Saha Ionisation Equation

Ionisation energies of atoms are of the order of tens of eV (for hydrogen, it is 13.6 eV). The
equivalent temperature is ∼ 104 − 105 K. Such high temperatures are easily realised in various
astrophysical situations (stars, in particular). Hence, it is of importance to astrophysical studies.
Thermodynamically, ionisation can be looked upon as a chemical reaction where one reactant
transforms into two new ones (or vice-versa). Such a treatment was first developed by Meghnad
Saha, and the essence is captured in the famous Saha Ionisation equation.
In general, a process that can change the number of particles of a certain kind, is called a
chemical reaction. A generic reaction is written as,
X
R j α j = 0, (105)
j

where, R j and α j are the chemical symbol and the no. of particles of species j taking part in the
reaction. Since, reactions take place in equilibrium with their surroundings, temperature and
pressure remain constant, same as that of the surroundings. The appropriate thermodynamics
potential for the overall system is then the Gibb’s potential with the equilibrium condition given
by,
dG = −S dT + VdP = 0 . (106)
(It is, of course, conceivable to have reactions taking place under different circumstances. An
appropriate thermodynamic potential need to be used to describe such reactions.) Now, for each
particle species we have,

dG j = µ j dN j , with dT, dP = 0 , (107)

which, in equilibrium, implies,

X X
dG = dG j = µ j dN j = 0 . (108)
j j

However, the number of particles, of a particular species, would change strictly according to the
specific reaction being considered. Each time the reaction happens, the number of each reactant
R j changes only by n j . Thus, the total change, dN j must be proportional to α j . At equilibrium
then, the chemical potentials of different reactants satisfy the relation
X
µ j α j = 0, (109)
j

which is the law of chemical equilibrium. It should be remembered that we need to use the
total chemical potential (inclusive of the rest mass of the particle) here. Because, this condition
implies that X
j α j = 0,
µNR (110)
j
P
only if m j α j = 0, i.e., the total rest mass is invariant through the reaction in consideration
j
(µNR = µ − mc2 ). Since, that is not true of all reactions (notably, the fusion reactions in stars) it
is not appropriate to use µNR in general.

20
Let us now suppose that the reactants can be treated as classical ideal gases. The single particle
partition function for an ideal gas is given by,
!3/2
mkB T V
Q1 = V = , (111)
2π�2 ΛT

where ΛT can be thought of as the thermal de Broglie volume associated with a particle. Then
the partition function for an N-particle idea gas is,
!N
1 V
Q= , (112)
N! ΛT

and the grand partition function is

X QN
Z= eβµN . (113)
N=0
N!

Now,
∂ ln Z
P = kB T |T,µ
∂V
eβµ
= . (114)
βΛT
Using the ideal gas equation of state PV = NkB T = N/β we find

N eβµ
n= = . (115)
V ΛT
If the particles of the reactants (assumed to behave like classical, ideal gases) have internal
degrees of freedom, this formula is modified to

qeβµ
n= , (116)
ΛT
where q is the contribution to the partition function due to the internal degrees of freedom,
implying !
1 nΛT
µ = ln . (117)
β q
Using this in the equation for chemical equilibrium we have,
X ! !
n jΛ j αj q j αj
α j ln = 0, ⇒ Π j n j = Π j . (118)
j
qj Λj

Here Λ j is the quantum volume for the j-th species. The contribution of the internal degrees of
freedom to the partition function can be written as
0
q = q0 e−βǫ , (119)

where ǫ 0 is the energy of the ground state of the internal degrees of freedom and q0 is the internal
partition funtion for zero ground state energy. If there are only one kind of particles, then one

21
can set ǫ 0 = 0, but it can not be done when different species of particles are involved and we
have
 −βǫ 0 α j
 q0 e j 
Π j n j j = Π j   .
α j
(120)
Λ j 

For example, consider

X + Y ⇄ XY (121)
where X and Y denote two kinds of particles and XY denote their bound state. Then in equilib-
rium, the number densities of X, Y and XY are governed by,

nX nY q0X q0Y ΛXY h

0
i
= 0 exp β(ǫXY − ǫX0 − ǫY0 ) . (122)
nXY qXY ΛX ΛY

The quantity multiplying β in the exponent is the binding energy of the molecule XY if the
atoms X and Y react to form it.
Ionisation : Consider an ionisation equation of the following form

X + + e− ⇄ X , (123)

where X can be looked upon as the bound state of the ion X + and the electron. As the masses of
the ion and the atom are almost equal we have ΛX + ≃ ΛX . Therefore,

nX + ne− q0X + q0e− e−βǫ

= . (124)
nX q0X Λe

The first factor on the right hand side usually contains the degeneracy factors due to spin etc. If
we define the fraction of the particles in the following manner,
n x+
n X + n X + = nt , f = , (125)
nt
then, with charge neutrality (nX + = ne )

f2 q0 + q0− e−βǫ
= X 0e . (126)
1− f q X nt Λe

This is the Saha Ionisation Equation. In the NR limit Λ ∝ T −3/2 , so at high temperatures, the
right hand side becomes very large. This means at high temperatures the ionisation fraction is
very large. As T → 0, we have f → 0 although the validity of this equation should be suspect
in this limit because the gases may not behave as classical ideal gases.

22