Reproducible and well defined small hysteresis loops of rhomboid shape were measured on thin uniaxial magnetic garnet films. Observation of the domain structure showed that the change in the magnetic moment described by these loops was due solely to elementary translations of domain walls. The authors have called the half-width of these loops the domain wall coercive field, Hdw; this parameter differed substantially from the half-width, Hc of the major saturation-to-saturation hysteresis loops. They called the rhomboid small loops the domain wall coercive loops (DWCL) and showed that Hdw obtained from DWCL was identical to the coercive field value measured by the low frequency domain wall oscillation method. The slope of DWCL coincided with the initial susceptibility of stripe domain structures and thus made it possible to calculate the basic parameters of the sample, giving results in agreement with other established methods of measurement of these parameters.

The coercive properties of magnetically uniaxial liquid-phase epitaxy garnet films were investigated between 10 K and the NCel temperature ( TN<500 K). Two independent methods, the results of which are nearly identical (magnetical response of oscillating domain walls and the method of coercive loops measured in a vibrating sample magnetometer), were used. Besides the usual domain-wall coercive field, Hdw, the critical coercive pressure, pd,,.t was also introduced as it describes in a direct way the interactions of the domain walls with the . . wall-pmnmg traps. Both Hdw and pdw were found to increase exponentially with decreasing temperature. Three different types of wall-pinning traps were identified in the sample and their strength, their rate of change with temperature, and their temperature range of activity were determined. I. INTRODUCTIONEpitaxial magnetic garnet films belong to the most perfect single-crystalline materials, as their growth by liquidphase epitaxy (LPE) technology and the control of their magnetic properties are very well-established processes. Their magnetic domain structure is usually a very simple one: stripes or bubbles. Due to the large growth-induced uniaxial anisotropy, KU, which keeps both the magnetization in the domains and the 180" domain walls (DW) normal to the sample surface, there are no closure domains in the sample. This simple DW configuration makes it easy to investigate interactions of the moving DWs with the material, i.e., to investigate the domain-wall coercive field, Hdw, the nonzero value of which originates from local variations of the DW energy as the walls sweep through the defects in the sample.The domain-wall coercive field Hdw is defined as the minimum magnetic field normal to the sample surface which initiates an irreversible DW motion. A large number of theories (for a review see, e.g., Ref. 1) have attempted to calculate the coercive field from the material parameters and from an assumed distribution of the material defects. Some of the models met the experimental observations to a fair extent. The temperature dependence of Hdw, however, has been studied only to a very limited extent. A considerable increase of the coercive field with decreasing temperature has usually been reported.2-8 The aim of the present paper is to measure the domain-wall coercive field, H dw, from the Ntel temperature, TN, down to about T = 10 K and to suggest a mathematical expression describing the temperature dependence of Hdw within the whole temperature range of the existence of the ferrimagnetic order in our samples.

The saturation magnetization, cubic anisotropy constants K1, K2, FMR linewidth, and g‐factor of thin epitaxial films of Y3Fe5−x−y−zCoxTiyGazO12 with x = 0 to 0.11, z = 0; 0.5 are measured and compared with previous results on the system Y3Fe5−x−y−zCoxGeyO12. The measured data are used to discuss the distribution of Co ions, spin canting, nonlinear dependence K1(x), influence of the constant K3, and the presence of a dynamic shift of the resonance field.

No abstract

No abstract