Selection rules are conditions that determine whether a transition between states of a quantum system can occur through a specified interaction. In quantum mechanics, they usually express when a transition matrix element must vanish because of symmetry or the mathematical form of the interaction. They are fundamental to spectroscopy, where they help explain why some spectral lines are strong, others weak, and others absent. A transition described as forbidden is forbidden under particular assumptions; it may still occur through a weaker interaction, a higher-order process, or a mechanism that relaxes those assumptions.
Mathematical basis
Let and denote the initial and final states, represented by wave functions, and let be the operator responsible for the transition. The relevant transition amplitude contains the matrix element
If symmetry requires , the transition is forbidden for that operator. A nonzero matrix element means that the transition is allowed, but does not by itself establish that the transition is strong or experimentally observable.
In time-dependent perturbation theory, a weak time-dependent contribution to the Hamiltonian produces a first-order transition amplitude
where is the reduced Planck constant. The transition probability depends on the squared magnitude of this amplitude.
For transitions into a continuum of final states, Fermi’s golden rule often gives a rate of the form
with energy conservation understood and the density of accessible final states. Selection rules constrain ; energy conservation, state populations, and the availability of final states impose additional conditions.
Symmetry and conservation laws
Selection rules arise because states and transition operators transform in definite ways under symmetry operations. For example, if an integrand changes sign under a symmetry operation while the integration domain remains unchanged, its integral is zero.
In the language of group theory, a transition can be nonzero only if the product
contains the totally symmetric representation. Here , , and describe the symmetry properties of the initial state, final state, and transition operator.
This formulation encompasses angular-momentum rules, parity rules, and molecular point-group rules. Symmetry establishes necessary conditions for a nonzero amplitude, not a guarantee: additional cancellations can make an otherwise allowed matrix element vanish.
Electric-dipole transitions in atoms
For an atom interacting with radiation whose wavelength is much larger than the atom, the leading interaction is usually the electric-dipole interaction,
where is the electric-dipole operator and is the radiation’s electric field. Transitions produced by this term are called electric-dipole, or E1, transitions.
Angular momentum
For states labeled by total angular momentum and its projection , the E1 rules are
and
The allowed magnetic-sublevel transitions depend on the polarization of the radiation relative to the chosen quantization axis. In a spherical-component description, the component couples states satisfying .
These conditions are consequences of the Wigner–Eckart theorem and the vector character of the dipole operator. They also express the angular-momentum constraints associated with absorption or emission of a photon.
Parity
The electric-dipole operator is odd under spatial inversion. Consequently, an E1 transition must connect states of opposite parity:
This statement assumes that the states possess definite parity. An external environment that mixes opposite-parity states can give intensity to transitions that would otherwise be parity-forbidden.
For a one-electron atom in a central potential, the orbital rule is
where labels the electron’s orbital angular momentum. There is no general electric-dipole restriction requiring a particular change in the principal quantum number .
Spin and coupling approximations
The nonrelativistic electric-dipole operator does not act on spin. In the approximation of Russell–Saunders, or , coupling, the usual additional rules are
with excluded.
These rules depend on the quality of the -coupling approximation. Spin–orbit interaction can mix states of different , allowing weak intercombination lines that violate the approximate rule. The total- and parity rules remain applicable when those quantum numbers remain well defined.
Higher multipoles and forbidden lines
The electric-dipole interaction is only the leading term in a multipole expansion of the radiation–matter interaction. A transition forbidden as E1 may be allowed as magnetic-dipole (M1), electric-quadrupole (E2), or a still higher multipole.
For a photon multipole of order , angular momentum requires
The parity conditions are
Thus E1 transitions change parity, whereas M1 and E2 transitions preserve it. These are necessary conditions; further restrictions can arise from electronic configuration, spin, or the detailed operator.
In the long-wavelength regime, higher-multipole transitions are generally weaker than comparable E1 transitions. States whose rapid decay channels are forbidden can therefore be metastable.
Forbidden lines are important in astrophysics. In sufficiently dilute gases, an excited state can survive long enough to emit through a slow radiative channel rather than losing its excitation through a collision. The term “forbidden line” consequently describes a transition classification, not an impossibility.
Molecular spectroscopy
In a molecule, electronic, vibrational, and rotational motion introduce additional quantum numbers and symmetries. The applicable rules depend on the type of spectroscopy and on approximations used to separate these motions.
Rotational transitions
For a linear rigid rotor, pure electric-dipole rotational absorption or emission requires a permanent molecular dipole moment and obeys
Consequently, a free homonuclear diatomic molecule has no ordinary electric-dipole pure rotational spectrum. Other mechanisms, including electric-quadrupole transitions and collision-induced absorption, can produce much weaker rotational features.
Symmetric and asymmetric tops have additional rotational labels. Their rules depend on which molecular-axis components of the dipole moment are nonzero. Nuclear exchange symmetry can also restrict the rotational levels available to identical nuclei.
Vibrational infrared transitions
A normal mode is active in infrared spectroscopy when the molecular dipole moment changes to first order along its normal coordinate :
For a harmonic oscillator with a dipole moment linear in , the vibrational rule is
Anharmonicity and higher-order terms in the dipole-moment expansion allow overtone and combination bands, usually with lower intensities.
Rotational structure accompanying a vibrational transition produces branches conventionally labeled P, Q, and R for , respectively. Whether a Q branch exists depends on the molecular structure and the symmetry of the transition.
Raman transitions
In Raman spectroscopy, the leading interaction involves the molecular polarizability rather than the permanent dipole moment. A vibration is Raman-active when
where is the polarizability tensor.
For ordinary pure rotational Raman scattering by a linear molecule, the shifted rotational lines obey . The contribution belongs to the unshifted Rayleigh component.
For molecules with an inversion center, the mutual-exclusion rule states that a fundamental normal mode cannot be both electric-dipole infrared-active and ordinary Raman-active. Infrared-active modes have odd inversion symmetry, whereas Raman-active modes have even inversion symmetry. This rule does not apply generally to molecules without inversion symmetry.
Magnetic resonance
In nuclear magnetic resonance, an oscillating magnetic field induces transitions between nuclear spin states. For an isolated spin in the usual high-field, first-order description, a transverse magnetic-dipole interaction gives
For ordinary absorption in this regime, only the transition to the higher-energy state is driven resonantly.
In electron paramagnetic resonance, the simplest allowed transitions satisfy
These rules assume states that are approximately separable into electron and nuclear spin components. Hyperfine mixing, anisotropic interactions, and more complex excitation sequences can make nominally forbidden transitions observable. Multiple-quantum coherence in magnetic resonance must be distinguished from a directly allowed one-photon transition.
Solids and nuclear transitions
Optical transitions in solids
In crystals, selection rules depend on the symmetry of electronic states and of the optical transition operator. In the electric-dipole approximation, photon momentum is often negligible on the scale of the Brillouin zone, so optical transitions approximately conserve crystal momentum:
modulo a reciprocal lattice vector.
In a semiconductor with an indirect band gap, a transition between the valence-band maximum and conduction-band minimum generally requires a phonon, disorder, or another source of momentum transfer. Symmetry and polarization further determine whether a particular interband transition has a nonzero optical matrix element.
Nuclear transitions
Electromagnetic transitions of an atomic nucleus obey angular-momentum and parity rules analogous to atomic multipole transitions.
For beta decay, the relevant transition operators arise from the weak interaction. In the leading allowed approximation:
- Fermi transitions have , including , with no parity change.
- Gamow–Teller transitions have , excluding , also with no parity change.
“Forbidden” beta transitions involve higher-order terms with different angular-momentum and parity conditions. The terminology is analogous to optical spectroscopy, but the operators and physical mechanisms differ.
Interpretation and limitations
Selection rules must always be stated together with the interaction and assumptions to which they apply.
- Exact and approximate rules differ. A rule derived from an exact symmetry remains valid while that symmetry holds. Rules based on approximate quantum numbers can fail when neglected interactions mix the states.
- Allowed does not mean intense. Small spatial overlap, cancellation, low initial-state population, and limited phase space can suppress an allowed transition.
- Forbidden does not mean absent. Higher multipoles, multiphoton processes, collisions, external fields, and symmetry-breaking environments can provide alternative transition amplitudes.
- Experimental geometry matters. Polarization and observation direction can suppress particular components even when the transition is allowed in another geometry.
- Transition order matters. Two-photon transitions involve products of interaction operators and sums over intermediate states, so they obey different rules from one-photon transitions. The hydrogen decay is a standard example: it is forbidden as a one-photon E1 transition but proceeds predominantly by two-photon emission in an isolated atom.
The observation of an apparently forbidden feature can therefore reveal state mixing, local symmetry breaking, or an interaction omitted from the simplest model.