arXiv subjects

Dong Won Jeon

Publications and source records attributed to Dong Won Jeon.

3 recordsLinked to original sources

arXiv2026

Topological-Insulator Heterophase Gate Stacks for Transistor Electrostatics

Conventional gate-stack scaling reduces dielectric thickness and increases permittivity while largely treating the position and electronic character of the gate-side screening boundary as fixed. As equivalent oxide thickness is reduced, however, finite interfacial responses can increasingly constrain gate control [1-5]. Here we show that this screening boundary can itself be engineered by converting the surface of the topological insulator Bi2Se3 into insulating high-kappa$ BiF3. Position-resolved calculations reveal a gap-opened immediate amorphous-BiF3/crystalline-Bi2Se3 interface and a reconstructed gap-closed Bi2Se3-derived state in the adjacent subinterface layer, accompanied by a localized interfacial dipole. Independently, capacitor measurements resolve a finite series response consistent with the electronic compressibility of this buried boundary, which reduces rather than enhances the nominal stack capacitance. Despite this capacitance penalty, MoS2 transistors with closely matched BiF3 thicknesses and a common BiF3/MoS2 channel-side material interface exhibit near-thermionic switching, negligible hysteresis, and approximately sevenfold lower drain-induced barrier lowering than BiF3-only controls. These results identify the position and electronic character of the gate-side screening boundary as additional design variables for transistor electrostatics beyond nominal dielectric capacitance.

physics.app-ph↗

arXiv2026

Oxygen stoichiometry directs rutile-anatase phase selection through kinetic control of nucleation

Synthesis of a target polymorph remains more empirical than predictive because crystallization often selects the most accessible nucleation pathway rather than the thermodynamically most stable phase. Here, we show that oxygen stoichiometry converts this empirical synthesis variable into a kinetic control parameter for anatase-rutile selection in TiO$_{2-x}$. Enhanced-sampling simulations reveal that oxygen content alters the nucleation-barrier landscape, switching the relative accessibility of anatase and rutile, even while rutile remains thermodynamically favored. Molecular dynamics simulations show the presence of a diffuse intermediate shell around the nucleus, where oxygen deficiency alters Ti-O coordination and connectivity and drives shell-local motif evolution from anatase-like toward rutile-like environments. A coupled-flux model that integrates barrier competition with shell-mediated attachment/exchange yields a relative nucleation-rate map consistent with reported oxygen-dependent synthesis trends. These results establish stoichiometry-controlled intermediate-shell motif evolution as a kinetic origin of polymorph selection and provide a framework for predicting target phases in composition-coupled crystallization.

cond-mat.mtrl-sci↗

arXiv2025

Towards Fully-Automated Materials Discovery via Large-Scale Synthesis Dataset and Expert-Level LLM-as-a-Judge

Materials synthesis is vital for innovations such as energy storage, catalysis, electronics, and biomedical devices. Yet, the process relies heavily on empirical, trial-and-error methods guided by expert intuition. Our work aims to support the materials science community by providing a practical, data-driven resource. We have curated a comprehensive dataset of 17K expert-verified synthesis recipes from open-access literature, which forms the basis of our newly developed benchmark, AlchemyBench. AlchemyBench offers an end-to-end framework that supports research in large language models applied to synthesis prediction. It encompasses key tasks, including raw materials and equipment prediction, synthesis procedure generation, and characterization outcome forecasting. We propose an LLM-as-a-Judge framework that leverages large language models for automated evaluation, demonstrating strong statistical agreement with expert assessments. Overall, our contributions offer a supportive foundation for exploring the capabilities of LLMs in predicting and guiding materials synthesis, ultimately paving the way for more efficient experimental design and accelerated innovation in materials science.

cs.CL↗