Neutron scattering experiments are sometimes used to study atomic diffusion in solid materials. My intuition is that the typical speeds of the atoms diffusing in the material--say, the diffusion constant (in cm^2/s) divided by the atomic spacing (in angstroms)--must be comparable to the speed of the incident neutrons for any information about diffusion to be gained. Much faster neutrons should see only a snapshot of the target atoms frozen in time, while much slower neutrons should see only average densities of the target atoms.
Is this reasoning correct? (Can anyone point me to a good introduction?)