Daggolu, Parthiv2014-02-052014-02-052013-12https://hdl.handle.net/11299/162449University of Minnesota Ph.D. dissertation. December 2013. Major: Chemical Engineering. Advisor: Jeffrey Jay Derby. 1 computer file (PDF); x, 131 pages.Horizontal ribbon growth (HRG) promises the growth of crystalline silicon at rates that are orders of magnitude greater than vertical ribbon growth technologies. If successful, this process would enable the production of higher-quality, near-single-crystalline silicon wafers at fraction of the cost of current production techniques. This fascinating process was first conceived by Shockley in late 1950's for silicon growth and was practiced by Bleil in the late 1960's for germanium growth. Large-scale development efforts were sub- sequently carried out by Kudo in Japan in the late 1970's and by the Energy Materials Corporation in the US in the early 1980's. However, after encouraging early results, experimental advances and process development efforts stalled, and this technique was abandoned in favor of growth methods that were easier to develop.Unlike vertical meniscus-defined crystal growth processes, such as edge-defined film- fed growth (EFG), which are inherently stable, there are many failure modes that must be avoided in the HRG process. We argue that its successful operation will rely on a thorough understanding of system design and control-issues that are perhaps only feasibly addressed via computational modeling of the system. Towards these ends, we present a comprehensive thermal-capillary model based on finite-element methods to study the coupled phenomena of heat transfer, fluid mechanics and interfacial phenom- ena (solidification and capillarity) in the HRG process. Bifurcation analysis coupled with transient computations using this model reveals process limitations that manifest as failure mechanisms, such as bridging of crystal onto crucible, spilling of melt from the crucible, and undercooling of melt at the ribbon tip, that are consistent with prior experimental observations and suggests operating windows that may allow for stable process operation. Further, coupled impurity transport calculations reveal interesting and potentially beneficial redistribution mechanisms at the solidification interface that lead to an inherent purification of the majority of the growing crystal ribbon.en-USComputer SimulationEdge defined film fed growthFluid flowsHeat transferSemiconducting siliconSolar cellsThesis or Dissertation