Quantum Mirage
There are some spooky goings-on in the atomic world. Get them under control and we'll be on our way to making a computer with atoms that aren't really there. Robert Irion investigates

SOMETHING bizarre has popped up in the laboratory of physicist Donald Eigler, and no one is quite sure what to make of it. It's the world's smallest ghost, a phantom on an atomic scale. Eigler and colleagues at IBM's Almaden Research Center in San Jose, California, call it a "quantum mirage" because it projects an atom's electronic signature to another place while the atom itself stays put.

The projection spans only a tiny distance: about 10 nanometres, or 10 000 times narrower than a human hair. Still, it offers the tantalising promise of transferring information within tiny circuits of the future in which wires are obsolete and the components are single atoms.

The eerie image may also let physicists probe an atom without disturbing it directly. That's an intriguing prospect in a miniworld where even a few photons of light can alter the states of particles. Eigler and his team think it may even be possible to forge a chemical bond with the mirage by moving a compatible atom next to it. That would create a weird hybrid molecule that IBM physicist Hari Manoharan calls "half real and half ghost".

The mirage is yet one more curious manifestation of the quantum world, in which electrons and other subatomic flecks behave neither like particles nor like waves, but as a mixture of the two. In this case, the researchers confined a sea of electrons on a copper surface within a "quantum corral", a picket fence built of several dozen closely spaced cobalt atoms. The corral kept the surface electrons from flitting about the surface of the metal with their normal freedom. What's more, Eigler's corral is in the shape of a near-perfect ellipse. That special geometry compelled some of the electrons to cluster around two spots: the left and right foci of the ellipse.

Using the precise motions of a scanning tunnelling microscope (STM), Eigler's team placed an atom of cobalt at one focus. In the frigid temperature and ultra-high vacuum of the STM chamber, the atom stuck there like a refrigerator magnet. When the physicists probed the cobalt atom with the STM, they saw a swarm of electrons around it. Then, when they scanned the other focus, they saw an unmistakable signature. The swarm surrounding the real atom was mirrored at the empty focus, even though there was no atom there. The reflection wasn't complete, as the STM detected the mirage atom at only one-third the intensity of the real one. Still, it was a startling apparition.

So what is it? Of one thing Eigler is sure: it's not just an illusion. "There are real electrons there," he says. "It's a physical object." But when it comes to the nitty-gritty physics of describing the genesis of the mirage and its implications, Eigler and his team admit they're puzzled. "What exactly are we doing? What properties are really there?" asks Manoharan.

It all seems rather surreal, and that other-worldliness is heightened by the manner in which Manoharan, physicist Christopher Lutz and Eigler use exaggerated vertical relief and garish colours to depict their quantum corral and the mirage it contains. There it was, on the cover of the 3 February issue of Nature: a ring of aggressive yellow atomic spikes enclosing a wavy orange-and-green sea of electrons and two bright violet islands, the atom and its mirage. One almost could envision the cursive writing of René Magritte under the image: "Ceci n'est pas un atom."

The quantum mirage is the natural denouement of progress made during the past decade along three fronts: STM technology, quantum corrals, and recognition of a peculiar state called a "Kondo resonance" around a magnetic atom.

The basic concept of the STM hasn't changed since IBM physicists Gerd Binnig and Heinrich Rohrer developed the apparatus in Zurich in 1981. Researchers use a fine wire--in Eigler's case, made of pure iridium--that tapers to a single atom at the point. When this tip approaches a conducting surface, such as a metal, the physicists apply a small voltage. This induces electrons to "tunnel" across the gap between the tip and the metal. By keeping the tunnelling current constant as the tip is scanned across the surface, the researchers create a topographic map of its atomic peaks and valleys. An STM can work at room temperature, but Eigler's machine has to operate at just four or five kelvin to keep the atoms in place.

Eigler's group was the first to use an STM to drag atoms into desired spots. The researchers raised the voltage in the tip, and brought it close enough to the surface to form brief chemical bonds with single atoms. Eigler and IBM colleague Erhard Schweizer made an international splash in 1990 by spelling "I-B-M" with 35 atoms of xenon atop a layer of nickel.

Three years later, Eigler and Lutz teamed up with IBM physicist Michael Crommie--now at the University of California at Berkeley--to build the first quantum corrals. Their circular barrier of iron atoms on a copper surface caused another sensation, for it revealed in dramatic fashion the wavelike behaviour of the electrons trapped within. In this case, the electron densities peaked sharply in the centre of the ring. Surrounding that peak were concentric rings of lower electron densities, looking for all the world like ripples from a pebble plopped into a pond.

The third advance that paved the way towards quantum mirages came in 1998 when physicists first observed the Kondo resonance. This effect was proposed in 1964 to explain the strange way that metals interact with atom-sized magnetic impurities, but until the invention of the STM it was impossible to verify. Two years ago separate teams led by Schneider and Crommie found unambiguous evidence that the Kondo resonance actually occurs.

The surface of a conducting metal, such as copper, is covered with a sheen of electrons that swarm freely across it--which is why they conduct electricity so well. Place a single atom of cobalt or another magnetic element on this surface, and its electrons disrupt the smooth flow of the conduction electrons. "You can think of the tightly bound cobalt electrons as a little hard ball," says Crommie. "The copper electrons swimming around are repelled from the ball." In addition, a tiny cloud forms, and within this the copper electrons spin in the opposite direction from those around the cobalt atom. This alignment effectively screens the disruptive magnetic field of the intruder atom.

At the atomic scale, nothing looks quite like the Kondo resonance. That's why Eigler's team chose the cobalt-on-copper system. Project a Kondo resonance to a remote location and you have the unmistakable signature of a quantum mirage.

Another critical factor that Eigler exploited is the special geometry of an ellipse. At school, you probably learned to draw an ellipse by sticking two tacks on a board--the foci of the ellipse--then forming a loop by tying a length of string around them. A pencil that pulls the loop tight traces out an ellipse. In the atomic world, this means that if a signal starts at one focus and bounces off the ellipse wall toward the other focus, it travels the same distance no matter which direction it goes.

The consequences for electrons trapped within an elliptical corral are fascinating. "One path length connects the atoms at each focus," Manoharan says. "It changes the two-dimensional problem to a one- dimensional problem. The entire ellipse acts like a single wire connecting the atoms at the two foci." An electronic disturbance at one focus, such as the Kondo resonance, echoes off the walls of the corral and appears, ghostlike, at the other.

Whispering galleries

For those who can't wrap their brains around quantum mechanics, acoustic analogies abound. "Whispering galleries" are often elliptical chambers, in which sounds from a speaker at one focus echo into the ears of someone standing at the other. Many musical instruments also feature resonant chambers in which waves vibrate with characteristic modes. "To some extent, what happens within our corral is nothing more than what happens when you have waves inside a resonant structure," Eigler says. "Where it is special is that the elliptical structure helps to make the effect we are observing very obvious in one special spot."

But acoustic descriptions don't quite illustrate what exists at the mirage site and what is illusory. For that aspect, Lutz's favourite analogy is inspired by the image of a solar eclipse in a pinhole camera. "It's not the Sun itself, but it has some characteristics of the Sun," he says. "It's made of photons that really came from the Sun, and it shows the changing shape of the Sun. You could even start a fire with it if you intensified the image with a lens." One can learn some of the Sun's properties, but by no means all, by studying the projection. The same may be true of the quantum mirage, Lutz believes.

For now, the only properties of the real atom that show up conclusively in the mirage are the energy states of its electrons. A spectrum taken with the STM reveals how many electrons exist at different energy levels around the atoms. That analysis shows that both the real atom and its evanescent twin display the distinctive Kondo resonance. The team is optimistic that other properties will also come into view. For instance, the spins of the electrons surrounding the cobalt atom also form a unique pattern, but the STM isn't yet sensitive to those signatures. Manoharan thinks it may also be possible to place a simple molecule, such as carbon monoxide, at one focus, and detect the characteristic vibrations of its atoms at the other.

It's one of the golden rules of physics that you can't probe the properties of an atom without simultaneously affecting that atom in some way. Does the discovery of mirages change all that? Might it be possible to sense some properties of the real atom by scanning the mirage, leaving the atom itself unmolested? Sadly, the answer seems to be no. Whenever physicists investigate the mirage site, about 10 billion electrons flow through the STM tip into the corral each second, flooding the ellipse and influencing the real atom. "Anything you do at one focus is felt by the other," observes Paul McEuen of the University of California, Berkeley. "You're not getting a free lunch." Still, remote sensing disturbs the real atom much less than placing the STM tip directly above it.

For IBM, Eigler's work has important implications in a different direction. The company has its sights set on one day building electronic circuit out of assemblies of atomic structures rather than wires, and for this the quantum corrals show obvious promise. As Eigler points out, the corral transmits information. That could be very useful when it comes to achieving the electronics industry's goal of building components measuring a mere 100 angstroms (10 nanometres) across. This, Eigler says, is still "many generations" away.

"How do I build something that adds two numbers together, has input and output channels, and fits in a 100-angstrom package?" Eigler asks. Eliminating wires would be a great start, as IBM has not been slow to point out. Its news release on the Nature paper touted the discovery of a "nanotech communication method". That's rather premature, as the researchers can't yet code information in a way that would allow it to be sent from one focus of the ellipse to the other. But there appear to be plenty of possibilities. "We could use the location of the atom, the energies of the electrons, the location of the electrons, the spins of the electrons, or a complex mixture of those," says Lutz.

Upping the level

Other STM physicists were impressed when they learned of this latest work from Eigler's lab. "My first reaction was that this is simply fantastic," says Wolf-Dieter Schneider. "It's a marvellous and novel aspect of how magnetic atoms behave." Paul McEuen says that Eigler's team has "upped the level of the game" for precise manipulation of atoms. "This shows that you can tailor the interactions between different atoms with an exquisite degree of control," he says.

In the past few months, Manoharan and his colleagues have been doing what he calls "all kinds of weird experiments." Their tinkering was motivated by the musings of physicist Charles Rettner, a colleague at Almaden, that perhaps another atom could form a molecule with the mirage. After all, the mirage is the projection of an electronic structure. "We know that those structures underlie chemical bonding," Manoharan explains. "If we brought a real atom next to the mirage, would it make a chemical bond?"

Though this initially struck Manoharan as unlikely, subsequent research has muted his scepticism. In one experiment the team observed what appears to be a magnetic interaction between two real atoms within the corral. Manoharan says that this interplay is "a step below chemical interactions", suggesting that a real chemical bond might be possible.

The team has not yet attempted to form a full chemical bond between a real atom and the mirage. "It might work, depending on how much of the electronic structure we're projecting," says Manoharan. "But we are not projecting the nucleus of the atom or the orbitals of the electrons."

Ultimately, turning quantum corrals into electronic components will require mass production on some sort of atomic assembly line. Eigler's group doesn't plan to try that, at least not soon. But across the continent at the National Institute of Standards and Technology in Gaithersburg, Maryland, another STM group is moving in that direction. Joseph Stroscio and his colleagues have completed a machine based on a cryogenic microscope that will assemble corrals and other structures containing many thousands of atoms. The system will start by randomly depositing atoms on a surface. Then, armed with blueprints of the desired structures, a computer will direct the STM tip to build them in the most efficient way. "It will do it overnight while we're sleeping," Stroscio says.

Once it's working, this set-up will let physicists produce countless atomic structures with subtly different configurations, allowing them to investigate how their behaviour varies. Stroscio's team also plans to apply strong magnetic fields to alter the dynamics of electrons in their corrals, which he expects will be 10 times the size of those in Eigler's lab. The hope is that physicists will be able to use these "quantum laboratories" to work out how electrons behave in different environments.

It's not easy to surprise physicists who routinely play with single atoms as if they were marbles. Yet the new frontier of manipulating the electronic properties of atoms has energised the STM community. "I'm amazed each day that I can build these structures," Manoharan says. "The prospect of what we will discover excites me to no end. We can go in any direction we want." *

Robert Irion is a freelance science journalist based in Santa Cruz, California

From New Scientist magazine, 08 July 2000.