Sputter-engineering a first-order magnetic phase transition in sub-15-nm-thick single-crystal FeRh films

Equiatomic FeRh alloys undergo a fascinating first-order metamagnetic phase transition (FOMPT) just above room temperature, which has attracted reinvigorated interest for applications in spintronics. Until now, all attempts to grow nanothin FeRh alloy films have consistently shown that FeRh layers tend to grow in the Volmer-Weber growth mode. Here we show that sputter-grown sub-15-nm-thick FeRh alloy films deposited at low sputter-gas pressure, typically $\sim 0.1$ Pa, onto (001)-oriented MgO substrates, grow in a peening-induced Frank-van der Merwe growth mode for FeRh film thicknesses above 5 nm, circumventing this major drawback. The bombardment of high-energy sputtered atoms, the atom-peening effect, induces a rebalancing between adsorbate-surface and adsorbate-adsorbate interactions, leading to the formation of a smooth continuous nanothin FeRh film. Chemical order in the films increases with the FeRh thickness, $t_{\text{FeRh}}$, and varies monotonically from 0.75 up to 0.9. Specular x-ray diffraction scans around Bragg peaks show Pendellösung fringes for films with $t_{\text{FeRh}}\ge 5.2$ nm, which reflects in smooth well-ordered densified single-crystal FeRh layers. The nanothin film's roughness varies from 0.6 down to about 0.1 nm as $t_{\text{FeRh}}$ increases, and scales linearly with the integral breadth of the rocking curve, proving its microstructured origin. Magnetometry shows that the FOMPT in the nanothin films is qualitatively similar to that of the bulk alloy, except for the thinnest film of 3.7 nm.

Figure 1

X-ray reflectometry (XRR) scans for nanothin FeRh alloy films with FeRh thickness $t_{\text{FeRh}}=5.2$ nm and 11.6 nm. $t_{\text{FeRh}}$ was determined from the Kiessig fringe spacings [28]. The blue line corresponds to the XRR scan for the nanothin film with $t_{\text{FeRh}}=3.7$ nm. Because of the absence of Kiessig fringes, its thickness was subsequently determined by using scanning transmission electron microscopy. The inset shows a linear fit of the Kiessig fringe spacing analysis for the films with $t_{\text{FeRh}}=5.2$ nm (full dots) and 11.6 nm (empty dots).

Figure 2

HAADF-STEM images of a cross-sectional TEM lamella showing the FeRh thin film (middle of each image) grown on the MgO substrate (left), with an average thickness of 3.7 nm, and protective Pt layer (right). (a) High resolution HAADF-STEM image showing the localized structure of the FeRh and its epitaxial interface with the MgO substrate, as well as an average thickness profile (inset). HAADF-STEM images showing (b) a zone with uniform thickness and (c), (d) others with nonuniform FeRh thin film, the discontinuous growth features denoted with red arrows.

Figure 3

Specular x-ray diffraction (XRD) $\theta -2\theta$ scans for (001)-oriented FeRh alloy films grown onto (001)-oriented MgO single crystal substrates for FeRh thickness (a) $t_{\text{FeRh}}=14.1$ nm, where the inset shows a close-up of the XRD scan around the (001) Bragg peak as a function of the scattering wave-vector $\mathit{Q}$. Well-developed Pendellösung fringes up to fifth order are labeled. (b) As above for $t_{\text{FeRh}}=3.7$, 5.2, and 9.1 nm, where the inset displays a close-up of the XRD scan around the (001) Bragg peak.

Figure 4

Rocking curve analysis. (a) Dependence of the full-width-half-maximum (FWHM), measured over the rocking curve scan collected on the (002) FeRh Bragg peak, with the FeRh thickness, $t_{\text{FeRh}}$. The line is a guide to the eye. The inset shows the rocking curve scan collected on the film with $t_{\text{FeRh}}=3.7$ nm. The line corresponds to a Gaussian fit with $\text{FWHM}=0.7^{\circ }$ (b) Coherent FeRh thickness, $t_{\text{FeRh}}^{\text{coh}}$, determined as $2\pi /\mathrm{\Delta }Q$ [30, 31], where $\mathrm{\Delta }Q$ is the spacing between adjacent Pendellösung fringes, as a function of the experimental $t_{\text{FeRh}}$ values. The dashed line corresponds to a plot where $t_{\text{FeRh}}^{\text{coh}}$ = $t_{\text{FeRh}}$ and the continuous one is a guide to the eye.

Figure 5

HAADF-STEM analysis. Left panel: HAADF-STEM image showing the MgO substrate (left-hand portion of the image), FeRh layer (middle), and protection Pt layer (right) for the film with $t_{\text{FeRh}}=3.7$ nm. Right panel: Chemical composition mapping for oxygen, rhodium, and iron, obtained using EELS (pixel lateral size is 5 Å) over the area enclosed by the green box (long-side length is $\sim 15$ nm) in the left panel. The 1 nm-thick Fe-enriched layer is denoted by yellow dashed lines.

Figure 6

Atomic force micrographs taken on nanothin FeRh alloy films with FeRh thickness, $t_{\text{FeRh}}=14.1$ nm, (a) and (d), $t_{\text{FeRh}}=9.2$ nm, (b) and (e), and $t_{\text{FeRh}}=3.7$ nm, (c) and (f). AFM scan size is $50\times 50\mu {\mathrm{m}}^2$ (top row) and $0.5\times 0.5\mu {\mathrm{m}}^2$ (bottom row); micrographs edges are aligned along the [110]$\mathrm{Mg}\mathrm{O}\parallel$[100]FeRh and [$\overline{1}$10]$\mathrm{Mg}\mathrm{O}\parallel$[010]FeRh directions. Height scales have been set to optimize image contrast.

Figure 7

Structural analysis. (a) Average root-mean-square roughness, $R_{\text{rms}}$, versus full-width-half-maximum (FWHM) of the rocking curve. The line corresponds to a linear fit, where the intercept is zero and the slope is to $0.845\pm 0.056$ nm/deg. Standard deviation values for the $R_{\text{rms}}$ are smaller than or similar to the dots size. The inset shows the variation of $R_{\text{rms}}$ with $t_{\text{FeRh}}$. (b) Thickness dependence of the nanothin FeRh alloy films order parameter, $S$, determined according to Ref. [35].

Figure 8

Magnetic properties. (a) Magnetization as a function of temperature for FeRh alloy films with thickness, $t_{\text{FeRh}}=9.2$ (black), 8.1 (blue), 5.2 (red) and 3.7 nm (green) for an applied magnetic field, ${\mu }_0{H}_{\text{app}}=0.1$ T so that H$\parallel$[110]. (b) $M-H$ loops for H$\parallel$[110] (blue squares) and H$\parallel$[001] (black dots) for the thinnest film in the FM phase.