LAMOST J040901.83+329355.6 -- a new Galactic star with Wolf--Rayet characteristics on a post-AGB to CSPN transitional stage

The similarity in physical conditions in winds of low-mass post-asymptotic giant branch stars and evolved massive stars leads to the appearance of an interesting phenomenon of spectral mimicry. Due to that the discovery of every new star with Wolf--Rayet spectrum requires special study of its evolutionary status before it may be included in the list of Galactic Wolf--Rayet (WR) stars. A couple of years ago LAMOST J040901.83+323955.6 (hereafter J0409+3239) was selected as a WR star in LAMOST spectroscopic database by machine learning methods. In this work we investigated its evolutionary status. Analyzing the spatial location of J0409+3239, in the Galaxy and its position in the color-magnitude diagram we concluded what J0409+3239, is instead a low mass object with WR phenomenon. Its luminosity is $L*=1000 L_\odot$ and effective temperature $T_{\rm eff}$=40,000 K. Using new and archival photometric data we detected irregular variability on time scales from hours to tens of days with amplitude up to $\approx0.2$ mag. Comparison of the spectrum obtained in 2022 with the one from 2014 also shows an evidence of spectral variability. The absence of clearly detected circumstellar nebula does not allow to classify J0409+3239, as [WR], i.e. a central star of planetary nebula (CSPN). However, its position in Hertzsprung--Russell diagram suggests that J0409+3239, is a low mass star caught in rare transitional phase to CSPN. Estimation of J0409+3239, mass based on evolutionary tracks shows that it is less than $0.9~M_\odot$, and thus the age of the Galaxy is barely enough for the star to evolve to its current stage.


INTRODUCTION
Classical Wolf-Rayet (WR) stars are the final points in the life of massive stars, which massive stars may reach either through the Conti scenario (in the case of single stars, Conti (1975Conti ( , 1984))), or through the mass transfer in binary systems (Paczyński 1967).WR stars are the stripped cores of evolved massive stars that lost their external envelopes, have spectra enriched with CNO-cycle products, and burning helium in their cores.Despite being divided into three distinct groups (nitrogen rich WN, carbon rich WC and rare oxygen rich WO), in Hertzsprung-Russell (HR) diagram all WR stars are located in a small area in upper left corner.
WR stars are known for more than 150 years, and their properties are sufficiently well studied (Crowther 2007).However there are still open questions about how the formation of WR stars is affected by the metallicity, and the percentage of binary evolution products among them (Neugent & Massey 2019;Shenar et al. 2020).Comparison of observed WN/WC ratio with theoretical predictions of single and ★ E-mail: olga.maryeva@asu.cas.czbinary evolutionary scenarios is a good way to test the theory of stellar evolution (Maeder & Meynet 1994;Massey et al. 2015;Rosslowe & Crowther 2015;Neugent & Massey 2019).
In our Galaxy there are 669 WR stars currently discovered 1 (Rosslowe & Crowther 2015), while the theory based on current Milky Way star formation rate and duration of WR phase predicts the numbers of WR stars ∼ 1200 (Rosslowe & Crowther 2015).It means that significant part of WR stars are still not discovered, and the WN/WC ratio may change significantly.WR stars in the Galaxy align with spiral arms and the regions of star formation.Thus, the discovery of new WR stars through optical observations has been significantly limited due to the presence of dust extinction.Although the probability of finding a WR star during optical spectroscopic surveys remains (Maíz Apellániz et al. 2016;Zhang et al. 2020), the majority of discoveries of new WR stars now happens through the observations in infrared (IR) range.For example, Mauerhan et al. (2011) identified 60 Galactic WR stars: candidates were selected using the photometry from Spitzer and Two Micron All Sky Survey (2MASS) databases and confirmed by near-IR spectroscopy.Shara et al. (2012) expanded the list by adding 71 more stars, also preliminary selected from  and  band 2MASS photometry.
Moreover, the search for new WR stars is complicated by the effect of spectral mimicry -in addition to classical WR stars there is also WR phenomenon, happening when fast-moving, hot plasma is expanding around a hot star (Gräfener et al. 2011;Vink 2015).The mass loss rate or rather the density of the expanding envelope and the temperature are the critical basic parameters which are responsible for the WR phenomena (van der Hucht et al. 1981), and due to that not only evolved massive stars during WR stage may show it.Besides classical WR stars, WR phenomenon is also observed in young supernovae (SN) Gal-Yam et al. (2014) and [WR] objectscentral stars of planetary nebulae (CSPNe) (van der Hucht et al. 1981;Marcolino et al. 2007;Todt 2009;Todt & Hamann 2015), hydrogen deficient post-AGB stars, coming through "born-again" evolutionary scenario (Iben et al. 1983;Werner & Herwig 2006) and typically having luminosities less than 25000 L ⊙ (Weidmann et al. 2020).The most striking example of the object with WR phenomenon is the famous SS 433 object -Galactic microquasar, binary system with black hole candidate (Fabrika 2004), with a Bowen C III /N III blend which is typical for WR stars clearly visible in the spectrum.
The [WR] objects masquerading as classical WR stars is a particular case of such mimicry between low mass stars that already passed through asymptotic giant branch (post-AGB), and hypergiants.Similarity of wind properties (luminosity  * to mass  * ratio of a star, mass loss rate ) is the reason for spectral mimicry as it was discussed a lot in the works of V. Klochkova and E. Chentsov (Klochkova et al. 1997, 2007, 2014), as well as by Lamers et al. (1998).Thanks to the results of Gaia mission (Gaia Collaboration et al. 2016) we now have reliable estimations of distances, and it helps to quickly understand the nature of individual objects and separate low-mass evolved stars from massive stars, and, moreover, to find the objects of unusual origin among low mass stars.For example, based on Gaia distance estimation Gvaramadze et al. (2019) demonstrated that IRAS 00500+6713 is a product of merging of two white dwarfs, while the star has WO-type spectrum.Reverse example is PMR 5 star classified as [WN6] by Morgan et al. (2003) that was recently reclassified as a classical WN (Todt 2023).
The present paper is devoted to study of LAMOST J040901.83+323955.6 (hereafter J0409+3239) which is a star initially classified as WN by Škoda et al. (2020).However, close distance to it and its low brightness at same time suggest that J0409+3239 is most probably the low mass object with WR phenomenon.The paper is organised as follows.In Section 2 we present the object of studies, in Sections 3 and 4 we describe the archival and newly acquired spectroscopic and photometric data, as well as variability of J0409+3239 .In Section 5 we consider spatial position of J0409+3239 in the Galaxy and its location in the HR diagram, discuss its initial mass and current age.Last Section 6 presents the conclusion.

HISTORY OF STUDIES OF J040901.83+323955.6
The first spectral observation of J0409+3239 was done with Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) during LAMOST Spectroscopic Survey of the Galactic Anticentre (Yuan et al. 2015).Škoda et al. (2020)  The object was also selected in several studies based on photometric data.Sesar et al. (2017) added J0409+3239 to the list of RR Lyrae stars using machine-learning identification method and multi-epoch, asynchronous multi-band photometric data from Panoramic Survey Telescope and Rapid Response System (Pan-STARRS).For J0409+3239 Sesar et al. (2017) found the period P=0.2847409137 d.Moreover, J0409+3239 was included in the first catalog of variable stars of The All-Sky Automated Survey for Supernovae (ASAS-SN) survey Jayasinghe et al. (2018).There the star is mentioned as a non-periodic object with =14.48mag and amplitude of 0.39 mag.
Table 1 presents the coordinates of the object, distance estimations according to Gaia third Data Releases (DR3; Bailer-Jones et al. ( 2021)), photometric data from different catalogs and an estimation of interstellar reddening  ( − ) in the direction of J0409+3239 .For estimation of  ( − ) we used the distance from Gaia DR3,  and a three-dimensional map of dust reddening2 (Green et al. 2019).Taking into account the distance and  ( −) we estimated absolute visual magnitude   of J0409+3239 and determined its location in the colour-magnitude diagram (Figure 1). Figure 1 clearly shows that J0409+3239 lies in the region of low mass stars and the object significantly differs from WR stars by magnitude.Exotic products of binary evolution mentioned above like galactic microquasar SS 433 (Fabrika 2004) or two white dwarfs merger product IRAS 00500+6713 (Gvaramadze et al. 2019) are also significantly more luminous.Such a large difference in luminosity cannot be explained by interstellar absorption, because the total absorption in the Galaxy in this direction does not exceed 1 mag (Schlegel et al. 1998;Schlafly & Finkbeiner 2011).As we will show in Section 5.2, incorrect determination of the distance also cannot explain it.Therefore we may conclude that J0409+3239 is not a bona fide Wolf-Rayet star, but a low mass object with WR phenomenon.

SPECTROSCOPY
New low resolution spectrum of J0409+3239 was obtained in August 24, 2022 with the Russian 6-m telescope with the SCORPIO-2 multimode focal reducer3 (Afanasiev & Moiseev 2011) with the VPHG 1200@540 grism, which provided a spectral range of 3650-7200 Å and a spectral resolution    ≈ 4.5Å with 1 arcsec slit width.New E2V CCD261-84 (2K×4K) was used as a detector (Afanasieva et al. 2023).The spectrophotometric standard star G191B2b was observed for flux-calibration on the same night.Data reduction was done using IDL-based packages as it was described in our previous papers (e.g.Maryeva et al. 2022).Figure 2 shows two spectra of J0409+3239 -the new one and previous spectrum obtained with LAMOST on January 3rd, 2014.The joint observational log is given in Table 2.
The spectrum of J0409+3239 contains strong emissions of C II 4267 and C III 4647, 4650, 4652 lines.The last one were identified as N III by Škoda et al. (2020) and as C III +N III by Sun et al. (2021), however due to the absence of other nitrogen lines we unambiguously identify the bump at 4650 Å as C III lines.The He II 4686 is the strongest line in the spectrum, it is significantly stronger than another He II 5412 line.Also broad emission lines of C IV 5801, 5812 and weak He I are visible.Presence of helium and carbon lines in the spectrum is a clear evidence for far evolved stage of the star.Besides carbon and helium, there are also hydrogen lines of both Balmer and Paschen series in the spectrum.H has double-peaked profile in the LAMOST spectrum.Due to lower resolution of SCORPIO-2 we do not resolve this feature in the 2022 spectrum.We will discuss the nature of hydrogen lines in Section 5.In the spectrum there are also narrow absorptions of Ca II H/K and Na I D formed in the interstellar medium (ISM).Comparison of the spectra obtained in 2014 and in 2022 shows a clear difference -C III 5696 line weakened in 2022 in respect to the spectrum of 2014.Hajduk et al. (2015).The latter spectra are also continuum normalized, and shifted vertically for clarity.
To better understand the nature of an object we carried out its spectral classification.As we see that the star is an evolved lowmass object with WR phenomenon, for its spectral classification we used the criteria of Weidmann et al. (2020) who suggested the list of key spectral lines (Table 3) for preliminary determination of sub-type of CSPNe4 .Since most of the spectral features in the J0409+3239 's spectrum are observed in the emission, and there are no oxygen O and nitrogen N lines, we classified J0409+3239 as [WC] star.For more accurate classification we used the criteria from Crowther et al. (1998) based on equivalent width ratios of C IV 5801, 5812/C III 5696 and C IV 5801, 5812/C II 4267.According to these equivalent width ratios J0409+3239 is a [WC8-9].
Although according to formal criteria we may classify J0409+3239 as [WC8-9] stars, its spectrum significantly differs from the spectra of typical [WC9] stars, for example, Hen 2-142 and BD+303639 (Figure 3).The main difference is in the intensities of emission lines, that may be interpreted as a significant difference in mass loss rates.Moreover, nebular forbidden lines such as [O III] 4959, 5007 and [N II ] 5755 are absent in the J0409+3239 's spectrum.

PHOTOMETRY
In order to assess the photometric variability of J0409+3239 we utilized the data from Zwicky Transient Facility (ZTF) (Bellm et al. 2019) survey.From measurements collected between March 27, 2018 and February 19, 2023 in ZTF  and  filters photometric variability of J0409+3239 is clearly seen.We extended the light curve further back in time to better characterize the long-term photometric behaviour of J0409+3239 by combining ZTF data with the measurements from Pan-STARRS1 Survey, Asteroid Terrestrial-impact Last Alert System (ATLAS) project and ASAS, in the same way as we did it in Maryeva et al. (2022).Total light curve covering last 12 years is shown in Figure 4.The star does not display any significant change in its photometric behaviour in respect to what we see in ZTF data.
In order to better constrain the short time scale variability of the object color, we also initiated a series of dedicated observations of J0409+3239 on FRAM-ORM robotic telescope (Janeček et al. 2019), which is a 10-inch Meade f/6.3 Schmidt-Cassegrain telescope with custom Moravian Instruments G2-1600 CCD installed in the Roque de Los Muchachos Observatory, La Palma.The observations were performed since July till October 2022.The data were acquired in several sets every night, with several 120 s exposures in Johnson-Cousins  and  filters in every set, and then automatically processed by a dedicated Python pipeline based on the STDPipe package (Karpov 2021), which includes bias and dark current subtraction, flat-fielding, cosmic ray removal, astrometric calibration, aperture photometry, and photometric calibration using the catalogue of synthetic photometry based on Gaia DR3 low-resolution BP/RP spectra (Gaia Collaboration et al. 2022).The resulting light curve is shown in Figure 5.Its behaviour is essentially the same as in ZTF data, with irregular variability on the time scale of days to weeks and While the light curve is clearly variable in all data sets we collected, we were unable to find any clear periodicity in it.The autocorrelation functions of the data from different instruments computed using a method of Kreutzer et al. (2023) suitable for non-uniform time series are shown in Figure 7. Their shapes are slightly different due to different time intervals and temporal samplings used, but all clearly shows characteristic variability time scales of several tens of days.
We also constructed the color-magnitude diagrams using the data from ZTF and FRAM-ORM (see Figure 8).While not conclusive, they both show slight dependence of the bluest band ( for ZTF and  for FRAM-ORM) on the color, while reddest band is not correlated with it.Such behaviour suggests that the amplitude of variability is slightly larger in bluest bands, which is consistent with the root mean square values for corresponding light curves shown in Figure 5.

Location in Hertzsprung-Russell diagram and evolutionary status of J0409+323
We constructed the qualitative model for J0409+3239 atmosphere using cmfgen non-LTE radiative transfer code (Hillier & Miller 1998).To determine the effective temperature  eff we used intensities of C III 4647 and C IV 5801, 5812 lines, as well as ones of He I and He II .The model spectrum correctly reproduces the set of lines seen in the data, with the intensities close to observed ones, as well as overall spectral energy distribution (SED) (Figure 9).However, as we do not know the exact wind parameters, its terminal velocity, and chemical composition (such as He/H ratio and abundances of CNO elements) on such advanced evolutionary stage, we did not try to fit the exact profiles of individual spectral lines, and only estimated temperature of the star ( eff = 37000 − 41000 K) and its luminosity ( * = 900 − 1000 L ⊙ ).
By comparing an overall shape of the model spectrum with the photometric data acquired from several sky surveys (see Figure 9) in optical and near-infrared ranges we also estimated the reddening towards J0409+3239 to be  ( − )=0.6.It is 2.6 times larger than 06:00 07:00 08:00 09:00 10:00 11:00 12:00 Time, UT the total Galactic reddening in this direction (see Table 1).It suggests that additional extinction occurs in the circumstellar material around J0409+3239 .The presence of such material is also consistent with the infrared excess apparent in the SED as shown in Figure 9.To describe this infrared excess we added two blackbody components with  = 2700 K and  = 150 K on top of the cmfgen model.Thus we conclude that J0409+3239 is surrounded by dust envelopes, both warm and cold, likely formed due to past eruptive activity.
Figure 10 shows the location of J0409+3239 in HR diagram and evolutionary tracks for low mass stars on post-AGB stage (Miller Bertolami 2016).Numerical simulations of a post-AGB evolution show that stars become significantly hotter during approximately 1000 years after leaving the asymptotic giant branch, and rapidly move leftward in HR diagram (Miller Bertolami 2016) before finally settling down as slowly cooling CSPNe.Low luminosity of J0409+3239  * ≈ 1000 L ⊙ suggests that J0409+3239 is a very low mass star, near the limit for CSPN formation during the lifetime of the Galaxy.Indeed, the stars with initial mass less than 0.88 M ⊙ did not yet have enough time to leave the main sequence and evolve towards CSPNe (Saracino et al. 2016;Weidmann et al. 2020).
The spectrum of J0409+3239 displays prominent hydrogen lines (see Section 3).They are typical for late-type WN stars (Hamann et al. 2006;Crowther 2007), and thus were in agreement with the star' classification of Škoda et al. (2020).On the other hand, they are not consistent with our new classification of J0409+3239 as [WC8-9].However, the two-peaked shape of H with 5Å separation suggests that these lines are formed in the hydrogen-rich circumstellar shell surrounding the star that had already been ejected but did not yet expand significantly to be observable as a protoplanetary nebula like e.g. in Gómez-González et al. (2022).Shell expansion velocity derived from peak separation is about 120 km s −1 , which is significantly higher than the ones seen in nebular lines in planetary nebulae (20-30 km s −1 ), suggesting that the shell is still much smaller and denser.Therefore, we interpret the presence of hydrogen lines and absence of nebular lines in the spectrum of J0409+3239 as one more evidence that this object is in transition into [WR] stage right now.Indeed, it is already hot enough to be bona fide CSPN, but its spec-  trum still lacks nebular emission lines.Also, no signs of the nebula around it are seen in direct images in IR range, while the SED shows warm and cold thermal components most probably corresponding to the dust surrounding the star.

Position in the Galaxy
Our initial hypothesis about J0409+3239 being a low mass object with WR phenomenon was based on the distance estimation.As mentioned before, according to Gaia DR3 the distance to J0409+3239 is 2.5 kpc, which means that the object lies in the Perseus Galactic arm (Figure 11).There is no reason to believe that this distance estimation is wrong -indeed, while there is another spiral arm, an Outer Arm, in that direction, moving J0409+3239 there would only increase its luminosity by approximately 4 times, which is not enough to make it classical WR star.It is interesting to note that J0409+3239 is located just 21 degrees from Galactic anti-center direction, in a sector between 138.9 • -227.8 • Galactic longitude where no WR were ever found5 , which also indirectly supports that J0409+3239 is a low mass object with WR phenomenon.We considered the distance of J0409+3239 from the Galactic plane and compared its location with the distribution of Galactic WR and [WR] stars (Figure 12).While ∼ 50% of [WR]s are located above the Galactic plane (farther than ±0.35 kpc), only ∼ 3% of WRs are located outside of the thin Galactic disk.For J0409+3239 the distance above Galactic plane is -0.725 kpc, i.e. the star is a member of Galactic halo.

Variability
Analysis of photometric data in Section 4 clearly shows that J0409+3239 displays significant aperiodic variability on time scales from hours till tens of days, with amplitude slightly larger in bluer bands.We had not been able to find similar photometric behaviour being mentioned for other known [WR] stars in the literature.Thus, we checked the variability of CSPN stars classified as [WR] from the list of Weidmann et al. (2020) in ZTF data archive.We found the light curves having more than 100 photometric measurements for 19 stars including V348 Sgr, the variable classified as R Coronae Borealis (RCB) type.The latter is the only significantly variable object in the sample -the rest does not show any signs of a systematic variability comparable in amplitude or pattern to the one we see for J0409+3239 .On the other hand, the variability with similar amplitudes on similar time scales and with similar color-magnitude behaviour was previously detected in carbon-rich protoplanetary nebulae (Hrivnak et al. 2022), where it is attributed to the opacity variations of the circumstellar dust.
Spectral variability of CSPNe stars usually manifests in changes of emission line intensities and changes of P Cygni profiles.It was detected in both ultraviolet (Feibelman et al. 1992;Patriarchi & Perinotto 1995;Guerrero & De Marco 2013) and optical (Arkhipova et al. 2013;Kondratyeva et al. 2017) ranges.Unfortunately, for J0409+3239 no UV data are available, and the only variation in the optical spectra separated by 8 years is the change in the intensity of C III 5695.92line.Thus, we cannot yet relate it to either observed photometric variability or the evolutionary changes on the time scales of 8 years between the spectra, and additional spectral monitoring is definitely required.In general, we suggest that this peculiar variability, both photometric and spectral, confirms the transitional evolutionary stage of J0409+3239 .

CONCLUSION
We performed detailed study of J0409+3239 , originally selected from LAMOST spectral archive and classified as a WN by Škoda et al. (2020).Based on its luminosity, location in the Galaxy and position in color-magnitude and HR diagrams we speculate that J0409+3239 is on a transitional post-AGB stage evolving towards CSPN, with actually spectral type [WC8-9].The object displays peculiar pattern   of aperiodic variability with characteristic time scales from hours to tens of days which is not observed in other known CSPN, which we attribute to its earlier evolutionary stage.The absence of forbidden lines, presence of warm and cool thermal components in the SED, and double-peaked H also suggest that J0409+3239 very recently made transition between post-AGB and CSPN, did not yet develop a nebula, but is already surrounded by a shell-like structure which is typical for the protoplanetary nebulae.The very low initial mass of the object (less than 0.9 M ⊙ ) is on the limit of CSPN formation during the life time of the Galaxy.

Figure 2 .
Figure 2. Continuum normalized spectra of J0409+3239 obtained in January 2014 and August 2022.The spectra are vertically shifted for clarity.

Figure 4 .Figure 5 .
Figure 4.The light curve of J0409+3239 over last 12 years using the data from ZTF, ATLAS, ASAS-SN and Pan-STARRS sky surveys.

Figure 7 .
Figure7.Autocorrelation function of J0409+3239 lightcurves from different datasets constructed using the method ofKreutzer et al. (2023) as described in Section 4. The data from FRAM-ORM is best sampled and most uniform, and thus better captures short-term correlations.

Figure 10 .
Figure 10.Location of J0409+3239 in HR diagram.Filled orange squares mark the positions of known [WO] CSPNe, red ones -[WC] CSPNe, taken from Weidmann et al. (2020).Black solid lines show evolutionary tracks for objects of three different initial masses on H-burning post-AGB stage, while blue dash-dotted lines are corresponding isochrones for different ages since the moment of leaving the AGB, defined as log  eff = 3.85.Evolutionary tracks and isochrones are from Miller Bertolami (2016).

Figure 12 .
Figure 12.Distribution WRs and [WRs] from galactic plane.Orange squares are Galactic [WR] stars (list taken from (Weidmann et al. 2020)) distances for them taken from DR3, while grey circles -WR stars, according Galactic Wolf-Rayet Catalogue -distances for them taken from same database.Dashed line shows centre of Galactic plane, while dashed-doteed lines mark borders of the think disc at 0.35 kpc.

Table 1 .
selected J0409+3239 in machine learning based search for emission-line stars in LAMOST archive and classified it as a WN star.Independently Sun et al. (2021) Parameters of J040901.83+323955.6.

Table 3 .
Weidmann et al. (2020)pectral classification of CSPNe, according toWeidmann et al. (2020).The lines are marked as either absent in the spectrum of J0409+3239 , present in emission, or in a broad emission.
Color-magnitude diagrams constructed from ZTF (upper panels) and FRAM-ORM (lower panels) light curves.Slight trends are noticeable in the bluest ( and ) bands, while reddest ( and ) are uncorrelated.