Microscale Reactor CFD Model Validation Using Direct Numerical Simulations, High-Speed MicroPIV, and Reactive Laser-Induced Fluorescence

Iowa State University, Ames IA

Investigators

Abstract

Rodney O. Fox 0730250 Microscale reactors operating in the turbulent flow regime are used in Flash NanoPrecipitation" to achieve the required mixing times for the production of uniform-sized nanoparticles of hydrophobic organic compounds. Because the overall process is controlled by mixing and kinetics, as opposed to thermodynamics, significant quantities of precisely controlled nanoparticles can be rapidly manufactured at the industrial scale in a continuous process without the need for long batch times and handling large quantities of solvents. However, the understanding of rapid precipitation needed to design a microscale reactor requires a detailed understanding of the role of macro-, meso- and micromixing on the development of supersaturation. More generally, some of the key questions in developing microscale reactors for Flash NanoPrecipitation are: (i) How efficiently will the microscale reactor operate at different flow velocities and stream ratios? (ii) How do multi-injector microscale reactors compare to reactors with two impinging streams? (iii) What are the flow regimes observed in microscale reactors and how can they be optimized for rapid mixing? To answer these questions, computational fluid dynamics (CFD) can be employed, provided that it has been adequately validated against experimental data over the range of operating conditions of interest for design. Currently, microscale reactor design is done by experimental trial and error; hence the availability of a predictive computational tool would transform how research is done in this field. A CFD model based on low-Reynolds-number turbulence models has been developed at Iowa State and partially validated using outlet conversion data collected at Princeton for confined impinging jet (CIJ) and multi-inlet vortex mixer (MIMV) microscale reactors. While the results of these validation studies are extremely promising and suggest that CFD will indeed be a transformative tool for microscale reactor design, the indirect dependence of the outlet conversion data on the details of the flow field in the microscale reactor leaves many open questions concerning the true predictive capability of the CFD model. This unsatisfactory situation provides the motivation for developing direct numerical simulation (DNS) and microscale experimental tools that can provide the local data for instantaneous velocity and scalar fields in microscale reactors that are required for rigorous CFD model validation. To meet this need, the two PIs are collaborating on the development of microscale particle image velocimetry (microPIV) for measuring the instantaneous velocity field in microscale reactors operating in the turbulent flow regime. Recent microPIV measurements of turbulent flow in a microscale CIJ reactor have demonstrated the feasibility of such measurements using the state-of-the-art equipment in their laboratory. Moreover, they have recently demonstrated the ability to control experimental uncertainty to the level required for quantitative comparisons with CFD predictions. The primary purpose of this project is to continue to improve this promising work to the point where the experimental data can provide for definitive validation of CFD models for microscale reactors. The second goal is to perform selected DNS of the fundamental governing equations for the fluid velocity in microscale reactors to complement the experimental microPIV data. These data will be employed in a systematic validation study to improve the CFD model, and to determine its range of validity for microscale reactor design. Intellectual Merit The development of experimental tools for the measurement of flow in microscale devices is an important technical and intellectual challenge. Indeed, in microscale flows the need to image small domains and the desire for quantitatively accurate measurements of unsteady velocity fields leads to very stringent experimental requirements as compared to macroscale flows. Likewise, the development and validation of predictive computational models for microscale reactors pushes the limits of current knowledge. For example, the prediction of the scalar dissipation rate in low-Reynolds-number, high-Schmidt-number flows is still an open problem, but one which lies at the heart of microscale reactor design. Broader Impact The production of uniform-sized nanoparticles of hydrophobic organic compounds by an economical, scalable process is a considerable challenge. It is motivated by the use and potential use of nanoparticles in drug delivery, especially poorly water soluble drugs, cosmetics, dyes, medical imaging and diagnostic, and pesticides. One of the most advanced processes to produce such nanoparticles is Flash NanoPrecipitation. The tools developed in this project will facilitate the design and optimization of the microscale reactors needed for Flash NanoPrecipitation, and will complement the detailed kinetic models for nanoparticles formation under development in other NSF-funded projects. The potential broader impact of this project could thus be to transform the field of microscale reactor design.