Bulk NdNiO2 is thermodynamically unstable with respect to decomposition while hydrogenation reduces the instability and transforms it from metal to insulator
Through CaH2 chemical reduction of a parent R3+Ni3+O3 perovskite form, superconductivity was recently achieved in Sr-doped NdNiO2 on SrTiO3 substrate. Using density functional theory (DFT) calculations, we find that stoichiometric NdNiO2 is significantly unstable with respect to decomposition into 1/2[Nd2O3 + NiO + Ni] with exothermic decomposition energy of +176 meV/atom, a considerably higher instability than that for common ternary oxides. This poses the question if the stoichiometric NdNiO2 nickelate compound used extensively to model the electronic band structure of Ni-based oxide analog to cuprates and found to be metallic is the right model for this purpose. To examine this, we study via DFT the role of the common H impurity expected to be present in the process of chemical reduction needed to obtain NdNiO2. We find that H can be incorporated exothermically, i.e., spontaneously in NdNiO2, even from H2 gas. In the concentrated limit, such impurities can result in the formation of a hydride compound NdNiO2H, which has significantly reduced instability relative to hydrogen-free NdNiO2. Interestingly, the hydrogenated form has a similar lattice constant as the pure form (leading to comparable XRD patterns), but unlike the metallic character of NdNiO2, the hydrogenated form is predicted to be a wide gap insulator thus, requiring doping to create a metallic or superconducting state, just like cuprates, but unlike unhydrogenated nickelates. While it is possible that hydrogen would be eventually desorbed, the calculation suggests that pristine NdNiO2 is hydrogen-stabilized. One must exercise caution with theories predicting new physics in pristine stoichiometric NdNiO2 as it might be an unrealizable compound. Experimental examination of the composition of real NdNiO2 superconductors and the effect of hydrogen on the superconductivity is called for.