Aleš Heyberger

Mathematical model of extraction of citric acid with amine

Abstract Various mathematical models of amine extraction of tribasic weak acid were developed and compared with equilibrium data on the distribution of citric acid between water and solution of trialkylamine in MIBK. It has been found that the two suggested mechanisms of extraction (distribution of non-dissociated portion of acid × distribution of dissociated portion of acid) are equivalent, as far as the fit of the model is concerned the only difference being in the values of equilibrium constants. Two amine : acid complexes in the organic phase have been assumed (1 : 1, 2 : 1), their hydration numbers being 4 and 1, respectively. The non-ideality of the aqueous phase was neglected in the concentration range of interest. The non-ideality of the organic phase was expressed both in terms of aggregation and hydration (exponential correction using hydration numbers). The best fit was obtained using a combination of the aggregation and hydration corrections.

Extraction of citric acid with tertiary amine—third-phase formation

Abstract The conditions of third-phase formation in the system made up of aqueous solution of citric acid-solution of trialkylamine in a binary diluent were studied. As binary diluents mixtures of octanol-1 or chloroform with n -heptane were used. The experiments were carried out in water-free systems, at 25°C under the limiting condition of amine fully loaded with acid, thus simulating extraction form highly concentrated aqueous phase. The data could be presented as pseudo-ternary equilibria and were correlated using the extended method, devised by Rod ( Chem. Engng J. 1976, pp. 105–110).

Extraction Equilibrium of Dicarboxylic Acids with Tertiary Amine in Single and Binary Diluents

Abstract The extraction equilibria of aliphatic dicarboxylic acids with solutions of trialkylamine (TAA) in binary diluents were measured and compared. Oxalic, succinic, DL‐malic and D‐tartaric acids were used. Trialkylamine with C7–C9 carbon atom chains was dissolved in 1‐octanol, methyl‐isobutyl ketone (MIBK), and chloroform, or in their binary mixtures with n‐heptane. Physical extraction with pure active diluents also was measured. The contribution of physical extraction was significant only in the case of succinic and oxalic acids and with octanol‐1 and MIBK as diluents. In all systems, formation of the (1:1) acid:amine complex was found dominant at low aqueous‐phase equilibrium acid concentrations. The values of the respective equilibrium constants grew in the same order as the first dissociation constants of these acids. In the region of high acid concentrations, overloading of amine was observed, indicating formation of (2:1) and (3:1) complexes. Here, the extraction grew with increasing lyophilicity of the acids. Various modifications of a mathematical model, incorporating reactions of formation of these complexes and physical extraction, were used for correlating the equilibrium data.

Correlation of ternary liquid-liquid equilibria in system isobutyl acetate-acetic acid-water

A method of correlating ternary liquid-liquid equilibrium data based on the model presented by Rod (1976, Chem. Engng J.11, 105–110) is discussed and compared with other methods available in the literature. Several variants of the method are defined and corresponding computation algorithms are described. Their properties are tested on equilibrium data for the system isobutyl acetate-acetic acid-water at 25°C presented by Heyberger et al. (1977, Collection Czech. Chem. Commun.42, 3355–3362). The accuracy of the individual variants and the stability of the respective algorithms are demonstrated. A procedure for stabilizing the optimization and finding a unique solution is developed. The present correlation procedure compares favourably with other published methods.

Parameter evaluation of an extractor with back mixing

Abstract From measured longitudinal solute concentration profiles in a vibrating plate extraction column parameters of a stagewise model with back mixing were evaluated as functions of the axial coordinate of the column. In the model the parameters were represented by polynomials of the stage number and the dependence of model error on the degrees of the polynomials was studied. Very good agreement of measured and calculated concentration profiles was obtained with polynomials of not higher than 4th degree, whereas the model with constant parameters in most cases failed to describe adequately the profiles and the exit concentrations. As a possible cause of the variability of parameters along the column the existence of hold-up profiles is sugges