The Theory of Supernovae in Massive Binaries Philipp Podsiadlowski (Oxford)
• the majority of massive stars are in interacting binaries • the large diversity of observed supernova types and (sub-)types is almost certainly caused by the diversity of binary interactions
I. Binary Evolution and the Progenitor Appearance II. The Fates of Stars in Binaries (vs. Single Stars) III. Neutron-Star and Black-Hole Formation IV. [Exotic Supernovae] Binary Interactions
• most stars are members of binary Classification of Roche-lobe overflow phases systems (Paczynski) R/ R carbon ignition P = 4300 d • a large fraction are members of O. 1000 interacting binaries (30 − 50 %) 45 % M = 5 M Sana et al. (2012): 1 O. M / M = 2 1 2 Case C % > ⊙ 70 for O stars with M ∼ 15 M radius evolution • note: mass transfer is more likely for 100 helium ignition post-MS systems P = 87 d • binary interactions 45 % ⊲ common-envelope (CE) evolution ⊲ stable Roche-lobe overflow Case B ⊲ binary mergers 10 ⊲ tidal interactions (de Mink; → P = 1.5 d Marchant) Case A 10 % ⊲ P = 0.65 d wind Roche-lobe overflow (case D) main sequence (Podsiadlowski & Mohamed 2007; → 7 Pols) 0 5 10 (10 yr) Supernova lightcurves (core collapse) LIGHTCURVES OF CORE-COLLAPSE SUPERNOVAE • central explosion may be very similar in all cases (with E ∼ 1051 ergs) SN 1969L • variation of lightcurves/supernova subtypes mainly due to varying envelope properties SN 1987A ⊲ envelope mass: determines thermal diffusion time and length/existence of plateau ⊲ envelope radius: more compact progenitor → more expansion work SN II-b required → dimmer supernova SN II-L SN II-P • binary interactions mainly affect stellar envelopes Sequence of Mass Loss SN II-P → II-L → IIb → Ib → Ic
Hsu, Ross, Joss, P. Stable Mass Transfer Unstable Mass Transfer
• mass transfer is ‘largely’ • dynamical mass transfer → conservative, except at very common-envelope and spiral-in phase mass-transfer rates (mass loser is usually a red giant) • mass loss + mass accretion ⊲ mass donor (primary) engulfs • the mass loser tends to lose most of secondary its envelope → formation of helium ⊲ spiral-in of the core of the primary stars → hydrogen-deficient and the secondary immersed in a supernovae (II-L, IIb, Ib, Ic) common envelope • the accretor tends to be rejuvenated • if envelope ejected → very close binary (i.e. behaves like a more massive star (compact core + secondary) with the evolutionary clock reset) • otherwise: complete merger of the • orbit generally widens binary components → formation of a single, rapidly rotating star PTF Rates
Arcavi et al. (2010) Podsiadlowski (1989, 1992) Stripped Supernovae How much mass remains after mass transfer? • stable mass transfer ⊲ mass transfer ends when donor shrinks below Roche lobe
→ remnant H envelope (up to ∼ 1 M⊙; [Joss & Podsiadlowski 1988; PhP1992]) → extended, stripped supergiant (SN II-L, ex- tended IIb [SN 93J]) ⊲ Problem: not enough? (Podsiadlowski+ 1992; Claeys+ 2011) ⊲ Solution: extend mass transfer parameter space: wind RLOF (case D) • CE ejection ⊲ small amount of H left ⊲ stellar wind can remove H (metallicity depen- dence; compact SNe IIb vs. SNe Ib) The SN Ic problem Caution: there are different types of SNe Ic (broad-lined [engine]; normal SNe Ic’s) solar metallicity
• large fraction of SNe Ic compared to . O
SNe Ib (latest: comparable numbers single stars [Modjaz+ 2015]) faint/no SN • II−P are they He free? Ib • Hachinger & Mazzali (2012): < . ⊙ 0 1 M 500 1000 1500 2000 Ic maximum radius (R ) • incompatible with all models 10 15 20 25 30 35 40 • He non-thermally excited by Ni Initial Mass (M ) decay: different mixing in O. supernovae? (PhP+ 2016)
PhP, Mazzali, Justham Binary Accretion On Main Sequence (Case A) • rejuvenation (8 %; de Mink; → Schneider) • further evolution similar to single star Post Main Sequence (Case B/C) • change core to envelope mass ratio → favours blue solutions (no red supergiant phase) (Podsiadlowski & Joss 1989)
Podsiadlowski & Joss (1989) Maund et al. (2004) Slow Mergers
Dynamical Mergers
Nandez, Ivanova & Lombardi (2014) Also collisional mergers in clusters Ivanova (2002) Example: SN 1987A LBV Supernovae from Massive Binary Mergers Justham, Podsiadlowski & Vink (2014)
• large number of O-star binary mergers (Sana et al. [2012]: 20–30%) • for sufficiently small core mass fraction ⊲ He burning in blue-supergiant phase (cf. Podsiadlowski & Joss 1989) ⊲ with relatively low-mass loss rate ⊲ transition to the red only after He-core burning → possibility of SN explosion in LBV phase (with various amounts of H envelope masses) Justham et al. (2012) • even relatively massive stars may produce neutron stars rather than black holes (low entropy, plus core erosion) • large diversity of outcomes: yellow supergiants, SNe IIn, interaction SNe, PISNe
Justham et al. (2014) Core Collapse Supernovae Iron core collapse Electron-capture supernovae • • inert iron core (> MCh) occurs in degenerate ONeMg core collapses ⊲ at a critical density − ⊲ presently favoured model: (4.5 × 109 g cm 3), corresponding delayed neutrino heating to a critical ONeMg core mass to drive explosion (1.370 ± 0.005M⊙), electron captures onto 24Mg removes electrons (pressure support!) → triggers collapse to form a low-mass neutron star
ν note: essentially the whole core ν ν ν ν ν Iron Core ν ν collapses ν NS ν ν → ν easier to eject envelope/produce ν Collapse ν ν ν ν ν supernova → no significanct ejection of heavy Kifonidis elements Second dredge−up in AGB stars (around 10 Msun)
with H envelope without H envelope