7. Internal structure.

II
 Internal Structure II.
1. Equations of solar evolution
2. Scaling laws
3. Nuclear reactions: p-p chain
4. Convective energy transport
5. Solar neutrino problem
Energy transfer and balance equations
The total energy flux, L = 4π r 2 F , integrated over a sphere of radius r :
16π acT 3 dT
L=− .
L+dL 3κρ dr

If ε is the energy release per unit mass then

ε the energy flux change in a shell dr is:
L
dL = ερ 4π r 2 dr
dL
= 4πρ r 2ε
dr
This is the equation for conservation
of energy (energy balance).
 Equations of the stellar structure
dm
= 4πρ r 2 (1) These equations describe the
dr structure of stellar radiative
dP Gm ρ zones. In the convection zone
=− 2 (2)
dr r Eq.(4) is replaced by an
dL equation of convective energy
= 4πρ r 2ε (3) transport, e.g. mixing length
dr theory.
dT κρ
=− 2 3
L ≡ −F (4)
dr 16π r acT
ℜρ T
P= (5)
µ A numerical code for solving
1 these equations is available in
µ= (6) the book: C.J. Hansen, S.D.
2 X + 43 Y + 12 Z Kawaler, Stellar Interiors.
ε = ε 0 X 2 ρT 4 (7) Physical Principles, Structure
and Evolution, Springer,
κ = κ 0 ( X + 1) Z ρT −3.5 (8) 1995.
Kramer’s opacity law
 Scaling Laws
Simple estimates can be obtained without solving the equations for solar
structure numerically.
Temperature inside the Sun can be roughly estimated from the equation
of hydrostatic balance and the equation of state.

Using Eq.(1-2) and (4) we obtain the following relations:

dm M
= 4πρ r 2 ρ∼ 3
dr R
dP Gm ρ P GM 2
=− 2
∼
dr r R R5
ℜρT ℜρ T
P= P ∼ .
µ µ 0

The molecular weight for X ∼ 0.7 , Y ∼ 0.28 , and Z ∼ 0.02 is:

µ0 ∼ 0.6.
GM µ0 6.67 × 10−8 ⋅ 2 × 1033 ⋅ 0.6 7
Then, T ∼ ∼ 7 10
∼ 1.4 × 10 K. (9)
Rℜ 8.31 × 10 ⋅ 7 × 10
More general relations among the global properties of stars can be
represented in terms of scaling laws.
Consider homologous transformations: m = Mmɶ ( r /R ) , ρ = ρ0 ρɶ (r / R )
etc, where M , R , ρ0 etc are scaling factors.
Then Eq.(1)-(8) require the scaling factors to satisfy the relations:
dm M
= 4πρ r 2
⇒ ρ0 ∼ 3 (10)
dr R
dP Gmρ P GM 2
=− 2 ⇒ ∼ 5
(11)
dr r R R
2 4
dL L M T ε0
= 4πρ r 2ε , ε = ε 0 X 2 ρT 4 ⇒ ∼ 4
(12)
dr R R
dT κρ T κ0M 2L
=− 2 3
L, κ = κ 0 ( X + 1) Z ρT −3.5
⇒ ∼ 8 6.5 (13)
dr 16π r acT R RT

From equations (12) and (13) we find:

8/ 7
2/7 M
T ∼ (ε 0κ 0 ) 20 / 7
. (14)
R
Then, from Eqs 9 and 14:

GM µ0
T∼ (9)
Rℜ
M 8/ 7
T ∼ (ε 0κ 0 ) 2 / 7 (14)
R 20 / 7

we find the relationship between the mass and radius of a star:

(ε 0κ 0 ) 2 /13 1/13
R∼ 7 / 13
M . (15)
(G µ 0 )

Similarly, we get the temperature-mass relation:

T ∼ M 12 /13 ,
and the luminosity-mass relation:
L ∼ M 71/13 ∼ M 5.46 .
The relations between the global properties and the element abundances can
be obtained if we use power-law approximations for µ0 , ε 0 , and κ 0 :
µ0 ∼ X γ ,
d log µ 5X
where γ = ≈− ≃ −0.54
d log X 5X + 3
for X = 0.7 .

Then, κ 0 ∼ ( X + 1) ∼ X β , where β ∼ 1 ≃ 0.59 ; ε ∼ X 2 .

0
X +1

Using these relations and Eq.(12-13) one can determine how the luminosity
of the star depends on the hydrogen abundance and mass:
L ∼ X −4.78 M 5.46 .
It shows that the luminosity increases when the abundance of hydrogen
decreases.
Exercise: Using the relation between the luminosity and the effective
temperature: L = 4π R 2σ Teff4 estimate the slope of the Main sequence on
d log L
the HR diagram, − .
d log T
eff
 Evolution on the Main Sequence
The change of the hydrogen abundance can be described by the energy-balance
equation which states that the luminosity of the Sun, L , is equal the energy release
dX
per unit mass, E , times the decrease rate of the hydrogen mass, − M :
dt
dX
L = − EM , (16)
dt
where E ≈ 0.007c 2 (the resulting helium has 0.7% less mass
than the original hydrogen).
 Lifetime on the Main Sequence
Equation for the hydrogen abundance:
dX
L = − EM , (16)
dt
Luminosity-abundance scaling law: L ∼ X −4.78 M 5.46 .
Approximate the relationship between the luminosity and hydrogen abundance as:
α
 X 
L = L0  , (17)
X
 0
where α ≈ −4.78 .
α
dX L  X  X L0
=− 0   , y = , a =
dt EM  X 0  X0 EMX 0
dy dy (1−α )
= − ayα , ⇒ α
= − a ⋅ dt ⇒ y = a (1 − α ) (τ n − t )
dt y
τ n is the intergration constant
(1−α )
 X   t 
  = a (1 − α )τ n 1 −  , a (1 − α )τ n = 1 if X = X 0 at t = 0
 X0   τn 
 EMX 0 
τn =  
 (1 − α )L0 
 Lifetime on the Main Sequence
Finally, the solution is:
1
 t  α
1−
X = X 0 1 −  ,
 τn 

α
 t 1−α MEX 0
L = L0 1 −  , where τ n = . α ≈ −4.78 .
 τn  L0 (1 − α )

L0
Then, L= 0. 8
. (19)
(1 − t /τ n )

A finite solution exists only for t < τ n .

MEX 0 0.007c 2 M
τ n ≈ 0.17 ≈ 0.17 ≃ 5.4 × 1017 s ≈ 1.7 × 1010 years
L0 L
is a characteristic time of the Sun’s evolution on the Main Sequence (‘nuclear time’).
 Nuclear Energy Sources: p-p chains
The main source of the solar energy is the proton-proton (pp) reaction which
converts hydrogen to helium
The p-p chain
Nuclear potential
 Nuclear Reactions
Nuclear reactions on the Sun are slow. They have low
probabilities because of the high electrostatic barrier of the
order of 1 MeV compared to the thermal energy of about
1 keV.
Zβe2/r
Particle flux: nαv Zα
Reaction cross- 1 keV 1 MeV
section: σ

Reaction rate: Coulomb barrier

rαβ=nαv σ nβ
Attractive nuclear
potential
density of particles β 25 MeV

Nuclear Barrier.
Consider the reaction rate between particles α and β per unit mass:
rαβ ∼ nα nβ < σ v >,
where σ is the reaction cross-section, and v is the relative velocity
(the reaction rate is the product of the flux, nα v , of particles α ,
density nβ of the target particles and the reaction cross-section σ );
< ... > in an average over the particle velocity (or energy)
distribution:
dn
< σ v >= ∫ σ v ,
n
where dn/n is a fractional number of particles of in an energy
interval [ E , E + dE ] .
For Maxwell-Boltzmann ideal gas:
dn 2 1 E
− kT 1/ 2
= 3/ 2
e E dE .
n π ( kT )
E
− kT
Then, rαβ ∝ ∫ e Eσ dE
We estimate the reaction cross-section qualitatively assuming
that the nuclear interaction happens at the scale of DeBroglie
length,
λ p = hp = h
2 mE ,
where h is Plank’s constant, p is the particle momentum,
m is the mass, and E energy

(compare this with the photon wavelength

λ = c/ν = (hc) / (hν ) = (hc ) /E = h/ ( E /c) = h/p )
The cross-section is proportional to the squared DeBroglie length, λ p2 , and
the probability the Coulomb barrier penetration, exp( − Ee /E ) , where
Zα Z β e2
Ee = is the height of the barrier, Zα and Z β are the particle charges.
λp
Zα Z β e2

2
−
λpE λ p = hp = h
2 mE
Then, σ ∼λ ep ,
Zα Z β e2 b
h 2
− 2m
1 − E1/ 2
or σ∼ e h E
∝ e ,
2mE E
where b is a constant.
dr/dE
−E− b
Then, the reaction rate: r ∝ ∫ Ee
kT E1/ 2
dE .
The integrand has a sharp peak at E = (bkT )3/ 2
- ‘Gamov peak’.
− a
1 T 1/ 3
Therefore, r∝ 2/ 3
e ,
T EGamov E
where a is a constant.
The energy release in the nuclear reactions is:
E ∼ ΣQαβ rαβ ∝ ρT n ,
where n = a − 23 , or more precisely
( kT )1/3
1/ 3
 π Zα Z β e mαβ 
2 2 2 4
2
n= 2  − ,
 2h kT  3
mα mβ
where mαβ = mα + mβ
.
The power index, n , is high for most reactions.
 Nuclear reactions of the pp chain. The rates of the reactions depend on
element abundances, density and temperature, e.g. for a reaction involving
particles a and b , λab ∝ ρ 2 X a X bT n . The power law index n is shown in the
figure.
Abundances of the elements in the pp-chain are determined from the balance
equation, e.g. for hydrogen abundance, X ≡ X 1 :
dX
ρ = ρ 2 ( −3λ11 X 2 + 2λ33 X 32 − λ34 X 3 X 4 ),
dt
where X 3 is the abundance (mass fraction) of 3 He, X 4 ≡ Y is the 4 He abundance.

Estimates from the balance

equations
Balancing the main terms in the
nuclear reaction equations we
obtain relations among various
elements.
ppI:
ρ 2 X 12T 4 ∝ ρ 2 X 32T 16

X 3 ∝ X 1T −6
 X 3 ∝ X 1T −6

Abundances of hydrogen, X 1 and 3 He ( X 3 ), and the the total energy-

generation rate, ε . The 3 He production in the ppI chain increases with the
radius following the simple power law, but drops to zero outside the energy-
generating core because the balance equation is invalid there (the reaction is
not in equilibrium).
ppII: ρ 2 X 4 X 3T 17 ∝ ρ X 7 neT −1/ 2
ρ X 4 X 3T 11.5 (1 − X ) X 11.5
X7 ∝ ∝ T
ne 1+ X
- abundance of 7 Be.
ppIII: ρ 2 XX 7T 13 ∝ ρ X 8
(1 − X ) X 2 24.5
X8 ∝ ρ T
1+ X
- abundance of 8 B which is the main source
of high-energy neutrinos is very sensitive
to temperature T the core.
 Convective Instability
Consider a displacement, δ r , of a small fluid element along the radius. If the
density inside the displaced element, ρ + δρi is smaller the density of surrounding
plasma, ρ + δρ , then the element will continue moving up under the buoyancy
force. Therefore, the condition of the convective instability is:
∆ρ ≡ δρi − δρ < 0.
Physical conditions inside the element obey the adiabatic
law because the characteristic time for heat exchange is
much longer than the dynamic time. Then,
ρ+δρi  dρ  ρ  dP 
δρi =   δ r =   δ r,
δr  dr ad γ P  dr 
ρ
where γ is the adiabatic exponent.
The density variation in the surrounding plasma is:
dρ
δρ = δ r.
dr
Finally, the instability condition is:
1 d log P d log ρ
A∗ ≡ − < 0.
γ d log r d log r
Parameter A∗ is called the Ledoux parameter of
convective stability.
 Pµ
By using the equation of state, ρ = , the condition of instability
ℜT
can be expressed in terms of temperature gradients:
 dT  dT
  − > 0, or ∇ad − ∇ < 0,
 dr ad dr
 d log T 
where ∇ad ≡   is the ‘adiabatic gradient’,
 d log P ad
 d log T 
∇≡  is the ‘local (ambient) gradient’.
 d log P 
For an ideal gas,
P = ℜρ T / µ , P ∝ ρ γ
γ −1
γ
T∝P .
γ −1
∇ ad ≃ ≈ 0.4
γ
 Convective Energy Transport. Mixing-Length Theory
Consider the momentum equation of a fluid element:
∆ρ ≡ δρi − δρ
d 2δ r  d ρ  dρ 
ρ+δρi
ρ 2 = − g ∆ρ = − g   − δ r =
 ad
δr
ρ dt dr dr 
 d log ρ  d log ρ 
= − g ρ   −  δr =
 dr ad dr 
 d log T  d log T  1
= − g ρ   −  δ r,
 d log P  ad d log P H
 P
1 d log P ℜρ T
where =− is the pressure scale height; P =
HP dr µ
Another form of
d 2δ r g 2 Brunt-Vaisala frequency:
Finally, 2
= − ( ∇ ad − ∇ )δ r = − N δ r, (22)
dt HP
g  1 dP 1 d ρ 
2 N2 = g − .
where N = (∇ad − ∇) is the Brunt-Väisälä frequency.  γ P dr ρ dr 
HP
If N 2 < 0 then the medium is convectively unstable.
If N 2 > 0 then it is convectively stable.
In this case Eq.(22) has a solution δ r ∝ sin( Nt ) , that is the fluid elements oscillate with
frequency N . These oscillations are called internal gravity waves - g-modes.
If we multiply Eq.(22)

d 2δ r g
2
= − (∇ad − ∇)δ r = − N 2δ r,
dt HP
by 2 dδ r and then integrate over t we get
dt
2
 dδ r  g
  = (∇ − ∇ad )δ r 2 . (23)
 dt  H P

The mixing-length theory assumes that the convective elements travel without
destruction a distance l - ‘mixing length’.

Then, the characteristic velocity, v , of these elements can be estimated from Eq.(23):
g
v2 ∼ (∇ − ∇ad )l 2 .
HP
3/ 2
 
2 3 g 2
Then, the convective energy flux is: Fc = ρ v ⋅ v = ρ v = ρ 

(∇ − ∇ ad )l  .
HP 

 3/ 2
 
3 g
 2
The convective energy flux is: Fc = ρ v = ρ  (∇ − ∇ad )l  .

HP 
 

In the convection zone, ∇ is slightly greater than ∇ad . This is sufficient to carry the
energy flux; the convective velocity is small in this regime, which is called ‘efficient
convection’.

However, near the solar surface where the density, ρ , is small much higher, near sonic
velocity, is required to transport the solar energy. In this case, ∇ >> ∇ad . This near surface
zone is called a ‘superadiabatic zone’.

The mixing length is usually defined in terms of the pressure scale height:
ł = α HP,
where α is called ‘the mixing length parameter’.
Comment on numerical modeling of the stellar structures.

The stellar structure equations are solved for a star of mass M to match
the observed (given) radius R and luminosity L .
The radius depends mostly on the mixing length parameter, α , and the
luminosity depends mainly on the abundance of hydrogen, X , (or
helium Y ). Usually, α and Y are considered as free parameters to
match R and L .

Online stellar modeling:

http://mesa-web.asu.edu/index.html
Standard solar model
Standard solar model
 Gravitational Settling
It is assumed that initially, at the ZAMS, the Sun was chemically homogeneous because of mixing
during the collapse phase. During the evolution on the main sequence it is well mixed in the
convection zone. However, in the radiative zone there is element separation due to gravity.
Elements heavier than hydrogen ‘settle’ under gravity.

Estimate the rate of the gravitational settling (diffusion). The settling speed can be estimated as:
vD ∼ gτ D ,
1
where τ D is the mean collision time: τ D ∼ , where n is the plasma density, σ is the
nσ vT
collision cross-section, and vT is the thermal velocity of ions.

g
Thus, vD ∼ .
nσ vT
For g ∼ 104 cm/s 2 , n ∼ 1023 cm −3 , vT ∼ 106 cm/s, and σ ∼ 10−16 cm −2 , we get vD ∼ 10−8 cm/s.
The characteristic diffusion length during the previous solar evolution is: L ∼ vD t⊙ ∼ 1.5 × 109 cm.
This is significantly less than the radius of the radiative zone, ∼ 5 × 1010 cm.

Nevertheless, taking into account the gravitation settling is important for an accurate modeling the
solar structure. The abundance of helium in the convection zone has decreased due the
gravitational settling from 0.28 to 0.25, by almost 10%.
 Solar Neutrino Flux
The total neutrino flux can be estimated from the nuclear reaction rate:
2 L⊙ 1 10 −2 −1
Φ⊙ = 2
≈ 6. 51 × 10 cm s ,
Q − 2 Eν 4π D
where D ≈ 1.5 × 1013 cm is the distance from the Sun, Q = 26.733 MeV is the energy output of the
reaction 4 1 H →4 He; each reaction produces 2 p-p neutrinos, Eν ≈ 0.265 MeV is the energy of
neutrino in each of p-p chains.
The total neutrino flux consists mainly of p-p, 7 Be and 8 B neutrinos:
Φ ⊙ ≈ Φν (p − p) + Φν ( 7 Be) + Φν (8 B).
Estimate the flux of 8 B neutrinos by counting the relative rate of the ppIII chain:
Φν (8 B) ∼ 2 ⋅ 0.015 ⋅ 0.0002Φ ⊙ ≃ 6 × 10−5 Φ ⊙ ≃ 4 × 106 cm −2s−1.
The 8 B neutrinos are observed through the reaction:
ν e + 37 Cl → e − + 37 Ar.
Φ obs =< σ Ar Φν > N ≃ 4.5 × 10−36 Ns −1,
where the cross-section of the reaction σ Ar ≈ 1.11 × 10−42 cm 2 , N is the number of the target atoms
( 37 Cl).
For a reference number of the target atoms, N = 1036 :
Φ obs ≈ 4.5 SNU,
where SNU is a ‘solar neutrino unit’ - the number of the reaction per sec per 1036 atoms.
Sensitivity of various neutrino experiments
37Cl detector
 GALLEX experiment
Result: 77.5 8 SNU

Standard Solar Model prediction:

129.6 SNU

νe + 71Ga → 71Ge + e–
Superkamiokande (Japan)
Neutrino image of the Sun
 Sudbury Neutrino Observatory (SNO)
measured the total flux of electron, muon and tau neutrinos
and confirmed the neutrino transition from one type to another on their
way from the Sun to Earth
νe + 37Cl → 37Ar + e–. νe + 71Ga → 71Ge + e–
 MSW (Mikheyev-Smirnov-Wolfenstein) effect

The MSW effect states that any coherent forward scattering of electron
neutrinos in electronic matter results in a density-dependent effective
mass. This means that electron neutrinos which travel through an
inhomogeneous medium (like the density gradients in the sun or the
earth) have some probability of changing their effective mass and hence
their flavor.

The masses of two flavors of neutrino as a

function of density.

The probability of transition between the states

is high in the avoided crossing region
(MSW-effect).
Mixing angle: ν e = ν e sin θ + ν µ cos θ .
The mixing angle should be small to explain
the deficit of ν e .