论文标题
通过深层生成模型对新型2D材料的数据驱动发现
Data-driven discovery of novel 2D materials by deep generative models
论文作者
论文摘要
生成具有良好稳定性特性的候选晶体结构的有效算法可以在数据驱动的材料发现中起关键作用。在这里，我们表明，晶体扩散变化自动编码器（CDVAE）能够生成高化学和结构多样性和形成能的二维（2D）材料，反映了训练结构。具体而言，我们在2615 2D材料上训练CDVAE，其能量高于凸壳$ΔH_ {\ Mathrm {hull}} <0.3 $ EV/ATOM，并生成5003材料，我们使用密度功能理论（DFT）放松放松。我们还通过系统的元素取代训练结构生成14192个新晶体。我们发现，生成模型和晶格装饰方法是互补和产量材料，具有相似的稳定性，但晶体结构和化学成分非常不同。总共我们发现11630个预测了新的2D材料，其中8599个具有$ΔH_ {\ mathrm {hull}} <0.3 $ ev/Atom作为种子结构，而2004年在凸壳的50 meV之内，可以合成潜在的。所有材料的松弛原子结构都可以在开放的计算2D材料数据库（C2DB）中获得。我们的工作将CDVAE确定为有效且可靠的晶体生成机器，并显着扩大了2D材料的空间。
Efficient algorithms to generate candidate crystal structures with good stability properties can play a key role in data-driven materials discovery. Here we show that a crystal diffusion variational autoencoder (CDVAE) is capable of generating two-dimensional (2D) materials of high chemical and structural diversity and formation energies mirroring the training structures. Specifically, we train the CDVAE on 2615 2D materials with energy above the convex hull $ΔH_{\mathrm{hull}}< 0.3$ eV/atom, and generate 5003 materials that we relax using density functional theory (DFT). We also generate 14192 new crystals by systematic element substitution of the training structures. We find that the generative model and lattice decoration approach are complementary and yield materials with similar stability properties but very different crystal structures and chemical compositions. In total we find 11630 predicted new 2D materials, where 8599 of these have $ΔH_{\mathrm{hull}}< 0.3$ eV/atom as the seed structures, while 2004 are within 50 meV of the convex hull and could potentially be synthesized. The relaxed atomic structures of all the materials are available in the open Computational 2D Materials Database (C2DB). Our work establishes the CDVAE as an efficient and reliable crystal generation machine, and significantly expands the space of 2D materials.
