Impurities in silicon solar cells

Abstract

The effects of various metallic impurities, both singly and in combinations, on the performance of silicon solar cells have been studied. Czochralski crystals were grown with controlled additions of secondary impurities. The primary dopants were boron and phosphorus while the secondaires were: A1, B, C, Ca, Co, Cr, Cu, Fe, Mg, Mn, Mo, Nb, P, Pd, Ta, Ti, V, W, Zn, and Zr. Impurity concentrations ranged from 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">10</sup> to 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">17</sup> /cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> . Solar cells were made using a conventional diffusion process and were characterized by computer reduction of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I-V</tex> data. The collected data indicated that impurity-induced performance loss was primarily due to reduction of the base diffusion length. Based on this observation, an analytic model was developed which predicts cell performance as a function of the secondary impurity concentrations. The calculated performance parameters are in good agreement with measured values except for Cu, Ni, and Fe, which at higher concentrations, degrade the cell substantially by means of junction mechanisms. This behavior can be distinguished from base diffusion length effects by careful analysis of the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I-V</tex> data. The effects of impurities in n-base and p-base devices differ in degree but submit to the same modeling analysis. A comparison of calculated and measured performance for multiple impurities indicates a limited interaction between impurities, e.g., copper appears to improve titanium-doped cells.