Download files
Abstract
The focus of this thesis is on the optimisation of metallisation for 10 µm thick laser-crystallised polycrystalline silicon thin film solar cells with glass (LPCSG). These cells originate from the crystalline silicon on glass (CSG) technology. University of New South Wales has been developing new processes for deposition (by e-beam evaporation) and crystallisation (by a line focus diode laser) of thin Si films. There still remain several engineering challenges in the cell such as the high fractions of power loss from existing metallisation scheme and the degradation of open circuit voltage (V_oc) and fill factor (FF). The thesis contributes to LPCSG cell development by optimising the metallisation scheme and investigating the benefits of selectively doped absorber contacts.
The first section of the thesis focuses on optimisation of the point-contact metallisation scheme. The parameters such as size and number of contact openings are optimised by the fractional power loss modelling. This contact optimisation of metallisation scheme results in an improved FF, from 56% to 69.9%.
In the second section, the thesis develops and applies a method for creating a locally diffused boron doped region under the absorber contacts to prevent degradation of V_oc in the completed devices as well as to improve contact resistance. The work in the thesis demonstrates to be ascribed to the interaction between the aluminium and lightly doped p-type silicon after heat treatment for contact-bake.
One way of eliminating the degradation is selective doping of the absorber contacts. The result shows that it eliminates the degradation of both V_oc and FF. However, it can introduce localised ohmic shunt was found in the cell. Dark lock-in thermography images confirm that there is a significant ohmic shunt near the openings which are patterned for the selective boron doping. Reducing the size of the openings helps to minimise the shunt, which results in V_oc of 510 mV and J_sc of 23 mA/〖cm〗^2.
Lastly, boron diffusion temperature for selectively boron doping is tested to find its effects on the device performance. A higher annealing temperature leads to higher doping concentration and deeper dopant diffusion. Four different diffusion temperatures (840 ºC, 860 ºC, 880 ºC and 900 ºC) were tested. As a result, the cells diffused at 900 ºC show high V_oc (543.4 mV). The cells diffused at 880 ºC show high V_oc (530 mV) and relatively high shunt resistance (582 Ω). In contrast, the cells diffused at 840 ºC and 860 ºC result no significant improvement. Two higher diffusion temperature cells (880 ºC and 900 ºC) are chosen to be metallised and investigated. The cell which was processed at 900 ºC showed a significant improvement in V_oc of 559 mV, however, no improvement in FF was seen due to the low shunt resistance.
Further investigation of higher diffusion temperature (e.g. 920 ºC and 940 ºC) may help to identify the effect on the performance parameters such as V_oc, FF and shunt resistance. The investigation of the selectively doped absorber contacts technique can lead to the stable cell performance and the higher cell efficiency up to 9%.