Element selection for functional materials discovery by integrated machine learning of atomic contributions to properties

Abstract

At the high level, the fundamental differences between materials originate from the unique nature of the constituent chemical elements. Before specific differences emerge according to the precise ratios of elements (composition) in a given crystal structure (phase), the material can be represented by its phase field defined simply as the set of the constituent chemical elements. Classification of the materials at the level of their phase fields can accelerate materials discovery by selecting the elemental combinations that are likely to produce desirable functional properties in synthetically accessible materials. Here, we demonstrate that classification of the materials’ phase field with respect to the maximum expected value of a target functional property can be combined with the ranking of the materials’ synthetic accessibility. This end-to-end machine learning approach (PhaseSelect) first derives the atomic characteristics from the compositional environments in all computationally and experimentally explored materials, and then employs these characteristics to classify the phase field by their merit. PhaseSelect can quantify the materials’ potential at the level of the periodic table, which we demonstrate with significant accuracy for three avenues of materials’ applications: high-temperature superconducting, high-temperature magnetic and targetted energy band gap materials.

The most compact search space is not always the most efficient: A case study on maximizing solid rocket fuel packing fraction via constrained Bayesian optimization

“…unitless simulations) [37,38] (referred to as "size"), a set of parameters that is represented as a composition or formulation (i.e. Al 2 O 3 ≡ 0.4Al + 0.6O where 0.4 + 0.6 = 1.0) [39][40][41][42][43][44][45][46][47][48][49][50] (referred to as "comp"), or a set of parameters that exhibits permutation invariance (e.g. [36,40,51] (referred to as "order").…”

Effect of reducible and irreducible search space representations on adaptive design efficiency: a case study on maximizing packing fraction for solid rocket fuel propellant simulations

“…unitless simulations) [25,26] (referred to as "size"), a set of parameters that is represented as a composition or formulation (i.e. Al 2 O 3 ≡ 0.4Al + 0.6O where 0.4 + 0.6 = 1.0) [27][28][29][30][31][32][33][34][35][36][37][38] (referred to as "comp"), or a set of parameters that exhibits permutation invariance (e.g. [24,28] (referred to as "order").…”

Element selection for functional materials discovery by integrated machine learning of elemental contributions to properties

“…The large number of such phase fields among the top-performing candidates with respect to the measure of synthetic accessibility provides verification of the developed models and demonstrates that highly ranked candidates are likely to produce thermodynamically stable materials observed experimentally (See Figure 4a-c). We report the full list of likely candidates for novel superconducting materials among the phase fields that have been reported to form stable compounds in ICSD, but were not investigated from the perspectives of superconducting applications (See the full list of candidates in 22 and its excerpt in Supplementary Table 7).…”

“…While these predictions may align well with the human experts' understanding of chemistry, hence emphasizing the models' capability to infer complex elemental characteristics and materials-properties relationship from historical data, the models can also be used to identify unconventional and rare prospective elemental combinations as well as to rank the attractive candidate materials for experimental investigations. Such less expected examples include combinations of elements that do not exhibit ambient pressure or pressure-induced superconductivity as elemental solids 23 , exclude Fe, Cu and rare-earth metals, known for forming families of superconducting materials, but are classified as high-temperature (>10K) superconductors, when combined: C-Mg-Rb, Cr-K-N and As-C-Na among other 125 ternaries 22 .…”

“…For magnetic applications, all unexplored phase fields that were classified to exhibit magnetism at Tc > 300K contain known magnetic elements Fe, Co, Ni, or Mn. However, we can highlight the combinations that are interesting also from the perspective of synthetic accessibility, including 4001 Fe-free ternaries 22 , such as Co-Ti-Zr, Mo-Co-B and Hf-Mn-Ti. Among 13070 large bandgap dielectric phases 22 not involving oxides nor fluorides we can highlight Te-S-I and Ga-S-Cl.…”