Valley splitting engineering in nuclear spin-free SiGe for silicon qubits

• Applicants: Dr. Kevin-Peter Gradwohl; Dr. Lars Reiner Schreiber
• Subject Area: Experimental Condensed Matter Physics
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 554676597

Solid-state-based quantum computing using single-electron spin qubits in silicon (Si) is a highly promising technology due the small footprint of qubits, potentially enabling the integration of millions of qubits on a single wafer, taking advantage of the elaborated fabrication techniques from the silicon industry. However, respective qubit integration is limited by yet unmet material challenges, in particular, a high degree of crystalline quality of SiGe quantum well layer stacks, the use of isotopically purified materials, and a reliable enhancement of the local conduction-band valley splitting on the length-scale of a quantum chip. In this proposal, these problems are tackled by growth of isotopically purified, nuclear spin-free 28Si76Ge material stacks by molecular beam epitaxy, as well as subsequent state-of-the-art spin qubit fabrication and characterization techniques. We investigate the introduction of innovative nuclear spin-free germanium (Ge) concentration profiles in the 28Si quantum well heterostructures, providing a new degree of freedom and innovative possibilities to improve valley splitting in a deterministic and reliable way. The optimized material properties are continuously benchmarked on a lateral scale by coherent shuttling of spin qubits in conveyor-mode, which is a new functional element enabling scalable quantum computing architectures. Here, we explore three fundamentally different types of such heterostructure designs to boost the local valley splitting. These types are increased barrier heights via increased Ge-content at the quantum well interfaces, a constant but arbitrary distribution of Ge atoms within the quantum well, and an oscillation of the Ge concentration in the quantum well layer with wavelengths of 1.80 and 0.31 nm. Central is the control and suppression of Ge segregation and intermixing with Si, which will be fundamentally explored by state-of-the-art materials characterization and controlled to realize optimal heterostructures for qubit application. In order to validate the applicability of the heterostructure designs for scalable spin quantum computing, we benchmark spin dephasing time of the single electron spin qubits, the electron g-factor, and map the local potential disorder and the valley splitting energy. For the mapping of the local material parameters, we employ an electron-spin qubit as local probe by shuttling it in a conveyor-mode to the site of interest. We aim at boosting the global minimum of the valley splitting above 25 µeV while keeping the spin coherence time long. This would pave the way to scalable quantum computing architectures based on shuttling techniques in SiGe.

• DFG Programme: Research Grants