COMPUTATIONALLY-BASED DEVELOPMENT OF A MICROFLUIDIC PLATFORM FOR NANOPARTICLE SYNTHESIS: FROM DESIGN TO PRODUCTION

Nanomedicine is a new branch of medicine in which nanotechnologies are employed to prevent and treat diseases in an improved and personalized way. Drug-loaded nanoparticles (NPs) fall in this category and have become especially interesting following the COVID-19 pandemic, during which lipid-based NPs served as carriers to deliver pharmaceutical payloads (mRNA) into patients, improving efficiency and reducing side effects through targeted delivery. However, the synthesis of NPs is currently mainly done by bulk methods, where small variations in environmental conditions can lead to sub-optimal characteristics and high inter-batch variability. Since NPs are formed by the mixing of different fluids, microfluidic technologies are being investigated as promising alternatives to produce polymeric and lipidic NPs. The micrometric dimensions of the channels and the laminarity of the flows employed in these techniques enable highly repeatable production processes, ensuring stable environmental parameters both locally and globally, resulting in uniform conditions throughout the entire volume involved. In this context, engineering strategies can be exploited to describe and deepen what happens in the fluid-dynamic process during the formation of NPs, revealing peculiar fluidic characteristics that may help researchers in the optimization and scaling-up of the manufacturing process. Numerical strategies (e.g., computational fluid dynamics simulations, CFD) have proved to be able to provide valuable insight into the mixing patterns involved in the formation phases. However, when optimizing new formulations, the classical characterization methods lack the mechanistic aspect of NPs assembly. In this context, the first part of my PhD path focused on the adoption of numerical strategies to reproduce, understand, and predict the mixing process and NPs formation within microfluidic channels. To achieve this aim, three crucial aspects were addressed: i) constant comparison of numerical results with qualitative and quantitative experimental evidence; ii) introduction of novel mechanistic evidence connecting NPs formation with the constitutive elements of binary mixtures; iii) design of novel numerical variables to give insight into the impact of fluid-dynamic working conditions on the mixing process. From a practical perspective, the accurate handling of fluids is essential to guarantee a reproducible manufacturing process able to synthetize homogeneous NPs. Hence, in recent years, microfluidic-based production stations have become accurate and reliable instruments for NPs synthesis. Even though the promising results allow rapid market growth, the high costs and limited customization of commercial devices hinder widespread adoption of such devices. In this PhD project, different microfluidic-based platforms were developed adopting syringe and peristaltic pumps. Since each commercial microfluidic platform employs proprietary chip designs, the development of new NPs formulations becomes instrument- and chip-specific, limiting any possible transfer of know-how between laboratories. Hence, a new Design of Experiments (DoE) should be implemented, requiring extensive experimental effort and, in some cases, a novel experimental set-up to achieve an equivalent formulation. In this PhD thesis, this problem was addressed by adopting a novel numerical framework to translate an optimized liposome formulation between two distinct microfluidic chip geometries without requiring extensive experimental investigation. This approach enabled an increase in the total flow rate (TFR), thereby enhancing production throughput while preserving excellent and comparable morphological characteristics of the synthesized NPs.