Vibrational disorder and densification-induced homogenization of local elasticity in silicate glasses

Abstract

We report the effect of structural compaction on the statistics of elastic disorder in a silicate glass, using heterogeneous elasticity theory with the coherent potential approximation (HET-CPA) and a log-normal distribution of the spatial fluctuations of the shear modulus. The object of our study, a soda lime magnesia silicate glass, is compacted by hot-compression up to 2 GPa (corresponding to a permanent densification of ~ 5%). Using THz vibrational spectroscopic data and bulk mechanical properties as inputs, HET-CPA evaluates the degree of disorder in terms of the length-scale of elastic fluctuations and the non-affine part of the shear modulus. Permanent densification decreases the extent of non-affine elasticity, resulting in a more homogeneous distribution of strain energy, while also decreasing the correlation length of elastic heterogeneity. Complementary 29Si magic angle spinning NMR spectroscopic data provide a short-range rationale for the effect of compression on glass structure in terms of a narrowing of the Si–O–Si bond-angle and the Si–Si distance.

The boson peak in silicate glasses: insight from molecular dynamics

“…Not only that but also the increase of the BP intensity (see the ESI,† Fig. S2–S4 36 ) is an indication of the enhancement of the soft elastic modes, 71 which were shown in previous reports to introduce higher correlation length. 72 In addition, the dropping of the correlation length from 3.5 nm in v-SiO 2 to ≈1.3 nm and ≈1.3 nm in BTS and CAS, respectively, indicates the reduction of the poorly packed structural domains enclosed in the network of silicate glass.…”

“…This expectation was confirmed previously by Ando et al 7 for binary SiO 2 -Al 2 O 3 , aluminosilicate and other glassy systems studied by Schroeder et al 69 and in other works including densified silicate glass. 70,71 Fig. 4, represents the dynamic correlation length of all studied glasses with varied concentrations of intermediates and modifiers alkali oxides.…”

Pressure‐driven homogenization of lithium disilicate glasses

“…The pressure effect on the ring‐size distribution in the glass could be explained by the presence of topological variation in the structure, which was related to the type of modifier present in the glass 42,62 . We notice that the mean ring size decreases with increasing pressure, indicating the dominance of small rings in the distributions; the dominance of the small rings ($$<$$ 6‐membered) suggests the existence of small Li‐rich regions in these glasses, which we can consider it as the first indication of Li homogenization in the glasses with increasing pressure 63–65 …”

“…42,62 We notice that the mean ring size decreases with increasing pressure, indicating the dominance of small rings in the distributions; the dominance of the small rings (< 6-membered) suggests the existence of small Li-rich regions in these glasses, which we can consider it as the first indication of Li homogenization in the glasses with increasing pressure. [63][64][65] The pressure-induced homogenization of the structure was shown using the clustering ratio and the density maps of the Li in the simulation box at different pressures (see Figures 11 and 12). The increase of the homogenization of the Li distribution is mainly due to an increasing repulsion between the Li atoms when the pressure increases, which is confirmed by the analysis of the CSRO parameter, 58 where for the glass cooled under no external pressure, the Li atoms are more attracted toward each other, and this attraction decreases with increasing the pressure (see Figure 11b).…”

Thermal conductivity of glasses: A structural point of view

“…Other experiments have found types of clustering in different oxide glasses, e.g., finding that three-and four-coordinated borons tend to cluster separately in modified borate glasses, yet that such clustering decreases upon increasing pressure [41]. Interestingly, such arguments of inhomogeneity have also been found to apply to pure oxides, e.g., SiO 2 , yet here with a less clear specification of its structural origin [42][43][44]. Other common ways to characterize MRO rely on characterizing the ring-type structures found in oxides upon pressure addition.…”