Mechanistic modeling of primary drying of frozen particulates

Thin-film freeze-drying (TFFD) produces discrete, millimeter-scale frozen particles under highly controlled conditions, yielding beds with reproducible structure and distinctive drying kinetics. In this study, we present an integrated, mechanistic modeling framework for the primary drying of frozen particulate beds, with a particular emphasis on TFFD products. The model incorporates experimentally measured porosity, tortuosity, and particle geometry from micro-computed tomography (μ-CT) scans and resolves the dynamic interplay of heat and mass transfer at both the particle and bed scale. Heat transfer is described using the Zehner–Bauer–Schlünder (ZBS) model for effective thermal conductivity in porous media, while mass transfer is governed by the dusty gas model and a shrinking core model. The particle bed is discretized using a finite volume method into horizontal layers, and model predictions of temperature, ice saturation, and sublimation kinetics are directly compared to experimental drying data for beds composed of spheres, domes, and discs. Results demonstrate that particle geometry and packing architecture exert a significant influence on drying rates and spatial temperature profiles, with the model predicting drying endpoints within 3% of experimental values and accurately capturing the temporal evolution of temperature and ice saturation. The framework enables rational optimization of primary drying protocols by linking measurable structural parameters to drying performance, and its predictive accuracy positions it as a valuable tool for future parametric studies and process design. This approach provides new insight into the design of next-generation lyophilized products and establishes a foundation for predictive modeling across a wide range of particle morphologies and drying conditions.