Rates of Diffusion-Limited Reactions in Stirred Microfluidic Flows
Bradley Morgenfeld1, Joseph D. Kirtland2, Christina D. Shah2, and Abraham D. Stroock2. (1) School of Chemical and Biomolecular Engineering, Cornell University, 120 Olin Hall, Ithaca, NY 14853, (2) Department of Physics, School of Chemical and Biomolecular Engineering, 360 Olin Hall, Cornell Univeristy, Ithaca, NY 14583
A challenge in the development of microchemical systems is the induction of mixing flows to homogenize solutions of reagents. Microfluidic processes occur at low to moderate Reynolds numbers (Re < 100), such that the flows are laminar, and high Peclet numbers for mass transfer (Pe > 100). This situation implies that solutions introduced into a microfluidic flow will typically only achieve a fully mixed state after traveling a long distance (~Pe*H, where H is the typical cross-sectional dimension of the channel). In this default situation, microfluidic reactors can become impractically long. This technological context demands a new investigation of the fundamental processes by which laminar mixing flows affect the rate of homogeneous chemical reactions. We will present our experimental work on the characterization of the rate of a second order, diffusion-limited reaction in three different classes of laminar stirred flows. The flows considered are: 1) a Poiseuille flow in microchannel with smooth walls and periodically spaced corners that induce three-dimensional Dean flows at moderate values of Re. 2) A flow generated by a uniform topographical pattern in one wall of a channel; this flow is three-dimensional, but does not exhibit Lagrangian chaos. 3) A flow generated by an axially varying topographical pattern in one wall of a channel; this flow does exhibit Lagrangian chaos. The reaction studied generates a fluorescent product. We use fluorescence microscopy and digital image analysis to follow the progress of the local mean concentration of the product as a function of axial position within the channel. We operate at Peclet numbers ranging from 300 to 30000. Based on these experimental results, we will discuss important qualitative and quantitative features of the reaction dynamics in these three classes of flows. In the two cases of non-chaotic flows, we will explain the importance of weak Dean flows in controlling the progression of the reaction. In the case of chaotic flow, we will compare our results to predictions of a simple theoretical model of convection-diffusion-reaction. We will finish with conclusions regarding reactor design based on our findings.