Details
Focus
This topic centers on three tasks: extracting actionable insights from large-scale experimental data and computational screening, building predictive models for the structural and thermodynamic stability of oxide semiconductors, and converting these findings into practical, high-performance thin-film transistor (TFT) fabrication strategies.
Methods
- Knowledge Structuring: Organize comprehensive experimental datasets using Large Language Model (LLM)-powered extraction to identify strategies to achieve high performance.
- Modeling for Synthesizability: Evaluate thermodynamic phase diagrams and the effects of epitaxial strain to predict the feasibility of synthesizing heterostructural alloys.
- Design-Space Prioritization: Prioritize optimal fabrication parameters and architectures—such as Plasma-Enhanced Atomic Layer Deposition (PEALD), in situ capping layers, and precise cation doping—to overcome the inherent mobility-stability trade-off.
What we deliver
- Recipe Candidates: Ranked synthesis conditions, including specific sub-cycle ratios for Atomically-Ordered IGZO, optimal post-deposition annealing windows, and strategic doping compositions (e.g., Ytterbium-doped IZO) for targeted high-mobility and extreme reliability.
- Decision Support: Interpretable thermodynamic indicators and quantitative visualizations linking input variables (such as atomic structure, strain, and gate insulator selection) to expected device stability, subgap defect suppression, and carrier transport.
- Faster Iteration: A streamlined framework connecting AI-driven literature reviews and ab initio computational design directly to experimental validation, enabling the rapid development of next-generation semiconductor devices.
Related Papers
- AI-driven quantitative review of mobility-stability trade-off in oxide semiconductors (Nano Convergence, 2026)
- Ultra-High Mobility Atomically-Ordered InGaZnO Transistors Through Atomic Layer Deposition (Advanced Electronic Materials, 2025)