Protocol Dependence of Spatially Inhomogeneous Magnetic Systems

Abstract

The work presented in this dissertation studies the effects of measurement protocol on magnetic systems. There are two broad classes of metallic systems which will be described -- ferromagnets and spin glasses. The research presented demonstrates a clear protocol dependence for both for nonequilibrium dynamical magnetic materials such as spin glass and metallic ferromagnetic systems. Thus, for experimental reproducibility, it is necessary for one to specify the protocol used to prepare and take measurements. In a single crystal of CuMn 7.92 at.%, the out-of-equilibrium dynamics of aging, rejuvenation, and memory are explored. By using a double-waiting time protocol and quenching to the measuring temperatures, the underlying dynamics of the memory effect are able to be observed with no finite cooling rate effects. After quantifying the memory loss seen in glassy systems, previously proposed explanations were experimentally tested and a quantitative model developed. We find that coincident growth of spin glass correlations reduces the amount of memory retained, and that there is an additional length scale present whose ratio with size of the original correlated regions controls the severity of the memory loss. The effects of a finite cooling rate in spin glasses are then investigated as the temperature is swept continuously. We quantitatively find competing effects of aging and rejuvenation. This implies that the growth rates of glassy order between protocols utilizing quenches cannot naively be compared to protocols which use a finite rate of cooling. Additionally, we determine that the slower growth observed in the finite cooling rate protocols is due to rejuvenation, rather than cumulative aging. In four metallic ferromagnets, four paths to a net-zero magnetzation state are explored. This was done by demagnetizing samples using four different methods and then conducting the same measurements afterwards. The results indicated that the path to zero magnetization changes the behavior of the system, and thus the preparation of the initial state affected subsequent measurements.