Presolar Grains in Meteorites: An Overview
and Some Implications
Ulrich Ott, Max-Planck-Institut für Chemie (Otto-Hahn-Institut),
D-55128 Mainz (email@example.com)
An overview is given of the types and properties of grains of presolar origin that have been identified in primitive meteorites. Some of the inferences that can be drawn from their isotopic compositions are discussed.
I. Before the Beginning
The Solar System as we know it is the result of the collapse of a cloud of interstellar gas and dust. The chemical elements locked up in the dust component have largely been synthesized in stars: astrophysics and chemical evolution tell us that only H, He and some Li are the result of Big Bang Nucleosynthesis, with all elements from carbon on upwards in the chart of the nuclides the result of nuclear processes like those going on in our Sun (Clayton, 1983). Dust that formed in the outflows (winds) of stars or from material explosively ejected during violent events such as supernova explosions carries a signature of the nuclear processes going on in those stars. As a lot of work performed within the past decade has shown, several types of such "circumstellar grains" have survived within primitive meteorites and are available for us to study with all the modern analytical tools available in the laboratory (Bernatowicz and Zinner, 1997). Because it is the deviation from the normal (solar system) isotopic composition that is the diagnostic tool to recognize a grain as being undoubtedly of presolar origin, mass spectrometric analyses have played the most essential part in the analyses so far. Noble gases have had a special role in this context: they are rare in solid materials, thus hints for the presence of presolar grains in meteorites were first seen in the isotopic composition of noble gases, where the effects from the presolar grains "shone through" even in samples largely diluted with isotopically normal solids that formed within the Solar System. The isotopic results allow to draw conclusions regarding the stellar sources and the nucleosynthesis processes responsible for the observed effects plus, in some cases, physical conditions in stars where these processes occurred (Hoppe and Ott, 1997) as well as details of Galactic Chemical Evolution (e.g. Nittler and Alexander, 1999). Two examples will be discussed below (Sec. III). In addition, the simple fact of the survival of the grains puts constraints on the conditions they have encountered in the interstellar medium and in the Solar System.
II. Inventory of Presolar Grains
An overview of the various types of presolar grains identified so far in primitive meteorites, including their physical properties and a list of the elements for which isotopic abundance anomalies have been detected is given in Table 1. Figs. 1 and 2 are TEM and SEM pictures of examples of the most abundant grain types, diamond and silicon carbide (SiC).
Listed abundances are from the compilation of Huss (1997) and refer to the CI meteorite Orgueil (exception: silicon nitride, where the number is for Murchison). Generally the abundances scale with the matrix fraction of the meteorites and decrease with increasing metamorphism (Huss, 1997). Size ranges are from Zinner (1997) and may be biased against small grains.
acontains subgrains of: TiC, ZrC, MoC
Presolar diamonds.The most abundant, but least understood, presolar phase is nanometer-sized diamond (Fig. 1) which occurs in the most primitive meteorites on a level of nominally more than one permill.
We cannot rule, however, that most of these nanodiamonds are actually of local, not presolar, origin. This is, because the structural element, carbon has an overall isotope ratio 12C/13C (~92) that is within the range of normal compositions and at least with current techniques it is not possible to analyze individual diamond grains, which on the average consist of only about 1000 atoms of 12C (and hence 11 atoms of 13C) for their isotopic composition.
Fig.1: TEM photograph of a presolar diamond grain. Courtesy F. Banhart.
The presolar origin for at least a fraction of thediamond grains is indicated by the occurrence within them of components of isotopically strange trace elements (see below) such as xenon (Xenon-HL; Lewis et al., 1987) and tellurium (Te-H; Richter et al., 1998), but these trace elements occur on such a low level, that there is only one atom of them per about a million diamond grains.
The situation is different for the other grain types in Table 1, the size of which is on the order of a fraction of a Ám up to several Ám (Fig. 2). Grains larger than ~0.5 Ám have been individually analyzed for the isotopic composition of the structural and several of the more abundant trace elements and show isotope abundance anomalies in many elements (Table 1). The next generation of ion microprobes which should come into operation within the near future will extend the range where individual analysis is possible, to even smaller sizes (Zinner, 1997).
Presolar silicon carbide. The origin of the presolar silicon carbide grains is the most clear-cut of all presolar phases. For the vast majority of grains (the "mainstream" grains as well as subtypes A, B, Y and Z) an origin from carbon stars is indicated (Hoppe et al., 1994; Hoppe and Ott, 1997). Only for the type X grains a different origin, namely from supernovae, is indicated (Amari and Zinner, 1997; Hoppe and Ott, 1997).
Fig. 2: SEM photograph a (large) presolar silicon carbide grain. Courtesy P. Hoppe.
The isotopic compositions of the major elements C and Si as well as the important trace nitrogen in SiC show large variations: 12C/13C ratios vary between 2 and 7000 (solar system: 89) and 14N/15N ratios from 7 to 19000 (solar system: 272), i.e., both ratios vary by more than three orders of magnitude. The mainstream grains are marked by 12C/13C = 10-100 and 14N/15N = 50-19000, with most of them (» 70%) having enrichments in 13C and 14N with 12C/13C ratios between 40 and 80 and 14N/15N ratios between 500 and 5000. That the range of 12C/13C observed in individual mainstream grains is similar to the range observed in the atmospheres of such stars (Alexander, 1993; Hoppe and Ott, 1997), is one of the strong arguments for the carbon star origin of most SiC grains. Another one is the fact that the heavy trace elements show the signature of the slow neutron capture (s-) process of nucleosynthesis (Lewis et al., 1990; Hoppe and Ott, 1997). The latter is thought to occur in the He burning shell of Red Giant stars during the asymptotic giant branch (AGB) phase, with transport of He shell material to the surface by "third dredge up" episodes (Hoppe and Ott, 1997; Lattanzio and Boothroyd, 1997; Gallino et al., 1997). The observed nitrogen isotopic compositions, the detection of extinct 26Al in the form of 26Mg excesses, as well as the presence of almost pure 22Ne (Ne-E(H)), which, as the s-process nuclides, is thought to originate from the He-burning shell are consistent with this interpretation (Gallino et al., 1990; Hoppe and Ott, 1997; Gallino et al., 1997). The opposite signature is shown by the X grains of probably supernova origin, most of which are characterized by enrichments in 12C and 15N, strong enrichments of 28Si, much higher abundance of now-extinct 26Al as well as evidence for the former presence of now-extinct 44Ti (T1/2 = 60 a).
Graphite. Graphite has probably the most diverse range of sources among the identified presolar grains. Presolar graphite grains (Amari et al., 1990; Hoppe et al., 1995) probably originated from all four sources that potentially can provide carbon-rich dust: carbon stars, novae, Wolf-Rayet stars and supernovae. The isotopic ratio 12C/13C varies among the analyzed single round grains by a factor ~3500 from ~2 to ~7300 and, based on the distribution of 12C/13C ratios, the grains can be assigned to four groups with the distribution different for grains of different effective density.
It is primarily the presence of anomalous noble gases that points to an origin for a significant number of grains from novae and from carbon stars (Amari et al., 1995): Ne-E(L), monoisotopic 22Ne present in about 30% of the grains (Nichols et al., 1994), in all likelihood is the decay product of 22Na (T1/2 = 2.6 a) thought to be synthesized in novae and supernovae, while the presence of s-process-Kr and Xe shows the need for a contribution from carbon stars. Other unusual isotopic compositions point to a supernova origin for a large fraction (probably >50%) of the individually analyzed graphite grains. Among these are large silicon isotopic anomalies and very high inferred 26Al/27Al ratios that range up as high as 0.1 (orders of magnitude higher than the "canonical value" of 5x10-5 generally considered typical of the early solar system: see review by MacPherson et al., 1995). Also found are the decay products of other extinct radionuclides: 41Ca (T1/2 = 0.1 Ma) and 44Ti (T1/2 = 60 a) which show up in a number of grains in the form of large excesses of 41K and 44Ca (Amari et al., 1996; Nittler et al., 1996). The results point to the need for complex mixing processes in supernova ejecta so that nuclides produced in the inner regions of the supernova end up being trapped in carbonaceous materials originating farther out (Amari and Zinner, 1997).
Aluminum oxide. Characteristic signatures are large variations in the isotopic composition of oxygen and large excesses of 26Mg corresponding to initial 26Al/27Al ratios of up to 0.016 (Nittler et al., 1997). The most likely sources of the vast majority of corundum grains are red giant stars, but some recently identified grains may have a supernova origin (Nittler et al., 1998; Choi et al., 1998), It is worthy of note, however, that the detailed compositions of oxygen measured in presolar grains in the laboratory span a much wider range than spectroscopical observations of stars (Zinner, 1997).
Silicon nitride. Only very few grains have been identified so far (Amari and Zinner, 1997). They bear similarities to SiC grains of type X with which they probably share an origin from supernovae. Characteristic isotopic features are excesses relative to solar of 28Si and 15N, both excesses and deficits of 13C and extremely high inferred 26Al/27Al (up to 0.2).
III. Presolar grains on the origin of heavy elements.
The light (and more abundant) elements can be made in a variety of contexts and nuclear processes in stars. In contrast, the elements heavier than iron are predominantly made by two types of neutron capture processes: one where the time scale for neutron capture is long compared to typical half lives of b -unstable nuclei (slow neutron capture, s-process) and one where the situation is just the opposite (rapid neutron capture, r-process).
Silicon carbide and the s-process. The most conspicuous feature in the isotopic compositions of the heavy elements in presolar SiC is the relative overabundance of the isotopes exclusively or predominantly manufactured in the s-process (Fig. 3). Several of the isotopic ratios in the s-process components of elements for which data are available are sensitive to the physical parameters that characterize the s-process: the abundances of isotopes that have small cross sections for neutron capture (mostly at closed neutron shells, i.e. magic neutron numbers) are sensitive to neutron dose ("exposure"), while at "branchings", where both neutron capture and b -decay occur in non-negligible proportion, the effects of neutron density and in some cases, of effective temperature and mass density are recorded.
Fig. 3. Comparison of normal isotopic composition of Nd with that of s-process Nd found in presolar SiC grains.
Based on the approach of the classical model (a phenomenological approach in which the physical parameters are held constant; see Käppeler et al (1989; 1990) "effective" values for neutron exposure, neutron density and mass density have been calculated from the SiC measurements (Hoppe and Ott, 1997): a neutron exposure t o that is distinct from the one that characterizes the solar system abundance distribution (0.14-0.17 mb-1 vs. 0.30 mb-1 for kT =30 keV), an effective neutron density of 3.2x108 cm-3, and an upper limit for the mass density at the s-process site of 3.3´ 103 g/cm3.
Diamonds and the r-process. While there is no doubt that the isotopic features in the heavy trace elements in SiC are the result of the s-process , the case is less
clear-cut for the presolar diamonds and their association with the r-process. The key has been the isopically strange xenon component Xe-HL, with about a factor of two enhancements of the light (p-process only) and heavy (r-process only) isotopes (Fig. 4).
A natural explanation for the enhancement of the r-only isotopes is an overabundance of r-process products, but there is the puzzling fact that the exact ratio of the excesses in the r-only 134Xe and 136Xe isotopes does not correspond to the "average" r-process ratio (Ott, 1996) - otherwise 134Xe and 136Xe would show the same relative enhancement in Fig 4.
Fig. 4: Isotopic composition of Xe-HL compared with that of solar wind xenon.
A "neutron burst" occurring in a supernova has long been considered the most likely source of the heavy isotope enrichment, but the agreement between predicted Xe compositions from that model and observations has been poor. An alternative model has been suggested (Ott, 1996) because of this, among other things, which suggests that within hours after termination of an "average" r-process, before complete decay of the radioactive precursors from the process into stable end products, an early separation occurred. With the addition of a few percent of fully decayed precursors, the model provides an exact match to the Xe-H spectrum. It also successfully predicts the observed virtual absence of Ba-H (Ba-H/Xe-H <10-3: Lewis et al., 1991), and that Xe-H should be accompanied by Te-H consisting of virtually 128Te and 130Te only. If further substantiated, the model may provide important constraints on the dynamics and on mixing / separation processes in supernova ejecta.
IV. Missed Opportunities (?)
It is important to note that the interstellar material from which the solar system formed, must have had contributions from a variety of stellar sources, with the sum of these contributions yielding the composition that we define as "normal". The "safe" identification via large isotope abundance anomalies thus will not identify presolar grains that show isotopic signatures typical at the time and location of solar system formation. What we do identify rather are "circumstellar" grains that formed from a single stellar source and the isotopic composition of which is dominated by (a) special process(es) of nucleosynthesis.
While this may be a reason for the fact that no additional presolar grain types have yet been identified, another may be that - following the lead of the isotopically anomalous noble gases - only acid-resistant phases have been isolated so far: diamond, SiC and graphite as actual carriers of these gases, others by the pure chance that they too are chemically resistant and ended up in the same residues. In situ search for silicates has been unsuccessful so far, but, apart from the fact that even in the most primitive meteorites silicates may have largely a solar system origin, this may be a problem related to small grain sizes and progress along these lines will have to await the advent of a new generation of ion microprobes with higher spatial resolution (Zinner, 1997). In the meantime, another promising approach to tracing down acid-soluble presolar grains - currently employed in the search for the carrier of excess 54Cr in primitive meteorites - may be that of stepwise dissolution of minerals and analysis of the solutions (Podosek et al, 1997; Ott et al., 1997).
Alexander C.M.OD. (1993) Geochim. Cosmochim. Acta 57, 2869-2888;
Amari S, Anders E., Virag A. and Zinner E. (1990) Nature 345, 238-240;
Amari S., Lewis R.S. and Anders E. (1995) Geochim. Cosmochim. Acta 59, 1411-1426;
Amari S., Zinner E. and Lewis R.S. (1996) Ap. J. 470, L101-L104;
Amari S. and Zinner E. (1997) in Astrophysical Implications of the Laboratory Study of Presolar Materials (T.J. Bernatowicz and E. Zinner, eds), 287-305;
Bernatowicz T.J. and Zinner E. (eds, 1997) Astrophysical Implications of the Laboratory Study of Presolar Materials (AIP Conf. Proc 402);
Choi B.-G., Huss G.R. and Wasserburg G.J. (1998); Meteoritics Planet. Sci. 33, Suppl., A32;
Clayton D.D. (1983) Principles of stellar evolution and nucleosynthesis, University of Chicago Press;
Gallino R., Busso M., Picchio G. and Raiteri C.M. (1990) Nature 348, 298-302;
Gallino R., Busso M. and Lugaro M. (1997) in Astrophysical Implications of the Laboratory Study of Presolar Materials (T.J. Bernatowicz and E. Zinner, eds), 115-153;
Hoppe P., Amari S., Zinner E., Ireland T. and Lewis R.S. (1994) Ap. J. 430, 870-890;
Hoppe P., Amari S., Zinner E. and Lewis R.S. (1995) Geochim. Cosmochim. Acta 59, 4029-4056;
Hoppe P. and Ott U. (1997) in Astrophysical Implications of the Laboratory Study of Presolar Materials (T.J. Bernatowicz and E. Zinner, eds), 27-58;
Huss G.R. (1997) in Astrophysical Implications of the Laboratory Study of Presolar Materials (T.J. Bernatowicz and E. Zinner, eds), 721-748;
Käppeler F., Beer H. and Wisshak K. (1989) Rep. Prog. Phys. 52, 945-1013;
Käppeler F., Gallino R., Busso M., Picchio G. and Raiteri C.M. (1990) Astrophys. J. 354, 630-643;
Lattanzio J.C. and Boothroyd A.I. (1997) in Astrophysical Implications of the Laboratory Study of Presolar Materials (T.J. Bernatowicz and E. Zinner, eds), 85-114;
Lewis R.S., Tang M., Wacker J.F., Anders E. and Steel E. (1987) Nature 326, 160-162;
Lewis R.S., Amari S. and Anders E. (1990) Nature 348, 293-298;
Lewis R.S., Huss G.R. and Lugmair G. (1991) Lunar Planet. Sci. 22, 807-808;
MacPherson G.J., Davis A.M. and Zinner E.K. (1995) Meteoritics 30, 365-386;
Nichols R.H.Jr., Kehm K., Brazzle R., Amari S., Hohenberg C.M. and Lewis R.S. (1994) Meteoritics 29, 510-511;
Nittler L.R., Amari S., Zinner E., Woosley S.E. and Lewis R.S. (1996) Ap. J. 462, L31-L34;
Nittler L.R., Alexander C.M.O'D., Gao X., Walker R.M. and Zinner E. (1997) Ap. J. 483, 475-495;
Nittler L.R., Alexander C.M.O'D., Wang J. and Gao X. (1998) Nature 393 222;
Nittler L.R. and Alexander C.M.OD. (1999) Ap.J. 526, 249-256;
Ott U. (1996) Ap. J. 463, 344-348;
Ott U., Specht S. and Podosek F.A. (1997) Lunar Planet. Sci. 28, 1053-1054;
Podosek F.A., Ott U., Brannon J.C., Neal C.R., Bernatowicz T.J., Swan P. and Mahan S. E. (1997) Meteoritics Planet. Sci. 32, 617-627;
Richter S., Ott U. and Begemann F. (1998) Nature 391 261-263;
Zinner E. (1997) in Astrophysical Implications of the Laboratory Study of Presolar Materials (T.J. Bernatowicz and E. Zinner, eds), 3-26.