Richard D. Braatz

Massachusetts Institute of Technology (MIT)

Wednesday, October 07, 2026, 16:15 - 17:30

KG I, Aula, Platz der Universität 3, FreiburgKollegiengebäude I, Aula, Platz der Universität 3 , D-79098 Freiburg, Germany

Abstract: This talk describes a new two-stage process for the automated, controlled continuous manufacturing of mRNA-loaded lipid nanoparticles. The talk begins with a summary of the commercial technologies for the manufacturing of these nanoparticles and their weaknesses in terms of mixing time scales. A two-stage process is defined which substantially improves micromixing while providing an additional variable for the control of the nanoparticle size and morphology. After demonstrating enhanced control including the use of Bayesian optimization, the talk concludes with progress in the building of mechanistic and machine learning-based models for the continuous manufacturing of lipid nanoparticles. While the experimental results are for mRNA-loaded lipid nanoparticles, the new process applied to the loading of any biotherapeutic molecule or macromolecular assembly, and the modeling and control methods are largely applicable to the manufacturing of other types of nano- and microparticles. This work is in collaboration with Profs. Allan Myerson and Bernhardt Trout at MIT.