Imaging mechanism for hyperspectral scanning probe microscopy via Gaussian process modelling

Maxim Ziatdinov,Stephen Jesse,Dohyung Kim,Sergei V Kalinin,Mahshid Ahmadi,Rama K Vasudevan,Sabine Neumayer,Liam Collins

doi:10.1038/s41524-020-0289-6

Abstract

We investigate the ability to reconstruct and derive spatial structure from sparsely sampled 3D piezoresponse force microcopy data, captured using the band-excitation (BE) technique, via Gaussian Process (GP) methods. Even for weakly informative priors, GP methods allow unambiguous determination of the characteristic length scales of the imaging process both in spatial and frequency domains. We further show that BE data set tends to be oversampled in the spatial domains, with ~30% of original data set sufficient for high-quality reconstruction, potentially enabling faster BE imaging. At the same time, reliable reconstruction along the frequency domain requires the resonance peak to be within the measured band. This behavior suggests the optimal strategy for the BE imaging on unknown samples. Finally, we discuss how GP can be used for automated experimentation in SPM, by combining GP regression with non-rectangular scans.

Highlights

Over the past three decades, scanning probe microscopy (SPM) has emerged as a primary tool for characterization of structure and functionality at the nanometer and atomic scales
The spectroscopic modes define the sampling of the parameter space of interest at each spatial location, for example, classical force–distance and current–voltage curve mappings in atomic force microscopy[9] and scanning tunneling microscopy, and complex time and voltage spectroscopies in piezoresponse force microscopy (PFM) and electrochemical strain microscopy[10,11,12]
BE PFM on BiFeO3 film The BE PFM measurements were performed on an Asylum Cypher microscope using an in-house built BE controller

Summary

Introduction

Over the past three decades, scanning probe microscopy (SPM) has emerged as a primary tool for characterization of structure and functionality at the nanometer and atomic scales. This method allows exploration of the data structure from purely information theory perspective simultaneously in the spatial and parameter space.

Results

Conclusion