Volume 1, Issue 4
Quantum Computing's Magnetic Attraction
Tomorrow's nanocomputers may be incredibly powerful, but they'll also be mind-bogglingly strange. Instead of binary numbers, these machines will speak a language of "quantum bits" that can exist as zeros, ones, or both at the same time. By harnessing the unusual properties of quantum physics, quantum computers will perform calculations up to a billion times faster than today's silicon-based processors and store data in the spin of individual magnetic atoms.
Michael Crommie beside his customized Scanning Tunneling Microscope (STM).
UC Berkeley physicist Michael Crommie is bringing us closer to the next computer revolution by understanding and manipulating magnetism at the atomic level.
"The magnets that most of us know are simple things with a north and south pole," says Crommie, also a faculty scientist at Lawrence Berkeley National Laboratory (LBNL). "But when you shrink a magnet down to the size of a single atom or molecule, the properties change considerably. There's a whole world down there with a wealth of different phenomena that depend on very subtle interactions."
At the atomic scale, magnetism is known as "spin." An electron's spin is similar, at least conceptually, to the direction of a rotating top. It can either be "spin-up" or "spin-down" or, Crommie says "kind of up and kind of down." This spin state, called a "superposition" state, can be altered by a magnetic field.
In a futuristic quantum computer, Crommie explains, the direction of an electron's spin could be used as a quantum bit, or "qubit." The power of quantum computers lies in the qubit's ability to exist in a superposition state, representing multiple values at one time. This quantum weirdness is what enables quantum computers to process so much data at once. As more qubits are strung together, the power of the quantum processor grows exponentially.
"The big questions though are can we precisely control the quantum state of these microscopic structures, measure them, and connect them to create an extended circuit of quantum mechanically-interacting bits?" Crommie says.
To tackle these challenges, Crommie is collaborating with colleagues from the Physics Department, College of Chemistry, and College of Engineering on a $4.5 million grant funded by the National Science Foundation. The aim is to evaluate whether different schemes for the seemingly far-fetched technology will actually work.
In 1993, Crommie was part of a team that discovered a new method for confining electrons inside a "quantum corral" of iron atoms. (courtesy the researchers)
Crommie and collaborators Yossi Yayon, Xinghua Lu, and Andre Wachowiak are exploring ways to reliably control spin states of single atoms and molecules in the laboratory. The tool of the trade is the scanning electronic microscope (STM). Unlike the lens of an optical microscope, an STM has an extremely sharp tip as its probe. As the stylus scans across a sample, the amount of electrical current flowing between the tip and surface is measured. This enables a profile of the sample to be generated and visualized with atomic-scale resolution. STMs can also be used to push atoms around into desired structures. Last spring, Crommie and his collaborators used their STM to change the electrical properties of a single buckminsterfullerene molecule ("buckyball") by moving it on a surface where it picked up potassium atoms one at a time.
As a buckyball acquires potassium atoms, its energy state changes, causing it to "light up" in this STM image. (courtesy the researchers)
In their spin experiments, the researchers sandblast a crystal surface in a vacuum chamber until it's perfectly flat and clean. They then heat up a chunk of iron until the atoms jump off onto the crystal surface where the STM can be wielded to drag them into desired geometries. The researchers customized their STM with a magnetic tip that allows them to measure the spin of the atoms.
"Typically, the electrons jumping off an STM tip to measure the properties of atoms have spins that are arranged randomly," Crommie says. "But if the electrons coming off the tip can be made to have a net spin orientation, then you can detect the spin orientation of what's on the surface."
Most recently, the researchers used the STM to study clusters of magnetic molecules. The molecules consist of one or more magnetic atoms surrounded by organic molecular structures such as benzene rings or alkane chains. These outside structures can link one molecule's magnetic state with other neighboring molecules. Potentially, coupling the molecules could lead to a technique for controlling the spin states within the clusters.
For example, Crommie says, it may be possible to attach a magnetic molecule to tiny carbon nanowires so that the flow of electrons through a circuit is controlled magnetically rather than electrostatically, the basis of traditional electrical engineering. Spintronics, as the new paradigm is known, has already been used in a coarser way in computer hard disk drive heads to increase storage density. Nanocomputers, however, remain a quantum leap away.
"We're not building the world's smallest disk drives or the next quantum computer," Crommie says. "But we're figuring out what may be possible and the knowledge we collect might have applications for those future technologies."