In general, EELS capability can be improved by:
i.a) Use small camera length (0.20 m) to get better signal-to-noise ratio, improving EELS sensitivity.
i.b) Increasing the probe size, which deteriorates the spatial resolution, increases the beam current to improve the signal-to-noise ratio (SNR) .
i.c) Detection limit depends strongly on the background (B). As the concentration of an element decreases or TEM sample thickness increases, its signal-to-background ratio, S/B, decreases. In this case, there can only be confidence about the signal contained within a few eV beyond the preceding ionization edge.
ii) Spatial resolution improvement:
ii.a) Spectroscopy can be done with either TEM imaging mode or TEM diffraction mode. For instance, a recommended TEM diffraction mode is STEM mode. Imaging mode gives better energy resolution while diffraction mode
gives better spatial resolution.
ii.b) Use small probe size in STEM mode.
ii.c) The spatial resolution can be improved by using a compromise for these parameters combined with averaging of repeated line scans from the same sample, normally resulting in ≤ 0.5 nm.
iii) Energy resolution improvement:
iii.a) Perform EELS measurements in TEM imaging mode.
iv) Information improvement:
iv.a) The recorded electrons in the zero loss peak have passed the sample without energy loss or with energy changes too small to be measurable from phonon excitations (a few 10 meV).
 Heiko Stegmann and Ehrenfried Zschech, Compositional analysis of ultrathin silicon oxynitride gate dielectrics by quantitative electron energy loss spectroscopy, Applied Physics Letters, 83 (24), (2003) 5017.