Machine learning provides workout for new polymers

Doctors Ryo Tamura, Kenji Nagata and Takashi Nakanishi, based at the National Institute for Materials Science in Tsukuba, developed their method on homo-polypropylene polymers, applying X-ray diffraction patterns of them under different conditions to provide detailed information.

This approach, outlined in Science and Technology of Advanced Materials, offers a potential alternative to the standard practice of manually testing polymer strength and flexibility, which has the disadvantage of being both costly and destructive. 

“Machine learning can be applied to data from existing materials to predict the properties of unknown materials. However, to achieve accurate predictions, it’s essential to use descriptors that correctly represent the features of these materials,” they stated.

Thermoplastic crystalline polymers, such as polypropylene, have a complex structure that is further altered during the process of molding them into the shape of the end product. It was, therefore, important for the scientists to adequately capture the details of the polymers’ structure with X-ray diffraction and to ensure that the machine learning algorithm could identify the most important descriptors in that data.

The trio analysed two datasets using Bayesian spectral deconvolution, to extract patterns from data: X-ray diffraction data from 15 types of homo-polypropylenes subjected to a range of temperatures; and data from four types of homo-polypropylenes that underwent injection molding. Properties analysed included stiffness, elasticity, the temperature at which the material starts to deform, and flexibility.

The team found that the machine learning analysis accurately linked features in the X-ray diffraction imagery with specific material properties of the polymers. Some of the mechanical properties were easier to predict from the X-ray diffraction data, while others, such as the stretching break point, were more challenging.

“We believe our study, which describes the procedure used to provide a highly accurate machine learning prediction model using only the X-ray diffraction results of polymer materials, will offer a nondestructive alternative to conventional polymer testing methods,” wrote the researchers.

The team also suggested that their approach could be applied to other data, such as X-ray photoelectron spectroscopy, and used to understand the properties of other materials, both inorganic and organic.