AI 中文总结
该研究提出AdsZeo坐标解析数据集，包含191种沸石拓扑的甲烷吸附构型数据，可用于吸附分析及相关机器学习模型构建。
AI 中文摘要
沸石是广泛应用于吸附、分离和催化过程的结晶纳米多孔材料。分子模拟常被用于预测吸附性质，但大多数高通量吸附数据集仅报告负载量或等温线等系综平均量，而非获得这些平均值的分子构型。本文提出AdsZeo，这是一个针对含铝、含钠沸石骨架中甲烷吸附平衡构型的坐标解析数据集。本次发布的处理后数据包含来自191种沸石拓扑结构的4775个骨架实例，每个骨架实例在298 K下，于0.1至100 bar之间的13个甲烷压力下，采用巨正则蒙特卡洛模拟进行计算，共得到62075次生产模拟。除标量吸附记录外，该数据集还在处理后的DuckDB数据库中存储了生产帧甲烷伪原子坐标、可移动Na⁺阳离子坐标、骨架原子坐标、逐帧负载量和能量统计数据以及模拟元数据。本次发布包含12415000条已保存的生产帧记录和1245376215条已保存的粒子坐标记录。AdsZeo提供了跨骨架拓扑结构、铝含量及分布、钠阳离子排列、压力和甲烷负载量变化的坐标解析吸附数据，可复用用于吸附分析、空间统计、密度估计以及带电沸石孔中分子构型生成的机器学习模型。
英文摘要
Zeolites are crystalline nanoporous materials widely used in adsorption, separation, and catalytic processes. Molecular simulations are commonly used to predict adsorption properties, but most high-throughput adsorption datasets report only ensemble-averaged quantities such as loadings or isotherms, rather than the molecular configurations from which these averages are obtained. Here, we present AdsZeo, a coordinate-resolved dataset of equilibrium methane adsorption configurations in aluminium-substituted, sodium-containing zeolite frameworks. The processed release contains 4,775 framework realisations derived from 191 zeolite topologies. Each framework realisation was simulated at 13 methane pressures between 0.1 and 100 bar at 298 K using grand canonical Monte Carlo simulations, giving 62,075 production simulations in total. In addition to scalar adsorption records, the dataset stores production-frame methane pseudo-atom coordinates, mobile Na$^+$ cation coordinates, framework atomic coordinates, per-frame loading and energy statistics, and simulation metadata in a processed DuckDB database. The release contains 12,415,000 saved production-frame records and 1,245,376,215 saved particle-coordinate records. AdsZeo provides coordinate-resolved adsorption data across variations in framework topology, aluminium content and distribution, sodium cation arrangement, pressure, and methane loading, enabling reuse for adsorption analysis, spatial statistics, density estimation, and machine-learning models for molecular configuration generation in charged zeolite pores.
