Influence of the distribution of the inherent ordering temperature on the ordering in layered magnets

We study the influence of gradients in the inherent ordering temperature of coupled layered magnets on the overall magnetic ordering. The gradients were accomplished by growing Fe(001) layers with thicknesses ranging from two to three monolayers, all separated by seven monolayers of V(001). Two types of gradient superstructures were grown: one with the highest and one with the lowest inherent ordering temperature in the center of the samples. The superstructure with the thinnest outermost Fe layers exhibits lower ordering temperature, demonstrating the importance of the sequence of the layers. Both these structures order at temperatures significantly lower than a superlattice with a constant thickness of the Fe layers (three monolayers). The results highlight the intricate collective aspects of the magnetic ordering in layered magnets, which are not captured by current models in magnetism research.

Figure 1

The thickness profiles of the magnetic layers. The inherent ordering temperature of each Fe layer is proportional to its thickness. (a) illustrates a regular superlattice structure with constant thickness of all the layers while (b) resembles the building block used for obtaining a distribution in $T_c$. (c) illustrates the structure with the lowest inherent $T_c$ in the center of the sample while (d) has the highest inherent ordering temperature in the center.

Figure 2

(Color online) X-ray reflectivities obtained from the samples. The topmost curve is a representative fit for one of the samples ($\text{GR}-$, red). The superlattices peaks are well displayed at $Q_1$ and $Q_2$. The intensities are shifted for clarity.

Figure 3

X-ray diffraction spectra of the samples. The main peak $Q_0$ is related to the average lattice constant. The peaks $Q_{-1}$ and $Q_{+1}$ are the markers of the superlattice structure.

Figure 4

Magnetization as a function of temperature for all the samples. The NG measurements are incomplete because of the limited temperature range of the SQUID.

Figure 5

(Color online) Neutron data and fits for the sample NG. The dots represent the raw data whereas the lines represent the fits.

Figure 6

(Color online) Reflectivity measurements at 12 K for the series of samples. The measurements are shifted in intensity for clarity.

Figure 7

(Color online) Reflectivity measurements at 12 K for the samples $\text{GR}+$ and $\text{GR}-$. The measurements are shifted in intensity for clarity.

Figure 8

Hysteresis loops for the three samples recorded at $\approx 0.45t$ (see text). The data sets are normalized.

Figure 9

MOKE results of the temperature dependence of the remanent magnetization. The Kerr rotation is normalized at 80 K for easing the comparison of the results.

Figure 10

Real part of the complex susceptibility as a function of $t$. For details, see text.