Adsorption processes are increasingly important in modern filtration systems for the removal of pollutants, odors, and chemical contaminants in applications such as water treatment, air purification, carbon capture, and industrial filtration. To support simulation-based analysis and design of adsorption-based filters, a new adsorption modeling capability has been integrated into the GeoDict simulation platform and validated in a joint project with MANN+HUMMEL GmbH, focusing on toluene adsorption on activated carbon.

The methodology couples fluid flow and mass transport simulations to describe adsorption phenomena across multiple length scales. Molecular transport is modeled using an Euler–Lagrange particle-tracing approach, while adsorption equilibria are calculated using established isotherm models, including Langmuir, Toth, and Freundlich. This combined framework enables the prediction of adsorption dynamics and breakthrough curves for a wide range of adsorbates.

Two complementary modeling approaches are implemented within the platform. A tracer-based transport method iteratively solves adsorption equilibria to predict breakthrough behavior in large, non-resolved structures such as cabin air filters or honeycomb geometries. A field-based approach targets smaller, fully resolved microstructures, such as packed beds, where solute transport is represented by a transient concentration field.

For validation, MANN+HUMMEL GmbH provided a µCT scan of a filter medium containing activated carbon. The image data were processed in GeoDict to generate a 3D-representation of the filter structure for adsorption and desorption simulations. Material parameters of the activated carbon, including porosity and tortuosity of the unresolved particles, as well as adsorption isotherms for toluene and corresponding experimental datasets, were also supplied. Based on the experimental conditions, both adsorption and desorption simulations were performed. The tracer-based toluene adsorption simulations showed very good agreement with experimental breakthrough data, while desorption simulations were consistent with physical expectations.

The validated simulation approach enables identification of critical transport and adsorption effects, such as hot spots and poorly utilized regions, and supports systematic optimization of adsorption-based filtration systems. By providing insight into adsorption mechanisms from microstructure to component scale, the approach reduces reliance on extensive experimental prototyping and supports efficient digital development of advanced filtration technologies.

Future work will focus on extending the framework to additional adsorption isotherms, multi-species adsorption, and further improvements in computational efficiency to address emerging challenges in next-generation adsorption-based filtration technologies.