The effect of sample inhomogeneity on the ferromagnetic resonance linewidth is determined by diagonalization of a spin wave Hamiltonian for ferromagnetic thin films with inhomogeneities spanning a wide range of characteristic length scales. A model inhomogeneity is used that consists of size D grains and an anisotropy field H(p) that varies randomly from grain to grain in a film with thickness d and magnetization M(s). The resulting linewidth agrees well with the two-magnon model for small inhomogeneity, H(p)D<>piM(s)d, the precession becomes localized and the spectrum approaches that of local precession on independent grains.
Detailed ferromagnetic resonance measurements on thin Ni80Fe20 films are described that determine characteristics of intrinsic damping and the effects of intentionally created defects on linewidth. Measurements are made as a function of frequency with magnetization oriented in-plane and normal to the plane. For nominally uniform films, the ferromagnetic resonance linewidth is linear in frequency for both in-plane and normal magnetization and is well described by Landau–Lifshitz damping with a constant damping parameter. The effect of inhomogeneity was investigated by depositing the Ni80Fe20 films on specially prepared substrates, one with parallel, random grooves and one with polycrystalline, antiferromagnetic NiO. The linewidth in these samples is generally larger with magnetization in-plane than it is with magnetization normal. In the NiO/Ni80Fe20 film, the in-plane linewidth is linear in frequency with a significant zero-frequency linewidth. In the Ni80Fe20 sample on the grooved substrate, the linewidth has a more complicated frequency dependence.
Ferromagnetic resonance linewidth measurements are reported for thin films of Ni80Fe20 deposited on substrates with 320 nm period grooves or parallel aperiodic scratches. These films are model systems for studying the effects of inhomogeneity on ferromagnetic resonance. With the grooved substrate, the ferromagnetic resonance spectra show that interactions split the resonant frequencies of interacting modes leading to two resonances for strong interactions and broadened resonances for weak interactions. The results obtained with the scratched substrate are in agreement with the two-magnon model of linewidth, showing sharply increased linewidth when the scratches are oriented to allow coupling to degenerate spinwave modes. The results are not in agreement with a simple model of linewidth as a direct measure of the spread in local resonance fields.
Cr(001)/CoCrPt (112̄0) films, used in magnetic hard disks, exhibit in-plane magnetic anisotropy when grown over a substrate patterned with shallow grooves. In this experiment, Cr/CoCrPt films are grown over substrates that have been patterned using lithographic techniques to produce well-controlled topography. The in-plane anisotropy, which is of the order of 104 J m−3, increases with the groove height and frequency. This magnetic anisotropy is a result of anisotropic in-plane crystallographic orientation and anisotropic stress. Stress-induced anisotropy accounts for up to 25% of the total in-plane anisotropy, but the major contribution to the anisotropy is a preferential Co c-axis alignment parallel to the groove direction.
Topographically induced in-plane magnetic anisotropy has been observed in CoCrTa and CoCrPt films deposited onto oxidized silicon substrates which are lithographically patterned with fine grooves of period 200–320 nm and amplitude 20–50 nm. The coercivity and remanence are higher parallel to the grooves. Anisotropy has been observed in both rf- and dc-magnetron sputtered films with a (112̄0) preferred orientation, which is achieved by growth at elevated temperature on a (200)-oriented Cr underlayer at low base pressures. Anisotropy increases with the amplitude of the grooves in the silica substrate.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.