粗粒晶体图神经网络用于网状材料设计
Vadim Korolev1,2, Artem Mitrofanov1,2
1Department of Chemistry, Lomonosov Moscow State University, Moscow 119991, Russia.
Journal of chemical information and modeling
|March 8, 2024
概括
这项研究引入了用于网状材料的粗粒度晶体图神经网络,以较低的计算成本提供准确的属性预测. 这种方法挑战了材料设计中的以原子为中心的方法.
科学领域:
- 材料科学 材料科学 材料科学
- 计算化学计算化学
- 人工智能的人工智能
背景情况:
- 网状材料如金属有机框架 (MOF) 和共价有机框架 (COF) 有多种应用.
- 由于巨大的化学空间,预测这些材料的性能是复杂的.
- 当前的人工智能模型使用原子级图形,这些图形在计算上可能很昂贵,并且包括多余的功能.
研究的目的:
- 开发一种更有效的材料表示,用于属性预测.
- 在网状材料设计中克服原子级图形神经网络的局限性.
- 为反向材料设计引入粗粒度晶体图方法.
主要方法:
- 开发了一种粗粒度晶体图表表示,专注于分子构建单元.
- 评估神经网络的性能,使用基于构成,晶体结构意识和粗粒度模型.
- 评估不同表示的预测准确性和能源效率.
主要成果:
- 粗粒晶体图神经网络表现出具有竞争力的准确性,计算成本明显降低.
- 提出的方法是原子级图形神经网络的可行替代方案.
- 模型成功地集成到一个反向材料设计管道中.
结论:
- 粗粒晶体图框架为网状材料属性预测提供了一种高效和准确的方法.
- 这种方法挑战了材料设计中传统的以原子为中心的观点.
- 它为加速发现新型网状材料提供了宝贵的工具.
相关概念视频
Ionic Crystal Structures
Ionic crystals consist of two or more different kinds of ions that usually have different sizes. The packing of these ions into a crystal structure is more complex than the packing of metal atoms that are the same size.
Most monatomic ions behave as charged spheres, and their attraction for ions of opposite charge is the same in every direction. Consequently, stable structures for ionic compounds result (1) when ions of one charge are surrounded by as many ions as possible of the opposite...
Most monatomic ions behave as charged spheres, and their attraction for ions of opposite charge is the same in every direction. Consequently, stable structures for ionic compounds result (1) when ions of one charge are surrounded by as many ions as possible of the opposite...
Network Covalent Solids
Network covalent solids contain a three-dimensional network of covalently bonded atoms as found in the crystal structures of nonmetals like diamond, graphite, silicon, and some covalent compounds, such as silicon dioxide (sand) and silicon carbide (carborundum, the abrasive on sandpaper). Many minerals have networks of covalent bonds.
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
Neural Circuits
Neural circuits and neuronal pools are two of the main structures found in the nervous system. Neural circuits are networks of neurons that work together to carry out a specific task or process. They consist of interconnected neurons and glial cells, which provide structural and metabolic support.
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
Crystal Density
The crystal lattice structure of a material allows us to determine how many molecules exist in its unit cell. With this information, alongside the unit-cell parameters - three distance parameters (a, b, c) and three angular parameters (α, β, γ).Density (ρ) = (Z × M) / (a × b × c × NA)where:Z is the number of formula units per unit cellM is the molar mass of the substancea, b, and c are the edge lengths of the unit cellNA is Avogadro’s numberFor a simple cubic lattice, atoms are located only at...
Lattice Energies of Ionic Crystals
Lattice energy represents the energy released when gaseous cations and anions combine to form an ionic solid, reflecting the strength of electrostatic interactions within the crystal. This process is fundamentally governed by Coulombic attraction between oppositely charged ions, where the potential energy varies inversely with the interionic distance and directly with the product of ionic charges. As ions approach one another, the electrostatic energy becomes increasingly negative, indicating a...
Imperfections in Crystal Structure: Non-Stoichiometric Defects
Non-stoichiometric defects refer to a type of defect in the crystal structure of a compound where the ratio of its constituent elements deviates from the ideal stoichiometric ratio. There are two main types of non-stoichiometric defects: metal excess defects and metal deficiency defects.Metal excess defects occur when there is a slight surplus of metal ions than what is required by the stoichiometric ratio of the compound. For example, heating a sodium chloride crystal in sodium vapor results...


