Department of Mechanical Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802, USA.
J Chem Phys. 2019 Sep 21;151(11):114701. doi: 10.1063/1.5118888.
A simple mean-field model of carbon-water interactions was developed to predict the binding energy in classical simulations for graphene and graphite surfaces. Using this model, analytical expressions were derived to link microscopic parameters (such as the binding energy) with macroscopic wetting behavior (work of adhesion). Adding these expressions to an optimized mean-field model of wettability, the empirical relationship between the binding energy and the work of adhesion in classical simulations was formally explained. An orientation dependent mean-field model and the insight gained from mean field modeling of the binding energy were used to develop a method to optimize comprehensive carbon-water interaction potentials, where molecular orientation is taken into account using data from state-of-the-art high-resolution multibody electronic structure methods. This method eliminates the ambiguity of finding a set of four parameters by informing on the bounds for the parameter-search process using physics-informed constraints.
开发了一个简单的碳-水相互作用的平均场模型，用于预测经典模拟中石墨烯和石墨表面的结合能。利用该模型，推导出了一些解析表达式，将微观参数（如结合能）与宏观润湿行为（粘附功）联系起来。将这些表达式添加到优化的润湿平均场模型中，就可以从经典模拟中正式解释结合能与粘附功之间的经验关系。基于取向相关的平均场模型和从结合能的平均场建模中获得的深入理解，开发了一种方法来优化综合的碳-水相互作用势能，其中使用最先进的高精度多体电子结构方法的数据来考虑分子取向。该方法通过使用物理信息约束来告知参数搜索过程的边界，从而消除了寻找一组四个参数的歧义。
