Context. Extremely metal-poor (EMP) stars provide us with indirect information on the first generations of massive stars. The TOPoS survey has been designed to increase the census of these stars and to provide a chemical inventory that is as detailed as possible. Aims. Seven of the most iron-poor stars have been observed with the UVES spectrograph at the ESO VLT Kueyen 8.2 m telescope to refine their chemical composition. Methods. We analysed the spectra based on 1D LTE model atmospheres, but also used 3D hydrodynamical simulations of stellar atmospheres. Results. We measured carbon in six of the seven stars: all are carbon-enhanced and belong to the low-carbon band, defined in the TOPoS II paper. We measured lithium (A(Li) = 1.9) in the most iron-poor star (SDSS J1035+0641, [Fe/H] <−5.2). We were also able to measure Li in three stars at [Fe/H] ~−4.0, two of which lie on the Spite plateau. We confirm that SDSS J1349+1407 is extremely rich in Mg, but not in Ca. It is also very rich in Na. Several of our stars are characterised by low α-to-iron ratios. Conclusions. The lack of high-carbon band stars at low metallicity can be understood in terms of evolutionary timescales of binary systems. The detection of Li in SDSS J1035+0641 places a strong constraint on theories that aim at solving the cosmological lithium problem. The Li abundance of the two warmer stars at [Fe/H] ~−4.0 places them on the Spite plateau, while the third, cooler star, lies below. We argue that this suggests that the temperature at which Li depletion begins increases with decreasing [Fe/H]. SDSS J1349+1407 may belong to a class of Mg-rich EMP stars. We cannot assess if there is a scatter in α-to-iron ratios among the EMP stars or if there are several discrete populations. However, the existence of stars with low α-to-iron ratios is supported by our observations.

Manganese (Mn) is a key Fe-group element, commonly employed in stellar population and nucleosynthesis studies to explore the role of SN Ia. We have developed a new non-local thermodynamic equilibrium (NLTE) model of Mn, including new photo-ionisation cross-sections and new transition rates caused by collisions with H and H− atoms. We applied the model in combination with one-dimensional (1D) LTE model atmospheres and 3D hydrodynamical simulations of stellar convection to quantify the impact of NLTE and convection on the line formation. We show that the effects of NLTE are present in Mn I and, to a lesser degree, in Mn II lines, and these increase with metallicity and with the effective temperature of a model. Employing 3D NLTE radiative transfer, we derive a new abundance of Mn in the Sun, A(Mn) = 5.52 ± 0.03 dex, consistent with the element abundance in C I meteorites. We also applied our methods to the analysis of three metal-poor benchmark stars. We find that 3D NLTE abundances are significantly higher than 1D LTE. For dwarfs, the differences between 1D NLTE and 3D NLTE abundances are typically within 0.15 dex, however, the effects are much larger in the atmospheres of giants owing to their more vigorous convection. We show that 3D NLTE successfully solves the ionisation and excitation balance for the RGB star HD 122563 that cannot be achieved by 1D LTE or 1D NLTE modelling. For HD 84937 and HD 140283, the ionisation balance is satisfied, however, the resonance Mn I triplet lines still show somewhat lower abundances compared to the high-excitation lines. Our results for the benchmark stars confirm that 1D LTE modelling leads to significant systematic biases in Mn abundances across the full wavelength range from the blue to the IR. We also produce a list of Mn lines that are not significantly biased by 3D and can be reliably, within the 0.1 dex uncertainty, modelled in 1D NLTE.

Context. Current theory regarding heavy element nucleosynthesis in metal-poor environments states that the r-process would be dominant. The star HD 140283 has been the subject of debate after it appeared in some studies to be dominated by the s-process. Aims. We provide an independent measure of the Ba isotope mixture using an extremely high quality spectrum and an extensive χ 2 analysis. Methods. We have acquired a very high resolution (R ≡ λ/Δλ = 95 000), very high signal-to-noise (S /N = 1110 around 4554 Å, as calculated in IRAF) spectrum of HD 140283. We exploit hyperfine splitting of the Ba ii 4554 Å and 4934 Å resonance lines in an effort to constrain the isotope ratio in 1D LTE. Using the code ATLAS in conjunction with KURUCZ06 model atmospheres we analyse 93 Fe lines to determine the star's macroturbulence. With this information we construct a grid of Ba synthetic spectra and, using a χ 2 code, fit these to our observed data to determine the isotopic ratio, f odd , which represents the ratio of odd to even isotopes. The odd isotopes and 138 Ba are synthesized by the r-and s-process while the other even isotopes ( 134,136 Ba) are synthesized purely by the s-process. We also analyse the Eu lines. Results. We set a new upper limit of the rotation of HD 140283 at v sin i ≤ 3.9 km s −1 , a new upper limit on [Eu/H] < −2.80 and abundances [Fe/H] = −2.59 ± 0.09, [Ba/H] = −3.46 ± 0.11. This leads to a new lower limit on [Ba/Eu] > −0.66. We find that, in the framework of a 1D LTE analysis, the isotopic ratios of Ba in HD 140283 indicate f odd = 0.02 ± 0.06, a purely s-process signature. This implies that observations and analysis do not validate currently accepted theory. Conclusions. We speculate that a 1D code, due to simplifying assumptions, is not adequate when dealing with observations with high levels of resolution and signal-to-noise because of the turbulent motions associated with a 3D stellar atmosphere. New approaches to analysing isotopic ratios, in particular 3D hydrodynamics, need to be considered when dealing with the levels of detail required to properly determine them. However published 3D results exacerbate the disagreement between theory and observation.

Context. The stellar abundances observed in globular clusters show complex structures, currently not yet understood. Aims. The aim of this work is to investigate the relations between the abundances of different elements in the globular cluster M4, selected for its uniform deficiency of iron, to explore the best models explaining the pattern of these observed abundances. Moreover, in turnoff stars, the abundances of the elements are not suspected to be affected by internal mixing. Methods. In M4, using low and moderate resolution spectra obtained for 91 turnoff (and subgiant) stars with the ESO FLAMES-Giraffe spectrograph, we have extended previous measurements of abundances (of Li, C and Na) to other elements (C, Si, Ca, Sr and Ba), using model atmosphere analysis. We have also studied the influence of the choice of the microturbulent velocity. Results. Firstly, the peculiar turnoff star found to be very Li-rich in a previous paper does not show any other abundance anomalies relative to the other turnoff stars in M4. Secondly, an anti-correlation between C and Na has been detected, the slope being significative at more than 3σ. This relation between C and Na is in perfect agreement with the relation found in giant stars selected below the RGB bump. Thirdly, the strong enrichment of Si and of the neutron-capture elements Sr and Ba, already observed in the giants in M4, is confirmed. Finally, the relations between Li, C, Na, Sr and Ba constrain the enrichment processes of the observed stars.Conclusions. The abundances of the elements in the turnoff stars appear to be compatible with production processes by massive AGBs, but are also compatible with the production of second generation elements (like Na) and low Li produced by, for example, fast rotating massive stars.

Context. Recent developments in the three-dimensional (3D) spectral synthesis code Linfor3D have meant that for the first time, large spectral wavelength regions, such as molecular bands, can be synthesised with it in a short amount of time. Aims. A detailed spectral analysis of the synthetic G-band for several dwarf turn-off-type 3D atmospheres (5850 T eff [K] 6550, 4.0 ≤ log g ≤ 4.5, −3.0 ≤ [Fe/H] ≤ −1.0) was conducted, under the assumption of local thermodynamic equilibrium. We also examine carbon and oxygen molecule formation at various metallicity regimes and discuss the impact it has on the G-band. Methods. Using a qualitative approach, we describe the different behaviours between the 3D atmospheres and the traditional onedimensional (1D) atmospheres and how the different physics involved inevitably leads to abundance corrections, which differ over varying metallicities. Spectra computed in 1D were fit to every 3D spectrum to determine the 3D abundance correction. Results. Early analysis revealed that the CH molecules that make up the G-band exhibited an oxygen abundance dependency; a higher oxygen abundance leads to weaker CH features. Nitrogen abundances showed zero impact to CH formation. The 3D corrections are also stronger at lower metallicity. Analysis of the 3D corrections to the G-band allows us to assign estimations of the 3D abundance correction to most dwarf stars presented in the literature. Conclusions. The 3D corrections suggest that A(C) in carbon-enhanced metal-poor (CEMP) stars with high A(C) would remain unchanged, but would decrease in CEMP stars with lower A(C). It was found that the C/O ratio is an important parameter to the G-band in 3D. Additional testing confirmed that the C/O ratio is an equally important parameter for OH transitions under 3D. This presents a clear interrelation between the carbon and oxygen abundances in 3D atmospheres through their molecular species, which is not seen in 1D.

Context. The pursuit of more realistic spectroscopic modelling and consistent abundances has led us to begin a new series of papers designed to improve current solar and stellar abundances of various atomic species. To achieve this, we have begun updating the three-dimensional (3D) non-local thermodynamic equilibrium (non-LTE) radiative transfer code, MULTI3D, and the equivalent one-dimensional (1D) non-LTE radiative transfer code, MULTI 2.3. Aims. We examine our improvements to these codes by redetermining the solar barium abundance. Barium was chosen for this test as it is an important diagnostic element of the s-process in the context of galactic chemical evolution. New Ba II + H collisional data for excitation and charge exchange reactions computed from first principles had recently become available and were included in the model atom. The atom also includes the effects of isotopic line shifts and hyperfine splitting. Methods. A grid of 1D LTE barium lines were constructed with MULTI 2.3 and fit to the four Ba II lines available to us in the optical region of the solar spectrum. Abundance corrections were then determined in 1D non-LTE, 3D LTE, and 3D non-LTE. A new 3D non-LTE solar barium abundance was computed from these corrections. Results. We present for the first time the full 3D non-LTE barium abundance of A(Ba) = 2.27 ± 0.02 ± 0.01, which was derived from four individual fully consistent barium lines. Errors here represent the systematic and random errors, respectively.

Context. Atomic diffusion and mixing processes in stellar interiors influence the structure and the surface composition of stars. Some of these processes cannot yet be modelled from the first principles, and they require calibrations. This limits their applicability in stellar models used for studies of stellar populations and Galactic evolution. Aims. Our main goal is to put constraints on the stellar structure and evolution models using new refined measurements of the chemical composition in stars of a Galactic open cluster. Methods. We used medium-resolution, 19 200 ≤ R ≤ 21 500, optical spectra of stars in the open cluster NGC 2420 obtained within the Gaia-ESO survey. The sample covers all evolutionary stages from the main sequence to the red giant branch. Stellar parameters were derived using a combined Bayesian analysis of spectra, 2MASS photometry, and astrometric data from Gaia DR2. The abundances of Mg, Ca, Fe, and Li were determined from non-local thermodynamic equilibrium (NLTE) synthetic spectra, which were computed using one-dimensional (1D) and averaged three-dimensional (3D) model atmospheres. We compare our results with a grid of Code d’Evolution Stellaire Adaptatif et Modulaire (CESTAM) stellar evolution models, which include atomic diffusion, turbulent, and rotational mixing. Results. We find prominent evolutionary trends in the abundances of Fe, Ca, Mg, and Li with the mass of the stars in the cluster. Furthermore, Fe, Mg, and Ca show a depletion at the cluster turn-off, but the abundances gradually increase and flatten near the base of the red giant branch. The abundance trend for Li displays a signature of rotational mixing on the main sequence and abrupt depletion on the sub-giant branch, which is caused by advection of Li-poor material to the surface. The analysis of abundances combined with the CESTAM model predictions allows us to place limits on the parameter space of the models and to constrain the zone in the stellar interior, where turbulent mixing takes place.

Aims. One of the primary objectives of the TOPoS survey is to search for the most metal-poor stars. Our search has led to the discovery of one of the most iron-poor objects known, SDSS J092912.32+023817.0. This object is a multiple system, in which two components are clearly detected in the spectrum. Methods. We have analysed 16 high-resolution spectra obtained using the UVES spectrograph at the ESO 8.2 m VLT telescope to measure radial velocities and determine the chemical composition of the system. Results. Cross correlation of the spectra with a synthetic template yields a double-peaked cross-correlation function (CCF) for eight spectra, and in one case there is evidence for the presence of a third peak. Chemical analysis of the spectrum obtained by averaging all the spectra for which the CCF showed a single peak found that the iron abundance is [Fe/H] = −4.97. The system is also carbon enhanced with [C/Fe] = +3.91 (A(C) = 7.44). From the permitted oxygen triplet we determined an upper limit for oxygen of [O/Fe] < +3.52 such that C/O > 1.3. We are also able to provide more stringent upper limits on the Sr and Ba abundances ([Sr/Fe] < +0.70, and [Ba/Fe] < +1.46, respectively).