Astron. Astrophys. 342, 823-830 (1999)

1. Introduction

NGC 2346 is a much studied bipolar planetary nebula at a distance pc (Acker et al. 1992) ¹. At its center lies a binary system, formed by a main-sequence star of spectral type A5V, with mass M, temperature K and luminosity L (Méndez and Niemela 1981; Walsh 1983) and hot star, not detected in the visual, with T K (Méndez 1978), which excites the nebula. Its luminosity is very uncertain, as we will discuss in Sect. 4. Estimates in the literature give L L (Méndez 1978; Calvet and Peimbert 1983).

In the optical, the nebula has a butterfly shape (Balick 1987; Walsh et al. 1991), with well developed bipolar lobes and a bright torus which surrounds the central star. The temperature and density of the ionized gas in the torus have been estimated to be K and cm^-3, respectively (Liu et al. 1995; McKenna & Keenan 1996). The nebula contains a large amount of material in the form of molecular gas, as revealed by the CO observations of Knapp (1986), Huggins & Healy (1986), Healy & Huggins (1988). Bachiller et al. (1989) have mapped the entire nebula in the two CO lines J=1-0 and J=2-1; the morphology of the molecular gas follows very closely that of the ionized gas, showing a clumpy, inhomogeneous torus, tilted with respect to the line of sight by about 56^o, which is expanding outward. Scaled to our adopted distance pc, the torus has a radius of 0.05 pc, and mass 0.26 M, much larger than the mass of ionized gas (0.01 M, Walsh 1983). The radial velocity of the most intense CO condensations is of the order of 15-35 km s^-1, which results in a dynamical age of about 2500 yr (but see also Walsh et al. 1991).

NGC 2346 is a Type I nebula, originated by a massive progenitor (Calvet and Peimbert 1983). As many PNe of similar type, NGC 2346 is detected in the vibrationally excited lines of H₂ (Webster et al. 1988). Zuckerman & Gatley (1988) have mapped the nebula in the 1-0S(1) line at 2.12 µm using a single-beam 12 arcsec spectrometer with resolution 200. The morphology of the nebula in this line is again very similar to the morphology observed in the optical lines and in CO. This result was confirmed more recently by the images obtained in the same line with much better spatial resolution (about 1-2 arcsec) by Kastner et al. (1994) and Latter et al. (1995).

The excitation mechanism of the vibrationally excited H₂ lines in this, as in other PNe, is still uncertain. Zuckerman & Gatley (1988) discuss the possibility that they form in a shock driven by the fast wind emitted by the central star. Kastner et al. (1994) surveyed a sample of bipolar planetary nebulae (including NGC 2346); they conclude that the H₂ emission very likely originates in thermally excited (possibly shocked) molecular gas. Recently, Natta & Hollenbach (1998; hereafter NH98) have computed theoretical models of the evolution of PN shells and predicted, among others, the intensities of the most commonly observed H₂ vibrationally excited lines (namely, the 1-0S(1) at 2.12 µm and the 2-1S(1) at 2.25µm). They consider the emission of the photodissociation region (PDR) formed by the UV photons emitted by the central star impinging on the shell, including in the calculations time-dependent H₂ chemistry and the effects of the soft X-ray radiation emitted by the central star, which are important in sources like NGC 2346 where T K. NH98 compute also the emission of the shocked gas at the interface between the shell and the wind ejected by the central star in its previous red giant phase. They point out that both mechanisms (PDR and shocks) can produce lines of similar intensity, with reasonable values of the model parameters.

The PN properties that determine the intensity of the H₂ lines are very different in the two cases. As discussed in NH98, if the emission is produced in the warm, neutral PDR gas, the line intensity depends mostly on the stellar radiation field which reaches the shell and, to a lower degree, on the density of the neutral gas itself. If the emission is produced in the shocked gas, then the line intensity does not depend directly on the properties of the central star or of the PN shell, but only on the shock velocity and on the rate of mass-loss of the precursor red-giant. It is therefore clear that, before attributing any diagnostic capability to the H₂ lines, we need to understand which of the possible excitation mechanisms dominate the PN emission.

This paper is a first attempt to understand the H₂ emission of a well-studied PN in a quantitative way, i.e., by comparing the observations to detailed models of PDR and shock emission, such as those discussed in NH98. To this purpose, we have collected new near-IR broad and narrow-band images of NGC 2346 as well as K band spectra with resolution 1000. These observations are described in Sect. 2. The results are described in Sect. 3 and compared to the predictions of PDR and shock models in Sect. 4. A discussion of the results follows in Sect. 5; Sect. 6 summarizes the main conclusions of the paper.

© European Southern Observatory (ESO) 1999

Online publication: February 23, 1999
helpdesk.link@springer.de