A Connected Convolutional Neutral Network Protocol for Design of Two-Dimensional Materials Based on Modified Graphdiyne,The Journal of Physical Chemistry Letters

当前位置： X-MOL 学术 › J. Phys. Chem. Lett. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

A Connected Convolutional Neutral Network Protocol for Design of Two-Dimensional Materials Based on Modified Graphdiyne
The Journal of Physical Chemistry Letters ( IF 4.8 ) Pub Date : 2024-07-25 , DOI: 10.1021/acs.jpclett.4c01485
Junqing Li ₁ , Ziyi Liu ₁ , Zhehuan Zhao ₂ , Dongqi Wang ₁

Affiliation

In materials science, doping plays a crucial role in manipulating the electronic properties of materials. Conventional screening via a trial-and-error strategy is challenging owing to the enormous chemical space. We proposed a connected convolutional neutral network (CCNN) for quick screening of boron nitrogen (B–N) codoped graphdiyne in terms of band gap. A paired-atomic localized matrix (PALM) descriptor was designed to describe the local chemical environment of materials with the matrix form adapted to a neutral network. An attribution analysis was conducted, and a quantitative relationship between structure and band gap is proposed, which reveals more significant influence of B–N doping at sp² hybridized sites than at sp hybridized sites on broadening of the band gap of GDY. The accuracy and efficiency of the proposed approach implicate its potential in promoting the design of graphdiyne-based optoelectronic devices and catalysts with expected electronic properties, opening a new avenue for rational design of novel materials.

中文翻译：

基于改性石墨炔的二维材料设计的连接卷积神经网络协议

在材料科学中，掺杂在控制材料的电子特性方面起着至关重要的作用。由于化学空间巨大，通过试错策略进行的传统筛选具有挑战性。我们提出了一种连接卷积神经网络（CCNN），用于根据带隙快速筛选硼氮（B-N）共掺杂石墨炔。设计了配对原子局域矩阵（PALM）描述符，以适应中性网络的矩阵形式来描述材料的局域化学环境。进行了归因分析，提出了结构与带隙之间的定量关系，揭示了 sp ²杂化位点的 B-N 掺杂对 GDY 带隙展宽的影响比 sp 杂化位点的影响更显着。该方法的准确性和效率意味着其在促进具有预期电子性能的基于石墨二炔的光电器件和催化剂的设计方面具有潜力，为新型材料的合理设计开辟了新途径。

更新日期：2024-07-25

点击分享查看原文

点击收藏

阅读更多本刊新发论文本刊介绍/投稿指南