Examining the metal-to-insulator transitions in Li1+xTi2-xO4 and LiAlyTi2-yO4 with a Quantum Site Percolation model

TL;DR

This study uses a quantum site percolation model to analyze metal-insulator transitions in lithium titanate compounds, finding discrepancies with experimental data that suggest the importance of strong electronic correlations.

Contribution

The paper applies a quantum site percolation model to cation-substituted lithium titanate, highlighting its limitations and the potential role of electronic correlations in these transitions.

Findings

Model predicts higher critical Al concentration than observed experimentally.

Discrepancies suggest strong correlations are significant.

Model captures configurational disorder effects but not electron correlations.

Abstract

We have studied the composition-induced metal-to-insulator transitions of cation substituted Lithium Titanate, in the forms Li1+xTi2-xO4 and LiAlyTi2-yO4, utilising a quantum site percolation model, and we argue that such a model provides a very reliable representation of the noninteracting electrons in this material if strong correlations are ignored. We then determine whether or not such a model of 3d electrons moving on the Ti (corner-sharing tetrahedral) sublattice describes the observed metal-to-insulator transitions, with the critical concentration defined by the matching of the mobility edge and the chemical potential. Our analysis leads to quantitative predictions that are in disagreement with those measured experimentally. For example, experimentally for the LiAlyTi2-yO4 compound an Al concentration of y_c approximately 0.33 produces a metal-to-insulator transition, whereas our analysis of a quantum site percolation model predicts y_c approximately 0.83. One hypothesis that is consistent with these results is that since strong correlations are ignored in our quantum site percolation model, which includes the effects of configurational disorder only, such strong electronic correlations are both present and important.