Chemical Composition of Giant Stars in the Open Cluster IC 4756

A&A 615, A165 (2018) Astronomy https://doi.org/10.1051/0004-6361/201832695 & © ESO 2018 Astrophysics

Chemical composition of giant stars in the open cluster IC 4756?,?? Vilius Bagdonas1, Arnas Drazdauskas1, Gražina Tautvaišiene˙1, Rodolfo Smiljanic2, and Yuriy Chorniy1

1 Institute of Theoretical Physics and Astronomy, Vilnius University, Sauletekio˙ al. 3, 10257 Vilnius, Lithuania e-mail: [email protected] 2 Nicolaus Copernicus Astronomical Center, Polish Academy of Sciences, Bartycka 18, 00-716 Warsaw, Poland

Received 24 January 2018 / Accepted 29 March 2018

ABSTRACT

Context. Homogeneous investigations of red giant stars in open clusters contribute to studies of internal evolutionary mixing processes inside stars, which are reflected in abundances of mixing-sensitive chemical elements like carbon, nitrogen, and sodium, while α- and neutron-capture element abundances are useful in tracing the Galactic chemical evolution. Aims. The main aim of this study is a comprehensive chemical analysis of red giant stars in the open cluster IC 4756, including determinations of 12C/13C and C/N abundance ratios, and comparisons of the results with theoretical models of stellar and Galactic chemical evolution. Methods. We used a classical differential model atmosphere method to analyse high-resolution spectra obtained with the FEROS spectrograph on the 2.2 m MPG/ESO Telescope. The carbon, nitrogen, and oxygen abundances, 12C/13C ratios, and neutron-capture element abundances were determined using synthetic spectra, and the main atmospheric parameters and abundances of other chemical elements were determined from equivalent widths of spectral lines. Results. We have determined abundances of 23 chemical elements for 13 evolved stars and 12C/13C ratios for six stars of IC 4756. The mean metallicity of this cluster, as determined from nine definite member stars, is very close to solar – [Fe/H] = 0.02 0.01. Abun- dances of carbon, nitrogen, and sodium exhibit alterations caused by extra-mixing: the mean 12C/13C ratio is lowered− to±19 1.4, the C/N ratio is lowered to 0.79 0.05, and the mean [Na/Fe] value, corrected for deviations from the local thermodynamical equilibrium± encountered, is enhanced by±0.14 0.05 dex. We compared our results to those by other authors and theoretical models. Conclusions. Comparison of the ±α-element results with the theoretical models shows that they follow the thin disc α-element trends. Being relatively young ( 800 Myr), the open cluster IC 4756 displays a moderate enrichment of s-process-dominated chemical elements compared to the Galactic∼ thin disc model and confirms the enrichment of s-process-dominated elements in young open clusters compared to the older ones. The r-process-dominated element europium abundance agrees with the thin disc abundance. From the comparison of our results for mixing-sensitive chemical elements and the theoretical models, we can see that the mean values of 12C/13C, C/N, and [Na/Fe] ratios lie between the model with only the thermohaline extra-mixing included and the model which also includes the rotation-induced mixing. The rotation was most probably smaller in the investigated IC 4756 stars than 30% of the critical rotation velocity when they were on the main sequence. Key words. stars: abundances – open clusters and associations: individual: IC 4756 – stars: horizontal-branch – stars: evolution

1. Introduction of red giant branch stars in open clusters from high-resolution spectra by analysing the open cluster IC 4756. Our Galaxy has always occupied an important place in the field Stars in open clusters are known to be born from the same of astrophysics. In today’s era of space- and ground-based sur- molecular cloud at roughly the same time and distance. Sharing veys like Gaia (Gaia Collaboration 2016), Gaia-ESO (Gilmore the same primordial material, the open cluster stars have sim- et al. 2012), APOGEE (Nidever et al. 2012), GALAH (De Silva ilar metallicity and chemical composition. New open clusters et al. 2015), and RAVE (Kunder et al. 2017), there is an abun- are constantly identified and the available sample of these stellar dance of data for studying the formation and evolution of the associations expands in variety of ages, heliocentric and galac- Milky Way. However, even massive surveys do not include every tocentric distances, and metallicities. Stars in open clusters have object, or every aspect of selected objects. That is why employ- a significant advantage over the field stars. Precise ages that can ing smaller telescopes and individual observations of selected be determined for open clusters allow studies of how the proper- targets is still important, providing valuable information that is ties of the Galactic disc, such as chemical patterns for example, otherwise lost due to various constraints that are often present change with time, providing important constraints for Galactic in massive surveys. In this paper, we continue our investigation modelling. Furthermore, while members of open clusters have a common origin, they still have a different initial mass, which ? Based on observations collected at the European Organization for Astronomical Research in the Southern Hemisphere under ESO pro- predetermines the lifetime of a star. For this reason, stars in open gramme 085.D-0093(A). clusters are crucial tools in understanding how photospheric ?? The full Table 2 is only available at the CDS via anonymous ftp to chemical composition varies due to stellar evolutionary effects. cdsarc.u-strasbg.fr (130.79.128.5) or via Every chemical element in stellar atmospheres provides valu- http://cdsarc.u-strasbg.fr/viz-bin/qcat?J/A+A/615/A165 able information. Some of them, such as carbon and nitrogen

Article published by EDP Sciences A165, page 1 of 13 A&A 615, A165 (2018) & A A proofs: manuscript no. output

as investigated in this(RGB) work, - called being the susceptible ﬁrst dredge-up to (Iben stellar 1965), evolu- when the con- tionary effects, showvective alterations envelope of expands their observable and connects the abundances inner layers contain- 6 in photospheres of evolveding the nuclear stars. processed The standard material stellar and allows evolution it to rise to the surface. However, as observational evidence suggests (Smiljanic model only predictset one al. 2009; mixing Mikolaitis event et on al. the 2010, red 2011a,b; giant Drazdauskas branch et al. (RGB) – called the ﬁrst2016a,b; dredge-up Tautvaišien (Ibene˙ et 1965 al. 2016), when and references the convec- therein), after tive envelope expandsthe and luminosity connects bump the on inner the RGB, layers there containing is a further decrease 8 12 14 the nuclear processedof materialC and increase and allows of N it in to low-mass rise to the stars, surface. which cannot be explained by standard stellar evolution models. For example, However, as observationalin stars of evidence solar metallicity suggests and about(Smiljanic 2.5 solar et mass, al. the car-

2009; Mikolaitis et al.bon 2010 abundance, 2011a may,b; decreaseDrazdauskas by 0.3 dex, et al. and 2016a the nitrogen,b; abun- V Tautvaišiene˙ et al. 2016dance and may references increase by 0.3 therein), dex (Charbonnel after the & lumi- Lagarde 2010). 1 0 nosity bump on the RGB,The exact there causes is a of further this effect decrease are not yet of fully12C understood and and 14 there are models which predict different extra-mixing effects de- increase of N in low-masspending on stars, stellar which turn-off cannotmass and be metallicity explained (Chanamé by et al. standard stellar evolution2005; Charbonnelmodels. For 2006; example, Cantiello in & Langer stars of 2010; solar Charbonnel metallicity and about& 2.5 Lagarde solar 2010; mass, Denissenkov the carbon 2010; abundance Lagarde et al. may 2011, 2012; 1 2 decrease by 0.3 dex,Wachlin and the et al. nitrogen 2011; Angelou abundance et al. 2012; may Lattanzio increase et al. 2015). In recent years there has been an increase in studies of car- by 0.3 dex (Charbonnel & Lagarde 2010). The exact causes of bon and nitrogen in field and open cluster stars. Together with 0 . 0 0 . 5 1 . 0 1 . 5 2 . 0 this effect are not yettheoretical fully modelsunderstood (Charbonnel and & there Lagarde are 2010; models Lagarde et al. ׍ which predict different2012) extra-mixing we try to constrain effects possible depending mechanisms on stellar governing the B V extra-mixing in evolved stars. In our work, in addition to carbon A colour-magnitude diagram of the open cluster IC 4756. The turn-off mass and metallicity (Chanamé et al. 2005; Charbonnel12 13 Fig.Colour-magnitude 1. diagram of the open cluster IC 4756. The and nitrogen ratios, we look at carbon isotopic ratios ( C/Fig.C), 1.UBV photometry of the open cluster is from Alcaino (1965). The stars 2006; Cantiello & Langerwhich are 2010 less;susceptible Charbonnel to systematic & Lagarde errors 2010 in stellar; atmo-UBV photometryinvestigated in of this the work open are marked cluster with isfrom circles. Alcaino The green(1965 circles). rep- The stars Denissenkov 2010; Lagardespheric parameters et al. 2011 and, 2012 provide; Wachlin an even et better al. 2011 insight; intoinvestigated the resent the in stars this with work uncertain are marked membership with status. circles. The green circles Angelou et al. 2012; extra-mixingLattanzio et phenomenon. al. 2015). represent the stars with uncertain membership status. In recent years thereAbundances has been of an other increase chemical in studies elements of like car- oxygen and α- and iron-peak elements play an important role when trying Jacobson & Friel 2013; Overbeek et al. 2016) provides impor- bon and nitrogen into field constrain and opentheoretical cluster models stars. of Galactic Together chemical with evolution.abundancestant information by a standard about stellar LTE evolution abundance and about analysis the Galaxy (Reddy as & theoretical models (CharbonnelOxygen and α-elements & Lagarde aremainly 2010; produced Lagarde via et Type al. II super-Lamberta whole. 2017). Overall, a comparison of n-capture chemical 2012) we try to constrainnovae explosions possible and mechanisms the timescales governing of elemental injectionthe element into abundancesWe compare our not results only with with theoretical age, but models, with other which important de- the interstellar medium are short. Fe-peak elements, produced pict aspects of both Galactic and stellar chemical evolution. extra-mixing in evolved stars. In our work, in addition to carbon parametersModels such by Magrini as metallicity et al. (2009), or Pagel galactocentric & Tautvaisiene distance (1997), (e.g. by Type Ia supernovae, on the other hand,12 have longer13 interstel- and nitrogen ratios, welar mediumlook at enrichment carbon isotopic timescales ratios (Tinsley ( 1979;C/ C Matteucci), Jacobson & and &Maiorca Frielet 2013 al. (2012); Overbeek trace the et Galactic al. 2016 evolution) provides as a func- important which are less susceptibleGreggio to 1986; systematic Wyse & Gilmore errors 1988; in stellar Yong atmo-et al. 2005).information As tion of the about galactocentric stellar evolution distance, metallicity, and about and age,the respec- Galaxy as a tively, whereas models by Lagarde et al. (2012) provide insight spheric parameters and[Fe/H] provide and [α/Fe] an steadily even better increase insight and decrease into over the time,whole. respectively, the ratios of these elements are good age or place- into the stellar evolution and alterations of light elements in at- extra-mixing phenomenon.of-birth indicators of specific objects of the Galaxy (MatteucciWemospheres compare of evolved our results stars. with theoretical models, which Abundances of other1992). It chemical is also established elements that like the scatter oxygen of abundances and depict of aspects of both Galactic and stellar chemical evolution. [α/Fe] at the same [Fe/H] value can help us distinguish between α- and iron-peak elements play an important role when trying Models2. by Cluster Magrini IC 4756 et al.(2009), Pagel & Tautvaisiene(1997), to constrain theoreticalthinmodels and thick discs of Galactic (Mikolaitis chemical et al. 2014; evolution. Masseron & Gilmoreand Maiorca et al.(2012) trace the Galactic evolution as a 2015). IC 4756 is a relatively young open cluster in our Galaxy. Like Oxygen and α-elementsAbundances are mainly of produced neutron-capture via Type (n-capture) II super- chemicalfunction ele- most of young the open galactocentric clusters, IC 4756 distance, is located metallicity, in the inner part and age, novae explosions andments the alsotimescales deserve further of elemental investigations. injection A recent into study ofrespectively, low- of the Galactic whereas disc models (the galactic by Lagarde coordinates etl al.= (362012◦.381,) provideb the interstellar mediummass are stars short. by D’Orazi Fe-peak et elements, al. (2009) showcased produced an by interestinginsight= into5◦.242), the stellarat a galactocentric evolution distance and alterationsRgc of around of light 8.1 kpc elements Type Ia supernovae,phenomenon: on the other the hand, abundance have of longer the slow interstellar neutron capturein pro- atmospheres(Gilroy 1989). of evolved In order to stars. separate cluster members from field cess (s-process) element barium seems to increase with decreas- stars, the first photometric study for this cluster was made by medium enrichmenting timescales cluster age. The (Tinsley barium abundances1979; Matteucci for young open & clusters Kopff (1943), and later by Alcaino (1965), Seggewiss (1968), Greggio 1986; Wyseexhibited & Gilmore the enhancements 1988; Yong up to et0.6 al. dex.2005 While). As later2. stud- Clusterand Herzog IC 4756 et al. (1975). There are several cluster age deter- [Fe/H] and [α/Fe] steadilyies confirmed increase this finding and for decrease barium,∼ to over the same time, or a lesser minations: an age of 0.82 Gyr was derived by Alcaino (1965), respectively, the ratiosextent, of these the abundance elements enrichment are good patterns age for or other place- s-processIC el- 47560.8 is Gyr a by relatively Gilroy (1989), young 0.79 open Gyr by cluster Salaris et in al. our (2004), Galaxy. 0.8 Like ements in young open clusters are still debatable (e.g. Maiorca 0.2 Gyr by Pace et al. (2010), and most recently 0.89 0.07 Gyr± of-birth indicators ofet specific al. 2011; Yongobjects et al. of 2012; the D’OraziGalaxy et (Matteucci al. 2012; Jacobsonmost & youngwas determined open clusters, by Strassmeier IC 4756 et al. (2015). is located in± the inner part 1992). It is also establishedFriel 2013; thatMishenina the etscatter al. 2013, of 2015; abundances Reddy & Lambert of 2015;of the GalacticThere are disca few studies (the galactic of this cluster coordinates in which metallicityl = 36◦.381, [α/Fe] at the same [Fe/H]D’Orazi value et al. can 2017). help The us suggested distinguish explanations between for higherb =s-5◦was.242 determined), at a galactocentric from photometry. distance Smith (1983),Rgc of using around a DDO 8.1 kpc thin and thick discs (Mikolaitisprocess element et al. yields 2014 include; Masseron higher production & Gilmore rates of these cyanogen colour excess parameter, calculated the cluster metal- elements than previously thought by very old, low-mass( AGBGilroylicity 1989 [Fe)./H] In= 0 order.04 0. to05. separate The most recent cluster comprehensive members study from field 2015). stars (e.g. Maiorca et al. 2011), additional contribution bystars, the theof this first cluster photometric was± made by study Strassmeier for this et al. cluster (2015). was In their made by Abundances of neutron-capturei-process (Mishenina (n-capture) et al. 2015), or chemical overestimation ele- of bariumKopff(work,1943 an), averagedand later metallicity by Alcaino from several(1965), CMD Seggewiss diagrams( was1968), ments also deserveabundances further investigations. by a standard LTE A abundance recent study analysis of (Reddyand & Herzog0.03 et0. al.02( dex1975 and). a There combined are metallicity several cluster from spectroscopy age determina- Lambert 2017). Overall, a comparison of n-capture chemical el- and− photometry± was 0.08 0.06 dex. low-mass stars by D’Oraziement abundances et al.(2009 not) showcased only with age, an but interesting with other importanttions: an ageGilroy of 0.82(1989), Gyr using− was± high-resolution derived by Alcaino spectroscopy(1965 from), 0.8 Gyr phenomenon: the abundanceparameters of such the as slow metallicity neutron or capturegalactocentric process distanceby (e.g. Gilroyseven(1989 giant), stars, 0.79 determined Gyr by aSalaris cluster metallicity et al.(2004 [Fe),/H]=0.08.04 0.2 Gyr (s-process) element barium seems to increase with decreasing by Pace et al.(2010), and most recently 0.89 0.07± Gyr± was cluster age. The bariumArticle number,abundances page 2 of for 14 young open clusters determined by Strassmeier et al.(2015). ± exhibited the enhancements up to 0.6 dex. While later studies There are a few studies of this cluster in which metallic- confirmed this finding for barium, to∼ the same or a lesser extent, ity was determined from photometry. Smith(1983), using a the abundance enrichment patterns for other s-process elements DDO cyanogen colour excess parameter, calculated the cluster in young open clusters are still debatable (e.g. Maiorca et al. metallicity [Fe/H] = 0.04 0.05. The most recent comprehensive 2011; Yong et al. 2012; D’Orazi et al. 2012; Jacobson & Friel study of this cluster was± made by Strassmeier et al.(2015). In 2013; Mishenina et al. 2013, 2015; Reddy & Lambert 2015; their work, an averaged metallicity from several CMD diagrams D’Orazi et al. 2017). The suggested explanations for higher was 0.03 0.02 dex and a combined metallicity from spec- s-process element yields include higher production rates of these troscopy− and± photometry was 0.08 0.06 dex. elements than previously thought by very old, low-mass AGB Gilroy(1989), using high-resolution− ± spectroscopy from stars (e.g. Maiorca et al. 2011), additional contribution by the seven giant stars, determined a cluster metallicity [Fe/H] = 0.04 i-process (Mishenina et al. 2015), or overestimation of barium 0.07. In a later study of this cluster performed by (Luck 1994±),

A165, page 2 of 13 V. Bagdonas: Chemical composition of giant stars in the open cluster IC 4756

Table 1. Analysed stars and their parameters.

ID RA(2000) Dec(2000) VB V Date Obs. Exp. time Rad. vel. S/N − deg deg mag mag seconds km s 1 (at 6400 Å) − ∼ 12 278.9476 +5.3381 9.54 1.03 2010-06-28 600 25.25 140 14 278.9936 +5.4167 8.86 0.86 2010-06-30 600 −24.82 ∼180 28 279.1384 +5.2119 9.01 1.32 2010-06-30 600 −25.26 ∼160 38 279.2717 +5.2921 9.83 1.10 2010-06-27 900 −25.78 ∼140 42 279.3365 +5.8953 9.46 0.97 2010-06-30 600 −24.92 ∼140 44 279.3762 +5.2044 9.77 1.08 2010-06-30 900 −25.00 ∼140 52 279.3992 +5.2605 8.06 1.37 2010-06-30 120 −25.21 ∼120 69 279.5215 +5.4094 9.24 1.05 2010-06-27 600 −27.70 ∼150 81 279.5865 +5.4340 9.46 1.00 2010-06-30 600 −23.25 ∼140 101 279.6824 +5.2389 9.36 1.02 2010-06-30 600 −24.70 ∼140 109 279.7205 +5.3379 9.05 1.05 2010-06-30 300 −25.25 ∼120 125 279.8245 +5.2302 9.36 1.01 2010-06-30 600 −24.00 ∼140 164 280.0771 +5.3144 9.27 1.08 2010-06-30 600 −25.51 ∼150 − ∼ Notes. The IDs are adopted from Kopff(1943), the UBV photometry are taken from Alcaino(1965), the radial velocities as determined in this work. four giant stars gave a metallicity [Fe/H] = 0.03 0.05. Slightly Silla, between June 27 and 30, 2010. The FEROS covers a sub-solar metallicities of [Fe/H] = 0.22− 0.12± (8 stars) and whole visible range of 3500–9200 Å over 39 orders with the [Fe/H] = 0.15 0.04 (6 stars) were− derived± using moderate- resolving power of about 48 000. The FEROS Data Reduction resolution− CCD± spectra by Thogersen et al.(1993) and Jacobson System pipeline within MIDAS was used for spectral reductions. et al.(2007), respectively. A later study by Twarog et al.(1997), Depending on magnitudes of the observed objects, exposure using the DDO photometry from Smith(1983) and data from times where chosen to achieve signal-to-noise ratios (S/Ns) Thogersen et al.(1993), provided the weighted average metal- higher than 120 at about 6400 Å. The most luminous object was licity of 0.06 0.10 dex. The recent study by Smiljanic et al. observed for 120 s, while others have observation times ranging − ± (2009) determined metallicities of 0.04 0.03 dex. It is worth from 300 to 600 s with a maximum of 900 s. A total of 13 stars in ± mentioning that Santos et al.(2009), using a line-list by Sousa the cluster IC 4756 were observed. We list them in Table1. The et al.(2008), determined metallicity of 0.02 0.02 dex, while identiﬁcations of stars were adopted from Kopff(1943). A total ± for the same stars a value of [Fe/H] = 0.08 0.01 was obtained of 13 stars in the cluster IC 4756 were observed (Fig.1). using a line-list by Hekker & Meléndez(2007±). The most recent entirely spectroscopic study of 12 giant stars in this cluster 3.2. Membership was published by Ting et al.(2012), in which a metallicity of 0.01 0.1 dex was determined. The red giants analysed here were selected for observations − However,± these works investigated relatively small samples based on the results of the radial velocity monitoring programme of stars and were mostly dedicated to α- and iron-peak chem- of Mermilliod et al.(2008). All the 13 giants in our sample ical elements. A more comprehensive study of C, N, and O were considered to be likely cluster members. Moreover, all were abundances is required, as these abundances were previously found to be single stars, apart of star 69 which is a single lined determined only by Smiljanic et al.(2009) for three stars, and spectroscopic binary. Mermilliod et al.(2008) found a mean radial velocity of 25.16 0.25 (error) 1.10 (rms) km s 1 for an upper value of oxygen abundance was evaluated also from − ± ± − three stars by Pace et al.(2010). The 12C/13C ratios for this IC 4756. Recently, Casamiquela et al.(2016) reported a mean radial velocity of 24.7 0.7 km s 1 for the cluster based on cluster were also derived only by Gilroy(1989) and Smiljanic − ± − et al.(2009). A deeper analysis for the n-capture elements is also the analysis of 7 red giants. The radial velocities determined in needed. The only determinations of abundances for these ele- our work are within 2 σ (rms) of these two mean values for all ments were performed by Luck(1994; Ba, Y, Zr, Nd, and Eu 12 single giants of the sample. Therefore, as far as radial veloc- abundances presented), by Maiorca et al.(2011; provided abun- ities are concerned, there is no reason to doubt the membership dances of Y, Zr, La, Ce) and by Ting et al.(2012; abundances of of any of the giants that were observed. only Ba determined). Thus, the deﬁciency of studies of neutron Star 69 is a system with a long period of 1994 days, and a capture elements for this cluster and also its young age, make very circular orbit, with eccentricity of 0.05 (Mermilliod et al. IC 4756 the important object to analyse when trying to constrain 2007; Van der Swaelmen et al. 2017). It was included in the s- and r-process element enrichment patterns in young ages. sample analysed here to follow up on the results of Smiljanic et al.(2009), who found star 69 to have very different C and N abundances with respect to other cluster giants (in fact, only 3. Observations and method of analysis limits were determined: [C/Fe] 0.60 and [N/Fe] +0.55). ≤ − ≥ 3.1. Observations As suggested in Smiljanic et al.(2009), and further discussed in Van der Swaelmen et al.(2017), star 69 is likely a post-mass- The spectra of our programme stars were observed using the transfer system. bench-mounted, high-resolution astronomical echelle spectro- Membership probabilities based on proper motions from graph Fiber-Fed Extended Range Optical Spectrograph (FEROS; UCAC4 (Zacharias et al. 2013) have been determined for stars in Kaufer et al. 1999) at the 2.2 m MPG/ESO Telescope in La IC 4756 by Dias et al.(2014) and Sampedro et al.(2017). Stars 69,

A165, page 3 of 13 A&A 615, A165 (2018)

81, 101, 109, and 125 are part of the Dias et al.(2014) catalogue atmospheres (Gustafsson et al. 2008) from the MARCS stellar and all five have membership probability of 99%. Only stars 81 model atmosphere and flux library1. Atomic oscillator strengths and 109, are part of the Sampedro et al.(2017) catalogue. The for the main lines used in this work were taken from the two have very high membership probabilities (>95%) according inverse solar spectrum analysis performed by Gurtovenko & to all three methods used in that work. Kostyk(1989). Using the g f values and solar EWs from Six of our sample giants were included in the work of Gurtovenko & Kostyk(1989), we computed the solar elemental Frinchaboy & Majewski(2008). These authors determined mem- abundances used for the differential determination of elemental bership probabilities using Tycho-2 proper motions, radial veloc- abundances in the programme stars. For the Sun, we used the ities and considering the spatial distribution of the stars. Four main atmospheric parameters Teff = 5777 K, log g = 4.44, and 1 stars were considered to be members of the cluster (stars 69, the microturbulent velocity value 0.8 km s− , as in our earlier 81, 101, and 109) while stars 44 and 52 (TYC 455-00950-1 studies (e.g. Tautvaišiene˙ et al. 2000) where the same method of and TYC 455-00136-1) were found to be non-members (with analysis was applied. membership probabilities of 13% and 47%, respectively). Abundances of Na, Mg, Al, Si, Ca, Sc, Ti, Cr, and Ni were Two of our stars (numbers 14 and 109) were included in determined using the EWs method. The EWs of spectral lines a sample analysed by Baumgardt et al.(2000). These authors are presented in an online table (see Table2 for an example). determined membership probabilities for stars in open clusters The number of lines for each element slightly varied among using HIPPARCOS parallaxes and proper motions combined with the stars, as every line was inspected individually and some of ground-based data (photometry, radial velocity, proper motion, them were excluded due to contamination of cosmic rays or other distance from the cluster centre). The membership probabilities observational effects. For sodium, we applied non-LTE (NLTE) determined by Baumgardt et al.(2000) for stars 14 and 109 were corrections as described by Lind et al.(2011). The corrections 59% and 80%, respectively. range from 0.08 to 0.11 dex. Another work determining membership probabilities for The spectral− synthesis− method was used to derive C, N, stars in IC 4756 was conducted by Herzog et al.(1975) using and O, as well as neutron-capture element abundances. We proper motions. Ten of our sample giants were included in that used the forbidden line at 6300.3 Å for the oxygen abundance work (all except stars 12, 14, and 42). Only stars 28 and 52 were determination. The gf values for 58Ni and 60Ni isotopic line found to be non-members with membership probabilities of 0% components, which blend the oxygen line, were taken from and 29%, respectively. For star 44, Herzog et al.(1975) found a Johansson et al.(2003). Two C 2 molecular bands at 5135 and membership probability of 96%, which disagrees with the 13% 5635 Å were used to determine carbon abundances and up to probability found by Frinchaboy & Majewski(2008) using the eight 12CN molecular lines in the region of 7980–8005 Å for the Tycho-2 data; the latter probably being more robust. For star nitrogen abundances. We used the same molecular data of C2 as 14, the only available study (Baumgardt et al. 2000) indicates in Gonzalez et al.(1998) and the CN molecular data provided a somewhat low membership probability. Stars 12 and 42 were by Bertrand Plez. The Vienna Atomic Line Data Base (VALD; not included in any of the membership works that used proper Piskunov et al. 1995) was used for preparation of input data used motions mentioned here. in the calculations. In order to check correctness of the input Summarizing, there seems to be some evidence that stars 14, data, synthetic spectra of the Sun were compared to the solar 28, 44, and 52 might be non-members of the cluster. In this atlas of Kurucz(2005) with solar abundances of (Grevesse & work, we check how their chemical composition agrees with Sauval 2000) and necessary adjustments were made to the line composition of the high-probability members of the cluster. atomic data. Since carbon and oxygen are bound together by the molec- 3.3. Atmospheric parameters and elemental abundances ular equilibrium, in order to correctly measure abundances of these elements, we investigated them in unison. How abundance We used a standard spectroscopic method, differential to the changes in one of the elements affect abundances in the others Sun, for the determination of atmospheric parameters. Effec- is shown in Table3. As we can see, the most sensitive case is tive temperatures (Teff) were derived by minimizing a slope between nitrogen and carbon, where 0.1 dex change in carbon between abundances of Fe I lines with different lower-level exci- coincides with a change in the nitrogen abundance by the same tation potentials (χ). The step of our Teff determination was 5 K, amount. We adopt a procedure of a few back and forth iterations which corresponds to the slope change d [Fe/H]/d EP = 0.001. between these elements to achieve a combination of these three Surface gravities (log g) were determined using the iron ioniza- abundances, until all of them match the features in the observed tion equilibrium. Since our step in log g determination was 0.1, spectra. we allowed abundances of Fe I and Fe II to differ by no more The synthetic spectra method was also used for the determi- than 0.02 dex. The final iron abundance was based on the neutral nation of (Mn, Co, Y, Zr, Ba, La, Ce, Pr, Nd, and Eu abundances. iron lines. In order to find microturbulence velocities (vt), a min- Cobalt abundances were determined from the 5280.62, 5301.03, imization of scatter in abundances from Fe I lines was employed, 5352.05, 5530.78, 5647.23, 6188.98, 6455, and 6814.95 Å lines. as well as minimization of the iron abundance trend with regard For the analysis of lines at 5301.03 and 5530.78 Å we applied to the Fe I line equivalent widths (EWs). Equivalent widths of hyperfine structure (HFS) data from Nitz et al.(1999), while about 35–38 Fe I and 4–5 Fe II lines were used for the derivation for the remaining Co I lines the HFS data were taken from of stellar metallicities, as well as other atmospheric parameters. Cardon et al.(1982). Manganese was investigated using lines For the measurement of EWs, we used a SPLAT-VO programme at 6013.49, 6016.64, and 6021.80 Å with the HFS data taken package (Škoda et al. 2014). from Den Hartog et al.(2011). Yttrium abundances were deter- The EQWIDTH and BSYN software packages (developed mined from the Y II lines at 4883.69, 4900.12, 4982.14, 5200.41, at the Uppsala Observatory) were used to derive elemental and 5402.78 Å; zirconium from the Zr I lines at 5385.10 and abundances from EWs and synthetic spectra, respectively. We 6127.50 Å; lanthanum from the La II lines at 5123.01, 6320.41, have taken a set of plane-parallel, one-dimensional, hydrostatic, constant flux local thermodynamical equilibrium (LTE) model 1 http://marcs.astro.uu.se/

A165, page 4 of 13 V. Bagdonas: Chemical composition of giant stars in the open cluster IC 4756

Table 2. Spectral line equivalent widths in stars of IC 4756.

Elem. Wavelength EW (mÅ) (Å) 12 14 28 38 42 44 52 69 81 101 109 125 164 Na I 5148.84 33 51 42 – 32 31 52 40 30 35 39 28 33 6154.22 67 91 93 67 71 68 107 78 65 68 80 69 74 6160.75 89 110 109 85 87 85 123 93 81 84 94 89 94 Mg I 6318.70 55 67 67 56 55 53 70 55 46 54 57 55 61 ......

Notes. The full table is available at the CDS.

Table 3. Effects on derived abundances and isotopic ratios for the target Table 4. Effects on derived abundances, ∆[A/H], resulting from model star IC 4756 12, resulting from abundance changes of C, N, or O. changes for the star IC 4756 12.

∆ ∆ ∆ ∆Teff ∆ log g ∆vt ∆[Fe/H] Species C N O Species 1 Total 100 K 0.3 0.3 km s− 0.1 0.1 dex 0.1 dex 0.1 dex ± ± ± ± ± ± ± C (C2) 0.03 0.02 0.00 0.01 0.04 ∆C– 0.01 0.02 N (CN) 0.10 0.03 0.01 0.01 0.11 ∆N 0.10± – ±0.05 ∓ ± O ([O I]) 0.01 0.14 0.00 0.01 0.14 ∆O 0.01 0.01 – 12C/13C 1 1 0 0 1.4 ∆C/N ±0.19 ±0.16 0.01 ± ∓ ∓ Na I 0.06 0.01 0.00 0.00 0.06 ∆12C/13C 2 2 0 − ± ± Mg I 0.05 0.00 0.01 0.01 0.05 − Al I 0.06 0.00 0.01 0.01 0.06 − Si I 0.01 0.04 0.00 0.01 0.04 and 6390.48 Å. For the analysis of the La II 5123.01 Å and Ca I 0.08 0.01 0.01 0.01 0.08 − − 6390.48 Å lines, we applied the HFS data from (Ivans et al. Sc II 0.01 0.13 0.00 0.03 0.13 − 2006). We were not able to find the HFS data for the La II Ti I 0.11 0.01 0.00 0.01 0.11 − − 6320.41 Å line, however it seems that the HFS influence is Ti II 0.01 0.13 0.01 0.03 0.13 − − small for this line since lanthanum abundances were very similar Fe I 0.09 0.01 0.00 0.01 0.09 from all three lines. Cerium abundances were determined from Fe II 0.07 0.14 0.01 0.04 0.15 − − the Ce II lines at 5274.22, 6043.00 Å. Neodymium abundances Cr I 0.10 0.01 0.01 0.01 0.10 − − − were derived from the Nd II lines at 5092.80, 5293.20, 5319.80, Mn I 0.12 0.04 0.11 0.01 0.17 − 5356.97, and 5740.86 Å with the HFS adopted from Den Hartog Co I 0.10 0.03 0.02 0.01 0.11 et al.(2003). Barium and europium abundances were determined Ni I 0.06 0.03 0.00 0.01 0.07 from single lines at 5853.67 and 6645.10 Å, respectively. For the Y II 0.01 0.11 0.25 0.04 0.27 − Ba II line, the HFS data were taken from McWilliam(1998) and Zr I 0.15 0.00 0.02 0.00 0.15 for the Eu II line from Lawler et al.(2001). The HFS was also Ba II 0.05 0.11 0.36 0.01 0.38 − taken into account for the determination of praseodymium abun- La II 0.02 0.13 0.01 0.03 0.13 − dance from Pr II lines at 5259.7 and 5322.8 Å (Sneden et al. Ce II 0.01 0.13 0.05 0.04 0.14 − 2009). All log g f values were calibrated to fit the solar spectrum Pr II 0.01 0.13 0.01 0.03 0.13 − by Kurucz(2005) with solar abundances provided by (Grevesse Nd II 0.03 0.14 0.05 0.03 0.15 − & Sauval 2000). Several examples of the synthetic spectra fits Eu II 0.01 0.11 0.00 0.03 0.11 − for some of the lines are presented in Fig.2. As stellar rotation (v sin i) changes the shape of lines, it is also an important factor in abundance determinations from spectral We evaluated uncertainties of atmospheric parameters using syntheses. We obtained v sin i by fitting iron lines of different the whole sample of IC 4756 member stars. The average of strengths in our investigated spectral regions. slopes between [Fe/H] abundance and the excitation potential There are two categories of uncertainties that should be con- (EP) for these stars d [Fe/H]/d EP = 0.0008 0.008. The error sidered. Firstly, there are random errors that affect each line of 0.008 corresponds to the temperature change± of 30 K. We independently and originate from the local continuum place- can assume this as an error in our temperature estimation. The ment, from each line’s fitting variations, EW measurements, or change of log g by 0.1 dex alters the Fe I and Fe II equilib- from uncertainties in atomic parameters. Uncertainties coming rium by 0.04 dex, which is larger than our acceptable tolerance from the atomic data are minimized when a differential analysis of 0.02 dex. Therefore, the error in log g determination is less is performed. Since all cluster-member stars have relatively sim- than 0.1 dex. For the microturbulence velocity we minimized ilar parameters in our case, we can compare measurements of a slope between the Fe I abundances and the EWs. The mean EWs for the same line in different stars. The mean scatter of EWs d [Fe/H]/d EW = 0.0001 0.0003. This error corresponds to the − 1 ± for iron lines was 4.3 mÅ. An approximate value for the random change of 0.05 km s− , which can be assumed as our error for the errors in the abundance determinations is given by the scatter of microtubulence velocity determination. the derived abundances from individual lines for all elements in The second type of uncertainties are systematic; they are all stars. The mean scatter in our sample of stars is 0.06 dex. influenced by uncertainties of atmospheric parameters and affect

A165, page 5 of 13 A&A proofs: manuscript no. output

Table 1. A list of analysed stars and their parameters.

ID R.A.(2000) DEC(2000) VB V Date Obs. Exp. time Rad. vel. S/N − deg deg mag mag seconds km s 1 (at 6400 Å) − ∼ 12 278.9476 +5.3381 9.54 1.03 2010-06-28 600 25.25 140 14 278.9936 +5.4167 8.86 0.86 2010-06-30 600 −24.82 ∼180 28 279.1384 +5.2119 9.01 1.32 2010-06-30 600 −25.26 ∼160 38 279.2717 +5.2921 9.83 1.10 2010-06-27 900 −25.78 ∼140 42 279.3365 +5.8953 9.46 0.97 2010-06-30 600 −24.92 ∼140 44 279.3762 +5.2044 9.77 1.08 2010-06-30 900 −25.00 ∼140 52 279.3992 +5.2605 8.06 1.37 2010-06-30 120 −25.21 ∼120 69 279.5215 +5.4094 9.24 1.05 2010-06-27 600 −27.70 ∼150 81 279.5865 +5.4340 9.46 1.00 2010-06-30 600 −23.25 ∼140 101 279.6824 +5.2389 9.36 1.02 2010-06-30 600 −24.70 ∼140 109 279.7205 +5.3379 9.05 1.05 2010-06-30 300 −25.25 ∼120 125 279.8245 +5.2302 9.36 1.01 2010-06-30 600 −24.00 ∼140 164 280.0771 +5.3144 9.27 1.08 2010-06-30 600 −25.51 ∼150 − ∼ The IDs are adopted from Kopﬀ (1943), the UBV photometry are taken from Alcaino (1965), the radial velocities as determined in this Notes. A&A 615, A165 (2018) work.

1 . 0 1 . 0 1 . 0

0 . 9 0 . 9

C 13 2 0 . 8 [O l] Sc ll CN y y y t t t i i i s s 0 . 8 s n n n e e 0 . 9 0 . 7 e t t t n n n i i i

e e e v v v i i i t t t a a a 0 . 6 l l l 0 . 7 e e e 12 12 12

R CN CN CN R R Fe l 0 . 5

0 . 6 0 . 8

12 Fe l IC 4756 12 0 . 4 IC 4756 12 Fe l CN IC 4756 12

5 6 3 4 5 6 3 5 5 6 3 6 6 3 0 0 . 0 6 3 0 0 . 5 6 3 0 1 . 0 6 3 0 1 . 5 8 0 0 3 8 0 0 4 8 0 0 5 Wavele ngth, Å Wavele ngth, Å Wavele ngth, Å 1 . 0 1 . 0 1 . 0

0 . 9

0 . 8 0 . 9 Z r I P r I I 0 . 8 y y y t t i t i C e I I i s s s n n n e e e t 0 . 7 t t 0 . 8 n n i n i i

0 . 6

e e e v v i v i i t t

t C r I I a a l a 0 . 6 F e I l l e e e R R R

0 . 7 0 . 5 0 . 4 N i I F e I F e I 0 . 4 F e I I C 4 7 5 6 1 2 I C 4 7 5 6 6 9 F e I I C 4 7 5 6 3 8 0 . 6

5 3 2 1 5 3 2 2 5 3 2 3 5 2 7 3 5 2 7 4 5 2 7 5 6 1 2 7 6 1 2 8 6 1 2 9 W a v e l e n g t h , Å W a v e l e n g t h , Å W a v e l e n g t h , Å

Fig. 2. Examples of of the the synthetic synthetic spectrum spectrum fits fits to to various various lines lines for for the the stars stars IC IC 47561 47561 12, 12, IC 47563IC 47563 8, and 8 and IC 4756IC 4756 69. 69. The The blue blue and and green green lines lines rep- Fig. 2. 12 13 representresent a change a change in abundance in abundance by by0.10dex.1 dex to the to the corresponding corresponding elements, elements, except except in the in thecase case of 12 ofC/13CC/ whereC where the thegreen green and and red redlines lines represent represent5. 5. ± ± ± ± all the lines simultaneously. Table4 shows sensitivity of abun- 4.1. Atmospheric parameters peraturesdance estimates (Teff) were to changes derived in by atmospheric minimizing parameters a slope between for the iron lines. In order to find microturbulence velocities (vt), a min- abundancesstar IC 4756 of12. Fe Changesi lines with in atmospheric different lower-level parameters excitation provide rel-po- imizationThe average of scatter metallicity in abundances for this cluster from Fe determinedi lines was in employed, this work from nine high-probability members is close to solar – [Fe/H] = tentialsatively small (χ). The abundance step of our deviations Teff determination from the initial was values.5 K, which The as well as minimization of the iron abundance trend with regard 0.02 0.01. Those stars have similar atmospheric parameters correspondslarger deviations to the are slope present change for d abundances [Fe/H]/d EP derived= 0.001. from Surface ion- to− the Fe± i line equivalent widths (EWs). Equivalent widths of gravitiesized lines, (log asg these) were elements determined respond using to the the iron log g ionizationvalue changes equi- aboutand are 35–38 in the Fe redi and clump: 4–5 the Fe ii averagelines wereTeff = used5124 for the58 derivationK, log g = 2.72 0.1, [Fe/H]= 0.02 0.01, v = 1.38 0.11± km s 1. librium.more strongly. Since our step in log g determination was 0.1, we al- of stellar± metallicities− as well± as othert atmospheric± parameters.− lowedAlong abundances with IC 4756of Fe stars,i and weFe ii performedto differ by our no atmospheric more than Our mean metallicity value agrees well with the majority 0.02parameter dex. The and final abundance iron abundance determination was based procedures on the onneutral the of the previous spectroscopic studies (Gilroy 1989, Luck 1994, Arcturus’s spectrum by (Hinkle et al. 2000; the results for Smiljanic et al. 2009, Santos et al. 2009, Ting et al. 2012), all of ArticleArcturus number, we present page 4 of together 14 with the IC 4756 results). Our which provide metallicities close to solar for this cluster. Several results for Arcturus agree well with the majority of recent studies studies by Thogersen et al.(1993) and Jacobson et al.(2007) of this star (Worley et al. 2009; Ramírez & Allende Prieto 2011; derived a slightly sub-solar metallicity of 0.22 0.12 dex and 0.15 0.04 dex, respectively. These discrepant− ± metallici- Abia et al. 2012; Jofré et al. 2015). − ± Several examples of the synthetic spectra fits in several ties were probably caused by different line lists and/or analysis IC 4756 stars for some of the lines are presented in Fig.2. techniques. The stars 14, 28, 44, and 52, as already discussed in Sect. 3.2, are considered as doubtful members according to the literature, 4. Results and discussion however, we chose to investigate them and compare their chemical composition to other stars of the cluster. We found that stars The determined stellar atmospheric parameters and rotational 14, 28, and 52 have slightly lower metallicities thus we sup- velocities for our IC 4756 programme stars are presented in port their doubtful membership. Star 44 has the same metallicity Table5, and the detailed chemical abundances are in Tables6 as the cluster average, thus we leave its membership question and7. Mean values of element-to-iron ratios were calculated open. taking just definite members of IC 4756. These values with corresponding scatters are listed in Table8 along with results from 4.2. 12C/13C and C/N ratios previous studies. Luck(1994) investigated one supergiant, one dwarf, and two giants. Since the investigated dwarf star was During a star’s lifetime, complex nuclear reactions take place rather peculiar, we did not take its abundances while calculating inside the core, producing different elements which, for the most the mean values presented in Table8. part, stay deep inside stellar interiors. However, when the star

A165, page 6 of 13 V. Bagdonas: Chemical composition of giant stars in the open cluster IC 4756

Table 5. Atmospheric parameters of the programme stars, the Sun, and Arcturus.

1 1 Star ID Teﬀ (K) log g vt (km s− ) v sin i (km s− ) [Fe/H] σ Fe I n σ Fe II n Members 12 5135 2.6 1.45 1.6 0.03 0.06Vilius 38 Bagdonas 0.02 et al.: Chemical 4 composition of giant stars in the open cluster IC 4756 − 38 5165 2.9 1.35 1.9Table 5. 0.00Atmospheric 0.07 parameters 38of the programme 0.05 stars, 5 the Sun, and Arcturus. 42 5165 2.7 1.25 3.0 0.01 0.04 36 0.06 5 − 1 1 69 5150 2.7 1.40 2.5 0.00Star 0.07 ID Te 38ﬀ (K) 0.04 log g vt 5(km s− ) v sin i (km s− ) [Fe/H] σ Fe i n σ Fe ii n 81 5180 2.8 1.15 3.7 0.01 0.06 37 0.06 5 Members − . 12 5135 2.6 1.45 1.6 0.03 0.06 38 0.02 4 101 5135 2.8 1.40 3.2 0 01 0.07 38 0.02 5 − 109 5000 2.6 1.50 2.9 −0.03 380.06 516538 0.04 2.9 5 1.35 1.9 0.00 0.07 38 0.05 5 42 5165 2.7 1.25 3.0 0.01 0.04 36 0.06 5 125 5150 2.8 1.45 2.6 −0.02 0.05 38 0.02 5 − − 69 5150 2.7 1.40 2.5 0.00 0.07 38 0.04 5 164 5040 2.6 1.45 2.2 0.03 810.06 518038 0.04 2.8 5 1.15 3.7 0.01 0.06 37 0.06 5 − 101 5135 2.8 1.40 3.2 −0.01 0.07 38 0.02 5 Doubtful members 109 5000 2.6 1.50 2.9 −0.03 0.06 38 0.04 5 14 4760 2.3 1.50 1.9 0.06 1250.07 515037 0.04 2.8 4 1.45 2.6 −0.02 0.05 38 0.02 5 − . 164 5040 2.6 1.45 2.2 −0.03 0.06 38 0.04 5 28 4650 2.1 1.45 2.2 0 10 0.08 38 0.06 5 − 44 5115 2.7 1.25 3.6 −0.02 0.06 37 0.07 5 Doubtful members 52 4500 1.9 1.60 2.8 −0.12 140.08 476035 0.05 2.3 5 1.50 1.9 0.06 0.07 37 0.04 4 28 4650 2.1 1.45 2.2 −0.10 0.08 38 0.06 5 − − Sun 5777 4.44 0.8 2 0.05*44 0.03 511538 0.01 2.7 5 1.25 3.6 0.02 0.06 37 0.07 5 52 4500 1.9 1.60 2.8 −0.12 0.08 35 0.05 5 Arcturus 4345 1.6 1.65 2.4 0.58 0.04 38 0.03 5 − − Sun 5777 4.44 0.8 2 0.05* 0.03 38 0.01 5 Arcturus 4345 1.6 1.65 2.4 0.58 0.04 38 0.03 5 Notes. (*) The solar iron abundance in this work is A(Fe) = 7.50 (Grevesse & Sauval 2000). −

Notes. * The solar iron abundance in this work is A(Fe) =7.50 (Grevesse & Sauval 2000).

leaves the main sequence and becomes a red giant, the convective 1 D U P S T ( C L 2 0 1 0 ) 2 . 5 1 D U P S T ( C L 2 0 1 0 ) envelope of the star deepens, and the processed material from T H ( C L 2 0 1 0 ) T H ( C L 2 0 1 0 ) the core is brought up to the surface. By looking at abun- 1 D U P S T ( L 2 0 1 2 ) 1 D U P S T ( L 2 0 1 2 ) dance alterations of certain elements in stellar photospheres, we 3 0 T H + V ( L 2 0 1 2 ) can analyse the efﬁciency of the transport mechanisms. Among 2 . 0 T H + V ( L 2 0 1 2 ) the most sensitive elements are carbon and nitrogen. 12C/13C and 12C/14N ratios are particularly good indicators of mixing C 3 1 N / in stars. / 1 . 5

C 2 0 C 2

Carbon and nitrogen abundances in evolved stars have 1 been investigated for more than 45 yr, since studies by Day et al.(1973); Tomkin & Lambert(1974); Tomkin et al.(1975); Dearborn et al.(1976) and others. It is well known that stars 1 . 0 undergo several mixing events during their evolution, however 1 0 only one, the first dredge-up, was predicted by the classical theory of stellar evolution until a star leaves the red clump position 0 . 5 on the HR diagram. As concerns studies of evolved stars in open 1 2 3 4 5 6 1 2 3 4 5 6 12 13 M / M clusters, inconsistencies between the C/ C ratios predicted by ؽ M / M ؽ the classical stellar evolution model and the observational results Fig. 3. The average carbon isotope ratios in clump stars of open clus- Fig. 4. The average carbon-to-nitrogen ratios in clump stars of open were clearly demonstrated by Gilroy(1989). The observational Fig. 3. tersAverage as a function carbon of stellar isotope TO mass. ratios Red in square clump indicates stars the of value open of clustersclusters as a function of stellar TO mass. In addition to symbols in Fig. results agreed relatively well with the classical model for stars as a functionIC 4756. of Red stellar open squares TO mass. indicate Red results square of previously indicates investigated the value3, here of we include the results from Tautvaišiene˙ et al. (2015) as red open IC 4756.open Red clusters open by squares Tautvaišien indicatee˙ et al. (2000, results 2005, of 2016); previously Mikolaitis investigatedet al. triangles. in open clusters with turn-off (TO) masses larger than 2.2 M . (2010, 2011a,b, 2012); Drazdauskas et al. (2016a,b). Other symbols ˙ However, for stars in open clusters with lower turn-off∼ masses open clustersinclude by results Tautvaišien from Gilroye et (1989) al.(2000 – pluses,, 2005 Luck, 2016 (1994)); –Mikolaitis open cir- et al. 12 13 (2010, 2011acles, Smiljanic,b, 2012 et); al. Drazdauskas (2009) – green crosses, et al.( Santrich2016a,b et). al. Other (2013) symbols – / 12 13 the observed C C ratios were decreasing with decreasing includeopen results diamond. from TheGilroy solid(1989 lines) (1DUP – pluses, ST) representLuck(1994 the )C –/ openC ra- circles,leaves the main sequence and becomes a red giant, the convec- TO mass. tios predicted for stars at the first dredge-up with standard stellar evo- tive envelope of the star deepens, and the processed material Smiljanic et al.(2009) – green crosses, Santrich et al.(2013) – openfrom dia- the core is brought up to the surface. By looking at abun- In order to explain discrepancies between theoretical lutionary models of solar metallicity by Charbonnel12 & Lagarde13 (2010) mond. The(black solid solid lines line) and (1DUP Lagarde ST) et al. represent (2012) (blue the solidC line)./ C Theratios short- predicteddance alterations of certain elements in stellar photospheres, we and observed abundances of mixing-sensitive chemical ele- for starsdashed at the line first (TH) dredge-up shows the with prediction standard when only stellar thermohaline evolutionary extra- modelscan analyse the efficiency of the transport mechanisms. Among ments, some other transport mechanisms besides the classical of solarmixing metallicity is introduced by Charbonnel (Charbonnel & & Lagarde 2010),(2010 and; black the long- solidthe line) most sensitive elements are carbon and nitrogen. 12C/13C + 12 14 convection had to be introduced. Nowadays, the most promising and Lagardedashed etline al. (TH(2012V) is;for blue the solidmodel that line). includes The both short-dashed the thermohaline line (TH)and C/ N ratios are particularly good indicators of mixing in and rotation-induced mixing (Lagarde et al. 2012). A typical error bar stars. of these seems to be a thermohaline-induced extra-mixing. We shows theis indicated prediction (Charbonnel when &only Lagarde thermohaline 2010; Smiljanic extra-mixinget al. 2009; Gilroy is introduced (Charbonnel1989). & Lagarde 2010), and the long-dashed line (TH+V)Carbon and nitrogen abundances in evolved stars have been compare our results with the first dredge-up model, a model with investigated for more than 45 years, since studies by Day et al. only the thermohaline mixing included (Charbonnel & Lagarde is for the model that includes both the thermohaline and rotation-(1973); Tomkin & Lambert (1974); Tomkin et al. (1975); Dear- 2010) and with a model where rotation and thermohaline- induced mixing (Lagarde et al. 2012). A typical error bar is indicated (Gilroy 1989; Smiljanic et al. 2009; Charbonnel & Lagarde 2010). Article number, page 7 of 14 induced mixing act together (Lagarde et al. 2012). Stellar rotation may modify an internal stellar structure even before the RGB phase, however results become visible only in later stages. The (Lagarde et al. 2014). We present the comparison of our results initial rotation velocity of the models on the zero age main and the models in Figs.3 and4. The TO mass of IC 4756 was sequence (ZAMS) was chosen at 30% of the critical velocity at determined using PARSEC isochrones (Bressan et al. 2012). The that point and leads to the mean velocity of about 120 km s 1 input metallicity of 0.02 dex was the one we determined in this − − A165, page 7 of 13 A&A 615, A165 (2018)

Table 6. Elemental abundances in member stars of IC 4756 and Arcturus.

El./Star 12 38 42 69 81 [C/Fe] 0.23 0.01 (02) 0.23 0.03 (02) 0.29 0.02 (02) – 0.29 0.02 (02) [N/Fe] −0.45 ±0.06 (08) −0.43 ±0.03 (08) −0.45 ±0.04 (08) – −0.42 ±0.05 (07) [O/Fe] 0.01± 0.09 ± 0.10± – 0.05± C/N 0.83 0.87 0.72− –− 0.78 12C/13C 19 20 – – 17 [Na/Fe] 0.23 0.04 (03) 0.17 0.01 (02) 0.25 0.02 (03) 0.33 0.04 (03) 0.19 0.02 (03) LTE ± ± ± ± ± [Na/Fe]NLTE 0.15 0.07 (03) 0.06 0.02 (02) 0.17 0.04 (03) 0.24 0.08 (03) 0.10 0.06 (03) [Mg/Fe] 0.08 ± 0.04 (02) 0.11 ± 0.01 (02) 0.15 ± 0.05 (02) 0.05 ± 0.04 (02) 0.02±(01) [Al/Fe] 0.10 ± 0.01 (02) 0.00 ± 0.02 (02) 0.01 ± 0.08 (02) 0.00 ± 0.07 (02) −0.03 0.08 (02) [Si/Fe] 0.00 ± 0.07 (16) 0.05 ± 0.08 (19) 0.05 ± 0.06 (19) 0.01 ± 0.06 (17) 0.01± 0.05 (18) [Ca/Fe] 0.06 ± 0.09 (06) 0.13 ± 0.08 (08) 0.11 ± 0.10 (07) 0.10 ± 0.08 (07) −0.10 ±0.06 (08) ± ± ± ± ± [Sc II/Fe] 0.08 0.05 (07) 0.03 0.02 (07) 0.06 0.05 (07) 0.02 0.05 (07) 0.00 0.06 (06) − ± ± − ± − ± ± [Ti I/Fe] 0.02 0.04 (13) 0.00 0.04 (16) 0.02 0.04 (14) 0.01 0.04 (15) 0.01 0.05 (14) − ± ± − ± − ± ± [Ti II/Fe] 0.11 0.07 (05) 0.01 0.02 (05) 0.10 0.05 (05) 0.04 0.07 (05) 0.08 0.04 (05) [Cr/Fe] −0.08 ±0.05 (08) −0.02 ±0.07 (09) −0.03 ±0.06 (09) −0.05 ±0.08 (10) −0.01 ±0.07 (10) [Mn/Fe] 0.08± 0.02 (03) 0.02± 0.01 (03) 0.09± 0.05 (03) 0.07± 0.06 (03) 0.19± 0.02 (03) [Co/Fe] −0.02 ± 0.03 (08) −0.03 ± 0.03 (08) −0.07 ± 0.02 (08) −0.03 ± 0.03 (08) −0.07 ± 0.02 (08) [Ni/Fe] −0.08 ± 0.09 (24) −0.06 ± 0.07 (26) −0.06 ± 0.10 (25) −0.06 ± 0.08 (26) −0.07 ± 0.07 (21) [Y/Fe] −0.05 ± 0.14 (05) −0.08 ±0.09 (05) −0.02 ±0.05 (05) −0.06 ±0.08 (05) −0.01 ± 0.08 (05) [Zr/Fe] −0.20 (01)± 0.24 ± 0.14 (02) 0.17 ± 0.02 (02) 0.14 ± 0.04 (02) −0.18 (01)± [Ba/Fe] 0.20 (01) 0.23 (1)± 0.24 (1)± 0.30 (01)± 0.22 (01) [La/Fe] 0.20 0.02 (03) 0.23 0.05 (03) 0.14 0.02 (03) 0.19 0.03 (02) 0.23 0.04 (03) [Ce/Fe] 0.21 ± 0.10 (02) 0.27 ± 0.11 (02) 0.18 ± 0.07 (02) 0.17 ± 0.06 (02) 0.18 ± 0.04 (02) [Pr/Fe] 0.21 ± 0.03 (02) 0.18 ± 0.07 (02) 0.10 ± 0.01 (02) 0.14 ± 0.06 (02) 0.16 ± 0.02 (02) [Nd/Fe] 0.08 ± 0.06 (05) 0.14 ± 0.06 (05) 0.03 ± 0.06 (05) 0.13 ± 0.04 (05) 0.11 ± 0.07 (05) [Eu/Fe] 0.02 (01)± 0.11 (01)± 0.03 (01)± 0.04 (01)± 0.03 (01)± El./Star 101 109 125 164 Arcturus [C/Fe] 0.23 0.03 (02) 0.24 0.05 (02) 0.26 (01) 0.28 0.02 (02) 0.03 0.01 (02) [N/Fe] 0−.46 ±0.04 (07) 0−.47 ±0.03 (08) −0.46 0.05 (08) −0.45 ±0.03 (08) 0.29 ± 0.02 (08) [O/Fe] 0.05± 0.02± 0.02± 0.02± 0.55 ± C/N 0.81 0.78 0.76 0.74 2.19 12C/13C – – 17 20 7 [Na/Fe] 0.19 0.06 (03) 0.24 0.04 (03) 0.19 0.04 (03) 0.22 0.02 (03) 0.20 0.05 (03) LTE ± ± ± ± ± [Na/Fe]NLTE 0.11 0.11 (03) 0.16 0.09 (03) 0.11 0.01 (03) 0.14 0.03 (03) 0.16 0.06 (03) [Mg/Fe] 0.03 ± 0.05 (02) 0.04 ± 0.02 (02) 0.11 ± 0.01 (02) 0.14 ± 0.02 (02) 0.43 ± 0.01 (02) [Al/Fe] 0.01 ± 0.05 (02) 0.05 ± 0.04 (02) 0.07 ± 0.04 (02) 0.06 ± 0.01 (02) 0.43 ± 0.07 (03) [Si/Fe] 0.06 ± 0.07 (18) 0.08 ± 0.08 (17) 0.06 ± 0.06 (18) 0.07 ± 0.07 (17) 0.25 ± 0.04 (18) [Ca/Fe] 0.09 ± 0.06 (07) 0.16 ± 0.09 (07) 0.12 ± 0.08 (07) 0.12 ± 0.06 (08) 0.18 ± 0.06 (08) ± ± ± ± ± [Sc II/Fe] 0.01 0.06 (06) 0.01 0.07 (07) 0.03 0.07 (07) 0.01 0.06 (07) 0.15 0.04 (07) ± ± ± ± ± [Ti I/Fe] 0.04 0.04 (13) 0.01 0.05 (12) 0.02 0.05 (15) 0.01 0.04 (13) 0.27 0.05 (16) − ± ± ± ± ± [Ti II/Fe] 0.02 0.06 (05) 0 0.05 (05) 0.02 0.02 (05) 0.04 0.04 (05) 0.23 0.05 (04) [Cr/Fe] 0.01± 0.04 (09) 0.±00 0.07 (09) −0.01 ±0.06 (10) −0.04 ±0.06 (09) 0.03± 0.07 (10) [Mn/Fe] −0.06 ± 0.05 (03) 0.07± 0.06 (03) 0.02± 0.03 (03) 0.04± 0.07 (03) −0.18 ± 0.08 (03) [Co/Fe] −0.06 ± 0.03 (08) −0.01 ± 0.03 (08) −0.02 ± 0.02 (08) −0.04 ± 0.04 (08) −0.2 0±.03 (08) [Ni/Fe] −0.05 ± 0.07 (25) −0.05 ± 0.08 (24) −0.04 ± 0.05 (24) −0.03 ± 0.05 (26) 0.02± 0.09 (24) [Y/Fe] −0.06 ±0.10 (05) −0.03 ±0.10 (05) −0.09 ±0.06 (05) −0.02 ±0.12 (05) 0.08± 0.16 (05) [Zr/Fe] 0.11 ± 0.10 (02) 0.22 ± 0.08 (02) 0.21 ± 0.07 (02) 0.19 ± 0.04 (02) −0.02 ± 0.11 (02) [Ba/Fe] 0.25 (01)± 0.18 ± 0 (1) 0.23 (01)± 0.18 (01)± −0.22 (01)± [La/Fe] 0.27 0.03 (03) 0.24 ± 0.03 (03) 0.26 0.01 (02) 0.24 0.01 (03) −0.07 0.05 (03) [Ce/Fe] 0.21 ± 0.06 (02) 0.24 ± 0.11 (02) 0.24 ± 0.08 (02) 0.21 ± 0.07 (02) 0.10± 0.11 (02) [Pr/Fe] 0.19 ± 0.03 (02) 0.22 ± 0.06 (02) 0.22 ± 0.04 (02) 0.19 ± 0.01 (02) 0−.19 ±0.04 (02) [Nd/Fe] 0.16 ± 0.07 (05) 0.16 ± 0.03 (05) 0.15 ± 0.06 (05) 0.13 ± 0.05 (05) 0.08± 0.04 (05) [Eu/Fe] 0.10 (01)± 0.08 (01)± 0.13 (01)± 0.14 (01)± −0.38 (01)±

Notes. The elemental abundance ratios are presented together with the abundance scatter from individual lines and a number of lines used for the analysis.

A165, page 8 of 13 V. Bagdonas: Chemical composition of giant stars in the open cluster IC 4756

Table 7. Elemental abundances in doubtful member stars of IC 4756.

El./Star 14 28 44 52 [C/Fe] 0.24 0.03 (02) 0.20 (01) 0.32 0.02 (02) 0.18 0.03 (02) [N/Fe] −0.50 ±0.03 (08) −0.37 0.02 (08) −0.42 ±0.04 (08) −0.49 ±0.05 (08) [O/Fe] 0.10± 0.09 ± 0.08± 0.14 ± C/N 0.72 – 0.72− 0.85 12C/13C 19 – – – [Na/Fe] 0.31 0.03 (03) 0.24 0.06 (03) 0.19 0.01 (03) 0.28 0.06 (03) LTE ± ± ± ± [Na/Fe]NLTE 0.23 0.07 (03) 0.16 0.03 (03) 0.10 0.05 (03) 0.20 0.04 (03) [Mg/Fe] 0.13 ± 0.03 (02) 0.14 ± 0.01 (02) 0.03 ± 0.06 (02) 0.12 ± 0.02 (02) [Al/Fe] 0.09 ± 0.04 (02) 0.14 ± 0.04 (02) 0.06 ± 0.05 (02) 0.19 ± 0.07 (02) [Si/Fe] 0.13 ± 0.1 (17) 0.16 ± 0.10 (17) 0.03 ± 0.08 (16) 0.18 ± 0.10 (13) [Ca/Fe] 0.14 ± 0Vilius.10 (07) Bagdonas et0. al.:08 Chemical± 0.09 composition(05) 0 of.12 giant± stars0.06 in the(07) open cluster0.17 IC 4756± 0.11 (08) ± ± ± ± [Sc II/Fe] 0.08 0.04 (06) 0.02 0.09 (06) 0.01 0.03 (05) 0.04 0.10 (06) Table 5. Atmospheric parameters± of the programme stars,± the Sun, and Arcturus. ± ± [Ti I/Fe] 0.02 0.05 (16) 0.02 0.07 (17) 0.01 0.06 (15) 0.00 0.08 (18) ± ± 1 1± ± [Ti II/Fe] 0.Star03 ID0.02Te(05)ﬀ (K) log g0.02 vt (km0 s.02− ) (05)v sin i (km0. s06− ) [Fe0.05/H](05)σ Fe i0.07n σ0Fe.03ii (05)n [Cr/Fe] 0.02 ± 0.08 (10) −0.03 ±0.07 (09) 0−.07 ±0.10 (10) 0.02 ± 0.07 (09) ± ± Members ± ± [Mn/Fe] 120.12 0.05 5135(03) 2.60.07 1.450.08 (03) 1.60.16 0.002.03(03) 0.060 38.07 0.020.08 (03) 4 [Co/Fe] −380.01 ± 0.02 5165(08) 2.9−0.02 ± 1.350.04 (08) 1.9−0.07 ± 0−. 0.0004 (08) 0.07−0.02 38±0 0.05.04 (08) 5 −42± 5165 2.7− ± 1.25 3.0− ± 0.01 0.04 36± 0.06 5 [Ni/Fe] 0.00 0.11 (21) 0.02 0.11 (24) 0.07 0−.07 (26) 0.03 0.12 (24) 69± 5150 2.7− ± 1.40 2.5− ± 0.00 0.07 38± 0.04 5 [Y/Fe] 0.8101 0.10 5180(05) 2.80.05 1.150.12 (05) 3.70.01 0.070.01(05) 0.060.05 370 0.06.15 (05) 5 [Zr/Fe] 0.10121 ± 0.06 5135(02) 2.80.15 ± 1.400.01 (02) 3.20.16 ± 0.09−0.01(02) 0.070.13 38± 0 0.02.02 (02) 5 − [Ba/Fe] 0.10915 (01)± 5000 2.60.19 (01)± 1.50 2.90.27 (01)± 0.03 0.060.19 38(01)± 0.04 5 125 5150 2.8 1.45 2.6 −0.02 0.05 38 0.02 5 [La/Fe] 0.16430 0.06 5040(03) 2.60.23 1.450.05 (03) 2.20.20 0.07−0.03(03) 0.060.30 380 0.04.06 (03) 5 [Ce/Fe] 0.24 ± 0.14 (02) 0.21 ± 0.13 (02) 0.20 ± 0.15− (02) 0.29 ± 0.12 (02) ± ± Doubtful members± ± [Pr/Fe] 0.1427 0.04 4760(02) 2.30.22 1.500.01 (02) 1.90.14 0.030.06(02) 0.070.31 370 0.04.06 (02) 4 28± 4650 2.1± 1.45 2.2 ± −0.10 0.08 38± 0.06 5 [Nd/Fe] 0.10 0.04 (05) 0.09 0.04 (05) 0.04 0.05− (05) 0.15 0.08 (05) [Eu/Fe] 0.4415 (01)± 5115 2.70.14 (01)± 1.25 3.60.04 (01)± 0.02 0.060.17 37(01)± 0.07 5 52 4500 1.9 1.60 2.8 −0.12 0.08 35 0.05 5 − Notes. The elemental abundance ratiosSun are presented 5777 together 4.44 with the 0.8 abundance scatter 2 from 0.05* individual 0.03 lines 38 and 0.01 a number 5 of lines used for the Arcturus 4345 1.6 1.65 2.4 0.58 0.04 38 0.03 5 analysis. − Notes. * The solar iron abundance in this work is A(Fe) =7.50 (Grevesse & Sauval 2000).

work, and we took the cluster age of 0.89 Gyr as determined 1 D U P S T by ( C L 2 0 1 0 ) 2 . 5 1 D U P S T ( C L 2 0 1 0 ) Strassmeier et al.(2015). We consider all the stars in our T analysisH ( C L 2 0 1 0 ) T H ( C L 2 0 1 0 ) as being in a He-core burning stage. Our determined TO mass of 1 D U P S T ( L 2 0 1 2 ) 1 D U P S T ( L 2 0 1 2 ) IC 4756 is around 2.2 M .3 0 T H + V ( L 2 0 1 2 ) The average 12C/13C value as obtained from three evolved 2 . 0 T H + V ( L 2 0 1 2 ) IC 4756 giants is 19 1.4 and the mean C/N ratio is equal to 0.79 0.05. The previous± analysis of C/N ratios were made C

± 3 1 N / by Luck(1994) and Smiljanic et al.(2009). Our result agrees / 1 . 5

C 2 0 C 2 well with the one by1 Luck(1994) however the mean C/N value reported by Smiljanic et al.(2009) is slightly larger and has a larger scatter (see Table8). From the same stars, Smiljanic et al.(2009) obtained a somewhat lower result for 1 . 0 12 13 C/ C and a larger scatter.1 0 The Luck(1994) result lies in the middle with a result from only one star with the value of 15. The mean 12C/13C ratio determined for this cluster by Gilroy(1989) 0 . 5 agrees well with our determination.1 2 3 4 5 6 1 2 3 4 5 6 M / M From the comparison of our result and the theoreticalؽ mod- M / M ؽ 12 /13 els we can see that theFig. mean 3. The values average carbon of isotopeC C ratiosand in C/Nclump ratiosstars of open clus- Fig. 4. The average carbon-to-nitrogen ratios in clump stars of open lie between the modelters with as a function only the of stellar thermohaline TO mass. Red extra-mixing square indicates the valueFig. of 4. Averageclusters as carbon-to-nitrogen a function of stellar TO mass.ratios In in addition clump to stars symbols of inopen Fig. clusters included and the modelIC which4756. Red also open includes squares indicate the rotation-induced results of previously investigatedas a function3, here we of include stellar the results TO mass. from Tautvaišien In additione˙ et al. (2015)to symbols as red open in Fig.3, open clusters by Tautvaišiene˙ et al. (2000, 2005, 2016); Mikolaitishere et al. wetriangles. include the results from Tautvaišiene˙ et al.(2015) as red open mixing. The rotation(2010, was 2011a,b,most probably 2012); Drazdauskas smaller et al.in (2016a,b). the inves- Other symbols include results from Gilroy1 (1989) – pluses, Luck (1994) – opentriangles. cir- tigated IC 4756 starscles, than Smiljanic 120 km et al. s− (2009)when – green they crosses, were Santrich on the et al. (2013) – ZAMS. When lookingopen at the diamond. C/N The values solid linesfrom (1DUP ours ST) and represent previous the 12C/13C ra- leaves the main sequence and becomes a red giant, the convec- 4.3. Sodiumtive envelope of the star deepens, and the processed material studies, it is obvioustios that predicted the pure for stars thermohaline at the first dredge-up extra-mixing with standard stellar evolutionary models of solar metallicity by Charbonnel & Lagarde (2010) from the core is brought up to the surface. By looking at abun- model is preferred for(black the solid open line) clustersand Lagarde with et al. (2012) the (blueTO solid masses line). TheSodium, short- dance besides alterations carbon of certain and nitrogen,elements in isstellar one photospheres, of the other we mixing- below 2.2 M . There weredashed some line (TH) debates shows the (see prediction Wachlin when et only al. thermohaline 2014 sensitive extra- can chemicalanalyse the e elements.fficiency of the Theoretical transport mechanisms. models Among by Lagarde mixing is introduced (Charbonnel & Lagarde 2010), and the long- the most sensitive elements are carbon and nitrogen. 12C/13C and references therein)dashed that line (TH thermohaline+V) is for the model convection that includesboth models the thermohalineet al.(2012and 12)C/ predict14N ratios a are significant particularly good increase indicators of of sodium mixing in abun- greatly overestimate theandeffects rotation-induced of such mixing mixing, (Lagarde and et al. some 2012). A other typical errordance bar stars. after the first dredge-up for stars with turn-off masses is indicated (Charbonnel & Lagarde 2010; Smiljanic et al. 2009; Gilroy explanation might be needed.1989). of aroundCarbon 2 M and, nitrogen and an abundances even larger in evolved overabundance stars have been if the investigated for more than 45 years, since studies by Day et al. (1973); Tomkin & Lambert (1974); Tomkin et al. (1975);A165, Dear- page 9 of 13 Article number, page 7 of 14 A&A 615, A165 (2018)

Table 8. Comparison of mean IC 4756 abundances determined in this work and previous studies.

Element /Resolution This study Luck(1994) Jacobson et al.(2007) Smiljanic et al.(2009) Pace et al.(2010) Maiorca et al.(2011) Ting et al.(2012) 48 000 18 000 15 000 48 000 100 000 100 000 30 000 [C/Fe] 0.26 0.03 0.27 0.05 ... 0.15 0.02 ...... − ± − ± − ± [N/Fe] 0.45 0.02 0.54 0.16 ... 0.43 0.07 ...... ± ± ± [O/Fe] 0.01 0.05 0.17 0.09 ... 0.01 0.01 <0.07 ...... ± − ± − ± C/N 0.79 0.05 0.80 0.20 ... 1.05 0.12 ...... ± ± ± 12C/13C 19 1.4 15 ... 13 3.1 ...... ± ± [Na/Fe] 0.22 0.05 0.14 0.01 0.57 0.06 ... 0.11 0.04 ... 0.21 0.15 LTE ± ± ± ± ± [Na/Fe] 0.14 0.05 ...... 0.01 0.04 ...... NLTE ± − ± [Mg/Fe] 0.08 0.05 0.20 0.14 ... 0.05 0.02 ...... 0.12 0.13 ± ± − ± ± [Al/Fe] 0.04 0.03 0.07 0.21 0.29 0.08 ... 0.11 0.05 ... 0.12 0.12 ± ± ± − ± ± [Si/Fe] 0.04 0.03 0.23 0.04 0.34 0.06 0.06 0.04 0.02 0.01 ... 0.13 0.13 ± ± ± ± ± ± [Ca/Fe] 0.11 0.03 0.07 0.13 0.07 0.08 0.02 0.04 0.02 0.03 ... 0.05 0.16 ± − ± ± ± − ± ± [Sc II/Fe] 0.01 0.04 0.03 0.12 ... 0.06 0.07 ...... − ± ± ± [Ti I/Fe] 0.00 0.02 0.09 0.13 ... 0.04 0.01 0.03 0.03 ...... ± − ± − ± ± [Ti II/Fe] 0.03 0.05 ...... 0.28 0.08 − ± ± [Fe I/H] 0.02 0.01 0.05 0.05 0.15 0.04 0.04 0.03 0.08 0.02 0.01 0.03 0.01 0.10 − ± − ± − ± ± ± ± − ± [Fe II/H] 0.02 0.02 0.04 0.04 ... 0.04 0.03 ...... 0.00 0.11 − ± − ± ± ± [Cr/Fe] 0.03 0.03 0.08 0.04 ... 0.04 0.05 0.00 0.03 ... 0.03 0.15 ± ± ± ± ± [Mn/Fe] 0.07 0.05 0.20 0.18 ...... − ± ± [Co/Fe] 0.04 0.02 0.32 0.16 ... 0.06 0.04 ...... − ± ± ± [Ni/Fe] 0.06 0.01 0.08 0.07 0.08 0.05 0.01 0.02 0.04 0.01 ... 0.03 0.13 − ± ± ± − ± − ± ± [Y/Fe] 0.03 0.04 0.37 0.03 ...... 0.11 0.01 ... ± ± ± [Zr/Fe] 0.18 0.04 0.25 0.13 ...... 0.09 0.02 ... ± ± ± [Ba/Fe] 0.23 0.04 0.01 0.21 ...... 0.00 0.14 ± ± ± [La/Fe] 0.22 0.04 ...... 0.19 0.01 ... ± ± [Ce/Fe] 0.21 0.03 ...... 0.16 0.02 ... ± ± [Pr/Fe] 0.18 0.04 ...... ± [Nd/Fe] 0.12 0.04 0.37 0.01 ...... ± ± [Eu/Fe] 0.08 0.04 0.16 ...... ± Notes. The abundances marked in bold were determined using spectral syntheses. The remaining abundances were derived from EWs. Along with the mean elemental abundances, values of root mean square between stars are presented. extra-mixing is taken into account. Our results and several average abundances were calculated simply by taking the aver- other recent NLTE determinations of sodium abundances in age of [Mg/Fe], [Ca/Fe], [Si/Fe], and [Ti/Fe] values. We find other studies (Drazdauskas et al. 2016b; MacLean et al. 2015; no α-element enhancement in our cluster, and our result shows Smiljanic et al. 2016) are displayed in Fig.5 together with the an almost perfect match with the theoretical model by Pagel & theoretical models by Lagarde et al.(2012). Following Smiljanic Tautvaisiene(1997). et al.(2016), only high-probability giant stars were extracted We take a look at oxygen and magnesium abundances sep- from the paper by MacLean et al.(2015). The theoretical mod- arately in Fig.7 and compare them with models by Magrini els show how sodium abundance increases in relation to the TO et al.(2009) and abundance results from other studies. Oxy- mass. At the TO mass of IC 4756, which is around 2.2 M , gen is one of the more difficult elements to precisely analyse the average NLTE sodium abundance, as in case of 12C/13 C due to the relations with carbon and nitrogen. Therefore, for and C/N, lies between the theoretical model which includes the the comparison we take the results from studies where the thermohaline- and rotation-induced extra-mixing and the model oxygen abundance was determined in the same way as in of the first dredge-up (or the model of pure thermohaline mixing this study (Mikolaitis et al. 2010, 2011a,b; Tautvaišiene˙ et al. since they both give very close [Na/Fe] values). These results 2015; Drazdauskas et al. 2016a,b). For the magnesium, we confirm a conclusion by other investigators who suggest that have taken results from the same authors as other α-elements the trend of sodium abundance increases in relation to a TO (Jacobson et al. 2008, 2009; Friel et al. 2010; Mikolaitis et al. mass and is most probably caused by internal stellar evolutionary 2010, 2011a,b; Reddy et al. 2012, 2013, 2015; Mishenina et al. processes (Smiljanic et al. 2016 and references therein). 2015). The models depict the oxygen and magnesium abundance trends in relation to the Galactocentric distance. Our 4.4. α-Elements results for IC 4756 follow the models, and at Rgc of 8.1 kpc show no visible over- or under-abundances for oxygen or mag- α-Elements are of particular interest when studying Galactic nesium. archaeology due to a different timescale of their production compared to iron-peak elements. These elements are mainly pro- 4.5. Iron-peak elements duced in massive stars, thus any visible enhancement in their abundances can reveal differences of star formation histories in The mean abundances of five iron-peak elements investigated in different parts of the Galaxy. our sample of IC 4756 RGB stars display no significant devia- Figure6 displays the determined average α-element abun- tions from the mean iron abundance. These results are consistent dance in IC 4756 together with results from other studies. The with previous abundance determinations for this cluster (see

A165, page 10 of 13 V. Bagdonas: Chemical composition of giant stars in the open cluster IC 4756 A&A proofs: manuscript no. output Vilius Bagdonas et al.: Chemical composition of giant stars in the open cluster IC 4756

0 . 5 M a g r i n i e t a l . ( 2 0 0 9 ) O M g 4.5. Iron-peak elements P r e s e n t 0 . 4 0.5 0 . 5 4 G y r s a g o The mean abundances of ﬁve iron-peak elements investigated in 0 . 3 ] e F

] our sample of IC 4756 RGB stars display no signiﬁcant devia- / ] i

0 . 2 e e T F ,

/ tions from the mean iron abundance. These results are consistent F 0 . 0

l a / 0 . 1 E a C [ ,

i 0.0 with previous abundance determinations for this cluster (see Ta- N [ 0 . 0 S , ble 8). Larger discrepancies are coming from the study by Luck 1 D U P ( L 2 0 1 2 ) g - 0 . 5

- 0 . 1 M T H + V ( L 2 0 1 2 ) [ (1994) who reported relatively large values and scatters for man- - 0 . 2 ganese and cobalt, which might be caused by uncertainties in - 0 . 3 −0.5 - 1 . 0 accounting for a hyperﬁne structure in these element lines. O M g 1 2 3 4 5 6 −0.5 −0.4 −0.3 −0.2 −0.1 0.0 0.1 0.2 0.3 M / M 0 . 5 ؽ [ F e / H ] 4.6. Neutron-capture elements / ] Fig. 5. The mean [Na Fe]NLTE abundances in open clusters compared H

Fig. 6. Mean/ α-element abundances in open cluster RGB stars. The Fig. 5. Mean [Na/Fe]NLTE abundances in open clusters compared to the- l 0 . 0

to theoretical models by Lagarde et al. (2012). The result obtained in E Along with our results, in Table 8 we present abundances of oretical models by Lagarde et al.(2012). The result obtained in thisresult for[ this study is indicated by the red circle. Results from Taut- this study is marked with the red square. The red open squares indicate vaišiene˙ et al. (2005); Mikolaitis et al. (2010, 2011a,b); Drazdauskas neutron-capture elements determined in IC 4756 by several other studyresults is marked from Drazdauskas with the et red al. (2016b). square. The The results red openfrom MacLean squares et indicate al. results from Drazdauskas et al.(2016b). The results from MacLean et al.et al. (2016b)- 0 . 5 are marked as the red open squares. The blue triangles in- studies. The results by Luck (1994) were averaged from one su- (2015) are indicated as the open grey circles and from Smiljanic et al. dicate results from Reddy et al. (2012, 2013, 2015); the green plus signs (2015(2016)) are are indicated shown as as the the green open open grey circles. circles and from Smiljanic et al.from Mishenina et al. (2015); and the black reverse triangles from Friel pergiant and two giants. Maiorca et al. (2011) investigated three A&A proofs: manuscript(2016) no. are output shown as the green open circles. dwarfs using an EW method, while Ting et al. (2012) from 12 et al. (2010);- 1 . 0 Jacobson et al. (2008, 2009). The black line represents the Galactic disc evolution6 model8 by1 0 Pagel1 &2 Tautvaisiene1 4 6 (1995).8 1 0 1 2 1 4 giant stars (10 of which are common with our work) determined born et al. (1976) and others. It is well known that stars undergo R , k p c R , k p c solely the barium abundance. 0 . 5 several mixing events during their evolution, however only one, g c g c the ﬁrst dredge-up, was predicted by the classical theory of stel- ter. The Luck (1994) result lies in the middle with a result from Y and Zr could be attributed to the elements that lie in the 0 . 4 Mean oxygen and magnesium abundances compared to the the- lar0.5 evolution until a star leaves the red clump position on the HR only oneFig. star 7. Mean with the oxygen value and of magnesium15. The mean abundances12C/13C ratio compared de- to the the- ﬁrst s-process peak and are referred to as the light s-process el- 0 . 3 ] oretical models by Magrini et et al. al.( (2009).2009). The The oxygen oxygen results results are are taken taken e diagram. As concerns studies of evolved stars in open clusters, termined for this cluster by Gilroy (1989) agrees well with our ements. Their production scenarios are similar. In the Sun, 74% F

/ 12 13 from Tautvaišiene˙ et al. (2005); Mikolaitis et al. (2010, 2011a,b); Taut- ] 0 . 2 i inconsistencies between the C/ C ratios predicted by the clas- determination.from Tautvaišiene˙ et al.(2005, 2015); Mikolaitis et al.(2010, 2011a,b); of yttrium and 67% of zirconium are produced via the s-process e T , vaišienDrazdauskase˙ et al. et (2015); al.(2016a Drazdauskas,b). Symbols et are al. the (2016a,b). same as inSymbols Fig.6. are the F sical stellar evolution model and the observational results were From the comparison of our result and the theoretical mod- a / (Travaglio et al. 2004). The main s-process component, which 0 . 1 same as in Fig. 6. 12 13 a

C / , clearly demonstrated by Gilroy (1989). The observational results els we can see that the mean values of C C and C/N ratios

i 0.0 gathers free neutrons from low-mass AGB stars, contributes to N [ 0 . 0 S

, agreed relatively well with the classical model for stars in open lie between the model with only the thermohaline extra-mixing

g 69% of Y and to 65% of Zr, while the weak s-process compo- 1 D U P ( L 2 0 1 2 ) clusters with turn-oﬀ (TO) masses larger than 2.2 M . How- included and the model which also includes the rotation-induced

- 0 . 1 M [ ∼ 69% of Y and to 65% of Zr, while the weak s-process component contributes only 5% of Y and 2% of Zr. The rapid neutron T H + V ( L 2 0 1 2 ) ever, for stars in open clusters with lower turn-off masses the ob- mixing.dance The increases rotation was in mostrelation probably to a TO smaller mass in and the inves- is most probably - 0 . 2 12 13 nent contributes only 5% of1 Y and 2% of Zr. The rapid neutron capture process (r-process) is responsible for 8% of Y and 15% served C/ C ratios were decreasing with decreasing TO mass. tigatedcaused IC 4756 by stars internal than stellar 120 km evolutionary s− when they processes were on (Smiljanic the et al. capture process (r-process)/ is responsible for 8% of Y and 15% of Zr. A responsible process for production of the remaining 18% - 0 . 3 −0.5In order to explain discrepancies between theoretical and ob- ZAMS.2016 When and looking references at the therein). C N values from ours and previ- served abundances of mixing-sensitive chemical elements, some ous studies,of Zr. itA is responsible obvious that process the pure thermohalinefor production extra-mixing of the remaining 18% of Y and Zr is debatable. Travaglio et al. (2004) called this pro- 1 2 3 4 5 6 other transport−0.5 − mechanisms0.4 −0.3 besides−0.2 − the0.1 classical0.0 convection0.1 0.2 had 0.model3 of is Y preferred and Zr is for debatable. the open clustersTravaglio with et the al.( TO2004 masses) called this process the lighter element primary process (LEPP) which possibly .M / M ؽ to be introduced. Nowadays, the most promising of these seems below 2.2 M . There were some debates (see Wachlin et al [ F e / H ] 4.4.cessα the-elements lighter element primary process (LEPP) which possibly acts in low-metallicity massive stars. Therefore, due to the simi- to be a thermohaline-induced extra-mixing. We compare our re- 2014acts and references in low-metallicity therein) that massive thermohaline stars. convection Therefore, mod- due to the similarities of production of Y and Zr, abundances of these elements The mean [Na/Fe] abundances in open clusters compared sults with the first dredge-up model, a model with only the ther- els greatly overestimate the effects of such mixing, and some Fig. 5. NLTE Fig.Fig. 6. Mean α-element abundances in open cluster RGB stars. stars. The The αlarities of production of Y and Zr, abundances of these elements should be similar. As displayed in Fig. 8, the clusters’ results mohaline mixing included (Charbonnel & Lagarde 2010) and other explanation-elements might are of be particular needed. interest when studying Galactic ar- to theoretical models by Lagarde et al. (2012). The result obtained in resultresult for for this this study study is is indicated indicated by by the the red circle.red circle. Results Results from Taut-from with a model where rotation and thermohaline-induced mixing chaeologyshould be duesimilar. to a As different displayed timescale in Fig. of8, their the productionclusters’ results com- from various studies of Y and Zr exhibit abundances roughly this study is marked with the red square. The red open squares indicate Tautvaišienvaišiene˙ ete˙ al. et al.(2005);(2005); Mikolaitis Mikolaitis et et al. al. (2010,(2010, 2011a 2011a,b);,b); Drazdauskas Drazdauskas act together (Lagarde et al. 2012). Stellar rotation may modify paredfrom various to iron-peak studies elements. of Y and These Zr exhibit elements abundances are mainly roughly pro- confined between 0 and 0.2 dex, with several values scattered results from Drazdauskas et al. (2016b). The results from MacLean et al. et al. (2016b) are marked as the red open squares. The blue triangles in-4.3. Sodium et al.an(2016b internal) are stellar marked structure as the even red before open the squares. RGB phase,The blue however triangles ducedconfined in betweenmassive stars,0 and thus 0.2 dex, any visiblewith several enhancement values scattered in their around this range. Our programme cluster IC 4756 displays av- (2015) are indicated as the open grey circles and from Smiljanic et al. indicatedicate results results from from Reddy Reddy et al. et (2012, al.(2012 2013,, 2013 2015);, 2015 the); thegreen green plus plus signs results become visible only in later stages. The initial rotation ve- Sodium,abundancesaround besides this carbon range. can revealand Our nitrogen, programme differences is one of ofcluster the star other formationIC mixing- 4756 displays histories aver- in erage abundance ratios of 0.03 dex for [Y/Fe], and 0.18 dex for (2016) are shown as the green open circles. signsfrom fromMishenina Mishenina et al. et(2015); al.(2015 and); the and black the black reverse reverse triangles triangles from from Friel locity of the models on the zero age main sequence (ZAMS) was sensitiveageff chemical abundance elements. ratios Theoreticalof 0.03 dex models for [Y/Fe], by Lagarde and 0.18 dex for [Zr/Fe]. Abundances of those elements are not similar, however, Frielet al. et(2010); al.(2010 Jacobson); Jacobson et al. et (2008, al.(2008 2009)., 2009 The). black The black line represents line repre- the di erent parts of the Galaxy. chosen at 30% of the critical velocity at that point and leads to et al.[Zr/Fe]. (2012) predict Abundances a significant of those increase elements of sodium are abundance not similar, however, they follow the modeled trends by Maiorca et al. (2012) quite sentsGalactic the Galacticdisc evolution disc evolution model by model Pagel by &1 Pagel Tautvaisiene & Tautvaisiene (1995). (1995). Figure 6 displays the determined average α-element abun- the mean velocity of about 120 km s− (Lagarde et al. 2014). We after the first dredge-up for stars with turn-off masses of around born et al. (1976) and others. It is well known that stars undergo dancethey follow in IC the4756 modeled together trends with results by Maiorca from etother al.( studies.2012) quite The well. present the comparison of our results and the models in Figs. 3 2 M ,well. and an even larger overabundance if the extra-mixing is several mixing events during their evolution, however only one, and 4. The TO mass of IC 4756 was determined using PAR- taken average into account. abundances Our results were and calculated several other simply recentby NLTE taking the av- Ba, La, and Ce belong to the second s-process peak. Accord- the first dredge-up, was predicted by the classical theory of stel- Tableter.SEC The8). isochrones Luck Larger (1994) discrepancies (Bressan result et lies al. are2012). in the comingThe middle input from with metallicity the a result study of from by erageBa, of La, [Mg and/Fe], Ce [Ca belong/Fe], to [Si the/Fe], second and s-process[Ti/Fe] values. peak. WeAccord- find ing to Arlandini et al. (1999), s-process contributes to Ba, La, 12 13 determinations of sodium abundances in other studies (Draz- lar evolution until a star leaves the red clump position on the HR Luckonly0( one1994.02 dex star) who was with thereported the one value we relatively determined of 15. largeThe in this mean values work, andC and/ scatters weC ratiotook for de-dauskasing etα to al. Arlandini 2016b; MacLean et al.(1999 et al.), 2015,s-process and Smiljanic contributes et al. to Ba, La, and and Ce production by 81%, 62%, and 77%, respectively. The − no -element enhancement in our cluster, and our result shows diagram. As concerns studies of evolved stars in open clusters, manganeseterminedthe cluster for and age this cobalt, of cluster 0.89 which Gyr by as Gilroy might determined (1989) be caused by agrees Strassmeier by well uncertainties et with al. our2016)anCe are almost production displayed perfect in Fig. by match81%, 5 together 62%, with with andthe the theoretical 77%, theoretical respectively. modelsmodel by The Pagel newer & newer study by Bisterzo et al. (2016) provided slightly different inconsistencies between the 12C/13C ratios predicted by the clas- indetermination. accounting(2015). We for consider a hyperfine all the stars structure in our in analysis these element as being lines. in a by (LagardeTautvaisienestudy by et al.Bisterzo 2012). (1997). Following et al.(2016 Smiljanic) provided et al. slightly (2016), different only results, results, indicating the s-process enrichment of Ba, La, and Ce by sical stellar evolution model and the observational results were He-coreFrom the burning comparison stage. Our of determined our result TOand mass the theoretical of IC 4756 is mod-high-probabilityindicating giant the s stars-process were enrichment extracted from of Ba, the paper La, and by Ce by 83%, 83%, 73%, and 81%, respectively. This strengthens a key fea- around 2.2 M . 12 13 MacLeanWe et al. take (2015). a look The theoreticalat oxygen models and magnesium show how sodium abundances sepa- clearly demonstrated by Gilroy (1989). The observational results els we can see that the mean values of C/ C and C/N ratios 73%, and 81%, respectively. This strengthens a key feature of the 4.6. Neutron-capture12 / elements13 rately in Fig. 7 and compare them with models by Magrini et al. ture of the elements in the second peak that all elements beyond agreed relatively well with the classical model for stars in open lie betweenThe average the modelC withC value only as the obtained thermohaline from three extra-mixing evolved abundanceelements increases in the in relation second to peak the TO that mass. all At elements the TO mass beyond A = 90, IC 4756 giants is 19 1.4 and the mean C/N ratio is equal to of IC(2009) 4756, which and abundance is around 2.2 resultsM , the from average other NLTE studies. sodium Oxygen is one A = 90, except for Pb, are mainly produced via the main s- Along with our results, in Table8 we present abundances of except for Pb, are mainly produced via the main s-process com- clusters with turn-off (TO) masses larger than 2.2 M . How- included0.79 and0.05. the The model previous± which analysis also of includes C/N ratios the were rotation-induced made by abundance,of the as more in case di offfi12cultC/13 elementsC and C/N, to lies precisely between analysethe the- due to the process component (Käppeler et al. 2011). The remaining abun- ∼ neutron-capture± elements determined in IC 4756 by several other ponent (Käppeler et al. 2011). The remaining abundances of ever, for stars in open clusters with lower turn-off masses the ob- mixing.Luck (1994) The rotation and Smiljanic was most et al. (2009).probably Our smaller result agrees in the well inves-oreticalrelations model which with carbon includes and the nitrogen.thermohaline- Therefore, and rotation- for the compar- dances of these elements are provided by the r-process. Thus, the 12 13 studies. The results by Luck(1994) were1 averaged from one these elements are provided by the r-process. Thus, the similar- served C/ C ratios were decreasing with decreasing TO mass. tigatedwith ICthe 4756 one by stars Luck than (1994) 120 however km s− thewhen mean they C/N werevalue re- on theinducedison extra-mixing we take the and results the model from of studies the first where dredge-up the (or oxygen abun- similarity of the Ba, La and Ce production should be reflected in supergiant and two giants. Maiorca/ et al.(2011) investigated ity of the Ba, La, and Ce production should be reflected in their In order to explain discrepancies between theoretical and ob- ZAMS.ported When by Smiljanic looking et al. at (2009) the C isN slightly values larger from and ours has a and larger previ-the modeldance ofwas pure thermohalinedetermined mixingin the samesince they way both as ingive this very study (Miko- their abundance similarities. By looking at Table 8, we see that threeousscatter studies, dwarfs (see it Table usingis obvious 8). an From EW that the themethod, same pure stars, thermohaline while Smiljanic Ting et et al.extra-mixing(2009) al.(2012)closeabundance [Na/Fe] values). similarities. These results By looking confirm at a conclusion Table8, we by see that the served abundances of mixing-sensitive chemical elements, some 12 13 laitis et al. 2010, 2011a,b; Tautvaišiene et al. 2015; Drazdauskas the average abundances of these elements are identical in our fromobtained 12 giant a somewhat stars (10 lower of which result for areC common/ C and awith larger our scat- work)otheraverage investigators abundances who suggest of these that the elements trend of˙ are sodium identical abun- in our work. other transport mechanisms besides the classical convection had model is preferred for the open clusters with the TO masses et al. 2016a,b). For the magnesium, we have taken results from work. determinedM solely the barium abundance. Nd and Pr are elements that have a mixed origin, as they to be introduced. Nowadays, the most promising of these seems belowArticle 2.2 number,. page There 8 of were 14 some debates (see Wachlin et al. the same authors as other α-elements (Mikolaitis et al. 2010, Nd and Pr are elements that have a mixed origin, as they are Y and Zr could be attributed to the elements that lie in the are produced via s- and r-processes in roughly equal fractions. to be a thermohaline-induced extra-mixing. We compare our re- 2014 and references therein) that thermohaline convection mod- 2011a,b; Reddy et al. 2012, 2013, 2015; Mishenina et al. 2015; produced via s- and r-processes in roughly equal fractions. As first s-process peak and are referredff to as the light s-process ele- As reported by Arlandini et al.(1999) and Bisterzo et al.(2016), sults with the first dredge-up model, a model with only the ther- els greatly overestimate the e ects of such mixing, and some Friel et al. 2010; Jacobson et al. 2008, 2009). The models de- reported by Arlandini et al. (1999) and Bisterzo et al. (2016), s- ments. Their production scenarios are similar. In the Sun, 74% s-process produces around 49% of Pr and 56% of Nd, while the mohaline mixing included (Charbonnel & Lagarde 2010) and other explanation might be needed. pict the oxygen and magnesium abundance trends in relation to process produces around 49% of Pr and 56% of Nd, while the of yttrium and 67% of zirconium are produced via the s-process remaining fractions of the elements are created via the r-process with a model where rotation and thermohaline-induced mixing the Galactocentric distance. Our results for IC 4756 follow the remaining fractions of the elements are created via the r-process (Travaglio et al. 2004). The main s-process component, which 51% and 44%, respectively. Rather similar abundances of Nd act together (Lagarde et al. 2012). Stellar rotation may modify models, and at R of 8.1 kpc show no visible over- or under- – 51% and 44%, respectively. Rather similar abundances of Nd gathers4.3. Sodium free neutrons from low-mass AGB stars, contributes to and− Pr in IC 4756gc confirm their formation scenario. an internal stellar structure even before the RGB phase, however abundances for oxygen or magnesium. and Pr in IC 4756 confirm their formation scenario. results become visible only in later stages. The initial rotation ve- Sodium, besides carbon and nitrogen, is one of the other mixing- A165, page 11 of 13 locity of the models on the zero age main sequence (ZAMS) was sensitive chemical elements. Theoretical models by Lagarde Article number, page 9 of 14 chosen at 30% of the critical velocity at that point and leads to et al. (2012) predict a significant increase of sodium abundance 1 the mean velocity of about 120 km s− (Lagarde et al. 2014). We after the first dredge-up for stars with turn-off masses of around present the comparison of our results and the models in Figs. 3 2 M , and an even larger overabundance if the extra-mixing is and 4. The TO mass of IC 4756 was determined using PAR- taken into account. Our results and several other recent NLTE SEC isochrones (Bressan et al. 2012). The input metallicity of determinations of sodium abundances in other studies (Draz- 0.02 dex was the one we determined in this work, and we took dauskas et al. 2016b; MacLean et al. 2015, and Smiljanic et al. − the cluster age of 0.89 Gyr as determined by Strassmeier et al. 2016) are displayed in Fig. 5 together with the theoretical models (2015). We consider all the stars in our analysis as being in a by (Lagarde et al. 2012). Following Smiljanic et al. (2016), only He-core burning stage. Our determined TO mass of IC 4756 is high-probability giant stars were extracted from the paper by around 2.2 M . MacLean et al. (2015). The theoretical models show how sodium

The average 12C/13C value as obtained from three evolved abundance increases in relation to the TO mass. At the TO mass IC 4756 giants is 19 1.4 and the mean C/N ratio is equal to of IC 4756, which is around 2.2 M , the average NLTE sodium 0.79 0.05. The previous± analysis of C/N ratios were made by abundance, as in case of 12C/13C and C/N, lies between the the- Luck± (1994) and Smiljanic et al. (2009). Our result agrees well oretical model which includes the thermohaline- and rotation- with the one by Luck (1994) however the mean C/N value re- induced extra-mixing and the model of the ﬁrst dredge-up (or ported by Smiljanic et al. (2009) is slightly larger and has a larger the model of pure thermohaline mixing since they both give very scatter (see Table 8). From the same stars, Smiljanic et al. (2009) close [Na/Fe] values). These results conﬁrm a conclusion by obtained a somewhat lower result for 12C/13C and a larger scat- other investigators who suggest that the trend of sodium abun-

Article number, page 8 of 14 A&A proofs:A&A 615,manuscript A165 (2018) no. output

Abundancesa relatively young of all investigated open cluster elements with an age in IC of 4756 around agree 0.8 well Gyr. withAbundances the model of byall Maiorcainvestigated et al. elements (2012) in for IC the 4756 Solar agree radius. well Thiswith is the applicable model by also Maiorca to barium et (nine al.(2012 IC 4756) forgiants the Solar provide radius. the averageThis is applicable [Ba/Fe]= 0 also.22 to0 barium.03), while (nine [Ba IC/Fe] 4756 in giants OC investigated provide the average [Ba/Fe] = 0.22± 0.03), while [Ba/Fe] in OC investigated in several other studies± at similar ages have values from 0.3 dex toin 0.6 several dex exceeding other studies predictions at similar of ages the model have values (see Fig. from 8). 0.3 dex to 0.6 dex exceeding predictions of the model (see Fig.8). 5. Conclusions 5. Conclusions Recognising the role that open clusters play in the establish- mentRecognising of Galactic the rolechemical that open evolution clusters models, play we in the performed establish- a detailedment of high-resolution Galactic chemical spectroscopic evolution analysis models, of we the performed open clus- a terdetailed IC 4756. high-resolution We determined spectroscopic the main atmospheric analysis of parameters the open clus- of 13ter giant IC 4756. stars, We with determined 9 of them the being main the atmospheric high-probability parameters mem- bersof 13 of thegiant cluster. stars, We with also 9 carried of them out being a comprehensive the high-probability analysis ofmembers 23 chemical of the elements cluster. made We in also various carried stellar out phases a comprehen- and processes.sive analysis The key of results 23 chemical of our analysis elements are made as follows: in various stellar phases and processes. The key results of our analysis are as follows:– IC 4756 has a metallicity close to Solar. [Fe/H] = 0.02 − ± – 0IC.01, 4756 whereas has a other metallicity iron-peak close elements to Solar. do not[Fe di/H]ffer= by0 more.02 than0.01 0.1, whereas dex from other Solar iron-peak values. elements do not differ by− more± α – Ourthan determined 0.1 dex from average Solar values.-element abundances show a slight – enrichmentOur determined of 0.07 average dex comparedα-element to iron, abundances however that show is ex- a pectedslight from enrichment the thin of disc 0.07 chemical dex compared evolution to model iron, by however Pagel &that Tautvaisiene is expected (1995). from the The thin results disc chemical for oxygen evolution and magne- model siumby Pagel show & no Tautvaisiene deviation from(1995 the). The models results by for Magrini oxygen et and al. (2009),magnesium also at show the given no deviation galactocentric from the distance models of by 8.1 Magrini kpc. The mean ratio of carbon and nitrogen, C/N = 0.79 0.05, – et al.(2009), also at the given galactocentric distance of and the carbon isotope ratio, 12C/13C = 19 1.41, are± altered 8.1 kpc. more than predicted in the first dredge-up model,± and lie be- – The mean ratio of carbon and nitrogen, C/N = 0.79 0.05, tween the model where only the thermohaline extra-mixing and the carbon isotope ratio, 12C/13C = 19 1.41± , are is included and the model which also includes the rotation- altered more than predicted in the first dredge-up± model, induced mixing (Lagarde et al. 2012). The value we obtained Fig.Fig. 8. 8.AveragedAveraged values values of of neutron-capture neutron-capture element element abundances abundances vs. vs. age age and lie between the model where only the thermohaline for sodium with NLTE corrections, [Na/Fe] = 0.14 0.05, inin IC IC 4756 4756 (red (red filled filled squares) squares) and and other other previously previously investigated investigated clusters. clusters. extra-mixing is included and the model which also includes The green plus signs indicate results by Reddy et al. (2012, 2013, 2015); is larger than the first dredge-up prediction, however± this is The green plus signs indicate results by Reddy et al.(2012, 2013, 2015); the rotation-induced mixing (Lagarde et al. 2012). The thethe blue blue triangles triangles – – results results by by MisheninaMishenina et et al. al.( (2015);2015); the the black black crosses crosses lower than the value predicted by the model of thermohaline- value we obtained for sodium with NLTE corrections, –– results results by by D’OraziD’Orazi et et al. al.( (2009)2009) and and Maiorca et al.(2011 (2011);); the the black black and rotation-induced extra mixing. The rotation was most [Na/Fe] = 0.14 0.05, is larger than the first dredge-up pre- reversereverse triangles triangles – – results results by by JacobsonJacobson & & Friel Friel( (2013)2013) and and Overbeek Overbeek probably smaller in IC 4756 stars than the 30% of the crit- diction, however± this is lower than the value predicted by the etet al. al.( (2016);2016); the the empty empty red red squares squares – – results results from from our our previous previous studies studies ical rotation velocity at the ZAMS. model of thermohaline- and rotation-induced extra mixing. byby TautvaišienTautvaišiene˙e˙ et et al. al.( (2000,2000, 2005 2005),), Mikolaitis Mikolaitis et et al.(2010 (2010,, 2011a 2011a,b),b) – Being relatively young, the open cluster IC 4756 displays a and Drazdauskas et al. (2016a). The adopted approximate age errors The rotation was most probably smaller in IC 4756 stars and Drazdauskas et al.(2016a). The adopted approximate age errors moderate enrichment of s-process-dominated chemical el- ofof 0.9 0.9 Gyr Gyr are are based based on on determinations determinations by by (Salaris (Salaris et et al. al. 2004).2004). The The than the 30% of the critical rotation velocity at the ZAMS. ements compared to the Galactic thin disc model (Pagel continuouscontinuous lines lines indicate indicate a a chemical chemical evolution evolution model model by by MaiorcaMaiorca et et al. al. – Being relatively young, the open cluster IC 4756 displays & Tautvaisiene 1997) and confirms the enrichment of s- (2012)(2012) at the Solar radius. a moderate enrichment of s-process-dominated chemi- process-dominated elements in young open clusters com- cal elements compared to the Galactic thin disc model pared to the older ones. Abundances of all investigated s- (Pagel & Tautvaisiene 1997) and confirms the enrichment Finally, europium is an almost purely r-process-dominated process-dominated elements in IC 4756 agree well with the Finally, europium is an almost purely r-process-dominated of s-process-dominated elements in young open clusters element, as only 6% of its production takes place via the s- model for the Solar radius by Maiorca et al. (2012). The r- element, as only 6% of its production takes place via the s- compared to the older ones. Abundances of all investigated process and the remaining 94% via the r-process (Bisterzo et al. process-dominated element europium abundance agrees with process and the remaining 94% via the r-process (Bisterzo et al. s-process-dominated elements in IC 4756 agree well with 2016). The europium abundance in IC 4756 is similar to the thin the thin disc abundance. 2016). The europium abundance in IC 4756 is similar to the thin the model for the Solar radius by Maiorca et al.(2012). The discdisc chemical chemical content content (e.g. (e.g. PagelPagel & Tautvaisiene 1997).1997). From the comparison of n-capture chemical element abun- Acknowledgements.r-process-dominatedThis research element has made europiumuse of the WEBDA abundance database agrees (oper- From the comparison of n-capture chemical element abun- ated atwith the Department the thin disc of Theoretical abundance. Physics and Astrophysics of the Masaryk dancesdances in in open open clusters clusters and and a a semi-empirical semi-empirical models models of of the the University, Brno), of SIMBAD (operated at CDS, Strasbourg), of VALD (Kupka GalacticGalactic thin thin disc disc chemical chemical evolution evolution by by PagelPagel & & Tautvaisiene Tautvaisiene et al. (2000)), and of NASA’s Astrophysics Data System. Bertrand Plez (Univer- Acknowledgements. This research has made use of the WEBDA database (oper- (1997),(1997), we we see see that thats-process-dominated s-process-dominated elements, elements, and and especially espe- sity of Montpellier II) and Guillermo Gonzalez (Washington State University) barium, have higher abundances for many open clusters. This wereated particularlyat the Department generous of Theoreticalin providing Physics us with and atomic Astrophysics data for CN of the and Masaryk C2, re- cially barium, have higher abundances for many open clusters. spectively.University, We Brno), thank of Laura SIMBAD Magrini (operated for sharing at with CDS, us Strasbourg), the Galactic of chemical VALD Thisphenomenon phenomenon can be can explained be explained by the by young the ageyoung of these age of clusters these evolution(Kupka et models. al. 2000 We), and also of thank NASA’s the referee Astrophysics for helpful Data suggestions System. Bertrand and com- Plez clustersand a larger and a contribution larger contribution of low-mass of low-mass asymptotic asymptotic giant branch giant ments(University that improved of Montpellier the quality II) of and this Guillermo paper. VB, Gonzalez AD, GT, (WashingtonYC were partially State branchstars in stars producing in producings-process s-process elements elements at the time at the these time clusters these supportedUniversity) by were the grant particularly from the generous Research in Council providing of Lithuania us with atomic (MIP-082 data/2015). for CN were formed (c.f. D’Orazi et al. 2009, Maiorca et al. 2011). At the RSand acknowledges C2, respectively. support We bythank the Laura National Magrini Science for Center sharing of with Poland us throughthe Galactic the clusters were formed (c.f. D’Orazi et al. 2009, Maiorca et al. grantchemical 2012 evolution/07/B/ST9 models./04428. We also thank the referee for helpful suggestions and (2011)).same time, At the the samer-process-dominated time, the r-process-dominated element europium element results eu- comments that improved the quality of this paper. VB, AD, GT, YC were partially ropiumagree well results with agree the thin well disc with model. the thin disc model. supported by the grant from the Research Council of Lithuania (MIP-082/2015). In Fig.8 we can see how abundances of n-capture element RS acknowledges support by the National Science Center of Poland through the In Fig. 8 we can see how abundances of n-capture element grant 2012/07/B/ST9/04428. abundancesabundances in in open open clusters clusters depend depend on on cluster cluster ages. ages. IC IC 4756 4756 is is References a relatively young open cluster with an age of around 0.8 Gyr. Abia, C., Palmerini, S., Busso, M., & Cristallo, S. 2012, A&A, 548, A55

ArticleA165, page number, 12 of page 13 10 of 14 V. Bagdonas: Chemical composition of giant stars in the open cluster IC 4756

References Maiorca, E., Magrini, L., Busso, M., et al. 2012, ApJ, 747, 53 Masseron, T., & Gilmore, G. 2015, MNRAS, 453, 1855 Abia, C., Palmerini, S., Busso, M., & Cristallo, S. 2012, A&A, 548, A55 Matteucci, F. 1992, Mem. Soc. Astron. It., 63, 301 Alcaino, G. 1965, Lowell Observatory Bulletin, 6, 167 Matteucci, F., & Greggio, L. 1986, A&A, 154, 279 Angelou, G. C., Stancliffe, R. J., Church, R. P., Lattanzio, J. C., & Smith, G. H. McWilliam, A. 1998, AJ, 115, 1640 2012, ApJ, 749, 128 Mermilliod, J.-C., Andersen, J., Latham, D. W., & Mayor, M. 2007, A&A, 473, Arlandini, C., Käppeler, F., Wisshak, K., et al. 1999, ApJ, 525, 886 829 Baumgardt, H., Dettbarn, C., & Wielen, R. 2000, A&AS, 146, 251 Mermilliod, J. C., Mayor, M., & Udry, S. 2008, A&A, 485, 303 Bisterzo, S., Travaglio, C., Wiescher, M., et al. 2016, J. Phys. Conf. Ser., 665, Mikolaitis, Š., Hill, V., Recio-Blanco, A., et al. 2014, A&A, 572, A33 012023 Mikolaitis, Š., Tautvaišiene,˙ G., Gratton, R., Bragaglia, A., & Carretta, E. 2010, Bressan, A., Marigo, P., Girardi, L., et al. 2012, MNRAS, 427, 127 MNRAS, 407, 1866 Cantiello, M., & Langer, N. 2010, A&A, 521, A9 Mikolaitis, Š., Tautvaišiene,˙ G., Gratton, R., Bragaglia, A., & Carretta, E. 2011a, Cardon, B. L., Smith, P. L., Scalo, J. M., Testerman, L., & Whaling, W. 1982, MNRAS, 413, 2199 ApJ, 260, 395 Mikolaitis, Š., Tautvaišiene,˙ G., Gratton, R., Bragaglia, A., & Carretta, E. 2011b, Casamiquela, L., Carrera, R., Jordi, C., et al. 2016, MNRAS, 458, 3150 MNRAS, 416, 1092 Chanamé, J., Pinsonneault, M., & Terndrup, D. M. 2005, ApJ, 631, 540 Mikolaitis, Š., Tautvaišiene,˙ G., Gratton, R., Bragaglia, A., & Carretta, E. 2012, Charbonnel, C. 2006, in EAS Pub. Ser., 19, 125 A&A, 541, A137 Charbonnel, C., & Lagarde, N. 2010, A&A, 522, A10 Mishenina, T., Korotin, S., Carraro, G., Kovtyukh, V. V., & Yegorova, I. A. 2013, Day, R. W., Lambert, D. L., & Sneden, C. 1973, ApJ, 185, 213 MNRAS, 433, 1436 De Silva, G. M., Freeman, K. C., Bland-Hawthorn, J., et al. 2015, MNRAS, 449, Mishenina, T., Pignatari, M., Carraro, G., et al. 2015, MNRAS, 446, 3651 2604 Nidever, D. L., Zasowski, G., Majewski, S. R., et al. 2012, ApJ, 755, L25 Dearborn, D. S. P., Eggleton, P. P., & Schramm, D. N. 1976, ApJ, 203, 455 Nitz, D. E., Kunau, A. E., Wilson, K. L., & Lentz, L. R. 1999, ApJS, 122, 557 Den Hartog, E. A., Lawler, J. E., Sneden, C., & Cowan, J. J. 2003, ApJS, 148, Overbeek, J. C., Friel, E. D., & Jacobson, H. R. 2016, ApJ, 824, 75 543 Pace, G., Danziger, J., Carraro, G., et al. 2010, A&A, 515, A28 Den Hartog, E. A., Lawler, J. E., Sobeck, J. S., Sneden, C., & Cowan, J. J. 2011, Pagel, B. E. J., & Tautvaisiene, G. 1995, MNRAS, 276, 505 ApJS, 194, 35 Pagel, B. E. J., & Tautvaisiene, G. 1997, MNRAS, 288, 108 Denissenkov, P. A. 2010, ApJ, 723, 563 Piskunov, N. E., Kupka, F., Ryabchikova, T. A., Weiss, W. W., & Jeffery, C. S. Dias, W. S., Monteiro, H., Caetano, T. C., et al. 2014, A&A, 564, A79 1995, A&AS, 112, 525 D’Orazi, V., Magrini, L., Randich, S., et al. 2009, ApJ, 693, L31 Ramírez, I., & Allende Prieto C. 2011, ApJ, 743, 135 D’Orazi, V., Biazzo, K., Desidera, S., et al. 2012, MNRAS, 423, 2789 Reddy, A. B. S., & Lambert, D. L. 2015, MNRAS, 454, 1976 D’Orazi, V., De Silva, G. M., & Melo, C. F. H. 2017, A&A, 598, A86 Reddy, A. B. S., & Lambert, D. L. 2017, ApJ, 845, 151 Drazdauskas, A., Tautvaišiene,˙ G., Randich, S., et al. 2016a, A&A, 589, A50 Reddy, A. B. S., Giridhar, S., & Lambert, D. L. 2012, MNRAS, 419, 1350 Drazdauskas, A., Tautvaišiene,˙ G., Smiljanic, R., Bagdonas, V., & Chorniy, Y. Reddy, A. B. S., Giridhar, S., & Lambert, D. L. 2013, MNRAS, 431, 3338 2016b, MNRAS, 462, 794 Reddy, A. B. S., Giridhar, S., & Lambert, D. L. 2015, MNRAS, 450, 4301 Friel, E. D., Jacobson, H. R., & Pilachowski, C. A. 2010, AJ, 139, 1942 Salaris, M., Weiss, A., & Percival, S. M. 2004, A&A, 414, 163 Frinchaboy, P. M., & Majewski, S. R. 2008, AJ, 136, 118 Sampedro, L., Dias, W. S., Alfaro, E. J., Monteiro, H., & Molino, A. 2017, Gaia Collaboration (Prusti, T., et al.) 2016, A&A, 595, A1 MNRAS, 470, 3937 Gilmore, G., Randich, S., Asplund, M., et al. 2012, The Messenger, 147, 25 Santos, N. C., Lovis, C., Pace, G., Melendez, J., & Naef, D. 2009, A&A, 493, Gilroy, K. K. 1989, ApJ, 347, 835 309 Gonzalez, G., Lambert, D. L., Wallerstein, G., et al. 1998, ApJS, 114, 133 Santrich, O. J. K., Pereira, C. B., & Drake, N. A. 2013, A&A, 554, A2 Grevesse, N., & Sauval, A. J. 2000, in Origin of Elements in the Solar System, Seggewiss, W. 1968, Mitteilungen der Astronomischen Gesellschaft Hamburg, Implications of Post-1957 Observations, ed. O. Manuel, 261 25, 172 Gurtovenko, E. A., & Kostyk, R. I. 1989, Fraunhofers Spectrum and a System of Smiljanic, R., Gauderon, R., North, P., et al. 2009, A&A, 502, 267 Solar Oscillator Strengths (Kiev: Nauk. dumka), 200 Smiljanic, R., Romano, D., Bragaglia, A., et al. 2016, A&A, 589, A115 Gustafsson, B., Edvardsson, B., Eriksson, K., et al. 2008, A&A, 486, 951 Smith, G. H. 1983, PASP, 95, 296 Hekker, S., & Meléndez, J. 2007, A&A, 475, 1003 Sneden, C., Lawler, J. E., Cowan, J. J., Ivans, I. I., & Den Hartog E. A. 2009, Herzog, A. D., Sanders, W. L., & Seggewiss, W. 1975, A&AS, 19, 211 ApJS, 182, 80 Hinkle, K., Wallace, L., Valenti, J., & Harmer, D. 2000, Visible and Near Infrared Sousa, S. G., Santos, N. C., Mayor, M., et al. 2008, A&A, 487, 373 Atlas of the Arcturus Spectrum 3727–9300 Å (San Francisco: ASP) Strassmeier, K. G., Weingrill, J., Granzer, T., et al. 2015, A&A, 580, A66 Iben, Jr. I. 1965, ApJ, 142, 1447 Tautvaišiene,˙ G., Edvardsson, B., Tuominen, I., & Ilyin, I. 2000, A&A, 360, Ivans, I. I., Simmerer, J., Sneden, C., et al. 2006, ApJ, 645, 613 499 Jacobson, H. R., & Friel, E. D. 2013, AJ, 145, 107 Tautvaišiene,˙ G., Edvardsson, B., Puzeras, E., & Ilyin, I. 2005, A&A, 431, Jacobson, H. R., Friel, E. D., & Pilachowski, C. A. 2007, AJ, 134, 1216 933 Jacobson, H. R., Friel, E. D., & Pilachowski, C. A. 2008, AJ, 135, 2341 Tautvaišiene,˙ G., Drazdauskas, A., Mikolaitis, Š, et al. 2015, A&A, 573, A55 Jacobson, H. R., Friel, E. D., & Pilachowski, C. A. 2009, AJ, 137, 4753 Tautvaišiene,˙ G., Drazdauskas, A., Bragaglia, A., Randich, S., & Ženoviene˙ R. Jofré, P., Heiter, U., Soubiran, C., et al. 2015, A&A, 582, A81 2016, A&A, 595, A16 Johansson, S., Litzén, U., Lundberg, H., & Zhang, Z. 2003, ApJ, 584, L107 Thogersen, E. N., Friel, E. D., & Fallon, B. V. 1993, PASP, 105, 1253 Käppeler, F., Gallino, R., Bisterzo, S., & Aoki, W. 2011, Rev. Mod. Phys., 83, Ting, Y.-S., De Silva, G. M., Freeman, K. C., & Parker, S. J. 2012, MNRAS, 427, 157 882 Kaufer, A., Stahl, O., Tubbesing, S., et al. 1999, The Messenger, 95, 8 Tinsley, B. M. 1979, ApJ, 229, 1046 Kopff, E. 1943, Astron. Nachr., 274, 69 Tomkin, J., & Lambert, D. L. 1974, ApJ, 193, 631 Kunder, A., Kordopatis, G., Steinmetz, M., et al. 2017, AJ, 153, 75 Tomkin, J., Lambert, D. L., & Luck, R. E. 1975, ApJ, 199, 436 Kupka, F. G., Ryabchikova, T. A., Piskunov, N. E., Stempels, H. C., & Weiss, Travaglio, C., Gallino, R., Arnone, E., et al. 2004, ApJ, 601, 864 W. W. 2000, Baltic Astron., 9, 590 Twarog, B. A., Ashman, K. M., & Anthony-Twarog, B. J. 1997, AJ, 114, 2556 Kurucz, R. L. 2005, Mem. Soc. Astron. It. Supp., 8, 189 Škoda, P., Draper, P. W., Neves, M. C., Andrešic,ˇ D., & Jenness, T. 2014, Astron. Lagarde, N., Charbonnel, C., Decressin, T., & Hagelberg, J. 2011, A&A, 536, Comput., 7, 108 A28 Van der Swaelmen, M., Bofﬁn, H. M. J., Jorissen, A., & Van Eck S. 2017, A&A, Lagarde, N., Decressin, T., Charbonnel, C., et al. 2012, A&A, 543, A108 597, A68 Lagarde, N., Anderson, R. I., Charbonnel, C., et al. 2014, A&A, 570, C2 Wachlin, F. C., Miller Bertolami, M. M., & Althaus, L. G. 2011, A&A, 533, Lattanzio, J. C., Siess, L., Church, R. P., et al. 2015, MNRAS, 446, 2673 A139 Lawler, J. E., Wickliffe, M. E., den Hartog, E. A., & Sneden, C. 2001, ApJ, 563, Wachlin, F. C., Vauclair, S., & Althaus, L. G. 2014, A&A, 570, A58 1075 Worley, C. C., Cottrell, P. L., Freeman, K. C., & Wylie-de Boer E. C. 2009, Lind, K., Asplund, M., Barklem, P. S., & Belyaev, A. K. 2011, A&A, 528, A103 MNRAS, 400, 1039 Luck, R. E. 1994, ApJS, 91, 309 Wyse, R. F. G., & Gilmore, G. 1988, AJ, 95, 1404 MacLean, B. T., De Silva, G. M., & Lattanzio, J. 2015, MNRAS, 446, 3556 Yong, D., Carney, B. W., & Teixera de Almeida, M. L. 2005, AJ, 130, Magrini, L., Sestito, P., Randich, S., & Galli, D. 2009, A&A, 494, 95 597 Maiorca, E., Randich, S., Busso, M., Magrini, L., & Palmerini, S. 2011, ApJ, Yong, D., Carney, B. W., & Friel, E. D. 2012, AJ, 144, 95 736, 120 Zacharias, N., Finch, C. T., Girard, T. M., et al. 2013, AJ, 145, 44

A165, page 13 of 13