RESEARCH ARTICLE Richard A. Truesdell J. W. Bartsch T. Buranda L. A. Sklar A. A. Mammoli Direct measurement of mixing quality in a pulsatile flow micromixer Received: 2 December 2004 / Revised: 6 June 2005 / Accepted: 8 June 2005 / Published online: 27 September 2005 Ó Springer-Verlag 2005 Abstract Pulsatile action can be used to mix two streams entering a tube from two separate branches of a bifur- cation at low Reynolds numbers. The pulsatile action is provided by two pinch valves, which deform flexible tubing immediately upstream of the connection. The pinch valve action is controlled using a master-slave pulse generator setup. The quality of mixing is evaluated directly by measuring the fluorescence that results from the chemical reaction of species transported in the two streams, one containing native biotin and the other, fluorescein biotin bound to streptavidin. The reaction kinetics are accounted for by normalization using fluo- rescence measurements on well mixed solutions at the same residence time. The results show that the pulsatile micromixer provides almost complete mixing. Further- more, the present measurements match results obtained in a previous experiment where flow visualization and image analysis were used to measure mixing quality in a scaled-up model. Keywords Laminar flow Mixing Channel flow Fluorescence Quenching 1 Introduction Recent trends in the field of flow cytometry and biode- tection are towards device miniaturization. With reduced dimensions, viscous forces dominate the flow in the fluid sample handling apparatus and hence turbulence is lar- gely absent. This creates difficulties when the mixing of fluid samples is required. While several techniques [Strook et al. (2002), Khakhar et al. (1987), Jackson et al. (2002a)] have been proposed to achieve mixing at low Reynolds numbers (Re) there are several physical constraints which hinder their application for the above purpose. In particular, the mixer should not allow suspended particles (beads, cells, etc.) to become trapped at a wall or obstacle. A promising device which is char- acterized by simplicity of operation and which satisfies many of the criteria for use in biodetection applications was proposed recently by Truesdell et al. (2003). In this, two miscible fluid streams, each carrying a reagent, merge at a three arm junction (a ‘Y‘ connection). The flow rate of the two incoming streams has a mean and a pulsating component. In the limiting case of zero mean flow, a deforming region of fluid in which material points are confined can be identified. Motion within this region is similar in nature to the well-know case of the driven cavity [Leong and Ottino (1989) and Ottino (1989)], with the difference that the boundaries of the cavity deform. Also, qualitative similarities between the pulsatile mixer flow and the highly idealized sine flow described by Liu et al. (1995) are obvious. In the former case, material points are confined to the deforming control volume, while in the latter case material points exiting the fixed control volume from one side re-enter from the opposite side. In both driven cavity and sine flow, chaotic advec- tion rapidly ensues with appropriate boundary condi- tions. Given that the flow for the pulsatile mixer considered here is a combination of the driven cavity and the sine flow, chaotic advection can be expected. Effective mixing is also shown in a similar mixer design by Selve- rov and Stone (2001), Glasgow and Aubry (2003a, 2003b), Glasgow et al. (2004a) and Bertsch et al. (2001). Finally, the chaotic nature of the flow at the Y junction is demonstrated experimentally by Truesdell et al. (2003). In the study of Truesdell et al. (2003) strong pulsa- tions in the incoming streams were generated by squeezing flexible tubing via two pinch valves immedi- ately upstream of each branch of the ‘Y’ connection. The pinch valves were controlled by pulse generators which allowed independent control of the pulse width, delay, and period as shown in Fig.1. R. A. Truesdell (&) J. W. Bartsch A. A. Mammoli Department of Mechanical Engineering, University of New Mexico, Albuquerque, NM, 87131 USA E-mail: rtrue@unm.edu T. Buranda L. A. Sklar Health Sciences Center, Department of Pathology, University of New Mexico, Albuquerque, NM, 87131 USA Experiments in Fluids (2005) 39: 819–827 DOI 10.1007/s00348-005-0015-7