# Fileset

[2025ISIJInt(Shi)_Spinel.pdf](https://mdr.nims.go.jp/filesets/61a2ff34-5c4b-414b-a277-ab7c99004f1a/download)

## Creator

[Chonglin Shi](https://orcid.org/0009-0008-3856-1387), [Yuxing Liu](https://orcid.org/0000-0002-1735-5663), [Kentaro Urata](https://orcid.org/0009-0000-3175-6407), [Shintaro Yasui](https://orcid.org/0000-0003-0524-9318), [Taichi Abe](https://orcid.org/0000-0002-5065-0939), Yoshinao Kobayashi

## Rights

[Creative Commons BY-NC-ND Attribution-NonCommercial-NoDerivs 4.0 International](https://creativecommons.org/licenses/by-nc-nd/4.0/)

## Other metadata

[Thermodynamic Investigation of MnAl<sub>2</sub>O<sub>4</sub> Spinel Precipitation Behavior in Manganese Containing-Slag Systems](https://mdr.nims.go.jp/datasets/dc52510f-674f-47fb-832f-eb54a734bfed)

## Fulltext

ISIJ Int. 65(13): 2269-2274 (2025)2269ISIJ International, Vol. 65 (2025), No. 13, pp. 2269–2274https://doi.org/10.2355/isijinternational.ISIJINT-2025-334* Corresponding author: E-mail: shi.c.aa@m.titech.ac.jp© 2025 The Iron and Steel Institute of Japan. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs license (https://creativecommons.org/licenses/by-nc-nd/4.0/).1.  IntroductionManganese (Mn) steel is widely used in automotive, railway, and structural applications due to its excellent mechanical properties.1,2) With the increasing demand for high-performance steel, the production and application of Mn steel have continued to grow. According to the USGS Mineral Commodity Summaries 2024, global Mn mine production in 2023 was approximately 20 million tons, of which about 90% was consumed by the steelmaking indus-try.3) However, because of the strong affinity of Mn with oxygen, a substantial portion of Mn is inevitably lost to the slag during steelmaking. Material flow studies in Japan have shown that the amount of Mn used as an alloying element is nearly equivalent to the amount ultimately contained in the steelmaking slag.4–6) Statistics indicate that basic oxygen furnace (BOF) slag typically contains 2–6 mass% Mn, while Thermodynamic Investigation of MnAl2O4 Spinel Precipitation Behavior in Manganese Containing-Slag SystemsChonglin SHI,1)*  Yuxing LIU,1)  Kentaro URATA,2)  Shintaro YASUI,2)  Taichi ABE3)  and  Yoshinao KOBAYASHI2)1)  School of Materials and Chemical Technology, Institute of Science Tokyo, 2-12-1 Ookayama, Meguro-ku, Tokyo, 152-8550 Japan.2)  Laboratory for Zero-Carbon Energy, Institute of Integrated Research, Institute of Science Tokyo, 2-12-1 Ookayama, Meguro-ku, Tokyo, 152-8550 Japan.3)  Research Center for Structural Materials, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba City, Ibaraki, 305-0047 Japan.(Received September 28, 2025; Accepted October 16, 2025; Advance online published October 29, 2025; Published December 15, 2025)In this study, the chemical equilibrium experiments were systematically conducted at 1 673 K under an oxygen partial pressure of 10−13 atm to investigate the precipitation of MnAl2O4 spinel in the CaO–SiO2–MgO–FeO–MnO–Al2O3 slags. The results showed that with increasing Al2O3 content, Mn gradually trans-ferred from the silicate liquid phase into the spinel phase. At 33.64 mol% Al2O3, the precipitation ratio of MnAl2O4 was approximately 80.32%, at which point the slag reached its solubility limit with respect to Al2O3. Therefore, 80.32% represented the maximum precipitation ratio attainable through Al2O3 control under the present experimental conditions. At this composition, comparison between water-quenching and slow cooling at 4 K/min revealed negligible differences in the crystal structure and composition of the spinel phase, indicating that MnAl2O4 precipitation predominately occurred during the high-temperature equilibration stage and remained stable upon cooling. In addition, with increasing Al2O3 content, the activ-ity of MnO decreased, indicating its progressive consumption from the silicate melt and incorporation into the spinel phase.KEY WORDS:  thermodynamic analysis; MnAl2O4 spinel; CaO–SiO2–MgO–FeO–MnO–Al2O3 slag; precipita-tion behavior; activity measurement.Short Articleelectric arc furnace (EAF) slag may contain 5–10 mass% Mn.7) In Japan, 98.4% of steelmaking slag is reused, mainly as road base and civil engineering materials.8) Nevertheless, the reutilized slags still contain considerable amounts of Mn. Previous studies have shown that the phase assemblage in which major elements ultimately occur largely determines the reusability of slag,9) as it governs its long-term stabil-ity and mechanical properties. Therefore, it is necessary to systematically investigate the Mn-bearing phases in steel slags and their crystallization behavior.Among the possible Mn-bearing phases, MnAl2O4 spinel not only forms readily in the slag but is also regarded as the most stable Mn-containing phase. Thus, clarifying the thermodynamic precipitation conditions of MnAl2O4 is particularly important, as it enables compositional control over its formation and enhances both the functionality and resource efficiency of slag reuse.Nevertheless, studies on MnAl2O4 in the steelmaking process have so far mainly focused on MnAl2O4 inclusion https://creativecommons.org/licenses/by-nc-nd/4.0/https://orcid.org/0009-0008-3856-1387https://orcid.org/0000-0002-1735-5663https://orcid.org/0009-0000-3175-6407https://orcid.org/0000-0003-0524-9318https://orcid.org/0000-0002-5065-0939ISIJ International, Vol. 65 (2025), No. 13©  2025  ISIJ 2270in steel,10–14) while systematic studies on MnAl2O4 in steel slags remain very limited. By contrast, Cr and V-bearing spinels in slags have been extensively investigated, pro-viding valuable insights. For Cr, the FeO content15,16) and basicity17) of the slag were found to strongly promote its incorporation into spinel, while Al2O3 additions could sta-bilize Cr in spinel.18) In addition, cooling treatment has also been reported to further enhance Cr spinel formation.19,20) For V, longer holding times and lower temperatures were shown to enhance the formation of V spinel in slags,21) while higher FeO/SiO2 ratios and MgO contents were found to promote MgV2O4 crystallization.22) Despite these studies on spinel formation in slags, no systematic investigation has yet addressed MnAl2O4 in steel slags. In light of these gaps, clarifying the thermodynamic precipitation conditions of MnAl2O4 in steel slags could enable composition-based control over its generation, thereby enhancing the function-ality and resource efficiency of slag reuse.Therefore, this study focuses on the CaO–SiO2–MgO–FeO–MnO–Al2O3 slags and aims to clarify the effects of varying Al2O3 content on the precipitation behavior of MnAl2O4 spinel through equilibrium experiments. On the basis of this, the influence of cooling conditions on the effect of Al2O3 addition was also investigated to comprehensively understand the precipitation behavior of MnAl2O4. In addi-tion, the activities of MnO were also measured as part of the effort to better understand spinel precipitation behavior.2.  Experimental2.1.  Experimental ApparatusThe experimental apparatus is schematically shown in Fig. 1. An electric resistance furnace was equipped with MoSi2 heating elements and an alumina tube reaction cham-ber. The isothermal zone was determined by the Pt/Pt-13Rh thermocouple with an accuracy of ±  1 K.The chemical equilibrium process was conducted with an iron crucible which was placed on the refractory bricks in the alumina tube. The gas flow, consisting of argon (Ar) as the carrier gas and CO–CO2 as reactive gases, together with the oxygen partial pressure, was regulated by the gas control system. Moreover, the Ar gas was further deoxidized using titanium in a deoxidizing furnace. The mixed gases were introduced from the top of the alumina tube and vented from the bottom.2.2.  Slag PreparationThe slag compositions in this study were designed based on a typical electric arc furnace (EAF) slag provided by a domestic Japanese steelmaking plant, with the initial chemical compositions summarized in Table 1, where Slag 2 represents the original slag. The Al2O3 content was varied while correspondingly adjusting the amounts of CaO and SiO2 to maintain a constant basicity (CaO/SiO2 molar ratio =  0.82).To prepare CaO, CaCO3 powder was calcined in an alu-mina crucible at 1 473 K (1 200°C) for 12 hours under an Ar atmosphere with a flow rate of 150 mL/min. Similarly, FeO was prepared by heating FeC2O4·2H2O at 1 273 K (1 000°C) for 6 hours under an Ar atmosphere with a flow rate of 300 mL/min.23,24) The phase purity of both CaO and FeO was confirmed by XRD (Bruker D2 Phaser, 2nd generation).The other raw materials, including MgO, MnO, SiO2, and Al2O3, were mixed with the prepared CaO and FeO powders in predetermined ratios and ground uniformly using a plan-etary mono mill (Fritsch P-6S classic line, agate grinding ball). This homogenized powder mixture was used as the material for the thermodynamic equilibrium experiments.2.3.  Experimental ProcedureThermodynamic equilibrium experiments were carried out to determine the activity of MnO in slags by equilibrat-ing the slag with an Ag–Mn alloy. According to preliminary experiments, 28 hours of the holding time was long enough to reach equilibrium state. Each experiment was conducted at 1 673 K (1 400°C), using 8 g of pre-melted slag and 4 g of electrolytic silver (Ag, 99.99% purity), placed in an iron crucible. After equilibration, the samples were retrieved from the furnace and water-quenched immediately to obtain the Ag–Mn alloy and slag for analysis.The oxygen partial pressure was precisely controlled at 10 −13 atm by introducing a mixture of CO and CO2 into the furnace chamber (at 1 673 K, PCO/PCO2  ≈ 151:1).25) This control method is based on the redox equilibrium between CO and CO2,26) represented by the following equations.  2 22 2CO g O g CO g� � � � � � � � ................... (1)  �G T1 561 911 170 46� �� � � � � �. /J mol ............. (2)  PGRTPPOCOCO2212��������� �������� ��exp�.................... (3)In Eq. (2), �G 1� ��  is the standard Gibbs energy change of reaction (1) at absolute temperature T. In Eq. (3), PO2, PCO and PCO2  are the partial pressures of O2, CO and CO2 respec-tively, and R is the gas constant (8.314 J/mol·K).Fig. 1.  Schematic diagram of the experimental apparatus.Table 1.  Initial slag compositions (mol%). (All values in the fol-lowing tables are given with four significant figures.)No. CaO SiO2 Al2O3 MgO FeO MnO1 18.97 23.06 13.64 7.420 28.24 8.6702 14.46 17.58 23.64 7.420 28.24 8.6703 12.21 14.84 28.64 7.420 28.24 8.6704 9.940 12.09 33.64 7.420 28.24 8.670ISIJ International, Vol. 65 (2025), No. 13©  2025  ISIJ2271Additionally, to further clarify the precipitation behavior of MnAl2O4, additional experiments were conducted under controlled cooling conditions. Slag 4 was selected and subjected to slow cooling at a rate of 4 K/min after isother-mal equilibration at 1 673 K for 28 hours. To maintain the oxygen partial pressure at 10 −13 atm throughout the cooling process, the partial pressure ratio of CO2 to CO was adjusted every 40 K (down to 1 273 K, since the slag had already solidified and further gas adjustment was unnecessary). This adjustment was determined by numerically integrat-ing Eq. (4) to maintain the average oxygen partial pressure at 10 −13 atm over the specified temperature range, and the corresponding partial pressure ratios of CO2 to CO are sum-marized in Table 2.pT TGRTP P dTOTT2 21 1 2���������� � � �� ��end startCO COstartenexp /�dd� ... (4)2.4.  AnalysisThe morphology of the slag after the equilibrium experiment was examined by scanning electron microscopy equipped with energy-dispersive X-ray spectroscopy (SEM-EDS, JEOL JCM-7000). The chemical compositions of the individual phases in the slag were determined by SEM-EDS, and the overall phase identification was conducted by X-ray diffraction (XRD). The Mn concentration in the Ag–Mn alloy was measured by inductively coupled plasma optical emission spectroscopy (ICP-OES, HITACHI PS7800) after dissolution in nitric acid.3.  Results and Discussion3.1.  Thermodynamic Equilibrium Calculation of the MnAl2O4 Precipitation in the SlagThe precipitation reaction of MnAl2O4 spinel in slag and the corresponding Gibbs free energy change can be described as the following equations.  MnO Al O MnAl O� �2 3 2 4 ..................... (5)  � �G G RTaa a5 52 42 3� � � ��� ��ln MnAl OMnO Al O............... (6)Since the activity of the precipitated spinel phase can be approximated as unity, the equilibrium driving force is mainly governed by the activities of MnO and Al2O3 in the slag. As what is shown in Eq. (6), increasing Al2O3 activ-ity lowers ΔG(5), shifting the equilibrium toward MnAl2O4 formation. This provides the thermodynamic basis for inves-tigating the effect of Al2O3 content on spinel precipitation.To evaluate this effect, equilibrium calculations were performed using FactSage 8.3 (FToxid 8.3 and FactPS 8.3 databases) under the same conditions as the experiments (1 673 K and an oxygen partial pressure of 10 −13 atm). In the calculations, the molar fractions of MgO, FeO, and MnO were kept constant (7.420, 28.24, 8.670 mol% respectively), while the basicity was fixed at 0.82 to match the designed slags in Table 1. The calculated activities of Al2O3 are sum-marized in Table 3. With increasing bulk Al2O3, its activity increased progressively and reached unity at 36.56 mol%, indicating that the solubility limit had been reached.The calculated Mn precipitation ratios are presented in Fig. 2(a). The precipitation ratio of MnAl2O4 was defined as the fraction of Mn incorporated into the spinel phase, relative to the total Mn content in the bulk slag. The addi-tion of Al2O3 increased its activity in the slag. As a result, the fraction of Mn incorporated into the spinel phase was significantly enhanced. Accordingly, Al2O3 content was selected as the key compositional variable for the subse-quent experiments.3.2.  Effect of Al2O3 Content on MnAl2O4 PrecipitationThe SEM micrographs of Slag 1 to Slag 4 are shown in Fig. 3. Based on EDS compositional analysis, it was preliminarily identified that Slag 1–Slag 3 (Al2O3 =  13.64 to 28.64 mol%) consisted of three phases: a silicate/glassy matrix, metallic iron (Fe), and spinel. In Slag 4 (Al2O3 = 33.64 mol%), an Al2O3 phase was also observed. The XRD patterns shown in Fig. 4 confirmed the above phase iden-tification. For Slag 1 to Slag 3, the main crystalline phases Table 2.  Partial pressure ratios of CO2 to CO required to maintain 10 −13 atm oxygen partial pressure at different tempera-ture intervals.Interval (K) PCO / PCO21 673–1 633 121.01 633–1 593 77.741 593–1 553 49.511 553–1 513 31.341 513–1 473 19.621 473–1 433 11.901 433–1 393 6.9401 393–1 353 3.8631 353–1 313 2.0621 313–1 273 0.4219Table 3.  Calculated activities of Al2O3 in slags.No. Al2O3 content (mol%)   aAl O2 31 13.64 0.36912 23.64 0.72033 28.64 0.80204 33.64 0.93745 36.56 1.000Fig. 2.  Effect of Al2O3 content on Mn distribution in slags: (a) calculated results; (b) experimental results. (Online ver-sion in color.)ISIJ International, Vol. 65 (2025), No. 13©  2025  ISIJ 2272were spinel and metallic Fe, whereas Slag 4 exhibited not only spinel and metallic Fe but also minor diffraction peaks of Al2O3, suggesting that the content of Al2O3 exceeded the solubility limit of Al2O3 under the present condition. Moreover, from Slag 1 to Slag 4, with the gradual increase of Al2O3 content, the broad amorphous hump between 25° and 37° (2θ), corresponding to the diffraction of the glassy matrix phase, gradually diminished and eventually disappeared, while the spinel peaks intensified, suggesting a relative decrease of the glassy matrix and enrichment of the spinel phase. Furthermore, the chemical compositions of the spinel and matrix phase, determined by EDS, are summarized in Table 4. For spinel phase, the cation ratio (Mg+Fe+Mn):Al was approximately 1:2, consistent with the stoichiometry of spinel-type oxides (AB2O4), confirm-ing that the spinel phase is a solid solution of MnAl2O4, FeAl2O4, and MgAl2O4. Moreover, silicon (Si) was detected only in the matrix phase, whereas Mn was found exclusively in the spinel and matrix phases (despite the presence of Mn in the Ag–Mn alloy, with a concentration of no more than 0.153 mol%, it was considered negligible and not included in the analysis). Accordingly, the relative amounts of matrix and spinel, as well as the precipitation ratio of MnAl2O4, were calculated based on elemental balances using the fol-lowing equations.  NNXmatrixSibulkSimatrix= .............................. (7)  NN N XXspinelMnbulkmatrix MnmatrixMnspinel�� � .................. (8) RN XNMnAl Ospinel MnspinelMnbulk2 4 �� ....................... (9)NMnbulk  and NSibulk  denote the total moles of Mn and Si in the bulk slag, while XMnmatrix , XSimatrix  and XMnspinel represent the average molar fractions of elements in the respective phases determined by EDS. Nmatrix and Nspinel are the relative amounts of the matrix and spinel phase, and RMnAl O2 4 is the precipitation ratio of MnAl2O4, defined as the fraction of Mn incorporated into the spinel phase relative to the total Mn content in the bulk slag. In this study, since the dissolution of spinel-rich slags was incomplete in ICP-OES analysis, the total moles of Mn and Si in the bulk slag used in the bal-ance calculations were taken directly from the initial charge.The ratio of spinel to the glassy matrix phase (Nspinel/ Nmatrix) shows an increasing trend with Al2O3 content, which is consistent with the XRD results. Mn distribution in slags at different Al2O3 contents is shown in Fig. 2(b). The results show that the precipitation ratio of MnAl2O4 increased with increasing Al2O3 content and reached a maximum of 80.32% at 33.64 mol% Al2O3, at which point the slag became saturated with Al2O3. Owing to this satura-tion effect, further additions of Al2O3 did not lead to any increase in the precipitation ratio. Therefore, under the present experimental conditions, 33.64 mol% Al2O3 can be regarded as the optimum composition for achieving the maximum precipitation ratio of MnAl2O4 spinel (80.32%). This behavior is consistent with the evaluation in Section 3.1: as Al2O3 content increases, its activity rises, shifting the equilibrium of reaction (5) toward spinel formation and thereby increasing the precipitation ratio of MnAl2O4 spi-nel. Although the activity of Al2O3 is difficult to measure directly due to its strong stability, the experimental obser-vation of Al2O3 saturation (Al2O3 activity→1) substantiates this interpretation. It should be noted that quantitative dis-crepancies remain between experiment and calculation. At the same bulk Al2O3 contents, the experimental precipitation ratios were generally higher than the calculated values, and the onset of Al2O3 saturation occurred earlier than predicted. Fig. 4.  XRD patterns of slags. (Online version in color.)Fig. 3.  SEM micrograghs for representive slag samples.Table 4.  Chemical composition of spinel and matrix phases deter-mined by EDS (mol%).No. Phase O Ca Si Al Mg Fe Mn1 Spinel 49.25 – – 34.23 10.28 2.131 4.110Matrix 51.93 12.00 14.84 12.02 2.544 1.686 4.9892 Spinel 48.42 – – 34.30 6.473 5.420 5.390Matrix 51.58 12.06 14.44 14.29 0.8600 2.846 3.9963 Spinel 50.55 – – 33.45 5.360 5.935 4.700Matrix 53.66 11.83 14.36 14.48 0.3770 2.681 2.6104 Spinel 48.89 – – 34.60 4.670 6.410 5.430Matrix 53.08 11.68 13.96 18.06 0.07800 1.170 1.970ISIJ International, Vol. 65 (2025), No. 13©  2025  ISIJ2273These deviations are mainly due to the lack of prior studies, which has limited the completeness of the FactSage ther-modynamic database for the present system. Nevertheless, the calculations reproduced the overall experimental trend, indicating their reliability. The present results can provide a thermodynamic supplement and extend the applicability of thermodynamic predictions to Mn-bearing spinel precipita-tion in this system.3.3.  Effect of Cooling Rate on MnAl2O4 PrecipitationIn industrial practice, steel slag is commonly subjected to various cooling methods, including slow cooling, water cooling, and air cooling.27,28) As noted in the introduction, previous studies have demonstrated that cooling conditions are known to affect spinel precipitation, such as crystal growth, crystallization sequence, and the precipitation ratio of specific element-bearing spinels.12,15–19) To examine whether such effects occur under the present conditions, the slag with 33.64 mol% Al2O3 was first equilibrated at 1 673 K for 28 h under an oxygen partial pressure of 10 −13 atm. The sample was then subjected to slow cool-ing at 4 K/min, and the results were compared with those obtained from the sample quenched immediately after equilibration. SEM, XRD, and EDS analyses revealed that the spinel phase obtained under both conditions exhibited nearly identical compositions (Table 5). Furthermore, the principal diffraction peaks of spinel in the XRD patterns showed almost no shift in 2θ positions (Fig. 5), confirming that the crystal structure was essentially unchanged with the cooling method.This phenomenon can be explained from both thermody-namic and kinetic perspectives. For the reaction given in Eq. (5), the Gibbs free energy change of MnAl2O4 formation,29) expressed in Eq. (10), decreases with decreasing tempera-ture, indicating that the spinel phase remains thermodynami-cally stable during cooling. Meanwhile, since most Mn in the liquid phase had already been consumed during the high-temperature equilibration stage and diffusion rates decreased significantly upon cooling, further nucleation of spinel was kinetically unfavorable.In summary, the cooling method (quenching vs. slow cooling) had no significant effect on the MnAl2O4 precipi-tated during the high-temperature equilibrium stage, which indicates that MnAl2O4 maintains its stability under different cooling conditions once equilibrium precipitation has been achieved.  �G T5 45116 11 81� �� � � � � �. /J mol ............. (10)3.4.  Effect of Al2O3 Content on the Activity of MnO in SlagThe activity of MnO is defined as its thermodynamically effective concentration in the slag melt. Unlike the simple concentration, activity reflects the actual chemical potential of MnO in the system and therefore determines its ability to participate in reactions. When the activity of MnO in the slag system decreases, it indicates that the free MnO in the melt is being progressively consumed, for example, through incorporation into the spinel phase. Therefore, mea-suring the activity of MnO under different Al2O3 contents can directly reveal the thermodynamic driving force for MnAl2O4 formation. In this study, the activities of MnO were determined by using the chemical equilibrium method with an Ag–Mn alloy as the reference metal at 1 673 K and 10 −13 atm oxygen partial pressure in the CaO–SiO2–MgO–FeO–MnO–Al2O3 slags with varying Al2O3 contents. These data not only provided valuable insights into the forma-tion mechanism of MnAl2O4 but also served as important thermodynamic supplements for this complex system, contributing to improved accuracy in future thermodynamic calculations. The activity of MnO, which is expressed as aMnO can be calculated from its oxidation reactions and the corresponding standard Gibbs energy changes,21) which are shown as the following equations.  Mn l O g MnO s� � � � � � � �122 .................. (11)  �G T11 400 995 84 27� �� � � � � �. /J mol ........... (12)  aGRTX PMnO Mn Mn O���������� � � �� ��exp� 11122� ........... (13)  log Mn in Ag l� � �� ���2 1321 088T. ................ (14)In Eq. (13), XMn is the mole fraction of Mn in the Ag–Mn Table 5.  Comparison of spinel compositions after slow cooling and quenching determined by EDS (mol%).Cooling method O Al Mg Fe Mnquenching 48.89 34.60 4.670 6.410 5.430slow cooling 48.24 35.18 4.506 6.572 5.502Fig. 5.  Comparison of XRD patterns of spinel phases under water-quenching (a) and slow cooling (4 K/min) (b). (Online version in color.)ISIJ International, Vol. 65 (2025), No. 13©  2025  ISIJ 2274alloy, and γMn denotes the activity coefficient of Mn in the Ag–Mn alloy. Under dilute conditions, γMn can be regarded as equivalent to the Raoultian activity coefficient of Mn at infinite dilution in liquid Ag, as given by Eq. (14).30) The activities of MnO calculated from the method described above along with the corresponding mole fractions of Mn in the Ag–Mn alloy were summarized in Table 6. With increasing Al2O3 content, the activity of MnO decreased, while, as observed in Section 3.2, the precipitation ratio of Mn spinel increased. This indicates that the free MnO dis-solved in the silicate melt (which solidified into the glassy matrix upon cooling) was continuously consumed to form the spinel phase. The variation trend of MnO activity pro-vides thermodynamic support for the formation of MnAl2O4.4.  ConclusionsIn the present work, thermodynamic equilibrium experi-ments were conducted at 1 673 K under an oxygen partial pressure of 10 −13 atm to investigate the precipitation behav-ior of MnAl2O4 spinel in the CaO–SiO2–MgO–FeO–MnO–Al2O3 slags. The main conclusions are as follows:(1)  With increasing Al2O3 content from 13.64 mol% to 33.64 mol%, Mn gradually transferred from the silicate liquid phase into the spinel phase. At 33.64 mol% Al2O3, the precipitation ratio of MnAl2O4 was approximately 80.32%, at which point the slag reached its solubility limit with respect to Al2O3. Therefore, the maximum precipitation ratio of MnAl2O4 was achieved under this condition.(2)  At 33.64 mol% Al2O3, comparison between quench-ing and slow cooling at 4 K/min revealed negligible differ-ences in the crystal structure and composition of the spinel phase, indicating that MnAl2O4 precipitation was mainly completed during the high-temperature equilibration stage and remained stable upon cooling.(3)  With increasing Al2O3 content, the activity of MnO decreased, indicating that MnO originally dissolved in the silicate melt (later solidified as the glassy matrix) was pro-gressively consumed and incorporated into the spinel phase.(4)  From a practical perspective, controlling the Al2O3 content effectively facilitates the precipitation of Mn in the form of MnAl2O4 spinel. This provides a thermodynamic basis and support for regulating the chemical state of Mn in slag and improving the utilization of steelmaking slags.Statement for Conflict of InterestThe authors declare that there are no conflicts of interest regarding this study.AcknowledgmentThe authors are grateful to the research group of “Analysis of the Existence State of Mn in Steelmaking Slags” at the Iron and Steel Institute of Japan (ISIJ) for their financial support.REFERENCES1)  R. Zellagui, L. Hemmouche, H. Ait-Sadi and A. Chelli: Arch. Metall. Mater., 67 (2022), 863. https://doi.org/10.24425/amm.2022.1396762)  S. J. Kim, J. Takekawa, H. Shibata, S. Y. Kitamura, K. Yamaguchi and Y. B. Kang: ISIJ Int., 53 (2013), 1325. https://doi.org/10.2355/isijinternational.53.13253)  United States Geological Survey: Manganese Statistics and Infor-mation, https://www.usgs.gov/centers/national-minerals-information-center/manganese-statistics-and-information, (accessed 2025-07-25).4)  K. Nakajima, K. Yokoyama and T. Nagasaka: ISIJ Int., 48 (2008), 549. https://doi.org/10.2355/isijinternational.48.5495)  S. Kim, H. Shibata, N. Maruoka, S. Kitamura and K. Yamaguchi: High Temp. Mater. Process., 30 (2011), 425. https://doi.org/10.1515/htmp.2011.0646)  S. Kim, J. Suzuki, X. Gao, S. Ueda and S. Kitamura: J. Sustain. Metall., 2 (2016), 141. https://doi.org/10.1007/s40831-016-0042-z7)  Y. Huang, G. Xu, H. Cheng, J. Wang, Y. Wan and H. Chen: Procedia Environ. Sci., 16 (2012), 791. https://doi.org/10.1016/j.proenv.2012.10.1088)  Nippon Slag Association: Iron and Steel Slag (2022), https://www.slg.jp/statistics/report.html, (accessed 2025-05-09).9)  L. Jiang, Y. Bao, Q. Yang, Y. Chen, G. Liu, F. Han, J. Wei, F. Engström and J. Deng: Steel Res. Int., 88 (2017), 1700066. https://doi.org/10.1002/srin.20170006610)  Z. Y. Deng and M. Y. Zhu: ISIJ Int., 53 (2013), 450. https://doi.org/10.2355/isijinternational.53.45011)  M. K. Paek, J. M. Jang, M. Jiang and J. J. Pak: ISIJ Int., 53 (2013), 973. https://doi.org/10.2355/isijinternational.53.97312)  L. Z. Kong, Z. Y. Deng and M. Y. Zhu: ISIJ Int., 57 (2017), 1537. https://doi.org/10.2355/isijinternational.ISIJINT-2017-11813)  Z. Y. Deng, Z. H. Liu, M. Y. Zhu and L. Q. Huo: ISIJ Int., 61 (2021), 1. https://doi.org/10.2355/isijinternational.ISIJINT-2020-35214)  Y. Wang, Y. Yang, Z. Dong, J. Park, Z. Mi, X. Mao and W. Mu: Metall. Mater. Trans. B, 53 (2022), 2182. https://doi.org/10.1007/s11663-022-02517-215)  J. Li, A. Xu, D. He, Q. Yang and N. Tian: Int. J. Miner. Metall. Mater., 20 (2013), 253. https://doi.org/10.1007/s12613-013-0720-916)  Q. Zeng, J. Li, Q. Mou, H. Zhu and Z. Xue: JOM, 71 (2019), 2331. https://doi.org/10.1007/s11837-019-03465-017)  X. Wu, X. Dong, R. Wang, H. Lv, F. Cao and X. Shen: J. Residuals Sci. Technol., 13 (2016), S57. https://doi.org/10.12783/issn.1544-8053/13/2/S1018)  Z. Wang and I. Sohn: Ceram. Int., 47 (2021), 10918. https://doi.org/10.1016/j.ceramint.2020.12.21119)  L. Cao, C. Liu, Q. Zhao and M. Jiang: J. Iron Steel Res. Int., 25 (2018), 1131. https://doi.org/10.1007/s42243-018-0058-720)  Y. Yu, J. Li, M. Li and Z. Xue: J. Min. Mater. Charact. Eng., 5 (2017), 132. https://doi.org/10.4236/jmmce.2017.5301121)  W. Huang and Y. Liu: ISIJ Int., 60 (2020), 2183. https://doi.org/10.2355/isijinternational.ISIJINT-2020-03922)  J. Diao, H. Tao, Z. Lai, Y. Qiu, L. Chen, H. Li and B. Xie: CrystEng-Comm, 26 (2024), 2248. https://doi.org/10.1039/d4ce00189c23)  Y. Shen, J. Chong, Z. Huang, J. Tian, W. Zhang, X. Tang, W. Ding and X. Du: Mater. (Basel), 12 (2019), 2562. https://doi.org/10.3390/ma1216256224)  Y. Gao, C. Duan, Y. Yang, D. Ruan, C. Yang and C. Hong: ISIJ Int., 55 (2015), 2273. https://doi.org/10.2355/isijinternational.ISIJINT- 2014-82625)  X. Lu, T. Miki and T. Nagasaka: Int. J. Miner. Metall. Mater., 24 (2017), 25. https://doi.org/10.1007/s12613-017-1375-826)  E. T. Turkdogan: Physical Chemistry of High Temperature Technol-ogy, Academic Press, New York, (1980), 7.27)  K. Schollbach, M. J. Ahmed and S. R. van der Laan: Int. J. Ceram. Eng. Sci., 3 (2021), 21. https://doi.org/10.1002/ces2.1007828)  Y. Li, K. Guo, J. Xiang, Y. Wang, W. Guo and S. Zhang: ISIJ Int., 62 (2022), 2197. https://doi.org/10.2355/isijinternational.ISIJINT- 2022-22929)  M. Timucin and A. Muan: J. Am. Ceram. Soc., 75 (1992), 1399. https://doi.org/10.1111/j.1151-2916.1992.tb04200.x30)  S. C. Shim, F. Tsukihashi and N. Sano: Metall. Trans. B, 24 (1993), 333. https://doi.org/10.1007/BF02659136Table 6.  Activities of MnO in slags.No. Al2O3 content  (mol%) XMn aMnO1 13.64 0.001530 0.041372 23.64 0.001170 0.031643 28.64 0.001052 0.028454 33.64 0.0007829 0.02117https://doi.org/10.24425/amm.2022.139676https://doi.org/10.2355/isijinternational.53.1325https://doi.org/10.2355/isijinternational.53.1325https://www.usgs.gov/centers/national-minerals-information-center/manganese-statistics-and-informationhttps://www.usgs.gov/centers/national-minerals-information-center/manganese-statistics-and-informationhttps://doi.org/10.2355/isijinternational.48.549https://doi.org/10.1515/htmp.2011.064https://doi.org/10.1515/htmp.2011.064https://doi.org/10.1007/s40831-016-0042-zhttps://doi.org/10.1016/j.proenv.2012.10.108https://doi.org/10.1016/j.proenv.2012.10.108https://www.slg.jp/statistics/report.htmlhttps://www.slg.jp/statistics/report.htmlhttps://doi.org/10.1002/srin.201700066https://doi.org/10.1002/srin.201700066https://doi.org/10.2355/isijinternational.53.450https://doi.org/10.2355/isijinternational.53.450https://doi.org/10.2355/isijinternational.53.973https://doi.org/10.2355/isijinternational.ISIJINT-2017-118https://doi.org/10.2355/isijinternational.ISIJINT-2020-352https://doi.org/10.1007/s11663-022-02517-2https://doi.org/10.1007/s11663-022-02517-2https://doi.org/10.1007/s12613-013-0720-9https://doi.org/10.1007/s11837-019-03465-0https://doi.org/10.12783/issn.1544-8053/13/2/S10https://doi.org/10.12783/issn.1544-8053/13/2/S10https://doi.org/10.1016/j.ceramint.2020.12.211https://doi.org/10.1016/j.ceramint.2020.12.211https://doi.org/10.1007/s42243-018-0058-7https://doi.org/10.4236/jmmce.2017.53011https://doi.org/10.2355/isijinternational.ISIJINT-2020-039https://doi.org/10.2355/isijinternational.ISIJINT-2020-039https://doi.org/10.1039/d4ce00189chttps://doi.org/10.3390/ma12162562https://doi.org/10.3390/ma12162562https://doi.org/10.2355/isijinternational.ISIJINT-2014-826https://doi.org/10.2355/isijinternational.ISIJINT-2014-826https://doi.org/10.1007/s12613-017-1375-8https://doi.org/10.1002/ces2.10078https://doi.org/10.2355/isijinternational.ISIJINT-2022-229https://doi.org/10.2355/isijinternational.ISIJINT-2022-229https://doi.org/10.1111/j.1151-2916.1992.tb04200.xhttps://doi.org/10.1007/BF02659136