Universality of the high-temperature viscosity limit of silicate liquids

Abstract

We investigate the high-temperature limit of liquid viscosity by analyzing measured viscosity curves for 946 silicate liquids and 31 other liquids including metallic, molecular, and ionic systems. Our results show no systematic dependence of the high-temperature viscosity limit on chemical composition for the studied liquids. Based on the Mauro-Yue-Ellison-Gupta-Allan (MYEGA) model of liquid viscosity, the high-temperature viscosity limit of silicate liquids is 10 −2.93 Pa·s. Having established this value, there are only two independent parameters governing the viscosity-temperature relation, namely, the glass transition temperature and fragility index. It is also critical for understanding the relaxation characteristics of liquids, as in the well-known Angell plot 1 where the logarithm of viscosity, log 10 η, is plotted as a function of the T g -scaled inverse temperature, T g /T. Here, T g is the glass transition temperature, defined as the temperature at which the liquid viscosity equals 10 12 Pa·s, and T is absolute temperature. The slope of the Angell curve at T g defines the fragility index m,Fragility is a common measure of the slowing down of liquid dynamics upon cooling through the glass transition. According to Angell, 1 liquids can be classified as either "strong" or "fragile" depending on whether they exhibit an Arrhenius or super-Arrhenius scaling of viscosity with temperature, respectively. The degree of non-Arrhenius scaling varies greatly among different glass-forming liquids and reflects the second derivative of the viscosity curve with respect to inverse temperature. With the assumption of a universal hightemperature limit of viscosity, η ∞ , Angell proposed that this non-Arrhenius character is directly connected to the fragility index, m, a first-derivative property of the viscosity curve at T g . 2 However, the assumption of a universal high-temperature limit of viscosity, which enables this direct connection between first-and second-derivative properties, has not yet been validated by a systematic analysis of experimental data.In this Brief Report, we analyze viscosity-temperature curves of 946 silicate liquids and 31 other liquids, including water and silica, as well as borate, metallic, molecular, and ionic liquids. Our results show that there is no systematic dependence of η ∞ on composition and point to a narrow spread around η ∞ = 10 −2.93 Pa·s for silicate liquids. This result implies the existence of a universal high-temperature limit of viscosity, indicating that the fragility index m does have a direct relationship to the non-Arrhenius scaling of liquid viscosity (a measure of curvature), at least for silicate liquids. Our results indicate that there are only two independent parameters governing the viscosity of silicate liquids: T g and m.This simplifies the process for modeling the composition dependence of liquid viscosity and is an indication of the universal dynamics of silicate liquids at the high-temperature limit.To evaluate η ∞ , we analyze experimental viscosity data using ...

Composition–structure–property relationships in boroaluminosilicate glasses

“…Moreover, the high-temperature viscosity limit has recently been shown to be composition independent and equal to approximately 10 − 3 Pa·s [34]. With these definitions of glass transition temperature, high-temperature viscosity limit, and fragility, Eq.…”

Relaxation and physical aging in network glasses: a review

“…An analysis of viscosity curves [19] using a convenient fitting formula for silicate liquids and other liquids including metallic, molecular, and ionic systems, has shown that the high-temperature viscosity limit of such liquids is about 10 −2.93 Pa.s [19]. As there seems to be no systematic dependence of η ∞ on composition, at least for the silicates given the narrow spread around the average value of 10 −2.93 Pa.s, it is believed that η ∞ has some kind of a universal character.…”

“…The labels on the x-axis refer to all possible triplets i0j between a central atom 0 and two neighbors i and j. For all systems, the PBADs relative to the Group IV (Si, Ge) atom have a low standard deviation σ θ , of the order of [10][11][12][13][14][15][16][17][18][19][20] o when the first four neighbours are considered. One finds e.g.…”

Topological Ordering and Viscosity in the Glass-Forming Ge–Se System: The Search for a Structural or Dynamical Signature of the Intermediate Phase

“…The model was used to fit the measured viscosity data for Se (Cukierman and Uhlmann, 1973;Koštál and Málek, 2010;Gueguen et al, 2011), Ge 0.10 Se 0.90 (Nemilov, 1964;Senapati and Varshneya, 1996;Gueguen et al, 2011), Ge 0.20 Se 0.80 (Nemilov, 1964;Gueguen et al, 2011), and Ge 0.25 Se 0.75 (Nemilov, 1964;Senapati and Varshneya, 1996;Gueguen et al, 2011) where two or more of the data sets are self-consistent, and the measured viscosity data for Ge 0.30 Se 0.70 (Gueguen et al, 2011) where only one data set is available. For a given composition, the logarithm of the high-temperature viscosity was treated as either a fitting parameter or a fixed parameter set at log 10 [η ∞ (Pa s)] = − 2.93 (Zheng et al, 2011). The fits corresponding to log 10 [η ∞ (Pa s)] = − 2.93 are shown in Figure 11, and give values of T g and m visc (Figure 12) that are within the spread of values reported in the literature from viscosity measurements ( Table 1).…”