Finite-state automata for exact matrix product operators of tight-binding Hamiltonians: fractals, quasicrystals, trees and hyperbolic lattices

Imagine trying to understand the behavior of electrons moving through a material. In the simplest models, scientists picture these electrons hopping from one atom to the next on a grid, much like a traveler moving between cities on a map. This "tight-binding" picture is a fundamental tool for understanding how electricity flows, how materials conduct heat, and why some substances become magnetic. For decades, this model worked best on regular, repeating patterns, like the perfect squares of a tiled floor. But nature is rarely so orderly. Electrons also travel through materials with strange, repeating-but-never-exactly-the-same patterns, or through shapes that repeat themselves at every scale, like a coastline or a snowflake. These complex geometries—fractals, quasicrystals, and tree-like structures—create exotic electronic behaviors that regular grids cannot explain. However, studying them has been a nightmare for computers. As the number of atoms in a simulation grows, the memory required to map out every possible path for an electron explodes, quickly overwhelming even the most powerful supercomputers.

A team of researchers has now found a way to tame this complexity. They have developed a new method to describe these complicated electron maps using a clever trick borrowed from the way computers process language. Instead of trying to store a massive, detailed list of every single connection between atoms, they built a compact, rule-based system that captures the entire structure of the material in a tiny amount of digital space. This system allows them to simulate materials with billions of atoms, a scale that was previously impossible to reach with exact calculations. By combining this new tool with a standard technique for analyzing energy levels, they were able to calculate the electronic properties of these strange lattices with high precision, revealing how electrons behave in environments that range from the self-similar patterns of fractals to the curved, infinite expanses of hyperbolic space.

The core of this breakthrough lies in how the researchers organized the information about the material's shape. In a standard computer simulation, the memory needed to describe a system usually grows in direct proportion to the number of atoms. If you double the size of the material, you roughly double the memory needed. But for the complex shapes studied here, the memory needed would typically grow so fast that a simulation of a large system would require more storage than exists on Earth. The researchers realized that these specific shapes—fractals, quasicrystals, and trees—are built using recursive rules. This means the whole structure is made of smaller copies of itself, stacked together in a predictable way. They encoded these rules into a simple machine, similar to a traffic light that changes its state based on the cars passing it, but in this case, the machine tracks the position of an electron as it moves through the layers of the material's construction.

This machine, which the authors call a finite-state automaton, acts as a blueprint. It does not list every single atom. Instead, it holds a small set of instructions that tell the computer how to generate the connections between atoms as it moves from one level of the structure to the next. Because the rules for building these shapes are finite and repeatable, the machine only needs a fixed, small number of internal states to describe the entire system, no matter how large it becomes. The researchers translated this machine into a mathematical structure known as a matrix product operator. In plain terms, this is a way of writing down the rules for the electron's movement as a chain of small, connected blocks. The size of these blocks depends only on the complexity of the shape's rules, not on the total number of atoms. This means that even if the simulation contains billions of sites, the computer only needs to store a manageable amount of data to describe the Hamiltonian, which is the mathematical object that dictates how the electrons move and interact.

To test their method, the team applied it to a wide variety of challenging geometries. They started with the familiar square and triangular grids to ensure their method worked correctly, then moved on to the Sierpiński carpet, a famous fractal shape where a square is repeatedly divided into nine smaller squares with the center one removed. They also simulated the Fibonacci chain, a one-dimensional line of atoms arranged in a pattern that never repeats but follows a strict rule, and its two-dimensional extension. Finally, they tackled Cayley trees, which are branching structures that grow outward from a central point, and hyperbolic lattices, which are grids that curve inward like a saddle, allowing for an exponential growth in the number of connections as you move away from the center. In every case, the method produced an exact description of the system's rules without ever needing to write down the full list of connections.

Once the rules were encoded, the researchers used a technique called the kernel polynomial method to calculate the spectral properties of these materials. This approach allows them to determine the energy levels available to electrons and how those electrons are distributed across the material, without having to solve the entire system at once. For the Sierpiński carpet, they simulated a system with over 200,000 sites and found that the electron waves remained spread out across the structure, showing no signs of getting stuck in one spot. They also looked at a Fibonacci chain with nearly 200,000 sites and observed the characteristic, intricate patterns of energy that are unique to quasicrystals, where the electron density varies in a complex, non-repeating way. In the case of the hyperbolic lattice, which they modeled as a 3-Cayley tree at generation n = 10, they were able to map out how disorder in the material affects the flow of electrons, showing that the symmetry of the structure is preserved even when the environment is chaotic.

The significance of this work is that it provides a unified way to study a broad class of materials that were previously too difficult to simulate exactly. Before this, researchers often had to rely on approximations or could only study very small versions of these systems, which might miss important large-scale behaviors. By using this new method, they can now explore the electronic properties of these complex lattices at sizes that are exponentially larger than what was possible before. The method is particularly powerful because it is exact; it does not throw away information or make guesses about the system's behavior. It simply finds a more efficient way to write down the rules. This opens the door to exploring how electrons behave in non-Euclidean geometries, which are becoming increasingly relevant as scientists build new types of quantum simulators using light, circuits, and arrays of atoms.

The researchers also noted that while their method works perfectly for the shapes they studied, it relies on the fact that these shapes are built from simple, repeating rules. If a material has a structure that does not follow such a clear pattern, this specific approach might not work as well. However, for the vast family of recursively structured lattices, the method offers a powerful new lens. It allows scientists to ask questions about the electronic properties of materials with billions of atoms, a scale that is relevant for understanding real-world devices and for designing new materials with tailored properties. The ability to simulate these systems exactly means that the predictions made by the computer are as reliable as the laws of physics themselves, providing a solid foundation for future experiments and discoveries in the realm of quantum matter.