LAMOST J045019.27+394758.7, with peculiar abundances of N, Na, V, Zn, is possibly a Sculptor dwarf galaxy escapee ^††thanks: Based on data collected using HCT/HESP

Carbon stars are a special class of objects characterised by strong molecular bands of carbon such as C₂, CH and CN in their spectra. Many surveys were conducted to explore carbon stars such as the First Byurakan Spectral Sky Survey (Gigoyan et al., 1998), infrared objective-prism surveys (Alksnis et al. 2001, Automatic Plate Measuring survey (Totten & Irwin, 1998; Ibata et al., 2001), and references therein), and the Large Sky Area Multi-Object Fibre Spectroscopic Telescope (LAMOST, Wu et al. 2011; Bai et al. 2016; Ji et al. 2016). Many studies have shown that a large fraction of iron-deficient stars exhibit enhancement of carbon as well as neutron-capture elements and are the ideal candidates to study the mechanism(s) that produce these elements. The chemical peculiarity of these objects promoted many spectroscopic studies of this group. One of our primary goals is therefore to explore carbon enhanced stars with signatures of enhanced abundance of heavy elements from large surveys for conducting such studies.

Following the low-resolution spectroscopic analysis, Goswami (2005); Goswami et al. (2007); Goswami et al. (2010), identified a substantial fraction ($\sim$ 30%) of CH stars from the list of faint high-latitude carbon stars of Christlieb et al. (2001). Follow-up detailed chemical analysis of some of these objects have shown many of them to be carbon-enhanced metal-poor (CEMP) stars in a metallicity range [Fe/H] $<$ $-$1, comprising of CEMP-r/s, CEMP-s and CEMP-no stars (Goswami et al., 2006; Purandardas et al., 2019b; Goswami et al., 2021; Purandardas & Goswami, 2021a). Ji et al. (2016) could identify 894 carbon stars from LAMOST DR2, from the measurement of the line indices of carbon molecular lines using spectra with R $\sim$ 1800, and S/N $>$ 10, and classified them into various spectral sub-classes such as C-N, C-R, and C-H. We performed a detailed chemical analysis for a few objects from this list and showed that the estimated abundances of the CEMP-r/s star LAMOSTJ151003.74+305407.3 (hereafter J151) could be well explained by the model yields ([X/Fe]) of i-process nucleosynthesis of heavy elements, and LAMOSTJ091608.81+230734.6 (hereafter J091) was identified as a CH giant (Shejeelammal et al., 2021). A parametric model based analysis have also shown that the model yields ([X/Fe]) of i-process nucleosynthesis of heavy elements could well explain the estimated abundance patterns of J151. The estimated neutron-density-dependent [Rb/Zr] ratio confirmed former low-mass asymptotic giant branch companions for these two objects. Our kinematic analysis have shown that J151 is a halo object and J091 belongs to the disk population (Shejeelammal et al., 2021).

We report here the results based on the high-resolution abundance analysis of another object J045 from Ji et al. (2016) which is classified as a carbon star with no identifiable sub-type. Xiang et al. (2019) derived the atmospheric parameters and elemental abundances for O, C, N, Mg, Na, Si, Ti, Ca, Cr, Co, Mn, Ni, and Ba based on low-resolution (R$\sim$ 1800) spectroscopic analysis and using data driven models. Based on near-infrared (15140-16940 Å  ) spectroscopy using a spectra at a resolution of R ${\sim}$ 22,500, Jonsson et al. (2020) determined the atmospheric parameters as well as these elemental abundances except for Ba using APOGEE Stellar Parameter and Chemical Abundance pipeline (ASPCAP). While Xiang et al. (2019) estimated the atmospheric parameters T_eff, log g, microturbulent velocity ${\zeta}$, and metallicity [Fe/H] for this object as (4627 K, 2.26, 0.72 km s^-1, $-$0.90), these estimates derived by Jönsson et al. (2020) are respectively (4250 K, 1.00, 1.55 km s^-1 , $-$0.50). The large difference in the estimates of the atmospheric parameters by these two groups have prompted us to re-visit this object. Although some of the aspects were discussed in these two studies, in this work, the results from a high-resolution spectroscopic analysis (R ${\sim}$ 60 000) of this object is presented for the first time. In addition to the elements for which the abundances are available in these two studies, we have determined the abundances for a few more elements such as Eu, Sm, Nd, Pr, La, Sr, Y, and Zn. Our analysis, however, shows that this object does not exhibit the spectral properties of carbon stars. From a visual inspection of its spectrum, the molecular bands due to carbon are found to be very weak, and also enhanced signature of neutron-capture elements were absent in the spectrum. We find a metallicity [Fe/H] = $-$1.05 for this object with elemental abundances that are quite unusual from that expected for Galactic metal-poor stars of similar metallicity. The observed abundance pattern is however found to be compatible with the abundance patterns observed in Sculptor Dwarf galaxy stars, indicating that the object J045 is possibly a Sculptor Dwarf Galaxy escapee. We have also examined for signatures of any internal mixing in this object using ¹²C/¹³C and [C/N] ratios that were not discussed in the previous works.

The paper is arranged as follows: Section 2 discusses observation and data reduction. Estimation of stellar atmospheric parameters, mass, and age determination are discussed in Section 3. A discussion on the determination of abundances is presented in Section 4. Uncertainty in the abundance estimates is discussed in Section 5. Results of the TESS photometry for the programme star is presented in Section 6. Results obtained from the kinematic analysis are discussed in Section 7. A discussion on the abundance analysis is given in Section 8 and the conclusions are presented in Section 9.

The object J045 is taken from the sample of carbon stars listed in LAMOST DR2 (Ji et al., 2016). The high-resolution (R $\sim$ 60,000) spectrum of the object is obtained from the Indian Astronomical Observatory (IAO), Hanle, using the HESP (high-resolution fiber-fed Hanle Echelle Spectrograph (HESP)) attached to 2m HCT (Himalayan Chandra Telescope). We have acquired two frames with exposure 2700s each. The spectra are combined to get the final spectrum with high S/N ratio. The spectrum covers from 3530 to 9970 Å in the wavelength range.

The spectrograph permits a resolution of 30,000 without slicer, and a resolution of 60,000 with slicer. The spectrum is recorded on a CCD with 4096$\times$4096 pixels, each pixel of 15 micron size. The data reduction is performed following the standard procedures using spectroscopic data reduction packages of IRAF¹¹1IRAF is distributed by the National Optical Astronomical Observatories, which is operated by the Association for Universities for Research in Astronomy, Inc., under contract to the National Science Foundation. Basic data of the programme star is presented in Table 1. Examples of sample spectra of J045 are shown in Figure 1.

Radial velocity of J045 is calculated from the shift in the observed wavelength from the lab wavelength for clean and unblended lines and is found to be $-$65.1$\pm$0.05 km s^-1 (Table 2). Jönsson et al. (2020) estimated a radial velocity of ${\sim}$ $-$39.9 km s^-1 for this object comparing its stellar spectrum with a grid of synthetic spectra. The estimated radial velocity of the object obtained using the LAMOST 1D pipeline by Luo et al. (2015) is ${\sim}$ $-$70.1 km s^-1. Our estimate of radial velocity ($-$65.1 km s^-1) derived using several clean unblended lines on the spectra of this object is distinctly different from the estimate of Jönsson et al. (2020) and closer to the estimate of Luo et al. (2015). The difference in the estimates of radial velocity indicates that the object could possibly be in a binary system.

Atmospheric parameters, the effective temperature T_eff, surface gravity, log g and microturbulent velocity ${\zeta}$ of J045 are estimated using the equivalent widths measured for unblended and clean lines of Fe I and Fe II lines (Table A1). We have made use of thirty four Fe I and two Fe II lines for our analysis. The line information are taken from linemake ²²2linemake contains laboratory atomic data (transition probabilities, hyperfine and isotopic substructures) published by the Wisconsin Atomic Physics and the Old Dominion Molecular Physics groups. These lists and accompanying line list assembly software have been developed by C. Sneden and are curated by V. Placco at https://github.com/vmplacco/linemake. (Placco et al., 2021), an atomic and molecular line database.

The effective temperature is fixed at the value by an iterative process for which the trend between the abundances derived using Fe I lines and the excitation potentials corresponding to those lines gives a slope close to zero. At this temperature, microturbulent velocity is adopted to be that value for which the Fe I lines abundances and the reduced equivalent widths do not exhibit any trends. At these values of effective temperature and microturbulent velocity, the surface gravity log g is fixed at the value, for which the abundances of Fe I and Fe II are nearly the same. The iron abundances of J045 as a function of excitation potential and as a function of equivalent widths are shown in Figure 2. The detailed procedure followed can be found in our earlier papers Purandardas et al. (2019b); Shejeelammal et al. (2021). We used MOOG (Sneden 1973, updated version 2013) for our analysis considering local thermodynamic equilibrium. We have selected the model atmospheres from the Kurucz grid of model atmospheres with no convective overshooting (http://kurucz.harvard.edu/grids.html). Solar abundances are taken from Asplund et al. (2009). The derived atmospheric parameters for J045 are presented in Table 3.

Elemental abundances are estimated from the equivalent widths measured for a number of good lines due to various elements. We have also performed spectrum synthesis calculations whenever found applicable. Various lines are identified from the overplot of the Arcturus spectrum on the spectrum of J045. A master line list was generated including the equivalent width measurements, lower excitation potential of the lines and log gf values obtained using linemake.

We have used only clean and symmetric lines for the measurement of equivalent widths. We could estimate the abundances for 25 elements as presented in the Table 4. We have performed the spectrum synthesis calculation for the elements that show hyperfine splitting, such as, Sc, V, Mn, Co, Ba, La, and Eu.

In the Table 4, we have presented the abundance results. The lines used for the determination of elemental abundances are tabulated in Table A2. Before presenting our discussion on elemental abundances, we present in the following, a discussion on verification of abundance estimates.

Random and systematic errors produce uncertainties in the estimated elemental abundances. Random errors are caused by the uncertainties in the line parameters like equivalent widths, blending of lines and oscillator strength. The systematic error is produced due to the uncertainties in the stellar atmospheric parameters adopted for the analysis.

We have calculated these uncertainties of our abundance estimates following the steps as discussed in Shejeelammal et al. (2021). The total uncertainties in the abundance estimates, log$\epsilon$ is given by

$\sigma^{2}_{log\epsilon}=\sigma^{2}_{ran}+(\frac{\partial log\epsilon}{\partial T})^{2}\sigma_{T_{eff}}^{2}+(\frac{\partial log\epsilon}{\partial logg})^{2}\sigma^{2}_{logg}+(\frac{\partial log\epsilon}{\partial\zeta})^{2}\sigma^{2}_{\zeta}+(\frac{\partial log\epsilon}{\partial[Fe/H]})^{2}\sigma^{2}_{[Fe/H])}$

where $\sigma_{ran}$ = $\sigma_{s}/\sqrt{N}$ . $\sigma_{s}$ indicates the standard deviation in the abundance estimates which is calculated from the number of lines, N, corresponding to that particular element. The $\sigma$’s represent the uncertainties in the adopted atmospheric parameters of the star, and are given by T_eff $\sim$ $\pm$ 100 K, log g $\sim$ $\pm$ 0.2 dex, $\zeta$ $\sim$ $\pm$ 0.2 km s^-1 , and [Fe/H] $\sim$ $\pm$ 0.1 dex. The uncertainty in [X/Fe] is calculated using the relation :

The differential elemental abundances obtained for J045 is presented in Table 8.

We have used TESS photometric data to search for variability in J045. The Transiting Exoplanet Survey Satellite (TESS), the first high-precision full-sky photometry survey in space observed this LAMOST target (TIC 187263688) from September 28, 2019 to December 23, 2019, in sector 19 with Camera 1 CCD 3, exposure time of 1426s and 30-miniutes cadence. We acquried TESS Full Frame Images using Lightkurve (Lightkurve Collaboration et al., 2018) which utilizes the TESSCut tool (Brasseur et al., 2019). The target is near the limiting magnitude of TESS and the data is quite noisy. Since our target is quite faint, we have created appropriate aperture as shown in Figure 8. To account for the instrumental and noise effects, we have applied regression correction to our light curve to remove trends, and set the light curve to the mean level. TESS scattered light is removed by subtracting model of the background that is built in Regression Correction. As seen from the Figure 9, we do not see any signatures of variability. Nonetheless, we employed Lomb-Scargle algorithm commonly used for detecting and characterising periodic signals in un-evenly sampled data (Lomb, 1976; Scargle, 1982). Lomb-Scargle periodogram is shown in Figure 9. The Lomb-Scargle algorithm performs computation of a Fourier-like power spectrum from un-evenly sampled data allowing determination of the period of oscillation. No period of oscillation could be detected from the limited TESS data set.

Kinematic analysis of J045 is performed following the detailed procedure as described in Purandardas & Goswami (2021b). The space velocity is derived using a right-handed coordinate system. In this system, U, V and W the three components of the space velocity are such that in the direction of the Galactic centre U is positive, V is positive in the direction of the Galactic rotation and W is positive in the direction of the North Galactic Pole (Johnson & Soderblom, 1987). The values of parallax and proper motion of the star are taken from Gaia Collaboration et al. (2016, 2018) and we have used our estimate of radial velocity for the calculation.

We have also checked whether the programme star belongs to the thin disk, thick disk or the halo population. We have followed the procedures as discussed in Reddy et al. (2006); Bensby et al. (2003, 2004); Mishenina et al. (2004), and the assumption that the Galactic space velocity of the star has Gaussian distribution:

Where $K=\frac{1}{(2\pi)^{\frac{3}{2}}\sigma_{U}\sigma_{V}\sigma_{W}}$
The values for characteristic velocity dispersion $\sigma_{U}$, $\sigma_{V}$ and $\sigma_{W}$ and the asymmetric drift $V_{asy}$ are adopted from Reddy et al. (2006). We have presented the results of the kinematic analysis of the programme star in Table 9. Chen et al. (2004), have shown that the thin disk objects have spatial velocity $<$ 85 km s^-1 and thick disk objects have spatial velocity in the range 85 - 180 km s^-1. Our estimated space velocity ${\sim}$ 78 km s^-1 therefore indicates that the object J045 to be a thin disk object. On the other hand, from high-resolution spectroscopic study of northern sky dwarf stars, Mikolaitis et al. (2019) have found that the stars with a lower combined velocity V_spa $<$ 50 km s^-1 (V_spa = (U_LSR+V_LSR+W_LSR)^1/2) are most probably the thin-disk stars, whereas those with v_spa $>$ 50 km s^-1 should belong to the thick-disk. Although following Mikolaitis et al. (2019)’s definition J045 is likely a thick disk object, if we consider the criteria that all stars with p_thick/p_thin $>$ 2 are thick-disk candidates and stars with p_thick/p_thin $<$ 0.5 are thin disk objects (Bensby et al., 2003, 2004), then, the J045 with p_thick/p_thin $<$ 0.5 seems to belong to thin-disk population.

The object J045 is included as a possible carbon star in LAMOST DR2 catalog, however, our detailed analysis shows that this object does not show any spectral characteristics of carbon stars and has a carbon abundance of near-solar value. The object is found to be a metal-poor giant with metallicity [Fe/H] $\sim$ $-1.05$ with unusual abundance for several key elements.

While the abundances of C is near-solar, N is found to be high with [N/Fe] = 0.61. One of the possible reasons for the observed enhancement of nitrogen in giant stars is the mixing which brings the CNO processed material to the surface that makes the atmosphere of the star "N rich" and "C poor" (Spite et al., 2005). However, as discussed below, the program star does not show any signatures of mixing based on [C/N] and ¹²C/¹³C ratios. However, if we consider the lower limit of error in [C/N] ratio, the star is well mixed. The origin of enhancement of nitrogen in a giant star can also be interpreted based on several scenarios. One such scenario is the accretion of N rich material from an AGB star (Lennon et al., 2003). However, the observed abundance patterns in the program star does not support any AGB companion as the possible source of the nitrogen enhancement. Early accretion of Globular Clusters or dwarf spheroidal galaxies can also lead to nitrogen enhancement (Fernández-Trincado et al., 2019). This scenario makes the star chemically distinguishable from the stars that belongs to disk, halo, or bulge population. The dynamical history of such stars is not yet clearly understood (Fernández-Trincado et al., 2019).

O is found to be marginally under-abundant with [O/Fe] $\sim$ $-$0.13, a value which is highly unusual for a Galactic star of metallicity [Fe/H] ${\sim}$ $-$1.0. None of the ${\alpha}$-elements Mg, Si and Ca are compatible with [${\alpha}$/Fe] - enhancement (i.e., 0.3 to 0.4 dex) characteristic of Galactic metal-poor stars of similar metallicity. J045 shows almost solar values with [Mg/Fe] = 0.08, [Si/Fe] = 0.18, and [Ca/Fe] = 0.02. Such abundance pattern is however compatible with stars of similar metallicity of Sculptor Dwarf Galaxy as shown in Figure 10 (upper panel). This comparison provides observational evidence to argue that the object could be a possible Dwarf Galaxy escapee in the Galaxy. Finding globular cluster (GC) escapees are not uncommon in the Galactic halo. While Martell et al. (2011) give an estimate of GC escapees in the halo to be $\sim$ 3% , an estimate of Dwarf galaxy escapees is not available in the literature.

Na is also enhanced with [Na/Fe] $\sim$ 0.68. Such high abundance values for Na are generally not found to be reported for normal metal-poor Galactic stars of similar metallicity (Figure 11). Among the Dwarf Sculptor Galaxy stars one object UET0130 with a metallicity of $-$2.2 is known to show very high Na abundance with [Na/Fe] = 0.84 (Figure 12).

Among the Fe-peak elements, the Mn abundance is found to be near-solar, whereas, Co and Ni are found to be marginally enhanced . Sc and Ti are found to be under-abundant with [Sc, Ti/Fe] ${\sim}$ $-$ 0.28 and $-$0.25 respectively. As seen from Figure 10 (upper panel) these values match closely with Sculptor Dwarf Galaxy stars.

The [$\alpha$/Fe] ratios observed in J045 are generally noticed in dwarf spheroidal galaxies at closer metallicity (Koch & McWilliam, 2008; Kirby et al., 2009). The observed [Sc/Fe] ${\sim}$ $-$0.28 and [Ti/Fe] ${\sim}$ $-$0.25 in J045 are also seen as typical values in many Sculptor Dwarf Galaxy stars (Skúladóttir et al., 2017; Hill et al., 2019).

The abundance of V with [V/Fe] = 0.51 is higher than generally noticed in metal-poor Galactic stars. In a recent work, Ou et al. (2020) have derived abundance of Vanadium from a sample of 255 metal-poor stars in the metallicity range $-$4.0 ${\leq}$ [Fe/H] ${\leq}$ $-$1.0, and found that [V I/Fe] = $-$0.10$\pm$ 0.01 from 128 stars and [V II/Fe] = 0.13$\pm$ 0.01 from 220 stars. The authors suspect this offset to be due to departures from LTE and recommend using [V II/Fe II] values which is enhanced relative to solar ratio. We could not compare our estimate of vanadium abundance with Sculptor Dwarf Galaxy stars as literature values are not available.

The derived abundance of Zn with [Zn/Fe] = $-$0.62 is found to be very different from its counterparts observed in Galactic giants and halo stars of similar metallicity. If we take account of the Non-LTE correction ($<$ 0.1 dex) zinc abundance will slightly enhance but still remains distinctly different from the Galactic trend. Saito et al. (2009) studied the trend of [Zn/Fe] with respect to metallicity in the range $-$4.2 $\leq$ [Fe/H] $\leq$ +0.5, considering a large sample of stars (434 stars) belonging to the population of Galactic thin and thick disk as well as halo, and confirmed a trend where [Zn/Fe] decreases with increase of [Fe/H]. [Zn/Fe] was found to follow a nearly flat trend in the range of metallicity $-$2.0 $\leq$ [Fe/H] $\leq$ +0.5, with a slight enhancement of $\sim$ +0.007 dex upto [Fe/H ] $<$ $-$1.0. The enhancement was found to be more pronounced in the range $-$1.0 $\leq$ [Fe/H] $\leq$ $-$0.5, and the trend is found to gradually decrease to solar value in [Fe/H] $>$ $-$0.5. This observed trend of [Zn/Fe] in the Galactic stars is very different from those observed in Sculptor Dwarf Galaxy (dSph) stars. As shown by Skúladóttir et al. (2017), the Sculptor Dwarf galaxy stars show a decreasing trend of [Zn/Fe] ratios at [Fe/H] $>$ $-$1.8 with a large star-to-star scatter ranging from $-$0.8 to $<$ 0.4. A comparison of the Zn abundance of J045 with the counterparts in dSph stars with similar metallicities (Figure 11) have shown a reasonably good match indicating that J045 could be a possible Sculptor Dwarf Galaxy escapee.

According to Skúladóttir et al. (2017), the [Zn/Fe] trend in the metallicity range $-$2.5 $<$ [Fe/H] $<$ $-$1.0 observed in the Sculptor Dwarf galaxy stars can be attributed to SNe Ia that causes the slope of [Zn/Fe] to be steeper than that in a lower metallicity. Since SNe Ia is not able to produce Zn, stars formed out of a medium contaminated by the ejecta of SNe Ia are expected to show low [Zn/Fe] values and in particular, the stars with [Zn/Fe] $<$ $-$0.5 are believed to be formed from a medium which is predominantly influenced by the ejecta of SNe Ia.

Based on chemodynamical simulations of dwarf galaxies, Hirai et al. (2018) suggested that the electron-capture supernovae (ECSNe) can be one of the sources that can produce Zn as they could not justify the observed trends if ECSNe is not considered as a source of Zn.

The abundance of neutron-capture elements Sr, Y, Ba, Pr, Sm are near-solar and La is marginally enhanced in J045. The abundance of Ce and Nd are marginally under-abundant, whereas Eu is mildly enhanced. In figures 10 and 13 we have compared the estimated abundance ratios of light and heavy elements in J045 with the corresponding ratios measured in normal Galactic giants and Sculptor dwarf galaxy stars. The observed abundance pattern in J045 is found to match well with that of Sculptor dwarf galaxy stars.

We have examined if the accretion of material from a binary companion could be a possible explanation for the non-typical abundance pattern observed in J045. The CH, CEMP-s, and CEMP-r/s stars are metal-poor stars that are believed to have been received gas transferred from an AGB companion. However, with an estimated [C/Fe] ${\sim}$ 0.11, the star is neither carbon-enhanced nor s-process enhanced with [X/Fe] ${\sim}$ solar values, where X stands for heavy elements.

This work presents the results of a detailed spectroscopic analysis of the object J045 using high resolution spectra. The program star is presented in the list of possible carbon stars of LAMOST DR2 catalog. However, from our analysis we show that the program star does not show any spectral properties of a typical carbon star. From our analysis, we find that J045 is a metal-poor giant of metallicity [Fe/H] $-$1.05 with very unusual elemental abundances. The observed abundance pattern is not compatible with abundances characteristic of Galactic metal-poor stars of similar metallicity, but, matches quite closely with those observed in Sculptor Dwarf galaxy stars. This observational evidences prompted us to argue that J045 is likely a Sculptor Dwarf Galaxy escapee.

While the carbon abundance is estimated to be near-solar, N is over abundant with [N/Fe] $\sim$ 0.61. Unlike in Galactic metal-poor stars, oxygen is found to be marginally under-abundant with [O/Fe] $\sim$ $-$0.13 and $\rm<[\alpha/Fe]>$ $\sim$ 0.09. Na is found to be slightly enhanced with [Na/Fe] $\sim$ 0.68. Sc and Ti are found to be under-abundant. While the Fe-peak element Mn is determined to be near-solar, Ni and Co are estimated to be marginally enhanced. Among the neutron-capture elements, Eu is marginally enhanced with [Eu/Fe] $\sim$ 0.29. Sr, Y, Ba, Pr, and Sm show near-solar values. While La is marginally enhanced, Ce and Nd are found to be marginally under-abundant. With the estimated [Ba/Eu] $\sim$ $-$ 0.38, the star falls into the category of neutron-capture-rich star of r-I type of Beers & Christlieb (2005).

The estimated luminosity is high with log(L/L_⊙) ${\sim}$ 2.27. Such high values would indicate that the star might have undergone internal mixing. However, the estimates of [C/N] and ¹²C/¹³C ratios show that the object has not undergone any significant internal mixing. We however note that, if we consider the lower limit of error in [C/N] ratio, the star is well mixed. From the kinematic analysis, we find that the program star is a thin disk object.