11. Nitrogen iso-electronic sequence
There are three main groups of 2s2 2p3 - 2s2p4 transitions which can be observed at SERTS-89 wavelengths, namely 4S - 4P, 2D - 2D and 2D - 2P. There are also three groups of weaker lines decaying down from the D, P and S levels to the P levels in the ground state, with the P - S transitions being the strongest of those.
11.1. Ne IV
The S - P and D - D transitions occur above 450 Å and so are not observed by SERTS-89. Subsequently, we only have two Ne IV lines identified (357.89 and 421.59 Å), both of which are weak. Dwivedi et al. (1997) suggest the identification of a line at 388.228 Å, and we give the details of the line fit and transition in Table 24.
CHIANTI predicts the strongest Ne IV line to occur at 358.68 Å, but it is not reported in the catalogue. However, a Fe XI line is identified at 358.67 Å and, as explained in Sect. 13.3, the intensity is too strong to be accounted for by Fe XI alone. We suggest that Ne IV is a significant component of this feature, and can estimate its contribution as follows. As shown in Sect. 10.2, Ne IV accounts for 2.5 erg cm-2 s-1 sr-1 of the blend with Ne V at 357.95 Å. Using this and the density insensitive ratio given in Table 13 for 357.89/358.68, we then get a Ne IV contribution of 5 erg cm-2 s-1 sr-1 to the 358.67 Å blend.
The 388.23/357.89 ratio is also density insenstive, and the revised value for the 357.89 Å intensity ensures that the ratio agrees well with theory (Table 13).
The line at 421.59 Å is a self-blend excited principally from the ground P levels and so shows density sensitivity relative to the 357.89 Å line up to around 10 cm-3 when the P levels come into Boltzmann equilibrium with the D levels. Although this ratio is more useful for lower density sources, we note that the intensity ratio observed by SERTS-89, as corrected above, is consistent with the expected high-density limit predicted by CHIANTI and shown in Table 14.
11.2. Mg VI
The three groups of strong lines mentioned in the introduction are all present in the SERTS-89 spectrum. In addition, Thomas & Neupert identified a SERTS-89 feature at 319.73 Å as the 2s2p D - 2p P transition which is not included in the CHIANTI atomic model. However, that line is principally excited by very weak transitions from the ground levels and is not likely to have a significant intensity, so we can rule out this identification.
There are two relatively strong P - D transitions predicted by CHIANTI at 387.79 Å and 388.01 Å, whereas Thomas & Neupert reported only one broad feature at 387.95 Å that they identified as the second of these lines. We now find that both lines can be measured in the SERTS-89 spectrum with wavelengths of 387.769 Å and 387.967 Å. Other details of the fits can be found in Table 24. The two lines are density insensitive relative to each other and also with the 349.16 Å line. We present comparisons with theory in Table 13, where good agreement is found.
The line at 403.30 Å is blended with a Ne VI transition (see Sect. 9.1), and we can use a Ne VI branching ratio to estimate the Mg VI contribution as around 27 erg cm-2 s-1 sr-1. We use this value in Table 13 where it shows excellent agreement with the 400.67 Å line. We also find the 399.28/400.67 ratio to be in agreement with theory.
The two D - P transitions at 269.04 Å and 270.40 Å are density insensitive relative to each other and show agreement with theory (Table 13). However when compared with the D - D transitions which are self-blended at 349.16 Å - we find a discrepancy as witnessed by the 270.40/349.16 ratio. The 349.16 Å self-blend seems too low in intensity by about 30 - 40%.
Taking the ratio of the D - D transitions to the S - P transitions potentially yields a good density diagnostic over 10 - 10 cm-3. However, if we take the 349.16/400.67 ratio, for example, the high density limit is around 2.3 and yet the observed ratio is , suggesting either that the S - P transitions are too weak in the SERTS-89 spectrum or that the 349.16 Å self-blend is too strong. (Note that the latter possibility is in direct conflict with the conclusion drawn in the previous paragraph.)
Of further interest regarding density diagnostics are two lines predicted by CHIANTI at 314.56 Å and 314.67 Å, which are P - S transitions and related to each other by the branching ratio given in Table 12. Both lines are strongly density sensitive relative to 349.16 Å up to 10 cm-3, but neither are identified in the catalogue. However we do find an unidentified line at 314.59 Å, which Dwivedi et al. (1997) identify as the CHIANTI 314.67 Å line. We note that the 314.67/349.16 ratio has an upper limit of 0.15, yet the observed 314.59/349.16 ratio is , marginally above this value. Another possibility is that the CHIANTI 314.56 Å and 314.67 Å lines are blended in the SERTS-89 spectrum, but the separation of the lines is governed by the separation of the ground configuration 2P levels which is strongly constrained by observations of ultra-violet lines at 1190.1 Å and 1191.7 Å. Thus SERTS-89 should be capable of resolving these lines. On account of the above we can not confirm the identification of Dwivedi et al. (1997), and we await further high resolution spectra of this wavelength region.
Table 12. Branching ratios for the N-like ions.
Table 13. Insensitive ratios for the N-like ions.
Table 14. Density sensitive line ratios for the N-like ions.
Dwivedi et al. (1997) give two additional Mg VI identifications at 291.359 Å and 293.146 Å, and we give details of the line fits and transitions in Table 24. Both lines are blends of two Mg VI transitions, and they are related to the 269.04 Å and 270.40 Å lines, respectively, via branching ratios. Comparisons with theory are presented in Table 13, where good agreement is found.
The problems noted in comparing the D - P, D - D and S - P transitions seem likely to be due to inaccuracies in the Mg VI atomic data adopted for CHIANTI. Since there are very few calculations of collision rates and transition probabilities available in the literature, future studies would be highly desirable for this ion.
11.3. Si VIII
The 338.38 Å line identified in the SERTS-89 catalogue is a 2s2p - 2p transition and so is not in the CHIANTI model of Si VIII. However, for the same reason as with the Mg VI 2s2p - 2p transition, we dismiss this identification. The other reported transitions come from the 4S - 4P and 2D - 2 D multiplets.
The 277.05 Å line is reported as blended with a Mg VII transition and from Sect. 10.3 we can estimate the Si VIII contribution to be around 60 erg cm-2 s-1 sr-1. Although not reported, the 276.85 Å line is expected to be blended with a Si VII transition and from Sect. 12.3 we estimate a Si VIII contribution of 45 erg cm-2 s-1 sr-1. As the two ground D levels come into Boltzmann equilibrium with the S level at different densities, the 276.85/277.05 ratio is sensitive to densities up to around 10 cm-3. With the "de-blended" intensities we find the ratio close to the high density limit and we give the density estimate in Table 14.
The three 4S - 4P transitions that occur at 314-320 Å are density insensitive relative to each other, and comparisons with theory provided in Table 13. Reasonable agreement is found.
Comparing either of the 2D - 2D lines to one of the 4S - 4P lines yields a density diagnostic valid over the range 10 - 10 cm-3. We choose the 277.05/319.84 ratio and use the "de-blended" intensity for the 277.05 Å line to give the density shown in Table 14.
The 2D - 2P transitions, not reported by Thomas & Neupert, occur at around 216 Å and so lie in the second order bandpass of SERTS. Dwivedi et al. (1997) report a new identification for the Si VIII 216.92 Å line (the strongest of the 2D - 2P transitions) and details of the fit are given in Table 24. This line is density insensitive relative to the 277.05 Å line and we compare the observed ratio with theory in Table 13, where we have used the above intensity estimate of 60 erg cm-2 s-1 sr-1 for the 277.05 Å line. The large error bars on the observations allow marginal agreement.
11.4. S X
Only the 4S - 4P transitions fall within the SERTS-89 bandpass and of these only two of the three transitions are reported in the catalogue, although Dwivedi et al. (1997) have since reported the identification of the remaining transition, for which the fit details are given in Table 24. The three lines are all density insensitive relative to each other with values given in Table 13. The 259.45/264.22 ratio shows only marginal agreement with theory, suggesting that the 259.45 Å line may be blended, although we note that Malinovsky & Heroux (1973) give a value of 1.06 for this ratio in their full-disk solar spectrum, consistent with the SERTS value.
The nitrogen sequence of ions potentially provide excellent density diagnostics, but for the lines in the SERTS-89 spectrum we are hampered by blending and inconsistencies in the atomic data.
© European Southern Observatory (ESO) 1998
Online publication: November 24, 1997