Improved bond valence models for ion transport pathways in glasses.
- PDF / 348,531 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 28 Downloads / 188 Views
1152-TT03-14
Improved bond valence models for ion transport pathways in glasses. Stefan N. Adams1 1
Department of Materials Science and Engineering, National University of Singapore, Singapore
117574, SINGAPORE
ABSTRACT Structure property function relationships provide valuable guidelines for a systematic development of functional materials. It is demonstrated how an augmented bond valence (BV) approach helps to establish such relationships in solid electrolytes. In principle it permits to identify mobile species, transport pathways and provides estimates for ion mobilities. In this work we discuss ion conduction pathways in glassy Lithium metasilicate as an illustrative example. The required representative local structure model is derived from Molecular Dynamics simulations, which provides the opportunity to compare the bond-valence-based predictions from a static structure model with a comprehensive analysis of a complete simulation trajectory. It is shown that understanding the bond valence mismatch as an effective Morse-type interaction opens up a way for systematically analyzing ion transport pathways and for a generally applicable method with improved reliability to predict ion transport characteristics in solid electrolytes from the structure model.
INTRODUCTION A fundamental understanding of ionic motion in disordered solids requires the knowledge of the dependence of mobility on the structural and energetic local environment of the mobile species. Structural information can be obtained experimentally by a wide range of techniques: the most detailed information on the average structure of crystalline compounds is available from neutron and x-ray diffraction, but also NMR, EXAFS and vibrational spectroscopy provide additional insight. For glassy solid electrolytes local structure models may be derived from diffraction data, but additional constraints based on density, chemical plausibility or spectroscopic information are necessary to generate static snapshot-type models that capture essential features of the local glass structure by reverse Monte Carlo (RMC) fitting.[1-4]. Still it has to be emphasized that RMC models are no unique structure solutions and their interpretation is thus limited to a statistical extraction of characteristic features. Alternatively the relation between structure and conductivity can be investigated using molecular dynamics (MD) simulations, which in principle allows to derive all the required structural and dynamical
information within the limitations imposed by the system size, the simulated period and the agreement of the employed model parameters with reality. Effectively the MD approach has shown to be a useful tool in obtaining insight into the conduction mechanism and its correlation to the atomic structure. In previous work we discussed how the bond valence (BV) method can be utilized to perform such a statistical analysis of ion transport pathways yielding predictions of ionic conductivity.[5-9] Here, the investigation is extended by discussing a way to optimize th
Data Loading...
