Single Alkali Glasses Research Articles

Mixed alkali effect in Li and Na aluminophosphate glasses: influence of the cation environment

This paper deals with the variation of the mixed alkali effect (MAE) in Li and Na aluminophosphate glasses. The general system investigated is: 0.46 [ xNa 2O(1− x)Li 2O], 0.54 [ yAl 2O 3(1− y)P 2O 5], with x varying between 0 and 1 to probe the MAE, and y between 0 and 0.08 to modify the cation environment and to try to understand how the amorphous host affects the ionic motions. We show that the dc conductivity minimum characteristic of the MAE effect decreases and almost disappears as Al 2O 3 is added to mixed phosphate glasses. On another hand, the dc conductivity of the single alkali glasses is almost not affected by Al 2O 3 addition. Dielectric relaxation characteristics also change with the Al 2O 3 concentration in mixed glasses and not in single ones. On another hand, the mixed alkali peak measured using mechanical relaxation appears to be almost not affected by the presence of Al 2O 3. This decoupling between the mechanical and dielectric relaxation processes appears to «through into question» the proposal that the mechanical relaxation in mixed alkali glasses accounts for sites reconfiguration to adapt to the different cations.

Mixed-alkali effect in Na–Rb borate glasses: A tracer diffusion and electrical conductivity study

The ionic transport in a series of 0.2 [X Na2O·(1 − X)Rb2O]·0.8B2O3 glasses has been studied by measuring both the electrical conductivity of these glasses and the tracer diffusivities of 22Na and 86Rb as functions of composition and temperature. At constant temperature (380 °C) the tracer diffusivities of 22Na and 86Rb decrease monotonically with decreasing Na- or Rb-content, respectively. The tracer diffusivities as functions of the composition parameter X intersect near X = 0.2 (‘crossover composition’). This behaviour gives rise to a minimum in the conductivity which appears near X = 0.4. At the crossover composition X = 0.2 the diffusivities of 22Na and 86Rb are practically the same within the measured temperature range resulting in similar Arrhenius parameters. For each other investigated composition the Arrhenius parameters for 22Na and 86Rb diffusion differ significantly. Comparing the 22Na-diffusivity in the pure Na-borate glass with the corresponding conductivity diffusion coefficient calculated via the Nernst–Einstein equation from the measured conductivity yields a temperature-independent Haven ratio. In contrast, a temperature-dependent Haven ratio was found in the pure Rb-borate glass. In order to compare the tracer diffusivities in mixed-alkali glasses with their ionic conductivity we introduce a ‘common Haven ratio’ for the mixed-alkali glasses which coincides with the conventional definition of the Haven ratio for single-alkali glasses. On the Rb-rich side the common Haven ratio decreases with temperature for Na–Rb borate glasses (X = 0.2 and X = 0.4) similar to the Haven ratio in the pure Rb borate glass.

Ionic conductivity and the mixed alkali effect inLixRb1−xPO3glasses

The temperature and frequency dependent ionic conductivity in ${\mathrm{Li}}_{x}{\mathrm{Rb}}_{1\ensuremath{-}x}{\mathrm{PO}}_{3}$ glasses has been studied using dielectric spectroscopy. The dc conductivity decreases by more than six orders of magnitude on mixing the alkali ions in the glass structure, that is, a strong mixed alkali effect on the dc conductivity was observed. The results show that the mixed alkali effect on the dc conductivity diminishes as the temperature is increased. An ac conductivity mixed alkali effect can be observed up to high frequencies, although it gradually becomes weaker as the frequency is increased. A quantitative analysis of the conductivity spectra shows that the number of mobile cations in the single alkali glasses is, within experimental uncertainty, temperature independent. The results are discussed in relation to a detailed microscopic structural model taking into account the random mixing of alkali ions [Phys. Rev. B 63, 132202 (2001)].

Mixed mobile ion effect in fluoride glasses

Certain ionic materials containing dissimilar mobile cations exhibit the nonlinear behavior of diffusion-related properties, called the mixed mobile ion effect. In these materials only cations participate in the diffusion process. In this paper we report on the mixed mobile ion effect in fluoride glasses, where anions also participate in the diffusion processes in addition to cations, unlike mixed alkali oxide glasses and crystals. We have provided several pieces of evidence in favor of this effect in fluoride glasses. This phenomenon was associated with the low dimensionality of the diffusion pathways in the mixed alkali fluoride glasses compared to single alkali glasses.

Dielectric modulus analysis of mixed alkali Li xRb 1− xPO 3 glasses

Dielectric data of mixed alkali Li x Rb 1− x PO 3 glasses have been analysed in the modulus formalism. The imaginary part of the complex modulus was fitted with a KWW function. The stretching parameter β is larger in the mixed glasses as compared to the single alkali glasses, which indicates that the mixed alkali glasses behave as single alkali glasses of effectively lower concentrations. The finding is consistent with the random ion distribution model, which assumes that the two kinds of alkali ions are moving in distinctly different conduction pathways.

Nonuniversal features of the ac conductivity in ion conducting glasses

Recently, Sidebottom [Phys. Rev. Lett. 82, 3653 (1999)] proposed a new scaling approach for the conductivity spectra of ion conducting glasses. This approach is based on the condition that the shape of the spectra is universal. In this Letter, we show that this condition is generally not fulfilled, but that the shape depends on the glass composition. In single alkali glasses, the frequency dependence of the conductivity varies with the alkali oxide content. Furthermore, the mixing of dissimilar alkali ions leads to pronounced changes in the shape of the conductivity spectra.

EXAFS study of a “mixed-alkali” type effect in sodium-calcium borate glass

Replacement of one alkali in a glass by another is known to have a profound effect on many physical properties in that glass. In the case of electrical conductivity, there is a nonlinear variation between the values of the two relevant single-alkali glasses, with a minimum at some mixed composition. This is referred to as the mixed-alkali effect (MAE), and has often been investigated using EXAFS, which shows a distinct change in the first-shell Debye—Waller factor at a composition usually corresponding to the minimum conductivity. However, investigations of the MAE have normally been restricted to alkali metal oxides, R2O, and other monovalent cation oxides. The authors have undertaken an EXAFS study of a series of glasses of general formula xNa2O·(1—x)CaO· 2B2O3 in order to ascertain whether a “mixed-alkali” type effect occurs between cations of differing charge, and whether any significant change in Debye—Waller factor occurs at the same percentage replacement as in glasses containing cations of similar charge.

Evidence for site memory effects in the ionic relaxation of (Li 2O) x(Na 2O) y(GeO 2) 1−x−y glasses

We report measurements of the dielectric response due to ionic relaxation in alkali germanate glasses (Li 2O) x (Na 2O) y (GeO 2) 1− x − y . The relaxation is investigated in both electric modulus and ac conductivity formalisms, between which important discrepancies are observed. In single alkali glasses, the width of the electric modulus decreases with decreasing alkali concentration. However, the power law dispersion of the ac conductivity scales regardless of the alkali concentration. This scaling property of the ac conductivity implies that inter-ionic interactions are not responsible for the non-Debye relaxation. In mixed alkali glasses, the exponent of the power law dispersion of the ac conductivity decreases and the dielectric relaxation strength increases relative to the single alkali glasses. With increasing total alkali concentration in the mixed alkali glasses, the frequency of maximum dielectric loss decreases, approaching frequencies comparable to that of the mixed alkali mechanical loss peak. These findings indicate a transition of the ionic relaxation toward a slower polarization process which is consistent with site memory mechanisms recently proposed for the mixed alkali effect.

Molecular dynamics simulations of structural changes in mixed alkali (Li–K) silicate glasses

The structure of Li-silicate, K-silicate and mixed (Li–K) silicate glasses, of composition 25R 2O·75SiO 2, has been investigated by molecular dynamics simulations. The environment of each type of alkali ion differs from that of the other type, regardless of whether in the single or mixed alkali glass, though there are structural changes. There exists a site mismatch energy, which can act as a barrier for diffusion of alkali ions. Alkali ions preferentially jump into sites that were previously occupied by the same type of alkali ions. We attribute the mixed alkali effect to site differences and a decrease in the available number of neighboring sites when the alkali ions are mixed. The alkali ions diffuse by various ways in the silicate glasses. Among them, vacancy and exchange mechanisms are directly observed in this study and they seem to have a resemblance to diffusion in a crystalline phase.

Influence of the spatial distribution of sites on the ion transport in glasses

This paper presents a computational model which mimics the ionic conductivity on single alkali glasses. A Monte Carlo simulation procedure is employed to investigate the transport of ions between sites with spatial distribution. The results of the simulation shows that the existence of a site dispersion plays an essential role and seems to be sufficient to produce the dependence of the conductivity on the modifier content which was found experimentally.

Ionic conduction and mixed cation effect in silicate glasses and liquids: 23Na and 7Li NMR spin-lattice relaxation and a multiple-barrier model of percolation

Abstract The mixed cation effect in glasses affects transport properties such as diffusion and ionic conductivity. The temperature dependence of 7Li and 23Na NMR spin-lattice relaxation times, T1, have been measured in a series of trisilicate (Na2Si3O7, NaRbSi3O7, NaCsSi3O7 and NaCa0.5Si3O7) and disilicate (Li2Si2O5, LiKSi2O5 and LiBa0.5Si2O5) glasses in order to understand the atomic scale mechanism of this effect. Diffusion-induced NMR spin-lattice relaxation of alkali cations is shown to be affected by a mixed cation effect which tends to shift the T1 minimum to a higher temperature and increases the slope of the T1 vs. 1000 T curve on the low temperature side of the minimum. These changes are consistent with an effective reduction in the number of hopping sites available to the alkali cation in a mixed cation glass. As a result the high energy cutoff for percolation in a disordered barrier height landscape increases significantly on going from a single alkali to a mixed alkali (or, alkali-alkaline earth) glass. The dc conductivities calculated on the basis of a random walk model with different percolation thresholds for single and mixed alkali glasses are in reasonably good agreement with the experimental results. This interpretation of the NMR spin-lattice relaxation results, thus, successfully combines the two recent models — that of Svare et al. on the general relationship between ionic transport and spin-lattice relaxation and of Maass et al. on mixed alkali effect resulting from strong energetic preference for hopping associated with percolation transition. In addition, the mixed alkali effect can be generalized to mixed alkali-alkaline earth systems.

Structure of mixed alkali glasses

Some theories for the anomalous conduction phenomenon of the mixed alkali glasses predict or rely on a particular type of glass structure; either a clustering tendency with positive enthalpy of mixing or an ordering tendency with negative enthalpy of mixing between two dissimilar alkali species. For example, according to a theory based upon the thermodynamic state of glasses suggested by the present author and his collaborator, the large static dielectric relaxation strength and corresponding low dc conductivity observed for mixed alkali glasses indicate a negative enthalpy of mixing. Independent experiments, however, have shown both positive enthalpy of mixing and negative enthalpy of mixing, Mixed alkali glasses have larger static dielectric constant but smaller high frequency dielectric constant than the corresponding single alkali glasses. This difference indicates that the mixed alkali glasses have a negative enthalpy of mixing under dc electric fields but a positive enthalpy of mixing under high frequency electric field. It is suggested that the structural studies made under dc electric field or involving ion-exchange are relevant to the mixed alkali effect.

Application of percolation theory in composites and glasses

We discuss percolation models for ionic transport in dispersed ionic conductors and in glassy materials. In dispersed ionic conductors, the model takes into account that the ionic conductivity is strongly enhanced along the interface between (bad) ionic conductor and dispersed insulator. This leads to the occurrence of highly conducting percolating pathways at intermediate insulator concentration, by which the anomalous physical properties of the mixture can be satisfactorily described. In glassy ionic conductors, the model is based on the experimental evidence that ions maintain their distinct local environments. This leads to the formation of fluctuating percolation pathways. The connectivity of these pathways determines the ion mobility in the glassy network, and leads to an understanding of the well known transport anomalies in single and mixed alkali glasses.

Relaxations in mixed alkali metaphosphate glass

A study of the materials response to time-dependent mechanical and electrical perturbations was performed on a mixed alkali, sodium and lithium, metaphosphate glass over a range of overlapping frequencies. The characteristic timescale that characterizes the mechanical relaxations, τ μ , was observed to be two orders of magnitude slower than the characteristic response, τ σ , of the system to electrical perturbations. This is unlike the situation in single alkali glasses where τ μ and τ σ are equivalent. Interestingly, it was also noted that the breadth of the distribution of relaxation times that characterize the response of the mixed mobile ion system is slightly narrower than that of the single alkali (Li) metaphosphate. A comparison of the characteristic timescales, τ 1, that describes the nuclear magnetic resonance spin-lattice relaxation time response revealed that τ 1 is comparable to τ σ . This is also by contrast with the single alkali case where τ σ is approximately one order of magnitude faster than τ 1. However, as observed in single alkali glasses, the breadth of the distribution of relaxation times that characterize the loss processes of the mechanical response is considerably larger than that of the electrical response.

Influence of alkali content and alkali mixing on the chemical durability of fluorozirconate glasses

The corrosion of alkali-containing fluorozirconate glasses in water and acidic and alkaline solutions was studied with glasses of the compositions (100−x) (0.6ZrF4·0.1AlF3·0.3BaF2)· xLiF and 48ZrF4·8AlF3·24BaF2·xLiF·(20−x)NaF. The corrosion of the glasses in deionized water and 0.1 n HCl solution is mainly controlled by diffusion with an inductive period due to the passage from reaction-controlled to diffusion-controlled mechanisms, and the weight loss of glass increases with increasing LiF content in the single-alkali glasses. There was no appreciable effect of alkali mixing on the corrosion of glasses in deionized water and 0.1 n HCl solution. The glasses exhibited good stability in alkaline solution. The weight loss due to corrosion of the glasses in acidic buffered solution increased exponentially with increasing LiF content.

Internal Friction of Mixed Alkali Aluminogermanate Glasses

The internal friction of alkali aluminogermanate glasses containing no nonbridging oxygen ions was measured by the free torsional vibration method. The internal friction as a function of temperature revealed two peaks in mixed-alkali glasses with low alkali contents. Those peaks increased rapidly in magnitude and came close to overlapping with increasing alkali content. The results indicate that both peaks can be assigned as mixed-alkali peaks, and that the mixed-alkali peak was closely related to the low-temperature peak in single alkali glasses. The following explanation is proposed. (1) There are two kinds of mechanism responsible for the mixed-alkali peak. (2) The mixed-alkali peak is essentially caused by the individual and duplex movement of two species of alkali ions which are affected by the interaction between different kinds of alkali ions. (3) When two mixed-alkali peaks overlap, there is little difference between the two mechanisms since the two species of alkali ions move with almost the same activation energy. (4) The rapid increase in magnitude of the mixed-alkali peak with increasing alkali content is probably attributed to a considerable increase in the number of mobile alkali ions which contribute to the mechanism of the mixed-alkali peak, and at the same time is also attributed to the increase in energy consumption during alkali ion migration.

Mechanism of the mixed alkali effect based upon the thermodynamic state of glass

A new mechanism for the mixed alkali effect of dc conductivity is proposed. It was shown earlier that the high frequency conductivity shows very little mixed alkali effect and that mixed alkali glasses show a much larger dielectric relaxation strength than single alkali glasses, indicating that the mixed alkali effect in dc conductivity is related to the difference in the dielectric relaxation strength. An expression for dielectric relaxation strength in terms of the thermodynamic state was experimentally verified for single alkali glasses. An analogous equation for dielectric relaxation strength of mixed alkali glasses was derived. Using the regular solution model, it was shown that the dielectric relaxation strength for mixed alkali glasses can show a large maximum if the enthalpy of mixing between the two alkalis is negative. A large negative enthalpy of mixing between the two alkali species was found experimentally from ion-exchange equilibrium, which, it is suggested, originates from the stress due to their ionic size difference. The proposed mechanism appears to be consistent with many experimental observations.

Change in boron coordination in alkali borate glasses, and mixed alkali effects, as elucidated by NMR

NMR with substantial enhancement in signal-to-noise ratio (primarily from gisnal averaging and fast passage techniques) and thermal analysis are utilized to reinvestigate binary alkali borate glasses. The experimental results for the Li, Na, K, Rb and Cs borate glasses at various alkali contents show a strong dependence of the fraction, N 4, of 4-coordinated borons on the particular alkali ion. NMR and thermal analysis experimental results for half-and-half mixed LiNa, LiK, LiRb and LiCs borate glasses at different total alkali content are also presented in this paper. It was found that the measured N 4 for the mixed alkali glass is less than the value expected from a simple superposition of the values of the constituent alkalis, and in some cases is even less than either of the N 4 values for the relevant single alkali glasses. 11B and 133Cs nuclear magnetic resonance results and T g measurements are used to strengthen the evidence supporting the pairing model suggested by several workers for the mixed alkali effect.

Haven ratio in mixed alkali glass

It is known that the Haven ratios, H R, in mixed alkali glasses increase anomalously. The authors have attempted to explain quantitatively such anomalous behavior by introducing the thermodynamic activity term in the Nernst-Planck equation, which results from the attractive interaction between dissimilar alkali ions via oxygen ion. As a result, it was found that the apparently large values of H R in the mixed alkali glasses can be reduced to the same as in the single alkali glasses when corrected by the thermodynamic term. This fact also indicates that the ionic diffusion mechanism in mixed alkali glasses does not differ from that in single alkali glasses.

Properties of mixed alkali fluoride glasses in the ZrF 4PbF 2AlF 3RF (R = Li, Na, K) system

The densities, refractive indices, thermal expansions, hardnesses and electrical properties of mixed alkali fluoride glasses of composition 40ZrF 4 · 20PbF 2 · 10AlF 3 · 30RF, where RF are the LiF-NaF, LiF-KF and NaF-KF pairs, have been measured and those and derived properties have been discussed in terms of the mixed alkali effect. It is found that the density, refractive index, molar volume, molar refractivity, thermal expansion coefficient and glass transition and softening temperatures do not exhibit the appreciable mixed alkali effect, while the electrical conductivity and dielectric loss show the appreciable mixed alkali effect as observed in silicate and other oxide glasses. It is found that there is a small mixed alkali effect in Vickers hardness. The magnitude of the mixed alkali effect in electrical conductivity and dielectric loss is larger when the difference of the activation energy for conduction between two constituent single alkali glasses is larger.