Relative Cross Sections for L- and M-Shell Ionization by Electron Impact

Abstract

Results from measurements and calculations of relative L-and M-shell ionization cross sections by electron impact are presented. Measurements were performed for elements Te, Au and Bi on an electron microprobe with specimens consisting of extremely thin ®lms of the studied element deposited on thin, self-supporting, carbon layers. The relative variation of the ionization cross section was obtained by counting the number of characteristic X-rays from the considered element and shell, for varying incident electron energies, from the ionization energy up to 40 keV. Measured data were corrected to account for the energy-dependent spread of the electron beam within the active ®lm and for the ionization due to the electrons backscattered from the carbon layer, using Monte Carlo simulation. Cross sections were evaluated in the Born approximation using an opticaldata model with numerically evaluated dipole photoelectric cross sections. Calculated ionization cross section were converted to vacancy production cross sections, which can be directly compared with our experimental data.Reliable cross sections for inner-shell ionization by impact of keV electrons are required for quantitative Auger electron spectroscopy (AES), electron probe microanalysis (EPMA) and electron energy loss spectroscopy (EELS). In spite of this need, simple systematic methods to calculate ionization cross sections from ®rst principles remain to be found. The usual practice consists in using semi-empirical formulae, which have limited ranges of validity and normally lead to signi®cantly different results. In the energy range considered here, 1 ± 50 keV, experimental measurements have been mostly carried out for Kshell ionization; cross section data for L-and M-shells are, unfortunately, very scarce. As far as we know, absolute L-shell ionization cross sections have been reported only for the elements P, S and Cl [1], Ar [2 ± 5], Kr [6], Xe [6,7] and Au [8 ± 11] (see also refs. [12,13]). Comprehensive reviews of experimental measurements (as well as theoretical calculations and semi-empirical formulae) of inner-shell ionization cross sections have been published by Powell [14,15,16]. Most of the reported cross sections were based on the measurement of characteristic X-rays which are emitted in the de-exitation of ionized atoms in the gas phase. Because of the lack of knowledge of the corresponding¯uorescence yields and CosterKronig decay yields, experimental results are affected by large uncertainties. Measurements using thick solid samples have also been performed, but the results are obscured by multiple scattering within the target. It is thus dif®cult to establish the reliability of semiempirical formulae and theoretical calculations by comparison with the available experimental data. This dif®culty is found not only regarding the absolute value of the cross section, but also for its energy dependence.The determination of the (relative) energy dependence of the ionization cross section is of interest for

Experimental and theoretical L -subshell ionization cross sections for 83Bi by electron impact from the L3 threshold to 100 keV

“…Furthermore, our relative L 3 ionization cross sections are in excellent accord with those of Ref. [22], and also with the calculations performed with the two theoretical formalisms studied here.…”

“…The (black) empty circles denote the data from Ref. Ll00 [22]. The continuous and dashed curves are, respectively, the normalized DWBA [19,50] and SCADW [17,18,20] L 3 ionization cross sections.…”

“…Another consequence of the higher atomic number of 83 Bi is that the L x-ray peaks in the spectra can be better resolved by the employed spectrometer, thereby the uncertainties associated with the peak area estimates should be smaller. Finally, the existing data for Bi L are limited to either 17-40 keV [22][23][24] or 60 keV, 100 keV, and above 200 keV [25][26][27][28], and only one of these publications [28] presents absolute Bi L 1 , L 2 , and L 3 subshell ionization cross sections.…”

Improved X-ray Spectrum Simulation for Electron Microprobe Analysis

“…m is a constant for each shell in the range 0.7-0.9, adjusted to fit the shape of the relationship to experiment; a similar approach has been used by Pouchou (1994), who found m to be a slowly varying function of atomic number. Direct measurements by Llovet (2000a) for the L and M shells and by Llovet (2000b) for the K shell also accord generally with this form. Finally, the multipliers q K , q L , and q M are set by experiment to match the observed intensities from each shell to one another and to those results for which the intensities are known in absolute terms.…”

Experimental x-ray-production cross sections for the M3, M4 , and M5 subshells of Pt and Au by electron impact

“…This is precisely the energy range of interest for the materials characterization techniques based on x-rayemission or Auger-electron spectroscopy. As pointed out by Merlet et al [9], the scarce measurements of M-shell x-ray production cross sections found in the literature were mostly performed at very-high-incident electron energies [10,11] or were mainly focused on the energy dependence of the cross section [12][13][14]. Nevertheless, absolute experimental determinations were recently carried out in the low-overvoltage range for Au, Bi [9], Pb [15], Th [16], and U [17].…”