A quantum computing platform that is capable of the simultaneous operation of multiple spin-based quantum bits (qubits) has been created by researchers in South Korea. Designed by Yujeong Bae, Soo-hyon Phark, Andreas Heinrich and colleagues at the Institute for Basic Science in Seoul, the system is assembled atom-by-atom using a scanning tunnelling microscope (STM).
While quantum computers of the future should be able to outperform conventional computers at certain tasks, today’s nascent quantum processors are still too small and noisy to do practical calculations. Much more must be done to create viable qubit platforms that can retain information for long enough for quantum computers to be viable.
Qubits have already been developed using several different technologies, including supercomputing circuits and trapped ions. Some physicists are also keen on creating qubits using the spins of individual electrons – but such qubits are not as advanced as some of their counterparts. However, that does not mean that spin-based qubits are out of the running.
“At this point, all existing platforms for quantum computing have major drawbacks, so it is imperative to investigate new approaches,” explains Heinrich.
Precise assembly
To create a viable spin-based processor, qubits must be assembled precisely, coupled together reliably, and operated in a quantum-coherent manner, all on the same platform. This is something that has so far eluded researchers, until now – according to the Seoul-based team.
The researchers created their multi-qubit platform with the help of an STM, which is a powerful tool for imaging and manipulating matter on atomic scales. When the conducting tip of an STM is brought very close to a sample surface, electrons are able to quantum-mechanically tunnel between the tip and the sample surface.
Since the probability of tunnelling strongly depends on the distance between tip and surface, an STM can map out the sample’s nanoscale topography by measuring the current of these tunnelling electrons. Individual atoms on surface can also be manipulated and assembled by pushing them around by the nanoscale forces applied by the tip.
Using these capabilities the team has “demonstrated the first qubit platform with atomic scale precision,” according to Heinrich. “It is based on electron spins on surfaces, which can be placed at atomically precise distances from each other.”
Sensor qubit
Using STM, the researchers assembled their system on the pristine surface of a magnesium oxide bilayer film. The system includes a “sensor” qubit, which is a spin-1/2 titanium atom that is located directly below the STM tip. The tip is coated in iron atoms, which means that it can be used to apply a local magnetic field (see figure).
To either side of the tip are a pair of “remote” qubits – also spin-1/2 titanium atoms. These are placed at precise distances from the sensor qubit, outside the region where electron tunnelling between atoms can occur.
To control the remote qubits simultaneously with the sensor qubit, the team created a magnetic field gradient by placing iron atoms nearby. The iron atoms behave as single-atom magnets because their spin relaxation times far exceed the operation times of individual qubits.
In this way, the iron atoms each act as a substitute for the STM tip in providing a static, local magnetic field for aligning the spins of each remote qubit. Transitions between the spin states of the qubits are done by using the STM tip to apply radio-frequency pulses to the system – a technique called electron spin resonance.
Addressed and manipulated
The team initialised their qubits by cooling them to 0.4 K, then applying an external magnetic field to bring them into the same spin state and coupling them together. Afterwards, the state of the sensor qubit depended reliably on the states of both remote qubits, but could still be addressed and manipulated individually by the STM tip.
The overall result was entirely new qubit platform that allowed multiple qubits to be operated simultaneously. “Our study has achieved single qubit, two qubit, and three qubit gates with good quantum coherence,” Heinrich says.
He adds that, “the platform has its pros and cons. On the pros, it is atomically precise and hence can be easily duplicated. On the cons, the quantum coherence is good but needs to be improved further.”
If these challenges can be overcome, Heinrich and colleagues see a bright future for their system.
“We believe that this approach can relatively easily be scaled to tens of electron qubits,” Heinrich says. “Those electron spins can also be controllably coupled to nuclear spins which might enable efficient quantum error correction and increase the available Hilbert space for quantum operations. We have just scratched the surface!”
The research is described in Science.