Characterization of small-scale surface topography using transmission electron microscopy

Multi-scale surface topography is critical to surface function, yet the very smallest scales of topography are not accessible with conventional measurement techniques. Here we demonstrate two separate approaches for measuring small-scale topography in a transmission electron microscope (TEM). The first technique harnesses 'conventional' methods for preparation of a TEM cross-section, and presents how these methods may be modified to ensure the preservation of the original surface. The second technique involves the deposition of the material of interest on a pre-fabricated substrate. Both techniques enable the observation and quantification of surface topography with Ångström-scale resolution. Then, using electron energy loss spectroscopy (EELS) to quantify the sample thickness, we demonstrate that there is no systematic effect of thickness on the statistical measurements of roughness. This result was verified using mathematical simulations of artificial surfaces with varying thickness. The proposed explanation is that increasing the side-view thickness of a randomly rough surface may change which specific features are sampled, but does not significantly alter the character (e.g. root-mean-square (RMS) values and power spectral density (PSD)) of the measured topography. Taken together, this work establishes a new approach to topography characterization, which fills in a critical gap in conventional approaches: i.e. the measurement of smallest-scale topography.