Roberto - Improved dynamics in self-consistent-field molecular-dynamics simulations of polymers

主要目标是开发一种新的分子模拟算法，该算法允许研究聚合物，特别是纳米复合材料中的现象，这些现象在科学上有趣，技术上相关，但迄今为止无法模拟。这是常见的情况，因为可用的计算能力不足以同时使用真实的模型，模拟足够大的样本，并模拟足够长的时间来观察所有相关的物理过程。我们提出的方法结合了两种最近发展起来的方法。第一个成分是米兰（萨莱诺）和川胜（仙台）所谓的自洽场MD方法（MD- scf）。它通过辅助场表示加速了计算密集型非键相互作用的计算。它可以用于原子和粗粒聚合物模型，并已被证明以低计算成本非常准确地再现聚合物材料的分子结构和热力学性质。然而，通过设计，它无法正确捕获聚合物的分子动力学和由此产生的所有特性，如扩散、熔体粘度和流变参数。这种失败是由于MD-SCF模型中不同链珠之间缺乏排除体积相互作用造成的。链交叉无法避免，链纠缠和重复缺失，导致定性错误的聚合物动力学结果。在这一点上，第二种成分进入。与Masubuchi（名古屋）合作，我们最近开发了聚合物链之间临时移动交联的概念，作为缠结的模型。我们已经成功地将这些所谓的滑动链接实现到聚合物的耗散粒子动力学（DPD）模型中。DPD也缺乏珠子之间的硬核排斥，并且与MD-SCF具有完全相同的聚合物动力学问题。然而，在引入滑链之后，DPD再现了聚合物熔体动力学的所有基本特征。因此，我们也将在MD-SCF方法中实施滑动连接，这是对DPD的改进，因为它允许更多的化学特异性。项目阶段为：(1)MD-SCF/滑环组合的开发和实施及其技术验证。(2)解决一些科学风险，这些风险可能导致该方法在某些情况下失败。例如，在哈密顿函数中引入滑簧可以将计算出的性质改变到一定程度，以至于必须以某种方式进行补偿。(3)找到了适用于聚合物动力学性质计算的新方法的操作参数范围。(4)最后，我们希望将新方法首次应用于最近涉及聚合物复合材料流变学和其他动力学特性的问题，这些问题可以用其他方法来解决，但很难解决，或者根本无法解决。

The main objective is to develop a new molecular-simulation algorithm, which allows the investigation of phenomena in polymer and especially nanocomposites that are scientifically interesting and technologically relevant, but which hitherto cannot be simulated. This is commonly the case, because available compute power is not sufficient to simultaneously use realistic models, model large enough samples, and simulate them long enough to observe all relevant physical processes.Our proposed approach combines two recently developed methods. The first ingredient is the so-called self-consistent-field MD method (MD-SCF) of Milano (Salerno) and Kawakatsu (Sendai). It accelerates the calculation of the compute-intensive nonbonded interactions by an auxiliary field representation. It can be used with both atomistic and coarse-grained polymer models and has been shown to reproduce molecular structure and thermodynamic properties of polymer materials very accurately at low computational cost. By design, however, it is unable to capture correctly the molecular dynamics of polymers and all properties resulting from it, such as diffusion, melt viscosity and rheological parameters. This failure results from the lack of excluded-volume interactions between beads of different chains in the MD-SCF model. Chain crossing is not prevented, chain entanglement and reptation are absent, and a qualitatively wrong polymer dynamics results.At this point, the second ingredient enters. Working with Masubuchi (Nagoya), we have recently developed the concept of temporary mobile cross-links between polymer chains as a mock-up for entanglements. We have implemented these so-called slip-links successfully into the dissipative-particle-dynamics (DPD) model for polymers. DPD also lacks a hard-core repulsion between beads and has exactly the same problems with the polymer dynamics as has MD-SCF. After the intro-duction of slip-links, however, DPD reproduced all essential features of polymer melt dynamics. We will therefore implement slip-links into the MD-SCF method as well, which is an improvement over DPD, because it allows for more chemical specificity.The project phases are: (1) Development and implementation of the MD-SCF/slip-link combination and its technical validation. (2) Addressing a number of scientific risks, which might cause the method to fail in certain cases. For example, introducing slip-springs into the Hamiltonian could shift the calculated properties to such an extent that they would have to be compensated in some way. (3) Finding a range of the operational parameters of the new method which is suitable for the calculation of dynamical properties of polymers. (4) Finally, we would like to do a first application of the new method to very recent problems involving rheological and other dynamical properties of polymer composites, which can be addressed by other methods only with difficulties or not at all.