In the PV sector, it pays to keep on top of all the latest technological developments, which is why PES has always published a wide range of whitepapers and research materials. We’re committed to continuing this with the introduction of a new, regular feature: Tech Dissemination. This issue discusses the improvements made possible by modelling.
Abstract
Technology Computer-Aided Design (TCAD) using integrated process and device simulation tools is widely used in the semiconductor industry to reduce development costs and time, and to enhance device performance. In the PV industry, TCAD has usually focussed on device simulation. This paper shows results of integrated TCAD applied to the Laser Grooved Buried Contact (LGBC) silicon solar cell process, using 2D simulation of process steps to predict the solar cell structure and 2D device simulation of the resultant cell operation. This enables direct assessment of the impact of changing fabrication steps on key cell parameters such as Voc, jsc, FF and efficiency. Results suggest integrated TCAD may significantly accelerate development of future PV processes.
Introduction
TCAD is widely used in the microelectronics industry to reduce process development time and costs, and to ensure optimal performance of fabricated devices. In recent years TCAD models, especially those used to describe processing steps, have matured considerably as a result of hundreds of person-years of research in universities, institutes and companies. Process models now require relatively little ‘calibration’ to specific Si-based technology processes. At the same time, industry-standard device models have been extended to include opto-electronic interactions, thus enabling application to PV technology.
In the PV industry the use of TCAD is rapidly increasing, but is not yet well established. As in the microelectronics industry one to two decades ago the focus has mainly been on device modelling. Little if any work is done using integrated TCAD to investigate the impact of changes in processing steps. Such investigations are done using in-line processing experiments, despite difficulties in picking out true optima because of numerous trade-offs between process parameters, and variations arising from process fluctuations and drift.
