Static Magnetic Proximity Effect in $\mathrm{Pt}/{\mathrm{NiFe}}_2{\mathrm{O}}_4$ and $\mathrm{Pt}/\mathrm{Fe}$ Bilayers Investigated by X-Ray Resonant Magnetic Reflectivity

The spin polarization of Pt in $\mathrm{Pt}/{\mathrm{NiFe}}_2{\mathrm{O}}_4$ and $\mathrm{Pt}/\mathrm{Fe}$ bilayers is studied by interface-sensitive x-ray resonant magnetic reflectivity to investigate static magnetic proximity effects. The asymmetry ratio of the reflectivity is measured at the Pt $L_3$ absorption edge using circular polarized x-rays for opposite directions of the magnetization at room temperature. The results of the 2% asymmetry ratio for $\mathrm{Pt}/\mathrm{Fe}$ bilayers are independent of the Pt thickness between $1.8$ and 20 nm. By comparison with ab initio calculations, the maximum magnetic moment per spin polarized Pt atom at the interface is determined to be $(0.6\pm 0.1){\mu }_B$ for $\mathrm{Pt}/\mathrm{Fe}$. For $\mathrm{Pt}/{\mathrm{NiFe}}_2{\mathrm{O}}_4$ the asymmetry ratio drops below the sensitivity limit of $0.02{\mu }_B$ per Pt atom. Therefore, we conclude, that the longitudinal spin Seebeck effect recently observed in $\mathrm{Pt}/{\mathrm{NiFe}}_2{\mathrm{O}}_4$ is not influenced by a proximity induced anomalous Nernst effect.

Figure 1

(a) XRR intensity $I$ and simulation for $\mathrm{Pt}/\mathrm{Fe}/\mathrm{MAO}$. (b) MOKE data ($\lambda =488\mathrm{nm}$) for $\mathrm{Pt}/\mathrm{Fe}/\mathrm{MAO}$. (c) XRR intensity $I$ and simulation for $\mathrm{Pt}/\mathrm{NFO}/\mathrm{MAO}$. (d) Vibrating sample magnetometry (VSM) and LSSE measurements ($\mathrm{\Delta }T=18\mathrm{K}$) for $\mathrm{Pt}/\mathrm{NFO}/\mathrm{MAO}$.

Figure 2

Fluorescence spectrum at the Pt $L_3$ edge of the $\mathrm{Pt}(3.4\mathrm{nm})/\mathrm{Fe}(9.1\mathrm{nm})$ sample and simulation, both normalized to the edge jump. $\mathrm{\Delta }\delta$ and $\mathrm{\Delta }\beta$ are derived from the simulation of the absorption resulting from a magnetic moment of $0.2{\mu }_B$ per Pt atom. XRMR was performed at the absorption maximum of 11 567.5 eV (dashed vertical line) yielding a ratio $\mathrm{\Delta }\beta /\mathrm{\Delta }\delta$ of 3.4 (black crosses).

Figure 3

(a) XRMR intensity $I_{\pm }$ for a positive and negative external field and left circularly polarized light (photon energy of 11 567.5 eV) for $\mathrm{Pt}(3.4\mathrm{nm})/\mathrm{Fe}(9.1\mathrm{nm})$. (b) Corresponding asymmetry ratio $\mathrm{\Delta }I(q)$ and simulated data. (c) Magnetooptic profile used for the asymmetry simulation.

Figure 4

XRR intensity $I$ and corresponding asymmetry ratios $\mathrm{\Delta }I$ for $\mathrm{Pt}(x)/\mathrm{Fe}(9.8\mathrm{nm})$ with (a),(b) $x=1.8\mathrm{nm}$ (average of eight curves), (c),(d) $x=5.9\mathrm{nm}$, and (e),(f) $x=20\mathrm{nm}$. For the XRR simulations the model in (g) is used and for the asymmetry simulations the magnetooptic profiles in (h) are used.

Figure 5

XRMR intensity $I_{\pm }$ and corresponding asymmetry ratios $\mathrm{\Delta }I(q)$ for $\mathrm{Pt}(3.2\mathrm{nm})/\mathrm{NFO}(\sim 900\mathrm{nm})$ [cf. XRR curve in Fig. 1]. (a) Magnetooptic profile used for the simulation in (b). (b) Asymmetry ratio $\mathrm{\Delta }I(q)$ (average of eight curves). (c) XRMR intensity $I_{\pm }$ (average of 52 curves) for the reflectivity from $q=0.2{\AA }^{-1}$ up to $q=0.6{\AA }^{-1}$. (d) Asymmetry ratio $\mathrm{\Delta }I(q)$ from the XRMR in (c) and simulation assuming 2% of the $\mathrm{Pt}/\mathrm{Fe}$ spin polarization.