A team of researchers at UC Santa Barbara has joined a new $27.9 million initiative aimed at overcoming a fundamental bottleneck in quantum computing, as Quantum Zeitgeist reported. The newly formed MARQUIS institute, supported by a five-year grant from the National Science Foundation, unites 24 laboratories across nine institutions to overhaul the microscopic layer inside Josephson junctions.
The initiative focuses on the Josephson tunnel junction, a component just a few atoms thick that forms the core of superconducting qubits. With historical obstacles like wiring losses and electromagnetic noise largely addressed, researchers now identify the junction as the primary limitation in scaling up quantum devices. Nobel laureate and UCSB physics professor Michel Devoret told Quantum Zeitgeist that the Josephson tunnel junction has finally become the weakest element in a superconducting quantum processor.
At UC Santa Barbara, the effort brings together multidisciplinary experts, including Devoret, Chris Van de Walle, and Chris Palmstrøm. Their work spans computational modeling of materials at the atomic scale, advanced material growth, and quantum device behavior. This collaborative structure provides the team with the freedom and tools to attack complex quantum manufacturing problems from fundamentally different directions.
The institute's primary objective is to move past incremental engineering tweaks and replicate the high standards of consistency achieved in traditional semiconductor manufacturing. Modern quantum algorithms require robust two-qubit error correction, which researchers calculate will demand a tenfold improvement in Josephson junction quality. By focusing on material science, scientists can test experimental designs without interference from other system flaws.
Van de Walle, a Herbert Kroemer Distinguished Professor at UCSB, is directing computational research to locate the microscopic origins of defects that degrade processor performance. His previous work on light-emitting diodes demonstrated that defects existing in ratios under one in a million atoms can measurably reduce efficiency. That insight is directly informing the team's effort to resolve energy-absorbing defects and microscopic two-level systems inside quantum materials.
To manufacture higher-quality materials, Palmstrøm’s research group is utilizing molecular beam epitaxy to fabricate single-crystal superconductors and dielectric heterostructures. Operating seven interconnected molecular beam epitaxy systems, his laboratory can transport samples under ultrahigh vacuum to layer atomic structures with precise control. This process allows researchers to grow exceptionally pure single-crystal silicon to potentially replace aluminum oxide, the traditional one-nanometer insulating layer in junctions that often suffers from structural disorder.
Replacing aluminum oxide with single-crystal silicon could permit a thicker dielectric layer, reducing microscopic performance variations caused by tiny fluctuations in thickness. Additionally, Palmstrøm's team is designing fabrication pathways that are compatible with existing 300-millimeter silicon manufacturing tools. Among their experimental designs is a structure featuring an etched silicon fin hosted within deposited superconducting material, a concept that could shrink the physical footprint of a transmon qubit by more than two orders of magnitude.
Researchers note that reliance on the same basic junction design for roughly 25 years demonstrates both its durability and the urgent need for structural innovation. By resolving other potential failure modes, scientists believe 2026 offers an unprecedented opportunity to focus entirely on building superior junctions. With Devoret establishing precise experimental guardrails, UCSB aims to leverage its history of converting laboratory breakthroughs into reproducible manufacturing technologies, paving the way for scalable quantum processing.
