APOGEE discovery of a chemically atypical star disrupted from NGC 6723 and captured by the Milky Way bulge
José G. Fernández-Trincado, Timothy C. Beers, Dante Minniti, Leticia Carigi, Vinicius M. Placco, Sang-Hyun Chun, Richard R. Lane, Doug Geisler, Sandro Villanova, Stefano O. Souza, Beatriz Barbuy, Angeles Pérez-Villegas, Cristina Chiappini, Anna. B. A. Queiroz, Baitian Tang, Javier Alonso-García, Andrés E. Piatti, Tali Palma, Alan Alves-Brito, Christian Moni Bidin, Alexandre Roman-Lopes, Ricardo R. Muñoz, Harinder P. Singh, Richa Kundu, Leonardo Chaves-Velasquez, María Romero-Colmenares, Penelope Longa-Peña, Mario Soto, Katherine Vieira
AAstronomy & Astrophysics manuscript no. NGC6723 © ESO 2021February 4, 2021
APOGEE discovery of a chemically atypical star disrupted fromNGC 6723 and captured by the Milky Way bulge
José G. Fernández-Trincado , (cid:63) , Timothy C. Beers , Dante Minniti , , Leticia Carigi , Vinicius M. Placco ,Sang-Hyun Chun , Richard R. Lane , Doug Geisler , , , Sandro Villanova , Stefano O. Souza , BeatrizBarbuy , Angeles Pérez-Villegas , Cristina Chiappini , , Anna. B. A. Queiroz , Baitian Tang , JavierAlonso-García , , Andrés E. Piatti , , Tali Palma , , Alan Alves-Brito , Christian Moni Bidin , AlexandreRoman-Lopes , Ricardo R. Muñoz , , Harinder P. Singh , Richa Kundu , , Leonardo Chaves-Velasquez , ,María Romero-Colmenares , Penelope Longa-Peña , Mario Soto and Katherine Vieira (A ffi liations can be found after the references) Received: XX / XX / XXXX; Accepted: XX / XX / XXXX
ABSTRACT
The central (‘bulge’) region of the Milky Way is teeming with a significant fraction of mildly metal-deficient stars with atmospheres that arestrongly enriched in cyanogen ( C N). Some of these objects, which are also known as nitrogen-enhanced stars, are hypothesised to be relicsof the ancient assembly history of the Milky Way. Although the chemical similarity of nitrogen-enhanced stars to the unique chemical patternsobserved in globular clusters has been observed, a direct connection between field stars and globular clusters has not yet been proven. In this work,we report on high-resolution, near-infrared spectroscopic observations of the bulge globular cluster NGC 6723, and the serendipitous discoveryof a star, 2M18594405 − / Fe] (cid:38) + .
94) is wellabove the typical Galactic field-star levels, and it exhibits noticeable enrichment in the heavy s -process elements (Ce, Nd, and Yb), along withmoderate carbon enrichment; all characteristics are known examples in globular clusters. This result suggests that some of the nitrogen-enhancedstars in the bulge likely originated from the tidal disruption of globular clusters. Key words. stars: abundances – stars: chemically peculiar – Galaxy: globular clusters: NGC 6723 – techniques: spectroscopic
1. Introduction
The great majority of stars in the halo of the Milky Way (MW)have elemental-abundance ratios, spanning a wide range ofmetallicities ( − . < [Fe / H] < .
0) that are similar to one an-other and overall track the general level of metallicity. However,there are a number of important exceptions, including carbon-enhanced ([C / Fe] (cid:38) + .
7) metal-poor (CEMP) stars, in par-ticular at very low metallicities ([Fe / H] < − . − . < [Fe / H] < − . / Fe] (cid:46) + . − < [Fe / H] < + .
1) with light-element abundances (e.g. N, Al, and Si) whose chemical com-positions mimic the abundance patterns observed in some of thestars in Galactic (and extragalactic) globular cluster (GC) pop-ulations (e.g. Fernández-Trincado et al. 2017; Schiavon et al.2017b; Bekki 2019; Hanke et al. 2020; Mészáros et al. 2020;Fernández-Trincado et al. 2020b).Multiple potentially pieces of evidence support the originof CEMP stars in low-mass galaxies that have been accretedby the MW (Yoon et al. 2019), either due to mass-transfer bi-naries (CEMP- s stars, Beers & Christlieb 2005) or from natalgas polluted by high-mass stars in the early Galaxy (CEMP-no stars, Beers & Christlieb 2005). However, the mildly metal-poor giants with stellar atmospheres that are strongly enriched in (cid:63) To whom correspondence should be addressed; E-mail:[email protected] and / or [email protected] C N (in cases with available data, with distinctive Al and Siabundances as well), are referred to as nitrogen-enhanced (N-rich) stars (Fernández-Trincado et al. 2016a, 2017; Schiavonet al. 2017b; Fernández-Trincado et al. 2019b) or NRS (Bekki2019), and defined in Johnson et al. (2007) as nitrogen-enrichedmetal-poor (NEMP) stars, which are CEMP stars with [C / N] < − . / Fe] > + .
5. They have resisted a unifying explana-tion of the nucleosynthetic processes responsible for their abun-dance anomalies. These stars share chemical-abundance pat-terns that are hypothesised to be associated with the so-calledsecond-generation stars in GCs (see e.g. Schiavon et al. 2017a;Fernández-Trincado et al. 2019d, 2020e; Mészáros et al. 2020),with only a handful of exceptions, such as the binary hypoth-esis (see e.g. Fernández-Trincado et al. 2019c). There are alsoa handful of known early asymptotic giant branch (early-AGB)stars in GCs (see, e.g. Mészáros et al. 2020), which exhibit a ni-trogen enrichment similar to that of the N-rich stars, but with amodest enrichment in carbon ([C / Fe] > + . ∼ Article number, page 1 of 16 a r X i v : . [ a s t r o - ph . GA ] F e b & A proofs: manuscript no. NGC6723
Fig. 1.
Properties of the extra-tidal star compared with likely members of NGC 6723. Panel (a): The metallicity – radial velocity plane for starsin the APOGEE-2 catalogue in the region of NGC 6723 (green triangles). The black open circles are potential members of NGC 6723; sizesreflect their V o magnitudes, with decreasing size for fainter magnitudes. The pink line and shaded region indicate the (cid:104) [Fe / H] (cid:105) ± σ [Fe / H] of theextra-tidal star. The black dotted lines indicate the nominal [Fe / H] and radial velocity of NGC 6723 from Harris (1996, 2010 edition). The redunfilled diamonds refer to probable members analysed in the literature (Crestani et al. 2019). The blue rectangle indicates the region adopted tochoose potential cluster members, based on radial velocity and metallicity (see Section 2). Panel (b): Positions for stars in the APOGEE-2 and
Gaia
EDR3 (grey dots) catalogues in the region of NGC 6723. The magenta cross symbol indicates the
Gaia astrometric uncertainty. The big redand shadow circumference show the cluster tidal radius ( r t = . ± . (cid:48) ) of NGC 6723 determined in this work from HST + Gaia
EDR3 data set(see Section 8). The orbital paths of the cluster (black line) and the extra-tidal star (magenta line) are shown, along with the proper motion vectors(black arrow) of the cluster from Baumgardt et al. (2019), and the extra-tidal star (magenta arrow); their lengths (scaled up for visibility) anddirections are essentially identical. Panel (c): Proper-motion plane for the selected candidates. Panels (d) and (e): V o versus (B-V) o and 2MASSH o versus (J-K s ) o Colour-Magnitude Diagram (CMD) centred on NGC 6723, showing all the stars within the cluster tidal radius, and the probablecluster members (cyan symbols) within a 2 σ deviation from the best isochrone fitting (Souza et al. 2020; Oliveira et al. 2020). The largest opencircle in all panels indicate the extra-tidal star. these stars play a crucial role in reconstructing the formation andevolutionary history of the MW.NGC 6723 is an old ( (cid:38) × M (cid:12) : Baumgardt et al. 2019) GCsituated in the bulge region ( d (cid:12) ∼ (cid:46) .
85 kpc from the Galactic Centre (Baum-gardt et al. 2019). It is currently located at a position where theforeground interstellar reddening is low, E(B-V) ∼ .
063 (Leeet al. 2014), making it an excellent target to identify and studythe family of N-rich stars in the bulge region of the MW, in orderto probe for a linkage between GC environments and field starswith nitrogen over-abundances.The Apache Point Observatory Galactic Evolution Experi-ment (APOGEE-2, Majewski et al. 2017) is a cornerstone mis-sion of the Sloan Digital Sky Survey IV (SDSS-IV, Blanton et al. 2017), which has been designed primarily to investigate thechemical history of the MW. APOGEE and APOGEE-2 have de-livered an exquisite data set by combining high-resolution near-IR spectroscopic observations from the Sloan 2.5m telescope atApache Point Observatory (APO) and the Irénée du Pont 2.5mtelescope at Las Campanas Observatory (LCO): the largest andmost precise census of more than 24 chemical species for hun-dreds of thousands of stars throughout the MW. These surveyshave revealed numerous giant stars with light-element abun-dances that deviate from the general patterns, and are at oddswith current chemical-evolution models. Their origins have re-mained obscure, due to limitations on the numbers of such starsand the precision of the previously available data.In this work, we take advantage of the APOGEE-2 data setto alleviate the lack of data towards the Galactic Bulge. We re-
Article number, page 2 of 16osé G. Fernández-Trincado et al.: A chemically atypical star being tidally disrupted from NGC 6723 port on the discovery of an extra-tidal star candidate in the vicin-ity of a GC (NGC 6723) toward the bulge region, with nitrogenover-abundances well above that of normal Galactic field stars,and with a modest carbon enrichment. The other properties ofthis star, including metallicity and kinematics, strongly suggestit was once a cluster member. This is the first demonstration of adirect link between a mildly metal-poor N-rich field star with amodest carbon enrichment and a parent GC in the bulge regionof the MW.
2. Data and sample selection
We present a spectroscopic study of the bulge GC NGC 6723,and its surrounding regions, based on high-resolution ( R ∼ H -band ( λ ∼ and , which contain reliable spectral infor-mation for 526 stars. Probable cluster members were selectedbased on the nominal radial velocity (RV) of the cluster (witha velocity di ff erence no larger than 15 km s − ), and metallicitywithin 0.25 dex from the value reported in Harris (1996, 2010edition). The initial search was made within the blue box high-lighted in panel (a) of Figure 1, and limited to spectra with aSNR >
60 pixel − .We decided to adopt the uncalibrated ASPCAP [M / H] scale,which tracks all metals relative to the Sun (listed in Table 1), as afirst guess to the stellar metallicity. This gives the overall metal-licity of the stars, as it is derived by fitting the entire wavelengthregion covered by the APOGEE-2 spectrographs. Finally, we useFe I lines to measure the [Fe / H].The final sample contains eight potential members ofNGC 6723. We find that seven of these stars are located insidethe tidal radius, as shown in panel (b) of Figure 1. There is onestar, 2M18594405 − o versus (B-V) o Colour-Magnitude Di-agrams (CMDs), as displayed in panels (d) and (e) of Figure 1,and have
Gaia proper motions similar to the nominal value forthe cluster.Following the same methodology and techniques de-scribed in several APOGEE papers (Fernández-Trincado et al.2016a; Hawkins et al. 2016; Fernández-Trincado et al. 2017,2019a,c,b,d, 2020e), we employed the
BACCHUS software(Masseron et al. 2016) to manually analyze each star of oursample, in order to re-examine the reliability of each atomicand molecular line present in each spectrum, and provide chem-ical abundances for 14 chemical species, listed in Table 1.These abundances have been computed adopting a line-by-line approach under the assumption of local thermodynamic equi-librium (LTE). With this independent methodology, we obtaincomplementary abundance ratios not provided by the
ASPCAP pipeline (García Pérez et al. 2016), in particular for the heavyneutron-capture ( s -process) elements (Ce II, Nd II, and Yb II).Figure 2 illustrates the best spectral-synthesis calculation onclean selected features for the extra-tidal star.Importantly, none of the newly identified stars in this workhave strong RV variability over the temporal span of theAPOGEE-2 observations (visit-to-visit variations, σ RV < − ), which were obtained during 2018-05-24 (visit 1) and 2018-05-25 (visit 2). Therefore, with the current data there is no evi-dence that any of these objects are members of binary systems.Furthermore, we found that our sample has a Gaia re-normalised unit weight error (
RUWE ) value < Gaia
Early Data Release 3 (
Gaia
EDR3) cata-logue (Gaia Collaboration et al. 2020).Panels (a) to (c) of Figure 1 show the metallicity versus kine-matic, sky positions, and astrometric properties of our sample,as well as the serendipitous discovery of an chemically atypicalextra-tidal star in the vicinity of NGC 6723; a result not previ-ously seen. We find that 5 of the 7 cluster members are stronglyenhanced in nitrogen, [N / Fe] > + .
86, well above the Galacticlevels (which is typically (cid:46) + .
3. Atmospheric parameters and elementalabundances
We used the
BACCHUS code (Masseron et al. 2016) to derivethe metallicity (from Fe I lines), broadening parameters, andchemical abundances for the stars in our sample, based on care-ful line selection, and carried out a detailed inspection of eachAPOGEE-2 spectrum in order to examine the reliability of theabundance ratios, in the same manner as described in Fernández-Trincado et al. (2019b). The adopted atmospheric parametersand the typical uncertainties are listed in Table 1.The
BACCHUS code relies on the radiative transfer code
Turbospectrum (Alvarez & Plez 1998; Plez 2012) and the
MARCS model atmosphere grid (Gustafsson et al. 2008). For eachelement and each line, the abundance determination proceeds asin previous APOGEE-2 works (Hawkins et al. 2016). In sum-mary, the steps are: ( i ) a spectrum synthesis, using the full set of(atomic and molecular) lines to find the local continuum level viaa linear fit; ( ii ) cosmic and telluric line rejections are performed;( iii ) the local signal-to-noise ratio (S / N) per element is estimated;( iv ) a series of flux points contributing to a given absorption lineare automatically selected; and ( v ) abundances are then derivedby comparing the observed spectrum with a set of convolvedsynthetic spectra characterised by di ff erent abundances.Four di ff erent abundance determinations are used: ( i ) line-profile fitting; ( ii ) core line-intensity comparison; ( iii ) globalgoodness-of-fit estimate; and ( iv ) equivalent-width comparison.Each diagnostic yields validation flags. Based on these flags, adecision tree then rejects or accepts the line, keeping the best-fit abundance. We adopted the χ diagnostic as the abundancedeterminant, because it is considered to be the most robust.However, we stored the information from the other diagnos- Article number, page 3 of 16 & A proofs: manuscript no. NGC6723 .
516 1 . µm )0 . . N o r m a li ze d F l u x K IS/N=184A(K) = 4.373 . . . µm )0 . . N o r m a li ze d F l u x Nd IIS/N=347A(Nd) = 1.566 .
000 0 . µm ) +1 . . . N o r m a li ze d F l u x C NS/N=434A(N) = 7.570 . . . µm )0 . . N o r m a li ze d F l u x Ti IS/N=142A(Ti) = 4.521 .
555 1 . µm )0 . . N o r m a li ze d F l u x Si IS/N=268A(Si) = 6.941 .
576 1 .
577 1 . µm )0 . . N o r m a li ze d F l u x Mg IS/N=200A(Mg) = 6.498 .
595 1 .
596 1 . µm )0 . . N o r m a li ze d F l u x Fe IS/N=151Al(Fe) = 6.346 .
619 1 . µm )0 . . N o r m a li ze d F l u x Ca IS/N=289A(Ca) = 5.297 .
637 1 . µm )0 . . N o r m a li ze d F l u x Ce IIS/N=129A(Ce) = 1.514 .
577 1 . µm )0 . . N o r m a li ze d F l u x C OS/N=52A(C) = 7.732 . . . . µm )0 . . N o r m a li ze d F l u x OHS/N=134A(O) = 8.047 . . . µm )0 . . N o r m a li ze d F l u x Al IS/N=254A(Al) = 5.402 .
000 0 . µm ) +1 . . . N o r m a li ze d F l u x Ni IS/N=129A(Ni) = 5.085 .
649 1 . µm )0 . . N o r m a li ze d F l u x Yb IIS/N=222A(Yb) = 0.97
Fig. 2.
High-resolution near-IR H -band spectrum of the extra-tidal star. The spectral regions are shown with orange squares. Superimposed is thebest-fit of a MARCS / BACCHUS spectral synthesis (black line). The light blue shaded regions show the strength of the molecules ( C O, C N, OH), and the atomic lines, namely the α -elements (Mg I, Si I, Ca I, Ti I), the odd-Z elements (Al I, K I), the iron-peak elements (Fe I, Ni I), andthe s -process elements (Ce II, ND II, Yb II), expressed in air wavelengths. The legends in each panel show the absolute abundance, A (X), of thespecies under consideration, and the signal-to-noise (S / N) in the regions of the features, respectively. tics, including the standard deviation between all four meth-ods. The line list used in this work is the latest internal DR14atomic / molecular linelist ( linelist.20170418 ), including the s -process elements (Ce II, Nd II, and Yb II)–(Hasselquist et al.2016; Cunha et al. 2017). For a more detailed description ofthese lines, we refer the reader to a forthcoming paper (Holtz-man et al. in preparation). In particular, a mix of heavily CN-cycled and α -poor MARCS models were used, as well as the same molecular lines adoptedby APOGEE-2 (Smith et al. 2013), in order to determine the C,N, and O abundances. In addition, we have adopted the C, N,and O abundances that satisfy the fitting of all molecular linesconsistently; that is to say, we first derive O abundances from OH lines, then derive C from C O lines, and N from
Article number, page 4 of 16osé G. Fernández-Trincado et al.: A chemically atypical star being tidally disrupted from NGC 6723 C N O M g A l S i K C a T i N i C e N d Chemical Species21012 [ X / F e ] a JC+2019 ( [Fe/H] 0.90)This work ( [Fe/H] 0.99)extra-tidal star ([Fe/H] 1.17) C N O M g A l S i K C a T i N i C e N d Chemical Species21012 [ X / F e ] d RGB GCs ( [Fe/H] = 1.07±0.07)extra-tidal star ([Fe/H] 1.17) C N O M g A l S i K C a T i N i C e N d Chemical Species21012 [ X / F e ] b M107 ([Fe/H] = 1.02)This work ( [Fe/H] 0.99)extra-tidal star ([Fe/H] 1.17) C N O M g A l S i K C a T i N i C e N d Chemical Species21012 [ X / F e ] e eAGB GCs ( [Fe/H] = 1.11±0.06)extra-tidal star ([Fe/H] 1.17) C N O M g A l S i K C a T i N i C e N d Chemical Species21012 [ X / F e ] c NGC 362 ([Fe/H] = 1.26)This work ( [Fe/H] 0.99)extra-tidal star ([Fe/H] 1.17) C N O M g A l S i K C a T i N i C e N d Chemical Species21012 [ X / F e ] f IRAS07134/post-AGB ([Fe/H] = 0.86)extra-tidal star ([Fe/H] 1.17)
Fig. 3.
Comparison of elemental-abundance ratios of likely member stars for NGC 6723 with other GCs of similar metallicity, and the extra-tidalstar with RGB and early-AGB (eAGB) cluster stars, and IRAS07134 star. Panel (a): Elemental-abundance patterns for our sample (black boxesand red dot) compared to the spectroscopic study of NGC 6723 (Crestani et al. 2019) (light blue boxplots). Panel (b): Comparison with M 107(Mészáros et al. 2020). Panel (c): Comparison with NGC 362 (Mészáros et al. 2020). Panel (d): Elemental-abundance patterns for the extra-tidalstar compared to those of RGB GC stars. Panel (e): Compared with early-AGB GC stars (Mészáros et al. 2020). Panel (f): Comparison with a fieldpost-AGB star, IRAS 07134 (De Smedt et al. 2016), with comparable atmospheric parameters. Boxes represent the inter-quartile ranges (IQR),whiskers the 1.5 × IQR limits, and blue lines the medians. C N lines; the C–N–O abundances were derived iterativelyto minimize the OH, C O, and C N dependences (Smithet al. 2013).It is important to perform consistent chemical-abundanceanalyses using atmospheric parameters determined indepen-dently, in order to check for any significant deviation in thederived abundances. To achieve this, the photometric e ff ectivetemperatures were calculated using the J − K s (2MASS) colour-relation methodology (González Hernández & Bonifacio 2009).For the extra-tidal star, we adopted the extinction correction ob-tained from the Rayleigh Jeans Colour Excess (RJCE) method(Majewski et al. 2011). Thus, the photometry is extinction cor-rected, adopting E(B-V) = = ff ected by extinction,given their position toward the Galactic bulge where the redden- ing variation may be substantial (see, e.g. Alonso-García et al.2017), even if the projected distance between the extra-tidal starand cluster is small.We assumed a surface gravity from the PARSEC isochrones(Bressan et al. 2012), using ≈ / H]), as derived by
ASPCAP / APOGEE-2 runs. The adopted stellar parameters are listed in Table 1.The adoption of a purely photometric temperature scale en-ables us to be somewhat independent of the
ASPCAP / APOGEE-2pipeline, which provides important comparison data for futurepipeline validation. The final results presented in this paper arebased on computations done with the
BACCHUS code using thementioned photometry and atmospheric parameters, as listed inTable 1.The abundance values are sensitive to all of the atmosphericparameters, depending on the chemical species. To estimate their
Article number, page 5 of 16 & A proofs: manuscript no. NGC6723 − [ A l / F e ] . . . . . . − [ A l / F e ] . . . . . − [ A l / F e ] . . . . . . . . . . . [ S i / F e ] . . . . . . . . . . . [ K / F e ] . . . . . . [ N / F e ] . . . . . . − [ A l / F e ] . . . . − . . . [ N i / F e ] . . . . . . . . [ T i / F e ] . . . . . Fig. 4.
Distributions of various elemental-abundance ratios from the APOGEE-2 survey for bulge field stars in close proximity to the GalacticCentre ( − ◦ < l < ◦ and | b | < ◦ ), and comparison with stars in NGC 6723. The field-star abundances are represented by orange iso-abundancecontours. The member stars of NGC 6723 and the extra-tidal star analyzed in this work are shown with black and red open circles, respectively.The circle sizes reflect their V o magnitudes, with decreasing size for fainter magnitudes. The blue cross symbol denotes the average (cid:104) [X i / Fe] (cid:105) andits associated star-to-star scatter ( σ (cid:104) [X i / Fe] (cid:105) ) of the abundance derived from the seven member NGC 6723 stars analyzed in this work. uncertainties, we have varied the atmospheric parameters one ata time by the typical values of ∆ T e ff = ±
100 K, ∆ log g = ± . ∆ ξ t = ± .
05 km s − , and then computed the abundancesfor all species for each of these possibilities for two stars – onecluster member (2M18594898 − − σ T e ff , which refers to its response to ∆ T e ff ; σ log g theresponse to ∆ log g ; σ ξ t the response to ∆ ξ t , and the uncertain-ties on the mean due to line-to-line scatter. The uncertaintiesare propagated in quadrature to compute the uncertainty of eachchemical species in the two stars. The computed uncertaintiesare listed in Table 2.Even though the BACCHUS code has its own procedure to in-clude or reject lines on a star-by-star basis, it is still importantto select the lines beforehand, due to the uncertainty related tothe synthesis approach, such as line saturation. All the selectedatomic and molecular lines were visually inspected to ensure thatthe spectral fit was adequate.The individual oxygen abundances are also listed in Table 1.We caution about the accuracy of [O / Fe], as they display largerscatter, which may be due to telluric features and uncertain de- terminations in the T e ff regime of our objects. As highlightedby previous works, the uncertainty arises because BACCHUS de-termines these abundances from the strengths of C N and C O lines, which become too weak for stars at relatively lowmetallicities ([Fe / H] (cid:46) − .
4. Results
For the selected stars, we derive abundance ratios covering themain element families, namely the light elements (C, N), the α -elements (O, Mg, Si, Ca, Ti), the odd-Z elements (Al, K), theiron-peak elements (Fe, Ni), and products of the slow neutron-capture ( s -) process (Ce, Nd, Yb), and have abundance deter-minations from spectra in the H -band of APOGEE-2 (Mészáros Article number, page 6 of 16osé G. Fernández-Trincado et al.: A chemically atypical star being tidally disrupted from NGC 6723
Fig. 5.
Predicted abundance ratios, as a function of the initial stellar mass, for low-and-intermediate-mass stars. [X i / Fe] values were obtainedfrom stellar yields by Karakas et al. (2014, 2018). For comparison, the red line and shaded red bands represent the observed abundance ratios andsensitivities for our extra-tidal star. et al. 2020). For the cluster members, we find metallicity consis-tent with previous estimates for this cluster, (cid:104) [Fe / H] (cid:105) = − . / H] behaviour.We obtained a mean nitrogen abundance ratio of [N / Fe] = + .
95, with an observed star-to-star scatter (0 .
34 dex) thatwell exceeds the observational uncertainties. The abundance ra-tios for the other elements up to the iron peak (O, Mg, Al, Si,K, Ca, Ti, and Ni) appear to be homogeneous within the permit-ted error variation, and replicate the chemical patterns observedfor Galactic GCs of comparable metallicity, as shown in pan-els (b) and (c) of Figure 3. The heavy s -process elements, suchas Ce and Nd, are lower ([Ce / Fe], [Nd / Fe] (cid:46) + .
29) comparedto M107, a GC with similar metallicity as NGC 6723, except-ing Yb, which we find to be higher in some NGC 6723 stars.It has been noted previously that s -process-element enhance-ments (cid:38) + . (cid:38) M (cid:12) ) GCs at all metallicities. There is thus strongevidence for an intrinsic spread in [N / Fe], including a clear C–Nanti-correlation. However, we find no clear Al–N correlation orAl–O anti-correlation (see Figure 4). In other words, NGC 6723exhibits MPs based on the N abundances, despite appearing tohave single populations in the Al abundances. This is similar toother Galactic GCs at this metallicity (Mészáros et al. 2020), forwhich it has been suggested that lower spreads in Al could be as-cribed to operation of a modest Mg-Al cycle. If the intra-clusterpolluters were in fact low-mass ( (cid:46) (cid:12) ) stars, we would expectlow Al production (Mészáros et al. 2020) in NGC 6723. The lowmean abundance and small Al spread ( (cid:104) [Al / Fe] (cid:105) = + . ± . / Fe] < − .
17, and display a similar C–N anti-correlation as M107, with (cid:104) [C / Fe] (cid:105) = − . ± .
16. Additionally, NGC 6723 exhibits a star-to-star [Mg / Fe] scatter with no significant [Al / Fe] spread. NoMg-Al anti-correlation is apparent, and the scatter is small. Inaddition, inspection of the [Si / Fe] ratio, as a function of [Al / Fe]or [Mg / Fe], for the few stars with reliable Si measurements, did
Article number, page 7 of 16 & A proofs: manuscript no. NGC6723 not reveal any clear trend. This suggests no net production ofthese elements, but rather is the likely result of the conversion ofMg into Al during the Mg-Al cycle (Mészáros et al. 2020).Figure 4 compares the C, N, O, Mg, Al, Si, Ca, Ti, Ni, andCe abundances of our sample with Galactic Bulge stars. Thebehaviour of those chemical species matches the mean [X i / Fe](with X i meaning the chemical species) of MW stars at similarmetallicity, while the [N / Fe] abundance ratios are clearly super-Solar. Our results for [Si / Fe] are in reasonable agreement withoptical observations (Rojas-Arriagada et al. 2016; Crestani et al.2019).NGC 6723 has a uniform and constant [Ca / Fe] abundanceratio ( (cid:104) [Ca / Fe] (cid:105) = + . ± . ff ected by the H-burning process (Mészáros et al. 2020), as Cais mostly produced by supernovae. For the odd-Z element K, wefind that NGC 6723 exhibits a weak K–Mg anti-correlation, with (cid:104) [K / Fe] (cid:105) = + . ± .
10, suggesting that this population mighthave formed from super-AGB ejecta (Ventura et al. 2012).For the remaining chemical species, we find that the aver-age (cid:104) [O / Fe] (cid:105) ( + ± (cid:104) [Ti / Fe] (cid:105) ( + . ± . (cid:104) [Ni / Fe] (cid:105) ( − ± (cid:104) [Ce / Fe] (cid:105) ( + . ± . (cid:104) [Nd / Fe] (cid:105) ( + . ± . (cid:104) [Yb / Fe] (cid:105) ( + . ± .
09) ratios are comparable to values forM107 and NGC 362, and in agreement with previous results(Rojas-Arriagada et al. 2016; Crestani et al. 2019). However,some di ff erences are noteworthy for Nd, compared to GCs atsimilar metallicity, but it does agree with the production of other s -process species such as Ba (Rojas-Arriagada et al. 2016), withsmall star-to-star scatter. Panels (d) and (e) of Figure 1 show the existence of an extra-tidal star in the smoothly curved morphology of the upper RGBin the V o versus (B-V) o (Lee et al. 2014), and H o versus (J-K s ) o diagrams, extending from the cluster centre out to ∼ ff set in all three diagrams fromthe equally bright RGB stars indicate that this extra-tidal star isin reasonable agreement with the expected behaviour for AGBstars in GCs. Figure 1 (a) also shows that this extra-tidal star hassimilar metallicity and radial velocity, and Figures 1 (b) and (c)demonstrate the similarity of the orbital paths and proper mo-tions to NGC 6723 member stars as well. Thus, these propertiesare vetted not only by chemical abundances, but by photometryand kinematics as well.The α − element (Mg, Si, Ca, and Ti), the odd-Z element(Al and K), and the iron-peak element Ni abundance ratiosfall between − .
09 and + .
52 for our extra-tidal star. This issimilar to the members of NGC 6723 and other GC stars withcomparable metallicity, within the uncertainties (see Table 2 andFigure 3). Figure 3 also shows that the extra-tidal star is stronglyenhanced in the s -process elements (Ce II, Nd II, and Yb II),which is in agreement with other GC stars at similar metallicity(Mészáros et al. 2020).Figure 2 reveal that the newly identified extra-tidal star hasa stellar atmosphere strongly enriched in C N features, whichindicates a high enrichment in nitrogen ([N / Fe] = + . / GCstars, but indeed is part of a second-type of N-rich star with mod-est carbon enrichment, likely a GC star with AGB-like chemicalpatterns. Figures 2 and 3 also clearly show that K I, and Ce IIenhancement of the extra-tidal star can be convincingly claimed.Thus, we conclude that the Ce, Nd, and Yb enhancement wasinherited from the initial gas composition of NGC 6723.We also find that the oxygen abundance is in reasonableagreement with levels seen for NGC 6723 stars (Rojas-Arriagadaet al. 2016; Crestani et al. 2019). From the strength of the C Oand OH molecular features (see Figure 2), one would expectthat more oxygen in the atmosphere means that more carbon islocked into C O, and is not available to form C N, as thestrength of the C N features is anti-correlated with [O / Fe].From Table 1, even a small change in the surface temperatureand log g requires some adjustment to the carbon and oxygenabundance, but [N / Fe] is relatively insensitive to small T e ff andsurface gravity changes, and we can be confident that this star isindeed nitrogen enriched.
5. Possible origins for the observed chemicalcomposition of the extra-tidal star
There are several competing scenarios that might explain the un-usual chemical abundances of our extra-tidal star, in the contextof an escaped member of NGC 6723 that was tidally disruptedand captured by the MW’s bulge.It appears likely that some stars (C-poor or C-enriched) be-longing to the family of mildly metal-poor N-rich stars (peak-ing at [Fe / H] ∼ − .
0) could be relics of initially massive clus-ters such as NGC 6723. It is reasonable to assume that some ofthose objects were formed in Galactic GCs and later dynami-cally ejected into the Galactic field when their parent GCs wereultimately disrupted and destroyed, possibly during disc–bulgecrossing events (see, e.g., Leon et al. 2000; Lane et al. 2010;Fernández-Trincado et al. 2015a,b, 2016b; Kundu et al. 2019b,a,2020; Hanke et al. 2020; Sollima 2020). However, this has notheretofore been directly demonstrated.With the newly analyzed extra-tidal star, we want to investi-gate the possible origin of neutron-capture ( s -process) enrich-ment, simultaneous with the observed enrichment in nitrogenand its apparent carbon enrichment.Figure 5 shows the observed photospheric abundances of ourtarget star, along with theoretical (C, N, Mg, Al, Ce, Nd, andYb) abundance ratios from nucleosynthetic AGB models closestto [M / H] ∼ − .
09. We obtained these models by interpolatingthe stellar yields from Karakas et al. (2014) and Karakas et al.(2018). From inspection, an over-abundance of the elements cre-ated by the s -process (Ce II, Nd II, and Yb II) support the ideathat AGB stars with initial masses of ∼ (cid:12) could reason-ably explain the observed s -process enrichment. As also seen inthis figure, the C and N enrichments could be explained, at ∼ σ ,with initial masses of ∼ ∼ (cid:12) , respectively, while Mgand Al suggest an initial mass of ∼ (cid:12) . Thus, most of thesespecies are in agreement with production by an AGB star havingan initial mass ∼ (cid:12) , which underwent a mass-transferevent, resulting in pollution of an intermediate-mass companionthat today has evolved to the tip of the RGB. We caution aboutthe accuracy of the estimated [O / Fe] values, as they display alarger scatter, possibly due to telluric features and uncertain de-terminations in the T e ff regime of our objects. As highlightedin previous works, the uncertainty arises because BACCHUS de-termines these abundances from the strengths of C N and
Article number, page 8 of 16osé G. Fernández-Trincado et al.: A chemically atypical star being tidally disrupted from NGC 6723 X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° ≠ bar = 33 km s ° kpc X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° ≠ bar = 43 km s ° kpc X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° ≠ bar = 53 km s ° kpc X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° X [ k p c ] ° ° Y [ k p c ] ° ° Z [ k p c ] ° ° N G C e x t r a -t i d a l N - r i c h s t a r a t d i s t a n ce e x t r a -t i d a l N - r i c h s t a r a t d i s t a n ce a b cd e fg h i N G C e x t r a -t i d a l s t a r a t d i s t a n ce e x t r a -t i d a l s t a r a t d i s t a n ce Fig. 6.
Thousand-orbit realisations of NGC 6723 and the extra-tidal star time-integrated for 2 Gyr. The dark colours correspond to the mostprobable regions of the space, which are crossed more frequently by the simulated orbits, assuming three di ff erent values of the angular velocityof the bar ( Ω bar = , , and 53 km s − kpc). Panels (a), (b), and (c) show the orbits of NGC 6723, using as initial conditions the observed valuesfrom Baumgardt et al. (2019); the middle and bottom rows show the orbits of the extra-tidal star, adopting the observed values from the Table 1,and an assumed heliocentric distance at 8.3 kpc (panels d, e, and f), and 6.24 kpc (panels g, h, and i). C O lines, which become too weak for stars at relatively lowmetallicities ([Fe / H] − . s -process enrichment. How-ever, such a range in mass would correspond to a very young star( ∼ Extrinsic mechanism (binary-mass transfer system). Theover-abundance of s -process elements could come from theaccretion of s -process-rich matter from a former thermally pulsing (TP)-AGB companion during its heavy mass-lossphase on the AGB (Brown et al. 1990), which has sinceevolved into a faint white dwarf (see, Fernández-Trincadoet al. 2019c, for instance). Figure 5 indicates that our mea-sured abundance ratios, from the light elements (C, N), the α -element (Mg), and the odd-Z element (Al), to the s -processelements (Ce, Nd, and Yb), could have originated from ma-terial previously enriched in a TP-AGB companion with aninitial mass of ∼ . ∼ . (cid:12) , respectively.Consequently, if the case of an extrinsic mechanism isfavoured, and the extra-tidal star is / was part of a binary sys-tem before / after leaving the cluster, it would be possible thatthe star accreted N-rich material from the companion whenthe latter reached the AGB phase with the inferred mass. Inthis case, the abundances that we measure are not its orig-inal ones, but they reflect the chemical composition of theinterior of the companion, plus some degree of dilution inthe convective envelope of the accreting star. In addition tothe nitrogen and s -process enrichment, carbon should alsobe over-abundant, as a result of a third dredge-up episode Article number, page 9 of 16 & A proofs: manuscript no. NGC6723 in the donor. Such a star could have polluted the extra-tidalstar with C, N, and s -process-rich material, but lost so muchmass that the binary may be disrupted, or become so widethat RV variations would be very small, which could explainthe lack of a detectable white dwarf companion. However,the current PMs and radial velocity of the extra-tidal star ap-pear to support that this object was detached from the clusterby tidal disruption, consistent with the low relative velocitybetween this cluster and the star during close encounters. Insuch a case, the possibility that our object was part of a bi-nary system before departure from the cluster would not besupported (see e.g. Fernández-Trincado et al. 2016a).The absence of RV variation over the temporal span ofthe APOGEE-2 observations ( < Gaia
EDR3( − . ± .
23 km s − ) provide no evidence that our object(the extra-tidal star) is currently a binary with a white dwarfcompanion. Long-term RV monitoring of this unusual objectwould naturally be the best course to rule out an origin bythe binary channel.If there is any possibility to support the binary hypothesis inthe future, perhaps from careful long-term RV monitoring,then it is worth mentioning that, if the donor material was lostin a wind from a ∼ (cid:12) AGB star, it must have beenshed very recently. However, this scenario is not compatiblewith its apparent brightness in Figure 1 (d) and (e), as theobject is intrinsically bluer and more luminous than otherobserved stars in NGC 6723. • Intrinsic mechanism. A second scenario is that our unusualobject could itself be an AGB star. In this case, it was en-hanced in s -process elements via the third dredge-up of nu-cleosynthetic products created in the stellar interior and shellhelium-burning (via thermal pulses), followed by subsequentmixing that can result in an over-abundance of C and N atthe surface (Karakas et al. 2014). Moreover, if the extra-tidalstar is an actual evolved object, possibly an early-AGB or anAGB star, it would have a mass in the range ∼ . ∼ . (cid:12) to explain the observed photospheric chemical composi-tion. As pointed out above, this would imply that a GC originof this star is ruled out. • A third possibility could be that our star is an s -process-enriched red giant that is the result of star formation in an already s -process-enriched medium, where AGB stars haveplayed a more dominant role in chemical evolution (Venturaet al. 2016). Hence, a valid possibility is that the extra-tidalstar inherited its present chemical composition while it wasstill a cluster member, likely from gas lost by a previous gen-eration of ∼ (cid:12) AGB stars. This mass range is alsofairly well-supported by the observed [Mg / Fe], and [Al / Fe]abundance ratios, as shown in Figure 5. • A fourth plausible interpretation is that N-rich stars towardsthe bulge region are among the oldest in the Galaxy, andtheir abundances are in fact the imprints of the very earlychemical enrichment by spinstars, metal-poor, fast-rotatingmassive stars (Chiappini et al. 2011; Barbuy et al. 2014),which polluted the interstellar medium from which bulgeGCs formed. An enhancement in N and modest enhancementin C, as well as some contribution to the neutron-capture el-ements ( s -process nucleosynthesis) might be expected. Thisscenario would predict that all of the stars in NGC 6723 (andindeed in all bulge GCs formed in situ) would be N-rich stars,whereas if the third possibility is correct, then there shouldexist first-generation stars in the cluster that do not exhibitenhanced nitrogen. We conclude that it is very likely that we have identified aformer cluster member (the extra-tidal star), with observed abun-dances consistent with production by AGB stars that played animportant role in the chemical enrichment of NGC 6723, whichis in-line with previous results for GCs at similar metallicity(Ventura et al. 2016; Mészáros et al. 2020).It is important to note that, at [Fe / H] ∼ − .
17, it seems likelythat the s -process would dominate the production of such el-ements, invoking the AGB progenitor (Kobayashi et al. 2020)rather that other possible sources such as binary neutron star(NS–NS) mergers (Wanajo et al. 2014). In addition, if we were toassume that the n -capture came from an NS-NS merger, then wewould need a second progenitor to account for the N and otherlight elements.We also note some di ff erences between the extra-tidal starand MW field stars. Figure 4 shows a comparison between thisstar (red symbol) and bulge field stars at similar metallicity fromthe APOGEE-2 survey. While this result reinforces that N-richstars of di ff erent generations (Martell et al. 2016) populate di ff er-ent regions of the canonical abundance planes, some similaritiesare noteworthy with Galactic field stars. The unique nature ofour star is clear in the [C / Fe]–[N / Fe], [N / Fe]–[Al / Fe], [O / Fe]–[Al / Fe], [Mg / Fe]–[Al / Fe], [Mg / Fe]–[Si / Fe], [Mg / Fe]–[K / Fe],and [Ca / Fe]–[Ti / Fe] planes, which slightly exceed the Galacticlevels, while some similarities remain for the α − elements Mg,Si, Ca, and Ti, the odd-Z elements Al and K, and the iron-peakelement Ti. In contrast to bulge field stars, the extra-tidal starclearly has higher [C, N, O / Fe] abundance ratios ( > + . ff erences (and similarities) suggest that the extra-tidal star ex-hibits the typical chemical patterns similar to those of evolvedstars in GC at [Fe / H] ∼ − . ff ects the measurement of the ironabundance in some way, causing a large o ff set with respect to thecluster mean. Another detailed photometric / spectroscopic analy-sis of this star is needed to investigate possible variability e ff ectson the metallicity derivation. We note that, according to its radialvelocity, proper motions, position in the CMD, and atmosphericparameters, this star is / was very likely a cluster member.
6. Expected Radial Profile of halo field stars with aglobular cluster-like abundance patterns towardNGC 6723
We compute the predicted number ( N gc ) of halo field stars withglobular cluster-like abundance patterns observed in APOGEE-2 (see, e.g., Fernández-Trincado et al. 2016a, 2017; Schiavonet al. 2017b; Fernández-Trincado et al. 2019b) towards the fieldof NGC 6723 using the smooth halo density relations presentedin Horta et al. (2021), and by adopting the same Monte Carlo im-plementation of the Von Neumann Rejection Technique (Presset al. 2002) as in Eq. 1 in Mateu et al. (2009) and Eq. 7 in(Fernández-Trincado et al. 2015b).We find the expected number of observed APOGEE-2 halofield stars with a possible GC origin in the range 5 . < d (cid:12) < . N gc < .
05 (from 1000 Monte Carlo realisations).
Article number, page 10 of 16osé G. Fernández-Trincado et al.: A chemically atypical star being tidally disrupted from NGC 6723
This indicates that the region around NGC 6723 is underdense,not overdense, in Galactic halo field stars with globular cluster-like abundance patterns, leaving enough room to favour the sce-nario of a genuine extra-tidal star associated with NGC 6723.
7. Galactic orbit
We used the
GravPot16 algorithm for the calculation of the or-bital path of both NGC 6723 and the extra-tidal star around theMW. The GravPot16 model is composed of multiple potentialsfor the Galactic disc and a box / peanut bulge. For the Sun’s posi-tion in the MW we assumed a distance to the Galactic Centre of R (cid:12) = U (cid:12) , V (cid:12) , W (cid:12) ) = (11 . , . , .
25) km s − (Brunthaler et al. 2011) and the ve-locity of the local standard of rest (LSR) ν LS R = . − ,based on R o and assuming the composite rotation curve of Sofue(2015).For NGC 6723 and the extra-tidal star, we integrated overa 2 Gyr timespan, and calculated their orbital paths and orbitalparameters using three di ff erent values of the angular velocityof the bar Ω bar = , , and 53 km s − kpc − . For NGC 6723,we adopted the observational parameters from Baumgardt et al.(2019); for the extra-tidal star, we adopted the observational pa-rameters given in Table 1, except for the distance, for which weassumed two di ff erent values, e.g., 8 . ± .
83 kpc, the same he-liocentric distance as NGC 6723 (Baumgardt et al. 2019), and6 . ± .
69 kpc, the estimated distance from the
StarHorse code (Anders et al. 2019) (the algorithm combines the
Gaia as-trometric information with multiband photometric informationand a number of Galactic priors), which have smaller uncertain-ties at large ranges, as previously expected, and which have beenproven to be an useful tool in confirming the dichotomy withAPOGEE-2 / DR16 and dissecting the bulge stellar populations(Queiroz et al. 2018, 2020).A simple Monte Carlo analysis was performed for the com-plete calculation, taking into account the errors in distance, RV,and proper motions, which were randomly propagated as 1 σ variations, assumed to follow Gaussian distributions. For bothobjects we computed 1000 orbits. Figure 6 shows the probabil-ity densities of the resulting orbits on the X–Y, X–Z, and Y–Zplanes in the non-inertial reference frame where the bar is at rest,where the dark area correspond to the most probable regions ofthe space, which are crossed more frequently by the simulatedorbits. We found that both cluster and extra-tidal star are situ-ated in the inner region of the bulge. For the full ensemble oforbits, we also calculated the deviations in the orbital parame-ters due to the di ff erent values of the angular velocity of the bar,( ∆ r peri , ∆ r apo , ∆ Z max , ∆ e ): NGC 6723 (0.02 kpc, 0.65 kpc, 0.37kpc, 0.01), extra-tidal star at d (cid:12) = . d (cid:12) = .
24 kpc (0.06 kpc,0.07 kpc, 0.05 kpc, 0.03). With the di ff erent values of Ω bar , wecan see that the predicted orbits are not significantly a ff ected bythe change of this parameter, and consequently, it does alter ourconclusions.The median orbital parameters, assuming Ω bar =
43 km s − kpc for NGC 6723 and the extra-tidal star, are as follows: ( r peri , r apo , Z max , e ): NGC 6723 (0 . ± .
08 kpc, 3 . ± .
12 kpc, 3 . ± .
12 kpc, 0 . ± . d (cid:12) = . . ± . . ± .
10 kpc, 3 . ± .
12 kpc, 0 . ± . d (cid:12) = .
24 kpc (1 . ± .
07 kpc, 4 . ± .
66 kpc, 2 . ± . . ± . https://gravpot.utinam.cnrs.fr found extra-tidal star possess trajectories indicating that they areconfined to the bulge region ( (cid:46)
8. Comments on the tidal radius
The tidal radius for NGC 6397 has been measured by severalstudies, resulting in di ff erent values. For instance, a recent workby de Boer et al. (2019) using Gaia
DR2 data determined atidal radius that peaks at r t = . ± .
50 pc ( ∼ . (cid:48) ) froma generalised lowered isothermal LIMEPY model. This is con-siderably larger than previously thought. The Baumgardt et al.(2019) value, using the same data set and a di ff erent analysismethod, adopted r t = .
95 pc ( ∼ (cid:48) ), which is substantiallylarger than the value of r t = . (cid:48) listed by Harris (1996, 2010,2010 edition).Moreover, simulations from Moreno et al. (2014) calculatedtidal radii between 29 – 31.4 pc using a 6-D dataset in an ax-isymmetric and non-axisymmetric Galactic potential, includinga rotating Galactic bar and a 3-D model for the spiral arms. Theirresults indicate good agreement with the tabulated value fromHarris’s compilation and the cluster Jacobi radii determined inPiatti et al. (2019). Furthermore, the observed cluster stellar-density maps from near-infrared observations (see, e.g., Chunet al. 2015) suggest the existence of over-density features ofsome number of stars from ∼ . (cid:48) (cid:46) r (cid:46) . (cid:48) , exhibiting az-imuthally irregular patterns. Our newly identified extra-tidal starlies in this region as shown in Figure 7, and placed well outsidethe Jacobi radius ( (cid:38) (cid:48) ; Piatti et al. 2019). This called our atten-tion to the recent derived slightly larger r t values from de Boeret al. (2019), based on Gaia
DR2. This value, should be viewedwith caution, as it has its own limitations toward the bulge re-gions, and could lead to fitting non-physically motivated coreKing (1962) and tidal radii.With the new improved data from
Gaia
EDR3 (Gaia Collab-oration et al. 2020), and the available
Hubble Space Telescope (HST) imaging data within 1.1 (cid:48) from the cluster center (Saraje-dini et al. 2007), we can determine updated structural parameterswith better precision. To construct number density profiles, wemake use of star counts by adopting the same technique as in Co-hen et al. (2020) to combine both surveys. Briefly, the stellar den-sity is found in annuli of varying radii around the centre of thecluster, and these density values are fitted by a King profile (King1966). The HST + Gaia
EDR3 data is fairly well-fit by the (King1966) profile. The stellar-density profile is shown in Figure 8.For NGC 6723 we have determined a radius r t = . ± . (cid:48) , which is in good agreement with the values reported in Har-ris (1996, 2010 edition), and by dynamical studies of NGC 6723(see, e.g., Moreno et al. 2014; Piatti et al. 2019).Our finding is in line with recent studies in the vicinity ofNGC 6723, for example, near-infrared J-, H-, and K-band obser-vations from the WFCAM camera on the 3.8 m UKIRT Chunet al. (2015), suggesting the existence of extra-tidal features (cid:38) (cid:48) (Fig. 7), and is in good agreement with a rather highdestruction ratio of ν tot /ν evap ∼ . − . ν tot is the total destructionrate of this cluster, and ν evap is the evaporation rate per Hub-ble time). The newly identified extra-tidal star lie near the weakextended substructure beyond the tidal radius in the southern re- Article number, page 11 of 16 & A proofs: manuscript no. NGC6723
Fig. 7.
Modified Figure 10 from Chun et al. (2015). The red ‘star’ symbol shows the position of the newly identified extra-tidal star. The orangecircle indicates the cluster tidal radius ( r t = . ± . (cid:48) , see Section 8), and the red arrow defines the proper motion of NGC 6723 from Baumgardtet al. (2019). The orange solid and dashed line shows the direction of the Galactic Centre and Galactic plane, respectively. The top-left panel is thestar-count map around the cluster; the top-right panel is the surface-density map, smoothed by a Gaussian kernel 0.07 ◦ ; the lower-left panel is thesame data, but smoothed with a 0.11 ◦ kernel; the bottom-right panel shows the distribution map of E(B-V). The iso-density contour levels shownare 2.0 σ , 2.5 σ , 3.0 σ , 4.0 σ , 5.0 σ , and 7.0 σ . The orbital path of the cluster (black line) and the extra-tidal star (purple line) is over-plotted. gion of NGC 6723 (see, e.g., Chun et al. 2015) as highlightedin Figure 7, suggesting that this star has left the GC’s potential.Thus, our findings provide strong support for the idea of linkingN-rich field stars to GCs in the bulge region of the MW (Bekki2019).
9. Concluding remarks
Our finding is in-line with recent studies in the vicinity ofNGC 6723, e.g., near-infrared J-, H-, and K-band observationsfrom the WFCAM camera on the 3.8 m UKIRT (Chun et al.2015), suggesting the existence of extra-tidal features aroundthis cluster (see Figure 7). Thus, our findings provide strong sup-port for the idea of linking some of the observed N-rich fieldstars to GCs in the bulge region of the MW (e.g., Schiavon et al.2017b; Fernández-Trincado et al. 2019b).We also conclude that it is very likely that we have identifieda former cluster member, with observed abundances consistentwith production by AGB stars that played an important role in the chemical enrichment of NGC 6723, which is in-line withprevious results for GCs at similar metallicity (Mészáros et al.2020).There are also several competing scenarios that might ex-plain the unusual chemical abundances of the newly identi-fied extra-tidal star, in the context of an escaped member ofNGC 6723 that was tidally disrupted and captured by the MW’sbulge. In summary: • Extrinsic mechanism : The over-abundance of the s -processelements in this star could come from the accretion of s -process-rich matter from a former thermally pulsing AGBcompanion during its heavy mass-loss phase on the AGB,which has since evolved into a faint white dwarf star. • Intrinsic mechanism : This star could be an intrinsic AGBstar, enhanced in s -process elements during the third dredge-up of nucleosynthetic products created in the stellar interiorand shell helium-burning (via thermal pulses), followed bysubsequent mixing that can result in an over-abundance ofcarbon and nitrogen at the surface. Article number, page 12 of 16osé G. Fernández-Trincado et al.: A chemically atypical star being tidally disrupted from NGC 6723
Fig. 8.
Stellar-density profile of NGC 6723 determined from theHST + Gaia
EDR3 data set. Small orange symbols show densities be-fore background subtraction. The-best fit King (1966) profile is shownin lime and cyan, respectively. The red dashed line show the determinedtidal radius ( r t = . ± . (cid:48) ) in this work, while the black line in 1.1 (cid:48) separate the HST from Gaia
EDR3 data set. • This star could be an s -process-enriched red giant result-ing from star formation in an already s -process-enrichedmedium, where low-mass AGB stars have played a dominantrole in chemical evolution.Finally, the newly identified extra-tidal star displays a [N / Fe]ratio similar to other metal-poor N-rich field stars, but whichare not carbon enriched (one of the typical signatures of second-generation GC stars)–(see e.g., Fernández-Trincado et al. 2016a,2017; Schiavon et al. 2017b; Fernández-Trincado et al. 2019b),suggesting that the extra-tidal star is not a genuine second-generation star, but it is part of a sub-family of the N-rich starswith modest carbon enrichment as that observed in some Galac-tic GC stars at similar metallicity, likely associated with theintermediate-mass ( (cid:46) (cid:12) ) population of early-AGB stars.
Acknowledgements.
We thank the anonymous referee for helpful comments thatgreatly improved the paper. We thank Dr. Lee Y.-W. J.G.F-T is supported byFONDECYT No. 3180210. T.C.B. acknowledge partial support for this workfrom grant PHY 14-30152; Physics Frontier Center / JINA Center for the Evolu-tion of the Elements (JINA-CEE), awarded by the US National Science Founda-tion. D.M. is supported by the BASAL Center for Astrophysics and AssociatedTechnologies (CATA) through grant AFB 170002, and by project FONDECYTRegular No. 1170121. D.G. gratefully acknowledges support from the ChileanCentro de Excelencia en Astrofísica y Tecnologías Afines (CATA) BASAL grantAFB-170002. D.G. also acknowledges financial support from the Dirección deInvestigación y Desarrollo de la Universidad de La Serena through the Programade Incentivo a la Investigación de Académicos (PIA-DIDULS). S.V. gratefullyacknowledges the support provided by Fondecyt reg. n. 1170518. S.O.S. ac-knowledges the FAPESP PhD fellowship 2018 / ◦ ffi / University ofTokyo, Lawrence Berkeley National Laboratory, Leibniz Institut für AstrophysikPotsdam (AIP), Max-Planck-Institut für Astronomie (MPIA Heidelberg), Max-Planck-Institut für Astrophysik (MPA Garching), Max-Planck-Institut für Ex-traterrestrische Physik (MPE), National Astronomical Observatory of China,New Mexico State University, New York University, the University of NotreDame, Observatório Nacional / MCTI, The Ohio State University, Pennsylva-nia State University, Shanghai Astronomical Observatory, United Kingdom Par-ticipation Group, Universidad Nacional Autónoma de México, University ofArizona, University of Colorado Boulder, University of Oxford, University ofPortsmouth, University of Utah, University of Virginia, University of Washing-ton, University of Wisconsin, Vanderbilt University, and Yale University.This work has made use of data from the European Space Agency (ESA) mis-sion
Gaia ( ), processed by the Gaia
DataProcessing and Analysis Consortium (DPAC, ). Funding for the DPAC has been provided bynational institutions, in particular the institutions participating in the
Gaia
Mul-tilateral Agreement.
Article number, page 13 of 16 & A proofs: manuscript no. NGC6723 T a b l e . T op t ab l e : F i n a l a bund a n ce s f o r t h e s t a r s a n a l y ze d i n t h i s w o r k , a dop ti ng s t e ll a r a t m o s ph e r i c p a r a m e t e r s fr o m pho t o m e t r y ( T e ff ) a nd l og g fr o m . G y r PARSEC i s o c h r on e s , a s d e t e r m i n e du s i ng l o ca lt h e r m odyn a m i ce qu ili b r i u m ( LTE ) m od e l a t m o s ph e r e s . M i dd l e t ab l e : T h e fi n a l a bund a n ce s a s d e t e r m i n e dby t h e ASPCAP p i p e li n e . B o tt o m t ab l e : L i s t o f t h e m a i nphy s i ca l p a r a m e t e r s o f ou r s t a r s . A P OG EE _ I D T e ff l og g [ M / H ] ξ t [ C / F e ][ N / F e ][ O / F e ][ M g / F e ][ A l / F e ][ S i / F e ][ K / F e ][ C a / F e ][ T i / F e ][ F e / H ][ N i / F e ][ C e / F e ][ N d / F e ][ Y b / F e ] T h i s w o r k ( K ) k m s − M − . − . . − . + . + . + . + . + . − . + . + . − . − . + . ...... M − . − . . − . + . + . + . + . + . + . + . + . − . − . + . ...... M − . − . . − . + . + . + . + . + . + . + . + . − . + . + . + . ... M − . − . . − . + . + . + . + . + . + . + . + . − . + . + . ...... M − . − . . − . + . + . + . + . + . + . + . + . − . − . + . ...... M − . − . . − . + . + . + . + . + . + . + . + . − . + . + . + . + . M − . − . . − . + . + . + . + . + . + . + . + . − . + . + . ... + .
76 2 M − . − . . + . + . + . + . + . + . + . + . + . − . + . + . + . + . A P OG EE _ I D T e ff l og g [ M / H ] ξ t [ C / F e ][ N / F e ][ O / F e ][ M g / F e ][ A l / F e ][ S i / F e ][ K / F e ][ C a / F e ][ T i / F e ][ F e / H ][ N i / F e ][ C e / F e ][ N d / F e ][ Y b / F e ] ASPCAP ( K ) k m s − M − . − . . − . + . + . + . − . + . + . + . + . − . − . − . ...... M − . − . . − . + . + . + . + . + . + . + . + . − . + . + . ...... M − . − . . − . + . + . + . + . + . + . + . ... − . + . + . ...... M − . − . . − . + . + . + . + . + . + . + . + . − . + . + . ...... M − . − . . − . + . + . + . + . + . + . + . + . − . + . + . ...... M − . − . . − . + . + . + . + . + . + . + . ... − . + . ......... M − . − . . − . + . + . + . + . + . + . + . ... − . + . + . ...... M − − . − . . + . + . + . + . + . + . + . + . ... − . − . ......... A P OG EE _ I D S / N RV σ RV E ( B - V ) V o ( B - V ) o H o ( J - K s ) o µ α c o s ( δ ) µ δ RUWE V i s it s k m s − k m s − ( m a s y r − )( m a s y r − ) CTIOCTIO2MASS2MASS M − − . . . . . . . . ± . − . ± . . M − − . . . . . . . . ± . − . ± . . M − − . . . . . . . . ± . − . ± . . M − − . . . . . . . . ± . − . ± . . M − − . . . . . . . . ± . − . ± . . M − − . . . . . . . . ± . − . ± . . M − − . ... . . . . . . ± . − . ± . .
212 2 M − − . . . . . . . . ± . − . ± . . N o t e s . N o t e : T h e s yn t h e ti c s p ec t r a w e r e b a s e don1 D L o ca l T h e r m odyn a m i c E qu ili b r i u m ( LTE ) m od e l a t m o s ph e r e s ca l c u l a t e du s i ng MARCS m od e l s ( G u s t a f ss on e t a l . ) a nd t h e S o l a r a bund a n ce s fr o m A s p l und e t a l . ( ) , e x ce p t f o r C e II , N d II , a nd Y b II , f o r w h i c h w e h a v ea dop t e d t h e S o l a r a bund a n ce s fr o m G r e v e ss ee t a l . ( ) . Article number, page 14 of 16osé G. Fernández-Trincado et al.: A chemically atypical star being tidally disrupted from NGC 6723
Table 2.
Abundance-ratio uncertainties for the model parameters with ∆ T e ff =
100 K, ∆ log g = ∆ ξ t = .
05 km s − , the uncertaintieson the mean due to line-to-line scatter ( σ [X / H] , mean ), and correspondingtotal error. σ [X / H] , log g σ [X / H] , T e ff σ [X / H] ,ξ t σ [X / H] , mean σ [X / H] , total − [dex] [dex] [dex] [dex] [dex] C O → C 0.065 0.018 0.004 0.110 0.129 C N → N 0.075 0.094 0.023 0.071 0.142 OH → O 0.018 0.097 0.014 0.068 0.121Mg I 0.005 0.089 0.022 0.052 0.106Al I 0.011 0.080 0.006 0.117 0.142Si I 0.045 0.023 0.006 0.094 0.107K I 0.025 0.044 0.009 0.137 0.146Ca I 0.023 0.023 0.007 0.085 0.091Ti I 0.007 0.070 0.033 0.246 0.258Fe I 0.016 0.042 0.005 0.154 0.161Ni I 0.027 0.047 0.010 0.155 0.165Ce II 0.123 0.067 0.025 0.125 0.189Nd II 0.075 0.071 0.023 0.130 0.168Yb II 0.092 0.024 0.004 0.020 0.097 σ [X / H] , log g σ [X / H] , T e ff σ [X / H] ,ξ t σ [X / H] , mean σ [X / H] , total − [dex] [dex] [dex] [dex] [dex] C O → C 0.035 0.064 0.009 0.052 0.090 C N → N 0.042 0.108 0.004 0.069 0.135 OH → O 0.049 0.139 0.003 0.035 0.152Mg I 0.025 0.073 0.016 0.017 0.081Al I 0.014 0.111 0.009 0.116 0.161Si I 0.033 0.007 0.003 0.079 0.086K I 0.038 0.009 0.004 0.014 0.042Ca I 0.037 0.033 0.003 0.026 0.056Ti I 0.067 0.070 0.043 0.199 0.225Fe I 0.036 0.072 0.008 0.116 0.141Ni I 0.011 0.021 0.003 0.070 0.074Ce II 0.059 0.013 0.004 0.069 0.092Nd II 0.049 0.043 0.005 0.043 0.078Yb II 0.150 0.027 0.002 0.047 0.159
Note:
The total error is defined as: σ [X i / Fe] , total = (cid:113) σ i / Fe] , log g + σ i / Fe] , T e ff + σ i / Fe] ,ξ t + σ i / Fe] , mean . References
Ahumada, R., Allende Prieto, C., Almeida, A., et al. 2020, ApJS, 249, 3Alonso-García, J., Minniti, D., Catelan, M., et al. 2017, ApJ, 849, L13Alvarez, R. & Plez, B. 1998, A&A, 330, 1109Anders, F., Khalatyan, A., Chiappini, C., et al. 2019, A&A, 628, A94Asplund, M., Grevesse, N., & Sauval, A. J. 2005, in Astronomical Society of thePacific Conference Series, Vol. 336, Cosmic Abundances as Records of Stel-lar Evolution and Nucleosynthesis, ed. I. Barnes, Thomas G. & F. N. Bash,25Barbuy, B., Chiappini, C., Cantelli, E., et al. 2014, A&A, 570, A76Baumgardt, H., Hilker, M., Sollima, A., & Bellini, A. 2019, MNRAS, 482, 5138Beers, T. C. & Christlieb, N. 2005, ARA&A, 43, 531Bekki, K. 2019, MNRAS, 490, 4007Blanton, M. R., Bershady, M. A., Abolfathi, B., et al. 2017, AJ, 154, 28Bowen, I. S. & Vaughan, A. H., J. 1973, Appl. Opt., 12, 1430Bressan, A., Marigo, P., Girardi, L., et al. 2012, MNRAS, 427, 127Brown, J. A., Wallerstein, G., & Oke, J. B. 1990, AJ, 100, 1561Brunthaler, A., Reid, M. J., Menten, K. M., et al. 2011, AstronomischeNachrichten, 332, 461Chiappini, C., Frischknecht, U., Meynet, G., et al. 2011, Nature, 472, 454Chun, S.-H., Kang, M., Jung, D., & Sohn, Y.-J. 2015, AJ, 149, 29 Cohen, R. E., Bellini, A., Libralato, M., et al. 2020, arXiv e-prints,arXiv:2012.00791Crestani, J., Alves-Brito, A., Bono, G., Puls, A. A., & Alonso-García, J. 2019,MNRAS, 487, 5463Cunha, K., Smith, V. V., Hasselquist, S., et al. 2017, ApJ, 844, 145de Boer, T. J. L., Gieles, M., Balbinot, E., et al. 2019, MNRAS, 485, 4906De Smedt, K., Van Winckel, H., Kamath, D., et al. 2016, A&A, 587, A6Dotter, A., Sarajedini, A., Anderson, J., et al. 2010, ApJ, 708, 698Fernández-Trincado, J. G., Beers, T. C., & Minniti, D. 2020a, A&A, 644, A83Fernández-Trincado, J. G., Beers, T. C., Minniti, D., et al. 2020b, ApJ, 903, L17Fernández-Trincado, J. G., Beers, T. C., Minniti, D., et al. 2020c, A&A, 643, L4Fernández-Trincado, J. G., Beers, T. C., Placco, V. M., et al. 2019a, ApJ, 886, L8Fernández-Trincado, J. G., Beers, T. C., Tang, B., et al. 2019b, MNRAS, 488,2864Fernández-Trincado, J. G., Chaves-Velasquez, L., Pérez-Villegas, A., et al.2020d, MNRAS, 495, 4113Fernández-Trincado, J. G., Mennickent, R., Cabezas, M., et al. 2019c, A&A,631, A97Fernández-Trincado, J. G., Minniti, D., Beers, T. C., et al. 2020e, A&A, 643,A145Fernández-Trincado, J. G., Robin, A. C., Moreno, E., et al. 2016a, ApJ, 833, 132Fernández-Trincado, J. G., Robin, A. C., Reylé, C., et al. 2016b, MNRAS, 461,1404Fernández-Trincado, J. G., Robin, A. C., Vieira, K., et al. 2015a, A&A, 583, A76Fernández-Trincado, J. G., Vivas, A. K., Mateu, C. E., et al. 2015b, A&A, 574,A15Fernández-Trincado, J. G., Zamora, O., García-Hernández, D. A., et al. 2017,ApJ, 846, L2Fernández-Trincado, J. G., Zamora, O., Souto, D., et al. 2019d, A&A, 627, A178Frebel, A., Christlieb, N., Norris, J. E., et al. 2006, ApJ, 652, 1585Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al. 2020, arXiv e-prints,arXiv:2012.01533García Pérez, A. E., Allende Prieto, C., Holtzman, J. A., et al. 2016, AJ, 151,144Gnedin, O. Y. & Ostriker, J. P. 1997, ApJ, 474, 223González Hernández, J. I. & Bonifacio, P. 2009, A&A, 497, 497Grevesse, N., Scott, P., Asplund, M., & Sauval, A. J. 2015, A&A, 573, A27Gunn, J. E., Siegmund, W. A., Mannery, E. J., et al. 2006, AJ, 131, 2332Gustafsson, B., Edvardsson, B., Eriksson, K., et al. 2008, A&A, 486, 951Hanke, M., Koch, A., Prudil, Z., Grebel, E. K., & Bastian, U. 2020, A&A, 637,A98Harris, W. E. 1996, AJ, 112, 1487Harris, W. E. 2010, arXiv e-prints, arXiv:1012.3224Hasselquist, S., Shetrone, M., Cunha, K., et al. 2016, ApJ, 833, 81Hawkins, K., Masseron, T., Jofré, P., et al. 2016, A&A, 594, A43Holtzman, J. A., Hasselquist, S., Shetrone, M., et al. 2018, AJ, 156, 125Horta, D., Mackereth, J. T., Schiavon, R. P., et al. 2021, MNRAS, 500, 5462Johnson, J. A., Herwig, F., Beers, T. C., & Christlieb, N. 2007, ApJ, 658, 1203Karakas, A. I., Lugaro, M., Carlos, M., et al. 2018, MNRAS, 477, 421Karakas, A. I., Marino, A. F., & Nataf, D. M. 2014, ApJ, 784, 32Karinkuzhi, D. & Goswami, A. 2014, MNRAS, 440, 1095King, I. 1962, AJ, 67, 471King, I. R. 1966, AJ, 71, 64Kobayashi, C., Karakas, A. I., & Lugaro, M. 2020, ApJ, 900, 179Kundu, R., Fernández-Trincado, J. G., Minniti, D., et al. 2019a, MNRAS, 489,4565Kundu, R., Minniti, D., & Singh, H. P. 2019b, MNRAS, 483, 1737Kundu, R., Navarrete, C., Fernández-Trincado, J. G., et al. 2020, arXiv e-prints,arXiv:2011.10525Lane, R. R., Kiss, L. L., Lewis, G. F., et al. 2010, MNRAS, 406, 2732Lee, J.-W., López-Morales, M., Hong, K., et al. 2014, ApJS, 210, 6Lee, Y. S., Beers, T. C., Masseron, T., et al. 2013, AJ, 146, 132Leon, S., Meylan, G., & Combes, F. 2000, A&A, 359, 907Majewski, S. R., Schiavon, R. P., Frinchaboy, P. M., et al. 2017, AJ, 154, 94Majewski, S. R., Zasowski, G., & Nidever, D. L. 2011, ApJ, 739, 25Martell, S. L., Shetrone, M. D., Lucatello, S., et al. 2016, ApJ, 825, 146Masseron, T., Merle, T., & Hawkins, K. 2016, BACCHUS: Brussels AutomaticCode for Characterizing High accUracy SpectraMateu, C., Vivas, A. K., Zinn, R., Miller, L. R., & Abad, C. 2009, AJ, 137, 4412Mészáros, S., Masseron, T., García-Hernández, D. A., et al. 2020, MNRAS, 492,1641Moreno, E., Pichardo, B., & Velázquez, H. 2014, ApJ, 793, 110Muñoz, C., Geisler, D., Villanova, S., et al. 2018, A&A, 620, A96Nidever, D. L., Holtzman, J. A., Allende Prieto, C., et al. 2015, AJ, 150, 173Oliveira, R. A. P., Souza, S. O., Kerber, L. O., et al. 2020, ApJ, 891, 37Piatti, A. E., Webb, J. J., & Carlberg, R. G. 2019, MNRAS, 489, 4367Placco, V. M., Beers, T. C., Santucci, R. M., et al. 2018, AJ, 155, 256Placco, V. M., Frebel, A., Beers, T. C., et al. 2014, ApJ, 781, 40Plez, B. 2012, Turbospectrum: Code for spectral synthesis
Article number, page 15 of 16 & A proofs: manuscript no. NGC6723
Press, W. H., Teukolsky, S. A., Vetterling, W. T., & Flannery, B. P. 2002, Numer-ical recipes in C ++ : the art of scientific computingQueiroz, A. B. A., Anders, F., Chiappini, C., et al. 2020, A&A, 638, A76Queiroz, A. B. A., Anders, F., Santiago, B. X., et al. 2018, MNRAS, 476, 2556Rojas-Arriagada, A., Zoccali, M., Vásquez, S., et al. 2016, A&A, 587, A95Sarajedini, A., Bedin, L. R., Chaboyer, B., et al. 2007, AJ, 133, 1658Schiavon, R. P., Johnson, J. A., Frinchaboy, P. M., et al. 2017a, MNRAS, 466,1010Schiavon, R. P., Zamora, O., Carrera, R., et al. 2017b, MNRAS, 465, 501Smith, V. V., Cunha, K., Shetrone, M. D., et al. 2013, ApJ, 765, 16Sofue, Y. 2015, PASJ, 67, 75Sollima, A. 2020, MNRAS, 495, 2222Souza, S. O., Kerber, L. O., Barbuy, B., et al. 2020, ApJ, 890, 38Ventura, P., di Criscienzo, M., Schneider, R., et al. 2012, MNRAS, 420, 1442Ventura, P., García-Hernández, D. A., Dell’Agli, F., et al. 2016, ApJ, 831, L17Wanajo, S., Sekiguchi, Y., Nishimura, N., et al. 2014, ApJ, 789, L39Yoon, J., Beers, T. C., Tian, D., & Whitten, D. D. 2019, ApJ, 878, 97Zasowski, G., Cohen, R. E., Chojnowski, S. D., et al. 2017, AJ, 154, 198 Institut Utinam, CNRS UMR 6213, Université Bourgogne-Franche-Comté, OSU THETA Franche-Comté, Observatoire de Besançon,BP 1615, 25010 Besançon Cedex, France Instituto de Astronomía y Ciencias Planetarias, Universidad de Ata-cama, Copayapu 485, Copiapó, Chile Department of Physics and JINA Center for the Evolution of theElements, University of Notre Dame, Notre Dame, IN 46556, USA Depto. de Cs. Físicas, Facultad de Ciencias Exactas, UniversidadAndrés Bello, Av. Fernández Concha 700, Las Condes, Santiago,Chile Vatican Observatory, V00120 Vatican City State, Italy Instituto de Astronomía, Universidad Nacional Autónoma de Méx-ico, A.P. 70-264, 04510, Ciudad de México, Mexico NSF’s National Optical-Infrared Astronomy Research Laboratory,Tucson, AZ 85719, USA Korea Astronomy and Space Science Institute, 776 Daedeokdae-ro,Yuseong-gu, Daejeon 34055, Republic of Korea Departamento de Astronomí a, Casilla 160-C, Universidad de Con-cepción, Concepción, Chile Departamento de Astronomía, Universidad de La Serena, 1700000La Serena, Chile Instituto de Investigación Multidisciplinario en Ciencia y Tec-nología, Universidad de La Serena. Benavente 980, La Serena, Chile Universidade de São Paulo, IAG, Rua do Matão 1226, Cidade Uni-versitária, São Paulo 05508-900, Brazil Instituto de Astronomía, Universidad Nacional Autónoma de Méx-ico, Apdo. Postal 106, 22800 Ensenada, B.C., México Leibniz-Institut für Astrophysik Potsdam (AIP), An der Sternwarte16, 14482 Potsdam, Germany Laboratório Interinstitucional de e-Astronomia - LIneA, Rua Gal.José Cristino 77, Rio de Janeiro, RJ - 20921-400, Brazil School of Physics and Astronomy, Sun Yat-sen University, Zhuhai519082, China Centro de Astronomía (CITEVA), Universidad de Antofagasta, Av.Angamos 601, Antofagasta, Chile Millennium Institute of Astrophysics, Santiago, Chile Instituto Interdisciplinario de Ciencias Básicas (ICB), CONICET-UNCUYO, Padre J. Contreras 1300, M5502JMA, Mendoza, Ar-gentina Consejo Nacional de Investigaciones Científicas y Técnicas (CON-ICET), Godoy Cruz 2290, C1425FQB, Buenos Aires, Argentina Universidad Nacional de Córdoba, Observatorio Astronómico deCórdoba, Laprida 854, 5000 Córdoba, Argentina Consejo Nacional de Investigaciones Científicas y Técnicas (CON-ICET), Godoy Cruz 2290, Ciudad Autónoma de Buenos Aires, Ar-gentina Universidade Federal do Rio Grande do Sul, Instituto de Física, Av.Bento Gonçalves 9500, Porto Alegre, RS, Brazil Instituto de Astronomía, Universidad Católica del Norte, Av. Ang-amos 0610, Antofagasta, Chile Departamento de Astronomía, Universidad de Chile, Camino delObservatorio 1515, Las Condes, Santiago, Chile Visiting astronomer, Cerro Tololo Inter-American Observatory, Na-tional Optical Astronomy Observatory, which is operated by the As- sociation of Universities for Research in Astronomy (AURA) un-der a cooperative agreement with the National Science Foundation,Chile Department of Physics and Astrophysics, University of Delhi,Delhi-110007, India European Southern Observatory, Alonso de Córdova 3107, 7630000Vitacura, Santiago, Chile University of Nariño Observatory, Universidad de Nariño, SedeVIIS, Avenida Panamericana, Pasto, Nariño, Colombia30