arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2404.02676cond-mat.soft
收藏

评估聚合物/增塑剂混溶性的高通量筛选方法框架

A Framework for a High Throughput Screening Method to Assess Polymer/Plasticizer Miscibility

Lois Smith, H. Ali Karimi-Varzaneh, Sebastian Finger, Giuliana Giunta, Alessandro Troisi, Paola Carbone

AI总结:

本文提出基于分子模拟与机器学习的聚合物增塑剂混溶性高通量筛选框架,利用拓扑和热力学描述符及决策树实现高效分类。

摘要:

聚合物复合材料需要软化,以降低其玻璃化转变温度并改善加工性能。为此,人们将增塑剂(一类有机小分子)加入聚合物基体中。这些增塑剂的混溶性对其有效性有很大影响,因此必须仔细考虑它们与聚合物基体的相互作用。许多增塑剂特征(包括其尺寸、拓扑结构和柔性)都会影响其混溶性,并且由于增塑剂的数量呈指数级增长,当前的试错方法效率极低。在本工作中,我们表明,利用包含48种增塑剂的小型数据集进行分子模拟,可以识别出作为其混溶性代理指标的拓扑和热力学描述符。借助计算成本相对较低的特设分子动力学模拟设置,我们建立了增塑剂的拓扑结构、内部柔性、聚集热力学与其混溶程度之间的关联,并使用这些描述符将分子分类为可混溶或不可混溶。利用所有可用数据,我们还构建了一个决策树模型,该模型通过重复的分层5折交叉验证取得了0.86 ± 0.01的F1分数,表明这种机器学习方法是实现筛选完全自动化的一条有前景的途径。通过评估各描述符的单独性能,我们证明该流程能够将测试空间缩小10倍,并为开发可高效筛选数千种具有各种特征的增塑剂的工作流奠定基础。

英文摘要:

Polymer composite materials require softening to reduce their glass transition temperature and improve processability. To this end, plasticizers, which are small organic molecules, are added to the polymer matrix. The miscibility of these plasticizers has a large impact on their effectiveness and therefore their interactions with the polymer matrix must be carefully considered. Many plasticizer characteristics, including their size, topology and flexibility, can impact their miscibility and, because of the exponentially large numbers of plasticizers, the current trial-and-error approach is very ineffective. In this work we show that using molecular simulations of a small dataset of 48 plasticizers, it is possible to identify topological and thermodynamic descriptors that are proxy for their miscibility. Using ad-hoc molecular dynamics simulation set-ups that are relatively computationally inexpensive, we establish correlations between the plasticizers' topology, internal flexibility, thermodynamics of aggregation and their degree of miscibility and use these descriptors to classify the molecules as miscible or immiscible. With all available data we also construct a decision tree model which achieves a F1 score of 0.86 +/- 0.01 with repeated, stratified 5-fold cross-validation, indicating that this machine learning method is a promising route to fully automate the screening. By evaluating the individual performance of the descriptors, we show this procedure enables a 10-fold reduction of the test space and provides the basis for the development of workflows which can efficiently screen thousands of plasticizers with a variety of features.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑