Branching junctions such as “T” and “Y” are common elements of fluidics networks such as HPLC and capillary electrophoresis, plus automated analysis systems. The simplicity of their design masks complicated fluid dynamics that can trap and delay particles in the liquid flow. Inhomogeneous flow might cause errors in analytical microsystems.
The phenomenon is seen as a dispersion of suspended particles when flowing from the base of the Y or T to the branches in square cut channels. Photomicrographs show flow profiles of a suspension that creates a vortex in the outlet arms that has a trapping region (TR). The majority of the flow is around the TR, as one would expect. However, in the TR, there are several regions. In one, the particle remains stationary for several milliseconds, until another particle comes in and bumps it out, much as a billiard ball. In another region, the particle moves slowly downstream and then stops. Some of the stopped particles slowly creep back to their origin over 50 ms.
The profile of the trap for a T junction resembles a classical ship’s anchor (Figure 1a and b). The flow regime is a mixture of the vortex with other flows. Thus, the flow velocity is much more nonuniform than one might expect. The authors report, “We are unaware of prior descriptions of such regions in the literature” (Oettinger, D.; Ault, J. et al. Invisible anchors trap particles in branching junctions. Phys. Rev. Lett. Aug. 3, 2018). The trapping regions are effectively invisible surfaces.
Figure 1 – Profile of trapping region for a Y branching fluidics junction with downward flow. Particles can be trapped within the region for many milliseconds. a) 3-D view of the two trapping flows created in an inverted T junction; (b) top-down view of the anchor trap; c) flow profile of particles in a V junction. Red lines show flow of nontrapped particles. Some particles (green) will remain in the trapping region and spiral into the extended vortex region. A very few will be trapped at P1 and not move.The experimental parameters are too complex for a blog post. The anchor-shaped trapping region can be one or two, depending on Reynolds number and dimensions of the junction. However, if you are working in microfluidic channels or with manifolds, you should consider the potential impact of the flow inhomogeneity on your experiment.
I asked the authors about the potential impact of interacting particles on the TR. This might be important with dynamic aggregates such as proteins or concentrated formulations.
Robert L. Stevenson, Ph.D., is Editor Emeritus, American Laboratory/Labcompare; e-mail: [email protected]
Related Products