emission lines of ni xviii in the solar euv spectrum

EMISSION LINES OF Ni XVIII IN THE SOLAR EUV SPECTRUM

F. P. KEENAN and V. J. FOSTERDepartment of Pure and Applied Physics, The Queen’s University of Belfast, Belfast BT7 1NN,

Northern Ireland

M. MOHANDepartment of Physics and Astrophysics, University of Delhi, Delhi–1110007, India

K. G. WIDINGCode 7674W, E.O. Hulburt Center for Space Research, Naval Research Laboratory,

Washington DC, 20375–5352, U.S.A.

(Received 13 August, 1996; in final form 8 November, 1996)

Abstract. Using electron excitation rates calculated with the R-matrix code, theoretical Ni XVIIIelectron-temperature-sensitive emission line ratios are presented forR1 = I(220:41 A)=I(320:56 A),R2 = I(233:79 A)=I(320:56 A), and R3 = I(220:41 A)=I(292:00 A). A comparison of these withobservational data for two solar flares, obtained by the Naval Research Laboratory’s S082A slitlessspectrograph on board Skylab, reveals good agreement between theory and observation for R1 andR2 in two spectra, which provides limited support for the accuracy of the atomic data adopted inthe analysis. However, several of the measured ratios are much larger than theory predicts, whichis probably due mainly to saturation of the strong 292.00 and 320.56 A lines on the photographicfilm used to record the S082A data. A comparison of our line ratio calculations with active regionobservations made by the Solar EUV Rocket Telescope and Spectrograph (SERTS) indicate that afeature at 236.335 A, identified as the Ni XVIII 3p 2P3=2 � 3d 2D3=2 transition in the SERTS data, isactually the Ar XIII 2s22p2 3P0 � 2s2p3 3D1 line. The potential usefulness of the Ni XVIII line ratiosas electron temperature diagnostics for the solar corona is briefy discussed.

1. Introduction

Emission lines arising from transitions in Na-like ions have been frequentlyobserved in the spectra of both laboratory and astrophysical plasmas (Sandlinet al., 1986; Wang et al., 1988). Flower and Nussbaumer (1975) pointed out thatthese transitions may be used to determine the electron temperature of the solartransition region/corona through diagnostic line ratios, although to calculate thesereliably accurate atomic data must be employed, especially for electron impactexcitation rates. In addition, Feldman and Doschek (1977) have shown that theNa-like emission line ratios are also sensitive to variations in the electron densitywhen Ne � 1013 cm�3, so that they may be used as Ne diagnostics in laboratoryplasmas such as tokamaks, where Ne is greater than this limit (see, for example,Keenan et al., 1991).

Recently, Mohan, Sharma, and Hibbert (1996) have calculated electron impactexcitation rates for transitions in Na-like Ni XVIII with theR-matrix code of Burkeand Robb (1975). In this paper we use the Mohan et al. rates to derive emissionline ratios for Ni XVIII applicable to the solar atmosphere, and compare these withflare observations from the S082A instrument on board Skylab.

Solar Physics 171: 337–343, 1997.c 1997 Kluwer Academic Publishers. Printed in Belgium.

GR: 201018825, PIP: 125694 SPACKAP

*125694* sola7089.tex; 13/03/1997; 9:15; v.7; p.1

338 F. P. KEENAN ET AL.

2. Atomic Data and Theoretical Line Ratios

The model ion for Ni XVIII consisted of the six energetically lowest LS states,namely 3s 2S, 3p 2P , 3d 2D, 4s 2S, 4p 2P , and 4d 2D, making a total of tenfine-structure levels. Energies of all these levels were taken from Corliss and Sugar(1981).

Electron impact excitation rates for transitions among the Ni XVIII levels dis-cussed above have been calculated by Mohan et al. (1996) using theR-matrix code(Burke and Robb, 1975). These data, which include the effects of resonance struc-ture in the collision strengths, are probably the most accurate currently available,and have therefore been adopted in the present analysis.

Einstein A-coefficients for allowed transitions in Ni XVIII were obtained fromSampson, Zhang, and Fontes (1990). Radiative rates for the forbidden 3s 2S�3d 2D

transitions, extrapolated from the results of Godefroid et al. (1985) for Ti XII, Cr XIV,and Fe XVI, were also included in the model ion, although we note that inclusionof these data do not have a significant effect on the derived emission line ratios.

As has been discussed by, for example, Seaton (1964), proton excitation maybe important for transitions with small excitation energies, i.e., fine structure trans-itions such as that in the 2s22p 2P ground term of boron-like ions (Foster, Keenan,and Reid, 1996). However, test calculations for Ni XVIII setting the proton rates for2P1=2 �

2 P3=2 and 2D3=2 �2 D5=2 equal to the electron rates, or 100 times these

values, had a negligible effect on the level populations, showing this atomic processto be unimportant, as found for other Na-like ions under solar plasma conditions(see Keenan et al., 1994a, and references therein).

Using the atomic data discussed above in conjunction with the statistical equi-librium code of Dufton (1977), relative Ni XVIII level populations and henceemission line strengths were calculated for a range of electron temperaturesabout that of maximum Ni XVIII fractional abundance in ionization equilibrium,logTe = logTmax = 6:5 (Arnaud and Rothenflug, 1985). Details of the proceduresinvolved and approximations made may be found in Dufton (1977) and Duftonet al. (1978).

In Figure 1 the theoretical emission line ratios

R1 = I(3p 2P1=2 � 3d 2D3=2)=I(3s2S � 3p 2P1=2) = I(220:41 A)=I(320:56 A)

and

R2 = I(3p 2P3=2 � 3d 2D5=2)=I(3s 2S � 3p 2P1=2) = I(233:79 A)=I(320:56 A)

are plotted as a function of electron temperature at an electron density of Ne =

1010 cm�3, although we note that the results are insensitive to variations in thedensity for Ne < 1017 cm�3 (Feldman and Doschek, 1977). The ratios

R3 = I(3p 2P1=2 � 3d 2D3=2)=I(3s 2S � 3p 2P3=2) = I(220:41 A)=I(292:00 A)

sola7089.tex; 13/03/1997; 9:15; v.7; p.2

EMISSION LINES OF Ni XVIII IN THE SOLAR EUV SPECTRUM 339

Figure 1. The theoretical Ni XVIII emission line ratios


and

R2 = I(3p 2P3=2 � 3d 2D5=2)=I(3s2S � 3p 2P1=2) = I(233:79 A)=I(320:56 A);

where I is in energy units, plotted as a function of logarithmic electron temperature (Te in K) at anelectron density of Ne = 1010 cm�3, although we note that the results are insensitive to variations inthe density for Ne < 1017 cm�3 (Feldman and Doschek, 1977).

and


have the same temperature dependence as R1 due to common upper levels, exceptthat

R3 = 0:473�R1 ;

R4 = 0:152�R1 :

As all of the emission lines in R1 to R4 have large A-values, they are in thecoronal approximation (Elwert, 1952), so that the theoretical values of the ratiosdepend principally on the ratio of the excitation rates of the 3d 2D and 3p 2P levelsfrom the 3s 2S ground state. These should be in error by less than 10% (Mohan

sola7089.tex; 13/03/1997; 9:15; v.7; p.3


Table INi XVIII emission line ratios

Solar flare R1 R2 R3 R1=R2 R3=R1

December 17, 1973: 00:44 UT 2.88–1a 2.26–1 9.13–2 1.27 0.32January 21, 1974: 23:24 UT 6.36–2 1.23–1 6.20–2 0.52 0.97January 21, 1974: 23:46 UT 1.06–1 1.25–1 6.98–2 0.85 0.66

Theoretical valueb 8.49–2 1.38–1 4.02–2 0.62 0.47

a A–B implies A �10�B .b Determined from Figure 1 at logTe = logTmax = 6:5 (Arnaud and Rothenflug, 1985).

et al., 1996), so that the theoretical line ratios should be accurate to better than15%.

3. Observational Data

The 220.41, 233.79, 292.00, and 320.56 A emission lines in Ni XVIII have beenobserved in solar flare spectra obtained with the Naval Research Laboratory’sXUV slitless spectrograph (S082A) on board Skylab (Dere, 1978). This instrumentoperated in the 171–630 A wavelength range in two sections (171–350 A and 300–630 A), and produced dispersed images of the Sun on photographic film with aspatial resolution of 200 and a maximum spectral resolution of�0.1 A. Details of theS082A spectrograph and reduction procedures can be found in Parker et al. (1976)and Tousey et al. (1977), the absolute calibration curves of Dere and Mango (1979)and Dufton et al. (1983) being adopted for wavelengths >260 A and �260 A,respectively.

In Table I we summarize measurements of the R1, R2, and R3 ratios for thesolar flares of December 17, 1973 at 00:44 UT (discussed in detail by Dere et al.,1979; Widing and Spicer, 1980; Widing and Cook, 1987), and January 21, 1974at 23:24 UT and 23:46 UT (Brueckner, 1976; Hiei and Widing, 1979; Widing andHiei, 1984). The S082A line ratios should normally be accurate to approximately�30% (Keenan et al., 1984; Widing, Feldman, and Bhatia, 1986). However theNi XVIII 233.79 A feature lies in the wing of Fe XV 233.90 A (Dufton, Kingston,and Widing, 1990), while the Ni XVIII 320.56 A flare image is projected on to theHe II 304 A disk. As a result, we believe the line ratio data presented here areprobably only accurate to typically �40%.

4. Results and Discussion

In Table I we list the theoretical values of R1, R2, and R3 at the temperature ofmaximum Ni XVIII fractional abundance in ionization equilibrium, logTmax = 6:5

sola7089.tex; 13/03/1997; 9:15; v.7; p.4


(Arnaud and Rothenflug, 1985). An inspection of the table reveals that, in viewof the errors discussed in Section 3, there is good agreement between theoryand observation for R1 and R2 in the January 21 flare, with discrepancies thataverage only �15% for both the 23:24 UT and 23:46 UT spectra. However, theother observed line ratios are consistently larger than theory predicts, by up to afactor of 3.4 in one instance (R1 in the December 17 flare). These disagreementscould be due to (i) blending of the 220.41 and 233.79 A lines (which wouldlead to their intensities being overestimated), (ii) the strong 292.00 and 320.56 Aresonance transitions being optically thick, or (iii) saturation of these lines on thephotographic film, both of which would lead to an underestimation of the lineintensity. To investigate the first of these possibilities, we summarize in Table I theobserved values ofR1=R2 = I(220:41 A)=I(233:79 A), along with the theoreticalestimate. An inspection of the table reveals generally good agreement (within30%) between observation and theory for these ratios, indicating that it is highlyunlikely that blends contribute a significant amount to the 220.41 and 233.79 A lineintensities in most instances. The exception to this is the December 17 flare, wherethe observed I(220:41 A)=I(233:79 A) ratio is more than twice the theoreticalvalue. It is possible that in this flare the 220.41 A feature is blended with the O V

line at 220.35 A, as this transition should be present in high electron density solarflares, such as that of December 17 (Keenan et al., 1994b).

To investigate possibilities (ii) and (iii), namely optical thickness or saturationof the 292.00 and 320.56 A lines, we list in Table I the measured R3=R1 =

I(320:56 A)=I(292:00 A) ratios, along with the theoretical estimate. Although theDecember 17 flare observation is within �30% of theory, the January 21 23:24 UTand 23:46 UT measurements are a factor of �2.1 and �40% larger than theorypredicts, respectively. This suggests that the observed intensity of the 292.00 A lineis too small (as the 320.56 A line should be free from blends; Dere, 1978), whichmay be due to either optical depth or saturation, both of which would preferentiallyeffect the stronger 292.00 A transition. It is likely that optical thickness of the292.00 A transition can be discounted, as the resonance lines in other Na-like ionshave been found to be optically thin under solar plasma conditions, in particularthose of Fe XVI (Keenan et al., 1994a). This element is a factor of �18 moreabundant than Ni in the solar corona (Meyer, 1985), and hence should be moresusceptible to optical depth effects. We therefore conclude that saturation of the320.56 and the 292.00 A lines (particularly the latter) on the photographic filmis probably the main cause of the observed discrepancies between theory andobservation.

As noted in Section 2, the theoretical Ni XVIII ratios depend principally on theelectron impact excitation rates from the 3s 2S ground level. Hence the limitedagreement found between theory and observation for R1 and R2 in the solar flarespectra provides some experimental support for the accuracy of the adopted atomicdata.

sola7089.tex; 13/03/1997; 9:15; v.7; p.5


Thomas and Neupert (1994) have identified an emission feature at 236.335 Ain an active region spectrum, obtained by the Solar EUV Rocket Telescope andSpectrograph (SERTS), as the Ni XVIII 3p 2P3=2 � 3d 2D3=2 transition. However,their intensity ratio for this feature relative to the 320.56 A line, R4 = 1:63� 0:73,is at least a factor of 70 larger than the theoretical value at log Tmax = 6:5 fromSection 2 (R4 = 0:0129). Hence, we believe that the line in the SERTS data at236.335 A is not due to Ni XVIII but is probably the Ar XIII 2s22p2 3P0�2s2p3 3D1

line, as originally suggested by Dere (1978). This identification is also supportedby S082A flare observations, where the intensity of this line is consistent with thatof the other Ar XIII features in the spectra (Keenan et al., 1993).

Finally, we note that the ratios in Figure 1 are quite sensitive to changes in theelectron temperature, with R1 and R2 varying by�60% between logTe = 5:8 and7.2. Hence in principle they may be useful as Te-diagnostics for the solar corona.However, for reliable temperatures to be derived, the observed ratios would needto be determined to a much higher degree of accuracy than the �40% achievedwith the S082A instrument (see Section 3). In the future accurate measurements forNi XVIII should be possible using the Coronal Diagnostic Spectrometer on boardSOHO (Harrison et al., 1995).

Acknowledgements

VJF acknowledges financial support from PPARC. This work was supported bythe Royal Society and NATO travel grant CRG.930722.

References

Arnaud, M. and Rothenflug, R.: 1985, Astron. Astrophys. Suppl. 60, 425.Brueckner, G. E.: 1976, Phil. Trans. Roy. Soc. London A281, 443.Burke, P. G. and Robb, W. D.: 1975, Adv. Atom. Molec. Phys. 11, 143.Corliss, C. and Sugar, J.: 1981, J. Phys. Chem. Ref. Data 10, 197.Dere, K. P.: 1978, Astrophys. J. 221, 1062.Dere, K. P. and Mango, S. A.: 1979, NRL Internal Report.Dere, K. P., Mason, H. E., Widing, K. G., and Bhatia, A. K.: 1979, Astrophys. J. Suppl. 40, 341.Dufton, P. L.: 1977, Comp. Phys. Comm. 13, 25.Dufton, P. L., Berrington, K. A., Burke, P. G., and Kingston, A. E.: 1978, Astron. Astrophys. 62, 111.Dufton, P. L., Kingston, A. E., Doyle, J. G., and Widing, K. G.: 1983, Monthly Notices Roy. Astron.

Soc. 205, 81.Dufton, P. L., Kingston, A. E., and Widing, K. G.: 1990, Astrophys. J. 353, 323.Elwert, G.: 1952, Z. Naturforsch. 7A, 432.Feldman, U. and Doschek, G. A.: 1977, J. Opt. Soc. Am. 67, 726.Flower, D. R. and Nussbaumer, H.: 1975, Astron. Astrophys. 42, 265.Foster, V. J., Keenan, F. P., and Reid, R. H. G.: 1996, Astron. Astrophys. 308, 1009.Godefroid, M., Magnusson, C. E., Zetterberg, P. O., and Joelsson, I.: 1985, Phys. Scripta 32, 125.Harrison, R. A. et al.: 1995, Solar Phys. 162, 233.Hiei, E. and Widing, K. G.: 1979, Solar Phys. 61, 407.

sola7089.tex; 13/03/1997; 9:15; v.7; p.6


Keenan, F. P., Kingston, A. E., Dufton, P. L., Doyle, J. G., and Widing, K. G.: 1984, Solar Phys. 94,91.

Keenan, F. P., McCann, S. M., Kingston, A. E., Barnsley, R., Dunn, J., and Peacock, N. J.: 1991,Phys. Rev. A44, 3831.

Keenan, F. P., Conlon, E. S., Foster, V. J., Aggarwal, K. M., and Widing, K. G.: 1993, Solar Phys.145, 291.

Keenan, F. P., Conlon, E. S., Foster, V. J., Tayal, S. S., and Widing, K. G.: 1994a, Astrophys. J. 432,809.

Keenan, F. P., Conlon, E. S., Warren, G. A., Pradhan, A. K., Zhang, H. L., Sampson, D. H., andWiding, K. G.: 1994b, Astrophys. J. 421, 851.

Meyer, J.-P.: 1985, Astrophys. J. Suppl. 57, 173.Mohan, M., Sharma, R., and Hibbert, A.: 1996, Phys. Scripta 54, 352.Parker, I. G., Patterson, N. P., Mango, S. A., and Tousey, R.: 1976, NRL Report 3410.Sampson, D. H., Zhang, H. L., and Fontes, C. J.: 1990, Atomic Data Nucl. Data Tables 44, 209.Sandlin, G. D., Bartoe, J.-D. F., Tousey, R., and Van Hoosier, M. E.: 1986, Astrophys. J. Suppl. 61,

801.Seaton, M. J.: 1964, Monthly Notices Roy. Astron. Soc. 127, 191.Thomas, R. J. and Neupert, W. M.: 1994, Astrophys. J. Suppl. 91, 461.Tousey, R., Bartoe, J.-D. F., Brueckner, G. E., and Purcell, J. D.: 1977, Appl. Optics 16, 870.Wang, J.-S., Griem, H. R., Hess, R., and Rowan, W. L.: 1988, Phys. Rev. A38, 4761.Widing, K. G. and Cook, J. W.: 1987, Astrophys. J. 320, 913.Widing, K. G. and Hiei, E.: 1984, Astrophys. J. 281, 426.Widing, K. G. and Spicer, D. S.: 1980, Astrophys. J. 242, 1243.Widing, K. G., Feldman, U., and Bhatia, A. K.: 1986, Astrophys. J. 308, 982.

sola7089.tex; 13/03/1997; 9:15; v.7; p.7

emission lines of ni xviii in the solar euv spectrum

Documents