L X-ray production cross-section ratios for protons incident on high-Z atoms: A test of ECPSSR theory and newly recommended vacancy de-excitation parameters

https://doi.org/10.1016/j.nimb.2007.01.079Get rights and content

Abstract

The X-ray intensity ratios Lβ/Lα and Lγ/Lα for six heavy elements have been measured using a Si(Li) detector coupled to both analog and digital signal processors. The proton energies were chosen to ensure that possible united atom (UA) and intra-shell (IS) effects would be small, thereby ensuring that the experiments were a direct test of the ECPSSR together with the L subshell fluorescence and Coster–Kronig probabilities. Careful attention was paid to reducing recognized uncertainties in the fitting of the spectra, with the desired intensity ratios being extracted from a procedure that assumed the DF radiative transition rates of Scofield to be correct. Experiment and ECPSSR theory agree closely when the fluorescence and CK probabilities from a recent critical review are adopted in preference to those commonly used in earlier work. This result provides a significant improvement in the database used for PIXE analysis, and it paves the way for more definitive studies at lower proton energies, aimed at testing predicted UA and IS corrections.

Introduction

A very large amount of experimental effort has been devoted to testing the predictions of the ECPSSR model for proton ionization cross-sections in the inner atomic shells. In the case of the L shell, a remarkable degree of scatter is observed when all the experimental cross-section data for a given element are collected together. This observation suggests the existence of hitherto unrecognized systematic errors in experimental measurements. Improved measurements and data analysis are critically needed for two reasons: (i) because the scatter in existing data reduces the sensitivity of tests of various corrections that have been proposed to refine the basic theory; and (ii) because better X-ray production cross-sections are required for PIXE analysis applications.

We report here new experiments on L subshell X-ray production by proton beams, wherein very careful attention has been paid to systematic error and its elimination. We examine the influence of different assumptions regarding the L subshell fluorescence and Coster–Kronig yields, values of which are required in the comparison of experimental X-ray production cross-sections and theoretical predictions. Our objectives are to determine accurate values for the Lβ/Lα and Lγ/Lα X-ray intensity ratios, to compare these with the full set of earlier results from the literature, and to test different sets of ancillary atomic data that must be assumed when converting theoretical ionization cross-sections into predicted X-ray production cross-sections.

Section snippets

Review of theory

The widely-cited ECPSSR model for the ionization of bound atomic electrons by incident light ions having MeV energies is basically a plane wave Born approximation (PWBA) calculation, with a number of refinements which evolved over several years during the pioneering work of Brandt and Lapicki [1]. The PSS correction accounts for the perturbation of the bound stationary state due to the influence of the projectile, and is expressed as an effective change in the electron binding energy. The

Review of experimental Lβ/Lα and Lγ/Lα ratios

The most recent collection of published experimental data is that of Strivay and Weber [8], which builds upon earlier work such as that of Orlic et al. [9], and thus enables us to demonstrate some of the issues of scatter mentioned in Section 1. We use gold as an example, because it is the element for which there exists the largest volume of experimental data.

In Fig. 1, we show the measured intensity ratios Lβ/Lα and Lγ/Lα for gold as a function of bombarding proton energy. Most of the

Basis of the present work

Proton energies of 2.0 and 2.5 MeV were selected, for two reasons. The predicted IS and UA effects decrease with increasing projectile energy, and at the chosen energies their combined effect is only a few percent for heavy elements. There is much merit in testing the ECPSSR and the associated atomic parameters as rigorously as possible at such energies, because only then does one have a secure basis for lower energy work directed at understanding the UA and IS effects. In addition, the energies

Experimental details of the present work

The targets used for the L X-ray studies were 50 μg/cm2 deposits of tungsten, platinum, gold, lead, thorium and uranium on 6.3 μm Mylar backings; they were supplied by the MicroMatter Corporation. These targets were supplemented by similar targets of nickel, copper, germanium, yttrium, rhodium and silver, whose K X-rays were used in determination of the detector’s lineshape.

The proton beams were supplied by an NEC Pelletron accelerator. They were directed into the analysis chamber of a

Fitting the spectra

Mainly because of the strong contributions of Lorentzian lineshapes to the background, we rejected the simple approach of defining energy windows on the Lα, Lβ and Lγ peaks, in favour of a more sophisticated least-squares fitting approach.

Results

The full set of measured and predicted Lβ/Lα and Lγ/Lα intensity ratios determined here is presented in Table 4. For each combination of Z and proton energy values, the experimental ratios have been averaged over several measurements using both the ASP (analogue signal processor) and the DSP (digital signal processor); no statistically significant differences emerged between the results obtained with these two processors. The main uncertainties are those in the L1M3, L2M4 and L3N5 peak areas,

Discussion

It is interesting to note in the particular case of gold – see Fig. 1 – that the Lβ/Lα and Lγ/Lα intensity ratios determined in this work do not follow the “average” trend of previously published experimental values. This tends to support our suggestion of common systematic errors, especially in spectrum fitting, in at least part of the published body of results. Incidentally, the choice of gold by so many experimenters (there are more data for gold than for any other element) is a questionable

Acknowledgements

This work was supported by the Natural Sciences and Engineering Research Council of Canada. Extensive discussions with T. Papp and J.A. Maxwell were of great assistance to the authors. We appreciate the help of W.J. Teesdale in running the accelerator.

References (31)

  • D.D. Cohen et al.

    Atom. Data Nucl. Data Tables

    (1985)
  • Z. Liu et al.

    Comp. Phys. Comm.

    (1996)
  • M.H. Chen et al.

    Atom. Data Nucl. Data Tables

    (1985)
    M.H. Chen et al.

    Atom. Data Nucl. Data Tables

    (1989)
  • M. Vigilante et al.

    Nucl. Instr. and Meth. B

    (1990)
  • D. Strivay et al.

    Nucl. Instr. and Meth. B

    (2002)
  • I. Orlic et al.

    Atom. Data Nucl. Data Tables

    (1994)
  • J.L. Campbell et al.

    Atom. Data Nucl. Data Tables

    (1989)
  • J.H. Scofield

    Phy. Rev. A

    (1974)
  • M.O. Krause

    J. Phys. Chem. Ref. Data

    (1979)
  • D.D. Cohen et al.

    Atom. Data Nucl. Data Tables

    (1986)
  • T. Papp et al.

    Nucl. Instr. and Meth. B

    (2004)
  • T. Papp et al.

    Nucl. Instr. and Meth. B

    (1996)
  • J.M. O’Meara et al.

    X-Ray Spectrom.

    (2004)
  • T. Papp, private...
  • J.L. Campbell et al.

    Atom. Data Nucl. Data Tables

    (2001)
  • Cited by (15)

    • L-shell X-ray production cross sections in metal oxide thin films due to <sup>12</sup>C, <sup>16</sup>O and <sup>28</sup>Si ion beams at MeV SIMS energies

      2019, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
      Citation Excerpt :

      The ECPSSR theory builds on one of the earliest models of particle induced inner shell ionisation, the plane wave Born approximation (PWBA) [10], by including effects of energy loss (E) and Coulomb deflection(C) of the incident ions, as well as accounting for the perturbation of the target atomic stationary states (PSS) by the projectile and the relativistic nature (R) of the inner-shell target electrons. The model is generally of good accuracy for light projectile ion and heavy target atom combinations (i.e. Zion ≪ Ztarget) and high (≥0.3 MeV/u) energies [11,12]. Further work done to improve the predictive accuracy of ECPSSR theory in the description of collisions involving heavier projectiles resulted in what is referred to as the United Atom (UA) correction by Sarkadi and Mukoyama [13].

    • Proton induced L<inf>1</inf>, L<inf>2</inf>, L<inf>3</inf>-sub-shell X-ray production cross sections of Hf and Au

      2015, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
      Citation Excerpt :

      In less than 15% of the publications, Lβ and Lγ cross sections are separated into sub-lines of their respective groups, while only five publications report cross sections for the weak Lη line [3]. In these fittings the intensities of the individual X-ray lines are obtained either adjusting them all [4], or by adjusting the main line of the sub-shell and calculating the other intensities based on the relative emission rates [5]. When the X-ray production cross sections of Lα, Lβ, and Lγ are to be compared with ECPSSR theory, it is necessary to regroup them using several constants, as the Coster–Kronig coefficients, fluorescence yields, and emission rates [6].

    • Cross section for induced L X-ray emission by protons of energy &lt;400 keV

      2014, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
      Citation Excerpt :

      Far-reaching investigations have been made for K-shell ionization and the mechanism is now reasonably well understood. However, the situation is not so clear for L-shell ionization because of lack of experimental data on L-shell X-ray production cross sections and large discrepancy between the experimental data and theoretical predictions [3,4]. The present work is a part of a series of measurements directed towards these studies by low energy proton bombardment [5,6].

    • Werner Brandt legacy to PIXE: Past and present perspectives

      2014, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
    • Emission rates of Gd L sub-shell proton induced X-rays as a function of projectile energy

      2013, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
      Citation Excerpt :

      Experimental results followed closely the theoretical predictions, especially for higher proton energies, although agreement was not as good for the low energies regime. As a further complementary study, in this work, the authors compare the relative intensity ratios of some selected L sub-shell emission lines (Lα, Lβ, Lγ and Ll) to achieve a better understanding of the emission rates behavior; in order to compare the experimental data with similar previous works, a parameter defined as the Reduced Velocity was used, allowing to display different sets of target-projectile in the same plot, thus forming a cloud with the experimental data available which was previously measured by different authors [4–10]. A theoretical prediction for the different emission rates were performed following the calculations presented in Ref. [1] and compared with a 4th degree polynomial fit of the full data cloud.

    View all citing articles on Scopus
    View full text