Even with stringent levels of detection the CDC estimates 48 million people a year suffer from food poisoning (1, 2). While it is true that deaths from food poisoning are rare, the economic impact from food poisoning is estimated at $77 billion dollars a year (3). Rapid and efficient detection in food safety applications has two primary challenges. First, the FDA detection level for bacterial contamination is one live cell per 20 grams of food material, a large volume. The practicality of finding one prokaryotic cell in the volume requires pre-concentration as a crucial step in improving the reliability of the test. Second, it is possible to find multiple strains of a given bacteria in a sample. Each of these strains will have similar morphology but different pathological results. Thus a system is required that is strain specific which can also process large quantities of sample.

Techniques for cell sorting in microfluidic systems have received a lot of attention recently. The most popular methods to separate the cells are based upon physical characteristics, such as size, shape, weight, density or elasticity. These systems often utilize passive structures that rely on flow conditions or size exclusion principles. However, sorting via this method is not useful for food safety applications because of the strain problem previously described. Labeling the targets with fluorescent or bead-based signals is inefficient due to the large sample volumes involved in food safety testing.

This work presents a system that captures dilute concentrations of particles from a bulk fluid (4, 5). The particles are captured by magnetic beads inside a microfluidic channel. The magnetic beads are actuated by rotating around soft magnetic features using an external magnetic field. The magnetic bead sweeps through the fluid, increasing the chances of the particles coming into contact with the magnetic bead by increased fluid mixing and flow agitation (6). Protein and antibody coatings on the magnetic bead then capture the target particles from the bulk fluid and trap it with the magnetic bead. This system is easily multiplexed across multiple channels to process large quantities of fluid sample and incorporated into a portable system.

The system described here uses commercially available magnetic beads (Dynabeads M-280, Life Technologies) and soft magnetic features of permalloy (80/20 NiFe). The soft magnetic features are made by electron beam evaporation onto a bare glass wafer. A standard photolithography process with wet etching of the excess metal was used to produce the soft magnetic features. A layer of silicon dioxide is deposited via plasma enhanced chemical vapor deposition (PECVD) over the entire wafer to protect the soft magnetic features from oxidation during experiments. Finally a PDMS microfluidic channel is created from standard PDMS molding and placed over the wafer to create a fully sealed channel. Magnetic beads are pumped into the array where the magnetic forces from the external magnetic field and the soft magnetic features distribute the beads across the entire array.

In this work fluorescent particles (Fluorspheres, 1 µm diameter) are pumped through the array of rotating magnetic beads at various conditions and images taken of the fluorescent particles capture in the array. These particles are analogous in size to common bacteria sought in food safety testing. The goal is to quantify the effectiveness of this system to capture targets from the bulk fluid under different flow and geometric conditions. The conditions considered in this work are flow rate, bead rotation speed, the ratio of magnetic bead diameter to channel height and the packing density of the magnetic features in the array.

References

1.         E. Scallan, P. M. Griffin, F. J. Angulo, R. V. Tauxe and R. M. Hoekstra, Emerging infectious diseases, 17, 16 (2011).

2.         E. Scallan, R. M. Hoekstra, F. J. Angulo, R. V. Tauxe, M.-A. Widdowson, S. L. Roy, J. L. Jones and P. M. Griffin, Emerg Infect Dis, 17 (2011).

3.         R. L. Scharff, Journal of Food Protection®, 75, 123 (2012).

4.         D. Owen, W. Mao, A. Alexeev, J. Cannon and P. Hesketh, Micromachines, 4, 103 (2013).

5.         M. Ballard, Z. Mills, D. Owen, S. Hanasoge, P. Hesketh and A. Alexeev, in APS Meeting Abstracts, p. 43002 (2015).

6.         M. Ballard, D. Owen, Z. G. Mills, S. Hanasoge, P. Hesketh and A. Alexeev, Bulletin of the American Physical Society, 60 (2015).