Marvin Hartwig Zoellner

CMOS compatible Si/SiGe and Ge/SiGe quantum-well (QW) structures with gate-defined quantum dots are promising material platforms to realize large scale fabrication of spin qubit devices. As the development of functional qubits rises and exceeds single qubit operation, large-scale integration becomes a challenging factor to be faced for semiconductor spin qubit devices. To enable multi qubit algorithms to run on large arrays of qubits with shared gate control, band edge energy levels should be well aligned on a scale of several microns. Furthermore, communication in larger architectures requires coherent shuttling of spin information sensitive to steep short range fluctuations of the potential energy.

Band structure potential fluctuations in the QW, causing uncontrolled transition and decoherence, may be related to the plastic relaxation process of the SiGe buffer below as well as the influence of elastic relaxation of the gate structure on top. To determine these fluctuations quantitatively, we leverage Scanning X-ray Diffraction Microscopy (SXDM) with a nanoprobe beam at ID01/ESRF to reveal the lattice homogeneity of the buried QW in a non-destructive way, with high strain sensitivity (Δa/a ~ 1E-5) and with a resolution on the scale of the electrode pitch (200 nm).

In particular, we present laterally resolved lattice strain and rotation maps of Si QW in electron shuttling devices with a 50 nm resolution. We found that the maximum strain fluctuation in Si/Si0.7Ge0.3 heterostacks is around several E-4 attributed i) to the semi-periodic appearance of the cross-hatch pattern across several microns caused by dislocation formation and simultaneously ii) stress by the electrodes inducing short-range strain fluctuations in the QW.

Strain profiles across electrodes are compared to thermomechanical Finite Element Method (FEM) simulations, which are translated to profiles of the band energy levels. We observe that the energy level variation reaches the meV range in the 45 nm deep buried Si QW, hence having the same order of magnitude as the charging energy.

Thus, our work provides important insights into the correlation of material and qubit properties, demonstrating that improved material control in terms of short and long range strain fluctuations is vital for the optimization of SiGe-based qubits.