RELATED
ARTICLES

Nanoimages

Why People are
Watching
the World's
Tiniest Tubes

Spreading it Around

Gearing up for
Nanomachines

Liquid Memory



LINKS

The
nanoManipulator

The Whole Elephant > 1 > 2 > 3 > 4

Pictures from the Dark Side

As Taylor and his mentors in computer science made progress on the nanoManipulator, Rich Superfine began to realize that he needed it. “It became clear to us,” he says, “that the manipulations we were trying would be very difficult to do sitting at a keyboard, using keyboard commands to push a particle around.”

When Superfine met him, Taylor was working on an “interface,” the computer program that would use images to help people visualize nanoscale microscopy in a more natural, intuitive way.

“Originally I was skeptical,” Superfine says. “You know the easy comment is that it’s just a bunch of pretty pictures. But after working on this project for all these years I’ve, as they say, ‘come over to the dark side.’ I’m now a real proponent of pretty pictures.”

The pretty pictures did more than make the work appealing. Painting an image of the data took advantage of the human aptitude for rapidly seeing and interpreting patterns. This aptitude is a product of human evolution, Superfine says. In the wild, understanding at a glance the subtleties of a facial expression, an animal’s markings, or the lay of the land can spell the difference between death and survival.

Computer scientists had been learning how to compute and render images of familiar kinds of spaces—the interiors of houses or offices or factories, for instance—realistically and rapidly enough that people could don a virtual-reality helmet and take a virtual tour. This was complicated aplenty, since a stroll through a virtual kitchen might require thousands of polygons (the geometric shapes from which virtual objects are assembled) and millions of computations to render. But here was a chance to draw and paint surfaces and objects so small that people hadn’t even seen them, whose dimensions and properties were still largely unknown. So the same system that would reveal the object’s properties would also have to render it on the screen. And all of this would have to happen immediately, in “real time,” because when you do experiments, it is often impractical to move something and then wait for a picture to see what happens next. As Sean Washburn puts it, “You have to respond on the fly.”

To engineer these attributes into the system, Washburn spent seven or eight hours with Russ Taylor every Tuesday for more than a year working on the nanoManip-ulator. The technical hurdles were enormous. Washburn, an expert in quantum transport (the behavior of electrons), had to work out the dynamics of how the tip interacted with various materials on the surface. He wanted the tip not only to record the surface, but to be able to move things around on it, as well. When the team added a “force-feedback” device, the scientists could feel, through changes in pressure from a joystick-like controller, the resistance of objects as the tip touched them.

Gradually, the system began to display the nanoworld as if lit by the sun, with highlights and shadows that conveyed texture, size, and shape. Strange new landscapes, full of rich colors and contours, appeared on the screen. The system made the work so vivid and immediate, it almost seemed like play. “It has a lot of the same things you see in fancy arcade games,” Washburn says.

Superfine says that this video-game quality helps sustain excitement about the work, especially for students.

“The experiments that have to get done these days at the highest level are very hard,” he says. “They take a lot of time and extended concentration. And it takes a lot of training and experience to really understand in a physical sense what the data mean. But if a student can sit down and see a picture, a representation of what’s happening on the surface, then that gets them excited. And that excitement helps us overcome some of the hurdles in training new scientists.”

But while graphics made it easier to sort information and recognize patterns, Superfine and Washburn had to keep reminding themselves that this gleaming new world on their computer screens was in large part illusion, a sort of visual fiction to help the brain arrive at something real. “The system is enormously powerful,” Washburn says, “so it’s easy to be seduced by the graphics, to get caught up in the way things look.”

For Russ Taylor, the test of the graphic interface was its usefulness. He avoided rendering extraneous detail, even if that meant painting data-sparse areas unnaturally flat and blank, like uncharted terrain.

From Slaughterhouse to Nanohouse

As the team beat a path between physics and computer science, the project attracted new attention and collaborators. By the mid 1990s, several dozen faculty members and students from various disciplines were part of the mix, struggling to learn how to talk to each other, to match up their interests.

This was not an easy way to work. Scientists typically prefer to immerse themselves in their science, to tune in to the music of their field and tune out the noise. But in the house of nanoscience, the team had to stay tuned to a lot of different channels. And they had to attend what Washburn calls “an infuriating number of meetings.”

Washburn, like the others, knew that the meetings were a necessary evil. The nano-Manipulator and the science around it were revolutionary. They were blurring academic boundaries, rewriting rules. There was no turning back.

“A lot of the stuff that you can do by yourself, sitting alone in a lab, has been done,” Superfine says. “The problems we’re dealing with today are so complex that there’s no single kind of training that will encompass them.”

Consider, for example, the group’s recent work with motor proteins. Motor proteins are like tiny trolley cars. In cells, they run along a tangled network of microtubules, carrying cargo. To the nanoguys, this elegant little transportation system suggested a world of possibilities for doing big jobs in small ways. They planned to combine motor proteins with carbon nanotubes, for instance, to see if they could make electrical circuits much smaller than the computer-chip circuits of today.

“Think about all of the disciplines involved in such an enterprise,” Superfine says. “You get together in a meeting with biologists to talk about how to get some motor proteins, and they’re saying something like, ‘Yeah, we got a call from the slaughterhouse, and next Wednesday it’s time to go out there with some ice.’ You know, it really starts there. Because the biologists know how to get the cow brains and purify them to the point where we’re ready to try and stick the motor protein on a surface. But then we have to image it on a surface, so you need computer graphics. You have to hook the protein up to a nanotube—well, we don’t make nanotubes, but Otto Zhou down the hall knows how to make them—and then we have to hook electric leads up to it, and the electrical properties are pretty complex. Well, Sean Washburn has studied electrical properties in semiconductors for twenty years. So it’s all these different kinds of expertise being brought together to study problems that are way too complicated for any one discipline to understand.”

> NEXT PAGE: The Forest and the Trees

© Copyright 2000 Endeavors magazine, The University of North Carolina at Chapel Hill. All rights reserved.

What do you think of this story?
Let us know.