(364e) Development of a Realistic 3D Model of Silica Monoliths for Cfd Simulations

Authors: 
Vasudevan, V., National University of Singapore
Loh, K. C., National University of Singapore


The full potential of therapeutic biomolecules can be realized only if large quantities are available for testing, characterization and quantitative structure?activity analysis. Some of the crucial factors for their successful and fast separation are a reduced mass transfer resistance within large open-ended pores or channels and a small residence time within the chromatographic column to preserve their bio-activity. This has traditionally been carried out using particulate HPLC columns in which separation efficiency increases with decreasing particle diameter. The trade-off between column back pressure and efficiency has resulted in the popular use of monolithic columns for fast and efficient purification. Reduced back pressure, minimal non-specific binding and low product degradation, without loss of resolution, make monoliths ideal candidates as chromatographic matrices for purification of large bio-macromolecules.

Direct visualization, believed to be the only way to capture the true morphology of porous materials, has traditionally been conducted using scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Rapid advances in non?invasive 3D scanning techniques have led to the use of laser scanning confocal microscopy (LSCM) to capture ?as is? the inherent morphologies of porous materials. Direct use of reconstructed 3D images as bounding geometries in an environment amenable to solving fundamental transport equations has obviated the need for fitting parameters in transport models, which limit the predictive value of the models. This research is aimed at elucidating the pore structure and transport mechanism characteristics of a silica monolith through CFD simulations by developing a realistic 3D model through image analysis of micro-CT scans of a monolith sample.

Reconstructed images from micro-CT scans of a silica monolith sample were processed to identify a representative pore volume. The images were thresholded and the computed porosity compared with that of the largest scanned image, to identify a representative sized domain. The surface mesh enclosing the flow domain (pore volume) was imported into an environment amenable to using the surfaces as physical boundaries for flow simulations. Simulations of flow hydrodynamics as well as dispersion characteristics in the porous sample were performed using the commercial CFD software ? FLUENT. These were verified for grid independence and other geometry-related parameters such as tortuosity and porosity. The effects of flow rate using water as mobile phase were validated against experimental data obtained from commercially available monoliths. Simulation results for dispersion of analyte molecules in non-retained and retained conditions will be discussed. The computing resources utilized in the study will also be highlighted in this presentation.