Chemical trend of exchange couplings in diluted magnetic II-VI semiconductors

TL;DR

This study investigates how chemical composition affects magnetic exchange interactions in Zn-based II-VI semiconductors doped with Co or Mn, highlighting the importance of electron correlation treatment for accurate predictions.

Contribution

It demonstrates that LSDA+U method improves agreement with experimental exchange parameters and reveals a crossover in hole coupling strength in ZnO:Mn.

Findings

LSDA+U yields better $J_{dd}$ agreement with experiments.

$N \alpha$ matches experimental data well.

Hole localization in ZnO:Mn explains experimental discrepancies.

Abstract

We have calculated the chemical trend of magnetic exchange parameters ($J_{dd}$, $N \alpha$, and $N \beta$) of Zn-based II-VI semiconductors ZnA (A=O, S, Se, and Te) doped with Co or Mn. We show that a proper treatment of electron correlations by the LSDA+$U$ method leads to good agreement between experimental and theoretical values of the nearest-neighbor exchange coupling $J_{dd}$ between localized 3$d$ spins in contrast to the LSDA method. The exchange couplings between localized spins and doped electrons in the conduction band $N \alpha$ are in good agreement with experiment as well. But the values for $N \beta$ (coupling to doped holes in the valence band) indicate a cross-over from weak coupling (for A=Te and Se) to strong coupling (for A=O) and a localized hole state in ZnO:Mn. That hole localization explains the apparent discrepancy between photoemission and magneto-optical data for ZnO:Mn.