Pulsed Laser Dielectric Ablation for Copper-plated Silicon Solar Cells

Copper-plated metallisation can act as an alternative to the dominant Ag screen-printing with a reduced material cost for Si solar cell manufacturing. However, concerns of laser-induced damage with carrier recombination and plated metal adhesion have contributed to a low market fraction of Cu-plated modules. This thesis aimed to further understand the contact formation process using light-induced plating (LIP) on selective openings formed by short pulse laser ablation of dielectrics on p-type Si solar cells including the passivated emitter and rear cell (PERC). A composite optical-thermal model was developed to explore the light-matter interaction between short laser pulses and textured silicon nitride (SiNx)/Si surfaces. It is shown that picosecond laser pulses of the longer wavelength of 532 nm result in melting depths of ~ 1 µm and the SiNx being primarily ablated by the indirect ‘spallation’ process, whereas for 266 nm pulses melting depths are predicted to be limited to < 150 nm and the direct ablation contributes to a greater extent in the removal of the SiNx antireflection coating. Longer nanosecond pulses result in less steep temperature gradients, and for 266 nm nanosecond pulses, an increased melting depth compared to picosecond pulses and hence are more appropriate for laser doping processes than dielectric ablation for contact formation. Surface chemistry and Ni nucleation were identified as key factors for the plated metal adhesion. Residual SiNx, resulting from incomplete laser dielectric ablation, was shown to hinder the formation of Ni silicides on both Al back surface field (BSF) and PERC cells and resulted in low busbar pull forces. Although busbar pull forces > 1 N mm-1 could be readily achieved on Al BSF cells with minimal electrical impact on the cells, PERC cells were more sensitive to laser damage and comparable busbar adhesion could not be achieved without significantly reduced VOC and pFF. It is shown that the use of pulsed LIP of Ni can be used to improve the busbar adhesion through enabling a greater density of Ni nucleates which can act as adhesive anchor points along busbars. However, the improvements in adhesion may not be sufficient to address the adhesion/electrical performance trade-off that appears to exist for higher efficiency PERC cells. Finally, this thesis shows that low pull forces can be measured even though bulk Si is fractured during 180° busbar pull tests. This suggests that an examination of the peeled interface is required before interpreting the busbar adhesion results, a finding that may also be relevant to the more dominant screen-printed cells produced industrially