Elsevier

Atomic Data and Nuclear Data Tables

Volumes 123–124, September–November 2018, Pages 313-328
Atomic Data and Nuclear Data Tables

The study of the core–valence and core–core correlation effects on the radiative properties along the magnesium isoelectronic sequence

https://doi.org/10.1016/j.adt.2018.01.003Get rights and content

Abstract

We investigated the effect of core–valence and core–core correlations on the energy levels and transition probabilities along the Mg isoelectronic sequence from Si2+ to U80+. In order to attain a certain accuracy for the atomic structure we considered configurations obtained by electron promotions from the n=3 up to n=7 valence shells. Core–valence and core–core correlations are systematically included in the model by allowing single and double electron promotions from the n=2 core up to the n=6 and n=5 shells with l4 and l2, respectively. The present results contain the energy of the levels arising from the valence configurations along with oscillator strengths and radiative rates corresponding to E1, M1, E2, M2 transitions between states arising from 3l3l with l,l2 and 3snl with n7 and l4 configurations. We conclude the data accuracy trend along the magnesium isoelectronic sequence as a result of the detailed study of the correlation effects on the radiative properties. The core–valence correlations generally improve both energy and radiative rate values, while the core–core correlations mostly give better oscillator strengths but may slightly increase the energy deviations. The calculations have been performed by employing the fully relativistic model-potential Flexible Atomic Code.

Introduction

Transitions belonging to the Mg-like ions are useful for performing diagnostics of astrophysical [[1], [2], [3]] and laboratory plasmas [[4], [5], [6], [7], [8]]. The necessity for accurate data led to a series of calculations for excitation energy, line strengths and transition probabilities by employing the multiconfiguration Hartree–Fock (MCHF) [[9], [10]], Dirac–Fock (MCDF) [11], Dirac–Hartree–Fock (MCDHF) [12], random-phase approximation [13] and configuration-interaction (CI) using Breit–Pauli (BP) [[14], [15], [16]] methods. The sequence itself is interesting from a theoretical standpoint since there are two electrons outside the complete neon core meaning that core–valence and core–core correlations play an important role in the atomic calculations [17] along with the many-body effects [18]. The 3s2(1S0), 3p2(1S0), 3d2(1S0) configuration interaction (CI) is dominant in the calculation and strong mixing also occurs between the 3p2(1D2) and 3s3d(1D2) terms, while the transition probabilities are significantly affected by the degree of electron correlation [19]. Resonant lines such as 3s2(1S0)3s3p(1P1o) play a crucial role in electron-impact excitation processes [20] and accurate calculation requires inclusion of more than valence–valence correlations [21].

In recent years the Mg-like ions have been subjected to different calculations using various methods. The R-matrix intermediate-coupling frame transformation (ICFT) approach [22] has been employed by L. Fernandez Menchero et al. [23] to compute the structure and electron-impact collision strengths for the magnesium like ions from Al II to Zn XIX. Only valence–valence correlations have been accounted for by using 283 levels arising from 3snl, 3pnl and 3dnl configurations with n5 and l4. Reasonable agreement with experimental results has been reached.

The MCDHF approach [24] has been used by F. Hu et al. [25] to compute energy levels and radiative transition probabilities for the Mg-like ions ranging from Z=19 to 92. The 1s22s22p6 neon core was kept fixed while the active set was systematically increased by including configurations belonging to the n=3 complex up to n=7, each time optimizing the orbitals. In order to account for core–valence correlations, the same procedure was applied using core-excited reference sets obtained by exciting an electron up to n4 from either 2p or 2s orbital. Good agreement between the computed atomic data for the n=3 to n=3 transitions and results from NIST [26] has been attained.

The MCDHF approach has been also employed recently by S. Gustafsson et al. [27] to accurately compute energy levels, lifetimes and transition rates belonging to the Ca IX to As XXII and Kr XXV Mg-like ions. The states arise from the multireference set containing the following configurations: 3l3l with l,l2, 3l4l with l2, l3 and 3s5l with l4. Single and double excitations to orbital sets with principal quantum number n=6,7,8 have been allowed while keeping the 1s shell closed. Only single excitations have been allowed from the 2p and 2s inner subshells.

A finite configuration interaction (CI) expansion can be chosen to well account for the near-degeneracy effect or static correlation, but in order to include the bulk of dynamic correlations one needs to consider a significantly large-scale CI expansion [28]. An alternative approach to adopting large-scale configuration interaction (CI) expansions is that of relativistic many-body perturbation theory (MBPT). The multi-reference Møller–Plesset (MR-MP) many-body perturbation theory [18] was used by J.A. Santana [29] to compute with unprecedented accuracy low-lying energy levels within the magnesium isoelectronic sequence from Z=12 to Z=100. The multi-reference configuration interaction (MR-CI) calculations have been performed for the 3sn13pn23dn3 configurations with i=13ni=3 accounting for the near-degeneracy effect while state-specific MR-MP calculations based on this CI expansion were performed in order to treat the residual electron correlation. Quantum electrodynamical effects were included by estimating the electron self-energy and the vacuum polarization.

The relativistic MBPT has been employed by U.I. Safronova et al. [30] to obtain excitation energies, radiative rates and lifetimes within the Mg-like ions for Z ranging from 13 to 100. The Breit interaction has been considered within the effective Hamiltonian and the electric dipole allowed transitions corresponding to the n=3 complex have been calculated in both length and velocity gauges.

Despite the necessity for knowledge of emission lines in plasma diagnostics, only few papers treat the forbidden transitions within the Mg-like ions. Y. Zou and C. Froese Fischer [12] used different optimization strategies for the intercombination 3s2(1S0)3s3p(3P1o) transition belonging to Mg-like Al II to S V ions. As a result, they managed to eliminate the transition probabilities dependence on optimization strategies and they further extended their calculations to magnetic dipole (M1), electric (E2) and magnetic (M2) quadrupole transitions between the 3s2(1S0), 3s3p(3P0,1,2o), 3s3p(1P1o) levels belonging to Mg-like ions with Z92 [31]. A comprehensive work on Na-like to Ar-like isoelectronic sequences performed by C. Froese Fischer et al. [19] also provides rates for selected M1, E2, M2 transitions belonging to Mg-like Mg I to Fe XV ions. More radiative rates for forbidden transitions are available from G. Tachiev and C. Froese Fischer [32].

There are no measured values available for M1, E2 and M2 transitions belonging to Mg-like ions to the authors knowledge. Forbidden M1 transition probabilities have been computed by A.M. Naqvi [33] by combining the Slater theory with the observed energy data. By comparing the empirically determined Slater parameters with the theoretically derived ones, he choose the values that resulted in better agreement between theory and experiment and then used them to compute the rest of the term intervals [34].

In a previous work, we computed the atomic structure and electron-impact collision strengths for the Mg-like S V ion while considering the effect of core–valence correlations by including 22 configuration-state function obtained by electron promotion from 2p subshell [35]. This improved the overall accuracy of the energy levels and transition probabilities as compared to other works from literature and led to the possibility to further investigate the core–valence and core–core correlations on the atomic structure of the Mg-like ions. In the case of Mg I, Al II, and Si III ions, an accurate treatment requires adopting a large-scale CI expansion containing mostly valence configurations instead of allowing addition of more core-excited CSFs [19]. Experimental values for the decay rates are sparse and obtaining good quality theoretical data represents a difficult task [[36], [37], [38], [39]].

This work aims to show the effects of accounting for core–valence and core–core correlation within the Mg-like ions from Si2+ to U80+ ions and related consequences on radiative properties while providing reliable data where no experimental measurements exist. This paper is structured as follows. Section 2 presents the method of calculation and the various CI expansions used in this work. Section 3 contains the data assessment for the results computed with the largest CI expansion and contains three separate subsections, the first one deals with the energy levels, the second and third treat the electric dipole allowed and forbidden M1, E2, M2 transitions, respectively. Section 4 gives the concluding remarks.

Section snippets

Method of calculation and results

The calculations are performed within the framework of the fully-relativistic model-potential Flexible Atomic Code (FAC) [40]. Quantum electrodynamic (QED) effects such as self-energy and vacuum polarization are treated within the screened hydrogenic approximation. The valence–valence correlations are accounted for by including [Ne]3snl, [Ne]3pnl, [Ne]3dnl configurations with 3n7 and 0l4 in the model, where [Ne] stands for the closed neon core. Further addition of states belonging to the n=8

Data assessment

The results presented in this section are obtained by employing the largest CI expansion considered in our work, containing both core–valence and core–core correlations denoted as “CC” in the previous section.

Conclusions

Systematic relativistic calculation of atomic structure for the magnesium-like ions accounting for core–valence and core–core correlations is performed in order to study the change in data accuracy. Inclusion in the CI expansion of core-excited states of type 213s2nl and 213s3pnl has a small and sometimes detrimental effect on the atomic structure. Further inclusion of core-excited configurations of type 213s3dnl and 213p2nl is shown to have a strong effect on the accuracy of the atomic

Acknowledgments

This work has been financed by the National Authority for Research and Innovation in the frame of Nucleus programme-contract 4N/2016. Partial financial support from the Institute of Atomic Physics under project number F01/2016 is also acknowledged.

References (66)

  • AggarwalK.M. et al.

    At. Data Nucl. Data Tables

    (2007)
  • ManiaA.J. et al.

    J. Quant. Spectr. Rad. Transfer

    (2009)
  • BadnellN.R.

    Comput. Phys. Comm.

    (2011)
  • JönssonP. et al.

    Comput. Phys. Comm.

    (2007)
  • HuF. et al.

    J. Quant. Spec. Rad. Transfer

    (2014)
  • SantanaJ.A.

    At. Data Nucl. Data Tables

    (2016)
  • IorgaC. et al.

    At. Data Nucl. Data Tables

    (2017)
  • LandiE.

    At. Data Nucl. Data Tables

    (2011)
  • Djeniz̆eS. et al.

    Spectrochim. Acta B

    (2010)
  • IrwinD.J.G. et al.

    Nucl. Instrum. Methods

    (1973)
  • Buchet-PoulizacM.C. et al.

    Nucl. Instrum. Methods

    (1982)
  • TrabertE. et al.

    Phys. Lett.

    (1982)
  • HuttonR. et al.

    Nucl. Instrum. Methods Phys. Res. B

    (1988)
  • CurdtW. et al.

    Astron. Astrophys.

    (2001)
  • FeldmanU. et al.

    Astrophys. J. Suppl.

    (1997)
  • CurdtW. et al.

    Astron. Astrophys. Suppl. Ser.

    (1997)
  • PutterichT. et al.

    Plasma Phys. Control. Fusion

    (2008)
  • JupenC. et al.

    Phys. Scr.

    (1990)
  • WyartJ.F.

    Phys. Scr.

    (1985)
  • StewartR.E. et al.

    J. Opt. Soc. Amer. B

    (1987)
  • SugarJ. et al.

    J. Opt. Soc. Amer. B

    (1989)
  • ChengK.T. et al.

    Phys. Rev. A

    (1977)
  • Froes FischerC. et al.

    Nucl. Instrum. Methods Phys. Res.

    (1982)
  • JohnssonP. et al.

    J. Phys. B: At. Mol. Opt. Phys.

    (1997)
  • ZouY. et al.

    Phys. Rev. A

    (2000)
  • HuangK.N. et al.

    Nucl. Instrum. Methods Phys. Res. B

    (1985)
  • ChenM.H. et al.

    Phys. Rev. A

    (1997)
  • KonovalovaE.A. et al.

    Phys. Rev. A

    (2015)
  • SantanaJ.A. et al.

    Phys. Rev. A

    (2015)
  • Froes FischerC. et al.

    At. Data Nucl. Data Tables

    (2006)
  • WallbankB. et al.

    Phys. Rev. A

    (2007)
  • SantanaJ.A. et al.

    J. Phys. B: At. Mol. Opt. Phys.

    (2010)
  • Fernandez-MencheroL. et al.

    Astron. Astrophys.

    (2014)
  • Cited by (0)

    View full text