Microstructure of absorber layers in CdTe/Cds solar cells
This work concerns the microstructure of CSS-grown CdTe layers used for CdTe/CdS solar cells. Particular attention is given to how the development of microstructure on annealing with CdCl(_2) may correlate with increases in efficiency. By annealing pressed pellets of bulk CdTe powder, it is shown that microstructural change does occur on heating the material, enhanced by the inclusion of CdCl(_2) flux. However, the temperature required to cause significant effects is demonstrated to be higher than that at which heavy oxidation takes place. The dynamics of this oxidation are also examined. To investigate microstructural evolution in thin-films of CdTe, bi-layers of CdTe and CdS are examined by bevelling, thus revealing the microstructure to within ~1 µm of the interface. This allows optical microscopy and subsequent image analysis of grain structure. The work shows that the grain- size, which is well described by the Rayleigh distribution, varies linearly throughout the layer, but is invariant under CdCl(_2) treatment. Electrical measurements on these bi-layers, however, showed increased efficiency, as is widely reported. This demonstrates that the efficiency of these devices is not dictated by the bulk microstructure. Further, the region within 1 µm of the interface, of similar bi-layers to above, is examined by plan-view TEM. This reveals five-fold grain-growth on CdCl(_2) treatment. Moreover, these grains show a considerably smaller grain size than expected from extrapolating the linear trend in the bulk. These observations are explained in terms of the pinning of the CdTe grain size to the underlying CdS, and the small grain size this causes. A simple model was proposed for a link between the grain-growth to the efficiency improvement. The study also examines the behaviour of defects within grains upon CdCl(_2) treatment provided the first direct evidence of recovery on CdCl(_2) treatment in this system. Finally, a computer model is presented to describe the evolution of microstructure during growth. This is shown to be capable of reproducing the observed variation in grain size, but its strict physical accuracy is questioned.