Molecular Modeling of Thermodynamic Properties of Fluids and Fluid Mixtures

我的研究目标是基于化学工程中出现的复杂流体系统的基本统计力学和量子力学描述的计算算法的开发和应用。目的是预测宏观热力学性质和阐明其潜在的分子基础。这种能力有助于改进基于流体的系统和过程的设计。宏观模型（例如，状态方程）通常用于基于计算机的过程模拟器的性质预测。虽然它们是为了满足热力学和其他科学定律的约束而构建的，但它们保留了重要的经验成分，并包含多个参数，这些参数的确定需要大量的实验数据。我们的方法是使用相对简单的（需要相对较少的参数）分子模型来描述潜在的系统分子间和分子内的力，并使用基本精确的统计力学机制进行研究。计算机硬件和软件能力的迅速提高和成本的降低使这种方法适用于日益复杂的系统，而且比单独使用实验研究要便宜得多。目标是使用这种方法来预测以前未研究过的化合物和过程的性质和行为，这些化合物和过程的实验数据要么不可用，要么很少。这样的预测为模型系统产生了精确的伪实验数据，这些数据可以在传统宏观热力学模型的背景下进行分析和解释。这产生了在过程模拟器中执行计算快速预测的能力。分子模拟方法更广泛应用的主要障碍是缺乏用于模拟某些性质和过程的计算效率高的算法。所需算法的开发需要综合对宏观化学工程热力学和复杂流体混合物统计力学的深刻理解，我的小组在这两方面都有相当的经验。该计划的重点是开发和应用模拟化学反应和相平衡的算法，以及水电解质系统的热化学性质的建模。前一个项目的目标是改进该领域现有的主要算法，该算法是由我的小组在20世纪90年代中期开发的，现在已被世界各地的研究人员广泛使用。本提案的应用涵盖了各种各样的过程和应用，包括传统的工程流体系统和生物学中出现的水系统。

The goal of my research program is the development and application of computational algorithms based on fundamental statistical mechanical and quantum mechanical descriptions of the complex fluid systems arising in chemical engineering. The objectives are the prediction of macroscopic thermodynamic properties and elucidation of their underlying molecular basis. Such capabilities are useful in facilitating the improved design of fluid-based systems and processes.Macroscopic models (e.g., equations of state) are conventionally used for property prediction in computer-based process simulators. Although constructed to satisfy the constraints of thermodynamic and other scientific laws, they retain a significant empirical component and contain multiple parameters, whose determination requires significant amounts of experimental data. Our approach is to use relatively simple (requiring relatively few parameters) molecular models describing the underlying system inter- and intra-molecular forces, coupled with their study using the fundamentally exact machinery of statistical mechanics. The rapidly improving capabilities and decreasing costs of computer hardware and software make this approach applicable to increasingly complex systems, and is much less expensive than the use of experimental investigations alone. The goal is to use this methodology to make predictions of the properties and behaviour of previously unstudied compounds and processes, for which experimental data is either unavailable or sparse. Such predictions generate exact pseudo-experimental data for the model systems that can be analyzed and interpreted in the context of conventional macroscopic thermodynamic models. This yields the ability to perform computationally rapid predictions within process simulators. The main impediment to the more widespread application of molecular simulation methodology is the lack of computationally efficient algorithms for the simulation of certain properties and processes. The development of the required algorithms requires the integration of deep understanding of both macroscopic chemical engineering thermodynamics and the statistical mechanics of complex fluid mixtures, in both of which my group has considerable experience.The proposed project focuses on the development and applications of algorithms for the simulation of chemical reaction and phase equilibria, and on the modeling of thermochemical properties of aqueous electrolyte systems. A goal of the former project is to improve upon the primary existing algorithm in that area, developed by my group in the mid 1990's and which is now in widespread use by researchers worldwide. Applications of this proposal encompass a wide variety of processes and applications, including both traditional engineering fluid systems and aqueous systems arising in biology.