Development of coarse-grained simulation methods for the study of soft matter systems

粗粒度分子模型已广泛用于软物质系统的模拟。分子系统的粗粒度建模背后的简单思想是开发一个比传统原子模型分辨率更低的模型，其目的仍然是捕获系统的关键化学和物理。粗粒度模型将允许模拟远远超出原子模型（通常为1000x-10000x）的时间和长度尺度。这使得传统的原子模拟难以研究的系统得以研究。例子包括：溶液中的自组装，肽折叠和聚集，脂质膜内的自组装，胶束的形成，表面活性剂相图的预测，复杂热致液晶相的阐明，不同相之间分子的分配，以及复杂聚合物体系的研究（仅举几例！）影响大多数粗粒度分子建模的关键问题有两个：可表征性和可转移性问题。前者关注的是粗粒度模型在被参数化的热力学状态点上表示物理性质的能力，后者关注的是同一模型在不同状态点（即在无法获得参数化数据的情况下）的预测能力。这个项目关注的是更好的粗粒度模型的开发，这将提高可移植性和可表示性。该项目旨在对目前粗粒度分子模型领域的可能性做出重大改变。它寻求开发比现有模型更可靠和更可转移的新模型，并且可以自信地用于解决软物质化学中的一系列复杂问题。

Coarse-grained molecular models have become widely used for the simulation of soft matter systems. The simple idea behind coarse-grained modelling for a molecular system is to develop a model at a lower resolution than a conventional atomistic model, with the aim of still capturing the key chemistry and physics of the system. The coarse-grained model will allow for simulations of time and length scales that are far beyond what is possible with an atomistic model (typically 1000x-10000x). This allows systems to be studied that would be intractable for a conventional atomistic simulation. Examples include: self-assembly in solution, peptide folding and aggregation, self-assembly within lipid membranes, the formation of micelles, prediction of surfactant phase diagrams, elucidation of complex thermotropic liquid crystal phases, partitioning of molecules between different phases, and the study of complex polymer systems (to name but a few!). There are two key problems that affect most coarse-grained molecular modelling: the representability and transferability problems. The former is concerned with the ability of a coarse-grained model to represent physical properties at the thermodynamic state point at which it is parametrised, the latter is concerned with the ability of the same model to be predictive at different state points, i.e. under conditions where parametrisation data was not available. This project is concerned with the development of better coarse-grained models, which will improve both transferability and representability. The project aims to make a sea-change in terms of what is currently possible in the realm of coarse-grained molecular models. It seeks to develop new models that are more reliable and more transferable than current models and can be used with confidence to tackle a range of complex problems in soft matter chemistry.