Czochralski growth of low-oxygen silicon single crystals for high-efficiency solar cell applications

In 2021 India announces $600 million incentives schemes for solar manufacturing. The Success project has therefore had very good response in India. The challenge in future fabrication of silicon for photovoltaic applications is to reduce the cost of solar power, were the improvement of crystal quality and yield is an important factor. The Czochralski process has been refined and optimized for mass production of silicon single crystals with diameters from 6" to 8" and charge weights up to 150 kg. In a Czochralski configuration the silicon feedstock is melted in a high purity quartz crucible and a cylindrical crystal is pulled from the free surface at rates of a few ten millimetres per hour. The heating of the melt gives rise to buoyant convection. Crucible and crystal are commonly rotated at steady rates in opposite directions, so that centrifugal forces counteract free convection. The fluid flow and heat transfer processes in the melt of a Czochralski configuration system is extremely complex and in large scale growth systems the flow is mostly turbulent. Since the silicon is melted in a quartz crucible, the growing crystal contains significant levels of oxygen due to dissolution of the quartz crucible. Solar cell manufactures attempting to limit negative impact of oxygen and related defects will ask for much lower oxygen levels in future. SUCCESS is a bilateral project between the Norwegian research organisation SINTEF and the Indian Institute of Technology Bombay. The main objective of SUCCESS is to develop a cost-efficient method/tool for controlling the heat and mass transport in Czochralski silicon growth aiming at a significant reduction of oxygen levels in the crystal through the application of a versatile technique where the fluid in the crucible is accelerated and decelerated periodically. By means of numerical simulation of the fluid flow in the crucible optimal rotation schemes will be revelead to supress the uptake of oxygen by the crystal. These findings will be validated by crystal growth experiments, and the impact of material improvements will be finally demonstrated at solar cell and module levels. Numerical modelling in CGSIM is used to generate boundary conditions for new3D models developed in Open foam. The industrial scale Czochralski crystal puller at SINTEF was modified and two ingots have been successfully grown. One reference ingot and one ingot with the osciallation rotation. The ingots have been processed at SINTEF and will be further cut into wafers at CEA Ines. The wafers will be sent to IIT. An online workshop on "SiliconCrystalGrowth" was organized jointly bt NCPRE, IIT Bombay and SINTEF as part of the Success project. More than 600 participants was registered for the conference leading to several meeting between Norwegian and Indian Industry for further cooperation and technology transfer.

In the project we have developed a new modelling framework, where results from a 2D model are used as input to an open source (OpenFoam ) 3D model. The model will be useful for more advanced modelling of the Czochralski process and other solidfication processes. The use of open source software allows for more efficient sharing and co development for academic and industrial research.

In conventional Czochralski growth of single-crystalline silicon, a high purity quartz crucible holds the silicon melt from which the crystal is grown. The Czochralski process, which is a typical batch process, has been refined and optimized for mass production of silicon single crystals with diameters from 6” to 8” and charge weights up to 150 kg for PV applications. Since the silicon is melted in a quartz crucible, the ingot contains significant levels of oxygen due to dissolution of the crucible. Solar cell manufactures attempting to limit negative impact of oxygen and related defects will ask for much lower oxygen levels in future. The SUCCESS project aims to reduce oxygen levels in the crystal by at least 50%. It will be achieved through external control of melt stirring and thus transport of oxygen by periodically variation of the crystal and/or crucible rotation. The basic rationale is that an optimal stirring in the melt can be achieved by changing the rotation rate of the crucible and/or crystal. Improvements in material performance will be demonstrated at solar cell and module level. This project builds on SINTEF's and IIT Bombay's internationally leading position in research on silicon for solar cell applications. Both research partners will build a relationship along the entire silicon photovoltaic value chain, where Norway has natural advantages on the upstream in silicon ingot and wafer manufacturing, and India on the downstream on solar cell and module production. The project is supported by Norwegian silicon ingot and wafer manufacturers, Norsun and Norwegian Crystals, and Indian solar cell and module manufacturer Mundra Solar, which is part of the Adani Group. This collaboration will lead to mutual exchanges of knowledge and contribute to further development of SINTEF and IIT Bombay in close collaboration with the industry.