Diffusion in Metallic Glasses and in Oxide Glasses - An Overview

2021 ◽  
Vol 413 ◽  
pp. 109-124
Author(s):  
Helmut Mehrer

We remind the reader to some common features of metallic and oxide glasses. We then introduce the radiotracer method for diffusion studies, which can be applied for both types of glasses. We provide an overview on diffusion in metallic glasses in which we consider both types of metallic glasses – conventional and bulk metallic glasses. In the last part we discuss diffusion and ionic conduction in oxide glasses. For ionic glasses, conductivity measurements are an important complement to tracer diffusion studies. We remind the reader to the method of impedance spectroscopy. We discuss results for soda-lime silicate glasses, single alkali borate glasses and mixed alkali borate glasses and present evidence for collective jump processes in glasses.

2011 ◽  
Vol 312-315 ◽  
pp. 1184-1197
Author(s):  
Helmut Mehrer

This paper reviews typical results of tracer diffusion and ionic conduction in soda-lime silicate glass and in single-alkali and mixed-alkali borate glass obtained in our laboratory and published in detail elsewhere. We have studied tracer diffusion of modifier cations and ionic conduction as functions of composition, temperature and, in the case of borate glass, also as function of pressure. We compare tracer diffusion with charge diffusion and in the case of soda-lime glass also with viscosity diffusion. The Haven ratios for soda-lime glass are temperature independent. For sodium borate glass the Haven ratio is almost temperature- and pressure-independent, whereas it decreases significantly with decreasing temperature and increasing pressure for rubidium borate glass. It also decreases with increasing alkali content. We attribute these facts to collective atomic jump events, in which several ions move simultaneously in a string-like or chain-like fashion. We also illustrate the mixed-alkali effect, which was studied by conductivity measurements and by tracer diffusion for mixed sodium-rubidium borate glasses.


2016 ◽  
Vol 6 ◽  
pp. 59-106 ◽  
Author(s):  
Helmut Mehrer

In this Chapter we review knowledge about diffusion and cation conduction in oxide glasses. We first remind the reader in Section 1 of major aspects of the glassy state and recall in Section 2 the more common glass families. The diffusive motion in ion-conducting oxide glasses can be studied by several techniques – measurements of radiotracer diffusion, studies of the ionic conductivity by impedance spectroscopy, viscosity studies and pressure dependent studies of tracer diffusion and ion conduction. These methods are briefly reviewed in Section 3. Radiotracer diffusion is element-specific, whereas ionic conduction is not. A comparison of both types of experiments can throw considerable light on the question which type of ions are carriers of ionic conduction. For ionic conductors Haven ratios can be obtained from the tracer diffusivity and the ionic conductivity for those ions which dominate the conductivity.In the following sections we review the diffusive motion of cations in soda-lime silicate glass and in several alkali-oxide glasses based mainly on results from our laboratory published in detail elsewhere, but we also take into account literature data.Section 4 is devoted to two soda-lime silicate glasses, materials which are commonly used for window glass and glass containers. A comparison between ionic conductivity and tracer diffusion of Na and Ca isotopes, using the Nernst-Einstein relation to deduce charge diffusivities, reveals that sodium ions are the carriers of ionic conduction in soda-lime glasses. A comparison with viscosity data on the basis of the Stokes-Einstein relation shows that the SiO2 network is many orders of magnitude less mobile than the relatively fast diffusing modifier cations Na. The Ca ions are less mobile than the Na ions but nevertheless Ca is considerably more mobile than the network.Section 5 summarizes results of ion conduction and tracer diffusion for single Na and single Rb borate glasses. Tracer diffusion and ionic conduction have been studied in single alkali-borate glasses as functions of temperature and pressure. The smaller ion is the faster diffusing species in its own glass. This is a common feature of all alkali oxide glasses. The Haven ratio of Na in Na borate glass is temperature independent whereas the Haven ratio of Rb diffusion in Rb borate glass decreases with decreasing temperature.Section 6 reviews major facts of alkali-oxide glasses with two different alkali ions. Such glasses reveal the so-called mixed-alkali effect. Its major feature is a deep minimum of the conductivity near some middle composition for the ratio of the two alkali ions. Tracer diffusion shows a crossover of the two tracer diffusivities as functions of the relative alkali content near the conductivity minimum. The values of the tracer diffusivities also reveal in which composition range which ions dominate ionic conduction. Tracer diffusion is faster for those alkali ions which dominate the composition of the mixed glass.Section 7 considers the pressure dependence of tracer diffusion and ionic conduction. Activation volumes of tracer diffusion and of charge diffusion are reviewed. By comparison of tracer and charge diffusion the so-called Haven ratios are obtained as functions of temperature, pressure and composition. The Haven ratio of Rb in Rb borate glass decreases with temperature and pressure whereas that of Na in Na borate glass is almost constant.Section 8 summarizes additional common features of alkali-oxide glasses. Activation enthalpies of charge diffusion decrease with decreasing average ion-ion distance. The Haven ratio is unity for large ion-ion distances and decreases with increasing alkali content and hence with decreasing ion-ion distance.Conclusions about the mechanism of diffusion are discussed in Section 9. The Haven ratio near unity at low alkali concentrations can be attributed to interstitial-like diffusion similar to interstitial diffusion in crystals. At higher alkali contents collective, chain-like motions of several ions prevail and lead to a decrease of the Haven ratio. The tracer diffusivities have a pressure dependence which is stronger than that of ionic conductivity. This entails a pressure-dependent Haven ratio, which can be attributed to an increasing degree of collectivity of the ionic jump process with increasing pressure. Monte Carlo simulations showed that the number of ions which participate in collective jump events increases with increasing ion content – i.e. with decreasing average ion-ion distance. For the highest alkali contents up to four ions can be involved in collective motion. Common aspects of the motion process of ions in glasses and of atoms in glassy metals are pointed out. Diffusion in glassy metals also occurs by collective motion of several atoms.Section 10 summarizes the major features of ionic conduction and tracer diffusion and its temperature and pressure dependence of oxide glasses.


2014 ◽  
Vol 1 ◽  
pp. 125-151 ◽  
Author(s):  
Helmut Mehrer

Firstly, this paper reminds the reader of some basic facts about the glassy state, then of the various ways to produce amorphous metals with particular emphasis on the route of vitrification from the melt. Vitrification of an undercooled melt is the most important route from the viewpoint of the application of metallic glasses. We compare diffusion in some metallic glasses with related crystalline metals. Glassy metals, also called metallic glasses, comprise conventional [1] and bulk metallic glasses [2,3]. We remind the reader of the major experimental techniques for diffusion studies in metallic glasses. The paper then reviews our current understanding of diffusion in glassy metals (see also [4,5,6]), including conventional as well as bulk metallic glasses and undercooled melts. We cover the temperature dependence of diffusion in metallic glasses and discuss the spectrum of activation parameters of glassy metals and its difference to the corresponding one of crystalline metals. We mention the pressure dependence and the isotope effect and we discuss tracer diffusion and viscosity diffusion for a bulk metallic glass and its undercooled melt. Finally we mention computer simulations of atomic jump processes. The diffusion mechanism in metallic glasses differs from that in crystalline metals and involves thermally activated, highly collective (chain-like or caterpillar-like) diffusion jumps. Finally, we mention diffusion along shearbands in a plastically deformed glassy metal.


1994 ◽  
Vol 369 ◽  
Author(s):  
Sanjeev Aggarwal ◽  
Rudiger Dieckmann

AbstractCation diffusion in the spinel solid solution (Fe1-xTix)3-δO4 (0≤ x ≤ 0.3) was investigated at 1200 ºC as a function of oxygen activity, aO2 and cationic composition, x. At different cationic compositions, cation tracer diffusion coefficients, D*Me of Me = Fe and Ti were measured as a function of oxygen activity. Plots of log DMe vs. loga0 show V-shaped curves, indicating that different types of point defects prevail at high anc low oxygen activities. Thermogravimetric experiments were conducted, using a high resolution microbalance, to determine the deviation from stoichiometry in (Fe1-xTix)3-δO4 at 1200 °C. δversus log aO2 curves are S-shaped. An analysis of the oxygen activity dependences of thecation diffusion coefficients and the deviation from stoichiometry with regardto the point defect structure suggests that at high oxygen activities cation vacancies are the predominant defects governing the deviation from stoichiometry and the diffusion ofcations. At low oxygen activities, and at small values of x, cation interstitials determine the deviation from stoichiometry, while they dominate for 0 ≤ x ≤ 0.3 inthe cation diffusion.


Sign in / Sign up

Export Citation Format

Share Document