Michael DeBar, a scientist with Eastman Kodak in Rochester, N.Y., examined the idea of a "thermocapillary pump" while at the University of California at Berkeley. He presented his work at Microelectromechanical Systems or MEMS 2002, which explores devices ranging in size from a small computer chip down to a few molecules.
DeBar's pump, built using standard computer chip-making processes, is based on a fluid effect caused by an imbalance in surface tension. Three heating elements straddle a channel about 100 microns across -- slightly wider than a human hair. To operate the pump, the middle heater boils a tiny amount of fluid and one of the other elements heats up slightly.
"When a vapor bubble is created and a temperature gradient is imposed, the resulting surface tension gradient will drive the flow from the hotter side of the bubble to the cooler side," DeBar said in his presentation.
The pump generates enough flow to deliver a useful dose of insulin, DeBar told United Press International. The pump could constantly send insulin into a diabetic's system or respond to signals from a blood sampling unit in an integrated medical device, he said.
"Either way, the whole idea is ... keeping the dosage level constant, which is better for the liver," DeBar said.
The project was sponsored by Kodak and the Defense Acquisition Research Project Agency, and explored the idea of a storable device for responding to chemical or biological attack, DeBar said. Antidotes would be in a powder form in the patch-like device; soldiers would add water and apply the patch.
Jesse Fowler, who studies the properties of microfluids at UCLA's mechanical and aerospace engineering department, told UPI liquid solutions such as insulin can sometimes have problems mixing at the microdose level.
A thermal-based pump could help maintain the proper mixture, Fowler said at the conference, but might present other problems. The heat necessary to create a vapor bubble in the insulin could destroy the essential molecules, he said.
DeBar said if heat becomes an issue in medical applications, the pump could be used to activate a gate controlling discharge from a separate reservoir, keeping the medicine cool. Since the pump can send fluid in two directions, one flow could act as a piston and force the gate open, while the opposite flow would close it.
The pump's other applications include inkjet printers, where very accurate flow control can reduce ink usage. Temperature damage is again an issue, but there are other ways around the problem, DeBar said.
"You could do electrolysis and get a bubble sitting there, that's actually more energy-efficient," DeBar told UPI. "When you're generating the temperature gradient (to cause a flow), you're not going up to a point where you're boiling the ink."
Another issue is the pump's response time, he said. The pump starts and stops within at least one-thirtieth of a second, DeBar said, but since that was the limit of the optical camera he used to monitor the pump, he is not sure how much faster it could work. Inkjet printers normally cycle on and off hundreds of times a second to form images.
While the pump is reliable and relatively easy to build with today's technology, DeBar said, it still has to be integrated into other parts of a drug delivery system. In particular, engineers are working out how to include micro-needles in such a device, he said.
Co-sponsored by the Institute of Electrical and Electronics Engineers and the Robotics and Automation Society, the MEMS 2002 conference displays the latest research into the molecule-sized systems. Scientists expect the technology to provide breakthroughs in
several areas, including medicine, computers and robotics, in the next few years.