Vanadium Oxidation States and Structural Role in Aluminoborosilicate Glasses: An Integrated Experimental and Molecular Dynamics Simulation Study

Abstract

Vanadium-containing glasses have aroused interest in several fields such as electrodes for energy storage, semiconducting glasses, and nuclear waste disposal. The addition of V 2 O 5 , even in small amounts, can greatly alter the physical properties and chemical durability of glasses; however, the structural role of vanadium in these multicomponent glasses and the structural origins of these property changes are still poorly understood. We present a comprehensive study that integrates advanced characterizations and atomistic simulations to understand the composition−structure− property relationships of a series of vanadium-containing aluminoborosilicate glasses. UV−vis spectroscopy, X-ray photoelectron spectroscopy, and X-ray absorption near-edge structure (XANES) have been used to investigate the complex distribution of vanadium oxidation states as a function of composition in a series of six-component aluminoborosilicate glasses. High-energy X-ray diffraction and molecular dynamics simulations were performed to extract the detailed short-and medium-range atomistic structural information such as bond distance, coordination number, bond angle, and network connectivity, based on recently developed vanadium potential parameters. It was found that vanadium mainly exists in two oxidation states: V 5+ and V 4+ , with the former being dominant (∼80% from XANES) in most compositions. V 5+ ions were found to exist in 4-, 5-, and 6-fold coordination, while V 4+ ions were mainly in 4-fold coordination. The percentage of 4-fold-coordinated boron and network connectivity initially increased with increasing V 2 O 5 up to around 5 mol % but then decreased with higher V 2 O 5 contents. The structural role of vanadium and the effect on glass structure and properties are discussed, providing insights into future studies of sophisticated structural descriptors to predict glass properties from composition and/or structure and aiding the formulation of borosilicate glasses for nuclear waste disposal and other applications.

Predicting initial dissolution rates using structural features from molecular dynamics simulations

“…ref. Lu et al 45 Lu et al Model fitting was performed using JMP version 14.0.0 (SAS Institute, Cary, NC). The model form (linear mixture) is described in the following equation:…”

“…The 28 glasses studied (referred hereafter as the V series, 44,45 Zr series, 33,46 and CJ series 22,32 ) were previously fabricated, characterized experimentally, and simulated using MD. Table 1 lists the glass compositions and the initial dissolution rates (r 0 ) both determined experimentally in this work and from references.…”

“…The potentials have been successfully used to simulate a range of borosilicate and boroaluminosilicate glasses such as ISG and related systems. 32,33,45 TA B L E 1 Target glass composition (mol%), initial dissolution rates experimentally measured at different pH values, and studies that characterized (Exp. ref.…”

“…The impacts of V 2 O 5 on the properties of ISG have been investigated previously, 44 where the addition of V 2 O 5 was found to decrease glass transition temperature ( T g ) and hardness while increasing the refractive index and crack initiation load. It was previously found that vanadium mainly exists in two oxidation states in ISG‐based glasses, ∼80% V 5+ (in four‐, five‐, and six‐fold coordination) and ∼20% V 4+ (mainly in fourfold coordination), by an integrated experimental and MD simulation study 45 …”

“…It was previously found that vanadium mainly exists in two oxidation states in ISG-based glasses, ∼80% V 5+ (in four-, five-, and six-fold coordination) and ∼20% V 4+ (mainly in fourfold coordination), by an integrated experimental and MD simulation study. 45 This paper is arranged as follows: methodology, including data collection, r 0 measurement, and model fitting; then the results on structural features obtained from MD, dissolution data, and a model to predict r 0 from the structural features (e.g., former linkages); and finally discussions relating the model to others in the literature and conclusions.…”