AUTHOR=Martin-Salvador P. , Verschueren R. H. , De Beer T. , Kumar A. TITLE=CFD-DEM model of a cold plasma assisted fluidized bed powder coating process JOURNAL=Frontiers in Chemical Engineering VOLUME=6 YEAR=2024 URL=https://www.frontiersin.org/journals/chemical-engineering/articles/10.3389/fceng.2024.1347313 DOI=10.3389/fceng.2024.1347313 ISSN=2673-2718 ABSTRACT=

Cold plasma coating technology for surface functionalization of pharmaceutical powder particles is a promising approach to introduce new characteristics such as controlled release layers, improved powder flow properties, stability coatings, and binding of active components to the surface. This is typically achieved in a fluidized bed reactor, where a jet containing the chemical precursor and the plasma afterglow is introduced through a nozzle while extra fluidization gas is injected from the bottom plate. However, the process requires proper mixing of the particles and precursor inside the plasma active zone to ensure a homogeneous coating of all particles. Therefore, such coating processes are challenging to optimize, given the complex phenomena involved in fluidization, plasma species reactions, and surface reactions. In this study, we use the CFD-DEM approach as implemented in the CFDEM®coupling package to model the process. The functionalization rate is modeled as mass transfer from the surrounding gas onto the particles, using a plasma coating zone where this transfer may happen. Mass transfer is switched off outside this zone. The DEM contact parameters and drag force are calibrated to our cellulose beads model powder using experimental tests composed by the FT4 rheometer and spouting tests. We show that while the chemistry can make or break the process, the equipment design and process conditions have a non-negligible effect on the coating metrics and thus must be considered. Cases where the fluidization flow is not high enough to produce good mixing have a high coefficient of variation of the coating mass, and therefore, they must be avoided. In addition, we also proposed an extrapolation procedure to provide results at longer coating times, showing that it is possible to predict coating performance even when simulations of the process for more than a minute are not computationally efficient.