Dynamical Spreading and Memory Retention Under Power Law Potential

Technical Summary: Dynamical Spreading and Memory Retention of Particle Suspensions under Power Law Potential

Problem Statement
While power law potentials ($U \propto 1/r^k$) are fundamental to numerous physical phenomena, existing literature has predominantly focused on equilibrium configurations and phase transitions. Far less attention has been devoted to the non-equilibrium dynamics of such systems, particularly the deterministic spreading of overdamped particle suspensions. This study addresses the gap in understanding how an initially concentrated suspension of particles evolves in free space under repulsive power law interactions, specifically investigating whether the system exhibits self-similar spreading and how the interaction range ($k$) relative to spatial dimension ($d$) influences the relaxation dynamics and memory retention of the initial state.

Methodology
The authors employ a multi-faceted approach combining analytical derivation, numerical simulation, and experimental validation:

1. Analytical Framework: The system is modeled as overdamped, where particle velocity is proportional to the net force ($v = \mu F$). The authors derive a continuum model using the continuity equation and a self-similar ansatz for the density profile, $\rho(r, t) = At^\gamma f(Br/t^\beta)$. They determine scaling exponents based on the power law potential $U \propto 1/r^k$.
2. Numerical Simulations: Molecular dynamics simulations were conducted using an eighth-order Runge-Kutta scheme with adaptive time steps. Simulations involved thousands of point-like particles interacting via dipolar repulsion ($k=3$) and various other power laws ($k = -1, 0, 1$) in 1D and 2D. The simulations tracked the evolution of density profiles, standard deviations, and collision dynamics between multiple suspensions.
3. Experimental Setup: Experiments utilized superparamagnetic colloids (polystyrene and silica) embedded with iron oxide nanoparticles, suspended in deionized water. The particles were magnetized perpendicularly using Helmholtz coils, inducing a dipole-dipole repulsion scaling as $1/r^3$ ($k=3$). The system was observed under a microscope to track particle positions and density evolution over time. Thermal diffusion was deemed negligible due to high Péclet numbers ($Pe \sim 100-300$).

Key Contributions and Results

• Self-Similar Spreading Dynamics: The study predicts and confirms that the suspension spreads in a self-similar form. The radius of the suspension grows as $R(t) \propto t^\beta$, where the exponent is $\beta = 1/(k+2)$. Crucially, this exponent is independent of the spatial dimension $d$; dimension only affects the density amplitude.

  • For the experimental case of magnetic colloids ($k=3$ in 2D), the radius grows as $t^{1/5}$, and density profiles at different times collapse onto a single curve when rescaled by $t^{2/5}$ and $r/t^{1/5}$.
  • The density profile is compact with a sharp boundary, distinct from the diffusive tails of passive thermal particles.

• Classification of Density Profiles: The dynamical evolution is categorized into three regimes based on the relationship between the interaction exponent $k$ and dimension $d$:

  1. Origin-centered ($k > d-2$): Particles remain concentrated near the origin (e.g., $k=3$ in 2D).
  2. Constant profile ($k = d-2$): The system exhibits a "Coulomb gas" or log-gas behavior.
  3. Boundary-centered ($k < d-2$): Particles accumulate at the perimeter, resembling the "coffee-ring effect" or "evaporation catastrophe."

• Memory Retention and Critical Threshold: A significant finding is the existence of a critical power $k_c = d-2$.

  • Above the threshold ($k \ge d-2$): When two suspensions collide, they merge rapidly into an isotropic state, losing memory of their initial configuration exponentially.
  • Below the threshold ($k < d-2$): The system exhibits anomalously slow relaxation. Instead of merging immediately, a persistent particle-free zone (void) forms between colliding suspensions. The initial distribution "seeds" the resulting pattern, encoding the future structure of the system. This memory retention follows a power law, with the time to reach a fully isotropic state approaching infinity as $k \to -2$.
  • Experiments and simulations demonstrated that for $k=-1$, a "word" written by the absence of particles (e.g., "WORLD") remains visible as the system expands, whereas for $k=3$, such patterns ("HELLO") vanish quickly.

Significance
The paper establishes that power law interactions lacking a typical length or time scale lead to universal self-similar spreading dynamics governed by the exponent $\beta = 1/(k+2)$. The primary significance lies in the discovery of a dynamical phase transition at $k = d-2$, which dictates whether a system rapidly forgets its initial conditions or retains a long-lived memory of them.

The authors suggest that this interaction-driven memory effect and the associated slow relaxation are relevant to understanding jammed and glassy materials. Furthermore, the findings may have implications for phoresis in synthetic microswimmers and communication mechanisms in robotic swarms, where the ability to maintain spatial patterns over time is crucial. The work bridges the gap between equilibrium statistical mechanics and non-equilibrium dynamics in systems governed by long-range repulsive forces.