27 June 2012
R. Colin Johnson
Separate labs in the U.S. and Europe recently reported progress in adapting solid-state materials to store spintronic quantum states, a critical hurdle on the path to using spintronics in quantum computing.
Many researchers believe that spintronics for quantum computing is the most promising way forward for future computer chips, but few have reliably cast them into solid-state materials. Unfortunately, the most successful experiments today use ultra-cold gases to store quantum spin-states. However, semiconductor R&D labs worldwide are aiming to recast spintronics into traditional solid-state materials.
Researchers at the City College of New York (CCNY) and the University of California-Berkeley (UCB) reported success using laser light to encode the spin-state of atomic nuclei on gallium arsenide chips. Using a technique whereby a scanning laser defines the spin-states on a gallium arsenide chip, the researchers claim they can set-up the initial conditions for a quantum computation that can be quickly reconfigured after completion.
The technique amounts to soft lithography, since it can reconfigure each quantum computation on-the-fly, according to the researchers. The group includes UC Berkeley professor Jeffrey Reimer and CCNY professor Carlos Meriles, along doctoral candiates Jonathan King of UC Berkeley and Yunpu Li of CCNY.
Such rewritable quantum computers would use the laser to encode their spin-states, thus suppressing the tendency of solid-state materials to lose their magnetization during computations. The researchers are currently experimenting with push-pull architectures that the laser could set in order to ensure that the quantum spintronic states remain stable until the end of a computation.
Separately, the current record holders for maintaining a quantum state in a solid-state material recently surpassed their own record, reporting encoded spin states that lasted over three minutes. The researchers at Simon Fraser University and Oxford University reported a 100-time improvement over their 2008 report of 1.75 seconds. Because their solid-state material is conventional silicon, professor Mike Thewalt at Simon Fraser (Canada) and professor John Morton at Oxford (U.K.) claim their technique could enable conventional CMOS manufacturing to eventually be harnessed for future quantum computers.
Both research groups encoded quantum states on the magnetic spin of atomic nuclei, on gallium arsenide and silicon chips respectively, rather than the more conventional approach of encoding spin states on electrons.