npj Computational Materials最新文献

Global optimization of crystal compositions is a significant yet computationally intensive method to identify stable structures within chemical space. The specific physical properties linked to a three-dimensional atomic arrangement make this an essential task in the development of new materials. We present a method that efficiently uses active learning of neural network force fields for structure relaxation, minimizing the required number of steps in the process. This is achieved by neural network force fields equipped with uncertainty estimation, which iteratively guide a pool of randomly generated candidates toward their respective local minima. Using this approach, we are able to effectively identify the most promising candidates for further evaluation using density functional theory (DFT). Our method not only reliably reduces computational costs by up to two orders of magnitude across the benchmark systems Si₁₆, Na₈Cl₈, Ga₈As₈ and Al₄O₆ but also excels in finding the most stable minimum for the unseen, more complex systems Si₄₆ and Al₁₆O₂₄. Moreover, we demonstrate at the example of Si₁₆ that our method can find multiple relevant local minima while only adding minor computational effort.

Ductile inorganic semiconductors have recently received considerable attention due to their metal-like mechanical properties and potential applications in flexible electronics. However, the accurate determination of slip pathways, crucial for understanding the deformation mechanism, still poses a great challenge owing to the complex crystal structures of these materials. In this study, we propose an automated workflow based on the interlayer slip potential energy surface to identify slip pathways in complex inorganic systems. Our computational approach consists of two key stages: first, an active learning strategy is utilized to efficiently and accurately model the interlayer slip potential energy surfaces; second, the climbing image nudged elastic band method is employed to identify minimum energy pathways, followed by comparative analysis to determine the final slip pathway. We discuss the validity of our selected feature vectors and models across various material systems and confirm that our approach demonstrates robust effectiveness in several case studies with both simple and complicated slip pathways. Our automated workflow opens a new avenue for the automatic identification of the slip pathways in inorganic materials, which holds promise for accelerating the high-throughput screening of ductile inorganic materials.

Tailoring material properties often requires understanding the solidification process. Herein, we introduce the geometric descriptor Soliquidy, which numerically captures the Euclidean transport cost between the translationally disordered versus ordered states of a materials. As a testbed, we apply Soliquidy to the classification of glass-forming metal alloys. By extending and combining an experimental library of metallic thin films (glass/no-glass) with the aflow.org computational database (geometrical and energetic information of mixtures) we found that the combination of Soliquity and formation enthalpies generates an effective classifier for glass formation. Such a classifier is then used to tackle a public dataset of metallic glasses showing that the glass-agnostic assumptions of Soliquity can be useful for understanding kinetically-controlled phase transitions.

Charge transport in organic amorphous systems has been considered to occur by intermolecular hopping. However, it has been difficult to reveal even the intra- and intermolecular structures because of their amorphous nature. Therefore, the details of charge transport at the molecular level have not been clarified. Here, we investigate a detailed molecular-level insight into the charge transport in an amorphous film by the analysis of multiscale simulation. The charge mobility is normally described by a constant value but is found to be widely distributed with two orders of magnitude even in the 100 nm neat film. From the detailed analysis at the molecular level, it becomes clear that there are three types of charge traps; in addition to (1) the well-known traps due to the site energy difference, we found (2) traps caused by the distribution of molecular packings in the aggregate, and (3) those by charge hopping against the electric field. These traps are the origins of the widely distributed mobilities and the understanding of these traps is important to improve mobility.

Wurtzite-type ferroelectrics are highly promising for next-generation microelectronic devices due to their ferroelectric properties and integration with exiting semiconductors. However, their high coercive fields, which are close to breakdown electric fields, need to be lowered. To deal with this issue and secure device reliability, much effort has been devoted to exploring novel wurtzite compounds with lower polarization switching barriers and implementing doping strategies. Here, we report first-principles calculations on polarization switching in cation-vacancy ordered wurtzite α-Al₂S₃, unveiling its uniaxial quadruple-well ferroelectricity and moderate switching barrier, 51 meV/cation, which is much lower than that of conventional wurtzite ferroelectrics. There are three important features relevant to the Al vacancies leading to the uncommon quadruple-well ferroelectricity and the moderate switching barrier: mitigation of cation-cation repulsion, structural flexibility that alleviates an in-plane lattice expansion, and formation of σ-like bonding states consisting of Al 3p_z and S 3p_z orbitals. Biaxial compressive strain and Ga doping lower the switching barriers by up to 40%. This study encourages experimental investigation of the ferroelectric properties for defective wurtzite α-Al₂S₃ as a new promising material with unconventional and intriguing ferroelectricity and suggests a potential strategy for reducing switching barriers in wurtzite ferroelectrics: introducing cation vacancies.

In the realm of lithium superionic conductors, pursuing higher ionic conductivity is imperative, with the variance in lithium-ion concentration playing a determining role. Due to the permanent and temporary site-blocking effects, especially at non-dilute concentrations, not all Li-ions contribute to ionic conductivity. Here, we propose a strategy to directly calculate effective mobile ion concentration in which multiple-ion correlated migration is considered in the percolation analysis with the input of Li-ion distributions and hopping behavior based on kinetic Monte Carlo simulation, termed P-KMC. We provide examples of two representative lithium superionic conductors, cubic garnet-type Li_xA₃B₂O₁₂ (0 ≤ x ≤ 9; A and B represent different cations) and perovskite-type Li_xLa_2/3−x/3TiO₃ (0 ≤ x ≤ 0.5), to demonstrate the direct dependence of the ionic conductivity on the effective mobile ion concentration. This methodology provides a robust tool to identify the optimal compositions for the highest ionic conductivity in superionic conductors.

Doping is critical for inducing ferroelectricity in hafnia films, yet the underlying mechanisms remain debated. Here, through first-principles studies, we elucidate the pivotal role played by the complex phase transition mechanisms under carrier doping in understanding the origin of hafnia ferroelectricity. Specifically, electron doping orchestrates a metastable polar phase to stable antipolar phase transformation, driven by strong screening effects and weakened nonpolar covalent bonds, making n-type dopants rare. Conversely, weak screening effect and enhanced polar covalent bonding strengthen robust ferroelectricity, enabling significant ground-state phase transitions from the monoclinic to the polar orthorhombic phase and finally to the cubic phase under hole doping, a phenomenon prevalent in hafnia-based films doped with p-type dopants. Furthermore, this hole-enhanced polar distortion also results in an inverse size effect in hafnia ferroelectric films, unlike perovskite ferroelectrics. Our findings offer new insights into the preparation of robust hafnia-based ferroelectric films through doping or interface engineering.

Uncertainty quantification (UQ) to detect samples with large expected errors (outliers) is applied to reactive molecular potential energy surfaces (PESs). Three methods–Ensembles, deep evidential regression (DER), and Gaussian Mixture Models (GMM)—were applied to the H-transfer reaction between syn-Criegee and vinyl hydroxyperoxide. The results indicate that ensemble models provide the best results for detecting outliers, followed by GMM. For example, from a pool of 1000 structures with the largest uncertainty, the detection quality for outliers is ~90% and ~50%, respectively, if 25 or 1000 structures with large errors are sought. On the contrary, the limitations of the statistical assumptions of DER greatly impact its prediction capabilities. Finally, a structure-based indicator was found to be correlated with large average error, which may help to rapidly classify new structures into those that provide an advantage for refining the neural network.

We demonstrate a multi-fidelity (MF) machine learning ensemble framework for the inverse design of photonic surfaces, trained on a dataset of 11,759 samples that we fabricate using high throughput femtosecond laser processing. The MF ensemble combines an initial low fidelity model for generating design solutions, with a high fidelity model that refines these solutions through local optimization. The combined MF ensemble can generate multiple disparate sets of laser-processing parameters that can each produce the same target input spectral emissivity with high accuracy (root mean squared errors < 2%). SHapley Additive exPlanations analysis shows transparent model interpretability of the complex relationship between laser parameters and spectral emissivity. Finally, the MF ensemble is experimentally validated by fabricating and evaluating photonic surface designs that it generates for improved efficiency energy harvesting devices. Our approach provides a powerful tool for advancing the inverse design of photonic surfaces in energy harvesting applications.

This paper describes a novel physics-informed Bayesian optimization approach that leverages prior physics knowledge, specifically Vegard’s law and the linear relationship between gas flow rate and composition in compound semiconductors. The methodology was applied to metal-organic chemical vapor deposition for III–V semiconductor growth. It resulted in the successful synthesis of III–V semiconductors with tailored band gap wavelengths and lattice constants in the region of growth conditions not included in the training data. Furthermore, it predicted hidden trends that Ga composition would be smaller than In composition in As-rich growth regions. This trend is not described by prior physics, demonstrating that statistical machine learning is effective not only for optimization but also for gaining a physical understanding of crystal growth mechanisms. The study demonstrates the potential to develop extrapolable machine learning models by integrating robust physics knowledge, which significantly enhances the efficiency of high-throughput and autonomous material synthesis.