Mining Materials Design Rules from Data: The Example of Polymer Dielectrics

Abstract

Mining of currently available and evolving materials databases to discover structure–chemistry–property relationships is critical to developing an accelerated materials design framework. The design of new and advanced polymeric dielectrics for capacitive energy storage has been hampered by the lack of sufficient data encompassing wide enough chemical spaces. Here, data mining and analysis techniques are applied on a recently presented computational data set of around 1100 organic polymers, organometallic polymers, and related molecular crystals, in order to obtain qualitative understanding of the origins of dielectric and electronic properties. By probing the relationships between crucial chemical and structural features of materials and their dielectric constant and band gap, design rules are devised for optimizing either property. Learning from this data set provides guidance to experiments and to future computations, as well as a way of expanding the pool of promising polymer candidates for dielectric applications.

Sn‐Polyester/Polyimide Hybrid Flexible Free‐Standing Film as a Tunable Dielectric Material

“…Recently, a rational co-design process of discovering new dielectric materials based on iterative ab initio computation suggests that the electronic polarization for non-polar materials like BOPP correlates inversely with the band gap and thus cannot be enhanced without affecting the latter. [3,4] On the contrary, such correlations are not observed for ionic polarizations found in organometallic polymers. [5] Organometallic polymers containing metal atoms in the backbone such as tin complexation systems were studied by our team with the aid of the rational co-design process.…”

“…However, low κ (≈2.1) and low operation temperature (85 °C) of BOPP limits its practical use in many high energy density and high temperature applications. Recently, a rational co‐design process of discovering new dielectric materials based on iterative ab initio computation suggests that the electronic polarization for non‐polar materials like BOPP correlates inversely with the band gap and thus cannot be enhanced without affecting the latter . On the contrary, such correlations are not observed for ionic polarizations found in organometallic polymers …”

Machine learning in materials informatics: recent applications and prospects

“…These polymers were then fingerprinted by keeping track of the occurrence of a fixed set of molecular fragments in the polymers in terms of their number fractions [63,71]. A particular molecular fragment could be a triplet of contiguous blocks such as -NH-CO-CH 2 -(or, at a finer level, a triplet of contiguous atoms, such as C 4 -O 2 -C 3 or C 3 -N 3 -H 1 , where X n represents an n-fold coordinated X atom) [72,73]. All possible triplets were considered (some examples are shown in Figure 4a), and the corresponding number fractions in a specific order formed the fingerprint of a particular polymer (see Figure 4b).…”

Soft Matter Informatics: Current Progress and Challenges

“…recently used chemical blocks of CH 2 , CO, NH, C 6 H 4 , C 4 H 2 S, CS, O to form 284 symmetrically unique 4‐block polymers . These studies utilized data‐mining of ab initio data, kernel ridge regression (KRR), and GA to design simple polymer dielectrics with higher energy densities than the current state‐of‐the‐art biaxially oriented polypropylene (BOPP) . Based on the data from the predictive model, a polyurea polymer, polyimide polymer, and polythiourea polymer were synthesized, characterized, and proven to have higher energy densities than BOPP, providing potential alternatives for polymer dielectrics beyond BOPP.…”

“…Based on the data from the predictive model, a polyurea polymer, polyimide polymer, and polythiourea polymer were synthesized, characterized, and proven to have higher energy densities than BOPP, providing potential alternatives for polymer dielectrics beyond BOPP. The successful predictions of organic polymers led the authors to expand their exploration of chemical space and include organometallic polymers, which indicated that the ionic contribution of polymers with metals such as Sn could provide significant improvements in polymer dielectric properties . However, combining organic and organometallic polymers into the same data set led to a decrease in prediction accuracy, due to the wider range of chemical space .…”