Defect passivation and activation of thin film CdSeTe/CdTe solar cells

Multiple sustainable energy sources are of paramount importance in today's ever power hungry-world. Solar photovoltaics have provided cleaner, sustainable energy for nearly 70 years. While the market was previously purely silicon solar devices, the low cost thin film processing of cadmium telluride devices has allowed it to become the cheapest power per watt material for solar modules.

The processing of the CdTe absorber has several important factors; firstly a CdCl₂ treatment to remove stacking faults, passivate grain boundaries and give CdTe it's photovoltaic ability, secondly, the addition of Se at the front interface, further passivating grain boundaries, increasing the lifetimes and grading the bandgap in the device, and also the production of large grain CdTe and the incorporation of other dopants such as arsenic.

The work presented in this thesis focuses on the effect of replacing Cl with Br in CdTe processing and separately, the pulsed DC sputtering (PDCMS) deposition method of CdSe as a Se source for CdTe devices.

CdCl₂ treatment of CdTe is a staple for almost all CdTe based photovoltaic devices. Studies of CdBr₂ treatment of CdTe have been performed to observe similarities and differences in the effects of the halogens. An initial study was performed with varied pressure, temperature and treatment times to observe if CdBr₂ would have an effect on CdTe doped with Cl. This was successful and led to an efficiency after CdBr₂ that was higher than an anneal without a halide salt. 

After this initial experiment, optimisation was attempted to find a temperature and time of CdBr₂ that would lead to the best results. Finally, a higher quality back contact method including Cu was applied to attempt to increase the efficiency further. 

CdBr₂ treatment of CdTe leads to similar effects as seen with CdCl₂, bromine decorating the grain boundaries is seen in high resolution compositional maps, stacking faults are removed and there is an increase in photoactivity. However, a major point is that bromine activation requires chlorine to be present in the bulk for it to work. This indicates two separate effects of the halides, one in the bulk with Cl creating p-type CdTe with A-centres, and one passivating the grain boundaries.

Further investigations on the effects of CdCl₂ treatment of CdSe polycrystalline films would be beneficial in the production of a CdSe absorber and full device. CdBr₂ of CdTe was not viable alone, however it raises the possibility of combining halogens in a device to get benefits from both halogens.

Sputter deposition of CdSe was optimised for high quality as deposited films, characterised by transmission electron microscopy and X-ray diffraction to observe the microstructural effects of the sputter conditions. After annealing these films, bubbles and blisters appeared. The sputter conditions were then optimised for reducing the subsequent voids and blistering after annealing. The bubbles and blisters were caused by sputter gas incorporation into the film, this is complemented by a theoretical molecular modelling study of the behaviour of Ar gas in CdSe films. After preventing the blisters and reducing the gas bubbles that would appear after anneal, the CdSe films were CdCl₂ treated to promote greater structural rearrangement and to observe any phase changes.  Cl was found to decorate the grain boundaries of the CdSe and promoting a reasonable photoactivity when compared to untreated CdSe. The CdCl₂ treatment also led to further Ar gas agglomerating into bubbles and blistering, while promoting grain growth and stacking fault removal. Evaporated CdSe was also analysed and this had none of the blister problems associated with sputter deposition, larger grains were formed after CdCl₂ treatment, and the photoactivity and lifetimes were much higher than those found in sputtered treated CdSe.