Search for: All records

Abstract Graph neural networks (GNNs) have excelled in predictive modeling for both crystals and molecules, owing to the expressiveness of graph representations. High-entropy alloys (HEAs), however, lack chemical long-range order, limiting the applicability of current graph representations. To overcome this challenge, we propose a representation of HEAs as a collection of local environment graphs. Based on this representation, we introduce the LESets machine learning model, an accurate, interpretable GNN for HEA property prediction. We demonstrate the accuracy of LESets in modeling the mechanical properties of quaternary HEAs. Through analyses and interpretation, we further extract insights into the modeling and design of HEAs. In a broader sense, LESets extends the potential applicability of GNNs to disordered materials with combinatorial complexity formed by diverse constituents and their flexible configurations. 

HfO2-based ferroelectrics show tremendous potential for applications in computing technologies, but questions remain as to what dictates the stabilization of the desired phase. Here, it is demonstrated that the substrate the film is grown on is more influential than factors such as thickness, defect content, and strain. The presence of different possible polymorphs of Hf0.5Zr0.5O2 are observed to vary widely for different substrate materials—with La0.67Sr0.33MnO3, (LaAlO3)0.3(Sr2AlTaO6)0.7, and Al2O3 being (more) optimal for stabilizing the ferroelectric-orthorhombic phase. This substrate effect is found to be more influential than any changes observed from varying the film thickness (7.5–60 nm), deposition environment (oxygen vs argon), and annealing temperature (400–600 °C) in vacuum (10−5 Torr). X-ray diffraction and scanning transmission electron microscopy verify the phases present, and capacitor-based studies reveal ferroelectric behavior (or lack thereof) consistent with the phases observed. First-principles calculations suggest that forming oxygen vacancies in Hf0.5Zr0.5O2 lowers its work function, driving electrons away and helping to stabilize the ferroelectric phase. Substrates with a high work function (e.g., La0.67Sr0.33MnO3) facilitate this electron transfer but must also have sufficient ion conductivity to support oxygen-vacancy formation in Hf0.5Zr0.5O2. Together, these observations help clarify key factors essential to the stabilization of HfO2-based ferroelectrics. 

The global instability index (GII) is a computationally inexpensive bond valence-based metric originally designed to evaluate the total bond strain in a crystal. Recently, the GII has gained popularity as a feature of data-driven models in materials research. Although prior studies have proven that GII is an effective predictor of structural distortions and decomposition energy when applied to small datasets, the wider use of GII as a global indicator of structural stability has yet to be evaluated. To that end, we compute GII for thousands of compounds in inorganic structure databases and partition compounds by chemical interactions underlying their stability to understand the GII values and their variations. Our results show that the GII captures relative chemical trends, such as electronegativity, even beyond the intended domain of strongly ionic compounds. However, we also find that GII magnitudes vary significantly with factors such as chemistry (cation–anion identities and bond character), geometry (connectivity), data source, and model bias, making GII suitable for comparisons within controlled datasets but unsuitable as an absolute, universal metric for structural feasibility. 

Electronic materials that exhibit phase transitions between metastable states (e.g., metal-insulator transition materials with abrupt electrical resistivity transformations) are challenging to decode. For these materials, conventional machine learning methods display limited predictive capability due to data scarcity and the absence of features that impede model training. In this article, we demonstrate a discovery strategy based on multi-objective Bayesian optimization to directly circumvent these bottlenecks by utilizing latent variable Gaussian processes combined with high-fidelity electronic structure calculations for validation in the chalcogenide lacunar spinel family. We directly and simultaneously learn phase stability and bandgap tunability from chemical composition alone to efficiently discover all superior compositions on the design Pareto front. Previously unidentified electronic transitions also emerge from our featureless adaptive optimization engine. Our methodology readily generalizes to optimization of multiple properties, enabling co-design of complex multifunctional materials, especially where prior data is sparse. 

Abstract The Rashba effect enables control over the spin degree of freedom, particularly in polar materials where the polar symmetry couples to Rashba‐type spin splitting. The exploration of this effect, however, has been hindered by the scarcity of polar materials exhibiting the bulk‐Rashba effect and rapid spin‐relaxation effects dictated by the D'yakonov–Perel mechanism. Here, a polar LiNbO₃‐typeR3cphase of Bi_1‐xIn_1+xO₃withx≈0.15–0.24 is stabilized via epitaxial growth, which exhibits a bulk‐Rashba effect with suppressed spin relaxation as a result of its unidirectional spin texture. As compared to the previously observed non‐polarPnmaphase, this polar phase exhibits higher conductivity, reduced bandgap, and enhanced dielectric and piezoelectric responses. Combining first‐principles calculations and multimodal magnetotransport measurements, which reveal weak (anti)localization, anisotropic magnetoresistance, planar‐Hall effect, and nonreciprocal charge transport, a bulk‐Rashba effect without rapid spin relaxation is demonstrated. These findings offer insights into spin‐orbit coupling physics within polar oxides and suggest potential spintronic applications. 

Abstract 2D polymers (2DPs) are promising as structurally well‐defined, permanently porous, organic semiconductors. However, 2DPs are nearly always isolated as closed shell organic species with limited charge carriers, which leads to low bulk conductivities. Here, the bulk conductivity of two naphthalene diimide (NDI)‐containing 2DP semiconductors is enhanced by controllably n‐doping the NDI units using cobaltocene (CoCp₂). Optical and transient microwave spectroscopy reveal that both as‐prepared NDI‐containing 2DPs are semiconducting with sub‐2 eV optical bandgaps and photoexcited charge‐carrier lifetimes of tens of nanoseconds. Following reduction with CoCp₂, both 2DPs largely retain their periodic structures and exhibit optical and electron‐spin resonance spectroscopic features consistent with the presence of NDI‐radical anions. While the native NDI‐based 2DPs are electronically insulating, maximum bulk conductivities of >10⁻⁴ S cm⁻¹are achieved by substoichiometric levels of n‐doping. Density functional theory calculations show that the strongest electronic couplings in these 2DPs exist in the out‐of‐plane (π‐stacking) crystallographic directions, which indicates that cross‐plane electronic transport through NDI stacks is primarily responsible for the observed electronic conductivity. Taken together, the controlled molecular doping is a useful approach to access structurally well‐defined, paramagnetic, 2DP n‐type semiconductors with measurable bulk electronic conductivities of interest for electronic or spintronic devices.