Diode laser processing of silicon thin-film solar cells

A large area cw diode laser is utilised in this thesis to develop three new processes to improve the efficiency and reduce the cost of silicon thin-film solar cells made on glass substrates. The first is a defect annealing process which replaces a belt furnace anneal for the removal of crystallographic defects and activation of dopants in solid-phase crystallised silicon. The highest substrate temperature required during the process sequence is reduced from 960C to 620C, which opens the door to the use of cheaper glass substrates. A peak 1-sun voltage of 492 mV is achieved on planar samples, which is an improvement of 30 mV over the belt furnace process. The films are shown to be partly recrystallised during the process, significantly improving the material quality of the melted section while retaining the dopant profile within the un-melted section. Completed devices are shown to match the performance of the optimised belt furnace process. The second is a complete laser crystallisation process applied to thick a-Si films. Here the film is molten across the entire thickness, and a regime is found whereby long linear grains are formed by continuous growth with the previous crystallised region acting as a seed for the newly crystallising material. Grains up to millimetres in length and hundreds of microns in width have been grown with virtually zero detectable intra-grain defects. A rear diffused emitter isused to create a p-n junction, which produces 1-sun voltages as high as 539 mV. IQE measurements show that the glass-side of the films is very well passivated and that the diffusion length is much longer than the device thickness. Cells up to 7.01% efficiency have been demonstrated without light trapping or optimisation of many of the key processes. The third new process is large area solid-state diffusionusing a diode laser to drive in spin-on dopants. This process has been used to form a junction and demonstrated some of the highest voltages on the diode laser crystallised silicon and to form a back surface field for solid-phase crystallised cells. A thermal/doping model is presented which predicts the depth of diffusion and its accuracy is confirmed by dopant profile measurements