Experimental investigation and optimization of integrated photovoltaic and photoelectrochemical hydrogen generation

This paper examines the transport phenomena and optimal performance of an integrated concentrated photovoltaic and photoelectrochemical hydrogen reactor. Individual components and the overall system are studied experimentally including the performance of the concentrator, spectrum-splitting mirror, electrolyser, reactor, and photovoltaic module. Integrating the solar concentration with a spectrum-splitting mirror allows simultaneous photovoltaic electricity generation and direct photonic energy conversion to produce hydrogen via electrolytic and photoelectrochemical water splitting. A multi-objective optimization of the integrated system is performed with machine learning and integration of a neural network. This yields a relationship between the system inputs and outputs. The neural network is used to optimize the overall system through a genetic algorithm. Numerical and experimental results are presented and discussed in the paper.