Free Access
Issue
A&A
Volume 612, April 2018
Article Number A62
Number of page(s) 15
Section Stellar atmospheres
DOI https://doi.org/10.1051/0004-6361/201732401
Published online 25 April 2018

© ESO 2018

1 Introduction

The hot white dwarf RE 0503−289 is unique because a large number of trans-iron elements were discovered in its photosphere (Werner et al. 2012b). Abundance analyses of 14 species with atomic numbers in the range Z = 30–56, i.e., zinc through barium, revealed high quantities of up to about 104 times the solar values (Rauch et al. 2017b, and references therein).

RE 0503−289 is a non-DAwhite dwarf (WD) with a helium-dominated atmosphere (spectral type hot DO, effective temperature Teff = 70 000 ± 2000 K, surface gravity logg = 7.5 ± 0.1; Rauch et al. 2016b). We concluded that the high heavy-metal abundances are probably caused by radiative levitation (Rauch et al. 2016a). If true, then we should find the same phenomenon among other hot DOs with similar temperatureand gravity. Therefore we performed ultraviolet (UV) spectroscopy of three other WDs with the Cosmic Origins Spectrograph (COS) aboard the Hubble Space Telescope (HST). Two of these WDs (PG 0109+111 and WD 0111+002, alias HS 0111+0012) are also of the hot DO spectral type. The third (PG 1707+427) was drawn from the PG 1159-type class, which are hot, non-DA (pre-) WDs with large amounts of atmospheric carbon and oxygen. Their surface chemistry is the result of a late He-shell flash that dredged up the intershell matter located between the H and He burning shells of the previous Asymptotic Giant Branch (AGB) evolutionary phase (Werner & Herwig 2006). PG 1159 stars without remaining hydrogen evolve into DO WDs as soon as gravitational settling starts to remove these species from the atmosphere.The choice of a PG 1159-type target was motivated by the fact that the trans-iron element abundances are not yet affected by diffusion and, hence, should reflect the abundances in the intershell layer created by the s-process during the AGB phase.

This paper is structured as follows. First, we introduce our program stars in Sect. 2, and proceed with a description of our new and archival observations (Sect. 3). We then present our spectral analysis to determine the basic atmospheric parameters Teff, log g, and metal abundances (Sect. 4). We discuss our results and conclude in Sect. 5.

2 The program stars

For our new observations we chose two hot DOs and a PG1159 star, which are bright enough to obtain UV spectra with a good signal-to-noise ratio enabling the identification of even weak metal lines. The location of the program stars in the gTeff diagram(already with our improved parameters for PG 0109+111) together with other hydrogen-deficient stars is shown in Fig. 1.

thumbnail Fig. 1

Location of our three program stars and the comparison star RE 0503−289 (red symbols) together with related objects (Hügelmeyer et al. 2006; Kepler et al. 2016; Reindl et al. 2014a; Werner & Herwig 2006) in the Teffg plane. The single triangle symbol denotes a helium-dominated object of the O(He) spectral class (Reindl et al. 2014b). Evolutionary tracks for H-deficient WDs (Althaus et al. 2009) labeled with the respective masses in M are plotted for comparison. Four variants of the PG 1159–DO transition limit (labeled A–D, Unglaub & Bues 2000) are indicated.

2.1 WD 0111+002 and PG 0109+111 (DO WDs)

The non-local thermodynamic equilibrium (NLTE) analysis of optical spectroscopy yielded Teff = 65 000 ± 5000 K, log g = 7.8 ± 0.3 for WD 0111+002 and Teff = 110 000 ± 10 000 K, log g = 8.0 ± 0.3 for PG 0109+111. A carbon abundance of C=0.003 in both stars was derived (Dreizler & Werner 1996); all abundances given in this paper are mass fractions. Dreizler (1999) used HST/GHRS spectra to assess metal abundances in WD 0111+002. With the exception of carbon lines, only a few weak Ni V lines were detected and an upper limit of Ni = 1.4 ×10−5 was established. Archival FUSE spectra of both stars remained hitherto not analyzed.

2.2 PG 1707+427 (PG1159 star)

This star is a well-studied object (Werner et al. 2015, and references therein). It has Teff = 85 000 ± 5000 K and log g = 7.5 ± 0.5, and is dominated by helium and carbon in roughly equal amounts. In the work cited, we assessed the abundances of metals up to the iron group using far-UV spectra taken with the Far Ultraviolet Spectroscopic Explorer (FUSE) and HST spectra of relatively low resolution taken with the Goddard High Resolution Spectrograph (GHRS). Our new HST/COS spectroscopy aimed at spectra of better quality to detect weaker lines, particularly of heavy metals.

3 Observations

In Table 1, we list the observation log of our new HST/COS data together with other archival UV HST and FUSE spectra used in our study. The datasets were retrieved from the Barbara A. Mikulski Archive for Space Telescopes (MAST). According to the COS Instrument Science Report 2009-01(v1)1, the line spread function (LSF) of the COS spectra deviates from a Gaussian shape. We therefore convolve all synthetic spectra with the correct LSF.

We used optical observations to constrain the surface gravity from the ionized helium lines. Optical spectra for WD 0111+002 and PG 1707+427 were obtained from the Sloan Digital Sky Survey (SDSS; observation IDs 1237663784742354959 and 1237665571981623391, respectively). We also used the spectra of PG 0109+111 from Dreizler & Werner (1996) and PG 1707+427 from Werner et al. (1991).

Table 1

Observation log of our three program stars.

3.1 Radial velocities

To shift theobservations to rest wavelength, we determined the stellar radial velocities vrad from the HST/COS spectra. To measure the observed wavelengths by Gaussian fits, we employed IRAF2 . The average values are vrad = 44.8 ± 5.4 km/s, vrad = 37.2 ± 4.7 km/s, and vrad = −33.8 ± 5.9 km/s for WD 0111+002 (from 11 lines), PG 0109+111 (11 lines), and PG 1707+427 (32 lines), respectively. Considering error limits, our value for WD 0111+002 is marginally in agreement with wavelength measurements of several lines in HST/GHRS spectra performed by Dreizler (1999).

3.2 Interstellar neutral hydrogen and reddening

We used theLy β line profile in the FUSE spectra to measure the interstellar neutral hydrogen column density NH (Fig. A.1). The results are for WD 0111+002, for PG 0109+111, and for PG 1707+427. Dreizler & Heber (1998) used the Ly α line in the HST/GHRS spectrum of PG 1707+427 and found 1.5 × 1020 cm−2, that is consistent with our value within its error limit.

We converted NH into reddening EB-V using the empirical relationship derived by Groenewegen & Lamers (1989): log (NHEB-V) = 21.58 ± 0.10. We obtained for WD 0111+002, for PG 0109+111, and for PG 1707+427. In Fig. A.2, we show our final models attenuated with these reddening values together with the observed continuum shape. The agreement is very good. To further improve the fits, marginally lower reddening values (within error limits) for WD 0111+002 and PG 1707+427 are required. Our result for PG 1707+427 agrees with the value of 0.02 derived by Dreizler & Heber (1998) from the continuum slope of the GHRS spectrum.

3.3 Unidentified lines

There are many unidentified photospheric lines in the HST/COS spectra of our program stars. Particularly conspicuous are features that are seen in both our spectrum of PG 0109+111 and in a spectrum of the comparison star RE 0503−289, taken with the Space Telescope Imaging Spectrograph (STIS) aboard HST and presented by Hoyer et al. (2017); see Fig. A.5. These features potentially stem from trans-iron elements and we therefore list their wavelength positions in Table A.1.

4 Spectral analysis

4.1 Model atmospheres and atomic data

To calculatechemically homogeneous, NLTE model atmospheres, we used the Tübingen model-atmosphere package (TMAP3 ; Werner et al. 2003, 2012a). It assumes plane-parallel geometry and radiative and hydrostatic equilibrium. Atomic data were compiled from the Tübingen Model-Atom Database (TMAD; Rauch & Deetjen 2003) that has been constructed as a part of the Tübingen contribution to the German Astrophysical Virtual Observatory (GAVO4 ). A summaryof the model atoms for the included light elements up to atomic number Z = 16, As and Sn are givenin Table 2.

For ironand nickel, we used a statistical approach (employing the IrOnIc tool; Rauch & Deetjen 2003) with seven superlevels per ion linked by superlines, together with an opacity sampling method. Ionization stages Fe III–VI (plus Fe VII for the PG 1707+427 models) andNi III–VI augmented by a single, ground-level stage were considered per species using the line lists of Kurucz (Kurucz 1991, 2009, 2011).

Motivated by the discovery of trans-iron elements in the DO RE 0503−289 (Werner et al. 2012b), respective transition probabilities have been calculated (for references see Table A.2) and are provided by the Tübingen Oscillator Strength Service (TOSS), which is a part of the GAVO project. For our analysis, these species were treated with the same statistical method as Fe and Ni (Rauch et al. 2015b, except for As and Sn, for which classical model atoms as for the other species were used), considering lines of Zn IV–V, Ga IV–VI, Ge IV–VI, AsIV–VII, Se IV–VI, Kr III–VII, Sr III–VII, Zr III–VII, Mo III–VII, Sn III–VI, Te IV–VI, I IV–VI, Xe III–VII, and Ba V–VII.

The model atmospheres for all three program stars include He, C, N, O, and Ne. The neon abundance cannot be determined, so we assumed the following values: Ne = 0.01 for PG 1707+427, which is a value that is typical for PG 1159 stars (Werner et al. 2004), and rather small values for WD 0111+002 (2.0 × 10−5) and PG 0109+111 (1.5 × 10−6). For the two hottest program stars (PG 0109+111 and PG 1707+427), we also included the following species into the model atmosphere calculations: F, Si, P, S, Zn, Ga, and Ge. To save computing time, smaller versions of the model atoms for all metals up to sulfur were used for the model atmosphere calculations. Subsequently, the large versions (whose numbers of levels and lines are listed in Table 2) were used to compute improved level occupation numbers while keeping fixed the model-atmosphere structure. All other species heavier than Ge were considered as trace elements, i.e., their backreaction on the model structure was neglected.

Table 2

Number of NLTE levels and lines in our model ions.

4.2 Effective temperature and surface gravity

To measure Teff, we evaluated the C III/C IV ionization equilibrium using the multiplets of C III at 1166, 1296, 1175 Å plus a singlet at 1247 Å, and the multiplets of C IV at 1108, 1169, 1198, 1230 Å (Figs. 2, A.3, and A.4). For the determination of g we used lines of the He II Pickering series (Fig. 3). Various other ionization equilibria (N III/N IV/N V, O III/O IV/O V/O VI, F V/F VI , S V /S VI, Fe V/Fe VI/Fe VII, Zn IV/Zn V Ge IV/Ge V, Ga IV/Ga V; not all balances available in all stars) confirm the results for Teff and log g found by this procedure.

For WD 0111+002, we found Teff = 58 000 ± 5000 K and log g = 7.7 ± 0.3, in agreement with previous results obtained from optical spectra alone (65 000/7.8; Dreizler & Werner 1996). In the case of PG 0109+111, we derived Teff = 70 000 ± 5000 K and log g = 8.0 ± 0.3. In comparison to previous results (110 000/8.0; Dreizler & Werner 1996), the temperature is significantly lower. In the work cited, it was noted that indeed the weak, observed He I lines are not compatible with the high-temperature model that displays no neutral helium lines at all. For PG 1707+427 our results (Teff = 85 000 ± 5000 K, log g = 7.5 ± 0.3) confirm previous analyses of UV and optical data (Werner et al. 2015, and references therein).

thumbnail Fig. 2

C III and C IV lines in the HST/COS spectrum of WD 0111+002 (black) compared with three log g = 7.7 models with different Teff: 53 000 K (blue dashed), 58 000 K (red), and 63 000 K (green dashed). Red wavelength labels indicate forbidden C IV line components (Werner et al. 2016). Also indicated are Zn lines near the C III multiplet at 1175 Å. The vertical bar indicates 20% of the continuum flux.

thumbnail Fig. 3

Observed spectra (black) around two lines of the He II Pickering series compared with synthetic spectra (red). From top to bottom: WD 0111+002 with three models (Teff = 58 000 K and log g = 7.6, 7.7, 7.8), PG 0109+111 with model 70 000/7.5, PG 1707+427 with model 85 000/7.5. The observations of PG 0109+111 and PG 1707+427 were smoothed with a low-pass filter (Savitzky & Golay 1964).

4.3 Element abundances

Element abundances were derived by detailed line-profile fits. Upper limits were established by reducing the abundances such that computed spectral lines become undetectable regarding the signal-to-noise ratio. We describe the procedure for the different species and display some representative fits. Results are given in Table 3 and Fig. 8. Typical errors are estimated to 0.3 dex. A comparison of the complete HST/COS spectra with the final models appears in Fig. A.5.

Table 3

Final adopted atmospheric parameters of the three program stars and the comparison star RE 0503−289 (Cols. 2–5), trans-iron abundances from literature for two cool DOs (HD 149499B and HZ 21) (Cols. 6 and 7); and solar abundances (last column).

thumbnail Fig. 4

N V resonance doublet in the program stars (black) compared to our final models (red).

thumbnail Fig. 5

Like Fig. 4, but for O III and O IV lines.

thumbnail Fig. 6

Synthetic line profiles for F, Si, P, S, Fe, and Ni in our final models (red) compared to observations (black) of our three program stars.

thumbnail Fig. 7

Like Fig. 6, but for Ga, Ge, As, Se, Sr, Sn, Te, and I.

thumbnail Fig. 8

Photospheric abundances of our three program stars and the comparison star RE 0503−289. [X] denotes log(abundance∕solar abundance) of species X. The dashed horizontal line indicates solar abundances.

Carbon, nitrogen, and oxygen

For carbon, we used the same lines as for the Teff determination (Figs. 2, A.3, and A.4). For nitrogen, the N V resonance doublet is the primary indicator (Fig. 4), and for oxygen the O III λλ1149.6, 1150.9, 1153.8 Å and O IV λλ1338.6, 1343.0, 1343.5 Å lines (Fig. 5).

Fluorine, silicon, phosphorus, and sulfur

For these species, we proceeded as described in detail for the analysis of the FUSE spectrum of PG 1707+427 (Werner et al. 2015). The following lines were used: F V λλ1087.8, 1088.4 Å, F VI λ1139.5 Å, Si IV λλ1066.6, 1122.5, 1128.3 Å, P V λλ1000.4, 1118.0, 1128.0 Å, S V λλ1128.7, 1128.8 Å, S VI λλ933.4, 944.5, 1117.8 Å. The fits to most of these lines are presented in Fig. 6.

Iron and nickel

Iron could not be detected in PG 0109+111, but a few weak Fe VI lines are detected in WD 0111+002. In the COS spectra of PG 1707+427, we see many lines from Fe VI and few from Fe VII. The identification of iron was uncertain in the FUSE spectra (Werner et al. 2015). We detect many nickel lines (Ni V) in the COS spectra of WD 0111+002 and PG 0109+111. In PG 1707+427, the presence of Ni VI lines in FUSE spectra is uncertain (Werner et al. 2015), but in the COS spectra we detect two lines at 1124.19 and 1154.60 Å (Fig. 6, Tables A.3A.5).

Zinc

Many Zn V lines are present in the COS data of our program stars; these lines are most prominent in WD 0111+002, followed by PG 0109+111, and are rather weak in PG 1707+427. Among the strongest features is a Zn V blend at 1177 Å adjacent to the C III λ 1175 Å multiplet (Figs. 2, A.3, and A.4). WD 0111+002 also shows Zn IV lines, for example, at 1349.88 Å.

Gallium and germanium

No lines are detectable in PG 1707+427, but we see a number of Ga IV and Ga V as well as Ge IV and Ge V lines in WD 0111+002 and PG 0109+111 (Fig. 7).

As, Se, Sr Sn, Te, and I

We used the As V λλ987.7/1029.48 Å resonance doublet in the FUSE spectra to determine the As abundance in WD 0111+002 and upper limits for the other two stars. In the case of PG 1707+427, ISM lines are located at the wavelength position of the As V doublet (Fig. A.1 in Werner et al. 2015), such that we can derive an upper limit only. Se V λλ1151.0, 1227.5 Å can be seen in both DOs but not in PG 1707+427. Lines from Sr, Sn, Te, and I areonly seen in PG 0109+111 (Sr V λ1154.9 Å, Sn IV λ1314.5 Å, Te VI λ1313.9 Å, I VI λ1153.3 Å) but not in the other two stars. For line fits, see Fig. 7.

In PG 0109+111, the fitted Sn IV λ1314.5 Å line is rather weak.A number of detected Sn V lines would be better suited for a more precise analysis, but oscillator strengths are unavailable. Sn V lines are not detected in PG 1707+427, but they would give a much tighter upper limit for the Sn abundance than that derived from the absence of the Sn IV line.

The tellurium abundance in PG 0109+111 is extremely high (5.6 dex oversolar). Its mass fraction of 6.2 × 10−3 means that it is the most abundant metal in the atmosphere. The abundance was derived from the Te VI λ1313.9 Å line (see Fig. 7). In fact, there are two much stronger Te VI lines, namely the resonance doublet at 951.0 and 1071.4 Å, but they are located in rather noisy portions of the FUSE and HST/COS spectra. Nevertheless, these lines can be detected in the observations (see the 1071.4 Å component in Fig. A.5) as very strong and broad features. They are among the strongest metal line features rivaling the prominent C III and C IV lines.

Kr, Zr, Mo, Xe, and Ba

No lines of these species could be identified in our program stars.

5 Summary and conclusions

We have analyzed new UV spectra taken with HST/COS of one PG 1159 star and two hot DO WDs. The primary aim of this study was the identification and abundance determination of trans-iron group elements to reach a conclusion about the effects of s-process and element diffusion. To this end, we computed line-blanketed NLTE model atmospheres and began with an assessment of the effective temperatures and surface gravities. For both DOs, the evaluation of metal ionization balances resulted in improved Teff determinations yielding lower values than found in previous analyses of optical spectra. In particular, Teff of PG 0109+111 is 70 000 K instead of 110 000 K.

PG 0109+111 and WD 0111+002 (DO WDs)

For the two DOs, we measured the trace-element abundances and compare these to the DO RE 0503−289, a hitherto unique DO in which a total of 14 trans-iron group elements (i.e., atomic number Z > 28) have been identified, with extreme overabundances (about 2–4.5 dex) relative to the solar values (Table 3 and Fig. 8). Both investigated DOs also exhibit trans-iron species, but there are fewer of these species (five and eight in WD 0111+002 and PG 0109+111, respectively). In both DOs we detect Zn, Ga, Ge, and Se, plus the heavier species Sr, Sn, Te, and I only in PG 0109+111, and As only in WD 0111+002. Generally, their abundances are (within about 1 dex) similar to RE 0503−289 with the exception of Te. This element is almost 6 dex oversolar, significantly exceeding the enrichment in RE 0503−289.

The general tendency that RE 0503−289 (Teff= 70 000 K, log g = 7.5) has the highest number of detected trans-iron species can be explained by the fact that radiative acceleration is probably more effective compared to PG 0109+111, which has a higher gravity (70 000/8.0), and compared to WD 0111+002 (with the lowest number of detected trans-iron species), which is significantly cooler and has a slightly higher surface gravity (58 000/7.7).

Chayer et al. (2005) detected trans-iron elements in two cool DO WDs, namely HZ 21 (Teff = 53 000 K, log g = 7.8; Dreizler & Werner 1996) and HD 149499B (Teff = 49 500 K, log g = 7.97; Napiwotzki et al. 1995). Their abundances are also given in Table 3. The numbers listed (mass fractions) were computed from the number ratios reported in Chayer et al. (2005) (averaged over results from several lines, if applicable), accounting for the fact that HD 149499B is a DOA WD with 6% hydrogen (by mass). Generally, the trans-iron element abundances are lower than in the hot DOs, probably owing to weaker levitation at lower temperatures.

As to the lighter elements, these are also generally less abundant in the two DOs investigated here compared to RE 0503−289, with the exception of P, that is 0.8 dex higher in WD 0111+002 than in RE 0503−289. A noteworthy commonality of all three DOs is the low Fe/Ni ratio of ≤1.6, in contrast to the solar ratio Fe/Ni = 18.

PG 1707+427 (PG 1159 star)

Our results for Teff and log g agree with those in Werner et al. (2015). The same holds, within error limits, for the element abundances. For the first time, however, weidentify nickel lines and find a solar Ni abundance. The most significant result from our new HST/COS spectroscopy concerns the trans-iron elements. While only high upper abundance limits could be derived previously, we now determined the Zn abundance and, for most other species, were able to reduce the upper abundance limits significantly compared to our previous work involving FUSE data alone. We find that Zn is solar and Ge is at most 0.8 dex oversolar, suggesting that there is no general overabundance of trans-iron elements. The upper limits for other trans-iron elements are still high, however, between about 2.8 and 4.5 dex oversolar. In comparison to the DOs, this confirms that radiative levitation does not affect the element abundances in the atmosphere of PG 1707+427. This is a plausible result for a PG 1159 star because, as outlined in the introduction, the star is located before the wind limit so that diffusion cannot affect the element abundances. In addition, from the helium and light metal abundances we concluded to see intershell matter composition (Werner et al. 2015).

It is expected from stellar evolution models that, as discussed in detail in Werner et al. (2015) based on the models of a 2 M star after the 30th thermal pulse (Gallino, priv. comm.; Karakas et al. 2007) s-process elements should be overabundant. But our results do not allow a quantitative comparison because the predicted enhancement of Zn and Ge to < 0.4 dex oversolar values is near the error margin of our analysis. Much stronger enhancements are predicted by stellar models for other trans-iron elements, but they are still below our upper limits derived. For example, we could reduce the upper detection limit for germanium by 1.6 dex down to Ge < 1.3 × 10−6, but the predicted value is Ge < 5.9 × 10−7, i.e., just barely below our detection threshold. The strongest enhancement among the investigated s-process elements is predicted for barium, namely Ba = 1.2 × 10−5. Again, although we were able to reduce the detection threshold by 1.8 dex, the current limit of Ba < 2.3 × 10−5 is slightly smaller. Further progress can only be achieved with UV spectra with even better signal-to-noise ratio.

Appendix A: Additional figures and tables

Table A.1

Wavelengths (in Å) of unidentified photospheric lines.

Table A.2

Ions of trans-iron elements with recently calculated oscillator strengths.

Table A.3

Fe V lines identified in the HST/COS spectrum of PG 1707+427.

Table A.4

Fe VI lines identified in the HST/COS spectrum of PG 1707+427.

Table A.5

Nickel lines identified in the HST/COS spectra of WD 0111+002 and PG 0109+111.

thumbnail Fig. A.1

Synthetic spectra around Ly β from our final models (black, dashed graphs) for the program stars compared with the FUSE observations (black). Red graphs: interstellar absorption with neutral hydrogen column densities as given in Sect. 3.2 applied. The adjacent dashed graphs indicate the error limits.

thumbnail Fig. A.2

Spectral energy distribution of our program stars compared to the final, reddened models (red graphs). Models of WD 0111+002, and PG 0109+111 were normalized to the 2MASS H magnitude and that of PG 1707+427 to the SDSSz magnitude. Magnitudes from GALEX, SDSS, and 2MASS are plotted in blue with their error bars.

thumbnail Fig. A.3

Like Fig. 2, but for PG 0109+111.

thumbnail Fig. A.4

Like Fig. 2, but for PG 1707+427.

thumbnail Fig. A.5

Complete HST/COS spectra of our three program stars (three bottom spectra in each panel) together with the HST/STIS spectrum of the comparison DO star RE 0503−289 (top spectrum). Overplotted (red graphs) are our final models and the final model for RE 0503−289 from Rauch et al. (2017b). Blue dashed graphs indicate ISM lines.

thumbnail Fig. A.5

continued.

thumbnail Fig. A.5

continued.

Acknowledgements

DH was supported by the German Aerospace Center (DLR, grant 50 OR 1501). The authors acknowledge support by the High Performance and Cloud Computing Group at the Zentrum für Datenverarbeitung of the University of Tübingen, the state of Baden-Württemberg through bwHPC, and the German Research Foundation (DFG) through grant no INST 37/935-1 FUGG. The TIRO (http://astro-uni-tuebingen.de/TIRO) TMAD (http://astro-uni-tuebingen.de/TMAD) and TOSS (http://astro-uni-tuebingen.de/TOSS) services used for this paper were constructed as part of the activities of the German Astrophysical Virtual Observatory. Some of the data presented in this paper were obtained from the Mikulski Archive for Space Telescopes (MAST). STScI is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS5-26555. Support for MAST for non-HST data is provided by the NASA Office of Space Science via grant NNX09AF08G and by other grants and contracts. This research has made use of NASA’s Astrophysics Data System and the SIMBAD database, operated at CDS, Strasbourg, France.

References

  1. Althaus, L. G., Panei, J. A., Miller Bertolami, M. M., et al. 2009, ApJ, 704, 1605 [NASA ADS] [CrossRef] [Google Scholar]
  2. Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, ARA&A, 47, 481 [NASA ADS] [CrossRef] [Google Scholar]
  3. Chayer, P., Vennes, S., Dupuis, J., & Kruk, J. W. 2005, ApJ, 630, L169 [NASA ADS] [CrossRef] [Google Scholar]
  4. Dreizler, S. 1999, A&A, 352, 632 [NASA ADS] [Google Scholar]
  5. Dreizler, S., & Heber, U. 1998, A&A, 334, 618 [NASA ADS] [Google Scholar]
  6. Dreizler, S., & Werner, K. 1996, A&A, 314, 217 [NASA ADS] [Google Scholar]
  7. Grevesse, N., Scott, P., Asplund, M., & Sauval, A. J. 2015, A&A, 573, A27 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  8. Groenewegen, M. A. T., & Lamers, H. J. G. L. M. 1989, A&AS, 79, 359 [NASA ADS] [Google Scholar]
  9. Hoyer, D., Rauch, T., Werner, K., Kruk, J. W., & Quinet, P. 2017, A&A, 598, A135 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  10. Hügelmeyer, S. D., Dreizler, S., Homeier, D., et al. 2006, A&A, 454, 617 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  11. Karakas, A. I., Lugaro, M., & Gallino, R. 2007, ApJ, 656, L73 [NASA ADS] [CrossRef] [Google Scholar]
  12. Kepler, S. O., Pelisoli, I., Koester, D., et al. 2016, MNRAS, 455, 3413 [NASA ADS] [CrossRef] [Google Scholar]
  13. Kurucz, R. L. 1991, in 341: Stellar Atmospheres - Beyond Classical Models eds. L. Crivellari, I. Hubeny, & D. G. Hummer NATO ASIC Proc., 441 [Google Scholar]
  14. Kurucz, R. L. 2009, in American Institute of Physics Conference Series, eds. I. Hubeny, J. M. Stone, K. MacGregor, & K. Werner, 1171, 43–51 [Google Scholar]
  15. Kurucz, R. L. 2011, Can. J. Phys., 89, 417 [NASA ADS] [CrossRef] [Google Scholar]
  16. Napiwotzki, R., Hurwitz, M., Jordan, S., et al. 1995, A&A, 300, L5 [NASA ADS] [Google Scholar]
  17. Rauch, T., & Deetjen, J. L. 2003, in Stellar Atmosphere Modeling, eds. I. Hubeny, D. Mihalas, & K. Werner, ASP Conf. Ser., 288, 103 [NASA ADS] [Google Scholar]
  18. Rauch, T., Werner, K., Biémont, É., Quinet, P., & Kruk, J. W. 2012, A&A, 546, A55 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  19. Rauch, T., Werner, K., Quinet, P., & Kruk, J. W. 2014a, A&A, 564, A41 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  20. Rauch, T., Werner, K., Quinet, P., & Kruk, J. W. 2014b, A&A, 566, A10 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  21. Rauch, T., Hoyer, D., Quinet, P., Gallardo, M., & Raineri, M. 2015a, A&A, 577, A88 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  22. Rauch, T., Werner, K., Quinet, P., & Kruk, J. W. 2015b, A&A, 577, A6 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  23. Rauch, T., Quinet, P., Hoyer, D., et al. 2016a, A&A, 587, A39 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  24. Rauch, T., Quinet, P., Hoyer, D., et al. 2016b, A&A, 590, A128 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  25. Rauch, T., Gamrath, S., Quinet, P., et al. 2017a, A&A, 599, A142 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  26. Rauch, T., Quinet, P., Knörzer, M., et al. 2017b, A&A, 606, A105 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  27. Reindl, N., Rauch, T., Werner, K., et al. 2014a, A&A, 572, A117 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  28. Reindl, N., Rauch, T., Werner, K., Kruk, J. W., & Todt, H. 2014b, A&A, 566, A116 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  29. Savitzky, A., & Golay, M. J. E. 1964, Anal. Chem., 36, 1627 [NASA ADS] [CrossRef] [Google Scholar]
  30. Scott, P., Asplund, M., Grevesse, N., Bergemann, M., & Sauval, A. J. 2015a, A&A, 573, A26 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  31. Scott, P., Grevesse, N., Asplund, M., et al. 2015b, A&A, 573, A25 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  32. Unglaub, K., & Bues, I. 2000, A&A, 359, 1042 [NASA ADS] [Google Scholar]
  33. Werner, K., & Herwig, F. 2006, PASP, 118, 183 [NASA ADS] [CrossRef] [Google Scholar]
  34. Werner, K., Heber, U., & Hunger, K. 1991, A&A, 244, 437 [NASA ADS] [Google Scholar]
  35. Werner, K., Deetjen, J. L., Dreizler, S., et al. 2003, in Stellar Atmosphere Modeling, eds. I. Hubeny, D. Mihalas, & K. Werner, ASP Conf. Ser., 288, 31 [NASA ADS] [Google Scholar]
  36. Werner, K., Rauch, T., Reiff, E., Kruk, J. W., & Napiwotzki, R. 2004, A&A, 427, 685 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  37. Werner, K., Dreizler, S., & Rauch, T. 2012a, TMAP: Tübingen NLTE Model-Atmosphere Package, Astrophysics Source Code Library [Google Scholar]
  38. Werner, K., Rauch, T., Ringat, E., & Kruk, J. W. 2012b, ApJ, 753, L7 [NASA ADS] [CrossRef] [Google Scholar]
  39. Werner, K., Rauch, T., & Kruk, J. W. 2015, A&A, 582, A94 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  40. Werner, K., Rauch, T., Hoyer, D., & Quinet, P. 2016, ApJ, 827, L4 [NASA ADS] [CrossRef] [Google Scholar]

2

IRAF is distributed by the National Optical Astronomy Observatory, which is operated by the Associated Universities for Research in Astronomy, Inc., under cooperative agreement with the National Science Foundation.

All Tables

Table 1

Observation log of our three program stars.

Table 2

Number of NLTE levels and lines in our model ions.

Table 3

Final adopted atmospheric parameters of the three program stars and the comparison star RE 0503−289 (Cols. 2–5), trans-iron abundances from literature for two cool DOs (HD 149499B and HZ 21) (Cols. 6 and 7); and solar abundances (last column).

Table A.1

Wavelengths (in Å) of unidentified photospheric lines.

Table A.2

Ions of trans-iron elements with recently calculated oscillator strengths.

Table A.3

Fe V lines identified in the HST/COS spectrum of PG 1707+427.

Table A.4

Fe VI lines identified in the HST/COS spectrum of PG 1707+427.

Table A.5

Nickel lines identified in the HST/COS spectra of WD 0111+002 and PG 0109+111.

All Figures

thumbnail Fig. 1

Location of our three program stars and the comparison star RE 0503−289 (red symbols) together with related objects (Hügelmeyer et al. 2006; Kepler et al. 2016; Reindl et al. 2014a; Werner & Herwig 2006) in the Teffg plane. The single triangle symbol denotes a helium-dominated object of the O(He) spectral class (Reindl et al. 2014b). Evolutionary tracks for H-deficient WDs (Althaus et al. 2009) labeled with the respective masses in M are plotted for comparison. Four variants of the PG 1159–DO transition limit (labeled A–D, Unglaub & Bues 2000) are indicated.

In the text
thumbnail Fig. 2

C III and C IV lines in the HST/COS spectrum of WD 0111+002 (black) compared with three log g = 7.7 models with different Teff: 53 000 K (blue dashed), 58 000 K (red), and 63 000 K (green dashed). Red wavelength labels indicate forbidden C IV line components (Werner et al. 2016). Also indicated are Zn lines near the C III multiplet at 1175 Å. The vertical bar indicates 20% of the continuum flux.

In the text
thumbnail Fig. 3

Observed spectra (black) around two lines of the He II Pickering series compared with synthetic spectra (red). From top to bottom: WD 0111+002 with three models (Teff = 58 000 K and log g = 7.6, 7.7, 7.8), PG 0109+111 with model 70 000/7.5, PG 1707+427 with model 85 000/7.5. The observations of PG 0109+111 and PG 1707+427 were smoothed with a low-pass filter (Savitzky & Golay 1964).

In the text
thumbnail Fig. 4

N V resonance doublet in the program stars (black) compared to our final models (red).

In the text
thumbnail Fig. 5

Like Fig. 4, but for O III and O IV lines.

In the text
thumbnail Fig. 6

Synthetic line profiles for F, Si, P, S, Fe, and Ni in our final models (red) compared to observations (black) of our three program stars.

In the text
thumbnail Fig. 7

Like Fig. 6, but for Ga, Ge, As, Se, Sr, Sn, Te, and I.

In the text
thumbnail Fig. 8

Photospheric abundances of our three program stars and the comparison star RE 0503−289. [X] denotes log(abundance∕solar abundance) of species X. The dashed horizontal line indicates solar abundances.

In the text
thumbnail Fig. A.1

Synthetic spectra around Ly β from our final models (black, dashed graphs) for the program stars compared with the FUSE observations (black). Red graphs: interstellar absorption with neutral hydrogen column densities as given in Sect. 3.2 applied. The adjacent dashed graphs indicate the error limits.

In the text
thumbnail Fig. A.2

Spectral energy distribution of our program stars compared to the final, reddened models (red graphs). Models of WD 0111+002, and PG 0109+111 were normalized to the 2MASS H magnitude and that of PG 1707+427 to the SDSSz magnitude. Magnitudes from GALEX, SDSS, and 2MASS are plotted in blue with their error bars.

In the text
thumbnail Fig. A.3

Like Fig. 2, but for PG 0109+111.

In the text
thumbnail Fig. A.4

Like Fig. 2, but for PG 1707+427.

In the text
thumbnail Fig. A.5

Complete HST/COS spectra of our three program stars (three bottom spectra in each panel) together with the HST/STIS spectrum of the comparison DO star RE 0503−289 (top spectrum). Overplotted (red graphs) are our final models and the final model for RE 0503−289 from Rauch et al. (2017b). Blue dashed graphs indicate ISM lines.

In the text
thumbnail Fig. A.5

continued.

In the text
thumbnail Fig. A.5

continued.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.