Polymer Genome: A Polymer Informatics Platform to Accelerate Polymer Discovery
Polymer Genome: A Polymer Informatics Platform to Accelerate Polymer Discovery
复制标题
DOI：
10.1007/978-3-030-40245-7_18
复制
发表时间：
2020-01-01
期刊：
影响因子：
--
通讯作者：
Ramprasad, Rampi
中科院分区：
文献类型：
--
作者：
Chandrasekaran, Anand;Kim, Chiho;Ramprasad, Rampi
The Materials Genome Initiative has brought about a paradigm shift in the design and discovery of novel materials. In a growing number of applications, the materials innovation cycle has been greatly accelerated as a result of insights provided by data-driven materials informatics platforms. High-throughput computational methodologies, data descriptors, and machine learning are playing an increasingly invaluable role in research development portfolios across both academia and industry. Polymers, especially, have long suffered from a lack of data on electronic, mechanical, and dielectric properties across large chemical spaces, causing a stagnation in the set of suitable candidates for various applications. The nascent field of polymer informatics seeks to provide tools and pathways for accelerated polymer property prediction (and materials design) via surrogate machine learning models built on reliable past data. With this goal in mind, we have carefully accumulated a dataset of organic polymers whose properties were obtained either computationally (bandgap, dielectric constant, refractive index, and atomization energy) or experimentally (glass transition temperature, solubility parameter, and density). A fingerprinting scheme that captures atomistic to morphological structural features was developed to numerically represent the polymers. Machine learning models were then trained by mapping the polymer fingerprints (or features) to their respective properties. Once developed, these models can rapidly predict properties of new polymers (within the same chemical class as the parent dataset) and can also provide uncertainties underlying the predictions. Since different properties depend on different length-scale features, the prediction models were built on an optimized set of features for each individual property. Furthermore, these models are incorporated in a user friendly online platform named Polymer Genome (www.polymergenome.org). Systematic and progressive expansion of both chemical and property spaces are planned to extend the applicability of Polymer Genome to a wide range of technological domains.
