NEUCOS, DESY, May 29–June 2 2017 Nuclear composi.on of GRB jets: sources and survival Shunsaku Horiuchi (Center for Neutrino Physics, Virginia Tech) Image credit: NASA/ESA Contents A. Heavy nuclei in massive stars B. The nuclei composiTon of jets – Stellar nucleosynthesis – IniTal loading – Explosive nucleosynthesis – Entrainment – In-situ jet nucleosynthesis – DestrucTon processes C. Acceleraon to UHECR D. Conclusion NEUCOS Shunsaku Horiuchi (Virginia Tech) 2 Stellar Nucleosynthesis Nucleosynthesis Iron nuclei abundance Massive stellar nucleosynthesis The final Fe core mass is typically some 1.5 results in the famous onion-shell Msun with radius of 108 cm or so structure: + : Fe core mass Core mass IniTal mass Woosley et al (2002) NEUCOS Shunsaku Horiuchi (Virginia Tech) 3 Explosive Nucleosynthesis Propagaon of shock wave Compression Explosive through the core & envelope and heang nucleosynthesis Energy injecon Mass cut defines final remnant Expansion Shock compression The rest and heang is ejected However, how much heavy nuclei is released is model dependent: 1. Some amount of energy injecTon 2. At some locaon 3. With some mass cut NEUCOS Shunsaku Horiuchi (Virginia Tech) 4 How much is possible? A lot of 56Ni (potenTally) With large CO core, large explosion energies, and small mass cut, up to ~10 Msun ~109 cm Core collapses typically observed ~0.1 Msun of 56Ni. Umeda & Nomoto (2008) (some may have more, e.g., hypernovae, superluminous supernovae; x100 needed if 56Ni) NEUCOS Shunsaku Horiuchi (Virginia Tech) 5 1. IniTal loading 2. Entrainment 3. In-situ nucleosynthesis 4. DestrucTon processes COMPOSITION OF JETS NEUCOS Shunsaku Horiuchi (Virginia Tech) 6 1. Ini.al loading External composiTon Depends on jet model and launch Tming, e.g., • Original stellar nuclei if launched prior to supernova shock revival • More nuclei if aer (20Msun) explosive nucleosynthesis Can nuclei survive? NEUCOS Shunsaku Horiuchi (Virginia Tech) 7 Survival in inial loading Photodisintegraon Thermal photons with T0 as: Spallaon Target ion/nucleons thermalized to T0 Nuclei survival OpTcal depths of destrucTve processes must be small 8 à Needs large r0 > 10 cm or 49-50 low Lrad < 10 erg/s à Low-luminosity or magnec models beer for survival than fireball NEUCOS Shunsaku Horiuchi (Virginia Tech) Horiuchi, Murase, et al (2012) 8 Simula.ons Loading simulaons 2D relavisTc MHD simulaon of jet induced collapse, mass fall back, and jet bulk acceleraon. Fireball • Fireball: heavy nuclei dissociated • MagneTc (parTally): parTal dissociaon of nuclei à Magne2c models (and low-Lrad) be6er for nuclei composion Magne2c magneTzaon NEUCOS Shunsaku Horiuchi (Virginia Tech) Shibata & Tominaga (2015) 9 2. Entrainment Entrainment of external nuclei into the jet medium: do nuclei survive? star jet nuclei Can it safely do this? NEUCOS Shunsaku Horiuchi (Virginia Tech) Zhang et al. (2003) 10 Survival in entrainment Cocoon The cocoon is made up of shocked Jet head Stellar maer ~ stellar and shocked jet material and has βh 0.1 (plenty of nuclei) expands non-relavisTcally into the stellar material. Jet has Γ >> 1 β(1) 0.01L3/8 r1/2 c ⇠ ke,50 9 (1) 3/16 1/2 nuclei-rich Tc 100L r9− keV ⇠ ke,50 à Cocoon can be nuclei rich (mixing Cocoon aided by instabilies) e.g., Aloy (2002) JET βc ~ 0.01 e±, γ, some T cocoon jet h nuclei nuclei cocoon Stellar maer Can it safely do this? NEUCOS Shunsaku Horiuchi (Virginia Tech) 11 Survival in entrainment Survival Demand the nuclei velocity is always below the spallaon threshold à requires the nuclei to be thermalized FASTER than it takes to move up the velocity gradient and its speed becomes too fast Rapid thermalizaon for Survival KEFe < 20 GeV Energy loss on electrons are fast (can à If velocity gradient is small, nuclei be further helped by pair thermalize before reaching the e+e- when T is high) and spallaon is slower spallaon threshold (Collimated jets can tolerate a higher gradient due to higher e- density) cocoon jet nuclei Conical jet (easier for collimated jet) Can it safely do this? NEUCOS Shunsaku Horiuchi (Virginia Tech) Horiuchi et al (2012) 12 3. In-situ jet nucleosynthesis GRB fireball: • IniTal radiaon temperatures of a few MeV à nuclei are dissociated (NSE) 5 • Large entropy (nγ/np ~ 10 ) • needed for saturaon Lorentz factor η > 100 • Rapid expansion Tme scales (~0.1 ms) • Electron fracTon probably close to 0.5 • but uncertain NEUCOS Shunsaku Horiuchi (Virginia Tech) 13 In-situ jet nucleosynthesis: fireball Nucleosynthesis Bokleneck to heavy nuclei synthesis due to easily destroyed deuterium. He d When T is low enough for the bokleneck to be broken, the Mass fracTon jet is too dilute for tripe- alpha to be fast enough. Freeze-out composiTon à Very few heavy nuclei are Radius [cm] made Pruet et al (2002), Lemoine (2002), Beloborodov (2003) NEUCOS Shunsaku Horiuchi (Virginia Tech) 14 Alterna.ves Importance of entropy and expansion Tmescale Lower entropy and larger Tmescale are conducive to nucleosynthesis à Deuterium producTon is more efficient and deuterium bokleneck is broken earlier while densiTes are sTll high. 1. MagneTc scenario Lrad can be low if the GRB can be powered magneTcally, e.g., for rapidly rotang proto-magnetar model e.g., Usov (1992) ˙ 49 4 2 6 1 Thompson (1994) E 10 Pms− B15R6 erg s− ⇠ e.g., Blandford & Other models, e.g., magneTzed disk winds or BH powered Znajek (1977) 2. Low-luminosity GRB and dirty (baryon) jets With lower Lrad by default (but enough for iniTal dissociaon). With large baryon loading by default NEUCOS Shunsaku Horiuchi (Virginia Tech) 15 Example: proto-magnetar scenario Proto-magnetar model: Neutrino-driven wind + The GRB jet fills a sweet spot; later, it becomes a magneTc driven ourlow pair-wind as the neutrino-driven wind subsides χ=0: aligned χ=π/2: oblique Pre-jet break phase Jet properTes GRB phase Pair wind phase “magneTzaon” E˙ σ0 = Time since core bounce Mc˙ 2 Metzger et al (2011) (=η if fully converted to KE) NEUCOS Shunsaku Horiuchi (Virginia Tech) 16 Nucleosynthesis yields Freezeout yield v EsTmate the mass fracTon of A≥56 (Xh) with analyTc wind v nucleosynthesis • τexp: Tmescale v • s: entropy Roberts et al (2010) • Ye: electron fracTon For an oblique* rotator: Pre-jet • Expansion Tme-scale [ms] break phase • Entropy [kB/baryon] • Electron fracTon? *Obliquity reduces ν-heang GRB phase Pair wind phase by centrifugal force Time since core bounce NEUCOS Shunsaku Horiuchi (Virginia Tech) Metzger, Giannios, Horiuchi (2011) 17 Nucleosynthesis yields Electron fracTon: evolves by neutrino irradiaon to 0.4 – 0.6, but may be different due to B field alignment ˙ νe + n e− + p n N⌫¯e σ⌫¯e L⌫¯e E⌫¯e ! + è h i ν¯e + p e + n p ! N˙ σ ⇠ L⌫ E⌫ ! ⌫e ⌫e e h e i Nucleosynthesis n-rich à Freezeout composion Pre-jet can be heavy-dominated p-rich, oblique during the GRB phase break phase GRB phase Especially for: • IniTally n-rich maer Pair wind phase • Oblique rotators (receive less ν-heang and hence have lower entropies) A ≥56 nuclei mass fracTon Time [s] NEUCOS Shunsaku Horiuchi (Virginia Tech) Metzger, Giannios, Horiuchi (2011) 18 Simula.ons Aligned rotator simulaons Force-free approximaon, aligned dipole field 1014-15 G, SkyNet nuclear reacTon network. Shows r-process nucleosynthesis across a range of rotaon periods Approximately 40% He, trace n and p The rest: abundance-weighed mean A alpha neutron proton NEUCOS Shunsaku Horiuchi (Virginia Tech) Vlasov et al (2017) 19 4. Collisions with neutrons Neutrons are collisionally coupled to the accelerang plasma: EM coulomb strong γ e p n $ $ $ But they lag behind if τcollision > τacc • Make sure the relave velocity ⌧coll 1 3 β˜ L− r ⌘ ⇠ ⌧acc / does not exceed the spallaon threshold Conical jet • Or, the neutrons decouple (easier for collimated jet) before the spallaon threshold is reached Horiuchi et al (2012) Horiuchi, Murase, et al (2012) à Nuclei survive unless η is very large NEUCOS Shunsaku Horiuchi (Virginia Tech) 20 5. Recollima.on shocks Jet collimaon by cocoon pressure à becomes a shock and the post-shocked jet moves at Γs 1/✓j 8 ⇠ ⇠ Bromberg et al (2011) Ions obtain random moTons with the pre- and post-shocked relave Lorentz factor Γ /2Γ 10 ⇠ j s ⇠ But the thermal temperature is relavely high (but not high enough for photodisintegraon) à conducive to pair-creaon à creates more target to thermalize with Conical jet à Check if thermalizaon is faster than spallaon Mizuta & Aloy (2009) à Will be a problem once shock grows Horiuchi, Murase et al (2012) NEUCOS Shunsaku Horiuchi (Virginia Tech) 21 Brief Summary Fireball GRB Magne2c GRB Low-luminosity GRB Source: stellar Y Y Y nucleosynthesis Survives: iniTal loading? N Y Y Survives: entrainment? gradient gradient gradient Source: jet nucleosynthesis N Y maybe Survives: n-collisions? Y Y Y Survives: oblique shocks? Only early If collimaon Only early magneTc “Y” means possible for canonical parameters; “N” is not possible; è In Fireball GRB, nuclei must be entrained beyond recollimaon shock radii è In magneTc GRB, mulTple opTons è In LL GRB, mulTple opTons NEUCOS Shunsaku Horiuchi (Virginia Tech) 22 ACCELERATION & SURVIVAL NEUCOS Shunsaku Horiuchi (Virginia Tech) 23 Proto-magnetar model Acceleraon: ~ 15-16 Demand acceleraon is Collision case: Rd 10 cm faster than cooling & GRB phase expansion Tmescales: Cooling limit Expansion limited Survival: Calculate the opTcal depth of photodisintegraon based on the Band funcTon for the Nuclei UHECR photon spectrum phase Demanding this is = 1 (or a few, allowing for a few destrucTons) defines the UHECR phase Metzger, Giannios, Horiuchi (2011) à There remains a window for