Quantum computers are a sort of holy grail of information science. Their inherent computational advantage comes from their fundamental computational unit, the quantum bit ("qubit"). Unlike a digital bit in a classical computer, which can take the form of either 0 or 1, a qubit can be both zero and one simultaneously, throwing open the door to vastly more powerful computation. And although a usable computer based on qubits remains a far-flung fantasy, investigators continue to make strides toward their realization.
New research from University College London, the University of Utah and Florida State University in Tallahassee demonstrates another step toward the development of a quantum computer. The team showed that by using a powerful magnetic field and very low temperatures, below –450 degrees Fahrenheit (–270 degrees Celsius), they could read the state of electrons in a silicon wafer, potential qubits, using electrical current, and were able to extend the usable lifetime of those qubits dramatically. The paper appears in the journal Physical Review Letters.
Reading out the state of the qubit, encoded in a property known as spin, is one of the main challenges to quantum computing, says Stephen Lyon, a professor of electrical engineering at Princeton University who did not participate in this study. Specifically, he says, researchers will eventually need to be able to measure the spin state of a single qubit, much as classical computers can read and write to individual bits.
One way to do that is to "map the quantum information onto a current flowing through a device," says study co-author Dane McCamey, a postdoctoral physicist at the University of Utah. In other words, the researchers track a current flowing through the device as the spins are manipulated. Quantum readout, Lyon says, "almost certainly must be done electrically to get the sensitivity we'll need."
The group's setup swaps out a silicon atom for a phosphorus one in a silicon wafer, introducing an extra "donor" electron that can be manipulated and measured. Then, McCamey says, the researchers use millimeter-wave radiation to tweak the spin of the electrons while monitoring the current flowing through. Overall, he says, there are a few thousand qubit electrons in the spin sample—the system is not yet sensitive enough to detect the spins of individual qubits or even small groups of them.
The advantage of using silicon is that "it's compatible with conventional semiconductor devices," McCamey says. "You can use a lot of the same processing techniques." Bruce Kane, a senior research physicist at the University of Maryland's NanoCenter, agrees: "The work is a significant step towards showing that electronic devices (like the transistors in conventional electronics) may one day be used to measure single spin qubits in silicon."
The problem, McCamey and his co-authors wrote, is that the spin of the donor electrons doesn't stay electrically readable for long, just two millionths of a second in previous studies using this kind of detection. The team extended the life of the spins 50 times by applying a strong magnetic field, some 25 times the strength of those used in previous experiments, to align the spins, along with lowering the temperature.
Kane stresses that the research doesn't yet put the field at the doorstep of practical quantum computing. "While this is an encouraging result," he says, "it is still a long way from what is necessary for measuring electron qubits in silicon." Scaling down to single-spin sensitivity, he says, will be a challenge, especially given the propensity for electric currents to interfere with the spins.
"We still have quite a ways to go before we can reliably read out single donors," Lyon notes, adding that this "is an important step along that path."