Tuning noncollinear magnetic states by hydrogenation

Two different superstructures form when atomic H is incorporated in the Fe monolayer on Ir(111). Depending on the amount of H provided, either a highly ordered $p(2\times 2)$ hexagonal superstructure or an irregular roughly square structure is created. We present here spin-polarized scanning tunneling microscopy measurements which reveal that in both cases the magnetic nanoskyrmion lattice state of the pristine Fe monolayer is modified. Our measurements of the magnetic states in these hydrogenated films are in agreement with superpositions of cycloidal spin spirals which follow the pattern and the symmetry dictated by the H superstructures. We thus demonstrate here the possibility to vary the symmetry of a noncollinear magnetic state in an ultrathin film without changing its substrate.

Figure 1

Overview of a hydrogenated sample. STM constant-current map of a hydrogenated Fe film on Ir(111). Hexagonal superstructures cover completely the monolayer (ML) and the double-layer (DL) areas. In the latter case, both the H1 and the H2 phases are visible. The image is partially differentiated along the scan direction (horizontal) to improve the visibility of the structures. Measurement parameters: 700 mV, 1 nA, 8 K, and 0 T.

Figure 2

The two phases of the hydrogenated Fe ML on Ir(111). (a) STM constant-current map showing the H-induced hexagonal superstructure. The darker area corresponds to the H1 phase of the hydrogenated Fe DL located on the adjacent terrace, which allows to identify the hexagonal structure in the monolayer as a $p(2\times 2)$ superstructure. The light gray defects correspond to H vacancies, whereas we assign the brighter ones to Fe adatoms expelled from the neighboring Fe double-layer areas. (b) Proposed lateral position of the H atoms in the hexagonal superstructure. The blue line marks the unit cell. (c) STM constant-current map showing the three rotational domains of the square superstructure; the domain boundaries are marked with the orange lines. The bright and dark dots are defects and impurities in the structure presumably originating from the sample preparation. (d) STM constant-current map showing a closer view of the square superstructure. The bright regions are aligned along the close-packed rows of the (111) surface and form an irregular roughly square pattern. The blue line marks the structural unit cell. (e) Sketch showing the position of the square superstructure on the atomic lattice. The dark areas in image (d) are marked in dark gray. A row of bright dots corresponds to two rows of atoms. The unit cell of the superstructure is indicated in blue. Measurement parameters: (a) 50 mV, 1 nA, and 8 K; (c) 800 mV, 1 nA, and 4 K; (d) 500 mV, 1 nA, and 4 K.

Figure 3

Magnetic state of the hexagonal phase of the hydrogenated Fe monolayer on Ir(111). (a) Spin-resolved differential conductance map showing an area of the hydrogenated Fe ML and the spin spiral in the neighboring H1 region of the DL. The spin spiral is used as a reference to determine the magnetization axis of the tip (here out of plane). (b) and (d) Differential conductance maps of two different sample areas showing the magnetic state of the hexagonal phase, measured with the same out-of-plane sensitive tip as in (a). The areas displayed in (b) and (d) correspond to opposite magnetic domains. The insets show magnified views of the data. (c) STM constant-current map of the region shown in (d). No magnetic contrast is visible, which allows for seeing the H superstructure and to conclude that the magnetic unit cell corresponds to a $p(4\times 4)$ structure. Measurement parameters: $-1\mathrm{V}$, 1 nA, and 4 K, Cr bulk tip. (e) Representation of a possible hexagonal nanoskyrmion lattice state. In this configuration, the points where the out-of-plane component of the magnetization is maximal are located on top sites of the atomic lattice (on-top state). The insets show SP-STM simulation of the two magnetic domains with an out-of-plane sensitive magnetic tip, displayed with the same lateral scale as the data in (b)–(d). The SP-STM simulations were performed with the following realistic parameters: tip-sample distance: 600 pm, work function: 4.8 eV, and spin polarization of sample ($s$) and tip ($t$): $P_s{P}_t=0.2$.

Figure 4

In-plane components of the magnetic state in the hexagonal phase. (a) Spin-resolved differential conductance map showing the two magnetic domains, the boundary between the domains coincides with the dark line. The image was measured with a tip sensitive slightly to the out-of-plane component of the sample magnetization and mostly to the in-plane component indicated by the arrow. The corresponding SP-STM simulations of the on-top state for both domains are shown in the insets. (b) Spin-resolved differential conductance map showing another in-plane component of the magnetic state and the corresponding SP-STM simulation of the on-top state as the inset. Measurement parameters: (a) $-700$ mV, 1 nA, 4 K, 0 T, Cr bulk tip; (b) $-1$ V, 1 nA, 8 K, 0 T, Fe-coated W tip. The SP-STM simulations were performed with the following realistic parameters: tip-sample distance: 600 pm, work function: 4.8 eV, and spin polarization of sample ($s$) and tip ($t$): $P_s{P}_t=0.2$.

Figure 5

Position of the H atoms in the hexagonal phase. (a) and (b) Spin-resolved differential conductance maps showing the magnetic state in the hexagonal phase, measured with an out-of-plane sensitive magnetic tip (same data as Fig. 3). Images (a) and (b) show the two opposite magnetic domains. Small darker dots are visible between the bright dots in image (a) and the dark dots in image (b) have a triangular shape. These features appear because the position of the H atoms is also visible in addition to the magnetic pattern. Measurement parameters: $-1\mathrm{V}$, 1 nA, 4 K, Cr bulk tip. (c) and (d) SP-STM simulations of the on-top state for an out-of-plane sensitive tip with an additional modulation reproducing the effect of the H atoms and allowing to reproduce the experimental data. The inset shows the pattern used to simulate the contribution of the H atoms. (e) Spin configuration of the on-top state showing the position of the H atoms which allows to reproduce the experimental data. The SP-STM simulations were performed with the following realistic parameters: tip-sample distance: 600 pm, work function: 4.8 eV, and spin polarization of sample ($s$) and tip ($t$): $P_s{P}_t=0.2$.

Figure 6

Magnetic state in the square phase. (a) Spin-resolved differential conductance map showing the magnetic state in the three rotational domains of the square H superstructure. The three small images show SP-STM simulations corresponding to each domain for the spin configuration shown in (b). Measurement parameters: 1.4 V, 3 nA, 4 K, 0 T, Cr bulk tip. The SP-STM simulations were performed with the following realistic parameters: tip-sample distance: 600 pm, work function: 4.8 eV, and spin polarization of sample ($s$) and tip ($t$): $P_s{P}_t=0.2$.