Density Functional Methods for Fast Screening of Metal Organic Frameworks for Hydrogen Storage
Density Functional Methods for Fast Screening of Metal Organic Frameworks for Hydrogen Storage
复制标题
DOI：
10.1021/jp505963m
复制
发表时间：
2015-03-12
影响因子：
3.7
通讯作者：
Wu, Jianzhong
中科院分区：
文献类型：
--
作者：
Fu, Jia;Liu, Yu;Wu, Jianzhong
经典密度泛函理论（DFT）通常用于通过气体物理吸附来表征多孔材料的孔径分布和比表面积。然而，其在气体储存材料的大规模筛选中的应用在很大程度上尚未得到探索，因为人们普遍认为 DFT 计算仅适用于一维系统，并且对自由能泛函中引入的近似高度敏感。在这项工作中，我们研究了非局域密度泛函的四个代表性版本，用于在广泛的温度和压力范围内使用狭缝孔模型和大型金属有机框架 (MOF) 库来预测 H2 吸附。 DFT 的四个版本共享修正的基本测度理论中的一个共同泛函，用于解释分子排除体积效应，同时在表示分子间吸引力的近似方面有所不同，即平均场近似、两种版本的加权密度近似 (WDA) 和二次泛函展开法。我们通过与实际感兴趣条件下氢气吸附的蒙特卡罗模拟数据进行广泛比较，测试了这些泛函。总体而言，与模拟结果相比，所有四个版本的 DFT 都相当准确。虽然密度展开法在 DOE 储氢目标条件下表现相当良好，但 WDA 方法在低温（通常用于材料表征的条件）下最为准确。除了快速预测吸附等温线之外，DFT 还能够生成分子密度分布，揭示有利的吸附位点等微观细节。从计算角度来看，DFT计算速度比传统模拟方法至少快1个数量级，有望大规模筛选用于气体储存的纳米结构材料。
Classical density functional theory (DFT) has been routinely used for the characterization of pore size distribution and specific surface area of porous materials by gas physisorption. However, its application to large-scale screening of materials for gas storage has been largely unexplored because it is commonly believed that the DFT calculations are applicable only to one-dimensional systems and highly sensitive to the approximations introduced in the free-energy functionals. In this work, we have investigated four representative versions of nonlocal density functionals for predicting H2 adsorption using both the slit pore model and a large library of metalorganic frameworks (MOFs) under a broad range of temperatures and pressures. The four versions of DFT share a common functional from the modified fundamental measure theory to account for the molecular excluded volume effects while differing in their approximations to represent the intermolecular attractions, viz., mean-field approximation, two versions of weighted-density approximations (WDA), and the quadratic functional expansion method. We have tested these functionals by extensive comparison with Monte Carlo simulation data for H2 adsorption at conditions of practical interest. Overall all four versions of DFT are reasonably accurate in comparison with the simulation results. While the density expansion method performs rather well at the DOE target condition for hydrogen storage, the WDA methods are found most accurate at low temperature, a condition typically used in materials characterization. In addition to rapid prediction of the adsorption isotherms, DFT is able to generate molecular density profiles revealing microscopic details such as favorable adsorption sites. From a computational perspective, the DFT calculation is at least 1 order of magnitude faster than conventional simulation methods, promising for large-scale screening of nanostructured materials for gas storage.