Download [OIII ] Electron Temperatures in Planetary Nebulae

J. Astrophys. Astr. (1989) 10, 147–150
[OIII ] Electron Temperatures in Planetary Nebulae
F. P. Keenan
The Department of Pure and Applied Physics, The Queen’s University
of Belfast, Belfast BT7 INN, N. Ireland
K.M.Aggarwal
The Department of Physics and Astrophysics, The University of
Delhi, Delhi 110007
Received 1988 October 3; accepted 1989 February 8
Abstract. Relative level populations for Ο III, derived using electron
impact excitation rates calculated with the R-matrix code, are used
to deduce the electron-temperature-sensitive emission-line ratio
R = I(2s22p2 lD–2s22p2 1 S)/I(2s2 2p2 3 P l,2 – 2s22p2 1D) = I(4363 Å)/I(4959
+ 5007 Å) for a range of Te = (7500–20000 K) applicable to planetary
nebulae. Electron temperatures deduced from the observed values of R in
several planetary nebulae are in excellent agreement with those determined
from T e-sensitive line ratios in other species, including C III]/C II, [ΝII] and
[Ar III], which provides support for the accuracy of the atomic data
adopted in the level population calculations.
Key
words:
emission-line diagnostics—electron temperature—planetary
nebulae
1. Introduction
Emission lines arising from transitions among the 2s22p2 3P, 1D and.1S levels of Ο III
are frequently observed in the visible (Aller & Keyes 1987), infrared (Dinerstein,
Lester & Werner 1985) and ultraviolet (Clegg et al. 1987) spectra of planetary nebulae.
However of all these lines probably the most important are 3 P1,2 –1D ) at 4959 Å and
5007 Å, and 1D-1S at 4363 Å, as they allow the electron temperature to be inferred
from the I(4363 Å)/ I(4959+5007 Å) intensity ratio (see, for example, Czyzak, Keyes &
Aller 1986). Under normal planetary nebulae conditions, where Ne<105 cm –3
(Barlow 1987), the 4363, 4959 and 5007 Å transitions are in the ‘coronal approximation.’ (Elwert 1952), so that their intensities depend almost exclusively on the
magnitude of the electron impact excitation rates for the lines. Hence to calculate the
I(4363Å)/ I(4959 + 5007Å) ratio reliably, it is important that accurate data are
employed for this atomic process.
Currently, the most accurate electron excitation rates available for Ο III are
probably those of Aggarwal (1983, 1985), which have been calculated with the Rmatrix code (Burke & Robb 1975; Berrington et al. 1978). In this paper we employ
these atomic data to derive the I(4363 Å)/I(4959 + 5007 Å) ratio and use it as a Te
diagnostic for several planetary nebulae whose electron temperatures have been
independently determined from line ratios in other species. This should allow the
validity of the Aggarwal excitation rates to be properly tested.
148
F. P. Keenan & Κ. Μ. Aggarwal
2. Theoretical ratios
Level populations for a wide range of electron densities and temperatures have been
published for O III by Keenan & Aggarwal (1987), who employed the electron impact
excitation rates of Aggarwal (1983,1985) in their calculations. Briefly, the lowest nine
LS states (2s22p2 3P, 1D, 1S; 2s2p3 5 S, 37), 3P, lD, 3S and 1P) were included in the model
ion, leading to fifteen levels when the splitting in the triplet states was taken into
account. Only collisional excitation and de-excitation by electrons and spontaneous
radiative de-excitation processes were considered, and the plasma was assumed to be
optically thin. Further details may be found in Keenan & Aggarwal (1987).
As noted by, for example, Keenan, Widing & McCann (1989), level populations may
be used to derive emission line ratios R through the expression:
(1)
where λij , λmn and I(λij), I(λmn) are the wavelengths and intensities (in energy units) of
the lines, respectively, Nj and Nn are the upper level populations of the relevant
transitions and Aji and Anm are the Einstein A-coefficients. In Fig. 1 we use the Keenan
& Aggarwal (1987) data to plot the emission line ratio R = I (1 D 1 S )/ I(3 P l,21 D )
= I(4363 Å)/I(4959 + 5007 Å) as a function of electron temperature at a density of Ne
= 103 cm–3, although we note that it is density insensitive for N e<10 5 cm–3.
Our calculations of R will lead to electron temperature estimates approximately
200 Κ higher than those that would be derived using the atomic data summarised by
Mendoza (1984) for the five energetically lowest levels in O III. This discrepancy is
partly due to the fact that in the present paper theoretical line strengths have been
derived with a fifteen level model ion, as opposed to the five level one necessary when
Figure 1. The theoretical [O III] emission line ratio R=I (2s22p2 1D–2s2 2p2 1S)/
I(2s22p2 3P l,2–2s22p2 1 D )=J(4363 Å)//(4959 + 5007 Å), where I is in energy units, plotted as a
function of electron temperature at an electron density of Ne = 103 cm–3.
[Ο III ]electron temperatures in planetary nebulae
149
Table 1. Observed values of the [O III] emission line ratio R=I(4363 Å)/ I(4959 Å+
5007 Å) in several planetary nebulae and the derived electron temperatures Te(R), plus the
mean temperatures Te deduced from the Te-sensitive line ratios in other ions (see text for
details).
using the Mendoza compilation. In addition, the electron impact excitation rate
calculations of Aggarwal (1983,1985) are an improvement over the results of Baluja,
Burke & Kingston (1980,1981) referenced by Mendoza, as discussed in detail by
Keenan & Aggarwal (1987).
3. Results and discussion
Many authors have determined the R ratios in planetary nebulae (see, for example,
Aller & Keyes 1987), although there are relatively few for which electron temperatures
have been independently estimated from a number of Te-sensitive ratios in other
species. In Table 1 we summarise R ratios for several planetaries, along with the
sources of the observational data. Also listed in the table are the electron temperatures
(Te(R)) derived from R and the mean values (Te) deduced from line ratios in ions that
have similar spatial distributions to O III, namely I(C III] 1909 Å)/I(C II 4267 Å)
discussed by Kaler (1986a), I(5755 Å)/ I(6548+6583 Å) in [N II] (Kaler 1986b) and
I(7136+7751 Å)/ I(5192Å) in [Ar III] (Keenan, Johnson & Kingston 1988). These
references also contain the sources of the observational data used in the electron
temperature determinations.
An inspection of Table 1 shows that the [O III] electron temperatures are inexcellent
agreement with the mean values determined from several independent Te diagnostics.
Discrepancies in the two sets of results are typically 1400 K, and only exceed 2000 Κ in
one case, namely that of NGC 6302. Εven in this instance we note that an increase in
the observed R ratio for NGC 6302 of only 0.1 dex will completely remove the
disagreement between Te(R) and Te. These results provide observational support for
both the accuracy of the atomic data adopted in the [O III] line ratio calculations, and
the techniques used to derive the theoretical values of R.
Acknowledgements
We would like to thank Professors P. G. Burke FRS and H. B. Gilbody for their
continued interest in this work. FPK is grateful to the United Kingdom Science and
Engineering Research Council for financial support.
150
F. P. Keenan & Κ. Μ. Aggarwal
References
Aggarwal, Κ. Μ. 1983, Astrophys. J. Suppl. 52, 387.
Aggarwal, Κ. Μ. 1985, Astron. Astrophys., 146, 149.
Aller, L. H., Czyzak, S. J. 1978, Proc. Nat. Acad. Sci. USA, 75, 1.
Aller, L. H., Czyzak, S. J. 1983, Astrophys. J. Suppl., 51, 211.
Aller, L. H., Keyes, C. D. 1987, Astrophys. J. Suppl, 65, 403.
Aller, L. H., Keyes, C. D. 1988, Proc. Nat. Acad. Sci. USA, 85, 2417.
Aller, L. H., Keyes, C. D., Czyzak, S. J. 1985, Astrophys. J., 296, 492.
Baluja, K. L., Burke, P. G., Kingston, A. E. 1980, J. Phys. B, 13, 829.
Baluja, K. L., Burke, P. G., Kingston, A. E. 1981, J. Phys. B, 14, 119.
Barker, T. 1986, Astrophys. J., 308, 314.
Barlow, Μ. J. 1987, Mon. Not. R. Astr. Soc, 227, 161.
Berrington, Κ. Α., Burke, P. G., Le Dourneuf, M., Robb, W. D., Taylor, K. T., Lan, V. K. 1978,
Comp. Phys. Comm., 14, 367.
Burke, P. G., Robb, W. D. 1975, Adv. At. Mol Phys., 11, 143.
Clegg, R. E. S., Harrington, J. P., Barlow, Μ. J., Walsh, J. R. 1987, Astrophys. J., 314, 551.
Czyzak, S. J., Keyes, C. D., Aller, L. H. 1986, Astrophys. J. Suppl, 61, 159.
Dinerstein, H. L., Lester, D. F., Werner, M. W. 1985, Astrophys. J., 291, 561.
Elwert, G. 1952, Z. Naturforsch, 7A, 432.
Kaler, J. B. 1986a, Astrophys. J., 308, 337.
Kaler, J. B. 1986b, Astrophys. J., 308, 322.
Keenan, F. P., Aggarwal, K. M. 1987, Astrophys. J., 319, 403.
Keenan, F. P., Johnson, C. T., Kingston, A. E. 1988, Astr. Astrophys., 202, 253.
Keenan, F. P., Widing, K. G., McCann, S. M. 1989, Astrophys. J., in press.
Mendoza, C. 1984, in IAU Symp. 103: Planetary Nebulae, Ed. D. R. Flower, D. Reidel,
Dordrecht, p. 143.
Shields, G. Α., Aller, L. Η., Keyes, C. D., Czyzak, S. J. 1981, Astrophys. J., 248, 569.