Добавил:
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:

Supersymmetry. Theory, Experiment, and Cosmology

.pdf
Скачиваний:
82
Добавлен:
01.05.2014
Размер:
13 Мб
Скачать
☆

Friedmann–Robertson–Walker Universes 459

A reference energy density at present time t0 is obtained from the Friedmann equation for vanishing cosmological (λ = 0) and flat space (k = 0):

 

 

3H2

 

kg m−3.

 

ρ

c ≡

0

= 1.9 10−26 h2

(D.31)

8πGN

 

0

 

 

This corresponds to approximately one galaxy per Mpc3 or 5 protons per m3. In fundamental units where h¯ = c = 1, this is of the order of 10−3eV 4. In the case of a vanishing cosmological constant, it follows from (D.27) that, depending on whether the present energy density of the Universe ρ0 is larger, equal or smaller than ρc, the present Universe is spatially open (k > 0), flat (k = 0) or closed (k < 0). Hence the name critical density for ρc.

It has become customary to normalize the di erent forms of energy density in the present Universe in terms of this critical density. Separating the energy density ρM 0 presently stored in nonrelativistic matter (baryons, neutrinos, dark matter, etc.) from the density ρR0 presently stored in radiation (photons, relativistic neutrino, if any), one defines:

ΩM ≡

ρM 0

 

ΩR ≡

ρR0

 

ΩΛ ≡

λ

 

Ωk ≡ −

k

(D.32)

 

,

 

,

 

,

 

.

ρc

ρc

3H02

a02H02

The last term comes from the spatial curvature and is not strictly speaking a contribution to the energy density.

Then the Friedmann equation taken at time t0 simply reads

ΩM + ΩR + ΩΛ + Ωk = 1.

(D.33)

Since matter dominates over radiation in the present Universe, we may neglect ΩR in the preceding equation. Using the dependence of the di erent components with the scale factor a(t), one may then rewrite the Friedmann equation at any time as:

 

3

4

 

a0

2

 

H2(t) = H02

ΩΛ + ΩM

a0

+ ΩR

a0

+ Ωk

,

(D.34)

a(t)

 

a(t)

 

a(t)

where a0 is the present value of the cosmic scale factor and all time dependences have been written explicitly. We note that, even if ΩR is negligible in (D.33), this is not so in the early Universe because the radiation term increases faster than the matter term in (D.34) as one gets back in time (i.e. as a(t) decreases). If we add an extra component X with equation of state pX = wX ρX , it contributes an extra term

ΩX (a0/a(t))3(1+wX) where ΩX = ρX /ρc.

Important information about the evolution of the Universe at a given time is whether its expansion is accelerating or decelerating. This is obtained by studying the

460 An introduction to cosmology

second derivative of the cosmic scale factor, which is easily extracted from (D.25) and (D.26):

a¨

= −

4πG

N

(3p + ρ) +

λ

(D.35)

 

 

 

.

a

3

 

3

The acceleration of our Universe is usually measured by the deceleration parameter q which is defined as:

aa¨

(D.36)

q ≡ − a˙ 2 .

Using (D.35) and separating again matter and radiation, we may write it at present time t0 as:

1

 

a¨

 

 

1

 

 

 

q0 = −

 

 

 

 

=

 

 

ΩM

+ ΩR − ΩΛ.

(D.37)

H02

a

t=t0

2

Once again, the radiation term ΩR can be neglected in this relation. We see that in order to have an acceleration of the expansion (q0 < 0), we need the cosmological constant to dominate over the other terms. We can also write the deceleration parameter (D.36) at a given time t as

 

1

 

1

 

3

 

a0

4

 

q(t) =

ΩM

 

a0

+ ΩR

− ΩΛ .

(D.38)

H2(t)

 

2

a(t)

 

a(t)

 

If we introduce an extra component X as above, it contributes a term ΩX (a0/ a(t))3(1+wX )(1 + 3wX )/2: only components with equation of state parameter wX <

−1/3 tend to accelerate the expansion of the Universe.

The measurement of the Hubble constant and of the deceleration parameter allows us to obtain the behavior of the cosmic scale factor in the last stages of the evolution of the Universe:

a(t) = a0

1 +

t − t0

−

q0

(t − t0)2

+

.

(D.39)

 

2 tH2 0

 

 

tH0

 

· · ·

 

D.2.2 Measure of distances

Measuring cosmological distances allows us to study the geometry of spacetime. Depending on the type of observation, one may define several distances.

In an expanding or contracting Universe, the light emitted by a distant source undergoes a frequency shift which gives a direct information on the time dependence of the cosmic scale factor a(t). To obtain the explicit relation, we consider a photon propagating in a fixed direction (θ and φ fixed). Its equation of motion is given as in special relativity by setting ds2 = 0 in (D.18):

c2dt2 = a2(t)

 

dr2

.

(D.40)

 

− kr2

1

 

 

Friedmann–Robertson–Walker Universes 461

Thus, if a photon (an electromagnetic wave) leaves at time t a galaxy located at distance r from us, it will reach us at time t0 such that

t0

cdt

= 0

r

 

dr

 

t

 

 

√

 

.

(D.41)

a(t)

 

1 − kr2

The electromagnetic wave is emitted with the same amplitude at a time t + T where the period T is related to the wavelength of the emitted wave λ by the relation λ = cT . It is thus received with the same amplitude at the time t0 + T0 given by

 

 

t0+T0

 

cdt

 

 

 

r

 

dr

 

 

 

 

 

t+T

 

 

 

= 0

 

√

 

 

 

,

 

(D.42)

a(t)

 

 

 

1 − kr2

the wavelength of the received wave being simply λ0 = cT0. Since T0, T

t0, t, we

obtain from comparing (D.41) and (D.42)

 

 

 

 

 

 

 

 

 

 

 

 

cT0

=

cT

,

 

i.e.

 

 

 

 

λ0

=

 

a0

.

(D.43)

 

a0

a(t)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

λ

a(t)

 

Defining the redshift parameter

 

z

as

the

fractional increase in

wavelength

z = (λ0 − λ)/λ, we have

 

 

 

1 + z =

 

a0

 

 

 

 

 

 

 

(D.44)

 

 

 

 

 

 

.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a(t)

 

 

 

 

 

 

 

One may thus replace time by redshift since time decreases monotonically as redshift increases. For example, the expression of the Hubble parameter in (D.34) can be turned into

Using

H2(z) = H02 ΩM (1 + z)3 + ΩR (1 + z)4 + Ωk(1 + z)2 + ΩΛ .

(D.45)

 

 

t0

 

 

cdt

 

 

a0 cda

a0 cda

z cdz

 

 

t

 

 

 

 

= a(t)

 

 

= a(t)

 

= 0

 

 

 

 

 

 

a(t)

aa˙

a2H

H(z)

 

we may extract from (D.34) and (D.41) the proper distance (D.20) at time t0:

 

a0 0r

√

dr

 

 

 

 

 

 

 

 

 

 

 

1 kr2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

−

 

 

 

 

 

 

 

 

 

 

 

 

 

sin

 

 

1 r

 

k =

+1

 

 

 

 

 

 

 

 

= a0 r

 

−

 

 

 

k =

0

 

 

 

 

 

 

 

(D.46)

 

sinh−1 r k =

1

 

 

 

 

 

 

 

 

t0

cdt

 

 

−

 

 

 

 

 

 

 

= a0 t

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a(t )

 

 

 

 

 

 

 

 

 

 

 

z

dz

= H0 0

 

[ΩM (1 + z)3 + ΩR (1 + z)4 + Ωk(1 + z)2 + ΩΛ]1/2

462 An introduction to cosmology

where H0 = cH0−1. We can also write the deceleration parameter in (D.38) as

q =

1

 

1

ΩM

(1 + z)3 + ΩR (1 + z)4 − ΩΛ .

(D.47)

H(z)2

 

2

This shows that the Universe starts accelerating at redshift values 1+z (2ΩΛ/ΩM )1/3 (neglecting ΩR ), that is typically redshifts of order 1.

If a photon source of luminosity L (energy per unit time) is placed at a distance r from the observer, then the energy flux φ (energy per unit time and unit area) received by the observer is given by

φ =

L

≡

L

(D.48)

 

 

.

4πa02r2(1 + z)2

4πdL2

The two powers of 1 + z account for the photon energy redshift and the time dilatation between emission and observation. The quantity dL ≡ a0r(1 + z) is called the luminosity distance.

If the source is at a redshift z of order one or smaller, then we can approximate the integral 0r dr/√1 − kr2 in (D.41) by simply r and this equation gives

t0 a0cdt

= a

a0 a0cda

a0

da

a0r t

 

 

 

 

H0 a

 

 

a(t)

aa˙

a [1 − q0H0(t − t0)]

where we have used the development (D.39) with tH0 = H0 /c = H0(t − t0) (a − a0)/a0 1 and a = a0/(1 + z), we obtain for z 1

a0r = H0 z 1 −

1

+ q

0

z + · · · .

 

2

 

(D.49)

H0−1. Using

(D.50)

This result can be obtained directly from (D.46) (see Exercise 3). Thus, the luminosity distance reads, for z 1,

d

 

=

z

1

−

1 + q0

z +

· · ·

(1 + z) =

 

z

1 +

1 − q0

z +

· · ·

. (D.51)

 

2

 

2

 

L

H0

 

 

 

 

H0

 

 

 

 

Hence measurement of deviations to the Hubble law (dL = H0 z) at moderate redshift allow us to measure the combination ΩM /2 − ΩΛ (see (D.37)).

Another distance is defined in cases where one measures the angular diameter δ of a source in the sky. If D is the diameter of the source, then D/δ would be the distance of the source in Euclidean geometry. In a Universe with a Robertson–Walker metric, it turns out to be a(t)r = a0r/(1 + z). This defines the angular diameter distance dA

dA = a(t)r =

dL

(D.52)

(1 + z)2 .

Friedmann–Robertson–Walker Universes 463

Several distance measurements tend to point towards an evolution of the present Universe dominated by the cosmological constant contribution3 and thus a late acceleration of its expansion.

The approach that has made the first strong case for such an hypothesis uses supernovae of type Ia as standard candles4. Two groups, the Supernova Cosmology Project [88] and the High-z Supernova Search [140] have found that distant supernovae appear to be fainter than expected in a flat matter-dominated Universe. If this is to have a cosmological origin, this means that, at fixed redshift, they are at larger distances than expected in such a context and thus that the Universe is accelerating its expansion.

More precisely, one uses the relation (D.48) between the flux φ received on Earth and the luminosity L of the supernova. Traditionally, flux and luminosity are expressed on a log scale as apparent magnitude mB and absolute magnitude M (magnitude is −2.5 log10 luminosity + constant). The relation then reads

mB = 5 log(H0dL) + M − 5 log H0 + 25.

(D.53)

The last terms are z-independent, if one assumes that supernovae of type Ia are standard candles; they are then measured by using low z supernovae. The first term, which involves the luminosity distance dL, varies logarithmically with z up to corrections which depend on the geometry, more precisely on q0 = ΩM /2 − ΩΛ for small z as can be seen from (D.51). This allows us to compare with data cosmological models with di erent components participating to the energy budget, as can be seen from Fig. D.1.

This can be turned into a limit in the ΩM − ΩΛ plane for the model considered here (see Fig. D.2).

Of course, such type of measurement is sensitive to many possible systematic e ects (extra corrections besides the phenomenological stretch factor applied to the lightcurves, presence of dust, etc.), and this has fueled a healthy debate on the significance of supernova data, as well as a thorough study of possible systematic e ects by the observational groups concerned.

Let us note that the combination ΩM /2 − ΩΛ is ‘orthogonal’ to the combination ΩM + ΩΛ measured in CMB experiments (see Section D.3.4). The two measurements are therefore complementary: this is sometimes referred to as ‘cosmic complementarity’.

Other results come from gravitational lensing. The deviation of light rays by an accumulation of matter along the line of sight depends on the distance to the source (D.50), and thus on the cosmological parameters ΩM and ΩΛ. As q0 decreases (i.e. as the Universe accelerates), there is more volume and thus more lenses between the observer and the object at redshift z. Several methods are used: abundance of multiplyimaged quasar sources [255], strong lensing by massive clusters of galaxies (providing multiple images or arcs) [356], and weak lensing [286].

3At least when analyzed in the framework of the model discussed in this section, i.e. including nonrelativistic matter, radiation and a cosmological constant. As discussed in Section 12.2.2 of Chapter 12, the cosmological constant may be replaced by a dynamical component.

4By calibrating them according to the timescale of their brightening and fading.

464 An introduction to cosmology

Type Ia Supernovae

 

 

26

 

 

 

 

 

 

 

 

 

 

24

 

 

 

Supernova

 

 

 

 

 

 

 

 

 

Cosmology

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fainter

 

22

 

 

 

Project

 

 

 

 

 

 

 

 

 

 

 

 

ΛCDM

 

20

25

 

 

 

 

 

 

 

 

Calan/Tololo

 

 

 

 

OCDM

 

 

 

Supernova Survey

0.2

0.4

0.6

1.0

SCDM

 

 

18

24

 

 

 

 

 

 

 

 

 

 

or TCDM

 

 

 

 

 

 

 

 

 

 

 

 

16

 

 

 

 

 

 

 

 

 

 

14

23

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Magnitude

0.01

0.02

0.04

0.1

 

 

 

 

 

 

22

 

Accelerating

 

 

 

 

 

 

 

 

 

universe

 

 

 

 

 

 

 

 

 

 

 

 

 

Decelerating

 

 

 

 

21

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

universe

 

 

 

 

20

 

0.2

 

0.4

 

0.6

1.0

 

 

 

 

 

 

 

 

 

 

 

 

 

Redshift

 

 

Fig. D.1 Hubble plot (magnitude versus redshift) for Type Ia supernovae observed at low redshift by the Calan–Tololo Supernova Survey and at moderate redshift by the Supernova Cosmology Project.

D.2.3 Age of the Universe

Since 1 + z = a0/a(t), we can write (D.45) as

 

1

 

dz

2

H2(z) =

= H02 ΩM (1 + z)3 + ΩR (1 + z)4 + Ωk(1 + z)2 + ΩΛ .

 

 

(1 + z)2

dt

(D.54) This is easily integrated to obtain the time–redshift relation. Sending z to infinity (or, for that matter, to the value of z corresponding to nucleosynthesis, up to which we think we understand the evolution of the Universe), we obtain the age of the Universe:

 

∞

dz

 

t0 = tH0

0

 

(D.55)

(1 + z) [ΩM (1 + z)3 + ΩR (1 + z)4 + Ωk(1 + z)2 + ΩΛ]1/2

(we have set t = 0 at the Big Bang singularity; hence t0 is the age of the Universe). One may neglect the radiation dominated era since it is short on the scale of the age of the Universe (see Table D.1 below). One can check, using (D.33) to express Ωk, that t0 increases with ΩΛ at fixed ΩM . Thus a nonvanishing cosmological constant helps to solve what was once known as the “age of Universe crisis”: the Universe should be older than the oldest globular clusters in the Milky way whose age is estimated at 10 Gyr. Present data on (ΩM ,ΩΛ) favors an age of the Universe around 15 Gyr.

The hot Big Bang scenario 465

Supernova Cosmology Project

Perlmutter et al. (1998)

3

 

 

 

 

 

 

 

 

 

 

No Big Bang

 

 

99%

 

 

 

 

 

 

 

 

95%

42 Supernovae

 

 

 

90%

 

 

 

 

 

 

 

 

 

 

2

68%

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1

 

 

 

 

 

 

 

 

 

Λ

 

 

 

 

 

 

 

 

 

Ω

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

er

 

 

 

 

 

 

 

 

rev

 

 

 

 

 

 

sfo

 

 

 

 

 

 

and

 

 

 

 

0

 

 

 

Exp

 

 

 

tually

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ven

 

 

 

 

 

 

es e

 

 

 

 

 

 

 

llaps

 

 

 

 

 

 

 

 

Reco

 

 

 

 

 

 

Flat

 

 

Closed

 

 

 

 

 

Λ = 0

 

 

 

 

 

 

–1

 

 

flat

 

 

 

 

 

Universe

 

 

 

 

 

 

 

open

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0

1

 

 

2

 

 

 

 

3

 

 

 

ΩΜ

 

 

 

 

 

Fig. D.2 Best-fit coincidence regions in the (ΩM , ΩΛ) plane, based on the analysis of 42 type Ia supernovae discovered by the Supernova Cosmology Project [88].

D.3 The hot Big Bang scenario

We summarize in Table D.1 the main stages of the evolution of the Universe which we will describe in this section. Before undertaking this task, let us recall a few facts about equilibrium distributions.

D.3.1 Equilibrium distributions

For a species in kinetic equilibrium, the phase space distribution functions are of the standard Fermi–Dirac or Bose–Einstein form5:

f (p) =

1

, E = "

 

,

(D.56)

p2 + m2

 

exp [E/(kT )] ± 1

where the + sign refers to Fermi–Dirac and the − sign to Bose–Einstein. One can easily infer the number density n, energy density ρ, and pressure p of the corresponding species:

n =

g

f (p) d3p,

 

 

 

(2πh¯)3

 

ρ =

g

 

E f (p) d3p,

(D.57)

 

(2πh¯)3

 

g

 

 

p2

 

p =

 

 

 

f (p) d3p,

 

(2πh¯)3

3E

 

5We disregard here any chemical potential.

466 An introduction to cosmology

 

t

kTγ (eV)

z

 

t0 15 Gyr

2.35 × 10−4

0

now

 

Gyr

10−3

4

formation of galaxies

trec

4 × 105 yr

0.26

1100

recombination

teq

4 × 104 yr

0.83

3500

matter-radiation equality

 

3 min

6 × 104

2 × 108

nucleosynthesis

 

1 s

106

3 × 109

e+e− annihilation

 

4 × 10−6 s

4 × 108

1012

QCD phase transition

< 4 × 10−6 s

> 109

 

baryogenesis

 

 

 

 

inflation

 

t = 0

 

∞

Big Bang

Table D.1 The di erent stages of the cosmological evolution in the standard scenario, given in terms of time t since the Big Bang singularity, the energy kT of the background photons, and the redshift z. The line following nucleosynthesis indicates the part of the evolution which has been tested through observation. The values (h0 = 0.7, ΩM = 0.3, ΩΛ = 0.7) are adopted to compute explicit values.

where g is the number of internal degrees of freedom (g = 1 for a neutrino or an antineutrino and g = 2 for a massless vector field such as the photon, for an electron or a positron). We will need the explicit form in the relativistic limit (kT m):

n =

3

(3)

 

 

kT

3

 

 

 

F

ζ

g

 

,

4

π2

h¯

ρ =

7

1

 

 

 

 

 

F

 

g aBB T 4,

(D.58)

8

2

p = ρ/3,

where the parenthesis (· · · ) indicates the extra factor to be taken into account in the

F

case of the Fermi–Dirac distribution and aBB is the blackbody constant (we restore here the powers of c):

 

π2k4

 

aBB ≡

15c3h¯3 = 7.56 × 10−16 J m−3 K−4 = 4.72 keV m−3 K−4.

(D.59)

For example we have at present time (T0 = 2.725 K) nγ0 = 411 cm−3.

In the nonrelativistic limit (kT m), n and ρ = mn are exponentially small:

n = g

mkT

3/2

 

exp [−m/(kT )] .

(D.60)

2πh¯2

bosons i
fermions i
bosons i
fermions i

The hot Big Bang scenario 467

We conclude that, as long as some species remain relativistic, the matter energy density

is dominated by this radiation and takes the form:

 

 

 

 

ρ

R

=

1 g

 

a

BB

T 4,

 

 

 

 

 

 

 

 

 

 

2

 

 

 

4

7

 

 

 

4

 

 

 

 

 

 

 

 

T

 

T

 

 

 

 

 

 

 

 

 

 

 

 

g =

 

gi

i

+

 

 

 

gi

i

,

(D.61)

T

8

T

where we have taken into account the possibility that the species i may have a thermal distribution at a temperature Ti di erent from the temperature T of the photons. If radiation dominates the energy density of the Universe, one obtains from (D.27)

H =

2π

 

π

1/2 (kT )2

(D.62)

 

 

 

 

g

 

.

3¯h

5

MP

It is possible to show, using the second law of thermodynamics (T dS = dE + pdV ),

that the entropy per unit volume is simply the quantity

 

s ≡

S

=

ρ + p

(D.63)

 

 

 

 

V

 

T

and that the entropy in a covolume sa3 remains constant. The entropy density is dominated by relativistic particles and reads

s =

2 g a

BB

T 3,

 

 

 

 

 

 

 

 

 

3 s

 

 

 

 

 

 

 

 

 

 

 

 

 

T

3

7

 

T

3

 

 

 

 

 

 

 

 

gs =

 

gi

i

+

 

 

 

gi

i

,

(D.64)

T

8

T

where the sum extends only to the species in thermal equilibrium. Note that, when all species have the temperature of the photons T , which is true through most of the history of the Universe, gs = g .

We deduce from the constancy of sa3 that gs (aT )3 remains constant. Hence the temperature T of the Universe behaves as a−1 whenever gs remains constant. This is so except when some species drop out of equilibrium. Indeed, a given species drops out of equilibrium when its interaction rate Γ drops below the expansion rate H. We will return to this question in Section D.3.3 and see for example that neutrinos decouple at temperatures below 1 MeV. Their temperature continues to decrease as a−1 and thus remains equal to T . However when kT drops below 2me, electrons annihilate against positrons with no possibility of being regenerated and the entropy of the electron– positron pairs is transferred to the photons. Since gs|γ,e± = 2 + 4 · 7/8 = 11/2 and gs|γ = 2, the temperature of the photons becomes multiplied by a factor (11/4)1/3. Since the neutrinos have already decoupled, they are not a ected by this entropy

release and their temperature remains untouched. Thus we have

 

 

T

=

11

1/3

 

 

 

 

1.40.

(D.65)

 

Tν

4

We can now compute the value of gs for temperatures much smaller than me: gs = 2 + (7/8)6(4/11) = 3.91. We deduce that, at the present time (T0 = 2.725 K), s0/k = 2890 cm−3.

468 An introduction to cosmology

D.3.2 The evolution of the universe

It follows from the Friedmann equation (D.27) that if the Universe is dominated by a component of equation of state p = wρ (as we have seen following (D.28) ρ(t) a(t)−3(1+w)), then the cosmic scale factor a(t) varies with time as t2/[3(1+w)]. We start at present time t0 with the energy budget discussed in Section D.2.2, say ΩM = 0.3, ΩΛ = 0.7, Ωk 0. Radiation consists of photons and relativistic neutrinos:

 

7

 

4

4/3

 

ρR(t) = ργ (t) 1 +

 

Nνrel(t) ,

(D.66)

 

 

8

11

where Nνrel(t) is the number of relativistic neutrinos at time t. We have Ωγ = ργ (t0)/ρc = 2.48 × 10−5 h−0 2 and the mass limits on neutrinos imply Nνrel(t0) ≤ 1. In any case,

ΩR ΩM .

For redshifts larger than 1, the vacuum energy is subdominant and the Universe is matter-dominated (a(t) t2/3). However radiation energy density increases more rapidly (as a(t)−4) than matter (a(t)−3) as one goes back in time (as a(t) decreases).

At time teq, there is equality. This corresponds to

 

 

 

1

=

a(teq)

=

1.68 Ωγ

=

4.17 × 10−5

,

(D.67)

 

1 + zeq

 

 

ΩM h02

 

 

a0

ΩM

 

 

where we have assumed three relativistic neutrinos at this time.

However, at trec > teq i.e. still in the matter-dominated epoch, electrons recombine with the protons to form atoms of hydrogen and, because hydrogen is neutral, this induces the decoupling of matter and photon: from then on (trec < t < t0), the Universe becomes transparent6. This is the important recombination stage. After decoupling the energy density ργ T 4 of the primordial photons is redshifted according to the law

 

T (t)

=

a0

= 1 + z.

(D.68)

 

a(t)

 

T0

 

 

One observes presently this cosmic microwave background (CMB) as a radiation with a blackbody spectrum at temperature T0 = 2.725 K or energy kT0 = 2.35 × 10−4 eV. The fact that the spectrum observed is very precisely a blackbody spectrum indicates that (D.68) applies even for periods earlier than recombination.

Since the binding energy of the ground state of atomic hydrogen is Eb = 13.6 eV, one may expect that the energy kTrec is of the same order. It is substantially smaller because of the smallness of the ratio of baryons to photons η = nb/nγ 5 × 10−10. Indeed, according to the Saha equation, the fraction x of ionized atoms is given by

npne

=

 

x2

4.05c3

 

me

 

3/2

(D.69)

 

 

 

η =

 

 

e−Eb/kT .

nH nγ

(1 − x)

π2

2πkT

Hence, because η 1, the ionized fraction x becomes negligible only for energies much

smaller than Eb. A careful treatment gives kTrec 0.26 eV. We return to CMB in a more detailed analysis in Section D.3.4 below.

6It is believed that hydrogen is later reionized by the photons produced by the first stars or quasars but the Universe is then su ciently dilute to prevent recoupling.