Reactive coarse-grained molecular dynamics simulations of nanoparticle/polymer nanocomposites

对新型高性能材料日益增长的技术需求对当今的材料科学家提出了重大挑战。合成纳米复合材料越来越多地用于必须承受极端机械载荷的条件。在这些材料的设计中，由于建模技术和计算能力的最新发展，计算机模拟技术变得不可避免。然而，当涉及到预测纳米复合材料的性能时，目前的技术要么考虑到足够大的系统（即原子模拟），要么考虑到正确的底层物理（即连续介质力学）。在这个雄心勃勃的项目中，它位于物理，化学和工程的联系，我们提出了一个解决这些问题的方法。在此基础上，我们提出了一种新的多尺度模拟协议，在此基础上对现有的一些技术进行了适应和改进，以建立纳米颗粒增强聚合物纳米复合材料的粗粒度模型。本项目的特点是利用全原子分子动力学模拟的映射程序推导反应性粗粒度力场。它可以模拟大型原子系统，准确描述化学键的断裂，这对于模拟复合材料的损伤机制是必不可少的。所提出的方法将应用于薄水铝石纳米颗粒增强环氧树脂，其作为纳米填料已显示出有希望的结果。我们将展示粗粒度力场在预测代表性体积元素中材料性能方面的强度，作为最典型的合成变量的函数，如聚合物基体交联程度、填料含量、填料的大小和分布。该项目还将总结出一套通用的模型开发规则，为粗粒度模拟奠定基础，使其成为未来纳米复合材料设计的主要计算工具。

The growing technological need for novel high-performance materials imposes major challenges on today’s materials scientist. Synthetic nanocomposites are increasingly used in conditions where extreme mechanical loads have to be withstood. In the design of such materials, computer simulation techniques are becoming inevitable thanks to recent developments of modeling techniques and computational capacities. However, when it comes to predicting nanocomposite properties, current techniques fall short either of considering large enough systems (i.e. atomistic simulations) or accounting for the right underlying physics (i.e. continuum mechanics). In this ambitious project, which lies at the nexus of physics, chemistry and engineering, we propose a solution to these issues. Here, we put forward a novel multiscale simulation protocol based on adapting and improving some existing techniques to develop coarse-grained models for nanoparticle reinforced polymer nanocomposites. The characterizing feature of this project is to derive reactive coarse-grained force fields using a mapping procedure from all-atom molecular dynamics simulations. It allows simulations of large atomistic systems with an accurate description of the breaking of chemical bonds, which is indispensable to model damage mechanisms in the composite materials. The proposed methodology will be applied to an epoxy reinforced with boehmite nanoparticles, which have shown promising results as nanofillers. We will demonstrate the strength of the coarse-grained force fields in predicting the material properties in representative volume elements as a function of the most typical synthesis variables, such as the degree of polymer matrix crosslinking, filler content, and size and distribution of fillers. This project will also conclude with a general set of model development rules, laying the groundwork for coarse-grained simulations to become a major computational tool in the design of future nanocomposites.