Gregg Nighswonger

November 1, 1999

15 Min Read
Micromachines:A Big Future for Small Devices

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI November 1999 Column

A quarter wafer containing 30 individual acoustic wave sensors fits neatly in an orange slice. How will methods for producing such microscale components influence the course of healthcare technology?

The continuing development of micromachining technology over the past few decades has had a profound effect on medical device development. Interest in microtechnology research has grown rapidly—in part because of the broad spectrum of potential engineering applications. Current research has seen significant advances made in chemical analysis systems, microinstrumentation for surgical use, physiological monitoring systems, and drug delivery technology. The long-range impact that such developments may have on healthcare has yet to be fully defined.

According to current estimates, 600 organizations are actively involved in development of microelectromechanical systems (MEMS) technology. One-quarter of those groups are also pursuing commercial distribution of finished devices or components. Although the global market for MEMS technology is believed to be approximately $8 billion per year, sales could reach as much as $20 billion by 2002. The market for microfluidics devices, which can be applied to a broad range of tasks such as medical diagnostic tests and screening of new drug candidates, appears to be undergoing particularly rapid growth, with increased use by pharmaceutical companies and researchers.


Microdevices are generally measured in terms of micrometers and are usually too small to be seen with the naked eye. Nanoscale devices, on the other hand, are built on the molecular or atomic scale and measured in nanometers. The creation of microscopic devices that can function as designed poses unique problems. Factors such as vibration and gravity that influence the performance of conventional motors and gear assemblies, for example, have virtually no influence over the operation of microdevices. The smaller devices, however, have a tendency to stick to one another, resulting in new problems related to the effects of stiction.

See sidebar: Microengineering Processes for Medical Technology


In addition to new diagnostic methods, one of the goals of micromedical research is to develop novel technologies for treating disease more effectively. According to an Ohio State University researcher, the same techniques being used to create computer microchips could one day provide the necessary technology for treating diseases such as insulin-dependent diabetes.

Mauro Ferrari, PhD, professor of biomedical engineering at Ohio State University (Columbus, OH), recently announced the development of a method for implanting silicon microcapsules—each one roughly the size of a pinhead—beneath a patient's skin. The capsules would be capable of carrying healthy transplant cells that would take the place of the patient's malfunctioning cells, producing needed chemicals for the body.

For example, says Ferrari, capsules carrying healthy pancreatic cells implanted in diabetics could allow those patients to produce insulin again. Conventional methods for treating diabetes using transplantation methods requires doctors to suppress the patient's immune system with drugs to keep antibodies from destroying the transplanted, healthy tissue. The technique, however, leaves patients open to infections. Ferrari states that hiss technique would eliminate the need for immunosuppression. In addition, it would eliminate the need to transplant an entire organ "when a few grams of healthy cells would do."

Insulin-dependent diabetes develops when the body fails to recognize cells in the pancreas that respond to glucose and produce insulin. Ferrari suggests that his technique would entail transplanting a few grams of healthy pancreatic cells by inserting a capsule containing the cells under the skin anywhere in the body Molecules of glucose would enter the capsule, causing the transplanted cells to release insulin that would subsequently flow back out into the patient's system.

Photonic lattice acts like a crystal in guiding light because of silicon "logs."

Ferrari explains, "If you can't beat the immune system, you hide from it." The technique he is developing with his colleagues entails using micromachining and nanotechnology methods to create microcapsules with holes that are "just the right size—big enough to let the needed chemicals get out, but small enough to keep the antibodies from getting in." Most recently, Ferrari and his colleagues were able to create 2.0-mm capsules containing "millions of channels" only 18.0 nm wide. His team simulated the capsule's use in treating insulin-dependent diabetes. He indicates that the 18.0-nm holes allowed glucose and insulin molecules to pass through, but successfully blocked the antibody immunoglobulin G (IgG), which attacks transplanted cells.

Ferrari notes that when other researchers have tried to construct capsules from plastic, the materials were incompatible with the body, and provided insufficient immunoisolation over long time periods. He and his colleagues chose silicon because of its biocompatibility, and because micromachining methods had already been developed for creating precise surface features with the material. A patented method of photolithography was employed to create holes that are precisely the same size, "down to an atom." Says Ferrari, "If we want to make 18-nanometer holes, we make 18-nanometer holes—millions and millions of them."

Three primary components of Sandia's Chemlab system fit inside a snow-pea pod. The components (from left to right) include a surface wave acoustic sensor array, a preconcentrator to absorb or adsorb chemical vapors, and a gas-chromatograph column.

Determining the optimum hole diameter was a challenge, according to Ferrari, because the actual size of the molecules involved was unknown. Initial tests revealed that both insulin and glucose could pass, though with difficulty, through 18.0-nm holes. The researchers then inserted a silicon membrane containing 18.0-nm holes between two chambers of liquid. One of the chambers contained IgG. The amount of IgG that penetrated the membrane was then measured over the next four days. After one day, the IgG concentration in the second chamber was less than 0.4%; the concentration grew to just over 1.0% after four days. Ferrari explains that this rate is several times slower than with capsules made of plastic, citing a similar study in which a perforated plastic membrane allowed an IgG concentration of 1.0% to accumulate in the second chamber after only 24 hours.

Ferrari emphasizes that keeping out 100% of IgG may be impossible because the molecule is not spherical. If the IgG molecule enters an opening with its narrow end, it can squeeze through. "These molecules twist and turn and find their way into small passageways—that's what makes them good antibodies."

Ferrari estimates that a single capsule should be capable of producing all the insulin a patient needs for an entire year and would cost about $20. Initial clinical trials using the new technique could begin within the next three years, he states.

The research is being supported in part by a recent $1.4 million grant from the State of Ohio Board of Regents to continue the work in a consortium. Consortium member institutions include Ohio State, Case Western Reserve University (Cleveland), University of Cincinnati, University of Akron (Akron, OH), the Cleveland Clinic Foundation (Cleveland), and the Battelle Memorial Institute (Columbus, OH).


At the recent Microsystems Expo hosted by Sandia National Laboratories (Livermore, CA), there was considerable attention focused on the development of lab-on-a-chip technology that has been made possible by the emergence of novel sensors. The completed system would, in essence, reduce the functionality of a complete chemistry lab to a single chip. Sandia indicates that initial applications of the device will undoubtedly be in various areas of national security, but future uses will encompass drug development and medical care.

Sandia's handheld µChemlab system is about the size of a palmtop computer but is capable of using both gas and liquid chromatographic techniques to separate the various constituents of complex chemical mixtures. In describing the system, Sandia explains that chromatography relies "primarily on differences in chemical interactions to separate different chemical species."

Chem Lab on a Chip: The top-level view shows push-button controls and viewing screen; midlevel encasements (left to right) perform computation and power management; and the bottom level includes (left to right) pumps, gas and liquid analysis channels, and batteries.

The device has a viewing screen and push-button controls. Samples drawn into the device by a micropump are concentrated on a spongelike structure. The sample is then routed through a series of channels where the mixture is separated into its basic components as it reacts with the walls or channel packing, according to Sandia. The result is that different compounds with the samples will come out of the channels at different times. A microcomputer analyzes and identifies the components and displays results on the device screen. Sandia indicates that the device can analyze the sample within about a minute and is capable of detecting samples weighing less than a single bacterium.

The system uses advanced microsensor technology to perform chemical analysis after separation of the mixture. Sandia has developed high-sensitivity detection technologies to provide analysis that has been found to be both versatile and accurate. The system incorporates miniaturized surface acoustic wave detectors, laser-induced fluorescence analysis, and electrochemical detection components.

In addition to creating miniaturized components and electronics, Sandia's research efforts have involved development of effective fluid-handling techniques, including extensive research into fluid dynamics within microsystems. Because fluids can exhibit unexpected behavior while being transported in microchannels, methods of precise controls are needed. The Sandia system uses electrophoresis, electric fields capable of precisely controlling the movement of fluids, as well as gases, in the microsystems.

Sandia's µChemlab is based upon use of engineered materials to provide the necessary functional characteristics. The system incorporates polymer materials with controlled porosity and high surface area and materials with controlled surface chemistry to provide the needed properties.


The increasing use of microscale devices for a variety of medical applications is being made possible by the advent of novel manufacturing techniques, including a range of laser ablation methods. According to Marvin Kilgo, PhD, senior research scientist at Tamarack Scientific Co. (Anaheim, CA), projection laser ablation is proving to be an extremely valuable enabling technology for commercial-scale manufacturing of the current generation of microfluidic devices. He explains that although such devices are now seeing use in a full spectrum of applications, their development has long been limited by difficulties in manufacturing. These limitations have been largely overcome, says Kilgo, but the challenge of commercializing these devices remains substantial.

Kilgo notes that excimer-laser ablation of polymeric materials has become a widely used technology for the generation of nozzles and through-holes for microfluidic devices. He adds, however, that ablation is also a viable process to create more-complex fluidic structures such as channels and manifolds. "Laser ablation is not limited to the creation of large arrays of small identical structures," says Kilgo. "The same technological strengths that make this technology useful for nozzle manufacture also can be applied to more-complex structures." Kilgo emphasizes that rapid throughput and a high degree of uniformity in producing either custom features or large-scale part production "make this a robust process for creating microscale devices in large quantities."

Kilgo states that material selection and mask design "have been identified as important preliminary components of process development for projection ablation." He adds that "the selection of target material is perhaps the most important aspect of ablation process design." Although material suitability is most often determined by a thorough evaluation of end-use considerations, the material chosen must have an easily attainable ablation threshold in order for projection ablation to be a viable process. Kilgo identifies polyimides and polyesters as commonly used, easily attainable materials. Polyimide is widely used, in part because it is opaque to UV light. Teflon, on the other hand, is not opaque. He notes that even polycarbonate has good and bad characteristics that must be considered when assessing its use for laser ablation applications.

When asked to identify potential challenges to projection laser ablation methods, Kilgo responds that, "in general, the only limitation for creating features of arbitrary complexity is the quality of the feature that is defined by the mask." He explains that any flaw in the mask will be reproduced precisely at scale in the finished component. That is to say, the better the mask, the better the finished device.

Looking to the future of the technique, Kilgo notes that projection laser ablation is an affordable technique that can provide high throughput for manufacturing microfluidic devices at commercial volumes. In addition, he suggests that the method is expected to prove useful in the manufacture of microassays and delivery systems.


Some of the research focusing on miniaturized systems for either therapeutic or diagnostic applications is attempting, to some degree, to imitate natural functions. Ferrari's implantable capsules, for example, are an attempt to use the body's own insulin-generation system to treat the effects of diabetes. Similarly, researchers at Boston College (Chestnut Hill, MA), have designed a working, chemically powered molecular motor.

The prototype molecular motor, designed and constructed by T. Ross Kelly, PhD, and coworkers, emulates processes used in nature to convert energy to motion. In humans, for example, molecular motors are used in a variety of functions, including muscle contraction, intracellular transport, and sperm motility.

"This is significant from two points of view," says John Schwab, PhD, of the National Institute of General Medical Sciences (NIGMS; Bethesda, MD), which supported the work. "One is that it proves we understand, at least in some sense, how nature might convert chemical energy into controlled motion."

Schwab emphasizes that Kelly's research represents an extraordinary example of miniaturization of technology. "This is going orders of magnitude beyond nanotechnology, which we can visualize using optical microscopy or electron microscopy. We're taking it all the way to the single molecule level, which is quite exciting," Schwab explains.

It is possible that the research could ultimately increase understanding of the molecular motors in structures such as muscles and cilia. The researchers suggest that it could improve knowledge of diseases in which molecular motors are faulty, such as certain cases of infertility, and specific respiratory and digestive disorders.

Kelly's nanomotor is an organic molecule made from fewer than 50 carbon atoms. Operating as a unidirectional ratchet, it is powered by carbonyl dichloride. In nature, molecular motors are larger protein molecules and are fueled by the universal unit of cellular energy called adenosine triphosphate (ATP). Kelly indicates that, because the device mimics the ability of molecular motors to convert chemical energy into ATP, it may eventually help researchers understand the natural process at an atomic level. The current research by Kelly and his group is focused on modifying the motor to operate faster and to run continuously.


There has been a noticeable increase in the efforts being made to bring functional lab-on-a-chip systems to market. Recently, Caliper Technologies Corp. (Mountain View, CA) and its commercial partner, Hewlett-Packard Co. (Palo Alto, CA), began marketing the first LabChip-technology-based microfluidic instrument system to serve the research products market.

The desktop HP2100 Bioanalyzer system has been designed initially to perform nucleic acid—analysis applications on a microfluidic chip and is the first commercial system developed by the Caliper and Hewlett-Packard collaboration. Using miniature, integrated biochemical processing systems etched into glass, silicon, quartz, or plastic, Caliper's lab chips allow the same steps customarily performed in conventional instruments to be done in minute quantities at small fractions of the usual elapsed time.

Says Dan Kisner, MD, Caliper's president and CEO, "We believe that our LabChip technology has the potential to significantly increase the productivity and quality of scientific experimentation across a broad range of industries, by miniaturizing, integrating, and automating critical functions."

The LabChip technology is based on use of miniature chips with micromachined fluid channels capable of performing a broad range of laboratory analyses. According to the firm, the system uses advanced miniaturization, integration, and microfluidic technologies to function like a liquid integrated circuit, performing multiple, integrated laboratory functions in seconds using nanoliter volumes of biological or chemical materials. By applying voltages to various channel intersections, the chip moves an analyte through the system, adjusting its concentration across three orders of magnitude, mixing it with buffers, separating out the constituents, adding fluorescent tags, and directing the sample past detection devices.

"It's really a microrobot," says Caliper's Mike Knapp. "It's actually doing the experiment. It's not a real-time replacement for a pipette, it's a robotic system working on a nanoscale." Caliper believes that the LabChip system could potentially be used to perform a wide variety of applications in various industries, including research products, pharmaceuticals, chemicals, diagnostics, and genomics.


The future of micromachining in medical device applications—ranging from catheters incorporating miniaturized monitoring probes to chip-based analyzers—will depend to a great measure on the degree of collaboration between research centers and industry.

Researchers at Sandia suggest that "intelligent integrated microsystems are the next logical step in the silicon revolution." They note that the integrated circuit industry of the past three decades has succeeded in achieving an exponential increase in the number of transistors that can be incorporated onto a tiny piece of silicon. "Over the next 30 years, the incorporation of new types of structures will enable the chip to sense, act, and communicate, as well as think. One key to this vision is the integration of all the machine functions onto a single device, mass produced, and using a single manufacturing process." They emphasize, however, that industry will be a key player in the development process. The reliability of microsystem components must ultimately be tested by industry. To do that, Sandia must work with industry to develop and commercialize microsystem technologies.

Other elements that could have a significant influence on microtechnology will be continued development of appropriate fabrication techniques, such as LIGA and laser ablation, and advances in materials. The successful integration of complete microsystems, however, has shown clear potential for changing the nature of both diagnostic and therapeutic medicine. Evidence of that potential, which has been demonstrated in a number of research laboratories, is now expected to establish an equal presence in the marketplace.

Gregg Nighswonger is executive editor of MD&DI.

Lead photo by Randy Montoya

Back to the table of contents

Copyright ©1999 Medical Device & Diagnostic Industry

Sign up for the QMED & MD+DI Daily newsletter.

You May Also Like