Scientists have demonstrated that a new form of qubit structure, which is easier to mass-produce, can perform as well as the currently leading qubit types. Through various mathematical evaluations, they have created a guide to streamline the production process, allowing for strong and dependable manufacturing of these fundamental components in quantum computing.
Researchers from the U.S. Department of Energy’s (DOE) Brookhaven National Laboratory have revealed that a new qubit design, better suited for mass production, can match the performance of the qubits that are presently the most prevalent in the industry. They have conducted extensive mathematical studies to outline a simplified qubit manufacturing approach that supports the efficient and consistent production of these key quantum computing elements.
This study is part of the Co-design Center for Quantum Advantage (C2QA), a DOE National Quantum Information Science Research Center led by Brookhaven Lab. It builds on many years of collaboration aimed at enhancing qubit performance for scalable quantum systems. Recently, researchers have been focused on prolonging the retention of quantum information in qubits, a quality termed coherence, which is closely related to the performance of the qubit’s junction.
Specifically, they have been concentrating on superconducting qubits that feature two superconducting layers separated by an insulator. This structure is referred to as an SIS junction, standing for superconductor-insulator-superconductor. However, the reliable production of these sandwich-like junctions proves to be challenging, especially when precision manufacturing is needed for the mass production of quantum computers.
“Creating SIS junctions requires great skill,” remarked Charles Black, a co-author of the paper published in Physical Review A and director of the Center for Functional Nanomaterials (CFN), a user facility at Brookhaven Lab operated by the DOE Office of Science.
Black and Mingzhao Liu, a senior scientist at CFN and the lead author of the paper, have been part of C2QA since its launch in 2020. While they have assisted quantum scientists in understanding the material properties of qubits to enhance their coherence, they have also become interested in the scalability of this qubit manufacturing technique and its alignment with the necessity to produce large-scale quantum computers.
Consequently, the team shifted their focus to qubit designs featuring superconducting junctions made of two layers linked by a thin superconducting wire, eliminating the insulating layer. Known as a constriction junction, this design lies flat rather than stacked like a sandwich. Importantly, the fabrication process for constriction junctions aligns well with standard techniques used in semiconductor manufacturing facilities.
“In our research, we looked into how this design change affects performance,” said Black. “Our aim was to evaluate the performance trade-offs involved in switching to constriction junctions.”
Addressing the challenges of increased current flow and linearity
The most common superconducting qubit design performs optimally when the junction between the two superconductors transfers a minimal amount of current. In the SIS junction, the insulator prevents almost all current passage but is thin enough to allow a small amount through a process called quantum tunneling.
“The SIS configuration is perfect for the superconducting qubits we have today, despite its complex production,” stated Black. “Yet, it seems counterintuitive to substitute the SIS with constriction, which inherently allows significantly more current to flow.”
Through their investigations, the researchers revealed that it is feasible to limit the current traversing a constriction junction to suitable levels for a superconducting qubit. However, achieving this requires using nontraditional superconducting metals.
“If we were to use aluminum, tantalum, or niobium, the constriction wire would have to be unrealistically thin,” Liu explained. “Conversely, employing other superconductors that conduct less efficiently would allow us to manufacture the constriction junction at practical sizes.”
However, constriction junctions act differently than SIS junctions. As a result, the team also studied the implications of this design shift.
For superconducting qubits to function, they require some degree of nonlinearity, which confines the qubit to operate between only two energy levels. Typically, superconductors do not exhibit nonlinear properties; this essential characteristic is introduced by the qubit junction.
Superconducting constriction junctions are naturally more linear than traditional SIS junctions, making them less suitable for qubit designs. Nevertheless, the researchers discovered that the nonlinearity of constriction junctions can be adjusted by selecting different superconducting materials and designing the junction’s size and shape appropriately.
“We are enthusiastic about this work because it directs materials scientists towards specific goals based on device requirements,” Liu elaborated. For instance, they identified that for qubits operating within the range of 5 to 10 gigahertz, which is typical in modern electronics, specific trade-offs must be considered between the material’s electrical conduction ability and the junction’s nonlinearity.
“Some material property combinations are not feasible for qubits operating at 5 gigahertz,” Black noted. However, with materials that match the criteria established by Brookhaven scientists, constriction junction qubits can achieve performance comparable to that of SIS junction qubits.
Liu and Black are currently collaborating with their colleagues in C2QA to research materials that fulfill the specifications laid out in their recent paper. They have particularly shown interest in superconducting transition metal silicides, as these materials are already employed in semiconductor production.
“Our study illustrated that we can address the concerning aspects of constriction junctions,” Liu said. “Now we can start to leverage the benefits of simpler fabrication processes for qubits.”
This research exemplifies the foundational co-design principle of C2QA, as Liu and Black investigated a qubit structure satisfying the requirements of quantum computing while corresponding with existing electronics manufacturing practices.
“These kinds of interdisciplinary partnerships will keep bringing us closer to making scalable quantum computers a reality,” Black concluded. “It’s almost surreal to think about the advancements in quantum computing we have achieved so far. We’re thrilled to contribute to the goals of C2QA.”
