Skip to main content

Microfluidics Turning Chips into Giant Chemical Labs

"Microfluidics," which aims to compress whole chemical engineering factories and analytical laboratories to the sizes of computer microchips, is becoming an experimental focus at a number of research universities, including Duke. The military hopes such a "wet" lab on a chip could someday be attached to a soldier's helmet, where it could perform instant laboratory analyses of suspicious airborne molecules "before you take your next breath," said Richard Fair, a Duke professor of electrical and computer engineering. With a lightning-quick shuffle of chemical reagents, dispensed in the form of microscopic droplets, the chip could identify lethal agents in a poison gas attack in time for another automated system to inject the appropriate antidote into the soldier's skin, Fair added. In a completely different scenario, a pharmaceutical company could use a similar kind of microfluidic array to derive test chemicals from rare plant extracts collected in a tropical jungle at great cost. After a Lilliputian-scale laboratory analysis that would have been prohibitively expensive had those extracts been tested in larger portions, scientists might discover a miracle drug. In a key innovation toward this futuristic technology, researchers at Duke's Pratt School of Engineering are perfecting electrical techniques that can already speed such microscopic droplets around the confines of tiny glass sandwiches at rates up to 1,000 Hertz, or cycles per second. Diameters of these droplets are sized at only 0.15 millimeters in diameter, and their volumes measure only about one billionths of a liter -- a "nano"liter -- each. "There is no reason we can't scale this down to "pico"liter (one trillionths of a liter) droplets, and propel them at hundreds of thousands of Hertz," said Fair, who is leading Duke's microfluidics research in collaboration with Nanolytics, a Research Triangle-area firm. An accounting of earlier research results, authored by Fair, his doctoral student Michael Pollack, and Alexander Shenderov of Nanolytics, appeared in the Sept. 11, 2000, edition of the research journal Applied Physics Letters. Readers can also get a look at such speeding droplets in motion by calling up the Web address www.ee.duke.edu/research/MONARCH/overview.html and clicking on the "movies" button near the bottom of the Web page. MONARCH, a three-year Defense Advanced Research Projects Administration-funded program that Fair directs at the Pratt School, is both evaluating the operating principles for building microfluidic systems with biomedical applications, and designing companion software systems to model these microfluidics architectures. Because these systems resemble the electronic arrays on microchips, other key investigators in the modeling studies include Don Rose, a veteran professor of computer science and mathematics; Krishnendu Chakrabarty, an assistant professor of electrical and computer engineering who is a very large scale integrated circuit designer; and Edward Shaughnessy, a professor of mechanical engineering and materials science who has expertise in fluidics. Like the general subject of nanotechnology (engineering at a scale of billionths of a meter), microfluidics is capturing attention as another effort to reap the technical advantages of smallness. It is because microelectronic circuits are very small that the fastest of today's computer workstations can perform nearly a billion operations a second. In an analogous way, the pace of laboratory analyses could be vastly sped up within a lot less space if sizes of fluid portions and the distances they had to move were ultracompressed. But microfluidics researchers are confronting a number of barriers, according to Fair. They have found that valves and pumps that efficiently move fluids around at normal dimensions do not scale down very well. Diffusion of liquids also becomes a problem, as does evaporation, the inability to precisely control apportionments of chemicals, and excess power consumption. A separate issue is the lack of standardization. "Today when people design microfluidics systems they do everything full custom," said Fair. "In other words, everything is done piece by piece, hand done pretty much. What we want to develop is a printed circuit board equivalent where microfluid droplets are moved around like bits of data." This approach, for which Fair said Nanolytics is seeking patent protection, uses no pumps, or valves but instead applies tiny, coordinated voltages of electricity to move microdroplets of fluids around on tiny arrays of electrodes. The advantage of that is that the engineers think they can use some well-tested principles of microelectronics to control where the tiny droplets go. In test apparatus installed in a laboratory near Fair's Hudson Hall office, microdroplets of a watery solution are surrounded by silicone oil to prevent evaporation. Both oil and droplets move within sandwiches of transparent glass. They follow channels bounded on the top and bottom by electrodes. Teflon insulators keep the electrodes out of direct contact with the fluids. The top electrode is a continuous strip of metal that connects the apparatus to the electrical ground. The bottom "control" electrodes are arranged as a series of separate units, each connected to an individual wire and each independently activated. The control electrodes are also positioned close enough to each other so that each microdroplet overlies more than one at a time. Engineers can manipulate the droplets by using electricity to modify their "surface tension," the property that can make a liquid behave as if it's encased within an elastic skin. Another key principal is that Teflon is a "hydrophobic" material -- one that water seeks to avoid contact with. When the power is off, each droplet maximizes its surface tension to minimize contact with the hydrophobic Teflon. In that case, the microdroplets "will bead up just like when you wax your car," Fair said. But turning on the electricity makes the space over an activated control electrode less hydrophobic. Since water "seeks" less hydrophic conditions, it will thus move toward that switched-on electrode. That means researchers can control the movements of the entire array's droplets by coordinating activations of all the control electrodes. Electrical control of surface tension, which the researchers call "electrowetting," also allows them to merge droplets together -- in essence mixing them -- by activating the appropriate electrodes. "This is a chemical engineer's dream," said Fair, because you have controllability, and instead of pouring in chemicals you are doing it drop by drop. This is drop by drop microchemical engineering."