钛基半导体的特征可解释学习和禁带预测

doi:10.11896/cldb.25040026

材料导报

2026, Vol. 40

Issue (7): 25040026-8 https://doi.org/10.11896/cldb.25040026

金属与金属基复合材料

钛基半导体的特征可解释学习和禁带预测

袁斌霞^*, 杨深奥, 刘宇豪, 钱红, 朱瑞

上海电力大学能源与机械工程学院,上海 201306

Interpretable Feature Learning and Band Gap Prediction for Titanium-based Semiconductors

YUAN Binxia^*, YANG Shen’ao, LIU Yuhao, QIAN Hong, ZHU Rui

College of Energy and Mechanical Engineering, Shanghai University of Electric Power, Shanghai 201306, China

摘要
参考文献
相关文章
推荐阅读
Metrics

下载:

全文 ( PDF ) ( 16555KB )
输出: BibTeX | EndNote (RIS)

摘要钛基半导体具有高化学稳定性和合适的带隙宽度。然而,由于材料种类繁多,传统的实验筛选方法效率低下。为了加快选择过程,本工作旨在构建一种基于可解释机器学习的钛基半导体带隙预测方法,以期为相关材料的筛选和优化提供理论支持。首先,通过筛选Materials Project数据库中的钛基化合物,并利用Magpie描述符提取元素特征。采用主成分分析(PCA)来降低数据的维度,创建一个具有代表性的数据集。同时,使用热图和SHAP方法揭示电负性、共价半径、周期数及晶胞体积等关键特征对带隙的影响机制,为理解材料内在属性与性能之间的关系提供直观依据。然后,综合比较不同的机器学习模型对带隙的预测性能,包括随机森林、支持向量机、线性回归和梯度增强回归,发现随机森林在带隙预测方面准确度最高。最后,通过参数调整提高了随机森林模型的预测性能。本研究为光催化和太阳能电池等领域的材料开发提供了强有力的数据支持和设计指导。

	服务
	把本文推荐给朋友
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	袁斌霞
	杨深奥
	刘宇豪
	钱红
	朱瑞

关键词: 钛基半导体禁带宽度特征提取预测随机森林

Abstract: Titanium-based semiconductors are known for their high chemical stability and suitable band gap widths. However, the conventional experimental screening methods are inefficient due to the wide variety of materials. To speed up the selection process, this work focuses on interpretable feature learning and band gap prediction for titanium-based semiconductors. First, titanium compounds were selected from the Materials Project database by machine learning, and elemental features were extracted using the Magpie descriptors. Then, principal component analysis (PCA) was applied to reduce the data dimensionality, creating a representative dataset. Meantime, heatmaps and SHAP (SHapley Additive exPlanations) methods were used to demonstrate the influence of key features such as electronegativity, covalent radius, period number, and unit cell volume on the bandgap, understanding the relationship between the material’s properties and performance. After comparing different machine learning models, including Random Forest (RF), Support Vector Machines (SVM), Linear Regression (LR), and Gradient Boosting Regression (GBR), the RF was found to be the most accurate for band gap prediction. Finally, the model performance was improved through parameter tuning, showing high accuracy. These findings provide strong data support and design guidance for the development of materials in fields like photocatalysis and solar cells.

Key words: titanium-based semiconductors band gap feature ertraction prediction random forest

发布日期: 2026-04-16

ZTFLH:

N32

基金资助: 国家自然科学基金(12172210)

引用本文:

袁斌霞, 杨深奥, 刘宇豪, 钱红, 朱瑞. 钛基半导体的特征可解释学习和禁带预测[J]. 材料导报, 2026, 40(7): 25040026-8.
YUAN Binxia, YANG Shen’ao, LIU Yuhao, QIAN Hong, ZHU Rui. Interpretable Feature Learning and Band Gap Prediction for Titanium-based Semiconductors. Materials Reports, 2026, 40(7): 25040026-8.

链接本文:

https://www.mater-rep.com/CN/10.11896/cldb.25040026 或 https://www.mater-rep.com/CN/Y2026/V40/I7/25040026

1 Bai X, Lu S, Song P, et al. Journal of Colloid and Interface Science, 2024, 664, 716.
2 Zhou J, Xu L, Gai H, et al. Angewandte Chemie International Edition, 2024, 63(36), e202409449.
3 Yang H, Che Y, Cooper A I, et al. Journal of the American Chemical Society, 2023, 145(49), 27038.
4 Li Y, Chen J, Cai P, et al. Journal of Materials Chemistry A, 2018, 6(12), 4948.
5 Wang R, Wang L, Liu R, et al. ACS Nano, 2024, 18(4), 2611.
6 Doonyapisut D, Kannan P K, Kim B, et al. Advanced Intelligent Systems, 2023, 5(8), 2300085.
7 Ma B, Zhang L, Wang W, et al. Green Energy & Environment, 2024, 9(5), 877.
8 Chen Z, Wang H, Liu B, et al. Physical Review B, 2024, 109(8), 085306.
9 Yang H, Che Y, Cooper A, et al. Journal of the American Chemical Society, 2023, 145(49), 27038.
10 Mai H, Le T C, Chen D, et al. Chemical Reviews, 2022, 122(16), 13478.
11 Zhao F, Gao A, Liao Q, et al. Advanced Functional Materials, 2024, 34(17), 2311812.
12 Wang M, Zhu H. ACS Catalysis, 2021, 11(7), 3930.
13 Tamtaji M, Guo X, Tyagi A, et al. ACS Applied Materials & Interfaces, 2022, 14(41), 46471.
14 Wang T, Pan R, Martins M L, et al. Nature Communications, 2023, 14(1), 4607.
15 Zhou Z, Tian D, Yang Y, et al. Current Research in Food Science, 2024, 8, 100679.
16 Grills D C, Ertem M Z, McKinnon M, et al. Coordination Chemistry Reviews, 2018, 374, 173.
17 Han Q, Jiao S, Liu X, et al. Journal of Materials Chemistry A, 2024, 12(24), 14679.
18 Cui Y W, Wang L, Zhang L C. Progress in Materials Science, 2024, 144, 101277.
19 Abd-Elaziem W, Darwish M A, Hamada A, et al. Materials & Design, 2024, 241, 112850.
20 Shokrani A, Dhokia V, Newman S T. Journal of Manufacturing Processes, 2016, 21, 172.
21 Tshephe T S, Akinwamide S O, Olevsky E, et al. Heliyon, 2022, 8(3), e09041
22 Lamiel C, Hussain I, Warner J H, et al. Materials Today, 2023, 63, 313.
23 Takata T, Jiang J, Sakata Y, et al. Nature, 2020, 581(7809), 411.
24 Borlido P, Schmidt J, Huran A W, et al. npj Computational Materials, 2020, 6(1), 1.
25 Furness J W, Kaplan A D, Ning J, et al. The Journal of Physical Che-mistry Letters, 2020, 11(19), 8208.
26 Moss B, Wang Q, Butler K T, et al. Nature Materials, 2021, 20(4), 511.
27 Tang H, Ju T, Dai Y, et al. Journal of Materials Chemistry A, 2022, 10(36), 18978.
28 Zhou J, Xu L, Gai H, et al. Angewandte Chemie, 2024, 136(36), e202409449.
29 Wang Y, Brocks G, Er S. ACS Catalysis, 2024, 14(3), 1336.
30 Benseghir Y, Solé-Daura A, Cairnie D R, et al. Journal of Materials Chemistry A, 2022, 10(35), 18103.
31 Guenther N, Schonlau M. The Stata Journal, 2016, 16(4), 917.
32 Yu C, Yao W. Communications in Statistics-Simulation and Computation, 2017, 46(8), 6261.
33 Bentéjac C, Csörgö A, Martínez-Muñoz G. Artificial Intelligence Review, 2021, 54, 1937.
34 Zhang Y, Xu W, Liu G, et al. PLoS One, 2021, 16(8), e0255637.
35 Wu L, Xiao Y, Ghosh M, et al. ES Materials & Manufacturing, 2020, 9(4), 34.
36 Huang Y, Yu C, Chen W, et al. Journal of Materials Chemistry C, 2019, 7(11), 3238.
37 Reyes-Coronado D, Rodríguez-Gattorno G, Espinosa-Pesqueira M E, et al. Nanotechnology, 2008, 19(14), 145605.
38 Yoshimatsu K, Sakata O, Ohtomo A. Scientific Reports, 2017, 7(1), 12544.
39 Harada S, Tanaka K, Inui H. Journal of Applied Physics, 2010, 108(8), 083703.

[1]	吴浪, 彭李琳, 王则乐, 李浩, 雷斌. 矿渣-水泥胶凝体系早龄期的化学收缩模型及预测研究[J]. 材料导报, 2026, 40(7): 25010012-7.
[2]	董俊涛, 安宗文, 马强, 白学宗. 改进的北方苍鹰算法优化BP神经网络在风电叶片复合材料疲劳预测中的应用[J]. 材料导报, 2026, 40(7): 25020176-7.
[3]	余婉秋, 吴凡, 姜攀星, 王林辉, 王雅东, 詹志刚. 结合机理信息和深度学习算法的质子交换膜燃料电池性能退化预测研究[J]. 材料导报, 2026, 40(7): 25040085-9.
[4]	魏俊飞, 杨成鹏, 贾斐. 不同密度2D-C/SiC复合材料的抗氧化性能研究[J]. 材料导报, 2026, 40(4): 25030254-8.
[5]	龙卓, 傅伊凡, 杨功记. 机器学习在非晶合金逆向设计中的应用[J]. 材料导报, 2026, 40(4): 25020053-9.
[6]	刘轶平, 余金桂, 章桥新. 橡胶的老化与寿命评估[J]. 材料导报, 2026, 40(3): 25030054-11.
[7]	李刊, 魏智强, 路承功, 蒲育. 纳米SiO₂改性聚合物水泥基复合材料孔隙结构演变特征及强度预测[J]. 材料导报, 2025, 39(6): 24070202-10.
[8]	段明翰, 覃源, 李阳, 耿凯强. 寒冷地区腈纶纤维混凝土力学性能及多层感知器神经网络预测[J]. 材料导报, 2025, 39(6): 23110143-9.
[9]	李自强, 崔素萍, 马忠诚, 王亚丽, 王晶, 刘云, 乔志杨. 机器学习算法用于水泥强度预测的研究进展[J]. 材料导报, 2025, 39(5): 23120133-12.
[10]	范明辉, 李薇, 焦宇浩, 任文渊, 罗滔, 赵丁筱烨. 基于SR-SSA-SVR的铁尾矿砂混凝土抗压强度预测[J]. 材料导报, 2025, 39(24): 24120042-8.
[11]	万思宇, 苏三庆, 曹振, 王照耀. 混杂纤维高强高延性水泥基复合材料弯曲性能及预测模型[J]. 材料导报, 2025, 39(23): 24120022-10.
[12]	田根, 朱甫宏, 王文宇, 王晓明, 赵阳, 韩国峰, 任智强, 朱胜. 基于机器学习的传感器监测在金属激光增材制造中的应用[J]. 材料导报, 2025, 39(2): 23080174-16.
[13]	冷建成, 赵雷, 张新, 许宏伟. 基于磁记忆在线监测的再制造毛坯疲劳寿命预测方法[J]. 材料导报, 2025, 39(2): 23040250-6.
[14]	黄耀英, 方晨, 邵成羽, 李泽鹏, 朱赵辉. 非饱和水工混凝土冻融劣化微观孔隙结构与宏观性能关系研究[J]. 材料导报, 2025, 39(19): 24080219-7.
[15]	黄斌彬, 曾磊, 汪超, 孙良福, 胡高兴. 基于Bi-LSTM与改进NSGAIII的混凝土配合比多目标优化[J]. 材料导报, 2025, 39(19): 24090069-8.

[1]	WANG Xu, HE Xin, XIE Zhipeng, ZHANG Da, HOU Shengping, WU Yue, DONG Peng, CHEN Jiale, LIANG Feng. Application of Low Temperature Plasma in the Anode Preparation and Modification of Sodium Ion Battery[J]. Materials Reports, 2026, 40(6): 25040155 -11 .
[2]	HE Geping, FU Zeguo, YANG Quan, BAO Chenhao, XU Rui, LI Mengxuan, JING Gege, GOU Wenhao. Preparation and Properties of Supercapacitor Electrode Material ZnWO₄@rGO Based on Polyvinylpyrrolidone Surfactant[J]. Materials Reports, 2026, 40(6): 25040062 -8 .
[3]	GAO Chenyu, WANG Yan, ZHANG Shaohui, LI Aoyang, WU Jie, HUO Yuren, LU Guannan. Modification of Waste Carbon Fiber Prepreg by Debonder and Its Effect on the Mechanical Properties and Electrical Conductivity of Concrete[J]. Materials Reports, 2026, 40(6): 25050007 -11 .
[4]	LI Yibo, ZHANG Ling, LIANG Zhixia, ZHANG Donghai. Research Progress on the Expansion Performance and Char Strength of Intumescent Fire-retardant Coatings[J]. Materials Reports, 2026, 40(6): 25040263 -12 .
[5]	MEI Shengqi, LIU Xiaodong, WANG Xingju, LI Xufeng, NIE Liangtao, KANG Xuejian. Fused Machine Learning Modeling of Concrete Creep with Strength Classification[J]. Materials Reports, 2026, 40(7): 24100121 -8 .
[6]	ZHANG Zeming, HUANG Cunsheng, DANG Dongying, ZHANG Dewei, WANG Chuansheng. Study on the Properties of BF/EPDM Composites Modified with Sodium Carboxymethyl Cellulose[J]. Materials Reports, 2026, 40(7): 24120160 -6 .
[7]	PENG Huijing, ZHANG Weimin, LI Qianming, XING Xinlong, FAN Jing. Preparation of Alkaline Titanate Material K-Ti and Its Adsorption Properties for Mn(Ⅱ)[J]. Materials Reports, 2026, 40(7): 24120197 -7 .
[8]	WU Lang, PENG Lilin, WANG Zele, LI Hao, LEI Bin. Chemical Shrinkage Model and Prediction for Early-age Slag-Cement Binder Systems[J]. Materials Reports, 2026, 40(7): 25010012 -7 .
[9]	PANG Xueyu, CHENG Guodong, HUANG Xianbin, BAI Yingrui, GAO Zhi, HAO Borong, LYU Kaihe, SUN Jinsheng. The Performance and Mechanisms of High-temperature Retrogression Resistance in Fly Ash and Slag-Modified Well Cementing Systems[J]. Materials Reports, 2026, 40(7): 25010110 -7 .
[10]	WANG Deteng, YANG Huiyong, YU Haijiang, WANG Luyan, WANG Lianyi, LUO Ruiying, HUANG Juntong. Research Status and Prospective of Zirconium Silicate[J]. Materials Reports, 2026, 40(7): 25010149 -14 .

Viewed

Full text

Abstract

Cited

Shared

Discussed