Assessment of the strain depth sensitivity of Moiré sampling Scanning Transmission Electron Microscopy Geometrical Phase Analysis through a comparison with Dark-Field Electron Holography

In this study, the Moiré sampling Scanning Transmission Electron Microscopy Geometrical Phase Analysis (or STEM Moiré GPA) strain characterization method is compared to the well-established Dark-Field Electron Holography technique on a thin film stack grown by Molecular Beam Epitaxy. While experimental data obtained with the two techniques are, overall, in good qualitative agreement, small statistically relevant differences are locally observed between the two methods. The results obtained from both techniques are further confronted with Finite Element Method (FEM) mechanical simulations modeling the strain relaxation phenomena from a thin lamella. The FEM simulation highlights a non-uniform deformation field along the beam propagation direction with a higher deformation level near the surface of the lamella compared to the center of the same lamella. The center-surface strain differences obtained from modeling are consistent with the experimentally derived differences accounting for the fact that the SMG method is sensitive to the strain state of the surface of the lamella with a very narrow depth-of-field, and the DFEH technique is measuring the strain state of the center of the same lamella averaging over a large section of the thickness. The depth-of-field difference between both methods can be reasonably related to their respective contrast mechanisms (STEM vs Conventional Transmission Electron Microscopy). As the SMG method is using a convergent probe, the narrow depth-of-field might be used to sense the deformation field over different sections of the lamella using the defocus and potentially retrieve the three-dimensional strain field.