Chen Zian, Li Haichao, Zhang Chen, Zhang Hongbin, Zhao Yongxiao, Cao Jian, He Tao, Xu Lina, Xiao Hongping, Li Yi, Shao Hezhu, Yang Xiaoyu, He Xiao, Fang Guoyong
Key Laboratory of Carbon Materials of Zhejiang Province, College of Chemistry and Materials Engineering, Wenzhou University, Wenzhou 325035, China.
College of Computer Science and Artificial Intelligence, Wenzhou University, Wenzhou 325035, China.
J Chem Theory Comput. 2024 Nov 12;20(21):9627-9641. doi: 10.1021/acs.jctc.4c01096. Epub 2024 Oct 25.
Crystal structure prediction (CSP) is an important field of material design. Herein, we propose a novel generative adversarial network model, guided by a data-driven approach and incorporating the real physical structure of crystals, to address the complexity of high-dimensional data and improve prediction accuracy in materials science. The model, termed GAN-DDLSF, introduces a novel sampling method called data-driven latent space fusion (DDLSF), which aims to optimize the latent space of generative adversarial networks (GANs) by combining the statistical properties of real data with a standard Gaussian distribution, effectively mitigating the "mode collapse" problem prevalent in GANs. Our approach introduces a more refined generation mechanism specifically for binary crystal structures such as gallium nitride (GaN). By optimizing for the specific crystallographic features of GaN while maintaining structural rationality, we achieve higher precision and efficiency in predicting and designing structures for this particular material system. The model generates 9321 GaN binary crystal structures, with 16.59% reaching a stable state and 24.21% found to be metastable. These results can significantly enhance the accuracy of crystal structure predictions and provide valuable insights into the potential of the GAN-DDLSF approach for the discovery and design of binary, ternary, and multinary materials, offering new perspectives and methods for materials science research and applications.
晶体结构预测（CSP）是材料设计的一个重要领域。在此，我们提出一种新颖的生成对抗网络模型，该模型以数据驱动方法为指导，并融入晶体的真实物理结构，以解决高维数据的复杂性并提高材料科学中的预测准确性。该模型称为GAN-DDLSF，引入了一种名为数据驱动潜在空间融合（DDLSF）的新颖采样方法，旨在通过将真实数据的统计特性与标准高斯分布相结合来优化生成对抗网络（GAN）的潜在空间，有效缓解GAN中普遍存在的“模式崩溃”问题。我们的方法引入了一种专门针对氮化镓（GaN）等二元晶体结构的更精细生成机制。通过在保持结构合理性的同时针对GaN的特定晶体学特征进行优化，我们在预测和设计该特定材料系统的结构方面实现了更高的精度和效率。该模型生成了9321个GaN二元晶体结构，其中16.59%达到稳定状态，24.21%被发现为亚稳态。这些结果可以显著提高晶体结构预测的准确性，并为GAN-DDLSF方法在二元、三元和多元材料的发现和设计中的潜力提供有价值的见解，为材料科学研究和应用提供新的视角和方法。
