Oxygen vacancy formation and electronic reconstruction in strained <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:msub><mml:mi>LaNiO</mml:mi><mml:mn>3</mml:mn></mml:msub></mml:math> and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:msub><mml:mi>LaNiO</mml:mi><mml:mn>3</mml:mn></mml:msub><mml:mo>/</mml:mo><mml:msub><mml:mi>LaAlO</mml:mi><mml:mn>3</mml:mn></mml:msub></mml:math> superlattices

Abstract

By using density functional theory calculations including a Coulomb repulsion term, we explore the formation of oxygen vacancies and their impact on the electronic and magnetic properties of strained bulk ${\mathrm{LaNiO}}_{3}$ and (${\mathrm{LaNiO}}_{3}{)}_{1}$/(${\mathrm{LaAlO}}_{3}{)}_{1}$(001) superlattices. For bulk ${\mathrm{LaNiO}}_{3}$, we find that epitaxial strain induces a substantial anisotropy in the oxygen vacancy formation energy. In particular, tensile strain promotes the selective reduction of apical oxygen, which may explain why the recently observed superconductivity in infinite-layer nickelates is limited to strained films. For (${\mathrm{LaNiO}}_{3}{)}_{1}$/(${\mathrm{LaAlO}}_{3}{)}_{1}$(001) superlattices, the simulations reveal that the ${\mathrm{NiO}}_{2}$ layer is most prone to vacancy formation, whereas the ${\mathrm{AlO}}_{2}$ layer exhibits generally the highest formation energies. The reduction is consistently endothermic, and a largely repulsive vacancy-vacancy interaction is identified as a function of the vacancy concentration. The released electrons are accommodated exclusively in the ${\mathrm{NiO}}_{2}$ layer, reducing the vacancy formation energy in the ${\mathrm{AlO}}_{2}$ layer by $\ensuremath{\sim}70%$ with respect to bulk ${\mathrm{LaAlO}}_{3}$. By varying the vacancy concentration from 0 to $8.3%$ in the ${\mathrm{NiO}}_{2}$ layer at tensile strain, we observe an unexpected transition from a localized site-disproportionated ($0.5%$) to a delocalized ($2.1%$) charge accommodation, a reentrant site disproportionation leading to a metal-to-insulator transition despite a half-filled majority-spin Ni ${e}_{g}$ manifold ($4.2%$), and finally a magnetic phase transition ($8.3%$). While a band gap of up to 0.5 eV opens at $4.2%$ for compressive strain, it is smaller for tensile strain or the system is metallic, which is in sharp contrast to the defect-free superlattice. The strong interplay of electronic reconstructions and structural modifications induced by oxygen vacancies in this system highlights the key role of an explicit supercell treatment beyond rigid-band methods and exemplifies the complex response to defects in artificial transition metal oxides.

Topics & Concepts