Radial pn Junction, Wire Array Solar Cells

Citation

Kayes, Brendan Melville (2009) Radial pn Junction, Wire Array Solar Cells. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/X9WE-V642. https://resolver.caltech.edu/CaltechETD:etd-09222008-173738

Abstract

Radial pn junctions are potentially of interest in photovoltaics as a way to decouple light absorption from minority carrier collection. In a traditional planar design these occur in the same dimension, and this sets a lower limit on absorber material quality, as cells must both be thick enough to effectively absorb the solar spectrum while also having minority-carrier diffusion lengths long enough to allow for efficient collection of the photo-generated carriers. Therefore, highly efficient photovoltaic devices currently require highly pure materials and expensive processing techniques, while low cost devices generally operate at relatively low efficiency.

The radial pn junction design sets the direction of light absorption perpendicular to the direction of minority-carrier transport, allowing the cell to be thick enough for effective light absorption, while also providing a short pathway for carrier collection. This is achieved by increasing the junction area, in order to decrease the path length any photogenerated minority carrier must travel, to be less than its minority carrier diffusion length. Realizing this geometry in an array of semiconducting wires, by for example depositing a single-crystalline inorganic semiconducting absorber layer at high deposition rates from the gas phase by the vapor-liquid-solid (VLS) mechanism, allows for a "bottom up" approach to device fabrication, which can in principle dramatically reduce the materials costs associated with a cell.

This thesis explores the potential of this design, first theoretically and computationally, and then by exploring the growth of structures with the proposed morphology via methods with the potential for low cost, and finally by the experimental characterization of cells.

Thesis Files