A homogeneous comparison between the chemical composition of the Large Magellanic Cloud and the Sagittarius dwarf galaxy
A. Minelli, A. Mucciarelli, D. Romano, M. Bellazzini, L. Origlia, F. R. Ferraro
DDraft version February 10, 2021
Typeset using L A TEX
RNAAS style in AASTeX63
A homogeneous comparison between the chemical composition of the Large Magellanic Cloud and theSagittarius dwarf galaxy ∗ A. Minelli,
1, 2
A. Mucciarelli,
1, 2
D. Romano, M. Bellazzini, L. Origlia, and F. R. Ferraro
1, 21
Dipartimento di Fisica e Astronomia
Augusto Righi , Universit`a degli Studi di Bologna, Via Gobetti 93/2, I-40129 Bologna, Italy INAF - Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Gobetti 93/3, I-40129 Bologna, Italy
ABSTRACTSimilarities in the chemical composition of two of the closest Milky Way satellites, namely the LargeMagellanic Cloud (LMC) and the Sagittarius (Sgr) dwarf galaxy, have been proposed in the literature,suggesting similar chemical enrichment histories between the two galaxies. This proposition, however,rests on different abundance analyses, which likely introduce various systematics that hamper a faircomparison among the different data sets. In order to bypass this issue (and highlight real similaritiesand differences between their abundance patterns), we present a homogeneous chemical analysis of 30giant stars in LMC, 14 giant stars in Sgr and 14 giants in the Milky Way, based on high-resolutionspectra taken with the spectrograph UVES-FLAMES. The LMC and Sgr stars, in the consideredmetallicity range ([Fe/H] > –1.1 dex), show very similar abundance ratios for almost all the elements,with differences only in the heavy s-process elements Ba, La and Nd, suggesting a different contributionby asymptotic giant branch stars. On the other hand, the two galaxies have chemical patterns clearlydifferent from those measured in the Galactic stars, especially for the elements produced by massivestars. This finding suggests the massive stars contributed less to the chemical enrichment of thesegalaxies with respect to the Milky Way. The derived abundances support similar chemical enrichmenthistories for the LMC and Sgr. Keywords:
Stars: abundances — techniques: spectroscopic — galaxies: Local Group — galaxies:evolution INTRODUCTIONAmong the nearby Local Group galaxies, the closest Milky Way (MW) satellites, namely the Sagittarius (Sgr) dwarfspheroidal galaxy, and the Large and Small Magellanic Clouds (LMC and SMC, respectively), provide the opportunityto investigate the evolution of stellar populations in interacting galaxies. LMC and SMC are two massive ( ∼ and10 M (cid:12) , respectively ) irregular galaxies orbiting each other, forming a triple system with the MW. Sgr is the remnantof a dwarf spheroidal galaxy still merging with the MW. The former case allows to study an ongoing merging eventbetween galaxies with comparable masses, while the latter is a spectacular case of a satellite almost totally disruptedby the tidal field of its (significantly more massive) parent galaxy.LMC and Sgr exhibit some similarities in terms of stellar populations, with their stellar content dominated byan intermediate-age population with similar metallicity. The metallicity distributions of these two galaxies are bothpeaked at [Fe/H] ∼ –0.5/–0.3 dex, as found by several spectroscopic works, see e.g. Pomp´eia et al. (2008), Lapennaet al. (2012), Van der Swaelmen et al. (2013), Song et al. (2017), Nidever et al. (2020) for LMC, and Monaco et al.(2005), Bellazzini et al. (2006), Sbordone et al. (2007), Carretta et al. (2010), McWilliam et al. (2013), Hasselquistet al. (2017), Mucciarelli et al. (2017a) for Sgr. The age range of their dominant populations is ∼ ∼ Corresponding author: Alice [email protected] ∗ Based on observations collected at the ESO-VLT under programs 071.B-0146, 072.B-0293, 072.D-0342, 074.D-0369, 076.D-0381, 078.B-0323,080.D-0368, 081.D-0286, 084.D-0933, 092.D-0244, 188.B-3002, 193.B-0936. a r X i v : . [ a s t r o - ph . GA ] F e b Minelli et al. component accounting for less than ∼
10% of the total stellar content (see e.g. Monaco et al. 2003; Cole et al. 2005;Hamanowicz et al. 2016; Nidever et al. 2020). However, it is important to note that the available spectroscopicmetallicity distributions of Sgr stars sample its central region where the massive metal-poor globular cluster M54 lies.Therefore these distributions are dominated by the stars of M54 at [Fe/H] < -1.2 dex (see, e.g., Mucciarelli et al. (2017a)for further discussions). Moreover, the presence of a metallicity gradient in the main body of Sgr (Hayes et al. 2020)does not allow to observe a representative sample of the whole galaxy focusing only in the central region.The violent interactions between LMC and SMC and between Sgr and MW have significantly impacted on thestellar populations of LMC and Sgr contributing to shape their star formation histories. LMC is likely at its firstperi-Galactic passage with the MW (Shuter 1992; Byrd et al. 1994; Besla et al. 2007) and it experienced significanttidal gas stripping only recently ( ∼ . ∼ ∼ α /Fe] abundance ratios that in the metal-rich stars of both galaxies are lower than those measured among theMW stars of similar [Fe/H], as expected for galaxies with lower star formation efficiencies (Matteucci & Brocato 1990).Also, sub-solar abundance ratios of some iron-peak elements and super-solar abundances for some neutron-captureelements are common features of the metal-rich stars of LMC and Sgr (Pomp´eia et al. 2008; Van der Swaelmen et al.2013). Their similar chemical patterns suggest that they have experienced analogous chemical enrichment historiesand that the progenitor of Sgr could be as massive as the LMC (Niederste-Ostholt et al. 2012; de Boer et al. 2014;Gibbons et al. 2017; Mucciarelli et al. 2017a; Carlin et al. 2018).However, in order to properly highlight similarities and differences between the chemical compositions of the twogalaxies one needs to compare sets of chemical abundances obtained under the same assumptions (see e.g. Reichertet al. 2020). In fact, the adopted model atmospheres, temperature scale, atomic data, solar reference abundances canlead to systematics among different chemical analyses, hampering the possibility of a fully meaningful comparison ofabundance patterns. The comparisons between the chemical patterns of LMC and Sgr performed so far are based onanalyses that adopted different physical assumptions, limiting our capability to highlight real differences or similaritiesand allowing us to provide only a qualitative comparison.In order to bypass this issue, in this study we present a homogeneous and self-consistent chemical analysis of high-resolution spectra for red giant branch (RGB) stars in LMC, Sgr and MW, with the twofold aim of comparing thechemical composition of LMC and Sgr, keeping the MW abundance pattern as a reference. This study is restricted tothe dominant stellar components of the two galaxies, therefore stars with [Fe/H] > –1.0 dex. In particular, we measuredchemical abundances for the main groups of elements (light, alpha, iron-peak, neutron-capture elements) to estimatethe role played to their chemical evolution by massive stars, exploding either as Type II Supernovae (SNe II) or moreenergetic hypernovae (HNe), degenerate binary systems, exploding as Type Ia Supernovae (SNe Ia) and AsymptoticGiant Branch (AGB) stars. SPECTROSCOPIC DATASETSThis paper presents the homogeneous chemical analysis of three samples of high-resolution spectra collected withthe optical spectrograph UVES-FLAMES (Pasquini et al. 2002) mounted at the Very Large Telescope of the EuropeanSouthern Observatory. The observations have been performed adopting the Red Arm 580 UVES setup, with a spectralresolution of 47000 and a spectral coverage between about 4800 and 6800 ˚A. All the spectra have been reduced withthe dedicated ESO pipelines , including bias subtraction, flat-fielding, wavelength calibration, spectral extraction and hemical composition of LMC and Sgr • LMC dataset — It includes 30 RGB stars belonging to the LMC. Eleven of these stars have been originally selectedas possible member stars of some LMC globular clusters but they revealed to be LMC field stars according totheir radial velocity and metallicity (both discrepant with respect to those of the close globular cluster). Thespectra of the other stars have been retrieved from the ESO archive, selecting UVES-FLAMES observationspointed toward the LMC and considering only giant stars with signal-to-noise ratio (SNR) per pixel larger than ∼
20 and with radial velocities between +170 and +380 kms − that is the range of radial velocities of the LMCstars (Zhao et al. 2003; Carrera et al. 2008). The LMC spectra have SNR ranging from ∼
20 to ∼
60 at 6000˚A.The final sample is composed by stars located in different regions of the galaxy, distributed between ∼ ◦ to ∼ ◦ from the LMC center (van der Marel & Cioni 2001). No significant metallicity gradient is expected amongthe LMC stars within this distance from the center because the mean metallicity of the LMC field stars remainsconstant within 6 ◦ from the LMC center (Carrera et al. 2011). • Sgr dataset — This dataset includes UVES-FLAMES spectra of 14 stars belonging to the upper RGB of the mainbody of Sgr. Twelve of these stars have been already discussed by Monaco et al. (2005) that, however, provideonly the abundances of Fe, Mg, Ca and Ti, while the remaining 2 stars are from the UVES-FLAMES sample byCarretta et al. (2010). The study of Monaco et al. (2005) included other 3 RGB stars with [Fe/H] between –1.5and –1.1 dex, all located within 3.2 ◦ from M54 center but only the most metal-poor considered as likely memberof M54. Our chemical analysis, however, suggests that these three stars are likely members of M54, in virtueof their strong enhancement of Na and Al abundances typical of second-generation stars observed in globularcluster-like systems (Bastian & Lardo 2018). Therefore we exclude these stars from our sample, focusing onlyon the metal-rich ([Fe/H] > –1.0 dex) component of Sgr. • MW dataset — We defined a reference sample of 14 giant/sub-giant MW stars selected from Soubiran et al.(2016) and Smiljanic et al. (2016) and covering the same range of metallicity of the LMC/Sgr targets. The starsbelong both to thin and thick disk of the Galaxy, and they have been selected in order to have observations withthe Red Arm 580 UVES setup available in the ESO archive and with low color excess (E(B-V) < eff ) and surface gravities (log g ) for the observed targets have been derivedby using the early third data release of the ESA/Gaia mission (Prusti et al. 2016; Gaia Collaboration et al. 2020) andthe near-infrared 2MASS survey (Skrutskie et al. 2006). Gaia eDR3 photometric parameters — T eff have been calculated by using the (BP − RP) -T eff transformation pro-vided by Mucciarelli & Bellazzini (2020) and based on the infrared flux method T eff estimated by Gonz´alez Hern´andez,& Bonifacio (2009). The transformation was calibrated on Gaia DR2 data, but it remains valid also for the new datarelease. The (BP-RP) colors have been corrected for extinction with an iterative procedure following the scheme pro-posed by Babusiaux et al. (2018). The color excess adopted for the Sgr targets is E(B-V)= 0.14 ± eff relations derived by Mucciarelli &Bellazzini (2020) have a dependence from the stellar metallicity, first we derived T eff adopting [Fe/H]=–0.5 dex for Minelli et al. all the stars (a reasonable value for the LMC/Sgr dominant stellar populations), and subsequently we refined T eff adopting for any star the appropriate metallicity obtained from the chemical analysis.Surface gravities have been calculated by adopting the photometric T eff described above, a stellar mass of 1 M (cid:12) (arepresentative value for the stellar mass of stars belonging to the main LMC and Sgr stellar populations) and theG-band bolometric corrections computed according to Andrae et al. (2018). To transform apparent magnitudes inabsolute magnitudes, we adopted the distance modulus of ( m − M ) = 17 . ± .
15 mag for Sgr (Monaco et al. 2004)and ( m − M ) = 18 . ± .
02 mag for LMC (Alves 2004). For the MW stars, their distances have been derived fromGaia eDR3 parallaxes corrected by the offset (+0.029 mas) provided by Helmi et al. (2018). Only for one star in theMW sample the ratio between parallax and its uncertainty is lower than 10, indicating that the distance errors arenot symmetrical (Bailer-Jones 2015). According to the typical parallax errors, the derived distance errors are of theorder of 0.10 pc. — For most of the targets we adopted the near-infrared photometry provided bythe 2MASS survey but for the LMC targets observed close to globular clusters, for which we used our own SofI@NTTphotometry (that is more precise than 2MASS photometry thanks to the higher spatial resolution) calibrated onto2MASS photometric system. T eff have been obtained using the ( J − K ) -T eff relation provided by Gonz´alez Hern´andez,& Bonifacio (2009) and defined onto 2MASS photometric system, and adopting the same color excesses discussed above.For log g the only difference with respect to the procedure based on the Gaia eDR3 photometry is the computation ofthe K-band bolometric corrections following the prescriptions by Buzzoni et al. (2010).The two sets of parameters are in good agreement for Sgr and MW stars. For the MW targets the mean differencesbetween 2MASS and Gaia eDR3 parameters are –136 K ±
40 ( σ = 150 K) and -0.01 ± σ = 0.09) respectively forT eff and log g, while for Sgr targets are –89 ±
20 K ( σ = 72 K) and –0.050 ± σ = 0.02). Instead, for the LMCtargets the mean differences are –149 ±
74 K ( σ = 405 K) and –0.13 ± σ = 0.31 K). Applying a 3- σ rejection,the mean difference between T eff from 2MASS and Gaia eDR3 decreases down to –100 ±
58 K ( σ = 310 K) but stillwith a significant scatter.An additional clue to validate the photometric parameters (and understand which set of parameters is more correct)is to use the standard spectroscopic constraints, namely, the excitation equilibrium to set T eff (all the Fe I lines providewithin the uncertainties the same abundances regardless of the excitation potential χ ) and the ionization equilibriumto set log g (neutral and single ionized Fe lines provide within the uncertainties the same average abundance). Asdemonstrated by Mucciarelli & Bonifacio (2020), the spectroscopic parameters derived following this approach wellagree with those derived from the photometry for [Fe/H] > –1.5 dex, while at lower metallicities the spectroscopicparameters are systematically biased and they should be avoided (or appropriately corrected following the relationsby Mucciarelli & Bonifacio 2020). All the stars discussed in this work have [Fe/H] > –1.1 dex, hence the spectroscopicmethod can be used to derive the parameters or to check the photometric ones. Therefore, correct parameters shouldprovide null (within the uncertainties) values for both the slope between the Fe I abundance and χ ( σ χ ) and thedifference between the average Fe I and Fe II abundances (∆Fe).T eff from Gaia eDR3 and 2MASS photometries provide values of σ χ that are null (within ± σ ) for almost all theMW and Sgr targets, indicating that the two photometric T eff are reliable. For the LMC stars, T eff from Gaia eDR3photometry are higher than the 2MASS T eff by about 200-250 K and providing significant values of σ χ (at a level of3-4 σ or more), at variance to 2MASS T eff that have σ χ null at a level of 1-2 σ . This difference with the spectroscopicT eff is found also when photometric T eff are estimated adopting the recent relation provided by Casagrande et al.(2020). This suggests that the Gaia eDR3 T eff are over-estimated, for the LMC targets only. We attribute thisdifferent behavior to the high stellar crowding conditions in the LMC, leading to possible problems in the backgroundsubtraction for LMC stars. Spectroscopic parameters — We decide to use spectroscopic parameters for the targets in all the three galaxies,necessary especially for LMC targets due to the issues with the Gaia eDR3 photometry and the large uncertainties inthe 2MASS photometry. In this way we guarantee a homogeneous approach in the determination of the atmosphericparameters for the three samples.An additional hurdle in the spectroscopic determination of the stellar parameters arises from the fact that in giantstars with T eff < eff than Fe I lines and ∆Fe is more sensitive to T eff rather The precise value of the adopted stellar mass does not significantly affect the derived log g because a variation of +1 M (cid:12) leads to a variationof +0.3 in log g . hemical composition of LMC and Sgr g . Therefore, the usual approach to derive T eff from excitation equilibrium and log g from ionizationequilibrium should be revised, because ∆Fe can be cancelled or reduced mainly with small changes in T eff (withoutsignificant changes in σ χ ) and not with large variations in log g. Starting from the photometric parameters, we changedT eff and log g in order to reduce the large ∆Fe observed in some stars and to have simultaneously a value of σ χ nullwithin ± σ .Finally, the microturbolent velocities ξ have been determined by minimizing the slope between the abundances fromFe I lines and the reduced equivalent widths.The final atmospheric parameters are listed in Table 1, together with the coordinates, the 2MASS/SofI and GaiaeDR3 photometry, the color excess and the measured metallicity.CHEMICAL ANALYSISThe lines used to derive the chemical abundances have been selected by comparing the observed spectra with syntheticspectra calculated with the code SYNTHE (Kurucz 2005) in order to evaluate the level of blending for each transition.The synthetic spectra have been calculated using the atomic and molecular data listed in the Kurucz/Castelli linelists and convoluted with a Gaussian profile in order to reproduce the observed broadening. Model atmospheres havebeen calculated for any star with the code ATLAS9 (Kurucz 1993, 2005) and assuming the stellar parameters derivedfrom the Gaia eDR3 (for Sgr and MW) or 2MASS/SofI (for LMC) photometry. Initially we assumed a metallicity of[Fe/H]=–0.5 dex for all the targets. Each linelist has been subsequently refined according to the metallicity and thestellar parameters obtained from the chemical analysis.Chemical abundances for species with unblended lines (Fe, Na, Al, Ca, Ti, Si, Cr, Ni, Zr, Y and Nd) have beenderived from the measured equivalent widths (EWs) of selected lines by using the code GALA (Mucciarelli et al. 2013).EWs have been measured with DAOSPEC (Stetson, & Pancino 2008) through the wrapper 4DAO (Mucciarelli 2013).A visual inspection on the fitted lines has been performed in order to identify possible lines with unsatisfactory fit.For these few lines (less than 1% of the total) the EWs have been re-measured using the IRAF task splot .For the species for which only blended lines (O, Sc, V, Mn, Co, Cu, Ba, La, Eu) or transitions located innoisy/complex spectral regions (Mg, Zn) are available, the chemical abundances have been derived with our owncode SALVADOR that performs a χ -minimization between the observed line and a grid of suitable synthetic spectracalculated on the fly using the code SYNTHE and varying only the abundance of the corresponding element.Atomic data (excitation potential χ , log gf, damping constants and hyperfine/isotopic splitting) for the used linesare from the Kurucz/Castelli database, improved for some specific transitions with more recent or more accurate data(see Mucciarelli et al. 2017b, for some additional references). Solar reference abundances are from Grevesse & Sauval(1998) but for oxygen for which the value quoted by Caffau et al. (2011) is adopted.In the following, we discuss in details the procedure adopted to derive chemical abundances for a few problematicspecies. • Oxygen : only the forbidden line at 6300.3 ˚A is available for this element in the optical range. This spectralregion is contaminated by several telluric lines. For each target we calculated a synthetic spectrum for the Earthtransmission using the code
TAPAS (Bertaux et al. 2014) and in case of contamination of the O line the observedstellar spectrum has been divided by the Earth atmosphere spectrum.Oxygen abundance is derived using spectral synthesis because the forbidden line is blended with a Ni line. Inprinciple, the oxygen abundance can be sensitive to the C and N abundances because of the molecular equilibrium.However, the UVES spectra do not allow to directly measure these abundances and the assumption of specific Cand N abundances for mixed RGB stars is sensitive to metallicity and stellar mass. We thus adopted solar-scaledC and N abundances but we checked how O abundance changes for different assumptions of C and N abundances.Indeed, according to the C and N abundances measured for RGB stars brighter than the RGB Bump in thesegalaxies (see, e.g., Smith et al. (2002) for the LMC, Hasselquist et al. (2017) for Sgr and Gratton et al. (2000)for MW), [C/Fe] is depleted and [N/Fe] is enhanced. Fig. 1 shows for a representative target star the variationof [O/H] as a function of [C/Fe] depletion and corresponding [N/Fe] enhancement. [O/H] is poorly dependenton [N/Fe], while a mild dependence with [C/Fe] is found. In particular a [C/Fe] depletion and a correspondingenhancement of [N/Fe]) by 0.5 dex decreases [O/H] by ∼ Minelli et al.
Figure 1.
Variation of [O/H] as a function of the adopted [C/Fe] and [N/Fe] for a representative star of our sample. • Magnesium : in the optical range the available Mg lines are those at 5528 and 5711 ˚A and the triplet at 6318-6319˚A . The first line is dominated by huge pressure-broadening wings, therefore excluded from our linelist. Thesecond line is often used in chemical analyses of giant stars. On the other hand, this line is heavily saturated(and often insensitive to the Mg abundance) at [Fe/H] > –1.0 dex and low T eff ( < ∼ < –0.9/–0.8 dex) the Mg line at 5711 ˚A is still sensitive to theabundance and it can be safely used. For all the other stars Mg abundances have been derived from the linesat 6318-6319 ˚A , using spectral synthesis because these transitions are located on the red wing of a broadauto-ionization Ca line that affects the continuum location. • Sodium : the two Na doublets used in this work (at 5682-88 ˚A and 6154-60 ˚A ) are both affected by departuresfrom local thermodynamic equilibrium. We applied the suitable NLTE corrections for each line by Lind et al.(2011), of the order of about –0.15 dex for the first doublet and about –0.05 dex for the second one. • Copper : the only available line is that at 5205.5 ˚A (the other optical Cu line, at 5782 ˚A lies in the gap betweenthe two chips of the 580 setup). At the metallicities/temperatures of our targets, the line is already on the flatpart of the curve of growth and basically insensitive to the abundance. Hence, we exclude the abundances of Cufrom our analysis and we discourage to use this Cu line for metal-rich giant stars similar to those analysed here. • Barium : three Ba II lines are available in the spectra, located at 5853.7, 6141.7 and 6496.9 ˚A . The lattertransition provides abundances systematically higher than the other two lines for all the targets. We check theatomic parameters of the three BaII lines on the solar-flux spectrum by Neckel & Labs (1984), and the line6496.9 ˚A provides Ba abundance 0.2 dex higher than the other lines, therefore it has been excluded. hemical composition of LMC and Sgr Figure 2.
Synthetic spectra calculated for a representative giant star with T eff = 4200 K, log g = 1.00 and ξ = 2.00 km/s atthree different metallicities ([Fe/H]=—1.0,–0.5,+0.0 dex, lower, middle and upper panels, respectively), around the Mg line at5711 ˚A and the Mg triplet at 6318-19 ˚A (left and right panels, respectively). For each metallicity, synthetic spectra have beencomputed with different Mg abundances, namely [Mg/Fe]=–0.2 (green lines), =0.0 (black lines), +0.2 (blue lines) and +0.4 dex(red lines). Error Estimates
Abundance uncertainties have been computed by summing in quadrature the error related to the measurementprocess and those arising from the adopted atmospheric parameters. The errors due to the measurement have beenderived according to the method adopted to obtain the abundances.Internal errors relative to the EW measurements have been estimated as the line-to-line scatter divided by the rootmean square of the number of used lines. For the elements for which less than 4 lines are available (namely Al, Na, Yand Zr) we adopt the standard deviation from Fe I lines as more realistic estimate of the line-to-line scatter.O, Mg, Sc, Co, V, Mn, Zn, Ba, La and Eu are the elements whose abundances are derived from spectral synthesis. Theuncertainties of their measurement have been estimated by resorting to Monte Carlo simulation. We created syntheticspectra with representative values for the atmospheric parameters of the analysed stars, and we injected Poisson noiseinto them, according to the SNR of the observed spectra. For each line, 200 noisy spectra have been generated andthe abundance derived adopting the same procedure used for observed spectra. Finally we calculated the internalmeasurement error as the standard deviation of the elemental abundance values derived from the 200 simulations.The uncertainties arising from the atmospheric parameters have been computed by varying one only parameterat a time, keeping the other ones fixed, and deriving the abundance variation. This method provides a conservativeestimate of the uncertainties because it does not take into account the correlations among the parameters. The appliedvariations are of 100 K, 0.1 dex, 0.1 km/s for T eff log g and ξ respectively. The variations correspond to the typicaluncertainties of the atmospheric parameters.Since our results are expressed as abundance ratios, also the uncertainties in the Fe abundance have been taken intoaccount. Therefore the final errors in [Fe/H] and [X/Fe] abundance ratios are calculated as follows: σ [ F e/H ] = (cid:115) σ F e N F e + ( δ T eff F e ) + ( δ log gF e ) + ( δ ηF e ) (1) σ [ X/F e ] = (cid:115) σ X N X + σ F e N F e + ( δ T eff X − δ T eff F e ) + ( δ log gX − δ log gF e ) + ( δ ηX − δ ηF e ) (2)where σ X,F e is the dispersion around the mean of the chemical abundances, N X,F e is the number of lines used toderive the abundances and δ iX,F e are the abundance variations obtained modifying the atmospheric parameter i . Minelli et al.
Table 1.
Main information about the stellar targets.
ID Ra Dec J K G BP RP E(B-V) T eff log g ξ [Fe/H](Degrees) (Degrees) (mag) (mag) (mag) (mag) (mag) (mag) (K) (km/s) (dex) LMC
NGC1754 248 73.58459 -70.43408 14.67 13.72 16.77 17.52 15.87 0.093 4030 1.00 1.5 -0.53NGC1786 2191 74.80183 -67.76463 13.70 12.95 15.53 16.16 14.76 0.074 4400 1.60 1.7 -0.29NGC1786 569 74.82006 -67.74430 14.76 13.88 16.68 17.40 15.85 0.068 4200 1.25 1.7 -0.50NGC1835 1295 76.28288 -69.39264 14.19 13.36 16.11 16.66 15.21 0.069 4200 0.85 1.7 -0.49NGC1835 1713 76.25366 -69.39896 14.12 13.25 16.07 16.73 15.21 0.069 4090 0.80 1.6 -0.58NGC1898 2322 79.16116 -69.65028 14.27 13.29 16.43 17.05 15.42 0.048 3920 0.80 1.5 -0.43NGC1978 24 82.19133 -66.24008 13.82 12.75 18.27 16.99 15.98 0.052 3960 0.60 1.7 -0.56NGC2108 382 86.00623 -69.18082 14.18 13.10 16.33 17.13 15.38 0.132 3920 0.70 2.0 -0.55NGC2108 718 85.96358 -69.19105 14.18 13.16 16.23 17.05 15.35 0.149 3930 0.75 2.1 -0.57NGC2210 1087 92.96237 -69.13304 13.78 12.93 15.79 16.53 14.96 0.062 4100 1.20 1.8 -0.522MASS J06112427-6913117 92.85120 -69.21990 14.33 13.48 16.36 17.12 15.48 0.074 4090 0.90 2.1 -0.982MASS J06120862-6911482 93.03606 -69.19669 14.40 13.38 16.38 17.13 15.55 0.077 4110 0.90 1.8 -0.912MASS J06113433-6904510 92.89313 -69.08083 14.44 13.57 16.34 17.08 15.51 0.060 4100 0.95 1.7 -0.562MASS J06100373-6902344 92.51558 -69.04289 14.49 13.57 16.44 17.20 15.59 0.058 4120 1.05 1.6 -0.622MASS J06122296-6908094 93.09576 -69.13594 14.50 13.63 16.33 17.00 15.55 0.062 4500 1.50 1.7 -0.952MASS J06092022-6908398 92.33421 -69.14439 14.53 13.50 16.58 17.40 15.71 0.065 4080 0.90 1.7 -0.452MASS J06103285-6906230 92.63706 -69.10633 14.53 13.83 16.53 17.11 15.67 0.064 4540 1.60 1.8 -0.332MASS J06122229-6913396 93.09298 -69.22767 14.55 13.56 16.65 17.42 15.77 0.071 4000 0.95 1.9 -0.692MASS J06114042-6905516 92.91859 -69.09769 14.56 13.69 16.61 17.37 15.75 0.060 4050 1.00 1.7 -0.752MASS J06110957-6920088 92.78991 -69.33578 14.58 13.64 16.53 17.26 15.70 0.076 4070 1.00 2.1 -0.632MASS J05244805-6945196 81.20025 -69.75546 14.59 13.58 16.72 17.14 15.68 0.063 4040 0.95 1.5 -0.842MASS J05235925-6945050 80.99690 -69.75140 14.73 13.86 16.78 17.57 15.90 0.049 4150 1.05 2.1 -0.352MASS J05225563-6938342 80.73190 -69.64287 14.78 13.96 16.77 17.29 15.88 0.036 4110 1.10 1.7 -0.262MASS J05242670-6946194 81.11131 -69.77203 14.87 14.01 16.87 17.65 16.02 0.046 4060 1.10 1.8 -0.362MASS J05225436-6951262 80.72653 -69.85732 14.88 14.24 16.97 17.61 16.06 0.091 4220 1.20 2.1 -0.572MASS J05244501-6944146 81.18757 -69.73737 14.96 14.15 16.88 17.64 16.07 0.064 4160 1.20 1.8 -0.722MASS J05235941-6944085 80.99753 -69.73572 15.00 14.51 17.07 17.67 16.29 0.049 4450 1.35 1.7 -0.432MASS J05224137-6937309 80.67245 -69.62527 15.13 14.16 16.94 17.55 16.14 0.030 4320 1.20 1.8 -0.582MASS J06143897-6947289 93.66241 -69.79135 15.51 14.63 17.29 17.96 16.53 0.072 4300 1.40 1.6 -0.332MASS J05224766-6943568 80.69869 -69.73249 15.57 15.19 17.07 17.50 16.35 0.053 4630 1.65 1.8 -0.33
Sgr MW HD749 2.90891 -49.65628 6.05 5.39 7.62 8.15 6.94 0.015 4680 2.70 1.2 -0.40HD18293 (nuHyi) 42.61800 -75.06707 2.53 1.80 4.33 5.01 3.55 0.047 4270 2.25 1.3 0.18HD107328 185.08612 3.31229 2.96 2.20 4.60 5.21 3.84 0.016 4550 2.45 1.8 -0.34HD148897 (* s Her) 247.63937 20.47890 2.95 1.97 4.80 5.50 3.98 0.052 4295 1.20 1.7 -1.08HD190056 301.08188 -32.05636 2.82 2.03 4.57 5.23 3.80 0.153 4375 2.20 1.1 -0.51HD220009 350.08609 5.38104 2.89 1.99 4.65 5.31 3.87 0.054 4410 2.25 1.1 -0.55GES J18242374-3302060 276.09888 -33.03495 10.11 9.42 11.77 12.35 11.04 0.164 4945 3.05 1.6 -0.02GES J18225376-3406369 275.72394 -34.11022 10.73 10.05 12.44 13.02 11.71 0.125 4870 2.95 1.2 -0.12GES J17560070-4139098 269.00287 -41.65274 11.08 10.45 12.94 13.48 12.12 0.204 5015 2.85 1.5 -0.27GES J18222552-3413578 275.60632 -34.23277 11.10 10.37 12.85 13.46 12.10 0.112 4715 3.00 1.2 -0.03GES J02561410-0029286 44.05890 -0.49131 11.49 10.90 13.13 13.60 12.41 0.055 4865 2.95 1.1 -0.71GES J13201402-0457203 200.05844 -4.95570 12.03 11.40 13.59 14.10 12.92 0.038 4875 3.00 1.0 -0.49GES J01203074-0056038 20.12810 -0.93438 12.30 11.56 14.02 14.62 13.28 0.029 4525 2.95 1.2 -0.25GES J14194521-0506063 214.93840 -5.10184 12.55 11.85 14.10 14.64 13.42 0.037 4720 3.10 1.0 -0.33 hemical composition of LMC and Sgr α -, iron-peak, neutron-capture elements) among the metal-rich stars in LMC, Sgr and MW. Althoughthese samples cannot be considered as fully representative of the metallicity distributions of the parent galaxies, inparticular because of some selection bias in their definition (see Section 2), this work has the main advantage to removemost of the systematics (i.e. solar abundances, atomic data, model atmospheres), affecting the comparison of theirabundances.Tables 2-4 list the measured values of the elemental abundances with their error. In Figs. 3-8 we show the resultsobtained for the three samples, together with the abundances in Galactic field stars from the literature (see captionof Figs. 3-8 for references). Only for the works that do not adopt solar values determined with their own linelist, were-scaled their abundances to our solar reference values. The latter measures are shown as a sanity check to verifythat our heterogeneous sample of MW stars reproduces the main MW chemical patterns. Also, the use of both dwarfand giant stars and of different assumptions in the chemical analyses (i. e. atomic data, solar reference values, modelatmospheres, among others) could hamper the direct comparison with the LMC and Sgr abundances derived here. Thecomparison between our abundances and those from the literature is satisfactory for almost all the elements, while wefound offsets of about 0.1-0.2 dex for Na, Al, Co, V and Eu. These differences are mainly explained by the differenttransitions, atomic parameters and (in the case of Na) NLTE corrections adopted by different authors. The existenceof these offsets enforces the importance of a homogeneous analysis for all the stars.In this section we also compare our results with the abundances available in literature, i.e. Pomp´eia et al. (2008),Lapenna et al. (2012), Van der Swaelmen et al. (2013), Nidever et al. (2020) for the LMC and Monaco et al. (2005),Sbordone et al. (2007), Carretta et al. (2010) and Mucciarelli et al. (2017a) for Sgr. Light elements: Na and Al
Na and Al are mainly synthesized in massive stars through the hydrostatic C and Ne burning and only a smallamount is produced during the H burning through the NeNa and MgAl cycles in AGB stars (Woosley & Weaver 1995).Stars in the LMC and Sgr have similar [Na/Fe] and [Al/Fe] abundance ratios that are significantly lower (by 0.5 dex)than those measured in the MW sample (Fig. 3). These low values could suggest that the contribution by massivestars is similar in the two galaxies but significantly lower than that in the MW.Low [Al/Fe] and [Na/Fe] abundances have been measured in Sgr stars also by Sbordone et al. (2007) and McWilliamet al. (2013), even if there are an offset of about -0.2 dex for Al and +0.3 dex for Na with respect to our values thatare likely attributable to the different log gf (as in the case of Al) or NLTE corrections (as in the case of Na). Instead,the Sgr stars analysed by Carretta et al. (2010) exhibit higher [Na/Fe] values. This difference can be only partiallyexplained by the different NLTE corrections for the Na lines. α -elements The α -elements are mainly produced in short-lived massive stars and released in the interstellar medium throughSNe II, with only a minor component produced in SN Ia that produce, instead, significant amounts of Fe on longtimescales. Therefore, [ α /Fe] ratios are used to trace the time-scales of the star formation in a given environment(Tinsley 1979; Matteucci & Brocato 1990; Gilmore & Wyse 1991). We grouped the measured α -elements according totheir formation mechanism: hydrostatic elements (O and Mg) that are synthesized via hydrostatic C and Ne burning,mainly in stars with masses larger than 30-35 M (cid:12) and without contributions by SN Ia, and explosive elements (Si, Caand Ti) that are synthesized via explosive O and Si burning, mainly in stars with masses of 15-25 M (cid:12) (Woosley &Weaver 1995), and in a smaller amount in SN Ia.Fig. 4 shows the behavior of the average abundance ratios of the two groups as a function of [Fe/H]. For both groupsof elements, LMC and Sgr agree each other but with values of [ α /Fe] lower than those measured in MW stars of similar[Fe/H]. This difference is more pronounced for the hydrostatic α -elements. Also, the hydrostatic α -elements show aclear decrease with increasing [Fe/H], reaching sub-solar values at [Fe/H] > –0.6 dex, at variance with the explosiveelements that display a less pronounced decrease by increasing [Fe/H]. It is worth noticing that most of the Sgr starshave [Fe/H] > –0.5 dex and only two stars with [Fe/H] between –1.0 dex and –0.5 dex are in the Sgr sample. However,the abundance ratios for these two stars well match with those of the LMC stars of similar [Fe/H].The low [ α /Fe] ratios measured in LMC/Sgr point out that these stars formed from a gas already enriched by SN Iaat [Fe/H] > –1 dex. Also, the larger difference between LMC/Sgr and MW measured for hydrostatic α -elements is0 Minelli et al.
Figure 3.
Behavior of the light elements [Na/Fe] and [Al/Fe] abundance ratios (left and right panel, respectively) as a functionof [Fe/H] for LMC sample (red circles), Sgr sample (light blue squares) and MW sample (gray triangles). Abundances of Galacticstars from the literature are also plotted as a reference: Edvardsson et al. (1993); Fulbright (2000); Reddy et al. (2003, 2006);Bensby et al. (2005) for both the elements, and Stephens & Boesgaard (2002); Gratton et al. (2003) for Na. consistent with galaxies having a lower number of stars more massive than ∼
30 M (cid:12) , for instance galaxies with a lowerstar formation efficiency (like LMC and Sgr).Comparing our abundances with the literature, no significant differences are found between the α abundances in theLMC sample and the ones derived by Pomp´eia et al. (2008), Lapenna et al. (2012) and Van der Swaelmen et al. (2013).Concerning Sgr, we find a general good agreement with the Mg, Ca and Ti abundances by Monaco et al. (2005) andwith the Mg and Ca abundances by Mucciarelli et al. (2017a). A nice agreement is found also with the abundances bySbordone et al. (2007) but Ti that is lower than our values by ∼ eff . Our O, Si and Ti abundances match those by Carretta et al. (2010), while their Mg are higherthan ours by ∼ α /Fe] ratios show a flat run with [Fe/H], compatible with our result for Si andCa but clearly different concerning Mg. The O and Mg abundances in our MW sample are slightly higher by ∼ Iron-peak elements
The iron-peak elements are the heaviest elements synthesized through thermonuclear reactions. They compose anheterogeneous group of elements in terms of nucleosynthesis. They form partly in massive stars, sometimes with asignificant contribution by HNe (that are associated to stars more massive than ∼ M (cid:12) and more energetic byat least one order of magnitude with respect to normal SNe II). Not negligible amounts of Fe-peak elements can beproduced also in SNe Ia (Leung & Nomoto 2018, 2020; Lach et al. 2020). Moreover, further complicating matters,some of the iron-peak elements have a strong dependence of their yields on the metallicity (see e.g. Romano et al.2010).LMC and Sgr stars exhibit similar abundance patterns for all the measured iron-peak elements, as shown in Fig. 5.Differences with respect to the MW stars are evident for Sc, V, Co, Ni and Zn abundances, showing in the casesof [Sc/Fe] and [Ni/Fe] a clear decrease of the abundance ratios by increasing [Fe/H]. A decreasing trend is also seen hemical composition of LMC and Sgr Figure 4.
Behavior of the hydrostatic and explosive [ α/F e ] abundance ratio (left and right panel, respectively) as a functionof [Fe/H]. Same symbols of Fig. 3. The MW literature data for both groups of elements are from Edvardsson et al. (1993);Gratton et al. (2003); Reddy et al. (2003, 2006); Bensby et al. (2005), while for the explosive elements additional data are fromFulbright (2000); Stephens & Boesgaard (2002); Barklem et al. (2005). in [Zn/Fe] for the LMC sample, but the small number of Sgr stars with Zn measures prevents to properly identify apossible trend with [Fe/H].The largest differences are observed for [V/Fe] and [Zn/Fe], whose values in LMC/Sgr stars are lower by 0.5-0.7dex respect to MW stars of similar metallicity. In contrast, [Cr/Fe] and [Mn/Fe] show values comparable betweenLMC/Sgr and MW stars.Even if the details of the nucleosynthesis of these elements are not fully known and for some of them the currentevolutionary chemical models are not even able to reproduce the observed MW trends (Romano et al. 2010), thechemical patterns obtained for the three samples provide a scenario coherent with that drawn above based on theabundances of light and α -elements. In fact, a large amount of these elements is produced by massive stars, viaSNe II, HNe and electron-capture SNe. The measured abundances in LMC and Sgr stars for most of the iron-peakelements are compatible with a scenario where the contribution by massive stars to the chemical enrichment of theparent galaxies is less important than in the MW. In particular, the low abundances of Zn would suggest a small orlacking contribution by stars more massive than ∼ M (cid:12) , because this element is almost totally produced by HNe(Nomoto et al. 2013), while its production in SNe Ia is probably negligible.As noted above, V and Zn exhibit the largest differences with respect to the MW stars with similar [Fe/H]. Theseabundance ratios are the most clean-cut chemical differences between LMC/Sgr and MW and in principle they couldbe used to distinguish, among the MW stars with [Fe/H] > –1 dex, those formed in smaller satellites that evolvedsimilarly to the LMC/Sgr and were subsequently accreted and disrupted by the MW tidal field. Zn abundances lowerthan those in MW stars of similar metallicity have been measured also in Sculptor (Sk´ulad´ottir et al. 2017) and inother dwarf galaxies (Shetrone et al. 2001, 2003), but at lower metallicities than those discussed here. Slow neutron-capture elements
Elements heavier than Fe are produced through neutron capture processes on seed nuclei (Fe and iron-peak elements),and subsequent β decays (Burbidge et al. 1957). According to the rate of neutron captures with respect to the time-scale of the β decays, we distinguish slow (s-) and rapid (r-)process elements. The s-process elements are groupedaround three peaks of stability corresponding to the neutrons magic numbers (N=50, 82, 126). These elements areproduced mainly by low-mass (1-3 M (cid:12) ) AGB stars (whose yields are strongly metallicity dependent) with only a minorcomponent produced in massive stars (see e.g. Busso et al. 1999).2 Minelli et al.
Figure 5.
Behavior of the iron-peak [Cr/Fe], [Mn/Fe], [V/Fe], [Zn/Fe], [Co/Fe], [Ni/Fe] and [Sc/Fe] abundance ratios as afunction of [Fe/H]. Same symbols of Fig. 3. The MW literature data are from the works of Edvardsson et al. (1993)(Ni),Fulbright (2000) (V, Cr, Ni), Stephens & Boesgaard (2002)(Cr, Ni), Gratton et al. (2003) (Sc, V, Cr, Mn, Ni, Zn), Reddy et al.(2003, 2006) (Sc, V, Cr, Mn, Co, Ni, Zn), Bensby et al. (2005)(Cr, Ni, Zn), Nissen et al. (2007)(Zn)
We measured Y and Zr abundances among the elements belonging to the first-peak. The elements of this group areproduced mainly in AGB stars with high metallicity, because the decrease of the number of neutrons per seed nucleusfavors the formation of the lightest s-process elements (ls). As shown in the first two panels of Fig. 7, the three samplesoverlap each other, even if the large scatter, particularly in [Y/Fe] among the LMC and Sgr stars, makes it hard tocompare these samples with the MW.For the second peak, the heavy s-process elements (hs), we measured Ba, La (that are produced mainly throughs-process) and Nd (that is produced by s-process for nearly 40% of the total, see e.g. Arlandini et al. 1999). The hemical composition of LMC and Sgr < –0.4 dex have [hs/Fe] compatible with thosemeasured in LMC stars, while at higher [Fe/H] these abundance ratios increase significantly, reaching values of about+1 dex. In Fig. 6 we show the profile of the Zr and Ba lines in two pairs of LMC/Sgr stars with similar parametersand metallicity: the stars in the upper panel have similar Zr and Ba abundances, as demonstrated by their similarline strengths, while the the Sgr star shown in the lower panel exhibit Zr and Ba lines stronger than the those of theLMC star with similar parameters and metallicity.The high heavy s-process element abundances measured in the most metal-rich Sgr stars seem to suggest a moresignificant contribution by metal-rich AGB stars in Sgr with respect to LMC. Also, LMC/Sgr stars have abundancesof [hs/Fe] higher than those measured in the MW, where the enhancement is moderate .Our abundances agree with those measured by Van der Swaelmen et al. (2013) for LMC stars and by Sbordone etal. (2007) for Sgr stars, despite some offsets due to the adopted atomic data.In the last panel of Fig. 7 we plot the heavy-to-light s-process abundance ratios as a function of [Fe/H] in order toevaluate the relative contribution of the two groups of s-process elements that mainly arise from AGB stars of differentmetallicity. All the three galaxies shows an increase of this ratio by increasing [Fe/H] with a trend that is steeper inLMC and Sgr. This behaviour points out that the production of s-process elements in these two galaxies is dominatedby AGB stars more metal-poor than in the MW. On the other hand, the production of heavy s-process elements isfavored in less massive AGB stars, while elements of the first peak are produced in a similar amount in AGB starsregardless of their mass (see AGB models of Lugaro et al. 2012; Karakas & Lattanzio 2014). Hence, the higher [hs/ls]ratios observed in LMC and Sgr with respect to the MW could suggest a lower contribution by the most massive AGBstars. Figure 6.
Comparison between the spectra of the two pairs of LMC and Sgr stars (red and blue lines, respectively) with similarstellar parameters and metallicities around the Ba II line at 6142 ˚A . The upper panel shows the comparison between two starswith similar Ba abundances (two Zr lines are also visible in the spectral range), while the lower panel shows the comparisonbetween two stars characterized by a strong difference in both Zr and Ba abundances. We note that in the MW sample, two stars (named HD749 and GES J14194521-0506063) are strongly enhanced in all the s-process elementsabundances. They could be formed through mass transfer in a binary system. The study of the 3D motion using the information from theGaia mission does not highlight anomalies in the kinematics of these stars. Minelli et al.
Figure 7.
Behavior of the slow neutron-capture [Y/Fe], [Zr/Fe], [Ba/Fe], [La/Fe] and [Nd/Fe] as a function of [Fe/H]. In thelast panel the comparison between ls and hs elements, where the ratio between the average value of Ba and La and the averagevalue of Y and Zr is represented as a function of [Fe/H]. Same symbols of Fig. 3. The MW literature data are from Edvardssonet al. (1993, Y, Zr, Ba, Nd), Burris et al. (2000, Y, Zr, Ba, La, Nd), Fulbright (2000, Y, Zr, Ba), Stephens & Boesgaard (2002,Y, Ba), Reddy et al. (2003, Y, Zr, Ba, Nd), Reddy et al. (2006, Y, Ba, Nd), Barklem et al. (2005, Ba), Bensby et al. (2005, Y,Ba), Forsberg et al. (2019, Zr, La). hemical composition of LMC and Sgr Rapid neutron-capture elements
Rapid neutron-capture processes produce an half of the heaviest elements (see e.g. the seminal paper by Burbidgeet al. 1957) but their precise sites of production are still debated, requiring neutron-rich, high energy environments.Among the possible sites, the most promising are low-mass SNII progenitors (in the range 8-10 M (cid:12) see e.g. Wheeleret al. 1998), the neutron star mergers (Pian et al. 2017) and the collapsars (Siegel et al. 2019). We measured theabundance of Eu that is an almost pure r-process element.As shown in the last panel of Fig. 8, both LMC and Sgr exhibit enhanced values of [Eu/Fe], comparable with thoseof the MW. The enhancement of [Eu/Fe] in LMC and Sgr in this range of metallicity has been already measured inprevious works in a few stars (Bonifacio et al. 2000; Van der Swaelmen et al. 2013; McWilliam et al. 2013). A possibledecrease of [Eu/Fe] by increasing [Fe/H] is visible among the LMC stars, while the same pattern is not clearly visiblein Sgr. Comparable enhanced values of [Eu/Fe] in the three samples seem to suggest a similar production of r-processelements in these galaxies, in particular a similar rate of neutron star mergers per unit stellar mass, if neutron starmergers are the main contributors to the Galactic Eu abundances (see e.g. Matteucci et al. 2014).Finally, we evaluate the abundance ratio between heavy s-process elements (considering the average of Ba and Laabundances) and Eu, in order to estimate the contribution of the r-process to the production of other neutron-captureelements. As shown in the last panel of Fig. 8, [hs/Eu] exhibits a rapid increase by increasing [Fe/H] in all the threesamples and in LMC/Sgr this increase occurs at lower metallicities that the MW. Theoretical models by Arlandini etal. (1999) and Burris et al. (2000) predict values of [Ba/Eu] of about –0.5 dex in case of pure r-process. The measured[hs/Fe] abundance ratios suggest that the role played by the r-process to the production of Ba and La decreases byincreasing [Fe/H] and that in the metal-rich stars of LMC and Sgr the production of Ba and La is dominated bys-processes. Figure 8.
In the left panel, behavior of the [Eu/Fe] abundance ratio as a function of [Fe/H]. In the right panel, the ratiobetween the hs elements (average value between Ba and La abundances) and the Eu abundances, as a function of [Fe/H]. Samesymbols of Fig. 3. The MW literature data are from Burris et al. (2000); Fulbright (2000); Reddy et al. (2003, 2006); Barklemet al. (2005); Bensby et al. (2005); Forsberg et al. (2019) for Eu.
CONCLUSIONSHigh-resolution UVES-FLAMES spectra of 30 LMC and 14 Sgr giant stars have been analysed, together with areference sample of 14 MW giant stars selected in the same metallicity range of the LMC/Sgr stars. The three sampleshave been analysed with the same procedure in order to erase the main systematics of the analysis (such as solarreference abundances, atomic data, temperature scales among others). The homogeneous analysis of different samplesof stars is a necessary step to highlight differences and similarities in the chemical compositions of these three galaxies.6
Minelli et al.
The metal-rich populations in LMC and Sgr show strong similarities in almost all the measured species, pointingout a similar chemical evolution. The main differences are related to the heavy s-process elements Ba and La, withthe stars of Sgr more enriched in both the abundance ratios with respect to LMC, suggesting a different contributionby AGB stars. Overall, their similar chemical compositions suggest similar chemical enrichment histories, coherentlywith a scenario where the progenitor of Sgr was a galaxy with a mass and a star formation rate similar to those of theLMC, as already suggested by different authors (see e.g. de Boer et al. 2014; Gibbons et al. 2017; Mucciarelli et al.2017a).The comparison between LMC/Sgr and MW samples reveals that the former galaxies have different chemical abun-dances with respect to the MW stars for almost all the species. This finding agrees with previous works about metal-richstars in LMC (Pomp´eia et al. 2008; Lapenna et al. 2012; Van der Swaelmen et al. 2013) and Sgr (Monaco et al. 2005;Sbordone et al. 2007; Mucciarelli et al. 2017a) but we stress that the present work is the first that allows to directlycompare the abundances of all the main groups of elements in these galaxies. The abundance ratios for elementsproduced by massive stars exploding either as core-collapse SNe or HNe are systematically lower in LMC/Sgr withrespect to the MW, pointing out that in these galaxies the contribution by massive stars to the chemical enrichment isless important. This can be explained in light of their low star formation rates, leading to a lower number of massivestars (poorly populating the IMF at the highest masses, see e.g. Yan et al. 2017; Jeˇr´abkov´a et al. 2018) and penalizingthe elements produced by very massive stars.Finally, we recall that, among the measured elements, the most evident differences between LMC/Sgr and MWstars are measured for [V/Fe] and [Zn/Fe], where LMC/Sgr stars have abundance ratios lower than the MW stars ofsimilar metallicity by as much as 0.5-0.7 dex. We suggest that these abundance ratios can be used to identify possibleextra-galactic interlopers among the Galactic disk stars with [Fe/H] > –1.0 dex, i.e. stars accreted from LMC and Sgror from galaxies that have experienced similar chemical enrichment histories. In other words, we suggest that [V/Fe]and [Zn/Fe] can be tools for a robust chemical tagging as powerful as the classical hydrostatic [ α /Fe] ratios.ACKNOWLEDGMENTSWe are grateful to the anonymous referee for his/her useful suggestions.This work has made use of data from the European Space Agency (ESA) mission Gaia
Gaia
Gaia
Multilateral Agreement.This research is funded by the project ”Light-on-Dark” , granted by the Italian MIUR through contract PRIN-2017K7REXT.A.Minelli would like to thank C. Fanelli for the useful discussions and support.DR benefited from discussions held at the International Space Science Institute (ISSI, Bern, CH) and the InternationalSpace Science Institute–Beijing (ISSI-BJ, Beijing, CN) thanks to the funding of the team “Chemical abundances inthe ISM: the litmus test of stellar IMF variations in galaxies across cosmic time”.REFERENCES
Alves, D. R. 2004, NewAR, 48, 659Andrae, R., Fouesneau, M., Creevey, O., et al. 2018, A&A,616, A8Arlandini, C., K¨appeler, F., Wisshak, K., et al. 1999, ApJ,525, 886. doi:10.1086/307938Bailer-Jones, C. A. L. 2015, PASP, 127, 994.doi:10.1086/683116Barklem, P. S., Christlieb, N., Beers, T. C., et al. 2005,A&A, 439, 129. doi:10.1051/0004-6361:20052967Bastian, N. & Lardo, C. 2018, ARA&A, 56, 83.doi:10.1146/annurev-astro-081817-051839 Bekki, K., Couch, W. J., Beasley, M. A., et al. 2004, ApJL,610, L93Bekki, K., & Chiba, M. 2005, MNRAS, 356, 680Bellazzini, M., Correnti, M., Ferraro, F. R., et al. 2006,A&A, 446, L1Bensby, T., Feltzing, S., Lundstr¨om, I., et al. 2005, A&A,433, 185. doi:10.1051/0004-6361:20040332Bertaux, J. L., Lallement, R., Ferron, S., et al. 2014, A&A,564, A46Besla, G., Kallivayalil, N., Hernquist, L., et al. 2007, ApJ,668, 949 hemical composition of LMC and Sgr Bonifacio, P., Hill, V., Molaro, P., et al. 2000, A&A, 359,663Bonifacio, P., Sbordone, L., Marconi, G., et al. 2004, A&A,414, 503Burbidge, E. M., Burbidge, G. R., Fowler, W. A., et al.1957, Reviews of Modern Physics, 29, 547.doi:10.1103/RevModPhys.29.547Burton, W. B., & Lockman, F. J. 1999, A&A, 349, 7Burris, D. L., Pilachowski, C. A., Armandroff, T. E., et al.2000, ApJ, 544, 302. doi:10.1086/317172Busso, M., Gallino, R., & Wasserburg, G. J. 1999, ARA&A,37, 239. doi:10.1146/annurev.astro.37.1.239Buzzoni, A., Patelli, L., Bellazzini, M., et al. 2010,MNRAS, 403, 1592Byrd, G., Valtonen, M., McCall, M., et al. 1994, AJ, 107,2055Caffau, E., Ludwig, H.-G., Steffen, M., et al. 2011, SoPh,268, 255Carlin, J. L., Sheffield, A. A., Cunha, K., et al. 2018, ApJL,859, L10Carrera, R., Gallart, C., Hardy, E., et al. 2008, AJ, 135, 836Carrera, R., Gallart, C., Aparicio, A., et al. 2011, AJ, 142,61. doi:10.1088/0004-6256/142/2/61Carretta, E., Bragaglia, A., Gratton, R. G., et al. 2010,A&A, 520, A95Casagrande, L., Lin, J., Rains, A. D., et al. 2020,arXiv:2011.02517Cole, A. A., Tolstoy, E., Gallagher, J. S., et al. 2005, AJ,129, 1465de Boer, T. J. L., Belokurov, V., Beers, T. C., et al. 2014,MNRAS, 443, 658de Boer, T. J. L., Belokurov, V., & Koposov, S. 2015,MNRAS, 451, 3489. doi:10.1093/mnras/stv946Edvardsson, B., Andersen, J., Gustafsson, B., et al. 1993,A&A, 500, 391Forsberg, R., J¨onsson, H., Ryde, N., et al. 2019, A&A, 631,A113. doi:10.1051/0004-6361/201936343Fulbright, J. P. 2000, AJ, 120, 1841. doi:10.1086/301548Gaia Collaboration, Prusti, T., de Bruijne, J. H. J., et al.2016, A&A, 595, A1Gaia Collaboration, Babusiaux, C., van Leeuwen, F., et al.2018, A&A, 616, A10Gaia Collaboration, Helmi, A., van Leeuwen, F., et al.2018, A&A, 616, A12Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al.2020, arXiv:2012.01533Gibbons, S. L. J., Belokurov, V., & Evans, N. W. 2017,MNRAS, 464, 794Gilmore, G. & Wyse, R. F. G. 1991, ApJL, 367, L55.doi:10.1086/185930 Gonz´alez Hern´andez, J. I., & Bonifacio, P. 2009, A&A, 497,497Gratton, R. G., Sneden, C., Carretta, E., et al. 2000, A&A,354, 169Gratton, R. G., Carretta, E., Claudi, R., et al. 2003, A&A,404, 187. doi:10.1051/0004-6361:20030439Grevesse, N., & Sauval, A. J. 1998, SSRv, 85, 161Guglielmo, M., Lewis, G. F., & Bland-Hawthorn, J. 2014,MNRAS, 444, 1759Hamanowicz, A., Pietrukowicz, P., Udalski, A., et al. 2016,AcA, 66, 197Harris, J., & Zaritsky, D. 2009, AJ, 138, 1243Hasselquist, S., Shetrone, M., Smith, V., et al. 2017, ApJ,845, 162Hayes, C. R., Majewski, S. R., Hasselquist, S., et al. 2020,ApJ, 889, 63. doi:10.3847/1538-4357/ab62adJeˇr´abkov´a, T., Hasani Zonoozi, A., Kroupa, P., et al. 2018,A&A, 620, A39. doi:10.1051/0004-6361/201833055Karakas, A. I. & Lattanzio, J. C. 2014, PASA, 31, e030.doi:10.1017/pasa.2014.21Koribalski, B., Johnston, S., & Otrupcek, R. 1994,MNRAS, 270, L43Kurucz, R. L. 1993, VizieR Online Data Catalog, VI/39Kurucz, R. L. 2005, Memorie della Societa AstronomicaItaliana Supplementi, 8, 14Lach, F., Roepke, F. K., Seitenzahl, I. R., et al. 2020,arXiv:2010.14084Lapenna, E., Mucciarelli, A., Origlia, L., et al. 2012, ApJ,761, 33Laporte, C. F. P., Minchev, I., Johnston, K. V., et al. 2019,MNRAS, 485, 3134. doi:10.1093/mnras/stz583Layden, A. C., & Sarajedini, A. 2000, AJ, 119, 1760Leung, S.-C. & Nomoto, K. 2018, ApJ, 861, 143.doi:10.3847/1538-4357/aac2dfLeung, S.-C. & Nomoto, K. 2020, ApJ, 888, 80.doi:10.3847/1538-4357/ab5c1fLind, K., Asplund, M., Barklem, P. S., et al. 2011, A&A,528, A103Lugaro, M., Karakas, A. I., Stancliffe, R. J., et al. 2012,ApJ, 747, 2. doi:10.1088/0004-637X/747/1/2Matteucci, F. & Brocato, E. 1990, ApJ, 365, 539.doi:10.1086/169508Matteucci, F., Romano, D., Arcones, A., et al. 2014,MNRAS, 438, 2177. doi:10.1093/mnras/stt2350McWilliam, A., Wallerstein, G., & Mottini, M. 2013, ApJ,778, 149. doi:10.1088/0004-637X/778/2/149Monaco, L., Bellazzini, M., Ferraro, F. R., et al. 2003,ApJL, 597, L25Monaco, L., Bellazzini, M., Ferraro, F. R., et al. 2004,MNRAS, 353, 874 Minelli et al.
Monaco, L., Bellazzini, M., Bonifacio, P., et al. 2005, A&A,441, 141Mucciarelli, A. 2013, arXiv e-prints, arXiv:1311.1403Mucciarelli, A., Pancino, E., Lovisi, L., et al. 2013, ApJ,766, 78Mucciarelli, A., Bellazzini, M., Ibata, R., et al. 2017, A&A,605, A46Mucciarelli, A., Monaco, L., Bonifacio, P., et al. 2017,A&A, 603, L7Mucciarelli, A., & Bellazzini, M. 2020, Research Notes ofthe American Astronomical Society, 4, 52Mucciarelli, A., & Bonifacio, P., 2020, A&A, 640, A87Neckel, H. & Labs, D. 1984, SoPh, 90, 205.doi:10.1007/BF00173953Nidever, D. L., Hasselquist, S., Hayes, C. R., et al. 2020,ApJ, 895, 88. doi:10.3847/1538-4357/ab7305Niederste-Ostholt, M., Belokurov, V., & Evans, N. W.2012, MNRAS, 422, 207Nissen, P. E., Akerman, C., Asplund, M., et al. 2007, A&A,469, 319. doi:10.1051/0004-6361:20077344Nomoto, K., Kobayashi, C., & Tominaga, N. 2013,ARA&A, 51, 457.doi:10.1146/annurev-astro-082812-140956Pasquini, L., Avila, G., Blecha, A., et al. 2002, TheMessenger, 110, 1Pian, E., D’Avanzo, P., Benetti, S., et al. 2017, Nature,551, 67. doi:10.1038/nature24298Pomp´eia, L., Hill, V., Spite, M., et al. 2008, A&A, 480, 379Reddy, B. E., Tomkin, J., Lambert, D. L., et al. 2003,MNRAS, 340, 304. doi:10.1046/j.1365-8711.2003.06305.xReddy, B. E., Lambert, D. L., & Allende Prieto, C. 2006,MNRAS, 367, 1329. doi:10.1111/j.1365-2966.2006.10148.xReichert, M., Hansen, C. J., Hanke, M., et al. 2020, A&A,641, A127. doi:10.1051/0004-6361/201936930Romano, D., Karakas, A. I., Tosi, M., et al. 2010, A&A,522, A32Rubele, S., Kerber, L., Girardi, L., et al. 2012, A&A, 537,A106. doi:10.1051/0004-6361/201117863Ruiz-Lara, T., Gallart, C., Bernard, E. J., et al. 2020,Nature Astronomy, 4, 965. doi:10.1038/s41550-020-1097-0 Sbordone, L., Bonifacio, P., Buonanno, R., et al. 2007,A&A, 465, 815Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103Shetrone, M. D., Cˆot´e, P., & Sargent, W. L. W. 2001, ApJ,548, 592. doi:10.1086/319022Shetrone, M., Venn, K. A., Tolstoy, E., et al. 2003, AJ, 125,684. doi:10.1086/345966Shuter, W. L. H. 1992, ApJ, 386, 101Siegel, D. M., Barnes, J., & Metzger, B. D. 2019, Nature,569, 241. doi:10.1038/s41586-019-1136-0Skowron, D. M., Skowron, J., Udalski, A., et al. 2020,arXiv:2006.02448Skrutskie, M. F., et al., 2006, AJ, 131, 1163Sk´ulad´ottir, ´A., Tolstoy, E., Salvadori, S., et al. 2017, A&A,606, A71. doi:10.1051/0004-6361/201731158Smiljanic, R., Romano, D., Bragaglia, A., et al. 2016, A&A,589, A115Smith, V. V., Hinkle, K. H., Cunha, K., et al. 2002, AJ,124, 3241Song, Y.-Y., Mateo, M., Walker, M. G., et al. 2017, AJ,153, 261Soubiran, C., Le Campion, J.-F., Brouillet, N., et al. 2016,A&A, 591, A118Stephens, A. & Boesgaard, A. M. 2002, AJ, 123, 1647.doi:10.1086/338898Stetson, P. B., & Pancino, E. 2008, PASP, 120, 1332Tinsley, B. M. 1979, ApJ, 229, 1046van der Marel, R. P. & Cioni, M.-R. L. 2001, AJ, 122, 1807.doi:10.1086/323099Van der Swaelmen, M., Hill, V., Primas, F., et al. 2013,A&A, 560, A44Wheeler, J. C., Cowan, J. J., & Hillebrandt, W. 1998,ApJL, 493, L101. doi:10.1086/311133Woosley, S. E. & Weaver, T. A. 1995, ApJS, 101, 181Yan, Z., Jerabkova, T., & Kroupa, P. 2017, A&A, 607,A126. doi:10.1051/0004-6361/201730987Zhao, H., Ibata, R. A., Lewis, G. F., et al. 2003, MNRAS,339, 701. doi:10.1046/j.1365-8711.2003.06208.x hemical composition of LMC and Sgr T a b l e . L M C a ndS g r c h e m i c a l a bund a n c e s I D [ F e H ] e rr [ F e II H ] e rr [ N a F e ] e rr [ A l F e ] e rr [ O F e ] e rr [ M g F e ] e rr [ S i F e ] e rr [ C a F e ] e rr [ T i F e ] e rr [ S c F e ] e rr [ V F e ] e rr L M C N G C - . . - . . - . . - . . . . - . . . . - . . - . . - . . - . . N G C - . . - . . - . . - . . . . - . . . . - . . - . . - . . - . . N G C - . . - . . - . . - . . . . . . . . - . . . . - . . - . . N G C - . . - . . - . . . . . . . . . . . . - . . - . . - . . N G C - . . - . . - . . - . . . . . . . . - . . - . . - . . N G C - . . - . . - . . - . . . . - . . . . - . . - . . - . . - . . N G C - . . - . . - . . - . . . . - . . . . . . . . - . . - . . N G C - . . - . . - . . . . . . - . . . . . . . . - . . - . . N G C - . . - . . - . . . . . . . . . . . . - . . - . . - . . N G C - . . - . . - . . - . . . . . . - . . - . . - . . - . .
20 2 M A SS J - - . . - . . - . . - . . . . . . . . . . - . . - . . . .
18 2 M A SS J - - . . - . . - . . - . . . . . . . . . . - . . - . . - . .
17 2 M A SS J - - . . - . . - . . . . . . . . . . - . . - . . - . . - . .
16 2 M A SS J - - . . - . . - . . - . . . . . . . . - . . - . . - . . - . .
17 2 M A SS J - - . . - . . - . . . . . . . . . . . . - . . . .
12 2 M A SS J - - . . - . . - . . - . . . . - . . . . . . . . - . . - . .
18 2 M A SS J - - . . - . . - . . - . . . . . . . . . . - . . - . . - . .
13 2 M A SS J - - . . - . . - . . . . . . . . - . . - . . - . . - . .
21 2 M A SS J - - . . - . . - . . - . . . . - . . . . . . - . . - . . . .
18 2 M A SS J - - . . - . . - . . . . . . . . . . - . . - . . - . .
17 2 M A SS J - - . . - . . - . . . . . . . . - . . . . - . . - . .
19 2 M A SS J - - . . - . . - . . . . . . - . . . . . . . . - . . - . .
16 2 M A SS J - - . . - . . - . . . . - . . . . - . . - . . - . . - . .
18 2 M A SS J - - . . - . . - . . . . . . . . . . - . . - . . - . .
17 2 M A SS J - - . . - . . - . . - . . . . - . . . . - . . - . . - . . - . .
18 2 M A SS J - - . . - . . - . . - . . . . . . - . . - . . - . . - . .
19 2 M A SS J - - . . - . . - . . - . . . . - . . . . - . . - . . - . . - . .
18 2 M A SS J - - . . - . . - . . - . . . . - . . . . . . - . . - . . - . .
16 2 M A SS J - - . . - . . - . . - . . . . . . - . . . . - . . - . .
15 2 M A SS J - - . . - . . - . . . . - . . . . . . - . . - . . - . . S g r - . . - . . - . . - . . . . . . . . . . . . - . . - . .
19 2300196 - . . - . . - . . - . . . . . . . . . . - . . - . .
18 2300215 - . . - . . - . . - . . - . . - . . . . . . - . . - . . - . .
16 2409744 - . . - . . - . . - . . . . . . . . . . - . . - . .
20 3600230 - . . - . . - . . - . . . . - . . - . . - . . - . . - . .
18 3600262 - . . - . . - . . - . . . . - . . . . . . - . . - . . - . .
18 3600302 - . . - . . - . . - . . . . - . . . . . . - . . - . . - . .
19 3800318 - . . - . . - . . . . . . - . . . . . . . . - . . - . .
17 3800558 - . . - . . - . . - . . . . - . . . . . . - . . - . . - . .
10 4214652 - . . - . . - . . - . . . . - . . - . . . . - . . - . . - . .
17 4303773 - . . - . . - . . . . - . . . . - . . - . . - . . - . .
18 4304445 - . . - . . - . . - . . . . - . . . . - . . - . . - . . - . .
17 4402285 - . . - . . - . . - . . . . . . . . - . . - . . - . . - . .
17 4408968 - . . - . . - . . - . . - . . - . . . . - . . - . . - . . - . . Minelli et al. T a b l e . L M C a ndS g r c h e m i c a l a bund a n c e s I D [ C r F e ] e rr [ M n F e ] e rr [ C o F e ] e rr [ N i F e ] e rr [ Z n F e ] e rr [ Y F e ] e rr [ Z r F e ] e rr [ B a F e ] e rr [ L a F e ] e rr [ N d F e ] e rr [ E u F e ] e rr L M C N G C - . . - . . - . . - . . - . . - . . - . . . . - . . . . . . N G C - . . - . . - . . - . . - . . - . . - . . . . . . . . . . N G C - . . - . . - . . - . . - . . - . . . . . . . . . . . . N G C - . . - . . - . . - . . - . . - . . . . . . . . . . N G C - . . - . . - . . - . . - . . . . . . - . . . . . . N G C - . . - . . - . . - . . - . . - . . . . . . . . . . N G C . . - . . - . . - . . - . . . . . . . . . . . . . . N G C . . - . . - . . - . . - . . . . . . . . . . . . N G C - . . - . . . . - . . - . . - . . - . . . . . . . . . . N G C - . . - . . - . . - . . - . . - . . . . . . . . . . . . M A SS J - - . . - . . - . . - . . - . . - . . . . - . . . . . . . . M A SS J - - . . - . . - . . - . . - . . - . . . . . . . . . . . . M A SS J - - . . - . . - . . - . . - . . - . . - . . . . . . . . . . M A SS J - - . . - . . - . . - . . - . . - . . - . . . . . . . . . . M A SS J - . . - . . - . . - . . . . . . . . . . . . M A SS J - . . - . . - . . - . . - . . . . . . . . . . . . M A SS J - - . . - . . - . . - . . - . . - . . - . . . . . . . . . . M A SS J - - . . - . . - . . - . . - . . - . . - . . - . . . . . . . . M A SS J - - . . - . . - . . - . . - . . - . . - . . . . . . . . . . M A SS J - - . . - . . . . - . . - . . . . . . . . . . . . M A SS J - . . - . . - . . - . . - . . . . . . . . . . . . M A SS J - . . - . . - . . - . . - . . . . - . . . . . . M A SS J - - . . - . . - . . - . . - . . - . . . . . . . . . . M A SS J - - . . . . - . . - . . - . . . . . . . . . . . . M A SS J - - . . - . . - . . - . . - . . . . . . - . . . . . . M A SS J - - . . - . . - . . - . . - . . . . . . . . . . . . M A SS J - . . - . . - . . - . . . . . . . . . . . . M A SS J - - . . - . . - . . - . . . . . . . . . . . . M A SS J - - . . - . . - . . - . . - . . . . . . . . M A SS J - . . - . . - . . - . . . . . . . . S g r - . . - . . - . . - . . - . . . . . . . . . . . . . . - . . - . . - . . - . . . . . . . . . . . . . . . . - . . - . . - . . . . . . . . . . . . . . . . - . . - . . - . . . . . . . . . . . . . . - . . - . . - . . - . . - . . . . . . . . . . . . - . . - . . - . . - . . - . . - . . . . . . . . . . . . . . - . . - . . - . . - . . . . . . . . . . . . . . . . . . - . . - . . . . . . . . . . . . . . - . . - . . - . . - . . - . . - . . - . . . . . . . . . . - . . - . . - . . - . . - . . - . . . . . . . . . . . . - . . - . . - . . - . . - . . - . . - . . . . - . . . . . . - . . - . . - . . - . . - . . - . . . . . . . . . . . . - . . - . . - . . - . . - . . - . . . . . . . . . . . . - . . - . . - . . - . . - . . . . - . . . . . . . . . . hemical composition of LMC and Sgr T a b l e . M W c h e m i c a l a bund a n c e s I D [ F e H ] e rr [ F e II H ] e rr [ N a F e ] e rr [ A l F e ] e rr [ O F e ] e rr [ M g F e ] e rr [ S i F e ] e rr [ C a F e ] e rr [ T i F e ] e rr [ S c F e ] e rr [ V F e ] e rr H D - . . - . . - . . . . . . . . . . . . . . . . . . H D ( nu H y i ) . . . . - . . . . . . . . - . . - . . . . - . . H D - . . - . . - . . . . . . . . . . . . . . . . H D ( * s H e r) - . . - . . - . . . . . . . . . . . . . . . . . . H D - . . - . . - . . . . . . . . . . . . . . . . H D - . . - . . - . . . . . . . . . . . . . . . . G E S J - - . . - . . - . . . . . . - . . - . . . . - . . . . . . G E S J - - . . - . . - . . . . . . . . . . . . . . . . . . G E S J - - . . - . . - . . . . . . . . . . . . . . . . . . G E S J - - . . - . . - . . . . . . . . . . . . - . . . . . . G E S J - - . . - . . - . . . . . . . . . . . . . . . . . . G E S J - - . . - . . - . . . . . . . . . . . . . . . . . . G E S J - - . . - . . - . . . . . . . . . . . . . . . . G E S J - - . . - . . . . . . . . . . . . . . . . . . I D [ C r F e ] e rr [ M n F e ] e rr [ C o F e ] e rr [ N i F e ] e rr [ Z n F e ] e rr [ Y F e ] e rr [ Z r F e ] e rr [ B a F e ] e rr [ L a F e ] e rr [ N d F e ] e rr [ E u F e ] e rr H D . . - . . . . . . . . . . . . . . . . . . . . H D ( nu H y i ) - . . . . . . . . - . . . . - . . . . . . . . . . H D - . . - . . . . - . . - . . - . . - . . - . . - . . - . . . . H D ( * s H e r) - . . - . . . . - . . - . . - . . . . . . . . . . . . H D - . . - . . . . . . . . - . . . . . . . . . . . . H D - . . - . . . . - . . . . - . . . . . . . . . . . . G E S J - - . . - . . . . - . . - . . . . . . . . . . . . . . G E S J - - . . - . . . . - . . . . - . . . . . . . . - . . . . G E S J - - . . - . . . . - . . . . . . - . . - . . - . . . . G E S J - - . . - . . . . - . . - . . - . . - . . . . . . . . . . G E S J - - . . - . . . . - . . . . . . . . . . . . . . . . G E S J - . . - . . . . . . . . - . . . . . . . . . . . . G E S J - - . . - . . . . . . . . . . . . . . . . . . . . G E S J - - . . - . . . . . . . . . . . . . . . . . . . ..