Quantum many-body scars

In the vast landscape of quantum mechanics, isolated, interacting systems are expected to inevitably descend into a state of thermal equilibrium, a process known as thermalization. This principle, formalized by the Eigenstate Thermalization Hypothesis (ETH), suggests that nearly all quantum states become indistinguishable from a high-entropy "thermal soup," erasing any memory of their origin. However, recent discoveries have unveiled a fascinating exception to this rule: quantum many-body scars. These are rare, anomalous states that stubbornly resist thermalization, holding onto their initial structure and coherence within a sea of chaos. This article addresses the puzzle posed by these non-thermal states, exploring their fundamental nature and potential utility. First, in "Principles and Mechanisms," we will examine the defining characteristics of scars, such as low entanglement and quantum revivals, and uncover the hidden symmetries that protect them from chaos. Subsequently, "Applications and Interdisciplinary Connections" will explore how these peculiar properties can be harnessed for advancements in quantum control, metrology, and computation, bridging fundamental physics with cutting-edge technology.

Imagine you're at a pool table. You take a powerful break shot, and the perfectly ordered triangle of balls explodes into a frenzy of motion. After a few moments of frantic collisions, the energy you imparted is shared among all the balls, which now move around randomly. This inevitable descent into disorder and uniform energy distribution is a cornerstone of physics, a concept we call ​​thermalization​​. For a long time, we've expected a similar behavior in the quantum world. An isolated system of many interacting quantum particles, like atoms or spins, should act as its own heat bath, with any initial arrangement eventually evolving into a featureless, high-entropy "thermal soup." This very reasonable idea is formalized in a powerful principle known as the ​​Eigenstate Thermalization Hypothesis (ETH)​​. The ETH suggests that, in a complex, chaotic quantum system, nearly every possible stationary state (an "eigenstate") is itself a microscopic version of this thermal soup.

But nature, it turns out, is more subtle and more beautiful than that. In certain systems, physicists discovered something astonishing: embedded within this vast sea of chaotic, thermal states were a few exceptional, anomalous eigenstates that stubbornly refused to thermalize. They were simpler, more ordered, and possessed a "memory" of where they came from. Like an intricate pattern that persists on the surface of a boiling pot, these states were dubbed ​​quantum many-body scars​​. They are a beautiful imperfection in the otherwise uniform fabric of quantum chaos.

So, do these scar states completely shatter the idea of quantum thermalization? The answer is a fascinating "no," and it reveals a crucial subtlety. The Eigenstate Thermalization Hypothesis actually comes in two flavors: strong and weak. The ​​strong ETH​​ proposes that every single eigenstate in a given energy range must look thermal. The very existence of scar states, which clearly look non-thermal, is a direct violation of this stringent rule.

However, the ​​weak ETH​​ makes a more modest claim: it only requires that almost all eigenstates are thermal. Here, "almost all" has a precise mathematical meaning. In a system of $L$ particles, the total number of states grows exponentially, something like $\exp(L)$. The number of scar states, by contrast, grows much more slowly, perhaps as a polynomial like $L^p$. As the system gets larger, the fraction of scar states compared to the total number of states becomes vanishingly small: $\frac{L^p}{\exp(L)} \to 0$. As a result, the scars become a "measure zero" exception in the vast landscape of the system's possibilities. This means that while strong ETH fails, weak ETH can remain perfectly valid.

This has a profound consequence: for a generic initial configuration, the system will still thermalize as expected, because a random state is overwhelmingly likely to be composed of the thermal eigenstates. The scars are hidden, and you won't even know they're there. To see their effects, you need to prepare the system in a very special, non-generic initial state—one that has a significant overlap with the scar states. Scars don't break the universal tendency toward thermalization; they just provide a secret, state-dependent loophole.

If scars are so rare, how do we find them? Like a detective hunting for clues, we look for two telltale signs that distinguish them from their thermal brethren.

First and foremost is their remarkably low ​​entanglement​​. Entanglement is the web of quantum connections between different parts of a system. A typical thermal eigenstate is a chaotic mess of connections; any part of the system is intricately linked to any other part. For a subsystem of size $L/2$, this entanglement grows in proportion to its size—a "volume law" of entanglement. It's like a dense, tangled ball of yarn. Scar states, by contrast, are shockingly simple. Their entanglement is tiny and often constant, independent of the system's size—an "area law" or even less. A famous scar state in the PXP model, which we'll discuss later, is an equal superposition of two "Néel states" (alternating up-down patterns of spins). For a bipartition of the system, this state, despite being at a high energy, has a constant entanglement entropy of exactly $S_A = 1$ (in bits), no matter how large the spin chain becomes. This is like finding a single, elegant thread instead of a tangled mess.

The second, and perhaps most spectacular, signature is the phenomenon of ​​quantum revivals​​. If you prepare the system in a special state that is a superposition of several scar states (for example, the simple $|101010...\rangle$ Néel state), something incredible happens. Instead of dissolving irreversibly into the thermal soup, the system evolves, becomes complex, and then—almost perfectly—returns to its initial state at regular intervals. It's as if you could unscramble an egg. The system's "fidelity," which measures its similarity to the initial state, shows periodic, sharp peaks, indicating persistent revivals. This behavior is completely at odds with thermalization and is a dead giveaway that the dynamics are being governed by a hidden, regular structure.

These bizarre properties are not magic; they are the consequence of a beautiful, hidden order lurking beneath the surface of a seemingly complex system.

One of the most prominent models exhibiting scars is the ​​PXP model​​, which describes a chain of Rydberg atoms where no two adjacent atoms can be excited at the same time—a "Rydberg blockade." The Hamiltonian seems complicated and non-integrable, promising chaos. However, for the subspace of scar states, the dynamics simplify dramatically. The complicated many-body problem can be mapped onto an effective model of a single, large quantum spin (with total spin $j = (M-1)/2$, where $M$ is the number of scar states) precessing in a magnetic field. The Hamiltonian in this subspace effectively becomes $H_{\text{eff}} = \Omega J_x$, where $J_x$ is a spin operator. This textbook problem has eigenvalues that are perfectly, equally spaced, like the rungs of a ladder. It is this regular "tower" of energy levels that orchestrates the perfect revivals, just as a series of harmonically related notes creates a clear musical chord. This underlying ​​algebraic structure​​ is the secret that protects the scar states from the surrounding chaos.

There is an even deeper and more poetic origin for scars, rooted in the connection between the quantum and classical worlds. Many quantum systems have a classical analogue (e.g., a collection of particles obeying Newton's laws). If the classical system is chaotic, most trajectories are erratic. However, there are always some special, ​​unstable periodic orbits​​—trajectories that perfectly repeat themselves, like a planet in a perfect elliptical orbit, but are unstable, meaning the slightest nudge sends them off into chaos. Semiclassical physics predicts that these fleeting classical orbits can leave a permanent imprint, or "scar," on the wavefunctions of the quantum counterpart. The energy spacing of the quantum scar tower is directly proportional to the "Lyapunov exponent" of the classical orbit, a measure of how quickly nearby trajectories diverge from it. In a way, the quantum scars are the ghosts of these lost classical paths, a beautiful echo of order from a forgotten world.

To fully appreciate what scars are, it's crucial to understand what they are not. There is another, more robust way for a quantum system to escape thermalization: ​​Many-Body Localization (MBL)​​. MBL is a phase of matter that can arise in the presence of strong, quenched disorder (a "bumpy" energy landscape). In an MBL system, all eigenstates at all energies are non-thermal and have low, area-law entanglement. The system is a perfect insulator and never thermalizes, regardless of its initial state. It is essentially a quantum traffic jam, frozen in place by the disordered environment.

Quantum scars are fundamentally different. They typically appear in clean, disorder-free systems. The ETH violation is weak and state-dependent: only a tiny fraction of states are scarred, and the vast majority are thermal. MBL is escaping chaos by being stuck in a gridlock; scarring is like finding a secret, perfectly paved side-road that only a few special travelers (initial states) know how to access, while everyone else is caught in the chaotic traffic of the main highway.

This also points to the ultimate fate of scars: they are fragile. Unlike MBL, which is a robust phase of matter, the special algebraic structure behind scarring is delicate. A small, generic perturbation to the system—a little bit of disorder, or an extra interaction—can break the hidden symmetry. This couples the scar state to the vast sea of thermal states surrounding it, causing it to "leak" and eventually decay. The scar's lifetime is determined by how weakly it couples to the thermal bath, a coupling that is often exponentially small in the system size but is not zero. Scars are not an eternal refuge from thermalization, but rather a long-lived, metastable phenomenon. They are a testament to the fact that even in the relentless march toward quantum equilibrium, the universe finds ways to create moments of breathtaking, transient order.

In our previous discussion, we uncovered the strange and beautiful phenomenon of quantum many-body scars. We saw that in the chaotic realm of interacting quantum particles, where we expect any initial pattern to quickly dissolve into a featureless thermal equilibrium, certain special states defiantly refuse. They behave like ghosts in the machine, orchestrating periodic revivals of quantum information that should, by all rights, be lost forever. A natural, and very physicist-like, question arises: "This is all very curious, but what is it good for?"

It turns out that these exceptions to the rule are more than just theoretical curiosities. They are not merely flaws in the tapestry of statistical mechanics, but rather golden threads pointing toward new scientific frontiers and technological capabilities. The very properties that make them anomalous—their persistent coherence, their special structure, their violation of thermalization—make them a powerful resource. In this chapter, we will explore how we can put these rebellious states to work, bridging the gaps between condensed matter physics, quantum computation, and precision measurement.

The most direct application of quantum scars lies in the realm of quantum control. If a system can stubbornly remember its initial state, it means we have a handle on its evolution, a way to preserve coherence in a complex, interacting environment. The hallmark of scarring is the phenomenon of ​​fidelity revivals​​. Imagine preparing a system of interacting atoms in a simple, patterned state, like the antiferromagnetic Néel state $|1010...\rangle$. You let it evolve, and as expected, the system scrambles into a complicated mess. But then, you wait. And after a precise period of time, like a perfect echo, the initial Néel state reappears with remarkable fidelity, as if no time had passed at all.

This isn’t magic; it’s a consequence of the initial state having a special relationship with the underlying scar eigenstates. The Néel state, for instance, isn't itself a scar, but it has a very large overlap, or projection, onto a small family of scar states that conspire to produce these revivals. This provides us with a powerful recipe: by preparing simple, accessible states, we can tap into the protected coherence of the hidden scar subspace.

Of course, this protection is not absolute. The scar states exist within a vast sea of "thermal" states that are always trying to pull the system into equilibrium. This process can be pictured as a "leakage" from the coherent scar subspace into the incoherent thermal bath. In a simplified but insightful picture, we can model the dynamics with just three levels: an initial state, a scar state, and a "thermal gateway" state. The system primarily oscillates between the initial and scar states, but a small coupling allows it to leak into the thermal-state representative, causing the revivals to eventually decay. Understanding and controlling this leakage is a key challenge, one that touches upon the central themes of quantum error correction.

What makes this pursuit so exciting is its universality. The physics of scarring is not confined to a single, esoteric setup. The same constrained PXP model that describes scarring in arrays of optically trapped Rydberg atoms also emerges as an effective description for chains of nitrogen-vacancy (NV) centers in diamond and arrays of superconducting transmon qubits used in quantum computing. This means that insights and techniques developed in one platform can be transferred to others, a beautiful example of the unity of physical laws.

If scars provide a sheltered pocket of coherence, it's natural to ask if we can use that pocket to build something. The answer appears to be a resounding yes, opening doors to new quantum technologies.

One of the most promising avenues is ​​quantum metrology​​, the science of making ultra-precise measurements. The fundamental limit to precision is set by the Quantum Cramér-Rao bound, which relates the measurement uncertainty to a quantity called the Quantum Fisher Information (QFI). A higher QFI means a better potential sensor. Because scar states are highly structured and energetically isolated in their own way, they can be exquisitely sensitive to small perturbations. A tiny change in an external field or an internal parameter $\mu$ can cause a significant change in the scar's wavefunction. By preparing a system in a scar-like state and measuring this change, one can perform a high-precision measurement of the parameter. Even in a simple toy model, a scar-like state can possess a remarkably high QFI, suggesting that these non-thermal states could serve as powerful probes for quantum sensing.

Pushing the technological frontier even further, one can envision using scars to power an ​​atom laser​​. An ordinary laser produces a coherent beam of photons. An atom laser, its matter-wave counterpart, produces a coherent beam of atoms. The coherence of the laser beam is inherited from the coherence of its source, or "gain medium." What if the source was a Bose-Einstein condensate prepared in a superposition of scar states? By continuously outcoupling atoms from this long-lived, coherent many-body state, one could potentially create an ultra-coherent matter-wave beam. The properties of such a beam would be directly imprinted with the quantum correlations of its scar source. The coherence could be tested using techniques like Ramsey interferometry, where the visibility of interference fringes would depend directly on the energy difference and coupling strengths of the scar states involved in the process. While still a futuristic vision, it highlights how a fundamental discovery in many-body physics could translate into a novel quantum device.

Beyond direct applications, the study of quantum scars builds powerful intellectual bridges between seemingly disparate fields of science.

At a fundamental level, scars force us to refine our understanding of quantum statistical mechanics. The cornerstone of this field is the ​​Eigenstate Thermalization Hypothesis (ETH)​​, which posits that individual eigenstates of a chaotic many-body Hamiltonian look thermal. Scars are glaring violators of the ETH. A key signature of this violation lies in their ​​entanglement structure​​. Thermal eigenstates are maximally messy; their entanglement entropy scales with the size of the subsystem (a "volume law"). Scar states, by contrast, are far more orderly. For certain models connected to the physics of Mott insulators, scar states can be constructed that exhibit an entanglement entropy $S_A$ that grows only as the logarithm of the system size, $S_A = \ln L$. This profound lack of entanglement is what allows them to store and revive local information, and it provides a sharp, quantifiable way to distinguish them from the teeming multitude of thermal states.

This distinction is not just a theoretical nicety; it has practical implications for finding these states. The Hilbert space of a many-body system is astronomically vast. How can we find these rare scar states in a haystack of a billion billion thermal ones? ​​Computational physics​​ provides a strategy. One can use numerical algorithms to "deflate" a Hamiltonian, systematically projecting out the uninteresting states to reveal the anomalous ones. A useful criterion for distinguishing states is the Inverse Participation Ratio (IPR), which measures how localized a wavefunction is. Thermal states are spread out and have low IPR, while scarred or localized states are concentrated and have high IPR. By computing the eigenstates and filtering them based on IPR, one can computationally isolate the scar states from the thermal background.

An even more tantalizing connection emerges with ​​quantum computation​​. Can we use a quantum computer to find a scar? The problem can be cleverly framed as a search within the constrained Hilbert space where scars live. If we view the special scar state as a "marked item" in a database of all allowed quantum states, we can adapt algorithms like Grover's search to find it. The search algorithm itself must be modified to respect the system's kinetic constraints (e.g., that no two adjacent atoms can be excited). This approach not only provides a potential quantum advantage for discovering scars but also beautifully illustrates the deep connections between the structure of many-body Hamiltonians and the logic of quantum algorithms.

In the end, the story of quantum many-body scars is a perfect illustration of how science progresses. We start with a rule—that complex systems thermalize. We find an exception. By studying that exception, we don't just find a curiosity; we find a key that unlocks new possibilities for control, new proposals for technology, and deeper connections between the fundamental theories that govern our world. The scars on the otherwise smooth landscape of statistical mechanics are not blemishes, but treasure maps.