Detailed Abundance Determination of Metal-Poor Stars with X-Shooter I. Unusual Chemistry in Halo Stars

TL;DR

This work provides a detailed chemical census of 16 metal-poor stars using VLT/X-Shooter spectra and 1D LTE spectral synthesis with Korg, deriving metallicities in the range $[ extrm{Fe/H}] oughly -2.3$ to $-3.8$ and abundances for 16 elements. It confirms prior metallicity estimates and identifies four EMPs, including two likely Gaia-Sausage-Enceladus members with $[ extrm{Fe/H}]$ as low as $-3.82$; the analysis reveals a GSE NEMP star with extreme Na and N and a Li enhancement, a Sr-rich r-II star with mixed r-/s-process signatures, and an N-depleted halo star, underscoring the chemical diversity of the early Milky Way. The study demonstrates that medium-resolution X-Shooter data can yield high-quality abundance patterns comparable to high-resolution studies, enabling broader, fainter-star surveys to probe Galactic assembly and nucleosynthetic channels. Together, these results illuminate the varied origins of the halo and its accreted components, and they show the potential for upcoming large spectroscopic surveys to map the chemical landscape of the early Galaxy.

Abstract

We present a detailed chemical analysis study of 16 candidate metal-poor stars, previously identified with 2dF+AAOmega, using X-Shooter spectra and the Korg 1D LTE spectral synthesis code. We generally confirm the earlier metallicity estimates and reveal four EMP ([Fe/H] $< -3$) stars in the current sample. Two of these stars, including the most metal-poor at [Fe/H] = $-3.83 \pm 0.07$, are kinematically associated with the GSE accretion event, increasing the number of known GSE stars with [Fe/H] $< -3.5$ to eight. From the X-Shooter spectra we also determine abundances for 16 elements, with the element-to-iron abundance ratios generally consistent with high-resolution studies of Galactic halo stars. Within the sample, we identify three peculiar stars: the first is a GSE nitrogen enhanced metal poor (NEMP; [N/Fe] $= 1.62 \pm 0.10$ and [C/Fe] $= 0.27 \pm 0.08$) star with unusually high Na ([Na/Fe] $= 2.28 \pm 0.07$) and Li (A(Li)$_{\rm 3DNLTE} = 1.90 \pm 0.08$) abundances, but which lacks any enhancement in [Al/Fe] or [Mg/Fe]. The second is a halo r-II star significantly enhanced in Sr ([Sr/Ba] $= 0.37 \pm 0.08$), suggesting mixture of r-process and s-process enrichment, uncommon for r-II stars. Whilst the third is a halo star very depleted in N ([N/Fe] $< -1.13$), with low C ([C/Fe] $= -0.36 \pm 0.08$) and otherwise 'normal' [X/Fe] abundances, suggesting enrichment with Type II supernova that proceeds enrichment from massive AGB stars. This study reveals the substantial degree of chemical diversity in the stellar populations which assembled the early Milky Way.

Figures (15)

• Figure 1: Upper panels: Metallicity comparison of the values from the X-Shooter spectra, against the 2dF+AAOmega values from lowe_rise_2025. 1:1 line shown by diagonal dashed line. Shaded region refers to our measured metallicities with $\left[\textrm{Fe/H}\right] \leq -3.0$. Points are colour-coded by their kinematic groupings (see Table \ref{['tab:star properties']}). Lower panels: The difference (2dF+AAOmega minus X-Shooter) between metallicities derived from the X-Shooter and 2dF+AAOmega spectra; the dashed line is for zero difference.
• Figure 2: Chemical abundances for our sample against high-resolution literature values (grey points) yong_abunds_2013jacobson_high-resolution_2015marino_keck_2019yong_high-resolution_2021. Each panel represents a different element measured. Stars are colour-coded based on their orbital classification. Those with upper-limits are shown by downward-facing arrows. The number of stars with measurements, alongside their mean and standard deviations, are given on each panel. For C, the abundance measurements have been evolutionary-corrected (to be consistent with literature values). For CaII, we show the NLTE-corrected abundances, while the literature values were measured from CaI (not NLTE-corrected).
• Figure 3: Abundance patterns $\left[\textrm{X/Fe}\right]$ for each measured element for our 16 metal-poor stars. In each panel, the black dots correspond to the mean literature values yong_abunds_2013jacobson_high-resolution_2015marino_keck_2019yong_high-resolution_2021, with the star symbols corresponding to the abundance value for that given star. Those coloured blue indicate measurements within $2\sigma$ of the mean value, otherwise they are red. Downward-facing arrows are upper-limits, and if these are $2\sigma$ away from the mean, they are also shown in red. Otherwise they are black. The grey shaded region is the standard deviation of the literature values for the particular element. Horizontal grey dashed lines are given at $\left[\textrm{X/Fe}\right] = 0$ and $\left[\textrm{X/Fe}\right] = \pm 1.0$. The star's $T_{\textrm{eff}}$, $\log g$ and $\left[\textrm{Fe/H}\right]$ is given in each panel.
• Figure 4: Evolutionary-corrected C abundances for our sample, with literature shown by the light grey points. Plot is separated into three regions: C-rich ($\left[\textrm{C/Fe}\right] > 0.7$), C-normal ($0 < \left[\textrm{C/Fe}\right] \leq 0.7$) and C-poor ($\left[\textrm{C/Fe}\right] \leq 0$). For the stars with detection: one star is possibly C-rich (two within errors), one is C-poor (one within errors), with the rest C-normal. For stars with evolutionary corrections, their raw measured value is shown by the black arrows.
• Figure 5: Identifying stars as NEMPs by comparing $\left[\textrm{N/Fe}\right]$ with $\left[\textrm{C/N}\right]$ (see Table \ref{['tab:cn values']} for their values). For consistency, we use the uncorrected C abundances when determining $\left[\textrm{C/N}\right]$. Literature is plotted underneath in light grey. Those with upper-limits in $\left[\textrm{N/Fe}\right]$ (represented by leftward-facing arrows), but detections in $\left[\textrm{C/Fe}\right]$ have lower-limits shown for $\left[\textrm{C/N}\right]$ (represented by upward-facing arrows). Stars with upper-limits in both $\left[\textrm{N/Fe}\right]$ and $\left[\textrm{C/Fe}\right]$ are not included. Stars are assigned as NEMP if $\left[\textrm{N/Fe}\right] > 0.5$ and $\left[\textrm{C/N}\right] < -0.5$johnson_search_2007, as shown by the dashed rectangle on the plot. From this, we identify two possible NEMP stars: one likely due to it having high evolutionary mixing corrections for $\left[\textrm{C/Fe}\right]$, raising C and lowering N ($0.53$ dex), with the other unlikely given it has low evolutionary corrections ($0.02$ dex).