1.3 Image Normalisation

The normalization of ADF data requires two values to be known, the D.C. offset and the effective detector sensitivity. The effective detector sensitivity is the combined effect of the intrinsic sensitivity of the detector and the amplifier gain.

In general, both these parameters are measured from so called 'detector scans'. There are various methods of recording these scans and care is required to make these as accurate as possible. Experimental ADF detectors often present a significant degree of variability and this needs to be reflected in the analysis. A discussion of detector variability can be found in:

• Katherine E. MacArthur, Lewys Jones and Peter D. Nellist “How Flat is Your Detector? Non-Uniform Annular Detector Sensitivity in STEM Quantification”, Journal of Physics: Conference Series 522 012018 (2014) doi:10.1088/1742-6596/522/1/012018.

After the detector scan has been recorded there are two methods of utilising these which account for detector asymmetry in different ways. These two choices are:

• Perform simple averaging of the detector sensitivity across its whole area and use a detector profile in comparator image simulations, or
• Perform electron flux weighted average of the detector and use a 'perfect' detector in the comparator image simulations.