Predicted annual energy yield of III-V/c-Si tandem solar cells: modelling the effect of changing spectrum on current-matching.

Ian Mathews,Shenghui Lei,Ronan Frizzell

doi:10.1364/oe.384620

Abstract

High efficiencies of >30% are predicted for series-connected tandem solar cells when current-matching is achieved between the wide-bandgap top cell and silicon bottom cell. Sub-cells are typically optimized for current-matching based on the standard AM1.5G spectrum, but in practice, the incident radiation on a solar cell can be very different from this standard due to the effects of the sun's location in the sky, atmospheric conditions, total diffuse element etc. The resulting deviations in spectral content from optimum conditions lead to current mismatch between tandem cell layers that adversely affects the device's performance. To investigate the impact of this issue the energy yield (%) of tandem solar cells comprising a III-V wide-bandgap solar cell connected electrically and optically in series with a silicon bottom cell was simulated over a full year using measured spectral data from Denver, CO. Top cells with bandgaps from 1.5-1.9 eV were modelled using an external radiative efficiency method. The predicted annual energy yields were as high as 31% with an optimum 1.8 eV top cell, only 2.8% lower (absolute) than the AM1.5G predicted efficiency. The annual energy yield of tandem cells with no current-matching constraint, i.e. parallel-connected devices, was also simulated. Here the difference between series and parallel connections were only significant for non-optimum bandgap combinations. Our results indicate that AM1.5G based optimization of sub-cells can be effectively employed to achieve high energy yields of >25% for III-V/Si tandem solar cells in mid-latitude US locations, despite the continuous variation in spectra throughout a calendar year.

Highlights

Silicon solar cells first demonstrated by Bell Laboratories in the 1960’s have become the dominant PV technology of the last number of decades
To check the suitability of AM1.5G based optimisation of tandem cells, we model the efficiency of silicon-based tandem solar cells comprising a III-V top cell using this standard spectrum and compare it to the predicted annual energy yield considering real spectral data measured in Denver, Colorado
Before calculating the annual energy yield from the various bandgap combinations, the top cell thickness was optimised to current match with a silicon bottom cell, with the optimisation carried out considering the AM1.5G spectrum

Summary

Introduction

Silicon solar cells first demonstrated by Bell Laboratories in the 1960’s have become the dominant PV technology of the last number of decades. A number of new cell architectures are under investigation to boost silicon solar cells performance above 30% and significantly reduce $/W The majority of these look to combine an additional pn-junction(s) with the primary silicon cell to form tandem cell architectures, mimicking the high-performance III-V multi-junction solar cells used in concentrated photovoltaic systems (CPV) [6,7]. Strategies to form these cells typically provide a wide-bandgap top cell, on the front surface of the silicon sub-cell, in the form of a-Si [8], Perovskite [9], III-V solar cells [10] or nanowires [11]. Significant efficiencies of close to 30% under 1-sun illumination have been achieved with dual-junction cells that combine a III-V top cell and silicon bottom cell [10]

Methods

Results

Conclusion