Quantum Materials for Energy Efficient Neuromorphic Computing: Defective La_1-xSr_xCoO_3-d

We present a first principle study of La_1-xSr_xCoO_3-d, a promising candidate material to build devices for neuromorphic computers. In particular we investigate how the generation of oxygen vacancies can drive a metal-to-insulator transition in the system, with focus on the interplay between structural, electronic and magnetic properties.

Using DFT+U and the Quantum Espresso package (https://github.com/QEF/q-e/releases/tag/qe-6.1.0), we found that increasing the oxygen vacancy concentration from 0 % to ~11 % leads to a change of the cobalt’s oxidation state and of the magnetic state of the system, accompanied by a variation of the octahedral tilt angle and lattice parameters. Altogether these changes determine the opening of the fundamental gap. For x = 0 (no Sr doping) and d varying from 0 to 0.5, we observe a structural transformation from a perovskite phase (semiconductor) to a brownmillerite phase (insulator), consistent with experiments.