Tiny Metal Libraries a Big Tool for Materials Scientists

July 6, 1995

Mike Wooldridge,

BERKELEY, CA--The pace at which researchers can create and test new advanced materials should be quickened from the current rate of about one compound a day to as many as 10,000 with a technique developed by scientists at the Ernest Orlando Lawrence Berkeley National Laboratory.

Working at the Berkeley National Lab's Molecular Design Institute, the researchers have found a way to deposit thousands of distinct combinations of metal-oxide molecules onto an area the size of a checkerboard square, thereby creating the first "combinatorial" library of advanced materials. The technique uses "masks" to stencil thin layers of metal-oxide ingredients onto different columns and rows of a inch-square grid. It has already been used to create libraries of high-temperature superconductors.

The research was described in a cover story in the June 23 issue of the journal Science by Berkeley National Lab physicist Xiao-Dong Xiang and chemist Peter Schultz, who is also a professor of chemistry with the University of California at Berkeley.

The combinatorial strategy is a departure from so-called rational design, where researchers try to predict beforehand which molecular structures will yield desired properties. The new strategy relies more on sheer numbers -- making a myriad of theoretical candidates and sorting through the bunch to find a solution.

The combinatorial approach has been used successfully by researchers in the life sciences to screen for potential drugs. This is the first time scientists have applied it to solid-state materials.

A high-volume strategy makes sense considering the wide range of possibilities scientists are faced with when concocting new materials. Just the five-element combinations available on the periodic chart represent millions of potential compounds. Metals can be combined at different ratios; they also form different molecular structures depending on the temperatures or pressures at which they are treated. The result is an enormous advanced materials "universe," much of which scientists still know very little about.

"In the vast majority of cases, the important discoveries in materials science are serendipitous." Xiang says. "It is very difficult to predict how these complex structures will behave before you make them." Until now, scientists have had to resort to making new materials by trial and error one at a time.

The combinatorial approach has long been known as a way in which the human body solves problems. The immune system, for instance, has a library of about one trillion differently shaped antibodies, each made up of different combinations of protein chains. When faced with an invading agent, such as a virus, the immune system selects the antibodies from the trillion that happen to bind to the virus. Their numbers are multiplied to fight the infection.

The key point is that there is no rational design of antibodies to fit a particular virus. The strategy is to create a wide variety of combinations that you can select from after the fact.

Previously, Schultz has borrowed the immune system's combinatorial strategy to invent "catalytic" antibodies -- antibodies that, because of their shape, promote certain chemical reactions. Schultz receive the 1995 Wolf Prize in Chemistry for his discovery of catalytic antibodies. Similarly, biotechnology researchers have used a combinatorial approach to select potentially useful drugs from large libraries of randomly generated molecules.

To apply the same philosophy to materials design, the researchers turned to thin-film technology, a method in which very small amounts of a complex metal material can be manufactured quickly. Metal components are laid down atop one another, each layer 10 to 100 angstroms in thickness. The layers are heated to mix the metal elements and create a stable, composite compound.

The researchers create arrangements of different metal combinations by depositing the thin films through masks. A so-called primary mask, with square openings like those on a screen door, is used to lay the metals down as a grid of separate squares. A secondary mask, when laid atop the primary mask, serves to block out specific rows or columns of the grid . By sending the metals through different secondary masks, each can be deposited on particular sections of the grid.

To test the technique, the researchers first created a small, 16-member library of copper-oxide high-temperature superconductors. The researchers laid down a base layer of copper through the primary mask then deposited layers of four other metals -- oxides of bismuth, calcium, lead, and strontium -- through different secondary masks. Because of the sequence of masks, each site on the grid received a different combination of metal oxides and every possible combination was represented once.

After heat treatment, the researchers tested the electrical characteristics of the materials in their library. Results showed the copper-oxide thin-films had the same resistive characteristics as superconductors created by large-scale means. The researchers went on to successfully make denser, 128-site superconductor libraries with combinations of seven metals.

Tests with decreasing mask sizes showed that the method could produce working superconductor films as small as 200 microns by 200 microns. At scales below 200 microns, metal-oxide molecules in the thin layers begin to evaporate.

"With 200-micron sites, we can realistically deposit 10,000 materials in a square inch area," Schultz says. "That would effectively increase the rate at which we can test new materials by a factor of 10,000."o

"We are pleased that the approach works so well for high-temperature superconductors," Xiang says. "Now we are also applying it to important classes of magnetic and optical materials."

Lawrence Berkeley National Laboratory is a U.S. Department of Energy national laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California.