Modelling Graphene-based Transparent Electrodes for Si Solar Cells by Artificial Neural Networks

Abstract

Transparent electrodes based on conductive transparent oxides (TCO) are increasingly invading the photovoltaic (PV) field because of their unique ability to reconcile high transparency with good electrical conductivity. The TCO market is dominated by the Indium oxide doped with Tin (ITO) with a resistivity of 30-80 Ω/sq and a transmittance of 90 % in the visible range. Yet, its cost is rising due to the high indium content, is one of the reason that encouraging research on alternative materials essential for the development of PV technologies. It is in this theme that graphene, a material with exceptional properties, is tested as a design material for transparent electrodes for Si solar cells. In this paper, we optimized optically and electronically the graphene-based transparent electrodes (G-TE) by proposing a model of simulation based on artificial intelligence and specifically artificial neural networks (ANN) which is the ANN-model. Therefore, to achieve an appropriate characterisation of a behaviour of G-TE for the Si solar cells, the ANN model has been performed to simulate and optimise different parameters of the G-TE, by controlling graphene layer number, tuning graphene work function, and deduce the suitable transmittance and resistivity in order to have a complete adjustment for these parameters. Our study mentioned that a G-TE with three layers of graphene and a work function of 4.75 eV leads for a sheet resistance of 50 Ω/sq and transmittance of 91.4 %; these results suggest that G-TE is a promising candidate in the TCO field.