the formation of neutron stars (and black holes) in binaries · the formation of neutron stars (and...
TRANSCRIPT
The Formation of Neutron Stars (and Black Holes)
in Binaries
Philipp Podsiadlowski (Oxford)
• the majority of massive stars are in interacting binaries
• the final structure and fate of massive stars is very different in
binary systems
I. Binary Interactions
II. The Fates of Stars in Binaries (vs. Single Stars)
III. Supernova Kicks
IV. Black-Hole Formation
Binary Interactions
• most stars are members of binary
systems
• a large fraction are members of
interacting binaries (30− 50%)
Sana et al. (2012):
70% for O stars with M ∼> 15M⊙
• note: mass transfer is more likely for
post-MS systems
• mass-ratio distribution:
⊲ for massive stars: masses correlated
⊲ for low-mass stars: less certain
• binary interactions
⊲ common-envelope (CE) evolution
⊲ stable Roche-lobe overflow
⊲ binary mergers
⊲ wind Roche-lobe overflow
R/ R .
radius evolution
O
Classification of Roche-lobe overflow phases
M = 5 M
2M / M = 2
1
O.1
45 %
45 %
10 %
helium ignition
carbon ignition P = 4300 d
P = 0.65 d
P = 1.5 d
P = 87 d
Case C
main sequence
Case A
(Paczynski)
100
10
1000
10 (10 yr)750
Case B
Stable Mass Transfer
• mass transfer is ‘largely’
conservative, except at very
mass-transfer rates
• mass loss + mass accretion
• the mass loser tends to lose most of
its envelope → formation of helium
stars → hydrogen-deficient
supernovae (IIb, Ib, Ic)
• the accretor tends to be rejuvenated
(i.e. behaves like a more massive star
with the evolutionary clock reset)
• orbit generally widens
Unstable Mass Transfer
• dynamical mass transfer →
common-envelope and spiral-in phase
(mass loser is usually a red giant)
⊲ mass donor (primary) engulfs
secondary
⊲ spiral-in of the core of the primary
and the secondary immersed in a
common envelope
• if envelope ejected → very close binary
(compact core + secondary)
• otherwise: complete merger of the
binary components → formation of a
single, rapidly rotating star
The Formation of NS-NS (NS-BH) Binaries
X X
XX
v∆
v∆
X
X
X
X
HeNS
Core Collapse Supernovae
Iron core collapse
• inert iron core (> MCh)
collapses
⊲ presently favoured model:
delayed neutrino heating
to drive explosion
νν ν
ν ν νν
νν
ννννννν
νν
Kifonidis
NS
Iron Core
Collapse
Electron-capture supernovae
• occurs in degenerate ONeMg core
⊲ at a critical density
(4.5× 109 g cm−3), corresponding
to a critical ONeMg core mass
(1.370± 0.005M⊙), electron
captures onto 24Mg removes
electrons (pressure support!)
→ triggers collapse to form a low-mass
neutron star
note: essentially the whole core
collapses
→ easier to eject envelope/produce
supernova
→ no significanct ejection of heavy
elements
����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������
����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������
������������������������������������������������������������������������
������������������������������������������������������������������������
−−> electron−capture supernova
without dredge−up−−> larger CO core mass
in ONeMg coreo lower explosion energyo lower supernova kickso NS mass: 1.25 Msun
Second dredge−up in AGB stars (around 10 Msun)
with H envelope without H envelope
AGB envelope
dredge−up of the He core−−> lower CO core masses−−> ONeMg WD
CO core
(Podsiadlowski et al. 2004)
Binary Evolution Effects
• dredge-up in AGB phase may prevent
ONeMg core from reaching Mcrit →
ONeMg WD instead of collapse
• can be avoided if H envelope is removed
by binary mass transfer
→ dichotomous kick scenario
(Podsiadlowski, Langer, et al. 2004)
⊲ e-capture SN in close binaries → low
kick
⊲ iron core collapse → high kick
Subsequent Work: Single Stars
Arend Jan Poelarends (PhD Thesis):
• examined conditions for e-capture SNe on metallicity, wind mass loss, dredge-up
efficiency in AGB stars
• best model: no e-capture SN at solar Z
Recent Work: Binary Stars → Thomas Tauris
�������������������������
�������������������������
������������������������������������������
������������������������������������������
helium burningconvective core(growing)
No H−burning shell
shrinkingHe−burning core
without H envelope
H−burning shell
with H envelope
−−> larger CO cores with lowerC/O ratio −−> no convective carbon burninghigher entropy (more massive) iron cores
−−−> BLACK HOLE
smaller CO cores with higher
lower entropy (mass) iron cores
He−core−burning stars (M > 20 − 25 Msun)
(Brown, Lee, Heger)−−−> NEUTRON STAR (60/70 Msun?)
C/O ratio −−> convective carbon burning
Brown, Heger, Langer et al. (2001)
Carbon Burning and Final Fe CoreMasses
(Brown et al. 2001)
• late He-core burning: 12C + α becomes
dominant and determines the final 12C
fraction
⊲ stars with H-burning shell: injection of
fresh He → long 12C + α phase → low final
C fraction
⊲ stars without H-burning shell: short12C + α phase → higher final C fraction
• C-core burning:
⊲ high C fraction → convective C burning
→ higher neutrino losses → lower-entropy
cores → lower-mass O and ultimately Fe
cores → neutron stars
⊲ low C fraction → radiative C burning →
lower neutrino losses → higher-entropy
cores, etc. → black holes
Petermann, Langer & Podsiadlowski (2015)
Petermann, Langer & Podsiadlowski (2015)
LBV Supernovae from MassiveBinary 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
⊲ 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)
Justham et al. (2014)
Neutron Star Birth Kicks
• single radio pulsars have large space velocities (Lyne &
Lorimer; Hobbs et al. 2005): σv = 265kms−1 without evi-
dence for a low-velocity component
Evidence for Low Supernova Birth Kicks
• neutron star retention in globular clusters (e.g. Pfahl,
Ivanova)
• the existence of wide Be/X-ray binaries with low eccen-
tricities (e.g. X Per) (Pfahl)
• DNSs with low eccentricities (van den Heuvel, Dewi)
• the spin period – eccentricity relation of DNSs (Dewi)
• preference for low-kick NSs in binaries?
The origin of supernova kicks
• dramatic recent progress in neutrino-driven
core-collapse simulations
• supernova kicks produced by standing accretion shock
instability (SASI) (Blondin, Mezzacappa, Foglizzo,
Janka)
• driven by advective-acoustic instability
• l = 1 instability in two flavours:
⊲ sloshing instability (m = 0)
⊲ spiral mode (m = ±1)
• can produce kicks of a few 100 km s−1 if the collapse
phase lasts ∼> 500ms (many growth timescale)
• prediction (Podsiadlowski, Langer, et al. 2004):
⊲ large kicks for slow explosions (standard Fe cores),
⊲ low kicks for fast explosions (e-capture SN, low-mass
Fe cores)
Sloshing Instability
(l = 1, m = 0)
(Janka, Scheck, Foglizzo)
Iwakami et al. (2008)
Spiral Mode
(l = 1, m = ±1)
(Blondin, Mezzacappa)
Podsiadlowski, Langer et al. (2004)
The Bimodal Spin Distribution of NSs in Be X-Ray BinariesKnigge, Coe & Podsiadlowski (2011)
Knigge, Coe & Podsiadlowski (2011)
• spin period may be a better proxy
for NS formation channel (?)
• comparable numbers of Fe core
collapse and e-capture NSs
• Puzzle: Why is the equilibrium spin
period different for the different NS
formation channels?
⊲ different B fields? (fast vs. slow
explosion phase)
⊲ different (accretion) geometry?
(e.g. misalignment of spin)
• Be X-ray binaries may be useful for
constraining NS formation and the
formation of double NS binaries
Black-Hole Formation
• single stars (case C binaries):
⊲ fallback black holes (up to 40M⊙?;
Fryer): faint supernova, NS-type kick
⊲ ‘prompt’ black holes: no supernova,
no kick
• binaries:
⊲ may need case C (i.e. very late) mass
transfer
⊲ uncertain range of case C for massive
stars (radius evolution, wind mass
transfer?)
• accretion-induced collapse of
accreting NS in I/LMXB
⊲ for efficient accretion efficiency →
form low-mass black holes in
binaries
⊲ not observed?
• accretion-induced collapse during
in-spiral in massive envelope (e.g.
Chevalier, Fryer, Brown)
• Demorest et al. (2010): PSR
1614-2230
⊲ MNS = 1.97± 0.04M⊙,
MWD = 0.5M⊙
⊲ massive WD requires
intermediate-mass
progenitor (Li et al. 2011;
Tauris et al. 2011)
→ relatively massive NS at
birth (> 1.6M⊙)
Li, Rappaport, Podsiadlowski (2011)
• spiraling-in NS in massive envelope,
collapse to black hole if in
neutrino-dominated regime
(Chevalier, Fryer, Brown,)
• form NS+BH binaries instead of
NS+NS binary
→ implications for direct
gravitational-wave detections with
aLIGO (direct test!)
• perhaps not!
⊲ recent study by MacLeod &
Ramirez-Ruiz of Hoyle-Lyttleton
accretion with large density
gradients
⊲ streams deflected due to angular
momentum → low accretion
efficiency (∼< 0.1M⊙) → NS
survival likely
MacLeod & Ramirez-Ruiz (2015)
.Initial binary: M
1
= 14M
�
,
M
2
= 9M
�
, P
orb
= 190 d
Stable non- onservative Case
B mass transfer leaving a
helium star with M
A
He
= 4M
�
and M
0
2
= 11M
�
, P
orb
= 350 d
After �rst supernova (with
ki k v
ki k
= 50 kms
�1
):
M
0
A
= 1:337M
�
, M
0
2
= 11M
�
,
P
orb
= 8:8 yr, e = 0:82,
�v
A
sys
= 13 kms
�1
High-mass X-ray binary phase
leading to unstable mass
transfer and a
ommon-envelope and
spiral-in phase and leaving
M
0
A
= 1:337M
�
,
M
B
He
= 2:4M
�
, P
orb
= 2:8 hr
Helium star mass transfer
phase (+ spin-up of neutron
star) leaving M
A
= 1:338M
�
,
M
He
= 1:559M
�
, P
orb
= 2:6 hr
Immediately after se ond
supernova: M
A
= 1:338M
�
,
M
B
= 1:249M
�
, P
orb
= 3:3 hr,
e = 0:12, �v
B
sys
= 35 kms
�1
`Standard' Channel
X X
XX
v∆
v∆
X
X
X
X
HeNS
Double-Core Channel
v∆
v∆
X
X
X
X
He CO
Initial binary: M
1
= 11:5M
�
,
M
2
= 11M
�
, P
orb
= 3:1 yr
Unstable Case C mass
transfer: se ondary expands
to �ll its Ro he lobe
Double- ore ommon-envelope
and spiral-in phase leaving a
CO star with M
CO
= 3:0M
�
and a He star with
M
He
= 2:4M
�
, P
orb
= 3:8 hr
After �rst supernova (with
ki k v
ki k
= 300 kms
�1
):
M
0
A
= 1:337M
�
,
M
0
He
= 2:4M
�
, P
orb
= 3:3 hr,
e = 0:33, �v
A
sys
= 230 kms
�1
Helium star mass transfer
phase (+ spin-up of neutron
star) leaving M
A
= 1:338M
�
,
M
He
= 1:559M
�
, P
orb
= 2:6 hr
Immediately after se ond
supernova: M
A
= 1:338M
�
,
M
B
= 1:249M
�
, P
orb
= 3:3 hr,
e = 0:12, �v
B
sys
= 35 kms
�1
Justham et al. (2010)
Importance of Late MassTransfer (Case C)
⊲ core evolution fixed → BH/NS
formation
⊲ late spin-up, no spin-down →
GRB progenitors
but: narrow range of predicted periods
(∼ 1%)
• at low metallicity ∼ 10%