Abstract

A new material design approach using first-principles density functional theory (DFT) calculations with acceptable precision can elucidate many critical features in emerging high-performance lithium-ion battery fields. However, the dominant impact of transition metals (TM) as the major component and/or modification dopant element with localized d-electrons in this material system, which needs so-called Hubbard correction U, limits DFT to many extents. As the U correction is not completely transferable, one may need to recalculate it for the same input structure with structural or elemental changes. While the accurate calculation of U is costly, it is worth investigating its sensitivity to input system parameters, such as cell size, structure, and chemical composition, to dismiss unessential recalculations, especially in high-throughput schemes. Furthermore, implementing DFT + U to get the expected properties is computationally expensive too. The necessity of including the U correction for expected properties or its small variation should be investigated. In this study, Hubbard correction for TM using a density functional perturbation theory (DFPT) approach was considered to investigate the electronic structure and structural stability of LiNiO2, LiCoO2, and LiNi0.75Co0.167Al0.083O2. The effect of system parameters (cell size, structure, and chemical composition) on U values was considered for different systems. The dependency of calculated electronic densities of states, lattice parameters, preferred doping sites, and vacancy formation energy, as well as charge voltage, was investigated, and low and high sensitivities were discussed. Furthermore, the effect of different doping scenarios with Co and Al on structural stability was studied with the efficient DFT + U approach. The results showed that the structural variation due to doping and vacancy site formation has a negligible effect on calculated U values. Significant correlation of predicted electronic properties with U for all systems was observed, while calculated lattice parameters, as well as preferred doping sites, reflected very low dependency on U variations.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call