About RC-like contacts in deep level transient spectroscopy and Cu(In,Ga)Se 2 solar cells

Abstract

The low temperature Deep-Level Transient Spectroscopy (DLTS) signal of two Cu(In, Ga)Se 2 samples on glass with different buffer layers is subjected to a thorough study. A similar signal is observed in the DLTS and admittance spectra of many solar cells and is usually labeled as N1. The standard DLTS theory assumes the investigated device to be a Schottky or p-n diode with an ohmic back contact, and relates the spectral components to capture or emission of free carriers by defect levels in the structure. It is well-known, though, that Cu(In, Ga)Se 2 thin film solar cells deviate from this ideal structure. However, even for a device like this, where advanced numerical modeling is necessary to describe the equilibrium charge distribution as a function of applied bias, a change in the free carrier concentration at a certain position of the device as a result of capture or emission by deep defect levels should satisfy the detailed balance equation. The DLTS experiment performed with conventional and complemental settings for the reverse and pulse bias voltages (V r < V p < 0 and V p < V r < 0, respectively) exhibit characteristics that cannot be explained using free carrier transfer between deep levels-in the bulk or at an interface-and the conduction (electrons) or valence (holes) band of a semiconductor as a model. On the other hand, we show that for the solar cells studied here, the N1 signals follow the behavior predicted for a non-ohmic RC-like contact, as established in our recent paper (J. Lauwaert et al. Journal of Applied Physics 2011) closely. Copyright © 2012 John Wiley & Sons, Ltd.