8 The central star

As mentioned in Sect. 2, the exciting star of NGC 6818 is the northern component of a visual binary system. The WFPC2 frames taken through the broad-band filter F555W ( $\lambda_{\rm c}= 5407$ Å, $\rm bandwidth= 1236$ Å) give m_V = 17.06($\pm0.05$) for the central star and $m_{V} = 17.73(\pm0.05$) for the southern red companion, where the unknown star colors are the main source of inaccuracy. Please note that for D = 1.7 kpc we have (M_V)₀(red companion) $\simeq$ +5.75, as expected of a late spectral type (G8 to K0) Main Sequence star: a further, although weak, sign in favour of the adopted distance.

According to Tylenda et al. (1993) and Feibelman (1994), the exciting star of NGC 6818 presents a weak emission line spectrum in both the UV and the optical regions. In order to derive the H I and He II Zanstra temperatures, $T_{\rm Z}$H I and $T_{\rm Z}$He II respectively, we have obtained the total H$\beta$ and $\lambda$4686 Å nebular fluxes from the overall line profile at each PA, assuming a circular symmetry of the spectral image. We find log $F({\rm H}\beta)_{\rm obs}=-10.49(\pm0.03$) mW m^-2 and $F(\lambda 4686~\AA)/F({\rm H}\beta)=0.65(\pm0.04$), in excellent agreement with the values reported by Aller & Czyzak (1979), Kohoutek & Martin (1981), Webster (1983), Acker et al. (1991), and Hyung et al. (1999).

The resulting Zanstra temperatures are log( $T_{\rm Z}{\rm H~I})= 5.20(\pm0.04$) and log( $T_{\rm Z}{\rm He~II})=5.24(\pm0.04$). We note in passing that $T_{\rm Z}{\rm H~I}\simeq T_{\rm Z}$He II; this is quite surprising for a high excitation, optically thin (almost thin in some directions) PN. Previous determinations (Harman & Seaton 1966; Martin 1981; Gathier & Pottasch 1988; Mal'kov 1997) span the range 4.82 to 5.20 for log($T_{\rm Z}$H I), and 5.03 to 5.30 for log($T_{\rm Z}$He II). Moreover Pottasch & Preite-Martinez (1983) and Preite-Martinez et al. (1991) find log T_*=4.93 and 5.02-5.05, respectively, from the energy balance of the nebula.

The stellar luminosities are log  $L_*/L_\odot(T_{\rm Z}{\rm H~I})=3.0 (\pm0.1$) and $\log L_*/L_\odot(T_{\rm Z}{\rm He~II})= 3.1(\pm0.1$) (using the bolometric corrections by Schönberner 1981, and adopting D = 1.7 kpc).

NGC 6818 results to be a rather young PN surrounding a very hot star of relatively low luminosity. This suggests that:
- the central star is a hydrogen-burning post-AGB star (the helium-burning ones evolve too slowly);
- the stellar mass, M_*, is larger than the average value ($\simeq$0.60 $M_\odot$) of the PNe nuclei (Blöcker 1995 and references therein).
The detailed comparison with the theoretical evolutionary tracks by Schönberner (1981, 1983), Iben (1984), Wood & Faulkner (1986), Blöcker & Schönberner (1990), Vassiliadis & Wood (1994) and Blöcker (1995) confirms that no solutions are possible for the He-burning post-AGB stars (they are too slow), whereas the H-burning nuclei give $M_*=0.625{-}0.65~M_\odot$.

Moreover, the position of the central star of NGC 6818 in the log L-log T diagram coincides with the beginning of the luminosity decline at the end of the shell nuclear burning. In this evolutionary phase all models present a large luminosity gradient with a deep minimum at $\log L_*/L_\odot\simeq 3.0$, i.e. the value obtained for the central star of NGC 6818. For example the 0.625 $M_\odot$ H-burning nucleus by Blöcker (1995) (showing an astonishing resemblance with our star) has $\log L_*/L_\odot=3.387$ at t=3457 yr, and $\log L_*/L_\odot=2.846$ at t=3585 yr, i.e. it drops in luminosity by a factor of 3.5 in 128 yr.

And indeed, the central star of NGC 6818 is clearly visible in the earliest photographic reproduction of the nebula (Pease 1917; plate taken on July 12, 1912 with the Mount Wilson 60-inch reflector); the rough photometric sequence defined by the field stars gives $m_{B}=15.2(\pm 0.4$). More stellar magnitude estimates reported in the literature are: $m_{\rm pg}=14$ (Curtis 1918), $m_{\rm pg}=15$ (Berman 1937), $m_{\rm pv}= 15.6$ (Cudworth 1974), $m_{V}\simeq 14.9$ (Martin 1981), m_B> 15 (Shaw & Kaler 1985), m_B=16.97 and m_V=17.02 (Gathier & Pottasch 1988), and m_V=17.06 (this paper, referring to the epoch 1998-2000).

In addition, the first spectroscopic observation of NGC 6818 (star+nebula, Palmer 1903, referring to 1901) indicates "a pretty strong continuum spectrum extending out to $\lambda$372, with the following bright lines: H$\beta$, H$\gamma$, H$\delta$...''. Palmer observed a large sample of nebulae with the Crossley reflector+plate spectrograph, reporting that "the faintest continuum spectrum photographed was that of a fifteenth magnitude star''.

Figure 8: The interstellar absorption-distance relation for the field stars of NGC 6818. Three symbols are used, referring to stars at different angular separation from the nebula.

In spite of the photometric heterogeneity of these data, the suspicion of a "historical'' decline for the nucleus of NGC 6818 appears legitimate (and robust). Also note that the magnitudes of the star just before the luminosity drop were: $m_{V}\simeq m_{\rm pv}\simeq 14.5$ and $m_{B}\simeq m_{\rm pg}\simeq 14.0$ (for D=1.7 kpc and $c({\rm H}\beta)=0.37$).

In summary, all the evidence is that:
- the quite massive central star is rapidly fading,
- the thin-thick transition is hanging over the nebula.
Within this scenario the following caveat is in order: because of the luminosity drop of the fast evolving star, the ionization and thermal structure of NGC 6818 are out of equilibrium. The recombination processes prevail, being faster in the high ionized species (Tylenda 1986; Marten & Szczerba 1997); this qualitatively explains the observed, unexpected closeness of the H I and He II Zanstra temperatures: the nebula is still optically thin (almost thin in some directions), but H$\beta$ maintains a "longer memory'' than $\lambda$4686 Å of the past stellar luminosity. Although the large $N{\rm e}$ of the gas implies a short "time lag'' of the main nebula (a few dozen years), nevertheless both $T_{\rm Z}$He II and, even more, $T_{\rm Z}$H I are slightly over-valued. In other words: NGC 6818 is over-luminous with respect to the present UV stellar flux.

The overall picture of our nebula shows a remarkable series of analogies with NGC 6565. According to Paper IV, NGC 6565 is a young (2300 yr), optically thick ellipsoid embedded in a large cocoon of neutral, dusty gas. It is in a deep recombination phase (started about 400 yr ago), caused by the luminosity drop of the massive powering star ( $M_*\simeq 0.65~M_\odot$), which has almost reached the white dwarf locus ( $\log L_*/L_\odot\simeq 2.0$, $\log T_* \simeq 5.08$).

All this suggests the evolutive contiguity of the two PNe, NGC 6818 being a bit older, but also a bit less "mature'' than NGC 6565, i.e. the central star of the latter is more massive. No doubt NGC 6818 will develop in the near future most of the observational characteristics already present in NGC 6565 (and in the other recombining PNe contained in Tylenda 1986 and Corradi et al. 2000).

We have also carried out a search for other PNe in the peculiar evolutionary phase of NGC 6818 (e.g. at the very beginning of the thin-thick transition) adopting the following criteria: high temperature and quite low luminosity of the star, different [O III] and [N II] nebular morphology, high excitation class and relatively weak low ionization emissions. Among the dozen or so candidates we identified, the most promising are NGC 6326, NGC 6884, NGC 7354 and IC 4663.