Is the susceptibility minimum of glass-forming liquids due to the β-relaxation of the mode coupling theory or caused by the ubiquitous constant loss?

Observation of a minimum in the susceptibility spectra of glass-forming liquids has often been used as evidence of the fast β-relaxation in the mode coupling theory (MCT). However, recent experiments showing the absence of the knee predicted by MCT and the presence of the fast relaxation in crystalline as well as in glassy and supercooled ortho-terphenyl has cast doubt on the cause of the fast relaxation. A constant (approximately frequency independent) loss contribution to the dielectric susceptibility also can generate a minimum. The constant loss contribution is ubiquitous. I show that it is related to the fast process observed by light and neutron scattering and is possibly the cause of the susceptibility minimum. The properties of the constant loss have led to the suggestion that it originates from vibrational relaxation, a conclusion reached also by recent work on crystalline ortho-terphenyl. This suggestion is made plausible by theoretical results from studies of relaxation of non-linear (anharmonic) vibrations. Finally I show that if the constant loss contribution is the cause of the susceptibility minimum, then an explanation of the entire spectrum including the structural relaxation at all temperatures can be obtained by simply incorporating the constant loss into the coupling model of the author.