R-process cosmochronometer

Transcription

1 R-process cosmochronometer Dr. B. Pfeiffer Inst. f. Kernchemie, Mainz Determining the age of the Universe has long been considered the holy grail of cosmology BD st VISTARS Workshop on Nuclear Astrophysics Russbach, March 14 19, 2004

2 Different ways to determine the age of the Universe Not only astronomers have tried to determine the age of the Universe. Theologians added dates in the Scripture to calculate the moment of Creation: B.C. J. Ussher, The Annals of the World (1658) Age of the bodies in the Universe Earth by radioactive dating of rocks (better primitive meteorites) Stars from stellar evolution models Globular Clusters from turn-off point Cooling rate of White Dwarfs Zircon crystal, 4.3 Ga Dating the Big Bang from back-extrapolation of the expansion of space-time Determining the onset of nucleosynthesis of the heavy elements in the r- and s-process One should not forget the future fate of the Universe. But this is another topic.

3 The nuclides in the left column are suitable for cosmological age determinations, whereas the right ones are markers for the formation of stars and planetary systems. The Solar system matter is the product of many cycles of nucleosynthesis events. Limits on the age of the elements can only be derived with assumptions on the Galactic chemical evolution as rate of star formation, infall of gas in disc, etc. Different scenarios for chemical evolution

4 Composition of extraterrestrial matter Old and new observables for chemical evolution Optical spectroscopy: : Bunsen/Kirchhoff Birth of astrophysics 1863: 9 elements detected in Aldebaran and Betelgeuse Spectroscopy of ultra-metal metal-poor Halo stars Interstellar matter (ISM) 13 Gy ago Elemental and isotopic composition of meteorites Interstellar medium 4.6 Gy ago Pre-solar grains in meteorites and high atmosphere ISM prior to formation of Solar system Ultra-heavy component of cosmic rays Recent nucleosynthesis events STARDUST at comet 81P/Wild 2 (Jan. 2, 2004) Pre-solar SiC grain

5 Ultra-metal metal-poor Halo stars Th-U cosmochronometer Th-U chronometer ideal for dating of solar system. Lower limit for age of the Universe requires models of Galactic chemical evolution. High-resolution optical spectroscopy of ultra-metal metal- poor, very old Halo red giant stars opens new perspectives: One (or( few) nucleosynthesis events seeded ISM. Scaled solar system r-process abundances for 56 Z 79. Radioactive dating requires production ratios. Th/U ratio (hopefully( hopefully) less affected by extra- polations of nuclear structure Abundances in BD+17 o 3248 Calculated abundances of the long-lived lived actinides test for short-lived isotopes expected in Galactic cosmic rays.

6 Astronomical observations Francesca Primas presented us the astronomical observation techniques and data reduction for determining Th and even U abundances in very old ultra-metal-poor Halo stars. HD BPS CS Contrary to terrestrial rocks and meteorites in stars often the Pb/Bi endproducts of the α-decay chains from U/Th are not observed simultaneously. Therefore, the production yields of the cosmochronometers have to be calculated.

7 Nucleosynthesis Calculations In the early 50`s, the development of the nuclear shell-model (Göppert-Mayer and Jensen, Haxel, Suess) was the basis for the understanding of nucleosynthesis. Several groups worked out a detailed picture, as Suess and Urey A.G.W. Cameron C.D. Coryell B 2 FH B 2 FH, the Bible of Nuclear Astrophysics The basic ideas are still valid. The rapid development of computing power now enables more refined calculations based on much more experimental data.

8 R-process at magic numbers B 2 FH predicted correctly even the detailed behaviour at N=50, 82 and 126. The climb up the staircases, the major waiting-point nuclei involved, as well as the break-through pairs, and their association with the rising sides of major peaks in the abundance curve for the r-process are still today important properties to be studied experimentally and theoretically. B 2 FH concluded that a reasonable but not exact agreement with observed abundances is obtained. It is essential that the r-process boulevard comes close to the valley of stability at the magic numbers giving the opportunity to obtain experimental nuclear physics data. Fig. VII,3 of B 2 FH: Classical static r-process calculation compared to observed abundances of Suess and Urey.

9 Nuclear physics input data Data for extremely neutron-rich rich nuclei, mostly out of reach of actual experimental techniques: nuclear masses β-decay properties: : T 1/2, P n, β-delayed fission (n,γ), (γ,n),( n-induced and spontaneous fission rates Approximation to full network calculations already give insights in decisive data: canonical r-process (n,γ) (γ,n),n) equilibrium ( waiting-point ) β-flow equilibrium Fe seed nuclei Y(Z) λ β = const. In the simplest case, S n and T 1/2 (by given neutron number and temperature) are sufficient

10 Importance of experimental data The r-process r abundances of the actinides depend also on nuclear-structure data at the magic neutron-numbers numbers,, nuclei which act as bottle-necks for the matter flow. Up to present, only neutron-rich rich nuclei with N=50 and 82 could be observed. Future RIB-facilities hopefully will make N=126 (and beyond?) accessible for experiments. r-process path at N=82

11 New nuclear structure at drip-lines, as shell quenching? Defiencies prior to the main peaks were attributed by our group to nuclear structure effects: - too strong shell strength for extremely neutron-rich magic nuclei far from stability - nuclear models adjusted to stable nuclides - new physics far from stability? N/Z 70 Ongoing discussion: Can one learn neutron-dripline physics from astrophysical observables? K.-L. Kratz, Ap.J (1993) 216 New magic numbers? B. Pfeiffer et al., Acta Physica Polonica B27 (1996) 475

12 Post - B 2 FH calculations R-process cannot be described with one single parameter set. With fixed temperature at least four sets of neutron densities and time durations are necessary: three corresponding to the abundance maxima at neutron magic numbers 50, 82 and 126 one for the actinide region

13 Influence of neutron number density on individual components

14 Classical Mz-Bs calculations Fit to solar r-process r isotopic abundances obtained as superposition of 16 canonical n n -components. Important is a good fit to the r-process r Pb and Bi contributions after summing up the α-decay chains of heavier nuclei. Observed neutron-capture elemental abundances in an ultra-metal metal-poor halo- star (blue squaresand dots) compared to scaled solar values (green dots) ) and our calculated r-abundances (red line). Pb -, -decays ETFSI-Q CS fission Whereas there is a nearly perfect agreement with the solar system isotopic abundances, the abundances for low-z elements in UMP stars are lower than the calculated values.

15 Two r-process components? Recent results indicate to two r-process components: extinct radionuclides in the early Solar system abundance anomalies in presolar diamonds elemental abundances in UMP stars below second peak In all UMP stars analyzed up till now, the elemental abundance pattern for elements from Ba to the Pt-group elements can be reproduced as scaled Solar system r-process abundances. weak r-process component This main r-process component reproduces also the Pb/Bi abundances. Therefore, the calculated production ratios of the cosmochronometers are unaffected by a weak component. B. Pfeiffer et al., Nucl. Phys. A688 (2001) 575 main r-process component

16 Influence of lower limit for summation of n n -components The mass region from the second peak up to the actinide cosmochronometers shows no appreciable dependence on the low n n components.

17 Progenitors of the Actinides After freeze-out out, the extremely neutron-rich rich nuclei undergo β-, β-del. n-, n β-del. fission-, sp. fission decay New attempts to include β-del. fission decay by Schatz et al., I.V. Panov et al., Extended studies underway Progress in mass modeling required: Production ratios even of close-lying lying nuclides as 232 Th and 238 U differ considerably when applying different mass models. (ETFSI-Q versus HFBCS-1) H. Schatz et al., Ap.. J. 579 (2002) 628; I.V. Panov et al., Proc. Pont d Oye V (2004) 3

18 Radioactive age determinations for Halo stars In some r-process enriched, ultra-metal-poor Halo stars abundances of Th could be measured together with rare-earth elements, as the (nearly) pure r-process element Eu. Most of the mass models give Th-ages in accordance with recent (independent) age determinations for the Universe. The value for the FRDM model (with excellent global results) is determined by deficiencies near closed shells.

19 Influence of mass models on Th/U production ratios Models, which reproduce the Pb/Bi abundances, yield Th/U ratios around 1.7. The high value for the HFBCS-1 model results from different nuclear shapes around A=230.

20 In old Halo stars only 232 Th and in a few cases 238 U have survived. Cosmic rays propagate for some tens of million years until they are destroyed by spallation/fission or are scattered out of the Galaxis. Short-lived actinides could be observed if the sources of the cosmic rays lie close to the Solar System and theythey are freshly synthesized.

21 Local Superbubble: : Gould s s Belt Cosmic rays propagate in a huge volume (about 1 kpc) of rarefied ISM. The outer expanding shell is marked by star-forming regions with young,, hot OB associations (Gould s s Belt, Linblad Ring). The expansion is fuelled by recent SN explosions (over some ten million years). The Sun does not belong to this structure. It just passes through. In this close-up scheme of our Galactic neighbourhood (500 pc), the outer edge of the Linblad Ring is shown in the lower left corner. The region is dominated by a multi- tude of expanding shells of hot gases enclosing bubbles of rarefied ISM (~0.06 atoms/cm 3 instead of mean value ~0.5 atoms/cm 3 ).

22 Abundance predictions for actinides The predictions on the left (Weaver,, Westphal) are based on our calculations from 1997, whereas the right ones were calculated by Goriely and Arnould The right figure is based on one single SN event, whereas the left one assumes a uniform synthesis model in the frame of the superbubble concept.

23 Superbubble origin of cosmic rays actinide nuclei Whereas in general the composition of the cosmic rays does not allow to distinguish between an origin in the general Galactic ISM or in the superbubbles, there ought to be strong differences for the abundances of short-lived actinide nuclei, as the cosmic rays are lost to the Halo by diffusion or are destroyed in collisions with the ISM in million years. The detection of short-lived actinides in the cosmic rays is a crucial test for the superbubble model. In the rarefied ISM of the superbubbles cosmic rays have a mean free path of about 1 pc. The bulk of the particles is still located within about 0.4 kpc of their birth place after 1 mill. years. Short-lived nuclei must be of local origin. Higdon & Lingenfelter, Ap.J. 582 (2003) 330

24 Predictions for mean actinide abundance ratios Mean actinide abunance ratios from r-process yields of core collapse supernovae in their accumulating ejecta averaged over various time intervals, assuming a constant supernova rate during those intervals. The typical cosmic-ray acceleration time span in the supernova- active cores of superbubbles of roughly 50 Myr is indicated by the dashed line (SB). R.E. Lingenfelter,, J.C. Higdon,, K.-L. Kratz and B. Pfeiffer, Ap. J. 591 (2003) 228

25 Predictions for superbubble core metallicity of actinides In the cores of superbubbles, SNe are concentraded in space (about 1 kpc) ) and time (about( 40 Myr). Freshly synthesized matter of high metallicity is deluted into old ISM with metallicity slightly higher than the Solar system value. CR Metallicity from low-z cosmic rays (CR) indicated in red. Expected cosmic-ray actinide abundance ratios as a function of superbubble core metallicity.. Also indicated are the estimated U/Th and Actinide/PtGroup(75< Z<79) ratios in the present interstellar medium (ISM) and in the supernova-enriched enriched protosolar material. R.E. Lingenfelter,, J.C. Higdon,, K.-L. Kratz and B. Pfeiffer, Ap. J. 591 (2003) 228

26 UHCR measurements Elemental abundances of UHCR were measured with track-etch detectors (as TREK on MIR): Pt-group and Pb Actinide nuclei: TREK: 6 nuclei (Z>88) actinide/pt Pt-group = 0.034(18) UHCRE on LDEF: 30 nuclei actinide/pt Pt-group = 0.028(8) Predicted value in superbubble model: actinide/pt Pt-group = 0.027(5) Etch pit in glass detector See, e.g. Proc.. 27th Int. Cosmic Ray Conf. ICR2001; Hamburg (2001)

27 Fe-60 in Deep-Sea Sediments An excess of 60 Fe has been detected in 2 to 5 Myr old deep-sea sediments. The upper Scorpius subgroup of the Sco-Cen OB association passed 40 pc from the Solar system about 2 Myr ago when several SNe exploded in this group. Dust particles are still passing the Solar system - detectors on Galileo, Ulysses, - STARDUST collected dust on way to 81P/Wild-2 Up till now,, no detection of of 244 Pu 244 Pu Some speculate that these close SNe might have had an influence on the development of early hominides at the Pliocene-Pleistocene Pleistocene transition. Others discuss relations between flux of CR and climate.

28 New interest in fission: termination of r-process fission recycling Parisinus graecus 2327, fol. 279 Bibl. Nat., Paris

29 A.G.W. Cameron, Ap. J. 587 (2003) 327 P. Möller M et al., Nature 409 (2001) 785 Recently, new interest in termination of r-process by fission Experimental evidence for termination of r-process by fission: No isotopes heavier than A=257 observed in subterrranean thermonuclear explosions. New scenarios as jets from SNe New (higher) fission barrier calculations as - Mamdouh et al., Nucl. Phys. A644 (1998) ongoing work by Peter Möller New fission rate calculations see, e.g. I.V. Panov et al. P. Demetriou et al. (Proc. Seminar on Fission Pont d Oye V, in print) Fission barriers for 234 U

30 Examples for influence of fission Attempt to include spontaneous and β-delayed fission modes very old data set no mass distribution of fission fragments no fission recycling ff 136 Te ff 136 Te No strong influence for SNII site for r-process Neutron star merger scenario has higher fluxes: Mass flow to fission region and recycling See, e.g. F.-K. Thielemann, J. Metzinger, H.V. Klapdor,, Z. Phys. A309 (1983) 301

31 Site of the r-process? In the supernova scenarios, the neutron flux is not sufficient to drive the path deep into the fission region: about 1-2 % fissions Contrary, in the neutron-star merger scenario, high neutron fluxes lead to the fission of substantial parts of the nuclei, which then act as new seed nuclei: fission recycling Pulsar PSR J and neutron-star companion

32 K. Farouqi et al., Ann. Rep. 2003, Inst. f. Kernchemie, Mainz Network calculations The approximatons in the canonical model can now be refined by network calculations using the charged particle network of Thielemann and the r-process code of Freiburghaus (ApJ 516 (99) 381). The seed nuclei composition after an α-rich freezeout lies beyond the first bottle-neck at N=50 enabling a very fast r-process with time duration in the order of 200 ms. First results look promising. The fit to the Solar distribution (blue) with the superposition of only 2 entropy components can reproduce the observations beyond A 120, the region of the main r-process component.

33 S-process cosmochronometers Very long-lived isotopes are also produced in the s-process. (We will hear in detail the next days.) In contrast to the UMP stars, a modeling of the history of Galactic star formation is needed. The mean life-times of s-process cosmochronometers as 176 Lu or 187 Re depend on the environment. In a stellar environment the isomeric 1 - state can be populated reducing the life-time considerably. The β-decay rate of 187 Re depends on the ionisation: The neutral atom has a half-live of 50 Gyr, the naked ion of 33 yr only. For age determinations one has to model the time the nuclide was in a star (astration) and in a dust particle. Age estimates of 15 Gyr are in the range from r-process cosmochronometers.

34 Independent age determinations The globular clusters in the Halo belong to the oldest objects in the Milky Way. Th/Eu ages were determined for UMP stars in M15. The mean age of (14±4) Gyr is in accordance with independent age estimates for globular clusters from stellar evolution theories and cooling rates of White Dwarfs. The r-process chronometers up till now have a quite large error margin. But the fact that they are in accordance with totally unrelated methods as the ones based on cosmology shows that the basic assumptions are right.

35 Future developments r-process calculations - inclusion of termination of r-process by fission - fission barriers, rates, mass distributions - dynamical network calculations ultra-metal-poor stars - sky surveys - extension of [Fe/H] systematics - isotopic abundances cosmic-ray experiments - fission track detectors: ECCO, on ISS? - ultra-long duration balloon flights: TIGER - dedicated satellites: Heavy Nuclei Explorer Mission interstellar matter - marine sediments: 60 Fe, 244 Pu? - sampling of stardust by high-flying air planes - STARDUST probe: comet 81P/Wild-2 Nucleus of 81P/Wild 2 1 st VISTARS Workshop on Nuclear Astrophysics Russbach, March 14 19, 2004