Chapter/Index: Introduction | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | Appendix
| Various sources of instability in the microscope contribute to this effect: High voltage fluctuation, magnetic field creep (or stray magnetic field) etc. This places a fundamental limit on the useful exposure time. Experimental efforts in electron spectroscopy consist largely in reducing unexpected electrical noise and specimen drift. If all these factors are stable, the acquisition time can be very long, e.g. >20 seconds [1]. In practice, EELS measurements with sufficient SNR at a common beam current of 10-100 pA and EELS collection optics of ~10-30% efficiency (without aberration correction) requires an acquisition time of ~0.1-1 s per pixel. Spectrum drift and energy drift in EELS measurements may be induced by various instabilities. It was proposed that the energy drift from low-frequency instabilities can be corrected by software techniques [2, 3], while the energy drift from high-frequency energy instabilities can be corrected by high speed acquisition technique [4]. In TEM and EELS measurements, energy instabilities normally originate from the fields in the frequency range of 50 or 60 Hz. To detect the 50-Hz instability, the time resolution of the detector needs to be below 10 ms, however, conventional CCDs cannot be used to acquire an EEL spectrum with such a short exposure time.
[1] Optimization of the Signal to Noise Ratio in EFTEM Elemental
Maps with Regard to Different Eonization Edge Types
G. Kothleitner and F. Hofer, Micron Vol. 29, No. 5, pp. 349–357, (1998).
|