Spatially resolved photoluminescence lifetime mapping in the vicinity of extended defects in semiconductors using a time correlated single photon counting system and confocal photoluminescence microscopy
Photoluminescence mapping and time resolved photoluminescence imaging have previously been used to study charge carrier dynamics in bulk semiconductors. Recently, our research group has used spatially resolved PL mapping to examine and model carrier diffusion and recombination in the vicinity of extended defects in GaAs. For the present work a system for obtaining both time and spatially resolved PL images has been developed for closer examination of these phenomena. The system was constructed using a Horiba LabRam 800 confocal micro-Raman system, a time correlated single photon counting (TCSPC) system, and a fast detector which employs a hybrid photomultiplier tube. The system was used to collect photoluminescence lifetime curves for a series of points from a 1-D scan across the defect site. Lifetime data was extracted from single exponential fits of the curves and showed a significant decrease in photoluminescence lifetime in the center of the dislocation. Outside the defect, a slight increase in photoluminescence lifetime and intensity above the background was observed for points very close to the defect (within 2-5 microm). Subsequent CW PL maps made at the same low excitation used for the lifetime measurements revealed a 'halo' around an elongated defect, which was in agreement with the intensity distributions seen in the 1-D lifetime scans from the same region. Lifetime and intensity data are used to construct plots and simple 1-D maps of these quantities vs. position. Carrier diffusion lengths are calculated from carrier lifetimes and the results are compared to diffusion length and carrier recombination data previously obtained by spatially resolved CW PL mapping, where the carrier diffusion length was assumed to be spatially invariant. Measured photoluminescence lifetimes and corresponding calculated diffusion lengths were found to be much shorter than previously measured values. Alteration of material properties near the defect from high power laser exposure and the low excitation used in making the lifetime measurements are discussed as possibly contributing to both short photoluminescence lifetimes as well as the appearance of the halo effect.