Comment on ''Paramagnetic and ferromagnetic resonance studies on dilute magnetic semiconductors based on GaN'' (Phys. Status Solidi A 205, 1872 (2008))

Recently, Kammermeier et al. (KD) reported on electron paramagnetic resonance (EPR) and ferromagnetic resonance (FMR) studies in the diluted magnetic semiconductor GaN:Gd (Kammermeier et al., Phys. Status Solidi A 205, 1872 (2008) 1). They observed several broad resonance signals at temperatures below 50 K in samples grown by molecular beam epitaxy on 6H SiC. The angular dependence of two FMR-like lines was measured and reveal uniaxial behavior with opposite anisotropy. From the temperature dependence of the resonance field and the line width as well as the uniaxial behavior of the line positions the authors suggest a ferromagnetic origin. The discussed resonance lines are supposed to originate from GdN clusters with the same orientation to the crystal environment. Several observed weaker signals which were not discussed in detail were attributed to Gd clusters of different orientations. Further, it is believed that signatures of isolated, paramagnetic Gd impurities are not visible in the resonance experiment, in particular, because the resonance field is presumable covered by the observed cavity and substrate signals. The authors did not perform a rigorous analysis of the observed angular dependence of the EPR line positions and their intensities with the corresponding Hamiltonian for the considered defect. In addition they overlooked some well-known findings on the feature of high-spin systems in strong crystal fields. Most of their conclusions and interpretations turn out to be erroneous. Here, we show by computer simulation of the EPR spectrum that all observed Gd-related lines can be successfully explained by isolated Gd3+ on the two inequivalent Ga sites of the hexagonal GaN lattice considering the fine-structure of the S = 7/2 system and the strain in the epitaxial layer.