Scattering in quantum mechanics begins with a physical model based on the Schrodinger wave equation for probability amplitude ${\displaystyle \psi }$ : $${\displaystyle -{\frac {\hbar ^{2}}{2\mu }}\nabla ^{2}\psi +V\psi =E\psi }$$  where ${\displaystyle \mu }$  is the reduced mass of two scattering particles and E is the energy of relative motion. For scattering problems, a stationary (time-independent) wavefunction is sought with behavior at large distances (asymptotic form) in two parts. First a plane wave represents the incoming source and, second, a spherical wave emanating from the scattering center placed at the coordinate origin represents the scattered wave:^[2]: 114 $${\displaystyle \psi (r\rightarrow \infty )\sim e^{i\mathbf {k} _{i}\cdot \mathbf {r} }+f(\mathbf {k} _{f},\mathbf {k} _{i}){\frac {e^{i\mathbf {k} _{f}\cdot \mathbf {r} }}{r}}}$$  The scattering amplitude, ${\displaystyle f(\mathbf {k} _{f},\mathbf {k} _{i})}$ , represents the amplitude that the target will scatter into the direction ${\displaystyle \mathbf {k} _{f}}$ .^[3]: 194 In general the scattering amplitude requires knowing the full scattering wavefunction: $${\displaystyle f(\mathbf {k} _{f},\mathbf {k} _{i})=-{\frac {\mu }{2\pi \hbar ^{2}}}\int \psi _{f}^{*}V(\mathbf {r} )\psi _{i}d^{3}r}$$  For weak interactions a perturbation series can be applied; the lowest order is called the Born approximation.

For a spherically symmetric scattering center, the plane wave is described by the wavefunction^[4]

${\displaystyle \psi (\mathbf {r} )=e^{ikz}+f(\theta ){\frac {e^{ikr}}{r}}\;,}$ 

where ${\displaystyle \mathbf {r} \equiv (x,y,z)}$  is the position vector; ${\displaystyle r\equiv |\mathbf {r} |}$ ; ${\displaystyle e^{ikz}}$  is the incoming plane wave with the wavenumber k along the z axis; ${\displaystyle e^{ikr}/r}$  is the outgoing spherical wave; θ is the scattering angle (angle between the incident and scattered direction); and ${\displaystyle f(\theta )}$  is the scattering amplitude.

The dimension of the scattering amplitude is length. The scattering amplitude is a probability amplitude; the differential cross-section as a function of scattering angle is given as its modulus squared,

${\displaystyle d\sigma =|f(\theta )|^{2}\;d\Omega .}$ 

When conservation of number of particles holds true during scattering, it leads to a unitary condition for the scattering amplitude. In the general case, we have^[4]

${\displaystyle f(\mathbf {n} ,\mathbf {n} ')-f^{*}(\mathbf {n} ',\mathbf {n} )={\frac {ik}{2\pi }}\int f(\mathbf {n} ,\mathbf {n} '')f^{*}(\mathbf {n} ,\mathbf {n} '')\,d\Omega ''}$ 

Optical theorem follows from here by setting ${\displaystyle \mathbf {n} =\mathbf {n} '.}$ 

In the centrally symmetric field, the unitary condition becomes

${\displaystyle \mathrm {Im} f(\theta )={\frac {k}{4\pi }}\int f(\gamma )f(\gamma ')\,d\Omega ''}$ 

where ${\displaystyle \gamma }$  and ${\displaystyle \gamma '}$  are the angles between ${\displaystyle \mathbf {n} }$  and ${\displaystyle \mathbf {n} '}$  and some direction ${\displaystyle \mathbf {n} ''}$ . This condition puts a constraint on the allowed form for ${\displaystyle f(\theta )}$ , i.e., the real and imaginary part of the scattering amplitude are not independent in this case. For example, if ${\displaystyle |f(\theta )|}$  in ${\displaystyle f=|f|e^{2i\alpha }}$  is known (say, from the measurement of the cross section), then ${\displaystyle \alpha (\theta )}$  can be determined such that ${\displaystyle f(\theta )}$  is uniquely determined within the alternative ${\displaystyle f(\theta )\rightarrow -f^{*}(\theta )}$ .^[4]

In the partial wave expansion the scattering amplitude is represented as a sum over the partial waves,^[5]

${\displaystyle f=\sum _{\ell =0}^{\infty }(2\ell +1)f_{\ell }P_{\ell }(\cos \theta )}$ ,

where f_ℓ is the partial scattering amplitude and P_ℓ are the Legendre polynomials. The partial amplitude can be expressed via the partial wave S-matrix element S_ℓ (${\displaystyle =e^{2i\delta _{\ell }}}$ ) and the scattering phase shift δ_ℓ as

${\displaystyle f_{\ell }={\frac {S_{\ell }-1}{2ik}}={\frac {e^{2i\delta _{\ell }}-1}{2ik}}={\frac {e^{i\delta _{\ell }}\sin \delta _{\ell }}{k}}={\frac {1}{k\cot \delta _{\ell }-ik}}\;.}$ 

Then the total cross section^[6]

${\displaystyle \sigma =\int |f(\theta )|^{2}d\Omega }$ ,

can be expanded as^[4]

${\displaystyle \sigma =\sum _{l=0}^{\infty }\sigma _{l},\quad {\text{where}}\quad \sigma _{l}=4\pi (2l+1)|f_{l}|^{2}={\frac {4\pi }{k^{2}}}(2l+1)\sin ^{2}\delta _{l}}$ 

is the partial cross section. The total cross section is also equal to ${\displaystyle \sigma =(4\pi /k)\,\mathrm {Im} f(0)}$  due to optical theorem.

For ${\displaystyle \theta \neq 0}$ , we can write^[4]

${\displaystyle f={\frac {1}{2ik}}\sum _{\ell =0}^{\infty }(2\ell +1)e^{2i\delta _{l}}P_{\ell }(\cos \theta ).}$ 

The scattering length for X-rays is the Thomson scattering length or classical electron radius, r₀.

The nuclear neutron scattering process involves the coherent neutron scattering length, often described by b.

A quantum mechanical approach is given by the S matrix formalism.

The scattering amplitude can be determined by the scattering length in the low-energy regime.

• Levinson's theorem
• Veneziano amplitude
• Plane wave expansion