Four-State Active Potts Model

• The four-state Active Potts Model is a framework for studying nonequilibrium dynamics in self-propelled particle systems with four discrete internal directions.
• It employs local alignment, biased hopping, and cyclic flipping rules to reveal discontinuous flocking transitions, coexistence phases, and band-to-lane reorientation.
• Hydrodynamic theories and coarse-grained descriptions provide actionable insights into anisotropic diffusion, pattern formation, and phase stability in active matter.

The four-state Active Potts Model (APM) generalizes classic equilibrium Potts models to nonequilibrium, self-propelled particle systems with four discrete internal states. In the "standard" APM, each particle on a lattice (or in continuous space with discrete orientations) has an internal variable corresponding to one of four movement directions, local alignment interaction, and self-propulsion via biased hopping. The q = 4 variant exhibits a spectrum of nonequilibrium macroscopic phenomenology absent in both the equilibrium Potts model and in models with different internal state cardinality, including discontinuous flocking transitions, stable coexisting phases, and a robust reorientation from moving "bands" to "lanes" as activity is tuned. Dynamical extensions employing cyclic flipping rules produce additional multi-color coexistence and non-trivial spatiotemporal pattern formation.

1. Microscopic Definition and Model Variants

The basic four-state APM is constructed on a two-dimensional periodic square lattice or in continuous space with discrete orientations. Each of the N particles is assigned an internal state (labeled $\sigma \in \{1,2,3,4\}$ or, equivalently, $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$), corresponding to four orthogonal directions. For lattice models:

• Any number of particles may occupy a given site (bosonic occupancy, no exclusion).
• Particles interact via on-site or neighborhood Potts-like ferromagnetic alignment, favoring coinciding internal states:

$$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$

where $\rho_i$ is the total number of particles at site $i$, and $J>0$.

• Internal-state updates (spin flips) occur with rate:

$$W_{\mathrm{flip}}(\sigma, \sigma') = \gamma\exp\left[-\frac{4\beta J}{\rho_i}(n_i^\sigma - n_i^{\sigma'} -1)\right]$$

with effective temperature $T=1/\beta$.

• Self-propulsion is implemented as biased hopping. A particle in state $\sigma$ attempts to hop in direction $p$ with rate:

$$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$0

so hopping in the preferred direction (p = σ) is at rate $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$1; otherwise, $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$2. The self-propulsion velocity is $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$3, with $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$4.

For continuous-space ("clock" or "ACM") variants, positions are real-valued, and orientational degrees of freedom are discretized into four equally spaced angles; alignment and propulsion rules are analogously discretized, with short-range interaction neighborhoods (Mangeat et al., 2020, Chatterjee et al., 2022).

Dynamical extensions, including cyclic flipping or state-dependent flipping energies, generate a broader landscape of nonequilibrium behavior and are critical in the emergence of spatiotemporal multistability, wave, and "cycling" regimes (Noguchi, 2024, Noguchi, 30 Jun 2025).

2. Hydrodynamic Theory and Coarse-Grained Descriptions

Coarse-graining the microscopic dynamics yields coupled PDEs for local densities $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$5 (for each state $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$6) and the polarization field $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$7. The general structure is: $$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$8

$$\theta \in \{0, \frac{\pi}{2}, \pi, \frac{3\pi}{2}\}$$9

The equations contain

• anisotropic diffusion coefficients: $$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$0, $$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$1,
• velocity $$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$2,
• cubic nonlinear alignment terms, and
• coupling to higher-rank (nematic) order parameters in clock models.

Unlike the Toner–Tu equations of the Vicsek (continuous) model, the discrete (q=4) symmetry produces explicit anisotropic gradients and nontrivial cubic nonlinearities, precluding direct mapping except in certain limits. Stability analysis on homogeneous or banded solutions quantifies the onset of order and the transitions between coexisting dynamic regimes (Mangeat et al., 2020, Chatterjee et al., 2022, Chatterjee et al., 2019).

3. Phase Behavior and Reorientation Transitions

The q=4 APM exhibits a discontinuous (first-order) flocking transition for any non-zero velocity bias ($$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$3), distinguishing it from the continuous (second order) transition of the passive q=4 Potts model and the equilibrium clock model. The phase diagram includes:

• Gas phase: disordered, low density ($$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$4) or high noise; order parameter $$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$5
• Polar liquid: homogeneous, ordered, $$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$6 at high density/low noise
• Coexistence (macrophase separation): formation of a single, propagating dense band (liquid), phase separated from a gaseous background. The band width $$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$7 grows linearly with $$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$8 via the lever rule: $$H_i = -\frac{J}{2\rho_i} \sum_{k\neq\ell} (4\,\delta_{\sigma_i^k, \sigma_i^\ell} - 1)$$9 (Chatterjee et al., 2022).

Within the coexistence regime, a sharp band-to-lane reorientation transition emerges as $\rho_i$0 increases:

• At low $\rho_i$1, bands move transversely (orthogonal to their elongation).
• For $\rho_i$2, bands become stationary and particles stream longitudinally (along the band axis), forming "lanes".
• The transition is driven by the vanishing of the transverse diffusion $\rho_i$3 as $\rho_i$4, with the threshold approximately $\rho_i$5 near the critical temperature (Mangeat et al., 2020, Chatterjee et al., 2022, Chatterjee et al., 2019).

Monte Carlo simulations confirm the theoretical phase boundaries, band velocities, and discontinuity at the transition (Mangeat et al., 2020).

4. Dynamical Regimes and Spatiotemporal Patterns

Introducing cyclic, asymmetric, or extended flipping rules transforms the dynamical phase space. Pattern taxonomy includes:

• HC4 (homogeneous cycling of 4 states): At low activity or bias, the system cyclically switches between uniform occupation by each state, via nucleation-and-growth processes. Skipping of intermediates (k→k+2) may occur at higher bias (Noguchi, 30 Jun 2025, Noguchi, 2024).
• W4 (four-state coexisting waves): At higher bias, domain boundaries between successive states (k and k+1) propagate ballistically, producing a patchwork of coherently moving interfaces ("waves"). True spiral cores are absent at J₂ = 0, but emerge at J₂ < 0 (contact energy for diagonal pairs negative) (Noguchi, 30 Jun 2025).
• Q (quad-phase coexistence): For sufficiently large flipping bias, all four states coexist in space with fluctuating, disordered fragments and traveling boundaries (no spiral tip stability) (Noguchi, 2024).
• D₀,₂ and D₁,₃ (diagonal two-phase coexistence): Asymmetry (e.g., $\rho_i$6) stabilizes coexistence of diagonal pairs (0,2) or (1,3); one may observe shrinking circular domains of one diagonal state embedded in the other, with curvature-driven boundary dynamics (Noguchi, 2024).
• HC2 and M2 (two-state cycling and mixed): For positive next-nearest neighbor interaction (J₂ > 0), alternating "homogeneous cycling" (HC2) and mixed two-state (M2) phases supplant the four-state cycling as dominant phases, with a continuous transition at $\rho_i$7 (h = 1) (Noguchi, 30 Jun 2025).
• HC3 and SW (with three-state cycling): Allowing direct transitions such as $\rho_i$8 enables competition or coexistence of three- and four-state cycling, with associated spiral wave patterns similar to rock–paper–scissors (Noguchi, 2024).

5. Order Parameters and Diagnostic Measures

Characterization of phases and transitions relies on a set of global and local order parameters:

Order Parameter	Mathematical Expression	Phase Sensitivity
Global polarization, $\rho_i$9	$i$0, $i$1	Detects flocking symmetry breaking
Band orientation	$i$2	Band vs. lane (transverse vs. longitudinal)
Magnetizations, $i$3	$i$4, $i$5, $i$6	$i$7: four-state order; $i$8: diagonal-pair order (Noguchi, 30 Jun 2025)
Binder cumulant, $i$9	$J>0$0	Criticality of transitions
Phase occupancy probabilities	$J>0$1, $J>0$2, $J>0$3	Stability of single/multi-phase regimes (Noguchi, 2024)
Net cycle flow	$J>0$4	Persistence of cycling (Noguchi, 2024)

Phase boundaries are validated by locating sharp changes, discontinuities, and critical crossings in these observables in finite-size or ensemble-averaged data.

6. Broader Implications and Connections

The four-state APM provides a robust platform to investigate how discrete symmetry, active self-propulsion, and nonequilibrium cyclic kinetics generate new classes of phase separation, order–disorder transitions, and spatiotemporal organization unreachable in either equilibrium models or continuous-symmetry (q→∞) analogues. The sharp band-to-lane reorientation, absence of microphase ordering at q=4, and stabilization/competition of multi-phase cyclic and diagonal regimes all signal the importance of symmetry class and dynamical rules. Embedding cyclic kinetic motifs (e.g., three-state cycles) yields spirals and additional route to multistability and pattern selection (Mangeat et al., 2020, Chatterjee et al., 2022, Noguchi, 2024, Noguchi, 30 Jun 2025).

A plausible implication is that generalizations of the APM will yield tractable frameworks for understanding active matter with finite symmetries and multi-channel interaction kinetics, connecting to biological patterning, driven soft matter, and catalysis at nonequilibrium steady states.

7. Key References

• "Flocking with a $J>0$5-fold discrete symmetry: band-to-lane transition in the active Potts model" (Mangeat et al., 2020)
• "Polar flocks with discretized directions: the active clock model approaching the Vicsek model" (Chatterjee et al., 2022)
• "Spatiotemporal Patterns in Active Four-State Potts Models" (Noguchi, 2024)
• "Dynamic modes of active Potts models with factorizable numbers of states" (Noguchi, 30 Jun 2025)
• "Flocking and reorientation transition in the 4-state active Potts model" (Chatterjee et al., 2019)