Skip to content

Scattering Applications

Scattering is symmetry plus boundary conditions. The incoming and outgoing states are not square-integrable bound states, but the Hamiltonian can still commute with rotations, parity, spin operators, channel projectors, or total angular momentum. Those symmetries organize the scattering matrix and reduce the number of independent amplitudes.

This page is an application map. The canonical calculations live in the approximation and scattering volume. Here the goal is to identify which symmetry label controls which scattering simplification.

A symmetry-first scattering workflow is:

Hamiltonian and asymptotic channels
-> conserved angular momentum, spin, parity, and internal labels
-> block structure of the scattering matrix
-> amplitudes, phase shifts, and cross sections
-> selection rules or polarization observables

Always separate two questions:

  • What states or channels are allowed by energy and boundary conditions?
  • Which transitions among those channels are forbidden or related by symmetry?

For elastic scattering from a central potential V(r)V(r), the Hamiltonian is rotationally invariant:

[H,L2]=0,[H,Lz]=0.[H,L^2]=0, \qquad [H,L_z]=0.

The incoming plane wave can be decomposed into angular-momentum channels. For an incoming wave along zz,

eikz=∑ℓ=0∞(2ℓ+1)iℓjℓ(kr)Pℓ(cos⁡θ).e^{ikz} = \sum_{\ell=0}^{\infty} (2\ell+1)i^\ell j_\ell(kr) P_\ell(\cos\theta).

Rotational symmetry makes different ℓ\ell sectors independent in the simplest one-channel central problem. The scattering information is then encoded in phase shifts δℓ\delta_\ell, or equivalently

Sℓ=e2iδℓS_\ell=e^{2i\delta_\ell}

for elastic scattering in each partial wave.

With the common convention used in the scattering volume,

f(θ)=12ik∑ℓ=0∞(2ℓ+1)(e2iδℓ−1)Pℓ(cos⁡θ).f(\theta) = \frac{1}{2ik} \sum_{\ell=0}^{\infty} (2\ell+1) \left(e^{2i\delta_\ell}-1\right) P_\ell(\cos\theta).

The canonical derivation is Partial-Wave Expansion, with convention details in Scattering Convention Dictionary.

For a parity-invariant central potential,

V(−r)=V(r),V(-\mathbf r)=V(\mathbf r),

each partial wave has parity

πℓ=(−1)ℓ.\pi_\ell=(-1)^\ell.

Parity can prevent mixing between even and odd sectors when the rest of the interaction also respects parity. In one-dimensional scattering, parity is often used to separate even and odd scattering solutions. In three-dimensional central scattering, it is already built into the partial-wave structure.

The symmetry-side foundations are Parity and Parity as Spatial Inversion.

When spin is present, orbital angular momentum alone may not be conserved. A rotationally invariant spin-dependent interaction can preserve total angular momentum

J=L+S\mathbf J=\mathbf L+\mathbf S

while mixing orbital and spin components inside a fixed JJ sector.

For two spin-1/21/2 particles, the total spin can be singlet or triplet:

S=0orS=1.S=0 \quad\text{or}\quad S=1.

If the interaction is spin-independent, singlet and triplet channels may scatter similarly except for exchange and statistics constraints. If the interaction includes spin–orbit, tensor, magnetic, or spin-exchange terms, different total-spin and total-JJ channels can acquire different phase shifts and mixing angles.

The relevant algebra is in Singlet and Triplet States and Spin–Orbit Coupling. Channel bookkeeping is previewed in Multichannel Scattering.

A scattering matrix is easiest to read after choosing asymptotic channel labels:

∣α;E,J,M,π,…⟩.\lvert \alpha; E,J,M,\pi,\ldots\rangle.

Here α\alpha may include particle species, internal states, thresholds, spin coupling, or other channel labels. If the Hamiltonian is rotationally invariant, JJ and MM constrain the scattering matrix. If it is parity invariant, π\pi constrains it. If a channel quantum number is conserved only approximately, the block structure is approximate too.

Schematically,

⟨β;E,J′,M′,π′∣S∣α;E,J,M,π⟩∝δJJ′δMM′δππ′\langle \beta;E,J',M',\pi'|S|\alpha;E,J,M,\pi\rangle \propto \delta_{JJ'}\delta_{MM'}\delta_{\pi\pi'}

when those symmetries are exact and the basis labels are chosen compatibly. Dynamics still determine the nonzero block entries.

In high-energy and relativistic scattering, it is often more natural to label states by helicity:

h=J⋅p∣p∣.h = \frac{\mathbf J\cdot\mathbf p}{|\mathbf p|}.

Helicity is angular momentum projected along the direction of motion. For massless particles it is especially robust; for massive particles it is frame-dependent. This page does not develop relativistic scattering. The bridge is From Angular Momentum to Helicity.

TaskStart with
Convert rotational invariance into partial wavesPartial-Wave Expansion
Interpret phase shiftsPhase Shifts
Compute cross sectionsDifferential and Total Cross Sections
Check normalization conventionsScattering Convention Dictionary
Add spin channel labelsMultichannel Scattering
Relate angular momentum to helicityFrom Angular Momentum to Helicity
  • Assuming angular momentum conservation means every ℓ\ell is conserved after spin-dependent forces are included.
  • Forgetting that scattering states use flux or delta-function normalization, not bound-state normalization.
  • Mixing scattering-amplitude conventions without tracking factors of kk, 2π2\pi, and signs.
  • Applying parity selection rules while the interaction or external setup breaks parity.
  • Treating helicity as identical to spin projection for massive particles.
  • Interpreting a zero amplitude as purely symmetry-forbidden when it may be a dynamical cancellation.
  1. Why do partial waves decouple for one-channel elastic scattering from a central potential?
Solution

A central potential is rotationally invariant, so the Hamiltonian commutes with L2L^2 and LzL_z. The Hilbert space decomposes into irreducible angular-momentum sectors labeled by ℓ,m\ell,m. In the simplest spinless one-channel central problem, the radial equation depends on ℓ\ell but not on mm, and different ℓ\ell sectors do not mix.

  1. What is the parity of a partial wave with orbital angular momentum ℓ\ell?
Solution

The angular part is a spherical harmonic Yℓm(r^)Y_\ell^m(\hat{\mathbf r}). Under spatial inversion, Yℓm(−r^)=(−1)ℓYℓm(r^)Y_\ell^m(-\hat{\mathbf r})=(-1)^\ell Y_\ell^m(\hat{\mathbf r}). Therefore the partial wave has parity (−1)ℓ(-1)^\ell.

  • R. G. Newton, Scattering Theory of Waves and Particles, 2nd ed., Springer, 1982.
  • J. R. Taylor, Scattering Theory: The Quantum Theory of Nonrelativistic Collisions, Dover, 2006.
  • M. L. Goldberger and K. M. Watson, Collision Theory, Dover, 2004.
  • L. D. Landau and E. M. Lifshitz, Quantum Mechanics: Non-Relativistic Theory, 3rd ed., Pergamon, 1977.
  • J. J. Sakurai and J. Napolitano, Modern Quantum Mechanics, 2nd ed., Cambridge University Press, 2017.