Big Data Approach on Polymer Materials: Fundamental, Progress and Challenge

Abstract

Big data approach, a new paradigm for data-driven wisdom paces together with conventional experimental, theoretical and simulation ways. The core in the application of big data study in material researches is at the composition-process-structure-property-performance relationship (CPSPPr). The digitalization and computational efforts, concepts, tools and resources to enable the subterms in the CPSPPr to cover fundamental research and scalable production will be presented. The big data approach can fully utilize the merits for the “black-box” of emerging machine learning algorithms, which allows for the construction of much more quantitative correlations beyond conventional rationality. It provides fantastic wisdom in reward for the discovery and the manufacture of new materials from the inherently frustrated multiple scales, broadly distributed, weak but accumulatively-strong response of polymers. We hereby reviewed such representative progresses in the innovation of polymer materials wholly or partially using big data approach in the last four years. These progresses are grouped as: polymer synthesis and self-assembly, mechanical and thermal properties, optic-electric-magnetic-acoustic properties and membranes for separation. The overall progress for the application of big data approach in the research of polymer materials is lagging in the comparison with that in inorganic or small-molecular materials. We then enumerated a number of challenges and possible short-term breakthroughs before the dawn of burst for big data in the reshape of the research and the production of polymer materials.