Angular Systems and Rotors

Angular systems make geometry part of the quantum problem. A coordinate may be periodic rather than unbounded, a wavefunction may live on a curved configuration space, and the probability measure may contain a geometric Jacobian. These facts determine the operator domain, the allowed quantum numbers, and the interpretation of angular plots.

This chapter begins with a particle on a circle, moves to a particle on a sphere, and then reinterprets the spherical model as a rigid molecular rotor. It also connects spherical-harmonic wavefunctions to rotational spectra and shows how an external field converts degeneracy into a diagnostic of broken symmetry.

Scope and ownership

The chapter owns the coordinate-space wave mechanics of the following ideal models:

• a particle constrained to the circle $S^1$;
• a particle constrained to the sphere $S^2$;
• a spinless linear rigid rotor with one fixed moment of inertia;
• the first interpretation of angular probability distributions;
• the ideal pure rotational line ladder;
• the first weak-field picture of Stark and Zeeman splitting.

The mathematical theory of spherical harmonics belongs to Spherical Harmonics. The operator algebra of rotations, angular-momentum addition, tensor operators, and representation theory belongs to Symmetry, Angular Momentum, and Spin. Detailed molecular structure, vibration-rotation coupling, nuclear-spin statistics, hyperfine structure, and precision spectroscopy require molecular-physics treatments beyond the ideal rotor. Systematic weak-field calculations belong to perturbation theory.

The baseline rotor is deliberately scalar and nonrelativistic. It does not silently include spin, electronic angular momentum, centrifugal distortion, or exchange restrictions. Those effects refine the model only after its configuration space, Hamiltonian, and symmetry assumptions have been stated.

Compact configuration spaces

The particle on a line has an unbounded configuration coordinate. Angular models instead use compact spaces:

$$S^1 = \{e^{i\theta}:0\leq\theta\lt2\pi\},$$

and

$$S^2 = \{\mathbf n\in\mathbb R^3:\lvert\mathbf n\rvert=1\}.$$

Compactness has two immediate consequences. First, coordinates cover the space with identifications or coordinate singularities: $heta=0$ and $heta=2\pi$ are the same point on a circle, while the azimuthal angle is undefined at the poles of a sphere. Second, a regular self-adjoint kinetic operator on a compact space has a discrete spectrum with finite-dimensional eigenspaces.

The governing data are not merely a differential expression. One must specify

$$(\mathcal Q,d\mu,\mathcal H,D(H),H),$$

where $\mathcal Q$ is configuration space, $d\mu$ its invariant measure, $\mathcal H=L^2(\mathcal Q,d\mu)$, and $D(H)$ contains the global boundary or regularity conditions. The same local differential equation with a different domain can describe a different quantum theory.

For the standard free angular models, the comparison is

Model	Configuration space and measure	Hamiltonian	Labels	Energy
Particle on a ring	$S^1$, $d\theta$	$-\hbar^2\partial_\theta^2/(2I)$	$m\in\mathbb Z$	$\hbar^2m^2/(2I)$
Particle on a sphere	$S^2$, $d\Omega=\sin\theta\,d\theta\,d\phi$	$-\hbar^2\Delta_{S^2}/(2I)$	$\ell,m$	$\hbar^2\ell(\ell+1)/(2I)$
Linear rigid rotor	$S^2$, $d\Omega$	$\mathbf J^2/(2I)$	$J,M$	$\hbar^2J(J+1)/(2I)$

The particle-on-a-sphere and linear-rotor rows are mathematically isospectral in this idealization. Their physical meanings differ: one tracks the position of a constrained particle, while the other tracks the orientation of an extended object.

The particle on a ring

Let a particle of mass $M$ move at fixed radius $R$. Its moment of inertia is

$$I=MR^2,$$

and the standard Hamiltonian is

$$H = -\frac{\hbar^2}{2I} \frac{d^2}{d\theta^2}.$$

The physical identification of the endpoints requires periodicity. For the standard domain,

$$\psi(\theta+2\pi)=\psi(\theta), \qquad \psi'(\theta+2\pi)=\psi'(\theta).$$

The normalized eigenfunctions are the Fourier modes

$$u_m(\theta) = \frac{e^{im\theta}}{\sqrt{2\pi}}, \qquad m\in\mathbb Z,$$

with

$$L_z u_m = \hbar m u_m, \qquad E_m = \frac{\hbar^2m^2}{2I}.$$

The integer label is a global consequence of single-valued periodic states, not a local consequence of the second-order differential equation alone. Except for $m=0$, the levels are paired:

$$E_m=E_{-m}.$$

The two states carry probability around the ring in opposite directions. For a normalized mode, the probability flow past a fixed angle is

$$J_\theta[u_m] = \frac{\hbar}{I} \operatorname{Im} \left( u_m^*\frac{du_m}{d\theta} \right) = \frac{\hbar m}{2\pi I}.$$

Thus equal energy does not imply identical values of every observable. Particle on a Ring develops the domain, completeness, current, and magnetic-flux shift. The flux problem is especially instructive because it can be represented either by a vector potential or by a twisted boundary condition; the gauge-invariant phase around the entire circle is what matters.

The particle on a sphere

For a particle constrained to a sphere of radius $R$, the angular kinetic energy is generated by the Laplace–Beltrami operator:

$$H = -\frac{\hbar^2}{2I}\Delta_{S^2}, \qquad I=MR^2.$$

In standard spherical coordinates,

$$\Delta_{S^2} = \frac{1}{\sin\theta} \frac{\partial}{\partial\theta} \left( \sin\theta \frac{\partial}{\partial\theta} \right) + \frac{1}{\sin^2\theta} \frac{\partial^2}{\partial\phi^2}.$$

Its regular single-valued eigenfunctions are the spherical harmonics:

$$-\Delta_{S^2}Y_\ell^m = \ell(\ell+1)Y_\ell^m,$$

where

$$\ell=0,1,2,\ldots, \qquad m=-\ell,-\ell+1,\ldots,\ell.$$

The energies are

$$E_\ell = \frac{\hbar^2}{2I}\ell(\ell+1),$$

and each fixed-$\ell$ eigenspace has dimension

$$g_\ell=2\ell+1.$$

This degeneracy is enforced by rotational invariance. The Hamiltonian depends on the magnitude of angular momentum but not on its component along an arbitrarily chosen axis. The constant harmonic $Y_0^0=1/\sqrt{4\pi}$ is a zero-energy ground state; no angular gradients means no kinetic energy.

Particle on a Sphere gives the coordinate-space derivation and explains why apparent coordinate singularities at the poles are not physical boundaries.

From a sphere to a rigid rotor

A linear rigid rotor consists of two masses with fixed separation $R_e$. After removing center-of-mass motion, its moment of inertia is

$$I=\mu R_e^2,$$

where $\mu$ is the reduced mass. The remaining coordinate specifies the molecular-axis orientation, and the ideal rotational Hamiltonian is

$$H_0 = \frac{\mathbf J^2}{2I}.$$

Spectroscopy convention usually relabels the sphere quantum numbers as

$$\ell\longrightarrow J, \qquad m\longrightarrow M.$$

The eigenstates and energies are therefore

$$\lvert J,M\rangle \longleftrightarrow Y_J^M(\theta,\phi),$$

and

$$E_J = B_EJ(J+1), \qquad B_E=\frac{\hbar^2}{2I}.$$

The same formula can describe several physical systems, but its assumptions matter. A simple polar heteronuclear diatomic molecule has an oriented molecular axis and fits the $S^2$ picture directly. For identical nuclei, exchange symmetry and nuclear spin can restrict which $J$ sectors occur or how they are weighted. The elementary scalar model omits those restrictions rather than predicting them.

Rigid Rotor is the canonical model page. It separates the ideal rotational degree of freedom from vibration, centrifugal stretching, electronic structure, and spin couplings.

Angular wavefunctions and probability

A wavefunction is normalized with the measure belonging to its configuration space. On a ring,

$$1 = \int_0^{2\pi} \lvert\psi(\theta)\rvert^2,d\theta.$$

On a sphere,

$$1 = \int_0^{2\pi} \int_0^\pi \lvert\psi(\theta,\phi)\rvert^2 \sin\theta,d\theta,d\phi.$$

Consequently, $\lvert\psi\rvert^2$ on the sphere is a density per unit solid angle, not a density per unit $\theta$. The polar-angle marginal is

$$P(\theta) = \sin\theta \int_0^{2\pi} \lvert\psi(\theta,\phi)\rvert^2,d\phi.$$

For the rotationally invariant state $Y_0^0$,

$$P(\theta)=\frac12\sin\theta.$$

The marginal peaks at the equator even though the state has no preferred direction. A band near the equator simply contains more solid angle than an equally wide band near a pole. Uniform direction is uniform in $\cos\theta$, not in $\theta$.

Complex harmonics $Y_\ell^m$ diagonalize $L_z$ and carry a phase $e^{im\phi}$. Real linear combinations are often more convenient for visualizing nodal surfaces or chemical orbitals, but they generally do not have definite $m$. Changing basis inside a degenerate $\ell$ multiplet changes the plotted lobes without changing the eigenspace or its energy.

Spherical Harmonics as Wavefunctions explains amplitudes, phases, nodes, and basis choices. Angular Probability Distributions is the canonical home for measures, marginals, and visualization cautions.

Rotational spectra

For the ideal linear rotor,

$$E_J=B_EJ(J+1).$$

An electric-dipole transition between adjacent rotational levels has energy

$$\Delta E_J = E_{J+1}-E_J = 2B_E(J+1).$$

The ideal absorption lines therefore occur at

$$2B_E,\ 4B_E,\ 6B_E,\ldots$$

They are equally spaced even though the energy levels are not. The angular part of an electric-dipole matrix element gives the familiar rule

$$\Delta J=\pm1,$$

with allowed $\Delta M$ determined by polarization and the chosen quantization axis. A pure rotational electric-dipole spectrum also requires a permanent molecular dipole; the energy ladder alone does not guarantee observable lines.

Since $B_E=\hbar^2/(2I)$, line positions reveal the moment of inertia and, within a stated structural model, a bond length. Line intensities contain additional information about populations, degeneracies, dipole matrix elements, and instrumental conditions. Rotational Spectra keeps these logical layers separate and states where the ideal model begins to fail.

External fields and broken symmetry

The field-free rotor has full rotational symmetry, so its energy is independent of $M$. A uniform external field along the laboratory $z$ axis leaves only axial rotations as a continuous spatial symmetry:

$$SO(3)\longrightarrow SO(2).$$

For a polar rotor with body-fixed dipole magnitude $d$ in a static electric field $\boldsymbol{\mathcal E}=\mathcal E\hat{\mathbf z}$,

$$H = B_E\frac{\mathbf J^2}{\hbar^2} -d\mathcal E\cos\theta.$$

The interaction preserves $M$ but couples neighboring field-free multiplets:

$$\Delta J=\pm1, \qquad \Delta M=0.$$

Thus $M$ remains an exact label for this axially symmetric Hamiltonian, while $J$ does not. Parity makes the diagonal matrix element of $\cos\theta$ vanish in an isolated field-free rotor state, so the simple $J=0$ level has no first-order Stark shift. Its leading shift is

$$\Delta E_{0,0}^{(2)} = -\frac{d^2\mathcal E^2}{6B_E}.$$

A magnetic coupling can behave differently because it may distinguish $M$ from $-M$. The precise Zeeman Hamiltonian depends on the actual magnetic moment and cannot be inferred from the rigid-rotor spectrum alone.

Rotor in External Fields: First Encounter develops this symmetry logic and the weak electric-field example. The durable method is to identify the residual symmetry, compute which operators commute with the full Hamiltonian, and only then decide which quantum numbers remain good.

Model comparison

Question	Ring	Sphere	Linear rotor	Rotor in an axial electric field
Configuration variable	One periodic angle	Direction on $S^2$	Molecular-axis direction	Molecular-axis direction relative to field
Exact continuous symmetry	$SO(2)$	$SO(3)$	$SO(3)$	$SO(2)$
Natural exact labels	$m$	$\ell,m$	$J,M$	$M$ plus a field-dressed level label
Generic degeneracy	$m\leftrightarrow-m$	$2\ell+1$	$2J+1$	Reduced; often $M\leftrightarrow-M$ remains
Main global issue	Periodic domain	Regularity and invariant measure	Model assumptions and orientation	Level mixing and residual symmetry

Three distinctions prevent many errors:

1. A coordinate label is not itself a probability density; the measure completes the statement.
2. A familiar quantum number remains exact only if its operator commutes with the full Hamiltonian.
3. Equal differential expressions do not erase physical differences among constrained particles, molecular orientations, and field-dressed systems.

Reading route

For a first pass, follow this sequence:

1. Study Particle on a Ring to see how topology and the Hamiltonian domain quantize a generator.
2. Move to Particle on a Sphere for curved configuration space, spherical harmonics, and rotational degeneracy.
3. Reinterpret the same angular equation in Rigid Rotor.
4. Use Spherical Harmonics as Wavefunctions and Angular Probability Distributions to learn what angular plots and marginals actually mean.
5. Continue to Rotational Spectra and then Rotor in External Fields: First Encounter for observable line structure and symmetry breaking.

Readers already comfortable with angular momentum may begin with the rigid rotor, but the ring remains the cleanest place to see why global boundary conditions matter.

Page map

Page	Canonical role
Particle on a Ring	Periodic domain, integer angular momentum, spectrum, current, and flux preview
Particle on a Sphere	Angular Laplacian, spherical-harmonic eigenstates, and $2\ell+1$ degeneracy
Rigid Rotor	Molecular moment of inertia, rotational Hamiltonian, and ideal spectrum
Spherical Harmonics as Wavefunctions	Normalization, phases, nodes, real bases, and plots
Rotational Spectra	Rotational constants, dipole selection rules, line positions, and model limits
Angular Probability Distributions	Circle and sphere measures, marginals, and visualization pitfalls
Rotor in External Fields: First Encounter	Residual symmetry, field-induced mixing, and degeneracy breaking

Common mistakes

• Solving the ring equation without imposing periodicity.
• Treating $m$ as nonnegative because the ring energy depends on $m^2$.
• Concluding that the $m$ and $-m$ ring states are physically identical because they are degenerate.
• Using the flat measure $d\theta\,d\phi$ on a sphere.
• Interpreting the coordinate poles as physical boundaries.
• Saying that an isotropic state is uniform in the coordinate $\theta$.
• Confusing $\lvert Y_\ell^m\rvert^2$ with a three-dimensional radial probability density.
• Treating real and complex spherical harmonics as different energy levels rather than basis choices in one multiplet.
• Using $I=MR^2$ for a two-body rotor without reducing the center-of-mass motion.
• Assuming the ideal scalar rotor includes vibration, spin, centrifugal distortion, or nuclear exchange effects.
• Inferring observable electric-dipole lines from the energy ladder without checking the dipole matrix element.
• Assuming a selected field axis leaves $J$ exact merely because $J$ labels the field-free states.
• Applying a generic Zeeman shift without specifying the magnetic moment that couples to the field.

References

• D. J. Griffiths and D. F. Schroeter, Introduction to Quantum Mechanics, 3rd ed., Cambridge University Press, 2018.
• C. Cohen-Tannoudji, B. Diu, and F. Laloë, Quantum Mechanics, Wiley, 1977.
• J. J. Sakurai and J. Napolitano, Modern Quantum Mechanics, 3rd ed., Cambridge University Press, 2020.
• R. N. Zare, Angular Momentum: Understanding Spatial Aspects in Chemistry and Physics, Wiley, 1988.
• P. W. Atkins and R. S. Friedman, Molecular Quantum Mechanics, 5th ed., Oxford University Press, 2011.
• G. Herzberg, Molecular Spectra and Molecular Structure I: Spectra of Diatomic Molecules, 2nd ed., Van Nostrand, 1950.
• C. H. Townes and A. L. Schawlow, Microwave Spectroscopy, Dover, 1975.

Exercises

1. Starting from the periodic ring eigenfunction $u_m=e^{im\theta}/\sqrt{2\pi}$, show that $m$ must be an integer and compare the probability currents of $u_m$ and $u_{-m}$.

Solution

Periodicity requires

$$u_m(\theta+2\pi) = u_m(\theta),$$

so

$$e^{i2\pi m}=1.$$

Therefore $m\in\mathbb Z$. The angular probability current is

$$J_\theta[u_m] = \frac{\hbar}{I} \operatorname{Im} \left( u_m^*\frac{du_m}{d\theta} \right) = \frac{\hbar m}{2\pi I}.$$

Hence

$$J_\theta[u_{-m}] = -J_\theta[u_m],$$

even though $E_{-m}=E_m$. The degeneracy pairs opposite circulation directions.

2. Count the number of particle-on-a-sphere states with $0\leq\ell\leq L$, where $L$ is a nonnegative integer.

Solution

For fixed $\ell$, the allowed values of $m$ give $2\ell+1$ states. Therefore

$$N(L) = \sum_{\ell=0}^{L}(2\ell+1) = (L+1)^2.$$

This count concerns a truncated collection of angular eigenspaces. It is not the degeneracy of one energy level; the degeneracy at fixed $\ell$ is only $2\ell+1$.

3. The state $Y_0^0=1/\sqrt{4\pi}$ is isotropic. Derive its polar marginal $P(\theta)$ and explain why its maximum at $\theta=\pi/2$ does not indicate equatorial localization.

Solution

The marginal includes the spherical measure:

$$\begin{aligned} P(\theta) &= \sin\theta \int_0^{2\pi} \left\lvert Y_0^0\right\rvert^2,d\phi \\ &= \sin\theta \int_0^{2\pi} \frac{1}{4\pi},d\phi \\ &= \frac12\sin\theta. \end{aligned}$$

The state has constant probability per unit solid angle. A fixed-width interval in $\theta$ near the equator contains more solid angle than one near a pole, so the coordinate marginal is larger there. No spatial direction is preferred.

4. For an ideal rigid rotor, derive the spacing between successive allowed absorption lines and state two reasons why the measured line intensities are not fixed by the energy formula alone.

Solution

With

$$E_J=B_EJ(J+1),$$

and the electric-dipole rule $\Delta J=+1$ for absorption,

$$\Delta E_J = E_{J+1}-E_J = 2B_E(J+1).$$

Neighboring line energies differ by

$$\Delta E_{J+1}-\Delta E_J = 2B_E.$$

Thus the ideal line ladder is equally spaced. Intensities also depend on thermal populations and dipole matrix elements. They may additionally depend on sublevel degeneracies, polarization, and experimental response. A molecule with no permanent electric dipole does not display the ideal pure rotational electric-dipole series even though it has rotational energy levels.