The Birth of a Giant
Imagine a grid of squares where each square is randomly filled with probability p. At low p, you see scattered isolated dots and tiny clusters. As p increases, clusters grow and merge. Then, at a precise critical value p_c = 0.5927, something dramatic happens: a single giant cluster suddenly connects one side of the grid to the other. This is percolation — one of the most fundamental phase transitions in physics.
Site Percolation on a Square Lattice
In site percolation, each lattice site is independently occupied with probability p and empty with probability 1-p. Two occupied sites are connected if they are nearest neighbors (up, down, left, right). A cluster is a maximal set of connected occupied sites. The central question is: at what p does a cluster first span the entire system from top to bottom?
Critical Phenomena
Near the percolation threshold, the system exhibits remarkable critical phenomena. The cluster size distribution follows a power law. The correlation length (typical cluster diameter) diverges. The probability of belonging to the giant cluster scales as P ~ (p - p_c)^(5/36). These are universal properties — they depend only on the dimensionality of the lattice, not its specific geometry.
Try It Yourself
Slowly sweep the occupation probability from 0 to 1. Watch how isolated sites merge into clusters. At p near 0.59, the largest cluster (shown in red) suddenly spans the grid. The chart below shows the order parameter curve with the critical threshold marked. Notice how the transition becomes sharper as you increase the grid size.