Public Release: 

Aluminum shows strange behavior; research solves old mystery

Ohio State University




Full size image available through contact

COLUMBUS, Ohio -- Aluminum -- one of nature's best conductors of electricity -- may behave like a ceramic or a semiconductor in certain situations, according to an Ohio State University scientist and his colleagues.

Among the findings that appear in the current issue of the journal Science: When it comes to forming tiny structures in computer chip circuits and nanotechnology, aluminum may endure mechanical stress more than 30 percent better than copper, which is normally considered to be the stiffer metal.

The research expands scientists' fundamental understanding of aluminum on the atomic level, said Ju Li, assistant professor of materials science and engineering at Ohio State. It also solves a decades-old mystery about the metal's behavior.

Li began this work while a research scientist at the Massachusetts Institute of Technology (MIT). His coauthors on the paper include Shigenobu Ogata, associate professor of mechanical engineering and systems at the University of Osaka, Japan (then a visiting scientist at MIT), and Sidney Yip, professor of nuclear engineering and materials science and engineering at MIT.

The scientists used quantum mechanical equations to model the behavior of thin layers of aluminum and copper atoms under a condition called pure shear strain, when one layer of atoms slides over another.

Large shear strain is a common concern in very small electronic devices, where large temperature fluctuations cause materials to expand and contract. How materials deal with the strain in turn affects the reliability and durability of the devices.

While scientists have a relatively good understanding of how materials deform on larger scales, nobody can yet directly measure what happens at the most fundamental level -- when a one-atom-thick layer of aluminum slides over another, for example. Yet this information is critical for a deeper understanding of the material, Li said.

In their calculations, Li and his colleagues found that one layer of copper slid essentially horizontally over another layer below, agreeing with the classical picture of how shear strain occurs in this type of metal. But aluminum atoms tended to hop across, rather than slide. There were also related movements in the bottom layer of atoms, as if they were somehow connected to the atoms on top by an invisible set of hinges.

One explanation could be that aluminum atoms have a kind of "directional bonding" with each other, in which different atoms share sets of electrons, Li said. Directional bonding is observed in ceramics and in semiconductors such as silicon, but not in highly malleable metals such as aluminum.

Aluminum showed a definite edge over copper in the simulations. It proved to be 32 percent stronger than copper, and it endured much larger shear strains before it began to soften.

"These are pretty shocking conclusions," Li said. "We know copper is three times heavier than aluminum, and significantly stiffer than aluminum under normal conditions. But when we looked at large shear strains, aluminum won hands down. Copper started out stiffer, but it softened earlier than aluminum."

"Still, you wouldn't expect to find significant directional bonding in a metal like aluminum. This could mean that aluminum behaves more like ceramics in certain ways than anyone had previously thought," Li continued.

The calculations may also solve a long-standing mystery of so-called abnormal intrinsic stacking fault energy in aluminum. Metallurgists have long known that this abnormality has profound consequences for the mechanical behavior of aluminum, because it controls the structure of an important class of material defects called dislocations. But understanding of this behavior at the quantum-mechanical level is lacking, Li said. Directional bonding between the atoms could be responsible.

"From what we've seen, the high strength that aluminum shows under uniform shear strain and its abnormal intrinsic stacking fault energy may be two sides of the same coin," Li said.

This research could prove important for nano-indentation experiments, where scientists press a tiny diamond shard into materials in order to gauge how the material responds to extreme forces. Complex calculations are then needed in order to interpret the measured data.

"There are some approximations, or shortcuts, that make the calculations easier. Now it seems those approximations will likely give the wrong interpretation for aluminum and copper," Li said.

Finally, the work opens the door to a more accurate model of mechanical behavior in structures for nanotechnology. Tiny devices are currently under development which would sense electrical current and resonate in response -- a situation that would create a great deal of strain on the thin layers of material involved.

Li and his colleagues are continuing this work, and they've started to find that nickel exhibits behavior similar to aluminum, but to a lesser degree. "It's as if nickel occupies a spot on a spectrum halfway between aluminum and copper," Li said.

Meanwhile, Li has written AtomEye, a software program for visualizing the atomic structure of materials, particularly the changes in structure that accompany material strains. AtomEye images, including animations, can be seen on the Web at http://164.107.79.177/Archive/Graphics/A/.

###

Written by Pam Frost Gorder, (614) 292-9475; Gorder.1@osu.edu

Editor's note: an AtomEye image in JPG format is available to the news media by contacting Pam Frost Gorder.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.