Welcome on the ECCE-6 CDROM.

Conference logo

European Congress of Chemical Engineering - 6
Copenhagen 16-21 September 2007

Abstract 815 - Transport phenomena inside liquid chromatographic columns and CFD modelling based on X-ray CT measurements

Transport phenomena inside liquid chromatographic columns and CFD modelling based on X-ray CT measurements

Multi-scale and/or multi-disciplinary approach to process-product innovation

Innovative Process Equipment-Operation Design & Analysis (T3-10)

Ing Florian Lottes
University of Erlangen-Nürnberg
Chair of Separation & Technology, Chemical and Bioengineering
Egerlandstr. 3
91058 Erlangen
Germany

Ing Dirk-Uwe Astrath
University of Erlangen-Nürnberg
Chair of Separation & Technology, Chemical and Bioengineering
Egerlandstrasse 3
91058 Erlangen
Germany

Prof Wolfgang Arlt
Universität Erlangen-Nürnberg
Lehrstuhl für Thermische Verfahrenstechnik
Egerlandstr.3
D-91058 Erlangen
Germany

Prof Erling Stenby
Technical University of Denmark
IVC-SEP, Department of Chemical Engineering
Building 229
DK-2800
Lyngby
Denmark

Keywords: Chromatography, x-ray CT, CFD

It is well known that the efficiency of liquid chromatographic (LC) columns strongly depends on the homogeneity of the packed bed inside. As the separation takes place in closed columns, the measurement of local system parameters related to the column efficiency cannot be achieved without deterioration of the flow field. X-ray CT provides an opportunity to measure local packing properties in a non invasive way.
Chromatographic glass columns of ID 26 mm and packed with different porous stationary phases were investigated with a medical CT device. Based on the attenuation of the X-rays, which are passing through the column, a two dimensional image of the column cross section was reconstructed. The degree of attenuation, which is material-dependent, is expressed by means of the so-called CT-number. By using potassium iodide as a contrast medium the measurement of intra-column breakthrough curves was possible. Iodide was used in these frontal analysis experiments due to its high atomic mass and thus a high attenuation for X-rays, leading to a qualitatively good contrast within the images. The columns were scanned at four axial positions, which were evenly distributed along the column length. To get radial resolved parameters an additional subdivision of the images into ten concentric annuli was necessary. From these images the information about the local column properties, like porosity or flow velocity can be extracted. Additionally the visualisation of locally resolved column efficiency was possible through calculation of local axial Dispersion coefficients.

The results show a clear existence of a homogeneous core region and an inhomogeneous wall region with different packing properties. In this wall region the interstitial porosity decreases compared to the core and additionally the fluid velocity drops due to a smaller permeability. These effects lead to a considerably decreased efficiency in these parts of the column. The influence of applying different pressures during the packing procedure and using differently shaped stationary phases on the overall and local column efficiency were intensively studied.

Subsequently the results of the CT experiments, e.g. permeability and Dispersion coefficients were implemented into a more dimensional CFD model. The CFD results show better agreement with experimental data than commonly used one dimensional models, e.g. the Equilibrium-Dispersive-Model of Chromatography.

These results show that it is worth to take local resolved parameters into account. This work gives clear evidence that connecting X-ray CT and CFD is a good approach in order to get a more precise performance prediction of LC-columns. Work in progress covers the modelling of the specific zones, like the near-wall region using different numerical methods and finding the reasons for the formation of an inhomogenous bed plus appropriate methods to avoid the latter.

Presented Thursday 20, 12:00 to 12:20, in session Innovative Process Equipment-Operation Design & Analysis (T3-10).

Conference logo