Abstract

3D printing is a rapidly developing and valuable tool in medicinal research. Sadia et al. (2018) developed 3D printed tablets with perforating channels and concluded that larger channels in greater quantity dissolved more rapidly than smaller and nonexistent channels. Building on this, a computational approach to analyse the drug release rate limiting steps is developed. A range of 1, 2, and 3 dimensional geometries are used to obtain an approximation of drug release based on diffusion models (Siepmann, 2015), using both implicit and explicit numerical methods. Fickian diffusion is modelled under the assumptions that the polymer relaxation time is much greater than solvent diffusion time, and the formulation is uniform in drug content. The diffusion model can be rescaled according to tablet dimensions (often diameter, d, is used), and the diffusion coefficient D, to define the amount of molar flux through a surface area.

Insertion of channels into the geometries tested shows that the increased drug release rates found experimentally is due to two factors; increased tablet surface area and decreased mass. Solving the diffusion equation showed that the drug release rate increase is largely related to the increased surface area, though it is inaccurate to compare the channelled tablets to whole tablets of a larger mass.

When the tablet geometry set-up was run and rescaled to match the experimental data, the formulation was found to have a diffusion coefficient D of approximately D=6x10-10m2 /s. However, the x-axis rescaling factor (tD/d2 ) applied to the diffusion equation are not applicable as the experimental data does not also follow this analytically derived rule. Thus, the standard Fickian diffusion model in sink conditions with accurate geometry is not sufficient to describe the drug release. Further in situ research is recommended in surface erosion and bubble formation in channels, alongside redesigning the experimental methodology to ensure uniformity where possible.