Dual-Wavelength Time-Resolved Photoluminescence Study of CdSe<sub>x</sub>Te<sub>1-x</sub> Surface Passivation via Mg<sub>y</sub>Zn<sub>1-y</sub>O and Al<sub>2</sub>O<sub>3</sub>
Deborah L. McGott, Brian Good, B. Fluegel, Joel N. Duenow, Colin A. Wolden, Matthew O. Reese
Abstract
Voltage loss is currently one of the biggest challenges facing cadmium telluride (CdTe) based photovoltaics. Determining the location(s) of major voltage loss within the device stack (e.g., front/back interface, grain boundaries) is therefore of primary interest. Here, we present a custom-built time-resolved photoluminescence system with two excitation wavelengths—670 (standard) and 405 nm—to probe the device stack at depths of approximately 130 and 35 nm, respectively; their comparison helps differentiate interface and bulk contributions to carrier lifetime. We apply this system to examine the passivation effect of two significant recent advances in CdTe: the incorporation of Se to form graded CdSe <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</sub> Te <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-x</sub> and the replacement of CdS with Mg <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">y</sub> Zn <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-y</sub> O. It is found that <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</i> = 0.2 Se is required to obtain lifetime improvements, primarily in the bulk. Additionally, evidence for trapping at the Mg <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">y</sub> Zn <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-y</sub> O/CdSe <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</sub> Te <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-x</sub> interface was observed. This indicates further work is required to sufficiently passivate the front interface.