Pore accessibility in CADET

Hi all,

I have been thinking recently of the necessity of differential pore accessibility between solutes (e.g. salt and proteins) in chromatography modeling and how best to account for it. Of course, if one would perform tracer experiments (under non-binding conditions) on a column packed with chromatographic resin containing typically sized pores (e.g., agarose with ~30 nm radius pores - Capto/Sepharose base matrix), it is expected that proteins (e.g. antibody with 4.5 nm hydrodynamic radius) and salt would have different retention. This would result in the protein and salt having different distribution coefficients or effective intraparticle porosities (say Ep = 0.9 for salt and Ep = 0.7 for protein). This implies that each component balance should have a distinct Ep value, depending on their size. Despite this, most modeling efforts use a single value of Ep for all species - either that for the salt component or that for the protein component. It is clear that both cases are incorrect. Although this would only cause the retention volume of the individual components to be off by less that one CV, in some cases this may be non-negligible (e.g. for very sharp elution peaks - 2 or 3 CV pool volume). This could be addressed by including a pore accessibility term or simply having the option to use a different Ep for each component in their respective balance equations.

One quick fix could be simply adjusting the lateral position of the elution peak by the difference in the exclusion volumes between the model porosity and the actual porosity of each component - CV*(Et,model - Et,actual).

A curious point tied to this would be if the isotherm parameters that have a basis of volume of solid phase (e.g. Langmuir qmax) will have to be adjusted for the protein components to account for the adjusted porosity. This introduces some additional complexity but should only require some conversions of the parameter with respect to the change in basis of the accessible volume.

From what I can tell in the CADET documentation, there is a pore accessibility term - but, it is only available for the size-exclusion case, where there is no adsorption. This seems limiting because the aforementioned problematic assumption applies for both binding and non-binding scenarios.

Would it be possible to have the pore accessibility term available under binding conditions or have a intraparticle porosity multiplex term that allows for component-specific porosities? I figure both of these should achieve the same result, so whichever is simpler would be fine to go with.

I am curious to hear the community’s thoughts on this topic, both from a theoretical and development point of view.



Thank you Scott for bringing this up.

As you have mentioned, the current model doesn’t work under binding conditions. We are actually working on an advanced model but it is too early to reveal detailed concepts and results.

It should be clear that under binding conditions the functionalized surface area needs to be partitioned into regions that are available to different numbers of components, and this will certainly make the unit system more complicated.