Uncovering the Hidden Disorder in Silicon Quantum Computers

Researchers map the variation in materials to the atom level to scale up silicon qubits

August 4, 2026

Estimated Read Time min

The Chicago Quantum Computing Testbed at Argonne National Laboratory is the first full-stack, solid-state qubit testbed at a U.S. National Laboratory. Researchers used it to measure industrial-grade qubit wafers and reveal the origin of qubit failure.

Image courtesy of Argonne National Laboratory

The Science

Qubits are the quantum counterpart to the bits used in conventional computers. Bits have a “0” and “1” state that is defined by electric charge. In a type of qubit called a silicon spin qubit, the “0” and “1” states are defined by electron spin. This spin can point either up or down in a magnetic field, analogous to a tiny compass needle. Scientists build silicon spin qubits by trapping a single electron inside a thin layer of silicon. The thin layer of silicon (called a quantum well) is sandwiched between another semiconductor material. In addition to spin, electrons in silicon also have a quantum property called a valley state. The energy difference between these valley states is called valley splitting. Valley splitting competes with the spin states used for computation. If the valley splitting is too small, the electron can leak into unwanted valley states. This leakage causes errors and loss of fidelity. In this study, researchers examined how the quantum well affected valley splitting.

The Impact

Because silicon spin qubits build on the same technology that underpins today’s semiconductor industry, they are one of the most promising platforms for scalable quantum computing. For years, researchers knew that defects and inconsistencies in the materials used in silicon quantum devices reduce valley splitting and cause failure. This valley splitting has long been known to vary from device to device, but its origin remained unclear. This study revealed that disorder on the atomic scale in the quantum well is the main source of variability of valley splitting. By identifying this root cause, the work turns a long-standing challenge into a tractable materials problem. It gives industry and National Laboratories a clear path towards building more reliable, higher-fidelity silicon qubits.

Summary

This study was enabled by a unique partnership in which Argonne National Laboratory used the Chicago Quantum Computing Testbed (managed by the Q-NEXT National Quantum Information Science Research Center) to study and measure an industrially fabricated 12-qubit class silicon quantum dot processor from Intel. It combined National Laboratory measurement and analysis expertise with Intel’s manufacturing capability. Using the Chicago Quantum Computing Testbed, researchers studied industrial-grade silicon quantum devices to determine what limits valley splitting in practice.

The team used a sensitive electrical spectroscopy method to measure the valley splitting while shifting the position of a quantum dot along the quantum well. This allowed them to build a nanoscale map of how valley splitting changes across the material. By analyzing how these variations correlate with distance, the researchers identified random atomic-scale fluctuations in the alloyed quantum well as the dominant source of valley splitting variability.

This work transforms valley splitting from an unexplained obstacle into a materials engineering challenge with clear paths toward improved silicon qubits.

Contact

Jonathan Marcks
Ass. Staff Scientist, Q-NEXT National Quantum Information Science Research Center, Argonne National Laboratory
jmarcks@anl.gov

James S. Clarke
Director of Quantum Hardware, Intel Corporation
james.s.clarke@intel.com

Martin Holt
Director, Q-NEXT National Quantum Information Science Research Center, Argonne National Laboratory
mvholt@anl.gov

Funding

Funding was provided by the U.S. Department of Energy, Office of Science, National Quantum Information Science Research Centers as part of the Q-NEXT center.

Publications

Marcks, J.C., et al., “Valley splitting correlations across a silicon quantum well containing germanium.” Nature Communications (2025). [DOI: 10.1038/s41467-025-67325-z]