Liquid-liquid extraction in a batch agitated vessel with a non-Newtonian dispersed phase

Author

Charlotte M. Farmer

Date of Award

12-1996

Degree Type

Thesis

Degree Name

Master of Science

Major

Chemical Engineering

Major Professor

Trueman D. Parish

Committee Members

Donald C. Bogue, George C. Frazier

Abstract

Chemical industry utilizes solvent extraction as an alternative to conventional distillation, specifically for the separation of liquid mixtures with unfavorable vapor-liquid equilibria relationships and/or thermally unstable components. Common to most extraction applications (i.e., mixer settler, multistage tower extractors, rotating disc contactors, etc.) is the transfer of material taking place to or from drops of one liquid dispersed in another. Thus, a more complete knowledge of the factors affecting transfer to and from drops is essential to further understanding extraction and the development of improved design principles. An extensive literature search indicates that much work has been done over the years to develop mass transfer coefficient models for both continuous-phase (k_c) and Dispersed-phase (k_d) controlling systems. However, all of the liquid-liquid systems studied to date have included Newtonian materials with aqueous continuos phases. Industrial streams tend to be solutions containing several components with a range of Newtonian to non-Newtonian flow characteristics; thus rendering them significantly different from systems featured in mass transfer studies to date. This study is unique in that it centers on purification of a non-Newtonian industrial stream containing three major components (cellulose acetate, acetic acid, water) and a solute (sodium sulfate). Before dispersing in a predominantly aqueous solution, a solvent (isopropyl acetate) was added to the stream to prevent precipitation of the cellulose acetate (this solution is later referred to as solvated dope). The composition of the contacting solution was determined experimentally. To this end, a quaternary phase diagram was constructed for cellulose acetate, acetic acid, water, and isopropyl acetate to identify the feasible operating region where liquid-liquid phases exist without formation of a solid cellulose acetate phase. A key requirement was to operate on a tie line, in the phase diagram described above, to deter transfer of nonsolutes across phase boundaries. This research measured the dispersed-phase mass transfer coefficient, and the corresponding drop sizes in the system described above where the dispersed phase, solvated dope, was a viscous fluid. This was achieved by measuring solute concentration in the continuous-phase over time analytically with discrete samples. A mass balance calculation on the system was used to calculate the disappearance of solute from the dispersed-phase. In turn, the solute concentration data was used to estimate the dispersed-phase mass transfer coefficient. The resulting correlation for dispersed-phase mass transfer was compared with expressions derived in the literature (for systems similar to the basic assumptions of this study). In conclusion: 1. Both the Penetration Theory (Higbie, 1935) and the Skelland and HuXien Model (1990) underestimate the dispersed-phase mass transfer coefficient in this kind of system. Therefore these expressions will yield a conservative extractor design. 2. The difference between the measured value of (k_c) and that estimated from Penetration Theory could be accounted for by errors in the physical and transport property estimations especially diffusivity, D_AB which is difficult to estimate for the liquid-liquid system described herein. 3. Experimentally estimated drop sizes generally agree with the Hong and Lee (1985) Sauter-mean diameter model.

Recommended Citation

Farmer, Charlotte M., "Liquid-liquid extraction in a batch agitated vessel with a non-Newtonian dispersed phase. " Master's Thesis, University of Tennessee, 1996.
https://trace.tennessee.edu/utk_gradthes/10830