Zhou Jie;Yang Bin;Song Xingcheng;Xu Baoqiang;Chen Xiumin;Tang Duzuo;Li Linfei;Sun Ronggang;Yuan Haibin;Li Zhilu;

• 1:School of Metallurgical and Energy Engineering
• 2: Kunming University of Science and Technology
• 3:Yunnan Tin Industry Group (Holding) Co.
• 4: Ltd.

Abstract：

The research on the carbothermal reduction mechanism of tin dioxide is of great significance to solve the problems of low tin recovery rate,high energy consumption and production efficiency optimization in tin smelting process.However,because the carbothermal reduction reaction is carried out at high temperature,the mechanism of the reaction process is complex,involving multiphase interface transfer and multi-step reaction,and accompanied by the volatilization behavior of stannous oxide(SnO) intermediate products,the experimental research methods are limited.In order to transform a "black box" type of reaction mechanism research into a "white box" system that can be understood,predicted and actively regulated,based on the thermodynamic analysis,with the help of deep learning technology,the deep neural network potential function model was trained with the first principles molecular dynamics(FPMD) simulation results of tin dioxide carbothermal reduction reaction as the data set.Based on the deep neural network potential function model,the dynamics simulation of tin dioxide carbothermal reduction reaction with different carbon content was carried out at 1073～1873 K in nanosecond time scale and tens of thousands of atoms space scale.The thermodynamic analysis results showed that the thermodynamic temperature range in which intermediate substances such as C_3O_2(g) and C_2O(g) could exist stably coincides with the thermodynamic temperature range in which SnO could exist stably in the carbothermal reduction reaction system of tin dioxide.The micro mechanism of tin dioxide carbothermal reduction reaction was obtained by molecular dynamics simulation of deep neural network potential function model.The direct reduction,indirect reduction and the mixed mechanism of the two in the carbothermal reduction process of tin dioxide were clarified,as well as the mechanism of low valence oxides of tin and carbon in the reduction process.The effect of temperature on the morphology and structure of carbon,and the effect of carbon morphology and activity on the direct and indirect reduction process were obtained.The direct reduction mechanism of tin dioxide carbothermal reduction reaction was that the end group carbon with high activity on the carbon chain reacted with oxygen with high activity on the surface of tin dioxide due to lattice distortion to form carbon oxides.However,the electronic structure characteristics of tin dioxide and carbon made the affinity between them poor,and the reduction could only occur when the end group carbon contacts with active oxygen.In the indirect reduction mechanism,carbon dioxide was not easy to react with carbon to generate carbon monoxide due to the reduction of carbon activity due to the easy aromatization of carbon at high temperature,resulting in the formation rate of carbon dioxide in the boudor reaction was much faster than that of carbon monoxide.Therefore,the rate control step of the carbon thermal reduction of tin dioxide was the formation of CO,regardless of whether the process involved direct or indirect reduction;the molecular dynamics simulation results of the deep neural network potential function model also showed that there were intermediate substances such as C_3O_2,C_3O_3,Sn_2O,Sn_3O_2,SnO in the carbothermal reduction process,which was consistent with the thermodynamic calculation results.The simulation results showed that these intermediate substances had high reaction activity,so promoting the formation of these intermediate substances was conducive to the carbothermal reduction process.The research results of the influence of temperature and carbon ratio on the carbothermal reduction reaction rate of tin dioxide showed that the carbothermal reduction reaction rate of tin dioxide was significantly accelerated when the temperature was higher than 1673 K.The reason was that when the temperature was lower than 1673 K,the carbon chain had a strong adsorption capacity for the generated CO,which would affect the progress of the boudor reaction.Therefore,the carbon ratio had a significant impact on the carbothermal reduction process of tin dioxide.Considering the reaction kinetics and energy consumption,the reduction temperature was 1473～1673 K,and the carbon ratio was 1:2,which was the better reduction condition.The above research results clearly demonstrated that to effectively facilitate the reduction of tin dioxide,it was essential to select carbon sources with high reactivity and conduct targeted physical processing.Such treatments could disrupt the inherent structure of carbon materials,expose more active sites,and enhance the reactivity of carbon chains during the reaction.Additionally,favorable conditions should be created to encourage the oxidation of carbon into carbon monoxide(CO),which served as a crucial reducing agent.Alternatively,suitable additives could be introduced to promote the conversion of tin dioxide into gaseous tin oxide,thereby increasing the proportion and rate of indirect reduction.The findings not only provided a theoretical basis for the process optimization of tin smelting reaction,but also established a research paradigm of deep learning technology to drive the research on the mechanism of carbothermal reduction reaction in tin metallurgical process and promote the process optimization.

Key Words： tin dioxide;carbothermal reduction;reaction mechanism;rate-controlling steps;deep learning

Abstract:

Keywords:

Foundation: 国家重点研发计划项目(2019YFC1904203);; 云南省科技厅重大科技专项计划项目(202302AB080014);云南省科技厅科技人才与平台计划项目(202205AD160028)资助

Authors: Zhou Jie;Yang Bin;Song Xingcheng;Xu Baoqiang;Chen Xiumin;Tang Duzuo;Li Linfei;Sun Ronggang;Yuan Haibin;Li Zhilu;

DOI: 10.13373/j.cnki.cjrm.XY26020008

References：

• [1] Chen L S, Li P C, Wu M Z, Su R C, Fan T W, Zhang Z, Liu Q D. Occurrence forms and separation-enrichment of tin and indium in tin smelting flue dust[J]. Mining and Metallurgical Engineering, 2025, 45(6):124.（陈丽诗，李鹏程，伍美珍，苏瑞春，范桐文，张璋，刘庆东. 锡冶炼烟尘中锡、 铟赋存形态及分离富集研究[J]. 矿冶工程， 2025, 45(6): 124.）
• [2] Yang B, Dai Y N, Zhang G J, Wu K H. Study on vacuum distillation separation of tin-antimony alloy[J]. Journal of Kunming University of Science and Technology （Science and Technology）, 1998, 23(3): 104.（杨斌， 戴永年， 张国靖， 吴昆华 . 真空蒸馏分离锡锑合金的研究[J]. 昆明理工大学学报： 理工版， 1998, 23(3): 104.）
• [3] Yang J H, Zhang Z X, Wang H B, Gao X S, Xiong K, Niu Y K, Yang G S. Geological characteristics, genesis and prospecting potential of the Bereko gold deposit in the Siguiri Basin, Guinea[J]. Gold, 2023, 44(9): 103.（郭景会， 张泽夏， 王滑冰， 高旭升， 熊凯， 牛玉凯， 杨桂松. 几内亚锡吉里盆地Bereko金矿床地质特征、 矿床成因及找矿前景分析[J]. 黄金， 2023, 44(9): 103.）
• [4] Huang W S. Tin[M]. Beijing: Metallurgical Industry Press, 2000. 91.（黄位森 . 锡[M]. 北京： 冶金工业出版社， 2000. 91.）
• [5] Song X C. Tin Metallurgy[M]. Beijing: Metallurgical Industry Press, 2011. 23.（宋兴诚. 锡冶金[M]. 北京： 冶金工业出版社， 2011. 23.）
• [6] Zhang Y B, Li G H, Jiang T, Guo Y F, Huang Z C. Reduction behavior of tin-bearing iron concentrate pellets using diverse coals as reducers[J]. International Journal of Mineral Processing, 2012, 110: 109.
• [7] Padilla R, Sohn H Y. The reduction of stannic oxide with carbon[J]. Metallurgical Transactions B, 1979, 10(1): 109.
• [8] Mitechll A R, Parker R H. The reduction of SnO_2 and Fe_2O_3 by solid carbon[J]. Minerals Engineering, 1988, 1(1): 53.
• [9] Zhang Y B, Su Z J, Zhou Y L, Li G H, Jiang T. Reduction kinetics of SnO_2 and ZnO in the tin, zincbearing iron ore pellet under a 20%CO-80%CO_2 atmosphere[J]. International Journal of Mineral Processing, 2013, 124: 15.
• [10] Debroy T, Patankar A, Simkovich G. Fuming of stannous oxide from silicate melts[J]. Metallurgical Transactions B, 1990, 21(3): 449.
• [11] Nover J C, Richardson F D. Stannous oxide and the solubility of oxygen in tin[J]. Transactions of the Institute of Metals, 1972, 81(787): 63.
• [12] Seetharaman S. On the standard Gibbs energy of formation of SnO_2[J]. Scandinavian Journal of Metallurgy, 2015, 6: 143.
• [13] Chen G F. Metallurgy for Heavy Metals[M]. Beijing: Metallurgical Industry Publishing company, 1992. 1.（陈国发 . 重金属冶金学[M]. 北京： 冶金工业出版社， 1992. 1.）
• [14] Hu Y, Zhong H, Tang K, Yang X M, Guo C J, Yang B. Influence of the corrosion resistance and high temperature oxidation resistance of Cu-15Ni-8Sn alloy with Co or P additions[J]. Nonferrous Metals Science and Engineering, 2025, 16(4): 585.（胡阳， 钟皓， 汤焜， 杨喜明， 郭诚君， 杨斌 . Co、 P 元素添加对 Cu-15Ni-8Sn 合金耐腐蚀性和高温抗氧化性能的影响[J]. 有色金属科学与工程， 2025, 16(4): 585.）
• [15] Liu H, Gong S P, Huang L H, Hu Y X, Zhou D X. Preparation of nanosized tin oxide powders by polymer network method[J]. Chinese Journal of Rare Metals, 2005, 29(5): 676.（刘欢， 龚树萍， 黄利华， 胡云香， 周东详 . 聚合物网络法制备纳米氧化锡粉体[J]. 稀有金属， 2005, 29(5): 676.）
• [16] Drevillon B, Kumar S, Rocai Cabarrocas P, Sciefeert J M. In situ investigation of the optoelectronic properties of transparent conducting oxide/amorphous silicon interfaces[J]. Applied Physics Letters, 1989, 54(21): 2088.
• [17] Unuma H, Tonooka K, Suzuki Y, Furusaki T, Kodaira K, Matsushita T. Preparation of transparentamorphous tungsten trioxide thin films by a dipcoating method[J]. Journal of Materials Science Letters, 1986, 5(12): 1248.
• [18] Xiong L B, Guo Y X, Wen J, Liu H R, Yang G, Qin P L, Fang G J. Review on the application of SnO_2 in perovskite solar cells[J]. Advanced Functional Materials, 2018, 28(35): 1802757.
• [19] Liu Y P, Li Y B, Zhang H, Chu K. Boosted electrocatalytic N_2 reduction on fluorine-doped SnO_2 mesoporous nanosheets[J]. Inorganic Chemistry, 2019, 58(15): 10424.
• [20] Liu G B, Li Z H, Shi J J, Sun K, Ji Y J, Wang Z G, Qiu Y F, Liu Y Y, Wang Z J, Hu P A. Black reduced porous SnO_2 nanosheets for CO_2 electroreduction with high formate selectivity and low overpotential[J]. Applied Catalysis B: Environmental, 2020, 260: 118134.
• [21] Salvini C, Re Fiorentin M, Risplendi F, Raffone F, Cicero G. Active surface structure of SnO_2 catalysts for CO_2 reduction revealed by ab initio simulations[J]. The Journal of Physical Chemistry C, 2022, 126(34): 14441.
• [22] Aseeri J A, Amer M S, Peramaiah K, Huang W H, Al-Mayouf M A. Electrochemical reduction of CO_2 by In doped-SnO_2 nanoparticles with high stability and activity toward selective formic acid formation[J]. Journal of CO_2 Utilization, 2025, 99: 103174.
• [23] Li F W, Chen L, Knowles G P, MacFarlane D R, Zhang J. Hierarchical mesoporous SnO_2 nanosheets on carbon cloth: a robust and flexible electrocatalyst for CO_2 reduction with high efficiency and selectivity[J]. Angewandte Chemie International Edition, 2016, 56(2): 505.
• [24] Bejtka K, Zeng J Q, Sacco A, Castellino M, Hernández S, Farkhondehfal M A, Savino U, Ansaloni S, Pirri C F, Chiodoni A. Chainlike mesoporous SnO_2 as a well-performing catalyst for electrochemical CO_2 reduction[J]. ACS Applied Energy Materials, 2019, 2(5): 3081.
• [25] Ning S L, Wang J G, Xiang D, Huang S B, Chen W, Chen S W, Kang X W. Electrochemical reduction of SnO_2 to Sn from the bottom: in-situ formation of SnO_2/Sn heterostructure for highly efficient electrochemical reduction of carbon dioxide to formate[J]. Journal of Catalysis, 2021, 399: 67.
• [26] Chang S S, Yoon S O, Park H J. Characteristics of SnO_2 annealed in reducing atmosphere[J]. Ceramics International, 2005, 31(3): 405.
• [27] Kim B S, Lee J C, Yoon H S, Kim S K. Reduction of SnO_2 with hydrogen[J]. Materials Transactions, 2011, 52: 1814.
• [28] You F F, Zhou Y N, Li D Y, Zhang H, Gao D W, Ma X H, Hao R, Liu J Z. Construction of a flower-like SnS_2/SnO_2 junction for efficient photocatalytic CO_2 reduction[J]. Journal of Colloid and Interface Science, 2023, 629: 871.
• [29] Abanades S. CO_2 and H_2O reduction by solar thermochemical looping using SnO_2/SnO redox reactions: thermogravimetric analysis[J]. International Journal of Hydrogen Energy, 2012, 37(10): 8223.
• [30] Jiang L, Xu Q. Reactions of the small tin clusters with carbon monoxide: infrared spectra and DFT calculations of the Sn_nCO （n=2-5） and Sn_2(CO)_2 molecules in solid argon[J]. Bulletin of the Chemical Society of Japan, 2006, 79(6): 857.
• [31] Zhao L B, Cheng J, Wang C J, Zhao S P, Yang H Y. First-principles calculation of mechanical properties of （Ti_(0.4)W_(0.15)Mo_(0.15)Cr_(0.15)）_(100-x)Alx high-entropy alloys[J]. Chinese Journal of Rare Metals, 2026, 50 （1）: 37.（赵凌波， 程军， 王川杰， 赵胜培， 杨皓元 . （Ti_(0.4)W_(0.15)Mo_(0.15)Cr_(0.15)）_(100-x)Al_x系高熵合金力学性能的第一性原理计算[J]. 稀有金属， 2026, 50(1): 37.）
• [32] Wang X Q, Huang Q H, Zhang X N, Liu Y Y, Xiao W, Wang X M. Thermophysical properties of Yb-Hf co-doped Gd_2Zr_2O_7 ceramics studied from first principles and experiment[J]. Chinese Journal of Rare Metals, 2026, 50(1): 79.（王星奇， 黄清荷， 张欣楠， 刘宇阳， 肖伟， 王星明 . Yb-Hf 共掺杂 Gd_2Zr_2O_7 基陶瓷热物理性能第一性原理与实验研究[J]. 稀有金属， 2026, 50(1): 79.）
• [33] Hafner J. Ab initio calculation of the pressure-induced A4 → A5 → A2 （distorted）→ A3 phase transitions in tin[J]. Physical Review B, 1974, 10(10): 4151.
• [34] Guo J Y, He S Q, Jie Y, Song H T, Lu H, Xu X Y, Zhao J, Zhang Y F, Hu C X, Lu J, Yan H. Mechanism of autocatalytic reduction of CO_2 over MgCO_3 to high value-added chemicals: a DFT & AIMD study[J]. Langmuir, 2024, 40(33): 17796.
• [35] Yin Q, Xu P L, Chen X M, Liu L, Liu B, Yang L F, Zhao X R, Sheng Z, Li D Q, Yang B, Xu B Q, Jiang W L, Liu D C. Mechanism and experimental study on preparation of high-quality vanadium nitride by one-step vacuum carbothermal reduction nitridation method[J]. Vacuum, 2023, 208: 111672.
• [36] Liu B, Chen X M, Zhang E H, Zhou J, Wu H P, Chen Y M, Yang B, Xu B Q, Jiang W L. Formation mechanism and luminescence properties of silicon carbide nanowires with core-shell structure[J]. Ceramics International, 2024, 50(22): 48312.
• [37] Bale C W, Bélisle E, Chartrand P, Decterov S A, Eriksson G, Gheribi A E, Hack K, Jung I H, Kang Y B, Melan??on J, Pelton A D, Petersen S, Robelin C, Sangster J, Spencer P, Van Ende M A. Reprint of: factsage thermochemical software and databases, 2010-2016[J]. Calphad, 2016, 55: 1.
• [38] Biovia D S. Materials studio. R2 Dassault Systèmes BIOVIA, San Diego, 2017.
• [39] Blochl P E. Projector augmented-wave method[J]. Phys. rev. B, 1994, 50(24): 17953.
• [40] Kresse G, Furthmüller J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set[J]. Computational Materials Science, 1996, 6(1): 15.
• [41] Hafner J. Ab-initio simulations of materials using VASP: density-functional theory and beyond[J]. Journal of Computational Chemistry, 2008, 29(13): 2044.
• [42] Zhou K, Liu B. Control Techniques of Molecular Dynamics Simulation[M]. Amsterdam: Elsevier, 2022. 67.
• [43] Nosé S. A molecular dynamics method for simulations in the canonical ensemble[J]. Molecular Physics, 2002, 100(1): 191.
• [44] Wang H, Zhang L F, Han J Q, E W N. DeePMD-kit: a deep learning package for many-body potential energy representation and molecular dynamics[J]. Computer Physics Communications, 2018, 228: 178.
• [45] Chen X M, Chen Y M, Zhou J. Short bond evaluation method for rapidly assessing the generalization ability of deep neural network potential function models and its effectiveness verification[J]. npj Computational Materials, 2026, 12: 90.
• [46] An G M, Yan Y Q, Tao Y C, Sun Q J, Wang Y R, Zhang Y, Chen H L, Li Tao. A molecular dynamics-based approach to the crystallization of bulk water in the presence of an electric field[J]. Physica B: Condensed Matter, 2024, 674: 415555.
• [47] Drici N. Hydration shell properties and dynamic behaviour of hydrogen bonds in some aqueous solutions of amino acids[J]. Journal of Solution Chemistry, 2025, 54(12): 1690.