(254c) The Horizontal Ribbon Growth of Solar Silicon Crystals: Process Analysis, Stability, and Control | AIChE

(254c) The Horizontal Ribbon Growth of Solar Silicon Crystals: Process Analysis, Stability, and Control


Daggolu, P. - Presenter, University of Minnesota
Yeckel, A. - Presenter, University of Minnesota
Derby, J. J. - Presenter, University of Minnesota
Bleil, C. - Presenter, Ribbon Technology International

Present-day photovoltaic technology is predominantly based on the fabrication of solar cells from silicon. The cost and efficiency of silicon-based solar cells depend primarily on the growth rate and degree of single-crystallinity attained by the growth process. Traditional silicon crystal growth methods are either very expensive or compromise on efficiency. The horizontal ribbon growth (HRG) process is a promising crystal growth technology that has the potential to overcome these limitations. William Shockley first conceived of the HRG process in late 1950's, with subsequent efforts by Kudo in Japan in the late 1970's and by Energy Materials Corporation in the US in the early 1980's. After encouraging initial development, these efforts stalled owing to various technical difficulties.

We are applying a comprehensive thermal-capillary model to study the coupled phenomena of heat transfer and interfacial phenomena (solidification and capillarity) in the HRG process. This model accounts for heat transfer in the melt-crystal-crucible domains with radiative heat loss from high temperature surfaces, melt convection due to buoyancy and surface-tension forces, and the self-consistent determination of melt-crystal, melt-ambient and crystal-ambient interface shapes. This moving-boundary problem is solved numerically by the Galerkin finite element method, using elliptic mesh generation.

Extensive parametric sensitivity studies are conducted to identify the variables that impact the process. The effect of pull rate, pulling angle, melt height, crucible geometry and furnace heat transfer are investigated. Transient simulations are performed to analyze the dynamics of the growth process and to determine its stability. Preliminary results indicate that quasi-steady-states exist for this system but that most are unstable.  In this light, the dynamic analysis of this system will be require to identify attainable stable states and possible strategies to stabilize unstable ones.