Complementary polarized neutron and resonant x-ray magnetic reflectometry measurements in Fe/Gd heterostructures: Case of inhomogeneous intralayer magnetic structure

A unified approach combining polarized neutron and resonant x-ray magnetic reflectometry has been applied to determine the magnetic structure in an ${\left[\text{Fe}\left(35\text{Å}\right)/\text{Gd}\left(50\text{Å}\right)\right]}_5$ multilayer as a function of temperature and magnetic field. Simultaneous self-consistent refinement of neutron and x-ray data made it possible to resolve the element-specific magnetization profile in the multilayer with unprecedented accuracy. It is shown that the small number of bilayer periods together with the asymmetric termination (Fe-top, Gd-bottom) lead to unique low-temperature magnetic phases characterized by significant twisting of Fe and Gd magnetic moments and nonuniform distribution of vectorial magnetization within Gd layers. A twisted magnetic state was found to be stable at small magnetic fields and at a low temperature of 20 K, which is well below the compensation temperature of this artificial ferrimagnetic system.

Figure 1

(Color online) Experimental (circles) and corresponding theoretical (line) x-ray reflectivity curves taken from the ${\left[\text{Fe}\left(35\text{Å}\right)/\text{Gd}\left(50\text{Å}\right)\right]}_5$ multilayer (a) at off-resonance conditions $\left(E=6500\text{eV}\right)$ and (b) at energy $E=7929\text{eV}$ ($\text{Gd}{L}_2$ resonance). Well-defined Bragg peaks up to the fifth order suggest highly correlated layer structure throughout the multilayer.

Figure 2

Temperature dependence of the magnetic moment per unit area for ${\left[\text{Fe}\left(35\text{Å}\right)/\text{Gd}\left(50\text{Å}\right)\right]}_5$ multilayer measured under applied magnetic fields (Ref. 30) of 10 and 50 mT. The minimum of $M\left(T\right)$ dependence is at about 90 K for $H=50\text{mT}$ and at about 65 K for $H=50\text{mT}$. The compensation temperature $T_{\text{comp}}$ for zero magnetic field is then estimated to be about 60 K.

Figure 3

Hysteresis loops for the ${\left[\text{Fe}\left(35\text{Å}\right)/\text{Gd}\left(50\text{Å}\right)\right]}_5$ multilayer measured in the Fe-aligned phase (140 K), at the compensation temperature (60 K), and at low temperature far from the compensation (20 K).

Figure 4

(Color online) $\text{Gd}{L}_2$ edge XMCD measured at temperature $T=20\text{K}$ for different values of the applied magnetic field and at temperature 140 K and magnetic field $H=-50\text{mT}$. The integrated XMCD signal was used as a measure of the relative net Gd moment in the multilayer.

Figure 5

(Color online) (a) Experimentally derived (circles) and tabulated (lines) (Ref. 39) charge resonant scattering factors across the $\text{Gd}{L}_2$ absorption edge. (b) Experimentally derived Gd magnetic scattering factors across the $\text{Gd}{L}_2$ absorption edge.

Figure 6

(Color online) (a) Resonant x-ray magnetic reflectivity spectra (asymmetry ratio) measured at the $\text{Gd}{L}_2$ edge $\left(E=7929\text{eV}\right)$ at $T=140\text{K}$ in magnetic fields of 50 and $-50\text{mT}$. The signals measured at opposite fields are symmetrical, which indicates their magnetic origin. (b) Experimental (circles) and fitted (line) RXMR spectra for $T=140\text{K}$, $H=50\text{mT}$. (c) Experimental (points) and fitted (lines) PNR spectra for $T=140\text{K}$, $H=50\text{mT}$. Since there is negligible signal in the spin-flip neutron channel, all the magnetic moments in the system are aligned along or opposite to the applied magnetic field.

Figure 7

(Color online) Refined magnetization profile within the Fe/Gd heterostructure (a) in the high-temperature Fe-aligned state ($T=140\text{K}$, $H=50\text{mT}$); in twisted low-temperature $\left(T=20\text{K}\right)$ states at (b) $H=50\text{mT}$ and (c) $H=500\text{mT}$. The Fe-aligned state is described in terms of the depth-dependent value of magnetic moment. The twisted states are described in terms of the depth-dependent angle between the direction of magnetic moment and applied magnetic field.

Figure 8

(Color online) (a) Resonant x-ray magnetic reflectivity spectra (asymmetry ratio) measured at the $\text{Gd}{L}_2$ edge $\left(E=7929\text{eV}\right)$ at $T=20\text{K}$ in magnetic fields 50 and $-50\text{mT}$. The signals measured at opposite fields are symmetrical, which indicates their magnetic origin. (b) Experimental (circles) and fitted (line) RXMR spectra for $T=20\text{K}$, $H=50\text{mT}$. (c) Experimental (points) and fitted (lines) non-spin-flip PNR spectra for $T=20\text{K}$, $H=50\text{mT}$.

Figure 9

(Color online) (a) Resonant x-ray magnetic reflectivity spectra (asymmetry ratio) measured at the $\text{Gd}{L}_2$ edge $\left(E=7929\text{eV}\right)$ at $T=20\text{K}$ in magnetic fields 500 and $-500\text{mT}$. The signals measured at opposite fields are symmetrical, which indicates their magnetic origin. (b) Experimental (circles) and fitted (line) RXMR spectra for $T=20\text{K}$, $H=500\text{mT}$. (c) Experimental (points) and fitted (lines) non-spin-flip PNR spectra for $T=20\text{K}$, $H=500\text{mT}$.