Multi-Messenger Signatures of the R-Process

Transcription

1 Multi-Messenger Signatures of the R-Process Brian Metzger Columbia University In Collaboration with Rodrigo Fernandez, Eliot Quataert, Geoff Bower, Dan Kasen (UC Berkeley) Andrey Vlasov (Columbia), Almudena Arcones, Gabriel Martinez-Pinedo (Darmstadt) Edo Berger, Wen-Fai Fong (Harvard), Tony Piro, Dan Perley (Caltech) Dimitrios Giannios (Purdue), Shunsaku Horiuchi (VA Tech) Institute for Nuclear Theory R-Process Workshop, August 1, 2014

2 Where, Oh Where Has the R-Process Gone? Galactic r-process rate: M A >130 ~ 5 "10 #7 M! yr -1 (Qian 2000) Snedan et al fractional contribution to the r-process Mergers SNe f NSM ~ f SN ~ # N NSM &# M ej & % (% ( $ 10 "4 yr -1 ' $ 10 "2 M! ' # N SN &# M ej & % $ 10 "2 yr -1 (% ( ' $ 10 "4 M! ' Hypernovae (MHD SNe?) f HNe ~ # N HNe &# M ej & % $ 10 "4 yr -1 (% ( ' $ 10 "2 M! '

3 Where, Oh Where Has the R-Process Gone? Galactic r-process rate: M A >130 ~ 5 "10 #7 M! yr -1 (Qian 2000) Snedan et al fractional contribution to the r-process Mergers SNe f NSM ~ f SN ~ # N NSM &# M ej & % (% ( $ 10 "4 yr -1 ' $ 10 "2 M! ' # N SN &# M ej & % $ 10 "2 yr -1 (% ( ' $ 10 "4 M! ' Hypernovae (MHD SNe?) f HNe ~ # N HNe &# M ej & % $ 10 "4 yr -1 (% ( ' $ 10 "2 M! '

4 Neutrino Driven Wind Neutrinos heat proto-ns atmosphere (e.g. ν e + n p + e - ) drives outflow behind outgoing supernova shock (e.g. Qian & Woosley 96) Before SN Shock Launch After Shock Launch Neutrino-Heated Wind Burrows, Hayes, & Fryxell 1995

5 Conditions for Third Peak R-Process No Symmetry energy shift 1) need Y e < 0.5 Entropy S (kb baryon -1 ) Roberts et al (see also Martinez-Pinedo et al. 2012) 2) α-rich freeze-out of requires Expansion Time τ (s) Thompson et al. 2001

6 Conditions for Third Peak R-Process No Symmetry energy shift 1) need Y e < 0.5 Entropy S (kb baryon -1 ) What is Missing? Roberts et al (see also Martinez-Pinedo et al. 2012) 2) α-rich freeze-out of requires Expansion Time τ (s) Thompson et al. 2001

7 Magnetars Wind Profile for B dip = G Dipole field B ~ G Spin periods ~ 3-10 s (birth periods unconstrained) Birthrate ~10% CC-SNe (massive progenitors > 40 M )

8 rotational energy Are Magnetars Born Rapidly Spinning? "E rot = B2 8# $ 4# 3 R 3 ns E rot = 1 *2 $ P ' 2 I"2 ~ 3 #10 51 & ) ergs % 3 ms( ' "& * % B eq ~ ), (&/2+ Field amplification via MRI or convective dynamo (Thompson & Duncan 1993) 2 ' P * ), ( 3 ms+ -2 G V c ΔΩ Dessart et al Vink & Kuiper 2006 P 0 > 4 ms

9 Effects of Strong Magnetic Fields Helmet - Streamer Microphysics (EOS, ν Heating & Cooling) Important for B G (Duan & Qian 2005) Ω

10 Effects of Strong Magnetic Fields Helmet - Streamer Ω Microphysics (EOS, ν Heating & Cooling) Important for B G (Duan & Qian 2005) Magneto-Centrifugal Slinging (Weber & Davis 1967; Thompson, Chang & Quataert 2004) Outflow Co-Rotates with Neutron Star when B 2 8" > 1 2 # v 2 r Magneto-Centrifugal Acceleration ( Beads on a Wire ) R heat Top View R A

11 Neutrino Winds from Rotating Proto-Magnetars Vlasov, Metzger & Thompson 2014 Velocity θ = θ max θ = 0 Spherical Entropy Spherical θ = θ max Force-free magnetosphere: geometry depends on spin period P, but not on B-field strength

12 M = 1.4 M, R = 12 km Factor ~4 improvement for P~2-5 ms Additional GR boost by factor ~4 (not included) (e.g. Cardall & Fuller 1997) Vlasov, BDM, Thompson

13 Moderately-Rapidly Rotating Proto-Magnetars as R-Process Source Required ejecta mass per event Magnetars have massive progenitor stars more massive proto-ns higher S 3 /τ. Birth periods P > 4 ms consistent with energetics of magnetar supernova remnants. Rapidly rotating stars more prevalent early in chemical evolution of galaxy. - Growing evidence for super-luminous SNe powered by magnetars in low metallicity galaxies (e.g. Kasen & Bildsten 2010) Even if not dominant source, can nucleosynthesis constrain the birth periods of magnetars?

14 X-ray Decay Lines from R-Nuclei in SN Remnants Ripley, BDM, Arcones & Martinez-Pinedo 2014 (see also Qian et al. 1998, 1999) Supernova Remnants Ave. mass per event

15 X-ray Decay Lines from R-Nuclei in SN Remnants Ripley, BDM, Arcones & Martinez-Pinedo 2014 (see also Qian et al. 1998, 1999) Magnetars Remnants (assume M ej = 10-2 M ) Ave. mass per event

16 Where, Oh Where Has the R-Process Gone? Galactic r-process rate: M A >130 ~ 5 "10 #7 M! yr -1 (Qian 2000) Snedan et al fractional contribution to the r-process Mergers SNe f NSM ~ f SN ~ # N NSM &# M ej & % (% ( $ 10 "4 yr -1 ' $ 10 "2 M! ' # N SN &# M ej & % $ 10 "2 yr -1 (% ( ' $ 10 "4 M! ' Hypernovae (MHD SNe?) f HNe ~ # N HNe &# M ej & % $ 10 "4 yr -1 (% ( ' $ 10 "2 M! '

17 Gravitational Waves " 1 P dp dt = 48 5 Binary Neutron Star Mergers G 3 M 2 c 5 a 4 NS Ω a NS Hulse-Taylor Pulsar 10 Known Galactic NS-NS Binaries (Lorimer 2008) T merge = 300 Myr N merge ~ 10-5 "10-4 yr -1 (e.g., Kalogera et al. 2004, Belczynski et al. 2002)

18 Binary Neutron Star Mergers Advanced LIGO / Virgo Range ~ Mpc Detection Rate ~ yr -1 LIGO (North America) Sky Error Regions ~ deg 2 Nissanke et a Virgo (Europe)

19 Gamma-Rays Radio Swift BAT FOV ~ 15% XRT slews in ~min GBM FOV ~ 60% Optical ( Now ) Soon: ZTF LOFAR FOV ~50% Optical (Future) Large Synoptic Survey Telescope (LSST) ~All sky mab<24.5 every ~3 d - Online >~2020

20 Neutron Star Binary Mergers Advanced LIGO/Virgo (>2016) Range ~ Mpc Detection Rate ~ yr -1 NS Ω NS BH NS LIGO (North America) Sky Error Regions ~ deg 2 ~ galaxies Nissanke et a Virgo (Europe)

21 Numerical Simulation - Two 1.4 M NSs Courtesy M. Shibata (Kyoto)

22 Electromagnetic Counterparts of NS-NS/NS-BH Mergers Short GRB Kilonova Metzger & Berger 2012

23 Electromagnetic Counterparts of NS-NS/NS-BH Mergers Short GRB Kilonova Metzger & Berger 2012

24 Numerical Simulation - Two 1.4 M NSs Courtesy M. Shibata (Tokyo U)

25 Remnant Accretion Disk (e.g. Ruffert & Janka 1999; Shibata & Taniguchi 2006; Faber et al. 2006; Chawla et al. 2010; Duez et al. 2010; Foucart 2012; Deaton et al. 2013) Lee et al Disk Mass ~ M & Size ~ km Hot (T > MeV) & Dense (ρ ~ g cm -3 ) Neutrino Cooled: (τ ν ~ ) Equilibrium e + + n " # e + p vs. Y e ~ 0.1 Accretion Rate " M t visc ~ 0.1 % $ ' # 3M! & 1/ 2 M ~ 10 "2 "10M! s -1 " ( % $ ' # 0.1& )1 3 / " R d % 2 " H /R% $ ' $ ' # 100 km& # 0.5 & e " + p # $ e + n )2 s Short GRB Engine?

26 Short GRBs are Rare in the LIGO Volume θ j Detection Rate >z (yr -1 )!!! Swift Short GRBs Metzger & Berger 2012 Redshift z Detectable fraction by all sky γ-ray telescope f " ~ 3.4 # $ 2 j 2 ~ 0.07 % $ ( j ' * & 0.2) 2

27 Electromagnetic Counterparts of NS-NS/NS-BH Mergers Short GRB Kilonova Metzger & Berger 2012

28 Neutron-Rich Ejecta Dynamical Tidal Tails (e.g. Janka et al. 1999; Lee & Kluzniak 1999; Ruffert & Janka 2001; Rosswog et al. 2004; Rosswog 2005; Shibata & Taniguchi 2006; Giacomazzo et al. 2009; Duez et al. 2010; East et al. 2012; Hotokezaka et al. 2013) Full GR / Simple EOS / Circular M ej ~ M Y e " n p n p + n n < 0.1 Model M ej (10-3 M ) Hotokezaka et al Newtonian / Realistic EOS / Eccentric

29 Neutron-Rich Ejecta Dynamical Tidal Tails (e.g. Janka et al. 1999; Lee & Kluzniak 1999; Ruffert & Janka 2001; Rosswog et al. 2004; Rosswog 2005; Shibata & Taniguchi 2006; Giacomazzo et al. 2009; Duez et al. 2010; East et al. 2012; Hotokezaka et al. 2013) Full GR / Simple EOS / Circular M ej ~ M Y e " n p n p + n n < 0.1 Model M ej (10-3 M ) Hotokezaka et al Newtonian / Realistic EOS / Eccentric Disk Outflows Neutrino-Powered (Early) (e.g. McLaughlin & Surman 05; Surman+08; BDM+08; Dessart+09) Recombination-Powered (Late) (e.g. Beloborodov 08; BDM+08, 09; Lee+09; Fernandez & BDM 13) Y e ~??? M ej = f w M d ~ (f w /0.1) M

30 as used in Metzger et al (movie courtesy G. Martinez-Pinedo, A. Arcones) R-Process Network (neutron captures, photo-dissociations, α- and β-decays, fission)

31 Final Abundance Distribution 2nd Scaled solar abundances 3rd BDM et al peaks at A ~ 130, 195

32 Radioactive Heating of Merger Ejecta (BDM et al. 2010; Roberts et al. 2011; Goriely et al. 2011; Korobkin et al. 2012; Bauswein et al. 2013) Y e = 0.1 t -1.2 Metzger et al Dominant β-decays at t ~ 1 day: 132,134,135 I, 128,129 Sb, 129 Te, 135 Xe Relatively insensitive to details (Y e, expansion history, NSE or not)

33 How Supernovae Shine (Arnett 1982; Li & Paczynski 1998) ejecta with mass M, velocity v, thermal energy E = f Mc 2, opacity κ M }R V R = v t " = M 4# 3 R 3 " ~ #$R t diff ~ " R c ( ) "1/ 2 % M M! Peak (t = t diff ) t peak ~ 2 weeks v 10 4 km s -1 L peak ~ E(t peak)t ~ ergs s -1 ( f )( v ) 1/ 2 peak 10-5 $ M 10 4 km s -1 M! # $ " # 1/ 2 & # ( % ) ' $ ) Fe 1/ 2 % " ' ( % $ ' & # ( Fe & Type Ia Supernova: v ~10 4 km s -1, M ej ~ M, f Ni Co ~ 10-5 t peak ~ week, L ~ erg s -1 NS Merger: v ~ 0.1 c, M ej ~ 10-2 M, f ~ t peak ~ 1 day, L ~ erg s -1 1/ 2 & ( ' )1/ 2

34 Radioactive Heating of Merger Ejecta (BDM et al. 2010; Roberts et al. 2011; Goriely et al. 2011; Korobkin et al. 2012; Bauswein et al. 2013) Metzger et al Dominant β-decays at t ~ 1 day: 132,134,135 I, 128,129 Sb, 129 Te, 135 Xe Relatively insensitive to details (Y e, expansion history, NSE or not)

35 Bolometric Luminosity kilo-nova Color Evolution Fe Opacity Blackbody Model BDM et al. 2010

36 High Opacity of the Lanthanides (Kasen et al. 2013; Tanaka & Hotokezaka 2013) lanthanides Fe Kasen et al. 2013

37 Bolometric Luminosity kilo-nova Color Evolution Fe Opacity Blackbody Model BDM et al. 2010

38 EM Counterpart Search following a GW Trigger Range of kilonova models with different ejecta mass M ej ~ M and velocity v ~ c BDM & Berger 2012 Requires depth J ~ and short cadence

39 Bolometric Luminosity kilo-nova Color Evolution Fe Opacity Blackbody Model BDM et al. 2010

40 Bolometric Luminosity kilo-nova Color Evolution Fe Opacity Blackbody Model BDM et al Y e > 0.25 Y e < 0.25

41 Neutron-Rich Ejecta Dynamical Tidal Tails (e.g. Janka et al. 1999; Lee & Kluzniak 1999; Ruffert & Janka 2001; Rosswog et al. 2004; Rosswog 2005; Shibata & Taniguchi 2006; Giacomazzo et al. 2009; Duez et al. 2010; East et al. 2012; Hotokezaka et al. 2013) Full GR / Simple EOS / Circular M ej ~ M Y e " n p Y e " 0.05 n p + n n < 0.1 Model M ej (10-3 M ) Hotokezaka et al Newtonian / Realistic EOS / Eccentric Disk Outflows Neutrino-Powered (Early) (e.g. McLaughlin & Surman 05; Surman+08; BDM+08; Dessart+09) Recombination-Powered (Late) (e.g. Beloborodov 08; BDM+08, 09; Lee+09; Fernandez & BDM 13) Y e ~??? M ej = f w M d ~ (f w /0.1) M

42 Two Component Blue Light Curve Red Barnes & Kasen 2013 Grossman+14 or disk winds: r-process or Fe? tidal tail: r-process

43 Remnant Torus Evolution (Fernandez & Metzger 2012, 2013) P-W potential with M BH = 3,10 M hydrodynamic α viscosity NSE recombination 2n+2p 4 He Equilibrium Torus M t ~ M R 0 ~ 50 km uniform Y e = 0.1 run-time Δt ~ t orb neutrino self-irradiation: light bulb + optical depth corrections: BH neutrino emission pattern R [2,2000] R g N r = 64 per decade N θ = 56

44

45 Delayed Disk Outflows M BH M out (with " recombination) M out (NO " recombination) Fernandez & BDM 13 Time (s) unbound outflow powered by viscous heating and α recombination neutrino heating subdominant outflow robust M ej ~ 0.05 M t V ej ~ 0.1 c compare with Just et al. (2014), who find M ej ~ 0.2 M t

46 Y e Freeze Out Outflow Composition t = 0.3 s t = 3 s t = 0.03 s still produces lanthanides Composition (Preliminary) Mass per bin (M )

47 Y e Freeze Out Outflow Composition Compare with: Just et al t = 0.3 s t = 3 s t = 0.03 s still produces lanthanides Composition (Preliminary) Mass per bin (M )

48 Two Component Blue Light Curve Red Barnes & Kasen 2013 Grossman+14 or disk winds: r-process or Fe? tidal tail: r-process

49 Two Component Blue Light Curve Red Barnes & Kasen 2013 Grossman+13 disk winds: r-process or Fe? tidal tail: r-process

50 (Tanvir+13; de Ugarte Postigo+13 ; Fong+13 ) Tanvir et al M 0.01 M

51 Two Component Blue Light Curve Red Barnes & Kasen 2013 Grossman+13 disk winds: r-process or Fe? tidal tail: r-process

52 Two Component Blue Light Curve Red Barnes & Kasen 2013 Grossman+13 or disk winds: r-process or Fe? tidal tail: r-process

53 Two Component Blue Light Curve Red tidal tail: r-process disk winds: r-process or Fe? ν ν NS ν ν " e + n #e $ + p Barnes & Kasen 2013 Grossman+13

54 Effect of Hypermassive Neutron Star Metzger & Fernandez 2014, see also Perego et al. 2014

55 Distribution of Ejecta Y e for Different Collapse Times t collapse = 0 t collapse = 30 ms t collapse = 100 ms

56 Imprint of the HMNS Lifetime equatorial 30 ms 100 ms 300 ms infinity strength of blue bump may encode HMNS lifetime ejecta mass up to ~10 times higher than prompt BH case

57 Stable Merger Remnant? (e.g. BDM+08; Ozel et al. 2010; Bucciantini et al. 2012; Zhang 13; Yu et al. 2013; Giacomazzo & Perna 13; Siegel 2014) Requires: low total mass binary, stiff EOS*, and/or mass loss during merger *supported by recent discovery of 2M NS by Demorest et al Rotating at centrifugal break-up limit with spin period P ~ 1 ms Magnetic field amplified by rotational energy + convection Magnetar? Giacomazzo & Perna 2013

58 Short GRBs with Extended Emission 1/5 Swift Short Bursts have X-ray Tails GRB Rapid Variability Ongoing Engine Activity Energy up to ~30 times Burst Itself! Extended Emission GRB S EE /S GRB ~ 30 Perley, BDM et al BATSE Examples (Norris & Bonnell 2006)

59 Stable Merger Remnant? (e.g. BDM+08; Ozel et al. 2010; Bucciantini et al. 2012; Zhang 13; Yu et al. 2013; Giacomazzo & Perna 13; Siegel 2014) spin-down luminosity spin-down time : L sd = µ2 " 4 % P ( # 6 $10 49 ' * c 3 & 1 ms) : " sd = E rot L sd # 5 $ P 0 ' & ) % 1 ms( % B dip ( ' & * erg s -1 G) $ B dip ' & % ) G( -2 min Magnetar wind confined by merger ejecta Bucciantini, BDM et al Jet Magnetar Wind Merger Ejecta Theoretical Light Curves vs. Observed X-ray Tails (magnetar wind model from Metzger et al. 2011) P 0 = 1.5 ms, B dip = G

60 Radio constraints on stable merger remnants (BDM & Bower 2013) Rotational energy eventually transferred to ISM via relativistic shock bright radio emission We observed 7 short GRBs with VLA on timescales ~1-3 years after burst NO DETECTIONS stable remnant disfavored in 2 GRBs with high ISM densities Additional JVLA observations now would be much more constraining Upcoming radio surveys (e.g. ASKAP) will strongly constrain stable NS merger remnants indirectly probes EoS 1.4 GHz Luminosity (erg s -1 ) Rest-Frame Time Since GRB (years)

61 Timeline of Binary NS Mergers 1. Chirp enters LIGO Bandpass t (minus) ~ mins 2. Last Orbit, Plunge & Dynamical Ejecta t ~ ms 3. BH Formation ~ ms - 4. Accretion of Remnant Disk, Jet Formation (GRB) ~ s 5. He-Recombination + Disk Evaporation ~ s outflow Y e depends on NS collapse time 6. R-Process in Merger Ejecta ~ few s 7. Jet from Magnetar (X-rays) ~ min (or longer) 8. Disk Wind Kilonova prompt BH formation Y e < 0.25 (NIR, L ~ erg s -1 ) ~ week delayed BH formation Y e > 0.25 (Optical, L ~ erg s -1 ) ~ day stable magnetar (Optical, L ~ erg s -1 ) ~ day 9. Tidal Tail Kilonova (IR) ~ week 10. Ejecta ISM Interaction (Radio) ~ years Much brighter if stable magnetar

62 Conclusions Strong B fields and rapid rotation alter the properties of proto-neutron star winds due to magneto-centrifugal acceleration. Both the dynamical timescale and entropy increase, increasing S 3 /t exp by a factor ~4. Moderately rapidly rotating proto-magnetars thus represent a promising r-process site, consistent with observations. The first direct detection of gravitational waves will likely be a binary NS merger, within the next ~3 years. Identifying an EM counterpart will be essential to maximize the scientific impact of this discovery. The most promising isotropic counterpart is an optical/ir transient ( kilonova ) powered by the radioactive decay of r-process nuclei. The radioactive heating of the ejecta is now well understood, but the photon opacity of r-process ejecta remains uncertain. The sensitive dependence of opacity on ejecta composition (lanthanide fraction) make kilonova colors a sensitive probe of physical processes at work during the merger, such as the delay until black hole formation.