Effective Range, Shallow Poles, and Universality Windows

The scattering length is only the first coefficient in the low-energy amplitude. Effective range and shape parameters determine how quickly zero-range universality fails and how a shallow pole’s energy and residue depart from their leading forms. A universality window is therefore a momentum interval, not simply the statement that $|a|$ is large.

Required background. Short-Range Scattering Data as Many-Body Inputs fixes the amplitude and pole conventions.

Helpful background. Effective Field Theory as a Controlled Expansion explains order-by-order uncertainty.

Effective-range expansion

For elastic $s$ -wave scattering below the first inelastic threshold,

k\cot\delta_0(k) =-\frac1a+\frac12r_ek^2+P k^4+O(k^6R^5),

and

\mathcal A(k) =\frac{4\pi/m}{k\cot\delta_0(k)-ik}.

Here $r_e$ is the effective range and $P$ a shape parameter with dimension length cubed. Analyticity requires $kR\ll1$ , where $R$ is set by the nearest left-hand singularity or inelastic threshold. A large $|a|$ does not enlarge this analytic radius.

If the effective-range term is treated perturbatively, its relative size near momentum $Q$ is approximately $Q|r_e|$ at unitarity or $|r_e|Q^2/|a^{-1}+iQ|$ more generally. Resumming a truncated effective-range denominator can introduce spurious deep poles; whether to resum is a power-counting decision.

Shallow pole and residue

For a bound state put $k=i\kappa$ , $\kappa>0$ . Through effective range,

-\frac1a-\frac12r_e\kappa^2+\kappa=0.

When $|r_e|/a\ll1$ and $a>0$ ,

\kappa=\frac1a+\frac{r_e}{2a^2} +O\!\left(\frac{r_e^2}{a^3},\frac{P}{a^4}\right).

The derivative of the inverse amplitude fixes the pole residue. In the energy plane it contains the wave-function factor

Z_{\mathrm{pole}}\propto\frac{1}{1-r_e\kappa+O(P\kappa^3)}.

Thus matching the pole energy alone does not fix its coupling to external probes. A residue near a singular denominator also warns that higher effective-range terms cannot be neglected.

Defining the universality window

For a many-body observable characterized by $Q\sim k_F$ , $\lambda_T^{-1}$ , a binding momentum, or a probe momentum, leading zero-range universality requires

Q R\ll1, \qquad Q|r_e|\ll1,

along with any channel-specific conditions. A next-to-leading result retaining $r_e$ must show that the shape correction $|P|Q^3$ is smaller than the required accuracy. Near a narrow resonance $|r_e|$ can be parametrically larger than the microscopic range, making $Q|r_e|$ the limiting parameter.

For finite-range, energy-independent potentials, causality constrains possible effective ranges at a fixed range. Such Wigner-type bounds do not apply unchanged to explicitly energy-dependent or coupled-channel models; the degrees of freedom and regulator must be stated. The analytic and causality restrictions are reviewed in Hammer and Lee 2010, §§II–III.

Uncertainty from the first omitted term

Suppose the amplitude is used through $r_e$ . Vary the shape coefficient over a natural range tied to $R^3$ , or compare fits with and without the $k^4$ term over the actual energy interval. Propagate the resulting amplitude change through the many-body calculation. Quoting only the experimental error on $a$ understates the theory error whenever $Qr_e$ or $QR$ is comparable.

Common pitfalls

Using $|a|\gg R$ as the only check. The observable also has a momentum $Q$ ; universality requires $QR\ll1$ .

Resumming a polynomial beyond its domain. A truncated effective-range expression is local information near threshold and can create unphysical remote poles.

Ignoring the residue. Pole energy and pole strength are independent pieces of shallow-state phenomenology once range corrections enter.

Exercises

Shift the shallow pole

Solve the pole equation perturbatively through first order in $r_e/a$ .

Solution

Write $\kappa=a^{-1}+\delta$ . Keeping only linear order gives $-a^{-1}-(r_e/2)a^{-2}+a^{-1}+\delta=0$ , hence $\delta=r_e/(2a^2)$ .

Compare two error parameters

At unitarity, an observable probes $Q=0.1/R$ and $r_e=3R$ . Is leading zero range accurate at the ten-percent level?