Optimization of p-type conductivity in lithium niobate by co-doping strategy

This is the authors' abstract. We don't add key points for this paper.

The development of lithium niobate (LiNbO3, LN)-based active devices has been limited by the absence of stable p-type conductivity. Overcoming this challenge is critical for unlocking the full potential of LN in optoelectronic applications. Through first-principles calculations, we have identified nitrogen (N) as the most effective p-type dopant among elements near oxygen (O) in the periodic table. However, mono-acceptor N doping creates defect levels 0.415 eV above the valence band maximum (VBM). We propose to passivate the N dopant by introducing a small amount of impurity magnesium (Mg). This doping process involves two steps: first, N–Mg co-doping with equal amounts creates fully occupied impurity bands; second, excess N doping the fully occupied impurity bands introduces shallow defect levels. Quantitative results show that in 2N+Mg co-doped LN the defect level is only 0.115 eV (about 0.3 eV lower than that of mono-acceptor N-doping) above the VBM. Correspondingly, the ɛ(0/−1) transition level is found at 0.067 eV (about 0.243 eV lower than that of mono-acceptor N-doping), indicating a typical shallow acceptor. Co-doping also lowers the defect formation energy by ∼1.9 eV, boosting dopant solubility. These key data demonstrate that the donor and excess acceptor co-doping strategy effectively converts the defect levels into shallow levels, reduces defect formation energy, and enhances local doping stability in p-type LN, thereby laying the foundation for the development of LN-based p–n junction and active optoelectronic devices.