收稿日期: 2020-08-10
网络出版日期: 2020-09-09
基金资助
上海市科委基金资助项目(19DZ2270200);省部共建高品质特殊钢冶金与制备国家重点实验室、上海市钢铁冶金新技术开发应用重点实验室自主课题资助项目(SKLASS 2019-Z024)
First-principles computation and machine learning of the energies and structures of spinel oxides
Received date: 2020-08-10
Online published: 2020-09-09
正尖晶石氧化物AB$_{2}$O$_{4}$结构, 可通过对A和B位点分别置换73种元素产生5 329种原子构型. 利用高通量第一性原理方法计算了这5 329种正立方尖晶石氧化物结构的形成能和晶格常数, 并提出了“中心-环境”(center-environment, CE)特征模型, 构建同时包含局部成分和结构信息的特征作为机器学习(machine learning, ML)方法的输入变量. 结合第一性原理计算数据, 使用随机森林算法开发了机器学习模型, 准确有效地预测了尖晶石氧化物结构的形成能和晶格常数. 通过比较机器学习预测的假想结构与实验结构的形成能, 预测出了361种更稳定的新尖晶石氧化物结构. 讨论了与尖晶石氧化物稳定性相关的“好”与“坏”的构成元素, 有助于指导实验合成新的、稳定的尖晶石氧化物.
关键词： 密度泛函理论; 机器学习; 特征工程; “中心-环境”特征模型; 尖晶石氧化物
李一航, 肖斌, 唐宇超, 刘馥, 王小梦, 刘轶 . 尖晶石氧化物能量和结构的第一性原理计算和机器学习[J]. 上海大学学报(自然科学版), 2021 , 27(4) : 635 -649 . DOI: 10.12066/j.issn.1007-2861.2251
The formal spinel oxides AB$_{2}$O$_{4}$ can have 5 329 configurations by substituting A and B with 73 elements. The first-principles method was applied to calculate the formation energies and lattice constants of 5 329 spinel oxides. To develop efficient machine learning (ML) methods, centre-environment (CE) feature models were proposed to construct the input variables of the ML methods containing local composition and structure information. Based on the first-principles computational data, random forest algorithm was used to develop an ML model to predict the formation energies and lattice constants of spinel oxides. By comparing the formation energies of hypothetical and experimental structures predicted by ML, 361 new and more stable spinel oxides were discovered. The “good” and “bad” stabilisation elements were disscussed, which helped in guiding theexperimental synthesis of novel stable spinel oxides.
| [1] | Xu Y L, Wang X M, Li X, et al. New materials band gap prediction based on the high-throughput calculation and the machine learning[J]. Scientia Sinica Technological, 2018, 49(1):44-54. |
| [2] | Jain A, Voznyy O, Sargent E H. High-throughput screening of lead-free perovskite-like materials for optoelectronic applications[J]. The Journal of Physical Chemistry C, 2017, 121(13):7183-7187. |
| [3] | Sun S J, Noor T P, Hartono Z D R, et al. Accelerated development of perovskite-inspired materials via high-throughput synjournal and machine-learning diagnosis[J]. Joule, 2019, 3(6):1437-1451. |
| [4] | Li G Z, Hu Z, Hou F, et al. Machine learning enabled high-throughput screening of hydrocarbon molecules for the design of next generation fuels[J]. Fuel, 2020, 265:116968. |
| [5] | Abe O, Hosono Y, Sano T, et al. A mechanochemical route to synthesize LiMn$_{2}$O$_{4}$ as a cathode material for lithium batteries[J]. Transactions of the Materials Research Society of Japan, 2008, 33(4):941-944. |
| [6] | Cui X L, Feng H X, Liu J L, et al. Porous LiMn$_{2}$O$_{4}$ nano-microspheres as durable high power cathode materials for lithium ion batteries[J]. Russian Journal of Electrochemistry, 2019, 55(5):351-357. |
| [7] | Suellen B, Caroline R, Eric D C S, et al. Synjournal of zinc aluminate (ZnAl$_{2}$O$_{4})$ spinel and its application as photocatalyst[J]. Materials Research, 2014, 17(7):734-738. |
| [8] | Zhang T, Zhu H B, Croué J P, et al. Production of sulfate radical from peroxymonosulfateinduced by a magnetically separable CuFe$_{2}$O$_{4}$ spinel in water: efficiency, stability, andmechanism[J]. Environmental Science & Technology, 2013, 47(6):2784-2791. |
| [9] | Suzuki Y, Van-Dover R B, Gyorgy E M, et al. Structure and magnetic properties of epitaxial spinel ferrite thin films[J]. Applied Physics Letters, 1996, 68(5):714-716. |
| [10] | Jiang W, Huang H Q, Liu F, et al. Magnetic Weyl semimetals with diamond structure realized in spinel compounds[J]. Physical Review B, 2020, 101(12):121113-121116. |
| [11] | 向勇, 谢道华. 尖晶石结构功能材料的新进展[J]. 磁性材料及器件, 2001, 32(3):21-26. |
| [11] | Xiang Y, Xie D H. The development of spinel structure functional materials[J]. Journal of Magnetic Materials and Devices, 2001, 32(3):21-26. |
| [12] | 郑弘泰. 第一原理理论计算作为科学发现的工具: 磁铁矿 (magnetite) 之电荷-轨道秩序 (charge-orbital ordering) 和 Verwey 相变[J]. 物理双月刊, 2005, 27(4):5. |
| [12] | Jeng H T. First principles calculations as a tool of scientific discovery: charge-orbital ordering of magnetite and Verwey phase transition[J]. Physics Bimonthly, 2005, 27(4):5. |
| [13] | 戴道生, 钱乾坤 铁磁学[M]. 北京: 科学出版社, 2000: 32-200. |
| [13] | Dai D S. Ferromagnetics [M]. Beijing: Science Press, 2000: 32-200. |
| [14] | Kurt E, Sickafu S, John M W. Structure of spine[J]. Journal of the American Ceramic Society, 1999, 82(12):3279-3292. |
| [15] | Robert E W, Hakim I, Larry A C, et al. Thermodynamic stability of low- and high-index spinel LiMn$_{2}$O$_{4}$ surface terminations[J]. ACS Applied Materials & Interfaces, 2016, 8(17):11108-11121. |
| [16] | Gabriel R S, Carlos M A, Adalberto F. Exploring two-dimensional materials thermodynamic stability via machine learning[J]. ACS Applied Materials & Interfaces, 2020, 12(18):20149-20157. |
| [17] | Kresse G, Furthmüller J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set[J]. Physical Review B, 1996, 54(16):11169-11186. |
| [18] | Baikov A A. Database on properties of chemical elements [EB/OL]. [2020- 08- 03]. http://phases.imet-db.ru/elements/mendel.aspx?main=1. |
| [19] | Villars P, Cenzual K, Okamoto H, et al. Materials platform for data science[C]// European Materials Modelling Council. 2017: 89921. |
/
| 〈 |
|
〉 |
