Laser Annealing of CuInSe2 Electrodeposited Precursors as the Absorber Layer in Thin Film Solar Cells

This thesis demonstrates a method by which a non-vacuum deposited CuInSe2 precursor is transformed into a semiconductor absorber layer, suitable for completion into a photovoltaic device, using 1 s laser annealing. A final device produced using this absorber fabrication method
gave a 1.6 % best conversion efficiency. This represents the first reported working device originating from electrodeposition - laser annealing. In this thesis, the steps taken to achieve this result are analysed in detail. This work is split into four parts.
Firstly, the precursor structure and the optical properties of its component materials were investigated to ensure optimal interaction between the laser beam and sample. Using the selected codeposited CuInSe2 precursor it was shown that it is possible to stimulate grain growth and atomic diffusion by using a 1064 nm Nd:YAG laser with only a 1 s annealing time. Secondly, it is shown that even on these rapid annealing time-scales the thermodynamic equilibrium reactions between the ternary CuInSe2 and its constituent materials must be considered. This is proven by the formation of the first device from a laser annealed absorber only when a sufficiently elevated partial pressure of Se is supplied during the annealing process. Absorbers formed under lower Se activities led to devices which were shunted or showed negligible efficiency. A correlation was identified between the Se partial pressure and the absorber properties including its Se content, grain size and PL yield.
Thirdly, a finite element model capable of predicting the film temperature during laser annealing
is demonstrated. By considering the Gaussian nature of the irradiating beam it is seen
that the peak temperature of the CuInSe2 film, which formed the 1.6 % device, fluctuated spatially by > 300 oC during processing. Variations in temperature led to different rates of atomic diffusion and grain growth and resulted in lateral inhomogeneities in the absorber. The low device efficiency is believed to be partially caused by these variations.
Finally, the chemical composition of the film and its effect on absorber properties is established.
Increasing the Cu/In ratio of CuInSe2 precursors and incorporating Na into CuInSe2 absorbers is shown to increase their crystal coherence length. However, the positive effects related to these elements must be balanced against their impact on the optoelectronic properties of the absorber when present in high concentrations.
Therefore it is hoped that this initial ‘first device’ efficiency validates this fabrication route
and motivates future research on this interesting topic.