Possible explanation of intermediate frequency absorption/loss in polar liquids through a non-linear itinerant oscillator model

Most of the previous work on the itinerant oscillator (IO) assumes that the non-linear cosine potential which couples a molecule of the liquid to its surroundings can be approximated by the equivalent harmonic-oscillator potential, i.e. Cos Θ≈Θ. Such an assumption simplifies the subsequent analysis but ignores the inherent non-linearity of the potential. In this paper two different ways of including non-linearities into the existing framework of the model are described in brief and some theoretical loss spectra are compared with loss spectra of the linear model.