One of the key engineering activities to achieve qubits that have the longest coherence times is materials engineering to produce purified substances that will help reduce the impact of things that will degrade the coherence time of a qubit. For spin qubit technologies a key material that is important is the starting silicon wafer upon which the qubit structures are built. Natural Silicon consists of three isotopes including Si-28 (92.2%) which has zero nuclear spin, Si-29 (4.7%) which has a spin state of 1/2, and Si-30 (3.1%) which has also has zero spin. Natural silicon is perfectly fine for semiconductor processing since transistors are not sensitive to nuclear spins. But for spin qubits built on silicon wafers, the Si-29 is a problem since the 1/2 spin state will hurt the coherence times.

So Australian company, Silex Systems Limited, has construction of a pilot demonstration facility that uses a SILEX laser isotope separation (LIS) platform technology to remove the Si-29 and provide customers with starting silicon that has no residual spin. They are collaborating with the University of New South Wales (UNSW) and commercial company Silicon Quantum Computing (SQC) and are receiving support from the Australian federal government. They are now producing small quantities of this silicon, which they call ZS-SI. The pilot facility will have the capacity to produce 5 kilograms of ZS-SI per year and the company has the eventual goal of scaling it up for larger quantities once the technology has been optimized. Some of the initial production was used in the demonstration by SQC of their analog quantum processor chip last month.

Additional information about Silex and the completion of their pilot facility is available in a press release located here and a web page that describes the overall project here.

July 21, 2022