Nanoscale etching characteristics for controlled AFM tomography of amorphous chalcogenide thin films

Amorphous chalcogenide materials are of increasing interest for phase-change memory and threshold-switching device applications. However, the fast-switching behavior in the nanoscale regime is difficult to confirm and characterize in terms of changes in conductivity. Understanding these electrical characteristics requires three-dimensional tomographic imaging capable of resolving structural and compositional changes across the film. Atomic force microscopy (AFM) is a powerful tool for obtaining such 3D tomographic imaging by alternating etching and imaging steps, and it also enables both electrical measurement and biasing during etching, ​which is essential given the volatile nature of Ovonic Threshold Switching (OTS) behavior. However, achieving accurate AFM-based etching for tomography requires a quantitative understanding of the tip-induced wear resistance of the chalcogenide film, since amorphous materials, lacking long-range structural order, make their wear resistance difficult to predict from local bonding configurations and surface-mediated contact mechanics alone. Here, we study a first-order framework for estimating the relative wear resistance of amorphous chalcogenide thin films from Raman spectroscopy and for selecting initial load​ in AFM etching and test it on depth-controlled AFM tomography. Etched depth as a function of applied load was measured for three amorphous binary films (Ge-Te, Si-Te, and Ge-Se) using AFM with a diamond-coated probe, and the probe state was monitored by force–distance pull-off measurements before and after etching. The load–depth data were analyzed within an Archard–DMT (Derjaguin-Muller-Toporov) framework to obtain a composite wear parameter ( \:Wᵣ ) that combines the wear coefficient and the effective hardness, allowing relative comparison of the films under nanoscale AFM contact conditions. This parameter follows the order Ge-Te > Si-Te > Ge-Se in both the low- and high-load ranges. To explain this order without independent nanoindentation, we introduce an effective bond enthalpy calculated from the composition and the Raman-resolved bonding configurations, which decreases monotonically with the wear parameter. Applied to an amorphous ternary Si-Ge-Te film, this relationship was used to select an etching load that yielded layer-by-layer tomographic imaging with an etched depth close to the target. These results suggest the Raman-derived effective bond enthalpy as a practical descriptor for selecting initial load​ in AFM-based nanoscale etching of amorphous chalcogenide films.