Comment on "Electronic structure and orbital ordering of SrRu$_{1-x}$Ti$_x$O$_3$: GGA+U investigations"

TL;DR

This paper critiques a previous study on SrRu$_{1-x}$Ti$_x$O$_3$, arguing that the metal-insulator transition is driven by bandwidth changes and disorder, not Coulomb repulsion, emphasizing the importance of bulk spectra in analysis.

Contribution

It highlights the discrepancy between surface-sensitive calculations and bulk experimental spectra, emphasizing the role of bandwidth and disorder over Coulomb interactions in the MIT.

Findings

Bulk spectra show finite density of states at the Fermi level for x ≥ 0.5.

Metal-insulator transition is primarily driven by bandwidth W.

Disorder plays a dominant role in the transition.

Abstract

In the paper, PRB {\bf 77}, 085118 (2008), the authors conclude that the observation of Ti-doping induced half-metallicity in SrRu$_{1-x}$TI$_x$O$_3$ within the limit of local density approximations is not valid as the experimental results indicate insulating behavior. It was described that the metal-insulator transition (MIT) at x = 0.5 observed in this system appears due to the enhancement of on-site Coulomb repulsion strength, U with Ti substitutions, x. The MIT primarily depends on U and partially on x and/or disorder. All these conclusions are in sharp contrast to the experimental observations, which predicted Anderson insulating phase at x = 0.5 (finite localized density of states at the Fermi level). The hard gap due to electron correlation appears at much higher x (~ 0.8). In addition, it is well established that homovalent substitution has negligible influence on on-site $U$ (a local variable). These inconsistencies appear due to the fact that the calculated results representing the bulk electronic structure are compared with the experimental results dominated by surface contributions. The experimental bulk spectra, already available in the literature exhibit finite density of states at the Fermi level even for x >= 0.5 sample, which suggests that the conclusion of half metallic phase in an earlier study is reasonable. Thus, the insulator to metal transition in these systems is driven by bandwidth, W rather than U and disorder plays dominant role in the metal-insulator transition. One needs to consider the bulk spectra to reproduce numerically the bulk electronic structure.