DMREF: Collaborative: Computationally Driven Discovery and Engineering of Multiblock Polymer Nanostructures Using Genetic Algorithms

*技术摘要*这个计算驱动的发现计划旨在打破多块材料设计的现状。研究集中在伪谱自洽场理论(SCFT)和实空间遗传算法(GAS)的结合上，并与实验合成和表征紧密耦合。虽然SCFT和GAS分别广泛应用于聚合物科学和生物信息学，但以前还没有集成到一起来解决嵌段聚合物的发现和设计问题。这里采用的方法解决了大参数空间和多态的挑战。它还解决了识别能够产生所需纳米级形态的多块序列和组合物的“逆问题”，而无需求助于在参数空间中进行穷尽的、无指导的搜索。计算工作与包括最先进的合成、处理和表征工具的实验计划协同并迭代地结合在一起。实验工作将验证计算方法，包括模型的参数化，并为有吸引力的和可综合访问的设计目标提供灵感。这种结合的方法将大大缩短新的多嵌段聚合物作为先进功能材料的发现、设计和部署的时间。*非技术摘要*加州大学圣巴巴拉分校和明尼苏达大学的研究人员之间的合作将开发发现工具，使多嵌段聚合物的合理、计算辅助设计在医学、微电子、分离以及能源生产和存储等领域的应用成为可能。这类软材料的复杂因素是决定分子结构、嵌段序列和相互作用的无数参数，以及可能的广泛的自组装纳米结构。通过理论、模拟和实验的协调和迭代结合，将设计和验证全局优化工具来预测聚合物结构和纳米结构之间的正反向关系。在该计划中开发的发现工具将通过明尼苏达超级计算机研究所托管的基于网络的作业提交程序向工业和学术聚合物材料社区广泛提供，并将根据项目过程中产生的结构/序列/形态图构建一个可搜索的数据库。通过利用UCSB(复杂流体设计联盟)和UMN(IPrime)的成熟和非常成功的工业联合体，将实现与行业的接触。该项目的人员将在UMN和UCSB现有的MRSEC提供的丰富的多学科研究环境中接受培训并得到加强。该奖项由材料研究部(DMR)和数学科学部(DMS)资助。

****Technical Abstract****This computationally driven discovery program aims to disrupt the status quo for the design of multiblock materials. The research centers on the marriage of pseudo-spectral self-consistent field theory (SCFT) and real-space genetic algorithms (GAs), with a tight coupling to experimental synthesis and characterization. While widely used in polymer science and bioinformatics, respectively, SCFT and GAs have not been previously integrated to tackle problems of block polymer discovery and design. The approach adopted here addresses the challenges of large parameter spaces and polymorphism. It also solves the "inverse problem" of identifying multiblock sequences and compositions that can produce a desired nanoscale morphology, without resorting to exhaustive, unguided searches through parameter space. The computational effort is synergistically and iteratively combined with an experimental program that includes state-of-the-art synthesis, processing, and characterization tools. The experimental work will validate the computational methodology, including parameterization of the models, and provide inspiration for attractive and synthetically accessible design targets. This combined approach will dramatically reduce the timescale for discovery, design, and deployment of new multiblock polymers as advanced functional materials.****Non-Technical Abstract****This collaborative effort between researchers at the University of California, Santa Barbara and the University of Minnesota will develop discovery tools that will enable the rational, computationally-assisted design of multiblock polymers for applications in medicine, microelectronics, separations, and energy production and storage, among others. Complicating factors in this class of soft materials are the myriad parameters that dictate molecular architecture, block sequence, and interactions and the wide range of self-assembled nanostructures that are possible. Through a concerted and iterative combination of theory, simulation, and experiment, global optimization tools will be devised and validated to predict the forward and reverse relationship between polymer architecture and nanostructure. The discovery tools developed in this program will be made widely available to the industrial and academic polymer materials community through a web-based job submission program hosted at the Minnesota Supercomputer Institute, and a searchable database will be constructed from the structure/sequence/morphology maps that result over the course of the project. Outreach to industry will be accomplished by leveraging the established and highly successful industrial consortiums at UCSB (Complex Fluids Design Consortium) and UMN (IPrime). Personnel on the project will be trained in and enhance the rich multidisciplinary research environments afforded by the existing MRSECs at UMN and UCSB.This award is funded by the Division of Materials Research (DMR) and the Division of Mathematical Sciences (DMS).