Nanoporous characterization
Bioseparation media typically have nominal pore ratings in the range of tens to hundreds of nanometers rather than the 500 µm to 200 µm designed channels as examined here . Each gyroid material phase has an internal porosity that provides the required surface area for product or impurity binding in the diffusive domain, analogous to internal chromatography bead structure. Imaging a small piece of 3D printed gyroid at a pixel size of 16 nm using nanoscale X-ray CT enabled fine features to be resolved in Figure 6a, however the volume generated was deemed to be insufficient due to a greatly restricted field of view.
Therefore a pixel size of 63 nm was selected for imaging as a suitable balance of high resolution whilst generating a larger volume for analysis and flow simulation. Figure 6b displays an equivalent slice at the lower resolution that still demonstrates good signal to noise ratio and clear differentiation between the black voidage phase and the gray material. Figure 6c shows a much larger field of view at the same resolution, revealing the intricate and detailed substructure within the material phase that is further demonstrated in the 3D render of Figure 6d.
Tortuosity factor was simulated across all planes within a binarised cube, resulting in an average measurement of 2.52 ± 0.33. Tortuosity is an important factor in governing mass transfer properties, therefore the internal nanostructure in the material phase provides another length scale that can be optimized for the product and process of interest . By obtaining digital representations of real porous material then more accurate measurements can be made in comparison to conventional techniques. Equation based derivation using porosity as an input has historically been applied, however does not consider the internal geometry of materials being estimated for tortuosity . Diffusive bulk flow and streamlines are provided in Figure 7, suggesting a complex yet reasonably consistent geometry without any noticeable areas of fluid bypass. A bulk solution diffusivity of 7.00 x10-11m2s-1 resulted in a simulated material diffusivity of 2.17 x10-11m2s-1 ± 0.16 x10-11 m2s-1.
The binarised volume enabled pore analysis of the voidage phase. Figure 8a displays a pore network map. Pores can be observed to have a range of sizes and also connecting with other pores as represented by sticks between the spheres. The pore size distribution in Figure 8b measured an average pore diameter of 793 nm ± 315 nm. This pore size distribution suggests larger diameters compared to a previous study that measured an average of 289 nm, however that approach relied on 2D scanning electron microscopy measurements on the sample surface with a different material composition .
By applying a design cycle approach that uses imaging data to inform decisions then pore size characteristics can be specifically tailored to the needs of the product and bioprocess across both permeable and diffusive domains by analyzing real materials. Ensuring that porogens are entirely removed from the material phase through washing steps is necessary to minimize nanostructure variability.