Optoelectronic Device Modeling of GaAs Nanowire Solar Cells

Abstract

Nanowire solar cells have great potential as candidates for high efficiency, next-generation solar cell devices. To realize their potential, accurate and efficient modeling techniques en- compassing both optical and electrical phenomena must be developed. In this work, a coupled optical and electronic model of GaAs nanowire solar cells was developed, with the goal of building a platform for automated, algorithmic device optimization. Significant work was done on the optical portion of model, with the goal of reducing run- times and improving the level of automation. Enhancements were made to an open-source implementation of the Rigorous Coupled Wave Analysis method for solving Maxwell’s equations, to make it more accurate for modeling nanowire solar cells. Its accuracy and efficiency were thoroughly investigated, and with the enhancements presented here it was shown to be an effective technique for rapid optical modeling of nanowire devices. Purely optical optimizations of a sample AlInP-passivated GaAs nanowire on a GaAs substrate were performed to demonstrate the efficacy of the technique using a Nelder-Mead simplex optimization of device geometry. The optical model was then coupled into a finite volume method based electrical model implemented in TCAD Sentaurus, to compute device efficiencies and ultimately optimize electrical device performance. As a first step, an algorithmic optimization of a p-i-n nanowire solar cell consisting of an AlInP-passivated GaAs nanowire on a Si substrate was performed using the generation rates computed by the enhanced RCWA implementation. The overall geometry was fixed to the result of the optical optimization, and only internal electrical parameters were optimized. The results showed that significant performance improvements can be obtained with the right choice of doping levels and doping region configurations, even without optimizing the global device geometry.