M5 = NGC 5904 (nearest “intermediate-metallicity” globular cluster

accessible from a northern hemisphere site)

Harris (2003, Feb version)• 7.5kpc from the Sun, 6.2kpc from gal. centre• [Fe/H] = -1.27 (= -1.40 in ZW84)• E(B-V) = 0.03, high galactic latitude (+46.8 deg)• HB index (B-R)/(B+V+R) = 0.31, c = 1.83

By Ivans, Kraft, Sneden, Smith, Rich, Shetrone (2001) vs. M4

- 36 luminous giants and AGB starsBy Cohen, Briley, & Stetson (2002) vs. M71 (nearest from n. h.)

- C & N variations at the base of the RGBBy Ramirez & Cohen (2003) vs. M71 (c.f. 47 Tuc)

- 25 stars covering a wide range in luminosity

CMD of M5 by Sohn & Lee (2000)

-metallicity via photometry-HB morphology

B:V:R = 75:95:56(B-R)/(B+V+R)=0.08

B:V:R = 92:40:32(B-R)/(B+V+R)=0.37

Ivans et al. (2001) - 36 luminous giants and AGB stars

QuickTime™ and aGIF decompressor

are needed to see this picture.

CMD of M5,

showing the positions of program stars on the AGB and RGB.

Symbols given at lower right correspond to the observatory and resolution used for each observation.

Ivans et al. (2001)

<[Fe/H]> = -1.21 based on FeII, adopting non-LTE precepts<[Fe/H]> = -1.34 based on FeI

e.g., M4, <[Fe/H]> = -1.08, redetermined from -1.18

Ivans et al. (2001)

QuickTime™ and aGIF decompressor

are needed to see this picture.

Box plot of the M5 giant star element abundances. For each abundance ratio the "box" contains the middle 50% of the data (i.e., the interquartile range) and the horizontal line inside the box indicates the median value of a particular element. The tails vertical extending from the boxes indicate the total range of abundances determined for each element, excluding outliers. Mild outliers (those between 1.5 and 3 times the interquartile range) are denoted by open circles. Severe outliers (those greater than 3 times the interquartile range) are denoted by filled circles.

Ivans et al. (2001)

QuickTime™ and aGIF decompressor

are needed to see this picture.QuickTime™ and aGIF decompressor

are needed to see this picture.

O and Na anticorrelation, Na and Al and CN correlation, seen in previous studies of other globular clusters, expected proton-capture nucleosynthesis (e.g., CN & ON cycle)

Ivans et al. (2001)

QuickTime™ and aGIF decompressor

are needed to see this picture.

QuickTime™ and aGIF decompressor

are needed to see this picture.

lower log g: higher O and lower Na,contrary to evolutionary scenario

opposite of M13

CN strong vs. weak -- primordial

scatter at given CN -- deep mixing

Ivans et al. (2001)

QuickTime™ and aGIF decompressor

are needed to see this picture.

<More enhanced in clusters>

Ivans et al. (2001)

QuickTime™ and aGIF decompressor

are needed to see this picture.

QuickTime™ and aGIF decompressor

are needed to see this picture.

QuickTime™ and aGIF decompressor

are needed to see this picture.

Proton-capture

s, r-process

heavier a-and Fe peak

Ivans et al. (2001)

QuickTime™ and aGIF decompressor

are needed to see this picture.

HB Index (B-R)/(B+V+R)Lee, Demarque, Zinn (1994)

M71 = -1.00M4 = -0.07

M5 = 0.37

M3 = 0.08M10 = 0.94M13 = 0.97

M2 = 0.96NGC6752 = 1.00

Blue tail vs. Super O-poor??

Cohen, Briley & Stetson (2002) - C & N variations at the base of the RGB

QuickTime™ and aGIF decompressor

are needed to see this picture.

Signal is too low for a detailed analysis

Keck LRIScoverage from 3600 to 4800A

CN band at 3885AG band of CH at 4300A

Main sample is LLG stars:(similar evolutionary state)

Cohen, Briley, & Stetson (2002)

QuickTime™ and aGIF decompressor

are needed to see this picture.

Arbitrarily divided CH-strong and CH-weak sample

Cohen, Briley, & Stetson (2002)

QuickTime™ and aGIF decompressor

are needed to see this picture.

strong anticorrelationamong SGBs

Cohen, Briley, & Stetson (2002)

QuickTime™ and aGIF decompressor

are needed to see this picture.

Briley et al. (1992): more luminous giants

Cohen, Briley, & Stetson (2002)

QuickTime™ and aGIF decompressor

are needed to see this picture.

Requires nearly a 0.75 dex star-to-star variation in [C/Fe] among SGB stars

Cohen, Briley, & Stetson (2002)

QuickTime™ and aGIF decompressor

are needed to see this picture.

C & N anticorrelation among SGB stars

No systematic trends with either luminosity or temperature are apparent in the abundances.

Cohen, Briley, & Stetson (2002)

QuickTime™ and aGIF decompressor

are needed to see this picture.

Sum of the derived C and N abundances is plotted as a function of the C abundance.

Large filled circle marks the location for both C and N depleted by a factor of 16, adopting the abundance of M5 of [Fe/H] = -1.2 dex, with C/N at the solar ratio.

Horizontal line extending to the left of that represents the locus of points for C gradually being converted into N, with the left end of the line having C/C0 = 0.1.

“The range of variation of the N abundances is very large, and the sum of C+N increases as C decreases. To reproduce this requires the incorporation

not only of CN but also of ON-processed material.”

Ramirez & Cohen (2003) - 25 stars covering a wide range in luminosity

Keck (HIRES)wavelength range (6000 ~ 8000A)(c.f., Ivans et al. (5400 ~ 6700A))

Photometry from Stetson et al. (1998)

QuickTime™ and aGIF decompressor

are needed to see this picture.

Ramirez & Cohen (2003)

QuickTime™ and aGIF decompressor

are needed to see this picture.

[Fe/H] from Fe I (top) and Fe II (bottom) against photometric Teff.

The solid lines are linear fits weighted by the errors.

The dashed lines indicate the mean [Fe/H] with their respective error plotted as an error bar at 4000 K.

In both cases, [Fe/H] shows no dependence on Teff.

Note that <[Fe/H](Fe I)> = -1.30 ± 0.02 and <[Fe/H](Fe II)> = -1.28 ± 0.02.

“at this metallicity, non-LTE effects are not important…”

“Thévenin & Idiart (1999) found that non-LTE corrections become more important as [Fe/H] decreases, being about 0.2 dex for stars with [Fe/H] about -1.25 dex, and that ionized lines are not significantly affected by non-LTE.”

Ivans et al. (2001)

<[Fe/H]> = -1.21 based on FeII, adopting non-LTE precepts<[Fe/H]> = -1.34 based on FeI

e.g., M4, <[Fe/H]> = -1.08, redetermined, -1.17 if based on FeIc.f., M71, <[Fe/H](Fe II)> = -0.84 ± 0.12 and <[Fe/H](Fe I)> = -0.71 ± 0.08

Ramirez & Cohen (2003)

QuickTime™ and aGIF decompressor

are needed to see this picture.

Abundance ratios of O and Na with respect to Fe against Teff.

The solid line is a linear fit weighted by the errors.

The dashed line indicates the mean abundance ratio with its respective error plotted as an error bar at 4000 K.

The open triangle corresponds to the abundance determined from the summed spectra of the six m.-s. stars.

Arrows represent upper limits for the oxygen abundance ratio.

Stars G18450_0453 and G18564_0457 with similar Te, part of whose spectra are shown in Fig. 11 (next slide), are marked with squares in the [Na/Fe] plot (bottom).

->The scatter shown by [Na/Fe] is due to real abundance variations

among stars of similar Teff.

Ramirez & Cohen (2003)

QuickTime™ and aGIF decompressor

are needed to see this picture.

G18450_0453 (5170 K, [Na/Fe] = +0.30)

G18564_0457 (5400 K, [Na/Fe] = -0.27).

Ramirez & Cohen (2003)

QuickTime™ and aGIF decompressor

are needed to see this picture.

Summary of abundance ratios in M5.

The thick line on the left side of the box is the predicted error (expected for the interquartile range), which includes the dependence on the stellar parameters and the equivalent-width determination.

Ramirez & Cohen (2003)

QuickTime™ and aGIF decompressor

are needed to see this picture.

QuickTime™ and aGIF decompressor

are needed to see this picture.

Ivans et al. (2001)

O“The difference of +0.24 dex in the mean O abundance

presumably reflects our inability to detect weak O lines

in the O-poor low-luminosity part of our sample in M5,

assuming they are actually present there.”

Al“although the 6696, 6698 Å Al I doublet is

the most useful feature of that element in this spectral region,

we could not get it to fit into a single HIRES setting

together with the O lines.

Ramirez & Cohen (2003)

Ramirez & Cohen (2003) - [Na/Fe] against [O/Fe] for our sample of M5 stars.

QuickTime™ and aGIF decompressor

are needed to see this picture.

QuickTime™ and aGIF decompressor

are needed to see this picture.

Arrows represent upper limits for the [O/Fe] abundance ratio.

The open triangle corresponds to the mean abundance of the six main-sequence stars.

[Na/Fe] against [O/Fe] for stars in M5 from our analysis, showing clear detections (filled triangles), mean m.-s. stars (open triangle), and others from the literature.

The dashed line corresponds to the Na-O anticorrelation present in M4 from the analysis of Ivans et al. (1999), shown as a fiducial line.

Ramirez & Cohen (2003) – very similar abundance ratios btw. M5 & M71

QuickTime™ and aGIF decompressor

are needed to see this picture.

Comparison of the abundance ratios for all elements common to our analysis of similar data for 25 stars in M71 (Ramírez & Cohen 2002; triangles) and in M5 (squares).

may be due to difficulties in the analysis

Cu in M5: based on single line

Briley, Cohen & Stetson (2004, astro-ph/0312315) -- M13Briley, Harbeck & Smith (2004, astro-ph/0312316) -- 47 Tuc

“pollution/accretionvia AGB ejecta”

“But rather than simple surface pollution, a substantial fraction of the present stars’ massesmust be involved.”

C depletion do appear smaller in accord with the prediction of AGB ejecta models of Ventura et al. (2001)…accretion of C-poor materials….to explain the gap & similar spread of [N/Fe], [O/Fe] and [Na/Fe] are needed….