Engineering topics

Rea, S. N.

Publications and source records attributed to Rea, S. N..

4 records

NASA NTRS1979

LSA Large Area Silicon Sheet Task Continuous Czochralski Process Development

A commercial Czochralski crystal growing furnace was converted to a continuous growth facility by installation of a small, in-situ premelter with attendant silicon storage and transport mechanisms. Using a vertical, cylindrical graphite heater containing a small fused quartz test tube linear from which the molten silicon flowed out the bottom, approximately 83 cm of nominal 5 cm diamter crystal was grown with continuous melt addition furnished by the test tube premelter. High perfection crystal was not obtained, however, due primarily to particulate contamination of the melt. A major contributor to the particulate problem was severe silicon oxide buildup on the premelter which would ultimately drop into the primary melt. Elimination of this oxide buildup will require extensive study and experimentation and the ultimate success of continuous Czochralski depends on a successful solution to this problem. Economically, the continuous Czochralski meets near-term cost goals for silicon sheet material.

Rea, S. N.↗

NASA NTRS1978

LSSA large area silicon sheet task continuous Czochralski process development

A Czochralski crystal growing furnace was converted to a continuous growth facility by installation of a premelter to provide molten silicon flow into the primary crucible. The basic furnace is operational and several trial crystals were grown in the batch mode. Numerous premelter configurations were tested both in laboratory-scale equipment as well as in the actual furnace. The best arrangement tested to date is a vertical, cylindrical graphite heater containing small fused silicon test tube liner in which the incoming silicon is melted and flows into the primary crucible. Economic modeling of the continuous Czochralski process indicates that for 10 cm diameter crystal, 100 kg furnace runs of four or five crystals each are near-optimal. Costs tend to asymptote at the 100 kg level so little additional cost improvement occurs at larger runs. For these conditions, crystal cost in equivalent wafer area of around $20/sq m exclusive of polysilicon and slicing was obtained.

Rea, S. N.↗

NASA NTRS1977

Large area Czochralski silicon

The overall cost effectiveness of the Czochralski process for producing large-area silicon was determined. The feasibility of growing several 12 cm diameter crystals sequentially at 12 cm/h during a furnace run and the subsequent slicing of the ingot using a multiblade slurry saw were investigated. The goal of the wafering process was a slice thickness of 0.25 mm with minimal kerf. A slice + kerf of 0.56 mm was achieved on 12 cm crystal using both 400 grit B4C and SiC abrasive slurries. Crystal growth experiments were performed at 12 cm diameter in a commercially available puller with both 10 and 12 kg melts. Several modifications to the puller hoz zone were required to achieve stable crystal growth over the entire crystal length and to prevent crystallinity loss a few centimeters down the crystal. The maximum practical growth rate for 12 cm crystal in this puller design was 10 cm/h, with 12 to 14 cm/h being the absolute maximum range at which melt freeze occurred.

Rea, S. N.↗

NASA NTRS1976

Large area Czochralski silicon for solar cells

A detailed model of a typical Czochralski silicon crystal puller is utilized to predict maximum crystal growth rate as a function of various furnace parameters. Results of this analysis, when combined with multiblade slurry sawing, indicate that the Czochralski process is highly attractive for achieving near-term cost reduction of solar cell silicon.

Rea, S. N.↗