Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://star.arm.ac.uk/preprints/2010/570.pdf Äàòà èçìåíåíèÿ: Tue Jan 4 17:26:56 2011 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 02:56:16 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï

Mon. Not. R. Astron. Soc. 000, 1­13 (2010)

Printed 4 January 2011

A (MN LTEX style file v2.2)

A detailed spectroscopic analysis of the open cluster NGC 5460
L. Fossati,1,2 C.P. Folsom,3 S. Bagnulo,3 J.H. Grunhut,4 O. Kochukhov, J.D. Landstreet,3,6 C. Paladini,1 G.A. Wade,4 1
2 3 4 5 6

5

Institut fur Astronomie, Universitat Wien, Turkenschanzstrasse 17, 1180 Wien, Austria ¨ ¨ ¨ Department of Physics and Astronomy, Open University, Walton Hall, Milton Keynes MK7 6AA, UK Armagh Observatory, College Hill, Armagh BT61 9DG, Northern Ireland, UK Physics Dept., Royal Military College of Canada, PO Box 17000, Station Forces, K7K 4B4, Kingston, Canada Department of Physics and Astronomy, Uppsala University, 751 20, Uppsala, Sweden Department of Physics & Astronomy, University of Western Ontario, London, N6A 3K7, Ontario, Canada

ABSTRACT

Within the context of a large project aimed at studying early F-, A- and late B-type stars we present the abundance analysis of the photospheres of 21 members of the open cluster NGC 5460, an intermediate age cluster (log t 8.2) previously unstudied with spectroscopy. Our study is based on medium and high resolution spectra obtained with the FLAMES instrument of the ESO/VLT. We show that cluster members have a nearly solar metallicity, and that there is evidence that the abundances of magnesium and iron are correlated with the effective temperature, exhibiting a maximum around T eff =10500 K. No correlations are found between abundances and projected equatorial velocity, except for marginal evidence of barium being more abundant in slower than in faster rotating stars. We discovered two He-weak stars, and a binary system where the hotter component is a HgMn star. We provide new estimates for the cluster distance (720 ± 50 pc), age (log t=8.2 ± 0.1), and mean radial velocity (-17.9 ± 5.2 km s-1 ). Key words: open clusters and associations: individual: NGC 5460 ­ stars: abundances

1

INTRODUCTION

The photospheres of main sequence early F-, A- and late B-type stars display the signatures of different physical effects of comparable magnitudes, including large and relatively simple magnetic fields, strong surface convection, the presence of emission lines, pulsation, diffusion of chemical elements under the competing influences of gravity and radiative acceleration, and various kinds of mixing processes, from small-scale turbulence to global circulation currents. Detailed studies of these stars are useful to explore and understand the physics of these various phenomena, most of which certainly play a role in the atmosphere of all the other kinds of stars across the Hertzsprung-Russell diagram, but are often more difficult to observe. This work is a part of a large project aimed at setting constraints to the diffusion theory, by means of a systematic study of the abundances of the chemical elements of early F-, A- and late Btype star that are cluster members. Cluster members are specially interesting mainly for two reasons: 1) stars belonging to the same cluster were born from the same initial composition, and the comparison of the abundances of the chemical elements that compose

their atmosphere may help to characterise how diffusion works for different stellar masses and effective temperatures; 2) the age of cluster members can be determined with much higher accuracy than for field stars, hence a systematic study of the abundances of the chemical elements of F-, A- and B-type stars belonging to different clusters may help to identify the characteristic time-scales of diffusion. Using several instruments (FLAMES at ESO/VLT, FIES at the Nordic Optical Telescope, ELODIE and SOPHIE at the Observatorie de Haute Provence, and ESPaDOnS at the CHFT), we have obtained low, medium, and high resolution spectra for about one thousand stars in the fields of view of a dozen of open clusters with age ranging from log t=6.8 to 8.9, and with distance modulus between 6.4 and 14.2 (see Fossati et al. 2008a, for the full list of observed open clusters). One of our ongoing projects based on these data is to perform accurate abundance analysis of early G, F, A, and late B-type cluster members, searching for a correlation between age and chemical composition of the stellar photospheres. Analysis of individual clusters will be used to search for correlations between chemical composition, stellar temperature and stellar rotation. This project was started with a detailed study of the Praesepe cluster (Fossati et al. 2007, 2008b, 2010). Similar works performed by other groups include the analysis of Coma Berenices (Gebran & Monier 2008), the Pleiades (Gebran et al. 2008), the Hyades (Gebran et al. 2010) and NGC 6475 (Villanova et al. 2009).

Based on observations made with ESO Telescopes at the Paranal Observatory under programme ID 079.D-0178A l.fossati@open.ac.uk


2

L. Fossati et al.
ment dedicated pipelines. In this work we used the low and midresolution GIRAFFE data as reduced by ESO1 . Due to the high precision required for our investigation, we decided to repeat the data reduction of the UVES data. For this purpose we used the UVES standard ESO-pipeline vers. 2.9.7 developed on the MIDAS platform vers. 08SEP patch level pl1.1. The pipeline included all the main steps of calibration: bias subtraction, flat field correction and wavelength calibration.

The analysis of the open cluster NGC 5460 is the subject of this paper.

2

OBSERVATIONS

NGC 5460 is an intermediate age open cluster (log t=8.20±0.20, Ahumada & Lapasset 2007)) that includes a large number of earlytype stars, covering a large range in stellar mass (up to about 4 M ). NGC 5460 is spectroscopically almost completely unstudied, except for a few polarimetric observations (Bagnulo et al. 2006). NGC 5460 was observed in service mode with FLAMES, the multiobject spectrograph attached to the Unit 2 Kueyen of the ESO/VLT, on May 22 2007. The FLAMES instrument (Pasquini et al. 2002) is able to access targets over a field of view of 25 arcmin in diameter. With its MEDUSA fibers linked to the GIRAFFE spectrograph, FLAMES can acquire low or medium resolution spectra (R = 7 500 - 30 000) of up to 130 stars. GIRAFFE low resolution spectra cover wavelength ranges of 500 up to 1200 å, while mid resolution spectra cover wavelength intervals of 170 ­ 500 å, all within the visible range 3700­9000 å. With the UVES fibers linked to the red arm of the UVES spectrograph, FLAMES can simultaneously obtain high resolution spectra (R = 47 000) for up to 8 stars, covering an interval range of about 2000 å, centred about 5200, or 5800, or 8600 å. 2.1 Instrument setups

2.2

Target selection

Taking into account that for the stellar parameter determination it is valuable to have observations of the hydrogen lines, and that the number of spectral lines (and the number of chemical elements that may be analysed) increases towards shorter wavelengths, for FLAMES/UVES we adopted the 520 nm setting, which covers the wavelength region 4140­6210 å. The telescope was guided at the central wavelength of this setting (5200 å). To minimise light losses due to the atmosphere differential refraction, for GIRAFFE we were forced to choose a setting with central wavelength not too far from 5200 å. We selected the HR09 setting, which, with a resolving power of 25 900, covers the wavelength region 5139­5355 å. Spectra obtained with this setting cover lines of a number of Fe-peak elements, such as iron, titanium and chromium, in addition to the Mg triplet at 5167, 5172 and 5183. UVES and GIRAFFE spectra were being taken simultaneously. However, due to its lower spectral resolution, the GIRAFFE setting required a substantially shorter exposure time than the UVES setting (for a given signal to noise ratio per spectral bin). While UVES shutter was still open, the extra-time that was left available for observations with GIRAFFE was used to acquire an additional mid resolution spectrum, using setting HR11, and a low resolution spectrum, adopting setting LR03. The setting HR11 covers the wavelength range 5592­5838 å with a resolving power of 24 200. Spectra obtained with this setting cover iron, sodium and scandium lines, where the latter are important to classify Atype stars as Am chemically peculiar stars. Setting LR03 covers the spectral range 4500­5077 å with a resolving power of 7 500. This setting was selected mainly to cover the H line, used as primary indicator for the atmospheric parameters. As to make sure to avoid saturation of the brightest targets, UVES observations were divided into four sub-exposures, while for each of the GIRAFFE settings we obtained two sub-exposures. Our data were obtained in service mode, and the package released to us included the products reduced by ESO with the instru-

Our target list is the result of three selection processes. The first selection was performed when the FLAMES observing blocks were prepared. The position of the Medusa fibers was determined by the Fiber Positioning System Software (FPOSS), which was fed with an input catalogue extracted from the UCAC2 catalogue (Zacharias et al. 2004) centred about (RA DEC) = (14:07:20 -48:20:36). FPOSS maximises the number of observed stars, offering to the user the possibility to set different priorities to the targets. To the input catalogue we attached a priority list based on information obtained from the literature. We tried to give highest priority to stars of spectral type earlier than late G5, and they were deemed as probable cluster members. Cluster membership was originally assessed using the works by Hog et al. (2000) Kharchenko et al. (2004), Kharchenko et al. (2005), and Dias et al. (2006), all based mainly on the star's apparent proper motions. UVES fibers were positioned on four stars that, from previous literature (e.g. Landstreet et al. 2007), we suspected to be chemically peculiar with low sin i values. The remaining four UVES fibers were placed to obtain sky spectra, as requested by FPOSS. We note that the number of MEDUSA fibers was larger than the number of targets in our list of probable members, therefore we decided to place all remaining free fibers on targets more or less randomly selected in the instrument field of view. For this reason the final number of observed targets was larger than our list of probable members After the observations were obtained, we performed a preliminary spectral analysis of all observed targets, and derived an estimate of the effective temperature, sin i, and radial velocities (see Sect. 3) from a quick analysis of the H line observed with the GIRAFFE LR03 setting. We then considered all targets with T eff >5400 K and radial velocity between -35 and 5 km s-1 , and which were included in our original higher priority list because, based on proper motion criteria, were deemed probable cluster members. The radial velocity range [-35,-5] km s-1 was chosen based on the previously published value of the mean cluster radial velocity (-12.70 ± 2.18 km s-1 Kharchenko et al. 2005) and on our preliminary r measurements of the observed brightest stars (considered fiducial cluster members). Out of 104 observed targets, we were left with a list of 38 stars. Finally, we cross-checked our target list with the list of probable cluster members compiled by Claria et al. (1993) on the basis of photometric measurements, and we removed another 14 stars from our list. In this process we considered as non cluster members also the stars not listed in Claria's work. Table 1 lists the targets that have survived our selection criteria. In this paper we report the spectral analysis performed on the

1 see the FLAMES web site (http://www.eso.org/sci/facilities/ paranal/instruments/flames/) for more details on the reduction of GIRAFFE data.


A detailed spectroscopic analysis of the open cluster NGC 5460

3

Table 1. List of programme stars. The last three entries refer to SB2 stars for which no abundance analysis could be performed. The V magnitude is taken from Claria et al. (1993). The symbol * in the spectral type column (No. 4) means that this was inferred from our determined effective temperature. SNR (col. 10) is the signal-to-noise ratio per spectral bin obtained with the GIRAFFE HR09/HR11/LR03 or the UVES 520 nm settings.

STAR HD 122983 HD 123097 HD 123182 HD 123183 HD 123184 HD 123201B HD 123202 HD 123226 HD 123269 CD-47 8868 CD-47 8869 CD-47 8889 CD-47 8905 CPD-47 6379 CPD-47 6385 UCAC 11104969 UCAC 11105038 UCAC 11105106 UCAC 11105176 UCAC 11105213 UCAC 11105379 HD 123225 Cl NGC 5460 BB 338 UCAC 11105152

Claria et al. (1993) ident. # 142 131 87 86 81 85 79 61 31 141 136 97 58 54 59 133 125 89 91 94 36 55 60 96

V 9.87 9.30 9.88 10.11 9.54 9.55 9.04 9.15 9.55 10.66 10.94 10.76 10.78 10.17 10.46 12.01 11.59 11.69 11.33 11.83 11.75 9.01 11.80 11.55

Sp. Type B9 B9 B9 A0 A0 B8 B9 B8 B8 A0 A5* A0 A0 A0 A0 F1* A8* A8* A6* A8* A8* B9 F* A*

Proper motion Dias et al. (2006) [mas/yr] -5.6 -4.9 -5.8 -6.6 -4.9 -3.6 -4.0 -6.3 -7.1 -6.4 -4.9 -6.7 -5.4 -5.3 -6.6 -5.3 -7.2 -7.4 -3.8 -6.8 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.9; -0.8 1.1; -3.1 1.1; -3.2 1.3; 0.8 0.9; 2.4 1.3; -5.4 1.0; -1.9 1.2; -4.4 1.0; -2.8 1.4; 2.8 1.7; -0.6 1.4; -3.0 1.4; -3.2 - 1.6; -2.5 2.1; -1.4 1.9; -0.2 1.4; 0.4 1.4; -1.0 2.9; -4.9 2.1; 2.4 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 1.0 1.2 1.2 1.3 0.9 1.3 0.9 1.1 1.2 2.4 1.4 2.2 1.4 1.4 1.7 1.4 1.4 2.7 1.4 1.4

Membership Dias et al. (2006) [%] 60 43 44 58 51 20 57 31 44 56 60 42 44 52 57 60 52 50 30 56 52 48

Proper motion Hog et al. (2000) [mas/yr] -5.3 -4.6 -5.7 -6.7 -4.8 -5.0 -4.2 -5.9 -7.0 -6.3 -4.6 -7.9 -5.2 -5.3 -7.9 -4.8 11.1 -6.6 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 1.1; -0.3 1.2; -2.4 1.2; -3.8 1.8; -0.4 1.2; -3.7 1.7; -5.6 1.1; -2.7 1.6; -6.3 1.2; -2.0 1.7; 2.4 1.7; 0.4 1.7; -5.1 2.1; -3.1 - 2.0; -0.2 2.0; -0.6 1.9; 1.3 2.2; 0.6 2.3; 1.1 - 2.3; 2.1 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 1.1 1.1 1.4 1.8 1.1 1.7 1.1 1.5 1.2 1.6 1.7 1.7 2.0 1.9 2.0 1.9 2.2 2.2

Pkin/Pph/Psp Kharchenko et al. (2004) 0.69/0.66/1 0.90/1.00/1 0.98/1.00/1 0.56/0.93/1 0.63/1.00/1 0.79/1.00/1 0.68/1.00/1 0.66/1.00/1 0.56/1.00/1 0.66/1.00/1 0.89/1.00/1 0.64/0.67/1 0.97/0.64/1 - / - /0.91/0.95/1 - / - /- / - /- / - /- / - /- / - /- / - /0.99/1.00/1 - / - /-

r [km s -1 ] -24 ± -20 ± 0±4 -19 ± -15 ± -20 ± -19 ± -12 ± -20 ± -10 ± -16 ± -19 ± -18 ± -24 ± -19 ± -17 ± -20 ± -16 ± -17 ± -21 ± -21 ± 2 2 7 3 5 11 1 2 9 5 4 5 7 2 3 4 3 5 3 2

SNR 159/178/213 209/233/275 148/165/195 144 [UVES] 195 [UVES] 199 [UVES] 247/275/323 231/258/306 185/211/243 104/116/137 96/106/125 99/110/128 100/111/131 120/134/158 123/135/162 58/63/76 69/78/94 66/73/82 78/87/104 62/69/79 65/72/87 207 [UVES] 49/55/69 70/79/93

-5.4 ±

± 2.1

-4.9 ± 1.2; -2.9 ± 1.0 -2.2 ± 1.4; -0.1 ± 1.4

-6.0 ± 1.2; -2.4 ± 1.3 -

-2 ± 11 / - 53 ± 2 -23 ± 6 / - 23 ± 1 -23 ± 4 / - 15 ± 1

24 targets of Table 1, which are cluster members according to considerations based on proper motion, radial velocity, and photometry, and have Teff >5400 K. We note that some early-type cluster members may have been missed while preparing the observations (either by accident, or because outside of the instrument field of view), and our selection based on radial velocity has left out some binaries that are potentially cluster members. Therefore, Table 1 cannot be considered a complete list of cluster members with spectral type earlier than G5.

3

ABUNDANCE ANALYSIS

Model atmosphere calculations were performed with the LLMODELS stellar model atmosphere code (Shulyak et al. 2004), assuming Local Thermodynamical Equilibrium (LTE), plane-parallel geometry, and modeling convection theory according to the Canuto & Mazzitelli (1991, 1992) model of convection (see Heiter et al. 2002, for more details). We used the VALD database (Piskunov et al. 1995; Kupka et al. 1999; Ryabchikova et al. 1999) as a source of atomic line parameters for opacity calculations. We derived the fundamental parameters for each analysed star mainly by fitting synthetic line profiles, calculated with SYNTH3 (Kochukhov 2007), to the observed hydrogen lines: H and H for the stars observed with FLAMES/UVES, and H for the stars observed with FLAMES/GIRAFFE. FLAMES/UVES spectra cover each the two hydrogen lines on more than one order, therefore we merged the unnormalised orders before the normalisation, which was performed with a low degree polynomial function. The H lines, observed with FLAMES/GIRAFFE, were normalised with a low degree polynomial function (usually of degree 4) to carefully selected continuum points. Our hydrogen line normalisation was performed on the basis of continuum points which extend for sev-

eral Angstroms both blueward and redward of the hydrogen line wings. This ensures a high quality continuum normalisation of the hydrogen lines, thanks also to the absence of instrumental artifacts, such as scattered light, which would strongly reduce the precision of the normalisation. Figure 1 shows a comparison between observed and synthetic H line profiles for seven stars of various Teff observed with FLAMES/GIRAFFE. Due to the very broad temperature range covered by the analysed stars, the hydrogen lines were used as both temperature and gravity indicators, because at lower temperatures hydrogen lines are more sensitive to temperature variations, while at higher temperatures gravity has the larger influence. This is also reflected on our error bars on the fundamental parameters. Since with the fitting of the hydrogen line wings it was not always possible to derive simultaneously both Teff and log g, we adopted also two other spectroscopic indicators based on the analysis of metallic lines. In particular, T eff was determined eliminating the correlation between line abundance and the line excitation potential (excit ) for a given ion/element and the surface gravity was determined with the ionisation balance of several elements. In particular we adopted mainly Fe lines to derive both Teff and log g from metallic lines, adopting then Ti, Ca and Si lines (the latter for the hotter stars of the sample) as secondary indicators. Given the fact that the number of measurable lines decreases with increasing sin i, for stars with a high projected rotational velocity we ended up adopting only Fe lines. Non-LTE effects might play an important role when measuring log g by imposing ionisation equilibrium. Fossati et al. (2009) concluded that for solar metallicity early-type stars non-LTE effects for Fe are most likely to be negligible. Mashonkina (2011) performed non-LTE calculations for Fe adopting the most complete iron model atom currently existing, and concluding that for A-type


4

L. Fossati et al.
UCAC 11104969

1.8

UCAC 11105106

1.6

UCAC 11105176

Table 2. Atmospheric parameters for the programme stars. The last six entries refer to SB2 stars for which only very few parameters could be estimated, and no abundance analysis was performed. The log g values given for the two components of HD 123225 are assumed and not measured, therefore it is not possible to give a realistic uncertainty to T ff as well. Due e to the very few measurable spectral lines, mic for HD 123202 is assumed. Star name HD 122983 HD 123097 HD 123182 HD 123183 HD 123184 HD 123201B HD 123202 HD 123226 HD 123269 CD-47 8868 CD-47 8869 CD-47 8889 CD-47 8905 CPD-47 6379 CPD-47 6385 UCAC 11104969 UCAC 11105038 UCAC 11105106 UCAC 11105176 UCAC 11105213 UCAC 11105379 HD 123225A HD 123225B NGC 5460 BB 338A NGC 5460 BB 338B UCAC 11105152A UCAC 11105152B T eff [K] 10700 12000 10800 10900 10900 11700 11400 12400 11600 10400 8400 10300 9700 11000 9350 7050 7650 7500 8100 7600 7700 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 200 300 200 250 250 250 250 300 250 200 200 200 200 250 200 150 200 200 200 200 200 log g [cgs] 4.00 3.96 4.25 4.05 4.25 3.95 3.70 4.04 3.94 4.20 3.60 4.15 4.10 4.30 4.12 4.00 4.00 4.00 4.00 4.00 4.00 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.15 0.10 0.15 0.15 0.15 0.15 0.15 0.10 0.15 0.15 0.20 0.15 0.20 0.15 0.20 0.20 0.20 0.20 0.20 0.20 0.20 mic [km s-1 ] 1.0 1.5 0.0 0.6 0.0 0.0 0.0 0.0 0.0 0.9 1.2 0.5 1.6 1.7 1.6 1.4 3.4 2.5 1.8 1.9 2.1 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.2 0.6 0.4 0.8 0.4 0.8 1.0 0.2 0.2 0.7 0.7 0.6 0.4 0.6 0.3 0.9 0.4 0.6 0.4 0.3 0.2 sin i [km s-1 ] 35 133 81 275 60 202 301 17 25 250 230 113 160 136 55 260 85 195 125 57 44 12 20 140 <11 190 < 11 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 3 5 6 14 6 11 17 1 2 10 10 3 6 7 5 15 5 10 7 3 2 1 2 7 1 10 1

CPD-47 6385

1.4

normalised flux

CD-47 8868

1.2
HD 123202

1

HD 123226

0.8

0.6 observed spectrum synthesised spectrum

0.4

4800

4850

wavelength [A]
Figure 1. Comparison between the observed H line profile (thick lines) and synthetic line profile (thin lines) for seven stars observed with FLAMES/GIRAFFE. From top to bottom the stars here displayed are: UCAC 11104969 (Teff = 7050 K), UCAC 11105106 (Teff = 7500 K), UCAC 11105176 (Teff = 8100 K), CPD-47 6385 (Teff = 9350 K), CD47 8868 (Teff = 10400 K), HD 123202 (Teff = 11400 K) and HD 123226 (T eff = 12400 K). Profiles are vertically offset for display purpose.

o

4900

13100 ± 8000 ± -

3.8 ± 4.0 ± -

main-sequence stars non-LTE effects for FeI are less than 0.1 dex and negligible for FeII. Both Ti, Ca and Si are subject to non-LTE effects, but we considered these elements only as secondary parameter indicator, giving most of the weight to what derived from Fe lines. For the cooler stars we were not able to apply the method based on line profile fitting of gravity-sensitive metal lines with developed wings, usually adopted to derive log g for late-type stars (see e.g. Fuhrmann et al. 1997), because the low resolution of the FLAMES/GIRAFFE spectra caused too heavy blending of those lines. For stars with a low sin i ( sin i< 30 km s-1 ) LTE abundance analysis was based on equivalent widths, analysed with a modified version (Tsymbal 1996) of the WIDTH9 code (Kurucz 1993), while for fast rotating stars we derived the LTE abundances by line profile fitting of carefully selected lines (see Fossati et al. 2007, for more details). The microturbulence velocity (mic ) was determined by minimising the correlation between equivalent width and abundance for several ions, for the slowly rotating stars for which it was possible to measure the line equivalent widths, or following the procedure described in Fossati et al. (2008b) for fast rotators, except for HD 123202, for which we were not able to measure mic , which we then assumed to be 0.0 km s-1 , with an uncertainty of 1.0 km s-1 .

Both the projected rotational velocity ( sin i) and r were determined by fitting synthetic spectra of several carefully selected lines on the observed spectrum. Table 2 lists the fundamental parameters derived for each analysed star in NGC 5460. For some stars a preliminary spectral classification is given here for the first time on the basis of the derived T eff . The uncertainties of T eff and log g, listed in Table 2, were determined from the hydrogen line fitting, our main source for the parameter determination, therefore the given error bars take into account the quality of the observed spectra and the sensitivity of the hydrogen line wings to both Teff and log g at different temperatures. Table 2 also shows that the uncertainties on Teff and log g depend on the stellar sin i. This is due to the fact that we also adopted metallic lines to determine the atmospheric parameters and the precision with which we determined the element/ion abundances decreases with increasing sin i, for example because of the decreasing number of measurable lines for each ion. The abundances obtained for each analysed star are shown in Table 3. The error bars are the standard deviations of the mean abundance derived from the individual line abundances, which include also the error bars in oscillator strengths determination given for the laboratory data (Fossati et al. 2009). The standard deviation of the mean underestimates the actual abundance error bars, since uncertainties on the atmospheric parameters should also be taken into account. In particular for the


A detailed spectroscopic analysis of the open cluster NGC 5460
fast rotating stars, the uncertainty on mic has a large impact on the total abundance uncertainty. For two stars with very similar atmospheric parameters, but different sin i values (CPD-47 6385 and CD-47 8905) we determined the element abundances adopting model atmospheres for which we increased each atmospheric parameter one at a time, by an amount representative of the whole sample of stars: 250 K for Teff , 0.2 for log g and 0.7 km s-1 for mic . Table 4 shows the results of this test. For the slow rotator, the abundance error bar is highly dominated by the uncertainty on Teff , while for the fast rotator, the uncertainty on mic plays a crucial role. In both cases the uncertainty on log g is almost negligible. This test also shows the dependency of the abundance error bars on the lines selected to perform the abundance analysis: for the slow rotator the abundance error bars of Mg, Ti and Ba, due to the uncertainty on mic , are rather large. This is due to the fact that most of the lines selected to measure the abundance of these three elements are strong (e.g. the Mg triplet at 5167, 5172 and 5183), being therefore very sensitive to mic variations. Anyway, this effect is mitigated by the fact that for the slow rotators the uncertainty on mic is not as large as 0.7 km s-1 , value chosen here to perform this test. Using the propagation theory in Table 4 we considered the situation where the determination of each fundamental parameter is an independent process, which in reality is not true, since there is a level of correlation between the various parameters. This makes the global uncertainties, given in columns five and nine of Table 4, upper limits. In conclusion, following the results of this test and of the ones performed by Fossati et al. (2008b) and Fossati et al. (2009), we can safely set a mean abundance uncertainty of 0.2 dex, for the slow rotators, which increases linearly with sin i, due to the increasing importance of the mic uncertainty in the total abundance error bar budget. The mean abundance error bars, adopted for each star, are shown in the last column of Table 3. The impact of systematic uncertainties is different for different elements, but it is out of the scope of this work to go in such details, since here we are mainly interested in general trends and only marginally in the detailed abundance characteristics of each single star, which should require spectra with much higher resolution and signal-to-noise ratio (SNR), as demonstrated by Fossati et al. (2009).

5

Figure 2. Histogram of the difference of the abundances obtained with the secondary and the standard analysis, divided by the error bar of the standard analysis, listed in the last column of Table 3. The dashed lines correspond to a 3- difference.

4

THE IMPACT OF SYSTEMATIC ERRORS

The scattering of the element abundance obtained from different lines provides only a lower limit for the actual error in the abundance determination. Systematic effects introduced by the normalisation to the continuum, and inaccurate Teff and log g estimates may have a large impact. To explore the reliability of the error estimates of Table 3 we re-analysed a number of stars in a completely independent way, using the ZEEMAN (Landstreet 1988; Wade et al. 2001) spectrum synthesis code. This `secondary' analysis helps to ensure our results and estimated uncertainties are robust. The ZEEMAN code performs polarised radiative transfer, under the assumption of LTE, using plane-parallel model atmospheres. Input LTE model atmospheres were calculated using ATLAS9 (Kurucz 1993), assuming solar abundances, which are a reasonable starting point as none of the stars in NGC 5460 show extreme chemical peculiarities. Input atomic data was drawn from the VALD database, using an `extract stellar' request, with effective temperature and surface gravity tailored to the specific star, and solar abundances. Initial values for temperatures and surface gravities were

found by fitting synthetic Balmer lines to the observations. These values were confirmed, and when possible refined, by fitting metallic lines with a range of ionisation states and excitation potentials. The Balmer line fitting was performed by eye, though the metallic line fitting was done using 2 minimisation. Abundances, sin i and mic were determined by fitting a synthetic spectrum to the observation, using a Levenberg-Marquardt 2 minimisation procedure. This was performed for 3 segments of spectrum between 100 å and 500 å long in the 4500­6000 å region. The final best fit values reported here are averages over the individual segments. Uncertainties are generally taken to be the standard deviation. For elements with less then 3 useful lines, an uncertainty was estimated by eye, taking into account the scatter between lines, blending, and potential normalisation errors. Best fit results for the stars analysed with this method are presented in Table 5. Figure 2 shows the histogram of the difference of the abundances obtained with the secondary and the standard analysis, divided by the error bar of the standard analysis (last column of Table 3). Generally the results are very consistent, with the abundances being within uncertainty of each other, at least within a 1.5 limit. This secondary analysis supports also the conclusion that HD 123182 is chemically peculiar. A large Mn overabundance is confirmed, though no Sc or Hg abundances could be derived. Our secondary analysis confirms also that HD 123183 is a chemically normal star, with a very high sin i, despite one published marginal magnetic field detection. In both HD 123201B and HD 123184 discrepancies in sin i were again detectable for lines with high excitation potentials, particularly He lines. The secondary analysis confirms both these stars are chemically normal.


6

L. Fossati et al.

Table 3. Element abundances, in log(NX /Ntot ), of the analysed cluster member stars. The estimated internal errors in units of 0.01 dex and the number of lines selected for each element are given in parenthesis. For comparison, the last row shows the solar abundances by Asplund et al. (2005). Abundances obtained from just one line have no error (-;1). The column labelled as `Uncertainty' shows the abundance uncertainty for each star, calculated following Fossati et al. (2008b), and which is taken into account in the discussion of the results.
Star HD 122983 HD 123097 HD 123182 HD 123183 HD 123184 HD 123201B HD 123202 HD 123226 HD 123269 CD-47 8868 CD-47 8869 CD-47 8889 CD-47 8905 CPD-47 6379 CPD-47 6385 UCAC 11104969 UCAC 11105038 UCAC 11105106 UCAC 11105176 UCAC 11105213 UCAC 11105379 sun Star HD 122983 HD 123097 HD 123182 HD 123183 HD 123184 HD 123201B HD 123202 HD 123226 HD 123269 CD-47 8868 CD-47 8869 CD-47 8889 CD-47 8905 CPD-47 6379 CPD-47 6385 UCAC 11104969 UCAC 11105038 UCAC 11105106 UCAC 11105176 UCAC 11105213 UCAC 11105379 sun Star HD 122983 HD 123097 HD 123182 HD 123183 HD 123184 HD 123201B HD 123202 HD 123226 HD 123269 CD-47 8868 CD-47 8869 CD-47 8889 CD-47 8905 CPD-47 6379 CPD-47 6385 UCAC 11104969 UCAC 11105038 UCAC 11105106 UCAC 11105176 UCAC 11105213 UCAC 11105379 sun He -2.00(-;1) -0.90(-;1) -1.48(-;1) -0.90(05;3) -1.00(03;6) -1.02(05;2) -0.98(-;1) -3.50(-;1) -3.47(-;1) -3.60(-;1) C O -3.41(-;1) -3.41(-;1) -3.40(11;4) -3.28(01;3) -3.28(-;1) -3.41(01;2) Na Mg -5.22(-;1) -4.55(36;3) -4.56(01;2) -4.30(16;3) -4.35(16;5) -4.29(10;5) -4.64(08;3) -4.58(16;3) -3.80(-;1) -5.27(16;2) -4.20(06;3) -4.03(11;3) -4.10(28;3) -4.28(02;3) -5.11(32;2) -4.86(07;3) -4.62(21;4) -4.45(03;3) -4.87(06;3) -4.69(03;3) -4.51 Cr -5.27(14;19) -6.78(12;2) -5.90(23;9) -5.81(01;2) -6.15(13;10) -5.99(14;4) -5.94(10;2) -6.63(12;6) -6.48(13;9) -6.40(25;4) -6.58(11;4) -6.16(09;5) -6.39(10;4) -5.15(15;4) -6.04(09;17) -6.76(15;6) -6.50(15;16) -6.81(09;10) -6.33(08;10) -6.27(15;24) -6.39(12;22) -6.40 Ba -8.64(-;1) -9.81(-;1) -9.34(-;1) -9.09(-;1) Al -6.19(-;1) -5.85(-;1) -5.43(-;1) -5.56(-;1) -5.70(-;1) -5.77(17;2) -5.76(-;1) Si -4.41(-;1) -4.54(01;2) -4.58(09;2) -4.37(19;2) -4.39(10;4) -4.52(01;2) -4.62(-;1) -4.63(-;1) -4.42(-;1) -4.52(-;1) -4.34(-;1) -4.52(17;2) -4.26(09;2) -4.77(45;2) -4.71(24;2) -3.58(-;1) -4.37(15;3) -4.09(21;2) -4.21(-;1) -4.27(13;3) -4.37(09;3) -4.53 Fe -3.81(12;100) -4.60(05;12) -4.20(14;60) -4.37(22;11) -4.29(14;57) -4.47(14;20) -4.50(07;5) -4.75(15;54) -4.57(16;60) -4.25(21;13) -4.50(10;15) -4.27(17;22) -4.43(15;21) -4.20(23;27) -4.38(09;59) -4.62(05;12) -5.08(13;91) -5.03(15;54) -4.80(12;61) -4.61(16;133) -4.74(16;129) -4.59 Nd -7.43(10;8) S

-4.77(17;6) -4.45(-;1)

-5.52(-;1)

-3.27(13;5) -3.47(-;1) -3.38(14;3) -3.24(04;2) -3.49(12;3) -3.42(05;3) -3.65 Sc -7.48(20;4) -5.29(-;1) -5.77(-;1) -9.21(-;1)

-5.35(-;1)

-5.78(03;2) -6.13(06;2) -5.69(04;2) -5.84(05;2) -5.87 V

-1.11 Ca

-3.38 Ti -7.37(13;3) -7.31(10;2) -6.87(08;5) -6.51(-;1) -6.88(14;8) -6.54(18;5) -7.27(-;1) -7.26(25;2) -7.04(15;2) -6.67(08;2) -7.07(13;3) -7.00(28;4) -7.12(22;4) -6.85(13;12) -7.48(32;5) -7.64(17;19) -7.44(14;10) -6.97(19;7) -7.16(04;13) -7.22(05;15) -7.14 Sr

-5.67 Mn

-4.37(-;1) -4.58(-;1) -4.90 Ni

-5.19(23;9) -6.10(-;1)

-5.86(-;1) -5.76(-;1)

-5.35(-;1)

-9.68(-;1)

-6.82(-;1)

-5.30(09;4) -5.69(17;4) -6.23(09;8) -5.67(07;8) -5.86(06;4) -6.03(25;19) -5.89(14;15) -5.73 Cu

-8.81(04;3) -9.43(-;1) -9.17(19;3) -8.56(22;2) -8.82(11;2) -9.23(09;4) -9.11(11;4) -8.99 Zn

-7.76(-;1) -8.04 Y

-5.99(03;2) -6.34(08;3) -6.83(14;3) -6.83(08;3) -6.35(17;3) -6.75(22;5) -6.78(12;4) -6.65 Pr -7.57(20;3)

-5.55(30;2) -5.65(65;5) -6.09(18;9) -6.04(04;3) -5.74(16;5) -5.67(16;15) -6.11(18;14) -5.81 Uncertainty 0.20 0.26 0.23 0.34 0.22 0.30 0.36 0.20 0.20 0.33 0.32 0.25 0.28 0.26 0.21 0.34 0.23 0.30 0.26 0.22 0.21

-9.70(-;1)

-8.64(03;2) -6.79(-;1) -7.15(-;1) -7.75(-;1) -7.44 -9.64(16;3) -9.87(13;2) -9.83

-7.83

-9.12

-8.17(11;2) -8.55(-;1) -9.18(16;2) -9.63(-;1) -10.10(-;1) -8.51(16;2) -9.35(16;2) -9.87

-11.33

-10.59


A detailed spectroscopic analysis of the open cluster NGC 5460
Table 4. Error sources for the abundances of the chemical elements of CPD-47 6385 and CD-47 8905. Both stars have very (T eff 9500 K, log g4.1 and mic 1.6 km s-1 ), but rather different sin i values, as shown in the second table line. Columns in abundance estimated by increasing Teff by 250 K, log g by 0.1 dex, and mic by 0.7 km s-1 , respectively. Columns 5 and 9 2 applying the standard error propagation theory on the systematic uncertainties given in Col. 2-4 and 6-8, i.e., (syst.) abn 2 ( abn mic ). CPD-47 6385 sin i=55 km s- El. C Mg Al Si Ca Sc Ti Cr Mn Fe Ni Ba
abn

7

similar atmospheric parameters 2-4 and 6-8 give the variation give the mean error calculated = 2 (T eff ) + 2 (log g) + abn abn

1

CD-47 8905 sin i=160 km s- abn (syst.) (dex) 0.15 0.27 0.07 0.10 0.27 0.15 0.19 0.16 0.23 0.15 0.10 0.55 abn (T eff ) (dex) 0.31 0.05 abn (log g) (dex) -0.07 -0.01 0.05 0.06 0.04

1

(T eff ) (dex) 0.15 0.21 0.00 0.09 0.25 0.14 0.08 0.13 0.23 0.13 0.10 0.26

abn (log g) (dex) -0.04 -0.07 0.06 -0.03 -0.04 0.04 0.01 0.03 -0.01 0.00 -0.03 -0.07

abn (mic ) (dex) -0.01 -0.15 -0.03 -0.04 -0.09 -0.03 -0.17 -0.08 -0.03 -0.08 0.00 -0.48

abn (mic ) (dex) -0.28 -0.06 -0.31 -0.19 -0.16

abn (syst.) (dex) 0.42 0.08

0.17 0.17 0.09

0.36 0.26 0.19

Table 5. Best fit parameters from the comparison analysis with ZEEMAN. Chemical abundances are in units of log(N /Ntot ). Elements marked with an asterisk X are based on less than 3 useful lines and have uncertainties estimated by eye. HD 123182 T eff [K] log g [cgs] sin i [km s-1 ] mic [km s-1 ] Vr [km s-1 ] He C O Na Mg Al Si S Ca Sc Ti V Cr Mn Fe Ni Cu Zn Y Ba 10800 ± 300 4.25 ± 0.2 81 ± 4 0 -1 ± 6 -1.48 ± 0.4 * -3.54 ± 0.1 -4.5 ± 0.2 -6.1 ± 0.5* -4.7 ± 0.2 -5.68 ± 0.4 * -6.88 ± 0.1 -6.24 ± 0.1 -5.52 ± 0.1 -4.31 ± 0.1 -7.3 ± 0.4 * -6.9 ± 0.4 * -4.4 ± 0.2 -3.5 ± 0.2 * -4.9 ± 0.4 * -4.5 ± 0.2 * HD 123183 10500 ± 700 4 . 3 ± 0. 3 270 ± 15 0 -18 ± 20 HD 123184 10900 ± 120 4.35 ± 0.2 63.1 ± 1.5 0.4 ± 0.22 -14 ± 1 -0.9 ± 0.2 * -3.4 ± 0.2 * -3.3 ± 0.12 -4.56 ± 0.14 -5.6 ± 0.2 * -4.4 ± 0.11 -4.4 ± 0.3 * -5.52 ± 0.27 -8.8 ± 0.1 -7.16 ± 0.12 -8.4 ± 0.4 * -5.99 ± 0.11 -6.4 ± 0.3 * -4.41 ± 0.14 HD 123201B 12000 ± 450 4 . 2 ± 0. 3 203 ± 6 0 -25 ± 13 -1.05 ± 0.2 * -3.45 ± 0.2 * -4.49 ± 0.24 -5.85 ± 0.4 * -4.71 ± 0.27 HD 123226 12140 ± 160 4.27 ± 0.1 18.1 ± 1.0 0.5 ± 0.5 -12 ± 1 -0.9 ± 0.2 * -3.57 ± 0.2* -4.87 -5.65 -4.57 -4.92 ± ± ± ± 0.06 0.15 0.08 0.3* -3.4 ± 0.21 -4.47 ± 0.17 -5.68 ± 0.3 * -4.85 ± 0.15 * -5.67 ± 0.3 * -8.48 ± 0.2 * -6.87 ± 0.08 -5.53 ± 0.11 -4.16 ± 0.05 -3.3 ± 0.15 * -5.48 ± 0.2 * -4.40 ± 0.05 -4.59 ± 0.1 * -5.45 ± 0.15 -8.81 ± 0.09 -6.88 ± 0.06 -6.21 ± 0.14 -4.35 ± 0.15 CPD-476379 10100 ± 300 4.15 ± 0.2 140 ± 8 1.02 ± 0.45 -30 ± 3 CPD-476385 9360 ± 150 4.12 ± 0.2 56.1 ± 2.5 1.8 ± 0.2 -21 ± 1

-7.26 ± 0.15 -6.67 ± 0.2 -4.66 ± 0.19

-7.4 ± 0.1 -6.53 ± 0.05 -6.72 ± 0.3* -4.73 ± 0.05

-9.6 ± 0.4 * -9.6 ± 0.4 *

-9.19 ± 0.4 * -8.85 ± 0.4 *

5

RESULTS

Our r measurements allowed us to derive the mean cluster radial velocity: -17.9 ± 5.2 km s-1 , obtained without taking into account the double line spectroscopic binaries (SB2) stars. This value agrees rather well with -12.70 ± 2.18 km s-1 , published by Kharchenko et al. (2005), in particular taking into account that the latter value is the average of the r of only three stars.

We cannot exclude that some of the analysed stars are single line spectroscopic binaries (SB1), but this was impossible to establish with the available spectra.

In the following we describe more in detail the results obtained for the most interesting objects.


8
5.1

L. Fossati et al.
HD 122983
1

The star HD 122983 was observed by Bagnulo et al. (2006) to measure the longitudinal magnetic field, but no field was detected. The abundance pattern shows that HD 122983 is a He-weak star, but it was not possible to determine the kind of chemical peculiarity (e.g. HgMn star). Manganese lines are too shallow to be measured and none of the covered mercury line is visible. A spectrum covering the spectral region around 3900 å would be extremely valuable to detect the presence of Hg overabundance (Hubrig & Mathys 1996). We looked also for the overabundance of elements characteristic of HgMn stars, such as P and Xe, but without results. 5.2 HD 123182

0.9

0.8

normalised flux

0.7

0.6

0.5

0.4

observed spectrum Teff: 13100 K; logg: 3.8 Teff: 8000 K; logg: 4.0 sum

0.3 4850 4855 4860 4865 4870 4875

The abundance pattern of HD 123182 reveals that the star is most likely a mild He-weak star, in particular it could be a hot Am star or a cool HgMn star. Sc lines are too weak to be measurable, so it is not possible to say whether it is underabundant or not. The not extreme overabundance of Cr and Fe (as expected for the rather high sin i), and the fact that Hg, P and Xe are not visible (although lines would be covered) are in agreement with an Am classification. On the other hand the clear Mn overabundance would tend more to a HgMn classification. On the basis of the available data we can only conclude that HD 123182 is a He-weak star of an unclear type. This star could also be a single line spectroscopic binary, due to the high radial velocity, compared to the one of the other cluster stars. 5.3 HD 123183

wavelength (A)
Figure 3. Fit of the observed H line profile for HD 123225. The red line shows the observed spectrum, while the dashed and dotted lines are the synthetic spectra of the two components assuming Teff and log g of 13100 K and 3.8 for the hotter star and 8000 K ad 4.0 for the cooler star. The black solid line shows the sum of the two synthetic spectra.

o

The star HD 123183 was observed with FORS1 by Bagnulo et al. (2006) who published a marginal magnetic field detection of Bz = -440 ± 146 G. The abundance pattern does not reveal any chemical peculiarity that could strengthen this detection. The very high sin i of 275 ± 14 km s-1 also goes against a classification of this object as chemically peculiar. For this reason we conclude that HD 123183 is a chemically `normal' early-type star. Given the very high sin i, it would be very valuable to perform new magnetic field measurements. 5.4 HD 123225

are present in the primary star, we looked for typical signatures present in HgMn stars. The spectrum reveals strong Mn lines in addition to evidences of Hg lines, allowing to conclude that the primary component is a HgMn star. This conclusion is strengthened by the fact that we found several strong lines of both Ne, P and Xe, typical of HgMn stars. For the secondary star it was not possible to determine whether it shows Am chemical peculiarities since the Sc lines are too shallow to be visible. In any case, the fact that the star seems to have Fe lines compatible with a solar iron abundance, let us believe that the secondary component is not an Am star. HD 123225 was observed with FORS1 by Bagnulo et al. (2006) and no magnetic field was detected.

5.5 The FLAMES/UVES observations of HD 123225 revealed that this is a SB2 system, where both components appear to have a rather low sin i. The spectral lines of the two components are well separated: r of the first component is -2 km s-1 , while r of the second component is -53 km s-1 . With only one available spectrum it was not possible to perform a precise abundance analysis, but the large separation of the two components allowed to roughly estimate their T eff : 13100 K for the primary and 8000 K for the secondary star. These values were obtained from a fit of synthetic spectra on the observed H profile (see Fig. 3), taking into account the expected relative difference in flux due to the different stellar radii of the two components and assuming that both are main sequence stars (log g was set to 3.8 for the hotter star and to 4.0 for the cooler star). The sin i value of the primary star is 11 km s-1 , while for the secondary star we derived a sin i value of 20 km s-1 . Given the roughly determined Teff of the two components and their sin i, it is possible that both stars developed chemical peculiarities and in particular HgMn peculiarities for the primary star and Am peculiarities for the secondary star. To reveal whether HgMn peculiarities

HD 123201B and HD 123184

The stars HD 123201B and HD 123184 are both chemically `normal' early-type stars and share the common property of showing a different sin i for the high excit lines, compared to the other spectral lines. The sin i of HD 123201B, measured on low excit lines, is 202 km s-1 , while He lines are best fit with a sin i of 150 km s-1 . The same occurs for HD 123184, for which sin i is 60 km s-1 , but He lines are best fit with a sin i of 50 km s-1 . Figure 4 shows the comparison between the observed spectrum and two synthetic line profiles with sin i = 60 and 50 km s-1 in the region of the FeII line at 5018 å and of the HeI multiplet at 5875 å for HD 123184. The plot shows that for the HeI multiplet the adopted stellar sin i of 60 km s-1 is far too high, while a sin i of 50 km s-1 is necessary to fit the observed line profile. In contrast, the region around the FeII line shows that a sin i of 60 km s-1 is required to fit this line. For both stars the He abundance was measured with the sin i determined for the He lines. This effect is most likely a consequence of the fact that both stars are fast rotators with a rather small inclination of the rotation axis relative to the line of sight, leading to a temperature difference between the poles (hotter) and the equator (cooler) (von Zeipel 1924).


A detailed spectroscopic analysis of the open cluster NGC 5460
abundance relative to the Sun
Fe II 1 He I 1.2 1 0.8 0.6 0.4 0.2 0 A-type stars B-type stars F-type stars

9

normalised flux

0.98

0.96

-0.2 -0.4 -0.6 -0.8 He C O Na Mg Al Si S Ca Sc Ti V Cr Mn Fe Ni Zn Sr Y Ba

0.94 observation Vsini = 50 km/s Vsini = 65 km/s 5017 5018 5019
o

0.92 5020

5874

5875

5876

wavelength [A]
Figure 4. Comparison between thetic spectra calculated with 50 km s-1 (thick dashed line) in plot) and of the HeI multiplet at

wavelength [A]

o

5877

element
Figure 5. Comparison between the mean abundances obtained for B-, Aand F-type stars indicated by an open square, circle and triangle respectively. The error bars are the standard deviations from the calculated mean abundances. The abundance patter of the mean F-type stars is corresponds to the abundance pattern of UCAC 11104969, since it is the only F-type stars of the sample with measured abundances.

the observed spectrum (thin line) and synsin i = 60 km s-1 (thick line) and sin i = the region of the FeII line at 5018 å (left 5875 å (right plot), for HD 123184.

5.6

UCAC 11105038 metallicity should be calculated on the basis of the abundances of the cool stars (more mixed) and that only one F-type star is present in our sample, we decided to derive Z from the mean abundance of UCAC 11104969 and of the A-type stars with Teff 8500 K. The cluster metallicity (Z ) is defined as follows: Z
cluster

This is an A-type star with a sin i of 85±5 km s-1 and a relatively high mic of about 3.4 km s-1 . Both sin i and mic could indicate a possible Am chemical peculiarity of this star, in particular sin i is below 90 km s-1 (Charbonneau & Michaud 1991), and mic is typical of Am stars (Landstreet 1998; Fossati et al. 2008b). Both Ca and Sc abundances are below the solar values, as well as the Ba overabundance is indicative of Am chemical peculiarities, but the underabundance of the Fe-peak elements goes against an Am classification.

=

a3 a1

ma 10log( ma 10log(

Na /Ntot ) Na /Ntot )

,

(1)

6 6.1

DISCUSSION Abundance patterns in NGC 5460: from F- to B-type stars

where a is the atomic number of an element with atomic mass ma . Making use of the mean abundances, we derived a metallicity of Z = 0.014 ± 0.003, consistent with the solar Z (Z =0.012, Asplund et al. 2005), adopting solar abundances by Asplund et al. (2005) for all elements that were not analysed. The Z value adopted by several other authors and used to characterise isochrones is calculated with the following approximation: Z
clu ster

Figure 5 shows the comparison among the mean abundance pattern obtained for B-, A- and F-type stars here analysed and considered members of the cluster. The error bars are the standard deviation from the mean abundance. In the whole sample, after the assignment of the cluster membership and binarity, only one F-type single star, UCAC 11104969, was left in the sample, therefore the mean abundance of the F-type stars corresponds to the abundance pattern of UCAC 11104969. Figure 5 shows that the three abundance patterns are quite similar and concentrated around solar values except for a sulfur overabundance, which cannot be ascribed to non-LTE effects (Fossati et al. 2007), and for an apparent Am peculiarity of some elements in UCAC 11104969, excluded by the high sin i value (260±15 km s-1 ). An important result of this work is the estimation of the cluster metallicity (Z ), spectroscopically never measured before. For the elements that most affect Z , such as C, O, Si, and Fe, we registered a rather good agreement among the three stellar types, with the exception of the Si abundance of UCAC 111049692 , which was not taken into account in the determination of Z . Since the cluster
the star has a high sin i and the Si abundance was derived only from one line, making the Si abundance value rather uncertain
2

10([

Fe/H ]F star s -[Fe/H ] )

·Z ,

(2)

assuming Z =0.019, which is favored by the theoretical stellar structure models based on (Anders & Grevesse 1989) solar abundances. We recalculated the Z value of NGC 5460, according to this approximation, obtaining Z = 0.013+0..008 , only about con-0 005 sistent with a solar metallicity. This value was derived from the mean Fe abundance of the seven stars taken into account ([Fe/H] = -0.18±0.22 dex). We believe that a solar metallicity would be anyway more appropriate for NGC 5460. The Fe underabundance of -0.18 dex is biased by two stars (out of a sample of seven) which show a considerable underabundance ([Fe/H] -0.46 dex). On the other hand, for the other five stars the iron abundance is [Fe/H]=-0.06±0.12 dex, which leads to Z = 0.017+0..005 , that, given -0 004 the uncertainties, is perfectly consistent with a solar metallicity. In addition, Fig. 5 shows that the mean Fe abundance of the analysed cluster stars is well centered around the solar value. 6.2 HR diagram

With the obtained effective temperatures and the magnitudes provided by Claria et al. (1993), we built the Hertzsprung-Russell (HR) diagram of the cluster (Fig. 6), adopting a distance modulus of 9.44 mag, a reddening of 0.092 mag (both values given by


10
2.5

L. Fossati et al.
2.4 2 1.6 1.2 0.8 0.4 0 -0.4 -0.8 -1.2 2.4 2 1.6 1.2 0.8 0.4 0 -0.4 -0.8 -1.2 8000
4 3.95 3.9 3.85

2

log(L/Lsun)

Al abundance

1.5

1

"normal" He-weak Z=0.019; Z=0.013; Z=0.019; Z=0.019; Z=0.019; 4.1

stars stars log(t)=8.20 log(t)=8.20 log(t)=8.10 log(t)=8.25 log(t)=8.30 4.05

Teff [K]

10000

12000

8000

Teff [K]

10000

12000

log(Teff)
Figure 6. Hertzsprung-Russell diagram of NGC 5460. The crosses and filled circles show the position on the HR diagram for the chemically peculiar and chemically normal stars, respectively. The error bar in luminosity is 0.10 dex. The full and dotted lines show isochrones from Marigo et al. (2008) for an age of log t = 8.20 with solar metallicity (Z = 0.019) and Z = 0.013, respectively. The short dashed, thin dot-dashed line and long dashed line show solar metallicity isochrones of three different ages: log t = 8.10, 8.25 and 8.30, respectively.

Figure 7. Abundances relative to the Sun (Asplund et al. 2005) of C, Mg, Al and Si as a function of Teff for B-type stars (open squares), A-type stars (open circles), F-type stars (open triangles) and He-weak stars (crosses).

Table 6. log L/L , logT eff , M/M and fractional age () with associated error bars for the stars of the NGC 5460 open cluster. Star HD 122983 HD 123097 HD 123182 HD 123183 HD 123184 HD 123201B HD 123202 HD 123226 HD 123269 CD-47 8868 CD-47 8869 CD-47 8889 CD-47 8905 CPD-47 6379 CPD-47 6385 UCAC 11104969 UCAC 11105038 UCAC 11105106 UCAC 11105176 UCAC 11105213 UCAC 11105379 log L/L 1.88 2.25 1.89 1.80 2.06 2.10 2.33 2.35 2.11 1.50 1.25 1.47 1.41 1.76 1.52 0.72 0.92 0.88 1.06 0.80 0.88 logT
eff

M/M 2.76 3.40 2.80 2.73 3.02 3.17 3.45 3.60 3.15 2.30 1.94 2.35 2.20 2.71 2.27 1.48 1.66 1.60 1.78 1.59 1.64 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.14 0.17 0.14 0.14 0.21 0.16 0.24 0.18 0.16 0.16 0.14 0.16 0.11 0.14 0.16 0.07 0.08 0.08 0.09 0.08 0.08 0.37 0.67 0.38 0.36 0.48 0.55 0.70 0.79 0.54 0.22 0.14 0.23 0.19 0.35 0.21 0.06 0.09 0.08 0.11 0.08 0.08

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.15 0.27 0.16 0.15 0.19 0.22 0.28 0.32 0.22 0.09 0.05 0.09 0.08 0.14 0.09 0.03 0.04 0.03 0.04 0.03 0.03

4.029 4.079 4.033 4.037 4.037 4.068 4.057 4.093 4.064 4.017 3.924 4.013 3.987 4.041 3.971 3.848 3.884 3.875 3.908 3.881 3.886

the small differences are due to the adoption of a different source for the stellar magnitudes. We adopted the magnitudes given by Claria et al. (1993), while Landstreet et al. (2007) adopted the magnitudes published by Claria (1971). Figure 6 shows that assuming a metallicity of 0.013 would have little impact on the estimate of the cluster age, but just of the cluster distance, decreasing it by about 100 pc. Therefore, we can safely conclude that the distance of the cluster is within the range 670­770 pc. Out HR diagram allows also to better constrain the cluster age. The age of the solar metallicity isochrone fitting the cluster main sequence is log t=8.10, which we can then consider as the minimum cluster age. The maximum age instead depends strongly on which stars of the upper main sequence are blue strugglers. If none of our analysed cluster stars is a blue struggler, the maximum cluster age is most likely to be log t=8.25, while if HD 123226 (the hottest star in our sample) is a blue struggler the maximum cluster age is more likely to be log t=8.30. On the basis of these considerations, our best estimate of the cluster age is log t=8.20±0.10, in comparison to log t=8.20±0.20 previously given by Ahumada & Lapasset (2007)). 6.3 Abundances vs. T and sin i

eff

WEBDA, Mermillod & Paunzen 2003), and the bolometric correction from Balona (1994). Luminosities are listed in Table 6. With a distance modulus uncertainty of 0.20 mag, an uncertainty in the bolometric correction of about 0.07 mag, and a reddening uncertainty of 0.01 mag, we estimate the typical uncertainty in Mbol being about 0.28 mag, corresponding to an uncertainty in log L/L of about 0.10 dex. Two stars of our sample, HD 122983 and HD 123183, are included in the work by Landstreet et al. (2007) and there is a good agreement between the obtained log L/L and M/M values, where

From Fig. 7 to Fig. 9 we display the abundances of chemical elements versus Teff . For some elements, such as Mg and Fe, it is possible to notice a correlation of the abundances with Teff , with an abundance increase with T eff up to T eff 10500 K and an abundance decrease for higher temperatures. A linear fit shows that these correlations are significant for both Mg and Fe (see Table 7). These correlations could in principle not be real, but due to other factors, such as mic , log g and sin i. We excluded that both mic and log g produced such trends, since the abundance uncertainties take into account the mic uncertainty, and the effect of log g variations on the abundances is much smaller than the deviations obtained here. In addition, we did not find any clear correlation between the abundances and sin i, as shown in Fig. 10, except for a decreasing Ba abundance with increasing sin i up to sin i values of about 125 km s-1 , in agreement with predictions by Turcotte & Charbonneau (1993). We find also unlikely that our LTE approximation could be the source of the abundance correlations with Teff , since for both Mg and Fe the expected non-LTE corrections are

Si abundance

Mg abundance

C abundance


A detailed spectroscopic analysis of the open cluster NGC 5460
Ca abundance Al abundance C abundance
2.4 2 1.6 1.2 0.8 0.4 0 -0.4 -0.8 -1.2 2.4 2 1.6 1.2 0.8 0.4 0 -0.4 -0.8 -1.2 8000

11
Sc abundance Si abundance Mg abundance

2.4 1.6 0.8 0 -0.8 2.4 1.6 0.8 0 -0.8 2.4 1.6 0.8 0 -0.8 50 100 150 200 250 300 50 100 150 200 250 300

Ca abundance

Teff [K]

10000

12000

8000

Teff [K]

10000

12000

Cr abundance

Ti abudnance

Sc abundance

Vsini [km/s]
Ni abundance Mn abundance Ti abundance
2.4 1.6 0.8 0 -0.8 2.4 1.6 0.8 0 -0.8 2.4 1.6 0.8 0 -0.8 50 100 150 200 250 300 50

Vsini [km/s]
Ba abundance Fe abundance Cr abundance
100 150 200 250 300

Figure 8. Same as Fig. 7, but for Ca, Sc, Ti and Cr.

2.4 2 1.6 1.2 0.8 0.4 0 -0.4 -0.8 -1.2 2.4 2 1.6 1.2 0.8 0.4 0 -0.4 -0.8 -1.2 8000

Mn abundance

Ba abundance

Ni abundance

Fe abundance

Vsini [km/s]

Vsini [km/s]

Teff [K]

10000

12000

8000

Teff [K]

10000

12000

Figure 10. Abundances relative to the Sun (Asplund et al. 2005) as a function of sin i for B-type stars (open squares), A-type stars (open circles), F-type stars (open triangles) and He-weak stars (crosses).

Figure 9. Same as Fig. 7, but for Mn, Fe, Ni and Ba.

small (Przybilla et al. 2001; Fossati et al. 2009) and their dependence on Teff is not expected to be as pronounced as shown by our results. As further check for the obtained abundance correlations with T eff , we looked for similar trends in other consistently analysed sample of stars covering a large temperature range. Erspamer & North (2003) analysed the spectra of 140 A- and F-type stars, present in the ELODIE archive3 , obtaining for each star funda3

http://atlas.obs-hp.fr/elodie/

Table 7. Correlations and relative uncertainties of the Mg and Fe element abundance as a function of Teff , in two temperature ranges: Teff <10500 K and T eff >10500 K. The last two rows show the correlations obtained for Mg and Fe from the Erspamer & North (2003) data. Element Mg Mg Fe Fe Mg Fe T
eff

slope K K K K -1.71 1.88â10-4 -4.03â10-4 7.23â10- 1.21â10-
5 4



slo pe

<10500 >10500 <10500 >10500

3.04â10-4

â10-4

0.42 1.08 0.47 0.82

â10- â10- â10- â10-

4 4 4 4 5 4

<10500 K <10500 K

1.80â10- 0.21â10-

mental parameters and abundances of several elements. Figure 11 shows a comparison between the results we obtained in NGC 5460 and what published by Erspamer & North (2003) for Mg and Fe. We register an agreement for both elements, in particular for Fe (see Table 7), although the trends appear less pronounced in the results published by Erspamer & North (2003). Unfortunately the sample of stars analysed by Erspamer & North (2003) did not cover stars with temperatures higher than 10500 K, so that it was not possible to check the presence of the abundance drop, present in our results. Diffusion processes could be responsible for the obtained abundance correlations with Teff . With the increase of the effective temperature the hydrogen and helium ionisation zones shift more and more towards the stellar surface, and as a consequence they become thinner and less turbulent. In this way diffusion processes become more and more effective. For example, it is well known that Ba is overabundant in early-type stars, which is a clear signature of an influence of diffusion even in chemically normal stars (Fossati et al. 2009). Another example is the presence of a correlation between strontium and oxygen or strontium and magnesium, obtained for chemically normal B-type stars by Hempel & Holweger (2003) (strontium is radiatively driven outwards, Mg and O sink down). The efficiency of diffusion processes is reduced by rotation. In particular, simulations carried out by Turcotte & Charbonneau (1993) show that for rotational velocities higher than 125 km s-1 the envelope is completely mixed. This would suggest that correlations between abundances of diffusion indicator elements (such as


12
2.4 2 1.6 1.2 0.8 0.4 0 -0.4 -0.8 -1.2 2.4 2 1.6 1.2 0.8 0.4 0 -0.4 -0.8 -1.2

L. Fossati et al.
the abundance analysis, we built the HR diagram of the cluster. From a comparison between the cluster HR diagram and isochrones by Marigo et al. (2008) we obtained that the cluster distance lies in the range 670­770 pc and that the cluster age is log t=8.20 ± 0.10, where the maximum age depends anyway on which stars of the upper main sequence are blue strugglers. We plotted the abundances of the elements analysed in most of the stars of our sample as a function of sin i and Teff . We did not find any clear abundance trend with respect to sin i, except for barium, for which we register a decrease of the abundance with increasing sin i, up to a sin i of about 125 km s-1 . For some elements, particularly Mg and Fe, we found trends between abundance and Teff ; their abundances increase up to a Teff value of about 10500 K, then decrease with increasing Teff . This abundance trend is marginally present also in the sample of stars analysed by Erspamer & North (2003), but their results do not cover the hotter region, where we found the abundance decline. We suggest that these trends might be signatures of diffusion in early-type stars.
Fe abundance
8000

Mg abundance

Mg abundance

Teff [K]

10000

12000

8000

Teff [K]

10000

12000

Figure 11. Comparison between the results obtained in this work (upper panels) and by Erspamer & North (2003) (lower panels), for Mg (on the left side) and Fe (on the right side). B-type stars are denoted with open squares, A-type stars with open circles, F-type stars with open triangles and He-weak stars with crosses.

Fe abundance

ACKNOWLEDGMENTS Sr, O, Mg, and Ba) and sin i should be obtained even for chemically normal star. But correlations between abundance and sin i are not found here, except for Ba, one of the diffusion indicators, for which the abundance decreases mildly with sin i, up to sin i values of 125 km s-1 , in agreement with what predicted by Turcotte & Charbonneau (1993). Probably a much larger sample of stars, which includes chemically normal low sin i stars could show the presence or not of these correlations. Based on observations made with ESO Telescopes at the Paranal Observatory under programme ID 079.D-0178. Astronomy research at the Open University is supported by an STFC rolling grant (L.F.). C.P. acknowledges the support of the project P19503N16 of the Austrian Science Fund (FWF). O.K. is a Royal Swedish Academy of Sciences Research Fellow supported by grants from the Knut and Alice Wallenberg Foundation and the Swedish Research Council. LF is deeply in debt with Denis Shulyak for the calculation of the LLMODELS stellar model atmospheres. We also acknowledge the use of cluster facilities at the Institute for Astronomy of the University of Vienna. We thank the anonymous referee for the many useful comments which improved the original manuscript.

7

CONCLUSIONS

We obtained low and mid-resolution spectra of several stars in the field of the NGC 5460 open cluster with the FLAMES spectrograph at the ESO/VLT, and we selected 24 probable cluster members with T eff > 7000. This cluster was never previously spectroscopically studied. We calculated the fundamental parameters and performed a detailed abundance analysis for all the stars of the sample, except for three objects (HD 123225, Cl NGC 5460 BB 338, and UCAC 11105152), which turned out to be double line spectroscopic binaries. Two stars, HD122983 and HD 123182, are Heweak stars, and we suspect that the hotter component of the HD 123225 SB2 system is a HgMn chemically peculiar star. Bagnulo et al. (2006) obtained a marginal magnetic field detection for HD 123183, but the star has remarkably high sin i for an Ap star, and no clear chemical peculiarities. The mean abundance pattern obtained for B-, A- and F-type stars considered members of the cluster concentrate around solar values. We determined the cluster metallicity from the cooler stars of our sample (Teff <8500 K). The resultant metallicity is Z = 0.014 ± 0.003. We rederived Z also from the mean iron abundance ([Fe/H]= -0.18 ± 0.22 dex), adopting the approximation given in Eq. 2, obtaining Z = 0.013+0..008 . We believe anyway that this metal-0 005 licity is underestimated, due to two stars of the sample which show a particularly low iron abundance, and, without taking into account these two objects, the metallicity turned out to be almost solar (Z = 0.017+0..005 ). -0 004 For each star of the sample we derived photometrically log L/L and, adopting the spectroscopic temperatures derived for

REFERENCES Ahumada, J. A., & Lapasset, E. 2007, A&A, 463, 789 Anders, E. & Grevesse, N. 1989, Geochimica et Cosmochimica Acta, 53, 197 Asplund, M., Grevesse, N., & Sauval, A. J. 2005, in ASP Conf. Ser. 336: Cosmic Abundances as Records of Stellar Evolution and Nucleosynthesis, ed. Thomas G. Barnes III & Frank N. Bash, 25 Bagnulo, S., Landstreet, J.D., Mason, E., et al. 2006, A&A, 450, 777 Balona, L. A. 1994, MNRAS, 268, 119 Canuto, V. M., & Mazzitelli, I. 1991, ApJ, 370, 295 Canuto, V. M., & Mazzitelli, I. 1992, ApJ, 389, 724 Charbonneau, P., & Michaud, G. 1991, ApJ, 370, 693 Claria, J. J. 1971, AJ, 76, 639 Claria, J. J., Lapasset, E., & Bosio, M. A. 1993, A&AS, 99, 1 Dias, W. S., Assafin, M., Florio, V., Alessi, B. S., & L´bero, V. ´ i 2006, A&A, 446, 949 Erspamer, D., & North, P. 2003, A&A, 398, 1121 Fossati, L., Bagnulo, S., Monier, R., et al. 2007, A&A, 476, 911 Fossati, L., Bagnulo, S., Monier, R., et al. 2008a, Contributions of the Astronomical Observatory Skalnate Pleso, 38, 123 Fossati, L., Bagnulo, S., Landstreet, J, et al. 2008b, A&A, 483, 891


A detailed spectroscopic analysis of the open cluster NGC 5460
Fossati, L., Ryabchikova, T., Bagnulo, S., et al. 2009, A&A, 503, 945 Fossati, L., Mochnacki, S., Landstreet, J., & Weiss, W. 2010, A&A, 510, A8 Fuhrmann, K., Pfeiffer, M., Frank, C., Reetz, J., & Gehren, T. 1997, A&A, 323, 909 Gebran, M., & Monier, R. 2005, Memorie della Societa Astronomica Italiana Supplement, 8, 200 Gebran, M., & Monier, R. 2008, A&A, 483, 567 Gebran, M., Monier, R., & Richard, O. 2008, A&A, 479, 189 Gebran, M., Vick, M., Monier, R., & Fossati, L. 2010, A&A, in press, arXiv: 1006.5284 Heiter, U., Kupka, F., van't Veer­Menneret, C., Barban, C., et al. 2002, A&A, 392, 619 Hempel, M., & Holweger, H. 2003, A&A, 408, 1065 Hog, E., Fabricius, C., Makarov, V. V. et al. 2000, A&A, 355, L27 Hubrig, S., & Mathys, G. 1996, A&AS, 120, 457 Iliev, I. K., & Budaj, J. 2008, Contributions of the Astronomical Observatory Skalnate Pleso, 38, 129 Kharchenko, N. V., Piskunov, A. E., Roser, S., Schilbach, E., & ¨ Scholz, R.-D. 2004, AJ, 325, 740 Kharchenko, N. V., Piskunov, A. E., Roser, S., Schilbach, E., & ¨ Scholz, R.­D. 2005, A&A, 438, 1163 Kochukhov, O. 2007, Spectrum synthesis for magnetic, chemically stratified stellar atmospheres, Physics of Magnetic Stars, 109, 118 Kupka, F., Piskunov, N., Ryabchikova, T. A., Stempels, H. C., & Weiss, W. W. 1999, A&AS, 138, 119 Kurucz, R. 1993, ATLAS9: Stellar Atmosphere Programs and 2 km/s grid. Kurucz CD-ROM No. 13 (Cambridge: Smithsonian Astrophysical Observatory) Landstreet, J. D. 1998, A&A, 338, 1041 Landstreet, J. D. 1988, ApJ, 326, 967 Landstreet, J. D., Bagnulo, S., Andretta, V., et al. 2007, A&A, 470, 685 Marigo, P., Girardi, L., Bressan, A., et al. 2008, A&A, 482, 883 Mashonkina, L. 2011, Magnetic Stars-2010, Proceedings of an international conference held in N. Archyz, Russia, 27-30 August 2010 (in press) Mermilliod, J.-C., & Paunzen, E. 2003, A&A, 410, 511 Pasquini, L., et al. 2002, The Messenger, 110, 1 Piskunov, N. E., Kupka, F., Ryabchikova, T. A., Weiss, W. W., & Jeffery, C. S. 1995, A&AS, 112, 525 Przybilla, N., Butler, K., Becker, S. R., & Kudritzki, R. P. 2001, A&A, 369, 1009 Rentzsch-Holm, I. 1996, A&A, 312, 966 Robichon, N., Arenou, F., Mermilliod, J.-C., & Turon, C. 1999, A&A, 345, 471 Ryabchikova, T. A., Piskunov, N. E., Stempels, H. C., Kupka, F., & Weiss, W. W. 1999, Phis. Scr., T83, 162 Ryabchikova, T., Fossati, L., & Shulyak, D. 2009, A&A, 506, 203 Shulyak, D., Tsymbal, V., Ryabchikova, T., Stutz Ch., & Weiss, ¨ W. W. 2004, A&A, 428, 993 Tsymbal, V. V. 1996, in ASP Conf. Ser. 108, Model Atmospheres and Spectral Synthesis, ed. S. J., Adelman, F., Kupka, & W. W., Weiss, 198 Turcotte, S., & Charbonneau, P. 1993, ApJ, 413, 376 Villanova, S., Carraro, G., & Saviane, I. 2009, A&A, 504, 845 von Zeipel, H. 1924, MNRAS, 84, 684 Wade, G. A., Bagnulo, S., Kochukhov, O., Landstreet, J. D., Piskunov, N., & Stift, M. J. 2001, A&A, 374, 265 Zacharias, N., Urban, S. E., Zacharias, M. I. et al. 2004, AJ, 127, 3043

13

This paper has been typeset from a T X/ LTEX file prepared by the EA author.