We compare dielectric relaxation $\mathrm{\tau (T)}$ data of several low molecular weight glass-forming liquids with the predictions of the Adam–Gibbs theory using experimental data for the configurational entropy $S_c\mathrm{(T).}$ Combination of Adam–Gibbs and Vogel–Fulcher equations yields an expression for $S_c\mathrm{(T)}$ which can be compared with experimental data. Good agreement is found for a range of temperatures near $T_g{\mathrm{<T<T}}_B$ which depends on the fragility of the liquid and on the presence of a β relaxation. For fragile liquids, $T_B$ coincides with a qualitative change in the temperature dependence of the relaxation time scale or viscosity, with the temperature $T_{\beta },$ where Johari–Goldstein-type β processes tend to merge into the α process, and with other crossover temperatures. For nonfragile liquids, $T_B{\mathrm{/T}}_g$ increases, and the deviations from the Adam–Gibbs equation weaken or disappear altogether. The significance of $T_B,$ and of the $S_c\mathrm{(T)}$ temperature dependence implied by the Vogel–Fulcher theory fitting, are discussed in terms of the landscape paradigm.

We have measured the dielectric relaxation times of the α process of phenyl salicylate (salol) covering 11 decades in frequency. Being representative for the class of low molecular weight organic glass forming materials, the highly resolved temperature dependence of the dynamics in salol does not follow a particular function like the Vogel–Fulcher–Tammann (VFT) law over the entire accessible range of temperatures. In order to conduct a detailed and unambiguous analysis of the temperature dependence, we take advantage of the derivatives of the experimental log(fmax) values with respect to temperature, which allow us to either linearize the frequently used temperature laws or to resolve subtle changes in fmax(T) by decreasing the number of free parameters. In this manner we observe that none of the common routes for rationalizing the dynamics like Arrhenius, VFT, Souletie scaling, and idealized mode-coupling theory account for the experimental findings properly. However, we do observe a VFT behavior within the limits 265 K≤T≤320 K, i.e., for temperatures ranging from significantly above the glass transition at Tg=220 K to far above the melting point.

We have studied the temperature dependence of dielectric relaxation times in terms of the peak frequency f_max(T) of dielectric loss ε″(ω) and the dc‐conductivity σ_dc(T) of several glass‐forming liquids, covering 12 decades in the peak frequency f_max and 9 decades in σ_dc. Although dc‐conductivity samples the mobility of ionic tracers, its variation with temperature is similar to that of f_max(T). The f_max(T) and σ_dc(T) are analyzed using the temperature‐derivative method and compared to the viscosity data η⁻¹(T). While most liquids reveal a common Vogel–Fulcher–Tammann (VFT) behavior for f_max, σ_dc and η⁻¹ in an extended temperature range T≥T_m, some liquids deviate from this behavior by displaying a crossover at T=T_A to an Arrhenius regime. In these cases the quantity f_max(T) decouples from the common curves for σ_dc(T) and η⁻¹(T) and attains activation energies in excess (∼40% for alcohols) of those related to translational processes. For many samples a departure from the VFT behavior occurs at lower temperatures T_B<T_m which tends to retard the glass transition. The onset of this qualitative change in the temperature dependence at T_B turns out to be a characteristic temperature also in other experiments.

We have measured the dielectric relaxation of the glass-former 1-propanol for temperatures between 65 and 350 K in the frequency range 10−2 to 2⋅1010 Hz and the photon correlation spectro-scopy decays near Tg. Attributing the strong Debye-type process of 1-propanol to distinct -OH group effects leaves two faster processes related to the structural relaxation which can be identified as α-relaxation and Johari–Goldstein type β-relaxation characteristic of nonhydrogen-bonding supercooled liquids. From the temperature dependent relaxation times τ(T) regarding the three distinct loss peaks, we can specify an α-β-bifurcation temperature Tβ, which coincides with characteristic qualitative changes in the τ(T) behavior, as also observed for ortho-terphenyl and other glass-forming liquids. This assignment is confirmed by the correla-tion times derived from incoherent quasielastic light-scattering data obtained from the simultaneously measured photon-correlation spectroscopy.

For nonpolymeric supercooled liquids, the empirical correlation m=56TgΔCp(Tg)∕ΔHm provides a reliable means of correlating dynamic and thermodynamic variables. The dynamics are characterized by the fragility or steepness index m and the glass transition temperature Tg, while thermodynamics enter in terms of the heat capacity step ΔCp at Tg and the melting enthalpy ΔHm. The combination of the above correlation with the 2∕3 rule for the Tg∕Tm ratio yields an expression, m=40ΔCp(Tg)∕ΔSm, which was rationalized as the correlation of the thermodynamic and kinetic fragilities. Defining a thermodynamic fragility via ΔCp(Tg)∕ΔSm also reveals that the slopes in Kauzmann’s original ΔS(T)∕ΔSm versus T∕Tm plot reflect the fragility concept [Chem. Rev. 43, 219 (1948)], so long as Tm∕Tg=1.5. For the many liquids whose excess heat capacity is a hyperbolic function of temperature, we deduce that the fragility cannot exceed m=170, unless the Tg∕Tm=2∕3 rule breaks down.