MXgap: A MXene Learning Tool for Bandgap Prediction
MXgap is a Python-based program designed to accelerate bandgap prediction of MXenes using Machine Learning (ML). Instead of performing costly hybrid functional calculations, MXgap predicts bandgaps from features extracted from standard PBE calculations. The program automatically processes VASP outputs to generate descriptors and employs trained ML models for bandgap prediction. Its default model combines a Classifier (to identify metallic vs. semiconducting MXenes) with a Regressor (to predict the bandgap of semiconductors), offering both accuracy and efficiency.
You can download the software here
Reference:
– D. Ontiveros, S. Vela, F. Viñes, C. Sousa, ACS Catal. 2025, 15, 16, 14403–14413
Authors: Diego Ontiveros, Sergi Vela, Francesc Viñes, Carme Sousa