A team of quantum computing engineers from UNSW and the University of Melbourne has broken a new record in precision manufacturing, in the quest to fabricate the silicon chips of future quantum computers.

In a paper published in Advanced Quantum Technologies recently, the team showed that phosphorous ions can be implanted in silicon crystal with 99.95 confidence, while at the same time being precisely located within the chip.

The achievement is significant because the manufacturing method, based on ion implantation, is the same as that already used today to make the binary bits in conventional computers. It thus offers a flexible and scalable way to fabricate donor qubits for quantum computers without having to ‘reinvent the wheel’.

The team lead by Professor Andrea Morello and Professor David Jamieson, leads the world in the development of silicon quantum bits based on the spin of individual phosphorus atoms in silicon. These are exceptionally good qubits – they support high-fidelity quantum operations, hold quantum information for long times, etc – but fabricating them reliably poses a technical conundrum.

In order to have high confidence that the atom has been placed in the crystal, one wants to accelerate it into the chip with a high energy. On the other hand, a high-energy atom shot into the chip will have a large uncertainty in the position where it finally stops.

Dr Holmes, posing as the stationary phosphorus atom and Professor Morello (left) and Benjamin Willhelm representing the fluorine atoms diffusing away.

The new paper, led by Dr Danielle Holmes, shows a way to resolve the conundrum. Instead of implanting a phosphorus ion by itself, the team implanted the molecular ion PF2+, where phosphorus (P) is accompanied by two fluorine (F) atoms. The molecule is shot into the chip at a high energy, which allows detecting the impact with 99.95 per cent confidence. However, less than half of that energy belongs to the P atom. The molecule breaks apart as soon as it hits the surface of the chip, and from there, the P atom travels slowly and comes to a rest in a well-defined position.

Furthermore, the researchers were able to prove that “the two stray fluorine atoms do not pose any problems to the operation of the phosphorus quantum bits”, according to Benjamin Wilhelm, a UNSW PhD student who performed the quantum measurements. After the implantation, the device is subjected to a quick (5 seconds) very high-temperature process (1000 degrees Celsius), which makes the F atoms diffuse away, whereas the P atom remains in place.

By increasing the placement precision without compromising on the detection signal, “we can have our cake and eat it too”, concludes Dr Holmes.