the$radiaon,dominated$era - caltech astronomygeorge/ay127/hirata_lec3_radera.pdf · =1.95(1+z)$k.$...
TRANSCRIPT
The Radia)on-‐Dominated Era
Ay 127 April 9, 2013
1
Outline
1. Radia)on-‐Dominated Plasmas, Expansion of the Universe
2. Neutrino Decoupling 3. Synthesis of the Light Elements
4. Observa)ons of D, He, Li
2
Radia)on-‐Dominated Plasmas
• Consider blackbody spectrum:
g = number of polariza)ons (2 for photons) − = bosons; + = fermions
• Energy density
g* = g (bosons) or ⅞ g (fermions)
€
Iν =ghν 3
c 21
ehν / kT ±1
€
u =4πc
Iν dν =π 2
303c 3g*(kT)
40
∞
∫
3
Thermodynamic proper)es
• Effec)ve ma[er density:
• Pressure:
• Entropy density:
€
p = 13 u =
π 2
903c 3g*(kT)
4
€
ρ =uc 2
=π 2
303c 5g*(kT)
4
€
s =duT0
u∫ =
2π 2
453c 3g*k
4T 3
4
Expansion of Universe
• Friedmann equa)on rela)ng H to T:
• Solu)on: H ~ T2 ~ a-‐2 a ~ t1/2 H=1/(2t).
• More convenient form:
€
H 2 =83πGρ =
4π 2G453c 5
g*(kT)4
€
t =3 53 / 2c 5 / 2
4πG1/ 2g*1/ 2(kT)2
€
t =1g*
1.56 MeVkT
⎛
⎝ ⎜
⎞
⎠ ⎟
2
sec5
What is g*?
• Consider the Universe at t~1 second. • Photons only: g*=2. • Also neutrinos: 3 flavors, ×2 for an)neutrinos: ⅞ × 3 × 2 = 21/4.
• At 3kT≥mec2, e+e− are “massless.” 2 spins each: ⅞ × 2 × 2 = 7/2.
• At MeV temperatures, g* = 43/4. 6
Evolu)on (slightly idealized)
7
g*
kT 10 MeV 100 MeV 1 GeV 10 GeV 100 GeV
t
γ, ν, e+e−
10¾ 10
100
+μ, π 17¼
−π+udsg 61¾
+c, τ 75¾
+b 86¼
+WZHt 106¾
??? + MSSM 228¾
quark-‐gluon plasma hadrons
1 ms 1 μs 1 ns
Neutrino Decoupling
• Neutrinos have weak interac)ons with e+e−:
• Typical E ~ 3kT so can find the number of neutrino interac)ons in the life)me of the Universe:
• At kT~1.6 MeV, universe transi)ons from being neutrino-‐opaque to transparent.
8
€
ne+e- ~ 5 ×1031(kT)MeV3 cm−3
€
σweak ~ 10−44 EMeV2 cm2
€
nσct ~ 0.1(kT)MeV5
Epoch of e+e− Annihila)on
• At kT<mec2 almost all the electrons and positrons annihilate. (~1ppb more e− than e+. These few e− survive.)
• Energy of annihila)on heats the photons but not the neutrinos, so Tγ>Tν.
• Can do computa)on assuming annihila)on is adiaba)c (conserves entropy of e+e−γ).
9
€
e+ + e− →≥ 2γ ~ 100%ν e + ν e (almost) neglgible⎧ ⎨ ⎩
Annihila)on Part 2
• Entropy before annihila)on:
• Entropy ayer annihila)on:
• Equa)ng gives:
(Remains true since both temperatures scale as 1/a.) 10
€
s(e±γ) =2π 2
453c 3g*(e
±γ)k 4Tν3 =
11π 2
453c 3k 4Tν
3
€
s(γ) =2π 2
453c 3g*(γ)k
4Tγ3 =
4π 2
453c 3k 4Tγ
3
€
Tγ = 114
3 Tν
Post-‐annihila)on
• Can combine neutrino and photon energy densi)es to get radia)on energy density:
• Factor of 1.68 from neutrinos. Affects BBN, CMB, etc.
• Equivalent to “effec)ve” g*=3.36. • From temperature ra)o, Tν=1.95(1+z) K.
• They’re s)ll here today: ~ 300/cm3. 11
€
u =π 2
303c 32(kTγ )
4 + 214 (kTν )
4[ ] =1.68π 2(kTγ )
4
153c 3
Nucleosynthesis
• Universe has net baryon number:
• We don’t know why. • At high temperature the thermodynamically favored state of baryons is p + n.
• At low temperature the favored state is 56Fe. • Nucleosynthesis is process of forming the heavier nuclei from p+n. Started in Big Bang, con)nues today in stars.
12
€
η =baryons− antibaryons
photons~ 6 ×10−10
Primordial nucleosynthesis
• The process: 1. n/p ra)o determined at neutrino decoupling 2. Universe expands/cools, some neutrons decay 3. Assembly of D, 3He, 4He, 7Li
• Nuclear processing is not finished ayer BBN, e.g. heavy elements are produced in stars. This more recent processing must be corrected/avoided to infer primordial abundances.
13
The ini)al n/p ra)o
• Before neutrino decoupling, neutrons are kept in equilibrium:
• Equilibrium ra)o
where Q = (mn-‐mp)c2 = 1.293 MeV. • n:p=1:1 at high T, then starts decreasing.
14
€
n + e+ ↔ p+ + ν en + ν e ↔ p+ + e−
€
n : p = e−Q / kT
The ini)al n/p ra)o (Part II)
• Neutrino decoupling at kT ~ 1.6 MeV would imply n:p=0.4.
• More detailed calcula)on gives n:p≈0.15.
• Neutron has a half-‐life of ~11 minutes.
• Therefore the neutron abundance declines:
15
€
n : (n + p) ≈ 0.152t /11min€
n→ p+ + e− + ν e
Deuterium (Part I)
• Deuterium is the simplest nucleus. Binding energy is B = 2.22 MeV.
• Saha-‐like equa)on for its abundance:
• Compare with total baryon abundance:
16
€
n + p+ ↔D+ + γ
€
n(D+)npnn
=34
2π2
mredkT⎛
⎝ ⎜
⎞
⎠ ⎟
3 / 2
eB / kT
€
nb = 5.6 ×1020Ωbh2T9
3 cm−3
Deuterium (Part II)
• Using nota)on Xi=ni/nb, and assuming the universe is mostly p (85%) and n (15%) we get
• Deuterium abundance grows exponen)ally. Would become of order unity at T9~0.7 (t ~ 3 min) except that deuterium can burn to heavier nuclei.
17
€
X(D+) ≈ 5 ×10−14Ωbh2T9
3 / 2e25.8 /T9
Helium
• Deuterium consump)on:
• When deuterium reaches ~1%, these reac)ons package most of the available neutrons into the very )ghtly bound 4He.
18
€
D + p↔ 3He + γ
D + D↔ 3H + pD + D↔ 3He + n
A=23
€
3He + p↔ 3H + n3H + D↔ 4He + n3He + D↔ 4He + p
A=34
What happens next
• At t ~ 3 minutes, 13% of baryons are neutrons. Packaged into 4He, we get Y~0.26 (4He frac)on by mass). Most of the rest is 1H.
• No new D is produced once free neutrons are gone. Deuterium is burned via D+D and D+p. A small amount, XD~2×10−5, survives.
• Some 3H and 3He are ley over (total X3~10−5). 19
End Game
• Some A=7 nuclei (X7~4×10−10) are produced by 3He+4He and 3H+4He.
• 6Li produced in much smaller amounts since it is fragile (X6 ~ 10−14).
• Heavy element produc)on e.g. 3α12C negligible at BBN densi)es (10−5 g/cm3).
• Radioac)ve nuclei (n, 3H, 7Be) decay.
20
21
1H
n
2H 3H
3He 4He
7Be
7Li 6Li
(n,γ)
(p,γ)
(D,p)
(p,γ) (D,n) (n,p)
(D,p)
(D,n)
(3H,γ)
(3He,γ) (n,p)
Stable isotope
Radioac)ve isotope
Reac)on
Radioac)ve decay
(3H,n) (3He,p) (D, γ)
22
BBN Yield Predic)ons
Cyburt, Fields & Olive 2003 Phys Le> B 567,227
4He frac)on by mass
Abundance rela)ve to H by number
Dependence on Cosmology
• 4He: Determined by n:p ra)o, so very sensi)ve to physics at neutrino decoupling, e.g. new massless par)cles. Slightly sensi)ve to Ωbh2.
23
u(T) increased
H(T) increased
Reduced nσct
ν decoupling at higher T
Larger n:p raMo
More 4He
New massless parMcle
Dependence on Cosmology
• 2H and 3He: Leyovers of incomplete H burning. Decrease for larger baryon density Ωbh2.
• 7Li: Non-‐monotonic behavior in Ωbh2: – Low Ωbh2: direct produc)on of 7Li, destroyed by
7Li+p at high densi)es. – High Ωbh2: 7Be can be produced, decays to 7Li.
• 6Li: Primordial contribu)on should be undetectable.
24
4He
• Predic)on: Y = 0.2482±0.0003±0.0006 • Usually es)mate He abundance from H II region spectra (H I and He I recombina)on lines). – Determine temperature (self-‐consistent or using [O III]). – Model collisional excita)on, fluorescence. – Op)cal depth effects. He I 23S1 is metastable. – Model reddening. – Model stellar He I absorp)on. – Ioniza)on correc)on factors (H II and He I can coexist). – Extrapolate to zero metallicity to get primordial value.
25
Helium
• Predic)on: Y = 0.2482±0.0003±0.0006 • Usually es)mate He abundance from H II region spectra (He I recombina)on lines). Extrapolate to zero metallicity to get primordial value.
26
0.232 < Yp < 0.258 Olive & Skillman 2004 ApJ 617,29
✔
predicted
Examples of Deuterium Measurements
LocaMon D/H (ppm) Method Earth 150
Venus 16000±2000 Mass spec. Donahue et al 1982
Mars 900±400 IR lines (HDO/H2O) Owen et al 1988
Jupiter 22—50 50±20
IR lines (CH3D/CH4) Kunde et al 1982 Mass spec. Niemann et al 1996
Saturn 4—29 IR lines (CH3D/CH4) Cour)n et al 1984
Local ISM ~ 7 – 20 Absorp)on lines in stellar spectra. Depends on line of sight; see tabula)on by Linsky et al 2006
Lyman-‐α absorbers (IGM)
28±4 H vs D Lyman absorp)on lines in QSO spectra Kirkman et al 2003
27
Warnings on D/H
• Not all D/H is the primordial value, or even that at the forma)on of the solar system.
• Problems: – Astra)on: burning of D to 3He etc. in stars. – Chemical frac)ona)on: At low temperatures D binds more )ghtly to molecules than H due to vibra)onal zero point energy.
– Frac)ona)on due to depth of planetary gravity well. (This is why Jupiter was once used for BBN D/H.)
28
XH+ + D XD+ + H
Intergalac)c D/H
• Probably most reliable method is low-‐metallicity quasar absorp)on line systems. – Li[le processing by stars – Less opportunity to hide deuterium in molecules or dust grains
• Reduced mass of e−D+ is greater than e−p+ by 1 part in 3600, rescaling all hydrogenic energy levels. Equivalent to velocity shiy of c/3600 = 80 km/s.
• IGM value 28±4 ppm agrees with predicted 25.7(+1.7)(−1.3) ppm.
29
✔
30
Kirkman et al 2003 ApJS 149, 1 Q1243+3047 zabs = 2.53
3He • Difficul)es:
– Can be both produced and destroyed in stars. – No IGM measurement.
• Most(?) accepted measurement is 3He+ hyperfine line (λ=3.46cm) low-‐metallicity H II regions. Bania et al (2002) find 3He/H < 15 ppm. – Directly observed 11±2 ppm, argue that stars would lead to net increase in 3He.
• Predicted from WMAP baryon abundance: 3He/H = 10.5±0.3±0.3 ppm. – Aside: Jupiter atmosphere ra)o 3He:4He = (1.1±0.2)×10−4 [Niemann et al 1996]. 31
✔?
Lithium
• Can be measured in low-‐metallicity stars. – Two mul)plets available for Li I: 6708Å (2s—2p) and 6104Å (2p—3d).
– 7Li destroyed at high T, 7Li + p 24He. Avoid low-‐mass stars due to deep convec)ve zone.
– Slight isotope shiy of 6Li vs. 7Li. • Li can also be produced via cosmic rays: – Spalla)on of CNO elements (also gives Be, B) – 4He + 4He collisions at low energies.
32
Li abundance evolu)on
33 Asplund et al (2006) ApJ 644, 229
12 + log10 XLi
The Lithium Problem(s)
• 7Li: Predicted (5.2±0.7)×10−10. – See update from Cyburt, Fields, Olive 2008.
• Observa)ons give lower values – e.g. (1.1—1.5) ×10−10 from Asplund et al. 2006. – Other determina)ons are typically ~(1—2)×10−10.
• 6Li: there have been reports of a plateau at ~6×10−12, although the accuracy of 6Li/7Li extracted from line profiles has been debated.
34
✗
✗?
Solu)ons
• The 7Li problem: – Errors in nuclear reac)on rates? – Deple)on in convec)on zone? – Stellar atmosphere modeling? – Destroyed in early genera)on of stars? – New physics?
35