Researchers Pioneer Techniques To Lubricate Microdevices

Until now, scientists couldn't accurately measure the friction that plagues miniature motors, pumps, and gears -- mechanisms that could one day move inside microscopic medical implants in the body, for example. Such devices -- microelectromechanical systems (MEMS) -- contain parts so small they are measured in nanometers, or billionths of a meter. Without knowing the friction present in these devices, scientists can't design the right micro-lubricants.

Led by Bharat Bhushan, Ohio Eminent Scholar and Howard D. Winbigler Professor of mechanical engineering at Ohio State, researchers here have pioneered the first direct method for measuring the friction of these tiny parts as they rub together -- with results twice as accurate as any previous indirect method could provide.

They also found a way to bake lubricant onto the surface of microdevices at temperatures as high as 150°C to oil the tiny moving parts.

An overview of this research appears in the Proceedings of the NATO Institute on Tribology this month, and two related papers will appear in the July-August 2001 issue of the Journal of Vacuum Science and Technology.

"Before we can design microdevices that actually work, we must understand how friction, wear, and other powerful forces operate on a tiny scale," said Bhushan.

Bhushan's cohorts on these projects include Ohio State mechanical engineering doctoral student Sriram Sundararajan, Huiwen Liu, and Wolfgang Eck and Volker Stadler, both physical chemists from the University of Heidelberg, Germany.

Normally, Bhushan's laboratory studies the texture of micro-thin coatings on computer hard disks and data tapes -- items for which friction, wear, and lubrication dominate performance.

He first began studying microdevices after organizing a 1997 National Science Foundation (NSF) conference on tribology and MEMS. NSF has since funded his research in this area.

In 1999, scientists at the Laboratory for Analysis and Architecture of Systems in Toulouse, France, asked Bhushan to help lubricate a new silicon micromotor they were developing for biomedical applications. Friction was preventing the motor's tiny eight-arm rotor, a kind of miniature propeller, from spinning around its central hub.

For this work, Bhushan and his colleagues found a new application for a commercially available tool already sitting in their laboratory: an atomic force microscope (AFM), which records the shape of an object by dragging a tiny needle over its surface.

"Since we were already using the AFM to investigate data storage devices, it seemed natural to try new applications," Bhushan said. "Friction and wear are big problems for MEMS, so I thought, why not us an AFM to study them?"

A typical AFM needle has a radius of only 50 to 100 nanometers -- a fraction of the width of a human hair. To the needle's sensitive touch, bumpy landscapes only a few atoms thick feel as expansive as mountains and valleys.

The AFM needle detected bumps on the surface of the rotor and the surrounding casing. The bumps, which ranged in size from 11 nanometers to 100 nanometers, were a normal result of the chemical process that shapes the tiny parts out of silicon, Bhushan said. Bumps on the rotor were rubbing against bumps on the casing, causing friction.

These measurements were twice as accurate as any produced by other means, Bhushan said.

The researchers were also able to gauge the amount of friction inside the motor by measuring the force required for the needle to nudge the central rotor into motion. Other researchers have tried calculating friction from indirect measurements, such as the loss of electric current flowing over a MEMS structure, but this is the first time anyone has directly measured the frictional forces at play inside a working microdevice.

To grease the motor, Bhushan first tried flooding the device with Z-DOL, a commercially available synthetic lubricant that costs nearly $1,000 per pound. Z-DOL is manufactured by chemical company Monti Edison in Milan, Italy, and Bhushan had already purchased a supply of it for his hard disk research.

Liquid Z-DOL only gummed up the motor, however, so the researchers tried something different -- they baked a 1-nanometer-thick coating of lubricant onto the surfaces of all the moving parts. At 150°C, the Z-DOL solidified into a smooth layer that allowed the components to move more freely.

When the researchers again nudged the rotor into motion, they found that the solid Z-DOL coating reduced friction by half. These results will appear in detail in one of Bhushan's two Journal of Vacuum Science and Technology (JVST) papers.

The second JVST paper concerns Bhushan's collaboration with physical chemists at the University of Heidelberg.

The Heidelberg researchers had already developed a thin coating of diamondlike carbon molecules to lubricate MEMS, and they were wondering whether a more complex -- and more expensive -- arrangement of molecules would result in a better coating. The arrangement is referred to as "cross-linked," since the atoms that form the molecules are linked together with strong bonds.

The Ohio State researchers found that the fancier coating wasn't better. By dragging the AFM needle across coated samples of silicon, they found that the simpler coating reduced friction twice as well as the cross-linked coating.

Bhushan will continue this work. His latest project involves measuring the amount of force tiny construction beams of silicon can withstand before they break. Such beams might support complex MEMS structures in the future.

This research is supported by NSF's Division of Electrical and Communications Systems.