By Dr. Andre Saraiva, UNSW

For a new proposal to make it out of the starting line in the race of quantum computing, every engineer and scientist asks the same question: what is your qubit coherence time? This tells them how long the qubits can hold in a superposition of 0s and 1s before they become as dull as a flipped coin. A study led by Prof. David Awschalom from the University of Chicago showed a technique to improve this time by a factor of 10,000.

The qubits studied at the Awschalom group were never too bad. Using spins of divacancies, which are natural defects in silicon carbide (pairs of missing atoms in the SiC crystal), they have been getting few-microseconds-long coherence times for the past decade. This means that it takes a few microseconds for any small fluctuations in the qubit frequency to build up to a significant error.

The coherence time of most qubits gets much better if a technique called dynamical decoupling is used – the spin is periodically turned upside down and if it rotates too fast or too slow, it will compensate by rotating now in the opposite direction. (View this video for an explanation of dynamical decoupling.)

In this study, the first author Dr. Kevin Miao and collaborators went a step beyond – now the qubits are no longer defined as a spin pointing up or down. Instead, their qubits are constantly being flipped by an external microwave field, and the definition of a 0 or 1 is based on whether they rotate in one direction or the other. In this case, the qubit is made from a spin and the microwave field that drives it, which in the language of quantum electromagnetism is called a dressed spin.

This is an idea that was explored in other qubit systems as well (see here, here, and here), but never using a clock transition. Clock transitions are a property of the defects that makes them insensitive to fluctuations in electromagnetic fields. This means that their resonance frequency remains very stable – a quality that has been exploited for accurate time keeping, hence the name. It also means that the coherence of the qubit can be preserved over much longer times. In fact, the Awschalom group showed that their dressed qubits sport coherence times 10000 times longer than their bare spins.

This is an important accomplishment for this technology as a memory qubit. Silicon carbide divacancy spin qubits are operated at the relatively high temperature of 5 kelvin (compared to the millikelvin regime of most qubits), have a very small footprint (few nanometers) and are based on a solid-state matrix, which fuels aspirations for scalability.

So, what’s next?

For SiC divacancy technology to take off, the main challenge is to make them interact strongly with photons in a cavity. This way, conversion between this memory register and photons can be used to generate long range coupling of divacancies or to couple divacancies to some other qubit platform, such as superconductors. The Awschalom group has made some recent progress in this direction, reaching the necessary long lifetime for the spin states in a cavity, but the holy grail of one-to-one conversion of spin qubits to photons is still to be seen.

For additional information, click on the link to view an abstract of the paper titled “Universal coherence protection in a solid-state spin qubit”.

Dr. Saraiva has worked for over a decade providing theoretical solutions to problems in silicon spin quantum computation, as well as other quantum technologies. He works on MOS Silicon Quantum Dot research and commercially-oriented projects at the University of New South Wales (UNSW).

September 6, 2020