Experimentally Validated Multiscale Modeling of Continuous Manufacturing of Complex Injectables

Purpose:
Liposomes and polymeric micelles are the nanoparticles used as pharmaceutical drug delivery systems to improve the solubility and chemical stability of drug molecules. In the pharmaceutical industry, both nanoparticles are typically manufactured by batch processing but with some disadvantages such as scalability and reproducibility issues which have resulted alternative approaches. Thus, continuous processing as one approach can benefit liposomes and polymeric micelles manufacturing in the aspects of higher-throughput, increased productivity, and reduced energy requirements. An innovative continuous processing system was developed at UConn, where two liquid flows are mixed in a highly controlled manner via a co-axial turbulent jet, producing monodispersed nanoparticles. In this study, we researched the impact of intermolecular forces such as dipole-dipole forces among the various components (ethanol, lipid/polymer and water) in nanoscale along with employing the principles of fluid dynamics. Hence, multi-scale computational approach was utilized to study the liposome and polymeric micelle formation process as a coaxial turbulent jet flow to probe the underlying mechanism, and to quantitatively predict the formation of the nanoparticles.

Methods:
Both computational fluid dynamics (CFD), as a meso-scale simulation, and, coarse-grained molecular dynamics (CG-MD), as a micro-scale investigation, have been conducted to reveal the detailed mechanism of nanoparticle formation. The CG-MD simulation trajectories and their analysis were carried out using GROMACS software package. The Large Eddy Simulation (LES) model in COMSOL Multiphysics incorporating energy equation, in CFD simulations, were implemented with a high-resolution mesh in the mixing area. The simulation results were further validated with experiments, using flow patterns, temperature profiles from the CFD studies, and micelle size and drug encapsulation from the MD studies.

Results:
In liposome studies, the CG-MD simulations reveal that liposomes have formed successfully in co-solvent solution. The evaluation of bilayer thickness of liposomes at different temperatures shows a good agreement with the trend of thickness changes from previous experimental studies. In polymeric micelles studies, multiple micelles were formed successfully in our CG-MD simulations, and the predicted particle size in a good agreement with those obtained in the actual experiment. The structure of the micelles was inspected, in which increasing drug concentration (at fixed polymer concentration) resulted changes in micelle structure, transitioning from spherical to worm-like micelles. In the CFD simulations, by adjusting Smagorinsky coefficient (Cs) of LES model companied with energy and mass transfer equations, the unique flow patterns of ethanol/water were reproduced successfully and the temperature variations within the co-axial turbulent jet were predicted accurately.

Conclusions:
The liposome and polymeric micelle formation processes in micro- and meso-scales were detailed by associating complementary discrete and continuum computational modeling work. The results from CG-MD precisely describe the internal structure view, bilayer thickness, and formation mechanisms of liposomes and micelles, which are challenging to observe and study through experimental work. Moreover, the influence of process temperatures and flow rates to the jet flow have been evaluated by CFD simulations. The integrative discrete CG-MD and continuum CFD computational modeling of nanoparticle formation were found to be two viable complementary approaches which cover micro-scale and meso-scale, respectively.