the thermal decomposition of CaCO₃ or Ca(OH)₂ existing in the outer layer of quicklime particles.^28,29,30) This layer is formed after the calcination of limestone. Therefore, the extent of this layer and thus, the degree it can contribute to the first foaming depends on the storage conditions of the quicklime samples after calcination. In this study, the spherical quicklime particles were weighed under the natural convection suppressed condition around the sample after calcination, still hot from the furnace, and then immediately cooled down under vacuum conditions and subsequently stored under the same condition. Though the samples were taken out just before the dissolution experiment and used immediately, the first foaming was observed. This fact indicates that the rate of the carbonation or hydroxylation reaction in quicklime is fast. It is desirable that the extent of the CaCO₃ or Ca(OH)₂ layer around the quicklime sample is as small as possible because these thermal decomposition reactions are endothermic reactions.

Fig. 7.

Effect of the CO₂ content in spherical quicklime samples on the duration of the first and second foaming.

On the contrary, the duration of the second foaming was clearly proportional to the CO₂ content in the spherical quicklime samples. The second foaming is caused by the thermal decomposition of the CaCO₃ core existing in the center part of the quicklime samples.³⁰⁾ The value of the CO₂ content corresponds to the CO₂ present in the core, because the CO₂ content was determined when the spherical quicklime particles were still hot. The thermal decomposition of CaCO₃ is an endothermic reaction. It is considered that the heat transfer from the slag to the CaCO₃ core is the rate-controlling step.

Figure 8 shows the relation between the start time of the second foaming and the thickness of decomposed CaO layer as shown in Fig. 3. The start time of the second foaming increased with the thickness of the CaO layer. This result indicates that the heat transfer in the CaO layer around the CaCO₃ core is the rate-controlling step for CO₂ formation until second foaming starts.

Fig. 8.

Relation between the thickness of decomposed CaO layer and onset of the second foaming.

4.2. Dissolution Rate of Spherical Quicklime Particles

The dissolved weight of a spherical quicklime particle during from t–Δt to t, W_lime,t−W_lime,t−Δt, can be calculated by the following mass balance equations:^19,22,30)   

$$\begin{array}{l}{W}_{\text{salg, }t}\times {\left(\%\text{CaO}\right)}_{\text{slag, }t}=W_{\text{salg, }t-\mathrm{\Delta }t}\times {\left(\%\text{CaO}\right)}_{\text{slag, }t-\mathrm{\Delta }t}\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+\left(W_{\text{lime, }t-\mathrm{\Delta }t}-W_{\text{lime, }t}\right)\times {\left(\%\text{CaO}\right)}_{\text{ lime}}\end{array}$$(2)

  

$$W_{\text{salg, }t}=W_{\text{slag, }t-\mathrm{\Delta }t}+\left(W_{\text{lime, }t-\mathrm{\Delta }t}-W_{\text{lime, }t}\right)-W_{\text{sampled}}$$(3)

where W_slag,t is the slag weight present in the crucible at the sampling time t, W_lime,t is the weight of the undissolved quicklime at sampling time t, and W_sampled is the averaged sampled weight of the slag at each sampling, which is assumed to be 3.0 g. These relations allow to calculate the slag weight present in the crucible at the sampling time. The average dissolution rate of quicklime from the initial to a steady state is defined as given in the following equation:   

$${\left(\frac{\text{d}{r}_{\text{lime}}}{\text{d}t}\right)}_{\text{ave}}=-\frac{W_{\text{lime,0}}-W_{\text{lime,end}}}{t_{\text{end}}-t_{\text{0}}}\frac{1}{{\rho }_{\text{lime}}{A}_{ave}}\hspace{1em}\left[\mathrm{c}\mathrm{m}/\mathrm{m}\mathrm{i}\mathrm{n}\right]$$(4)

where r is the radius of a quicklime sample. Subscripts 0 and std denote the initial value (i.e., t=0) and the value from the last sample of each experimental condition, respectively. A_ave is the average surface area of the initial and end state. Figure 9 shows the influence of the CO₂ content in spherical quicklime samples on the average dissolution rates in the molten slag. The average dissolution rate increased with the CO₂ content in spherical quicklime particles when the CO₂ content was lower than approximately 12 mass%. The average dissolution rate scattered and then decreased for CO₂ contents over 12%. In these cases, a reduction of the brightness of the slag around the quicklime sample was observed during the dissolution test. This phenomenon indicates that the temperature around the quicklime particle decreased because of the thermal decomposition of the CaCO₃ core. This result shows that the suitable CO₂ content in quicklime is 12% under the experimental condition that the diameter of spherical quicklime particle is 13 mm. The average dissolution rate of 12%-CO₂ quicklime was eight times higher than the average dissolution rates of quicklime samples with a CO₂ content of less than 0.2 mass%, whose rates correspond to the base rate of quicklime dissolution in which no second foaming is observed.

Fig. 9.

Influence of the CO₂ content of spherical quicklime samples on the average dissolution rate in the molten slag.

Next, the average dissolution rate of quicklime during the second foaming was calculated based on Eq. (4). Figure 10 shows the dissolution rates of the spherical quicklime particles with varying CO₂ contents during the second foaming. The rates during the second foaming increased with the increasing CO₂ content of the spherical quicklime particles; and the rate had a maximum value when the CO₂ content was 12 mass%. This trend is similar to the trend observed in the average dissolution rates, as shown in Fig. 9. Thus, the second foaming dominantly controls the dissolution rate of quicklime.

Fig. 10.

Influence of the CO₂ content on the average dissolution rate of quicklime in the molten slag during the second foaming.

4.3. Acceleration of the Quicklime Dissolution by the Second Foaming

The equation for the lime dissolution rate reported by Matsushima et al.¹⁰⁾ is generally used for analyzing a steelmaking process.   

$$-\frac{dr}{dt}\cong \frac{k_{\text{CaO}}{\rho }_{\text{CaO}}}{100{\rho }_{\text{slag}}}\left({(\%\text{CaO})}_{sat}-{(\%\text{CaO})}_{\text{slag}}\right)$$(5)

where (%CaO)_sat is the saturated composition of CaO in the liquid slag phase, which is assumed to be 37 mass%.¹⁹⁾ k_CaO is the mass transfer coefficient of CaO in the liquid slag, and Matsushima et al. reported that k_CaO is in the range of from 5.9×10⁻⁴ to 4.0×10⁻³ (cm/s) depending on the slag composition and relative velocity¹⁰⁾ (for conditions the most similar with this study, the value of k_CaO is 1.07×10⁻³). If this equation can be applied to this study, only the parameter k_CaO is affected by the changing CO₂ content in the quicklime samples, because the other parameter should be same under the applied experimental conditions. In this analysis, the quicklime particle added into molten slag assumed to keep a single sphere without splitting. Figure 11 shows the relation between average mass transfer coefficient during dissolution test, k_av, and the CO₂ content in the spherical quicklime samples. The average mass transfer coefficient, k_av, increased with the CO₂ content and then decreased when the CO₂ content exceeded 12 mass%, a behavior which was the same as the trend observed for the dissolution rate. The average mass transfer coefficient, k_av, shows the same order of magnitude as the value reported by Matsushima et al.,¹⁰⁾ when the CO₂ content was zero. In their study, a cylindrical rod of sintered lime was used, i.e., the lime was completely calcined. Therefore, their reported experiment corresponds to the conditions where the CO₂ content was lower than 0.2 mass% in this study. The value obtained in this study seems reasonable, though the experimental method was different.

Fig. 11.

Relation between the CO₂ content in spherical quicklime samples and mass transfer coefficients of CaO.

The second foaming enhanced the dissolution rate of quicklime by the gas emission due to the thermal decomposition of the CaCO₃ core. The mass transfer coefficient also increased as shown in Fig. 11. An acceleration factor α can be defined by the following equation.   

$$k_{av}=\left(1+\alpha \right)k_0$$(6)

where k₀ is the mass transfer coefficient when CO₂ content is zero. Figure 12 shows the relationship between the acceleration factor and the CO₂ content of the spherical quicklime samples. The acceleration factor has a linear relationship with the CO₂ content of the spherical quicklime samples where the CO₂ content is lower than 12 mass% and thus, the following empirical equation for quicklime with 13 mm in diameter is obtained.   

$$\alpha =0.53\times (\%{\text{CO}}_{\text{2}})$$(7)

Fig. 12.

Relation between the CO₂ content of spherical quicklime particles and the acceleration factor of the quicklime dissolution.

The advantage of Matsushima’s study¹⁰⁾ is that the relative velocity between the lime sample and slag can be precisely controlled by changing the rotation rate of the lime rod. Though the stirring power for the slag can be changed by changing the Ar gas flow rate in this study, the relative velocity between the lime sample and slag remains unclear. The mass transfer coefficient without acceleration by second foaming, k₀, can be determined by Matsushima’s equation as shown in Eq. (5) based on the properties of the slag, e.g., its composition, temperature, viscosity, density, and so on, while the acceleration factor α, taking into account the foaming due to the thermal decomposition of CaCO₃ core, can be determined as demonstrated in this study. It should be noted that acceleration factor α includes both effects of not only the promotion of relative velocity between quicklime and molten slag, but also the increase of effective area for dissolution by splitting of quicklime due to the internal gas emission.

Hence, by combining the equations obtained in Matsushima’s study and those in this study, the dissolution rate of quicklime used in actual operation can be estimated by the following equation.   

$$\begin{array}{l}-\frac{dr}{dt}\cong \left(1+\alpha \right)\frac{k_{\text{CaO}}{\rho }_{\text{CaO}}}{100{\rho }_{\text{slag}}}\left({(\%\text{CaO})}_{\text{sat}}-{(\%\text{CaO})}_{\text{slag}}\right)\\ \cong \left(1+0.53\times (\%{\text{CO}}_{\text{2}})\right)\frac{k_{\text{CaO}}{\rho }_{\text{CaO}}}{100{\rho }_{\text{slag}}}\left({(\%\text{CaO})}_{\text{sat}}-{(\%\text{CaO})}_{\text{slag}}\right)\end{array}$$(8)

The above equation can be applied for CO₂ contents below 12 mass%. As shown in Fig. 11, the mass transfer coefficient decreased for CO₂ contents of more than 12 mass%. Thus, a CO₂ content of more than 12 mass% has the disadvantage of a rapid dissolution of the lime. Deng et al. measured the dissolution rate of limestone (CaCO₃) in steelmaking slag.^33,34) They concluded that the addition of bulk limestone to the molten slag instead of quicklime is unbeneficial, because the dissolution rate of limestone is low in spite of gas formation from limestone (in which the CO₂ content is 44 mass%). The results in this study agree with their results. Excess CO₂ content in quicklime causes a decrease in the dissolution rate of quicklime into the slag. As a reason for this behavior, it is considered that the temperature in the surrounding of the quicklime sample is decreased by the thermal decomposition of the CaCO₃ core. Furthermore, there is the possibility that the contact between quicklime and a molten slag is obstructed due to the strong gas formation when the CO₂ content constitutes more than 12 mass%. The basis for this assertion is that the CaO content in the slag increased after the second foaming in the case of a CO₂ content of 25 mass%. Therefore, it is concluded that, for a rapid dissolution of quicklime in the molten slag, the suitable amount of CaCO₃ should exist as a core in the center of a quicklime particle. In the case of quicklime with 13 mm in diameter, the suitable CO₂ content is approximately 12 mass%, and the optimum content in quicklime might be different depending on the quicklime diameter.

5. Conclusion

The effect of CO₂ content in quicklime particle on the dissolution rates of quicklime particles in molten slag was investigated by using spherical quicklime with individually controlled CO₂ content, and the following findings were obtained:

(1) The dissolution of the spherical quicklime particles in the molten slag proceeds with two occurrences of the gas formation which are due to the thermal decomposition of CaCO₃ and/or Ca(OH)₂.

(2) The start time of the second foaming decreased while the duration of the second foaming elongated with increasing CO₂ content. Contrarily, the CO₂ content has little effect on the duration of the first foaming.

(3) The second foaming enhanced the dissolution of the quicklime with a CO₂ content lower than 12 mass%.

(4) The mass transfer coefficient of CaO in the slag for quicklime with foaming can be expressed by using an acceleration factor, which described an acceleration effect of 0.53 times per mass% CO₂ for the quicklime with 13 mm in diameter.

(5) The equation for the quicklime dissolution with gas formation can be expressed by combining the traditional dissolution equation proposed by Matsushima et al. and the acceleration factor.

Acknowledgements

This study was supported by the 23rd ISIJ Research Promotion Grant. The authors are grateful to all members of the ISIJ Research Committee in Slag formation with high speed lime dissolution for helpful discussions. Quicklime used in this study was supplied by Yoshizawa Lime Industry CO., LTD.

References

• 1)   J.  Yang,  M.  Kuwabara,  T.  Asano,  A.  Chuma and  J.  Du: ISIJ Int., 47 (2007), 1401.
• 2)   Y.  Satyoko and  W. E.  Lee: Br. Ceram. Trans., 98 (1999), 261.
• 3)   S.  Amini,  M.  Brungs and  O.  Ostrovski: ISIJ Int., 47 (2007), 32.
• 4)   N.  Dogan,  G. A.  Brooks and  M. A.  Rhamdhani: ISIJ Int., 49 (2009), 1474.
• 5)   S.  Kitamura,  H.  Shibata and  N.  Maruoka: High Temp. Mater. Process., 31 (2012), 195.
• 6)   Y.  Ogawa and  N.  Maruoka: Tetsu-to-Hagané, 100 (2014), 434.
• 7)   F.  Pahlevani,  S.  Kitamura,  H.  Shibata and  N.  Maruoka: Steel Res. Int., 81 (2010), 617.
• 8)   Z. S.  Li,  M.  Whitwood,  S.  Millman and  J.  van Boggelen: Ironmaking Steelmaking, 41 (2014), 112.
• 9)   R.  Sarkar,  U.  Roy and  D.  Ghosh: Metall. Mater. Trans. B, 47 (2016), 2651.
• 10)   M.  Matsushima,  S.  Yadoomaru,  K.  Mori and  Y.  Kawai: Trans. Iron Steel Inst. Jpn., 17 (1977), 442.
• 11)   M.  Umakoshi,  K.  Mori and  Y.  Kawai: Trans. Iron Steel Inst. Jpn., 24 (1984), 532.
• 12)   S.  Taira,  K.  Nakashima and  K.  Mori: Tetsu-to-Hagané, 81 (1995), 16.
• 13)   T.  Hamano,  M.  Horibe and  K.  Ito: ISIJ Int., 44 (2004), 263.
• 14)   S.  Jansson,  V.  Brabie and  P.  Jonsson: Ironmaking Steelmaking, 33 (2006), 389.
• 15)   F.  Kirihara,  S.  Sukenaga,  N.  Saito and  K.  Nakashima: Tetsu-To-Hagané, 100 (2014), 1361.
• 16)   E.  Cheremisina,  J.  Schenk,  L.  Nocke,  A.  Paul and  G.  Wimmer: ISIJ Int., 57 (2017), 304.
• 17)   W.  Yan,  W.  Chen,  Y.  Yang,  X.  Zhao and  A.  McLean: Iron and Steel Technology Conf. (AISTech) Proc., AIST, Warrendale, PA, (2015), 2162.
• 18)   N.  Maruoka,  A.  Ishikawa,  H.  Shibata and  S. Y.  Kitamura: Proc. Unified Ine. Technical Conf. on Refractiories (UNITECR) 2011 Cong.: 12th Biennial Worldwide Conf. on Refractories - Refractories-Technology to Sustain the Global Environment, American Ceramic Society, Westerville, OH, (2011), 590.
• 19)   N.  Maruoka,  A.  Ishikawa,  H.  Shibata and  S.  Kitamura: High Temp. Mater. Process., 32 (2013), 15.
• 20)   F. X.  Huang,  J.  Liu,  N.  Maruoka,  S.  Kitamura and  A.  Ishikawa: Int. J. Appl. Ceram. Tecnol., 12 (2015), 1239.
• 21)   N.  Maruoka,  A.  Ishikawa,  H.  Shibata and  S.-y.  Kitamura: J. Tech. Assoc. Refract. Jpn., (Taikabutsu Overseas), 34 (2015), 3.
• 22)   N.  Maruoka,  J.  Liu,  H.  Shibata and  S.-y.  Kitamura: 5th Baosteel Biennial Academic Conf., Baosteel, Shanghai, (2013), B116.
• 23)   F.  Huang,  C.  Liu,  N.  Maruoka and  S. Y.  Kitamura: Ironmaking Steelmaking, 42 (2015), 553.
• 24)   X.  Guo,  Z. H. I.  Sun,  J.  Van Dyck,  M.  Guo and  B.  Blanpain: Ind. Eng. Chem. Res., 53 (2014), 6325.
• 25)   X.  Guo,  M.  Guo,  Z.  Sun,  J.  Van Dyck,  B.  Blanpain and  P.  Wollants: Materials Science and Technology Conf. and Exhibition 2010 (MS&T ʼ10), TMS, Pittsburgh, PA, (2010), 1739.
• 26)   Z. H. I.  Sun,  X. L.  Guo,  J.  Van Dyck,  M. X.  Guo and  B.  Blanpain: AlChE J., 59 (2013), 2907.
• 27)   N.  Maruoka and  H.  Nogami: Iron and Steel Technology Conf. (AISTech) Proc., AIST, Warrendale, PA, (2015), 2117.
• 28)   N.  Maruoka and  H.  Nogami: Metall. Mater. Trans. B, 48 (2017), 113.
• 29)   N.  Maruoka and  H.  Nogami: AIST Trans., 13 (2016), 164.
• 30)   N.  Maruoka,  A.  Ito,  M.  Hayasaka,  S.-y.  Kitamura and  H.  Nogami: ISIJ Int., 56 (2017), No. 10, 1677.
• 31)   A. W. D.  Hills: Chem. Eng. Sci., 23 (1968), 297.
• 32)   N.  Maruoka,  A.  Ishikawa,  H.  Shibata and  S.-y.  Kitamura: Refactories (Taikabutsu), 65 (2013), 161.
• 33)   T. F.  Deng,  P.  Nortier,  M.  Ek and  S.  Du: Metall. Mater. Trans. B, 44 (2013), 98.
• 34)   T. F.  Deng and  S. C.  Du: Ironmaking Steelmaking, 41 (2014), 75.