Mechanism, influencing factors exploration and modelling on the reactive extraction of 2-ketogluconic acid in presence of a phase modifier
•Analysis of 2-ketogluconic acid reactive extraction with Amberlite LA2 and octanol.•Correlation of extraction mechanism and influencing factors in three solvents.•Complex stoichiometry 2-ketogluconic acid:extractant was 1:1 regardless of the solvent.•1-Octanol addition in all solvents improved extr...
Gespeichert in:
Veröffentlicht in: | Separation and purification technology 2021-01, Vol.255, p.117740, Article 117740 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | •Analysis of 2-ketogluconic acid reactive extraction with Amberlite LA2 and octanol.•Correlation of extraction mechanism and influencing factors in three solvents.•Complex stoichiometry 2-ketogluconic acid:extractant was 1:1 regardless of the solvent.•1-Octanol addition in all solvents improved extraction yield.
Taking into account that limited research has been carried out on the reactive extraction of 2-ketogluconic acid, this study was focused on analyzing the pH dependent extraction performance and the molar ratios of acid and extractant (Amberlite LA-2) dissolved in three solvents with 1-octanol as phase modifier. Back extraction was successfully performed using NaOH solutions. The mechanism of the interfacial reaction in the presence of 1-octanol, pointed out that, indifferent of the pH value and solvent polarity, only one molecule of 2-ketogluconic acid and one of extractant react at the interface. The positive effect of 1-octanol on extraction efficiency was quantified by means of the amplification factor, its maximum values being 2.43 for dichloromethane, 3.67 for butyl acetate and 3.64 for n-heptane. In addition, the process was modelled using statistical regression and Artificial Neural Networks (ANNs) determined with chaos based Differential Evolution algorithm. The best ANN model had a mean squared error for the testing phase of 0.19 and modeled the process with an acceptable error. |
---|---|
ISSN: | 1383-5866 1873-3794 |
DOI: | 10.1016/j.seppur.2020.117740 |