Optimization of Si/SiGe Heterostructures for Large and Robust Valley Splitting in Silicon Qubits

Abstract:Small and device-dependent valley splittings remain a key challenge for electron spin qubits in silicon (Si), directly limiting qubit fidelity, device uniformity, and the scalability of Si-based quantum processors. In silicon-germanium (SiGe) heterostructures, this problem can be addressed through engineering of the epitaxial layer stack. Several heuristic strategies have been proposed to enhance the energy gap between the two nearly degenerate valley states in strained Si/SiGe quantum wells (QWs), e.g., sharp Si/SiGe interfaces, Ge spikes, or oscillating Ge concentrations within the QW. Here, we develop a systematic variational optimization approach to compute optimal Ge concentration profiles that enhance selected properties of the intervalley coupling matrix element. Our free-shape optimization framework is augmented by realistic technological constraints to ensure feasibility of the resulting epitaxial profiles and is based on an effective-mass envelope-function theory accounting for strain and compositional alloy disorder. Previously proposed heterostructure designs are recovered as special cases of the constrained optimization problem. Our main result is a novel heterostructure design, which we refer to as the "modulated wiggle well", providing both a large deterministic enhancement of the valley splitting and a strong suppression of disorder-induced variability. In addition, this design enables wide electrical tunability of the valley splitting - from approximately $200\,\mu \text{eV}$ to above $1\,\text{meV}$ - offering new opportunities for engineering robust and switchable silicon qubits.

Mesoscale and Nanoscale Physics (cond-mat.mes-hall); Optimization and Control (math.OC); Applied Physics (physics.app-ph); Quantum Physics (quant-ph)

arXiv:2512.18064 [cond-mat.mes-hall]

(or arXiv:2512.18064v2 [cond-mat.mes-hall] for this version)

https://doi.org/10.48550/arXiv.2512.18064 Focus to learn more arXiv-issued DOI via DataCite

https://doi.org/10.1103/pd5d-sddp Focus to learn more DOI(s) linking to related resources

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.