Molecular energy levels and spectroscopy

1. Translational energy levels

The translational energy levels of a molecule are usually taken to be those of a particle in a
three-dimensional box:

h2 ⎛nx2 ny2 nz2 ⎞

E(nx,ny,nz) = 2m ⎜ + + ⎟
⎝ lx2 ly2 lz2 ⎠

In general, the separation of the translational energy levels is many orders of magnitude
smaller than kT, even for a box with dimensions on the order of molecular sizes. The
translational behaviour of molecules therefore appears to be classical.

2. Rotational energy levels – diatomic molecules

Diatomic molecules are often approximated as rigid rotors, meaning that the bond length is
assumed to be fixed. Solving the Schrodinger equation for a rigid rotor gives the following
energy levels:

E(J) = B J(J+1)

In this equation, J is the quantum number for total rotational angular momentum, and B is the
rotational constant, which is related to the moment of inertia , I = µr2 (µ is the reduced mass
and r the bond length) of the molecule.

h
B=
8π2cI

B is usually given in units of cm-1, and in order to end up with these units, the speed of light in
this equation, c, needs to be in cm s-1 rather than ms-1. Using B in cm-1 will yield the energy in
the same units.

[Note: cm-1 are common units in spectroscopy for historical reasons, even though they are
non-SI. Units of m-1 are never used, so don’t be tempted to convert, however stupid it might
seem...]

The equation above for E(J) is only approximate. In order to provide a better description of
the energy levels of a diatomic molecule, a centrifugal distortion term is added to the energy.
This takes account of the fact that as a real molecule rotates faster and faster (i.e. with
more energy), the bond stretches a little (if you swing a weight attached to a piece of elastic
in a circle, you can observe the same effect on a macroscopic scale).

E(J) = B J(J+1) – DJJ2(J+1)2

DJ is called the centrifigual distortion constant, and is several orders of magnitude smaller
than B.

Separations of rotational energy levels correspond to the microwave region of the

electromagnetic spectrum. In order for a molecule to absorb microwave radiation, it must
have a permanent dipole moment. i.e. µ ≠ 0. This condition is known as the gross selection
rule for microwave, or pure rotational, spectroscopy. Specific selection rules arise largely
from conservation of angular momentum, and generally involve statements of the allowed
changes in quantum number. In general, the selection rule for changes in rotational angular
momentum following absorption of a photon is ∆J = 0,±1. However, for pure rotational
transitions (we will cover mixed rotation-vibration transitions later), ∆J=0 does not
correspond to a transition since the particle stays in the same state, so the specific selection
rule for a pure rotation spectrum is ∆J = ±1.

The transition energies may be calculated from the energy equation above. For a transition
from a level J to a level J+1, we have:

∆E(J) = E(J+1) – E(J)

= B(J+1)(J+2) – DJ(J+1)2(J+2)2 – B J(J+1) + DJJ2(J+1)2
= B(J2 + 3J + 2 – J2 – J) + DJ(J+1)2[J2 – (J+2)2]
= 2B (J+1) - 4DJ(J+1)3

Neglecting the centrifugal distortion term, the spacing between lines in a spectrum will be

∆E(J+1) – ∆E(J) = 2B(J+2) – 2B(J+1) = 2B

i.e. lines in a pure rotational spectrum are (almost) equally spaced by 2B (almost because we
have neglected the effects of centrifugal distortion). This result is often useful if we are
attempting to assign lines in a spectrum in terms of the quantum numbers of the states
involved. The separation between a pair of lines allows us to estimate B (i.e. B =
separation/2), and we can use ∆E(J) ≈2B(J+1) to determine J.

3. Rotational energy levels – polyatomic molecules

Polyatomic molecules may rotate about the x, y or z axes, or some combination of the three.
They have moments of inertia Ix, Iy, Iz associated with each axis, and also corresponding
rotational constants A, B and C [A = h/(8π2cIx), B = h/(8π2cIy), C = h/(8π2cIz)]. The general
equation for a moment of inertia is:

I = Σi mi ri2

Where the sum is over the atoms in the molecule, mi is the mass of atom i and ri is its
distance from the rotation axis.
[Note: this explains why only two rotation axes are counted in a diatomic molecule. Rotation
about the principal axis would have I=0, since all the atoms lie along the axis, and the
associated rotational constant and energy would therefore be undefined]

In general, the rotational constants A, B, and C may all be different, and a molecule for which
this is true is called an asymmetric top. The spectroscopy of such molecules is quite
complicated, and beyond the scope of the course. Instead, we will look at symmetric top
molecules, such as CH3F, CH3Cl, CCl3H or benzene, for which two of the rotational constants
are equal.

Symmetric tops may be divided into prolate tops, for which B > A,C (e.g. CH3F, CH3Cl) and
oblate tops, for which B < A,C (e.g. CCl3H, benzene). The rotational motion of a symmetric
top molecule is described by three quantum numbers:

J – the total angular momentum

K – the projection of J onto the principal axis of the molecule
MJ – the projection of J onto some chosen lab-frame z axis (often just written M)

Regardless of which category the molecule falls into, the rotational energy is given
approximately by:

E(J,K) = B J(J+1) + (A-B)K2

When centrifugal distortion is included, this becomes

E(J,K) = B J(J+1) + (A-B)K2 - DJJ2(J+1)2 – DJKJ(J+1)K2 – DKK4

The selection rules for a symmetric top molecule are ∆J = ±1, ∆K = 0.

Transition energies and line separations may be calculated in a similar way to that worked
through above for diatomic molecules, by determining E(J+1,K) – E(J,K) etc.

4. Vibrational energy levels

To a first approximation, molecular vibrations can be approximated as simple harmonic

oscillators, with an associated energy

E(v) = (v + ½)hν

In the above expression, E(v) is in units of Joules. Usually, the energies are given in cm-1, in
which case hν is replaced by ωe, the vibrational wavenumber (i.e. this is simply the harmonic
vibrational frequency or energy expressed in units of cm-1).
 E(v) = (v + ½) ωe

This is usually a fairly good approximation near the bottom of the potential well, where the
potential closely resembles that of a harmonic oscillator. However, real molecules have
anharmonic potentials, and this is taken into account by rewriting the energy as a series
expansion, with successively higher order correction terms:

E(v) = (v + ½)ωe – (v+ ½)2 xeωe + (v+ ½)3yeωe + ...

Often, only the first correction term, involving xeωe, will be used, but note that the source of
differences in calculated and measured properties involving vibrational energy levels is
frequently neglect of higher order terms. Note that every vibration has a zero point energy
found by substituting v=0 into the above equation, ZPE = E(0) = ½ωe – ¼xeωe

The energy levels in a harmonic oscillator are equally spaced by an energy ωe. The effect of
anharmonicity is that the levels get increasingly closer together as the vibrational quantum
number v increases, until eventually there is no difference in energy between successive
quantum states and the energy levels form a continuum. The energy at which this occurs
(known as the dissociation energy) coincides with the point at which the potential energy
function reaches zero and the molecule is no longer bound (this is the reason why the energy
levels are no longer quantized). The dissociation energy may be measured either from the
bottom of the potential, in which case it is called De, or from the zero point energy, in which
case it is called D0.

dE
To determine the dissociation energy, we set equal to zero to determine the maximum
dv
value of the vibrational quantum number v.

dE ωe
dv = 0 = ωe - 2(vmax+½)xeωe ⇒ (vmax+½) = 2x ω
e e

Substituting back into the energy expression above gives

ωe2
De = E(vmax) = 4xeω
e

De and D0 are related by

D0 = De – E(0)

5. Rovibrational spectroscopy

Ignoring electronic excitation, the total internal energy of a molecule is the sum of its
vibrational and rotational energy.
 E(v,J) = (v+½)ωe – (v+½)2xewe + BvJ(J+1) – DJJ2(J+1)2

Transitions between the E(v,J) levels in which v changes correspond to absorption of energy
in the infrared region of the spectrum. The gross selection rule for a vibrational transition is
dµ
dt ≠ 0. i.e. the molecule must have a dipole moment that changes during the course of the
vibration (important note: this requirement may be satisfied without the molecule having a
permanent dipole). If the vibrations are approximated as simple harmonic oscillators, the
specific selection rule is ∆v = ±1. Since real vibrations are anharmonic, in reality there are
contributions from ∆v = ±2, ∆v = ±3, etc, particularly for transitions originating in highly
anharmonic regions of the potential. A change in vibrational quantum number v is often
accompanied by a change in rotational quantum number J, according to the selection rule ∆J =
0, ±1, so we talk about rovibrational (or vibrotational) transitions, rather than simply
vibrational transitions. In a linear molecule, ∆J = 0 is only allowed if there is an angular
momentum perpendicular to the axis. This could arise from a bending vibration or from the
electronic angular momentum of an unpaired electron (e.g. OH, NO). A transition between
two vibrational states gives rise to a vibrational band, made up of P, Q and R branches,
corresponding to transitions between rotational states with ∆J = -1, 0 (if allowed) and 1.

5. Populations of energy levels and spectral line intensities – the Boltzmann distribution

The intensities of lines in a spectrum depend upon a linestrength factor related to the
overlap between the wavefunctions of the two states, the light intensity, and the population
of the lower state. Within a given rotational or vibrational spectrum, the lower state
populations are usually the dominant factor. Under thermal conditions, these are given by
the Boltzmann distribution.

ni
= gi exp(-Ei/kT) ,
n0

where gi is the degeneracy and Ei the energy (relative to the ground state) of state i. As a
consequence, we can predict the most intense peak in a rotational or rovibrational spectrum
by maximising the Boltzmann distribution to find the most highly populated rotational state
at a given temperature. For rotational lines, the above equation becomes

nJ
= (2J+1) exp[-BJ(J+1)/kT]
n0

1 2kT 1
Setting the first derivative equal to zero and solving for Jmax gives Jmax = 2 B - 2.