Applied

Spectroscopy

Vibrational-Rotational
Spectroscopy

Recommended Reading:
Banwell and McCash Section 3.2, 3.3
Atkins Section 16.12
 Harmonic oscillator vibrations have the
∆ v = ±1
exact selection rule:

and the energy spacing is constant, the

same for all energy levels: ∆~
ν = G(v + 1) − G(v) = ~
ν

This would give rise to very boring spectra, since the lines for all
possible vibrational transitions would lie at the same energy (or
wavenumber of frequency)
This does not occur for two reasons
1) No molecule is a perfect Harmonic Oscillator; as discussed before
we must use an anharmonic potential as a better representation of
the true molecular potential (e.g. Morse potential)

νn→ n+1 = G(n + 1) − G(n) = ~

∆~ ν0 χ e (n + 1)
ν0 − 2 ~

so transitions shift to lower energy as n increases

 2) Absorption or Emission of light MUST be accompanied by a
change in angular momentum of the molecule because of the
gain/loss of the photon’s angular momentum.
All vibrational spectra MUST be Vibration-Rotation Spectra and
the rotational component for the transition must obey the usual
rotational selection rule ∆ J = ± 1.

IR spectrum of CO2 In Solution: Low resolution, see two

absorption bands

In Gas Phase: High resolution Rotational fine structure

Relative energies of transitions

Erot < Evib < Eel

 The Born-Oppenheimer Approximation
Typical rotational energies for a diatomic molecule are of the order
of 1 to 10 cm-1, while the separation of the vibrational energy levels
are typically 1000 to 5000 cm-1.

Since the energies of the two motions are so different we can

assume, as a first approximation, that the rotational and vibrational
motions of the diatomic molecule are independent of each other.

This is an example of the Born-Oppenheimer approximation, and is

equivalent to assuming that the combined rotational-vibrational
energy of the molecule is simply the sum of the separate energies.

E total = Erot + E vib Joules or ε total = ε rot + ε vib cm -1

We have already derived expressions for the rotational and
vibrational energies of the molecule
ε rot = F( J) = BJ (J + 1) + DJ2 (J + 1)2
and 2
 1 ~  1
ε vib = G(n) =  n +  ν −  n +  χ e ~
ν
 2  2
 The Born-Oppenheimer Approximation
Another way to understand the Born-Oppenheimer Approx is to
consider typical times involved in rotational and vibrational motions.

Rotational Motion:
Rotational Constants B are of the order of 1 - 100 cm-1.
Taking B = 1.0 cm-1 gives
I = 2.8 x 10-47 kg.m2 and E(J = 1) = 3.97 x 10-23 J.
Since E = ½ Iω2 this gives ω = 1.7 x 10-12 rad.s-1 = 2.7 x 1011 rot.s-1.

Vibrational Motion:
Vibrational energies are typically 500 - 5000 cm-1. Taking 2000 cm-1
gives a vibrational frequency of 6.0 x 1013 s-1.

This shows that the rotational and vibrational motions take place on
significantly different time scales, (the above example shows 222
vibrations occurring during a single rotation of the molecule).
So it is a good approximation to assume that these motions are
decoupled and, to a first approximation, do not influence each other.
 3

2 n =1
J’ =
Energy level diagram 1
of a diatomic 0
molecule showing
the n = 0 and n = 1
vibrational energy
levels and associated
rotational states.

Note this diagram is 3

not to scale. - The
vibrational states are 2 n=0
J’’ =
typically 500 - 5000 1
cm-1 apart while the 0
rotational levels have ν (1 − 2 χ e ) cm −1
~
typical separations of
1 - 100 cm-1

P3 P2 P1 Q0 Q1 Q2 cm-1
 Spectral Branches
The appearance of the vibration-rotation spectrum of a diatomic
molecule can then be discussed in terms of the combined
vibration rotation term S(n,J). S(n, J) = G(n) + F( J)
If we ignore the centrifugal distortion then
2
 1  1
S(n, J) =  n + ~ ν + BJ(J + 1)
ν - n +  χe~
 2  2
vib. term + anharm term + rot. term
When a vibrational transition occurs n → n+1 it must be accompanied
by a change in the rotational state in order to conserve angular
momentum so J → J ± 1 (Note that in some cases ∆ J = 0 is allowed)
The transitions then fall into three groups called branches of the
spectrum. Designating the rotational quantum numbers in the n = 0
state as J’’ and in the n = 1 state as J’ then an analytical expression
for the spectrum may be obtained by applying the selection rules to
the energy levels. Considering the n = 0 to n = 1 transition:
ν (1 − 2 χ e ) + B(J'− J' ')(J'+ J' '+1) cm −1
∆ε J,v = ε J',n=1 − ε J'',n=0 = ~
 P and R Spectral Branches
We now have three cases to consider ∆ J = - 1, ∆ J = +1 and ∆ J = 0
which can occur in some molecules.
Case 1: P Branch, ∆ J = - 1 that is J'' = J'+1 or J' - J'' = -1; hence

ν (1 − 2 χ e ) − B(J'+1) cm −1
∆ε J,n = ~ J' = 0, 1, 2, Κ
This corresponds to a vibrational transition in which the rotational
energy of the molecule decreases by one unit of angular momentum
⇒ spectral lines at ~ ν − 4B, Κ cm −1
ν − 2B, ~ with an intensity
distribution reflecting (I) the population of the rotational levels and
(2) the magnitude of the J → J-1 transition moment

Case 2: R Branch, ∆ J = + 1 that is J' = J’’+1 or J' - J'' = +1; hence

ν (1 − 2 χ e ) + B(J' '+1) cm −1
∆ε J,n = ~ J' ' = 0, 1, 2, Κ
This corresponds to a vibrational transition in which the rotational
energy of the molecule decreases by one unit of angular momentum
⇒ spectral lines at ~ ν (1 - 2χ e ) + 4B, Κ cm −1
ν (1 - 2χ e ) + 2B, ~ again,
with an intensity distribution reflecting (I) the population of the rotational
levels and (2) the magnitude of the J → J+1 transition moment.
 These two expressions may be combined into a single expression
∆ε J,n = ~ ν (1 − 2 χ e ) + 2Bm cm -1
νspect = ~ m = ± 1, ± 2, ± 3, Κ
where positive values of m correspond to ∆J = +1 transitions while
negative values of m correspond to ∆J = -1 transitions.

Case 3: Q Branch, ∆ J = 0 that is J'' = J’ or J' - J'' = 0; hence

ν (1 − 2 χ e ) cm −1
∆ε J,n = ~

The frequency ν0 = ν ̃(1-2χe) is the band origin or band centre

The Q-branch is usually not observed in diatomic molecules.

 3

2 n =1
J'' =
1
0

2 n=0
J' =
1
separation = 2B 0
ν (1 − 2 χ e ) cm −1
~
separation = 2B

P3 P2 P1 Q0 Q1 Q2 cm-1
separation = 4B
Vibration Rotation Spectra
 Vibration Rotation Spectrum of CO
Fundamental transition ∆n = 1.

Note second band,

centred at 2100 cm-1
due to the 1% of 13CO.

First overtone ∆n = 2.
 Vibration Rotation Spectrum of CO
Recall from Rotation Spectroscopy that the maximum population
level occurs at a J value of
kT 1 And since m = J+1 we see
-
2Bhc 2 that at the maxima in the P
and R branches
kT 1
m=± +
2Bhc 2
The maximum intensity values are at
 kT 1
νmax . int = ν (1-2 χ e ) ± 2B
~ ~ + 
 2Bhc 2 
The separation between the maxima is:
8kT B
∆~
ν= + 2B
hc
or since B is small compared with ∆~
ν we can write

∆~
ν=
8kT B
⇒ B=
hc (∆~ν )2
cm -1
hc 8kT
Can estimate rotational constant B from the vibrational spectrum.
Combination Differences

In general the rotational constant B1 of a

vibrationally excited state is slightly
smaller than the rotational constant of the
ground vibrational state B0, because the
vibration causes a more extended bond in
the upper state. As a result, in the
anharmonic oscillator:
(i) the Q band, if it exists, consists of a
series of closely spaced lines
(ii) the lines of the R branch converge
slightly as J increases and
(iii) those of the P branch diverge.
We can quantify this as follows:
The ground vibrational state n = 0, has a rotational constant B0 and
therefore its term values are
ν χe~
~ ν
S(0, J) = - + B0 J(J + 1)
2 4
The first excited vibrational state n = 1, has a rotational constant B1 and
its terms are
ν 9χ e ~
3~ ν
S(1, J) = - + B1J(J + 1)
2 4
The P-branch of the fundamental vibration now becomes

νP (J) = S(1, J − 1) − S(0, J) = ~

~ ν (1 - 2χ e ) − (B1 + B0 )J + (B1 − B0 )J2
Note that since B0 is always greater than B1 the last term will always be
negative, showing that lines in the P-branch diverge as J increases.

Similarly, the R-branch of the fundamental vibration is

νR (J) = S(1, J + 1) − S(0, J) = ~

~ ν (1 - 2χ e ) + (B1 + B0 )(J + 1) + (B1 − B0 )(J + 1)2
Again because the last term is always negative, the lines in the R-branch
will tend to converge as J increases
 In molecules where the Q-branch exists, we find

νQ (J) = S(1, J) − S(0, J) = ~

~ ν (1 - 2χ e ) + (B1 + B0 )J(J + 1)

We can determine the rotational constants individually using the

method of Combination Differences. A technique widely used in all
branches of spectroscopy to determine information about a particular
state.

It involves setting up expressions for the

difference in wavenumbers of transitions
to a common state.

From diagram we see that the transitions

νR (J − 1)
~ and νP (J + 1)
~

have a common upper state and from the

above equations we see that
 1
νR (J − 1) − νP (J + 1) = 4B0  J + 
~ ~
 2
Therefore a plot of the combination difference
~ν (J − 1) − ~ν (J + 1) against J+1/2
R P
should be a straight line of slope 4B0, so the rotational constant of
the n= 0 state can be determined. Any deviation from a straight line
can be attributed to centrifugal distortion so it can also be
investigated.
Similarly, we can look at transitions originating from a common
ground state. For example
νR (J)
~ and νP (J)
~

have a common lower state, their combination difference gives

 1
νR (J) − νP (J) = 4B1 J + 
~ ~
 2
A plot of this combination difference against (J+1/2) gives a
straight line of slope 4B1, so the rotational constant of the n = 1
vibrational state can be determined.
When applied to the vib-rot spectra of 1H35Cl this technique gives B0
= 10.440 cm-1 and B1 =10.136 cm-1 .