Using Feature‐Assisted Machine Learning Algorithms to Boost Polarity in Lead‐Free Multicomponent Niobate Alloys for High‐Performance Ferroelectrics

Abstract

To expand the unchartered materials space of lead‐free ferroelectrics for smart digital technologies, tuning their compositional complexity via multicomponent alloying allows access to enhanced polar properties. The role of isovalent A‐site in binary potassium niobate alloys, (K,A)NbO3 using first‐principles calculations is investigated. Specifically, various alloy compositions of (K,A)NbO3 are considered and their mixing thermodynamics and associated polar properties are examined. To establish structure‐property design rules for high‐performance ferroelectrics, the sure independence screening sparsifying operator (SISSO) method is employed to extract key features to explain the A‐site driven polarization in (K,A)NbO3. Using a new metric of agreement via feature‐assisted regression and classification, the SISSO model is further extended to predict A‐site driven polarization in multicomponent systems as a function of alloy composition, reducing the prediction errors to less than 1%. With the machine learning model outlined in this work, a polarity‐composition map is established to aid the development of new multicomponent lead‐free polar oxides which can offer up to 25% boosting in A‐site driven polarization and achieving more than 150% of the total polarization in pristine KNbO3. This study offers a design‐based rational route to develop lead‐free multicomponent ferroelectric oxides for niche information technologies.

Contact engineering for 2D Janus MoSSe/metal junctions

“…Finally, the features generated by SISSO are ranked according to the root mean square error (RMSE) to identify the optimal set of descriptors. 55 This approach enables us to gain a deeper understanding of the complex factors that contribute to the SBH, which is critical for developing high-performance MSJs.…”

Explicating the irreversible electric-field-assisted ferroelectric phase transition in the otherwise antiferroelectric sodium niobate for energy storage systems

“…, BaTiO 3 and KNbO 3 where the intra-octahedral distortions occur dominantly due to the softening of the dominant ferroelectric polar Γ − 4 mode. 61–65 This implies that the complex phase transition process in NaNbO 3 is expected to arise from an “orchestra” of different unstable phonon modes – each working together to drive the symmetry-lowering phase transitions in a non-trivial way. Details of the DFT-calculated relative energies for each phase with respect to cubic Pm 3̄ m NaNbO 3 are tabulated in Table S1 of the ESI †…”

“…In other words, there appear to be multiple dynamical instabilities in the cubic phase of NaNbO 3 when compared to the archetypical ferroelectric perovskites e.g., BaTiO 3 and KNbO 3 where the intra-octahedral distortions occur dominantly due to the softening of the dominant ferroelectric polar G À 4 mode. [61][62][63][64][65] This implies that the complex phase transition process in NaNbO 3 is expected to arise from an ''orchestra'' of different unstable phonon modes -each working together to drive the symmetry-lowering phase transitions in a non-trivial way. Details of the DFT-calculated relative energies for each phase Fig.…”

Machine Learning-Accelerated Discovery of A2BC2 Ternary Electrides with Diverse Anionic Electron Densities

“…In recent years, machine learning (ML) has been successfully applied in many fields of materials research, such as the design of crystal structures, − the development of interatomic potentials, , and the prediction of material properties. − In particular, ML-based searches for new materials that meet specific requirements in a huge chemical space have proved effective for different material systems. ,− Compared to HT ab initio computations, ML subverts the need for exhaustive quantum mechanical calculations, rendering combinatorial chemical spaces tractable. In turn, ML can provide insights into the complex relationships between composition and properties. − However, this approach requires a sufficiently large training data set (consisting of both electrides and non-electride compounds) to obtain a reasonable surrogate model.…”