Interfacial quantum dynamics and AI-driven engineering of CdS quantum dot-sensitized solar cells based on GO-TiO

The progress of quantum dot-sensitized solar cells (QDSSCs) is still constrained by inefficient interfacial charge separation and the lack of predictive models that directly link nanoscale quantum interactions to photovoltaic performance. In this work, we address these challenges by fabricating CdS QDSSCs with a graphene oxide (GO)-modified TiO2 photoanode and developing a hybrid theoretical-AI framework. Incorporation of GO improves electron mobility and charge transfer, with the best device (0.12 g GO) delivering a short-circuit current density (Jsc) of 2.03 mA cm-2 and an open-circuit voltage (Voc) of 0.43 V. To interpret these results, we establish an interfacial Hamiltonian model that provides an analytical photocurrent expression accounting for quantum coupling at the CdS/GO-TiO2 interface. Complementarily, artificial neural networks (ANNs) trained on experimental J-V data accurately predict photocurrent behavior under varying conditions. By addressing both the mechanistic understanding and predictive capability gaps, this hybrid physics-AI strategy provides a novel and robust pathway for the rational optimization of graphene-based QDSSCs.