7 Conclusion

In this paper we have analysed new optical integral-field spectroscopy of the PN SuWt 2 to study detailed ionized gas properties, and to infer the properties of the unobserved hot ionizing source located in the centre of the nebula. The spatially resolved emission-line maps in the light of $[$N II$]$ $\lambda$6584 have described the kinematic structure of the ring. The previous kinematic model (Jones et al., 2010) allowed us to estimate the nebula's age and large-scale kinematics in the Galaxy. An empirical analysis of the emission line spectrum led to our initial determination of the ionization structure of the nebula. The plasma diagnostics revealed as expected that the inner region is hotter and more excited than the outer regions of the nebula, and is less dense. The ionic abundances of He, N, O, Ne, S and Ar were derived based on the empirical methods and adopted mean electron temperatures estimated from the observed $[$O III$]$ emission lines and electron densities from the observed $[$S II$]$ emission lines.

We constructed photoionization models for the ring and interior of SuWt 2. This model consisted of a higher density torus (the ring) surrounding a low-density oblate spheroid (the interior disc). We assumed a homogeneous abundance distribution consisting of eight abundant elements. The initial aim was to find a model that could reproduce the flux intensities, thermal balance structure and ionization structure as derived from by the observations. We incorporated NLTE model atmospheres to model the ionizing flux of the central star. Using a hydrogen-rich model atmosphere, we first fitted all the observed line fluxes, but the time-scale of the evolutionary track was not consistent with the nebula's age. Subsequently, we decided to use hydrogen-deficient stellar atmospheres implying a VLTP (born-again scenario), and longer time-scales were likely to be in better agreement with the dynamical age of the nebula. Although the results obtained by the two models of SuWt 2 are in broad agreement with the observations, each model has slightly different chemical abundances and very different stellar parameters. We found a fairly good fit to a hydrogen-deficient central star with a mass of $\sim 0.64 {\rm M}_{\bigodot}$ with an initial (model) mass of $\sim 3 {\rm M}_{\bigodot}$. The evolutionary track of Blöcker (1995) implies that this central star has a post-AGB age of about 25000 yr. Interestingly, our kinematic analysis (based on $v_{\rm exp}$ from Jones et al., 2010) implies a nebular true age of about 23400-26300 yr.

Table 6 lists two best-fitting photoionization models obtained for SuWt 2. The hydrogen-rich model atmosphere (Model 1) has a normal evolutionary path and yields a progenitor mass of $3{\rm M}_{\bigodot}$, a dynamical age of 7,500 yr and nebular ${\rm N}/{\rm O}= 0.939$ (by number). The PG 1159 model atmosphere (Model 2) is the most probable solution, which can be explained by a VLTP phase or born-again scenario: VLTP $\rightarrow$ [WCL] $\rightarrow$ [WCE] $\rightarrow$ [WC]-PG1159 $\rightarrow$PG1159 (Werner & Herwig, 2006; Herwig, 2001; Blöcker, 2001; Miller Bertolami & Althaus, 2006). The PG1159 model yields ${\rm N}/{\rm O}=0.816$ and a stellar temperature of $T_{\rm eff}=160$ kK corresponding to the progenitor mass of $3{\rm M}_{\bigodot}$ and much longer evolutionary time-scale. The VLTP can be characterized as the helium-burning model, but this cannot purely explain the fast stellar winds ( $V_{\infty}=2000$kms$^{-1}$) of typical [WCE] stars. It is possible that an external mechanism such as the tidal force of a companion and mass transfer to an accretion disc, or the strong stellar magnetic field of a companion can trigger (late) thermal pulses during post-AGB evolution.

The abundance pattern of SuWt 2 is representative of a nitrogen-rich PN, which is normally considered to be the product of a relatively massive progenitor star (Becker & Iben, 1980; Kingsburgh & Barlow, 1994). Recent work suggests that HBB, which enhances the helium and nitrogen, and decreases oxygen and carbon, occurs only for initial masses of $\geq$5 ${\rm M}_{\bigodot}$ ($Z=0.02$; Karakas et al., 2009; Karakas & Lattanzio, 2007); hence, the nitrogen enrichment seen in the nebula appears to result from an additional mixing process active in stars down to a mass of 3 ${\rm M}_{\bigodot}$. Additional physical processes such as rotation increase the mass-loss rate (Paxton et al., 2013) and nitrogen abundance at the stellar surface (end of the core H- and He-burning phases; Ekström et al., 2012). The mass-loss via RLOF in a binary (or triple) system can produce a helium-rich outer layer (Benvenuto & De Vito, 2005; Chen & Han, 2002), which significantly affects other elements at the surface.