Modeling of microwave-assisted extraction of natural dye from seeds of Bixa orellana (Annatto) using response surface methodology (RSM) and artificial neural network (ANN)

With ever increasing demand for eco-friendly, non-toxic colorants, dyes derived from natural sources have emerged as a potential alternative to relatively toxic synthetic dyes. In the present work, microwave-assisted extraction of yellow-red natural dye from seeds of Bixa orellana (Annatto) was studied. Response surface methodology (RSM) and artificial neural network (ANN) were used to develop predictive models for simulation and optimization of the dye extraction process. The influence of process parameters (such as pH, extraction time and amount of Annatto seeds used in extraction) on the extraction efficiency were investigated through a two level three factor (23) full factorial central composite design (CCD) with the help of Design Expert Version 7.1.6 (Stat Ease, USA). The same design was also used to obtain a training set for ANN. Finally, both the modeling methodologies (RSM and ANN) were statistically compared by the coefficient of determination (R2), root mean square error (RMSE) and absolute average deviation (AAD) based on the validation data set. Results suggest that ANN has better prediction performance as compared to RSM.

► Microwave assisted extraction of natural colorant from seeds of Bixa orellana. ► RSM and ANN models used for simulation and optimization of the extraction process. ► Both RSM and ANN models provided good correlation to the experimental data. ► Statistical comparison showed that ANN has better prediction performance than ANN.