Computational modelling of interaction effects in Fe3O4 nanoparticle systems for comparison with experiments

Fine particle magnetism is employed in a wide range of applications ranging from magnetic data recording to cancer therapies. Characterisation of nanoparticles is important for improving their applicability. This is a complex task, especially if magnetostatic interactions are to be considered. Here we have developed a methodology to investigate the inverse problem, which consists of extracting the magnetic properties such as anisotropy, size or saturation magnetisation from experimental magnetisation curves. For each set of magnetic properties a magnetisation curve can always be obtained, but from a magnetisation curve the parameters cannot always be uniquely determined. If interactions are significant the issue becomes complicated and the question of whether the parameters can be uniquely identified arises. To study this we simulated the magnetic behaviour of interacting nanoparticles with Monte-Carlo techniques and applied two different methods for studying the inverse problem. This allows to show that a unique extraction of model parameters is indeed possible only in a certain range of magnetic nanoparticle concentrations and temperatures. Using simulations we investigated the inverse problem for two parameters, anisotropy and saturation magnetisation, at different temperatures. At low temperature both parameters can be well determined, but the errors and the parameters correlation is dependent on the strength of the magnetostatic interaction. In the high temperature case, due to superparamagnetic behaviour, only the saturation magnetisation can obtain using the inverse problem approach. The methodology was also tested for a set of experimental measurements done on magnetite nanoparticles.