Combining experiments and CFD-based compartment models for the scale-up of counter-current solid-liquid extraction

Abstract

Continuous counter-current solid-liquid extraction is a promising approach for the recovery of natural products, particularly when target compounds are present at low concentrations in the solid phase. However, reliable process design and scale-up remain challenging due to complex hydrodynamics and limitations of simplified modeling approaches. In this work, a combined experimental, CFD-based (using Computational Fluid Dynamics), and compartment modeling strategy is developed for an extruder-based extraction process for the recovery of artemisinin from Artemisia annua L. Residence time experiments performed with water in a transparent laboratory extruder are used to characterize the hydrodynamic behavior and to provide input data for simplified CFD-simulations without mass transfer. From these simulations, phase volumes and inter-compartment flows are derived and used to parameterize a compartment model. By incorporating mass transfer between extract and raffinate, the model reproduces the residence time distribution of the extract phase with excellent agreement, although these data were not used for parameterization. The approach is subsequently transferred to a toluene-based extraction system and scaled up to a larger apparatus by numbering-up the compartments. While a previously proposed short-cut model reproduces experimental extraction yields with high accuracy, it relies on simplified assumptions. The CFD-informed compartment model provides a more physically consistent description, resulting in slightly more conservative predictions but correctly capturing trends across operating conditions, with deviations below 19%. The results demonstrate that combining experiments, CFD-simulations, and compartment modeling enables a robust and physically interpretable framework for the design and scale-up of counter-current extraction processes.