Benchmarking Visible Spectral Line Data for Highly Charged Tungsten Using an EBIT and GRASP Calculations

Roger Hutton,Tomas Brage
2016-01-01
Abstract:In this contribution we will discuss a single spectral line in W 27+ an M1 line between the two ground state fine structure levels 4d 10 4f 2 F7/2 and 2 F5/2. This is the only transition within the ground configuration of this ion and therefore a challenge to identify spectroscopically. We have studied this line both experimentally using an EBIT and theoretically using the GRASP2K code [1]. These investigations were done independently and the results only compared when both methods had reached a conclusive wavelength, from careful and systematic considerations. One of the motivation for this work was the fact that earlier predictions for this transition showed a spread of over 400 Å for the wavelength [2, 3]. Our experiments were performed using the Shanghai permanent magnet EBIT and an Andor 303 Shamrock spectrometer equipped with an Andor Newton CCD camera. Tungsten was injected into the EBIT using the vaporous compound W(CO)6. Spectra were recorded at several electron beam energies to isolate the W 27+ line. The spectra were wavelength calibrated using a number of calibration lamps and the final experimental wavelength was determined to be 3377.43 ± 0.26 Å. To confirm the line was really from W 27+ we determined the lifetime of the upper level of the 2 F term and compared our value with one from a calculation. As the line is from an M1 transition the lifetime of the upper level it is trivial to calculate the rate and thereby the lifetime. Our theoretical studies are systematic in two senses – first we investigate the contributions to the wavelength from different core subshells. These studies implies it is important to include core valence correlation even between the valence 4f and deep subshells – as a matter of fact, the correlation contribution from 3d is more pronounced than from 4d. The second dimension of the systematic approach is an increase of the active set of orbitals, and thereby the space of configuration state functions (CSFs) to show convergence of the final results.
What problem does this paper attempt to address?