Next Contents Previous

6. THE ABUNDANCE OF ELEMENTS IN THE UNIVERSE

6.1. Experiments and Observations

The essence of every science is the ability to validate theories by comparison to experiments. The ability to perform guided experiments is somewhat limited in astrophysics because of the large scales and extreme conditions involved. Nevertheless, there is a wealth of data to be exploited. Atomic transitions and certain nuclear reactions (Käppeler et al. 1998) can be studied in the laboratory with current methods.

Since the s-process (Sect. 4.5.1) involves mainly stable nuclei, an important experimental contribution is the one studying neutron capture at low energy. Special focus is put on reactions at s-process branchings. For the majority of unstable nuclei reached through the s-process, β-decays are much faster than the neutron captures. At several places in the nuclear chart, however, the s-process path encounters long-lived nuclei for which the decay rate becomes comparable to the neutron capture rate. This leads to a splitting of the path: a fraction of the s-process flow proceeds through neutron capture, the other through the decay, bypassing certain isotopes. A comparison of abundances of nuclei reached through one or the other branch provides information on the relation between decay and capture rate. Due to the different temperature dependence of capture and decay, the branching ratio is dependent on temperature, becoming a sensitive s-process "thermometer." The capture rate also depends on the neutron density and thus such a branching can also be used as a neutron pycnometer. This yields detailed information on the conditions inside an AGB star. To be useful, neutron capture cross sections have to be known with an accuracy of better than 1% below about 50 keV neutron energy. This has been achieved for some target nuclei but remains a challenge for others. High-resolution time-of-flight experiments are most promising to give the required accuracy and provide cross sections within the required energy range. The cross sections have to be converted to reaction rates to be applied in astrophysical models. The Karlsruhe group (Käppeler et al. 1989) has had a leading role in directly determining rates by using a neutron spectrum created by the 7Li(p, n)7Be reaction. Through this trick, the resulting energy distribution of the released neutrons is very similar to the one of neutrons in stellar plasma with thermal energy of 25 keV, coinciding with s-process conditions. The limitation of this technique is that it can only provide the spectrum at this energy.

The predictions of reaction models giving cross sections relevant to the production of p-nuclei and of the nuclear properties required in such calculations can also be tested by using neutron and charged particle reactions on stable targets (Descouvemont and Rauscher 2006). Some of the accessible reactions are directly important in the nucleosynthetic processes while other experiments can only serve as tests of the theoretical approaches (Kiss et al. 2008). The reactions can either be studied in online beam experiments detecting directly emitted γ-rays or particles. Another important type of experiment is that of activation (see, e.g., Gyürky et al. 2006, and references therein). A material sample is activated by neutron, proton, or α beams at the energy of interest and the long-term radiation is counted over an extended period of time. Alternatively, the amount of the nuclei produced by the activation can be measured by the very sensitive accelerator mass spectrometry (AMS), which has become an important tool also for astrophysical measurements. It is especially well suited to study neutron-induced reactions producing different isotopes of the same element.

The advent of radioactive ion beam (RIB) facilities in nuclear physics allows studying the properties of and reactions with unstable nuclei (Käppeler et al. 1998; Thielemann et al. 2001b; Rauscher and Thielemann 2001). In addition to the few, already existing, smaller RIB facilities in Europe, Japan, and the USA, one large-scale facility is under construction at GSI Darmstadt, Germany, and another large-scale facility has recently been funded in the USA, the Facility for Rare Isotope Beams (FRIB) at Michigan State University. Both the GSI FAIR (Facility for Antiproton and Ion Research) and FRIB will allow producing highly unstable nuclides, both on the proton- and neutron-rich side of the chart of nuclides. For the first time, this will enable us to study nuclear properties of the p- and rp-processes directly, and also close to the r-process path. These investigations will largely improve the understanding of explosive nucleosynthetic processes.

Plasma physics experiments describe the properties of hot and thin plasmas. Utilizing data from laser-induced plasmas or nuclear testing allows drawing conclusions on the behavior of matter under conditions that are to a certain extent close to those found in stellar environments. Hydrodynamic simulations can also be validated against test cases drawn from experiments and terrestrial experience. However, large nuclear reaction networks involving highly unstable nuclei, extended stellar atmospheres with complicated mixing processes, or macroscopic amounts of matter at and beyond nuclear densities are only accessible by theoretical methods. The models have then to be tested ultimately against astronomical data. The latter have almost exclusively been observations in the electromagnetic spectrum, starting from the ancient observations of visible light coming from the Sun and the stars, to modern satellite observatories also exploring other frequency ranges and studying emissions of compact objects, faint galaxies, accretion disks, quasars, and the echo of the Big Bang, the cosmic microwave background. It is amazing how much has already been learned about the structure and history of the whole Universe by just examining the faint light reaching the surface of our tiny planet. The upcoming new missions of ground- and space-based observatories guarantee an increasing inflow of data, securing the development of the related fields and ensuring that this research field stays exciting and is still able to provide new insights.

In addition to the observation in the electromagnetic spectrum, other means of obtaining astrophysically relevant information become increasingly important. Among those are measurements of cosmic rays, on the surface and in the atmosphere of Earth, as well as in low-earth orbits (Westphal et al. 1998, 2001). Such investigations provide insights regarding the particle flux in our solar system, originating from the Sun and from galactic sources. In the future, it will be possible to study another type of radiation in addition to the electromagnetic one: gravitational waves emitted, e.g., by neutron star mergers, black hole formation, and other interactions between highly massive or relativistic objects (Sathyaprakash and Schutz 2009). Even more important for studying nucleosynthesis and stellar evolution are the isotopic ratios found in certain meteoritic inclusions (Lugmair et al. 1983; Lewis et al. 1987; Hoppe and Zinner 2000). Using advanced chemical extraction methods, these data can be utilized to deduce the composition of the material in the early solar system. Because of its growing importance for nucleosynthesis studies, the field is introduced in a separate subsection (Sect. 6.3).

6.2. Solar Abundances

As might have become evident from the previous sections, the origin of the elements can only be understood in detail when the physics of the nucleosynthesis sites is also understood. Complete simulations of nucleosynthetic events that account for all data are required, not just those giving details of elemental or isotopic abundances. Nevertheless, the determination of abundances remains central if one wants to study nucleosynthesis.

By inspection of the absorption lines in stellar spectra, it is possible to measure the contents of the stellar atmosphere. A theoretical model has to explain how many of those nuclides were inherited from the proto-cloud from which the star formed and what amount was produced in the central, nuclear burning regions of the star itself and brought up by convection.

Similar considerations apply to the observation of absorption and emission lines from other objects, such as supernova ejecta, planetary nebulae, and interstellar clouds. Some methods and results were already presented in Sect. 3.3.

On the way to understand the origin of elements on Earth, the abundances in the Sun have to be explained first because the planets and the Sun formed from the same interstellar cloud. Due to their low gravitation, the smaller planets subsequently lost those light elements that were not chemically bound in their crust whereas the Sun was able to retain more or less the original composition. In planets, physical and chemical fractionation processes then separated certain elements or isotopes and concentrated them in different regions, leading to the heterogeneous distribution found, e.g., in geological surveys. The solar composition is shown in Fig. 13. As an example of how the solar composition affects nucleosynthesis studies is shown in Fig. 12. In order to constrain the relative contributions of the s- and r-process, the solar abundances are used to represent the current composition of the local interstellar medium.

The accurate determination of solar abundances is, therefore, central to all investigations of nucleosynthesis. This is reflected in the recent commotion caused by a new study of solar abundances (Asplund et al. 2006a, b), revising the previously widely used tables of Anders and Grevesse (1989). The new abundances are based on modern three-dimensional model atmospheres (describing the region in the Sun where line absorption occurs). The content of elements beyond H and He was found to be lower by a factor of two compared to the previous study. This finding impacts all kinds of comparative nucleosynthesis studies but mainly those involving C, N, and O, which are the most abundant (apart from H and He). The new abundances resolved previous problems regarding the consistency of solar abundances and those of the solar neighborhood. On the other hand, they challenge current models of the solar interior, especially regarding the comparison of predictions of the local sound speed (which is also dependent on the abundances of C, N, O, Ne) and helioseismological results.

6.3. Meteoritic Inclusions

In addition to abundance determinations from stellar spectra, another way to obtain information about the composition of the early solar system as well as of environments that are more distant has become increasingly important in the last years. Certain types of meteorites contain inclusions wherein the composition of the early presolar cloud is conserved. This also enables one to study some isotopic compositions that cannot be extracted from the solar spectrum. More surprisingly, some meteorites contain so-called presolar grains, which are supposed to have been formed from material of other stars. The grains traveled through interstellar space with their inherent speed at formation and were incorporated into the protosolar cloud from which the Sun and its planets formed. Some of them survived the formation of the solar system and also the fall as a meteorite. This requires that their host material never experienced temperature above about 1,000 K. Various types of meteorites, most prominently carbon-rich ones (carbonaceous chondrites) carry such nm- to μm-sized inclusions, which were incorporated into the presolar cloud and were since then shielded from any chemical or physical fractionation and mixing processes occurring during planet formation or in the Sun.

Literally being "stardust", presolar grains provide information not just on the isotopic composition of other stars that cannot be determined through their spectra. Additionally, they may show the composition of different layers of a star depending on their formation process. There are a number of excellent reviews on the topic (e.g., Nittler 2003; Clayton and Nittler 2004; Zinner et al. 2006; and the book by Lugaro 2005). Therefore, here only a summary of the most important aspects regarding types and origins of presolar grains and the methods to analyze them (see also Chap. 54 of Vol. 5) is presented.

Presolar grains are foremost identified by their nonsolar isotopic composition. They are subsequently classified by the mineral phase carrying the isotopes and by the isotopic ratios of certain elements (mainly C, N, O, Si, Al, and Fe). Most abundant but the least understood are nanodiamonds. Best studied are the second most abundant SiC grains. Further phases include, in order of abundance, graphite, TiC, ZrC, MoC, RuC, FeC, Fe–Ni metal, Si3N4, corundum, spinel, hibonite, and TiO2.

The well-studied SiC grains can be subdivided into different classes according to the isotopic anomalies (relative to solar) they exhibit. The bulk of 90% is made up of so-called mainstream grains, which are thought to originate from AGB stars, showing almost pure s-process isotopic ratios. They are inferred to have formed in the winds of AGB stars or planetary nebulae. Therefore, they are a snapshot of the surface composition of the star but AGB stars, contrary to other types of stars, have strong convection, carrying freshly synthesized nuclides from the burning zone deep inside the star to the surface (Sect. 4.5.1). Much of the recent progress on the nucleosynthetic details of the s-process in recent years is due to the analysis of presolar grains.

A small subclass of SiC grains, the so-called type X grains, contain a large 26Mg/24Mg ratio and large excess of 44Ca, pointing to a core-collapse supernova origin. The radionuclides 26Al and 44Ti are concurrently produced only in such supernovae and decay to 26Mg and 44Ca, respectively. The SiC X grains are thought to be formed in supernova ejecta. As these consist of a large fraction of the progenitor star, grains can also condensate from material of inner layers or from a mixture of different layers of the star. There is some success in reproducing X grain compositions by mixing abundances predicted by current stellar models.

The origin of other types of grains is still debated and a unique identification with a site is not always possible. They could have been produced in supernovae, AGB stars, or novae.

The analysis of the content of a grain requires a combination of chemical and physical methods to separate the grain from the surrounding meteoritic material and to determine the isotopic abundances contained within. Luckily, mineral phases condensing in the vicinity of stars contain acid-resistant phases, which can be separated from the meteorite by essentially dissolving everything else away. Until recently, this was the way to go but it has the disadvantage that other, less acid-resistant, presolar phases, e.g., silicates, may be lost in the process. Once the grain material has been isolated and concentrated, standard mass spectrometric methods (with AMS being the most sensitive) can be applied. A complete analysis of the presolar content, including more easily dissolvable materials, requires new analytic methods, currently under development. Among those is resonance ionization mass spectrometry (RIMS), which allows measuring ppm-level trace elements in μm-sized grains with elimination of isobaric interferences. Even more promising is the NanoSIMS, an ion microprobe with high sensitivity and high spatial resolution (Stadermann et al. 1999; Marhas et al. 2008). The NanoSIMS allows studying a grain in a slice of a meteorite to obtain isotopic abundances with information on their location within the sample. This enables studies of layered grains in their meteoritic matrix and also grains made up of easily dissolvable phases. This is superior to TEM (transmission electron microscope) analysis, which requires ultrathin samples cut with diamond knives and losing some depth information. Nevertheless, the TEM can sometimes be complementary to a NanoSIMS analysis.

Presolar grains open a new, promising window into the Universe by enabling us "hands on" analysis of nonsolar, stellar matter. With improved preparation and analysis methods, this line of research will remain important for many types of nucleosynthesis studies, even directly impacting the theory of stellar structure and evolution.

6.4. Galactical Chemical Evolution: Putting it all Together

Although the Sun is considered to exhibit a composition typical of the disk of the Galaxy, it has to be realized that its abundances are only a snapshot in time. As is evident from the discussion in the preceding sections, the Sun contains elements, which have been produced in other stars or galactic sites and the solar abundances (Fig. 13) are only a snapshot of the composition of the interstellar medium. After the first stars in the galaxies lit up, numerous generations of stars have contributed to the elemental contents of the interstellar material from which new stars form. The current stars, on the other hand, are building up material that will be incorporated in future stellar generations. All this is illustrated by the fact that a gradient in metallicity is observed depending on the age of the star (see Sect. 3.3). Older stars contain less "metals", i.e., elements other than H and He, because the interstellar medium from which they formed contained less. Therefore, it becomes obvious that the interstellar medium in a galaxy becomes enriched with elements over time. The general picture is that of a cycle of matter within a galaxy as shown in Fig. 14.

Figure 14

Figure 14. Schematic view of the cycle of matter in a galaxy.

The primordial galactic material is processed and reprocessed in star-forming regions many times. Indeed, a general enrichment can actually be found when comparing galaxies of different ages.

We do not know how many generations of stars have contributed to the solar abundances and how well the products were mixed into the proto-solar cloud. Therefore, a comparison with abundances in old stars allows drawing conclusions on the relevant processes. For example, recent observations in stars found in the halo of the Galaxy show that the relative r-process abundances are very similar to the ones in the Sun, although very much depleted (Sneden et al. 2000; Cayrel et al. 2001; Frebel et al. 2005). This indicates that the r-process seems to be robust, i.e., occurring under almost the same conditions and giving almost the same elemental yield in each event.

For a complete understanding of the chemical evolution of a galaxy it is necessary to integrate over the yields of all possible contributors. With the advent of advanced stellar models, galactic chemical evolution has become a field of its own, providing further constraints to the nucleosynthetic models. Many considerations enter, such as the amount and composition of ejecta per event, frequency of events, and mixing processes that distribute matter within a galaxy (Pagel 1997). Thus, all available knowledge is combined to reach an improved level of understanding. However, the young field of galactic chemical evolution still faces major difficulties due to the timescales involved, the limitations in observations and models, and the impossibility of accurately dating stars and galaxies. One of the big questions is how material is mixed and transported. Nevertheless, recent promising trends in modeling galactic evolution might even provide constraints, e.g., on individual supernova models rather than only on global properties of SN II and SN Ia. The reason for this possibility is the fact that there is no instantaneous mixing of ejecta with the interstellar medium, and therefore early phases of galactic evolution can present a connection between low metallicity star observations and a single supernova event (Argast et al. 2000).

To summarize, it is worth emphasizing again the tremendous achievements obtained over the last decades. Accurate and detailed studies made it possible that, coming from more simple observations of the Sun and nearby stars, one reached the stage where details of the origin of chemical elements and their isotopes on our planet as well as the evolution of their abundances in entire galaxies and in the early Universe can be studied. It is especially amazing that all this knowledge was gathered without really or just barely leaving the surface of our planet.

Future efforts in nuclear physics and astronomy ensure that the stream of data will not be cut off and will greatly improve our detailed knowledge not only of the origin of the elements but also of the position of our Galaxy, our planet, and ourselves within a vast, evolving Universe.

Next Contents Previous