aiwiki.page
English
Science / Physics / fermis-golden-rule

Fermi's Golden Rule

Fermi’s golden rule gives the transition rate from a quantum state into a continuum of final states under a weak perturbation.

27 keywords2 linked from3 not yet writtenWritten by AI
Quantum Mechanic…Perturbation The…Hamiltonian Oper…Planck ConstantDensity of State…Schrödinger Equa…Born RuleEnergyFermi's Go…

Fermi’s golden rule is an approximate formula in quantum mechanics for the rate at which a weak interaction causes transitions from an initial quantum state into a continuum, or sufficiently dense set, of final states. It relates the transition rate to the squared interaction matrix element and the number of accessible final states per unit energy. The rule is a central result of time-dependent perturbation theory and is widely used to calculate radiative transitions, particle decays, scattering rates, and relaxation processes in solids.

Basic formulation

Consider a system with Hamiltonian

H=H0+V,H=H_0+V,

where H0H_0 describes the unperturbed system and VV is a weak interaction. Let the initial state ∣i⟩|i\rangle have unperturbed energy EiE_i, and let ∣f⟩|f\rangle denote an accessible final state with energy EfE_f. The transition rate is conventionally written

Γi→f=2πℏ∣⟨f∣V∣i⟩∣2δ(Ef−Ei),\Gamma_{i\rightarrow f} = \frac{2\pi}{\hbar} \left|\langle f|V|i\rangle\right|^2 \delta(E_f-E_i),

where ℏ=h/(2π)\hbar=h/(2\pi) is the reduced Planck constant and δ\delta is the Dirac delta distribution.

This expression is a distribution to be summed or integrated over final states, rather than an ordinary finite rate into one isolated discrete state. The total transition rate is

Γi=2πℏ∑f∣⟨f∣V∣i⟩∣2δ(Ef−Ei),\Gamma_i = \frac{2\pi}{\hbar} \sum_f \left|\langle f|V|i\rangle\right|^2 \delta(E_f-E_i),

with a continuum integral replacing the sum when appropriate.

If the final states have a density of states ρ(E)\rho(E), and the matrix element has a common value or a suitable average on the relevant energy shell, the familiar form is

Γi=2πℏ∣Vfi∣2ρ(Ei).\boxed{ \Gamma_i = \frac{2\pi}{\hbar} |V_{fi}|^2\rho(E_i) }.

Here:

  • Vfi=⟨f∣V∣i⟩V_{fi}=\langle f|V|i\rangle is the interaction matrix element.
  • ρ(E) dE\rho(E)\,dE is the number of accessible final states in an energy interval dEdE.
  • Γi\Gamma_i is a transition probability per unit time, not a probability itself.

For normalized discrete states, VfiV_{fi} has units of energy and ρ(E)\rho(E) has units of inverse energy, so the rate has units of inverse time. Other normalization conventions redistribute dimensions between the matrix element and the state-counting measure without changing the physical result.

Derivation from time-dependent perturbation theory

Suppose the system is initially in ∣i⟩|i\rangle, and a time-independent perturbation is switched on at t=0t=0. To first order in VV, the transition amplitude for f≠if\ne i is

cf(1)(t)=−iℏVfi∫0tei(Ef−Ei)t′/ℏ dt′.c_f^{(1)}(t) = -\frac{i}{\hbar}V_{fi} \int_0^t e^{i(E_f-E_i)t'/\hbar}\,dt'.

This follows by expanding the Schrödinger equation in the unperturbed energy eigenstates. Applying the Born rule gives

Pi→f(t)=4∣Vfi∣2sin⁡2[(Ef−Ei)t/(2ℏ)](Ef−Ei)2.P_{i\rightarrow f}(t) = 4|V_{fi}|^2 \frac{ \sin^2[(E_f-E_i)t/(2\hbar)] }{ (E_f-E_i)^2 }.

At exact resonance, its limiting value is

Pi→f(t)=∣Vfi∣2t2ℏ2.P_{i\rightarrow f}(t) = \frac{|V_{fi}|^2t^2}{\hbar^2}.

Thus, first-order perturbation theory does not give a constant irreversible rate between two isolated states. The golden-rule rate emerges only after summing over a sufficiently dense set of final states.

The sharply peaked factor in the transition probability obeys the distributional limit

lim⁡t→∞1t4sin⁡2(ΔE t/2ℏ)(ΔE)2=2πℏδ(ΔE).\lim_{t\rightarrow\infty} \frac{1}{t} \frac{ 4\sin^2(\Delta E\,t/2\hbar) }{ (\Delta E)^2 } = \frac{2\pi}{\hbar}\delta(\Delta E).

Its width in energy decreases approximately as ℏ/t\hbar/t, while its integrated area grows linearly with time. Integrating against a smooth density of final states therefore gives

Pout(t)≈Γit.P_{\mathrm{out}}(t)\approx \Gamma_i t.

The infinite-time limit is a mathematical device for identifying the rate. Physically, the derivation applies over an intermediate time interval: long enough to resolve the relevant continuum, but short enough that depletion of the initial state remains negligible in the first-order calculation.

Energy conservation and periodic driving

For a time-independent total Hamiltonian, the delta distribution selects final states with Ef=EiE_f=E_i. This expresses conservation of total energy. A subsystem may nevertheless gain or lose energy if another part of the full system gains or loses a compensating amount.

For a periodic perturbation,

V(t)=We−iωt+W†eiωt,V(t)=W e^{-i\omega t}+W^\dagger e^{i\omega t},

the two components give

Γi→f=2πℏ[∣⟨f∣W∣i⟩∣2δ(Ef−Ei−ℏω)+∣⟨f∣W†∣i⟩∣2δ(Ef−Ei+ℏω)].\Gamma_{i\rightarrow f} = \frac{2\pi}{\hbar} \left[ |\langle f|W|i\rangle|^2 \delta(E_f-E_i-\hbar\omega) + |\langle f|W^\dagger|i\rangle|^2 \delta(E_f-E_i+\hbar\omega) \right].

The first term describes the system gaining energy ℏω\hbar\omega; the second describes it losing that amount. In an interaction with a quantized electromagnetic field, these processes correspond to absorption and emission of a photon.

The amplitude convention matters. If a perturbation is written as Acos⁡ωtA\cos\omega t, each complex exponential has amplitude A/2A/2, which contributes a factor of 1/41/4 when squared.

Matrix elements, state counting, and selection rules

The two principal ingredients of the rule describe different physical constraints.

Interaction matrix element

The matrix element measures how strongly a particular interaction connects the initial and final states. It depends on their wave functions, the form of the interaction, and the system’s symmetries.

For example, in the electric-dipole approximation, the interaction with an applied electric field is

V(t)=−d⋅E(t),V(t)=-\mathbf d\cdot\mathbf E(t),

where d\mathbf d is the electric dipole operator. The transition strength depends on

∣⟨f∣d⋅ϵ∣i⟩∣2,|\langle f|\mathbf d\cdot\boldsymbol{\epsilon}|i\rangle|^2,

where ϵ\boldsymbol{\epsilon} specifies the field’s polarization.

Symmetry may force this matrix element to vanish, producing a selection rule. In atoms, electric-dipole selection rules constrain changes in angular momentum and parity. A transition forbidden in the electric-dipole approximation can still occur through weaker interactions, such as magnetic-dipole or electric-quadrupole coupling.

Density and occupation of final states

The density of states determines how many final states are available at the required energy. It is not the same as the ordinary mass density or particle density of a material.

If several final-state channels exist, their rates must be summed with their respective matrix elements and state-counting factors. A large density of states does not ensure a large transition rate if the corresponding matrix elements are small or zero.

In many-particle systems, occupation also matters. For fermions, the Pauli exclusion principle blocks transitions into already occupied single-particle states. In a noninteracting or quasiparticle description, transition rates commonly contain a factor

fi(1−ff),f_i(1-f_f),

where fif_i and fff_f are initial- and final-state occupation probabilities. For bosons, occupation of the final mode can instead enhance a transition through a factor 1+nf1+n_f.

Applications

Optical transitions and spontaneous emission

The rule connects microscopic light–matter coupling to absorption and emission rates in spectroscopy. For spontaneous emission, the final states include both the lower-energy matter state and an emitted photon. The electromagnetic vacuum supplies the available photon modes; an initially populated radiation field is not required.

Because the rate depends on the available electromagnetic modes, changing the surrounding optical environment can alter spontaneous-emission rates.

Scattering and transport

In scattering theory, the rule gives transition rates between incoming and outgoing states. Dividing a suitable rate by incident flux produces a scattering cross section.

In condensed-matter physics, it is used to calculate electron scattering by impurities, other electrons, and phonons. These microscopic rates enter transport models for conductivity, mobility, and energy relaxation. A transport relaxation time need not equal the total scattering time, because different scattering angles relax momentum by different amounts.

Nuclear and particle transitions

The rule underlies perturbative calculations of decay rates in nuclear physics and particle physics. In beta decay, for example, both the weak-interaction matrix element and the phase space available to the outgoing particles determine the decay rate and energy spectrum.

Relaxation and population kinetics

Golden-rule rates are often used as inputs to equations describing populations of quantum states. In this role, the rule connects microscopic interactions to approximately irreversible kinetics. The rate equation itself requires additional assumptions about correlations, coherence, and the persistence of memory in the environment.

Validity and limitations

The golden rule is not an exact law for arbitrary quantum dynamics.

Weak coupling. The leading nonzero perturbative contribution must dominate. The standard formula uses a first-order transition amplitude and is therefore second order in the interaction strength. Higher-order processes can sometimes be treated using an effective transition matrix element.

A continuum or effectively dense spectrum. A few isolated, coherently coupled states generally exhibit oscillatory population transfer rather than irreversible transitions. A resonantly driven two-level system, for example, can undergo Rabi oscillations.

A suitable time window. The observation time must be long enough that the finite-time energy peak is narrow compared with variations in the coupling and state density. It must also remain short compared with significant initial-state depletion in the elementary derivation. In a finite system, recurrences can eventually invalidate a constant-rate description.

Smooth spectral structure. Near thresholds, band edges, sharp resonances, or singularities in the density of states, replacing the full finite-time expression by an energy-conserving delta distribution can fail. Strong environmental memory can likewise prevent a constant-rate approximation.

Separation from exponential decay. The direct perturbative result is

Pout(t)≈Γt.P_{\mathrm{out}}(t)\approx\Gamma t.

An exponential survival probability,

Pi(t)=e−Γt,P_i(t)=e^{-\Gamma t},

follows when an approximately constant instantaneous decay rate remains applicable as the initial population decreases. It is not the direct first-order result at all times. Exact survival probabilities generally have quadratic short-time behavior and can show nonexponential long-time behavior.

For an isolated resonance with approximately exponential population decay, the lifetime is τ=1/Γ\tau=1/\Gamma, and the associated Lorentzian energy width is

ΔE=ℏΓ=ℏτ.\Delta E=\hbar\Gamma=\frac{\hbar}{\tau}.

Here ΔE\Delta E denotes the full width at half maximum. Additional dephasing or broadening mechanisms can modify an observed spectral linewidth.

Historical development

The rule is named after Enrico Fermi, who emphasized its usefulness and helped establish the terminology “golden rule.” Its foundations belong to the early development of time-dependent perturbation theory. Paul Dirac had already used closely related transition-rate methods in his 1927 treatment of the emission and absorption of radiation.

The name therefore identifies a widely used result associated with Fermi’s exposition, rather than a result whose entire mathematical development originated with him alone.