Nuclear Physics and the Interpretation of R-Process Observables
Total Page：16
File Type：pdf, Size：1020Kb
Nuclear physics and the interpretation of r-process observables Rebecca Surman University of Notre Dame INT20-1b Online Pre-Workshop Institute of Nuclear Theory 7 April 2020 solar system abundances Burbidge, Burbidge, Fowler, and Hoyle 1957 r-process from Smithsonian Frank Timmes Hendrik Schatz Artemis Spyrou R Surman INT20-1b 7 April 20 Burbidge, Burbidge, Fowler, and Hoyle 1957 https://www.youtube.com/watch?v=nyvSoZe_q9A R Surman INT20-1b 7 April 20 r-process nucleosynthesis solar system r-process residuals Arnould+2007 (n,γ) β proton number Z (γ,n) neutron number N € R Surman INT20-1b 7 April 20 € € r-process observables: abundance patterns solar system r-process residuals Arnould+2007, Hotokezaka+2018 elemental abundances from r-process-enhanced metal-poor stars R Surman INT20-1b 7 April 20 Electromagnetic counterpart to 16 Barnes et al.r-process observables: kilonovae the neutron star merger GW signal 17 Kilpatrick+2017, Kasen+2017, etc. − 16 GW170817 − 15 kilonova SSS17a bolometric light curve − 14 − SSS17a bolometric AB 13 ) Mej = 0.01M , vej = 0.1c J − ( 12 M = 0.05M , v = 0.1c M ej ej − 11 Mej = 0.05M , vej = 0.2c − bolometric compilation: Waxman+ 2017 10 models: Kasen+2017 − Mej = 0.1M , vej = 0.2c 9 − Mej = 0.1M , vej = 0.1c 8 lanthanide rich − 100 101 Days lanthanide poor Barnes+2016 Figure 17. Absolute (AB) J -band light curves for several ejecta models.sGRB The130603B: excess IRTanvir+2013 flux (gold star) suggests an ejected mass 2 1 between 5 10− and 10− M . ⇥ winds. Our mass estimate here is an improvement over earlier work which neglected detailed thermalization, and gives R Surman INT20-1b 7 April 20 substantially di↵erent results. For example, Piran et al. (2014) suggested Mej 0.02M , less than half our new value. However, we have⇠ not accounted Material for viewing with an- significant opacity is the best fit to the data Slide credit: Dan gle e↵ects. If the ejected material is mainly confined to the equatorial plane, the emissionKasan will beSuggests brighter when lanthanides were made in the r-process. the system is viewed face-on (Roberts et al. 2011), which would reduce the inferred mass somewhat. If the ejecta is oblate, thermalization will also be more efficient, which could have a small impact on mass estimates. Radia- tion transport simulations in three dimensions with time- dependent thermalization models will further constrain Mej. 6.4. Late-time light curve Figure 16. Synthetic bolometric light curves for our fiducial ejecta model, calculated with Sedona for three di↵erent treatments Late time kilonova light curves may probe the history of thermalization: full thermalization (blue curve); Sedona’s origi- of r-process nucleosynthesis in CO mergers. At 2days nal thermalization scheme, which deposits charged particle energy after merger, fission ceases to be important, and⇠↵- and but explicitly tracks the deposition of γ-ray energy (lime curve); β-decay dominate the kilonova’s energy supply. Energy and the time-dependent ftot(t)fromournumericalsimulations(red curve). Accounting for time-dependent thermalization efficiencies from ↵-decay is transferred entirely to fast ↵-particles, has a significant impact on kilonova luminosity, particularly for which thermalize fairly efficiently out to late times. Beta models with lower masses and higher luminosities. For our fiducial particles thermalize with similar efficiency, but carry only model, the predicted luminosity is lower by a factor of . 2atpeak, a fraction ( 25%) of the total β-decay energy, with the and by 10 days is lower by an factor of 5. rest lost to⇠ neutrinos and γ-rays. A kilonova’s late-time luminosity will therefore depend on the relative impor- 2013; Berger et al. 2013). Tanvir et al. (2013)deter- tance of ↵-versusβ-decay. Because only nuclei with mined that the source of the flux had an absolute AB 200 . A . 250 undergo ↵-decay, the late time kilonova magnitude in the J -band of -15.35 at t 7 days. Having luminosity may diagnose the presence of heavy elements ⇠ incorporated ftot(t) into kilonova light curve models, we in the ejecta, and therefore constrain the neutron-rich can more confidently constrain the mass ejected in the conditions required for heavy element formation. kilonova associated with GRB 130603B. We gauge the relative strength of late-time kilonova In Figure 17, we compare the detected flux to J -band light curves for di↵erent Ye,0 by estimating the percent light curves for various ejecta models, and find the ob- of energy from the decay of r-process elements emitted 2 served flux is consistent with 5 10− M . Mej . as fission fragments, ↵-, and β-particles, time-averaged 1 ⇥ 10− M . This mass is higher than what is typically pre- over t = 10 100 days. (Note that while all energy from dicted for the dynamical ejecta from a binary neutron ↵-decay emerges− as ↵-particles, β-particles receive only star merger, suggesting that if the kilonova interpreta- 25-30% of the energy from β-decay.) The results for our tion is correct, the progenitor of GRB 130603B was per- representative SPH trajectory, for a range of Ye,0 and haps a neutron star-black hole merger, or that the mass two nuclear mass models, are shown in Figure 18.The ejected was significantly enhanced by post-merger disk curves suggest that systems with Ye,0 . 0.17 have more 8 neutron star merger r-process environments t t = 0.3 ms t t = 0.6 ms ≠ mrg ≠ ≠ mrg prompt ejecta Ye 2 A. Perego et al. Radice+2019 ejecta from the accretion disk Perego+2014 t t = 1.2 ms t t = 2.5 ms R Surman≠ mrg INT20-1b ≠ mrg 7 April 20 Figure 2. Volume rendering of the electron fraction of the ejecta for the simulation SFHo M135135 M0.Theray-castingopacityislinear in the logarithm of the rest-mass density. From the top-left in clockwise direction, the transparency minimum – maximumFigure in the opacity 1. Left: sketch of the neutrino-driven wind from the remnant of a BNS merger. The hot hypermassive neutron star (HMNS) scale are (1011 1014)gcm 3,(108 1011)gcm 3,(108 1011)gcm 3,and(107 1011)gcm 3.Thelastpanelofthisfigureshouldand the accretion disc emit neutrinos, preferentially along the polar direction and at intermediate latitudes. A fraction of the neutrinos − − − − − − − − be compared with Fig. 14 where we plot a cut of the data in the xz-plane. is absorbed by the disc and can lift matter out of its gravitational potential. On the viscous time-scale, matter is also ejected along the equatorial direction. Right: sketch of the isotropised ⌫ luminosity we are using for our analytical estimates (see the main text for details). decompression of this initially cold and extremely neutron- et al. 2014) for the so-called “macro-” or “kilonovae” (Li rich nuclear matter had long been suspected to provide & Paczy´nski 1998; Kulkarni 2005; Rosswog 2005; Metzger favourable conditions for the formation of heavy elements et al. 2010a,b; Roberts et al. 2011), radioactively powered through the rapid neutron capture process (the “r-process”) transients from the decay of freshly produced r-process (Lattimer & Schramm 1974; Lattimer & Schramm 1976; elements. In particular, the delay of several days between Lattimer et al. 1977; Symbalisty & Schramm 1982; Eichler the sGRB and the nIR detection is consistent with the et al. 1989; Meyer 1989; Davies et al. 1994). While initially expanding material having very large opacities, as predicted only considered as an “exotic” or second-best model behind for very heavy r-process elements (Kasen et al. 2013). If core-collapse supernovae, there is nowadays a large litera- this interpretation is correct, GRB130603B would provide ture that –based on hydrodynamical and nucleosynthetic the first observational confirmation of the long-suspected calculations– consistently finds that the dynamic ejecta of a link between compact binary mergers, heavy elements neutron star merger is an extremely promising site for the nucleosynthesis and gamma-ray bursts. formation of the heaviest elements with A>130 (see, e.g., There are at least two more channels, apart from the Rosswog et al. 1999; Freiburghaus et al. 1999; Oechslin dynamic ejecta, by which a compact binary merger re- et al. 2007; Metzger et al. 2010b; Roberts et al. 2011; leases matter into space, and both of them are potentially Goriely et al. 2011a,b; Korobkin et al. 2012; Bauswein et al. interesting for nucleosynthesis and –if enough long-lived 2013; Hotokezaka et al. 2013; Kyutoku et al. 2013; Wanajo radioactive material is produced– they may also power et al. 2014). Core-collapse supernovae, on the contrary, additional electromagnetic transients. The first channel seem seriously challenged in generating the conditions that is the post-merger accretion disc. As it evolves viscously, are needed to produce elements with A>90 (Arcones et al. expands and cools, the initially completely dissociated 2007; Roberts et al. 2010; Fischer et al. 2010; H¨udepohl matter recombines into alpha-particles and –together with et al. 2010). A possible exception, though, may be magnet- viscous heating– releases enough energy to unbind an ically driven explosions of rapidly rotating stars (Winteler amount of material that is comparable to the dynamic et al. 2012; M¨osta et al. 2014). Such explosions, however, ejecta (Metzger et al. 2008; Beloborodov 2008; Metzger require a combination of rather extreme properties of the et al. 2009; Lee et al. 2009; Fern´andez & Metzger 2013). pre-explosion star and are therefore likely rare. The second additional channel is related to neutrino-driven Most recently, the idea that compact binary mergers are winds, the basic mechanisms of which are sketched in Fig. 1. related to both sGRBs and the nucleosynthesis of the This wind is, in several respects, similar to the one that heaviest elements has gained substantial observational emerges from proto-neutron stars.