Today, even the highest-density hard-disk drives can use a million magnetic atoms to store a single bit of information. Now IBM's Almaden Re- search Center (San Jose, Calif.) has measured the ability to store a bit on a single atom.
Simultaneously, IBM's Zurich Research Lab in Switzerland has successfully demonstrated a molecular switch with processors so small that a supercomputer could fit on a chip the size of a speck of dust.
While both achievements are perhaps a decade away from commercialization, they signal the future development of computers and storage devices that could enable applications once considered the pipe dreams of science fiction writers.
"I'm reminded of [theoretical physicist] Lawrence Krauss' observations in [the bestselling book] The Physics of Star Trek. There, Krauss says that even though it is theoretically possible, we can't build a tele-transporter, because there is no way to provide enough disk storage space to buffer the trillions-of-trillions of molecules you'd have to digitize for each person you beam somewhere," said Stephan Ohr, research director for semiconductors at Gartner Dataquest (Stamford, Conn.). "Krauss predicted that it would probably take until the 23rd century, when the Star Trek adventures were supposed to take place, to develop hard disks with enough capacity. But perhaps this IBM technology will cut down that development time [to a decade]."
Back on Earth, IBM's atomic-scale demonstration could comfortably pack 500 to 1,000 times as much information on a hard disk than is possible today--or 150 trillion bits of information per square inch, compared with 150 to 300 billion bits today. Hard disks of that density would be able to store 30,000 full-length movies or the entire contents of YouTube--approximately a petabit (1,000 terabits)--on a device the size of an iPod.
Regarding the Zurich lab's demonstration of a molecular switch, IBM claims the discovery heralds a future where today's semiconductors could be replaced by molecular devices so small as to be nearly microscopic. Such tiny computers are not only beyond the International Technology Roadmap for Semiconductors, but are literally beyond even the theoretical capabilities of any CMOS semiconductor.
"Since the invention of the semiconductor, we have relied on the ability to shrink their dimensions to improve performance," said Andreas Heinrich, manager of the scanning tun- neling microscopy lab at IBM's Almaden Research Center. "But the wavelength of an electron is about 10 nanometers, so you are never going to shrink semiconductors down to the size of single atoms, which are only about 1 angstrom [0.1 nm]. If you want to perform computations or data transmission at the atomic scale, you have to find an alternative to semiconductors, and that is what the Zurich lab is doing--jumping ahead to design a new building block for molecular-sized circuits that could completely replace both silicon circuitry and copper wiring."
Today the most exotic hard-drive architectures use what is called perpendicular recording, which relies on new magnetic media. Magnetic anisotropy--a measurement of the ability of a media type to retain a bit--is the most important parameter for next-generation perpendicular recording media.
Anisotropy has to be high enough to maintain a bit in a stable state indefinitely, but low enough that a hard-disk write head can quickly change its orientation.
"Perpendicular recording depends on crystalline magnetic anisotropy; that is the key property that makes perpendicular recording work," said Heinrich. "Even with a domain of a million atoms--as hard-disk media use today for a single bit--measuring magnetic anisotropy is very tricky and is a major achievement for disk drive researchers.
"Now, however, we have been able to measure the same property for a single magnetic atom. We can literally take one atom, measure its magnetic anisotropy, put another atom next to it, see how that affects the [first atom's] magnetic anisotropy, and from there learn how to develop a material with the ultrahigh data storage densities we are predicting.
"Now we have a unique tool--the only one in the world--capable of measuring the magnetic anisotropy of a single atom," he said.
The magnetic domains on hard disks today are about 20 nm long with a track width of 100 nm, or 2,000 nm2, compared with IBM's stated goal of 2 x 2-nm (4-nm2) magnetic domains.
"And these are conservative estimates," said Cyrus Hirjibeheden, a re- search staff member at the Almaden Research Center who worked on the project with Heinrich. "Our calculations include a safety margin of 10, because the atom itself is only about 2 angstroms [0.2 nm] across."
Next, the researchers plan to measure the anisotropy of different types of atoms at room temperature--rather than at half a degree above absolute zero, as in the present work--to find an ultradense material that can remain stable at room temperature for commercial hard drives.
"Our next step will be to find a material combination--a particular magnetic atom on a particular surface--that has the ability to stably maintain its magnetic orientation, plus the ability to switch between states, so we can quickly flip its spin from down to up and from up to down," said Hirjibeheden. "We hope to be able to demonstrate such a stable media material within the next couple of years."
The lab will use the magnetic tunnel junction on a real disk-drive read head to sense this advanced material's magnetic orientation, instead of the scanning tunneling microscope (STM) used in IBM's laboratory demonstration. (IBM scientists Gerd Binnig and Heinrich Rohrer won the Nobel Prize in Physics for inventing the STM in 1986.)
"Our final step would then be to move our demonstration out of the research laboratory and into a commercial manufacturing area, which will involve a lot of basic engineering effort, instead of the research breakthroughs we specialize in here at the lab," said Hirjibeheden.
Workings of a molecular switch
Meanwhile, at IBM's Zurich lab, a molecular electronic switch was developed completely separately from the magnetic-media work being done at the Almaden lab. The Zurich researchers have discovered a unique molecule that can switch data streams passing through it on and off without changing shape--a capability that the researchers say makes it a potential building block for real molecular-sized electronic circuits sans semiconductors.
"What our Zurich lab has found is a molecule consisting of just 50 or so atoms that they can flip like a switch by reorienting just two hydrogen atoms--the smallest atoms that exist--by changing their location in the center of that molecule," said Heinrich. "This leaves the entire framework of the molecule intact; it doesn't change the outside of the structure, but it does change the electronic properties of the molecule, potential- ly enabling it to switch the data trans- mission of electrons traveling through the molecule."
The molecule discovered by the lab is called naphthalocyanine and consists of eight benzene-like rings of carbon atoms formed into a cross shape. It has a cavity in the middle where two hydrogen atoms can be flipped between two possible attachment points that switch the molecule between conducting (on) and non-conducting (off), but without changing its outer dimensions.
"The biggest advantage of using naphthalocyanine as the switching elements of future molecular-sized circuits is that they don't change shape," said Hirjibeheden. "Consequently, it should be possible to build working circuits that retain their structural integrity while switching--unlike the other switching molecules to date, which change shape when they switch, thus potentially disconnecting any wiring you might try to attach to them."
The next step taken by the Zurich team, which the researchers estimate will be achieved within a year, will be to attach wires to the four ends of the cross-shaped molecule so that live signals can be switched through the device.