Computational Design of Magnetic Nanographenes through Ab Initio Methods and AI

The project focuses on a computational approach that integrates ab initio and machine learning (ML) methods to explore nanographene structures for quantum devices. These frames featuring armchair edges, present unpaired electrons that interact with each other, giving rise to multiple distinct spin states. These states can be tailored through chemical doping and functionalization, as well as tuned externally via light or electric fields. The main objective is to clarify these electronic interactions, model them using ab initio methods, and explore the nanographene chemical space to build a comprehensive database for training ML models. Ultimately, the project aims to develop a generative adversarial network (GAN) capable of designing frames with tailored properties.