Development of a New Algorithm to Quantify Local Strain in Scanning Tunneling Microscope Images

In scanning tunneling microscopy (STM) measurements, systematic effects such as piezoelectric drift results in locally distortion of the STM images. On top of those systematic effects, there may also be local variations of lattice constants related to strain. Furthermore the local strain may significantly affect superconductivity at the nanoscale (e.g. [1]). A common method used to correct for distortions relies on the Lawler-Fujita algorithm, where the displacement is measured by the local phase of the Bragg peak [2]. A subsequent spatial derivative of the distortion field allows access to the local strain of the sample. However, this derivative operation may amplify noise. In this talk, I will discuss the development of an algorithm to extract the strain values directly, based on methods used in computer vision. I will also present the application of this algorithm in STM drift correction and physical strain detection. We expect this method to have applicability to STM image correction protocols and in studies that aim to investigate the relation between physical strain and electronic properties.

[1] He Zhao et.al., Nature Physics, 17, 903 2021

[2] M. J. Lawler et.al., Nature, 466, 347 2010