Photoluminescence (PL) usually provides macroscopic, high quality spectroscopic data. Cathodoluminescence (CL), on the other hand, offers the same information with microscopic imaging. However, replicating PL signatures in a CL system is not straightforward since matching experimental conditions, such as temperature and excitation density, is difficult. The matter is further exacerbated by inherent differences in the nature of excitation: electrons versus photons. Our work with high purity semiconductors suggests that CL is generally more sensitive to excitation "circumstance" than PL. For example, electrons can cause sample charging and contamination-related phenomena that dramatically affect CL. Changes in surface attributes (e.g., by chemical passivation) also affect PL and CL signals differently. Here, we extend previous work on GaAs by exploring the role of surface topography (by atomic force microscopy) and temperature (1.8K-100K) on excitonic lineshapes. We find that topographic subtleties strongly influence the character of exciton-polariton luminescence. We interpret these changes in terms of non-classical scattering phenomena derived from microscopic roughness. These microscopic changes also influence the temperature behaviour of excitons in crystals. Specifically, we find that passivated samples are brighter partly because there is a corresponding reduction in the (Arrhenius) activation energy for excitonic processes. In summary, the changes in surface topography and corresponding recombination physics seem well correlated.
Myhajlenko, S.; Puechner, R. A.; Edwards, J. L.; and Davito, D. B.
"Can Photo- and Cathodoluminescence be Regarded as Complementary Techniques?,"
Scanning Microscopy: Vol. 1995
, Article 19.
Available at: https://digitalcommons.usu.edu/microscopy/vol1995/iss9/19