Enhancement in nanohardness of soda-lime-silica glass

For a thin commercial soda-lime-silica glass cover slip a significant increase in nanohardness (e.g. up to 74%) occurred with variations in the loading rate (10-20,000 μN·s^− 1) along with presence of serrations in load-depth plots and deformation band formation inside the nanoindentation cavity. These results are explained in terms of shear stress acting at positions of structural weakness close to the tip of the nanoindenter and the time scale of interaction between the nanoindenter and the local microstructure of the glass.     

Strengthening of soda-lime-silica glass by surface treatment with sol–gel silica

In this study, the failure resistance of soda-lime-silica glass was increased by surface treatment with sol–gel silica. Samples annealed and ion-exchanged in KNO₃ for 24 h at 450 °C were considered. Sol–gel silica coating was carried out by dipping the glass samples into a sol suspension prepared by hydrolysis of Si(OEt)₄ in ethanol/water solution. The deposited layer was consolidated in air for 24 h and subjected to mild thermal treatment at 300 °C for 1 h. The surface treatment increased the fracture resistance of annealed glass of about 35 MPa; conversely, ion-exchanged specimens showed an average increase of about 90 MPa. The strengthening effect induced by the surface treatment was attributed to the reduction of the effective crack length generated by the silica coating. The different strength increase between annealed and ion-exchanged samples is discussed in terms of fracture toughness which, for ion-exchanged glass, is not constant, due to the presence of the surface residual stresses and thus the reduction of the crack length due to the silica coating determines a higher strength increase than for annealed glass.     

Diffusion of the S isotope in soda-lime-silica and sodium trisilicate glass melts

The diffusivity of sulphur in nominal 10 Na₂O-16 CaO-74 SiO₂ (NCS) and 26 Na₂O-74 SiO₂ (NS3) melts was investigated in the temperature range 1273-1473 K using the ³⁵S radioactive isotope in a sandwich setup. Samples were sealed in platinum capsules and run with vertical alignment at 100 MPa confining pressure in an internally heated gas pressure vessel. Using the lowest diffusion coefficient D (m² s^− 1) for each temperature the Arrhenian relations logD = −(4.6 ± 0.3) − (216 ± 7) kJ mol^− 1/RT for NCS and logD = −(6.3 ± 0.6) − (167 ± 17) kJ mol^− 1/RT for NS3 were determined. Viscosity of the melts was used to test the applicability of the Stokes-Einstein (SE) and the Eyring (EY) equations to sulphur diffusion. The SE equation yields unrealistically low radii of the diffusing particles, a consequent of the non-molecular structure of the silicate melts. On the other hand, the Eyring relation yields reasonable jump distances of 550 pm (NCS) and 750 pm (NS3) by fitting the diffusion data to the EY equation. These large values imply that sulphate ions (ionic diameter = 290 pm) migrate as large entities through the silicate network.     

The optical properties of a soda-lime-silica glass in the region from 0.006 to 22 eV

From the measured absorption and reflection spectra, we have determined the optical properties of a well-characterized (with respect to impurities and homogeneity) high-purity 21.3 wt% Na₂O?5.2 wt% CaO?73.5 wt% SiO₂ glass over the energy range 0.006–22 eV. The origins of the absorption spectra are discussed.     

Precise XPS depth profile of soda-lime-silica glass using C60 ion beam

Ar ion sputtering is one of the most accepted techniques for depth profiling in practical X-ray photoelectron spectroscopy (XPS) analysis, while this technique is known to be inadequate for quantitative analysis of glass including mobile ions such as soda-lime-silica glass. For the precise depth profiling on XPS depth analysis, three methods, (i) coating of conductive thin film on glass surface, (ii) sample cooling at the temperature of −130 °C and (iii) buckminsterfullerene (C₆₀) ion sputtering, were investigated to suppress the migration of mobile ions in glass. In our best knowledge, it is the first time to succeed the precise XPS depth analysis of soda-lime-silica glass (70.4SiO₂, 0.9Al₂O₃, 7.3MgO, 7.8CaO, 13.6Na₂O in mol%) without compositional change by using C₆₀ ion sputtering. The precise analysis revealed that the ion implantation during Ar ion sputtering principally resulted in the compositional change due to the migration of mobile ions.     

Physicochemical model for predicting molten glass density

Models have been developed since the 1970s to predict the composition- and temperature-dependent density of silicate melts in which the molten glass is considered as a mixture of virtually ideal solutions. Published data were compiled to set bounds on the partial volumes of each constituent oxide of radioactive waste containment glass. A model based on the data is proposed to predict the density of complex molten borosilicate glass formulations between 900 and 1300 °C. The model is limited to compositions with silica concentrations between 40 and 80 mol%, alkali oxide concentrations between 5 and 50 mol%, and SiO₂/B₂O₃ molar ratios exceeding 0.9. Within this composition range the model is capable of reproducing the experimental data within 4%. One application of this approach is to construct an interpolation equation suitable for use in models simulating the thermal, rheological, electrical, and magnetic conditions of vitrification processes.     

Ion exchange equilibria between glass and molten salts

It is shown that interaction between the ions undergoing ion exchange leads to a sigmoidal equilibrium curve. A negative interaction in the glass causes the ratio of concentration of the two ions to be more nearly equal to unity in the glass than it is in the molten salt. It is further shown that the Rothmund-Kornfeld parameter is a measure of the non-ideality of the system. A value of the parameter greater than unity indicates a negative interaction energy.     

Density and surface tension of molten calcium fluoride

The density and the surface tension of molten calcium fluoride have been measured in the temperature range from 1690 to 1790 K by an improved Archimedian method and a ring depressing technique (J. Crystal Growth 187 (1998) 391), respectively. The ring depressing technique was demonstrated as an effective technique to measure the surface tension in comparison with the conventional ring pulling technique. The density varied with the temperature change corresponding to a linear relationship: ρ=3.767−6.94×10⁻⁴T (K). The density of the CaF₂ melt at the melting point is 2.594 g/cm³, which is equal to the result obtained by Shiraishi and Watanabe (Bull. Res. Inst. Miner. Dressing Metal, Tohoku Univ. 34 (1978) 1), but the temperature coefficient of the density is different from the results obtained by other investigators. The thermal expansion coefficient of calcium fluoride melt linearly increases with temperature heating. The surface tension of molten calcium fluoride indicates a negative linear relationship as a function of the melt temperature: γ(T)=442.4−0.0816×T(K) (mN/m). The surface tension measured using the ring depressing technique is larger than those results obtained by other techniques.     

Kinetic fragility of hydrous soda-lime-silica glasses

The effect of hydration on the kinetic fragility of soda-lime-silica glasses was investigated by viscometry in the glass transition range. Water-bearing glasses were prepared from industrial float glass (FG) and a ternary model glass (NCS = 16Na₂O 10CaO 74SiO₂ in mol%) by bubbling steam through the melt at 1480 °C and up to 7 bar. Additionally, a sodium borosilicate glass (NBS = 16Na₂O 10B₂O₃ 74SiO₂ in mol%) was hydrated under equal conditions. As detected by infrared spectroscopy water dissolves in the glasses exclusively as OH-groups. The hydration resulted in a total water content C_W up to ≈ 0.2 wt% for FG, NCS and NBS glasses. Kinetic fragility, expressed by the steepness index m, was determined from the temperature dependence of η at the glass transition. Viscosity data from previous studies on hydrous float glasses (C_W > 1 wt%) were surveyed together with literature data on the (H₂O)–Na₂O–CaO–SiO₂, (H₂O)–Na₂O–SiO₂ and (H₂O)–SiO₂ systems to expand the range of water concentration and bulk composition. We could demonstrate that m decreases for all glasses although water is dissolved as OH and should depolymerize the network. An empirical equation of the general type m = a ? b logC_W where a, b are fitting parameters, enables m to be predicted, for each glass series as function of the water content C_W. The enlarged data base shows that the parameter B of the Arrhenius viscosity-temperature relation decreases much stronger than the isokom temperature at the glass transition.     

Field-assisted ion exchange between molten alloys and soda lime glass

Using molten alloys, the electrolysis of soda lime glass has been studied. Ions from a molten anode can be driven into soda lime glass by applying a modest electric potential. The anode current densities have been measured. The temperature and potential dependences have also been measured. The current-time behavior depends on the ion exchange. In a molten alloy the most easily oxidized metal is suited for exchange.     

Redox behavior and diffusion of copper in soda-lime-silica melts

Glasses with the base mol% composition xNa₂O · 10CaO · (90 ? x)SiO₂ with x = 10, 16, 20 and 26 were investigated at high temperatures using square-wave voltammetry. The recorded voltammograms exhibit two peaks. That at less negative or (depending on temperature and glass composition) even positive potential is attributed to the reduction of Cu²⁺ to Cu⁺, while that observed at more negative potential is caused by the reduction of Cu⁺ to metallic copper. For both redox steps, the peak potentials decrease linearly with temperature. Those of the composition with 10 mol% Na₂O show the most negative values. The diffusion coefficients can be fitted to Arrhenius equation. If referenced to the same viscosity, the diffusion coefficients decrease with increasing Na₂O-concentration. The effect of composition on the thermodynamics as well as on diffusivities is explained by the incorporation of the copper ions into the melt structure.     

Conversion of batch to molten glass, I: Volume expansion

Batches designed to simulate nuclear high-level waste glass were compressed into pellets that were heated at 5 K/min and photographed. Three types of batches were prepared, each with different amounts of nitrates and carbonates. The all-nitrate batches were prepared with varying amounts of sucrose. The mixed nitrate-carbonate batches were prepared with silica particles ranging in size from 5 to 195 μm. One batch containing only carbonates was also tested. Sucrose addition had little effect on expansion, while the size of silica was very influential. Sucrose addition reduced primary foam for batches containing 5-μm silica, but had no effect on batches containing larger particles. Excessive amounts of sucrose increased secondary foam. The 5-μm grains had the strongest effect, causing both primary and secondary foam to be generated, whereas only secondary foam was produced in batches with grains of 45 μm and larger. We suggest that the viscosity of the melt and the amount of gas evolved are the main influences on foam production. As more gas is produced in the melt and as the glass becomes less viscous, gas bubbles coalesce into larger cavities until the glass can no longer contain the bubbles and they burst, causing the foam to collapse.     

Interaction of silica glass with molten mixtures of lead and potassium nitrates

The formation of lead–silicate coatings onto the surface of silica glass after its treatment in molten mixtures of Pb(NO₃)₂–KNO₃ and Pb(NO₃)₂–Cu(NO₃)₂–KNO₃ at 420–520 °C was investigated. The coating formation was supported by the appearance and adsorption/sedimentation of lead–oxide cations and nano-particles, from the salt melt, onto the treated glass surface. These particles, deposited onto the substrate, interacted with the structure of silica glass and formed an amorphous lead–silicate coating exhibiting a chemical composition gradient from PbO to SiO₂. The admixture of Cu(NO₃)₂ in the abovementioned salt mixture favored higher rates of lead oxide sedimentation (co-sedimentation with copper oxide). The external potion of this coating was progressively crystallized during treatment in the molten mixture of Pb(NO₃)₂–Cu(NO₃)₂–KNO₃ due to the nucleation of Cu₂O crystals supporting the formation of Pb₈Cu(Si₂O₇)₃, Pb₃O₄ and 2PbO · SiO₂.     

Effect of phase separation on weathering of soda-lime-silica glasses

Accelerated weathering tests on phase-separated and non-separated soda-lime-silica glasses have been made under static conditions of 75% relative humidity at 70°C. Detailed studies of the change of weatherability show that the rate of weathering is reduced if phase separation results in a connected structure of a silica-rich phase, whereas droplets of a silica-rich phase are more liable to weathering. The weathering products were examined by an X-ray wavelength dispersive spectrometer.     

Surface damage of soda-lime-silica glasses: indentation scratch behavior

Contact mechanics problems are of fundamental interest both to understand the process of surface damage and matter removal in brittle materials, and to develop a method to evaluate their scratch resistance. In order to get insight into these problems in the case of soda-lime-silica glasses, a classical indentation apparatus was used, and an original scratch experimental setup was designed, allowing for a monotonic loading (or unloading) of the indenter combined with a controlled sliding of the specimen beneath the indenter. The influences of the normal load, the moisture level and the glass composition have been studied, and clear relationships were established between the glass compositions and the indentation-scratching behavior. The indentation and scratching characteristics such as the critical-crack-initiation loads and the transition loads between the different scratch regimes were correlated and interpreted in the light of the brittleness index and structural considerations.     

Conversion of batch to molten glass, II: Dissolution of quartz particles

Quartz dissolution during the batch-to-glass conversion influences the melt viscosity and ultimately the temperature at which the glass forms. Batches to make a high-alumina borosilicate glass (formulated for the vitrification of nuclear waste) were heated at 5 K min^− 1 and quenched from temperatures of 400 to 1200 °C at 100 K intervals. The batches contained quartz as a silica source, with particles ranging from 5 to 195 μm in diameter. The content of unreacted quartz in the samples was determined with X-ray diffraction. Most of the fine quartz dissolved during the early batch reactions (at temperatures < 800 °C), whereas coarser quartz dissolved mostly in a continuous glass phase via diffusion. The mass-transfer coefficients were assessed from the data as functions of the initial particle sizes and the temperature. A series of batches were also tested that contained nitrated components and additions of sucrose, known to accelerate melting. While sucrose addition had no discernible impact on quartz dissolution, nitrate batches melted somewhat more slowly than batches containing carbonates and hydroxides in addition to nitrates.     

The influence of polyvalent metal cations on the corrosion rate of molybdenum in molten glass

The corrosion behavior of molybdenum in a silicate melt has been studied using electrochemical techniques. The aim was to examine whether or not polyvalent metal ions such as iron (Fe³⁺), nickel (Ni²⁺) and cobalt (Co²⁺) are able to influence the corrosion behavior of molybdenum and, if so, to which extent. In addition the impact of the temperature of the melt on the corrosion rate has been addressed.     

Undoped and calcium doped borate glass system for thermoluminescent dosimeter

Borate glasses present an absorption coefficient very close to that of human tissue. This fact makes some borates ideal materials to develop medical and environmental dosimeters. Glass compositions with calcium tetraborate (CaB₄O₇) and calcium metaborate (CaB₂O₄), such as the xCaB₄O₇ − (100−x)CaB₂O₄ system (0 ⩽ x ⩽ 100 wt%) were obtained by the traditional melting/quenching method. A phenomenon widely known as the ‘boron anomaly’ was observed in our thermal analysis measurements, as indicated by the increase of T_g and the appearance of a maximum value in the composition with 40 wt% of CaB₂O₄. The Dy doped and Li co-doped 80CaB₄O₇–20CaB₂O₄ (wt%) glass samples were studied by the thermoluminescence technique. The addition of Dy improved the signal sensitivity in about three times with respect to the undoped glass sample. The addition of Li as a co-dopant in this glass caused a shift to a lower temperature of about 20 °C in the main glow peak. The structural analysis of the 80CaB₄O₇–20CaB₂O₄ (wt%) undoped and doped samples were studied through infrared absorption. We have noted an increase in the coordination number of the boron atoms from 3 to 4, i.e., the conversion of the BO₃ triangular structural units into BO₄ tetrahedra.     

Doping of a high calcium oxide metaphosphate glass with titanium dioxide

This study investigates the effect of doping a high calcium oxide containing metaphosphate glass series (CaO)₄₀(Na₂O)₁₀(P₂O₅)₅₀ with TiO₂ (1, 3, and 5 mol%). TiO₂ incorporation increased the density and glass transition temperature while reduced the degradation rate (5 mol% in particular) by twofold compared with (CaO)₃₀ system reported previously. This has been confirmed by ion release and the minimal pH changes. TiP₂O₇, NaCa(PO₃)₃ and CaP₂O₆ phases were detected for all TiO₂-containing ceramics. XPS showed that the surface is composed of Ca, P, and Ti. Ti was recognized mainly as TiO₂, but its total amount was lower than theoretical values. ³¹P magic angle spinning (MAS) NMR showed a downfield shift of the ³¹P lineshape with increasing TiO₂, interpreted as an effect of the titanium cation rather than an increase in the phosphate network connectivity. FTIR showed that incorporation of TiO₂ increased the strength of the phosphate chains, and the O/P ratio while introducing more Q¹ units into the structure at the expense of the Q² units. There were no differences, however, in surface topography roughness and free energies between these glasses. These results suggested that TiO₂ and CaO were acting synergistically in producing glasses with controllable bulk and structural properties.     

The effect of impurity ions in molten salt KNO3 on ion-exchange and strengthening of glass

This paper deals with the particular diffusion behavior of the impurity ions in molten salt KNO₃ within glass, namely Ca, Sr, Ba and Na with different concentrations, and the effect of the behavior on K diffusion and the strengthening of ion-exchanged glass.