Abstract
Rutile (R) phase VO2 is a quintessential example of a strongly correlated bad metal, which undergoes a metal-insulator transition (MIT) concomitant with a structural transition to a V-V dimerized monoclinic (M1) phase below TMIT∼340K. It has been experimentally shown that one can control this transition by doping VO2. In particular, doping with oxygen vacancies (VO) has been shown to completely suppress this MIT without any structural transition. We explain this suppression by elucidating the influence of oxygen vacancies on the electronic structure of the metallic R phase VO2, explicitly treating strong electron-electron correlations using dynamical mean-field theory (DMFT) as well as diffusion Monte Carlo (DMC) flavor of quantum Monte Carlo (QMC) techniques. DMC calculations show a gap closure in the M1 phase when vacancies are present, suggesting that when vacancies are introduced in the high-temperature rutile phase, the dimerized insulating phase cannot be reached when temperature is lowered. Both DMFT and DMC calculations of nonstoichiometric metallic rutile phase shows that this tendency not to dimerize in the presence of vacancies is because VO's tend to change the V−3d filling away from its nominal half-filled value, with the eπg orbitals competing with the otherwise dominant a1g orbital. Loss of this near orbital polarization of the a1g orbital is associated with a weakening of electron correlations, especially along the V-V dimerization direction. This removes a charge-density wave (CDW) instability along this direction above a critical doping concentration, which further suppresses the metal-insulator transition. Our study also suggests that the MIT is predominantly driven by a correlation-induced CDW instability along the V-V dimerization direction.
Original language | English |
---|---|
Article number | 155129 |
Number of pages | 9 |
Journal | Physical Review B |
Volume | 101 |
Issue number | 15 |
DOIs | |
Publication status | Published - 23 Apr 2020 |
Publication type | A1 Journal article-refereed |
Publication forum classification
- Publication forum level 2