Anomalous Ionic Conductivity Increase in Li₂S + GeS₂ + GeO₂ Glasses

Abstract

Numerous studies of the ionic conductivities in oxide-doped chalcogenaide glasses have shown the anomalous result that the ionic conductivity actually increases significantly (by more than a factor of 10 in some cases) by the initial addition of an oxide phase to a pure sulfide glass. After this initial sharp increase, the conductivity then monotonically decreases with further oxide addition. While this behavior is important to the application of these glasses for Li batteries, no definitive understanding of this behavior has been elucidated. To examine this effect further and more completely, the ionic conductivities of 0.5Li2S + 0.5[(1 − x)GeS2 + xGeO2] glasses have been measured on disc-type bulk glasses. The ionic conductivity of the 0.5Li2S + 0.5GeS2 (x = 0) glass was observed to increase from 4.3 × 10-5 (Ω cm)-1 to 1.5 × 10-4 (Ω cm)-1 while the activation energy decreased to 0.358 eV from 0.385 eV by the addition of 5 mol % of GeO2. Further addition of GeO2 monotonically decreased the conductivity and increased the activation energy. On the basis of our previous studies of the structure of this glass system, the Anderson and Stuart model was applied to explain the decrease in the activation energy and increase in the conductivity. It is suggested that the "doorway" radius between adjacent cation sites increases slightly (from 0.29(±0.05) Å to 0.40(±0.05) Å) with the addition of oxygen to the glass and is proposed to be the major cause in decreasing the activation energy and thereby increasing the conductivity. Further addition of oxides appears to contract the glass structure (and the doorway radius) leading to an increase in the conductivity activation energy and a decrease in the conductivity. Numerous studies of the ionic conductivities in oxide-doped chalcogenaide glasses have shown the anomalous result that the ionic conductivity actually increases significantly (by more than a factor of 10 in some cases) by the initial addition of an oxide phase to a pure sulfide glass. After this initial sharp increase, the conductivity then monotonically decreases with further oxide addition. While this behavior is important to the application of these glasses for Li batteries, no definitive understanding of this behavior has been elucidated. To examine this effect further and more completely, the ionic conductivities of 0.5Li 2 S + 0.5[(1 -x)GeS 2 + xGeO 2 ] glasses have been measured on disc-type bulk glasses. The ionic conductivity of the 0.5Li 2 S + 0.5GeS 2 (x ) 0) glass was observed to increase from 4.3 × 10 -5 (Ω cm) -1 to 1.5 × 10 -4 (Ω cm) -1 while the activation energy decreased to 0.358 eV from 0.385 eV by the addition of 5 mol % of GeO 2 . Further addition of GeO 2 monotonically decreased the conductivity and increased the activation energy. On the basis of our previous studies of the structure of this glass system, the Anderson and Stuart model was applied to explain the decrease in the activation energy and increase in the conductivity. It is suggested that the "doorway" radius between adjacen...

Ionic Conductivity of Mixed Glass Former 0.35Na2O + 0.65[xB2O3 + (1 – x)P2O5] Glasses

“…At x = 0.4, B 4 makes up 37% of the SRO structural units and at x = 0.5 B 4 makes up 40% of the SRO structural units according to our previously reported NMR data. 51 However, at x = 0.4 and 0.5, phosphorus units make up 60 and 50% of the SRO units.…”

“…Therefore, the B 4 unit must bridge to more phosphorus units than boron units, especially at lower x values, where less boron is present. Because P 5+ has a higher charge density than B 3+ , P−O−B bonding would tend to polarize the electron cloud around the BO toward phosphorus and thereby decrease the net negative charge density of the B 4 unit. This decreased charge density means the Na + ion associated with a B 4 unit would be expected to be less tightly bound due to the smaller effective negative charge on the BO and hence produce a smaller binding energy for the Na + ion to hop to the next charge compensation site.…”

“…Such charge delocalization and resonance would necessarily increase the average bond distance of Na + ions and decrease the negative on the BOs in these glasses and lead to a decreasing columbic binding energy. The added effect here is that in the region of the conductivity activation energy minimum, x ≈ 0.4, the P 5+ bonding further reduces the Coulombic binding energy by pulling negative charge away from the BO 4 − groups. These correlations appear to be well-borne out in the data, as Figure 16 shows the composition dependence of the fraction B 4 units scaled to a linear trend between 0 for x = 0 and ∼0.5 for x = 1, the scaled composition dependence of the relative permittivity to a linear trend between the x = 0 sodium phosphate glass and x = 1 sodium borate glass, and finally the scaled σ dc values, again scaled to the linear trend between the x = 0 and 1 glass.…”

Mixed Barrier Model for the Mixed Glass Former Effect in Ion Conducting Glasses

“…(1) is only a function of E AB =E B and z (independent of p c ), E ðxÞ=E a ð1Þ ¼ fðE AB =E B ; zÞ, which allows us to determine these remaining two parameters by a least-square fit. Figure 1 shows such an analysis for the two systems [9]. In view of the scatter in the experimental data, the fitted curves are in fair agreement for parameter values z ' 2 and a barrier reduction E AB =E B ' 60%.…”

“…This phenomenon is commonly referred to as the ''mixed glass former effect'' (MGFE) [1]. The MGFE has been found in various glass systems, e.g., SiO 2 -B 2 O 3 [2], P 2 O 5 -B 2 O 3 [3,4], GeS 2 -SiS 2 [5], P 2 O 5 -TeO 2 [6], TeO 2 -B 2 O 3 [7], MoO 3 -TeO 2 [8], and GeO 2 -GeS 2 [9] mixed glass former systems.…”