Kilonova Emission from Compact Binary Mergers: Opaci-es of Lanthanide-rich and Lanthanide-free Ejecta

Size: px

Start display at page:

Download "Kilonova Emission from Compact Binary Mergers: Opaci-es of Lanthanide-rich and Lanthanide-free Ejecta"

Florence Gilmore
5 years ago
Views:

1 Kilonova Emission from Compact Binary Mergers: Opaci-es of Lanthanide-rich and Lanthanide-free Ejecta Masaomi Tanaka (Na-onal Astronomical Observatory of Japan) Daiji Kato, Gediminas Gaigalas, Pavel Rynkun, Laima Radziute, Shinya Wanajo, Yuichiro Sekiguchi, et al.

2 Kilonova Emission from Compact Binary Mergers: Opaci-es of Lanthanide-rich and Lanthanide-free Ejecta IntroducCon New opacity calculacons ApplicaCons to kilonova

3 2590 Merger S. Rosswog, T. Piran and E. Nakar Dynamical ejec-on Post-merger ejec-on < 10 ms ~< 100 ms r-process nucleosynthesis < 1 sec Rosswog+ 13 Shibata, Giacomazzo, Foucart, Just, Siegel, Tchekhovskoy, Perego, Wu, McLaughlin Radioac-ve decay => kilonova ~ days MT & Hotoke 13 This talk Korobkin and Baron

4 Radioac-ve hea-ng (decay of many r-process nuclei) NS merger ~ t hr 1 day 10 days erg s -1 supernova erg s -1 (for M = 0.01 Μsun) Metzger+10

5 Kilonova/Macronova" Ini-al works: Li & Paczynski 98, Kulkarni 05, Metzger+10, Goriely+11, High opacity: Kasen+13, Barnes & Kasen 13, MT & Hotokezaka 13, Timescale t peak = 3 M ej 4 cv 1/2 8.4 days M ej 0.01M 1/2 v 0.1c 1/2 10 cm 2 g 1 1/2 Luminosity L peak = L dep (t peak ) *assuming 50% thermaliza-on erg s 1 M ej 0.01M 0.35 v 0.1c cm 2 g

6 High opacity of NS merger ejecta Nd, Ce Mixture of r-process NSM-all NSM-Fe κ (cm 2 g -1 ) Fe, Os Wavelength (A) = = Figure 7. Wavelength-dependent line expansion opacities resulting from Kasen, Badnell, & Barnes 13 MT & Hotokezaka 13 κ ~ 10 cm 2 g -1

7 5 open s shell (l=1) E (ev) 4 3 open p-shell (l=2) 2 5 open d-shell (l=3) 1 Fe II E (ev) 2 1 Nd II 0 open f shell (l=4)

8 Light curve see Barnes & Kasen 13, MT & Hotokezaka 13 Type Ia SN NS merger a few x ~1 week Luminosity (erg/s) decay energy Mej = 0.01 Msun Days aeer the merger MT 16

9 Spectra See Eddie s talk Fe opacity Nd r-process Extremely red spectra Kasen+13

10 Spectra Observed magnitude (200 Mpc) g r i z NS-NS 1.5 days Type Ia SN SN 1998bw r-process mixture 5.0 days 10.0 days Absolute magnitude Wavelength (A) MT 16 Much redder than supernovae

Dynamical ejecta (~< 10 ms) Post-dynamical ejecta (~< 100 ms) Side view Side view lations tingui partic the sim tracer to the compo By tial di with v ing ou setup pared lations There tropy, flows

$Profiles of the electron number per baryon, Ye, (left in each and the specificfraction, entropy, s, (right in each may panel which potential, and contours of rest-mass density = in x-y (lower in each$ panel) and x-z (upper in each panel) planes.

panel) and x-z (upper in each panel) planes.

11 Dynamical ejecta (~< 10 ms) Post-dynamical ejecta (~< 100 ms) Side view Side view lations tingui partic the sim tracer to the compo By tial di with v ing ou setup pared lations There tropy, flows higher tion sh as in [ tion (n the hi ously e n FIG.panel) 2. Snapshots of electron normalized electron FIG. 2. Profiles of the electron number per baryon, Ye, (left in each and the specificfraction, entropy, s, (right in each may panel which potential, and contours of rest-mass density = in x-y (lower in each panel) and x-z (upper in each panel) planes.chemical The panels show the results for SFHo top9 three [10, 10, 10, 10, 10 ] g cm at t = 43 ms, when the disk Full (left), SFHo h (middle), and SFHo h (right) at 13 ms after the onset of the merger. The lower three panel has fully self-regulated itself to mild electron degeneracy. The [70] w show the results for DD h (left), DD h (middle), interior and DD h (right) at 10 ms after the onsetdividu of th of the BH horizon is masked (black). merger. shows throug imbalance: in regions of lower density, viscous heating proces 2 from estimated MHD driven andofenergy releaseatfrom the binaries, the typical ejecta mass would approach 10 M they theturbulence properties the ejecta. 5 mstoafte recombination of free nucleons into alpha particles exirrespective of the EOS employed. We note ethat the total the onset of the merger, perhaps because of their lent smaag ceeds cooling by neutrino emission, and the weak instars [ ejecta mass depends only weakly on the grid computational domain employed (L = 750 km). How + resolution teractions essentially freeze-out (although further mixas obs e as listed in Table I. ever, the would stillfunnel increase i ing can stillejecta changemass Ye ). In the polar these with out- time we rec flows an possess high-y high specific-entropy such early phase. Thisand could be one of the reason e (> 0.2) Top view Top view Sekiguchi+16 Rosswog+99, Lee+07, Goriely+11, Hotokezaka+13, Bauswein+13, Radice+16 - Mej ~ Msun - v ~ c n+ν - wide Ye n+e Siegel+17 Fernandez+13,15, Perego+14, Kiuchi+14,15, MarCn+15, Just+15, Wu+16, Siegel & Metzger 17 -> p + e -> ν + p - Mej >~ 10-3 Msun - v ~ 0.05 c - rela-vely high Ye

12 10 0 d p s f Se Ru Te Nd Er Ye = 0.30 Ye = 0.25 Ye = Mass fraction Atomic 1 day (Wanajo+14) Blue kilonova? SimulaCons with Fe opacity or gray opacity Metzger+14, Kasen+15, Fernandez & Metzger 16, Metzger 16 New opacity calcula-ons for Se (Z=34), Ru (Z=44), Te (Z=52), Nd (Z=60), Er (Z=68)

13 Kilonova Emission from Compact Binary Mergers: Opaci-es of Lanthanide-rich and Lanthanide-free Ejecta IntroducCon New opacity calculacons ApplicaCons to kilonova

14 Atomic structure calcula-ons HULLAC code (rela-vis-c, local radial poten-al, Bar-Shalom+99) Se I-III (Z=34, p), Ru I-III (Z=44, d), Te I-III (Z=52, p), Nd I-III (Z=60, f), and Er I-III (Z=68, f) GRASP code (rela-vis-c, e-e interac-on, Jonsson+07) Nd II-III (Z=60, f) and Er II-III (Z=68, f)

15 Line expansion opacity (from two codes) Nd II HULLAC GRASP2K Layer 0 GRASP2K Layer Er II HULLAC GRASP2K Layer 0 Opacity (cm 2 g -1 ) Opacity (cm 2 g -1 ) Wavelength (angstroms) Wavelength (angstroms) Consistent within a factor of 2 even for the worst case (Er II) Consistent with results by Kasen+13 (Autostruture code)

16 Line expansion opacity (for each element) Se (p) Ru (d) Te (p) Nd (f) Er (f) Fe (Kurucz) Se (HULLAC) Ru (HULLAC) Te (HULLAC) Fe (Kurucz) Nd (HULLAC) Er (HULLAC) Opacity (cm 2 g -1 ) Fe (d) Opacity (cm 2 g -1 ) Wavelength (angstroms) Wavelength (angstroms) κ (p shell) << κ (d shell) << κ (f shell) see Kasen+13, Fontes+17 MT+ in prep.

17 Kilonova Emission from Compact Binary Mergers: Opaci-es of Lanthanide-rich and Lanthanide-free Ejecta IntroducCon New opacity calculacons ApplicaCons to kilonova

18 3D Monte-Carlo -me/frequency-dependent radia-on transfer (MT & Hotokezaka 13, MT+14, MT 16) ~6,000,000 b-b transi-ons (out of ~70,000,000)

19 Bolometric luminosity (erg s -1 ) M = 0.01 Msun v = 0.1c f = 0.25 gray opacity 0.5 cm 2 g -1 3 cm 2 g -1 Ye = Ye = 0.25 Ye = cm 2 g -1 Ye = Ye = 0.25 Ye = 0.30 Lanthanide ~ 0.1-1% 0% gray opacity 0.5, 3.0, 10 cm 2 g -1 Simple model - Mej = 0.01 Msun - v = 0.1c - HeaCng rate ~ t Constant thermalizacon (0.25) Days after the merger Depends sensi-vely on Ye κ ~ 0.5 cm 2 g -1 for Lanthanide-free ejecta (Ye ~ 0.3)

Bolometric luminosity (erg s -1 ) 10 42 10 41 10 40 10 39 High therm. efficiency NS-NS (Ye = 0.10-0.40) Wind (Ye = 0.25) Wind (Ye = 0.30) Low hea-ng rate Low therm.

20 Bolometric luminosity (erg s -1 ) High therm. efficiency NS-NS (Ye = ) Wind (Ye = 0.25) Wind (Ye = 0.30) Low hea-ng rate Low therm. efficiency NS-NS - Mej = 0.01 Msun - v = 0.2c Hotokezaka+13, Sekiguchi+15,16 Wind - Mej = 0.01 Msun - v = 0.05c Days after the merger - HeaCng rate from nucleosynthesis calc. - ThermalizaCon (Barnes+16)

21 Op-cal (r-band) Mej = 0.01 Msun Absolute magnitude r-band NS-NS (Ye = ) Wind (Ye = 0.25) Wind (Ye = 0.30) Observed magnitude (200 Mpc) Days after the merger MT+ in prep. Wide variety even for the same ejecta mass => Accurate es-mate of Ye is crucial See Francois s, Oliver s, Sasha s, Albino s talks

the increase in Y e is due mostly to a change in the relative number of νe and νe absorptions in the outflows (or, equivalently, a change in the value of Y e at which the outflows are in equilibrium

$We note that the electron fraction of the polar outflows is largely set by neutrino emission and absorption very close to the compact neutron star core, where the temperature of [34,35,38,49].$

22 the increase in Y e is due mostly to a change in the relative number of νe and νe absorptions in the outflows (or, equivalently, a change in the value of Y e at which the outflows are in equilibrium with the neutrino radiation), rather than to additional absorptions of νe alone. We note that the electron fraction of the polar outflows is largely set by neutrino emission and absorption very close to the compact neutron star core, where the temperature of [34,35,38,49]. Our results show that, in the gray mation, the way in which we estimate the average the neutrinos can have important consequences magnitude of that effect. For the configuration s this work, evolving the neutrino number density to local estimate of the neutrino average energy mak that the polar ejecta are initially prevented from un strong r-process nucleosynthesis. This is a prerequ Blue component may be absorbed by dynamical ejecta? e.g., Kasen+15, Metzger 17 FIG. 17. Vertical slice through the numerical simulation 10 ms after merger. The color gradient shows the electron fraction o Dashed white lines show isodensity contours ρ0 ¼ 1010;11;12 g cm3. Arrows show the transport velocity in the fluid. The solid shows the boundary of the region in which the fluid is marked as unbound. All unbound material (i.e. fluid elements in the pola has a high electron fraction Y e > See Francois s talk Foucart High Ye in the polar region (< deg) => Blue emission may be able to escape Sekiguchi+16 FIG. 2. Profiles of the electron number per baryon, Ye, (left in each panel) and the specific entropy, s, (right n x-y (lower in each panel) and x-z (upper in each panel) planes. The top three panels show the results for S

23 NIR (J-band) Mej = 0.01 Msun -16 Absolute magnitude J-band GRB B (Tanvir+13, Berger+13) GRB B (Troja+16) Observed magnitude (200 Mpc) see also Kasliwal Days after the merger M ~ 0.06 Msun for GRB B Barnes+16

24 r-band 100 Mpc Mej = 0.01 Msun Absolute magnitude r-band Free n decay Cocoon emission (Metzger+15) (Gojlieb+17) NS-NS (Ye = ) Wind (Ye = 0.25) Wind (Ye = 0.30) 1m 4m 8m DECam HSC, LSST Observed magnitude (100 Mpc) Days after the merger

25 r-band 200 Mpc Mej = 0.01 Msun -16 Absolute magnitude r-band Free n decay Cocoon emission (Metzger+15) (Gojlieb+17) NS-NS (Ye = ) Wind (Ye = 0.25) Wind (Ye = 0.30) 1m 4m 8m DECam HSC, LSST Observed magnitude (200 Mpc) Days after the merger

26 Can we measure mass of r-process elements? Absolute magnitude Mej = 0.01 Msun r-band Ye = Ye = 0.25 Ye = Days after the merger Observed magnitude (200 Mpc) Rosswog+17 We need (1) mul--color observa-ons, and (2) good theore-cal models - mergers and nucleosynthesis (long-term simulacons) - heacng rate (nuclear physics) - radiacve transfer (atomic data, opacity)

27 Summary New opacity calcula-ons for Se, Ru, Te, Nd, and Er Opacity sensi-vely depends on composi-ons => Accurate es-mate of Ye is cri-cal κ ~ 0.5 cm2 g -1 for Ye ~ 0.3 (Lanthanide free) κ ~ 10 cm2 g -1 for solar abundance Wide variety depending on composi-ons OpCcal: mag for ~3 200 Mpc (0.01 Msun) NIR: mag for ~7 200 Mpc (0.01 Msun) Observa-onal prospects How to select NS mergers? AssociaCon w/ nearby galaxies, faintness, and rapid evolucon (possible diversity in color) ==> mulc-visit observacons Mass of r-process elements? ==> mulc-color observacons

Opaci-es of and light curves of kilonovae. Masaomi Tanaka (Na-onal Astronomical Observatory of Japan)

Opaci-es of and light curves of kilonovae Masaomi Tanaka (Na-onal Astronomical Observatory of Japan) r-band magnitude @ 100 Mpc Mej = 0.01 Msun -16 19 r-band Absolute magnitude -15-14 -13-12 -11 NS-NS