Particle swarm optimization–artificial neural network modeling and optimization of leachable zinc from flour samples by miniaturized homogenous liquid–liquid microextraction

• Zinc was extracted from flour sample using homogenous liquid–liquid microextraction.
• HLLME is a green extraction technique.
• This extraction method is simple, efficient and fast.
• PSO–ANN model was investigated to predict the extraction yield.
• Determination of the zinc in flour samples was successfully performed.

In this study, a new modeling method based on a particle swarm optimization (PSO)–three-layer artificial neural network (ANN) techniques has been employed to develop the ANN–PSO system for simulation and optimization of miniaturized homogenous liquid–liquid microextraction (HLLME) process for the extraction of zinc from flour samples and determination by atomic absorption spectrometry (FAAS). Morin was used as complexing ligand. Input variables of the model were pH of the solution, volume of morin, ultrasonic time and extracting solvent volume. After training using a back-propagation algorithm, the ANN model was able to predict the extraction efficiency of zinc ions. A tangent sigmoid transfer function (tansig) at hidden layer with 11 neurons and a linear transfer function (purelin) at output layer were used in the ANN model. Excellent linear regression was observed between the experimental data and the ANN predictions, showing a correlation coefficient (R2) of 0.9497. Using PSO method, the optimum operating conditions were determined. Under the optimum conditions, the detection limit (LOD) of the proposed procedure was calculated to be 0.8 ng g−1 with a relative standard deviation (RSD%) better than 3.8% (n = 10). The method was successfully applied to the separation, pre-concentration and determination of zinc in flour samples.