The Rigid Rotor as a Model for Molecular Rotation

Rotational spectroscopy built on the rigid rotor model remains one of the gold-standard techniques for gas-phase structure determination, because it is essentially free of the systematic errors that plague diffraction methods on complex molecules. Modern Fourier-transform microwave (FTMW) spectrometers, often combined with pulsed supersonic jet expansion to cool molecules to a few kelvin (collapsing the thermal population almost entirely into $J=0,1,2$), can measure rotational constants to eight or nine significant figures, resolving bond lengths and even subtle structural isomers or conformers that differ by hundredths of a degree in bond angle. In astrochemistry, the rigid rotor formula is the direct basis for identifying molecules in interstellar clouds and circumstellar envelopes: rotational transition frequencies computed from laboratory-measured $B$ values are matched against radio telescope observations, and to date well over 300 distinct molecular species — from simple diatomics like CO to complex prebiotic molecules — have been identified this way, largely through the Green Bank Telescope, ALMA, and related facilities. In atmospheric and environmental chemistry, rotational lines of trace gases (ozone, nitrous oxide, various volatile organics) are used for remote sensing from ground-based and satellite microwave instruments, exploiting exactly the $2B(J+1)$ spacing pattern derived here to identify and quantify pollutant and greenhouse gas concentrations from their pressure-broadened rotational absorption profiles. In pharmaceutical and materials chemistry, rotational spectroscopy combined with isotopic substitution (as in the ¹⁴N₂/¹⁵N₂ comparison above) is a standard technique for determining precise gas-phase structures of drug-like molecules, hydrogen-bonded clusters, and van der Waals complexes, feeding directly into force-field parameterization for molecular dynamics and structure-based drug design. The centrifugal distortion corrections that extend the rigid rotor model are essential in these applications, since real bonds stretch measurably as $J$ increases, providing an independent spectroscopic handle on bond stiffness and force constants that complements vibrational spectroscopy.

Ask the tutor: How does centrifugal distortion modify the rigid rotor energy levels, and how is the distortion constant $D$ measured from a real spectrum?

Picture two point masses, $m_1$ and $m_2$, joined by a perfectly rigid, massless rod of length $r$. Classically this dumbbell can spin about any axis perpendicular to the rod, and the energy of rotation is $\frac{1}{2}I\omega^2$ where $I$ is the moment of inertia and $\omega$ the angular velocity. Quantum mechanically, angular velocity is not a good quantum number — angular momentum is. The rigid rotor Hamiltonian contains no potential energy term at all (the bond length is fixed by assumption, so there is no bond-stretching potential to worry about); it is pure kinetic energy of rotation, and the entire problem is a statement about angular momentum operators. The crucial simplification is the two-body-to-one-body reduction. A free diatomic molecule has six degrees of freedom: three for the center of mass and three describing the relative position of the two nuclei. Because there is no external force, the center-of-mass motion is simply free-particle translation and separates out completely. What remains is the relative coordinate vector $\vec{r}$ connecting the two atoms, and the Schrödinger equation for this relative motion has exactly the form of a single particle of mass $\mu$ moving in the potential $V(r)$ that binds the two atoms. If the bond is treated as perfectly rigid, $V(r)$ is an infinitely steep well at $r = r_e$, freezing the radial coordinate and leaving only the two angular coordinates $\theta$ and $\phi$ free. That is precisely a particle confined to a sphere of radius $r_e$. The first figure below shows this reduction directly: two orbiting masses collapse into one effective particle on a sphere, with the reduced mass absorbing all the information about how "heavy" each atom is relative to the other.

Because there is no radial freedom left, the Hamiltonian for the relative motion is purely the angular kinetic energy operator, and its eigenfunctions are the spherical harmonics $Y_J^m(\theta,\phi)$ familiar from the particle-on-a-sphere problem and from atomic orbital shapes. The quantum number $J$ (used here instead of $l$ to signal "rotational" rather than "orbital" angular momentum) labels the total angular momentum, and $m_J$ labels its projection along a chosen axis, running from $-J$ to $+J$ in integer steps. Every rotational state with quantum number $J$ is therefore $(2J+1)$-fold degenerate: the molecule can point its angular momentum vector in $2J+1$ different directions relative to a lab axis, all with identical energy in the absence of an external field. The energy level diagram is the second essential piece of intuition. Unlike a harmonic oscillator, whose levels are evenly spaced, the rotor's levels spread apart quadratically with $J$, because the eigenvalue is proportional to $J(J+1)$ rather than to $J$ itself.

The starting point is the reduced two-body Hamiltonian. For nuclei of mass $m_1, m_2$ at separation $r$, define the reduced mass and moment of inertia:

where: $\mu$ is the effective single-particle mass carrying all information about the two real masses, and $I$ is the moment of inertia the rigid dumbbell would have if it were replaced by that single particle at distance $r$ from the rotation center. Reading: heavier atoms and longer bonds both increase $I$; a light atom paired with a heavy one (like H–Cl) gives $\mu$ close to the light atom's mass, so hydrogen effectively dominates the dynamics. The Hamiltonian for the relative motion, with the radial coordinate frozen at $r = r_e$, is purely rotational kinetic energy expressed through the angular momentum operator $\hat{L}^2$:

where: $\hat{L}^2$ is the total angular momentum operator, identical in form to the one governing atomic orbitals, and $I$ is the fixed moment of inertia defined above. Reading: this Hamiltonian has no explicit $\theta,\phi$-dependent potential; every term is angular momentum, so its eigenfunctions must be the same as those of $\hat{L}^2$ itself — the spherical harmonics. Since $Y_J^{m_J}(\theta,\phi)$ are eigenfunctions of $\hat{L}^2$ with eigenvalue $\hbar^2 J(J+1)$, the rigid rotor eigenvalue equation gives the energy directly:

where: $J$ is the rotational quantum number, and $\hbar^2/2I$ sets the overall energy scale. Reading: energies grow as $J(J+1)$, not linearly in $J$ — level spacing widens as $J$ increases, in sharp contrast to the evenly spaced vibrational ladder. Spectroscopists almost always work in the rotational constant $B$, expressed in wavenumbers (cm⁻¹), folding $\hbar^2/2I$ together with the speed of light and Planck's constant:

where: $h$ is Planck's constant, $c$ the speed of light, and $I$ the moment of inertia in SI units; $B$ absorbs all the molecule-specific structural information into one number extracted directly from experiment. Reading: once $B$ is measured from a spectrum, $I$ and hence $r_e$ can be backed out algebraically — this is the entire logic of rotational spectroscopy as a structural tool. Each level $J$ carries degeneracy $2J+1$, coming from the magnetic quantum number $m_J = -J,\dots,+J$:

where: $g_J$ counts the number of independent spatial orientations of the angular momentum vector that share the same energy. Reading: this degeneracy directly multiplies line intensities in rotational spectra and dominates the rotational partition function and heat capacity at moderate temperatures. Finally, the selection rule for electric-dipole–allowed pure rotational transitions, derivable from the transition dipole integral over spherical harmonics, is:

where: the nonzero permanent dipole moment is required for the molecule to couple to the oscillating electric field of microwave radiation at all. Reading: homonuclear diatomics like N₂ or O₂ have no permanent dipole and are microwave-silent in pure rotation, while heteronuclear diatomics like CO or HCl show rich rotational spectra.

Example 1 — Foundational: rotational constant and bond length of ¹H³⁵Cl. Take the experimentally known rotational constant of HCl, $B = 10.59\ \text{cm}^{-1}$, and the atomic masses $m_H = 1.008\ \text{u}$, $m_{Cl} = 34.97\ \text{u}$ (asymmetric, non-round values, as they must be for a real isotopologue). First compute the reduced mass in atomic mass units, then convert to kilograms using $1\ \text{u} = 1.6605\times10^{-27}\ \text{kg}$:

Now invert the rotational constant formula to get the moment of inertia. Starting from $B = h/(8\pi^2 c I)$,

Insert $h = 6.626\times10^{-34}\ \text{J s}$, $c = 2.998\times10^{10}\ \text{cm/s}$ (used in cm units to match $B$ in cm⁻¹), and $B=10.59\ \text{cm}^{-1}$:

Evaluate the denominator: $8\pi^2 = 78.96$; then $78.96 \times 2.998\times10^{10} = 2.3671\times10^{12}$; then $\times 10.59 = 2.5067\times10^{13}$.

Finally solve $I = \mu r_e^2$ for the bond length:

So $r_e \approx 127.5\ \text{pm}$, matching the accepted HCl bond length of about 127.5 pm essentially exactly — a striking confirmation that a purely spectroscopic measurement (line spacing in the microwave) reproduces a structural bond length to sub-picometer precision. Limiting case check: if the two atoms had equal mass ($m_1=m_2=m$), the formula gives $\mu = m/2$, the familiar reduced mass of two equal masses, and the derivation of $E_J$ above goes through unchanged — the physics does not depend on the masses being different, only the numerical value of $\mu$ does.

Example 2 — Applied: predicting the microwave spectrum of ¹²C¹⁶O. Carbon monoxide has $r_e = 112.8\ \text{pm}$, $m_C = 12.000\ \text{u}$, $m_O = 15.995\ \text{u}$. Predict the frequency (in GHz) of the $J=2\to3$ absorption line. Compute the reduced mass:

Compute the moment of inertia using $r_e = 1.128\times10^{-10}\ \text{m}$:

Compute $B$ in cm⁻¹ via $B = h/(8\pi^2 c I)$:

Denominator: $78.96 \times 2.998\times10^{10} = 2.3671\times10^{12}$; times $1.4485\times10^{-46} = 3.4291\times10^{-34}$.

This matches the well-known experimental value of CO, $B \approx 1.9313\ \text{cm}^{-1}$, to within rounding of the input bond length. Now use the transition formula for $J=2\to3$ ($J$ in the formula refers to the lower state, so $J=2$):

Convert to frequency using $\nu = c\tilde\nu$ with $c = 2.998\times10^{10}\ \text{cm/s}$:

This falls squarely in the microwave/millimeter-wave range routinely used in radio astronomy — indeed, rotational transitions of CO near this frequency region are the primary tracer used to map molecular clouds and star-forming regions in the interstellar medium. Limiting case check: as $J\to\infty$, $\tilde\nu = 2B(J+1)$ grows without bound in the ideal rigid rotor — but real molecules show this trend bending over and eventually reversing at very high $J$ due to centrifugal distortion (see below), which the perfectly rigid model cannot capture since it assumes $r$ never changes.

Example 3 — Exam/competition level: rotational Raman spectrum and isotope effect. Unlike infrared absorption, rotational Raman scattering obeys the selection rule $\Delta J = 0, \pm2$ (because the polarizability, not the dipole moment, drives the transition, and polarizability transforms as $Y_2^m$, an even-parity operator). Consider ¹⁴N₂, which has no permanent dipole and is microwave-silent, but is rotational-Raman active. Given $r_e = 109.8\ \text{pm}$ and $m(^{14}\text{N}) = 14.003\ \text{u}$, find the Raman shift (in cm⁻¹) of the $J=4 \to J=6$ (Stokes) line, then repeat for ¹⁵N₂ ($m = 15.000\ \text{u}$) and compute the isotope shift. For a homonuclear diatomic, $\mu = m/2$ exactly:

For the Raman transition $J \to J+2$, the energy change is

Expand: $(J+2)(J+3) - J(J+1) = (J^2+5J+6) - (J^2+J) = 4J+6$.

For $J=4$: $4J+6 = 22$, so

Now for ¹⁵N₂:

This roughly 3 cm⁻¹ isotope shift, despite an identical bond length assumed for both isotopologues (a good approximation since isotopic substitution barely perturbs the electronic potential energy surface), is exactly the kind of signature used to confirm isotopic identity and abundance ratios in astrochemical and atmospheric spectroscopy — the entire shift traces back to the change in $\mu$ alone, since $r_e$ is essentially isotope-independent. Limiting case check: setting $J=0$ in the Stokes formula gives $\tilde\nu = 6B$, the smallest allowed rotational Raman shift, consistent with the fact that $J=0\to2$ is the lowest-order Raman line possible (there is no $J=0\to1$ Raman line, since $\Delta J=\pm1$ is Raman-forbidden).

The rigid rotor model itself is a direct mathematical consequence of angular momentum quantization, but its experimental realization — microwave rotational spectroscopy — was pioneered in the 1930s–1940s, blossoming immediately after World War II when microwave generator technology developed for radar became available to chemists and physicists.

Townes's microwave spectroscopy program, and parallel work by groups such as those of Walter Gordy in the United States, transformed rotational spectroscopy from a theoretical curiosity following the successful quantum mechanical treatment of angular momentum into a routine, industrial-strength structural technique — one now extended to interstellar radioastronomy, where rotational transitions of species like CO, HCN, and complex organic molecules are used to map the chemistry of star-forming clouds.