Walker Pierre J, Zhao Tianpu, Haslam Andrew J, Jackson George
Department of Chemical Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, United Kingdom.
J Chem Phys. 2022 Apr 21;156(15):154106. doi: 10.1063/5.0087125.
A methodology for obtaining molecular parameters of a modified statistical associating fluid theory for variable-range interactions of Mie form (SAFT-VR Mie) equation of state (EoS) from ab initio calculations is proposed for non-associative species that can be modeled as single spherical segments. The methodology provides a strategy to map interatomic or intermolecular potentials obtained from ab initio quantum-chemistry calculations to the corresponding Mie potentials that can be used within the SAFT-VR Mie EoS. The inclusion of corrections for quantum and many-body effects allows for an excellent, fully predictive description of the vapor-liquid envelope and other bulk thermodynamic properties of noble gases; this description is of similar or superior quality to that obtained using SAFT-VR Mie with parameters regressed in the traditional way using experimental thermodynamic-property data. The methodology is extended to an anisotropic species, methane, where similar levels of accuracy are obtained. The efficacy of using less-accurate quantum-chemistry methods in this methodology is explored, showing that these methods do not provide satisfactory results, although we note that the description is nevertheless substantially better than those obtained using the conductor-like screening model for describing real solvents (COSMO-RS), the only other fully predictive ab initio method currently available. Overall, the reliance on thermophysical data is completely dispensed with, providing the first extensible, wholly predictive SAFT-type EoSs.
针对可建模为单个球形链节的非缔合物种,提出了一种从从头算计算中获取用于米氏形式变程相互作用的修正统计缔合流体理论(SAFT-VR Mie)状态方程(EoS)分子参数的方法。该方法提供了一种策略,可将从从头算量子化学计算中获得的原子间或分子间势映射到相应的米氏势,这些米氏势可用于SAFT-VR Mie EoS。包含量子和多体效应的修正后,能够对稀有气体的气液包络和其他体相热力学性质进行出色的、完全预测性的描述;这种描述的质量与使用传统方式根据实验热力学性质数据回归参数的SAFT-VR Mie所获得的描述相似或更优。该方法扩展到了各向异性物种甲烷,也获得了类似的准确度。探索了在该方法中使用精度较低的量子化学方法的效果,结果表明这些方法无法提供令人满意的结果,不过我们注意到,尽管如此,该描述仍比使用描述真实溶剂的导体类屏蔽模型(COSMO-RS)所获得的描述要好得多,COSMO-RS是目前唯一可用的另一种完全预测性的从头算方法。总体而言,完全摒弃了对热物理数据的依赖,提供了首个可扩展的、完全预测性的SAFT型EoS。