Optimization of fluoride removal by induced crystallization based on RSM methodology

In order to investigate the fluoride removal effect of calcium fluorophosphate induced crystallization, Box-Behnken design (BBD) was used for calcium fluorophosphate induced crystallization fluoride removal experiments to develop a quadratic regression model with the fluoride ion removal rate as the response value, and to compare the significance of the effects of three parameters〔n(Ca)/n(F), n(P)/n(F), crystalline seed dosage〕 on the fluoride ion removal rate by analysis of variance (ANOVA). The process parameters of calcium fluorophosphate induced crystallization for fluoride removal were optimized by RSM, and the model was validated by the experimental results under the optimal process condition. The optimal condition optimized by RSM were n(Ca)/n(F)=15, n(P)/n(F)=8.918, and the amount of crystalline seed dosage as 9.208 g/L, resulting in the theoretical optimal fluoride removal rate of 97.26%. Compared with the actual fluoride removal rate of 95.07%, the relative error was 2.3%. SEM showed that the surface morphology changed greatly before and after the reaction of the crystalline species, from regular crystalline particles to a layer of spherical clustered crystals completely covered. XRD revealed that the induced crystallization produced CaF2, Ca10(PO4)6(OH)2, Ca10(PO4) 6F2, and CaHPO4, indicating that there were multiple crystallization reaction pathways for the removal of fluorine by induced crystallization of calcium fluorophosphate.