Highlights from the AAAS: Batteries out of Paper, Order out of Chaos

I spent Saturday veering from science future, to science past and back again—learning about the ways nanotechnology could revolutionize energy generation and storage, delving into the history of how scientists made themselves more accountable to fact and then getting a peek at the connections researchers are trying to make between social policy and climate/energy technology.

Paper Batteries, and the Hunt for the Perfect Thermal Insulator

"We basically have perfect electrical insulators," said David Cahill, Ph.D., professor of engineering and materials science at the University of Illinois. "You can grab a live wire that's coated in insulating material and you don't die."

But there's no analog for that in thermal insulation, Cahill said. That fact that matters a lot if you care about improving energy efficiency in everything from engines to houses. Cahill is part of a team working to improve thermal insulation with nanotechnology. His goal: Create some kind of new material that will disrupt the transfer of heat energy between two objects. Getting it right would have big implications. For instance, we could drastically improve our ability to capture the waste heat from electrical generation and put it to use in other ways.

One possible solution is silicon nanowires. These structures are normally baby-butt smooth, but as you make their surfaces more and more rough, the nanowires conduct less and less thermal energy. Right now, it's not exactly clear why that trick works. But understanding it could put Cahill's team on the right path.

He's not the only one taking energy technology nano. Another researcher on the same panel, Yi Cui, Ph.D., of Stanford, is applying nanostructures to energy storage, in hopes of developing smaller batteries that can hold more power.

In fact, according to Cui, nanotech is absolutely essential to any future progress with batteries. Storage capacity for size has plateaued, he explained. To go further, we have to start making electrodes out of completely different—and probably completely new—materials.

Nanotech has even enabled Cui to make batteries out of paper and fabric. His team takes these ordinary materials and coats them in carbon nanotubes. The porous fibers give ions easy access to the conductive nanotubes. The more nanotubes you add, the more conductive the paper or fabric becomes. Layer many sheets of treated material together, and you get all-paper supercapacitors, which Cui's team has already built.

Do the same thing with the textiles—which are still stretchable and flexible—and you've got wearable electronics. Of course, there are limitations. In the Q&A, somebody asked Cui what would happen if you wore those fabrics on a static-y day. The reply was some awkward laughter and promise that the team is still working out the kinks.

I went to a lecture on the history of dealing with experimental error not really knowing what to expect: Boring round of scientific "Inside Baseball" or eye-opening fact fest? Thanks to historian Jed Z. Buchwald, the session definitely veered toward the latter.

Even the basic idea was fascinating. You probably know that modern scientists try to account for their own research mistakes by running an experiment or taking a measurement multiple times. Often, the results will all be slightly different, and scientists deal with that by taking an average—which is likely to represent the closest-to-correct answer.

But they didn't always do it that way. In fact, according to Buchwald, the first scientist to use averages to address error was Isaac Newton, who privately started taking averages in his notebooks in 1671.

Before that, obviously, scientists still made mistakes. Multiple measurements or experiments still yielded varying results. But they dealt with the variation in a very different way—they picked the answer they thought represented their best work.

To modern ears, that sounds like cheating—"You just randomly decided on the number you liked best? That's science?" But, at the time, it was perfectly logical. Historically, scientists viewed themselves as craftsmen, Buchwald said. If you were building a piece of fine furniture, you wouldn't make a bunch and pick the average to display. You'd choose the finished version that was the best, and best displayed your woodworking skill.

In fact, the whole reason Newton didn't publish his first research to use averages was because it would have made him look like a lousy scientist. After all, what kind of craftsman can't tell his own best work?