数据驱动辅助高通量筛选阴离子柱撑金属有机框架储氢

Hydrogen storage is a core issue that hinders the development of the hydrogen energy industry, and improving hydrogen storage density is a key technical difficulty. Metal-organic frameworks (MOFs) exhibit excellent hydrogen storage density, and have become one of the most promising energy storage materials. However, conventional computational approaches, including traditional molecular simulations and high-throughput screening methods, encounter significant limitations when applied to MOF hydrogen storage research. These methods are particularly constrained by their excessive computational time requirements and substantial resource consumption, which hinder efficient material discovery and optimization. To address these challenges, this study developed a data-driven high-throughput screening strategy for the rapid prediction of hydrogen storage performance in aniontemplated metal-organic frameworks (AP-MOFs). The proposed method achieved exceptional predictive accuracy, with R² values exceeding 0.99 for both the training and test sets, and required only 63 s of computation time. Through this approach, 20 AP-MOFs with a hydrogen storage density exceeding 5.5%(mass) at 77 K and 5 MPa were identified, surpassing the hydrogen storage target set by the United States Department of Energy. Among these, the ALFFIVE_2_Fe structure, which is potentially synthesizable, exhibited a remarkable hydrogen storage density of 9.75%(mass) under the same conditions, along with a deliverable hydrogen storage density of 3.05%(mass). The study also revealed that the volumetric porosity has the greatest impact on hydrogen storage performance, followed by gravimetric surface area, density and porosity. These findings provide theoretical insights for the future application of AP-MOFs in hydrogen storage technologies.