Abundances in M92

Center for Stellar and Planetary AstrophysicsMonash University

Summary prepared by John Lattanzio, Oct 2003

Abundances in M92

M92: Everything you need to know!

What’s special abut M92?

One of the most metal-poor:

[Fe/H] = -2.2

One of the oldest:

16Gyr

(according to Grundahl et al 2000)

Basic Parameters for M92

[Fe/H] = -2.2 Age = 16 Gyr C = 1.81 Distance = 27,000 ly Mass = 330,000 Msun

It all started with….Carbon et al 1982 Observed 71 red giants (above

HB)

Solar

It all started with….Carbon et al 1982 Observed 71 red giants, above

HB

Langer and Kraft 1984

Looked at C, N and C+N in various populations

M3 and M13 (same [Fe/H]) Field Giants M92 and M15 (same [Fe/H])

Langer and Kraft 1984: M3 and M13

Langer and Kraft 1984: Field giants

Langer and Kraft 1984: M92 and M15

Langer and Kraft 1984

Average abundances

Norris and Pilachowski 1985

C+N may be bimodal

N correlates with Na (in all 4 giants studied!)

Langer et al 1986 Clear decrease of C with L From as low as Mv=1.5

Pilachowski 1988

C, N and O in 6 giants C+N+O is very constant…

Sneden et al 1991

9 giants No variations in [Fe/H] from star to

star

Sneden et al 1991

Two groups: O-rich and O-poor?

Sneden et al 1991

Definite evidence for ON cycling!

Sneden et al 1991

Possible variation of O with L?

C, N, O, and Na

Clear decrease of C with L Corresponding increase of N with L C+N+O constant for some stars Large spread in C & N at any given L ON cycling has occurred in some

giants No O variation with L No Na variation with L

C variation

Clear decrease of C with L from Mv=1.5

Bellman et al 2001

Mv=1.5 is fairly low L…

Bellman et al 2001

The bump in the LF is at Mv=-0.4

Nearly 2 mag difference…

LF Bump

First Dredge-Up

Start

Finish

log L = 0.8 below bumpOr 2 mag!

C variation with L

So I think they just forgot about FDU!

FDU changes C from base of GB When L exceeds LF bump then

deep mixing continues (mu gradient removed)

But… Smith and Martell 2003 Measured values of d[C/Fe]/dMV

Get same value above and below LF bump…

C variation with L

Its not clear that FDU and deep mixing should change C at the same rate!!!

Needs work!

Heavy elements: More on Fe

King et al 1998 looked at 3 subgiants in M92:

Checked their data with a standard of similar [Fe/H] (HD 140283) and got same as everyone else…

Heavy elements: More on Fe King et al 1998: 3 subgiants in M92 were

not homogeneous: one was 0.15 dex different to the other two

Gravitational settling? Radiation effects? Richard et al 2002 expect factors of 2 or more in most metal poor systems

Later added data for 2 more subgiants: same average value of [Fe/H]

Heavy elements: More on Fe

Langer et al 1998 Used over 100

lines of various metals and looked at 3 bright giants Two are identical

One differs by 0.18 dex from the other two…

Heavy elements:Shetrone looked at Mg, Al, Eu in 6 giants

Intermediate [Fe/H]

High [Fe/H]

Low [Fe/H]

Heavy elements:Shetrone looked at Mg, Al, Eu in 6 giants

Heavy elements: Neutron capture stuffAronsky et al 1994: 9 giants

Heavy elements: Neutron capture stuffAronsky et al 1994: 9 giants

No variation from star-to-star No variation with evolutionary

state No variation with other elements

At last – something we understand

Heavy elements: 3 subgiants King et al 1998 looked at 3

subgiants: Field star

Mg depleted compared to field: just like giants in M92Na enriched compared to field: just like giants in M92Ba higher compared to field: just like giants in M92

Heavy Elements: Sneden et al 2000

34 giants in M92 (and 31 in M15)


Ca: no variation

Na: Large spread No variation with L

or Te Correlates with N

Ba: no variation


Ba and Eu are useful… Ba is lower in M92 than M15 (same [Fe/H]) Just like M4 and M5

[Ba/Eu] = -0.4 for pure r-process

[Ba/Eu] = -0.4 in M4

[Ba/Eu] = +0.2 in M5

[Ba/Eu] = -0.4 in M92(only 2 stars!)

[Ba/Eu] = -0.4 in M15 also


Si varies a lot from cluster to cluster

M92NGC6752NGC6723

M4M5

[Si/Fe] = +0.59[Si/Fe] = +0.23[Si/Fe] = +0.68[Si/Fe] = +0.55[Si/Fe] = +0.60

Si is primarily made in supernovae from starsWith M=20-25 Msun

Constraints from Li abundances?

Deliyannis et al 19951) 4 subgiants have A(Li) = 2 – 2.5

Boesgaard et al 19981) 7 subgiants have A(Li) = 2 – 2.62) Bonifacio reanalized these stars:

Claims A(Li) = 2.3 0.1 ie little spread


Subgiants now a problem1) They show Na and Al enhancements2) As expected from ON, NeNa and MgAl

cycle3) But Li not destroyed!4) Wherever the hot H burning happened,

the Li was added afterwards


Pilachowski et al 20001) 60 giants in M922) None have A(Li) > 0 (Te = 4500K) > 1 (Te = 5000K)Is this consistent with first dredge-up?Is it consistent with deep mixing? Should

make some Li!

Summary

Clear evidence for deep mixing on GB via the C and N variations

ON cycling has produced N and Na Some Al made and Mg destroyed at the same

time as the ON cycling and Na production Na (etc?) variations seen in subgiants also Some variation in Fe from star to star? From

giant to subgiant? Pure r-process in earlier life, no s-process Need mixing and primordial variations Is Li a problem? Or a useful constraint?

References

Armosky et al, 1994, AJ, 108, 1364 Bellman, et al, 2001, PASP, 113, 326 Boesgaard et al, 1998, ApJ, 493, 206 Bonifacio, 2002, A&A, 395, 515 Buonanno et al, 1985, A&A, 145, 97 Carbon et al , 1982, ApJS, 49, 207 Deliyannis et al, 1995, ApJ, 452, L13 Grundahl et al, 2000, AJ, 120, 1884 King et al, 1998, AJ, 115, 666 Langer & Kraft, 1984, PASP, 96, 339

Langer et al, 1986, PASP, 98, 473 Langer et al, 1998, AJ, 115, 685 Norris & Pilachowski, 1985, ApJ, 299,

295 Pilachowski, 1988, ApJ, 326, L57 Pilachowski et al, 2000, AJ, 119, 2895 Richard et al, 2002, ApJ, 580, 1100 Shetrone, 1996, AJ, 112, 1517 Smith & Martell, 2003, PASP, 115,

1211 Sneden et al, 1991, AJ, 102, 2001 Sneden et al, 2000, AJ, 120, 1351

Abundances in M92

Documents

Transcript of Abundances in M92