PFI:AIR - TT: Cost-Effective Membrane-Based Green Electrolytic Process for Solar and Semiconductor Grade Silicon Production

Trustees Of Boston University, Boston

Investigators

Abstract

This PFI: AIR Technology Translation project focuses on translating technology that will change the state-of-the-art silicon production process from an energy-intensive and environmentally detrimental one into a cost-effective green process. The new process is an order of magnitude better in energy cost than current practices, and also emits well below half the carbon dioxide (CO2) of the most efficient existing metallurgical processes. The goal of the project is to generate the necessary process data to evaluate scalability and cost-effectiveness of this solid-oxide-membrane-based green electrolytic process for semiconductor and solar grade silicon production. Upon successful implementation, this process will demonstrate several advantages over current processes. They include: very low energy usage relative to free energy required for silicon dioxide (SiO2) reduction; use of inexpensive raw materials that require little to no pre-treatment; no carbothermic reduction, which emits 10 kg of CO2 per kg silicon (Si) product and whose contaminants typically reduce purity of silicon from 99.6% in natural quartzite to 97-98% in metallurgical grade (MG) Si; possibility of inherent boron removal by borium triflouride (BF3) volatilization; absence of any carbon or chlorine in the process; and there are no anode effects resulting in perfluro and/or perchloro carbon emissions. Current methods for Si production include, fluidized bed reduction processes, carbothermic reduction of high-purity silica, slag/crystal refining, liquid Si electrorefining, and electrolysis of chlorides and fluorides. Some of the major limitations of these processes include extensive raw materials pre-processing, low yields, detrimental environmental impact and substantial energy requirements. In the proposed process, a one-end-closed oxygen-ion-conducting stabilized zirconia (SOM) tube will be used to separate pure silica (SiO2) dissolved in molten flux from an inert anode placed inside the SOM tube. To ensure product purity, a pre-reduction step using a secondary cathode at lower applied potentials will be employed to remove impurities that are more electronegative than Si. The impurity-laden secondary cathode will be removed, and then employing a liquid tin cathode the applied potential will be increased to reduce silica. The Si reduced will go into solution in the liquid tin cathode. Less electronegative impurity ions compared to Si will remain in the flux. Thus impurity oxides of both more and less electronegative impurities are not reduced along with silica. Si is over 95 atom% soluble in liquid tin at high temperatures but at lower temperatures pure Si and Sn are immiscible. This will allow directional solidification to be employed after electrolysis to produce high-purity Si ingots and demonstrate this as a cost-effective carbon-free method for mass production of Si from commercially available sources of silica. This project will provide research opportunities for graduate and undergraduate students to work with our industrial partners and move the technology closer towards commercialization. It will also provide a rich set of case-study materials for introduction into both undergraduate and graduate classroom teaching. Infinium, a clean metals company, and SunEdison Semiconductors, consumer of semiconductor grade silicon will be engaged in the research program to assess quality, scalability and cost-effectiveness of the green technology for mass production of silicon starting from commercially available sources of silica.