H. Gnaser 80 magnetic sector features a homogeneous magnetic field with inclined field boundaries. The mass resolution was M/ΔM ~ 300. The detection <strong>of</strong> secondary <strong>ions</strong> was done by a discrete-dynode electron multiplier, and count rates were limited to
<strong>Energy</strong> <strong>spectra</strong> <strong>of</strong> <strong>sputtered</strong> <strong>ions</strong> A ¼ R e jC e ja 2 (2) Calculated energy resolution ΔE (eV) FWHM Cs + energy spectrum (eV) 10 2 CA1 CA2 CA3 CA4 10 1 10 10 2 10 3 Exit slit width (µm) CA1 CA2 CA3 CA4 10 1 10 10 2 10 3 10 2 (a) Exit slit width (µm) (b) The radius <strong>of</strong> <strong>the</strong> electrostatic sector is R e = 85 mm, <strong>the</strong> constant C e = 3.5, and a is <strong>the</strong> angle to which <strong>the</strong> beam is constrained in <strong>the</strong> radial direction (a ~ 0.012). [21] With <strong>the</strong>se numbers, a value <strong>of</strong> A ~40mm is obtained. Eqn (1) may <strong>the</strong>n be used to estimate <strong>the</strong> <strong>the</strong>oretically predicted values <strong>of</strong> ΔE for <strong>the</strong> different S 1 and S 2 ; Fig. 3(b) shows such an evaluation which produces a qualitative agreement with <strong>the</strong> experimental data. The discrepancies that are found might be due to <strong>the</strong> presence <strong>of</strong> <strong>the</strong> o<strong>the</strong>r optical elements (<strong>the</strong> magnetic sector and severa lenses), which follow <strong>the</strong> energy analyzer and possible uncertainties associated with <strong>the</strong> specific values <strong>of</strong> C e and a. It was noted in <strong>the</strong> context <strong>of</strong> Fig. 1 that for wide exit slits, <strong>the</strong> energy <strong>spectra</strong> exhibit a box-type shape, at least in <strong>the</strong> lowenergy and <strong>the</strong> central parts. This feature is illustrated in Fig. 4, which depicts <strong>the</strong> distribut<strong>ions</strong> obtained for all four CAs and S 2 = 1400 mm. In a recent publication, [18] it was proposed that a box-shaped resolution function R(E) might be an appropriate description for <strong>the</strong> bandwidth in electric and magnetic sector-field instruments such as <strong>the</strong> one used in <strong>the</strong> present work. Specifically, <strong>the</strong> following normalized resolution function was applied RE ð Þ ¼ 1 4ΔE 1 þ erf E þ 0:5ΔE E pffiffiffi 1 erf 2 s 0:5ΔE pffiffiffi 2 s Eqn (3) constitutes <strong>the</strong> convolution <strong>of</strong> a rectangular box with a Gaussian <strong>of</strong> standard deviation s. [18] Such a box-type function R(E) is plotted in Fig. 4 (dashed line), using ΔE = 37.6 eV and s = 1.3 eV. Clearly, <strong>the</strong> function describes <strong>the</strong> experimental <strong>spectra</strong> quite well in <strong>the</strong> low-energy and <strong>the</strong> central parts. As suggested by Wittmaack, [18] <strong>the</strong> resolution function given in Eqn (3) could be used to compute energy <strong>spectra</strong> via a convolution using R(E). Their comparison with <strong>the</strong> corresponding experimental distribut<strong>ions</strong> would <strong>the</strong>n constitute a test as to <strong>the</strong> validity <strong>of</strong> <strong>the</strong> resolution function. This approach was tested for several measured energy <strong>spectra</strong>. Figure 5 displays normalized Cs + distribut<strong>ions</strong> (3) Figure 3. (a) The measured FWHM <strong>of</strong> <strong>the</strong> Cs + <strong>spectra</strong> as a function <strong>of</strong> <strong>the</strong> exit slit width for <strong>the</strong> four CAs. (b) The energy resolution ΔE calculated according to Eqn (1), with an aberration <strong>of</strong> A =40mm. <strong>of</strong> <strong>the</strong> CAs. In addition, optical aberrat<strong>ions</strong> may possibly contribute to <strong>the</strong> values <strong>of</strong> <strong>the</strong> FWHM for small slit sizes. Theoretically, <strong>the</strong> relative energy resolution E/ΔE <strong>of</strong> an electrostatic sector analyzer such as <strong>the</strong> one in <strong>the</strong> present SIMS instrument depends on <strong>the</strong> energy dispersion D E , <strong>the</strong> widths <strong>of</strong> <strong>the</strong> entrance and exits slits, S 1 and S 2 ,<strong>the</strong>analyzer’s magnification M, and some possible contribut<strong>ions</strong> due to optical aberration effects A [6] : E ΔE ¼ D E MS 1 þ S 2 þ A (1) Cs + intensity (normalized) 10 -1 1 10 -2 10 -3 S 2 = 1400 μm CA1 CA2 CA3 CA4 Box-type In <strong>the</strong> present case, S 1 is determined by <strong>the</strong> diameter <strong>of</strong> <strong>the</strong> CAs and M is unity because <strong>of</strong> <strong>the</strong> symmetry <strong>of</strong> <strong>the</strong> entrance and exit posit<strong>ions</strong> <strong>of</strong> <strong>the</strong> sector field. For an electrostatic sector, <strong>the</strong> contribution due to aberration in <strong>the</strong> radial plane is given by [19,20] -20 0 20 40 60 Emission energy (eV) Figure 4. Normalized Cs + energy <strong>spectra</strong> for <strong>the</strong> four different CAs and <strong>the</strong> exit slit S 2 = 1400 mm (solid symbols). The dashed line is <strong>the</strong> resolution function given in Eqn (3). 81 Surf. Interface Anal. 2013, 45, 79–82 Copyright © 2012 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/sia