• CFD
• cross-shaped channel
• cylindrical obstacles
• Damköhler number
• direct numerical simulations
• Flowsculpt
• genetic optimization
• inertial flow
• interface stretching
• laminar regime
• lipid particles
• micromixer
• microreactor
• mixing
• nanoparticle production
• pillar sequence
• reaction yield
• T-junction
• X-junction

Microfluidic reactor technology is receiving a growing interest, as it is believed to provide new opportunities for the chemical and process industries. The highly efficient mass and heat transfer combined with the unprecedented control over the tiny volumes has accelerated development in biochemical, biomedical, and pharmaceutical fields for laboratory automation and industrial manufacturing, especially after the global pandemic in 2020, where research and production of vaccines have been crucial.
The miniaturization of the processes allows the achievement of high reaction yields and high selectivities, together with the possibility of taming potentially dangerous reactions safely. The interest is further motivated by the remarkable advances in the field of microfabrication through additive manufacturing (3D printing) techniques.

The present thesis proposes an experimental study joined with numerical modeling of passive microreactors, i.e., where mixing is promoted without the help of external energy sources. The aim is to understand how the flow regimes can influence the mixing and the reaction yield in clever but simple geometries at different operating conditions related to working fluids, reaction kinetics, and flow rate.

The first part of the work foresees the detailed characterization - by using microscope imaging techniques and numerical simulations - of the complex fluid dynamics followed by an analysis of the mixing and the reaction yield in the simplest passive microdevices, i.e., the T- and X-shaped reactors. For this purpose, a test reaction is selected being suitable with the adopted optical techniques and with the chemical times of the same order as the residence time in the fluidic reactors.
For numerical modeling, grid refinement techniques have been implemented in Computational Fluid Dynamics (CFD), increasing the resolution of the calculation grid in areas with high reaction rates. The validation of the methodology is pursued through qualitative and quantitative data analysis techniques.
In particular, attention has been paid to the relationship between the mixing and the reaction timescales, identifying a semi-empirical correlation between the Reynolds and Damk\"{o}hler numbers. Furthermore, a detailed comparison of the performance of the two reactors is also discussed to show the most convenient for the process intensification and the flow chemistry.

The second part of the thesis concerns the optimization of the design of a sequence of mixing elements, called pillars, to enhance mixing in the inertial flow regime. An optimization design software is exploited to predict the combination of the pillars in sequence and improve mixing under narrow working conditions. Soft-lithography and advanced microscopy imaging are utilized to fabricate and explore optimally designed microreactors. The experimental data assisted by CFD simulations validate the innovative design strategy of microreactors.
To further investigate the limits of the optimization, the mixing performance achieved within the pillar sequences is tested with varying geometrical and operating parameters, finding one sequence to be notably more robust and efficient than the others.
Then, this microreactor is analyzed by mixing water and ethanol, as this process is used to generate, e.g., good-quality lipid nanoparticles for the encapsulation of active pharmaceutical ingredients (APIs). The findings suggest that the optimal sequences of pillars could satisfy the requirements of these new emerging engineering and bioengineering processes.