Fe dopant in ZnO: 2+ versus 3+ valency and ion-carrier s , p − d exchange interaction

TLDR: In this article, the authors analyzed the 3+ valency of transition metal ions in polycrystalline (Zn,Fe)O films with complementary theoretical and experimental studies.

Abstract: Dopants of transition metal ions in II-VI semiconductors exhibit native 2+ valency. Despite this, 3+ or mixed 3+/2+ valency of iron ions in ZnO was reported previously. Several contradictory mechanisms have been put forward for explanation of this fact so far. Here we analyze Fe valency in ZnO by complementary theoretical and experimental studies. Our calculations within the generalized gradient approximation (GGA+$U)$ indicate that the Fe ion is a relatively shallow donor. Its stable charge state is ${\mathrm{Fe}}^{2+}$ in ideal ZnO, however, the high energy of the (+/0) transition level enhances the compensation of ${\mathrm{Fe}}^{2+}$ to ${\mathrm{Fe}}^{3+}$ by nonintentional acceptors in real samples. Using several experimental methods like electron paramagnetic resonance, magnetometry, conductivity, excitonic magnetic circular dichroism, and magnetophotoluminescence we confirm the 3+ valency of the iron ions in polycrystalline (Zn,Fe)O films with the Fe content attaining 0.2%. We find a predicted increase of $n$-type conductivity upon the Fe doping with the Fe donor ionization energy of $0.25\ifmmode\pm\else\textpm\fi{}0.02$ eV consistent with the results of theoretical considerations. Moreover, our magneto-optical measurements confirm the calculated nonvanishing s,p-d exchange interaction between band carriers and localized magnetic moments of the ${\mathrm{Fe}}^{3+}$ ions in the ZnO, being so far an unsettled issue.