Closing the loop on valley splitting in <sup>28</sup>Si/SiGe: atom probe tomography, tightbinding, and cryomultiplexing
ORAL
Abstract
28Si/SiGe heterostructures provide a compelling host material for a scalable quantum computer, due to the long coherence times of spin qubits and compatibility with industry. Here we study 28Si/SiGe heterostructures with varying roughness of the critical Si/SiGe interfaces to understand the energy splitting of the lowest lying conduction valleys (valley splitting). To improve control of valley splitting in 28Si/SiGe, we implement a feedback cycle for materials stack engineering including several elements.
We use atom probe tomography to provide atomic 3D reconstruction of the material stack and statistical understanding of compositional variations at the Si/SiGe interface over nanoscale dimensions relevant for spin qubits. The resulting data is fed into a tight binding model to compute the valley splitting in real quantum wells with varying thickness, Si/SiGe interface width, and interface chemical roughness. We complete the cycle by comparing the simulation results with valley splitting measured in heterostructure field effect transistors and quantum dots, making use of cryomultiplexer technology to achieve statistically significant metrics.
We envision that such a feedback loop may help to engineer optimal stacks for large and controllable values of valley splitting in Si/SiGe.
We use atom probe tomography to provide atomic 3D reconstruction of the material stack and statistical understanding of compositional variations at the Si/SiGe interface over nanoscale dimensions relevant for spin qubits. The resulting data is fed into a tight binding model to compute the valley splitting in real quantum wells with varying thickness, Si/SiGe interface width, and interface chemical roughness. We complete the cycle by comparing the simulation results with valley splitting measured in heterostructure field effect transistors and quantum dots, making use of cryomultiplexer technology to achieve statistically significant metrics.
We envision that such a feedback loop may help to engineer optimal stacks for large and controllable values of valley splitting in Si/SiGe.
–
Presenters
-
Brian Paquelet Wuetz
- Delft University of Technology