CHAPTER 

6:  Ideal Diatomic Gas

[q q q q ]
N

Q ~ rot vib tr el for N indistinguishable molecules

N!
We will exam this more closely later
⎡ 2π ( m1 + m2 ) kT ⎤
3/2

qtr = ⎢ 2 ⎥ V
⎣ h ⎦

Actually, we also need to consider nuclear spin and its

ramifications for the rotational partition function.
Assuming a heteronuclear diatomic
−2θ r /T −6θ r /T
qrot = 1 + 3e + 5e + ...
T8π 2 IkT
high T replace sum by ∫ → qrot = =
θr h2
T ⎡ 1 θr 1 ⎛ θr ⎞ ⎤
2

qrot (T ) = ⎢1 + + + ...⎥ The extra terms result

θ r ⎢⎣ 3 T 15 ⎜⎝ T ⎟⎠ ⎥⎦ from doing a better sum
→ integral
Erot = NkT + ... transformation
CV ,rot = Nk + ...

T
q =
homonuclear diatomic rot 2θ keeping only the leading term
r

symm # (# indistinguishable
orientations)
 The total wave function is either antisymmetric (fermions) or symmetric
(bosons) wrt interchange of the nuclei

Ψtot = Ψtr Ψrot Ψvib Ψel Ψnuc

consider H2:  nuclear spins = 1/2 (fermions)

nuclear spin functions aβ − β a S
αβ + β a
αα T
ββ

nuclear spin          rot
para S (‐)              J even (+)
ortho T (+)             J   odd (‐) statistical weight 3

nuclei of spin I 2I + 1 spin states per nucleus

 qrot ,nucl ≠ qrot qnucl

if   θ r << T
1 1 ∞ T
∑
J even
≈ ∑
J odd
≈
2
∑
J
≈
2 ∫0
( 2 J + 1)e
−θ r J ( J +1) / T
dJ =
2θ r

( 2 I + 1)
2

qrot ,nucl ~ = qrot (T )qnucl need θ r / T < 0.2

2θ r to be valid 
T
( 2 I + 1)
2

2θ r
 1 +
In the above and following discussion, we assume a Σ g electronic state

Half‐integral spin
I ( 2 I + 1) antisymm nuclear spin function – even J
( I + 1)(2 I + 1) symm nuclear spin function – odd J
( I +1)( 2I +1) +I ( 2I +1) =( 2I +1)
2

Integral spin
I ( 2 I + 1) antisymm nuclear spin function – odd J
( I + 1)(2 I + 1) symm nuclear spin function – even J

Integral spin
qrot ,nucl = ( I + 1)( 2 I + 1) ∑ ( 2 J + 1)e −θr J ( J +1) / T
J even

+ I ( 2 I + 1) ∑ ( 2 J + 1) e
θ r J ( J +1) / T

J odd

prefactors switched for nuclei of ½ integral spin

 H2 θ r = 85.3 ( K ) largest θr value

qrot ,mol = ∑ ( 2 J + 1) e + 3∑ ( 2 J + 1) e
−θ r J ( J +1) / T −θ r J ( J +1) / T

J even J odd

para ortho

H2 is 100% para at T = 0
If at equilibrium
→ 25% para at high T

Measurements of Cv vs. T:  disagreement between theory and experiment

problem O ↔ P interconversion slow

so high temperature distribution locked in
as sample is cooled

Due to rotational
for most homonuclear diatomics don’t have to worry levels being
about this other than the factor of 2 symmetry #) much closer
together
 qe = ωe1e De / kT + ωe 2 e −ε 2 / kT + ...

De
using free atoms as the zero
of energy – generally need 
only retain the 1st term

⎛ 2π mkT ⎞ 8π 2 IkT − β hv /2
3/2

( )
− β hv −1
assumes no low‐lying
q=⎜ ⎟ V e 1 − e ω e De / kT
electronically states
⎝ h
2
⎠ σ h2 e1

E 5 hv hv / kT D 5 Nhv ⎛ hv ⎞
= + + hv / kT − e E= NkT + hv / kT − ⎜ De − ⎟ N
NkT 2 2kT ( e − 1) kT 2 e −1 ⎝ 2 ⎠
does not contribute
2
Cv 5 ⎛ hv ⎞ e hv / kT
to heat capacity
= +⎜ ⎟
Nk 2 ⎝ kT ⎠ ( e hv / kT − 1)2
 pV = NkT
<E>
S= +AnQ
T
S ⎡⎛ 2π mkT ⎞3/ 2 Ve5 / 2 ⎤ ⎛ 8π 2 IkTe ⎞
= An ⎢ ⎜ ⎟ ⎥ + An ⎜ ⎟
Nk ⎢⎣⎝ h 2
⎠ N ⎥⎦ ⎝ σ h 2
⎠

− An (1 − e − hv / kT ) + An ωel
hv / kT
+ hv / kT
e −1

μ0 ⎛ 2π mkT ⎞ 8π 2 IkT
3/ 2
hv
= − An ⎜ ⎟ kT − A n +
kT ⎝ h
2
⎠ σ h2 2kT

+ An (1 − e − hv / kT ) − e − An ωel
D
kT

For general polyatomic molecules we don’t have a closed
form for the rot. energy levels.  So it would appear that we
can’t simply extend the procedure followed for diatomics.