Silicon Production via Aluminothermic Reduction of Calcium Silicate Slags

Abstract

Silicon is in high demand around the world, thanks to its essential role in everything from solar cells to silicones. But producing silicon is no small feat—it consumes a lot of energy and produces a significant amount of emissions. At the same time, both the silicon and aluminum industries are looking for better ways to make use of by-products that are often wasted. This is particularly important in Europe, where supply chains for raw materials, including silicon, are becoming increasingly vulnerable due to shifts in global trade and rising geopolitical tensions. Enter the SisAl project, where one PhD student is tackling these challenges head-on. The research explores a novel method for producing silicon by reacting aluminum by-products with calcium silicate slags (CaO-SiO2), which are common in various industries. What's exciting about this process is that it doesn't produce direct CO2 emissions, and it only uses about one-third of the energy compared to traditional methods. The process has been successfully demonstrated on both lab and pilot scales at NTNU and Elkem in Kristiansand, respectively. It proved stable and adaptable, handling different materials and combinations, and making good use of by-products that usually go to waste. One of the key ingredients in this process is a troublesome by-product called dross, which Europe produces around 100,000 tons of each year. Remarkably, the process was able to produce silicon that matches the quality of silicon currently on the market, which is typically made using carbon as a raw material. But that's not all—the process also generates a slag that contains valuable Al2O3, which can be purified and used in making LEDs or lithium-ion batteries, a topic being explored by other project partners. These findings offer new insights for industrial manufacturers, helping them to design processes that optimize raw material usage, recycle slag internally, and produce specific product compositions. The study also deepens our understanding of the phenomena and kinetics behind the metallothermic reduction of molten slags. The research shows that this process is not only scalable but also potentially circular—meaning it could be largely self-sustaining—as long as temperature, aluminum, and by-product levels, as well as impurities, are carefully controlled for the intended use of the silicon. This work lays the groundwork for further process development that will be crucial for resource optimization and scaling up to an industrial level. The results are especially relevant for the silicon and aluminum industries, opening the door to a new potential symbiosis and collaboration between them. Future work is likely to focus on optimizing raw material flows, considering material costs, energy use, and CO2 emissions, while tailoring the process to produce specific end-use products like aluminum alloys or electronic-grade silicon.