Reducing background-induced bias in nickel oxide Ni 2 p XPS quantification with active-background Narrow-Shirley modeling

Nickel oxide (NiO) is a technologically relevant transition-metal oxide whose complex Ni 2p core-level spectrum provides a stringent test case for X-ray photoelectron spectroscopy (XPS) background modeling. The superposition of spin–orbit splitting, intense charge-transfer satellites, structured inelastic losses, and localized intrinsic background contributions makes fitted peak areas and derived stoichiometry highly sensitive to background selection. Here, we systematically evaluate five background models for Ni 2p quantification: standalone iterative Shirley, iterative Shirley combined with either a NiO-specific or 2-parameter Tougaard background, Shirley–Vegh–Salvi–Castle (SVSC) combined with a NiO-specific Tougaard background, and Narrow-Shirley (NS) combined with an NiO-specific Tougaard background (NS + Tougaard). Iterative Shirley and SVSC-based backgrounds reproduce the spectral envelope but require broad auxiliary components whose areas depend strongly on the adopted fitting constraints, leading to apparent stoichiometries as low as (NiO0.57-NiO0.74). In contrast, the NS + Tougaard framework constrains the extrinsic inelastic-loss contribution using a material-specific cross section while describing the localized intrinsic background intensity over a finite energy range, thereby reducing dependence on auxiliary components and analysis range. Combined with a four-component O 1s model and angle-resolved XPS Multilayer modeling, this approach yields NiO0.90±0.01. Validation against independent NiO spectra from the XPSSurfA and XPS Library further supports the portability of the fitted Ni 2p spectral model across different samples and instruments, establishing the NS + Tougaard framework as a physically grounded and reproducible strategy for reducing background-induced bias in the quantitative XPS analysis of complex transition-metal oxides.