270823 Multi-Layer Contactless Dielectrophoresis Using Thin Polyimide Films

Monday, October 29, 2012: 12:30 PM
406 (Convention Center )
Elizabeth Savage1, Michael B. Sano1, Alireza Salmanzadeh1,2, Eva M. Schmelz3 and Rafael V. Davalos1, (1)School of Biomedical Engineering & Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA, (2)Engineering Science and Mechanics, Virginia Tech, Blacksburg, VA, (3)Department of Human Nutrition, Foods, and Exercise, Virginia Tech, Blacksburg, VA

Multilayer Contactless Dielectrophoresis Using Thin Polyimide Films

Dielectrophoresis (DEP), the motion of particles in a suspending medium due to their polarization in a non-uniform electric field, is an effective means of microfluidic particle manipulation. Traditionally, DEP has been performed using electrodes patterned in the bottom of a sample channel, and has been successfully utilized for a variety of applications, ranging from separating live and dead cells to trapping cells, viruses, and DNA, to determining antigen expression. However, challenges in using this technique include Joule heating, bubble formation, and electrode delamination. Contactless dielectrophoresis (cDEP) addresses these issues by using conducting fluid electrodes in side channels along a central sample channel, separated from the sample channel by a thin insulating barrier. Metal electrodes are inserted into the conducting fluid and the electric field is then applied across the sample channel without direct contact between the electrodes and the sample. When applying an electric field across the device, there is a large voltage drop across the insulating barrier and a relatively small drop across the sample channel itself. In order to achieve a sufficient voltage drop across the sample channel, much larger voltages must be applied across the device. The maximum applied voltage is limited by instrumentation capabilities and the dielectric breakdown properties of the barrier material.

The insulating barrier behaves as a resistor and capacitor in parallel and analysis shows that a reduction in resistance and an increase in capacitance of the barrier decreases the voltage dropped across this barrier and enables cDEP devices to operate over a larger frequency spectrum. For mammalian cells with similar phenotype, the largest difference in the Clausius-Mossotti factor occurs close to their first crossover frequency. In low conductivity physiologically relevant buffers, this is usually between 5-100 kHz, a range reaching below the lower frequency limit of some previous single layer PDMS cDEP devices.

A multilayer device in which the fluid electrodes and sample channel are in separate stacked layers can help achieve a reduction in the voltage drop across the insulating barrier. Increasing the area of the electrode-sample channel interface reduces the resistance while increasing the capacitance of the barrier. In single-layer devices, this interfacial area is limited by the depth of the sample channel, which in turn is limited by the depth of the silicon wafer stamps whether using deep reactive ion etching (DRIE) or photoresist spin-coating (such as SU-8). However, a multilayer device enables the fluid electrode area to be increased by locating it above the sample channel; it is then limited only by the specified channel width.

In this work, we capitalized on the observation that the voltage drop from fluid electrode to sample channel can be further decreased by choosing a barrier material that has a higher relative permittivity than PDMS. Here, a commonly available thin polyimide tape was utilized as the barrier between the fluid electrode and sample channel. The use of a higher dielectric constant self-adhesive thin polyimide film with a multilayer cDEP device provides a simple and cost-effective means of significantly increasing the potential of cDEP for sorting of mammalian cells.

Master stamps were created for the fluid electrodes and sample channel on separate silicon wafers using deep reactive ion etching. Polymer replicates were then cast in PDMS with a 10:1 ratio of polymer to curing agent. After punching access holes to each of the fluid channels, a strip of polyimide tape was applied to the PDMS replicate containing the fluid electrode channels. The fluid electrode and sample channel substrates were bonded together after exposure to air plasma for two minutes such that only the electrode cannel was in contact with adhesive. The completed devices were stored under vacuum.

Prior to experimentation, the fluid electrode channels were filled with phosphate buffered saline with a conductivity of 1.4 S/m. Polystyrene microspheres were suspended in either DI water or a low conductivity sucrose solution with conductivities of 0.001 and 0.01 S/m, respectively. A high voltage amplifier and series transformer were used to apply voltages with amplitudes up to 300 VRMS between 10 and 800 kHz.  Preliminary experiments match theoretical predictions that these multilayer devices with polyimide film barriers induce a DEP response in particles over a wider frequency spectrum than single layer devices with PDMS barriers.

This work presents a simplified process for fabricating multilayer cDEP devices. The use of manufactured polyimide films ensures uniform thickness of the insulating barrier between batches and provides superior electrical performance compared to PDMS. The use of an adhesive film eliminates the need for multiple bonding steps and the multilayer fabrication process allows for the use of geometries which maximize the frequency response of these devices.

Extended Abstract: File Not Uploaded