Stellar Chemical Abundances Analysis

Stellar chemical abundances analysis is the basic means of exploring the evolution of stars, the Milky Way, and even the universe. By obtaining high-resolution spectra of stars, we can not only determine their chemical abundances and spatial distribution, but also know the chemical distribution of different stellar populaitons of the Milky Way. And constraints to the galactic formation, structure, and evolution model can be provided. By comparing the stellar evolution model with observation, one can trace the chemical history of the Milky Way, understand nucleosynthesis, and examine the current cosmology models.

Chemical Evolution of Different Stellar Populations: New Challenge to Existing Models

Large-scale structures of the universe and stellar evolution theory are well understood. But the galactic formation and evolution, the bridge connected the universe and stars, is still rather vague, which among the most important problems in modern astrophysics. The Milky Way, where out solar system located is an ideal sample of studying the formation and evolution of galaxies.

Metallicity of stars can be seen as "clock". Chemical abundances determination enable us to trace the history of galactic evolution. We selected three groups of stars with different metallicities to carry out large sample systematic analysis, which are:

  • Stars with metallicty([Fe/H]) range from -3.0 to -1.0[1][2], the representative of the extreme metal-poor, old halo stars.
  • Stars with metallicty([Fe/H]) range from -1.6 to -0.7 [3][4], which is the moderate metal-poor and might be thick disk stars.
  • Stars with metallicty([Fe/H]) range from -1.0 to +0.2 [5], the representative of the metal-rich, young disk population.

Through high-resolution and high signal to noise ratio spectra, we have identified abundance trends of more than 20 chemical elements, calculated their kinematic parameters, revealed characteristics of different stellare population, and finally describe a scenario of galactic chemical evolution.

In order to explain the chemical evolution of stellar population, we estabilished the "three components of the chemical evolution" model of the Milky Way[6], to reproduce the evolution trends of different elements, and found three types of abnormal stars which can not be interpreted in traditional theories. That is, the old metal-rich stars, kinematic anomaly stars and low [α/Fe] ratio stars. According to the traditional theory, early formed stars usually have low metallicities. While we found several old, but metal-rich stars[7], challenging the theories. The second one, kinematic anomaly stars, rotating reversely around the Milky Way, which may relfect the interaction between our Galaxy and others[8]. The third one is so-called low-[α/Fe] ratio stars. [α/Fe] means the relative ratio of α elements (such as oxygen, calcium, magnesium, etc.) and iron.

Satellite galaxies around our Milky Way are mainly dwarf galaxies, which evolve very slow. Due to the type Ia supernova, their [α/Fe] ratio are slightly lower than the Milky Way. We found the relevance of α abundances of halo stars and their kinetic parameters, provided evidence to two components model of halo. These observations and discoveries put forward a strong challenge to the traditional framework of chemical evolution of galaxies.

In addition, we also studied the relationship between red clump stars which often seen as "standard candles" and metallcities. Red clump stars are low-mass stars which are in core helium burning, have roughly the same brightness, and thus can be used as "standard candles" to determine the distances. We found that red clump stars can be divided into two types. One is metal-rich ones and the other is metal-poor ones. Further study showed that the relevance between I-band absolute magnitude and metallicity[9].

The Study of NLTE Effect: Resoving the Puzzle of r-/s- Process of Heavy Elements.

The vast majority of the current analysis of abundance are based on LTE(Local Thermodynamic Equilibrium) assumption. While this assumption is not suitable for each of the elements, and in some astrophysical environment which may result in serious deviations from the real situation, thus leading to wrong conclusions. In other words, we need NLTE(Non-Local Thermodynamic Equilibrium) effects of the amendments to further improve the accuracy of the abundance analysis.

We added a creative term that ralated to the stimulating potential and atomic principal quantum number and fit the optical-infrared neutral magnesium profiles very well. NLTE effects are particularly notable for metal-poor dwarfs and giants. The abundance corrections can reach 1.6 times. We also become one of few research team who can quantitatively calculate NLTE effects in the world.

Furthermore, we established atomic models of Li, Na, Al, K, Sc and Fe, and determined the NLTE effects in solar spectrum. The results of potassium abundance fit the that in meteorites very well[10]. Next, we applying those atomic model on some stars, for the first time found the different evolution patterns of different stellar populations in the Milky Way. In addition, using the [Na/Fe], [Mg/Fe] and [Al/Mg] abundance ratios via NLTE calculation, we successfully gave the criteria of halo, thick disk and thin disk stars[11][12][13]. Previous studies are not able to determine which stellar population belongs to for an individual star. However, our high-precision NLTE analysis made it possible.

Finally, NLTE study was successfully applied to solving the puzzle of r-/s- process of heavy elements. Generally accepted that the heavy elements in metal-poor stars are mainly come from rapid-neutron process (r-process) [14], such as barium (Ba) elements. While heavy elements in the sun are mainly produced in the slow-neutron process (s-process). This is the so-called r-/s- process puzzle. Previous studies are mainly restricted in LTE assumption. In recent years, with the development of NLTE studies, we were able to use it on resolving the controversy. Through a complete barium model, we analysed high-resolution spectra of 25 dwarfs, and found that with the decrease in metallicity, r-process contribution to barium ratio increased, and reached 33% in thick disk stars, and can achieve 67% in old halo stars, becoming a major contributor to nuclear synthesis[15]. Thus, by taking into account the hyperfine structure of spectral lines and NLTE effect, the r-/s- problem successfully resovled.


  • [1] Zhao G., Magain P., 1990, A&A, 238, 242.
  • [2] Zhang H.W., Zhao G., 2005, MNRAS, 364, 712.
  • [3] Zhang H.W., Zhao G., 2006, A&A, 449, 127.
  • [4] Qiu H.M., Zhao G., Takada-Hidai M. et al., 2002, PASJ, 54, 103.
  • [5] Chen, YQ, Nissen, PE, Zhao, G. et al. , 2000, A&AS, 141, 491.
  • [6] Liang,Y.C., Zhao,G., Zhang,B., 2000, A&A, 363, 555.
  • [7] Chen,Y.Q., Zhao,G., Nissen,P.E. et al., 2003, ApJ, 591, 925.
  • [8] Bai,G.S., Zhao,G., Chen,Y.Q. et al., 2004, A&A, 425, 671.
  • [9] Zhao,G., Qiu,H.M., Mao,Shude, 2001, ApJ, 551, L85.
  • [10] Zhang,H.W., Butler,K., Gehren,T., Shi,J.R., Zhao,G., 2006, A&A, 453, 723.
  • [11] Shi, J. R., Gehren, T., Zhao, G., 2004, A&A, 423, 683.
  • [12] Gehren,T., Liang,Y.C., Shi,J.R., Zhang,H.W., Zhao,G., 2004, A&A, 413, 1045.
  • [13] Gehren,T., Shi,J.R., Zhang,H.W., Zhao,G., Korn,A.J., 2006, A&A, 451, 1065.
  • [14] Truran, JW, 1981, A&A, 97, 391.
  • [15] Mashonkina & Zhao, 2006, A&A, 456, 313.