Microfluidic devices have long been a staple in research settings for performing a variety of cell manipulation applications or for determining such material characteristics as viscosity, pH, and chemical binding coefficients. However, university research groups and companies alike are struggling to adapt microfluidic technology to portable diagnostic devices.
Featuring one or more channels containing at least one dimension less than 1 mm in width, microfluidic products are used to move common fluids such as whole blood samples, bacterial cell suspensions, protein or antibody solutions, and various buffers. Their advantages are manifold. Because the volumes of fluids within microfluidic channels are generally in the nanoliter range, this technology minimizes the accompanying use of reagents and analytes—substances or chemical constituents that are determined in analytical procedures. In addition, microfluidic devices can be fabricated using relatively inexpensive production techniques capable of producing elaborate, multiplexed devices. Finally, microfluidic technologies enable the fabrication of highly integrated devices that can perform several different functions on the same substrate chip.
Microfluidics is a mature technology, but its use in medical applications is still evolving. Thus, while the development of microfluidics for point-of-care diagnostics remains a work in progress, new and promising approaches are in sight.
From the Lab to the Living Room
Among the research groups working to transform microfluidics into a core technology for point-of-care portable devices is the Center for Microfluidics and Medical Diagnostics at the University of Notre Dame (Notre Dame, IN), headed by engineering professor Hsueh-Chia Chang. “Microfluidics is particularly useful for global health and epidemic control because it offers portable diagnostic kits,” Chang comments. “These kits may not have the sensitivity of a lab-bound device, but they could be carried around in the field and could be used by nonprofessional personnel.” The team believes that the chip could aid in controlling such infectious diseases as tuberculosis, malaria, and AIDS in the Third World.
The group's nanosensor-based microfluidic chips feature hard-polymer substrates that contain embedded microelectrodes fabricated using standard microcircuit manufacturing methods. The technology employs ac electric fields to separate, concentrate, and detect molecules such as DNA, RNA, and peptide biomarkers on the chip. “Our chip technology is suitable for global health and epidemic control applications because it does not require laboratory-based fluorescent labeling or optical detection techniques,” Chang says.
One of the main challenges facing researchers is developing methods for pumping fluids through microfluidic chips. To that end, the chip produced by Chang's group transports fluids using either an internal or external pump driven by a small battery. “The fluid sample in my chip is often pumped by an ac electroosmotic-flow (EOF) pump developed by us at Notre Dame,” Chang states. “Hence, the chip does not require an external pump. However, it is designed so that a handheld syringe can also pump the sample.”
Another research group devoted to expanding the application range of microfluidic technology is centered at the University of Washington (UW; Seattle). “At heart, the common factor among microfluidic devices is small channels and small volumes of liquid being moved around,” remarks Paul Yager, professor and chair of UW's bioengineering department. “That sounds completely simple-minded, but lots of technologies have been developed in the last 16 or 17 years for really making that come true.”
In the biomedical sector, microfluidics first came into its own as a drug-development technology. But the UW group is focused on developing it for use in devices that will enable patients to perform their own tests, send results to their doctors, and bypass centralized laboratories for many routine diagnostic procedures. “The aim is to use microfluidics as a set of technologies that could allow people to make complicated tests that normally require a lot of equipment and expert people and move them out of the periphery,” Yager says. “And I think that’s the area in which they have the greatest long-term potential.”
Early microfluidic-based point-of-care diagnostic devices included glucose monitors, which have become a standard commodity because they are inexpensive to make and they work well, according to Yager. But his team's aim is to enable microfluidics to perform a greatly expanded range of tests for everything from infectious diseases to drug monitoring. “Pretty much anything you can make into a fluid without big chunks in it is suitable for microfluidics,” Yager says. “We’ve worked on a variety of samples, ranging from saliva, blood, cerebral spinal fluid, tears. Anything you can liquefy you can monitor using microfluidics.”
While early microfluidic prototypes were based on silicon and glass, the UW group is working on a technology that is based almost entirely on polymeric laminate technology. The team has experimented with nitrocellulose membranes and a variety of other materials for the primary liquid movement, but its current focus is the use of polymethylmethacrylate and Mylar for the contacting material, which is held together using pressure-sensitive adhesives. These technologies form the basis of what the research group calls the Dx Box, a microfluidic device designed to work with plastic cartridges for performing aminoassays and nucleic acid amplification assays.
To perform assays, you’ve got to convert some physical quantity into a number, or a signal, Yager notes. This can be done using an electrical or physical sensor. An alternative method is to use different colors to reflect quantitative changes and read the changes using an external instrument. That’s the primary method that Yager’s team has focused on: to convert chemical concentrations into a pattern of color intensities that can then be easily measured using some type of external camera. “It’s not quite the same as having a sensor,” Yager says. “There may be a place in the device where color changes occur, and I guess you can call that a sensor, but you don’t need to have wires, and you don’t need to have a meter.”
Like Chang’s group at the Center for Microfluidics and Medical Diagnostics, Yager 's team is also concerned with optimizing fluid movement in microfluidic devices. “We achieve fluid movement using two pressures and a vacuum generated by small quarter-bolt vacuum and pressure pumps,” he explains. Like other microfluidic devices, the UW group’s device incorporates valves that rely on an external power source. However, the researchers are trying to transcend the need for external power, enabling the fabrication of a completely disposable chip that needs nothing more than a piece of commonly available technology to make it fully capable of performing quantitative measurements.
Small Devices, Big Challenges
Although many universities are engaged in microfluidic research, manufacturers such as Dolomite Microfluidics (Royston, UK) are turning research into reality. Dolomite’s chips are used in point-of-care applications such as drug-delivery and clinical diagnostic devices, but the company is also interested in developing chips for devices that perform therapeutic effects similar to dialysis.
|Micronit's glass microneedle arrays have locally insulated electrodes for neural signal recording and stimulation. Equipped with needles up to 7.5 mm long, this array has a cross-section measuring 200 to 300 × 100 ?m.|
Fabricating microfluidic devices from glass, quartz, and polymers, the company's chips have a good surface finish and accurate feature alignment, according to Richard Gray, the company’s head of sales. They can also be fabricated with both shallow and deep features ranging in size from 250 nm to 1 mm. For customers that desire complete instruments, Dolomite also builds in pumps, sensors, and valves and can develop full systems. “Imagine a series of concentric circles of microfluidic functions around the chip itself,” notes Gray. “We can do as few or as many of those circles as our customers need.”
In addition, Dolomite makes connectors—otherwise known as macro-to-micro interfaces—and integrates electrodes and sensors into the chips. “We put a lot of attention into the interface—particularly the electrical and fluidic interfaces,” Gray says. “So, we think about the connectors, which I think can be a weak spot in other systems where you end up with a relatively ungainly or impractical interface. In contrast, we have technologies that make that a fairly straightforward process.”
Many of Dolomite’s chips have edge connections; instead of containing a hole in the surface for fluid inlets and outlets, the chip has channels that are fabricated right to the chip’s edge. This configuration is advantageous because it can help to reduce costs by eliminating the need to create holes at right angles to the channel direction. It can also reduce the shear stress on the fluid, which could be useful for handling such fluids as blood. In addition, edge connections enable the company’s chip designers to make straightforward connections to multiple channels simultaneously.
Optimizing pumps and valves is as much a challenge for Dolomite as for other microfluidics companies and university engineering departments. Hence, the company is collaborating with university researchers to commercialize novel miniature pumping technologies based on thermal or electroosmotic techniques, Gray states. And to overcome problems with fluid flow in microfluidic channels, the company is working to develop internal valves using a sandwich of rigid and flexible layers made from glass, elastomer, and rigid plastic in a diaphragm arrangement. “We use that configuration for valves, but you can also use it for a diaphragm pump as well,” Gray adds. “We also offer a range of EOF and miniature peristaltic pumps, which, while not in-chip, are very small and can then be placed on-chip or near-chip.”
A Little Goes a Long Way
Scientists and manufacturers are interested in microfluidics because it consumes small quantities of fluids and can be automated, reducing test costs. It also enables the analysis of tiny samples, allowing tests to be performed with the same or less fluid than other diagnostic methods. In addition, it offers rapid analysis times, improved data quality, multiparameter testing, and reliable parameter control.
Capitalizing on these features, Micronit Microfluidics (Enschede, Netherlands) designs microfluidic products for portable medical devices that can perform near-patient testing. The company’s chips, explains Harmen Lelivelt, manager of marketing and sales, are constructed from layers of glass. Each layer has channels, holes, or electrodes, or a combination of all three. Fabricated using lithographic techniques borrowed from the semiconductor industry, these microfluidic chips are made on 4- to 6-in. wafers and have features that typically range in size from 5 to 500 µm. Such minuscule features are made using chemical wet etching, micropowder blasting, or metal deposition techniques.
Measuring less the 1 cm2, Micronit’s chips are designed for one-time use, eliminating the need for cleaning and the risk of cross-contamination. They work on the principle of capillary electrophoresis, a technique for separating substances from a fluid substrate that does not entail the use of moving parts for pumping fluids.
Battery powered, the chips enable users to perform analyses in less than a minute using a single drop of fluid. First, they separate molecules according to their charge/size ratio. Then, the molecules pass by a detector that measures their conductivity or fluorescent signal, resulting in a quantitative value.
Sounding a familiar refrain, Lelivelt remarks that current-generation microfluidic devices are constrained by their need for external pumps. “External pumps such as high-performance liquid chromatography–type pumps work well in automated lab instruments but take up considerable space,” he remarks. “While they are a good choice for benchtop instruments, they are not an option for handheld microfluidic devices.”
For single-use devices, capillary forces are a potential option because they are compact and energy efficient. But like Gray from Dolomite, Lelivelt believes that EOF pumping could be a viable alternative to using capillary forces. Employed in fuel cells, EOF pumps do not have moving parts, scale well, and can be powered by batteries or even cell phones. Other solutions, according to Lelivelt, include micromembrane pumps and disk-shaped microfluidic devices that use centrifugal forces to pump liquids.
Micronit’s chips are used in a range of applications, such as drug-delivery systems. “For this market, we have developed etching technology that is used to fabricate flow restrictors,” Lelivelt comments. Employed in implantable medical devices to control drug release, flow restrictors consist of a pressurized reservoir containing a relatively high dose of a drug and a fluid-restricting component. “Patients suffering from chronic pain, for example, receive morphine dosed at a certain rate by the microfluidic device’s pumps,” Lelivelt says. “Too little morphine fails to relieve pain, while too much can be fatal.”
Drug delivery is just one of many possible microfluidic applications. “This technology allows people to monitor their health at home and allows professionals to make fast decisions in time-critical situations,” Lelivelt says. “Thus, the healthcare system and society will benefit significantly from its ability to provide early diagnosis and reduced sample logistics.”