Skyrmion Pinning Landscape in Magnetic Systems

• Skyrmion pinning landscape is the spatially inhomogeneous potential energy profile from engineered defects or intrinsic variations that dictates skyrmion equilibrium and motion.
• Theoretical models like the Thiele equation and first-principles simulations quantify depinning thresholds and force dynamics crucial for performance in magnetic devices.
• Engineering these landscapes with precise pinning centers enables optimization of memory, logic, and neuromorphic devices by tuning energy wells and barrier strengths.

A skyrmion pinning landscape refers to the spatially inhomogeneous potential energy profile that magnetic skyrmions experience due to engineered or intrinsic material variations, defects, or nanostructured features. The landscape determines both the equilibrium positions and the dynamical pathways of skyrmions under external drives such as spin-transfer torques or magnetic field gradients. Detailed understanding and control of the pinning landscape are essential for functional skyrmion-based devices, as the depinning behavior, directionality of motion, and stability are all encoded in the underlying energy and force topography.

1. Fundamental Models and Theoretical Framework

The microscopic and mesoscopic modeling of the skyrmion pinning landscape is typically rooted in a collective-coordinate (Thiele) description. For a single skyrmion, the generalized equation of motion reads: $$\alpha_d\,\mathbf{v} + \alpha_m\,\hat{z}\times\mathbf{v} = \mathbf{F}_\text{drive} - \sum_o \nabla U_o(\mathbf{r}-\mathbf{r}_o)$$ where $\alpha_d$ is a dissipative coefficient, $\alpha_m$ is the (dimensionless) Magnus term, $U_o(\mathbf{r})$ are the defect potentials (attractive or repulsive), and the drive can be a spin-current, field gradient, or other force (Basseto et al., 16 Oct 2025).

In the regime of dilute pinning, $U_o$ is well represented by Gaussian wells/hills (for pointlike pins), but the formalism is extensible to extended inhomogeneities or complex disorder (grain boundaries, modulated anisotropy, or thickness variations) (Reichhardt et al., 2021, Gross et al., 2017).

At the atomistic level, first-principles calculations based on extended Heisenberg Hamiltonians with parameters for exchange, DMI, anisotropy, and Zeeman coupling allow direct construction of pinning potentials arising from engineered defects or clusters (Arjana et al., 2020).

2. Classes and Origins of Pinning Centers

Pinning centers can be classified as attractive (trapping) or repulsive (blocking), determined by the local modification of micromagnetic parameters (Reichhardt et al., 2021):

• Point defects: Individual impurities (adatoms, substitutional atoms) modulate the exchange, DMI, or anisotropy. Ferromagnetic (e.g., Fe) or antiferromagnetic (e.g., Cr) adatom clusters on PdFe/Ir(111) are canonical examples, presenting either purely attractive, repulsive, or complex mixed (multi-domain) profiles depending on composition, geometry, and stacking (Arjana et al., 2020, Hanneken et al., 2016).
• Nanostructured features: Lithographically defined disks, notches, antidots, or thickness modulations create mesoscale energy wells or barriers (Gong et al., 2022, Pathak et al., 2021). The profile of the potential (e.g., symmetric central, off-center minimum, or ring-shaped) depends on both the property contrast and the wall structure of the skyrmion (thin- vs thick-wall).
• Intrinsic disorder: Polycrystalline grains or random thickness/anisotropy variations in multilayers introduce stochastic landscape features with characteristic length scales set by grain size or interface roughness. These are quantitatively characterized by extracting statistical distributions of well depths, correlation lengths, and spacing (Zeissler et al., 2017, Gross et al., 2017).

The table below summarizes typical pinning-site characteristics extracted from experimental and theoretical works.

Pinning Type	Length Scale	Typical Depth/Barrier
Atomic In-layer Defect	0.1–0.5 nm	~0.1–1 meV
Small Cluster (Co/Cr/Fe)	1–2 nm	5–10 meV
Nanodisk (modulated $A$, $D$)	10–100 nm	$\sim$meV to tens of meV
Disorder Pocket (grain)	10–20 nm	$10^{-23}$–$10^{-21}$ J
Lithographic Pocket	20–70 nm	$\alpha_d$0–$\alpha_d$1 J

(Arjana et al., 2020, Gong et al., 2022, Zeissler et al., 2017, Pathak et al., 2021, Hanneken et al., 2016)

3. Quantitative Characterization and Measurement

The local potential landscape $\alpha_d$2 is extracted experimentally by mapping the position-resolved occupation probability $\alpha_d$3 of thermally diffusing skyrmions and invoking the Boltzmann relation: $\alpha_d$4 This method provides energy-resolved landscapes with spatial resolution set by the imaging modality (typically $\alpha_d$550–300 nm for Kerr or x-ray microscopy) (Fröhlich et al., 17 Oct 2025, Gruber et al., 2022, Gruber et al., 18 Aug 2025).

In simulations, the landscape can be directly imposed or self-consistently generated by modulating micromagnetic parameters spatially. The extracted landscapes typically feature wells (traps) and barriers with amplitude and correlation length dictated by the underlying material structure. For example, in Pt/Co/Ir multilayers, disorder-derived pinning wells have widths $\alpha_d$610–15 nm, energy depths $\alpha_d$7–$\alpha_d$8 J, and pinning forces in the $\alpha_d$9–$\alpha_m$0 pN range (Zeissler et al., 2017).

Further, the dynamical response—diffusion coefficient, dwell-time at minima, and thermal hopping rates—delivers direct estimates of barrier heights and damping parameters, allowing full calibration of coarse-grained Thiele models to experiment (Fröhlich et al., 17 Oct 2025).

4. Pinning Landscape Effects: Depinning and Directional Locking

The pinning landscape critically determines the depinning threshold $\alpha_m$1 and the dynamical regimes accessible under external drives. In engineered periodic landscapes (mixed arrays of attractive and repulsive pins), directional locking emerges, quantizing the skyrmion Hall angle $\alpha_m$2 into discrete plateaus (e.g., $\alpha_m$3, $\alpha_m$4, $\alpha_m$5, $\alpha_m$6), with each plateau corresponding to a rational ratio $\alpha_m$7 set by the symmetry and spacing of the array (Basseto et al., 16 Oct 2025).

The depinning force $\alpha_m$8 in such arrays scales linearly with attractive well strength $\alpha_m$9 ($U_o(\mathbf{r})$0), is largely insensitive to repulsive hill strength $U_o(\mathbf{r})$1 (which instead expands the angular locking window), and decreases for shallower traps. The radius of pinning sites ($U_o(\mathbf{r})$2, $U_o(\mathbf{r})$3) selects which commensurability steps are prominent.

In random landscapes, the transition from pinning to flow is marked by a threshold force or current density, and the velocity–force relation exhibits either elastic ($U_o(\mathbf{r})$4) or plastic ($U_o(\mathbf{r})$5) behavior (Reichhardt et al., 2021).

Pinning also modifies the skyrmion Hall effect: the Hall angle increases with driving force and saturates to its intrinsic value only at high velocities. The Magnus force (gyrocoupling) allows skyrmions to be deflected around pointlike pins, lowering $U_o(\mathbf{r})$6 compared to the purely dissipative case, while for extended defects (e.g., grain boundaries), full pinning is restored (Basseto et al., 16 Oct 2025, Reichhardt et al., 2021).

5. Design and Engineering of Artificial Pinning Landscapes

Precise device functionalities (memories, logic gates, neuromorphic elements) can be implemented by deliberately engineering the skyrmion pinning landscape:

• Racetracks and memory arrays: Artificial arrays of Gaussian pins (wells/hills) can create pathways with deterministic write, read, and flow directions. The parameters $U_o(\mathbf{r})$7, $U_o(\mathbf{r})$8, $U_o(\mathbf{r})$9, $U_o$0, and lattice spacing are tunable knobs for setting depinning thresholds, locking angles, and operational regimes. For robust $U_o$1 channeling, $U_o$2, $U_o$3, and $U_o$4 are recommended (Basseto et al., 16 Oct 2025).
• Sub-nanometer precision: Atom-by-atom engineered clusters enable racetrack architectures with energy profiles, barrier heights, and domain patterns determined at the scale of the magnetic lattice constant (Arjana et al., 2020). Designer “Lego-style” lattice landscapes are constructed by alternating pinning and repulsive motifs, switching domain topologies and bit register performance.
• Boundary engineering: Edge decoration with ferromagnetic or antiferromagnetic rims tunes the confinement potential (e.g., “Lennard-Jones–like” edge potentials), stabilizing or repelling skyrmions at device boundaries, with field-tunable well depths and positions (Spethmann et al., 2021).
• Reservoir computing/Brownian tokens: For neuromorphic computing schemes, complex multi-well landscapes are rationalized and optimized using direct measurements of $U_o$5 and barrier distributions, with target complexity and memory times set by the number and depth of wells, and nonlinearity introduced via dynamic modulation (e.g., field oscillation-induced “shaking”) (Gruber et al., 2022, Gruber et al., 18 Aug 2025).

6. Stochastic Dynamics, Thermal Effects, and Quantum Phenomena

Thermal fluctuations cause skyrmions to hop between minima in the pinning landscape, with mean escape time $U_o$6 determined by the energy barrier $U_o$7 (Gong et al., 2022). At sub-micron scales and room temperature, typical $U_o$8 are a few $U_o$9, with observed dwell times from seconds to minutes (Gruber et al., 2022). Variations in field, size, or temperature provide real-time tuning of pinning efficacy and correspondingly of device switching rates.

Quantized skyrmion dynamics arise in ultrasmall systems or at low temperatures: the emergent magnetic field splits the lowest Landau level into quantized orbits in a pinning well, with level spacings $A$0 in the Lorentzian well model, and microwave transitions observable in the 10–100 GHz range (Lin et al., 2013).

7. Implications for Devices and Future Directions

The skyrmion pinning landscape is a central determinant in the engineering of efficient, robust, and reconfigurable skyrmionic devices. System design must balance stability (for long information retention) with mobility (for fast logic or memory operations), controlled via a hierarchy of pinning site depth, distribution, and periodicity (Basseto et al., 16 Oct 2025).

Key open challenges include:

• Quantitative inverse design: reconstructing or predicting the energy landscape from real material data and desired device operation.
• Control of multiskyrmion interactions within non-flat potential landscapes, including elastic and plastic flow regimes.
• Design of landscapes for optimal neuromorphic functionality, leveraging the landscape topology for robust, tunable reservoir properties.
• Thermally driven or quantum effects in sub-10 nm systems, and direct observation of quantum-level splitting in engineered pinning potentials.

Ongoing advances in experimental mapping, first-principles simulation, and stochastic Thiele-based modeling are converging to realize predictive, rational engineering of the skyrmion pinning landscape across scales, from atomically crafted lattices to macroscopic device arrays (Fröhlich et al., 17 Oct 2025, Reichhardt et al., 2021, Basseto et al., 16 Oct 2025).