Hyperquenched Glass Research Articles

Rejuvenation of granulated blast furnace slag (GBS) glass by ball milling

Differential scanning calorimetry, thermogravimetry and gas analysis were used to study the impact of high-energy mechanical milling (HEMM) on the rejuvenation of a granulated blast furnace slag (GBS). It is found that milled GBS – ground granulated blast furnace slag (GGBS) in practice – shows an enormous release of exothermic heat in the sub-Tg range upon heating at 10 K min−1, which increases with milling time and exceeds more than three times that of the unmilled GBS. Heating of GGBS revealed that the energy release of HEMM-activated glasses occurs at considerably lower temperatures than observed for thermally excited glasses (hyperquenched fibers). The difference in the characteristic release temperatures is assigned to two different relaxation mechanisms. The former (HEMM), is due to relaxation of structural defects on the surface of the powder particles, whereas the latter (hyperquenched glass), comprises relaxation of the entire glass structure throughout the volume.

Suppression of Orientational Correlations in the Viscous-Liquid State of Hyperquenched Pressure-Densified Glycerol.

Glycerol pressurized to 2kbar and hyperquenched from the bulk liquid at rates of about -10 000 K/s, has been frozen to an extreme out-of-equilibrium state. As compared to conventionally cooled melts, the resulting material exhibits lower orientational correlations, enabling the observation of a secondary relaxation peak in the ambient-pressure dielectric response. The hyperquenching rather than the pressurizing part of the preparation protocol induces the observed structural changes. These vanish entirely only well above the glass transition temperature of the equilibrium liquid and are evidence for strong similarities between hyperquenched and vapor-deposited glass formers.

Shadow glass transition as a thermodynamic signature of β relaxation in hyper-quenched metallic glasses.

One puzzling phenomenon in glass physics is the so-called ‘shadow glass transition’ which is an anomalous heat-absorbing process below the real glass transition and influences glass properties. However, it has yet to be entirely characterized, let alone fundamentally understood. Conventional calorimetry detects it in limited heating rates. Here, with the chip-based fast scanning calorimetry, we study the dynamics of the shadow glass transition over four orders of magnitude in heating rates for 24 different hyper-quenched metallic glasses. We present evidence that the shadow glass transition correlates with the secondary (β) relaxation: (i) The shadow glass transition and the β relaxation follow the same temperature–time dependence, and both merge with the primary relaxation at high temperature. (ii) The shadow glass transition is more obvious in glasses with pronounced β relaxation, and vice versa; their magnitudes are proportional to each other. Our findings suggest that the shadow glass transition signals the thermodynamics of β relaxation in hyper-quenched metallic glasses.

Cooling history of a wet-granulated blast furnace slag (GBS)

Granulated blast furnace slag (GBS) is a glassy by-product of the steel industry that is formed during quenching of the molten slag after the blast furnace in a water-jet. To elucidate the cooling history of GBS, calorimetric scanning in the glass transition range and viscometric experiments at temperatures above the liquidus were performed. The GBS studied was of industrial origin with a d50 = 700 μm and > 99% glassy. It is found that GBS glass is of high potential energy, showing a fictive temperature that is approx. 160 K higher than the glass transition temperature under standard cooling conditions. Using different viscosity models and the relationship between quench rate and shear viscosity at the fictive temperature it is calculated that GBS was formed at a cooling rate of approx. 2.6 × 105 K s−1 corresponding to a shear viscosity of 105.9 Pa s. Due to the high Tf and high amount of heat that is released during structural relaxation, GBS is assigned to the class of hyperquenched glasses.

Structural response to sub-Tg annealing in a hyperquenched SiO2-Al2O3 glass

We explore the connection between the structural change and the sub-Tg enthalpy relaxation in 61SiO2–39Al2O3 (SA) glass by using the hyperquenching-annealing-calorimetry approach, nuclear magnetic resonance (NMR) and high-resolution transmission electron microscopy (HRTEM). The differential scanning calorimetry (DSC) measurements show a sub-Tg relaxation peak with a long tail starting from the onset temperature. With increasing the degree of sub-Tg annealing, the tail evolves into a broad peak. The 5- and 6- fold coordinated Al species in the SA glass decrease with sub-Tg annealing, while the 4-fold coordinated Al species increases. The coordination change leads to the formation of the oxygen triclusters, which act as the nucleation sites for mullite formation. This is confirmed by the existence of ordered structural nanodomains (∼10 nm). The extremely unstable behavior of the SA glass against crystallization could be attributed to both the higher content of the 5-fold coordinated Al (∼64%) and the easily formed oxygen triclusters. This work provides insights into both the structural evolution during the sub-Tg relaxation, and the structural heterogeneity of an unstable glass system.

Quantification of the thermal and chemical inhomogeneity of glasses

AbstractAttempts to scientifically capture and quantify glass homogeneity have been based on optical methods, namely, on those based on the work of Christiansen, Raman, and Shelyubskii. In industrial practice, a number of further methods are in use. In this study, glass homogeneity is determined by the measurement of density fluctuation. The method of “density titration” proposed in earlier literature is used in an essentially improved procedure. In this method, the density distribution is determined for a representative ensemble of glass particles prepared from a bulk glass sample. The particles float on a heavy liquid which is diluted, step by step, by a miscible liquid of lower density. The particle fraction with density exceeding the density of the mixture sinks. When the method is performed on quenched or hyperquenched glass, then two types of inhomogeneities contribute to the overall homogeneity, i.e., thermal and chemical contributions. By two consecutive measurements, one performed on a sample of quenched glass, one on the same sample relaxed at , these contributions are clearly discriminated. Results are presented by an error function type distribution where the median marks the average density, and the standard deviation is adopted as a measure of glass homogeneity.

Comparing the atomic and macroscopic aging dynamics in an amorphous and partially crystalline Zr44Ti11Ni10Cu10Be25bulk metallic glass

Abstract

Sub-T g enthalpy relaxation in a milling-derived chalcogenide glass

In this work we succeeded in synthesizing a chalcogenide glass, that is, the Ag3PS4 glass, by milling the mixture of the crystalline Ag2S and P2S5 powder. In terms of potential energy, the glass was excited by the milling process. The evidence for the energy excitation was the occurrence of the sub-Tg relaxation recorded by a differential scanning calorimeter (DSC). The relaxation behavior in the milling-derived Ag3PS4 glass was found to significantly differ from that of a hyperquenched oxide glass. We observed two decoupled peaks of the energy release during a DSC scan, which are in contrast to the single asymmetric peak of the hyperquenched glass. The low-temperature peak could be attributed to the β-relaxation of the Ag–S ionic bonds, whereas the high-temperature peak could arise from the α-relaxation of the distorted rigid [PS4] units involving covalent bonds. The decoupling between the two peaks implies the fragile nature of the Ag3PS4 glass. Upon the dynamic heating and the sub-Tg annealing we found that the relaxation of the Ag3PS4 glass is of high nonexponentiallity, and hence, of high structural heterogeneity.

Pressure-induced transformations in glassy water: A computer simulation study using the TIP4P/2005 model.

We study the pressure-induced transformations between low-density amorphous (LDA) and high-density amorphous (HDA) ice by performing out-of-equilibrium molecular dynamics (MD) simulations. We employ the TIP4P/2005 water model and show that this model reproduces qualitatively the LDA-HDA transformations observed experimentally. Specifically, the TIP4P/2005 model reproduces remarkably well the (i) structure (OO, OH, and HH radial distribution functions) and (ii) densities of LDA and HDA at P = 0.1 MPa and T = 80 K, as well as (iii) the qualitative behavior of ρ(P) during compression-induced LDA-to-HDA and decompression-induced HDA-to-LDA transformations. At the rates explored, the HDA-to-LDA transformation is less pronounced than in experiments. By studying the LDA-HDA transformations for a broad range of compression/decompression temperatures, we construct a "P-T phase diagram" for glassy water that is consistent with experiments and remarkably similar to that reported previously for ST2 water. This phase diagram is not inconsistent with the possibility of TIP4P/2005 water exhibiting a liquid-liquid phase transition at low temperatures. A comparison with previous MD simulation studies of SPC/E and ST2 water as well as experiments indicates that, overall, the TIP4P/2005 model performs better than the SPC/E and ST2 models. The effects of cooling and compression rates as well as aging on our MD simulations results are also discussed. The MD results are qualitatively robust under variations of cooling/compression rates (accessible in simulations) and are not affected by aging the hyperquenched glass for at least 1 μs. A byproduct of this work is the calculation of TIP4P/2005 water's diffusion coefficient D(T) at P = 0.1 MPa. It is found that, for T ≥ 210 K, D(T) ≈ (T - T(MCT))(-γ) as predicted by mode coupling theory and in agreement with experiments. For TIP4P/2005 water, T(MCT) = 209 K and γ = 2.14, very close to the corresponding experimental values T(MCT) = 221 K and γ = 2.2.

Crystallization mechanisms occurring in the Se–Te glassy system

Crystallization kinetics, studied under non-isothermal conditions in dependence on particle size, is reviewed for the Se–Te glassy system, with particular focus on occurring macroscopic crystallization mechanisms. The observed complexity of the crystallization processes is explained in terms of the two previously identified mechanisms—bulk crystallization originating from nuclei randomly distributed in the volume of each glass particle (formation of these nuclei is consistent with the classical nucleation theory) and crystallization from surface/volume active centers, dislocations, structural defects, and heterogeneities originating from the mechanical damage of the material (e.g., during grinding and milling). In addition, a new crystallization mechanism, previously unobserved in this system, was revealed for hyperquenched Se95Te5 glass. The origin of this mechanism is suggested to be closely associated with the stress-induced defects contained in the highly strained structure of this glass. The strong compositional dependence of the occurrence of this mechanism then seems to be caused by the high cooperativity of the amorphous structure of the specific composition.

Calorimetric features of release of plastic deformation induced internal stresses, and approach to equilibrium state on annealing of crystals and glasses

• Stress release in a glass occurs at a faster rate than structural relaxation. • Plastically-deformed glass would show two exothermic minima, and no glass transition. • Enthalpy matching procedure would yield an inaccurate fictive temperature. • Complex heat capacity may distinguish plastically-deformed from quench-formed glass. Plastic deformation of crystals and glasses produces internal strains (stresses), which change their energy and other thermodynamic properties. On annealing, these stresses decrease at a rate faster than the structure relaxes toward the equilibrium state. Mechanism of such relaxations in crystals differs from that in glasses and it also differs for glasses of different types. In all cases, the energy related properties decrease with time isothermally and on heating, resembling the structure relaxation of a stress-free glass. We consider these features and argue that kinetics of enthalpy loss with time yields the rate constants of the stress release and of the structure change, and not the viscosity determining α-relaxation time. Since thermal cycling does not recover the enthalpy from internal stresses, a glass with stresses has neither a glass-softening temperature, T g , nor a fictive temperature, T f . Plastic deformation would not rejuvenate a physically aged glass to the properties of its un-aged state. The Prigogine–Defay ratio can be extended to all T f s, and used to investigate the effect of distribution of relaxation times on its value, but it can not be defined for an internally stressed glass. After discussing the effects of annealing on the heat capacity and DSC scans, we conclude that on slow heating, glass with deformation-induced stresses would show two exothermic minima, and normal glass would show only one such minimum. Temperature-modulated scanning calorimetry would also distinguish an internally stressed glass from an equally high-enthalpy, stress-free glass. Enthalpy matching procedure would yield inaccurate T f value of a plastically deformed glass, but not of a hyperquenched glass. Phenomenological analysis given here has consequences for interpretation of the calorimetric changes observed on heating the shear-amorphized, as well as pressure-collapsed states of crystals.

Aging and structural relaxation of hyper-quenched Mg65Cu25Y10 metallic glass

The structural relaxation, glass transition and crystallization processes of Mg65Cu25Y10 metallic glass are studied by Differential Scanning Calorimetry (DSC) and Mechanical spectroscopy. The relaxation model derived from the mechanical measurements is compared with the kinetics of these transformations obtained from the DSC curves. The structural relaxation kinetics is found to be controlled by the glassy dynamics following an Adams–Gibbs–Vogel function. The glass transition and crystallization kinetics are controlled by the dynamics of the supercooled melt following a Vogel–Fulcher–Tammann behaviour. The results suggest that the microscopic processes responsible of structural relaxation and aging below the glass transition correspond to the same processes generating the α-relaxation peak.

Sub-Tg enthalpy relaxation in an extremely unstable oxide glass and its implication for structural heterogeneity

We study the sub-Tg relaxation in an extremely unstable glass former, i.e., 65SiO2–35Al2O3, and its relation to structural heterogeneity (e.g., structurally ordered domains in glass matrix). This is done by hyperquenching (~106K/s) the liquid, then annealing the hyperquenched (HQ) glass below Tg and subsequently scanning the annealed HQ glass in a differential scanning calorimeter (DSC). The results show that structural ordering can take place even below Tg. An endothermic pre-peak is observed when the HQ sample is annealed at 0.75Tg for sufficiently long time, which is, however, much weaker compared to that of stable glass formers subjected to same annealing conditions. We also investigate the effect of the sub-Tg annealing on crystallization above Tg. The results imply that some structurally ordered domains exist already in the liquid state. The ordered domains lower the activation energy for further structural ordering during the sub-Tg annealing, and thereby lowers the activation energy for crystal growth above Tg.

Effect of the initial stage of annealing on modeling of enthalpy relaxation in a hyperquenched glass

One of the major challenges in glass relaxation study is to establish a universal model describing the enthalpy relaxation in both the hyperquenched glass (HQG) (i.e., far from equilibrium) and the partially annealed hyperquenched glass (AHQG). In particular, the detailed features of the enthalpy relaxation in both HQG and AHQG during the initial stage of both the sub-Tg and above-Tg annealing cannot be captured by the existing models. In this Letter we show that the combination of a modified stretched exponential equation [M. Peyron, et al., J. Magn. Reson, Series A 118 (1996) 214] and the recently proposed composite relaxation function [L. Hornboell, et al., Chem. Phys. Lett. 1-3 (2010) 37] is a reasonable approach for describing those features. In addition, our modeling results imply that the structural heterogeneity plays a crucial role in relaxation of HQG.

Thermodynamic anomaly of the sub-Tg relaxation in hyperquenched metallic glasses

Recently, we observed an unusual non-monotonic glass relaxation phenomenon, i.e., the three-step sub-T(g) relaxation in hyperquenched CuZrAl glass ribbons [L. N. Hu and Y. Z. Yue, Appl. Phys. Lett. 98, 081904 (2011)]. In the present work, we reveal the origin of this abnormal behavior by studying the cooling rate dependence of the sub-T(g) enthalpy relaxation in two metallic glasses. For the Cu46Zr46Al8 glass ribbons the sub-T(g) enthalpy relaxation pattern exhibits a three-step trend with the annealing temperature only when the ribbons are fabricated below a critical cooling rate. For the La55Al25Ni20 glass ribbons the activation energy for the onset of the sub-T(g) enthalpy relaxation also varies non-monotonically with the cooling rate of fabrication. These abnormal relaxation phenomena are explained in terms of the competition between the low and the high temperature clusters during the fragile-to-strong transition. By comparisons of chemical heterogeneity between Cu46Zr46Al8 and La55Al25Ni20, we predict that the abnormal relaxation behavior could be a general feature for the HQ metallic glasses.

Structural relaxation in annealed hyperquenched basaltic glasses: Insights from calorimetry

The enthalpy relaxation behavior of hyperquenched (HQ) and annealed hyperquenched (AHQ) basaltic glass is investigated through calorimetric measurements. The results reveal a common onset temperature of the glass transition for all the HQ and AHQ glasses under study, indicating that the primary relaxation mechanism is activated at the same temperature regardless of the initial departure from equilibrium. The analysis of secondary relaxation at different annealing temperatures provides insights into the enthalpy recovery of HQ glasses.

Pressure dependence of the boson peak for repulsive homogeneous potentials

Isotropic potentials used for studying glasses computationally are often positive homogeneous functions of the coordinates of the constituent particles or they reduce effectively to such forms at higher pressures. For such cases we show that the position and the intensity of the Boson peak, when it exists, scales asymptotically with pressure ($P$) as ${P}^{\ensuremath{\delta}}$ and ${P}^{\ensuremath{-}3\ensuremath{\delta}}$, respectively, where $\ensuremath{\delta}=(2+m)/[2(S+m)]$, $(\ensuremath{-}m)$ is the degree of homogeneity, and $S$ is the spatial dimension. In the same limit the shape of the entire vibrational spectrum saturates and the Boson peak frequency, the Debye frequency, and the average frequency become proportional to each other. Our numerical studies on single-component hyperquenched Lennard-Jones glass in three dimensions support the predicted scaling laws and are largely consistent with the predictions of one existing theory [W. Schirmacher et al., Phys. Rev. Lett. 98, 025501 (2007)] for the normalized boson peak intensity. However, the scaling exponent for the variation of the boson peak location and intensity with applied pressure is substantially different in this case from that found from laboratory experiments.

Abnormal sub-Tg enthalpy relaxation in the CuZrAl metallic glasses far from equilibrium

The sub-Tg enthalpy relaxation in glasses far from equilibrium (i.e., hyperquenched glasses) is usually manifested as a monotonic increase in both the onset temperature and the extent of enthalpy recovery with increasing the annealing temperature. Here we report an abnormal three-steplike relaxation pattern in hyperquenched Cu45Zr45Al10 and Cu48Zr48Al4 glasses below Tg, i.e., the usual monotonic increase is inverted when the annealing temperature increases to a critical value. This abnormal relaxation pattern could be attributed to the high extent of chemical heterogeneity in the CuZrAl supercooled liquids, and present a thermodynamic evidence for the dynamic fragile-to-strong transition in metallic glass-forming liquids.

Structural evolution during the sub-Tg relaxation of hyperquenched metallic glasses

We report the structural characteristics during the sub-Tg relaxation in hyperquenched La55Al25Ni20 glasses. The sub-Tg relaxation is associated with the structural change in intermediate range order, as manifested by the appearance of a prepeak in the x-ray diffraction spectrum. Such structural change could be the source of the Johari–Goldstein relaxation in metallic glasses. The mechanism governing the evolution of the prepeak is different between the glasses with the fictive temperature below 604 K and those above 604 K. Cooperative motion of atoms in La-centered clusters was further discussed in terms of the atomic bond deficiency model.

Cooling rates of basaltic hyaloclastites and pillow lava glasses from the HSDP2 drill core

Cooling rates have been determined for basaltic glasses from different depths of the submarine section of the drill core recovered in the 1999 phase of Hawaii Scientific Drilling Project (HSDP2). The glasses include degassed blocky hyaloclastite clasts and undegassed pillow rims. The degassed glassy clasts were generated in subaerial or shallow submarine environments, during explosive interactions between lava and seawater, before eventual deposition under water. The volatile contents of the glassy pillow rims are consistent with eruption and quenching in water several hundred metres deep. The cooling rates have been calculated from the calorimetric properties of the glass across the glass transition. The heat capacity ( c p ) of each sample was measured during several cycles of heating from room temperature to temperatures above their glass transition using a differential scanning calorimeter (DSC). Their compositions did not change during the thermal treatment, a prerequisite for successful c p measurements, although the glasses with higher H 2O contents became more opaque and their mid-IR spectra changed. Each c p - T path exhibits the now classic features of the glass transition; glassy and liquid states separated by a hysteresis marking the transition. After experiencing the same experimental thermal history the glass transition occurs at lower temperatures in glasses with higher H 2O contents. Except for one sample, the c p - T path measured on initial heating also releases energy stored during the natural quench, which is not recovered during subsequent experimental cooling. The energy stored in the HSDP2 glasses is much less than that observed in hyperquenched natural and synthetic glasses. Even so, the Tool–Narayanaswamy enthalpy relaxation geospeedometer, usually used to determine the cooling rates in volcanic glasses, is unable to deal with this energy release. For those samples that exhibit this feature an alternative method, developed for hyperquenched glasses, is applied. This uses the energy released to calculate T f , from which the cooling rate is calculated. The degassed blocky hyaloclastite clasts exhibit cooling rates 0.1–72.2 K s −1, while the undegassed pillow rims span 0.2–46.4 K s −1. The fastest cooling rates are consistent with the cooling of lava bodies in seawater. The wide variation for both types of glass could reflect quenching at different distances from the basalt-seawater interface. However, for the degassed hyaloclastite clasts the range could indicate that the clasts were generated by different processes operating during the explosive interaction between lava and seawater in the littoral zone. In the undegassed pillow lavas, glassy rims may have been reheated, giving rise to more complex, slower, thermal histories, as a result of latent heat released during the crystallisation of pillow interiors, or flow replenishment. Both types of glass may also have experienced reheating from succeeding flows or deposits. Compared to deep-sea limu o Pele hyaloclastite fragments, whose hyperquench rates indicate simultaneous cooling and fragmentation, the shallow blocky hyaloclastite clasts may have formed during post-cooling brittle fragmentation.