An extremely hot white dwarf with a rapidly rotating K-type subgiant companion: UCAC2 46706450
Klaus Werner, Nicole Reindl, Lisa Löbling, Ingrid Pelisoli, Veronika Schaffenroth, Alberto Rebassa-Mansergas, Puji Irawati, Juanjuan Ren
aa r X i v : . [ a s t r o - ph . S R ] S e p Astronomy & Astrophysicsmanuscript no. aa © ESO 2020September 14, 2020
An extremely hot white dwarf witha rapidly rotating K-type subgiant companion: UCAC2 46706450
Klaus Werner , Nicole Reindl , Lisa Löbling , Ingrid Pelisoli , Veronika Scha ff enroth ,Alberto Rebassa-Mansergas , , Puji Irawati , and Juanjuan Ren Institut für Astronomie und Astrophysik, Kepler Center for Astro and Particle Physics, Eberhard Karls Universität, Sand 1, 72076Tübingen, Germanye-mail: [email protected] Institut für Physik und Astronomie, Universität Potsdam, Karl-Liebknecht-Straße 24 /
25, Germany Departament de Física, Universitat Politècnica de Catalunya, c / Esteve Terrades 5, E-08860 Castelldefels, Spain Institut d’Estudis Espacials de Catalunya, Ed. Nexus-201, c / Gran Capità 2-4, E-08034 Barcelona, Spain National Astronomical Research Institute of Thailand, Sirindhorn AstroPark, Donkaew, Mae Rim, Chiang Mai 50180, Thailand Key Laboratory of Space Astronomy and Technology, National Astronomical Observatories, Chinese Academy of Sciences, Beijing100101, P. R. China3 June 2020 / ABSTRACT
The subgiant UCAC2 46706450 is a late-type star with an ultraviolet (UV) excess. It was considered as a candidate to establish asample of stars of spectral type F, G, and K with white dwarf (WD) companions that could be used to test binary evolution models.To verify the WD nature of the companion, UV spectroscopy has previously been performed by other authors. Via a detailed model-atmosphere analysis, we show that the UV source is an extremely hot WD with an e ff ective temperature of T e ff =
105 000 ± M / M ⊙ = . ± .
02, radius of R / R ⊙ = . + . − . , and luminosity of L / L ⊙ = + − , meaning that the compact object isjust about to enter the WD cooling sequence. Investigating spectra of the cool star ( T e ff = ±
250 K), we found that it is a K-typesubgiant with M / M ⊙ = . − . R / R ⊙ = . + . − . , and L / L ⊙ = + − that is rapidly rotating with v sin( i ) =
81 km s − . Optical lightcurves reveal a period of two days and an o-band peak-to-peak amplitude of 0.06 mag. We suggest that it is caused by stellar rotationin connection with star spots. With the radius, we infer an extremely high rotational velocity of v rot = + − km s − , thus markingthe star as one of the most rapidly rotating subgiants known. This explains chromospheric activity observed by H α emission andemission-line cores in Ca ii H and K as well as NUV flux excess. From equal and constant radial velocities of the WD and the Ksubgiant as well as from a fit to the spectral energy distribution, we infer that they form a physical, wide (though unresolved) binarysystem. Both components exhibit similar metal abundances and show iron-group elements with slightly oversolar (up to 0.6 dex)abundance, meaning that atomic di ff usion in the WD atmosphere is not yet active due to a residual, weak radiation-driven wind.Kinematically and from its height above the Galactic plane, the system belongs to the Galactic thick disk, indicating that it is an oldsystem and that the initial masses of both stars were close to 1 M ⊙ . Key words. stars: individual: UCAC2 46706450, stars: atmospheres, stars: abundances, stars: AGB and post-AGB, (stars:) whitedwarfs, (stars:) starspots
1. Introduction
Stellar multiplicity is an omnipresent outcome of the star-formation process. About every other solar-type (mass M ≈ . − . M ⊙ ) star is found in the binary system (see Duchêne &Kraus 2013 for a review). At solar metallicity, about one quarterof these stars is found in close binaries (orbital period P < d,separation a <
10 AU), and the close-binary fraction is foundto increase strongly with decreasing metallicity and may be ashigh as 53% at a metallicity of [Fe / H] = − ff the main sequence, such close systems will eventually inter-act with each other by exchanging mass and angular momentum(Willems & Kolb 2004). This in turn influences their evolutionand can lead to a broad range of astrophysical phenomena thatare absent for the life of single stars. Systems that will undergoa common-envelope event will end up as very close binarieswith final orbital periods typically between 0.1 and 10 d (NebotGómez-Morán et al. 2011), or even merge (Han et al. 2002). Systems that transferred mass via stable Roche-lobe overflowor wind accretion can be found at longer final periods of a few10 –10 d (Pastetter & Ritter 1989; Han et al. 2002; Nie et al.2012; Chen et al. 2013). Evolved systems with periods longerthan about 3000 d are thought to always remain detached (Vander Swaelmen et al. 2017).Detailed studies of evolved binaries are fundamental for var-ious reasons. The mass and period distribution of very shortorbital period binaries provides important observational con-straints on the poorly understood common-envelope phase (Rit-ter 1986; Schreiber et al. 2010; Davis et al. 2010; Zorotovic et al.2011). In addition, these binaries can be employed to search forand study supernovae type Ia progenitor candidates (Napiwotzkiet al. 2001; Geier et al. 2013; Santander-García et al. 2015;Rebassa-Mansergas et al. 2019; Reindl et al. 2020) or help us tounderstand how binary interactions alter the intrinsic propertiesof the stars (such as their atmospheric composition, rotationalrates, pulsations, mass loss, dust formation, and circumstellar-envelope morphology; Van Winckel 2018). Last but not least, Article number, page 1 of 13 & Aproofs: manuscript no. aa binaries that avoided mass exchange can be used to investigatethe initial-final mass relation, which is a key constraint on stel-lar evolution theory and important to understanding the chemicalenrichment and the e ffi ciency of star formation in galaxies (Sil-vestri et al. 2005; Catalán et al. 2008; Zhao et al. 2012b; Baxteret al. 2014; Andrews et al. 2015).Aiming to provide a large sample to test binary evolutionmodels and type Ia supernovae formation channels, Parsons et al.(2016) established a group of 934 main-sequence FGK starsfrom the Large Sky Area Multi-Object Fiber Spectroscopic Tele-scope (LAMOST, Zhao et al. 2012a) survey and the Radial Ve-locity Experiment survey (Kordopatis et al. 2013), which showexcess flux at ultraviolet (UV) wavelengths, and hence likelyhave a white dwarf (WD) companion. Such systems are stillvery rarely known in comparison to thousands of M stars withWD companions (Holberg et al. 2013; Rebassa-Mansergas et al.2016). For nine objects in their sample, Parsons et al. (2016) ob-tained follow-up spectroscopy with the Hubble Space Telescope (HST).One of these nine systems, UCAC2 46706450, is the subjectof this work. A spectroscopic analysis of the late-type star wasperformed by various authors revealing, for example, an e ff ectivetemperature of T e ff = ±
16 K and a surface gravity of log g = . ± .
04 (Ho et al. 2017), indicating that the star is an earlyK subgiant. Assuming that the observed Ly α line in the HSTspectrum is of photospheric origin, a model-atmosphere fit byParsons et al. (2016) to the Ly α profile yielded T e ff =
24 000 Kand a low surface gravity of log g ≈ .
0, indicating that the hotcomponent is a pre-WD object like a hot subdwarf.In the present paper, we analyze in detail the UV spectrumof the compact companion and show that its temperature wasstrongly underestimated, and that it is among the hottest knownWDs (Sect. 2). We then reassess the spectroscopic observationsto characterize the K-type star (Sect. 3), search for radial velocity(RV) variations (Sect. 4), and investigate its Galactic populationmembership (Sect. 5). We derive stellar parameters of both com-ponents (Sect. 6) and examine the optical photometric variability(Sect. 7). We conclude in Sect. 8.In the following, we refer to the spectroscopic binary com-ponents as UCAC2 46706450 − A and UCAC2 46706450 − B forthe K subgiant and the hot WD, respectively.
2. Analysis of UV spectroscopy of the hot whitedwarf
The star UCAC2 46706450 was observed on Oct. 30, 2014,with the Cosmic Origins Spectrograph (COS) aboard HST, us-ing the G130M grating centered on 1291 Å for a 2109 s expo-sure (dataset LCKY08010, PI: S. Parsons). The spectrum wasretrieved from the MAST archive. The approximate useful wave-length coverage is 1138–1278 Å and 1296–1426 Å. The spec-tral resolution is about 0.1 Å. For our analysis, we smoothed thespectra with a 0.1 Å-wide boxcar to increase the signal-to-noiseratio (SNR). For comparison with observations, model spectrawere convolved with a Gaussian (FWHM = iv , N iv - v , O iv - v ,Si iv - v , S vi as well as from iron-group elements, namely, Cr vi ,Mn vi , Fe vi - viii , and Ni vi . Evidently the spectrum is that of avery hot WD. Concerning the ionization stages observed in theUV, it appears similar to other hot objects, for example, the DAPG 0948 +
534 with T e ff =
105 000 K (Werner et al. 2019) and theDAO central stars of Sh 2-216 and EGB 6 with T e ff =
95 000 K and 105 000 K, respectively (Rauch et al. 2007; Werner et al.2018b).The broad Ly α profile is dominated by interstellar hydrogen,and the COS spectrum does not cover any helium line. Thus,the H / He ratio is unknown, and in principle the object could beeither a He-rich (DO or DOA) WD or a H-rich (DA or DAO)WD. Nevertheless, we show that the e ff ective temperature andmetal abundance measurements are rather independent of this.The photospheric lines are blueshifted by −
10 km s − . Ac-cordingly, the observed spectrum is shifted to rest wavelengthsin all figures presented here. Some of the most prominent inter-stellar lines were modeled. They are blueshifted between − −
35 km s − . From Ly α, we derived a neutral H column den-sity of log n H = . ± .
04 toward the WD. Comparing thecontinuum shape of our final model and observation, we found areddening of E B − V = . ± .
01. The models presented here areattenuated by this reddening value. We note the reddening de-rived by us also agrees with values reported in the 2D dust mapsof Schlegel et al. (1998) and Schlafly & Finkbeiner (2011), aswell as with the lower limit provided by the 3D dust map ofLallement et al. (2018).We used the Tübingen Model-Atmosphere Package(TMAP ) to compute non-LTE, plane-parallel, line-blanketedatmosphere models in radiative and hydrostatic equilibrium(Werner & Dreizler 1999; Werner et al. 2003, 2012). The modelsinclude H, He, C, N, O, Si, P, S, Cr, Mn, Fe, Co, and Ni. Theemployed model atoms are described in detail by Werner et al.(2018a). In addition, we performed line formation iterations(i.e., keeping the atmospheric structure fixed) for fluorine usingthe model atom presented in Werner et al. (2015). We started our analysis by fixing the value of the surface grav-ity to log g = T e ff ≈
100 000 K, this corresponds to aWD with a mass of ≈ ⊙ . We also fixed the H / He ra-tio to the solar value. We then varied T e ff and the metal abun-dances to obtain a good fit to the line features. We found T e ff =
105 000 ± g can be constrained and how the H / Heratio a ff ects the derived atmospheric parameters.The relatively small error in T e ff follows from the fact thatwe can exploit several ionization balances, first of all, iron. Wedo not see Fe v , imposing a strict lower limit of 100 000 K. Therelative strengths of Fe vi to Fe vii lines then yield the value of105 000 K (Fig. 2). In addition, at temperatures above 110 000 K,the only observed Fe viii line (at 1148.2 Å) becomes too strong(a problem with a blending Ni vi line is discussed in Sect. 2.2).Another sensitive indicator for T e ff is oxygen. With increasingtemperature, the O iv multiplets (Fig. 2) and O v v multiplet at 1420 Å be-comes stronger. The detection of N iv lines (around 1190 Å and1271 Å) rules out a T e ff value well over 105 000 K. The absenceof the C iii multiplet at 1175 Å, on the other hand, requires T e ff above 100 000 K.Next, we looked at how the surface gravity a ff ects the modelspectrum. Reducing log g from the fiducial value 7.4 to 7.1 re-sults in unobserved sharp line cores of the C iv doublet at 1352 Å.An increase to log g = iv http://astro.uni-tuebingen.de/~TMAP Article number, page 2 of 13. Werner et al.: An extremely hot white dwarf with a rapidly rotating K-type subgiant companion (C III)C IV C IVN IVO IVSiVF VI FeVI FeVIFeVIIFeVIII NiVI? ? ?01 1140 1150 1160 1170 1180 1190 1200Ly α C IV N IVN VO V SiV CrVIFeVIFeVIINiVI ? ? ? ?01 1210 1220 1230 1240 1250 1260 1270 r e l a t i v e f l u x o N VO V O VSiIV S VICrVIMnVI(FeV) (FeV)FeVI
Fig. 1.
HST / COS spectrum of the white dwarf in UCAC2 46706450 (gray) compared to a photospheric model spectrum (blue graph: T e ff =
105 000 K, log g = lines observed at 1178 Å and 1198 Å. We adopt log g = . ± . ff erentvalues of the H / He ratio, which remains undetermined at themoment. The reduction of helium from the assumed solar valuedown to 1 . × − (mass fraction) is a ff ecting the UV metalline strengths only marginally. The increase of helium to 98 % isslightly shifting the ionization balance of metals to lower stagessuch that an increase of T e ff by about 5000 K would be neces-sary to match the observed line strengths. Therefore, within er-ror ranges, the measured metal abundances do not depend on the assumed H / He ratio and hold irrespective of whether the star isa DA or a DO white dwarf.
We present in detail how elemental abundances were inferred.The employed lines are identified in Fig. 1. The resulting abun-dance values are listed in Tab. 1 and displayed in Fig. 4.
Article number, page 3 of 13 & Aproofs: manuscript no. aa T eff =
110 kK105 kK100 kK O IVFe VI Fe VIINi VI0.51.01.5 1162 1164 1166wavelength / A o r e l a t i v e f l u x T eff =
110 kK105 kK100 kK O IVFe VIFe VII1334 1338 1342wavelength / A o Fig. 2.
Details of observed WD spectrum compared with three models with di ff erent temperatures ( T e ff =
105 000 ± T e ff . Left: the O iv multiplet gets weaker with increasing T e ff , as well as lines from Fe vi and Ni vi , while Fe vii lines getstronger. Right: same behavior of oxygen and iron lines in another wavelength range. Ni VIFe VIII100 kK105 kK110 kK0.51.0 -1 0 1 ∆λ / A o r e l a t i v e f l u x Fig. 3. E ff ect of increasing temperature on the Ni vi / Fe viii line blendat 1148.2 Å ( T e ff = viii line profile. The nickel lines become weaker and the ironline becomes stronger. As a result, the line Fe + Ni blend stays almostconstant. The models have abundances of Fe = . × − and Ni = . × − by mass. Three C iv multiplets are detectable and were fit, namely 3d–4f,3d–4p, and 3p–4s, at 1169 Å, 1198 Å, and 1230 Å, respectively.It is worthwhile to note that two other, more highly excited, C iv multiplets are not detectable: 4p–7d at 1316 Å and the blend of4d–7f and 4f–7g at 1351 Å and 1353 Å. They serve to constrainthe upper limit on the C abundance and the lower limit on surfacegravity (see Sect. 2.1).Besides the strong N v resonance doublet, we observe a weakand highly excited N v multiplet (4s–5p) at 1390 Å, and severalN iv lines. There is a singlet at 1188 Å ( S– P o ), which, however, Table 1.
Atmospheric parameters of the WD in UCAC2 46706450. a T e ff / K 105 000 ± g / cm s − ) 7 . ± . X i [ X i ]C 2 . × − − . × − . × − − < . × − < . . × − . × − . × − . × − . × − < . × − < . . × − Notes. ( a ) Abundances given in mass fractions (column 2) and loga-rithmic abundances relative to solar value (column 3; solar abundancesfrom Asplund et al. 2009). Error limits for the abundances are ± . is blended by a Ni vi line and a multiplet at 1196 Å ( D– P o ).A triplet appears at 1271 Å ( P o – D), but two components areblended by Fe vi lines.Two prominent O iv multiplets are detected, namely a blendof components at 1164 Å ( D– F o ) and the three components ofthe P– D o transition at 1338.6 Å, 1343.0 Å, and 1343.5 Å. Wealso detect several O v lines: the prominent O v D– D o singlet at 1219 Å, and an accumulation of weak lines in theregion 1418–1426 Å, which stem from the transitions P o – Dand D– F o . We detect a very weak feature at the position of the F vi Article number, page 4 of 13. Werner et al.: An extremely hot white dwarf with a rapidly rotating K-type subgiant companion
H N F P Mn CoHe C O Si S Cr Fe Ni white dwarfK subgiantUCAC2 46706450 -50 5 10 15 20 25 atomic number l og ( m a ss f r a c t i on ) H N F P Mn CoHe C O Si S Cr Fe Ni
WD in UCAC2 46706450WD in PN EGB 6 -50 5 10 15 20 25 atomic number l og ( m a ss f r a c t i on ) Fig. 4.
Left: element abundances measured in the WD (red squares) and in the K subgiant (black dots) of the binary UCAC2 46706450. Right:abundances in the WD of UCAC2 46706450 (red squares), compared to the DAO WD in EGB 6 (black dots). The blue line indicates solarabundances. the blue edge of the COS spectrum. We therefore adopted thesolar abundance as an upper limit.Besides the Si iv resonance doublet at 1394 / v lines at 1152.0 Å and 1251.4 Å, which arecomponents of the P o – P and P o – D transitions, respectively.The Si iv doublet is blended by an ISM component that isblueshifted by −
25 km s − relative to the photospheric one. Aweak S vi line at 1419.7 Å is identified, which is the strongest ofthree components of the 4d–5p transition, allowing for a sulfurabundance measurement. A line of Cr vi is visible at 1417.7 Å. Another one, at 1261.1 Å, isa blend with an Fe vi line. A third one is seen at 1255.8 Å, but thisis blended with a broad, unidentified feature. Five weak Mn vi lines are detected at 1333.9 Å, 1345.5 Å, 1356.9 Å, 1391.2 Å,and 1396.7 Å.As mentioned previously, we identified iron lines from Fe vi - viii . Care must be taken when using the only Fe viii line (at1148.2 Å) as a temperature indicator (Fig. 3). Its strength in-creases with T e ff , but there is a blending Ni vi line whose strengthdecreases with T e ff . In e ff ect, the line blend stays constant in thetemperature regime discussed here. We do not see lines of Fe v .Some of the Fe v lines that we see in models at slightly lowertemperature in the wavelength region of 1373–1414 Å are indi-cated in Fig. 1.Our models predict many Co vi lines, mainly in the short-wavelength range of the COS spectrum, provided the cobaltabundance is high enough. We do not detect any such line andderive an upper abundance limit.Many Ni vi lines are found, mainly in the short wavelengthregion up to about 1210 Å. Similarly to iron, Ni v lines are absentdue to the high e ff ective temperature. A number of photospheric lines remain unidentified. Thestrongest appear at these wavelengths: 1173.6, 1191.0, 1195.7,1236.4, 1246.2, 1256.3, 1266.9, 1298.3, 1332.8, 1345.5, and1355.7 Å. These are indicated in Fig. 1 with a question mark.We also found these lines in other hot WDs or central stars ofplanetary nebulae (PNe) and suspect that some of them could beunknown Fe vii lines.
Within error limits, the abundances of light metals metals are so-lar (Tab. 1, Fig. 4). For the iron-group elements a slight enhance-ment (0.45–0.70 dex oversolar) is found. This overabundancemight point to radiative acceleration acting in the atmosphere,but we report below that the K-subgiant also exhibits an over-abundance of these elements (left panel of Fig. 4). Hence, the el-ement abundance pattern probably reflects the initial metallicityof both stars. In the right panel of Fig. 4, we compare the abun-dances found for UCAC2 46706450 − B with those of the WDcentral star of the PN EGB 6, which has the same T e ff and log g (Werner et al. 2018b). Apart from nitrogen, the abundance pat-terns are rather similar.
3. Optical and IR spectroscopy of the K star
Subgiant UCAC2 46706450 − A dominates the flux in the opticaland infrared (IR) wavelength range. Optical spectroscopy ( λ ≈ − R ≈ R ≈
22 500) IR ( λ ≈
15 100 −
17 000 Å)spectroscopy was obtained with the Apache Point Observa-tory Galactic Evolution Experiment (APOGEE; Majewski et al.2017). In the course of the Gaia mission, a medium-resolution( R ≈
11 700) spectrum over the wavelength range 8450–8720 Å(around the Ca ii triplet) was collected with the Radial VelocitySpectrometer.Additional follow-up spectroscopy were obtained at the 5.1-meter Palomar Hale telescope in California, USA, and the ThaiNational Telescope (TNT) at the National Observatory in Thai-land. One spectrum was obtained at the Palomar Hale Telescopeduring the night of the Jan. 20, 2014. We used the double spec-trograph together with the 1200 lines / mm grating in the red andthe 600 lines / mm grating in the blue. The observations were car-ried out using the long slit with 1 ′′ width. This configurationresulted in a wavelength coverage of ≈ − ≈ − R = R = softwares, respectively. Thir-teen additional spectra were obtained during the 2015–2017 pe-riod at the TNT using the Middle Resolution fiber-fed EchelleSpectrograph (MRES). The spectra covered the ≈ − Tom Marsh’s molly package is available at http://deneb.astro.warwick.ac.uk/phsaap/software
Article number, page 5 of 13 & Aproofs: manuscript no. aa
Table 2.
Atmospheric parameters, masses as obtained from evo-lutionary calculations, and radii and luminosities of the subgiantUCAC2 46706450 − A and the white dwarf UCAC2 46706450 − B.UCAC2 46706450 − A UCAC2 46706450 − B T e ff / K 4945 ±
250 105 000 ± g / cm s − ) 3 . ± .
25 7 . ± . M / M ⊙ . − . . ± . R / R ⊙ . + . − . . + . − . L / L ⊙ + − + − wavelength range at R =
15 000. The spectra were reduced andcalibrated using the DECH software package.The two LAMOST spectra from data release 2 (DR2) wereanalyzed by Luo et al. (2016), who derived T e ff , log g , [Fe / H],as well as the RV values from the individual observations. Thesespectra were also analyzed by Ho et al. (2017), who provide re-sults for T e ff , log g , [Fe / H], as well as the α -enhancement [ α / M].Two additional LAMOST spectra were obtained within the DR5that were analyzed (along with the previous spectra) by Luo et al.(2019). Abundances of several elements (C, Cl, N, O, Mg, Al,Si, P, K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, and Rb), T e ff , log g , andRV from the APOGEE spectrum are provided by the APOGEEStellar Parameter and Chemical Abundances Pipeline (ASPCAP,García Pérez et al. 2016). Some of these element abundancesrelevant for the comparison with the WD companion are de-picted in the left panel of Fig. 4. The iron-group elements areslightly enhanced, by 0.2–0.5 dex. An interesting point to no-tice is that ASPCAP predicts an unusually high rotational ve-locity of v sin( i ) = .
76 km s − . Atmospheric parameters fromthe APOGEE spectrum were also derived with the data drivenmethod called the Cannon (Ness et al. 2015; Casey et al. 2016).However, the χ of the Cannon fit is quite large, indicating thatthe parameters and abundances are not reliable, and thus we didnot consider these values further. Finally, the Gaia DR2 also pro-vides values for T e ff and RV.In Table 2, we provide the average of the T e ff and log g val-ues derived by the various analyses and give an estimate of theerror, which also includes the systematic error (estimated fromthe scatter of the results from the individual analyses). We notethat the fits to the optical spectrum do not consider the contribu-tion of the flux of the WD to the blue part of the spectrum (seeFig. 7). On the other hand, values from the fit to the LAMOSTspectrum do not di ff er too much from what is obtained from theAPOGEE and Gaia spectra.A closer inspection of the LAMOST spectra reveals that theline cores of the Ca ii H and K doublet (Fig. 5) appear in emis-sion. The double-peaked H α emission (also seen in the TNTspectra, Fig. 6), can be explained by the superposition of a verybroad emission line plus a photospheric absorption line. Thestrength of these emission lines is found to be time variable.Emissions in these lines are typical for chromospherically activestars (Wilson 1963, 1968; Gray & Corbally 2009).
4. Radial velocity variability
The scatter of the RV values from the three individual APOGEEspectra (which were all obtained within one month) is smallerthan 1 km s − , which already indicates that the system is verylikely not a close binary. We also searched for RV variations us-ing the values we measured from the LAMOST, Palomar, and r e l a t i v e f l u x C a II C a II λ / A o r e l a t i v e f l u x H α Fig. 5.
Time-variable emission line cores in the Ca ii H and K dou-blet and H α emission observed in the LAMOST spectra indicate thatUCAC2 46706450 − A is chromospherically active. The HJD of the ob-servations is provided. λ / A o r e l a t i v e f l u x H α Fig. 6.
Time-variable H α emission observed in TNT spectra indicatethat UCAC2 46706450 − A is chromospherically active. The HJD of theobservations is provided.
TNT spectra, as well as those provided in the catalogs for theAPOGEE and Gaia spectra (total time coverage is about fouryears, see Table 3). In order to check if the RV variations aresignificant or are merely produced by random fluctuations, wefollowed the approach as outlined in Maxted et al. (2001) andGeier et al. (2015). For this, we calculated the inverse-varianceweighted mean velocity from all RV measurements. Assumingthis mean velocity to be constant, we calculated the χ . Com-paring this value with the χ distribution for the appropriatenumber of degrees of freedom, the probability p of obtainingthe observed value of χ or higher from random fluctuationsaround a constant value can be calculated. We obtain a value oflog p = − . , and hence conclude that UCAC2 46706450 − A isnot significantly RV variable and very likely not a close binary. It is worth mentioning that Rebassa-Mansergas et al. (2017)claimed UCAC2 46706450 − A to be RV variable. They obtainedthe RV values by fitting the Na i absorption doublet at around5900 Å. However, revising these fits, we realized that some ofthe RVs they used were a ff ected by embedded emission (likelyresulting from the night sky) and are hence not reliable. By ex-cluding these RV values, UCAC2 46706450 − A does not showRV variation, in agreement with the results obtained here. Geier et al. (2015) considered objects with log p > . − . > log p > − . Table 3.
HJD and radial velocities of UCAC2 46706450 − A as measuredfrom the di ff erent spectra. HJD v rad [km s − ] survey2456424.148720 − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± .
15 APOGEE2456717.996280 − . ± .
12 APOGEE2456735.937290 − . ± .
17 APOGEE2457058.466008 − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − ± − . Consid-ering the gravitational redshift of v grav = . ± . − givesa RV value for the WD of v rad = − . ± . − (calcu-lated using the mass and radius from Sect. 6). This agrees verywell with the most precisely measured RV value of the K star( v rad = − . ± .
07 km s − from ASPCAP) and can be seenas a first hint that the system is indeed a physically connectedbinary and that the velocities of both stars represent the systemvelocity.We stress, however, that with the current set of RV mea-surements we are only able to detect RV amplitudes larger than ≈
15 km s − , and that in case of a low inclination angle or a longorbital period, the amplitude can be much smaller than this.
5. Distance and Galactic population membership
The distance of UCAC2 46706450 − A (the dominant source inthe optical) based on the Gaia DR2 parallax is provided inthe Bailer-Jones et al. (2018) catalog and amounts to d = + − pc. From this we calculate a height above the Galacticplane of z = + − pc, that is to say we find that it lies in aregion where the thick disk dominates.The Galactic population membership of UCAC2 46706450can be further constrained by looking at the chemistry ofUCAC2 46706450 − A, as well as the kinematics of the system.Both thick disk and thin disk stars are observed over a widemetallicity range ([Fe / H] between − . α -rich ata given metallicity compared to thin disk stars (e.g., Fuhrmann1998; Recio-Blanco et al. 2014). At metallicities above [Fe / H] = . / H] = .
09 dex (standard deviation 0 .
10 dex) and [ α / M] = .
02 dex(standard deviation 0 .
07 dex) reported for UCAC2 46706450 − Ain the LAMOST and ASPCAP catalogs, the star appears juston the borderline between the two populations (see, e.g., Fig. 1in Hayden et al. 2017). Therefore, it is worth also looking at the kinematics of the system. Since LAMOST, APOGEE, TNT,Palomar, Gaia, and HST / COS spectra do not indicate significantRV variations, we adopt the mean RV ( − . ± .
55, wherethe error represents the standard deviation of the individual mea-surements) as system velocity. With this and the proper motionand parallax from the Gaia DR2, we calculate space veloci-ties of U = . ± . − , V = − . ± . − , and W = − . ± . − ( ). The kinematic Galactic populationmembership can then be derived on the basis of the Mahalanobisdistance, which is the number of standard deviations between thespace velocities of UCAC2 46706450 and the velocity ellipsoidsof the thin disk, thick disk, and the halo (Gianninas et al. 2015).Using the velocity ellipsoid values from Kordopatis et al. (2011),we find that the kinematics of UCAC2 46706450 clearly point tomembership of the thick disk. A final, and probably most deci-sive, criterion that can be used to determine the Galactic popu-lation membership is the age of the system, and this is discussedfurther in the next section.
6. Radii, luminosities, masses, and ages
In order to determine the radii of the two stellar components,we performed a fit to the spectral energy distribution (SED),by varying their solid angles π ( R / d ) (which relates the astro-physical flux on the surface of the stars to what is receivedon Earth) until a good agreement of the combined syntheticflux and the observations was found. We assumed a redden-ing of E B − V = .
03 (see Sect. 2) and the distance providedby Bailer-Jones et al. (2018) based on the Gaia parallax. ForUCAC2 46706450 − B, we used our best fit model from Sect. 2.For UCAC2 46706450 − A, we used a NextGen model (Allardet al. 2012) with T e ff = g = .
0, and a metallicity of[Fe / H] = FUV and
NUV magnitudes were taken from Bianchiet al. (2014) and converted to fluxes as outlined in Reindl et al.(2016). B , V , g , r , and i magnitudes were taken from Hendenet al. (2015), 2MASS J , H , and K magnitudes from Cutri et al.(2003). To convert these magnitudes into fluxes, we used thezero-points provided in Holberg & Bergeron (2006). Wide-fieldInfrared Survey Explorer (WISE) W W W
3, and W − B,UCAC2 46706450 − A, and the combined best fit model, respec-tively. All model fluxes are corrected for interstellar reddening,using the reddening law of Fitzpatrick (1999) with E B − V = . / COS, co-addedLAMOST, and APOGEE spectra. Overall, the combined modelflux reproduces the observed SED well. Only the observed NUVflux exceeds the predicted flux by about 0.1 dex. We attribute thisto the chromospheric flux of the K-type star, as chromosphericfluxes are known to exceed the photospheric flux by severalorders of magnitude in chromospherically active stars (Stelzer Cardinal directions, with U being positive toward the Galactic center.Values have been adjusted for the Solar motion using the numbers ofTian et al. (2015). Article number, page 7 of 13 & Aproofs: manuscript no. aa -18-17-16-15-14-133.0 3.5 4.0 4.5 5.0 log ( λ / A o ) l og ( f λ / e r g s - c m - A o - ) COS LAMOST APOGEE
FUV NUV g r iB V HJ K W1 W2 W3 W4
Fig. 7.
Fit to SED of UCAC2 46706450. HST / COS, LAMOST (normalized to r-band magnitude), and APOGEE spectra are shown in gray. Themodel fluxes from the hot WD and the K-type subgiant are shown in light blue and green, respectively. The combined model flux is shown in red.The black dots indicate filter-averaged fluxes that were converted from observed magnitudes. Gray open circles indicate WISE fluxes for whichthe flux correction was not taken into account. et al. 2013, 2016; Dixon et al. 2020). Based on the NUV excessof 0.1 dex, we estimate that chromospheric flux of the K-typestar exceeds its photospheric flux by about two orders of magni-tude and which is also what is found for other such rapidly rotat-ing and chromospherically active giants (Dixon et al. 2020). TheWISE W W − A, we derive a radius of R A = . + . − . R ⊙ , and for the hot WD, R B = . + . − . R ⊙ . Using L = π σ R T ff , where σ is the Stefan-Boltzmann constant, wecalculate luminosities of L A = + − L ⊙ and L B = + − L ⊙ forUCAC2 46706450 − A and UCAC2 46706450 − B, respectively.These values are what is expected for a subgiant star and a hotWD that is just about to enter the WD cooling sequence. Hence,we conclude that the system is indeed a physical binary.With the spectroscopically determined surface gravities (av-erage value for the K star from Table 2) and using M = g R / G ,where G is the gravitational constant, we calculate masses of M A = . + . − . M ⊙ and M B = . + . − . M ⊙ for the K-type sub-giant and the hot WD, respectively. We note that the large errorson the masses are mainly a consequence of large uncertaintiesof the surface gravities of both stars. The masses can be con-strained more precisely when we compare their locations in theHertzsprung Russell diagram (HRD) with predictions from stel-lar evolutionary calculations. But this mass determination alsohas its limitations, as is outlined in the following. In the upperpanel of Fig. 8, we show the location of UCAC2 46706450 − A inthe HRD compared to stellar evolutionary tracks of Pietrinferniet al. (2004). The tracks are calculated for Z = .
01 ([Fe / H] = M A = . + . − . M ⊙ . In the lowerpanel of Fig. 8, we show the evolutionary tracks of Pietrinferniet al. (2004), which were calculated for a slightly higher metal- licity of Z = .
03 ([Fe / H] = ff ective temper-atures, and thus suggest a higher mass of M A = . + . − . M ⊙ . Thisshows that the mass of the K-type subgiant obtained from theHRD depends significantly on the chemistry used in the evolu-tionary calculations. We therefore infer a possible mass range of M A = . − . M ⊙ based on these evolutionary models.In Fig. 9, we show the location of UCAC2 46706450 − B inthe HRD compared to evolutionary calculations of H-shell burn-ing post-asymptotic giant branch (AGB) stars from Miller Berto-lami (2016). This suggests a mass of M B = . ± . M ⊙ and that the surface gravity of the WD might be rather atthe lower boundary of the error margin (log g = . ± .
1) de-termined spectroscopically. In gray, a He-shell burning verylate thermal pulse (VLTP) evolutionary track is shown. Thistrack extends toward higher T e ff at the luminosity found forUCAC2 46706450 − B, which supports the assumption that theWD is indeed H-rich (higher envelope mass; much lower massnot possible due to the limited age of the disk / Milky Way). InFig. 9, we also indicate the initial masses assumed for the com-putation of the evolutionary tracks. This suggests that if theinitial-to-final mass relation (IFMR) assumed in these calcula-tions are accurate and the system has not interacted in previ-ous evolutionary phases, the initial mass of the hot WD wasabout 1 M ⊙ , or only slightly higher. A low mass of the WD isalso supported by the absence of a PN. The tracks of MillerBertolami (2016) calculated for Z = .
01 predict a post-AGBage of about 72 kyrs for the 0.52 M ⊙ model and only about10 kyrs for the 0.56 M ⊙ ( Z = .
01) model. This means that ifUCAC2 46706450 − B should have ejected a PN on the AGB, andwe assume that a PN is typically visible for 30 kyrs, then, in thecase of a low mass post-AGB star, the PN should have fadedaway already, while for higher masses, the PN should still bevisible. Radii, luminosities, and masses of UCAC2 46706450 − Aand UCAC2 46706450 − B are summarized in Table 2.If we assume again that the system has not undergone masstransfer in its past and the initial mass of the WD was be-
Article number, page 8 of 13. Werner et al.: An extremely hot white dwarf with a rapidly rotating K-type subgiant companion
Z=0.03 R R
012 9000 8000 7000 6000 5000 T eff / K l og ( L / L ) Z=0.01 R R l og ( L / L ) Fig. 8.
Location of UCAC2 46706450 − A (red) in the HRD compared tostellar evolutionary tracks of Pietrinferni et al. (2004) for initial massesof 0.8, 1.0, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.2, and 2.4 M ⊙ . The upperpanel indicates evolutionary tracks calculated for Z = .
01 ([Fe / H] = Z = .
03 ([Fe / H] = R ⊙ . tween 1.0 and 1.25 M ⊙ , this would imply that the initial mass ofUCAC2 46706450 − A is only slightly lower than this. The totallifetime of stars in that mass range at solar metallicity stated byMiller Bertolami (2016) lies between 5.3 and 12.5 Gyrs. Thus,it is entirely possible that the system indeed belongs to the thickdisk, which is generally found to be exclusively old, meaningolder than nine Gyrs (Feltzing & Bensby 2009; Kilic et al. 2017;Hayden et al. 2017). We also note that in the course of theirstudy of main-sequence turno ff and subgiant stars from the AM-BRE:HARPS survey, Hayden et al. (2017) find stars at the metal-licity of UCAC2 46706450 − A that have ages between 8 and
M3B Z=0.02, 1.00/0.52 MM3B Z=0.01, 1.00/0.53 MM3B Z=0.02, 1.50/0.58 MM3B Z=0.01, 1.25/0.57 MVLTP, 1.00/0.52 M123 120 100 80 T eff / kK l og ( L / L ) Fig. 9.
Location of UCAC2 46706450 − B (red) in the HRD comparedto stellar evolutionary tracks of Miller Bertolami (2016) calculated for Z = .
01 (purple) and Z = .
02 (green). Initial and final masses of thetracks are indicated. Shown in gray is a VLTP evolutionary track. Theblack square indicates the position of EGB 6.
10 Gyrs. However, since the IFMR is poorly constrained at thelow-mass end, and mass loss on the RGB is not well understood,it is not possible to give a more precise estimate of the system’sage.
7. Light curve variability
We obtained the Catalina Sky Survey (CSS, Drake et al. 2009)DR2 V-band light curve of UCAC2 46706450 as well as c- ando-band light curves (e ff ective wavelengths 0 . µ m and 0 . µ m,respectively) from the Asteroid Terrestrial-impact Last AlertSystem (ATLAS, Tonry et al. 2018) DR1 (Fig. 10).We also observed the light curve of UCAC2 46706450 inthe V-band using the Tübingen 80 cm f / × ff erent aper-ture radii with respect to comparison star variances. We note thatthe data were obtained under non-photometric conditions, withthe weather conditions being worst in the third night. As there isonly one equally bright star in the field of view, we used only onecomparison star for the first and third nights. During the secondnight, the observing conditions were best, so a second compari-son star could be used.In addition, we used Transiting Exoplanet Survey Satellite (TESS) observations of UCAC2 46706450 (Fig. 11). The starwas in the TESS field of view during Sector 23 observations.
Article number, page 9 of 13 & Aproofs: manuscript no. aa
ATLAS c12.912.8 c [ m ag ] CSS V12.912.6 V [ m ag ] ATLAS o12.412.3 0.5 1.0 1.5 phase o [ m ag ] Fig. 10.
ATLAS c-band, CSS V-band, and ATLAS o-band phase-foldedlight curves. The CSS light curve is folded with the 1.97861678 d pe-riod. r e l a t i v e f l u x r e l a t i v e f l u x Fig. 11.
TESS (magenta) and Tübingen 80 cm telescope V-band (green)light curves. The V-band light curve has been shifted by one period.
As there is no 2-minute cadence light curve available, we per-formed photometry using the 30-minute cadence full-frame im-ages (FFIs) using lightkurve (Lightkurve Collaboration et al.2018). We used a relatively small 4 × P = . P = . − and 10 − , respectively. The derived periodfrom the o-band equals exactly half the period given in Heinzeet al. (2018), who classified the light curve of UCAC2 46706450as an irregular variable. This class serves in their catalog asa “catch-all” bin for objects that do not seem to fit into anyof their more specific categories. We also performed a gener-alized Lomb-Scargle search on the CSS light curve, and findthe strongest signal at P = . − . A possible explanation could bethat UCAC2 46706450 is indeed an irregular variable and thatits photometric period and amplitude change over time. Thus,in the much longer CSS light curve (time coverage ≈ = = = P = . = − ), close to the value found for theATLAS light curves. In Fig. 10, we show the phase-folded AT-LAS c- and o-band and CSS light curves. The black dashed linesare fits of a harmonic series to the phase-folded light curves (seeEquation (49) in Hartman & Bakos 2016) and predict peak-to-peak amplitudes of 0.08 mag, 0.10 mag, and 0.06 mag in the V-,c- and o-bands, respectively. We note, however, that due to thephotometric uncertainties (0.05 mag for CSS, and 0.01 mag forthe ATLAS light curves), the di ff erences in the amplitudes mightnot be real.For the TESS light curve, we derive a period of 1 . = − ). The light curve is shown in Fig. 11 along withthe V-band light curve obtained with the Tübingen 80 cm tele-scope, which is shifted for one period. We find that the phasesof both light curves match nicely, though the amplitudes di ff er.We note, however, that the amplitudes of both should be treatedwith caution. As mentioned above, the Tübingen 80 cm telescopelight curve was obtained under non-photometric conditions, anddue to the large pixel-size of TESS two neighboring nearby starsmay add additional continuum flux, which in turn may reducethe amplitude of the TESS light curve. What can also be seen inFig. 11 is that the shape of the TESS light curve changes slightly,which is typical for star spot evolution (K˝ovári et al. 2019).If we assume the observed period of the light curve varia-tion corresponds to an orbital period, and a mass of 0.54 M ⊙ for the WD and 1.0 M ⊙ of the K-type subgiant (Sect. 6), thenwe calculate via the third Kepler law, a separation of the twostars of a = . R ⊙ . In that case, a strong reflection e ff ect shouldbe noticeable. In such systems, a tidally-locked cool compan-ion is strongly irradiated by the hot primary, also causing si-nusoidal light curve variations at the orbital period of the sys-tem. Then, however, the amplitudes of the light curve variationsshould be increasing toward longer wavelengths, which is notobserved. In addition, we would expect to see all Balmer lines,the CNO line complex around 4560 Å, and possibly also someother lines in emission. These two observational properties areseen in BE UMa, which also has a T e ff =
105 000 K hot primary,an orbital period of 2 .
29 d, and a distance a = . R ⊙ from its Article number, page 10 of 13. Werner et al.: An extremely hot white dwarf with a rapidly rotating K-type subgiant companion cool companion (Ferguson & James 1994) and in many other re-flection e ff ect systems containing a very hot primary (e.g., Singet al. 2004; Exter et al. 2005; Nagel et al. 2006; Shimansky et al.2015; Mitrofanova et al. 2016).Thus, we conclude that the variations are caused by spotson the surface of the star and that the photometric period corre-sponds to the rotational period of UCAC2 46706450 − A. With aradius of 5 . + . − . R ⊙ and rotation period of 1.98 d, we calculate arotational velocity of v rot = + − km s − . If we assume that themass is close to 1 M ⊙ (Sect. 6), this value is close to the breakupvelocity. With v sin( i ) = .
76 km s − as suggested by the AS-PCAP fit of the APOGEE spectrum, one can estimate that theinclination of rotational axis is i = ◦ .If we assume a bound rotation (as seen in many RS CVnstars), that the inclination angle of the rotational axis is per-pendicular to the orbital plane, and the masses given in Tab. 2,then RV semi-amplitudes of 58–105 km s − and 23–41 km s − would be expected for the WD and subgiant, respectively. Thelatter is much larger than the standard deviation of the individ-ual RV measurements of UCAC2 46706450 − A (4.55 km s − , seeSect. 5) and, thus, again speaks against a close binary system.
8. Conclusions
From the SED fit, the fact that we do not find signifi-cant RV variability, that the RVs of UCAC2 46706450 − A andUCAC2 46706450 − B agree, and that no reflection e ff ect is no-ticeable in the light curves and spectra, we conclude the sys-tem is most likely a wide, physically connected binary system.UCAC2 46706450 − A is a subgiant just about to climb up theRGB, while UCAC2 46706450 − B is an extremely hot WD justabout to enter the WD cooling sequence.Kinematically, and from its height above the Galactic plane,the system belongs to the Galactic thick disk. This can also ex-plain the low mass of the WD without the need for binary in-teractions, as the thick disk is generally found to contain exclu-sively old, and by that low-mass stars.
We determined the atmospheric parameters of the hot WD inthe binary UCAC2 46706450. Within error limits, the light metal(C, N, O, F, Si, S) abundances are solar, whereas iron-group el-ements (Cr, Mn, Fe, Ni) are slightly enhanced (0.45–0.70 dexoversolar). The H / He ratio cannot be determined from the avail-able spectra. However, by comparison with another WD we canconclude that it is likely a DAO white dwarf. The e ff ective tem-perature ( T e ff =
105 000 K) is the same as that of EGB 6, a DAOcentral star of a PN (Werner et al. 2018b). The metal abun-dance pattern of EGB 6 is very similar, and its H / He ratio is solar(Fig. 4, right panel). It was concluded that EGB 6 is at the windlimit of its cooling track, shortly before gravitational settlingand radiative acceleration of elements begin to a ff ect the pho-tospheric abundance pattern. Before the wind limit, a radiation-driven wind prevents atomic di ff usion (e.g., Unglaub & Bues2000). Thus, UCAC2 46706450 − B is probably a DAO whitedwarf. The He abundance could be subsolar since the slightlyoversolar heavy-metal abundances indicate that di ff usion beganto act. On the other hand, heavy-metal abundances in the K-tpye subgiant are also oversolar (Fig. 4, left panel), thus, the el-ement abundance pattern probably reflects the initial metallicity The breakup velocity v breakup = √ GM / R of a 1.0 / M ⊙ star with aradius of 5.9 R ⊙ is 180 /
254 km s − , respectively. of both stars. UCAC2 46706450 − B could be just crossing thewind-limit. Observational evidence testifies that the wind limitfor H-rich WDs is indeed located at T e ff ≈
105 000 K (Werneret al. 2018b, 2019) and recent theoretical work confirms this re-sult (Krtiˇcka et al. 2020).We emphasize, however, that we cannot rule out that the WDis a helium-rich DO white dwarf. Before the wind limit, DOs canalso exhibit a near-solar metal abundance pattern (Werner et al.2017). To make a conclusive decision, an observation coveringthe position of He ii M = . + . − . M ⊙ , Fig. 9)than UCAC2 46706450 − B ( M = . ± . M ⊙ ), and thusEGB 6 evolved from the AGB to its current position morerapidly (post-AGB time is about log( t evol yr − ) = . + . − . )than UCAC2 46706450 − B (log( t evol yr − ) ≈ . − A no PN can be detected.
We reported the discovery of a time-variable emission of H α aswell as emission line cores in the Ca ii H and K doublet, whichis a strong hint to chromospheric activity. Further evidence forchromospheric activity can be seen from the NUV excess. Fur-thermore, we found that the object is photometrically variablewith a period of ≈ − A. With the radius from the SED fit, we findthat the star is rotating extremely fast with v rot = + − km s − .This marks UCAC2 46706450 − A as one of the most rapidly ro-tating subgiants known.The chromospheric activity, NUV excess flux, star spots, andvery rapid rotation is what is also seen in FK Comae Berenices(FK Com) stars, a small group (only a handful of objects known)of single, rapidly rotating, and very active GK-type subgiants,which are believed to be the product of a recent merger (Eggen& Iben 1989). The prototype FK Com is the undisputed “kingof spin” amongst these single red giants with an equatorial spinvelocity of v eq =
179 km s − , followed by HD 199178, the socalled “spin vice king” amongst giant stars (Costa et al. 2015),with v eq =
93 km s − (Strassmeier 2009). A related class is that ofthe more frequently known RS Canum Venaticorum (RS CVn)variable stars, which are close binary systems consisting of achromospherically active subgiant component, which exhibitsbrightness variations caused by large cool spots (Hall 1972), andwhich is also known to be fast rotating (Strassmeier 2009). Thereare also two other hot (pre-) WDs known, which are in a binarysystem with chromospherically active and rapidly rotating coolcompanions: LW Hya and LoTr 5. The latter is a wide binaryin a highly eccentric orbit ( e ≈ . P = v eq =
95 km s − , Strassmeier2009) and magnetically active G5 III giant that also shows H α line-profile variations (K˝ovári et al. 2019). LW Hya is a resolvedbinary system composed of a hot DAO WD (the ionizing star ofA 35, Ziegler et al. 2012; Löbling et al. 2019) and a magneticallyactive G8 III-IV-type companion ( v eq =
127 km s − , Strassmeier2009). For LW Hya, it is, however, not clear if the system is re-ally a physical binary, due to the mismatch of the spectroscopicdistance of the WD and the optical parallax. Article number, page 11 of 13 & Aproofs: manuscript no. aa
The driving mechanism for the chromospheric activ-ity and the reason for the rapid rotation observed inUCAC2 46706450 − A remains to be explored. In main sequencestars of spectral type F, G, K, and early M, a self-sustaining mag-netic dynamo driven by rotation and convection is believed tobe the source of chromospheric activity. Since due to magneticbraking, the rotational velocity of the star is expected to decreasewith age, the chromospheric activity is thought to decrease aswell. However, if angular momentum is sustained, by tidal in-teraction as in the case of short-period binaries (this holds forthe RS CVn stars), or maintained by convection, chromosphericactivity can be preserved (Zhao et al. 2011).Besides a recent stellar merger, sudden dredge-up of angu-lar momentum from the stellar interior (Simon & Drake 1989),and accretion of a substellar companion (Peterson et al. 1983;Carlberg et al. 2012) are considered as possible explanations forthe abnormally high rotation rates of FK Com stars. As men-tioned before, the lack of significant RV variations makes theclose-binary scenario for UCAC2 46706450 very unlikely. Onemay therefore speculate that the system used to be a hierarchi-cal triple system, and that UCAC2 46706450 − A was producedby the merger of the inner binary, as was suggested for IN Com(Jasniewicz et al. 1987; Malasan et al. 1991).
Recently, Tian et al. (2020) reported that based on common Gaiaparallaxes and proper motions, UCAC2 46706450 is in an ultra-wide binary system with Gaia DR2 1391040916768689280(alias SDSS J152844.16 + ′′ (111 634 AU). This fainter star has an absolute Gaia G-bandmagnitude of M G = .
76 mag and a color index of G BP − G RP = .
26 (Tian et al. 2020), thus it is likely a K-type main-sequencestar. That implies a low mass ( M ≤ . M ⊙ ) of the possi-ble ultra-wide companion, which would also be in line withthe lower limit of the (initial) masses of UCAC2 46706450 − Aand UCAC2 46706450 − B, meaning that they indeed could haveformed at the same time.
A future detailed spectral analysis of UCAC2 46706450 − A,which considers all spectral observations at once and also con-siders the flux contribution of the hot WD in the blue, would behighly desirable. This would help to better constrain the chem-ical composition and the mass of the star, as well as the age ofthe system. Near-UV spectroscopy could help to determine thehelium abundance of the WD and to investigate the nature of theNUV excess flux. Long-term, high-resolution monitoring, thatallows RVs to be measured with an accuracy of a few hundredm s − , will help to constrain the orbital period of the system andin turn makes it possible to conclude if the system underwentany mass transfer or wind accretion in its past, and to check ifthe merger scenario could be possible. Spectroscopic follow-upof SDSS J152844.16 + − A. Acknowledgements.
We thank the referee, Uli Heber, for his constructive re-port. We appreciate useful discussions with M3B and Stephan Geier. We thankMatti Dorsch and Boris Gänsicke for pointing out this system to us and use-ful discussions. ARM acknowledges support from the MINECO under theRamón y Cajal programme (RYC-2016-20254) and the AYA2017-86274-Pgrant, and the AGAUR grant SGR-661 / http://astro.uni-tuebingen.de/~TMAD ) used for this paper was constructed as part of the activities of the German Astrophysical Virtual Observatory. This work has madeuse of data obtained at the Thai National Observatory on Doi Inthanon, oper-ated by NARIT. Some of the data presented in this paper were obtained fromthe Mikulski Archive for Space Telescopes (MAST). This research has made useof NASA’s Astrophysics Data System and the SIMBAD database, operated atCDS, Strasbourg, France. This research made use of Lightkurve, a Python pack-age for Kepler and TESS data analysis (Lightkurve Collaboration, 2018). Thiswork has made use of data from the European Space Agency (ESA) mission Gaia ( ), processed by the Gaia
Data Process-ing and Analysis Consortium (DPAC, ). Funding for the DPAC has been provided by na-tional institutions, in particular the institutions participating in the
Gaia
Mul-tilateral Agreement. This work has made use of BaSTI web tools. This workincludes data from the Asteroid Terrestrial-impact Last Alert System (ATLAS)project. ATLAS is primarily funded to search for near earth asteroids throughNASA grants NN12AR55G, 80NSSC18K0284, and 80NSSC18K1575; byprod-ucts of the NEO search include images and catalogs from the survey area. TheATLAS science products have been made possible through the contributions ofthe University of Hawaii Institute for Astronomy, the Queen’s University Belfast,the Space Telescope Science Institute, and the South African Astronomical Ob-servatory. Guoshoujing Telescope (the Large Sky Area Multi-Object Fiber Spec-troscopic Telescope LAMOST) is a National Major Scientific Project built bythe Chinese Academy of Sciences. Funding for the project has been providedby the National Development and Reform Commission. LAMOST is operatedand managed by the National Astronomical Observatories, Chinese Academyof Sciences. Funding for the Sloan Digital Sky Survey IV has been providedby the Alfred P. Sloan Foundation, the U.S. Department of Energy O ffi / Universityof Tokyo, the Korean Participation Group, Lawrence Berkeley National Lab-oratory, Leibniz Institut für Astrophysik Potsdam (AIP), Max-Planck-Institutfür Astronomie (MPIA Heidelberg), Max-Planck-Institut für Astrophysik (MPAGarching), Max-Planck-Institut für Extraterrestrische Physik (MPE), NationalAstronomical Observatories of China, New Mexico State University, New YorkUniversity, University of Notre Dame, Observatário Nacional / MCTI, The OhioState University, Pennsylvania State University, Shanghai Astronomical Obser-vatory, United Kingdom Participation Group, Universidad Nacional Autónomade México, University of Arizona, University of Colorado Boulder, Universityof Oxford, University of Portsmouth, University of Utah, University of Virginia,University of Washington, University of Wisconsin, Vanderbilt University, andYale University.
References
Allard, F., Homeier, D., & Freytag, B. 2012, Philosophical Transactions of theRoyal Society of London Series A, 370, 2765Andrews, J. J., Agüeros, M. A., Gianninas, A., et al. 2015, ApJ, 815, 63Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, ARA&A, 47, 481Bailer-Jones, C. A. L., Rybizki, J., Fouesneau, M., Mantelet, G., & Andrae, R.2018, AJ, 156, 58Baxter, R. B., Dobbie, P. D., Parker, Q. A., et al. 2014, MNRAS, 440, 3184Bianchi, L., Conti, A., & Shiao, B. 2014, VizieR Online Data Catalog, 2335, 0Carlberg, J. K., Cunha, K., Smith, V. V., & Majewski, S. R. 2012, ApJ, 757, 109Casey, A. R., Hogg, D. W., Ness, M., et al. 2016, arXiv e-prints,arXiv:1603.03040Catalán, S., Isern, J., García-Berro, E., et al. 2008, A&A, 477, 213Chen, X., Han, Z., Deca, J., & Podsiadlowski, P. 2013, MNRAS, 434, 186Costa, A. D., Canto Martins, B. L., Bravo, J. P., et al. 2015, ApJ, 807, L21Cutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003, VizieR Online DataCatalog, II / Article number, page 12 of 13. Werner et al.: An extremely hot white dwarf with a rapidly rotating K-type subgiant companion
Fitzpatrick, E. L. 1999, PASP, 111, 63Fuhrmann, K. 1998, A&A, 338, 161García Pérez, A. E., Allende Prieto, C., Holtzman, J. A., et al. 2016, AJ, 151,144Geier, S., Kupfer, T., Heber, U., et al. 2015, A&A, 577, A26Geier, S., Marsh, T. R., Wang, B., et al. 2013, A&A, 554, A54Gianninas, A., Kilic, M., Brown, W. R., Canton, P., & Kenyon, S. J. 2015, ApJ,812, 167Gray, R. O. & Corbally, Christopher, J. 2009, Stellar Spectral ClassificationHall, D. S. 1972, PASP, 84, 323Han, Z., Podsiadlowski, P., Maxted, P. F. L., Marsh, T. R., & Ivanova, N. 2002,MNRAS, 336, 449Hartman, J. D. & Bakos, G. Á. 2016, Astronomy and Computing, 17, 1Hayden, M. R., Recio-Blanco, A., de Laverny, P., Mikolaitis, S., & Worley, C. C.2017, A&A, 608, L1Heinze, A. N., Tonry, J. L., Denneau, L., et al. 2018, AJ, 156, 241Henden, A. A., Levine, S., Terrell, D., & Welch, D. L. 2015, in American Astro-nomical Society Meeting Abstracts, Vol. 225, American Astronomical Soci-ety Meeting Abstracts / / ff enroth, V., Miller Bertolami, M. M., et al. 2020, A&A, 638,A93Ritter, H. 1986, A&A, 168, 105Santander-García, M., Rodríguez-Gil, P., Corradi, R. L. M., et al. 2015, Nature,519, 63Schlafly, E. F. & Finkbeiner, D. P. 2011, ApJ, 737, 103Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525Schreiber, M. R., Gänsicke, B. T., Rebassa-Mansergas, A., et al. 2010, A&A,513, L7 Schuh, S. L., Dreizler, S., Deetjen, J. L., & Göhler, E. 2003, Baltic Astronomy,12, 167Shimansky, V. V., Borisov, N. V., Nurtdinova, D. N., et al. 2015, AstronomyReports, 59, 199Silvestri, N. M., Hawley, S. L., & Oswalt, T. D. 2005, AJ, 129, 2428Simon, T. & Drake, S. A. 1989, ApJ, 346, 303Sing, D. K., Holberg, J. B., Burleigh, M. R., et al. 2004, AJ, 127, 2936Stelzer, B., Damasso, M., Scholz, A., & Matt, S. P. 2016, MNRAS, 463, 1844Stelzer, B., Marino, A., Micela, G., López-Santiago, J., & Liefke, C. 2013, MN-RAS, 431, 2063Strassmeier, K. G. 2009, A&A Rev., 17, 251Tian, H.-J., El-Badry, K., Rix, H.-W., & Gould, A. 2020, ApJS, 246, 4Tian, H.-J., Liu, C., Carlin, J. L., et al. 2015, ApJ, 809, 145Tonry, J. L., Denneau, L., Heinze, A. N., et al. 2018, PASP, 130, 064505Unglaub, K. & Bues, I. 2000, A&A, 359, 1042Van der Swaelmen, M., Bo ffi n, H. M. J., Jorissen, A., & Van Eck, S. 2017, A&A,597, A68Van Winckel, H. 2018, arXiv e-prints, arXiv:1809.00871Werner, K., Deetjen, J. L., Dreizler, S., et al. 2003, in Astronomical Society ofthe Pacific Conference Series, Vol. 288, Stellar Atmosphere Modeling, ed.I. Hubeny, D. Mihalas, & K. Werner, 31Werner, K. & Dreizler, S. 1999, Journal of Computational and Applied Mathe-matics, 109, 65Werner, K., Dreizler, S., & Rauch, T. 2012, TMAP: Tübingen NLTEModel-Atmosphere Package, Astrophysics Source Code Library [recordascl:1212.015]Werner, K. & Herwig, F. 2006, PASP, 118, 183Werner, K., Rauch, T., & Kruk, J. W. 2015, A&A, 582, A94Werner, K., Rauch, T., & Kruk, J. W. 2017, A&A, 601, A8Werner, K., Rauch, T., & Kruk, J. W. 2018a, A&A, 609, A107Werner, K., Rauch, T., & Kruk, J. W. 2018b, A&A, 616, A73Werner, K., Rauch, T., & Reindl, N. 2019, MNRAS, 483, 5291Willems, B. & Kolb, U. 2004, A&A, 419, 1057Wilson, O. C. 1963, ApJ, 138, 832Wilson, O. C. 1968, ApJ, 153, 221Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868Zechmeister, M. & Kürster, M. 2009, A&A, 496, 577Zhao, G., Zhao, Y.-H., Chu, Y.-Q., Jing, Y.-P., & Deng, L.-C. 2012a, Research inAstronomy and Astrophysics, 12, 723Zhao, J. K., Oswalt, T. D., Rudkin, M., Zhao, G., & Chen, Y. Q. 2011, AJ, 141,107Zhao, J. K., Oswalt, T. D., Willson, L. A., Wang, Q., & Zhao, G. 2012b, ApJ,746, 144Ziegler, M., Rauch, T., Werner, K., Köppen, J., & Kruk, J. W. 2012, A&A, 548,A109Zorotovic, M., Schreiber, M. R., Gänsicke, B. T., et al. 2011, A&A, 536, L3n, H. M. J., Jorissen, A., & Van Eck, S. 2017, A&A,597, A68Van Winckel, H. 2018, arXiv e-prints, arXiv:1809.00871Werner, K., Deetjen, J. L., Dreizler, S., et al. 2003, in Astronomical Society ofthe Pacific Conference Series, Vol. 288, Stellar Atmosphere Modeling, ed.I. Hubeny, D. Mihalas, & K. Werner, 31Werner, K. & Dreizler, S. 1999, Journal of Computational and Applied Mathe-matics, 109, 65Werner, K., Dreizler, S., & Rauch, T. 2012, TMAP: Tübingen NLTEModel-Atmosphere Package, Astrophysics Source Code Library [recordascl:1212.015]Werner, K. & Herwig, F. 2006, PASP, 118, 183Werner, K., Rauch, T., & Kruk, J. W. 2015, A&A, 582, A94Werner, K., Rauch, T., & Kruk, J. W. 2017, A&A, 601, A8Werner, K., Rauch, T., & Kruk, J. W. 2018a, A&A, 609, A107Werner, K., Rauch, T., & Kruk, J. W. 2018b, A&A, 616, A73Werner, K., Rauch, T., & Reindl, N. 2019, MNRAS, 483, 5291Willems, B. & Kolb, U. 2004, A&A, 419, 1057Wilson, O. C. 1963, ApJ, 138, 832Wilson, O. C. 1968, ApJ, 153, 221Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868Zechmeister, M. & Kürster, M. 2009, A&A, 496, 577Zhao, G., Zhao, Y.-H., Chu, Y.-Q., Jing, Y.-P., & Deng, L.-C. 2012a, Research inAstronomy and Astrophysics, 12, 723Zhao, J. K., Oswalt, T. D., Rudkin, M., Zhao, G., & Chen, Y. Q. 2011, AJ, 141,107Zhao, J. K., Oswalt, T. D., Willson, L. A., Wang, Q., & Zhao, G. 2012b, ApJ,746, 144Ziegler, M., Rauch, T., Werner, K., Köppen, J., & Kruk, J. W. 2012, A&A, 548,A109Zorotovic, M., Schreiber, M. R., Gänsicke, B. T., et al. 2011, A&A, 536, L3