Quantum computing holds great promise as a way to factor huge numbers, potentially breaking ultra-secure cryptographic codes unbreakable by traditional computers. However, this promise has historically been tempered by practical concerns: quantum computers rely on particles and molecules that are extremely sensitive to the environment; therefore, any such system only works for milliseconds, and the more particles and ions are added to a system, the quicker its ability fades.

But now researchers at the National Institute of Standards and Technology (NIST) have demonstrated, for the first time, that the lifetime of quantum-computing bits, known as qubits, can be extended using simple operations. In their experiment, they showed that by applying specially timed magnetic pulses to qubits, made of beryllium ions, they could prolong the life of the quantum bits from about one millisecond to hundreds of milliseconds. The work is described in this week's Nature.

"The worst thing about quantum information from an experimental perspective," says Michael Biercuk, a researcher at NIST, "is that even if you do nothing to your qubit, just its interaction with the environment does something to it." Qubits, he explains, are dependent on the quantum magic of superposition, in which certain properties of a quantum system exist in two or more distinct states at once. Superposition is a fleeting thing, and it quickly starts to break down, or decohere, due to noise such as random electrical fluctuations in the environment, says Biercuk. But what Biercuk and his colleagues John Bollinger and Herman Uys have done is "mitigate the effects of decoherence."

This means that the researchers have bought some time to do more complex experiments, such as modeling quantum states of large molecules, says Biercuk. It also means that they could add more qubits to the system, essentially providing more computational horsepower, and still have enough time to perform some experiments. Additionally, notes Biercuk, the team showed that it's possible to use the approach for different types of quantum-computing systems, such as those that are built in a semiconducting material like silicon. In other words, the researchers have provided a general solution to a problem that plagues all researchers who work on quantum computers.