Studying equilibrium properties of polymers in solution by atomistic simulations is a challenging task as the available computation time is often not sufficient to ensure representative sampling of the phase space. One approach to tackle this problem is to create a simulation scenario which is simple enough to enable adequate sampling of equilibrium states while it retains the essential parts of the physics of the polymer in solution. In this work, we present and test such a scenario, which is designed for studying whether a given polymer will aggregate or dissolve in a given solvent. Two periodic polymer molecules are simulated in the explicit solvent. The distance d between the polymer chains lends itself as an order parameter so that advanced sampling techniques, such as umbrella sampling, can be applied easily. A state corresponding to dissolved polymers (large d) and a state corresponding to aggregated polymers (small d) can be defined. The scenario misses the intramolecular collapse of the single chain, but it retains full atomistic detail regarding the polymer-solvent and the intermolecular polymer-polymer interactions. The thermoresponsive behavior of PNIPAM in water is studied with the new scenario, and it is shown that quantitative predictions of the experimental equilibrium data can be obtained after adjusting a single state-independent parameter in the force field.

通过原子模拟研究溶液中聚合物的平衡性质是一项艰巨的任务，因为可用的计算时间通常不足以确保对相空间进行有代表性的采样。解决此问题的一种方法是创建一种模拟方案，该方案足够简单，可以对平衡态进行足够的采样，同时保留溶液中聚合物物理的基本组成部分。在这项工作中，我们提出并测试了这种情况，该情况旨在研究给定的聚合物是否会聚集或溶解在给定的溶剂中。在显性溶剂中模拟了两个周期性的聚合物分子。距离d聚合物链之间的间隔本身就是一个有序参数，因此可以轻松应用高级采样技术，例如伞形采样。可以定义与溶解的聚合物相对应的状态（大d）和与聚集的聚合物相对应的状态（小d）。该方案错过了单链的分子内塌陷，但保留了有关聚合物-溶剂和分子间聚合物-聚合物相互作用的完整原子细节。在新的情况下研究了PNIPAM在水中的热响应行为，结果表明，在力场中调整单个与状态无关的参数后，可以获得对实验平衡数据的定量预测。