Self-avoiding walk

A self-avoiding walk (SAW) is formally defined as a path on a lattice, such as Z^d (d-dimensional integer lattice), that does not visit any vertex more than once. This constraint distinguishes it from a simple random walk, which can revisit vertices and edges. A self-avoiding polygon (SAP) is a closed SAW, returning to its origin without self-intersection.

From a purely mathematical perspective, rigorous results concerning SAWs are surprisingly scarce, especially concerning their asymptotic behavior. For instance, determining the precise growth rate of the number of SAWs of length N, or their average end-to-end distance, remains an open challenge. While exact analytical solutions are rare, physicists have developed numerous conjectures, strongly supported by extensive numerical simulations, regarding the properties of SAWs.

These simulations, often employing algorithms like the pivot algorithm, are crucial for exploring the behavior of SAWs, particularly in higher dimensions where analytical methods falter. The computational difficulty arises from the combinatorial explosion of possible paths and the need to enforce the self-avoidance constraint.

The concept of the self-avoiding walk was initially introduced by chemist Paul Flory in the 1940s to model the behavior of polymers. Flory recognized that the physical volume of polymer chains prevents them from overlapping, a principle known as the 'excluded volume' effect. A SAW provided a tractable mathematical model for this phenomenon, allowing for predictions about polymer configurations, solubility, and viscosity.

This physical motivation spurred interest in SAWs within the mathematics community, transforming them from a specialized tool in polymer physics into a subject of independent mathematical study. The transition highlights how practical problems in science can drive the development of abstract mathematical concepts, which then find broader applications across various fields.

Self-avoiding walks are fundamental objects in statistical mechanics, serving as idealized models for a wide range of physical systems. They are crucial for understanding the statistical properties of chain-like molecules, including proteins, DNA, and synthetic polymers. The excluded volume constraint is a key feature in modeling these systems, influencing their conformational entropy and phase behavior.

Furthermore, SAWs are believed to exhibit universal behavior, meaning their properties (like scaling exponents) are independent of the specific lattice details and depend only on the dimension of the space. This universality allows insights gained from studying SAWs on simple lattices to be applied to more complex, real-world systems. They also play a role in percolation theory and the study of critical phenomena, where the behavior of systems near a phase transition is investigated.

A remarkable property of self-avoiding walks is their fractal nature. As a SAW grows, its spatial extent increases with its length, N, according to a power law: <R_N^2> ~ N^(2ν), where <R_N^2> is the mean squared end-to-end distance and ν is a critical exponent. This exponent is related to the fractal dimension of the SAW.

For dimensions d=2 and d=3, the fractal dimensions are approximately 4/3 and 5/3, respectively. The dimension d=4 is known as the upper critical dimension, above which the excluded volume effect becomes negligible, and the SAW behaves like a simple random walk. The scaling behavior of SAWs is a major area of research, with exponents like ν being precisely determined through simulations and theoretical arguments, providing deep insights into the geometry of random structures.

Due to the difficulty in obtaining exact analytical results, computational methods are indispensable for studying SAWs. The pivot algorithm is a widely used Markov chain Monte Carlo technique that generates a new SAW from an existing one by randomly selecting a pivot point and applying a symmetry transformation (rotation or reflection) to the portion of the walk after the pivot. This ensures that the generated walks are sampled uniformly from the ensemble of all possible SAWs of a given length.

Other advanced concepts include the study of SAWs on different types of lattices, the relationship between SAWs and other stochastic processes, and the investigation of SAWs in the presence of external fields or interactions. Calculating the exact number of SAWs for a given length remains a computationally intensive problem with no known general formula.