Skyrmion morphology in ultrathin magnetic films

Nitrogen-vacancy magnetic microscopy is employed in the quenching mode as a noninvasive, high-resolution tool to investigate the morphology of isolated skyrmions in ultrathin magnetic films. The skyrmion size and shape are found to be strongly affected by local pinning effects and magnetic field history. Micromagnetic simulations including a static disorder, based on the physical model of grain-to-grain thickness variations, reproduce all experimental observations and reveal the key role of disorder and magnetic history in the stabilization of skyrmions in ultrathin magnetic films. This work opens the way to an in-depth understanding of skyrmion dynamics in real, disordered media.

Figure 1

(a) Principle of the experiment. A single NV defect placed at the apex of a diamond tip is employed as a noninvasive, scanning nanomagnetometer operating under ambient conditions. A microscope objective is used both to excite (green arrow) and collect the magnetic-field-dependent photoluminescence (PL) of the NV defect (red wavy arrows). The magnetic sample is a symmetric bilayer system with a stack of Pt(5 nm)/FM/Au(3 nm)/FM/Pt(5 nm), where $\text{FM}=\text{Ni}$(4 Å)/Co(7 Å)/Ni(4 Å). (b) PL quenching image recorded in zero external field ($B_z=0$).

Figure 2

Step-by-step formation of isolated skyrmions by applying an external out-of-plane magnetic field $B_z$. (a) PL quenching images recorded at zero field and (b) $B_z=3$ mT. (c), (d) Images recorded at $B_z=3$ mT after applying a 10-s field pulse of (c) 7 and (d) 10 mT. The bright PL spots correspond to particles on the sample which serve as position references.

Figure 3

(a) PL quenching images recorded above several isolated skyrmions at various positions in the sample. These experiments are performed at $B_z=3$ mT after applying a field pulse of 10 mT. The white dashed contours indicate the PL quenching ring from which the skyrmion area $\mathcal{A}$ is extracted. (b) Histogram of the effective skyrmion diameter $d_s$ extracted from measurements over a set of 27 skyrmions. The solid line is a fit with a Gaussian distribution. (c) Typical micromagnetic simulations of the skyrmion spin texture by including thickness fluctuations with a relative amplitude of $1\%$. (d) Histogram of the effective skyrmion diameter obtained for a large number of randomly picked disorder configurations. The blue dashed arrow indicates the skyrmion diameter for a disorder-free sample (860 nm) and the black dashed arrow shows the initial skyrmion size in the simulation (40 nm) before expansion in a 3-mT field.

Figure 4

(a), (b) Effective skyrmion size distribution obtained for different strengths of disorder after relaxation in a 3-mT field while starting with a skyrmion diameter of (a) 40 and (b) 1500 nm. The black dashed line in the middle indicates the skyrmion diameter for a disorder-free sample (860 nm). (c) Effective skyrmion size after relaxation vs initial size for different disorder strengths. The error bars represent the variance of the size distribution. The black dashed line corresponds to a disorder-free sample and the diagonal (black solid line) is a guide to the eye for the case without relaxation.