The recently proposed asymmetry model for the highly viscous flow suggests the equality of the dielectric spectrum with the retardation part of the shear spectrum. The equality was checked using literature data, taken under carefully controlled conditions to ensure the same samples and the same temperature control in both measurements. The relation is valid in two substances at all measured temperatures. In two other substances, one can argue that there are good reasons for the deviations that one finds. A promising approach for the study of the flow process in highly viscous liquids is its comparison [1][2][3][4][5][6] in different techniques. One usually finds the dielectric absorption peak close to the heat-capacity one, 1-4 but the shear modulus peak is about half a decade higher in frequency. In a broad distribution of relaxation times, a modulus peak always appears at a higher frequency than a susceptibility (compliance) peak.
7,8The question is whether this is the reason here.It is easy to convince oneself that the inversion of the dielectric susceptibility to a dielectric modulus does not solve the problem. If the difference between the static and the high-frequency dielectric susceptibility is small compared to the latter, as in the cases discussed in the present Brief Report, one gets practically no peak shift. If it is large, as in the cases of glycerol and propylene carbonate, 2 one gets a peak shift by more than two decades, much too large to explain the observed difference.The inversion of the shear modulus G(ω) is not trivial, because its low-frequency limit is zero. In fact, one has two equivalent textbook descriptions of the shear spectrum of a liquid, 7 a relaxation spectrum H (τ ) for the description of the complex shear modulus G(ω) (τ relaxation time),and a retardation spectrum L(τ ) for the description of the complex shear compliance,Here G is the infinite frequency shear modulus and η is the viscosity. A third material constant hidden in this equation is the recoverable compliance J 0 e , the elastic compliance plus the integral over the retardation processes,Equation (2) makes a separation of two independent contributions to the compliance: the retardation spectrum and the viscosity. From our gradually growing understanding of the highly viscous liquid, 9,10 we know that both parts must come from thermally activated transitions between inherent states, stable structures corresponding to minima of the potentialenergy landscape. The retardation spectrum is due to back jumps into the initial inherent state, the viscosity is due to no-return processes. The retardation description separates these two influences, the relaxation description does not.It has been argued that the Zwanzig-Mori formalism requires a relaxation description, 11 with the bath processes contributing both to relaxation and to viscosity. A different picture results from the asymmetry model, 12 based on experimental evidence for a strong asymmetry of the secondary relaxation. 13 In this picture, the viscous flow occurs when ...