Neutron specular reflectivity at soft interfaces provides sub-nanometre information concerning the molecular distribution of thin films, while the application of contrast variation can highlight the scattering from different parts of the system and lead to an overall reduction in fitting ambiguity. Traditional modelling approaches involve the construction of a trial scattering length density profile based on initial speculation and the subsequent refinement of its parameters through minimization of the discrepancy between the calculated and measured reflectivity. In practice this might produce an artificial bias towards specific sets of solutions. On the other hand, direct inversion of reflectivity data, despite its ability to provide a unique solution, is subject to limitations and experimental complications. Presented here is an integrated indirect Fourier transform/simulated annealing method that, when applied to multiple solvent contrast reflectivity data and within the limits of finite spatial resolution, leads to reliable reconstructions of the interfacial structure without the need for any a priori assumptions. The generality of the method permits its straightforward application in common experimental contrast-variation investigations at the solid/liquid and air/liquid interface.
Keywords: data fitting; neutron reflectivity; soft interfaces; structure recovery.