Modelling and Optimising GaAs-AlxGa1-xAs Multiple Quantum Well Solar Cells - Condensed Matter > Mesoscale and Nanoscale PhysicsReport as inadecuate




Modelling and Optimising GaAs-AlxGa1-xAs Multiple Quantum Well Solar Cells - Condensed Matter > Mesoscale and Nanoscale Physics - Download this document for free, or read online. Document in PDF available to download.

Abstract: The quantum well solar cell QWSC is a p - i - n solar cell with quantumwells in the intrinsic region. Previous work has shown that QWSCs have agreater open circuit voltage Voc than would be provided by a cell with thequantum well effective bandgap. This suggests that the fundamental efficiencylimits of QWSCs are greater than those of single bandgap solar cells. Thefollowing work investigates QWSCs in the GaAs-AlxGa1-xAs materials system. Thedesign and optimisation of a QWSC in this system requires studies of thevoltage and current dependencies on the aluminium fraction. QWSCs withdifferent aluminium fractions have been studied and show an increasing Voc withincreasing barrier aluminium composition. The QE however decreases withincreasing aluminium composition. We develop a model of the QE to test novelQWSC designs with a view to minimising this problem. This work concentrates ontwo design changes. The first deals with com- positionally graded structures inwhich the bandgap varies with position. This bandgap variation introduces anquasi electric field which can be used to increase minority carrier collectionin the low efficiency p and n layers. This technique also increases the lightflux reaching the highly efficient depletion regions. The second design changeconsists of coating the back of the cell with a mirror to exploit the portionof light which is not absorbed on the first pass. A model of the QE ofcompositionally graded QWSC solar cells with back surface mirrors is developedin order to analyse the effect of these design changes. These changes areimplemented separately in a number of QWSC designs and the resultingexperimental data compared with the model. An optimised design is thenpresented.



Author: James P. Connolly

Source: https://arxiv.org/







Related documents