betaFeedback

WorksPhysics and Astronomy

High-precision abundance analysis of a unique CEMP-no star in Sculptor

Abstract

The Carbon-Enhanced Metal-Poor stars (CEMP-no, low barium) are commonly accepted as the direct descendants of Population III (Pop III) stars which either ended their lives as faint supernovae (SNe; E_SN < 10^51 erg) or were fast-rotating stars (spinstars). While CEMP-no stars are common in the Milky Way (MW) and ultra-faint dwarf galaxies (UFDs), they are surprisingly rare in dwarf spheroidal galaxies (dSphs). This could hide an environmental dependence of the properties of Pop III stars, highlighting the importance of high-resolution studies of CEMP-no stars in dSphs. Here we present the first detailed study of a CEMP-no star with [Fe/H] < -3 in the Sculptor dSph. The high-resolution VLT/UVES spectra allowed us to derive one of the most precise abundance patterns obtained for a star belonging to a dwarf galaxy, providing chemical abundances for 17 elements and 4 upper limits. This star, DR20080, is extremely metal-poor, [Fe/H] = -3.26, and clearly enhanced in carbon, [C/Fe] = +1.60 (after correcting for internal mixing). Furthermore, it has low [Ba/Fe] = -1.28. The oxygen abundance of DR20080 ([O/Fe] = +1.53) is significantly higher than the average found in CEMP-no stars in the MW, while the iron-peak elements are under-abundant ([Mn/Fe] = -1.02; [Ni/Fe] = -0.45). The abundance pattern of DR20080 was compared to theoretical yield predictions of Pop III and Pop II SNe to infer the properties of its progenitor. The best result points toward a double Pop III progenitor scenario, with the major contributor to the abundance pattern being a Pop III faint SN with mass m_PopIII = 43 M_0 and explosion energy E_SN = 0.6 x 10^51 erg, while the minor contributor is a Pop III hypernova of m_PopIII = 19.4 M_0 and E_SN = 10 x 10^51 erg. The unique properties of DR20080 have never before been observed in Sculptor or in any other environment within the MW and its satellite system.

Introduction

According to cosmology and Big Bang nucleosynthesis, the first stars formed in a neutral environment composed only of hydrogen and helium. All the stars with such a metal-free composition are called Population III (Pop III) stars. Their impact on the evolution of the Universe was significant, as they were the first source of ionising photons, which started the reionisation of the Universe, as well as of metals, enriching the primordial surrounding environment. Due to their metal-free composition, Pop III stars are predicted to be more massive than present-day stars (e.g., [136, 75]). As a result, their initial mass function (IMF) is expected to be top-heavy, although different simulations yield varying conclusions (e.g., [95]). Recent observational evidence supports this picture: Rusta et al. [169] shows that the observed properties of the high-redshift Pop III galaxy Hebe ([122, 221]) are consistent with a top-heavy IMF extending beyond 150M⊙150M_{\odot}. Such an IMF implies short stellar lifetimes, which is consistent with the persisting non-detection of metal-free stars in the local Universe (e.g., [168]).

According to simple models for early chemical enrichment (e.g., [172, 212]) and numerical simulations resolving the first SN explosions (e.g., [31, 160]), even one Pop III SN could provide enough metals to allow for normal Pop II star formation (e.g., [160, 212]). This means that second-generation low-mass stars (≤0.8M⊙\leq0.8M_{\odot}) with lifetimes longer than the age of the Universe were able to form and survive to the present day (e.g., [150, 32, 121, 176, 177]). This is crucial because it provides an indirect method to study Pop III stars. Indeed, the atmospheres of second-generation low-mass stars retain almost unaltered, preserving the chemical yields released by the SNe of their Pop III progenitors, which primarily depend on the progenitor mass and explosion energy (e.g., [72, 212]).

Currently, the Milky Way (MW) stellar halo and its satellite galaxies, including dwarf spheroidal galaxies (dSphs) and ultra-faint dwarf galaxies (UFDs), represent the most promising environments for searching for surviving second-generation stars (e.g., [208, 173, 174]). Indeed, these are the only nearby systems where high-resolution spectroscopy on individual stars is feasible and whose stellar populations are predominantly ancient (e.g., [205, 28]). However, due to the relatively large distances of the MW satellites, high-resolution spectroscopy can only be carried out for the brightest stars, typically the red giant branch (RGB) stars.

One of the strategies for identifying second-generation stars is to look for the most metal-poor stars. Typically, stars are classified as very metal-poor when their iron abundance is [Fe/H] ≤−2\leq-2. Among them, the Carbon-Enhanced Metal-Poor stars (CEMP; [Fe/H] <−2< -2, [C/Fe] >+0.7> +0.7, Beers & Christlieb [18]; Aoki et al. [8]) are a particularly significant category, considering that their fraction grows towards lower metallicities, making up over 80% of the MW halo at [Fe/H] <−4< -4 (e.g., Lucatello et al. [118]; Lee et al. [107]; Placco et al. [155]; Yoon et al. [218]).

The CEMP stars are further divided into subgroups according to their neutron-capture element abundances. These subgroups are associated with different formation channels, which also lead to different origins for their enhanced carbon abundances. Those CEMP stars that also show an enrichment in heavy elements like barium, formed through the slow neutron-capture process (s-process), are classified as CEMP-s stars. The primary mechanism behind the C-enhancement of a CEMP-s star is the mass transfer from a binary companion that has entered the asymptotic giant branch (AGB) phase, which is the main production site of s-process elements (Aoki et al. [8]; Spite et al. [194]; Starkenburg et al. [199]; Hansen et al. [68]). On the other hand, CEMP-r stars are enriched in the rapid neutron-capture (r-process) elements like europium. It is important to notice that the s-process and r-process enrichment are not mutually exclusive. Stars enriched both in Eu and Ba are termed CEMP-r/s, and they could be linked to the intermediate neutron-capture process (i-process; Hampel et al. [66]). However, if a CEMP star shows no significant enrichment in neutron-capture elements, despite its high carbon content, it is classified as a CEMP-no star (e.g., Beers & Christlieb [18]).

The CEMP-no stars are commonly accepted as the direct descendants of those Pop III stars which either were fast-rotating stars (spinstars, e.g., Meynet et al. [138, 139]) or ended their lives as faint SNe (fSNe; e.g., Iwamoto et al. [83]; Vanni et al. [212]). These lower-energy SNe are believed to pollute their surrounding environments mainly with C and other light elements (Z<20Z < 20), resulting in very high [C/Fe] ratios. Depending on the explosion energy, other light elements, such as O, Na, Mg and Si, may also be enhanced (Norris et al. [147]). The study of CEMP-no stars is therefore of primary importance for attempting to characterise the properties of Pop III stars, such as their IMF and the typical energy distribution of their SNe (e.g., Koutsouridou et al. [100]).

However, CEMP-no stars, which are quite common in the MW and in UFDs, are surprisingly rare in dSphs (e.g., Starkenburg et al. [198]; Skúladóttir et al. [189, 190]; Lucchesi et al. [119], see, in particular, Fig. 14 of Lucchesi et al. [119]). The reason behind this distinction remains uncertain (e.g., Salvadori et al. [174]), which makes the study of CEMP-no stars discovered within dSph galaxies even more crucial. Such research could reveal whether these stars share common characteristics with those found in UFDs and in the MW halo or if they should be categorised into subpopulations based on their formation environments. This insight could help determine whether and how the chemical enrichment by Pop III stars is influenced by the environment in which they formed.

The Sculptor dSph, a satellite of the MW, was discovered by Shapley [179] and is located at a distance of 86 ±\pm 5 kpc (Pietrzyński et al. [151]). It has a total mass of (3.4±0.7)⋅108 M⊙(3.4 \pm0.7) \cdot10^{8}\ M_{\odot}, with a mass-to-light ratio of 158 ±\pm 33(M/L)⊙(M/L)_{\odot}, which reveals its dark matter dominated nature (Battaglia et al. [16]), and a mean systemic line-of-sight (los) velocity of vlos=111.2±0.3 km s−1v_{\mathrm{los}} = 111.2 \pm0.3\ \mathrm{km\ s^{-1}} (Arroyo-Polonio et al. [12], using data from Tolstoy et al. [206]).

Since its discovery, Sculptor has been the subject of numerous photometric and spectroscopic studies of its resolved stellar population. Particularly, an analysis of the colour-magnitude diagram reveals a star formation peak around 12 Gyr ago, followed by a gradual decline in the star formation rate until it ceased entirely around 8 Gyr ago (De Boer et al. [43]; Bettinelli et al. [27]). The stellar population of Sculptor is divided in two main populations, with the metallicity distribution functions showing peaks for [Fe/H] =−1.41= -1.41 and [Fe/H] =−2.00= -2.00, and a small sub-population at [Fe/H] =−2.9= -2.9 which may be the residual of a past merger with a UFD (Arroyo-Polonio et al. [12]).

During recent years, several studies in Sculptor focused on the carbon abundance of its stars (Frebel et al. [52]; Tafelmeyer et al. 2010; Starkenburg et al. [198]; Jablonka et al. [84]; Kirby et al. [94]; Simon et al. [182]; Chiti et al. [36]; Skúladóttir et al. [189, 190]). What was found is a general lack of CEMP stars, with the only exception of the low-resolution study from Chiti et al. [36]. However, it should be noted that, even in this case, no stars with [C/Fe] >+0.90> +0.90 were found. So far, the only star in Sculptor that was confirmed to be a CEMP-no star through high-resolution analysis is ET0097, with [Fe/H] =−2.03= -2.03 and [C/Fe] =+0.90= +0.90 (Skúladóttir et al. [189]).

Here we present the second high-resolution follow-up of a CEMP-no star in the Sculptor dSph, DR20080. This high-precision study aims to derive the chemical abundances of as many elements as possible, compare them with those of other metal-poor stars in Sculptor and across the MW and its satellite system, and fit them to theoretical yield models to constrain the properties of DR20080’s progenitors. @bibliography

Data acquisition

In this work we present the results of the high-resolution follow-up of DR20080 (RA 00h 57m 44.16s; DEC −34d 01m 14.3s), the only extremely metal-poor ([Fe/H] < −3) CEMP-no star discovered in the Sculptor dSph (see Figure 1). DR20080 is an RGB star, which was initially identified as a metal-poor star by a large survey that exploited the spectrograph FLAMES at the Very Large Telescope (VLT) to investigate the Ca ii NIR triplet (Tolstoy et al. [206]). It was followed up with the multi-wavelength medium-resolution spectrograph VLT/X-SHOOTER for the confirmation of the FLAMES measurement. These medium-resolution spectra, analysed by Skúladóttir et al. [190], allowed the authors to derive a metallicity of [Fe/H] = −3.00 ± 0.14 (which confirmed the metal-poor nature of DR20080), but also to add an estimate of the carbon abundance [C/Fe]corr_{\mathrm{corr}} = 1.02 ± 0.17 from the CH ∼ 4300 Å molecular band. These abundances, together with the upper limit [Ba/Fe] < −0.5, revealed the CEMP-no nature of DR20080, making it the second CEMP-no star confirmed in the Sculptor dSph and the most metal-poor of the two. However, further analysis outside of Fe, C and Ba was not feasible, due to both the low signal-to-noise ratio (S/NS/N) and the modest spectral resolution of the X-SHOOTER spectrum (S/NS/N ∼ 12 at 4400 Å and RR ∼ 5400, see Skúladóttir et al. [190]). Moreover, in the recent work by Arroyo-Polonio et al. [12], the authors showed that within the Sculptor dSph there is a sub-population of stars, which includes DR20080, that may have originated outside Sculptor and was then accreted through a past merger.

Member stars of the Sculptor dSph

Figure 1. Member stars of the Sculptor dSph from Battaglia et al. [17] based on positions from Gaia eDR3 (blue dots). Star symbols indicate known members with [Fe/H] ≤−3.0\leq-3.0 (data from Tolstoy et al. [206] and Skúladóttir et al. [190]). Red star: DR20080, the CEMP-no star analysed in this work; magenta circle: ET0097, the other CEMP-no star in the Sculptor dSph analysed in high-resolution (Skúladóttir et al. [189]).

In the present work, we performed a high-resolution follow-up of DR20080 that was carried out through a series of eight observations, between October and November 2022, using the high-resolution VLT/UVES spectrograph. Each of these eight observations was conducted with an exposure time of 3000 seconds, as listed in Table 1, and simultaneously covered three spectral bands: Blue (B) band, Lower Red (Rl) band, and Upper Red (Ru) band. The same table also reports the date of observations and the observational conditions. The mean airmass was calculated as the mean of the six airmass values of each exposure (one value for the start of the observation and another for the end, for all three spectral bands). The mean seeing value per observing night was derived as the average between the values on the three spectral bands.

DateExposure TimeAirmass (average)Seeing
[dd/mm/yyyy][s][″]
25/10/202230001.0210.58
29/10/202230001.0621.01
18/11/202230001.0201.17
18/11/202230001.0671.18
20/11/202230001.0220.78
21/11/202230001.0350.76
21/11/202230001.0980.75
22/11/202230001.0200.68

Table 1. Log of the VLT/UVES observations.

The instrumental configuration consisted of a 1.2″ slit and 2×2 pixel binning, with the DIC1 dichroic and the CD#2/390 and CD#3/564 settings for the blue and red arms, respectively. This configuration is a standard UVES setting centered on the wavelengths 3900 Å in the blue arm, and 5640 Å in the red arm. The spectral resolution for all three bands was RR ∼ 35000. The final wavelength ranges adopted for the subsequent analysis are listed in Table 2. The edges of the full wavelength coverage were trimmed in order to exclude regions with lower data quality, particularly at the boundaries of the spectral ranges. The 8 exposures per spectral band were then coadded to improve the S/NS/N, obtaining the final values reported in Table 2. During this process, we also performed a cosmic rays removal. The S/NS/N was determined in selected line-free regions of each spectral band, whose wavelength intervals are reported in Table 2.

Spectral BandWavelength [Å]S/NS/N [pxl−1^{-1}]Line-free Regions [Å]
B3700 − 4500164437.0 − 4441.0
Rl4630 − 5585405438.5 − 5444.0
Ru5688 − 6625525868.3 − 5874.3

Table 2. Wavelength ranges of all spectral bands and the corresponding S/NS/N.

Data analysis

Tools

The abundance analysis carried out in this work was conducted with the aid of ASTRA (Analysis of STellar Radiation and Atmospheres) from Romain Lucchesi, already used in Lucchesi et al. [119]. This tool wraps together the MARCS atmospherical models ([65]) and for the spectral analysis DAOSPEC ([200]), to perform the equivalent width (EWEW) method, and Turbospectrum (Plez [156]), to perform the spectral synthesis method.

We adopted the MARCS models with a standard chemical composition, corresponding to [α/Fe] = +0.4 at [Fe/H] = −3.00, assuming one-dimensional (1D) and Local Thermodynamic Equilibrium (LTE) approximations. The spectral analysis was performed through a combination of EWEW and spectral synthesis methods. The line list given to DAOSPEC is a compilation of those presented in Tafelmeyer et al. (2010); Van Der Swaelmen et al. [210]; Jablonka et al. [84], and is the same as that adopted by Lucchesi et al. [119] for the analysis of metal-poor stars. The literature line list used for the chemical abundances derivation through spectral synthesis was a combination of the VALD database1 (Piskunov et al. [152]; Ryabchikova et al. [170]; Kupka et al. [103, 104]) and, for molecular lines, we adopted the line lists from Sneden et al. [193] and Masseron et al. [134].

After our initial LTE analysis, we applied state-of-the-art 1D Non-LTE (NLTE) corrections (see Subsection 4.7) using the online tools from MPIA2 (Kovalev, M. et al. [102]) and Inasan3 (M. Mashonkina et al. [131, 128]), which take the stellar parameters as input and output line-by-line corrections for a specific ion.

Line-of-sight velocity

The membership of DR20080 in the Sculptor dSph had already been established through the study of proper motions conducted by Battaglia et al. [17] using Gaia eDR3. We performed los velocity (often called radial velocity) measurements (properly corrected for the heliocentric velocity) for all our 24 spectra Notes. Apparent magnitudes are taken from Gaia DR3, while the dust extinction E(B−V) is from dustmaps ([64]). The last three columns are the apparent magnitudes corrected for reddening (see Subsection 3.3).

across the three spectral bands (see Appendix § 9 for more details). The mean los velocity value of DR20080, vlos=120.15±0.27 km s−1v_{\rm los}=120.15\pm0.27\ \mathrm{km\ s^{-1}}, was measured as the arithmetic mean of all the velocity measurements, while the uncertainty was estimated as

Δvlos=σnobs−1,\Delta v_{\rm los}=\frac{\sigma}{\sqrt{n_{\rm obs}-1}},

where σ\sigma is the standard deviation between all the measurements and nobs=24n_{\rm obs}=24 is the total number of exposures. This value falls within 1σ1\sigma from the centre of the normalised distribution of los velocities of stars belonging to Sculptor (Tolstoy et al. [206]), which further verifies that DR20080 indeed belongs to the Sculptor dSph. Moreover, this is fully consistent with the value vlos=120.63±0.48 km s−1v_{\rm los}=120.63\pm0.48\ \mathrm{km\ s^{-1}} reported in the same paper and measured through the VLT/FLAMES spectra (obtained on 15/12/2018, program ID: 0102.B-0786), showing no substantial change of the los velocity with time. Although binarity cannot be excluded at this stage, there is currently no indication that DR20080 belongs to a binary system.

An arithmetic mean was also performed on the individual measurements for each band. The result for the B-band was vlosB=120.65±0.51 km s−1v_{\rm los}^{B}=120.65\pm0.51\ \mathrm{km\ s^{-1}}, for the Rl-band vlosRl=120.32±0.27 km s−1v_{\rm los}^{Rl}=120.32\pm0.27\ \mathrm{km\ s^{-1}}, and for the Ru-band vlosRu=119.47±0.51 km s−1v_{\rm los}^{Ru}=119.47\pm0.51\ \mathrm{km\ s^{-1}}; the first two results are consistent, while the result for the Ru-band, which had a significantly lower number of identified lines, is slightly inconsistent. However, the deviation was not considered significant enough to invalidate the overall mean result across the three bands.

Stellar parameters

The stellar parameters that characterise DR20080 were estimated through photometry based on Gaia DR3 ([60]). This choice was guided by the extremely low metallicity of this star. Indeed, as shown in Mucciarelli & Bonifacio (2020), spectroscopic parameters show an inconsistency at lower metallicities if compared with those derived by photometry, probably caused by metallicity-dependent 3D NLTE effects of Fe I (e.g., [126, 25], see also Fig. 10 in Koutsouridou et al. [101]). We used the three Gaia filters G (3300 − 10500 Å), BP (3300 − 6800 Å) and RP (6400 − 10500 Å), whose measured magnitudes are reported in Table 3, along with their values after reddening correction (G₀, BP₀ and RP₀), where the extinction value (E(B−V)) along the los was derived with dustmaps ([64]) and the extinction coefficients were calculated as reported in Gaia Collaboration et al. [61]. The uncertainties on the Gaia apparent magnitudes, being reported as ≲0.01\lesssim0.01 mag for each filter ([175]), are considered negligible compared to other sources of errors in this analysis.

Gaia IDRA [h : m : s]DEC [d : m : s]G [mag]BP [mag]RP [mag]E(B−V) [mag]G₀ [mag]BP₀ [mag]RP₀ [mag]
500318046507661337600 : 57 : 44.16−34 : 01 : 14.317.8418.4117.170.017817.8018.3617.14

Table 3. Apparent magnitudes of DR20080 in three spectral bands and corresponding reddening-corrected magnitudes.

Effective temperature

The photometric effective temperature, TeffT_{\rm eff}, was obtained using the Gaia calibration of Mucciarelli & Bellazzini [140]:

5040Teff=b0+b1C+b2C2+b3[Fe/H]+b4[Fe/H]2+b5[Fe/H]C,(1)\frac{5040}{T_{\rm eff}}=b_0+b_1C+b_2C^2+b_3[\mathrm{Fe}/\mathrm{H}]+b_4[\mathrm{Fe}/\mathrm{H}]^2+b_5[\mathrm{Fe}/\mathrm{H}]C , \tag*{(1)}

where bib_i are the coefficients for giant stars reported in Mucciarelli et al. [141], while CC is the selected photometric colour. Initially, the value [Fe/H] = −3 from Skúladóttir et al. [190] was assumed. Once the more accurate value of [Fe/H] = −3.26 was derived through abundance analysis, we performed a second iteration with the updated metallicity. The values reported below refer to this second iteration.

Using the reddening corrected magnitudes reported in Table 3, we applied Equation (1) using the three corrected colours (BP − RP)₀, (BP − G)₀ and (G − RP)₀. This led to the three effective temperature values reported in Table 4. The final estimate of the effective temperature, Teff=4672±94T_{\rm eff}=4672\pm94 K, was derived as the average and standard deviation of the values for these three temperatures.

StarTeff(BP−RP)0T_{\rm eff}(\mathrm{BP}-\mathrm{RP})_0 [K]Teff(BP−G)0T_{\rm eff}(\mathrm{BP}-\mathrm{G})_0 [K]Teff(G−RP)0T_{\rm eff}(\mathrm{G}-\mathrm{RP})_0 [K]TeffT_{\rm eff} [K]ΔTeff\Delta T_{\rm eff} [K]
DR200804658456447934672±94

Table 4. Values of the TeffT_{\rm eff} values obtained with Gaia photometry along with the mean TeffT_{\rm eff} and its standard deviation.

Surface gravity

The photometric surface gravity, log⁡g\log g, was determined through the standard relation:

log⁡g=log⁡g⊙+log⁡(M⋆M⊙)+4log⁡(TeffTeff,⊙)+0.4(Mbol,⋆−Mbol,⊙),\log g=\log g_{\odot}+\log\left(\frac{M_{\star}}{M_{\odot}}\right)+4\log\left(\frac{T_{\rm eff}}{T_{\rm eff,\odot}}\right)+0.4\left(M_{\mathrm{bol},\star}-M_{\mathrm{bol},\odot}\right),

based on the previously obtained TeffT_{\rm eff}, as well as the magnitude G₀ and the assumed stellar mass (M⋆=0.8±0.2M⊙M_{\star}=0.8\pm0.2M_{\odot}). The bolometric absolute magnitude (Mbol,⋆M_{\mathrm{bol},\star}) of DR20080 was derived through the equation:

Mbol,⋆=G0−(m−M)0+BC,M_{\mathrm{bol},\star}=G_0-(m-M)_0+BC ,

where we used the distance modulus of the Sculptor dSph galaxy ((m−M)0=19.67±0.16(m-M)_0=19.67\pm0.16 mag; [59]) and the bolometric correction, BCBC, calculated as indicated in the

Notes. The solar values are from Asplund et al. [13]. For Fe I and Fe II abundances and errors, [Fe/H] and Δ\Delta[Fe/H] values are listed instead of [X/Fe] and Δ\Delta[X/Fe]. Cobs_{\mathrm{obs}} is the observed C abundance, Ccorr_{\mathrm{corr}} when corrected for internal mixing (Placco et al. [155]).

Gaia data release documentation4. We adopted the solar parameters log⁡g⊙=4.44\log g_{\odot} = 4.44, Teff,⊙=5772T_{\mathrm{eff},\odot} = 5772 K and Mbol,⊙=4.74M_{\mathrm{bol},\odot} = 4.74 mag. The final value of DR20080 surface gravity results to be log⁡g=1.27±0.27\log g = 1.27 \pm0.27.

Microturbulence velocity

For the photometric microturbulence velocity, vturbv_{\mathrm{turb}}, the empirical relation presented by Anthony-Twarog et al. [5] was used:

vturb=2−0.2log⁡g.v_{\mathrm{turb}} = 2 - 0.2 \log g .

The scatter around this relationship is Δvturb,rms=0.17\Delta v_{\mathrm{turb,rms}} = 0.17 km s−1^{-1}, as reported in Carretta et al. [34], so the total uncertainty was estimated considering the two independent sources of error Δvturb,rms\Delta v_{\mathrm{turb,rms}} and Δlog⁡g\Delta\log g. The final value of the microturbulence velocity was found to be vturb=1.75±0.18v_{\mathrm{turb}} = 1.75 \pm0.18 km s−1^{-1}.

Chemical abundances

The analysis was carried out thanks to a combination of EW and spectral synthesis methods, giving as a result the abundances in 1D LTE approximation. Finally, state-of-the-art NLTE corrections were applied to all elements for which such corrections were available in the literature (see Subsection 3.1 and Subsection 4.7).

Notes. In the case of Fe I and Fe II, [Fe/H] and Δ\Delta[Fe/H] are shown.

To select which of the detected lines to use for the abundance derivation with EWs, we impose two main selection criteria: 1) 20 mÅ ≤EW≤\leq EW \leq 100 mÅ; 2) all those lines that presented an error ΔEWEW>25%\frac{\Delta EW}{EW} > 25\% were rejected, where ΔEW\Delta EW is the uncertainty on the individual EW derivation estimated by DAOSPEC (see Stetson & Pancino [200]). However, whenever a line presented a blend, or if a specific element presented only a few detectable lines, the analysis was carried out through spectral synthesis. For a complete list of the lines that were used and the method adopted to analyse each of them (EWs or spectral synthesis) see Table B.1 in Appendix B.

The abundance values are presented in Table 5, together with their associated uncertainties and the number of lines used to derive the abundance of each ion. All the [X/Fe] values are referred to the solar abundances from Asplund et al. [13]. For four of the elements, only the upper limits are available. This applied to all those elements whose absorption lines were too small to be directly detected but lay within the range covered by the three spectral bands. In this case, Table 5 displays the number of lines used to estimate the upper limit. In Table 6 are reported the mean NLTE corrections (δ\delta[X/H]NLTE_{\mathrm{NLTE}}) applied to obtain the final [X/Fe]NLTE_{\mathrm{NLTE}} values. When the available NLTE corrections in the literature do not cover all detected lines of a given element, we adopted the mean value of the available corrections. All the abundances [X/Fe]NLTE_{\mathrm{NLTE}} assume the NLTE metallicity [Fe/H]NLTE_{\mathrm{NLTE}}.

IonNLinesN_{\mathrm{Lines}}log⁡ϵ\log\epsilon[X/Fe]Δ\Delta[X/Fe]
Fe I61+4.24−3.26±0.12
Fe II5+4.25−3.25±0.13
Cobs_{\mathrm{obs}} (CH)1+6.24+1.07±0.21
Ccorr_{\mathrm{corr}} (CH)1+6.77+1.60±0.21
N (CN)1+6.29+1.72±0.26
O I1+6.96+1.53±0.25
Na I2+3.19+0.21±0.15
Mg I3+4.81+0.47±0.12
Al I1+3.02−0.17±0.29
Si I1+3.95−0.30±0.30
Ca I6+3.25+0.17±0.07
Sc II2−0.19−0.08±0.21
Ti I4+1.67−0.02±0.11
Ti II15+1.87+0.18±0.15
V I2< +0.57< −0.10−
Cr I4+2.15−0.23±0.16
Mn I2+1.15−1.02±0.18
Co I4+1.75+0.02±0.18
Ni I1+2.51−0.45±0.14
Cu I3< +1.33< +0.40−
Zn I2< +1.80< +0.50−
Sr II2−2.12−1.73±0.22
Ba II1−2.36−1.28±0.17
Eu II3< −2.24< +0.50−

Table 5. LTE abundances of DR20080.

IonNLinesN_{\mathrm{Lines}}δ\delta[X/H]NLTE_{\mathrm{NLTE}}[X/Fe]NLTE_{\mathrm{NLTE}}Δ\Delta[X/Fe]
Fe I25/61+0.19−3.08±0.12
Fe II3/5+0.01−3.24±0.13
Ccorr_{\mathrm{corr}} (CH)−−+1.42±0.21
N (CN)−−+1.54±0.26
O I−−+1.35±0.25
Na I2/2−0.26−0.23±0.15
Mg I3/3+0.04+0.33±0.12
Al I1/1+1.00+0.65±0.29
Si I1/1−0.09−0.57±0.30
Ca I6/6+0.23+0.22±0.07
Sc II−−−0.26±0.21
Ti I4/4+0.56+0.36±0.11
Ti II15/15+0.08+0.08±0.15
V I−−< −0.28−
Cr I4/4+0.55+0.14±0.16
Mn I2/2+0.17−1.03±0.18
Co I4/4+0.79+0.63±0.18
Ni I−+0.19−0.44±0.14
Cu I−−< +0.22−
Zn I−−< +0.32−
Sr II2/2+0.31−1.60±0.22
Ba II1/1+0.21−1.25±0.17
Eu II−−< +0.32−

Table 6. NLTE corrections and chemical abundances of DR20080.

The structure of the next Subsections is as follows. The LTE abundance analysis is discussed in Subsections 4.1 to 4.6. The NLTE corrections are described in Subsection 4.7. Finally, in Subsection 4.8 we present the method used to derive the uncertainties.

Metallicity [Fe/H]

In the DR20080 spectra, it was possible to identify 181 Fe i and 8 Fe ii lines. In this case, we imposed stricter selection criteria to maximise the accuracy and the precision of the abundance derivation. First of all, all the lines with excitation potential χexc<1.4\chi_{\mathrm{exc}} < 1.4 eV were rejected, due to the increase in the spread and the uncertainty of measurements at low χexc\chi_{\mathrm{exc}} values. Moreover, a σ\sigma clipping was imposed through an iterative process, with a rejection of all those lines that were situated at more than 2.5σ2.5\sigma from the [Fe/H] mean value. Finally, we only included Fe lines at λ>4000\lambda> 4000 Å to avoid the lowest S/NS/N region in the spectra.

Thanks to this selection process, of the original 181 detected Fe i lines, 61 (all analysed with EW method) were used to derive the final abundance [Fe/H] = −3.26±0.12-3.26 \pm0.12, as also reported in Table 5.

In the Fe ii case instead, the derived LTE metallicity resulted to be [Fe/H] = −3.25±0.13-3.25 \pm0.13. Even if Fe i and Fe ii values are fully consistent within the uncertainties, Fe i was assumed as the reference Fe abundance due to its best statistics. Nevertheless, both iron abundances confirm the extremely metal-poor nature of the star DR20080.

Finally, our derived abundances are consistent with the metallicity determinations previously reported in the literature for this star: [Fe/H] = −3.33±0.09-3.33 \pm0.09, in Tolstoy et al. [206], and [Fe/H] = −3.00±0.14-3.00 \pm0.14, in Skúladóttir et al. [190].

Carbon - nitrogen - oxygen

The only detected O i line is at 6300.304 Å (Figure 2b) and yields a clearly super-solar value of [O/Fe] = +1.53±0.25+1.53 \pm0.25. This value is consistent with the upper limit [O/Fe] <+2.10< +2.10, evaluated by using the 6363.776 Å line. To further test this result, we also analysed the available low-S/NS/N VLT/X-SHOOTER spectra (Skúladóttir et al. [190]) in the near-infrared (NIR) band (9940–21010 Å), searching for CO molecular lines. From these spectra, we derived an upper limit of [O/Fe] <+3< +3 using CO lines around 16000 Å commonly employed in APOGEE (e.g., Smith et al. [191]). The two absorption features observed at shorter wavelength in Figure 2b appear consistent with telluric lines, known to be in that part of the spectrum; however, there is no evident sign of blending with the O line.

The carbon abundance was derived through the CH molecular band (Figure 2a) around 4300 Å. The abundance derived from this band, however, shows a slight dependence on the oxygen abundance, due to the fact that carbon, in addition to being present in CH, can also be partially distributed into CO (and CN, though it has less of an impact on the abundance derived from CH). For this reason, [C/Fe] was derived upon completion of a second run of the abundance derivation, started only after oxygen abundance was derived and fixed at the value of [O/Fe] = +1.53+1.53. The resulting value was [C/Fe]obs=+1.07±0.21_{\mathrm{obs}} = +1.07 \pm0.21. However, this value represents the observed carbon abundance, which does not consider that the stellar atmosphere of an RGB star like DR20080 is affected by internal mixing. To take this process into account, we exploited the online tool5 based on Placco et al. [155], which generated the correction δ\delta[C/Fe] = +0.55+0.55, resulting in a final carbon LTE abundance of [C/Fe]corr=+1.60±0.21_{\mathrm{corr}} = +1.60 \pm0.21. This value is higher than what was found in Skúladóttir et al. [190] ([C/Fe]corr=+1.02±0.17_{\mathrm{corr}} = +1.02 \pm0.17), confirming that DR20080 is a CEMP star. Thanks to the high spectral resolution of our observations, we could also estimate

the isotopic ratio log⁡(12C13C)=0.9±0.2\log\left(\frac{^{12}\mathrm{C}}{^{13}\mathrm{C}}\right) = 0.9 \pm0.2 (see also Figure B.2 in Appendix B). This result is consistent to what was found in Gratton et al. [63] for stars with similar metallicity to that of DR20080. However, it is important to consider that the surface 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} ratio of DR20080 has been affected by internal mixing during its evolution and therefore does not directly reflect its initial isotopic composition.

The nitrogen abundance was derived through the CN molecular band around 3883 Å (see Figure B.1 in Appendix B). Due to the distribution of carbon among the CO, CH, and CN molecules described before, CN band was strongly affected by carbon abundance, so it was necessary to fix it to the observed value [C/Fe] = +1.07+1.07 before running the abundance derivation on the CN band. The LTE value of the nitrogen chemical abundance was found to be [N/Fe] = +1.72±0.26+1.72 \pm0.26. It has to be considered that, due to the before-mentioned internal mixing in RGB stars, the N surface abundance is expected to be enhanced with respect to its original value (Gratton et al. [63]; Spite et al. [195]; Placco et al. [155]). However, in the case of nitrogen, there is no tool available to correct for this effect.

All the following abundance values were derived after imposing a fixed value of [O/Fe] = +1.53+1.53, [C/Fe] = +1.07+1.07 and [N/Fe] = +1.72+1.72. This was done to count for possible blends of molecular lines (CH, CN) in the spectra.

alpha-elements

From the 9 detected Mg i lines, only 3 were kept to derive the abundance value of magnesium (one of them is plotted in Figure B.3 in Appendix B). The triplet in B-band (3829.355 Å, 3832.304 Å and 3838.294 Å) was detected, however it was located in a bad S/NS/N region, while three other detected lines were heavily affected by blends, thus we decide to exclude them from the analysis. The resulting Mg abundance is [Mg/Fe] = +0.47±0.12+0.47 \pm0.12.

Si i lines presented some challenges: the line at 3905.523 Å shows clear blends with CH lines, as can be seen in Figure 2c, while the redder line at 4102.936 Å is strongly affected by the wings of Hδ. Considering that we were able to identify and fit the CH lines (see the yellow spectrum in Figure 2c) that are blended with the first silicon line, we decided to assume the abundance value derived from this line, [Si/Fe] = −0.30±0.30-0.30 \pm0.30, as the final LTE Si i abundance of DR20080. An upper limit of [Si/Fe] <+0.30< +0.30 was estimated from the second Si line, which still indicates a low Si value, since the MW mean value is [Si/Fe] ≃+0.40\simeq+0.40 at this metallicity (Cayrel et al. [35]). It is worth noticing that sub-solar [Si/Fe] at these metallicities are extremely rare, especially in the case of CEMP-no stars (Cayrel et al. [35]; Norris et al. [147]).

In total, 10 calcium lines were identified in the observed spectra. Among these, 5 lines either met the selection criteria and/or were correctly synthesised showing no major blend, and were therefore used in the Ca i abundance derivation. Thus, the final estimate for the Ca abundance is [Ca/Fe] = +0.17±0.07+0.17 \pm0.07.

Finally, titanium is the only element, alongside iron, for which lines from two different ionisation states have been identified. Specifically, 4 Ti i and 15 Ti ii lines contributed to the final abundances, [Ti/Fe] = −0.02±0.11-0.02 \pm0.11 for Ti i and [Ti/Fe] = +0.18±0.15+0.18 \pm0.15 for Ti ii. As for the case of Fe i and Fe ii, Ti i and Ti ii abundances are fully consistent within the uncertainties. We chose Ti ii as the reference indicator for titanium chemical abundance due to its better statistics and because, as highlighted by Bergemann [20], Sitnova et al. [185] and Mallinson et al.

(2022), Ti ii lines in metal-poor stars are less affected by NLTE effects and their corresponding corrections are more reliable.

Odd elements

Only 2 sodium lines were detected in the spectra of DR20080, both in the Ru-band. The resulting Na abundance is [Na/Fe]=+0.21±0.15[\mathrm{Na}/\mathrm{Fe}] = +0.21 \pm0.15. The only aluminium line that was considered in this analysis was the one at 3961.520 Å. This yields an abundance of [Al/Fe]=−0.17±0.29[\mathrm{Al}/\mathrm{Fe}] = -0.17 \pm0.29. Even if also 3944.006 Å was detected, this line resulted to be heavily blended and, consequently, it was excluded. From the 3 scandium lines detected in the B-band spectrum of DR20080, 2 of them were kept in the final analysis. Thus, we derive a Sc abundance of [Sc/Fe]=−0.08±0.21[\mathrm{Sc}/\mathrm{Fe}] = -0.08 \pm0.21. No vanadium lines were strong enough to be detected in the spectra of DR20080. For this reason, we extracted the upper limit [V/Fe]<−0.30[\mathrm{V}/\mathrm{Fe}] < -0.30 from the V i line 4379.230 Å.

Iron-peak elements

From the 8 identified lines, 3 were considered to derive the final LTE Cr i abundance, yielding an abundance of [Cr/Fe]=−±0.07[\mathrm{Cr}/\mathrm{Fe}] = -\pm0.07. Regarding manganese, the chemical abundance was derived using the two bluer lines among the 4030–4034 Å triplet. The reddest line of the triplet was rejected due to heavy blending. Thus, the final estimate for the Mn abundance is [Mn/Fe]=−1.02±0.18[\mathrm{Mn}/\mathrm{Fe}] = -1.02 \pm0.18. Among the 5 detected Co i lines, all in the B-band, 4 were taken to derive the final abundance [Co/Fe]=+0.02±0.18[\mathrm{Co}/\mathrm{Fe}] = +0.02 \pm0.18, also reported in Table 5.

Despite the identification of 6 Ni i lines, only the line at 5476.904 Å was used to determine an abundance of [Ni/Fe]=−0.45±0.14[\mathrm{Ni}/\mathrm{Fe}] = -0.45 \pm0.14, i.e., a sub-solar value. Although the other lines were less reliable, either due to strong blends or to bad S/N, they were also consistent with a sub-solar abundance of nickel. This is a very unusual result, especially at these low metallicities where [Ni/Fe]∼0[\mathrm{Ni}/\mathrm{Fe}] \sim0 (e.g., Cayrel et al. [35]; Skúladóttir et al. [190]).

Finally, neither copper nor zinc lines were detected in the spectra. For Cu, three undetected Cu i lines were analysed to estimate the upper limit of its chemical abundance, [Cu/Fe]<+0.40[\mathrm{Cu}/\mathrm{Fe}] < +0.40. Regarding Zn, we estimated the upper limit in the same way, but from two lines, obtaining [Zn/Fe]<+0.50[\mathrm{Zn}/\mathrm{Fe}] < +0.50 (see Table B.1).

Neutron-capture elements

Both the detected strontium lines were considered when deriving the Sr i LTE abundance (Figure 2d). The final abundance, [Sr/Fe]=−1.73±0.22[\mathrm{Sr}/\mathrm{Fe}] = -1.73 \pm0.22, shows that strontium is not enhanced.

Barium abundance, as the representative element of the s-process, is crucial for distinguishing between CEMP-s stars and CEMP-no stars. The only identified Ba ii line, at 4934.076 Å (see Figure 2e), was analysed assuming solar isotopic fractions for Ba, yielding a very sub-solar abundance of [Ba/Fe]=−1.28±0.17[\mathrm{Ba}/\mathrm{Fe}] = -1.28 \pm0.17. This result is also consistent with the upper limit [Ba/Fe]<−0.5[\mathrm{Ba}/\mathrm{Fe}] < -0.5 found in Skúladóttir et al. [190]. Thus, we can conclude that the star DR20080 is clearly not a CEMP-s star.

No europium line was detected in the observed spectra of DR20080. Two Eu ii lines were analysed to estimate the upper limit, at 3907.106 Å and 4129.708 Å. The lowest of the two upper limits, [Eu/Fe]<+0.50[\mathrm{Eu}/\mathrm{Fe}] < +0.50, corresponds to the line at 4129.708 Å. This result indicates that europium is not enhanced, ruling out the hypothesis of a CEMP-r star. Together with the sub-solar barium abundance, this confirms the CEMP-no classification of DR20080.

NLTE corrections

We mainly used two online tools to retrieve NLTE corrections for our abundance derivations, MPIA and Inasan. We used the MPIA database for the corrections of Fe i and Fe ii (Bergemann et al. [25]), Mg i (Bergemann et al. [22]), Si i (Bergemann et al. [24]), Ca i (Mashonkina et al. [127]), Ti i and Ti ii (Bergemann [20]), Cr i (Bergemann & Cescutti [21]), Mn i (Bergemann & Gehren [23]) and Co i (Bergemann et al. [26]). We also checked O i corrections (Sitnova et al. [184]) using MPIA, but it proved negligible for our stellar parameters. The Inasan online tool was instead used to derive the corrections for Sr ii (Mashonkina et al. [129]) and Ba ii (Mashonkina & Belyaev [130]). The Na i correction (Lind et al. [115]) was retrieved using the tool Inspect-Stars.6

The NLTE corrections are provided by the respective tool for individual lines. If more than one line of the analysed element was present in the database, the final NLTE correction (δ[X/H]NLTE\delta[\mathrm{X}/\mathrm{H}]_{\mathrm{NLTE}}) was obtained by calculating the arithmetic mean of the corrections for the available individual lines.

In the case of Al, the NLTE correction came from Nordlander & Lind [144]. No grids exist for Ni, however the corrections are expected to be similar as those for Fe, thus the correction was assumed to be the same as that of Fe i (δ[Ni/H]NLTE=δ[Fe/H]NLTE=+0.19\delta[\mathrm{Ni}/\mathrm{H}]_{\mathrm{NLTE}} = \delta[\mathrm{Fe}/\mathrm{H}]_{\mathrm{NLTE}} = +0.19), as done, e.g., in Skúladóttir et al. [188]. For carbon, according to Popa et al. [157], the NLTE corrections are expected to be positive, in the range of 0.1–0.2 dex. However, this result is very uncertain, as performing NLTE calculations for a molecular band, such as the CH band, is very challenging. Furthermore, the 3D corrections are expected to be negative (Amarsi et al. [4]). Given the lack of accurate data, we have chosen to take a conservative approach and apply no corrections to the carbon abundance.

All the mean NLTE corrections applied, as well as the number of lines with respect to the total that presented such corrections, are reported in Table 6, together with the final 1D NLTE abundance values [X/Fe]NLTE[\mathrm{X}/\mathrm{Fe}]_{\mathrm{NLTE}}, all referred to the NLTE metallicity [Fe/H]NLTE=−3.08[\mathrm{Fe}/\mathrm{H}]_{\mathrm{NLTE}} = -3.08.

Uncertainties

The derivation of the uncertainties took into account both the contributions of the stellar parameters (Δ[X/Fe]param\Delta[\mathrm{X}/\mathrm{Fe}]_{\mathrm{param}}) and the random error affecting the EWs or spectral fitting (Δ[X/Fe]ran\Delta[\mathrm{X}/\mathrm{Fe}]_{\mathrm{ran}}). The latter uncertainties were automatically generated by ASTRA in the same way that is reported in Lucchesi et al. [119].

To estimate the contribution from the stellar parameters, we ran new iterations of the abundance derivation, varying only one parameter at a time. Specifically, six runs were executed, with each run setting one of the parameters to one of the two extremes defined by its uncertainty Teff=(4578;4766)T_{\mathrm{eff}} = (4578 ; 4766) K, log⁡g=(1.00;1.54)\log g = (1.00 ; 1.54) dex, and vturb=(1.57;1.93)v_{\mathrm{turb}} = (1.57 ; 1.93) km s−1^{-1}. Once all the abundances corresponding to the new stellar parameters were obtained, the uncertainty on mean abundances due to each individual parameter, e.g. TeffT_{\mathrm{eff}}, was calculated as

Δ[X/Fe]Teff=max⁡(∣[X/Fe]−[X/Fe]T4578∣;∣[X/Fe]−[X/Fe]T4766∣)(2)\Delta[\mathrm{X}/\mathrm{Fe}]_{T_{\mathrm{eff}}} = \max\left(\left|[\mathrm{X}/\mathrm{Fe}]-[\mathrm{X}/\mathrm{Fe}]_{T_{4578}}\right|;\left|[\mathrm{X}/\mathrm{Fe}]-[\mathrm{X}/\mathrm{Fe}]_{T_{4766}}\right|\right) \tag*{(2)}

and the same for the other stellar parameters, where [X/Fe] is the mean value reported in Table 5 and [X/Fe]param_{\mathrm{param}} indicates the abundance value obtained varying that specific parameter. The final uncertainties were thus determined by summing quadratically all the contributions, see Table 5.

There were a few exceptions that also took into account other terms to evaluate the final uncertainty. One of these special cases was carbon, whose chemical abundance was calculated after fixing the oxygen abundance value, due to its effect on CO and CH molecules. Thus, we ran the carbon abundance derivation varying the oxygen abundance to its extremes and then used an analogue of Equation (2) to calculate Δ[C/Fe]O\Delta[\mathrm{C}/\mathrm{Fe}]_{\mathrm{O}}. This term was then quadratically added to the overall uncertainty of carbon, but turned out to be negligible compared to the other contributions.

If calculated as described above, the silicon abundance should have an uncertainty of Δ[Si/Fe]=±0.21\Delta[\mathrm{Si}/\mathrm{Fe}] = \pm0.21. However, since both observed lines of this element exhibited significant and evident blends, it was deemed more conservative to assume the uncertainty of Δ[Si/Fe]=±0.30\Delta[\mathrm{Si}/\mathrm{Fe}] = \pm0.30.

Another exception was oxygen, for which the normal uncertainty derivation would yield a value of Δ[O/Fe]=±0.20\Delta[\mathrm{O}/\mathrm{Fe}] = \pm0.20. Instead, a value of Δ[O/Fe]=±0.25\Delta[\mathrm{O}/\mathrm{Fe}] = \pm0.25 was assumed to maintain a more cautious approach, given that we only have a single line, which is also very weak and therefore sensitive to small fluctuation of the noise.

It’s worth noting that the uncertainties on solar abundances (Asplund et al. [13]) proved to be negligible with respect to the other contributions considered. Moreover, considering that the majority of NLTE corrections in the literature are provided without uncertainties, we assumed the Δ[X/Fe]LTE\Delta[\mathrm{X}/\mathrm{Fe}]_{\mathrm{LTE}} values as the final errors also in the 1D NLTE case, as reported in the last column of Table 6.

DR20080 and the Local Group

When we compare the derived chemical abundances of DR20080 with other very metal-poor ([Fe/H]<−2[\mathrm{Fe}/\mathrm{H}] < -2) stars belonging to the Sculptor dSph, for which data are available in the literature, or to other environments in the MW satellite system, such as the MW Stellar halo, other dSphs or UFDs, we begin to reveal how unique this CEMP-no star really is.

Comparison Data

In this Subsection, we provide all the references for the comparison samples used in Subsections 5.2 and 5.3. Figure 3 to Figure 5 show the elemental abundances of DR20080 (red star symbol) compared with other metal-poor stars with similar metallicities (−4.0<[Fe/H]<−2.5-4.0 < [\mathrm{Fe}/\mathrm{H}] < -2.5) for which high-resolution analysis is available in the literature.

Comparison of elemental abundances for DR20080 and metal-poor stars

Figure 3. Comparison between [C/Fe] abundance values, as a function of [Fe/H], of DR20080 (displayed as a red star) and of metal-poor stars in four different environments: Sculptor dSph (blue stars), UFDs (yellow crosses), dSphs (orange diamonds), and MW halo (grey dots); see references in Subsection 5.1. Horizontal black solid line is the CEMP limit ([C/Fe] ≥ 0.7), gray dashed line is the solar value.

Figure 5. Same as Figure 3, but for n-capture elements, [Sr/Fe] and [Ba/Fe] (top panels), and [Sr/Ba] as a function of [Ba/H] (bottom panel). References in Subsection 5.1.

The Sculptor stars at these low metallicities (blue stars) are from: Frebel et al. [52]; Tafelmeyer et al. (2010); Kirby & Cohen [92]; Jablonka et al. [84]; Simon et al. [182]; Skúladóttir et al. [188, 190].

For the stars belonging to other dSphs (orange diamonds), in the case of carbon (Figure 3) we used the literature compilation from Lucchesi et al. [119], which includes data from: Tafelmeyer et al. (2010); Kirby & Cohen [92]; Venn et al. [213]; Susmitha et al. [203]; Hansen et al. [67]; Yoon et al. [219]; Lucchesi et al. [120, 119]; and Sestito et al. [178], integrated with the abundances retrieved from the SAGA database7 (e.g., Suda et al. [202, 201]), including data from: Fulbright et al. [58]; Cohen & Huang [41, 42]; and Honda et al. [77]. In Figure 4 and Figure 5, the data for dSphs stars were taken from Reichert et al. [159] for Mg, Sc, Ti, Cr, Zn. For the other elements (Na, Si, Ca, Mn, Co, Ni, Sr and Ba), the abundances were obtained through the SAGA database, which includes data from: Shetrone et al. [180]; Fulbright et al. [58]; Sadakane et al. [171]; Koch et al. [97]; Aoki et al. [7]; Cohen & Huang [41, 42]; Letarte et al. [110]; Tafelmeyer et al. (2010); Honda et al. [77]; Kirby & Cohen [92]; Venn et al. [213]; Lemasle et al. [109, 108]; Hendricks et al. [74]; Fabrizio et al. [45]; Tsujimoto et al. [207]; Ural et al. [209]; François et al. [49]; Norris et al. [149]; and Hansen et al. [67]. These data include stars belonging to the following galaxies: Canes Venatici I, Carina, Draco, Fornax, Sagittarius, Sextans and Ursa Minor.

Twelve abundance-ratio panels comparing DR20080 with Sculptor, dSphs, UFDs, and the Milky Way

Figure 4. Same as Figure 3, but for [O/Fe], [Na/Fe], [Mg/Fe], [Al/Fe], [Si/Fe], [Ca/Fe], [Sc/Fe], [Ti/Fe], [Cr/Fe], [Mn/Fe], [Co/Fe], and [Ni/Fe]. See Subsection 5.1 for references.

The MW compilation (grey dots) is the same used in Vanni et al. [212], which includes data from Cayrel et al. [35]; Christlieb et al. [38]; Norris et al. [145]; Caffau et al. [33]; Yong et al. [217]; Hansen et al. [69]; Keller et al. [91]; Roederer et al. [164]; Frebel et al. [50]; Li et al. (2015a); Bonifacio et al. [29, 30]; François et al. [48]; Starkenburg et al. [197]; Aguado et al. [2]; and Ezzeddine et al. [44], and has been extended with the recent studies done in Lombardo et al. [116, 117] and Fernandes De Melo et al. [47].

The data for stars belonging to UFDs (yellow crosses) are retrieved from Lai et al. [106]; Frebel et al. [57]; Roederer

& Kirby (2014); Kirby et al. [93]; Nagasawa et al. [143]; Ji et al. (2019); Marshall et al. [124]; Ji et al. (2020); Hansen et al. [70]; Verdiani et al. [214], and from the JINA database8 (Abohalima & Frebel [1]), covering data from: Feltzing et al. [46]; Norris et al. [148]; Simon et al. [181]; Frebel et al. [56] (2016); Gilmore et al. [62]; Ishigaki et al. [80]; Roederer et al. [163]; François et al. [49]; and Ji et al. (2016a,b). The sample includes stars from Boötes I, Boötes II, Canes Venatici II, Carina II, Carina III, Coma Berenices, Grus I, Grus II, Hercules, Horologium I, Leo IV, Reticulum II, Segue 1, Segue 2, Triangulum II, Tucana III, Tucana IV and Ursa Major II.

Given that NLTE corrections have been published for only a handful of these stars, and even then, typically for just a few elements, we focused on the LTE abundances for this comparison. son. All these abundances are referred to the solar abundances from Asplund et al. [13]. Additionally, carbon abundances taken into account were corrected for internal mixing (Placco et al. [155]). In the case of duplicate entries among the various references, the measurements from the original source were preferred.

Early chemical enrichment in different environments

Let us first examine the case of carbon. As shown in Figure 3, the [C/Fe][\mathrm{C}/\mathrm{Fe}] ratio for DR20080 lies well above the CEMP threshold ([C/Fe]≥0.7[\mathrm{C}/\mathrm{Fe}] \ge0.7), confirming its CEMP classification, as previously inferred from the medium-resolution measures presented in Skúladóttir et al. [190]. Moreover, it is the only star in Sculptor, at such low metallicity, exhibiting such behaviour, more closely resembling the trend observed among the members of the MW halo or of the UFDs reported in the figure. This supports the idea that dSphs experienced a very different early chemical enrichment history compared to UFDs and the MW halo.

Figure 4 shows the derived [X/Fe][\mathrm{X}/\mathrm{Fe}] abundances for elements from O to Ni (see also Figure C.1 in Appendix C for the plots of N and of the other analysed elements, for which we estimated an upper limit). The first panel of Figure 4 shows an enhancement in O relative to the few MW stars with available measurements in the literature. Indeed, this element is particularly challenging to measure at these [Fe/H][\mathrm{Fe}/\mathrm{H}] unless it is significantly enhanced, as also indicated by the lack of measurements at [O/Fe]<+0.5[\mathrm{O}/\mathrm{Fe}] < +0.5. As illustrated in the second panel, Na exhibits an enhancement in the DR20080 abundance pattern compared to the Sculptor average. A similar behaviour is observed for Al. In contrast, DR20080 is among the stars with the lowest abundance of Si within the considered metallicity range. Indeed, [Si/Fe][\mathrm{Si}/\mathrm{Fe}] is underabundant even when compared to stars from other environments (see Figure 4). Moreover, the iron-peak elements Mn and Co follow the lower edge of the distribution and Ni is found to be underabundant, contrasting with the general trend in Sculptor that aligns more closely with the populations of UFDs and the MW halo. On the other hand, the other elements shown in Figure 4, such as the α\alpha-elements or the iron-peak element Cr, are fully consistent with the overall trend in Sculptor, in dSphs, in UFDs and in the MW.

DR20080 shows no enhancement in n-capture elements, like strontium, barium and europium (see Table 5 and Table 6 and the top panels of Figure 5). As highlighted in Frebel & Norris [53], studies on UFDs have revealed that the abundances of Sr and Ba, along with their ratio [Sr/Ba][\mathrm{Sr}/\mathrm{Ba}], may serve as criteria to distinguish, at similar [Fe/H][\mathrm{Fe}/\mathrm{H}], which stars were formed from more pristine gas (e.g., Ji et al. (2019)). Specifically, lower abundances of n-capture elements seem to indicate a formation in an earlier gas cloud. Notably, the [Sr/Ba][\mathrm{Sr}/\mathrm{Ba}] ratio, displayed in the bottom panel of Figure 5 as a function of [Ba/H][\mathrm{Ba}/\mathrm{H}], appears to be underabundant with respect to the typical stellar populations represented in the plot. This seems to provide further evidence supporting the conclusion that DR20080 formed in a primitive environment.

Comparison to other CEMP-no stars

To further explore the unusual abundance pattern of DR20080, a comparison was made between this star and other CEMP-no stars from the MW halo and from UFDs with similar metallicities (Figure 6), i.e. ±0.75\pm0.75 dex from the value of DR20080. For this purpose, we considered the 39 MW CEMP-no stars with −4≤[Fe/H]≤−2.5-4 \leq[\mathrm{Fe}/\mathrm{H}] \leq-2.5 within the same literature compilation from Vanni et al. [212] used for Figure 3 to Figure 5 (see Subsection 5.1), which contains both main-sequence and RGB stars. We also considered 8 CEMP-no stars belonging to UFDs, identified within the same metallicity range in the same literature compilation described in Subsection 5.1. As for the previous case, we only considered 1D LTE abundance values for this analysis.

DR20080 LTE abundance ratios compared with mean abundances in MW and UFD CEMP-no stars

Figure 6. DR20080 LTE abundance ratios (red) compared with mean abundances in MW CEMP-no stars ⟨[X/Fe]⟩MW CEMP−no±σX\langle[\mathrm{X}/\mathrm{Fe}]\rangle_{\mathrm{MW\ CEMP-no}} \pm\sigma_{\mathrm{X}} (grey) and in UFD CEMP-no stars ⟨[X/Fe]⟩UFD CEMP−no±σX\langle[\mathrm{X}/\mathrm{Fe}]\rangle_{\mathrm{UFD\ CEMP-no}} \pm\sigma_{\mathrm{X}} (yellow). Only elements measured both in DR20080 and in the MW sample are displayed. Metallicity range of the MW and UFDs samples is −4≤[Fe/H]≤−2.5-4 \leq[\mathrm{Fe}/\mathrm{H}] \leq-2.5. The mean metallicity of MW CEMP-no stars is ⟨[Fe/H]⟩MW CEMP−no=−3.23±0.36\langle[\mathrm{Fe}/\mathrm{H}]\rangle_{\mathrm{MW\ CEMP-no}} = -3.23 \pm0.36, for the UFDs sample is ⟨[Fe/H]⟩UFD CEMP−no=−3.43±0.45\langle[\mathrm{Fe}/\mathrm{H}]\rangle_{\mathrm{UFD\ CEMP-no}} = -3.43 \pm0.45, while DR20080 has [Fe/H]=−3.26±0.12[\mathrm{Fe}/\mathrm{H}] = -3.26 \pm0.12. Compilation of Vanni et al. [212] for the MW sample, and of Norris et al. [148]; ); Frebel et al. [56, 57]; ); Ji et al. (2020) for the UFDs sample, see text for references. Figure C.2 shows the same for [X/H][\mathrm{X}/\mathrm{H}].

To make this comparison, we calculated the mean of the abundance ratios of each element X, ⟨[X/Fe]⟩CEMP−no\langle[\mathrm{X}/\mathrm{Fe}]\rangle_{\mathrm{CEMP-no}}, in the two samples, and we assumed the standard deviation, σX\sigma_{\mathrm{X}}, as the indication of the scatter of these values. These mean abundances, alongside the derived abundances of DR20080, are displayed in Figure 6. It is worth noticing that not all stars in the sample had abundance derivations for every element displayed, so each mean value is based only on the available data within the sample.

The first thing we notice is that carbon and nitrogen are well aligned with the typical MW CEMP-no stars, while the oxygen abundance in DR20080 is significantly enhanced. The carbon abundance is also consistent with that of CEMP-no stars in UFDs. In contrast, the iron-peak elements, namely manganese and nickel, are noticeably under-abundant relative to both populations (see also Figure 4). It is also worth mentioning silicon and cobalt, which, although consistent with the mean values of CEMP-no stars in the MW, appear to be slightly under-abundant. Moreover, while Co is consistent with the UFD sample, Si is significantly under-abundant relative to it.

Even though for single elements such as carbon, DR20080 appears to be more similar to what is observed in the MW halo or in UFDs than in Sculptor or other dSphs, as shown for instance in Figure 3, the overall abundance pattern reveals that DR20080 is intrinsically different even from the CEMP-no stars in the MW halo or in UFDs. This, in turn, suggests that its chemical enrichment history also differs from that of such an environment. Such observation may therefore support the hypothesis of different populations of CEMP-no stars, each significantly shaped by the environment in which they formed.

However, it is worth noting that, according to the recent study by Arroyo-Polonio et al. [12], DR20080 is part of a very metal-poor subpopulation of Sculptor’s stars that may not have formed in situ, but may instead be the remnant of a past merger with a UFD. This possibility is also predicted by cosmological models for the Local Group formation (e.g., Salvadori et al. [174]; ); Koutsouridou et al. [99]), as mergers are a natural consequence of hierarchical structure formation in the Λ\LambdaCDM cosmological model. However, quantifying the contribution of such events to the stellar population of a Sculptor-like galaxy is challenging, as individual UFDs are expected to contribute only a small fraction of their stellar mass, while the number of CEMPno stars accreted depends on the star formation history of the accreted UFD and its Pop III IMF (e.g., Rossi et al. [167]). According to this interpretation, if DR20080 indeed did not form in situ, it provides further indirect evidence of the rarity of CEMP-no stars in dSphs, still leaving open the question regarding the differences in their formation in this class of galaxies and the puzzle that distinguishes them from those in UFDs and in the MW halo.

Pop III progenitor of DR20080

Since Pop III stars cannot be observed directly, our understanding of them relies on indirect evidence, particularly from the chemical abundances found in the stellar photospheres of ancient, metal-poor stars that may be their direct descendants. Their chemical yields depend primarily on the stellar mass (mPopIIIm_{\mathrm{PopIII}}) and the explosion energy of its SN (ESNE_{\mathrm{SN}}), making the analysis of stellar abundances a valuable tool for constraining these factors and gaining insight into the elusive Pop III stars.

As discussed in § 5, DR20080 shows signs of having formed in a gas cloud mainly enriched by Pop III stars, as we can infer from its CEMP-no nature and from the low [Sr/Ba] and [Ba/Fe] ratios. Moreover, if we examine the chemical enrichment models from various types of SNe (including Pair-Instability SNe, PISNe), originating both from Pop III and Pop II stars, presented in Vanni et al. [211] and displayed in Figure 7, we can find more clues pointing towards the Pop III enrichment. In this plot, DR20080 (red star) occupies a region,

Abundance-ratio diagram showing SN yield models and DR20080

Figure 7. Red star represents DR20080’s NLTE abundance ratios, with their uncertainties, while coloured contours are SN yield models. Solid red and blue lines indicate environments enriched exclusively by PISNe or other Pop III SNe respectively, dashed red and blue primarily by PISNe or other Pop III SNe, but with some contribution from normal Pop II SNe, and dotted brown and grey mainly by normal Pop II SNe with minor contributions from PISNe or other Pop III SNe respectively. Adapted from Vanni et al. [211]. Models from Salvadori et al. [172] and Vanni et al. [212].

[Si/O]NLTE=−1.92±0.39[\mathrm{Si}/\mathrm{O}]_{\mathrm{NLTE}} = -1.92 \pm0.39 and [C/O]NLTE=0.07±0.32[\mathrm{C}/\mathrm{O}]_{\mathrm{NLTE}} = 0.07 \pm0.32, that is sparsely populated by stellar yield models. The only models that are consistent with the abundances derived for this CEMP-no star are those prevalently enriched by Pop III SNe. It is important to notice that the conclusion remains the same in the LTE case, with [Si/O]=−1.83±0.39[\mathrm{Si}/\mathrm{O}] = -1.83 \pm0.39 and [C/O]=0.07±0.32[\mathrm{C}/\mathrm{O}] = 0.07 \pm0.32.

Possible spinstar progenitor

This value of [C/O] is in good agreement with the predicted yields of spinstar progenitors, as shown, for example, in Figure 6 of Choplin & Meynet [37]. The abundances of n-capture elements [Sr/Fe], [Ba/Fe] and [Sr/Ba] also fall within the ranges predicted by the spinstar models presented in Rizzuti et al. [161]. However, log C/O = −0.19 falls outside the range covered by the models of Rizzuti et al. [161], while the Na and Mg abundances are more consistent with the non-rotating or slowly rotating models of Choplin & Meynet [37]. Moreover, Rizzuti et al. [161] point out that their spinstar models struggle to reproduce the abundances of C, N and O of CEMP-no stars in the metallicity range −4 < [Fe/H] < −2, tending to underestimate their abundances. The isotopic ratio 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} is generally considered one of the most powerful diagnostics for investigating a possible contribution from spinstar progenitors (e.g., Rizzuti et al. [161]). However, since DR20080 is an evolved giant, its present-day 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} ratio has been altered by internal mixing and therefore does not provide a direct indication of the isotopic ratio in the star at birth (see also subsection 4.2). Overall, although a spinstar origin for the progenitor of DR20080 cannot be ruled out.

Possible SN progenitors

To derive the properties of DR20080’s progenitor stars, we fitted the observed 1D NLTE abundance pattern using a χ2\chi^2 minimisation procedure applied to a grid of SN yield models, as also done in Koutsouridou et al. [98], using the general analytical calculations derived by Salvadori et al. [172] and Vanni et al. [212]. For Pop III stars, the type II and pair-instability SN yields were taken from Heger & Woosley [72, 73] and from Woosley & Weaver [216] and Limongi & Chieffi [114] for Pop II progenitors. For Type Ia SN yields (for both Pop III and Pop II stars), we adopted the models of Iwamoto et al. [82] and Leung & Nomoto [111].

In total, three fits were performed under different assumptions (see Figure 8): the first (blue line) assumes that DR20080 formed from gas enriched by a single progenitor, while the second (red line) and the third (green line) assume enrichment by two distinct progenitors. The latter two cases differ in the number of considered elements. In the first and second fits, all the measured elements from carbon to zinc were included. The only exception was nitrogen, which was always excluded due to the internal mixing that makes N abundance unreliable for this purpose. The third fit only included those elements whose abundance theoretical values are in best agreement with the observed ones for the overall stellar population (e.g., Kobayashi et al. [96]; ); Koutsouridou et al. [101]), namely O, Na, Mg, Al, Si, Ca, Cr, Mn, Fe, and Cu (indicated by the squares in Figure 8), which we will refer to as reliable elements.

Best-fit models of DR20080’s abundance pattern

Figure 8. Best-fit models of DR20080’s abundance pattern, all made considering both Pop III and Pop II SN yields models: 1) blue line, only one progenitor star (a Pop III fSN with mPopIII=11.4M⊙m_{\mathrm{PopIII}} = 11.4M_{\odot} and ESN=0.6⋅1051E_{\mathrm{SN}} = 0.6 \cdot10^{51} erg); 2) red line, it represents the fit with two progenitors (95% Pop III fSN with mPopIII=43M⊙m_{\mathrm{PopIII}} = 43M_{\odot} and ESN=0.6⋅1051E_{\mathrm{SN}} = 0.6 \cdot10^{51} erg; 5% Pop III HN with mPopIII=19.4M⊙m_{\mathrm{PopIII}} = 19.4M_{\odot} and ESN=10⋅1051E_{\mathrm{SN}} = 10 \cdot10^{51} erg.) which considers all the measured elements (squares and circles); 3) green line, it also represents the fit with two progenitors (85% Pop III fSN with mPopIII=43M⊙m_{\mathrm{PopIII}} = 43M_{\odot} and ESN=0.6⋅1051E_{\mathrm{SN}} = 0.6 \cdot10^{51} erg; 15% Pop III HN with mPopIII=50M⊙m_{\mathrm{PopIII}} = 50M_{\odot} and ESN=10⋅1051E_{\mathrm{SN}} = 10 \cdot10^{51} erg.), but only accounting for those elements considered as reliable (squares, see text for more details).

In all three cases, the best-fit model corresponds to a pure Pop III enrichment. Indeed, in the first case, the minimum χν,best2=4.54\chi^2_{\nu,\mathrm{best}} = 4.54 is obtained for a Pop III progenitor with a mass of mPopIII=11.4 M⊙m_{\mathrm{PopIII}} = 11.4\,\mathrm{M}_{\odot}, which exploded as a fSN with an energy of ESN=0.6⋅1051E_{\mathrm{SN}} = 0.6 \cdot10^{51} erg. In the second fit (χν,best2=2.17\chi^{2}_{\nu,\mathrm{best}} = 2.17), assuming two progenitors, the major contributor to the enrichment accounts for 95% of the abundance pattern, and corresponds to a Pop III fSN with mPopIII=43M⊙m_{\mathrm{PopIII}} = 43M_{\odot} and ESN=0.6⋅1051E_{\mathrm{SN}} = 0.6 \cdot10^{51} erg. The remaining 5% originates from a Pop III hypernova (HN) of mass mPopIII=19.4M⊙m_{\mathrm{PopIII}} = 19.4M_{\odot} and energy ESN=10⋅1051E_{\mathrm{SN}} = 10 \cdot10^{51} erg. Finally, the third fit (χν,best2=1.52\chi^{2}_{\nu,\mathrm{best}} = 1.52), which assumes two progenitors but only includes the 10 reliable elements, yields the same main progenitor as in the previous case, which now contributes 85% of the total abundance pattern. The remaining 15% comes from a Pop III HN with mPopIII=50M⊙m_{\mathrm{PopIII}} = 50M_{\odot} and ESN=10⋅1051E_{\mathrm{SN}} = 10 \cdot10^{51} erg. These results indicate that the two-progenitor solutions generally provide a better match to the observed abundances than the single-progenitor case. However, we should notice that some elements, such as Ti and Co, are still not well reproduced by the models. Still, these elements are not usually well reproduced by the theoretical yields, also when considering Pop II stars yields (e.g., [166, 96, 101]). In addition, there could also be some effects of underlying physics of Pop III stars that are not yet accounted for in current models.

Figure 8 presents the best-fit models, but it is also important to explore the parameter space in order to assess the possible degeneracies among the models. The fit results for the three studied cases are shown in Figure 9a to Figure 9c, where different combinations of mPopIIIm_{\mathrm{PopIII}} and ESNE_{\mathrm{SN}} are colour-coded according to their χν2/χν,best2\chi^{2}_{\nu}/\chi^{2}_{\nu,\mathrm{best}} values. For the two-progenitor cases, in Figure 9b and Figure 9c, only the parameter space for the main progenitor is plotted. For each primary contributor, the fitting procedure identifies the properties of the secondary progenitor that minimise χν2\chi^{2}_{\nu}. The best-fit solution (χν,best2\chi^{2}_{\nu,\mathrm{best}}) corresponds to the combination of primary and secondary progenitors that minimises the overall χν2\chi^{2}_{\nu}. From these two plots, it appears that, although restricting the fit to the most reliable elements yields a lower χν,best2\chi^{2}_{\nu,\mathrm{best}} value, the reduced number of constraints significantly increases the degeneracy of the final solution. However, it is worth noting that, despite the increased degeneracy, the result for the best fit main progenitor is the same in both cases, further reinforcing the robustness of this outcome. Regarding the secondary progenitor, being only a minor contributor the degeneracy of the possible solutions is higher, but it is remarkable that in both cases the large majority of solutions (i.e. those fits with χν2<1.5χν,best2\chi^{2}_{\nu} < 1.5\chi^{2}_{\nu,\mathrm{best}}; 97.5% accounting for the complete abundance pattern and 78.9% when only considering reliable elements) point towards a Pop III HN as the secondary progenitor of DR20080.

Parameter-space plots for Pop III progenitor fits

Figure 9. (a) Best fit: Pop III fSN with ESN=0.6⋅1051 ergE_{\mathrm{SN}} = 0.6 \cdot10^{51}\ \mathrm{erg}; mPopIII=11.4 M⊙m_{\mathrm{PopIII}} = 11.4\,M_{\odot}

(b) Best fit: 95% Pop III fSN with ESN=0.6⋅1051 ergE_{\mathrm{SN}} = 0.6 \cdot10^{51}\ \mathrm{erg}; mPopIII=43 M⊙m_{\mathrm{PopIII}} = 43\,M_{\odot}

(c) Best fit: 85% Pop III fSN with ESN=0.6⋅1051 ergE_{\mathrm{SN}} = 0.6 \cdot10^{51}\ \mathrm{erg}; mPopIII=43 M⊙m_{\mathrm{PopIII}} = 43\,M_{\odot}

Parameter space (progenitor mass, mPopIIIm_{\mathrm{PopIII}}, on the x-axis; SN energy, ESNE_{\mathrm{SN}}, on the y-axis) for the three fits displayed in Figure 8 ((a) blue fit, (b) red fit, and (c) green fit respectively), with every bin colour-coded according to the goodness of the fit (χν2/χν,best2\chi^2_\nu/\chi^2_{\nu,\mathrm{best}}). Darkest blue bins correspond to the best fits. The two progenitors plots only show the parameter space corresponding to the major contributor to the enrichment.

Notably, the secondary progenitor of DR20080 in our best-fit model is consistent with the main progenitor inferred for the most metal-poor star in Sculptor, AS0039: a Pop III HN with mPopIII≈20M⊙m_{\mathrm{PopIII}} \approx20M_{\odot} and ESN=10⋅1051E_{\mathrm{SN}} = 10 \cdot10^{51} erg ((Skúladdóttir et al. 2021, 2024)). This may indicate a connection between the formation environments. Particularly, it is possible that the same Pop III HN enriched both AS0039 and DR20080. Moreover, it should be noticed that both stars belong to the subpopulation identified by [12] and interpreted as the remnant of a past UFD merger. However, given the uncertainties on the secondary progenitor of DR20080, this interpretation should be treated with caution.

The results of this study strongly suggest that DR20080 is a strong candidate for being a direct descendant of Pop III stars. The evidence points to a Pop III star with a mass of mPopIII=43M⊙m_{\mathrm{PopIII}} = 43M_{\odot} that exploded as a fSN with an energy of ESN=0.6⋅1051E_{\mathrm{SN}} = 0.6 \cdot10^{51} erg as the main contributor to its chemical enrichment, with a minor contribution from a HN. This study confirms the crucial role of CEMP-no stars in trying to unveil the mystery behind the first stars ever formed in the Universe.

Discussion and Conclusions

DR20080 is a CEMP-no star within the Sculptor dSph, the first confirmed so far at extremely low metallicities ([Fe/H] ≤−3\le-3) and one of the very few identified in a dSph to date. Such stars are indeed very rare in dSphs, which can be the manifestation of an underlying difference in the star formation conditions in this class of galaxy. Moreover, according to the recent work of Arroyo-Polonio et al. [12], DR20080 belongs to a subpopulation of stars that may have formed in a UFD and were then accreted into Sculptor.

In this work, we conducted the analysis of the high-resolution spectra of this star, taken with VLT/UVES in three spectral bands, all within the visible range. These high-quality spectra allowed for the determination of chemical abundances for 17 elements and the estimation of upper limits for 4 others (see § 4). This makes the analysis presented here one of the most detailed studies of chemical abundances ever conducted in a dwarf galaxy at such low metallicities. In particular, the high abundance of carbon and the underabundance of barium, along with the non-detection of europium ([Eu/Fe] <+0.50< +0.50), confirm that this star belongs to the CEMP-no class. Subsequently, state-of-the-art corrections for NLTE effects were applied to all the abundances of elements for which such corrections were available.

When comparing the chemical abundances of DR20080 with other stars of similar metallicity observed in Sculptor, dSphs, UFDs, and the MW halo, it becomes evident that the derived pattern is unique, at least in terms of what has been observed in the literature so far, both among stars in Sculptor and among extremely metal-poor stars in general. This conclusion is particularly true in relation to other CEMP-no stars in the MW halo and in UFDs, and points towards the existence of a dependency between CEMP-no properties and the type of galaxy in which they formed (see § 5).

The high-resolution spectral analysis of DR20080 also gave us various hints about the extremely metal-poor gas from which this star formed. First of all, CEMP-no stars are generally thought to have formed in an environment enriched by Pop III stars exploded as fSNe. Furthermore, the low [Si/O] ratio indicates that Pop III stars are the most probable main contributors to the chemical composition of DR20080 atmosphere. Finally, by comparing the derived 1D NLTE abundances with theoretical SN yield models (considering both single- and two-progenitor scenarios, for Pop III and Pop II stars), the case of DR20080 is best reproduced by a two–Pop III-progenitor model. In this scenario, the dominant contributor is a Pop III fSN of mass mPopIII=43M⊙m_{\mathrm{PopIII}} = 43M_{\odot} and explosion energy ESN=0.6⋅1051 ergE_{\mathrm{SN}} = 0.6 \cdot10^{51}\ \mathrm{erg}, while the secondary progenitor is a Pop III HN with mPopIII=19.4M⊙m_{\mathrm{PopIII}} = 19.4M_{\odot} and ESN=10⋅1051 ergE_{\mathrm{SN}} = 10 \cdot10^{51}\ \mathrm{erg} (see § 6). Despite the degeneracies affecting the determination of the secondary progenitor, our best fit is consistent with the proposed progenitor of the star AS0039, the most metal-poor star known in Sculptor ([188, 190]), which may suggest a connection between the two formation environments. These results lead us to conclude that DR20080 is an excellent candidate for being a direct descendant of Pop III stars, expanding our view of the first stellar population.

Our results further highlight the importance of increasing the sample of stars observed in dwarf galaxies, particularly through high-resolution spectroscopy. Indeed, this would improve the statistics on the rarity of CEMP-no stars within dSphs and help to clarify the role of host galaxy characteristics on CEMP-no properties, thereby expanding our understanding of the possible environment dependence of this class of stars. In this context, the 4DWARFS survey (Skúladóttir et al. [187]), which saw its first light in October 2025, represents one of the most promising opportunities to shed light on these aspects by increasing the sample of observed stars in dwarf galaxies by orders of magnitude. Once a potential CEMP-no candidate is identified, this work also highlights how crucial it is to perform a high-quality follow-up, which can lead to significantly different results compared to low-resolution studies, and help break the degeneracy among progenitor models thanks to the larger number of detected elements.

Acknowledgements

This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 101117455). Based on observations carried out with ESO/VLT/UVES at the La Silla Paranal Observatory (program ID 110.2407, PI: Á. Skúladóttir).

Individual los velocities

ExposureDateB-band vlosv_{\mathrm{los}} [km/s]B-band Δvlos\Delta v_{\mathrm{los}} [km/s]Rl-band vlosv_{\mathrm{los}} [km/s]Rl-band Δvlos\Delta v_{\mathrm{los}} [km/s]Ru-band vlosv_{\mathrm{los}} [km/s]Ru-band Δvlos\Delta v_{\mathrm{los}} [km/s]
125/10/2022119.20±0.30\pm0.30119.25±0.24\pm0.24121.36±0.70\pm0.70
229/10/2022121.00±0.31\pm0.31120.33±0.26\pm0.26119.65±0.71\pm0.71
318/11/2022121.42±0.41\pm0.41120.66±0.31\pm0.31120.02±0.69\pm0.69
418/11/2022123.57±0.50\pm0.50121.68±0.27\pm0.27119.94±0.71\pm0.71
520/11/2022119.36±0.28\pm0.28120.07±0.24\pm0.24120.96±0.62\pm0.62
621/11/2022121.02±0.34\pm0.34120.83±0.25\pm0.25118.70±0.72\pm0.72
721/11/2022120.12±0.26\pm0.26120.24±0.22\pm0.22117.82±0.66\pm0.66
822/11/2022119.53±0.32\pm0.32119.47±0.26\pm0.26117.27±0.58\pm0.58

Table A.1. True los velocity value for each band and exposure.

The observed los velocity for each exposure, vlos,iobsv_{\mathrm{los},i}^{\mathrm{obs}}, where ii indicates ii-th exposure, was measured exploiting one of the features of the program DAOSPEC. This tool (see also Subsection 3.1) takes a literature list of lines of interest and an observed spectrum as input and, through a cross-check, outputs the observed los velocity associated with that specific spectrum, as well as its variance σvlos,iobs\sigma_{v_{\mathrm{los},i}}^{\mathrm{obs}} and the number of lines nlinesn_{\mathrm{lines}} that were used for the measurement.

The true los velocity of DR20080, vlos,iv_{\mathrm{los},i}, is the observed value corrected to the heliocentric reference frame, while the uncertainty was evaluated as

Δvlos,i=σvlos,iobsnlines,i−1.\Delta v_{\mathrm{los},i} = \frac{\sigma_{v_{\mathrm{los},i}}^{\mathrm{obs}}}{\sqrt{n_{\mathrm{lines},i}-1}} .

All the results are presented in Table A.1 and plotted in Figure A.1.

The mean value was estimated as an arithmetic mean of all the velocity measurements, while the uncertainty was estimated using the following equation

Δvlos=σnobs−1,\Delta v_{\mathrm{los}} = \frac{\sigma}{\sqrt{n_{\mathrm{obs}}-1}} ,

where σ\sigma is the standard deviation between all the measurements and nobs=24n_{\mathrm{obs}} = 24 is the total number of exposures. This method gave the result vlos=120.15±0.27v_{\mathrm{los}} = 120.15 \pm0.27 km s−1^{-1}.

Los velocities of all 24 exposures shown as a function of exposure number and observation date

Figure A.1. Los velocities of all 24 exposures after the heliocentric velocity correction, plotted as a function of the exposure number (top panel) and the observation date (bottom panel). Each band is represented by a different colour. The black dashed line is the arithmetic mean value of vlosv_{\mathrm{los}}, the grey ones are the two extremes defined by Δvlos\Delta v_{\mathrm{los}} and the green ones by the standard deviation, σ\sigma, from the mean.

Analysed lines

Line list

Table B.1 presents the line list, together with the corresponding EWEW value derived using ASTRA. The line list is a compilation of the VALD database ([152, 170, 103, 104]) and the works from [210, 84] (Tafelmeyer et al. 2010) (see also Subsection 3.1). When the EWEW value is reported in parentheses, the corresponding abundance ratio [X/Fe], listed in the fourth column, was derived through spectral synthesis rather than from the EWEW measurement. In these cases, the reported EWEW value should be regarded only as indicative and was not used for a reliable abundance determination (e.g., due to line blending). Moreover, for all individual lines for which a NLTE correction is available in the literature for our adopted stellar parameters, the corresponding correction is reported (Δ[X/H]NLTE\Delta[\mathrm{X}/\mathrm{H}]_{\mathrm{NLTE}}). In the same table are also reported the two molecular bands analysed in this work, namely CH and CN, together with their reference wavelength. Finally, we also report the upper limits derived for individual lines that were too weak to be detected, but correspond to elements whose abundance ratios are of particular interest.

Additional spectral fits

Figure B.1 reports the CN molecular band around 3883 Å used to derive the abundance of nitrogen. In Figure B.2 is reported one of the 13CH{}^{13}\mathrm{CH} absorption features (Masseron et al. [134]) used for the estimation of log⁡(12C/13C)\log\left({}^{12}\mathrm{C}/{}^{13}\mathrm{C}\right). Figure B.3 shows one of the three Mg i lines used to derive the abundance value of magnesium.

Ionλ\lambda [Å]χex\chi_{\mathrm{ex}} [eV]log⁡(gf)\log(gf)EWEW [mÅ][X/Fe] [dex]Δ[X/H]NLTE\Delta[\mathrm{X}/\mathrm{H}]_{\mathrm{NLTE}} [dex]
Cobs_{\mathrm{obs}} (CH)4300−−(−)+1.07−
N (CN)3883−−−+1.72−
O i6300.3040.000−9.750(12.5)+1.53+0.000
Na i5889.9510.0000.108(138.4)+0.33−0.313
Na i5895.9240.000−0.194(112.7)+0.09−0.198
Mg i5172.6842.712−0.450(155.2)+0.61+0.001
Mg i5183.6042.717−0.239(158.4)+0.48+0.002
Mg i5528.4054.346−0.498(35.0)+0.33+0.118
Al i3961.5200.014−0.323(125.0)−0.17+1.00
Si i3905.5231.909−1.041(−)−0.30−0.091
Ca i4226.7280.0000.244(146.7)+0.03+0.004
Ca i4425.4371.879−0.358(27.7)+0.31+0.297
Ca i4454.7791.8990.258(59.5)+0.22+0.308
Ca i6122.2171.886−0.316(27.7)+0.14+0.270
Ca i6162.1731.899−0.09045.8+0.28+0.264
Ca i6439.0752.5260.390(21.7)+0.04+0.248
Sc ii4246.8220.3150.24(100.5)−0.11−
Sc ii4226.7280.0000.244(146.7)+0.03−
Ti i4981.7300.8480.570(17.2)−0.25+0.495
Ti i4991.0660.8360.45027.4+0.11+0.528
Ti i5014.2760.8130.040(14.3)−0.06+0.569
Ti i5210.3840.048−0.820(17.3)+0.14+0.653
Ti ii3759.2930.6070.280(138.6)−0.28−0.112
Ti ii3761.3220.5740.180(120.0)−0.23−0.064
Ti ii3913.4611.116−0.360(88.9)−0.25+0.124
Ti ii4028.3381.892−0.92035.2+0.34+0.139
Ti ii4290.2151.165−0.870(84.6)+0.20+0.155
Ti ii4395.0311.084−0.540(93.5)+0.24+0.104
Ti ii4399.7651.237−1.20080.7+0.66+0.154
Ti ii4417.7131.165−1.190(61.6)+0.38+0.132
Ti ii4418.3311.237−1.99049.0+0.82+0.132
Ti ii4443.8011.080−0.71089.3+0.17+0.009
Ti ii4468.4931.131−0.630(100.8)+0.09+0.011
Ti ii5129.1561.892−1.340(16.0)+0.20+0.099
Ti ii5188.6871.582−1.05035.9−0.02+0.101
Ti ii5226.5391.566−1.26033.3+0.12+0.101
Ti ii5336.7861.582−1.60020.9+0.20+0.118
V i4379.2300.3010.580−< −0.10−
V i4384.7130.2870.410−< +0.70−
Cr i4254.3520.000−0.090(104.4)−0.25+0.644
Cr i5206.0230.9410.020(64.0)−0.20+0.533
Cr i5208.4090.9410.170(77.1)−0.19+0.526
Cr i5409.7841.030−0.670(18.2)−0.30+0.478
Mn i4030.7500.000−0.49(98.6)−0.94+0.153
Mn i4033.0600.000−1.817(76.6)−1.09+0.184

Table B.1. Line list (see text for references), derived LTE abundances and NLTE corrections for the star DR20080. Abundances of lines with the EWEW reported in parentheses were derived with spectral synthesis. The abundances are all referred to [Fe/H]mean=−3.26[\mathrm{Fe}/\mathrm{H}]_{\mathrm{mean}} = -3.26.

Analysed lines

Ionλ\lambda [Å]χex\chi_{\mathrm{ex}} [eV]log⁡(gf)\log(g f)EWEW [mÅ][X/Fe] [dex]Δ\Delta[X/H]NLTE_{\mathrm{NLTE}} [dex]
Fe I4021.8662.759−0.72936.5−0.04−
Fe I4032.6271.485−2.37725.6−0.17−
Fe I4062.4412.845−0.86222.6−0.12−
Fe I4067.9783.211−0.47224.9−0.02−
Fe I4076.6293.211−0.52921.8−0.05−
Fe I4107.4882.832−0.87936.7+0.19−
Fe I4132.8992.845−1.00634.9+0.29−
Fe I4134.6772.832−0.64931.8−0.15−
Fe I4147.6681.485−2.10443.1−0.09−
Fe I4153.8993.397−0.32130.1+0.16−
Fe I4154.4992.832−0.68833.8−0.07−
Fe I4175.6362.845−0.82740.2+0.22−
Fe I4181.7542.832−0.37147.4−0.11−
Fe I4187.0392.450−0.54856.9−0.21+0.170
Fe I4187.7952.425−0.55479.6+0.28−
Fe I4191.4302.469−0.66656.2−0.09−
Fe I4222.2132.450−0.96750.2+0.07+0.178
Fe I4227.4263.3320.26648.5−0.14−
Fe I4233.6032.482−0.60464.0+0.02−
Fe I4238.8103.397−0.23326.3−0.02−
Fe I4250.1192.469−0.40569.5−0.07−
Fe I4282.4032.176−0.77967.7−0.08+0.190
Fe I4442.3392.198−1.25559.8+0.20−
Fe I4447.7172.223−1.34238.2−0.09+0.197
Fe I4494.5632.198−1.13656.9+0.02+0.202
Fe I4859.7412.876−0.76436.8+0.05−
Fe I4871.3182.865−0.36353.4−0.06−
Fe I4872.1382.882−0.56742.7−0.03−
Fe I4890.7552.876−0.39443.2−0.20+0.179
Fe I4891.4922.851−0.11262.0−0.17+0.179
Fe I4903.3102.882−0.92622.4−0.09+0.185
Fe I4918.9942.865−0.34256.9−0.02−
Fe I4920.5022.8330.06879.4−0.02−
Fe I4938.8142.876−1.07722.0+0.04−
Fe I5006.1192.833−0.63847.6+0.06+0.197
Fe I5041.7561.485−2.20351.2+0.04−
Fe I5049.8202.279−1.35536.2−0.11+0.201
Fe I5068.7662.940−1.04223.6+0.11−
Fe I5171.5961.485−1.79381.7+0.21−
Fe I5191.4553.039−0.55132.4−0.08+0.191
Fe I5192.3442.998−0.42143.6−0.05−
Fe I5194.9411.557−2.09058.1+0.13+0.198
Fe I5202.3362.176−1.83825.5+0.01+0.197
Fe I5216.2741.608−2.15047.3+0.05+0.193
Fe I5232.9402.940−0.05857.7−0.23+0.182
Fe I5266.5552.998−0.38647.2−0.02−
Fe I5281.7903.039−0.83423.7+0.01+0.194
Fe I5283.6213.241−0.43227.7−0.06−
Fe I5324.1793.211−0.10345.7−0.08+0.217
Fe I5328.5311.557−1.85080.1+0.30−
Fe I5332.8991.557−2.77721.1+0.08−
Fe I5424.0684.3200.52023.0+0.16−
Fe I5572.8423.397−0.27529.8−0.00+0.210
Fe I6136.6152.453−1.40024.3−0.18+0.179
Fe I6137.6912.588−1.40324.6−0.00+0.175
Fe I6191.5582.433−1.41728.3−0.10+0.201
Fe I6230.7222.559−1.28132.1−0.01+0.203
Fe I6252.5552.404−1.68732.1+0.21+0.201
Fe I6393.6012.433−1.43224.9−0.17+0.209
Fe I6430.8452.176−2.00627.9+0.15+0.000
Fe I6494.9802.404−1.27347.5+0.05−
Fe II4233.1622.583−1.900(59.3)−0.04−0.006
Fe II4923.9212.891−1.32076.7+0.00+0.047
Fe II5018.4362.891−1.22086.6+0.10−
Fe II5197.5673.230−2.10022.0+0.09+0.000
Fe II5275.9973.199−1.940(24.5)−0.08−
Co I3894.0771.0490.100(64.5)+0.37+0.811
Co I3995.3070.923−0.22(59.7)−0.18+0.649
Co I4118.7731.049−2.011(37.8)−0.02+0.891
Co I4121.3180.923−1.489(55.5)−0.10+0.816
Ni I5476.9041.826−0.780(27.3)−0.45+0.185
Cu I5105.5371.389−1.54−< +0.40−
Cu I5700.2371.642−2.58−< +1.70−
Cu I5782.1271.642−1.91−< +1.00−
Zn I4722.1534.030−0.338−< +0.70−
Zn I4810.5284.078−0.137−< +0.50−
Sr II4077.7090.0000.167(70.9)−1.73+0.288
Sr II4215.5190.000−0.145(65.5)−1.73+0.329
Ba II4934.0760.000−0.15(36.1)−1.28+0.206
Eu II3907.1060.2070.171−< +1.00−
Eu II4129.7080.0000.22−< +0.50−
Eu II4205.0420.0000.21−< +0.50−

Table B.1 (continued).

CN molecular band spectrum

Figure B.1. As Figure 2, but for the CN molecular band (∼3883\sim3883 Å) − B-band.

CH molecular band spectrum

(a) The CH molecular band (∼4300\sim4300 Å) – B-band. Two abundances reported: the observed one and when corrected for mixing.

O I spectral line

(b) O I (6300.304 Å) – Ru-band.

Si I spectral line

(c) Si I (3905.523 Å) – B-band.

Sr II spectral line

(d) Sr II (4077.709 Å) – B-band.

Ba II spectral line

(e) Ba II (4934.076 Å) – Rl-band.

Figure 2. Fig. 2: Examples of key spectral lines. Black solid line is the observed normalised spectrum, green is the best fit, and orange is a reference synthetic spectrum in the absence of element X. In (b), (c), (d) and (e) the shaded green area indicates the uncertainty on the abundance measurement, and the magenta segment indicates the central wavelength of the transition.

:::

13CH absorption feature spectrum

Figure B.2. As Figure 2, but for the 13CH^{13}\mathrm{C}\mathrm{H} absorption feature at ∼4213\sim4213 Å − B-band.

Mg I absorption line spectrum

Figure B.3. As Figure 2, but for the Mg I absorption line at 5183.604 Å − Rl-band.

Abundance ratios

Figure C.1 shows the N abundance together with the upper limits derived for V, Cu, Zn and Eu in DR20080, compared with stars in the MW and its satellite system. The corresponding references are reported in Subsection 5.1 and are the same as for Figure 3 to Figure 5. For V, Cu and Eu, the MW comparison sample consists of a compilation of data from Lombardo et al. [116, 117] together with measurements retrieved from the JINA database, including data from: McWilliam et al. [137]; Norris et al. [146]; Westin et al. [215]; Aoki et al. [6, 10, 9]; Johnson [90]; Sneden et al. [192]; Honda et al. [79, 78]; Barbuy et al. [14]; Barklem et al. [15]; Masseron et al. [135, 133, 132]; Preston et al. [158]; Sivarani et al. [186]; Frebel et al. [51]; Cohen et al. [39, 40]; Lai et al. [105]; Hayek et al. [71]; Behara et al. [19]; Roederer et al. [165]; Hollek et al. [76]; Allen et al. (2012); Ishigaki et al. [81]; Placco et al. [154, 153]; Çalıskan et al. [220]; Mashonkina et al. [125]; Siqueira Mello et al. [183]; Spite et al. [196]; Hansen et al. [68]; Jacobson et al. [85]; and Li et al. (2015b).

In Figure C.2 are reported the same data as in Figure 6, but with the abundance ratios referred to hydrogen, [X/H], instead of iron.

Figure C.1. Same as Figure 3, but for [N/Fe][\mathrm{N}/\mathrm{Fe}], [V/Fe][\mathrm{V}/\mathrm{Fe}], [Cu/Fe][\mathrm{Cu}/\mathrm{Fe}], [Zn/Fe][\mathrm{Zn}/\mathrm{Fe}], and [Eu/Fe][\mathrm{Eu}/\mathrm{Fe}]. See Subsection 5.1 and Appendix C for references.

Figure C.2. As Figure 6, but the abundances are reported as [X/H][\mathrm{X}/\mathrm{H}].

References

  1. [1]Abohalima, A. & Frebel, A. 2018, ApJS, 238, 36
  2. [2]Aguado, D. S., González Hernández, J. I., Prieto, C. A., & Rebolo, R. 2019, ApJ, 874, L21
  3. [4]Amarsi, A. M., Nissen, P. E., Asplund, M., Lind, K., & Barklem, P. S. 2019, A&A, 622, L4
  4. [5]Anthony-Twarog, B. J., Delyannidis, C. P., Rich, E., & Twarog, B. A. 2013, ApJ, 767, L19
  5. [6]Aoki, W., Ando, H., Honda, S., et al. 2002a, PASJ, 54, 427
  6. [7]Aoki, W., Arimoto, N., Sadakane, K., et al. 2009, A&A, 502, 569
  7. [8]Aoki, W., Beers, T. C., Christlieb, N., et al. 2007, ApJ, 655, 492
  8. [9]Aoki, W., Beers, T. C., Lee, Y. S., et al. 2013, AJ, 145, 13
  9. [10]Aoki, W., Honda, S., Beers, T. C., et al. 2005, ApJ, 632, 611
  10. [12]Arroyo-Polonio, J. M., Battaglia, G., Thomas, G. F., et al. 2024, A&A, 692, A195
  11. [13]Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, ARA&A, 47, 481
  12. [14]Barbuy, B., Spite, M., Spite, F., et al. 2005, A&A, 429, 1031
  13. [15]Barklem, P. S., Christlieb, N., Beers, T. C., et al. 2005, A&A, 439, 129
  14. [16]Battaglia, G., Helmi, A., Tolstoy, E., et al. 2008, ApJ, 681, L13
  15. [17]Battaglia, G., Taibi, S., Thomas, G. F., & Fritz, T. K. 2022, A&A, 657, A54
  16. [18]Beers, T. C. & Christlieb, N. 2005, ARA&A, 43, 531
  17. [19]Behara, N. T., Bonifacio, P., Ludwig, H.-G., et al. 2010, A&A, 513, A72
  18. [20]Bergemann, M. 2011, MNRAS, 413, 2184
  19. [21]Bergemann, M. & Cescutti, G. 2010, A&A, 522, A9
  20. [22]Bergemann, M., Collet, R., Amarsi, A. M., et al. 2017, ApJ, 847, 15
  21. [23]Bergemann, M. & Gehren, T. 2008, A&A, 492, 823
  22. [24]Bergemann, M., Kudritzki, R.-P., Wü̈rl, M., et al. 2013, ApJ, 764, 115
  23. [25]Bergemann, M., Lind, K., Collet, R., Magic, Z., & Asplund, M. 2012, MNRAS, 427, 27
  24. [26]Bergemann, M., Pickering, J. C., & Gehren, T. 2010, MNRAS, 401, 1334
  25. [27]Bettinelli, M., Hidalgo, S. L., Cassisi, S., et al. 2019, MNRAS, 487, 5862
  26. [28]Bonifacio, P., Caffau, E., François, P., & Spite, M. 2025, A&A, 33, 2
  27. [29]Bonifacio, P., Caffau, E., Spite, M., et al. 2015, A&A, 579, A28
  28. [30]Bonifacio, P., Caffau, E., Spite, M., et al. 2018, A&A, 612, A65
  29. [31]Bromm, V., Coppi, P. S., & Larson, R. B. 2002, ApJ, 564, 23
  30. [32]Bromm, V., Ferrara, A., Coppi, P., & Larson, R. B. 2001, MNRAS, 328, 969
  31. [33]Caffau, E., Bonifacio, P., François, P., et al. 2011, Nat, 477, 67
  32. [34]Carretta, E., Bragaglia, A., Gratton, R. G., & Tosi, M. 2004, A&A, 422, 951
  33. [35]Cayrel, R., Depagne, E., Spite, M., et al. 2004, A&A, 416, 1117
  34. [36]Chiti, A., Simon, J. D., Frebel, A., et al. 2018, ApJ, 856, 142
  35. [37]Choplin, A. & Meynet, G. 2026, EPJA, 62, 43
  36. [38]Christlieb, N., Gustafsson, B., Korn, A. J., et al. 2004, ApJ, 603, 708
  37. [39]Cohen, J. G., Christlieb, N., McWilliam, A., et al. 2008, ApJ, 672, 320
  38. [40]Cohen, J. G., Christlieb, N., Thompson, I., et al. 2013, ApJ, 778, 56
  39. [41]Cohen, J. G. & Huang, W. 2009, ApJ, 701, 1053
  40. [42]Cohen, J. G. & Huang, W. 2010, ApJ, 719, 931
  41. [43]De Boer, T. J. L., Tolstoy, E., Hill, V., et al. 2012, A&A, 539, A103
  42. [44]Ezzeddine, R., Frebel, A., Roederer, I. U., et al. 2019, ApJ, 876, 97
  43. [45]Fabrizio, M., Nonino, M., Bono, G., et al. 2015, A&A, 580, A18
  44. [46]Feltzing, S., Eriksson, K., Kleyna, J., & Wilkinson, M. I. 2009, A&A, 508, L1
  45. [47]Fernandes De Melo, R., Lombardo, L., Alekstro Pulis, A., et al. 2024, A&A, 691, A220
  46. [48]François, P., Caffau, E., Wanajo, S., et al. 2018, A&A, 619, A10
  47. [49]François, P., Monaco, L., Bonifacio, P., et al. 2016, A&A, 588, A7
  48. [50]Frebel, A., Chiti, A., Ji, A. P., Jacobson, H. R., & Placco, V. M. 2015, ApJ, 810, L27
  49. [51]Frebel, A., Christlieb, N., Norris, J. E., et al. 2007a, ApJ, 660, L117
  50. [52]Frebel, A., Kirby, E. N., & Simon, J. D. 2010a, Nat, 464, 72
  51. [53]Frebel, A. & Norris, J. E. 2015, ARA&A, 53, 631
  52. [56]Frebel, A., Simon, J. D., Geha, M., & Willman, B. 2010b, ApJ, 708, 560
  53. [57]Frebel, A., Simon, J. D., & Kirby, E. N. 2014, ApJ, 786, 74
  54. [58]Fulbright, J. P., Rich, R. M., & Castro, S. 2004, ApJ, 612, 447
  55. [59]Gaia Collaboration, Babusiaux, C., van Leeuwen, F., et al. 2018a, A&A, 616, A10
  56. [60]Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al. 2021, A&A, 649, A1
  57. [61]Gaia Collaboration, Helmi, A., van Leeuwen, F., et al. 2018b, A&A, 616, A12
  58. [62]Gilmore, G., Norris, J. E., Monaco, L., et al. 2013, ApJ, 763, 61
  59. [63]Gratton, R. G., Sneden, C., Carretta, E., & Bragaglia, A. 2000, A&A, 354, 169
  60. [64]Green, M. G. 2018, JOSS, 3, 695
  61. [65]Gustafsson, B., Edvardsson, B., Eriksson, K., et al. 2008, A&A, 486, 951
  62. [66]Hampel, M., Stancliffe, R. J., Lugaro, M., & Meyer, B. S. 2016, ApJ, 831, 171
  63. [67]Hansen, C. J., El-Souri, M., Monaco, L., et al. 2018, ApJ, 855, 83
  64. [68]Hansen, T., Hansen, C. J., Christlieb, N., et al. 2015, ApJ, 807, 173
  65. [69]Hansen, T., Hansen, C. J., Christlieb, N., et al. 2014, ApJ, 787, 162
  66. [70]Hansen, T., Marshall, J. L., Simon, J. D., et al. 2020, ApJ, 897, 183
  67. [71]Hayek, W., Wiesendahl, U., Christlieb, N., et al. 2009, A&A, 504, 511
  68. [72]Heger, A. & Woosley, S. E. 2002, ApJ, 567, 532
  69. [73]Heger, A. & Woosley, S. E. 2010, ApJ, 724, 341
  70. [74]Hendricks, B., Koch, A., Walker, M., et al. 2014, A&A, 572, A82
  71. [75]Hirano, S., Hosokawa, T., Yoshida, N., et al. 2014, ApJ, 781, 60
  72. [76]Hollek, J. K., Frebel, A., Roederer, I. U., et al. 2011, ApJ, 742, 54
  73. [77]Honda, S., Aoki, W., Arimoto, N., & Sadakane, K. 2011a, PASJ, 63, 523
  74. [78]Honda, S., Aoki, W., Beers, T. C., & Takada-Hidai, M. 2011b, ApJ, 730, 77
  75. [79]Honda, S., Aoki, W., Kajino, T., et al. 2004, ApJ, 607, 474
  76. [80]Ishigaki, M. N., Aoki, W., Arimoto, N., & Okamoto, S. 2014, A&A, 562, A146
  77. [81]Ishigaki, M. N., Aoki, W., & Chiba, M. 2013, ApJ, 771, 67
  78. [82]Iwamoto, K., Brachwitz, F., Nomoto, K., et al. 1999, ApJS, 125, 439
  79. [83]Iwamoto, N., Umeda, H., Tominaga, N., Nomoto, K., & Maeda, K. 2005, Sci, 309, 451
  80. [84]Jablonka, P., North, P., Mashonkina, L., et al. 2015, A&A, 583, A67
  81. [85]Jacobson, H. R., Keller, S., Frebel, A., et al. 2015, ApJ, 807, 171
  82. [90]Johnson, J. A. 2002, ApJS, 139, 219
  83. [91]Keller, S. C., Bessell, M. S., Frebel, A., et al. 2014, Nat, 506, 463
  84. [92]Kirby, E. N. & Cohen, J. G. 2012, AJ, 144, 168
  85. [93]Kirby, E. N., Cohen, J. G., Simon, J. D., et al. 2017, ApJ, 838, 83
  86. [94]Kirby, E. N., Guo, M., Zhang, A. J., et al. 2015, ApJ, 801, 125
  87. [95]Klessen, R. S. & Glover, S. C. O. 2023, ARA&A, 61, 65
  88. [96]Kobayashi, C., Karakas, A. I., & Lugaro, M. 2020, ApJ, 900, 179
  89. [97]Koch, A., Grebel, E. K., Gilmore, G. F., et al. 2008, AJ, 135, 1580
  90. [98]Koutsouridou, I., Salvadori, S., & Skúladóttir, Á. 2024, ApJ, 962, L26
  91. [99]Koutsouridou, I., Salvadori, S., Skúladóttir, Á., et al. 2026, NERFERTI: Linking early galaxy formation to the assembly of the Milky Way, submitted to ApJ; ADS Bibcode: 2026arXiv260500990Karxiv.org/abs/2605.00990
  92. [100]Koutsouridou, I., Salvadori, S., Skúladóttir, Á., et al. 2023, MNRAS, 525, 190
  93. [101]Koutsouridou, I., Skúladóttir, Á., & Salvadori, S. 2025, A&A, 699, A32
  94. [102]Kovalev, M., Brinkmann, S., Bergemann, M., & MPIA IT-department. 2018, published: NLTE MPIA web server, [Online]. Max Planck Institute for Astronomy, Heidelberg.
  95. [103]Kupka, F., Piskunov, N., Ryabchikova, T. A., Stempels, H. C., & Weiss, W. W. 1999, A&AS, 138, 119
  96. [104]Kupka, F. G., Ryabchikova, T. A., Piskunov, N. E., Stempels, H. C., & Weiss, W. W. 2000, BaltA, 9, 590
  97. [105]Lai, D. K., Bolte, M., Johnson, J. A., et al. 2008, ApJ, 681, 1524
  98. [106]Lai, D. K., Lee, Y. S., Bolte, M., et al. 2011, ApJ, 738, 51
  99. [107]Lee, Y. S., Beers, T. C., Masseron, T., et al. 2013, AJ, 146, 132
  100. [108]Lemasle, B., De Boer, T. J. L., Hill, V., et al. 2014, A&A, 572, A88
  101. [109]Lemasle, B., Hill, V., Tolstoy, E., et al. 2012, A&A, 538, A100
  102. [110]Letarte, B., Hill, V., Tolstoy, E., et al. 2010, A&A, 523, A17
  103. [111]Leung, S.-C. & Nomoto, K. 2018, ApJ, 861, 143
  104. [114]Limongi, M. & Chieffi, A. 2018, ApJS, 237, 13
  105. [115]Lind, K., Asplund, M., Barklem, P. S., & Belyaev, A. K. 2011, A&A, 528, A103
  106. [116]Lombardo, L., Bonifacio, P., François, P., et al. 2022, A&A, 665, A10
  107. [117]Lombardo, L., Hansen, C. J., Rizzuti, F., et al. 2025, A&A, 693, A293
  108. [118]Lucatello, S., Beers, T. C., Christlieb, N., et al. 2006, ApJ, 652, L37
  109. [119]Lucchesi, R., Jablonka, P., Skúladóttir, Á., et al. 2024, A&A, 686, A266
  110. [120]Lucchesi, R., Lardo, C., Primas, F., et al. 2020, A&A, 644, A75
  111. [121]Maio, U., Ciardi, B., Dolag, K., Tornatore, L., & Khochfar, S. 2010, MNRAS, 407, 1003
  112. [122]Maiolino, R., Übler, H., Perna, M., et al. 2026, The search for Population III: Confirmation of a HeII emitter with no metal lines at z=10.6, submitted to MNRAS; ADS Bibcode: 2026arXiv260320362Marxiv.org/abs/2603.20362
  113. [124]Marshall, J. L., Hansen, T., Simon, J. D., et al. 2019, ApJ, 882, 177
  114. [125]Mashonkina, L., Christlieb, N., & Eriksson, K. 2014, A&A, 569, A43
  115. [126]Mashonkina, L., Gehren, T., Shi, J.-R., Korn, A. J., & Grupp, F. 2011, A&A, 528, A87
  116. [127]Mashonkina, L., Korn, A. J., & Przybilla, N. 2007, A&A, 461, 261
  117. [128]Mashonkina, L., Pakhomov, Y., Sitnova, T., et al. 2023, MNRAS, 524, 3526
  118. [129]Mashonkina, L., Pakhomov, Y. V., Sitnova, T., et al. 2021, MNRAS, 509, 3626
  119. [130]Mashonkina, L. I. & Belyaev, A. K. 2019, MNRAS, 509, 3626
  120. [131]Mashonkina, L. I., Belyaev, A. K., & Shi, J.-R. 2016, AstL, 42, 366
  121. [132]Masseron, T., Johnson, J. A., Lucatello, S., et al. 2012, ApJ, 751, 14
  122. [133]Masseron, T., Johnson, J. A., Plez, B., et al. 2010, A&A, 509, A93
  123. [134]Masseron, T., Plez, B., Van Eck, S., et al. 2014, A&A, 571, A47
  124. [135]Masseron, T., Van Eck, S., Famey, B., et al. 2006, A&A, 455, 1059
  125. [136]McKee, C. F. & Tan, J. C. 2008, ApJ, 681, 771
  126. [137]McWilliam, A., Preston, G. W., Sneden, C., & Searle, L. 1995, AJ, 109, 2757
  127. [138]Meynet, G., Ekström, S., & Maeder, A. 2006, A&A, 447, 623
  128. [139]Meynet, G., Hirschi, R., Ekström, S., et al. 2010, A&A, 521, A30
  129. [140]Mucciarelli, A. & Bellazzini, M. 2020, MNRAS, 452, 52
  130. [141]Mucciarelli, A., Bellazzini, M., & Massari, D. 2021, A&A, 653, A90
  131. [143]Nagasawa, D. Q., Marshall, J. L., Li, T. S., et al. 2018, ApJ, 852, 99
  132. [144]Nordlander, T. & Lind, K. 2017, A&A, 607, A75
  133. [145]Norris, J. E., Christlieb, N., Korn, A. J., et al. 2007, ApJ, 670, 774
  134. [146]Norris, J. E., Ryan, S. G., & Beers, T. C. 1997, ApJ, 488, 350
  135. [147]Norris, J. E., Yong, D., Bessell, M. S., et al. 2013, ApJ, 762, 28
  136. [148]Norris, J. E., Yong, D., Gilmore, G., & Wyse, R. F. G. 2010, ApJ, 711, 350
  137. [149]Norris, J. E., Yong, D., Venn, K. A., et al. 2017, ApJS, 230, 28
  138. [150]Omukai, K. 2000, ApJ, 534, 809
  139. [151]Pietrzyński, G., Gieren, W., Szewczyk, O., et al. 2008, AJ, 135, 1993
  140. [152]Piskunov, N. E., Kupka, F., Ryabchikova, T. A., Weiss, W. W., & Jeffery, C. S. 1995, A&AS, 112, 525
  141. [153]Placco, V. M., Frebel, A., Beers, T. C., et al. 2014a, ApJ, 781, 40
  142. [154]Placco, V. M., Frebel, A., Beers, T. C., et al. 2013, ApJ, 770, 104
  143. [155]Placco, V. M., Frebel, A., Beers, T. C., & Stancliffe, R. J. 2014b, ApJ, 797, 21
  144. [156]Plez, B. 2012, Astrophysics Source Code Library, ascl:1205.004
  145. [157]Popa, S. A., Hoppe, R., Bergmann, M., et al. 2023, A&A, 670, A25
  146. [158]Preston, G. W., Thompson, I. B., Sneden, C., Stachowski, G., & Shectman, S. A. 2006, AJ, 132, 1714
  147. [159]Reichert, M., Hansen, C. J., Hanke, M., et al. 2020, A&A, 641, A127
  148. [160]Ritter, J. S., Safranek-Shrader, C., Gnat, O., Milosavljević, M., & Bromm, V. 2012, ApJ, 761, 56
  149. [161]Rizzuti, F., Cescutti, G., Molaro, P., et al. 2025, A&A, 698, A118
  150. [163]Roederer, I. U., Mateo, M., Ili, J. I. B., et al. 2016, AJ, 151, 82
  151. [164]Roederer, I. U., Preston, G. W., Thompson, I. B., et al. 2014, AJ, 147, 136
  152. [165]Roederer, I. U., Sneden, C., Thompson, I. B., Preston, G. W., & Shectman, S. A. 2010, ApJ, 711, 573
  153. [166]Romano, D., Karakas, A. I., Tosi, M., & Matteucci, F. 2010, A&A, 522, A32
  154. [167]Rossi, M., QuercI, L., Ciabattini, S., et al. 2026, CRIMSONS: An Online Tool for Modeling Chemical Enrichment with Stochastic IMF Sampling, submitted to A&A; ADS Bibcode: 2026arXiv260925200Rarxiv.org/abs/2609.25200
  155. [168]Rossi, M., Salvadori, S., & Skúladóttir, Á. 2021, MNRAS, 503, 6026
  156. [169]Rusta, E., Salvadori, S., Maiolino, R., et al. 2026, ApJL, 1003, L14
  157. [170]Ryabchikova, T. A., Piskunov, N. E., Kupka, F., & Weiss, W. W. 1997, BaltA, 6, 244
  158. [171]Sadakane, K., Arimoto, N., Ikuta, C., et al. 2004, PASJ, 56, 1041
  159. [172]Salvadori, S., Bonifacio, P., Caffau, E., et al. 2019, MNRAS, 487, 4261
  160. [173]Salvadori, S., Schneider, R., & Ferrara, A. 2007, MNRAS, 381, 647
  161. [174]Salvadori, S., Skúladóttir, Á., & Tolstoy, E. 2015, MNRAS, 454, 1320
  162. [175]Sariya, D. P., Jiang, I.-G., Bisht, D., et al. 2021, AJ, 162, 146
  163. [176]Schneider, R. & Omukai, K. 2010, MNRAS, 402, 429
  164. [177]Schneider, R., Omukai, K., Bianchi, S., & Valiante, R. 2012, MNRAS, 419, 1566
  165. [178]Sestito, F., Vitali, S., Jofre, P., et al. 2024, A&A, 689, A201
  166. [179]Shapley, H. 1938, Nat, 142, 715
  167. [180]Shetrone, M. D., Côté, P., & Sargent, W. L. W. 2001, ApJ, 548, 592
  168. [181]Simon, J. D., Frebel, A., McWilliam, A., Kirby, E. N., & Thompson, I. B. 2010, ApJ, 716, 446
  169. [182]Simon, J. D., Jacobson, H. R., Frebel, A., et al. 2015, ApJ, 802, 93
  170. [183]Siqueira Mello, C., Hill, V., Barbuy, B., et al. 2014, A&A, 565, A93
  171. [184]Sitnova, T. M., Mashonkina, L. I., & Ryabchikova, T. A. 2013, AstL, 39, 126
  172. [185]Sitnova, T. M., Yakovleva, S. A., Belyaev, A. K., & Mashonkina, L. I. 2020, AstL, 46, 120
  173. [186]Sivarani, T., Beers, T. C., Bonifacio, P., et al. 2006, A&A, 459, 125
  174. [187]Skúladóttir, Á., Puls, A. A., Amarsi, A. M., et al. 2023, Published in The Messenger vol. 190, pp. 19-21, 3 pages
  175. [188]Skúladóttir, Á., Salvadori, S., Amarsi, A. M., et al. 2021, ApJ, 915, L30
  176. [189]Skúladóttir, Á., Tolstoy, E., Salvadori, S., et al. 2015, A&A, 574, A129
  177. [190]Skúladóttir, Á., Vanni, I., Salvadori, S., & Lucchesi, R. 2024, A&A, 681, A44
  178. [191]Smith, V. V., Bizyaev, D., Cunha, K., et al. 2021, AJ, 161, 254
  179. [192]Sneden, C., Cowan, J. J., Lawler, J. E., et al. 2003, ApJ, 591, 936
  180. [193]Sneden, C., Lucatello, S., Ram, R. S., Brooke, J. S. A., & Bernath, P. 2014, ApJS, 214, 26
  181. [194]Spite, M., Caffau, E., Bonifacio, P., et al. 2013, A&A, 552, A107
  182. [195]Spite, M., Cayrel, R., Plez, B., et al. 2005, A&A, 430, 655
  183. [196]Spite, M., Spite, F., Bonifacio, P., et al. 2014, A&A, 571, A40
  184. [197]Starkenburg, E., Aguado, D. S., Bonifacio, P., et al. 2018, MNRAS, 481, 3838
  185. [198]Starkenburg, E., Hill, V., Tolstoy, E., et al. 2013, A&A, 549, A88
  186. [199]Starkenburg, E., Shetrone, M. D., McConnachie, A. W., & Venn, K. A. 2014, MNRAS, 441, 1217
  187. [200]Stetson, P. B. & Pancino, E. 2008, PASP, 120, 1332
  188. [201]Suda, T., Hidaka, J., Aoki, W., et al. 2017, PASJ, 69, 76
  189. [202]Suda, T., Katsuta, Y., Yamada, S., et al. 2008, PASJ, 60, 1159
  190. [203]Susmitha, A., Koch, A., & Sivirani, T. 2017, A&A, 606, A112
  191. [205]Tolstoy, E., Hill, V., & Tosi, M. 2009, ARA&A, 47, 371
  192. [206]Tolstoy, E., Skúladóttir, Á., Battaglia, G., et al. 2023, A&A, 675, A49
  193. [207]Tsujimoto, T., Ishigaki, M. N., Shigeyama, T., & Aoki, W. 2015, PASJ, 67, L3
  194. [208]Tumlinson, J. 2006, ApJ, 641, 1
  195. [209]Ural, U., Cescutti, G., Koch, A., et al. 2015, MNRAS, 449, 761
  196. [210]Van Der Swaelmen, M., Hill, V., Primas, F., & Cole, A. A. 2013, A&A, 560, A44
  197. [211]Vanni, I., Salvadori, S., D’Odorico, V., Becker, G. D., & Cupani, G. 2024, ApJ, 967, L22
  198. [212]Vanni, I., Salvadori, S., Skúladóttir, Á., Rossi, M., & Koutsouridou, I. 2023, MNRAS, 526, 2620
  199. [213]Venn, K. A., Shetrone, M. D., Irwin, M. J., et al. 2012, ApJ, 751, 102
  200. [214]Verdiani, V., Skúladóttir, Á., Lucchesi, R., et al. 2026, A&A, 707, A329
  201. [215]Westin, J., Sneden, C., Gustafsson, B., & Cowan, J. J. 2000, ApJ, 530, 783
  202. [216]Woosley, S. E. & Weaver, T. A. 1995, ApJS, 101, 181
  203. [217]Yong, D., Norris, J. E., Bessell, M. S., et al. 2013, ApJ, 762, 26
  204. [218]Yoon, J., Beers, T. C., Dietz, S., et al. 2018, ApJ, 861, 146
  205. [219]Yoon, J., Whitten, D. D., Beers, T. C., et al. 2020, ApJ, 894, 7
  206. [220]Çalışkan, Ş., Caffau, E., Bonifacio, P., et al. 2014, A&A, 571, A62
  207. [221]Übler, H., Maiolino, R., Pérez-González, P. G., et al. 2026, GA-NIFS & JADES: Confirmation of pristine gas near GN-z11, accepted for publication in A&A; ADS Bibcode: 2026arXiv260320360U

Paper details

vlnc:W_dp7hsy9f
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 3

downloads · 1

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents