Light-Producing Nanotubes Could Mean Faster Chips

CannibalBob writes "From PCWorld: Researchers at IBM have used carbon molecules to emit light, a breakthrough that could replace silicon as the foundation of chips and lead to faster computers and telecommunication equipment. This is the first time light has ever been generated from a molecule by applying electricity. Read the article."

First Paragraph

[hendridm]

Was this a troll?

"Researchers at IBM have used carbon molecules to emit light, a breakthrough that could replace silicon as the foundation of chips and lead to faster computers and telecommunication equipment." (emphasis added)

It was also reported a year ago [slashdot.org] that they had created transistors [nytimes.com] using nanotubes, although not with light.

Getting the Facts on Light Emitting Carbon Nanotub

[jscribner]

Here's the detailed info on all this:
IBM Research Light Emitting Carbon Nanotube news release [ibm.com]

There's also an animation [ibm.com], but the pictures in the release are easier to follow.

Re:First Time...

[L7_]

a single molecule emits light whenever it makes a state transition... Its the phenomenon that caused the paradigm shift from classical to quantum mechanics.

That can't be the point.

Re:First time?

[f97tosc]

How do light bulbs work? True, they burn as a side effect of being heated, but you apply electricity, and you [eventually] get light. Then there's the the whole laser thing... Florcent tubes?

Well, any light form needs energy - and electricity is a common way of providing this.

Light bulbs emit light because they are heated by electricity. Unfourtunately, about 95% of the light emitted is not visible to humans, and thus wasted (human eyes are tuned to best view light from a certain body at about 6000 degrees, and this is much hotter than the light bulb - thus the inefficiency).

There are, however, ways to convert electricity to light without heating anything. LEDs do this - all energy is converted to light of a single certain frequency - which we can see. This is true for lasers also, but they go even further by not only having light of a single frequency but also aligning the light waves that compose the light.

But both light bulbs and LEDs are made of big crystals of metal / silicon (as opposed to molecules). What is new here is the atomic structure of the of the light emitting material; it is nanotubes which technically are big molecules. This is a major discovery - although it is probably too early to tell exactly what it will be useful for in the future.

Tor

The carbon nanotube...

[tellezj]

would constitute a single molecule. Applying electricity to it, as pointed out in the article, they were able to produce light (1.5 micron). An LED, tungsten wire, or burning lump of coal are not made up of a single molecule, no more so than an ice cube is a single molecule of water. What this constitutes is an engineering first. What is left to be seen is if they can find useful applications and mass produce it.

Re:Mass Production

[Anonymous Coward]

That's the major problem they've had...the things are pretty damn hard to manufacture with any quality.

Thankfully, some newish production methods are being put through, and a couple dedicated factories are being built. From what I read in new scientist, it should drop the price down to $5/kilogram instead of $500/milligram (or some similar outlandish figure)

Re:First Time...

[Hayzeus]

That would be from a SINGLE molecule; nanotubes are single molecules. More handy dandy info at IBMs nanonotube web site [ibm.com].

Re:First Time...

[BTM1001]

Well, some of us went onto college chemistry. a molecule is simply a group of more than one atom that are bonded together. Like say the standard Oxygen in the atmosphere - it is a molocule, O2 - 2 atoms of Oxygen linked. Not sure if Tungon is the same naturally.

Re:First Time...

[civilizedINTENSITY]

Yes, it is the point. Although it happens molecule by molecule when it happens in mass, there wasn't a way to control the applicatiion of a signal and stimulate emmision from a chosen molecule. Now there is. For the first time ever, *a* molecule is made to emit when electrically stimulated.

Posted on physicsweb

[parkanoid]

Three days ago [physicsweb.org]

Re:This is cool, but...

[Steffan]

I don't think Bill Joy's issue is with Nanotubes per se, but with self-replicating nanomachines that could, due to a 'programming' bug or something similar, not stop replicating even after their task is done, hence the 'gray-goo' you hear about. It's a legitimate concern, but IMO should not and does not justify the cessation for nanotechnology research.

Re:Mass Production (of nanotubes)

[wass]

yes, nanotech is currently one of the 'sexy' topics, and so every two months (or more often, usually) someone gets to publish their new fabrication or measurement technique.

I'm a graduate physics student (experimentalist), and I'll be working with nanotubes. But we're just building up our lab now (my advisor just arrived here only a few months ago). We'll be doing measurements with carbon nanotubes, initially continuing what we did last summer (at her old postdoc lab) by measuring superconducting nanowires. If you're curious, these nanowires are created by sputtering a superconducting alloy (MoGe) on top of a nanotube substrate. They're interesting because the system dimensions are small enough that the wires are effectively one-dimensional, which means they can't support long-range order and thus cannot allow Cooper-pair supercurrents to flow unimpeded through the wire.

It's hard to create nanotubes, and harder to put them where you want them. One way to create them is to use chemical vapor deposition (CVD), where you basically try to create a controlled environment where some hydrocarbon (eg methane) is ignited (the environment is somewhat oxygen-deficient so CO2 isn't the only carbon species produced) The 'soot' that is subsequently deposited on your substrate should contain nanotubes if the right conditions are met.

To get the tubes in certain places, sometimes little 'seeds' of iron particles are used, in hopes the nanotubes will grow/branch from them. It's hard to create good SWNT (Single-Walled Nanotubes), but easier to form 'ropes' of many nanotubes intertwined together.

Another difficult factor to control is the 'chirality' of the tube. Basically, a carbon nanotube is a rolled graphite sheet, but when the sheet is rolled, it can have certain 'twist' to it. For example, if you rolled lined paper into a cylinder, you can have zero helicity, in which case your lines will form independent circles. Or you can shift the lines by an integer number, in which case the lines will form helices of varying pitch. This factor in nanotubes determines the electronic band structure, which mandates whether the tubes are metallic or semiconducting. It would be highly desirable to be able to produce consistently tubes of the same chirality.

I hope this makes sense, I was up all night doing E&M homework (ya gotta love Jackson), so my brain is kinda fried right now.

Re:Graphite pencil leads

[pclminion]

What's done to prevent the rapid over-oxidation, especially in something that astonishingly thin?

In your experiment, the graphite glowed because it got hot. It also oxidized because it got hot. It's a simple case of resistive power dissipation.

In this case, the carbon is emitting light through a quantum process, not thermal radiation. The graphite doesn't get hot and therefore doesn't oxidize.

Re:Make humans glow!

[wass]

Couldn't we just apply electricity and make all the carbon's in our bodies glow?

I don't know how facetious you're being, but I'll answer anyway. The carbon atoms in a carbon nanotube are in a highly ordered arrangement (a nanotube is essential a crystal with well-defined point symmetry groups), which means the potential energy (ignoring end-effects of the tube) is invariant under certain symmetry operations, namely translation and rotation. These symmetries will manifest themselves when you solve Schrodinger's equation in some form of electronic band structure, probably as a splitting of their corresponding degenerate states. The resulting bandgap is what is most-likely being exploited to emit the photons.

Contrasted to the human body, in which case the carbon atoms don't have much ordering at all, and chemical reactions are constantly occurring. Hence the band structure would be a chaotic non-equilibrium mess.

Silicon Does Not Emit Light?

[Pooua]

Silicon, the main material used in semiconductors, does not emit light, and therefore can't be used in optoelectronic products, Avouris said.

I beg to differ. Silicon has been made to emit light in various ways for over a decade.

"Scientists at Surrey University, led by researcher Kevin Homewood, are showing off a prototype silicon-based light-emitting diode (LED) -- an invention that could be of significance to the whole electronics and communication industry.

"By enabling silicon to emit light, the scientists say they may have found a way to use light to efficiently transfer data around microchips. This could lead to smaller, more powerful computers and improve data communications significantly."

ZDNet UK: Light-emitting silicon boosts chip speeds: 8th March 2001 [zdnet.co.uk]

"The photoluminescence emanating from a regular array of 1.2 m sized dots composed of Si nanocrystals was studied with spatial, spectral and temporal resolution."

New Journal of Physics: Nanostructuration with visible-light-emitting silicon nanocrystals [iop.org]

"GENEVA, Switzerland -- STMicrolectronics claims to have achieved a breakthrough in the creation of light-emitting silicon and said it would have engineering samples of monolithic silicon devices based on the technology, combining electrical isolation and optical communication, before the end of 2002.

"The development allows silicon light emitters to match the efficiency of compound semiconductor materials such as gallium arsenide for the first time, the company said."

EE Times: STMicro claims light-emitting silicon breakthrough: October 28, 2002 [eetimes.com]

"The discovery of visible luminescence from porous silicon [1] has stimulated a large interest in this material. Numerous studies have demonstrated that it is possible to achieve efficient visible luminescence from porous silicon layers [2]. This material system has significant economic potential as efficient visible emitters could be fabricated on silicon wafers and incorporated with current microelectronic devices using existing silicon processing technologies."

[1] L. T. Canham. "Silicon quantum wire array fabrication by electrochemical and chemical dissolution of wafers." Appl. Phys.Lett., 1990, 57 1046 - 1048.

[2] For a recent review of the work in porous silicon see : Thin Solid Films, 1995, 225 and "Porous Silicon", edited by Z. Chuan and R Tsu, World Scientific, Singapore, 1995.

A Visible Large Area Light Emitting Diode Fabricated From Porous Silicon Using A Conducting Polyaniline Contact [dur.ac.uk]

BTW, technically, photocells are optoelectronic devices, as are LEDs.

Re:Maybe as a corollary?

[putigger]

In principle, yes. That's how solar cells and photodiodes work. Now, just because something is efficient at converting electricity into light, however, doesn't mean that it will be similarly efficient at converting light to electricity. Generally, devices are usually specialized toward one function or the other. I suspect that nanotubes wouldn't be that great as photocells. For one, you can already make high efficiency semiconductor photocells with large surface areas (important for light collection!), whereas, nanotubes wouldn't give you much of an active area to work with unless you manage to pack them densely on a surface.