Innovative separation model for boron removal from silicon during slag refining based on ion and molecule coexistence theory

How to reduce boron concentration to an extremely low level has become a major obstacle during silicon waste recycling. As a simple and efficient method of removing impurities, slag refining plays an increasingly important role for industrial silicon waste recovery. The deboration ability of slag is usually evaluated by boron distribution ratio L B . However, the development of a separation prediction model is challenged by the limitations on the recognition of impurity structure in slag and complexity of slag-silicon reactions. Herein, a novel thermodynamic model for predicting boron distribution ratio between CaO-SiO 2 based slags and molten silicon has been developed based on ion and molecule coexistence theory (IMCT), i.e., IMCT- L B model. The reaction abilities of structural units are represented by the calculated mass action concentration. The predictive boron distribution ratio between CaO-SiO 2 , CaO-Al 2 O 3 -SiO 2 , CaF 2 -CaO-SiO 2 slags and silicon correspond well with experimental data, reflecting the reasonability of IMCT- L B model. According to IMCT, deboration reactions between slags and silicon can be represented by ion couples (Ca 2+ + O 2– ), simple molecule Al 2 O 3 with SiO 2 to form eight deboration products as B 2 O 3 , 3CaO · B 2 O 3 , 2CaO · B 2 O 3 , CaO · B 2 O 3 , CaO · 2B 2 O 3 , 2Al 2 O 3 · B 2 O 3 , 9Al 2 O 3 · B 2 O 3 , CaO · 2SiO 2 · B 2 O 3 . With the developed IMCT- L B model, the respective boron distribution ratio L B , i of the eight deboration products can be determined quantitatively. Furthermore, the effect of temperature, mass ratio of CaO to SiO 2 (%CaO/%SiO 2 ), Al 2 O 3 and CaF 2 content of CaO-SiO 2 based slags on L B are fully discussed.