Abundances for M71

Fe-peak elements

[Fe/H] = -0.7 to -0.8
- There is very little variation from star to star
- Derived value depends on whether FeI or FeII is used
- Ramirez et al, 2001, AJ, 122, 1420
- Based on a sample from turnoff to tip GB
Other iron-peak elements show little variation
- Sc, V, Cr, Mn, Co and Ni follow Fe and show no variation with L or Te
- Ramirez and Cohen, 2002, AJ, 123, 3277
- Ramirez and Cohen plot for Sc, V, Cr, and Mn
- Ramirez and Cohen plot for Co. Ni, Cu and Zn
- Based on a sample from turnoff to tip GB

Alpha elements

Mg, Ca, Si and Ti are overabundant relative to Fe, as expected
- There is very little variation from star to star
- Ramirez and Cohen, 2002, AJ, 123, 3277
- Based on a sample from turnoff to tip GB
- <[Ti/Fe]> = +0.20 +- 0.08
- <[Si/Fe]> = +0.28 +- 0.14
- <[Ca/Fe]> = +0.43 +- 0.05
- No sign of variation with Mg
- Ramirez and Cohen plot for Mg, Si, Ca and Ti
Note that these values vary a little from those found by Sneden et al 1994 AJ, 107, 1773
- Based on a sample of 10 stars within 1 mag of RGB tip
- <[Ti/Fe]> = +0.48 +- 0.11 which is a fair bit higher...
- <[Si/Fe]> = +0.31 +- 0.11 which agress well
- <[Ca/Fe]> = +0.14 +- 0.10 which is a fair bit lower!

n-capture elements

Y, Zr, Ba, La and Eu show no trend with L or Te
little scatter, except for Zr, which may be observational (Zr is tough)
Ramirez and Cohen, 2002, AJ, 123, 3277
Based on a sample from turnoff to tip GB
Ramirez and Cohen plot for n-capture elements

C and N

Smith and Norris 1982 were first to find CN variation in 22 stars on GB (ApJ, 254, 159)
Smith and Penny 1989 looked at 16 HB stars: again found CN variation (AJ, 97, 1397)
- The relative fraction of CN strong stars is about the same on the GB and HB at ~65%
- Penny, Smith and Churchill showed that this fraction was also the same on the lower GB as on the HB. There was some evidence for a lower fraction at the tip of the GB. (MNRAS, 257, 89)
Smith and Lambert 1994 showed that the CN strong stars (ie with N up) have lower values of C12/C13, indicative of CN cycling. (ApJ, 424, L119, based on 5 giants)
- They also found the lower C12/C13 values correlated with lower [O/Fe] values, inplicating ON cycling as well
- Briley et al 1997 added another 5 giants and confirmed the trend (AJ 113, 306)
- Plot from Briely et al of [C12/C13 vs [O/Fe]
Briley et al 2001 looked at CN band strengths for 75 giants down to below the HB. (AJ 122 2561)
- They found that all the stars could be fitted with the same C and N abunds
- This means no variation along the GB, obviously
- Hence no deep mixing.....
- And at FDU the material dredged up has essentially the same C and N...
- This means pollution of the outer envelope is not responsible, as the internal mateiral has the same C and N as the surface...
- The "only" (?) possible solution is that the CN-strong stars formed from N-rich material...but that there were still stars made with the normal abundance as well
- This paper also discusses the yields of N and whetehr AGB stars can produce enough! DO we have to alter the IMF as well? This should be updated with recent, more accurate, yields.
- Briely and Cohen 2001 extended this to a sample of MS stars by Cohen (AJ 122 242) and found the same results, down to the MS.

O, Na and Al

Ramirez and Cohen, 2002, AJ, 123, 3277
Based on a sample from turnoff to tip GB
There is a spread of O abundances, but only a very small spread of Na
Ramirez and Cohen plot for C, O, Na and Al
Nevertheless, there seems to be a slight O-Na anti-correlation
This plot shows O-Na for M71 and other clusters.
Al was measured in only a subset of the Ramirez and Cohen sample.
- Al correlates with Na down to the the lowest L measured (MV=+1.8)
- There is a bit of scatter
- This plot shows Al-Na for M71 and other clusters.