Supernovae and their remnants
Why study Supernovae?
•Supernovae are of prime importance for the chemical evolution of the Universe
• Most important sources of energy for the interstellar medium. Part in the form of cosmic rays, which have an energy density of 1 —2 eV cm-3, thus containing one third of the energy density of ISM.
•standard candles for determining cosmological distances
•2—3 per century in a spiral galaxy like ours
2
Why study supernova remnants?
•Supernova remnants contain information about: •the supernova explosion that caused them •the circumstellar medium surrounding the progenitor
•Supernovae studied at large distances: SNRs in local neighbourhood!
•Supernova remnants: clues about ISM enrichment by Supernovae
•Supernova remnant likely account for cosmic rays in the Galaxy up to 3x1015 eV, requiring 10% acceleration efficiency
•Supernova remnant physics is rich: non-equilibrium ionisation and temperatures, shocks, highly magnetized plasmas, particle acceleration
3
Dawn of the scientific revolution
Tycho Brahe (1546-1601)       Johannes Kepler (1571 -1630) SN 1572 SN 1604
4
Historical Supernovae
SN 185: oldest source on a supernova (Chinese record)
SN 1006: brightest historical supernova (mv~-9 mag) recorded in China, the Arab world and Switzerland
SN 1054: recorded in Asia
SN 1185: observed by the Chinese
SN 1572: China, Europe - Tycho Brahe
SN 1604: China, Europe - Johannes Kepler
After that
No supernova spotted in the Milky Way Observational surveys of supernovae: expected 2-3 supernovae per century in the Milky Way
One supernova exploding every second in the Universe!
. .  Ml"!; .
Youngest Galactic SNRs
•G1.9+0.3 (Green+ '08) •Age: ~100 yr (Carlton+ '11)
•Cassiopeia A •Age: -330 yr
Supernovae
SN 1987A: February23, 1987
8
From Novae to Supernovae
Walther Baade Frits Zwicky
In addition, the new problem of developing a more detailed picture of the
happenings in a super-nova now confronts us. With all reserve we ad-vance the view that a super-nova represents the transition of an ordinary star into a neutron star, consisting mainly of neutrons. Such a star .may possess a very small radius and an extremely high density.  As neutrons
can be packed much more closely than ordinary nuclei and electrons, the "gravitational packing" energy in a cold neutron star may become very large, and, under certain circumstances, may far exceed the ordinary nuclear packing fractions. A neutron star would therefore represent the most stable configuration of matter as such. The consequences of this hypothesis will be developed in another place, where also will be mentioned some observations that tend to support the idea of stellar bodies made up mainly of neutrons.__
Supernova types
•Late 1930ies early 1940ies (Minkowski 1941): two recognized classes
•Type I: no hydrogen in spectra, also occur in elliptical galaxies, linear light curve
•Type II: with hydrogen in spectra, only in spiral galaxies •Since 1980ies: different types I: Type la, lb, Ic Type la: also in ellipticals -> exploding C/O white dwarfs
10
Supernova classification
Old Classification
11
Core Collapse Supernovae
11
Progenitors of core collapse SNe (II, Ibc)
Helium burning (T = 0.2x109K)
3 x4He -► 12C + y   ("triple a reaction")
i2C + 4He -► 160 + y Carbon burning (T = 2 x109 K)
12C + 12C -► 20Ne + 4He Neon burning    (T = 2x109 K)
20Ne+     160+ 4He Oxygen burning (T = 3.6 x109 K)
160+ 160 -► 28Si + 4He (takes -2 weeks) Silicon burning (T = 5x109K)
28Si + 4He -► 32S (takes -1 day!)
32S + 4He -► 32Ar ,etc. Important product 56Ni (-► 56Fe)
Core
i 1 i lí^ljj i l—r
-I-1-1-1-r
I    1    I    'I    1    I    1 I
Nuclear binding energy as a function of atomic number
i.i.i
i i i
i . i
20    40    60    80    100   120   140   160   180   200   220   240 260
12
Neutrino detection
n + e+ p + e-
P +ve n + ve
HO
V0'
2 * I 34 I
Kamiokande detector
.00
I.C
oeo
u
■H60
I«0 z
20
s
I
10
20. TIME (S«C)
■! 12x10* _
! g : .-o
i m •J0.6»K)*g
-20.0 -KJ.0 0.0 10.0 20.0 TME
(U.T.07:3S:3».2/23 87) (sec) FIG. 2 The ume sequence of oents in a 4S-sec interval centered on 07:35 35 UT. 23 February 1987. The vertical height «( each ine represents the relative energy of the event Solid lines represent low-energy electron events in units of the number of hit PMTv Vim (left-hand wale) Dashed lines represent muoci event* in units of the number of nhixoelcctrons (right-hand scale) tvents ul-i>4 arc ii ii.hi e>ents which precede the electron burst at time zero l"he upper right figure a the 0-2-sec time interval on an ci->anded scale
Kamiokande detected 12 neutrinos on Februari 23 1987 Confirmation of neutron star formation theory!
13
Neutron stars versus black holes
Heger+ 2003
Simplistic view:
< 25 Msun -► produce neutron star >25 Msun    produce black hole
14
Thermonuclear explosions (Type la)
• C/O white dwarf close to 1.4 MSUn
• Mass accretion: density/pressure increases
• Explosive fusion of C/O
• Different explosion models (deflagration, delayed detonation...)
• Problems single degenerate scenario: -only small range for stable accretion
-no donor stars found in Type la SNRs (Schaefer+ '12, Kerzendorf '14) -no population of accreting WDs in elliptical found (Gilfanov+Bogdan, '10)
• Popular alternative: double degenerate scenario (merging white dwarfs)
• allows for super-Chandrasekhar supernovae 15
Type la vs Core Collapse SNRs
s 1 2
•I—I
0.01 r
0.001
■ 1	i ■	1 i	1	i	i	i	i	i  1 i	i  1  i ■
-									/ \
				*'-					
				/ \					■: ■
; •				/					A :
					> \				
						\V"			
						\ •-•	""A	.................ii	
r •>						■.....	w—	.....• /	
:		\ *	•				.....■.		■ -
i	i		1	i	i	i	1		i  , i
6	8	10	12	14	16	18	20	22 24	26 28
8
Element Nr
Core collapse (black): rich in alpha elements (O, Ne, Mg,..)
Type la: Fe-group (coloured, different models)
6 -
©
2 4
—I
0
-I-1-1-1-1-1-1-
Oxygen yields
t-1-1-r
10
20 30 40
Main Sequence Mass (M0)
above 30Msun models predict different O yields
above 30Msun stellar cores may collapse into black holes and the amount of fallback is uncertain
16
From supernova to supernova remnant
Supernova lightcurves
Nickel decay
/_=3x108—6x109 Lsun
Cobalt decay
50      100      150      200      250      300      350 400
Day* after maximum light
freely expanding ejecta heated by 44Ti
shock-heated CSM ring
•56|\ii + e- -> 56Co* + ve (t=6.1 days) •56Co + e- —> 56Fe + ve (t=77.1 days) •SN expands -> ejecta cools (dust forms!!)
• Ejecta may still be warmed by late time radio-active heating (44Ti)
• Depending on the circumstellar density
- outer shock wave heats up a shell that may give rise to X-ray emission
- shock wave may accelerate particles -> relativistic electrons -> radio emission
ummary
•Supernovae come in two basic types 1 .thermonuclear supernovae
• exploding C/O white dwarfs
• ejecta mass ~ 1.4 Msun
• energy comes from nuclear fusion (C/0->56Ni)
• produce lots of Fe (-0.6 Msun, product of 56Ni)
• no stellar remnant remains 2.core collapse SNe
• imploding stellar cores (Fe-grp elements)
• stellar core becomes a neutron star
• energy source: gravitational energy
• most energy is neutrinos!
• nucleosynthesis yield: stellar fusion + explosive fusion
• yield dominated by oxygen
II
Supernova Remnants: Structure & Evolution
19
SNR evolutionary phases (simplified)
Four phases are recognised
1. Ejecta dominated phase (a.k.a. free expansion phase) •First 10-few 100 yr •Vs>3000 km/s
•Mej > Mswept
2. Adiabatic or Sedov-Taylor phase •Few 100 yr to few 5000 yr •200 km/s < Vs <3000 km/s
•Mej « Mswept
•Radiative losses unimportant
3. Snow plough phase
• 5000-50000 yr
• 20 km/s < Vs <200 km/s
• Radiative losses, momentum conserved
4. Disappearance phase
• Vs comparable to turbulent motions ISM • Different parts of SNR may be in different phases!!
20
Ejecta dominated phase
• Esn ~ 112 Me\ Vef ~ 1051 erg
• Fast moving ejecta vej~ 104 km/s (EsN/Msoi)1/3(pisivi/10-24g cm-3)-173 •shock-heating and sweeping up of the circumstellar medium
• heats the outer (cold) ejecta layers by reverse shock
• inner ejecta are cold
Adiabatic (non-radiative) Phase
Once Mswept > Mejecta, but Vs> 300 km/s a SNR is said to be in the adiabatic phase •(Almost) all energy is contained in the shock-heated plasma •Evolution is usually described by so-called Sedov self-similar solution:
i?s —
Po
dt
2(^)^-/5 2 5 V po / 5
material further out decelerates first, material further in starts to run into outer shells heating it up
22
Complications
•The environments of SNRs can be complex •Some of that due to progenitor:
- wind bubbles from various stellar phases
- complex surroundings: molecular clouds, stellar winds from other stars, etc.
• Possible presence of pulsar wind nebula
23
"Mature SNRs": snow-plough phase
• For T<5x105 K: cooling becomes very strong (oxygen line emission) •Cooled gas: bright optical emission from [Olll], [Nil],...
•This corresponds with Vs<200 km/s
• SNR no longer adiabatic. R~t°25 (momentum conservation)
24
SNR Types: shell-type
Cygnus Loop (ROSAT)
SNR Types: Plerion
Crab nebula .
3C58 (Chandra)
• Plerions are dominated by synchrotron emission from pulsar wind nebula
•They can still be considered SNRs as they have some ejecta components (in Crab nebula only seen in the optical)
26
SNR Types: Composite SNRs
Composite are a combination of a shell-type SNR and a pulsar wind nebula Not all core collapse SNRs are plerions/composites: neutron stars not always powerful pulsars!
27
Mixed-morphology/Thermal composite SNRs
• Mixed-morphology SNRs are shell-like in radio, and centrally bright in X-rays •X-ray emission is thermal
• Evidence for enhanced abundances •Are older SNRs
• Idea: shell too cool for X-rays, but center hot enough for X-rays (Cox+ '99)
• Many of the gamma-ray emitting mature SNRs are MM!!
28
Confirmation of SNR typing: light echoes
♦
'\ ISM dust cloud
Light-echo path t+At
|  0' ^
Direct light reaches us in time t.
Confirmation of SNR typing: light echoes
30
Ill
Emission from core collapse SNRs
31
Which elements, where?
wavelength (A)
50 30 20 10 5 2 1 Eb(eV) —1-1-1-1 1 1 1 1—i—'-'-1-r - o
X-ray lines between neutral fluorescent n—2-1, and H-like n= I -oo (think of the Bohr model!)
Linking SNRs with SN classes
XMM-M0S12
10 1
1 >	0.1
	
	
i W	0.01
W	0.001
o	
w	10
CD	
-*—' CO	1
u	
-~>	
C	
	0.1
o	
U	
	0.01
	0.001
1E0102.2-7219
'iii
XMM-RGS_ oe,
1 1 1 j cO
0519-69.0
Energy (keV)
10   0.6    0.8      1 1.2 Energy (keV)
n
o c 9
0) 0)
o
DO DO
Core collapse SNRs are rich in O, Ne, Mg Type la SNRs are iron-rich
e.g. Hughes '95, Flanagan+ '04, Kosenko+ '10
Cassiopeia A: (A)symmetries I
XMM-Newton based Doppler map + deprojection Cas A shows donut like shape
Willingale+ '02
see also Delaney et al '10
Evidence for fast Fe knots
Mg Fast moving Fe
Si (in Doppler space)
Fast moving Fe (projected, > 7000 km/s)
4
Fast moving "pure" Fe plasma (Hughes+ TO.Laming&Hwang 03)
Seen (to a lesser extent) in simulations (in Type Ib/c)
44Ti decay
EC (99.3%)
0*
ß+ decay(94%) Electron Capt (4.95%)
1	^78.4 kcVj
T	^ 67.9 keV
J (85,
10'4 r
4yr)
EC (0.7%)
T
>
CO
E
o
'</> c
o 10 Ü
1157.0 keV
(f
IO"7
i lili
Renaud, JV, et al. '06 (INTEGRAL-ISGRI)
■ *       ■      ■     i i
44
Ca
100
Energy (keV)
•44Ti exclusive explosive nucleosynthesis
• Decay time -86 yr
•Yield: sensitive to mass-cut, expansion, and asymmetries!
• Detected in Cas A
(lyudin+ 94, Vink+ '01, Renaud+ 06, Grefenstette+ 14, Siegert+ '15)
•Yield: (1.1-1.6)x104 MSUn
.0001
8x10"5
^-6x10~5 o
y 4x10"5 2x10"5 h 0
DiehlS Timmes 98
_ I  I  I  I  I  I  I  I  I  I I I II -   Type lb Supernova	i i i J i _
— /	M J li i *
"i #T i i i i  I i i i i I i	- 1 o i i i 1 i i i 1
5            10 15	20
Helium Core Mass (M0)	
37
44Ti map NuStar
Grefenstette+ 2014
•44Ti in blue
•Si/Mg ratio green
•Most 44Ti in unshocked interior
•Lines redshifted by 1000 km/s
44Ti in SN1987A
Boggs et al. 14
Energy (keV)
•Redshifted 44Ti in SN1987A -► asymmetric core •44Ti yield: 1.5 x 1CH Msun •Very similar to Cas A
39
IV
Emission from Type la SNRs
40
Type la supernova remnants
0509-675
0519-690
N103B
Dem L71
0534-699
Type la SNR in LMC
•Red:oxygen •Green: Fe-L • Blue: Mg/Si/S
Iron in center
•With age more Fe gets shocked by reverse shock
Light echo spectra exists for 0509
41
Type la vs Core Collapse SNe
Apart from obvious differences in composition:
- Type la are more regular stratified
- Have a more regular morphology ,0
Lopez et al. 2009/11
CD
E E
c/)
0
1
CD
E o
E
TD
C\l
<- 10'
10
10
CC SNe     G15.9+0.2 I ~i^49B
CasA
W49B
Kes 73 G292.0+1.8 •
Gl 1.2-0.3 -
B0453-685
0548-70.4
0509-67.5
Kepler •
RCW103 +
N132D
0519-69.0
Tycho
DEM L71
Type la
N103B
10
10 -7
P3/PQ(x10 7)
10
3rd moment (asphericity)
42
Non-X-ray emission
Type la: SNR0509-695 CC SNR: Cas A
Hubble Ha + Chandra Hubble Ha
In general:
•Type la have Ha associated with shock -► partial neutral medium •CC have not -► shock through fully ionised material Type la + partial neutral medium:
=>rules out supersoft sources as progenitor (progenitor not a source of
strong UV emission)!!
43
LMC SNR distribution
Likely Type la SNRs: blue
All other: cyan (may include Type la)
Likely core collapse SNRs: associated with HII regions
44
Cosmic Rays
Cosmic Ray Spectra of Various Experiments
*&-. 104 10
o o
>
(3
to 10*
* A
£ 10 4
10"7
10
10
10
13
10
•16
■19
10
io-22
10
25
10
-28
.......
i^.:..............:.............-...............:-"5.........
%     I (1 particle/m -sec)
LEAP - satellite Proton - satellite Yakustk - ground array <$>       Haverah Park - ground array O       Akeno * ground array a       AGASA - ground array
□ Fly's Eye • air fluorescence
^       HiResI mono - air fluorescence 0       HiRes2 mono - air fluorescence HiRes Stereo • air fluorescence
□ Auger ■ hybrid
109 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020
Energy (eV)
• Up to ~3x1018eV of Galactic origin
•Galactic CRs: likely powered by supernovae (Baade & Zwicky), as they provide sufficient power
•The "Knee" (1015eV): must be linked to a common property among Galactic accelerators
•Are particles mainly accelerated in supernova remnant phase?
•Alternatives:
-in SN phase, or <50 yr
-collective effects in superbubbles (Bykov, Parizot)
45
Why supernova (remnants) as sources?
• In normal spiral galaxies as the Milky Way: SNe most energetic sources
- Cosmic rays remain in Galaxy for tCr~107 yr
- Steady state/homogeneity requires Wir« 107 yr
- SNe rate is 2-3 per century
- SN explosion energy Ekin=1051 erg •SNe fulfil cosmic-ray energy requirements
- Energy density CRs uCr~1 eV/cm3
- Volume Galaxy: Vgai=n Rdisk2(2z) ~ 3x1011 pc3 ~1067 cm3
- Power needed: L=UcrVgai/tcr=5x1040 erg/s
- SN power: LSN=1051/tsN= 6x1041 erg/s
SNe provide enough power for cosmic rays if efficiency is -10% SNRs should be able to accelerate protons beyond 1015eV
46
Supernova remnant shock physics
• Atmospheric shocks: heating in shock due to particle-particle coll
• In astrophysical plasmas: density (n) is very low
• Mean free path= nov can be very long for particles
• Estimate of cross sections, two particle mi and nri2, charge Zi,Z2
• Impact parameter = b
• Relevant b: kinetic energy=potential energy
2 mi + 7712
1    mi 7722
ZxZ2e2 b
• For v^1000 km/s, n=1 cm3one finds for proton-proton
p
1020n * cm
• This is larger than the size of most supernova remnants!!
• Hence: shocks must be collisionless:
• Heating due to electric/magnetic fields & wavesU
Supernova remnant shocks:
the Rankine-Hugionot relations
■ shock
Rankine-Hugoniot relations: • mass-, momentum- & enthalpy-flux conservation
•do not depend on collisionlessness of shock!
P1V1 = P2V2
	region 1 J	region 2
	Vl=Vsh	V2=Vsh/X
		
		-►
Pi + P1V1 = P2 + P2V2
X =
El pi
El
V2
shock compression ratio
1 1
(Pi + Ui + -piv\)vi = (P2 +U2 + -p2vl)v2
Solutions for strong shocks:
X =
(7g + (7g - l)Ml + 2
4 for M{ =
2 = 1 Pi^i 7, Pi
■> 00
48
Diffusive shock acceleration theory
Particles scatter elastically (B-field turbulence)
Each shock crossing the particle increases its energy by a factor (3=1 +v/c After j crossings, a particle with energy Eowill have an energy E=Eo/3i
Resulting spectrum (e.g. Bell 1978):
dN/dE = C E-0+3/(x-D)
X shock compression ratio,
X=4    dN/dE = C E-2
Axford et al. , Blanford & Ostriker, Krymsky, and Bell (all 1977-78)
50
Diffusive shock acceleration theory
Smaller mean free path, smaller D, faster
acceleration
Böhm diffusion:
^mfp ^"gyro
1 cE
Acceleration of a 100 keV proton to 1015 keV by a Vs=5000 km s-1 shock requires 1400 crossings
Protons and nuclei can be accelerated to higher energies than the lighter electrons, because the loose less energy to synchrotron radiation
51
(X-ray) synchrotron emission from SNRs
52
First evidence for particle acceleration
•Since the 1950-ies SNRs associated with bright radio synchrotron sources •Synchrotron emission: relativistic electrons deflected in magnetic fields •Characteristic frequency
uch = 46-±- —— MHz
10 fiG VIGeV/
•Radio synchrotron    electrons with GeV energies •Brightness: number of rel. electrons + B-field •For power law electron distribution
Ne oc KE~q, Iv oc KB(q+V!2v-(q-V!2
•Relation electron and radio spectral index:
a = (q-l)/2 •Typical young SNRs in Radio: a=0.6   q= 2.2
X-ray synchrotron emission
• In 1995 ASCA showed that the X-ray emission from SN 1006 was a combination of thermal X-ray and synchrotron radiation (Koyama et al. 1995)
•X-ray synchrotron emission implies presence of 10-100 TeV electrons!!
hvch = 13.9(-^— - keV
c VlOO/xG/ V100 TeV /
54
X-ray synchrotron from young SNRs
Implications
• For B=10-100 mG: presence of 1013-1014 eV electrons
• Loss times are:
r    E  12 5 r E rv b-b r'-.T
56
X-ray synchrotron emission tells us that
- electrons can be accelerated fast
- that acceleration is still ongoing (loss times -10-100 yr)
- that particles can be accelerated at least up to 1014 eV
Young SNRs detect in TeV gamma-rays
57
Gamma-ray emission processes
-►   n /u+ -> e+ + ve +
Neutral pion production/decay    ** rr _   e- + -e + ^
if pion decay, then 1049 erg in high energy protons (compared to 1051 erg in the kinetic energy of the expansion of the remnant or 1053 erg of binding energy released in the formation of a neutron star)
Summary
• Supernovae come in two very different types:
• Core collapse SNe -► leave neutron star, oxygen rich •Thermonuclear/Type la -► exploding CO white dwarf
• Model uncertain: single- or double degenerate?
• single degenerate model has problems
• Supernova remnants phases:
• 1. Ejecta dominated;
• 2. Adiabatic phase;
• 3. Momentum conservation phase;
• 4. Disappearance phase
• Simple model may not be sufficient: effect of stellar wind bubbles
• Shock heating process:
• Shocks are collisionless
•Thermal X-ray emission from plasmas with kT>~106 K (0.1 keV)
Summary (continued)
• Type la can be distinguished from core-collapse SNRs:
• Type la: iron-rich; CC: oxygen-rich
• Type la: Ha emission
• Type la: layered structure; CC: chaotic
• Type la: symmetric morphology
• For core collapse SNRs: evidence for asymmetric explosions
• Jets in Cas A
• Donut like shapes
•Cosmic ray spectrum near power law over 12 orders of magnitude in energy
- most like origin in supernovae (other sources not energetic enough)
- requires SNRs to accelerate up to 3x1015eV (proton)
- requires that SNR put -10% of kinetic energy (1050erg) in cosmic rays - diffusive shock acceleration (1st order Fermi)
• Radio synchrotron: oldest evidence for particle (electron) acceleration in SNRs •X-ray synchrotron: identified since 1995 (SN1006 byASCA)
•X-ray synchrotron: -electron acceleration up to 10-100 TeV
-location @ shock + fast loss times -> shocks are responsible for acceleration •Cherenkov Telescopes -► detection of TeV gamma-rays from SNRs