A Comparative Study on Water and Ethanol Separation Mixture by Different Solvents: New Experimental and Correlation Data

Bacha, Oussama; Baka, Ouidad; Hasseine, Abedelmalek; Korichi, Mourad; Selloum, Djamel; Henni, Abdellah

doi:10.22036/pcr.2021.274620.1890

A Comparative Study on Water and Ethanol Separation Mixture by Different Solvents: New Experimental and Correlation Data

Document Type : Regular Article

Authors

Oussama Bacha ¹

Ouidad Baka ¹

Abedelmalek Hasseine ²

Mourad Korichi ²

Djamel Selloum ¹

Abdellah Henni ¹

¹ Laboratoire de Dynamique, Interactions et Réactivité des Systèmes, Faculty des Sciences Appliquées, Université Kasdi Merbah Ouargla, BP 511 Ouargla, Algeria

² Laboratoire de Recherche en Génie Civil, Hydraulique, Développement Durable et Environnement, Université Mohamed Khider, Biskra, Algeria

https://doi.org/10.22036/pcr.2021.274620.1890

Abstract

This paper describes our interest on the ethanol separation and the enhancement of the obtained interaction parameters. The binodal curves and tie-lines data for the ternary systems (Water +Ethanol + 1-Octanol), (Water + Ethanol +4-Methyl-2-pentanone) and (Water + Ethanol + Ethyl Acetate) at 293.15K were investigated. These ternary systems follow a type-1 Liquid-Liquid equilibrium diagram. Furthermore, the extraction capabilities of these solvents in this study were tested using the separation factor (S) which was found greater than 1 in all ternary systems (S varying between 3.52 and 18.37). Besides, the values of the distribution coefficient (D) of ethanol were also not constant over the entire composition of the two-phase region (D values varied between 0.777 and 0.928), which confirms the extraction of ethanol. The ternary system (Water + Ethanol + Ethyl Acetate) was found to be the best option in terms of the distribution coefficient and separation factor. Othmer-Tobias and Hand plots have properly tested the reliability of the experimental tie-line data. In addition, non-random two-liquid model was used to correlate the experimental data, the model parameters were obtained using genetic algorithm, and there was a strong agreement between the experimental and calculated tie-lines with RMSD values below 1.72%.

Graphical Abstract