Graph neural network guided design of novel deep-ultraviolet optical materials with high birefringence

IF 6.8 2区材料科学 Q1 MATERIALS SCIENCE, MULTIDISCIPLINARY

Science China Materials Pub Date : 2024-10-22 DOI:10.1007/s40843-024-3114-4

Ivan A. Kruglov (, ), Liudmila A. Bereznikova (, ), Congwei Xie (, ), Dongdong Chu (, ), Ke Li (, ), Evgenii Tikhonov (, ), Abudukadi Tudi (, ), Arslan Mazitov (, ), Min Zhang (, ), Shilie Pan (, ), Zhihua Yang (, )

{"title":"Graph neural network guided design of novel deep-ultraviolet optical materials with high birefringence","authors":"Ivan A. Kruglov \n (, ), Liudmila A. Bereznikova \n (, ), Congwei Xie \n (, ), Dongdong Chu \n (, ), Ke Li \n (, ), Evgenii Tikhonov \n (, ), Abudukadi Tudi \n (, ), Arslan Mazitov \n (, ), Min Zhang \n (, ), Shilie Pan \n (, ), Zhihua Yang \n (, )","doi":"10.1007/s40843-024-3114-4","DOIUrl":null,"url":null,"abstract":"<div><p>Finding crystals with high birefringence (Δ<i>n</i>), especially in deep-ultraviolet (DUV) regions, is important for developing polarization devices such as optical fiber sensors. Such materials are usually discovered using experimental techniques, which are costly and inefficient for a large-scale screening. Herein, we collected a database of crystal structures and their optical properties and trained atomistic line graph neural network to predict their Δ<i>n</i>. To estimate the level of confidence of the trained model on new data, D-optimality criterion was implemented. Using trained graph neural network, we searched for novel materials with high Δ<i>n</i> in the Materials Project database and discovered two new DUV birefringent candidates: NaYCO<sub>3</sub>F<sub>2</sub> and SClO<sub>2</sub>F, with high Δ<i>n</i> values of 0.202 and 0.101 at 1064 nm, respectively. Further analysis reveals that strongly anisotropic units with various anions and <i>π</i>-conjugated planar groups are beneficial for high Δ<i>n</i>.</p><div><figure><div><div><picture><source><img></source></picture></div></div></figure></div></div>","PeriodicalId":773,"journal":{"name":"Science China Materials","volume":"67 12","pages":"3941 - 3947"},"PeriodicalIF":6.8000,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Science China Materials","FirstCategoryId":"88","ListUrlMain":"https://link.springer.com/article/10.1007/s40843-024-3114-4","RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MATERIALS SCIENCE, MULTIDISCIPLINARY","Score":null,"Total":0}

引用次数: 0

Abstract

Finding crystals with high birefringence (Δn), especially in deep-ultraviolet (DUV) regions, is important for developing polarization devices such as optical fiber sensors. Such materials are usually discovered using experimental techniques, which are costly and inefficient for a large-scale screening. Herein, we collected a database of crystal structures and their optical properties and trained atomistic line graph neural network to predict their Δn. To estimate the level of confidence of the trained model on new data, D-optimality criterion was implemented. Using trained graph neural network, we searched for novel materials with high Δn in the Materials Project database and discovered two new DUV birefringent candidates: NaYCO₃F₂ and SClO₂F, with high Δn values of 0.202 and 0.101 at 1064 nm, respectively. Further analysis reveals that strongly anisotropic units with various anions and π-conjugated planar groups are beneficial for high Δn.

查看原文本刊更多论文

图神经网络引导的高双折射新型深紫外光学材料设计

寻找具有高双折射（Δn）的晶体，尤其是在深紫外（DUV）区域，对于开发光纤传感器等偏振设备非常重要。这类材料通常是通过实验技术发现的，而实验技术对于大规模筛选来说成本高、效率低。在此，我们收集了晶体结构及其光学特性数据库，并训练原子线图神经网络来预测它们的Δn。为了估算训练模型对新数据的置信度，我们采用了 D-optimality 准则。利用训练有素的图神经网络，我们在材料项目数据库中搜索了具有高Δn 的新型材料，并发现了两种新的 DUV 双折射候选材料：NaYCO3F2 和 SClO2F，在 1064 纳米波长处的Δn 值分别高达 0.202 和 0.101。进一步的分析表明，带有各种阴离子和 π 共轭平面基团的强各向异性单元有利于获得高 Δn。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Science China Materials Materials Science-General Materials Science

CiteScore

11.40

自引率

7.40%

发文量

949

期刊介绍： Science China Materials (SCM) is a globally peer-reviewed journal that covers all facets of materials science. It is supervised by the Chinese Academy of Sciences and co-sponsored by the Chinese Academy of Sciences and the National Natural Science Foundation of China. The journal is jointly published monthly in both printed and electronic forms by Science China Press and Springer. The aim of SCM is to encourage communication of high-quality, innovative research results at the cutting-edge interface of materials science with chemistry, physics, biology, and engineering. It focuses on breakthroughs from around the world and aims to become a world-leading academic journal for materials science.