On the potential and limits of large area seeding for photovoltaic silicon

Single crystal production of silicon for solar cell substrates has relied on the Dash neck technique developed more than 50 years ago. The technique is simple and repeatable and enables truly dislocation free crystal growth. It does have drawbacks, however, including limits on throughput and some structural difficulties. It has long been assumed that dislocation-free growth is not possible by any other method. In the ‘quasi-mono’ crystal growth technique, one of the key elements is the use of large area single crystal seeds. By melting the seeds at near-equilibrium conditions, it is feasible to avoid the production of dislocations during melting. We will review the dislocation relevant details of the large area seeding process and present best case results for dislocation density, including measured minority carrier lifetimes in excess of 1ms on p-type material. We will focus on dislocation density exclusive of seed boundaries, but we will also present a potential best-case limit for the technique. •We show feasibility to have very low dislocation densities without a Dash neck.•Abrupt increases in minority carrier lifetime are seen at start of crystal growth.•Measured lifetime in excess of 1ms on p-type directionally solidified material.