Fig. 2: Understanding the influence of detector efficiency and spatial resolution.
From: Quantifying short-range order using atom probe tomography
a,b, Influence of detection efficiency on the SRO parameter for random (a) and non-random (b) simulations. Averages and 95% confidence intervals are plotted from 100 simulations at each detection efficiency value. Annotations include the most common commercially available APT instruments. One example of the 95% confidence intervals is presented in the inset. The detailed maximum confidence intervals for b are presented in Supplementary Table 1. c,d, Influence of trajectory uncertainties (spatial resolution) on the SRO parameters at kNN = 12 for random (c) and non-random (d) simulations (arbitrary input, α = 0.041). e,f, Combined influence of 57% detection efficiency and trajectory uncertainties on the random (e) and non-random (f) cases (arbitrary input, α = 0.039) at kNN = 7. The SRO values are small across c and e, with the background colour near clear (white). The depth and lateral resolutions represent the standard deviation of Gaussian noise applied across the x–y (lateral) plane and z (depth) direction. The absolute values of the SRO are presented. The white dashed line indicates a contour where |α| ≤ 0.00022, the SRO threshold below which it is impossible to distinguish between random and non-random situations. This contour occurs at a lateral noise value of 0.95 nm in d and 0.85 nm in f. These tests have been repeated 100 times with similar trends. The standard deviation of these tests is plotted in Supplementary Fig. 1. g, Go/no-go threshold, with the standard deviation indicated by dashed lines, is presented on a logarithmic scale using SRO values ranging from ~0.100 to ~0.001 and x, y spatial noise. Below certain x, y spatial-noise levels, the SRO values are difficult to distinguish from random (|α| ≤ 0.00022).
