Controllable magnetic anisotropy and spin orientation of a prototypical easy-plane antiferromagnet on a ferromagnetic support

We report the magnetic properties of easy-plane antiferromagnetic NiO(111) thin films epitaxially grown on ferromagnetic Fe(110) layers. We find that the magnetic moments in NiO are rotatable within the NiO(111) sample plane and both magnetic anisotropy and orientation of antiferromagnetic NiO spins are determined by the magnetic properties of underlying ferromagnetic Fe layer. Specifically, the magnetic state and anisotropy of antiferromagnetic NiO can be modified by tuning the thickness of Fe, changing the temperature, or applying small external magnetic field.

Figure 1

LEED patterns of (a) uncovered Fe(110) and (b) NiO(111)/Fe(110) surfaces. Corresponding ball models of (c) Fe(110) and (d) NiO(111)/Fe(110) surfaces. The relative Fe and NiO in-plane directions as concluded from the LEED analysis are sketched in (e).

Figure 2

(a) Fe $L_{2,3}$ XMCD spectra from the NiO(111)/(50 Å Fe) and NiO(111)/(150 Å Fe) regions of the sample. LHP and RHP stand for circular left- and right-handed polarization, respectively. (b) Ni $L_2$ XMLD spectra acquired on two sample regions using x rays with linear polarization vector parallel to the in-plane Fe[$1\overline{1}0$] direction. (c, top panel) Fe $L_3$ XMCD-PEEM image of the SRT in NiO/Fe(110) at the 150-Å Fe (up)/50-Å Fe (down) boundary. The image contrast originates from the relative alignment of the photon propagation direction k and magnetization M, (c, center panel) corresponding Ni $L_2$ XMLD-PEEM image of the same area obtained with x-ray polarization along in-plane Fe[001] direction. In the bottom panel of (c) XMLD image obtained with x-ray polarization along the surface normal is shown. Modulations of the image intensity, appearing as intensity minima and maxima, affect the XMLD images, superimposed to the magnetic signal. These artifacts are caused by the inhomogeneity of the (improperly) focused x-ray-beam spot on the sample, which drifts during image acquisition. Scale bar in (c) corresponds to 5 μm. Intensity profiles across the domain boundary for both XCMD and XMLD images are presented in Supplemental Material [59].

Figure 3

(a) Schematic sketch of the angle-resolved XMLD measurements geometry. (b) $R_{L2}$ ratio dependence on the polar angle θ (fixed $\phi =0$) for 80 K (circles), 300 K (squares), and 350 K (triangles), for both studied Fe thicknesses, 50 Å (red) and 150 Å (black). (c) Temperature dependence of $R_{L2}$ ratio for NiO/(50 Å Fe) and NiO/(150 Å Fe) regions of the sample. (d) Temperature dependence of difference of $R_{L2}$ ratios from (c) defined as $\mathrm{\Delta }{R}_{L2(\theta =0)}=R_{L2}\left(50\AA \right)-R_{L2}\left(150\AA \right)$.

Figure 4

In-plane angular dependencies of (left) Fe magnetization in remanent state as determined from MOKE hysteresis loops and (right) XMLD magnitude defined as $\mathrm{\Delta }{R}_{L2\phi }=\left|R_{L2}\left(\theta =0^{\circ }\right)-R_{L2}\left(\theta ={60}^{\circ }\right)\right|$, as a function of the in-plane angle φ. Arrows at the bottom of the image schematically mark the easy axes of ferromagnetic Fe and antiferromagnetic NiO sublayers for both sample regions. Red and black colors correspond to 50- and 150-Å sample regions.

Figure 5

Element-sensitive XMCD (top) and XMLD (bottom) magnetic hysteresis loops of Fe and NiO sublayers, respectively.

Figure 6

Easy-axis magnetic hysteresis curves for NiO/(50 Å Fe) region as measured by MOKE at 300 and 80 K; external magnetic field was applied along Fe[$1\overline{1}0$] in-plane direction.

Figure 7

(a) Schematic sketch of the sample specially prepared in order to present tuning of SRT critical temperature in NiO/Fe bilayer. (b) Temperature dependence of XMLD ($R_{L2}$) shows SRT in AFM NiO during heating process for various Fe(110) thicknesses. Results were obtained in normal incidence geometry $\theta =0^{\circ }$ and with azimuthal angle $\phi ={90}^{\circ }$ which means E||Fe[001].