Assessment of non-isothermal pore network modeling of freeze-drying using 4D X-ray imaging

Freeze-drying (FD) is an important, but also expensive and slow conservation process for pharmaceuticals and foods. The interactions of coupled heat and mass transfer with the solid structure of the frozen product are complex and in most cases only addressed with macroscopic modeling approaches that heavily rely on averaged structural data. To enable a better physical understanding of primary FD, this work presents our non-isothermal pore network model (PNM) for primary FD in direct comparison to empirical data obtained with 4D X-ray imaging. For this, FD of maltodextrin (5 % w/w solid content) is experimentally tracked with highly time- and space- resolved operando X-ray imaging. The porous structure of the very same sample is afterwards reconstructed for PNM simulations. As the size of the sample is much larger than the geometric peculiarities of the sublimation front, it is for the first time possible to study correlations between individual pore temperatures and pressures, pore-scale heat and mass transfer and evolution of the sublimation front inside the amorphous microstructure numerically, guided by experimental 4D X-ray imaging of the FD process. Complementary to the experimental data, the time-dependent spatial temperature and pressure variations within the pore space could be explored with the PNM approach. Given the sublimation front widening during the process, significant temperature and pressure differences within this region could be demonstrated. Such information is generally not available from experimental measurements or any macroscopic model, which highlights the strengths of the PNM.