This paper presents a fully automated, hardware-accelerated code for high-throughput discovery of 2D ferromagnetic materials with high Curie temperatures. The code combines first-principles calculations with Monte Carlo simulations to estimate Curie temperature (Tc) from crystal structures. A high-throughput scan of 786 materials identified 26 with Tc exceeding 400 K. A machine learning model, trained on the results, further enhanced the discovery process, identifying additional high-Tc materials. This work significantly improves the computational toolbox for exploring 2D magnetism and its applications.

Arnab Kabiraj, Mayank Kumar, Santanu Mahapatra