无序碳单层的电子结构及氢原子吸附的第一性原理

本研究运用第一性原理计算方法,系统地研究了无序碳单层材料不同位点的电子结构及其析氢性能.计算结果显示无序结构中的C-C键相比于石墨烯中的C-C键在26.7%的范围内有不同程度的拉伸或压缩,使得C原子电荷在-0.17～+0.16个电子范围内变化,导致部分C原子电子局域化.电子的局域化增强了C原子的化学活性,从而表现出了较强的吸附性能.我们发现H原子与C原子的键合及析氢性能与C原子间的键角相关.对于三配位的碳原子,其中三个价电子通过sp~2杂化轨道与最邻近的碳原子结合形成较强的共价键,而余下的一个p_z轨道电子可以与H原子在垂直于原子层的方向形成较弱的化学键.无序结构可以打破三个sp~2杂化轨道的对称性,进而影响p_z轨道与氢的成键.本研究在一定程度上揭示了单层无序碳材料结构-性能的构效关系,为实验上设计特定性能的无序碳功能材料提供理论指导.     

Interfacial defect engineering and photocatalysis properties of hBN/MX2 (M = Mo,W, and X = S,Se heterostructures

Van der Waals (VDW) heterostructures have attracted significant research interest due to their tunable interfacial properties and potential applications in many areas such as electronics, optoelectronic, and heterocatalysis. In this work, the influences of interfacial defects on the electronic structures and photocatalytic properties of hBN/MX₂ (M = Mo, W, and X=S, Se) are studied using density functional theory calculations. The results reveal that the band alignment of hBN/MX₂ can be adjusted by introducing vacancies and atomic doping. The type-I band alignment of the host structure is maintained in the heterostructure with n-type doping in the hBN sublayer. Interestingly, the band alignment changed into the type-II heterostructrue due to V_B defect and p-type doping is introduced into the hBN sublayer. This can conduce to the separation of photo-generated electron-hole pairs at the interfaces, which is highly desired for heterostructure photocatalysis. In addition, two Z-type heterostructures including hBN(Be_B)/MoS₂, hBN(Be_B)/MoSe₂, and hBN(V_N)/MoSe₂ are achieved, showing the decreasing of band gap and ideal redox potential for water splitting. Our results reveal the possibility of engineering the interfacial and photocatalysis properties of hBN/MX₂ heterostructures via interfacial defects.     

基于材料基因工程的储氢材料数据库构建及其应用

高密度储氢材料的加速研发对于我国能源经济转型、早日实现双碳目标至关重要.集成高通量计算、数据库及机器学习预测的数据驱动材料研发新范式有望缩短研发周期并降低研发成本.由于组分、结构、工艺及形貌等多重复杂性,目前储氢材料相关的数据驱动性能预测研究较少,尚缺乏一个较为系统的性质性能数据库.因此,本文中我们开发了智能化的数据挖掘引擎,通过已发表的学术论文中发掘储氢材料热力学、动力学储氢性能数据,以及现有的材料基因工程数据库数据中获取含氢材料物理化学性质,并结合高通量第一性原理计算数据,构建了储氢材料性质性能数据集.基于所构建的数据集进一步建立了储氢材料数据库,并应用晶体图形神经网络等机器学习方法对储氢材料的吸放氢质量、吸放氢温度进行预测.相关工作将数据驱动的材料研发新模式与储氢材料相结合,为发展实用高效的新型储氢材料提供有效的平台支持、数据支撑、方法指引.