IRA FLATOW, host:
You're listening to Science Friday from NPR News. I'm Ira Flatow. We've been talking this hour about invention, and a lot of, you know, a lot of spectacular inventions have happened because of a simple flash of genius; boom, an idea hits you and it drives you to a new discovery. And my next guest says that's what happened to him. He was hit by that sudden inspiration to make tiny nanoelectronic chips by thinking about an Etch-A-Sketch. Interesting story, I can make that like an Etch-A-Sketch. Remember that little red plastic casing with the silver screen; you had two little knobs you twist to draw pictures on the screen? And the best part about it is if you drew something and you wanted to erase it, you could just back it up and shake the toy and start all over again.
Well, my next guest has found a way to sort of do the same thing, drawing tiny conducting nanowires and transistors on a little chip and then erasing them just as easily, all this at a scale much smaller than the silicon-based computer components that we use today, though these nanochips aren't quite ready for your computer. They're right at the beginning of development. And joining me now to talk about this research out this week in the journal Science is Jeremy Levy. He is professor of physics and astronomy at the University of Pittsburgh in Pennsylvania. Welcome to Science Friday, Dr. Levy.
Dr. JEREMY LEVY (Physics and Astronomy, University of Pittsburgh): It's a pleasure to be here, Ira.
FLATOW: Tell us about your flash of genius.
Dr. LEVY: Well, you explained it pretty well. I'd like to be able to say that, well, I was deriving some equations and I figured out some theorem that would tell me how to do this. But no, it was my toy that I experienced as a child that really was the basis of a discovery, not something that we use to spin the discovery afterwards.
FLATOW: Tell us what happened, where you were, how you said, boom, that's an Etch-A-Sketch.
Dr. LEVY: Well, even that was kind of circuitous. I was actually going to a conference that was at the same time as the World Cup in Germany, and there was about a two-week gap between the end of the conference and the final of the World Cup. My brother had gotten me, for about four birthday presents combined, a ticket to the final, and I had to spend about two weeks in Germany, basically, and justify my existence there, so I visited a lot of labs. And one of them was my then-to-be future collaborator Jochen Mannhart, and he was describing to me some of the really astounding results that they had discovered, where you take two materials that are normally insulating - one rather thick and the other one exceedingly thin - putting them together and they were able to make the interface, the reach in between those two materials, switch between a metal and an insulator. So, they could make it either conduct electricity or be a good insulator. And that was - when I heard about that, that's instantly when I began to think about the Etch-A-Sketch. And so, well, that began our collaboration, which, you know, ended up in this phone call.
FLATOW: Hmm. And so, take us further on that idea. You have this sandwich and how would you sketch it? What would you do with that?
Dr. LEVY: Right. So, the - well, the way that the Etch-A-Sketch works is - literally, the analogy works well. It's actually - there's a fine powder of aluminum encased in a - with a top glass surface. And the idea is that you - that the aluminum naturally sticks to the surface, and it's - it looks grayish, and then when you scrape it with a sharp probe, it looks clear or darker. And so, the same thing is happening in our system, we have a very sharp metal probe that's basically part of an atomic-force microscope, and we move across the top surface of this material, and we apply voltage to this, which is either positive or negative. And as you said, if it's positive, we end up making a wire at the interface, so below the surface by about 1.2 nanometers. And then if we reverse the sign of the voltage, we can erase it. So, what was really surprising was how small of structures that we were able to make.
FLATOW: So, you could actually physically, like the Etch-A-Sketch, you can move the probe back and forth and make a line, a conducting line, like, a conducting wire out of it?
Dr. LEVY: Precisely. We can make lines. We can make dots. So, a dot is nice because it could be used for memory. If you put a dot there, you have a little puddle of electrons, and then, they'll either be there or not and that can be a one or a zero. And the thing is, we can make that 1,000 times smaller than the smallest magnetic domains that are used in hard drives. And then by making wires, we, of course, can make the kinds of wires that you would have in electronics, and then more recently, what we recently published was the ability to make transistors, which is - it's been argued - been discussed as the most important invention of the 20th century.
FLATOW: And how - are you claiming, then, that you can make the smallest transistors in the world?
Dr. LEVY: Well, I think people may have made a one or two, some things that are smaller. I would say that if other methods - let say, putting a molecule in between two electrodes that are - happen to be very close to one another, it doesn't always work very well, and if you don't put - you can't always put the same molecule there in the same orientation.
Dr. LEVY: What we can do is basically, we can make lots of them and we can make them in a very reproducible fashion that - it's more deterministic than, for example, directed self-assembly, although those are very promising...
Dr. LEVY: Techniques of their own.
FLATOW: So, walk us through how you make a transistor.
Dr. LEVY: Well, we begin with the material itself, which was provided by my collaborator, and then we put it inside of our atomic-force microscope, and then we simply touch the surface with this very sharp probe. And then if we want to make a wire, we put a positive - take a battery and put the little plus tab on...
Dr. LEVY: You know, attach it to the atomic-force tip, and then we scan it across, and everywhere we go with that probe, we have a conducting region. And then if we want to then cut something, we can reverse the bias. It's actually really - it's quite straightforward.
FLATOW: Mm-hmm. And you could chop that wire up into little pieces?
Dr. LEVY: Exactly. And that's one of the ways that we actually can tell how small they are. We can make a wire, and then we can move across the wire and see how sharply it cuts. And we have very precise calibration of that.
FLATOW: Mm-hmm. And how complex a circuit can you make with this technology?
Dr. LEVY: That's a very good question. So, right now, we basically have - we can work in one-, two-dimensional plane, and so, we have to make sure that the wires can actually make a path out to the edge, for example. But we think that - I mean, we don't know the exact shape of the transistor that might make it into technology would look like and how that would be connected, but - so, really, we haven't been trying to make what I would call technology, things that you can...
Dr. LEVY: Put into a device, but rather demonstrate that we can make things that are exceedingly small. And by small, I mean a transistor that's 1,000 times smaller than the existing transistors that are used, the state-of-the-art transistors today, and 100 times smaller than the smallest ones that people think they can make with silicon.
FLATOW: And a transistor basically, if it's on, it's a one. If it's off, it's a zero, sort of.
Dr. LEVY: Well, that would be more like a memory-bit.
Dr. LEVY: But so the ideas that you can - so, a transistor, basically, has three terminals, and you have a channel, where you can - electrons can flow, and the idea is that there's a gate.
Dr. LEVY: And the gate is either open or closed, and that is, basically, is logic, which allows you to process information and not just store it.
FLATOW: Mm-hmm. Can you make the other electronic components the same way, like the resistors, the coils, things like that, that might go into the circuit?
Dr. LEVY: Well, yeah, you're describing analogue, for example, analogue components. I mean, yeah, these wires are themselves resistors naturally, and to make inductors - you're thinking about loops, we can sort of make ones that are flat, and so, it may be possible to make analogue-type devices as well. But you know, the other thing that's really amazing about this material is that we can do very different things with the same material. So, if we think about the way normal computers work, you have hard drives...
Dr. LEVY: Which are - which use magnetic materials for storage and then we have silicon for processing information. Here, we can do both with the same material, and the processes for making these things are a lot simpler. But there are even other things which, you know - moving away from this sort of technology towards science, the interface also has really interesting properties at low temperatures. So, for example, it can become superconducting, meaning that these electrons pair up and they move in such a way that there is no resistance whatsoever. And that's something that can be - that could be really useful for very futuristic things like quantum computers.
FLATOW: Could it help our electric grid?
Dr. LEVY: Uh, I'm not so sure about that...
(Soundbite of laughter)
Dr. LEVY: Because it's at very low temperature...
Dr. LEVY: Much lower - you would probably end up wasting a lot of energy just getting to those low temperatures.
FLATOW: 1-800-989-8255, talking with Jeremy Levy, who has come up - he and his partners have come up with a technique for using, like, an Etch-A-Sketch sort of microscope to scratch out nanotechnology transistors and little memory dots. What - is there a breakthrough that you need? Or what direction would you take this now?
Dr. LEVY: Well, I'd like to think that this was a breakthrough, and we - so, about a year ago, we showed that we could make dots and we could make wires. But then we said, OK, we have to be able to make transistors, and now we've made transistors.
FLATOW: Right, mm-hmm.
Dr. LEVY: And I think one of the challenges is going to be to see whether we can really get to the extreme, sort of, very practical limits of how small and efficient you can make these transistors. That's usually what determines whether technology is adopted, is how energy-efficient it is, because if you try to shrink things with - and they still consume the same amount of power, it basically turns into a heat source...
Dr. LEVY: And it will blow up.
FLATOW: Why would you - what's the advantage to being able to write and erase and rewrite using this technology, on the fly, so to speak?
Dr. LEVY: Yeah, well, I mean, there's certainly some practical advantages. Normally, the way that things are made in - with silicon or most semiconductors is that you have a process where you, let's say, you remove material or you deposit some material, and those involve many steps which are quite complex, and each one has a certain probability of success. And if you make a mistake, then you can't fix it; basically, you throw out the chip. And so, that's - from a practical point of view, that's something that we think is an advantage, but it may also be that simply being able to integrate very tightly memory and logic would provide a very new and different type of architecture. So, it's something that I think people haven't been thinking so much about because it wasn't - there was no...
Dr. LEVY: Tangible way of understanding how it might be implemented.
FLATOW: You know, we have things - we have, like, what we call firmware these days. If you're - like, if you have a telephone or something and you want to upgrade the software, you know, you download something into the computer into - of the software - into the software, of your phone and it upgrades it. Could you do the same thing with your kind of hardware?
Dr. LEVY: Yes. Yeah, firmware is sort of like nonvolatile memory, but there are also things related to, like, what you're saying, field-programmable gate arrays, and those are also - they're somewhat restrictive in their geometry, and so, it maybe that there would some advantages to being able to completely change the architecture. But you know, there's another thing about just the way in which we make it. Right now, of course, we have to use an atomic-force microscope, which is kind of big. But actually, major companies have in the past developed and then set aside technologies that involve using hundreds of thousands of these very sharp probes to store information. So, actually, IBM was one of the leaders in this, and they had this Millipede Project, but they put aside because it wasn't competitive against flash memory. So, it turned out that flash was getting - scaling so rapidly that they weren't going to win the race in some sense, and so, they put it aside. And many other companies that were developing similar technologies put those aside, but it may be that we want to sort of open those doors again because the - we can add to that logic and get processing capability, which was never on the table before and also being at a - about a factor of 100 times smaller.
FLATOW: Talking about nanotechnology with Jeremy Levy on Talk of the Nation: Science Friday from NPR News. You know, you talk about an electronic probe, a microscope, for some of the folks - Bunny on tweet - our Twitters are coming in and saying, you know, it sounds what you're doing more like a plotter, you know? You draw an architecture, and its goes out and plots, you know, a whole big design on the surface.
Dr. LEVY: It's very much like that, of course, but smaller.
Dr. LEVY: And yeah. So, basically, we can use, actually, the same kinds of plotting packages that are used for much bigger circuits, but we use them with our microscope.
FLATOW: And can you - and you can make - can you make a circuit like that by just plotting it?
Dr. LEVY: Yes, and that's literally what we do. And when we made the transistor, we sort of consciously tried to make something that was as simple as possible, three lines. It's hard to imagine something that's simpler than that and yet acts as a transistor.
FLATOW: Is this very expensive, this stuff?
Dr. LEVY: Very expensive? Well...
FLATOW: To make?
Dr. LEVY: The - I mean, the costs are - it's hard to compare it, for example, with a chip that is the result of 50 years of technological innovation.
Dr. LEVY: But let's say, the substrates, the - that - so, the sort of big part of the material that (unintelligible) is - can be actually pretty expensive, but it may be possible to grow those materials on silicon. So, actually, one of the things that we may be trying to do research on in the future is to see whether we can integrate this capability, these oxide hetero structures on silicon and put them on big silicon wafers, and that would make them a lot more cost-effective.
FLATOW: Mm-hmm. So, you'd create sort of as hybrid?
Dr. LEVY: Exactly.
FLATOW: Yeah. As you say, one of the big challenges is you have this tiny thing, how do you put the leads on it to get the stuff in and out?
Dr. LEVY: Precisely.
FLATOW: Yeah. The wires are much bigger than what you're drawing.
Dr. LEVY: Yes, that's true. And we do - but we do have the top side and the bottom side and...
(Soundbite of laughter)
Dr. LEVY: To work with.
FLATOW: Will we ever see a working model of this some place?
Dr. LEVY: A working model? Well, you can come to my lab.
(Soundbite of laughter)
Dr. LEVY: We have a working model.
FLATOW: Do you have a Web site we can look at it?
Dr. LEVY: Sure. Oh, well, I mean, to look at pictures...
(Soundbite of laughter)
Dr. LEVY: Yeah, I sort of have some pictures up - yeah, I have a Web site, epsilon.phyast.pitt.edu.
FLATOW: All right, well, we'll melt it down for you this afternoon when everybody goes to look at it.
Dr. LEVY: OK.
FLATOW: So, be ready for that. I want to thank you for taking time to be with us, Jeremy.
Dr. LEVY: Well, thank you, Ira.
FLATOW: And keep us abreast of any kind of new developments that you have, and we'll be happy to talk about where you go from here.
Dr. LEVY: Wonderful.
FLATOW: You're welcome. Jeremy Levy is a professor of physics and astronomy at the University of Pittsburgh in Pennsylvania, and he was talking about his Etch-A-Sketch ideas for making nanotransistors. That's about all the time we have for today. Our program is produced by Christopher Intagliata and senior producer Annette Heist. Charles Bergquist is our director, and Flora Lichtman is our producer for digital media. Our intern is Shelly DuBois. Neal Rausch is our technical director and at the controls here today is Manya. We also had help in Second Life from Lynn Cullins, Dave Andrews, Jeff Corbin and the University of Denver.
If you missed any of the stuff we talked about today, you might have a little trouble hearing some of the program today, we're going to have it up there on our Web - we apologize for that, some technical difficulties we had at the beginning of the program. We apologize, but our podcast will be up there. Hopefully, the whole thing will be ready and we can listen to it once we get it digitized. Also surf over to our Web site at ScienceFriday.com. We are Twittering, @SciFri is tweet address, and also we're podcasting, and folks in Second Life are gathering around as they do every Friday afternoon. I'm Ira Flatow in New York.
NPR transcripts are created on a rush deadline by a contractor for NPR, and accuracy and availability may vary. This text may not be in its final form and may be updated or revised in the future. Please be aware that the authoritative record of NPR’s programming is the audio.