Extreme ultraviolet and soft x-ray spectral lines in Rb XXIX
Khatri Indu1, †, , Goyal Arun1, Aggarwal Sunny1, 3, Singh A K2, Mohan Man1
Department of Physics and Astrophysics, University of Delhi, Delhi 110007, India
Department of Physics, D. D. U. College, University of Delhi, Delhi 110015, India
Department of Physics, Shyamlal College, University of Delhi, Delhi 110032, India

 

† Corresponding author. E-mail: indu.khatri.du@gmail.com

Abstract
Abstract

An extensive theoretical set of atomic data for Rb XXIX in a wide range with L-shell electron excitations to the M-shell has been reported. We have computed energy levels for the lowest 113 fine structure levels of Rb XXIX. The fully relativistic multiconfigurational Dirac–Fock method (MCDF) within the framework of Dirac–Coulomb Hamiltonian taking quantum electrodynamics (QED) and Breit corrections into account has been adopted for calculations. Radiative data are reported for electric dipole (E1), magnetic dipole (M1), electric quadrupole (E2), and magnetic quadrupole (M2) transitions from the ground level, although calculations have been performed for a much larger number of levels. To assess the accuracy of results, we performed analogous calculations using flexible atomic code (FAC). Comparisons are made with existing available results and a good agreement has been achieved. Most of the wavelengths calculated lie in the soft x-ray (SXR) region. Lifetimes for all 113 levels have also been provided for the first time. Additionally, we have provided the spectra for allowed transitions from n = 2 to n = 3 within the x-ray region and also compared our SXR photon wavelengths with experimentally recognized wavelengths. We hope that our results will be beneficial in fusion plasma research and astrophysical applications.

1. Introduction

As lines from fluorine-like ions have been observed in planetary nebulas,[1] solar flares[2,3] and in the spectra of laboratory plasmas,[4,5] atomic data are needed for diagnostic purposes. The factor which makes F-like configuration absolute for detailed study is the nearly full 2p shell which shows that the spectra are not too complex. The 2s22p5–2s2p6 transitions of F-like ions are observed in high-temperature plasmas and can thus be used for plasma diagnostic purposes. Magnetic dipole transitions between fine structure levels separated by spin–orbit interaction such as 2s22p5 2P3/22P1/2 transition in fluorine-like ions have important applications in astrophysics[6] and high-temperature plasma diagnostics.[7,8] The transition 2s22p5 2P3/22P1/2 is observed in tokamaks and is useful for measuring ion temperatures via Doppler broadening.[9] The emission line ratios involving the 2s22p5 2P3/2, 2s22p5 2P1/2, 2s2p6 2S levels of fluorine-like ions are useful for plasma diagnostics.[10] Forbidden lines of the fluorine isoelectronic sequence with configuration have been identified in tokamak discharges and in the solar corona and flares. As forbidden lines occur at long wavelengths, these are diagnostically important and hence accurate line profiles can be obtained.

The spectroscopy of highly charged ions is of considerable interest in both pure atomic physics and applications such as hot-plasma diagnostics, high temperature astrophysical plasma,[1113] x-ray laser development and for the study of high-field relativistic and quantum electrodynamic (QED) effects. Transitions in such ions are used to find temperature and density distributions and in calculating stellar envelope opacities.[14] Accurate measurements of energy levels provide tests of modern atomic-structure theory. Accurate atomic data such as transition probabilities, oscillator strengths, and lifetimes of excited levels are requisite in astrophysics, laser spectroscopy, laboratory plasma diagnostic, tokamak devices, and controlled thermonuclear reactions. Accurate atomic data are also needed for reliable interpretation of observed data.[15]

Astrophysical interest in the spectra of multiply charged ions has increased as the maximum number of lines emitted from high Z ions lie within extreme ultraviolet (EUV) 100–1200 Å and soft x-ray (SXR) 1–50 Å regions useful in providing detailed knowledge of coronal atmosphere. Recently, much progress has been made in the study of soft x-ray lasers. The interest in soft x-ray lasers motivates the study of the spectroscopy of highly charged ions.

F-like ions have been studied previously by various authors. Sampson et al.[16] calculated transition probabilities between levels of the 2s22p5, 2s2p6, 2s22p43l, and 2s2p53l configurations using the relativistic distorted wave method. Three lowest energy levels were calculated by Ivanova and Gulov[17] using the relativistic perturbation theory with a model potential. Kim and Huang[18] provided spin–orbit intervals in the ground state 2p5 2P3/22P1/2 of F-like ions for Z ≤ 56. Atomic binding energies for lithium to dubnium isoelectronic series in the Dirac–Fock approximation were calculated by Rodrigues et al.[19] Gu M F[20] calculated the level energies of 1s22lq (lq ≤ 8) states for ions with Z ≤ 60 using a combined configuration interaction and many-body perturbation theory approach. Cheng et al.[21] calculated energy levels, oscillator strengths, and transition probabilities for the first row atoms (Li through F) using the multiconfigurational Dirac–Fock method. Transition energies for ions from Z = 11 to Z = 33 were calculated by Edlén[22] using a semiempirical formula with three adjustable parameters. Some authors[23,24] using the CIV3 code obtained oscillator strengths of transition from the ground state of fluorine-like ions. Nahar[25] obtained atomic data for fluorine-like iron using relativistic Breit–Pauli approximation under the iron project. Fischer and Tachiev[26] obtained energy levels and transition rates for low lying states for ions up to Ti XIV using multiconfigurational Breit–Pauli wavefunctions. Jonauskas et al.[27] calculated energy levels and transition probabilities among levels of the ground and first 23 excited configurations of Fe XVIII using multiconfigurational Dirac–Fock method. Energies and E1, M1, E2 transition rates for 2s22p5 and 2s2p6 configurations states in F-like ions between Si VI and W LXVI were calculated by Jönsson et al.[28]

Feldman et al.[29] identified transitions of the type 2s–2p in the highly charged fluorine-like Mo, Cd, In, and Sn. Transitions of the type 2s–2p were also identified in F-like As, Se, Br, and Rb by Feldman et al.[30] Zigler et al.[31] presented wavelengths and identified 11 x-ray lines due to 2p–3d and 2p–3s transitions in Rb XXIX. The L-shell soft x-ray emission spectra of Cu, Zn, Ga, As, Br, Rb, and Sr from laser-produced plasma sources for 2p5–2p43s, 3d fluorine-like transitions were recorded by Hutcheon et al.[32] Denis-Petit et al. identified x-ray spectra in the Na-like to O-like Rubidium ions in the range 3.8–7.3 Å.[33]

The main focus of our work is to perform fully relativistic calculations in order to obtain transition energies, oscillator strengths, transition rates, and lifetimes for lowest 113 levels of the 2s22p5, 2s2p6, 2s22p43l, 2s2p53l, 2p63l, 2s22p44l, 2s2p54l, 2s22p45l, and 2s2p55l (l = 0,1,2,3) configurations of F-like Rb. In the present calculations, we have adopted the relativistic multiconfiguration Dirac–Fock (MCDF) approach. Such calculations including physical effects of the relativistic two-body Breit interaction and the QED corrections provide accurate tools to test the many-body atomic theory and for plasma diagnostics in high temperature plasmas. We have also provided a comparison of our results with the data compiled by NIST.[34,35] We theoretically identified EUV and soft x-ray transitions and plotted the Rb XXIX spectra. This work is an extension of our previous work[36] in which we investigated the spectra of Sr XXX.

2. Theoretical method

To perform large-scale calculations, the fully relativistic MCDF method originally developed by Grant et al.[37] and revised by Norrington[38] has been adopted which has successfully been applied earlier by ourselves and others.[3944] It takes into account all the relativistic effects as well as Breit and QED corrections important for highly ionized atoms.

In the MCDF method, the dominant interactions in an N-electron ion or atom are included in the Dirac–Coulomb Hamiltonian

where Ĥi denotes the Dirac Hamiltonian of each electron given by

Here, the first two terms denote the kinetic energy of the electron and the Coulomb potential of the nucleus is denoted by the last term, c is the speed of light, α and β represent 4 × 4 Dirac matrices, and Vnuc(ri) denotes the electron–nucleus Coulomb interaction.

The relativistic atomic state function of an atom or ion is described by an atomic state function (ASF) constructed approximately by a linear combination of configuration state functions (CSFs) with the same parity P and same angular momentum (J, M)

where nc is the number of CSFs considered in the evaluation of ASFs, Ci is the configuration mixing coefficient and denotes the number of electron–electron correlations to be considered, α denotes labeling of the configuration state function such as orbital occupancy, coupling scheme, etc, and |γr(PJM)〉 are CSFs specifying a particular state with a given parity and angular momentum (J,M).

The configuration state functions ϕ, are built as linear combinations of Slater determinants which, in the Dirac–Fock model, is a product of one-electron Dirac orbitals

where Pnk(r) and Qnk(r) denote the large and small component one-electron radial wavefunctions required to be orthonormal within each k symmetry. Here, k = ±(j + 1/2) for l = j±1/2, k is the Dirac angular quantum number and m is the projection of angular momentum j. In χkm(θ,φ, σ), σ is the electron spin z-projection quantum number, χkm is the spherical harmonics in the LSJ coupling scheme defined by

By the expectation value of Dirac Hamiltonian, the energy of the N-electron system is given as

where the Hamiltonian matrix HDC has elements

For the energy to be stationary with respect to variations in mixing co-efficients Ci(α) and using the orthonormality condition (Ci(α))+Cj(α) = δij

where I is the nc × nc unit matrix. Thus, equation (8) shows that atomic energy level Eα can now be taken to be eigenvalues of HDC.

CSFs are obtained from the antisymmetrized products of a common set of orthonormal orbitals which are optimized on the basis of Dirac–Coulomb Hamiltonian. By solving the MCDF equations, the radial functions are determined which depend upon varying the orbital radial functions to obtain an extremum of the energy functional. The wavefunction is expanded to thousands of CSFs and thus enables us to obtain a suitable description of level structure and transition properties for multiple and highly charged ions. Relativistic configuration interaction calculations are then carried out to evaluate CSF expansion coefficients by diagonalizing the Hamiltonian matrix. After obtaining self consistency, the contributions from the Breit interaction, vacuum polarization, self-energy and finite nuclear mass corrections are added as a first-order perturbation correction.

To calculate the radial wavefunctions, we have adopted the extended average-level (EAL) mode for self-consistent calculations in which a weighted trace of the Hamiltonian matrix is minimized. It can assign a configuration of a given weight automatically.

To assess the accuracy of our results from MCDF, we employ another fully relativistic code, the flexible atomic code (FAC)[45] based on Dirac equation. The atomic structure calculations are based on the relativistic configuration-interaction (CI) method. In FAC, the basis state functions are built by radial spinors obtained from a modified Dirac–Fock–Slater central field potential. The orbitals are optimized self-consistently where the average energy of a fictitious mean configuration with fractional orbital occupation numbers is minimized. This mean configuration stands for the average electron cloud of the configurations maintained in the configuration–interaction (CI) expansion. Using the Dirac–Coulomb Hamiltonian, relativistic effects are fully considered. Generally, there is no major discrepancy in the determination of energy levels and radiative data from FAC and MCDF. MCDF provides more options in terms of optimization schemes. The results from FAC will be helpful in assessing the accuracy of energy levels, radiative rates, and oscillator strengths.

3. Results and discussion
3.1. Energy levels

As velocity of electrons in the highly ionized atoms is fast and the nuclear potential energy is very strong, the relativistic effects must be taken into account. Therefore, in this work, calculations have been performed using fully relativistic MCDF approach. As an elaborate CI is needed to achieve a better accuracy, therefore we have included CI among 2s22p5, 2s2p6, 2s22p43l, 2s2p53l, 2p63l, 2s22p44l, 2s2p54l, 2s22p45l, and 2s2p55l configurations, where (l = s,p,d,f). We included correlation effects in the excitation up to the 5f orbital. In the last column of Table 1, we have also listed the mixing coefficients for these 113 levels. We find that some of the levels such as and 2s2p6 2S1/2 are almost pure. Due to clear dominance, some of the levels can be easily identified such as But some levels are strongly mixed, for example 2s22p4 (1S) 3p 2Po is strongly mixed with and with their compositions being 34%, 31%, and 28%. Therefore, identification of a particular level becomes difficult. Further, we notice that the levels 2s22p4 (1S) 3s 2S1/2 and can be interchanged due to strong mixing between them. As LSJ coupling is not accurate for highly mixed levels, the corresponding jj designations for the levels have also been provided under the column “JJ coupling”. The first number under this column represents the leading percentage of the level and p, p, d, d denote p1/2,p3/2,d3/2 and d5/2 respectively.

Table 1.

Level No., wave-function compositions and lifetimes of first 113 fine structure levels of Rb XXIX using MCDF.

.

In Table 2, we list our calculated energy levels of Rb XXIX, obtained from the MCDF wavefunctions with and without the inclusion of Breit and QED effects. We performed relativistic calculations and included 113 fine structure levels belonging to 2s22p5, 2s2p6, 2s22p43s, 2s22p43p, 2s22p43d, 2s2p53s, 2s2p53p, 2s2p53d, 2p63s, 2p63p, and 2p63d configurations. Our Breit and QED corrected energies are lower than the corresponding Coulomb energies by up to ∼ 0.33 Ryd. for a majority of levels. Our level energies without Breit and QED effects differ from the NIST values by (∼ 0.57 Ryd). Energies differ from NIST values by a maximum of (∼ 0.43 Ryd.) with the inclusion of Breit and QED effects, therefore we can deduce that higher relativistic effects for heavy ions are very important. Breit contribution to our results is up to 0.0216 Ryd. whereas QED contribution is up to 0.0258 Ryd. Further, it is seen that QED corrections are greater in magnitude for levels where single excitation from 2p to 3s occurs than excitations from 2p to 3p or 3d orbitals such as for The energy levels are arranged in ascending order. The present calculated values of the energy levels are compared with those calculated by Jönsson et al.[28] and with the compiled data from NIST.[34,35] To show the genuineness and credibility of our results, we performed another fully relativistic configuration interaction calculation important for highly ionized plasma using FAC. It is also a fully relativistic code based on jj coupling and results calculated are comparable to MCDF. In Table 2, we list the corresponding energies obtained from the FAC. All combinations of the 2s22p5, 2s2p6, 2s22p43l (l = 0,1,2), 2s22p44l (l = 0,1,2,3), 2s22p45l (l = 0,1,2,3,4), 2s2p53l (l = 0,1,2), 2s2p54l (l = 0,1,2,3), 2s2p55l (l = 0,1,2,3,4), 2p63l (l = 0,1,2), 2p64l (l = 0,1,2,3), 2p65l (l = and 0, 1, 2, 3, 4) configurations have been included in the calculations.

Table 2.

Dirac–Coulomb (DC), Breit and quantum electrodynamics (QED) contributions to the MCDF energy (in Ryd) for the lowest 113 levels for Rb XXIX. The sum total is compared with FAC and other available results. All energies are relative to the ground state.

.

A good agreement between the present calculations and the published data has been found. Taking NIST as a standard reference for comparison, we find that the maximum deviation between the present calculations and NIST energies and present results and Jönsson et al.[28] results is 0.84% for the level 2s2p6 2S1/2. The maximum disagreement of FAC calculations with our MCDF results is within 0.10%.

3.2. Radiative rates

For a transition ij, the relation between the absorption oscillator strength fi j (dimensionless), and radiative rate Aji (in s−1) is expressed as

In the above formula, m and e are the electron mass and charge respectively, c is the velocity of light, λji denotes the transition wavelength in Å, and wi and wj are the statistical weights of the lower i and upper levels j respectively. Line strength may be defined as

where

Here, ψi and ψj are the initial and final state wavefunctions, and Rij represents the transition matrix element of the appropriate multipole operator P. The oscillator strength fij and line strength Sij (in atomic units, a.u.) follow the following relations.

Table 3.

Transition data for E1, E2, M1, M2 transitions for Rb XXIX from lower level i, upper level j, wavelength λ (in Å), transition rate Aji (s−1) (length form), oscillator strength fij (length form), line strength Sij (in a.u.) (length form).

.

As the ratio of velocity/length is an indicator of accuracy of f- or A-values, we have also provided the ratios between length and velocity forms. A good agreement between the two forms is always desirable, but not a necessary condition for accuracy. The vel./len ratio in our case is generally within 10% of unity for most of the strong transitions (f ≥ 0.1). For some strong transitions such as the ratio is exactly unity. For weaker transitions (f ∼ 10−5 or less), such as the ratio is sometimes higher.

In Table 4, we have compared our calculated wavelengths of Rb XXIX with those compiled by NIST and other observed results by Denis–Petit et al.[33] We observe that our results agree well with both of them.

Table 4.

Comparison of computed wavelengths (in Å) with observed wavelengths from NIST and other available results for transitions from to selected upper levels for Rb XXIX.

.
3.3. Study of EUV and soft x-ray transitions

In the present paper, we have presented our calculated wavelengths of EUV and soft x-ray transitions in Table 3 by exciting a single electron from the L-shell. We have identified an EUV and 108 soft x-ray (SXR) transitions among the lowest 113 levels from the ground level. The transition lies in the EUV region and all other transitions 2s22p5 2Po–2s22p43l (l = s, p, d) are SXR.

In Table 3, there is a sudden decrease in the wavelength of the SXR photon from 49.32 Å to 6.767 Å which may be due to the increase in force of repulsion between inner and valence electrons thereby increasing transition energy. We have plotted gA spectra of Rb XXIX with SXR region for E1 and M2 transitions in Figs. 1 and 2 respectively, where g denotes the statistical weight of lower level.

Fig. 1. gA spectra of Rb XXIX within SXR region for E1 transitions.
Fig. 2. gA spectra of Rb XXIX within SXR region for M2 transitions.

In Fig. 3, we have plotted a comparison between our calculated SXR wavelengths and observed wavelengths from other available results. The horizontal axis titled ‘Line label’ represents the labeling of transitions as represented in the first column of Table 4.

Fig. 3. Circle and cross markers represent MCDF and observed wavelengths respectively.
4. Lifetimes

The lifetime (τ) for a level j for a transition ij is defined as follows:

In Table 1, we have provided lifetimes for all 113 levels from our calculations. These include A-values from all types of transitions, i.e., E1, E2, M1, and M2. It provides a check on the accuracy of the calculations, as it is a measurable quantity. To our knowledge, there are no results available for lifetimes for Rb XXIX to compare. We hope that the present results will be beneficial for future comparisons. These may encourage experimentalists to measure lifetimes, particularly for levels which have comparatively a larger value of lifetime.

5. Conclusion

In this paper, we have reported energy levels for the lowest 113 levels and radiative rates for E1, E2, M1, and M2 transitions from the ground level of Rb XXIX. Two independent calculations have been carried out using the MCDF and FAC codes. A comparison between the two sets of energy levels, as well as with the other available results, shows a satisfactory agreement. Wavefunction composition has been provided in both LSJ and JJ couplings. Though every care has been taken to identify the levels based on the strength of their eigenvectors, a possibility of their re-designation still remains. EUV and SXR transition wavelengths have been calculated using MCDF method. Comparisons have also been provided, for transition wavelengths for some selected transitions with NIST and other observed values are found to be in good agreement. The vel/len ratio is within 10% of unity, in particular the strong transitions with f ≥ 0.1 confirms the accuracy of our results.

Finally, the lifetimes for all 113 levels have been reported, although, due to paucity of data no comparisons with measured values are possible. However, our calculations will be helpful for an accuracy assessment for future measurement of lifetimes.

We have obtained new and previously unpublished energy levels, oscillator strengths, and transition probabilities for Rb XXIX. Hopefully, the present wide range of data will be useful in astrophysics, modeling of a variety of plasmas and for line identification in future experimental work.

Reference
1Aller L H1984Physics of Thermal Gaseous NebulaeDordrechtReidel
2Dere K P 1978 Astrophys. J. 221 1062
3Doyle J G 1983 Solar Phys. 89 115
4Stratton B CMoos H WSuckewer SFeldman USeely J FBhatia A K 1985 Phys. Rev. A 31 2534
5Suckewer SFonk RHinnov E 1980 Phys. Rev. A 21 924
6Edlén B 1969 Sol. Phys. 9 439
7Drawin H W 1981 Phys. Scr. 24 622
8Sudkewer S 1981 Phys. Scr. 23 72
9Suckewer SHinnov E 1978 Phys. Rev. Lett. 41 756
10Doschek G AFeldman U 1976 J. Appl. Phys. 47 3083
11Edlén B 1984 Phys. Scripta T8 5
12Träbert EHeckmann P HHutton RMartinson I 1988 J. Opt. Soc. Am B 5 2173
13Ellis D GMartinson ITräbert E 1989 Comm. At. Mol. Phys. 22 241
14Seaton M J 1987 J. Phys. B 20 6363
15Lambert D LWarner B 1968 Mon. Not. R. Astron. Soc. 140 197
16Sampson D HZhang H LFontes C J 1991 At. Data Nucl. Data Tables 48 25
17Ivanova E PGulov A V 1991 At. Data Nucl. Data Tables 49 1
18Kim Y KHuang K N 1982 Phys. Rev. A 26 1984
19Rodrigues G CIndelicato PSantos J PPatté PParente F 2004 At. Data Nucl. Data Tables 86 117
20Gu M F 2005 At. Data Nucl. Data Tables 89 267
21Cheng K TKim Y KDesclaux J P 1979 At. Data Nucl. Data Tables 24 111
22Edlén B 1983 Phys. Scr. 28 51
23Mohan MHibbert A 1991 Phys. Scr. 44 158
24Blackford H M SHibbert A 1994 At. Data Nucl. Data Tables 58 101
25Nahar S N 2006 Astronom. Astrophys. 457 721
26Fischer C FTachiev G 2004 At. Data Nucl. Data Tables 87 1
27Jonauskas VKeenan F PFoord M EHeeter R FRose S Jvan Hoof P A MFerland G JAggarwal K MKisielius RNorrington P H 2004 Astronom. Astrophys. 416 383
28Jönsson PAlkauskas AGaigalas G 2013 At. Data Nucl. Data Tables 99 431
29Feldman UEkberg J OSeely J FBrown C MKania D RMacGowan B JKeane C J 1991 J. Opt. Soc. Am. B 8 531
30Feldman USeely J FBehring W ERichardson M CGoldsmith S 1985 J. Opt. Soc. Am. B 2 1658
31Zigler AFeldman UDoschek G A 1986 J. Opt. Soc. Am. B 3 1221
32Hutcheon R JCooke LKey M HLewis C L SBromage G E 1980 Phys. Scr. 21 89
33Denis P DComet MBonnet THannachi FGobet FTarisien MVersteegen MGosselin GMéot VMorel PPain J-ChGilleron FFrank ABagnoud VBlazevic ADorchies FPeyrusse OCayzac WRoth M 2014 J. Quant. Spectrosc. Radiat. Transf. 148 70
34Kramida ARalchenko YReader J2015NIST ASD Team, NIST Atomic Spectra Database(Ver. 5.3), [Online]. Available http://physics.nist.gov/asd [2016, January 15] National Institute of Standards and Technology Gaithersburg, MD
35Sansonetti J E 2006 J. Phys. Chem. Ref. Data 35 301
36Goyal AKhatri IAggarwal SSingh A KMohan M 2015 J. Quant. Spectrosc. Radiat. Transf. 161 157
37Grant I PMckenzie B JNorrington P HMayers D FPyper N C 1980 Comput. Phys. Commun. 21 207
38Norrington P H2009http://www.am.qub.ac.uk/DARC/
39Quinet P 2012 J. Phys. B 45 025003
40Singh A KAggarwal SMohan M 2013 Phys. Scr. 88 035301
41Khatri IGoyal AAggarwal SSingh A KMohan M 2015 Chin. Phys. B 24 103202
42Aggarwal K MKeenan F PMsezane A Z 2003 At. Data Nucl. Data Tables 85 453
43Goyal AKhatri IAggarwal SSingh A KMohan M 2013 Can. J. Phys. 91 394
44Aggarwal SJha A K SKhatri ISingh NMohan M 2015 Chin. Phys. B 24 103202
45Gu M F 2008 Can. J. Phys. 86 675