Special Topic: Energy resolution improvement

1
Special Topic Energy resolution improvement
2

• Energy resolution
• in EELS depends on energy width of electron
  source
• Factors affect the energy resolution
• High-voltage stability
• Spectrometer design (aberration correction)
• Spectrometer power-supply stability
• Mechanical stability of the spectrometer
• Stray magnetic fields around spectrometer

3
1. Data sharpening algorithm
Factor by which the spectral resolution can be
increased by Fourier or by maximum-entropy
deconvolution
4
Lucy-Richardson deconvolution
Lucy-Richardson method (maximum-likelihood) with
15 iterations sharpens the zero-loss peak (from
FWHM 1.3 to 0.8 eV) and increases visibility of
fine structure. But it produces wings on the
zero-loss peak.
5
Lucy-Richardson deconvolution
background subtracted before sharpening
Wing problem is greatly reduced by removing the
zero-loss peak, for example by using
Fourier-log deconvolution before sharpening.
Comparison of the sharpened SSD (black) with the
sharpened PSD (red)
6
1500 iterations
7
2. Monochromated EELS
Gun monochromators Batson (IBM) 80meV
resolution at E0 100 keV fei Wien filter
(rainbow illumination) JEOL double Wien filter
(dispersion corrected) Zeiss omega filter
(dispersion corrected) In-column
monochromators JEOL-HREA80 (Terauchi et al.,
Tohoku U., 80keV) 200 meV at 300eV, 25 meV below
E 5 eV Berlin (Boersch et al. 1964) lt 6 meV at
E0 30 keV
8
Monochromation and aberration correction(Batson,
Ultramicroscopy 78, 33 - 42, 1999)
9
FEI monochromator (Tiejmeier et al.)
10
Additional fine structure seen with monochromation
Rez,2004
11
Maximum-likelihood deconvolution of Co L-edges
(Hofer et al., 2002)
Micron 34 (2003) 135