How ‘Clean’ Does a Quantum Computing Test Facility Need to Be? PNNL Scientists Show the Way

RICHLAND, Wash.— Now is the time to banish low-level radioactive energy sources from facilities that house and conduct experiments with superconducting qubits, according to a pair of recently published studies. Significantly improving quantum device coherence times is a key step toward an era of practical quantum computing.

Two complementary articles, published in the journals PRX Quantum and the Journal of Instrumentation, outline which sources of interfering ionizing radiation are most problematic for superconducting quantum computers and how to address them. The findings set the stage for quantitative study of errors caused by radiation effects in shielded underground facilities. 

A research team led by physicists at the Department of Energy’s Pacific Northwest National Laboratory, in collaboration with colleagues at MIT’s Lincoln Laboratory, the National Institute of Standards and Technology, along with multiple academic partners, published their findings to assist the quantum computing community to prepare for the next generation of qubit development.

Electronic ‘noise’ undermines efforts to extend qubit life

“The time is coming soon when advances in design and materials will result in qubits stable enough that environmental effects from stray radiation become the rate-limiting step to quantum coherence,” said physicist Brent VanDevender, one of the research team leads. VanDevender was one of the first scientists to identify natural ionizing radiation as a source of instability for the functioning of qubits, the basic unit of a quantum computer. 

Even the slightest interference can cause errors that lead superconducting qubits to lose their quantum state, a process known as decoherence. The research team found that cosmic radiation and naturally occurring isotopes, which emit low levels of ionizing radiation and are found in common materials, are about equally responsible for causing decoherence.

“Once we established the effect of ionizing radiation on superconducting qubits, we knew we needed to systematically and quantitatively identify sources of radiation in the environment,” said lead experimental physicist Ben Loer. “Our experience with measuring ultra-low levels of radiation in the laboratory led us to include radiation sources within the very experimental units, cryostats, where these experimental qubits are studied.”

“We found that a lot of the electrical connectors are just filthy dirty from the standpoint of acting as a radiation source,” VanDevender added. 

Cleaning up underground

Combined, the two studies point to effective measures for shielding sensitive experimental equipment from the effects of radiation exposure. 

In the Journal of Instrumentation, the research team describes the precautions taken to greatly reduce the potential for atmospheric and isotope radiation exposure in a shielded underground qubit testbed facility on the Richland, Washington campus of PNNL called the Low Background Cryogenic Facility. Constructed within an existing ultra-clean underground laboratory, the test bed includes a cryostat, also known as a dilution refrigerator, capable of cooling superconducting qubit devices to near absolute zero, a key to stabilizing quantum computing devices of this design. The research team reports that this lead-shielded cryostat could reduce the error rate by 20 times compared with the error rate seen in a typical above-ground, unshielded facility. 

In addition, the team reports that certain relatively simple precautions, such as eliminating natural sources of radiation within materials inside the dilution refrigerator, go a long way toward making quantum computing devices viable. These sources include metal isotopes—naturally occurring variants of elements that spontaneously eject radiation in the form of alpha, beta, and gamma rays—which can interfere with quantum devices.

In their exploration of these radiation sources within the lab, they used specialized ultrasensitive detection methods to identify contaminants in silicon, copper and ceramic electronic components such as circuit boards and cables used to collect data from instruments, and even the qubits themselves. To reduce the impact from these devices, the team advises using materials like brass instead of beryllium-copper alloys typically found in cables. Future goals of this research include testing the effectiveness of  “radiation-hardened” qubits that are less sensitive to the impact of radiation and studying low-background materials.

Transferring knowledge from sensitive detection technology

In the companion study, published in PRX Quantum, the research team directly measured ionizing radiation interactions on a superconducting sensor inside a cryocooler, a refrigerator that can reach ultra-cold cryogenic temperatures. They used simple radiation detection circuits printed on a piece of silicon similar to that used for qubits. Here they showed that stray radiation that interacts with a silicon circuit board and that could possibly cause decoherence in a qubit or other “unwelcome effects on circuit performance” matches well with the predicted rate and energy spectrum. 

The research team relied on expertise developed during design and building of double beta decay detectors, neutrino-physics”>neutrino detectors and dark matter detectors, which are similarly sensitive to low levels of radiation. The research team identified two complementary approaches to reducing the sensitivity of superconducting elements to stray radiation as first steps toward “radiation hardening:” isolating the superconducting elements on crystal “islands” and simply making the crystal substrate thinner. 

“We’ve demonstrated which sources of radiation are important, and we’re eager to see how new devices will perform in our low-background facility,” said Loer.

This research was supported by the Department of Energy, Office of Science, Nuclear Physics program’s Quantum Horizons program and an Early Career Award to Ben Loer, as well as the High Energy Physics program’s Quantum Information Science Enabled Discovery (QuantISED) program and internal investments by PNNL.