Search papers, labs, and topics across Lattice.
X-ray absorption spectroscopy (XANES) coupled with DFT+U calculations revealed that oxygen vacancies in NdBaCo$_2$O$_{5+δ}$ ($δ\sim 0.65$) are not confined to apical sites but also occur at basal sites, leading to magnetic inhomogeneity and variable range hopping conduction. The spectral features at the Co K, Ba, and Nd $L_{\mathrm{III}}$ edges suggest a $δ$ value closer to 0.75, with Ba confirmed to be in a +2 oxidation state. This counterintuitive vacancy distribution explains observed phenomena and offers a pathway to tailor oxygen-ion mobility for enhanced oxygen sensing.
Oxygen vacancies in layered cobaltates preferentially occupy basal sites, not just apical ones, fundamentally altering charge transport and magnetic properties.
The degree of oxygen vacancy in A-site-doped layered double perovskite oxides can lead to anion vacancy ordering, which subsequently results in distinct superstructures. We have studied oxygen deficient double perovskite oxide, NdBaCo$2$O$_{5+δ}$ ($δ\sim$ 0.65), utilising X-ray absorption near edge structure (XANES) spectroscopy complemented by multiple scattering and first-principles calculations based on density functional theory (DFT+U). The present results establish a direct correlation between oxygen-vacancy distribution, magnetic inhomogeneity, and charge transport in layered cobaltates. Our investigations on the Co K, Ba, and Nd $L_{\mathrm{III}}$ edges suggest that the near-edge spectra, based on the number and shape of the spectral features, indicate the value of $δ$ for the compound under study lies closer to 0.75. The analysis of the Ba $L_{\mathrm{III}}$ edge spectrum confirms the +2 oxidation state for Ba ions. The Co K-edge near-edge spectra suggest the possibility of oxygen vacancies arising at the basal sites in addition to the apical sites. This is counterintuitive to the common understanding that oxygen vacancies usually occur at the apical sites, i.e., in the lanthanide (Nd) layer in these compounds. Such a distribution of oxygen vacancies naturally explains the Griffith phase and the conduction mechanism of variable range hopping observed in the compound. The vacancy distribution, when at the optimum level, can also bring about easier oxygen ion mobility and a more sensitive and faster oxygen sensor. We believe that our results will help in tailoring oxygen-ion mobility and enhancing oxygen-sensing performance.