next previous
Up: Atomic data from the


2 The calculation

The basic atomic theory, the approximations and the computer codes employed in the IRON Project are described by Hummer et al. (1993). The target wavefunctions were constructed from 1s, 2s, 2p, 3s and 3p orbitals as given by Clementi & Roetti (1974), together with a 3d orbital optimised on the energy of the third $^{2}P^{\rm o}$ state (i.e. 3s3p53d), and a 4f correlation orbital optimised on the ground state. The optimizations were carried out using Hibbert's (1975) variational program CIV3. The radial parts of the Slater-type orbitals are
$\displaystyle 3{\rm d}(r)$ = $\displaystyle 141.9677864\exp(-8.0582535r)r^3$  
    $\displaystyle +58.2199125\exp(-4.3181676r)r^3$  
$\displaystyle 4{\rm f}(r)$ = $\displaystyle 183.85325\exp(-5.175)r^4.$  

All configurations were included with a minimum number of electrons in each shell of 1s22s22p63s03p2 and with a maximum of three electrons in the 3d shell and one electron in 4f. This correlation was necessary in order to converge the oscillator strength for the 3s23p5-3s3p6 transitions. Two complementary R-matrix calculations were carried out, as now described.

A 180-level BPRM calculation was used for the resonance energy region (up to 9.95 Ryd), with a maximum of 1104 channels in each partial wave. In order for this calculation to be computational feasable, the number of continuum terms was the minimum required to span this energy range: only five per channel, resulting in collisional Hamiltonian matrices of maximum order 6164. The purpose of this calculation was to obtain accurate collision strengths in the low-energy resonance region, and since resonances are important only for low partial waves the expansion was truncated at J=6.

A 31-level BPRM calculation was used to top-up both the energy and the partial-wave expansion, in order to calculate converged collision strengths to a high enough energy for collision rates to be obtained over a realistic temperature range. The 31-level calculation had only 158 channels, so 30 continuum terms per channel could be included and the partial-waves calculated up to J=56, enabling converged collision strengths to be calculated for the transitions 3s23p5-3s3p6, 3s23p5-3s23p43d and 3s3p6-3s23p43d (i.e. up to the lowest 31 levels), and the energy range extended from 9.95 to 600 Ryd.

Table 1 lists the energies of the lowest 31 levels calculated from the wavefunction used in the 181-level BPRM calculation, and compares with NIST reference data. Table 2 compares the oscillator strengths obtained using the 181-level BPRM wavefunction with those of Bhatia & Doshek (1995). The oscillator strengths from the two calculations are qualitatively similar but not in very good agreement. However, our results are much closer to a recent experiment by Träbert (1996) for the lifetime and branching ratio of the 3s3p6 2S $^{\rm e}_{1/2}$ level (Table 3).


 

 
Table 1: The lowest 31 energy levels (Ryd) for Fe X. For reference, level 32 (3s3p53d) is calculated at 6.284 Ryd. 180 levels were actually included in the calculation (see Fig. 1). "Expt'' is reference data from Sugar & Corliss (1985)
i State J Present Expt
1 $3{\rm s}^{2}3{\rm p}^5$ $^2{\rm P}^{\rm o}$ 3/2 .0000 .0000
2 $3{\rm s}^{2}3{\rm p}^5$ $^2{\rm P}^{\rm o}$ 1/2 .1346 .1429
3 $3{\rm s}3p^6$ $^2S^{\rm e}$ 1/2 2.631 2.636
4 $3{\rm p}^43{\rm d}$ $^4{\rm D}^{\rm e}$ 5/2 3.581 3.542
5 $3{\rm p}^43{\rm d}$ $^4{\rm D}^{\rm e}$ 7/2 3.582 3.542
6 $3{\rm p}^43{\rm d}$ $^4{\rm D}^{\rm e}$ 3/2 3.591 3.554
7 $3{\rm p}^43{\rm d}$ $^4{\rm D}^{\rm e}$ 1/2 3.603 3.568
8 $3{\rm p}^43{\rm d}$ $^4{\rm F}^{\rm e}$ 9/2 3.863 3.806
9 $3{\rm p}^43{\rm d}$ $^2{\rm P}^{\rm e}$ 1/2 3.875  
10 $3{\rm p}^43{\rm d}$ $^4{\rm F}^{\rm e}$ 7/2 3.906 3.853
11 $3{\rm p}^43{\rm d}$ $^4{\rm F}^{\rm e}$ 5/2 3.938 3.889
12 $3{\rm p}^43{\rm d}$ $^4{\rm F}^{\rm e}$ 3/2 3.947 3.903
13 $3{\rm p}^43{\rm d}$ $^2{\rm P}^{\rm e}$ 3/2 3.952 3.936
14 $3{\rm p}^43{\rm d}$ $^4{\rm P}^{\rm e}$ 1/2 4.014 3.962
15 $3{\rm p}^43{\rm d}$ $^2{\rm D}^{\rm e}$ 3/2 4.015 3.961
16 $3{\rm p}^43{\rm d}$ $^4{\rm P}^{\rm e}$ 3/2 4.057  
17 $3{\rm p}^43{\rm d}$ $^4{\rm P}^{\rm e}$ 5/2 4.079 4.026
18 $3{\rm p}^43{\rm d}$ $^2{\rm F}^{\rm e}$ 7/2 4.081 4.017
19 $3{\rm p}^43{\rm d}$ $^2{\rm D}^{\rm e}$ 5/2 4.097  
20 $3{\rm p}^43{\rm d}$ $^2{\rm G}^{\rm e}$ 9/2 4.172 4.108
21 $3{\rm p}^43{\rm d}$ $^2{\rm G}^{\rm e}$ 7/2 4.173 4.111
22 $3{\rm p}^43{\rm d}$ $^2{\rm F}^{\rm e}$ 5/2 4.195 4.126
23 $3{\rm p}^43{\rm d}$ $^2{\rm F}^{\rm e}$ 5/2 4.460  
24 $3{\rm p}^43{\rm d}$ $^2{\rm F}^{\rm e}$ 7/2 4.495 4.429
25 $3{\rm p}^43{\rm d}$ $^2{\rm D}^{\rm e}$ 3/2 4.735 4.664
26 $3{\rm p}^43{\rm d}$ $^2{\rm D}^{\rm e}$ 5/2 4.774  
27 $3{\rm p}^43{\rm d}$ $^2{\rm S}^{\rm e}$ 1/2 5.059 4.938
28 $3{\rm p}^43{\rm d}$ $^2{\rm P}^{\rm e}$ 3/2 5.255 5.141
29 $3{\rm p}^43{\rm d}$ $^2{\rm P}^{\rm e}$ 1/2 5.307 5.194
30 $3{\rm p}^43{\rm d}$ $^2{\rm D}^{\rm e}$ 5/2 5.351 5.221
31 $3{\rm p}^43{\rm d}$ $^2{\rm D}^{\rm e}$ 3/2 5.466 5.342



 

 
Table 2: gf-values for transitions from the 3p5 2P $^{\rm o}_{3/2,1/2}$ground levels to the the 3s3p6 and 3p43d levels for Fe X. BD = gf calculated by Bhatia & Doshek (1995); Present = gf calculated from the same wavefunction as used in the present 181-level BPRM collision calculation; A = present calculated A-value, s-1. The level indexing (i,i') is defined in Table 1
i-i' Ji' BD Present A, s-1
1-3 1/2 0.0716 0.1027 2.854E9
1-4 5/2 0.71E-4 2.08E-4 3.504E6
1-6 3/2 2.06E-4 2.02E-4 5.281E6
1-7 1/2 9.14E-5 5.59E-5 2.926E6
1-9 1/2 2.68E-3 8.12E-4 4.954E7
1-11 5/2 1.24E-3 1.49E-3 3.031E7
1-12 3/2 19.7E-3 9.03E-3 2.783E8
1-13 3/2 5.65E-5 2.10E-3 6.651E7
1-14 1/2 2.25E-3 3.56E-3 2.325E8
1-15 3/2 9.56E-3 4.85E-3 1.557E8
1-16 3/2 28.2E-4 4.17E-4 1.397E7
1-17 5/2 6.40E-3 4.64E-3 1.017E8
1-19 5/2 15.9E-3 3.67E-3 8.114E7
1-22 5/2 11.1E-4 6.82E-4 1.579E7
1-23 5/2 3.66E-3 4.60E-3 1.206E8
1-25 3/2 0.0131 0.0100 4.441E8
1-26 5/2 0.58E-3 2.45E-3 7.356E7
1-27 1/2 1.940 1.312 1.338E11
1-28 3/2 3.788 2.817 1.542E11
1-29 1/2 0.1863 0.3693 4.260E10
1-30 5/2 6.462 5.128 1.941E11
1-31 3/2 0.283 0.1710 1.013E10
2-3 1/2 0.0365 0.0513 1.285E9
2-6 3/2 31.9E-6 2.93E-6 7.094E4
2-7 1/2 13.6E-5 7.75E-5 3.757E6
2-9 1/2 72.0E-3 3.62E-3 2.055E8
2-12 3/2 7.51E-3 3.96E-3 1.137E8
2-13 3/2 4.16E-4 3.95E-4 1.169E7
2-14 1/2 6.77E-4 7.17E-4 4.382E7
2-15 3/2 8.56E-3 3.31E-3 9.929E7
2-16 3/2 15.9E-5 6.84E-5 2.149E6
2-25 3/2 11.7E-3 7.62E-3 3.193E8
2-27 1/2 0.2115 0.4553 4.403E10
2-28 3/2 0.2402 0.1052 5.468E9
2-29 1/2 1.833 1.091 1.196E11
2-31 3/2 4.004 3.213 1.810E11



 

 
Table 3: Predictions and measurement (Träbert 1996) on the line doublet 3s23p5 2P $^{\rm o}_{3/2,1/2}$- 3s3p6 2S $^{\rm e}_{1/2}$ for Fe X. BD = calculated from Bhatia & Doshek (1995); Present = calculated from the same wavefunction as used in the present 181-level BPRM collision calculation
  BD Present Measured
Lifetime ps. 344 242 $270\pm20$
Branch ratio 2.17 2.22 $2.4\pm0.3$



 

 
Table 4: A comparison of collision strengths for Fe X fine structure transitions as a function of electron impact energy (Ryd). The level indexing (i,i') is defined in Table 1; the numbers on the same line as the transition are collision strengths from the present calculation and the ones below labelled BD are from Bhatia & Doschek (1995) for the same transition
i-i' 18.0 27.0 36.0 45.0
1-3 .356 .363 .361 .361
BD .421 .441 .461 .479
2-3 .186 .188 .188 .188
BD .226 .239 .252 .262
1-4 .044 .028 .020 .015
BD .039 .026 .018 .013
2-4 .014 .0085 .0060 .0045
BD .012 .0079 .0056 .0042
3-4 .0020 .0013 .0010 .0008
BD .0020 .0013 .0009 .0007
1-5 .0662 .0423 .0293 .0213
BD .0593 .0383 .0265 .0193



next previous
Up: Atomic data from the

© ESO 2001