OPTIMIZATION OF THE SIGNAL-TO-NOISE RATIO IN EFTEM ELEMENTAL MAPS WITH REGARD TO DIFFERENT EONIZATION EDGE TYPES

An energy filtering transmission microscope can provide information ab out the spatial distribution of an element by selecting an energy loss corresponding to a characteristic inner-shell excitation. One paramet er affecting the obtainable resolution, as well as the detection limit of such core-loss images (elemental maps), is the signal to noise rat io. Statistical noise due to low inelastic cross-sections makes weak-c ontrast features invisible, leading to a loss of resolution and a dete rioration of the detection limits. The information content therefore i s largely determined by the noise level in the recorded maps. As a con sequence, improving the detection limits means first optimizing the si gnal to noise ratio (SNR). Since electron energy-loss spectroscopy (EE LS) is the basis for energy filtering, EEL spectra can be used as a ba sis to calculate optimal values of the SNR for different ionization ed ges and the results shall be directly applicable to core-loss imaging. In this paper we describe the procedures which have been employed for these calculations and discuss the results, giving useful information about some setup parameters for elemental mapping with different edge s. The calculated imaging parameters can then be interpreted in terms of the attainable spatial resolution in the elemental maps. (C) 1998 E lsevier Science Ltd. All rights reserved.