Studying sources of energy loss to make quantum computing gains

Scientists from Yale University and the Brookhaven National Laboratory of the US Department of Energy (DOE) have developed a systematic approach to understand how energy is lost from the materials that make up qubits. Energy loss inhibits the performance of these quantum computer building blocks, so determining their sources—and adjusting the materials as needed—could help bring researchers closer to designing quantum computers that could revolutionize several scientific fields.

With their new approach, the Yale scientists were able to design a compact device that could store quantum information for more than one millisecond.

This research, published in Nature communicationwas carried out as part of the Co-Design Center for Quantum Advantage (C²QA), a national research center for quantum information science led by Brookhaven Lab. Yale is an important partner of the center.

“An important hurdle we have to overcome is improving the ability of qubits to retain the quantum information in them. This is known as coherence,” explained Suhas Ganjam, who is the first author of the new paper. Ganjam did the research as a doctoral student at Yale and is now a research scientist at Google.

A few years ago, researchers from Princeton University – who at C²QA at its inception – designed qubits with a record-breaking coherence time of 0.3 milliseconds by replacing the traditionally used niobium or aluminum with a superconducting metal called tantalum. This indicated that the constituent materials of qubits directly affect their performance, but the reasons for this were still unclear.

So, scientists who contribute to C²QA began researching the various types of tantalum oxides that form on the surface of tantalum when it is exposed to air. They further improved the cohesion by coating tantalum with a thin layer of magnesium that prevented the oxidation of the material.

“Researchers have built devices with better coherence times. But there are so many different sources of energy loss, and we still couldn’t distinguish which one to improve,” said Ganjam. “So, we set out to distinguish between the different types of loss.”

Under the supervision of Robert Schoelkopf, a physicist at Yale University who leads the Devices Thrust of C²QA, Ganjam designed a device called a tripole strip line.

This new device consists of three superconducting thin-film strips patterned on a substrate, similar to other quantum devices. The strips were arranged in a special way so that the researchers could not only quantify energy lost, but also determine where it was lost by testing the device in three different modes – one for each pair of superconducting electrodes.

For example, the researchers could distinguish between surface loss and bulk dielectric loss by observing modes in which electromagnetic fields were either limited to the surface of the device or spread over the substrate. If they observed more loss from the mode in which electromagnetic fields were confined to the surface of the device, the loss was dominated by the surface contribution.

“Through our electromagnetic tests with the tripole strip line, we were able to observe that devices made with tantalum and aluminum lose different amounts of energy in different ways,” explained Ganjam.

In particular, the researchers found that using a thin tantalum film, instead of a thin aluminum film, reduced the surface loss. And the use of a fabrication technique called annealing, which involves heating a sapphire substrate and letting it cool slowly, reduced the bulk dielectric loss.

“We wanted to know why the different materials and manufacturing techniques affected losses like this,” Ganjam said. “So, we turned to our employees at the Center for Functional Nanomaterials.”

Quantum materials through the microscopy lens

The Center for Functional Nanomaterials (CFN) is a DOE Office of Science user facility at Brookhaven Lab with a state-of-the-art Electron Microscopy facility. Using transmission electron microscopy and scanning transmission electron microscopy to look at the microscopic structure of the materials, scientists at this facility can help other researchers, such as Ganjam and Schoelkopf, better understand the materials they work with.

“We suspect that qubit coherence is limited by energy loss caused by contaminants or defects in the materials,” explained Minghzao Liu, a senior scientist at CFN. “So, we analyze the quantum materials at CFN to look for these coherence-limiting features.”

Kim Kisslinger, an advanced engineering fellow at CFN, extracted microscopic cross-sections of the Yale scientists’ materials and devices and analyzed them at the atomic level.

“I look at projects like this through an electron microscopy lens,” Kisslinger said. “From crystallinity to chemical composition to epitaxy, which is related to the orientation of the crystal materials, I can tell our collaborators exactly what is going on with their materials and help them correlate these properties with the performance of the materials .”

Liu said, “Kim helps our collaborators better understand their materials, but he also helps them make meaningful improvements through an iterative process.”

Kisslinger added: “CFN is home to state-of-the-art equipment that can support the materials research needed for quantum devices. But we also have some of the most qualified scientists and specialists in the world. This combination of quality people and quality equipment is unique to CFN.”

Collaborative efforts yield improved devices

With a well-rounded understanding of the electromagnetic properties of their devices, as well as the material composition, the Yale researchers used an energy loss model that could predict the coherence of a device based on its constituent materials and the geometry of the circuit. And using this predictive model, they optimized circuit geometry to build a quantum device with a coherence time greater than one millisecond.

“This research marks an important milestone in the C²QA mission. Even beyond the longer coherence time, it shows a path forward to further improvements in coherence through the close collaboration of quantum devices and materials scientists,” said C.²QA Deputy Director Kai-Mei Fu.

The collaboration between the Schoelkopf lab’s qubit design experts and CFN material characterization experts, which began with the establishment of the center, embodies C²QA’s principle of “co-designing” materials and algorithms to achieve quantum computers that outperform classical computers.

“Collaborations like these are key to unlocking the best materials and optimal manufacturing processes that C²QA are realizing their goal,” Ganjam said.

“It has been quite rewarding to see these qubit design projects grow in scope and success over the years,” added Liu. “Scientific progress like this is not possible without collaboration.”