Electron probe microanalysis of light elements: Improvements in the measurement and signal extraction methods. (11th May 2022)

• Record Type: Journal Article
• Authors: Schweizer, Pia; Brackx, Emmanuelle; Jonnard, Philippe
• Is Part Of: X-ray spectrometry. Volume 51:Number 4(2022)
• Page Start: 403
• Page End: 412
• Publication Date: 2022-05-11
• Subjects: area/peak factors -- background subtraction -- electron probe microanalysis -- light elements -- micoanalyser parameters; X-ray spectroscopy -- Periodicals; 543.62
• DOI: 10.1002/xrs.3290
• Languages: English
• ISSNs: 0049-8246

Abstract

Abstract: Quantitative electron probe microanalysis (EPMA) by wavelength dispersive spectrometry of light elements is a significant challenge due to difficulties of signal extraction for low intensities. Therefore, a good choice of measurement parameters and a suitable signal treatment, accurate background description and interference correction are crucial. For light elements, peak to background ratios (P/B) can easily approach unity and special care needs to be paid to the setting of the pulse height analyser (PHA) to avoid noise contribution to the emission peak. Furthermore, the background may have a high curvature and the classical approach of a linear background fit is no longer valid. This work investigates the EPMA of light elements from Be to F with the aim to determine the best experimentation parameters (dispersive element, accelerating voltage, PHA). It also provides detailed information about peak shifts, peak broadening and the influence of different background fits on quantification. In total, 25 samples containing light elements were analysed and Monte Carlo simulations were carried out. These simulations show that the X‐ray intensity of the light element's characteristic line can be increased by working with low accelerating voltages around 5 kV. PHA should be used in automatic differential or integral mode. Our work also shows that a third degree linear polynomial describes the background better than an exponential or linear function. Consequently overestimation of P/B values leading to large quantification errors can be avoided. This paper provides a large database containing peak position and peak shifts, peak full width at half maximum, as well as P/B ratios and information about peak enlargement for different materials of various compositions.