For decades computer users have relied on hard disk drives to store their digital data. And thanks to our technologists, every year or so we get twice as many zeroes and ones onto the same amount of space on a disk surface.
Improvements in storage density are crucial not just to college students with enormous digital music and video collections, but also to companies such as Google Inc. that are building the physical infrastructure of the "cloud" in the form of massive data centers around the world. More storage density enables a better, bigger and cheaper cloud.
But in recent years, researchers have begun to forecast an end to the exponential gains seen in digital storage unless new methods can be developed to write the data on the drives.
That's where chemistry professor C. Grant Willson, along with chemical engineering professor Chris Ellison and a team of undergraduate and graduate students, may be able to help. They've developed a technique that could help increase the storage capacity of hard disk drives by a factor of five.
The Problem: Too Many Dots, Too Little Space
With current production methods, zeroes and ones the basic language of digital data are written as magnetic dots on a metal surface. The closer together the dots, the more information that can be stored in the same area.
But that tactic has been pretty much maxed out. The dots have now gotten so close together that any increase in proximity would cause them to be affected by the magnetic fields of their neighboring dots and become unstable.
"The industry is now at about a terabit of information per square inch," Willson says. "If we moved the dots much closer together with the current method, they would begin to flip spontaneously now and then, and the archival properties of hard disk drives would be lost. Then you're in a world of trouble. Can you imagine if one day your bank account info just changed spontaneously?"
But there's a quirk in the physics: If the dots are isolated from one another, with no magnetic material between them, they can be pushed closer together without destabilizing.
"That's the good news," Willson says. "The bad news is that these dots are really small. I mean really small. They are smaller than a virus. So standard processes cannot print them anymore. So how do you do it?"
One obvious place for Willson to begin was with nanoimprint lithography, an area in which he's been a pioneer. (In 2008 Willson was awarded a National Medal of Technology and Innovation for his work in this area; this year he's being honored for it, along with mechanical engineering professor S.V. Sreenivasan, as an OTC Inventor of the Year.)
The technique (which was described in an article in Science last month) is a nano-scale version of the process that has been used for centuries to reproduce lithographic prints. Etch a master plate, and then run off duplicates.
When Willson began working on this project, the only proven method of etching a master at such a small scale was to create the dots one by one with focused beams of electrons.
"It took about a month and cost about a million dollars," Willson recalls. "It wasn't workable from a production standpoint. What might be workable, however, was if you could write only every fourth dot and find something else that would fill in the rest of the pattern for you. Then you'd be down to a week of writing time."
The Answer: Directed Self-Assembly
This is where block copolymers come in. At room temperature, coated on a disk surface, they don't look like much. But if they're designed in the right way, and given the right prod, they'll arrange themselves into highly regular patterns of dots or lines. If their target surface already has some guideposts etched into it, the dots or lines will form into precisely the patterns needed for a hard disk drive.
This process, which is called directed self-assembly (DSA), was pioneered by engineers at the University of Wisconsin and the Massachusetts Institute of Technology.
So far so good, in theory. But when Willson, Ellison and their students began working with directed self-assembly, the best anyone in the field had done was to get the dots small enough to double the storage density of disk drives not enough of a gain for industry to invest in the complicated, messy and slow process.
The challenge has been to shrink the dots further and to find processing methods compatible with mass production. The team has made great progress on several fronts. They've synthesized block copolymers that self-assemble into the smallest dots in the world. In some cases they form into the right, tight patterns in less than a minute, which is also a record.
Most significantly, the team has designed a special top coat that goes over the block copolymers while they are self-assembling. It allows them to achieve the right orientation relative to the plane of the surface simply by heating.
Leon Dean, a senior chemical engineering major who's been working on the project since he was a sophomore, remembers the day when the top coat really came together.
"Honestly, I was getting frustrated," says Dean, a co-author on the Science paper. "We had tried numerous experiments, but none of them seemed to improve our results. Then one day we tried a new top coat, and I went downstairs to the scanning electron microscope to look at how the block copolymer had assembled. For the first time I saw perpendicular lines everywhere on the surface of the film, and I knew we were onto something pretty big. I had to run upstairs right away and tell Chris [Bates], who was the grad student supervising me. It was exciting."
For Willson, the significance of the research lies as much in the opportunity it's provided for students like Dean to shine as it does in the findings themselves.
"I am amazed that our students have been able to do what they've done," he says. "When we started, I was hoping we could get the processing time under 48 hours. We're now down to about 30 seconds. I'm not even sure how it is possible to do it that fast."
For the students, an article in Science is a rather extraordinary way to launch their careers. Dean plans to pursue a doctoral program in chemical engineering. He expects the publication and the years of lab experience will set him up to excel.
And then, of course, there are the hard drives. Julia Cushen, a doctoral student of Ellison's and one of the co-authors on the paper, is working with California-based HGST, one of the world's leading innovators in disk drives, to see whether the techniques developed in Austin can be adapted to their products and integrated into a mainstream manufacturing process.