Metallic glass seems an oxymoron. Even the chemistry challenged know that the atomic structures of metal and glass are quite different. Molten metal has an amorphous, or random, atomic structure, but its atoms rapidly clump into a latticework of crystals as the metal cools and solidifies. Molten glass is also amorphous but has exceedingly sluggish crystallization. So even if molten glass cools and hardens slowly, it solidifies in an amorphous state. Glass is truly a frozen liquid.
So metal was metal and glass was glass until the early 1950s, when German scientists made amorphous metals by rapidly cooling vaporized tin and lead at a rate of roughly 1 trillion degrees Fahrenheit per second, a speed that fixed the jumbled atoms in place before they had a chance to form crystals. But these metals were less than one-hundredth the thickness of aluminum foil, “and they crystallized again when warmed to –40°F,” says Johnson. “The aim wasn’t to make anything practical. It was driven by scientific curiosity about how to avoid crystallization.”
The next breakthrough came in 1959 when Pol Duwez, a Belgian-born materials scientist at Caltech, used rapid cooling to make a gold-silicon alloy that remained amorphous at room temperature. It was the first true metallic glass. Again, it was just a gossamer film, because anything thicker could not be cooled quickly enough to “frustrate the crystallization,” as materials scientists say.
By 1990 Japanese researcher Akahisa Inoue and his team at Tohoku University, building on the theories of Harvard University researcher David Turnbull, began casting metallic glasses up to one-quarter of an inch thick. They found they could make “bulk” metallic glass by using three or more elements that differ from one another in atomic size by at least 12 percent. Alloys that meet these conditions solidify according to what’s called the confusion principle: The widely differing atomic radii and the high number of different elements “confuse” the atoms so they don’t know where to go to form crystals as they cool.
Enter William Johnson.
He is 55 years old, soft-spoken and affable but with a stealthy intensity. “Once I get going, I can be kind of overwhelming,” he says, and watching him command grad students with a few quick words, it is clear who is in charge of his lab. As an inquisitive teenager growing up in Bowling Green, Ohio, he wanted to become a particle physicist and unscrew the twisted mysteries of the universe. But at Caltech he crossed paths with quirky physics genius Richard Feynman. “Feynman gave me a long lecture about how if he was a young guy like me, he’d find some other area of physics because the high-energy stuff was overcrowded.” Practical by nature, Johnson had already begun to wonder if he could actually make something with atoms, as opposed to merely puzzling over ghostly traces of their constituents on a linear accelerator’s plates.
“Turns out there was a guy in materials science named Pol Duwez who was making amorphous metals,” says Johnson. Intrigued, he joined Duwez’s lab in 1970 and embarked on a lifelong study of the strange materials.
Even then, it was becoming clear that metallic glasses had useful yet otherworldly properties. Duwez had already created extremely “soft” magnets, meaning their north and south poles could flip easily.
“Somebody realized that if you put them in transformer cores, you could reduce transmission losses” by as much as 50 percent, says Johnson. The Allied Signal Company began doing just that in the 1970s, making thin sheets of rapidly cooled amorphous metal and rolling them into cores. “Today when you drive down a road and see a transformer bucket attached to a power pole, that’s a bucket full of amorphous metal,” says Johnson. “So it’s already out there, quietly, in the world.”
What really excited Johnson was the potential of thick, or bulk, amorphous metal. Analysis of the thin ribbons hinted that a heavy hunk of material, thick enough to be formed into structural shapes, would be like nothing seen on Earth before. Conventional metals dent, tear, and rust because of defects known as grain boundaries and dislocations, in which the crystals are pushed out of alignment and provide entry points for oxidation. Amorphous metals have no crystals that could be affected by such imperfections and hence are springy, extremely strong, and corrosionproof. But a truly thick metal, one with little surface area relative to its mass, would require little or no rapid cooling.
“For years people thought you could only make amorphous metals via rapid cooling,” says Johnson. “Even the Japanese were using it to make their thicker metals. But maybe it wasn’t essential.”
