Diffusion-Limited Aggregation Theory

Introduction to Diffusion-Limited Aggregation

Diffusion-Limited Aggregation (DLA) is a process whereby particles undergoing Brownian motion cluster together to form aggregates of particles. This model was introduced by T.A. Witten Jr. and L.M. Sander in 1981 [1] and has since become a paradigm for understanding fractal growth phenomena in nature.

Mathematical Foundations

Random Walk Process

The foundation of DLA is the random walk, where particles move according to:

$$\mathbf{r}(t+1) = \mathbf{r}(t) + \mathbf{\xi}(t)$$

where $\mathbf{\xi}(t)$ is a random displacement vector chosen uniformly from the neighboring lattice sites.

Fractal Dimension

DLA clusters exhibit fractal geometry with a characteristic dimension $D_f \approx 1.71$ in 2D [2]. The mass (number of particles) scales with radius as:

$$M(R) \sim R^{D_f}$$

This sub-quadratic scaling means DLA clusters are less dense than compact objects but more space-filling than linear structures.

Growth Probability Distribution

The probability of a walker sticking at position $\mathbf{r}$ is proportional to the local electric field in the equivalent electrostatic problem:

$$P(\mathbf{r}) \propto |\nabla \phi(\mathbf{r})|^\eta$$

where $\phi$ satisfies Laplace’s equation $\nabla^2 \phi = 0$ with appropriate boundary conditions, and $\eta$ is the growth exponent [3].

Physical Principles

Screening Effect

A key feature of DLA is the “screening effect” or “shadowing” - particles are more likely to attach to protruding tips than to deep fjords. This occurs because:

1. Random walkers approaching from infinity have higher probability of encountering outer branches
2. Inner regions are “screened” by outer growth
3. This positive feedback creates the characteristic branching structure

Universality

DLA exhibits universal behavior independent of microscopic details:

• The fractal dimension remains $D_f \approx 1.71$ across different lattices
• Growth patterns show self-similarity across scales
• Statistical properties are robust to variations in the random walk rules

Applications in Nature

DLA patterns appear in numerous physical systems:

Electrodeposition

Metal ions in solution undergo random motion until depositing on an electrode, creating dendritic patterns similar to DLA [4].

Bacterial Colonies

Some bacteria colonies grow in DLA-like patterns when nutrients are limited and cells must search for resources [5].

Lightning and Dielectric Breakdown

Electrical discharge paths follow DLA-like patterns as charge carriers seek the path of least resistance [6].

Mineral Deposition

Manganese oxide dendrites and other mineral formations show DLA characteristics [7].

Computational Considerations

Algorithmic Efficiency

The naive DLA algorithm has time complexity $O(N^2)$ for $N$ particles. Optimizations include:

1. Off-lattice launching: Start walkers from a circle around the cluster
2. Variable step sizes: Use larger steps far from the cluster
3. Killing radius: Terminate walkers that wander too far

Noise Reduction

DLA clusters show significant variation due to their stochastic nature. Techniques for analysis include:

• Averaging over multiple realizations
• Measuring ensemble properties (fractal dimension, density profiles)
• Using larger clusters to reduce finite-size effects

Connection to Other Models

Eden Model

Unlike DLA where particles undergo random walks, the Eden model grows by randomly selecting perimeter sites. Eden clusters are compact with $D_f = 2$ [8].

Ballistic Aggregation

Particles move in straight lines rather than random walks, producing more compact, less branched structures [9].

Reaction-Limited Aggregation

Particles must attempt attachment multiple times before sticking, leading to more compact growth [10].

Quantitative Analysis

Density Profile

The average density $\rho(r)$ as a function of distance from the center follows:

$$\rho(r) \sim r^{D_f - d}$$

where $d$ is the embedding dimension (2 for planar DLA).

Growth Site Distribution

The distribution of growth probabilities follows a multifractal spectrum, characterized by the generalized dimensions $D_q$ [11].

Harmonic Measure

The growth probability distribution is related to the harmonic measure on the cluster boundary, connecting DLA to potential theory [12].

Open Questions

Despite extensive study, several aspects of DLA remain incompletely understood:

1. Exact fractal dimension: While numerically estimated as $D_f \approx 1.71$, no analytical derivation exists
2. Noise effects: The role of lattice anisotropy and finite-size effects on pattern formation
3. Three-dimensional DLA: Less studied than 2D, with $D_f \approx 2.5$ in 3D
4. Multi-particle DLA: Behavior when multiple particles aggregate simultaneously

References

[1] Witten Jr, T. A., & Sander, L. M. (1981). Diffusion-limited aggregation, a kinetic critical phenomenon. Physical Review Letters, 47(19), 1400.

[2] Meakin, P. (1983). Formation of fractal clusters and networks by irreversible diffusion-limited aggregation. Physical Review Letters, 51(13), 1119.

[3] Halsey, T. C. (2000). Diffusion-limited aggregation: a model for pattern formation. Physics Today, 53(11), 36-41.

[4] Brady, R. M., & Ball, R. C. (1984). Fractal growth of copper electrodeposits. Nature, 309(5965), 225-229.

[5] Ben-Jacob, E., & Garik, P. (1990). The formation of patterns in non-equilibrium growth. Nature, 343(6258), 523-530.

[6] Niemeyer, L., Pietronero, L., & Wiesmann, H. J. (1984). Fractal dimension of dielectric breakdown. Physical Review Letters, 52(12), 1033.

[7] Chopard, B., Herrmann, H. J., & Vicsek, T. (1991). Structure and growth mechanism of mineral dendrites. Nature, 353(6343), 409-412.

[8] Eden, M. (1961). A two-dimensional growth process. Proceedings of the Fourth Berkeley Symposium on Mathematical Statistics and Probability, 4, 223-239.

[9] Vold, M. J. (1963). Computer simulation of floc formation in a colloidal suspension. Journal of Colloid Science, 18(7), 684-695.

[10] Meakin, P., & Family, F. (1987). Structure and dynamics of reaction-limited aggregation. Physical Review A, 36(11), 5498.

[11] Halsey, T. C., Jensen, M. H., Kadanoff, L. P., Procaccia, I., & Shraiman, B. I. (1986). Fractal measures and their singularities. Physical Review A, 33(2), 1141.

[12] Hastings, M. B., & Levitov, L. S. (1998). Laplacian growth as one-dimensional turbulence. Physica D, 116(1-2), 244-252.

Further Reading

• Vicsek, T. (1992). Fractal Growth Phenomena. World Scientific.
• Meakin, P. (1998). Fractals, Scaling and Growth Far from Equilibrium. Cambridge University Press.
• Wikipedia: Diffusion-limited aggregation
• Scholarpedia: Diffusion-limited aggregation