Development of commercial high-efficiency solar cells incorporating selective emitters and passivated rear contacts

Abstract

To enable photovoltaics (PV) to become a leading source of energy in a low-carbon future, the cost of solar electricity needs to be reduced to a level that can compete with current electricity prices. From a technology perspective, this reduction can be achieved by increasing the solar cell’s conversion efficiency at similar or reduced manufacturing costs. This thesis describes the development of a high efficiency Si solar cell technology that has the potential to lower the cost per Watt of solar cells, through the use of laser-doped selective emitter and localised rear contacts.

This thesis begins with the development of laser-doped selective emitter (LDSE) technology to overcome the limitations of the front surface design associated with the industrial screen-printed solar cells. In this technology, laser doping is employed to simplify the formation of selective emitters that were conventionally fabricated by photolithographic patterning and long, high-temperature phosphorus diffusion. Defect studies and device loss analysis are performed to facilitate the selection of the most suitable laser parameters for the laser doping application. By using a continuous wave 532 nm laser, a sheet resistance as low as 2 Ω/ is achieved in the selective emitters with minimal laser-induced defects while maintaining a short process time of several seconds per wafer. Optimisation of the laser doping parameters leads to a cell efficiency of 19% being demonstrated on large area, commercial-grade CZ p-type wafers.

This thesis also explores the application of an identical LDSE cell structure on commercial-grade CZ n-type wafers. In this case, the selectively-doped n+ regions at the front surface act as a front surface field (FSF) while the screen-printed Al-alloyed region at the rear surface forms the emitter of the n-type solar cells. A study of the Al-alloyed emitter formation with respect to firing conditions is presented with the aid of SEM imaging. A final Voc approaching 650 mV and pFF of 83% demonstrate the successful formation of p+ emitter with minimum junction shunting. During the development of the n-type LDSE technology, an “apparent” shunting behaviour is observed in the cell areas that are not well-plated. Investigative work involving Photoluminescence imaging and PC1D modeling is detailed to further understand the cause of the “apparent” shunting and its impact on the cell performance. Through this investigation, solutions are devised and successfully implemented to overcome the “apparent shunting”. Further optimisation on the phosphorus FSF diffusion enables the achievement of a FF approaching 80% and a final cell efficiency of 18.7%.

The final stage of this work presents the development of a novel next-generation LDSE cell structure that focuses on improving the rear surface design while retaining the excellent front surface design of the LDSE cell. This cell structure employs a commercially-manufacturable thermal-SiO2/PECVDSiNx stack layer for rear surface passivation; and relies on the use of boron laser doping in conjunction with sputtered Al for localised rear contact formation. Test structures with an implied Voc over 700 mV (prior to laser doping) indicates high surface-passivation quality given by the SiO2/SiNx stack following thermal anneal. Demonstration of sheet resistance as low as 5 Ω / on the boron laser-doped p+ regions highlights the potential for low-temperature Al sintering to form an ohmic contact. Different geometries for the rear contact pattern are designed and implemented to establish optimum current collection. By improving the rear surface design, independently-confirmed cell efficiencies as high as 20% are achieved on large area, standard commercial-grade p-type wafers, with the potential for further efficiency increase exceeding 21%. These are believed to be record performance cell for this type of wafer. These results suggest that this new generation of LDSE cells have the potential of becoming an attractive cell technology in the PV industry.