14.5: Partial Waves in the Classical Limit - Hard Spheres

Consider the scattering of a small hard sphere (radius \(x_m\), mass \(m\)) by a large hard sphere (\(X_M, M\)). Firstly we transform the problem to the center of mass reference frame where it becomes that of a single effective particle of mass \(\mu = mM/(m + M)\) moving in a hard sphere potential \((V (r < r_H = X_M + x_m) = \infty )\). Thus the boundary condition is \(R_l(r_H) = 0\).

Consider the classical limit, where the sphere radius is much larger than the de Broglie wavelength, \(kr_H \gg 1\). Up to \(l = Kr_H\) the phase shift is enormous and \(\sin \delta_l\) could have any value. For \(l > Kr_H\) the impact parameter is so large that the particles miss and \(\delta_l = 0\). Thus we can write the scattering cross section:

\[\sigma = \frac{4\pi}{ K^2} \sum^{l = Kr_H}_{l = 0} (2l + 1)\frac{1}{2} \nonumber\]

where we replace \(\sin^2 \delta_l\) with its average value of \(\frac{1}{2}\).

Since \(Kr_H\) is large, we can replace the sum by an integral and take only the leading term; \((Kr_H)^2 \gg Kr_H\):

\[\sigma \approx \frac{2\pi}{K^2} \int^{l = Kr_H}_{l = 0} (2l + 1) dl \approx 2\pi {r_H}^2 \nonumber\]

This result should send us rushing back to look for the extra factor of 2, since the cross-section of a sphere might be expected to be \(\pi {r_H}^2\). In fact, though, the analysis is correct and closer analysis of the \(\theta\) dependence of the wavefunction shows that half the amplitude is diffracted into the classical ‘shadow’ of the sphere to cancel the amplitude of the unscattered wave there.