Electron impact multiple ionization of neon, argon and xenon atoms close to threshold: appearance energies and Wannier exponents

Abstract

We report the results of the experimental determination of the appearance energy values AE(Xn + /X) for the formation of multiply charged Ne, Ar and Xe ions up to n = 4 (Ne), n = 6 (Ar) and n = 8 (Xe) following electron impact on Ne, Ar and Xe atoms using a dedicated high-resolution electron impact ionization mass spectrometer. The data analysis uses the Marquart-Levenberg algorithm, which is an iterative, nonlinear least-squares-fitting routine, in conjunction with either a two-function or a three-function fit based on a power threshold law. This allows us to extract the relevant AEs and corresponding exponents for a Wannier-type power law from the measured near-threshold data. The values of the AEs determined in this work are compared with other available experimental and spectroscopic values of the AEs and the extracted exponents are compared with other available experimental data and with the predictions of the various Wannier-type power law models.

Electron impact ionization close to the threshold: classical calculations

“…Recent experiments (Gstir et al 2002) tend to give a slightly larger value, closer to 1.2. In figure 1 we plot our results for single ionization of Ar 0 and Ar 5+ .…”

“…Further to preliminary investigations corroborating the Wannier α value for simple ionization (Katsonis et al 2002b), we address here the question of the value of α for multiple ionization, which is still a matter of discussion (see Gstir et al 2002). …”

“…Basically, it has been confirmed, but research still focuses on the precise evaluation of exponent α. An updated bibliography can be found in Gstir et al (2002).…”

“…Both options have been investigated here. Ionization potentials were taken from (Moore 1949); they can also be found in table 2 of Gstir et al (2002).…”

“…This is not a problem for analytical calculations, where the limit to zero can always be taken, but raises a serious question for numerical or experimental work. The papers of Rost (1994), Koslowski et al (1987), and Gstir et al (2002) have shown that a safe choice for the width of the threshold region is 1 < E/V n < 1.1~1.2.…”