Predictive Modeling in High-Temperature Superconductors: Comparative Insights from Density Functional Theory and Machine Learning

Abstract

 Superconductivity, defined by the disappearance of electrical resistance below a critical temperature, holds significant promise for future technologies such as quantum computing, maglev transport, and energy-efficient power grids. Among the many superconducting materials, nickel and copper oxide stand out due to their comparable layered structures but distinct electronic properties. Understanding and predicting their superconducting behavior is essential for discovering new high-temperature superconductors. Density functional theory and machine learning have become indispensable tools in this effort. While DFT offers insights into band structure and orbital interactions, ML models enable high-throughput screening of potential materials. However, data scarcity, model interpretability, and limited generalizability remain significant barriers to progress. This review critically evaluates the effectiveness and limitations of these predictive techniques, identifies unresolved issues, and discusses integrative research strategies that combine theory, simulation, and experimentation to accelerate discoveries in high-Tc superconductivity. This comparative analysis offers a roadmap for building interpretable, efficient, and scalable predictive models in the next phase of high-Tc materials research.