© The Authors 2023

1. Introduction

Planetary nebulae (PNe) are among the most interesting objects in the Galaxy, not only for their complex morphologies, but also because they provide valuable information on the final stages of the stellar evolution. Low- to intermediate-mass stars end their lives with a phase of strong mass loss (Frew & Parker 2010). More precisely, observational data show that the mass range in which a star forms a planetary nebula is 0.8–5.0 M_⊙ (Weidmann et al. 2020). At the beginning of this phase, the star experiences thermal pulses on the asymptotic giant branch (AGB), where the stellar wind injects material into the interstellar medium, forming a circumstellar envelope. The ejected material is later compressed by the stellar remnant’s faster wind, and these compressed ejected layers become a planetary nebula (Kwitter et al. 2014).

Although the evolutionary paths (in the HR diagram) followed by the stars after this stage are not yet fully understood (e.g., Weidmann et al. 2020), we do know that they evolve towards higher effective temperatures until, finally, the nuclear energy sources are exhausted or quenched, and the star descends along a white dwarf (WD) cooling track. Nevertheless, it is also unclear whether all present WDs went through a PN phase. Nebula searches have been carried out around evolved stars, but with little success (e.g., Werner et al. 1997). We know that there are certainly some nebulae that are ionised by WDs; these are old nebulae (i.e. objects of large angular size, low surface brightness, and low electron density; Ahumada et al. 2019).

Understanding the evolution of a star implies, among other things, being able to reproduce with theoretical models the different changes that the star’s atmosphere undergoes in these few 20 000 years, the average lifetime of a PN. It is therefore crucial to increase the sample of WDs that are nuclei of PNe to characterise this population.

This is the second and last part of a project that was started in 2019 (Ahumada et al. 2019), whose goal was to double the known number of WDs that are also central stars of planetary nebulae (CSPNe).

The structure of the paper is as follows: In Sect. 2 we describe our sample and observational set-up; in Sect. 3 we present the spectral classification of the stars of our sample; in Sect. 4 we describe the stellar atmospheric models used to infer the atmospheric parameters of the stars; in Sect. 5 we present the colours and distances for the CSPNe derived from the Gaia DR3 survey; and in Sect. 6 we give the conclusions. Finally, in Appendix A we show the finding charts of the objects studied, and the spectral fits and Kiel diagrams in Appendix B.

2. Sample and observations

We want to identify white dwarfs (WDs) at the centre of PNe. To achieve this aim, we selected PNe with large angular size and low surface brightness (i.e. old PNe). All objects were extracted from the HASH database¹. The targets were previously identified as ‘blue’ (Parker et al. 2006; Miszalski et al. 2008). The coordinates of the nebulae reported in the catalogue match those of the corresponding stars. All PNe selected are classified as ‘true’ or ‘likely’ PNe, and all of them are poorly studied objects.

The targets in our sample were observed with the Gemini-South telescope in five programmes (Table 1). The Gemini-South Multi-Object Spectrograph (GMOS) was used with the grating of 600 groves mm⁻¹ blazed at 4610 Å (grating number G5323). In addition, we requested a 2 × 2 binning and 1.5 arcsec slits. The resulting spectra cover the wavelength range of 3500–6200 Å with a spectral resolution of 1 Å px⁻¹.

The observational strategy was to take at least three spectra per object. This allows us to improve the signal-to-noise ratio (S/N) and remove cosmic ray hits. However, in two objects (see Table 1) there were only two useful spectra; the remainder had poor S/N values. The stars were observed with an airmass lower than 1.5 and a slit orientation close enough to the parallactic angle. The seeing for each observation is indicated in Table 1. The raw data were processed with PyRAF². The reduction involves bias correction, overscan subtraction, flat-field correction, cosmic-ray subtraction, and sky subtraction.

3. Spectral classification

We were able to perform a spectral classification in nine objects (Table 2). However, we could not do so for the central star of PN G328.5+06.2, whose spectrum has the lowest S/N in our sample (Table 3). The full width at half maximum (FWHM) of the absorption lines, shown in the same table, suggests that all of the stars are evolved objects (see Table 3 in Ahumada et al. 2019). The spectra are presented in Figs. 1–4, and the individual descriptions follow.

Fig. 1. Normalised Gemini spectra of the DAO central stars of PNe in our sample. The Balmer lines Hβ, Hγ, and Hδ, together with the He II line at 4686 Å, are indicated. The interstellar D lines of Na I at 5890 and 5896 Å are not marked, but are clearly seen in all spectra.

Fig. 2. Normalised Gemini spectra of the WD(H) stars. The nebular emission lines of [O III] and Hβ in MPA J0704−2221 can be seen. These lines could not be removed.

Fig. 3. Normalised Gemini spectra of the PG 1159. The absorption of C IV at 4647–59 Å, typical of PG 1159 stars, can be seen. The interstellar absorption bands at 4428 Å, and the complexes at 5780 and 5890–6 Å are not indicated.

Fig. 4. Normalised Gemini spectra of the O(H) star. The absorption of N V at 4604–20 Å and the absence of He I, characteristics of early O(H) stars, can be seen. The interstellar absorption bands at 4428 Å, and the complexes at 5780 and 5890–6 Å are not indicated.

PN G015.5+02.8. No lines of He I or He II are seen in the spectrum (Fig. 2). Its S/N is one of the lowest among the spectra presented in this work, and therefore we cannot ensure that the object is in fact a DA (i.e. a WD showing only hydrogen absorption lines). Tentatively, we propose classifying this star as ‘WD(H)’. With this nomenclature, not previously used, we want to indicate that the star is a hydrogen-rich WD.

PN G023.8−06.2. The central star was classified as a possible WD by Gentile Fusillo et al. (2019). Our spectrum (Fig. 1) shows the wide absorption lines of the Balmer series, together with evident absorption of He II at 4686 Å. We classified this object as a DAO.

PN G231.1+03.9. The spectrum (Fig. 3) has a good S/N, and reveals lines of the He II and C IV that are not as wide as those of a WD. The FWHM of the line at 4861 Å is the same as that of the line at 4686 Å. This suggests that the line at 4861 Å is indeed He II. The absorption lines identified in the spectra are He II (3970, 4101, 4339, 4542, 4686, 4861, 5412 Å) and C IV (3933, 4441, 4554, 4647, 4659, 4786, 5019, 5471 Å). There is also a narrow emission of C IV at 5801–12 Å. In addition, there is an absorption line of O V at 5114 Å, which is the transition with the highest excitation energy. We classified this star as a PG 1159, according to the criteria of Werner (1992). Although the star can be classified within subtype A³ (the most numerous), the conspicuous emission of C IV at 5801–12 Å calls into question this subtype. Nevertheless, the definition of subtype A refers only to the C IV–He II line blend at 4686 Å. So, it is indeed A even if the C IV line at 5801–12 Å is in emission.

PN G234.3−07.2. This spectrum only displays two absorption lines (Fig. 2), that of He II at 4686 Å, and Hβ. We can only say that it is a hydrogen-rich WD (i.e. another WD(H)). In addition, the narrow nebular lines of Hβ and [O III] are clearly visible. It is unlikely that this is due to a poorly nebular subtraction component because the object has a large angular size. It is therefore possible that this nebular emission comes from a region very close to the central star.

PN G234.9−09.7. This spectrum displays clear and wide absorption lines of the Balmer series (Fig. 1). There is also absorption of He II at 4686 Å. We classified this star as DAO.

PN G237.3−08.4. Using Gaia photometry, two possible evolved states have been suggested for this CSPN. It is either an sdO star (Geier et al. 2019) or a WD (Gentile Fusillo et al. 2019). With our spectroscopic data, we can confirm that this CSPN is indeed a WD of subtype DAO (Fig. 1).

PN G247.5−04.7. This CSPN is relatively faint; Fesen et al. (1983) estimated photographic magnitudes m_r = 17.5 and m_b = 17.0. The spectrum (Fig. 3) displays absorption and emission lines. The emissions come from C IV at 5801–12 Å and O VI at 3811–34 Å. Unfortunately, the CCD gap covers the line of O VI at 5291 Å. Perhaps He II at 4686 Å is in emission as well. For the absorption lines, there is the He II series, as well as C IV at 4440, 4647, 4659, and 4786 Å. We classified this star as PG 1159 subtype ‘lgE’ following the Werner 1992 criteria.

PN G328.5+06.2. This spectrum has a low S/N, very likely because it was acquired under the less favourable conditions of this project. Moreover, the star is intrinsically faint (Table 4). Perhaps the line of He II at 4686 Å and Hγ are present. There is also an absorption, of unknown origin, at 5180 Å. We cannot confirm that we actually observed the central star.

PN G344.9+03.0. It is an object whose central star was identified with GEMINI images (Ahumada et al. 2019). The spectrum shows wide absorption lines of Hβ and He II at 4686 Å (Fig. 1). In addition, there is a possible absorption line at 5154 Å, whose origin is unknown. We classified this CSPN as a DAO.

PN G355.3+03.8. The rectification of this spectrum was a difficult task. Nevertheless, the spectrum has a good S/N and shows narrow lines of the Balmer and He II series (Fig. 4). The most relevant feature, however, is the absorption of N V at 4604–20 Å, which is typical of early O(H) stars. We classified this CSPN as O(H)3 III-V (Sota et al. 2011, 2014).

4. Spectral analysis of the central stars

4.1. PG 1159-type central stars

We used the Tübingen Model-Atmosphere Package (TMAP, Werner et al. 2003) to build grids of non-local thermodynamic equilibrium (NLTE) plane-parallel models in radiative and hydrostatic equilibrium. For the two PG 1159 stars we used models of the type introduced by Werner et al. (2014). In essence, they include H, He, C, and O. Since H is not detected in the observed spectra, we set the respective model H abundances to very low values. Best-fit models within this grid were identified to provide effective temperatures, surface gravities, and element abundances. The resulting fits are shown in Figs. B.1 and B.2, and the model parameters are listed in Table 2. Errors in atmospheric parameters were estimated by comparing synthetic spectra from our grid with the observations.

The positions of the two analysed PG 1159 stars in the Kiel (log g − T_eff) diagram are shown in Fig. B.5 together with all the other known objects in this class. Also depicted are the positions of two other helium-dominated classes, namely the O(He) stars and the hot DO white dwarfs. Both new PG 1159 stars are in the pre-WD stage of their evolution. PN G231.1+03.9 has T_eff = 120 000 K and is a relatively low-mass object (M = 0.53 M_⊙) approaching the maximum temperature of its evolution. PN G247.5−04.7 has a slightly higher mass (M = 0.57 M_⊙) and is significantly hotter (150 000 K). It also displays lines of highly ionised neon (not included in the model), as is usual in PG 1159 stars of lgE subtype (Werner et al. 2004, 2007). We see Ne VII 3644 Å in absorption and two Ne VIII lines in emission (at 4340 and 6068 Å). The abundances of He, C, and O of our new PG 1159 stars are typical for this spectral class (Werner & Herwig 2006). They do not exhibit nitrogen lines, indicating that they experienced a very late thermal pulse (VLTP) and not a late thermal pulse (LTP).

4.2. Hydrogen-rich central stars

The fitting procedure for the seven H-rich central stars was similar like for the PG1159 stars. We employed NLTE models composed of hydrogen and helium with different abundance ratios. For the DA white dwarfs, the He abundance fraction was set to 10⁻⁶. For the O(H) star, nitrogen line-formation calculations were performed to address the N abundance. Problems with the rectification of the spectra affect the higher-order Balmer lines of the two DA and four DAO white dwarfs, so the main emphasis for the line fitting was given to Hβ. For the DAO white dwarfs the helium abundances were derived from He II 4686 Å. The spectrum of the O(H) star is useful over a broader wavelength range. The N abundance of this object was obtained from a fit to the N V 4604/4620 Å doublet. The resulting spectral fits are shown in Figs. B.3 and B.4, and the respective model atmosphere parameters are given in Table 2. The positions of the analysed stars in the Kiel diagram are shown in Fig. B.6 together with other objects in the hydrogen-rich classes O(H), DA, and DAO.

The O(H) star PN G355.3+03.8 is a low-mass object with M = 0.54 M_⊙. It has a solar H/He ratio, and the nitrogen abundance is enriched to 1% by mass (about 14 times over solar). One of the two DA white dwarfs (PN G234.3−07.2) is close to the WD cooling track, and has a mass of 0.55 M_⊙. The other DA (PN G015.5+02.8) has a lower mass (0.49 M_⊙), but considering the error limits, it is still compatible with a post-AGB star. The same holds for the two of our four DAOs that also have a mass of 0.49 M_⊙ (PN G344.9+03.0 and PN G234.9−09.7). The third DAO (PN G023.8−06.2) has a mass of 0.53 M_⊙.

The remaining DAO star, PN G237.3−08.4, is an outstanding object. It cannot be clearly explained by post-AGB evolution; instead, we can identify it as a post-red-giant-branch (RGB) post-common envelope (CE) object with a mass of 0.41 M_⊙. The low surface gravity (and hence the low mass) derived from our spectroscopy is in agreement with the parallax distance measured by Gaia. The spectroscopic distance is obtained from the relation

$$\begin{array}{c}\hfill d[\mathrm{pc}]=7.11\times {10}^4\sqrt{H_v·M·{10}^{0.4V_0-logg}},\end{array}$$

where H_ν = 8.409 × 10⁻⁴ erg cm⁻² s⁻¹ Hz⁻¹ is the Eddington flux of the atmosphere model at 5400 Å, M is the stellar mass (in M_⊙), and V₀ = V − A_V is the dereddened visual magnitude V, with A_V the visual extinction. Using the values V = 17.634 (Wolf et al. 2018) and A_V = 0.5763 (Geier et al. 2019), we derived a distance $d=1.5_{-0.5}^{+0.8}$ kpc, in agreement with that obtained from the Gaia parallax (ϖ), d = 1/ϖ ± σ_ϖ/ϖ² = 1.9 ± 0.2 kpc (see Sect. 5).

5. Properties of the CSPNe derived from Gaia DR3 data

The survey Gaia DR3 (Gaia Collaboration 2016, 2023) provides the full astrometric solution (α, δ, parallax, and proper motion) for about 1.46 billion sources, down to a magnitude of G ≈ 21. It also gives the magnitudes in the G-, G_BP-, and G_RP-band for over 1.5 billion sources. In Table 4 the Gaia identifications, magnitudes, colours, and parallaxes are listed for our sample of CSPNe.

Table 4 also lists the star distances, estimated as follows. Luri et al. (2018) shows that the naive direct interpretation of the distance as the simple inverse of the parallax ϖ is only accurate when the relative error f = σ_ϖ/ϖ is at most 20%. Larger values of f require another, more careful analysis. They recommend tackling this problem as a question of inference, to be preferably handled with a full Bayesian approach. This method was treated in depth by Bailer-Jones (2015), and it involves the estimation of the posterior probability P(r|ϖ, σ_ϖ) over r, given the observables (ϖ, σ_ϖ):

$$\begin{array}{c}\hfill P(r|\varpi ,{\sigma }_{\varpi })=\frac{1}{Z}P(\varpi |r,{\sigma }_{\varpi })P(r).\end{array}$$(1)

Here P(ϖ|r, σ_ϖ) is the conditional probability of the observable parallax ϖ given r and σ_ϖ, P(r) is the prior probability, and Z is a normalisation constant. The estimate of the distance is then the mode of the pdf P(r|ϖ, σ_ϖ). For the measurement model used in the Gaia data processing, Bailer-Jones (2015) proposes

$$\begin{array}{c}\hfill P(\varpi |r,{\sigma }_{\varpi })=\frac{1}{\sqrt{2\pi }{\sigma }_{\varpi }}exp[-\frac{1}{2{\sigma }_{\varpi }^2}{(\varpi -\frac{1}{r})}^2],\end{array}$$(2)

which is a Gaussian distribution in ϖ but not in r. The prior expresses our knowledge of, or our assumptions about, the distance, independent of ϖ. Bailer-Jones (2015) and Astraatmadja & Bailer-Jones (2016a,b) discuss several priors, among them the exponentially decreasing space density

$$\begin{array}{c}\hfill P(r)=\{\begin{array}{ccc}\frac{1}{2L^3}{r}^2{e}^{-r/L}\hfill & \hfill & \mathrm{if}r>0,\hfill \\ \hfill \\ 0\hfill & \hfill & \mathrm{otherwise},\hfill \end{array}\end{array}$$(3)

with L the length scale. This is a simple and reasonable expression, and we used it to estimate the distances to our CSPNe, which are given with 90% confidence intervals, as recommended by Bailer-Jones (2015); we adopted the length L = 1.35 kpc, the value suggested in the cited works.

6. Discussion and conclusions

With the results of this paper plus those of Ahumada et al. (2019), we have increased by 26% the number of white dwarfs that are nuclei of planetary nebulae (Weidmann et al. 2020). On the other hand, we have identified two new PG 1159 stars, which brings the number of these objects to 67⁴. About one-third of them (24) are planetary nebula nuclei.

Moreover, we have identified a post-RGB central star; these stars are of great interest because they allow us to study CE evolution, one of the most poorly understood phases of close-binary evolution (e.g., Hall et al. 2013). However, only a few of them are known because they are difficult to identify. There are only four strong candidates for low-mass objects of this type, namely the central stars of ESO 330-9 (PN G331.0+08.4), Abell 46 (PN G055.4+16.0), HaWe 13 (PN G034.1−10.5), and Ou 5 (PN G086.9−03.4) (Hillwig et al. 2017; Jones et al. 2022, 2023). Their positions are also shown in the Kiel diagram (Fig. B.6). ESO 330-9 is regarded as one of the best post-RGB candidates, and its position in the Kiel diagram is identical to that of our DAO star. The study of the binary nature of PN G237.3−08.4 would therefore be of utmost importance. The optical spectrum does not exhibit emission lines from an irradiated cool companion. However, the bipolar shape of the PN (Miszalski et al. 2008) suggests that it was ejected during a previous CE phase (e.g., Boffin & Jones 2019).

In all cases it was possible to fit a model of stellar atmosphere to obtain the basic stellar parameters and the abundances. This allowed us to locate the objects on a Kiel diagram. Although the physical parameters of the H-poor stars are consistent with already known objects, in the case of the H-rich white dwarfs it is seen that our objects are hotter than those previously catalogued.

The mean mass of the central stars in our sample is 0.51 M_⊙. This is close to, but slightly lower than, the mean mass of white dwarfs (0.61 M_⊙, Kepler et al. 2016). In addition, the luminosities [L/L_⊙] that we get from the tracks in the Kiel diagrams are the following: PN G015.5+02.8 = 1.40 ± 0.28, PN G023.8−06.2 = 0.75 ± 0.80, PN G234.3−07.2 = 1.02 ± 1.42, PN G234.9−09.7 = 2.01 ± 0.49, PN G237.3−08.4 = $1.{36}_{-0.38}^{+0.68}$, PN G344.9+03.0 = 1.40 ± 0.28, PN G231.1+03.9 = $3.{46}_{-0.60}^{+0.63}$, PN G247.5−04.7 = $3.{42}_{-0.57}^{+0.61}$.

The post-AGB ages of the stars were estimated from the evolutionary tracks depicted in the Kiel diagrams. For the hydrogen-rich stars (except PN G237.3−08.4) with masses lower than the lowest-mass post-AGB track (0.534 M_⊙), we chose the stellar age at the point of this track that is located closest to the stars (PN G015.5+02.8 with 200 000 yr, PN G234.9−09.7 with 70 000 yr, and PN G344.9+03.0 with 200 000 yr). At face value, the ages are very large for a ‘usual’ PN. However, one should consider the following aspects. First, because of the error in the temperature and gravity determinations, the age estimates are rather uncertain. For example, the minimum ages of the oldest candidates PN G023.8−06.2 and PN G234.3−07.2 are about 150 000 and 44 000 yr, respectively. Second, the evolutionary rate of post-AGB stellar models is under debate. With new input physics, Miller Bertolami (2016) found that the models exhibit increased evolutionary speeds. Third, there are examples of PNe for which large ages were determined. For example, Napiwotzki (1999) found several objects with ages up to 660 000 yr.

Six objects in our sample are classified as likely PN (see Table 1). One possibility is that these nebulae are actually interstellar medium ionised by an evolved star. However, if the nebula has a bipolar morphology with a hot evolved star at its geometrical centre, it is a sign that the object is indeed a bona fide PN. In this sense, we propose that PN G023.8−06.2 and PN G237.3−08.4 are true planetary nebulae.

Acknowledgments

We thank the reviewer Dr. Albert Zijlstra whose very useful remarks helped us to improve this paper. KW wishes to thank Nicole Reindl (Potsdam) for useful discussions. Based on observations obtained at the Gemini Observatory and processed using the Gemini IRAF package. The Gemini Observatory is operated by the Association of Universities for Research in Astronomy, Inc., under a cooperative agreement with the NSF on behalf of the Gemini partnership: the National Science Foundation (United States), the National Research Council (Canada), CONICYT (Chile), Ministerio de Ciencia, Tecnología e Innovación Productiva (Argentina), and Ministério da Ciência, Tecnologia e Inovaao (Brazil). This research has made use of NASA’s Astrophysics Data System and the SIMBAD database, operated at CDS, Strasbourg, France. This work has made use of data from the European Space Agency (ESA) mission Gaia (https://www.cosmos.esa.int/gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement. This work was partially supported by grant PICT 2017-3301 awarded by Fondo para la Investigación Científica y Tecnológica (FonCyT).

References

Appendix A: Finding charts for the CSPNe

Fig. A.1. Finding charts, adapted from our r acquisition images, for the central stars of the planetary nebulae. The stars are indicated by arrows, and their coordinates are listed in Table 1.

Fig. A.2. Finding charts, adapted from our g acquisition images, for the central stars of the planetary nebulae. The stars are indicated by arrows and their coordinates are listed in Table 1.

Appendix B: Spectral fits and Kiel diagrams

Fig. B.1. Best-fit NLTE model (red) to the spectrum of the PG 1159-type central star of BMP J0739-1418 (blue). The significant lines due to principal ions in the model are identified, as are the location of calcium and sodium interstellar lines (i.s.), a diffuse interstellar band (DIB), and telluric absorption. The model atmosphere parameters are shown in Table 2.

Fig. B.2. Same as Fig. B.1, but for the PG 1159-type central star of FEGU 248-5.

Fig. B.3. Same as Fig. B.1, but for the four DAO central stars.

Fig. B.4. Same as Fig. B.1, but for the two DA central stars and the O(H) central star. The emission lines in MPA J0704−2221 are from the PN.

Fig. B.5. Positions and error bars of the two new PG 1159 stars in the Kiel diagram, together with known objects of this class plus O(He) stars and DO white dwarfs. The evolutionary tracks by Althaus et al. (2009) are labelled with the stellar mass in solar units.

Fig. B.6. Positions of seven new H-rich central stars (red symbols with error bars) in the Kiel diagram together with other H-rich objects. The blue symbols denote the strongest post-RGB, post-CE central star candidates known to date (Jones et al. 2023). H-rich central stars of PNe are shown as open circles, and ‘naked’ (i.e. with no PN) O(H)-type stars as filled circles. Evolutionary tracks for post-AGB remnants (solid lines) by Miller Bertolami (2016) and for post-RGB remnants (dashed lines) by Hall et al. (2013) are labelled with the stellar mass in solar units. This is an updated plot from Reindl et al. (2016) (see references in the caption of their Fig. 3), Jones et al. (2022) (see their Fig. 6), and Jones et al. (2023).

All Tables

All Figures

	Fig. 1. Normalised Gemini spectra of the DAO central stars of PNe in our sample. The Balmer lines Hβ, Hγ, and Hδ, together with the He II line at 4686 Å, are indicated. The interstellar D lines of Na I at 5890 and 5896 Å are not marked, but are clearly seen in all spectra.
In the text

	Fig. 2. Normalised Gemini spectra of the WD(H) stars. The nebular emission lines of [O III] and Hβ in MPA J0704−2221 can be seen. These lines could not be removed.
In the text

	Fig. 3. Normalised Gemini spectra of the PG 1159. The absorption of C IV at 4647–59 Å, typical of PG 1159 stars, can be seen. The interstellar absorption bands at 4428 Å, and the complexes at 5780 and 5890–6 Å are not indicated.
In the text

	Fig. 4. Normalised Gemini spectra of the O(H) star. The absorption of N V at 4604–20 Å and the absence of He I, characteristics of early O(H) stars, can be seen. The interstellar absorption bands at 4428 Å, and the complexes at 5780 and 5890–6 Å are not indicated.
In the text

	Fig. A.1. Finding charts, adapted from our r acquisition images, for the central stars of the planetary nebulae. The stars are indicated by arrows, and their coordinates are listed in Table 1.
In the text

	Fig. A.2. Finding charts, adapted from our g acquisition images, for the central stars of the planetary nebulae. The stars are indicated by arrows and their coordinates are listed in Table 1.
In the text

	Fig. B.1. Best-fit NLTE model (red) to the spectrum of the PG 1159-type central star of BMP J0739-1418 (blue). The significant lines due to principal ions in the model are identified, as are the location of calcium and sodium interstellar lines (i.s.), a diffuse interstellar band (DIB), and telluric absorption. The model atmosphere parameters are shown in Table 2.
In the text

	Fig. B.2. Same as Fig. B.1, but for the PG 1159-type central star of FEGU 248-5.
In the text

	Fig. B.3. Same as Fig. B.1, but for the four DAO central stars.
In the text

	Fig. B.4. Same as Fig. B.1, but for the two DA central stars and the O(H) central star. The emission lines in MPA J0704−2221 are from the PN.
In the text

	Fig. B.5. Positions and error bars of the two new PG 1159 stars in the Kiel diagram, together with known objects of this class plus O(He) stars and DO white dwarfs. The evolutionary tracks by Althaus et al. (2009) are labelled with the stellar mass in solar units.
In the text

Fig. B.6. Positions of seven new H-rich central stars (red symbols with error bars) in the Kiel diagram together with other H-rich objects. The blue symbols denote the strongest post-RGB, post-CE central star candidates known to date (Jones et al. 2023). H-rich central stars of PNe are shown as open circles, and ‘naked’ (i.e. with no PN) O(H)-type stars as filled circles. Evolutionary tracks for post-AGB remnants (solid lines) by Miller Bertolami (2016) and for post-RGB remnants (dashed lines) by Hall et al. (2013) are labelled with the stellar mass in solar units. This is an updated plot from Reindl et al. (2016) (see references in the caption of their Fig. 3), Jones et al. (2022) (see their Fig. 6), and Jones et al. (2023).

In the text