Ohio team studies friction shortfalls of MEMS

Ohio team studies friction shortfalls of MEMS
By R. Colin Johnson , EE Times
April 30, 2001 (7:35 a.m. EST)
URL: http://www.eetimes.com/story/OEG20010402S0092

COLUMBUS, Ohio — An engineer at Ohio State University recently demonstrated the world's first method for directly measuring friction between tiny microelectromechanical-system (MEMS) parts. Using an atomic-force microscope, Bharat Bhushan, the university's Howard D. Winbigler professor of mechanical engineering, characterized the tribological shortcomings of current MEMS and showed how to bake a lubricant onto MEMS surfaces to minimize friction.

"Tribological engineering — the science of friction, wear and other powerful forces like adhesion — must make major contributions to the design of MEMS devices, if the field is to be successful. We must gain an understanding of the tribology of MEMS on the nano scale," said Bhushan. He was assisted by Ohio State mechanical engineering doctoral student Sriram Sundararajan as well as by chemists Huiwen Liu at Ohio State and Wolfgang Eck and Volker Stadler at the University of Heidelberg, Germany .

Tribological designs minimize friction and wear by lubricating all surfaces that interact by virtue of relative motion. However, because of the nanometer scales on which MEMS devices are built, the rules are different, and normal lubricants just seem to gum up the works. The reason, according to the Ohio State researchers, is that the laws of nature operate differently at nanometer-sized scales.

In particular, as sizes shrink from 1 mm to 1 micron, the area only decreases by a factor of a million, but the volume decreases by a factor of a billion — a ratio of 1,000. Thus, the laws of nature relating to area — such as friction, adhesion, meniscus forces, viscous drag and surface tension — become 1,000 times more influential with regard to the laws of nature relating to volumes — such as inertial and electromagnetic forces — merely because of the 1,000 area-to-volume ratio.

For instance, the friction between two smooth surfaces sliding over each other opposes the inertia of the objects. On the macroscale, inertia is often used to make sure a part fully engages its latches, such as when shutting a car door. On the nanometer scale, however, parts will have 1,000 times less inertia with which to fully engage their latches, because the weight, inertia and other volume-related parameters of nanometer-scale parts are 1,000 times less powerful, relative to area-related parameters like friction. Consequently, friction is 1,000 times more likely to prematurely stop a part from completing its designed motion on the nanometer scale, due to weak inertia.

Likewise, the forces of adhesion that must be overcome to put nanometer-sized parts into motion are 1,000 times more powerful on the nanometer-sized scale, relative to volume-related parameters. For instance, the force needed to "unstick" two smooth surfaces so that they can slide against each other — stiction — is also 1,000 times more powerful on the nanometer-size scale.

"A much larger lateral force is required to i nitiate relative motion between two smooth surfaces when they are very small," said Bhushan, "as I already knew from my study of the tribology of heads for magnetic storage systems, including static friction, wear and surface contamination effects."

To study the nanometer-size tribology, the team accumulated data from all the commercially available MEMS devices, such as Analog Devices' capacitive silicon accelerometer (airbag trigger), Texas Instruments' spatial-light modulator (digital micromirror array), various MEMS hard-disk heads and servo-controlled microactuators for ultrahigh track density.

According to the researchers, the current $1 billion market is mostly for airbag sensors, but new applications are sprouting up everywhere. For automobiles, MEMS acceleration sensors for antiskid braking systems and four-wheel drives are planned, as well as pressure sensors for monitoring the compression ratio in automotive engines and the pressure in tires. Hydraulic, pneumatic and other consumer product s are being designed, and aircraft designers are designing MEMS into cockpit instrumentation. Also, medical instrumentation companies are designing various sensors, actuators, motors, pumps and switches, all of which are sorely in need of tribological evaluation, according to the researchers.

All of these new applications, the researchers said, are being undertaken with no special lubrication, because there are no nanometer-scale lubricants available. "We need to develop lubricants and identify new lubricating methods that work well with MEMS — we need [to do] component-level studies to better understand the tribological phenomena occurring in MEMS," said Bhushan.

The research group made the first direct measurement of the friction between nanometer-size parts by employing an atomic-force microscope (AFM). Bhushan had been asked to help lubricate a new silicon micromotor being developed for biomedical applications by the Laboratory for Analysis and Architecture of Systems in Toulouse, France. F riction had prevented the motor's micron-size rotor from spinning around its central hub. By dragging the AFM across the surface of the rotor, Bhushan detected bumps on its surface and the surrounding casing, which ranged in size from 11 to 100 nm. "These bumps, left over from the chemical process that sculpted the silicon tiny rotor, were rubbing against bumps on the casing, causing friction," he said.

By making measurements twice as accurate as prior attempts without the AFM, Bhushan characterized the amount of friction inside the motor and thus gauged the amount of force needed to overcome it. "Other researchers have tried to calculate friction indirectly, from measuring the loss of electric current flowing in a MEMS structure, but we are the first to directly measure the frictional forces at play inside an MEMS," he said.

Bhushan first tried the exotic $1,000-a-quart synthetic lubricants that he used in his hard-disk research, but they only gummed up the motor even worse. The meniscus and other s urface-tension-related forces governing liquids made the lubricant act more like glue.

After much experimentation, Bhushan discovered a way of baking on the lubricant to form a 1-nm-thick coating that was not subject to the forces affecting liquids, because it no longer flowed. The solution got the micromotor working. The Z-DOL lubricant, manufactured by Monti Edison in Milan, Italy, was baked on at 150 degrees C. By baking all MEMS surfaces in Z-DOL, friction was reduced by half in the micromotor's components.

The team also tried using a thin coating of diamondlike carbon molecules, but the more-expensive process only reduced friction by 25 percent.