Reliability of the quantification of eels measurements

Over the past few years at ORNL a significant effort has been expended on developing the transition metal boride TiB2 for applications that take advantage of its high hardness, high melting point, and chemical inertness. The major portion of this work has centered on the TiB2-Ni system. AEM has been used extensively to provide microstructural information for structure-property correlations. Particular emphasis has been placed on identifying phases that develop during high-temperature exposure. Because of the presence of boron and other low-atomic-number elements in these materials, electron energy loss spectroscopy (EELS) has been an important tool in these investigations. However, repeated analyses of secondary phases, as well as of TiB2 itself, lead to the conclusion that compositions determined by quantification of EELS spectra with the use of both K and L ionization edges were suspect. In an earlier paper the authors presented some of these results and suggested that the problem might lie in the computation of cross sections for L absorption edges. Since that time our interest in this problem has continued and the authors have concentrated on identifying the source of the inaccuracy with greater certainty. 12 references, 3 figures