Microfluidic devices take advantage of the chemical properties of liquids and gases and the electrical properties of semiconductors and combine them on a single chip. The minuscule systems, whose dimensions are measured in nanometers, perform tasks that range from detecting airborne toxins to analyzing DNA sequences. The devices offer several advantages over conventionally sized systems, including compact size, disposability, increased functionality, and they require smaller volumes of reagents and samples.
Microfluidics is booming in both academe and industry. "Most of the major research institutions -- including Berkeley, Caltech, M.I.T., the Univ. of Michigan, and Stanford -- have groups working in this area," said Harold Craighead of Cornell Univ. (Ithaca, NY), himself a researcher in the area. "We're also seeing commercialization of the technology for more rapid DNA sequencing, chemical analytical systems, manipulating cells, and doing general biological procedures."
A key factor in the optimism surrounding the future of microfluidics is that fabrication of the devices involves techniques compatible with those already used in batch processing standard semiconductors. "That should allow future microfluidic devices to be made quickly and cheaply in a microchip factory," said Carolyn Matzke of the Compound Semiconductor Research Lab at Sandia National Labs.
Matzke and her colleagues have developed and patented a technique for creating tiny hemispherically shaped canals through which microfluids can move on the surfaces of chips. The traditional method of producing canals involves using high heat to bond two pretrenched wafers. However, the high temperatures -- up to 1000 deg. C -- required for that process can damage other features on the chips.
The Sandia researchers' approach avoids excessive damage to other components on the chip and works more than ten times faster. They pattern a thin layer of photoresist on the surface of a wafer by shining light through a conventional photomask, and develop areas of the photoresist exposed to the light. That leaves a design of photoresist ridges on the wafer's surface that are destined to become the interiors of the canals. Then the team heats the wafer to a relatively benign 100 deg. C for about 20 seconds to cause the square-edged ridges to clump into hemispherical shapes. Finally, technicians deposit a 2-µm film of silicon oxynitride over the rounded photoresist and soak the entire wafer in acetone to dissolve the remaining photoresist.
The result is a series of hollow tunnels on the wafer's surface, each between 8 and 100 µm in diameter (see figure). The technique permits the researchers to twist the channels into curved shapes, including hairpin bends. The curves have a filtering effect; they help to separate gases and liquids into their individual components by allowing them to pass at different speeds. "You could envision an elaborate microfluidic network on a chip that sorts, detects, and identifies individual chemicals or proteins," said Sandia scientist Terry Michalske.
Typically, scientists use electricity to drive samples through the channels. In one form of stimulation, electro-osmosis, computer-driven power supplies located in reservoirs at each end of a channel are activated to generate electrical current through the channel. The current forces molecules with different electrical charges in fluids to travel through the channels at different rates; a typical speed is one millimeter per second. Another type of electrokinesis -- electrophoresis -- is simply a microminiaturized version of an analytical approach routinely used in chemistry laboratories. An electric field influences the movement of charged molecules in microfluids moving through the channels. Scientists can use electrophoresis to move molecules in solution or to separate molecules with very subtle differences.
Electrokinesis isn't the only method of choice for driving microfluids. For example, small amounts of pressure applied to microfluids traveling through channels create extremely predictable and reproducible flow. And last year Princeton Univ. researchers Sandra Troian and Dawn Kataoka (who has since moved to Sandia) showed that temperature gradients can move tiny volumes of liquids around nano-sized canals in silicon wafers. This occurs because the surface tension of the microfluids varies with temperature; a temperature gradient of just three or four deg. C is enough to cause a microfluid to seek a cold region in its pathway. Chemical modifications to the substrate applied by lithography amplify the effect by creating a series of chemical levees along the canals.
Microfluidic technology has made its most significant mark in biological chips designed to facilitate identification of DNA sequences. "The largest part of microfluidics is in DNA assays at present," said Carlos Mastrangelo of the Univ. of Michigan, who is currently on sabbatical at IntelliSense Corp. (Wilmington, MA). Craighead said, "We're seeing commercialization of more rapid DNA sequencing, chemical analysis systems, and systems for manipulating cells and doing other biological procedures."
The microfluidic approach has had relatively little influence on the Human Genome Project, the most ambitious application to date of DNA sequencing technology. But several companies are now developing microfluidic systems that will take on the next series of sequencing challenges, which will involve more detailed explorations of the human genome and others. Several young companies have already started to develop proprietary chips that will use microfluidics to face this daunting challenge.
Two relatively new firms illustrate the commercial promise. Caliper Technologies, Inc. (Mountain View, CA), which went public last year, has devised a "lab on a chip" technology that uses both computer-controlled pressure and electrokinetic forces to gain exact control over the flow of fluids in microfluidic channels. The company says that its complex channel designs permit standard laboratory bench procedures such as mixing, incubating, and separating to be carried out in volumes as small as a nanoliter. Cepheid, (Sunnyvale, CA), a company founded in 1996, is putting together a portfolio of proprietary technologies designed to automate complex DNA detection procedures, from preparing and processing samples to analysis and real-time detection. The company has recently received a joint award from the Defense Advanced Research Projects Agency to develop an integrated microfluidic system that will automatically detect biological agents and genetically engineered variants of them.
A former science editor of Newsweek, Peter Gwynne is a free-lance science writer based in Marstons Mills, MA.