Single-drop based modelling of solvent extraction in high-viscosity systems

• Modellierung von solventer Extraktion in hoch viskosen Systemen basierend auf Einzeltropfenexperimenten

Adinata, Donni; Pfennig, Andreas (Thesis advisor)

Aachen : Publikationsserver der RWTH Aachen University (2011)
Dissertation / PhD Thesis

Aachen, Techn. Hochsch., Diss., 2011

Abstract

The knowledge in the area of solvent extraction is limited for high-viscosity systems while the economic importance of such systems has always been known, because the chemical industry concerns chemicals processed at high viscosities such as extraction of pharmaceuticals from fermentation broth. In the future raw materials for chemicals will increasingly stem from renewable biological origin. These chemicals have often higher viscosities than conventional ones. Extraction may be more suitable separation process for process as compared to distillation. In this work the influence of high viscosity on extraction was researched. The extraction column behaviour for high-viscosity systems predicted and simulated by using the ReDrop programme. The test system used to investigate the extraction in high viscosity systems was polyethylene glycol (PEG) with water. PEG 4000 with water and PEG 600 with water respectively were used as continuous phase. Toluene and toluene with paraffin were used as disperse phase. Acetone was used as mass transfer component. The sedimentation velocity and the mass transfer of single drops in high-viscosity systems were investigated with laboratory scale experiments. Henschke’s model (2003) was validated with the experimental results. To increase the accuracy of the drop sedimentation velocity, Henschke’s model was modified by using a factor of 4 and = 5. It was even possible to obtain the general parameters, by fitting the experimental data with the results of Henschke’s model. For toluene the parameters are = 7.55 mm, = 0.085, and = 0.79, as well as for toluene with paraffin = 2.95 mm, = 0.085, and = 1.24. The sedimentation velocity and mass transfer of single drops is strongly influenced by the viscosity of the continuous phase. The effect of viscosity on the disperse phase is observable, but less significant. The viscosity of the continuous and the disperse phase affect the mass-transfer rate. The influence of viscosity in the continuous phase on the mass-transfer rate is considerably higher. The effect of viscosity in the disperse phase is observable, but less significant. Additionally, the diffusion coefficient has only a small influence on the error. The model of Wilke Chang may be applied to estimate in high viscosity systems. The behaviour of a pilot-plant scale pulsed sieve-tray extraction column was investigated. The extraction column in high viscosity systems could be better operated with a sieve-tray hole-diameter of 8 mm than with a diameter of 2 mm. The flow rates are by a factor of 20 lower as compared to low-viscosity systems. The separation performance with high viscosity system is about 4 times lower compared to low viscosity systems. The experimental data gained at the pulsed sieve-tray extraction column were compared with the simulation results of the ReDrop program. The extraction column behaviour in high-viscosity systems can be predicted well by using ReDrop program. A parameter study showed, that in contrast to, has a very strong influence on the number of theoretical stages Nth,sim. Different models determine values for Dax,c that differ by factor of 100, thus, reliable results for the mass-transfer performance in ReDrop are only able when employing sound models for Dax,c. The model of Henschke (2003) for the maximum stable diameter cannot be applied to liquid-liquid systems with a highly viscous continuous phase; therefore, the model for maximum stable diameter must be adjusted. In order to do so, more experimental effort should be put into understanding drop breakage in the presence of a high viscosity continuous phase. It was shown that in principle ReDrop may be used to simulate extraction columns with a high viscosity continuous phase. The mean relative deviation between experimentally found hydrodynamic parameters, such as hold-up and Sauter mean diameter is about 20%.