Study on polyvinyl butyral purification process based on Box-Behnken design and artificial neural network

Abstract

Polyvinyl butyral is a polymer resin with excellent mechanical properties, weather resistance and adhesive properties, which can be used in solar cell encapsulation films and automotive glass interlayer films. It is synthesized by the reaction between polyvinyl alcohol and n-butyraldehyde acetal, in which a certain amount of surfactant is added. Because it will cause the product to turn yellow, the surfactant needs to be removed by purification process. In this paper, polyvinyl butyral was synthesized using sodium dodecyl sulfonate as surfactant, and the purification process and the effect of surfactant residue in polyvinyl butyral on its yellowing degree were studied. Box-Behnken design and artificial neural network were used to optimize the purification process. The results showed that the influencing factors of the purification process were as follows: washing times>addition amount of sodium dodecyl sulfonate>washing temperature>stirring time. The correlation coefficients (R2) of Box-Behnken design and artificial neural network were 0.9943 and 0.9981 respectively, indicating that the artificial neural network was more accurate than Box-Behnken design. In conclusion, the results of this study provide an effective method for the purification process of polyvinyl butyral.