Perovskite Sr-doped LaCrO3 : an ab initio study

In this dissertation, the electronic and crystallographic structures of the perovskite Sr-doped LaCrO3 (La(1-x)Sr(x)CrO3) are theoretically investigated. It is a new p-type transparent conductive oxide (TCO), the Sr atoms replacing La ones to create holes as charge carriers. This material could lead to new routes for the design of future efficient p-type TCOs. Indeed, experimental measures show that the carrier transport may be due to small-polaron hopping in La(1-x)Sr(x)CrO3. This phenomenon is usually not wanted for practical applications because of the low effective mass of the carriers, but in this case the conductivity of the material keeps a high value thanks to the high doping levels that are achieved. The purpose of the ab initio computations was thus to find these polarons in the studied materials, and to characterize their electronic properties. Computations were realized using GGA+U with the ABINIT package. It appeared that polarons are indeed present in this material in the case of x = 0.25 and progressively disappear when x increases. There is a transition smearing temperature T∗, a priori different for each oxide, above which the material goes from the polaronic state to a metallic one. It has been possible to evaluate the difference of energy between the two states for x = 0.25 to about 63 meV (the polaronic state being the most stable of the two). This is nevertheless a rough approximation, and a more precise value could be obtained with a deeper analysis. The study led to the following hypothesis. Under a given fraction x∗ of Sr (close to 0.65 according to experiments), the polaronic state would be more stable than the metallic one. Above x∗ , the opposite would be true. In other words, a high concentration of Cr(4+) atoms would favour the metallic state (SrCrO3 having its Cr atoms in the Cr(4+) state, while LaCrO3 has its Cr atoms in the Cr(3+) state). This has been observed for x = 1. The difference of energy between the two structures in the case of SrCrO 3 has been approximated to 43 meV, the metallic one being the most stable state. In order to test this assumption (the existence of a x∗), many computations should be run for different values of x, in order to be able to compute the difference of energy between the insulating and the metallic states in each case.