Portland, Ore. - As microelectromechanical systems scale down in size, common forces such as surface tension and friction become more dominant. At Sandia National Laboratories (Albuquerque), friction at the nanoscale is getting close scrutiny that has turned up some surprises, including the discovery of a previously undetected adhesive force.
"Our goal is finite-element simulations of MEMS components that accurately predict response in the presence of adhesion and friction," said Dave Reedy, who heads the High-Fidelity Friction Models for the MEMS program at Sandia.
Sandia researcher Maarten de Boer recently announced the early results of a program that uses an "inchworm" (a common device for making precise measurements), downsized for MEMS, as a measuring stick to characterize friction between MEMS surfaces. De Boer found hitherto undetected adhesive forces that result from nanoscale features.
"It turns out that at very light loads there is an adhesive force distinct from the applied load that has to be taken into account," said de Boer. The force should be characterized and added to MEMS designers' engineering models, he said.
Macroscopic friction follows Amonton's Law-that friction is proportional to the force perpendicular, or normal, to a surface. But other forces of adhesion, aggravated by the larger surface-to-volume ratio of MEMS devices, become increasingly significant as devices shrink.
De Boer theorizes that the newly discovered adhesive force may be due to nanoscale asperities-high points where two surfaces actually touch, surrounded by gaps between surfaces as big as 100 to 200 nanometers. "Maybe there is some kind of collective action among the asperities,"he said.
De Boer is working with Rob Carpick, a nanotribology expert and professor at the University of Wisconsin, to associate Carpick's nanoscale measurements of asperities with the micron-scale measurements from de Boer's inchworm. The pair hope to explain the hundredfold difference between a standard MEMS model's predictions and the actual measured results with de Boer's inchworm.
In particular, the inchworm's measurement of "gross slip"-the distance two contacting surfaces will move before breaking free of static friction and sliding freely-should have been 1 to 2 nm using conventional models, but the measured gross slip was 100 to 200 nm.
"At this scale, a gross slip that moves something a hundred nm could be a big deal," said de Boer. "For instance, it could throw off your alignment if you are doing optical control."
According to de Boer, friction is usually considered a deleterious effect, especially at the nanoscale, where the surface-to-volume ratio increases as MEMS shrink.
"Anytime one MEMS surface rubs another MEMS surface, you get this extra drag force that you really don't want," he said.
To characterize how the forces of adhesion change friction for MEMS devices, de Boer's inchworm had to measure both high and low ranges of force. De Boer layered parallel MEMS plates, nudging one with a flex-drive stepper motor up to 80,000 times a second, each time moving the plate a single step. For the test regime, Sandia postdoctoral researcher Alex Corwin created a measurement methodology that "inched" the plate along in 40-nanometer steps at up to 3 mm/second.
Forces were measured by having the inchworm push against a load spring that became exponentially stiffer with each step, resulting in a tangential force of up to 2.5 millinewtons-250 times the force possible with a typical MEMS comb drive.
De Boer suspected that the leading clamp was not staying stationary but was corrupting the measurement with a gross slip that was larger than normal because of unknown adhesive forces. In general, such forces are called stiction: When two surfaces rest against each other, they tend to get stuck after a while.
After careful measurements, de Boer's research group characterized the new adhesive force as distinct from both the applied load and from van der Waals forces, and showed how the force becomes increasingly significant for light loads. Gross slip for light loads was measured to be 100 times bigger than the macroscale effect.
"The biggest thing that we would like to understand better is this issue of gross slip," said de Boer. He plans to characterize tangential loading more carefully with the hope of identifying the phenomenon that controls the 100x increase in that effect.
"We don't understand that right now," said de Boer. "We have some geometric arguments as to why it may be happening, and some materials-related arguments, and we are trying to develop tests that will allow us to figure out which is the most likely of these possible reasons."
Worm on the street
De Boer's inchworm has proved useful to other researchers in related fields, according to Sandia. For instance, Sandia staff member Marc Polosky made use of it in a prototype MEMS chip that performs mechanical-logic functions.
"On the MEMS scale, stiction forces can degrade system performance," said Polosky. "The inchworm provides a very high-output force and can be moved controllably both backward and forward."
Separately, Sandia staffer Dustin Carr has evaluated the inchworm for precise positioning and control of micro-optics.