Improved Hole Diffusion Lengths in Bulk n-Type GaAs for High Efficiency Solar Cells
- PDF / 284,936 Bytes
- 4 Pages / 420.48 x 639 pts Page_size
- 41 Downloads / 181 Views
IMPROVED HOLE DIFFUSION LENGTHS IN BULK n-TYPE GaAs FOR HIGH EFFICIENCY SOLAR CELLS D. WONG, T. E. SCHLESINGER AND A. G. MILNES Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh PA15213
ABSTRACT A method for wafer annealing which is effective in suppressing defects and raising minority carrier diffusion lengths in n-type bulk GaAs is described. The beneficial effect of the annealing is shown to be associated with the proximity surface with measurements of photoresponse as a function of depth. The concentration of the hole trap HCX (EV+0.2geV) varies as a function of depth from the surface, qualitatively, as might be expected of the concentration of the dominant recombination center in the material. The impact of improving the material in this manner on the performance of Zn diffused solar cells is demonstrated. INTRODUCTION Annealing can greatly alter the defect structure in bulk GaAs [1,2]. Given the correct anneal conditions, beneficial effects such as defect suppression [3], uniformity of electronic properties [4] and improved device performance [51]may be obtained. In earlier reports [6,71 we have shown that proximity wafer annealing of n-type GaAs in sealed quartz ampoules results in defect suppression and increases in hole diffusion length (L p)of up to a factor of 3. Such improvements can be obtained in Liquid encapsulated Czochralski (LEC) and horizontal Bridgman (HB) wafers, regardless of choice of dopant (Si or Te) and independantly of whether the material has been boule annealed. Deep level transient spectroscopy (DLTS) measurements of the defect structure in the specimens show that the dominant recombination center in bulk n-type GaAs manifests itself as a hole trap at Ev+0.29eV (HCX). In the present work we have further characterized the effects of proximity anneals in order to ascertain the nature of the mechanism responsible for the changes observed. EXPERIMENTAL PROCEDURE All the specimens discuessed here were cut from wafers from the same crystal of Te-doped (lxl017 cm3 ) LEC GaAs. Annealing was done in evacuated ampoules made from high purity quartz tubing. The proximity annealed specimens were placed with their polished faces together during annealing. We have described the specimens which were not stacked with their polished faces protected in this manner during annealing as having had "exposed surface" anneals. L in the specimens was determined using a cleaved diode electron beam induced current (EBIC) technique. Photoresponse measurements were made using the electrochemical cell unit from a commercial semiconductor profiler (Polaron 4100). Photocurrent was measured as a function of depth by making successive measurements on the same specimen following each of a sequence of chemical etches. For profiling of defect concentrations as a function of depth, several different pieces of each specimen were chemically etched for different lengths of time. p+n junctions for characterization of minority carrier traps by DLTS were then fabricated by solid phase diff
Data Loading...
