Reactive separation of gallic acid using trioctylamine in oleyl alcohol and dodecane mixtures: Equilibrium insights, modeling and optimization using RSM and ANN approaches

The aim of present study is to model and optimize gallic acid (GA) reactive extraction from aqueous solution with trioctylamine (TOA) dissolved in the biocompatible mixture of oleyl alcohol, a modifier and dodecane, an inert diluent using a statistical modeling approach, response surface methodology (RSM) with rotatable central composite design, and artificial neural network (ANN). The optimal values of input parameters for maximum extraction (86%) of GA are found as follows: (i) initial GA concentration ( ${\bar{C}}_{\mathrm{A},\mathrm{o}}$ = 0.387 mol/L); (ii) temperature, T = 26.5°C; (iii) initial TOA concentration ( $C_{\mathrm{H}\mathrm{G},\mathrm{o}}$ = 0.049 mol/L); (iv) modifier volume, V = 70% at 101.325 kPa pressure. Predicted responses in terms of extraction degree using RSM and ANN are found to be very close to experimental values. The overall apparent equilibrium constants (K_E = 37.24 and 22.38) and stoichiometry of reaction between acid and extractant (m =1.25 and 1.16) are determined considering negligible and 50% physical extraction, respectively. Individual equilibrium constants ( $K_{11}$ and $K_{21}$) for 1:1 and 2:1 GA:TOA complexes and their concentrations, C₁₁ and C₂₁, respectively, are calculated. Differential evolution approach, a bio-inspired algorithm, is used to investigate the insights of extraction mechanism at the optimum conditions.