Photocatalytic dye degradation and antibacterial activity of gold nanoparticles: a DFT and machine learning study
M. Yasmin Begum, Mukta Sharma, M. V. Arularasu, P. Vinitha, V. Vetrivelan, Geetha Kandasamy, A. Manikandan, Abinet Gosaye Ayanie, Lalitha Gnanasekaran, Ankush Mehta, Rupesh Gupta
Abstract
. To validate and interpret the experimental outcomes, density functional theory (DFT) simulations were performed, examining the Fermi level, HOMO-LUMO gap, work function, topological properties (ELF, LOL), and thermal stability of AuNPs. In parallel, machine learning (ML) models, including XGBoost, LightGBM, and Neural Networks, were employed to predict electronic band gaps. The XGBoost model showed the highest accuracy with a root mean square error of 0.9878 and a mean squared error of 0.00035, while other models also produced results consistent with DFT values. This combined experimental, theoretical, and data-driven approach highlights the promise of AuNPs for efficient dye degradation and antibacterial applications, offering sustainable solutions for environmental remediation.