Department of Mechanical and Industrial Engineering, Ryerson University, 350 Victoria St., Toronto, Ontario M5B2K3, Canada.
J Chem Phys. 2018 Jul 21;149(3):034502. doi: 10.1063/1.5031004.
Thermodiffusion or the Ludwig-Soret effect is known as the cross effect between the temperature gradient and induced separation of mixture species in multicomponent mixtures. The performance of the boundary driven non-equilibrium molecular dynamics enhanced heat exchange (eHEX) algorithm was validated by evaluating the sign and magnitude of the thermodiffusion process in methane/n-butane/n-dodecane (nC-nC-nC) ternary mixtures. The eHEX algorithm consists of an extended version of the HEX algorithm with an improved energy conservation property. In addition to this, the transferable potentials for phase equilibria-united atom augmented force field was employed in all molecular dynamics (MD) simulations to accurately represent molecular interactions in the fluid. Our newly employed MD algorithm was capable to appropriately reflect the thermophobicity concept and the coupled effect of relative density and mole fraction of the mixture species on the thermodiffusion process. The separation ratio of the ternary mixture for five different compositions (at 333.15 K and 35 MPa) showed good agreement with experimental data and better accuracy in predicting the sign change of the intermediate component (nC) as its concentration in the mixture increases, when compared to other MD models.
热扩散或路德维希-索雷特效应是指多组分混合物中温度梯度和混合物种诱导分离之间的交叉效应。通过评估甲烷/正丁烷/正十二烷（nC-nC-nC）三元混合物中热扩散过程的符号和大小，验证了边界驱动非平衡分子动力学增强热交换（eHEX）算法的性能。eHEX 算法由 HEX 算法的扩展版本组成，具有改进的能量守恒特性。此外，在所有分子动力学（MD）模拟中都使用了可传递相平衡-联合原子增强力场，以准确表示流体中的分子相互作用。我们新采用的 MD 算法能够适当反映疏液性概念以及混合物物种的相对密度和摩尔分数对热扩散过程的耦合效应。对于五种不同组成（在 333.15 K 和 35 MPa 下）的三元混合物的分离比与实验数据吻合较好，并且在预测中间组分（nC）的符号变化时，与其他 MD 模型相比，其准确性更高，因为它在混合物中的浓度增加。
