To study the application of different potential forms to describe phase equilibrium for Structure I gas hydrates, molecular computations and sensitivity analyses were performed, and results were compared. Using 3-phase monovariant pressure-temperature data, "site-site" and "molecule-molecule" Lennard-Jones and Kihara potential parameters were fit via explicit quadrature calculations for water clathrates with guest molecules of methane (CH_4), ethane (C_2H_6), and cyclopropane (C_3H_6). Although the Lennard-Jones and Kihara potential forms can be fit satifactorily to experimental P-T data for ethane and cyclopropane hydrate using the van der Waals and Platteeuw model, they fail to predict accurately cage occupacies of methane hydrates. To correct the inherent inconsistencies of using fitted potentials, an H_2O-CH_4 bimolecular potential developed and validated in our earlier ab initio study was employed. The intermolecular potential, mapped from first-principles calculations, captures the molecular interaction correctly when compared with experimental 2nd virial coefficient data. Predicted phase equilibria and cage occupancies for methane hydrates using the ab initio potential are in close agreement with experimental P-T data and measured cage occupancies. Moreover, use of this model independent potential helps to validate the van der Waals-Platteeuw statistical model, which is used almost exclusively to model the phase behavior of clathrate hydrates. Other potential forms including exp-6 were fit to the ab initio potential to predict phase equilibria, and Optimized Potential for Liquid Simulations (OPLS) was also used to predict cage occupancies. The comparison showed that only the first-principles ab initio potential is able to physically characterize both the microscopic and macroscopic behaviors of methane hydrates. The reasons for this are the following. (1) The ab initio potential is directly related to physical properties, whereas the parameters of pre-chosen potential forms are merely ab hoc fitted quantities. (2) Angular degrees of freedom need to be accurately included in "site-site" interaction potentials, and they are in the initio case, buy not in the case of most potential forms currently used for calculations of hydrates.
展开▼
机译：为了研究不同势能形式用于描述结构I气体水合物的相平衡的方法，进行了分子计算和灵敏度分析，并对结果进行了比较。使用三相单变量压力-温度数据，通过显式正交计算，用甲烷（CH_4），乙烷（C_2H_6）的客体分子对水包合物进行拟合，拟合了“现场”和“分子” Lennard-Jones和Kihara势参数，和环丙烷（C_3H_6）。尽管Lennard-Jones和Kihara势能形式可以使用范德华斯和普拉特约模型令人满意地拟合乙烷和环丙烷水合物的实验P-T数据，但它们无法准确预测甲烷水合物的笼式占有率。为了纠正使用拟合电势的固有矛盾，我们在较早的从头算研究中开发并验证了H_2O-CH_4双分子电势。从第一性原理计算得出的分子间电势与实验第二维里系数数据相比，可以正确捕获分子相互作用。使用从头算势预测的甲烷水合物的相平衡和笼占有率与实验P-T数据和测得的笼占有率非常吻合。此外，使用该模型的独立电势有助于验证van der Waals-Platteeuw统计模型，该模型几乎专门用于建模笼形水合物的相行为。包括exp-6在内的其他势能形式也适合从头算势，以预测相平衡，并且液体模拟的优化势（OPLS）也用于预测笼形占有率。比较结果表明，只有第一原理的从头算势能够从物理上表征甲烷水合物的微观和宏观行为。其原因如下。 （1）从头开始电势与物理性质直接相关，而预先选择的电势形式的参数仅是临时拟合量。 （2）角自由度需要准确地包含在“站点-站点”交互电位中，并且从一开始就是如此，对于当前用于计算水合物的大多数潜在形式，则不必购买。
展开▼
