Boron removal from metallurgical grade silicon by slag refining based on GA-BP neural network

• Published in: Rare metals Vol. 40; no. 1; pp. 237 - 242
• Main Authors: Yuan, Shi-Lai, Lu, Hui-Min, Wang, Pan-Pan, Tian, Chen-Guang, Gao, Zhi-Jiang
• Format: Journal Article
• Language: English
• Published: Beijing Nonferrous Metals Society of China 01.01.2021; Springer Nature B.V
• Subjects: Back propagation networks; Basicity; Biomaterials; Boron; Boron removal; Calcium oxide; Chemistry and Materials Science; Composition; Cryolite; Energy; Genetic algorithms; Genetic algorithm‐back propagation neural network; Materials Engineering; Materials Science; Metallic Materials; Metallurgical grade silicon; Metallurgy; Nanoscale Science and Technology; Neural networks; Oxidation; Physical Chemistry; Predictions; Silicon; Silicon dioxide; Slag; Slag system; Genetic algorithm-back propagation neural network; Metallurgical grade silicon; Slag system; Boron removal
• ISSN: 1001-0521; 1867-7185
• DOI: 10.1007/s12598-014-0373-5

In order to investigate the boron removal effect in slag refining process, intermediate frequency furnace was used to purify boron in SiO 2 -CaO-Na 3 AlF 6 -CaSiO 3 slag system at 1,550 °C, and back propagation (BP) neural network was used to model the relationship between slag compositions and boron content in SiO 2 -CaO-Na 3 AlF 6 -CaSiO 3 slag system. The BP neural network predicted error is below 2.38 %. The prediction results show that the slag composition has a significant influence on boron removal. Increasing the basicity of slag by adding CaO or Na 3 AlF 6 to CaSiO 3 -based slag could contribute to the boron removal, and the addition of Na 3 AlF 6 has a better removal effect in comparison with the addition of CaO. The oxidizing characteristic of CaSiO 3 results in the ineffective removal with the addition of SiO 2 . The increase of oxygen potential ( p O 2 ) in the CaO-Na 3 AlF 6 -CaSiO 3 slag system by varying the SiO 2 proportion can also contribute to the boron removal in silicon ingot. The best slag composition to remove boron was predicted by BP neural network using genetic algorithm (GA). The predicted results show that the mass fraction of boron in silicon reduces from 14.0000 × 10 −6 to 0.4366 × 10 −6 after slag melting using 23.12 % SiO 2 -10.44 % CaO-16.83 % Na 3 AlF 6 -49.61 % CaSiO 3 slag system, close to the experimental boron content in silicon which is below 0.5 × 10 −6 .