In a flurry of research reports during the past six months, physicists have proven that silicon, the basis of computers today, could also be the best platform for tomorrow's quantum computers.
Such computers would use the quantum properties of atoms or molecules to perform calculations in a fraction of the time it would take conventional computers. However, so far only rudimentary quantum computers have been built, comprising only a few quantum bits (qubits) and built in exotic systems such as ion traps, cryogenically cooled superconductors, and optical tweezers.
Silicon could provide a useful path to systems with 100 or more qubits, say some scientists, because it would make quantum computers easily compatible with conventional ones. The silicon solution originated in 1998, when Bruce Kane, a physicist at the University of Maryland, in College Park, suggested making a qubit from the nuclear spin—a quantum property similar to magnetic moment—of the phosphorus atoms with which silicon is often doped.
In the past few months, researchers have reported progress in using the phosphorus-in-silicon system. In the latest development, a team of physicists led by John Morton of the University of Oxford reported that by using bursts of radio waves, they have managed to entangle the spins of 10 billion pairs of electrons and nuclei in a crystal of phosphorus-doped silicon. Entanglement is a phenomenon that allows quantum particles to be interlinked even if they are separated. It is used in quantum computing, along with another quantum phenomenon called superposition, to create qubits that can exist in many different states at the same time. The experiment is being hailed in the quantum computing community as a promising step toward silicon-based quantum computers.
According to Morton, the main advantages of his group's design are that it integrates easily with ordinary silicon circuits and that it produces qubits that last for a few seconds. In many other quantum systems, qubits last only milli- or microseconds, which makes it difficult to perform calculations.
Dane McCamey of the University of Sydney, whose research involves a similar system, says that what is important about Morton's work "is the generation of a large number of identical entangled pairs." McCamey and other experts, such as Stephen Lyon of Princeton, say these pairs could pave the way to a form of quantum computing where large entanglements are generated and then a series of precise measurements on individual qubits lead to massively parallel processing.
Though the Oxford experiment produced 10 billion sets of entangled pairs, the number that were usable as qubits was small, and that is unlikely to change soon.
Raymond Laflamme, executive director of the Institute for Quantum Computing at the University of Waterloo, in Ontario, Canada, cautions: "Some people are making the leap that we will have silicon quantum computers soon. We are on the right track, but the track is a long one."
About the Author
Saswato R. Das, author of the Update story on spintronic advances, contributes frequently to IEEE Spectrum. For one assignment, he interviewed science fiction writer Arthur C. Clarke. That 2008 interview turned out to be the futurist's last. "He was in poor health, but his mind was still very sharp," Das says. Based in New York City, Das has written for The Economist, Scientific American, and New Scientist.