Skyrmion Lattice States: Mechanisms & Phases

• Skyrmion lattice states are arrays of nontrivial spin textures whose integer winding numbers define their topological character.
• They are stabilized by competing magnetic interactions, including Dzyaloshinskii–Moriya interactions, frustration, and external symmetry-breaking fields across diverse lattice geometries.
• Experimental techniques like neutron scattering and Lorentz TEM reveal distinct Bragg peaks and phase transitions that confirm the unique properties of skyrmion lattices.

A skyrmion lattice state is a magnetic phase characterized by a crystalline array of topologically nontrivial spin textures—skyrmions—whose cores possess integer winding number, and whose presence is encoded in reciprocal-space via multiple Bragg peaks at symmetry-related wave vectors. The stability of such lattices emerges from competing magnetic interactions, geometric frustration, spin-orbit coupling, or external symmetry-breaking fields. This article surveys the theoretical frameworks, stabilization mechanisms across lattice and symmetry types, phase transitions, and experimental observables for skyrmion lattice states in ferro- and antiferromagnetic, chiral, and centrosymmetric systems.

1. Theoretical Foundations and Classification

The emergence of a skyrmion lattice (SkL) is typically understood in the context of noncollinear spin Hamiltonians that support multiple-Q magnetic order. The canonical ferromagnetic model on a two-dimensional lattice with nearest-neighbor exchange $J$, Dzyaloshinskii–Moriya interaction (DMI) $D$, and Zeeman coupling $h$ takes the form

$$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$

In antiferromagnets and frustrated magnets, additional competing exchanges (e.g., up to third neighbors, $J_2$, $J_3$) or Rashba–Hund mechanisms further enrich the landscape.

Skyrmion lattices may be divided along several axes:

• Chiral (non-centrosymmetric) vs. centrosymmetric: Chiral materials host Bloch- or Néel-type SkL stabilized by DMI; centrosymmetric frustrated systems can support Bloch-type SkL via RKKY or biquadratic terms, even in absence of DM interaction (Kurumaji et al., 2018, Mohylna et al., 2022, Hayami et al., 2024).
• Ferromagnetic (FM) vs. antiferromagnetic (AFM): FM SkLs exhibit strong net magnetization; AF SkLs (AF–SkL) arise from superposed, often interpenetrating, skyrmion textures with vanishing or negligible net moment (Rosales et al., 2015, Mohylna et al., 2022).
• Lattice geometry: Triangular, square, and honeycomb lattices each yield distinct SkL topologies, periodicities, and magnon band structures (Okubo et al., 2011, Keesman et al., 2016, Ghader et al., 2 Jun 2025).
• Dimension and stacking: Layered, non-symmorphic, and curvilinear-defect arrays allow for integer, fractional, or layer-resolved skyrmion numbers; nonsymmorphic screw axes can stabilize SkLs without in-plane rotational symmetry (Hayami, 2022, Kravchuk et al., 2017).

2. Topological Characterization and Spin Texture

The defining property of a skyrmion lattice is its nontrivial topology, encoded in the winding number or skyrmion number per unit cell

$$Q = \frac{1}{4\pi} \int dx\,dy\, \mathbf S \cdot \left( \partial_x \mathbf S \times \partial_y \mathbf S \right).$$

In lattice models, this is discretized via sums over oriented triangle areas weighted by local scalar chirality, e.g.,

$$\chi_Q = \frac{1}{4\pi}\left\langle \sum_{i \in \text{sublattice}} \left[ A_i^{23}\, \text{sign}\bigl( \chi_{L,i,23} \bigr) + A_i^{45} \, \text{sign}\bigl( \chi_{L,i,45} \bigr) \right] \right\rangle,$$

converging to $Q$ for smooth textures (Rosales et al., 2015).

• AF SkL: On the triangular lattice, the AF–SkL state consists of three interpenetrating SkL (one per sublattice), each corresponding to a triple-Q noncoplanar texture with the same set of six wave vectors $$\{\pm \mathbf k_1^*, \pm \mathbf k_2^*, \pm \mathbf k_3^*\}$$ (Rosales et al., 2015).
• Integer and fractional SkL: In layered systems with screw axes, both conventional (integer total $D$0 per layer) and “fractional” SkLs (where only some layers are topological) arise, with $D$1 per 3-layer unit cell (Hayami, 2022).
• Skyrmion/antiskyrmion domain states: In certain triangular AF and frustrated systems, coexisting skyrmion and antiskyrmion lattices form mesoscopic domains, breaking mirror (Z2) symmetry only locally (Okubo et al., 2011, Mohylna et al., 2022).

3. Stabilization Mechanisms Across Lattice Types

The fundamental energetic drivers for SkL formation differ crucially by symmetry and Hamiltonian content:

Chiral Magnets (w/ DMI)

• Universal phase sequence: Helical → conical → SkL (“A-phase”) → field-polarized with increasing field, independent of metallic/insulating character (Bauer et al., 2016).
• Fluctuation-induced selection: In cubic B20 systems (MnSi, FeGe), the SkL pocket is stabilized by thermal fluctuations just below $D$2 (Brazovskii scenario), as mean-field theory alone predicts metastability (Bauer et al., 2016, Leonov, 2014).
• SkL ansatz: The SkL is realized as a phase-locked triple-superposition of helices; the real-space period (lattice constant) $D$3 with $D$4 set by the DMI/exchange ratio (Leonov, 2014).

Frustrated/Centrosymmetric Magnets

• Exchange frustration: Competing $D$5–$D$6–$D$7 Heisenberg interactions on the triangular or square lattice lead to multiple-Q degenerate minima in $D$8, enabling SkL formation under applied field, even without DMI (Mohylna et al., 2022, Keesman et al., 2016).
• Biquadratic ($D$9) and RKKY effects: Itinerant systems mediate a positive biquadratic term $h$0, favoring noncoplanar triple-Q SkLs with $h$1 or $h$2; stable in, e.g., Gd$h$3PdSi$h$4, GdRu$h$5Si$h$6 (Hayami et al., 2024).
• AFM SkL specifics: In pure AF Heisenberg triangular models under field and moderate frustration, a robust spontaneous AF–SkL/ASkL phase emerges, where vorticity and helicity are spontaneously selected, and net magnetization remains small (Mohylna et al., 2022).

Nonsymmorphic and Structured Substrates

• Screw-axis stabilized SkLs: Nonsymmorphic symmetries (e.g., screw axes in ABC-stacked triangular layers) allow SkLs even without in-plane threefold rotation, and enable fractional or coexisting SkX/spiral states per layer (Hayami, 2022).
• Substrate-induced molecular SkLs: On engineered triangular substrates, multi-skyrmion per well orderings—dimers, trimers, herringbone phases—appear, with superlattice order dictated by commensuration and skyrmion–skyrmion interactions (Souza et al., 2024).

Curvilinear Defect and Pinning

• Curvature-driven SkL: Gaussian “bump” arrays on perpendicularly magnetized films act as local pinning centers for discrete skyrmion bound states; periodic arrays generate zero-field SkLs with lattice constants determined by defect spacing (Kravchuk et al., 2017).

4. Phase Diagrams and Transitions

The generic magnetic-field–temperature ($h$7–$h$8) phase diagrams exhibit the following structure:

• AF triangular lattice ($h$9): At low $$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$0 and increasing $$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$1, spiral (single-Q) phase ($$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$2) undergoes a first-order transition to AF–SkX (triple-Q, $$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$3 with constant plateau), followed by a continuous transition to a paramagnetic or vortex-like regime ($$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$4 smoothly) (Rosales et al., 2015). Bragg peaks evolve from single to sixfold in $$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$5, $$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$6 (Rosales et al., 2015).
• Frustrated Heisenberg triangular lattice: The $$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$7 phase diagram includes collinear, single-Q, double-Q, spiral spin liquid, and robust triple-Q AF–SkL/ASkL phases in region C (Mohylna et al., 2022). The SkL phase is stabilized by noncoplanar triple-Q mixing via entropic order by disorder at moderate $$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$8 and $$\mathcal H = J\sum_{\langle i,j\rangle}\mathbf S_i\cdot\mathbf S_j + \sum_{\langle i,j\rangle} \mathbf D_{ij}\cdot(\mathbf S_i\times\mathbf S_j) - h\sum_i S_i^z.$$9.
• Square lattice AFM + DMI: The (B, D) plane is partitioned into AFM, spin-flop, spiral (1q), and 2q (skyrmion-like) regions, the last bounded on both sides by first-order transitions; in 2q, a square-lattice “skyrmion crystal” with alternating lobes in staggered field develops (Keesman et al., 2016).
• Chiral ferromagnets: The skyrmion lattice occupies a pocket just below $J_2$0 between $J_2$1 and $J_2$2 for typical DMI/exchange ratios (Han et al., 2010, Leonov, 2014).

5. Numerical and Experimental Diagnostics

Numerical Methodologies

• Monte Carlo simulations: Employed to map phase diagrams and thermal stability, utilizing heat-bath, Metropolis, over-relaxation, and parallel tempering algorithms on large $J_2$3 clusters ($J_2$4) (Rosales et al., 2015, Mohylna et al., 2022).
• Spherical (weak) constraint methods: Global spin-length constraint in Fourier-space enables identification of triple-Q minima, confirming thermodynamic SkL stability down to $J_2$5 (Rosales et al., 2015).
• Fast Fourier-space variational techniques: Enable free-energy minimization and constraints enforcement for large unit cells and arbitrary lattice symmetry (triangular, square, honeycomb); used also to identify metastable SkL patterns (honeycomb, square) (Balkind et al., 2019).

Structure Factor and Real-Space Observables

Phase	$J_2$6, $J_2$7	Skyrmion Number ($J_2$8)	Comment
Spiral (single-Q)	Peaks at $J_2$9	0	Simple modulated state
AF–SkL (triple-Q)	Sixfold peaks at $J_3$0	Plateau, constant over phase	Hexagonal SkL superlattice
Paramagnet	Diffuse/ring at $J_3$1	0	Featureless
Square 2q (AFM)	Four peaks at $J_3$2, $J_3$3	Finite (checkerboard of $J_3$4)	Square skyrmion-like sublattice

Bragg peak evolution and magnetization curves exhibit discontinuous jumps at first-order transitions and smooth changes at continuous transitions. Experimentally, neutron scattering yields sixfold symmetric Bragg peaks in the SkL phase, while Lorentz TEM and STM reveal real-space skyrmion arrays and allow determination of $J_3$5 profiles (Rosales et al., 2015, Kurumaji et al., 2018, Mohylna et al., 2022).

6. Finite-Temperature and Quantum Effects

• Thermal phase transitions: AF–SkL phases in the AF triangular model persist to $J_3$6; finite $J_3$7 “order by disorder” stabilizes noncoplanar triple-Q order out of degenerate spiral manifolds (Rosales et al., 2015, Mohylna et al., 2022, Okubo et al., 2011).
• Robustness: AF–SkL states, unlike DM-stabilized FM SkL, require only modest $J_3$8 or even none, in frustrated models; their topological charge and chirality remain robust under field quenches and over long Monte Carlo times.
• Melting and domain states: Upon heating, SkL phases can melt via first-order transitions to “polyhexatic” states characterized by orientational order in finite domains, preceding full melting to a skyrmion liquid (Garanin et al., 2022).
• Quantization and quantum entanglement: In quantum spin–½ models, a quantum SkL phase features enhanced entanglement entropy and localized quantum correlations across skyrmion domain walls, in contrast to classical product states; quantum SkL appears in the ground-state phase diagram for moderate DMI and field (Haller et al., 2021).

7. Implications and Extensions

The demonstration that skyrmion lattices can be stabilized through multiple, sometimes coexisting, mechanisms—including chiral DMI, geometric frustration, biquadratic exchange, lattice geometry, symmetry, and substrate design—vastly expands the playground for topological spin textures in condensed matter. Notably:

• Robust topological transport: Both FM and AF SkL generate topological Hall effects via emergent electromagnetic fields, independent of the underlying electronic structure (Kurumaji et al., 2018).
• Superlattice and molecular SkL: Engineered substrates enable control of skyrmion size, shape, and composite states (dimers, trimers), with superlattice order and variation in annihilation/deformation transitions (Souza et al., 2024).
• Fractional SkL and layer-selectivity: Nonsymmorphic, screw-axis systems admit fractional occupation of the SkL state, suggesting routes to tunable, non-uniform topological textures (Hayami, 2022).
• Metastable and defect-pinned SkL: Pinning centers, substrate patterning, and sample curvature provide means to stabilize otherwise metastable SkL patterns, including at zero field (Kravchuk et al., 2017).

In summary, the skyrmion lattice state exemplifies a generic, robust, and tunable class of topological matter, whose realization spans both chiral and frustrated regimes, multiple dimensions and symmetry classes, and persists across both classical and quantum scales (Rosales et al., 2015, Mohylna et al., 2022, Hayami et al., 2024, Bauer et al., 2016, Kurumaji et al., 2018, Hayami, 2022, Balkind et al., 2019).