Hund's Metals

In the quantum realm of solid materials, the collective behavior of electrons can lead to astonishingly complex phenomena. While simple metals are often understood as a free-flowing "sea" of electrons, strong interactions between them can cause a traffic jam, turning the metal into a Mott insulator. This is driven by the brute-force repulsion of electrons on the same atomic site. However, this leaves a gap in our understanding of a vast class of materials that are strongly correlated but remain metallic. These so-called "bad metals" defy simple pictures and hint at a more subtle organizing principle at play.

This article explores that principle, showing how a fundamental rule of atomic physics, Hund's coupling, gives birth to a bizarre and fascinating state of matter known as a Hund's metal. Across two chapters, you will discover a world governed by a strange form of electronic etiquette. The "Principles and Mechanisms" chapter will delve into how Hund's coupling creates incredibly sluggish electrons, leading to unique temperature-dependent behaviors and exotic orbital-selective phases. Following this, the "Applications and Interdisciplinary Connections" chapter will reveal how this seemingly abstract concept is the master key to understanding real-world phenomena, from the architecture of magnetism to the enduring mystery of high-temperature superconductivity.

Imagine the electrons in a simple metal, like copper. They form a sort of "electron sea," a bustling crowd where individuals move almost freely, jostling past each other. This freedom of movement is what allows copper to conduct electricity so well. Now, what happens if we start to crank up the interactions between these electrons? A standard approach starts with the most basic repulsion: two electrons, being of the same negative charge, loathe being in the same place at the same time. This is captured by a parameter called $U$, the on-site Coulomb repulsion. If $U$ is made enormously large compared to the electrons' kinetic energy (their desire to move), you can cause a colossal traffic jam. The electrons get stuck, one per atom, unable to move. The metal becomes an insulator. This state is known as a Mott insulator. It's a simple, brutal kind of correlation: a state of perfect gridlock.

But nature, as it turns out, is far more subtle and creative. There's another "social rule" for electrons, one that is less about brute force repulsion and more about a strange kind of etiquette. This rule is a cornerstone of atomic physics, and when it plays out on the grand stage of a crystalline solid, it gives birth to a bizarre and fascinating state of matter: the ​​Hund's metal​​.

Let's step back from the crystal for a moment and look at a single atom, say, an iron atom with multiple available "rooms" for its outer electrons—the famous $d$-orbitals. Imagine two electrons as unfriendly roommates. Their mutual repulsion ($U$) means they would vastly prefer to occupy different rooms. This much is obvious. But there's more. Hund's rules, discovered by Friedrich Hund in the 1920s, tell us about the preferred arrangement for electrons in these different rooms.

The most important of these, ​​Hund's first rule​​, states that the electrons will arrange their intrinsic magnetic moments, their "spins," to be parallel to each other. It's as if the roommates, despite disliking each other, find a strange comfort in sharing the same "political opinion." This parallel alignment gives the system an energy discount, an amount we call the Hund's coupling energy, $J_H$. This is a purely local, quantum mechanical effect. It's not about forming a large magnet by aligning with spins on distant atoms; it's an intimate piece of atomic etiquette that says, "If we must live on the same atom, let's at least point our spins in the same direction."

This simple rule has earth-shattering consequences when these atoms come together to form a metal.

In a metal, electrons are constantly trying to hop from atom to atom. This hopping is what defines the metallic state. But now, this hopping must contend with the local Hund's rule etiquette.

Imagine an electron trying to hop onto an atom that already hosts a few electrons, all with their spins happily aligned in a high-spin state thanks to $J_H$. The hopping electron faces a dilemma. If it tries to squeeze into an already occupied orbital, it pays a huge energy penalty $U$. If it hops into an empty orbital but with its spin pointing the "wrong" way (anti-parallel to the others), it disrupts the cozy high-spin arrangement, which is also energetically unfavorable. An electron hopping on needs to align its spin, but that requires a "spin flip," which is not always an easy process.

This creates a peculiar kind of traffic. It's not the complete gridlock of a Mott insulator, where the road is simply blocked by a giant $U$. It's more like driving in a city with incredibly complex and strict social rules. You can move, but every move is ponderous, requiring careful negotiation with the local environment. This is the heart of a Hund's metal: a state that remains metallic, but where the electrons are rendered incredibly sluggish. They behave as if they are much, much heavier than a free electron. We formalize this by saying they have a large ​​effective mass​​ ($m^*$) or, equivalently, a very small ​​quasiparticle weight​​ ($Z=m/m^*$). A quasiparticle is our name for an electron dressed by its interactions with the crowd; a small $Z$ means the "bare" electron is almost lost inside a heavy coat of many-body correlations.

In fact, one can show in a simplified model that the energy barrier an electron must overcome to create a charge fluctuation is directly proportional to the Hund's coupling, $\Delta E_{corr} = 3J_H$. This barrier, in turn, directly contributes to the enhanced effective mass. A larger $J_H$ literally makes the electrons heavier.

Most remarkably, this strong correlation and sluggishness can emerge even when the main repulsion $U$ is only moderate—not nearly strong enough to cause a Mott transition on its own. Hund's coupling provides a completely separate path to a strongly correlated state. This isn't just a precursor to a Mott insulator; it's a distinct state of matter. In many situations, particularly when the orbitals are not exactly half-filled with electrons, increasing $J_H$ actually makes the system more stable against a Mott transition by pushing the critical value of $U$ needed for localization to higher energies. The mechanism for strong correlation is not proximity to a charge-frozen Mott state, but what has been dubbed a ​​spin-freezing​​ mechanism: the kinetic energy of electrons is blocked by the presence of large, robust, and slowly fluctuating local spins stabilized by $J_H$.

This peculiar electronic state has a dramatic consequence for how it behaves as we change the temperature. At the very lowest temperatures, near absolute zero, the sluggish electrons can finally get their act together. They manage to coordinate their slow, heavy dance across the entire crystal, forming what physicists call a ​​coherent​​ Landau Fermi liquid. It's like a deep, slow-moving river. In this state, the electrical resistivity follows a characteristic $\rho(T) = \rho_0 + A T^2$ law, the hallmark of a clean metal where scattering between quasiparticles is the dominant process at low temperatures.

However, this delicate coherence is exceedingly fragile in a Hund's metal. The very same Hund's coupling that makes the quasiparticles heavy also makes them easy to disrupt. The energy scale that governs this coherence, known as the ​​coherence temperature​​ ($T_\text{coh}$), becomes drastically suppressed. It takes very little thermal energy to shatter the ordered dance of the heavy electrons.

What happens when you heat the system above $T_\text{coh}$? Chaos. The collective, coherent flow breaks down. The system enters an ​​incoherent​​ regime. The electrons lose their quasiparticle identity and behave as a gas of short-lived, individual particles scattering violently and frequently off the local spin configurations. The river turns into a turbulent, churning mess.

This coherence-incoherence crossover is not just a theoretical idea; it leaves dramatic fingerprints on measurable properties:

• ​​Electrical Resistivity​​: As the temperature rises through $T_\text{coh}$, the clean $T^2$ dependence of the resistivity gives way to a much weaker dependence, often nearly linear in $T$, i.e., $\rho(T) \propto T$. The resistivity continues to rise, often reaching values that would be absurdly high for a good metal. This is called ​​bad-metal behavior​​, where the electron's mean free path becomes as short as the distance between atoms, and the very concept of a particle-like electron starts to break down.
• ​​Photoemission Spectroscopy (ARPES)​​: This powerful technique is like taking a snapshot of the electrons' energy and momentum. Below $T_\text{coh}$, ARPES reveals sharp, well-defined peaks in the spectrum, the signature of long-lived quasiparticles. As the temperature rises above $T_\text{coh}$, these sharp peaks melt away. They broaden dramatically and lose their intensity, telling us that the quasiparticles are gone. The spectral weight lost from the coherent peak doesn't vanish; it gets transferred to form broad, incoherent "Hubbard bands" at higher energies, a direct signature of the strong local interactions $U$ and $J_H$.

So far, we have been playing a game with identical orbitals. But in real materials, like the fascinating iron-based superconductors, the different $d$-orbitals are not born equal. Due to their different shapes and orientations, they interact differently with their neighbors. Some may have strong overlap, leading to a large kinetic energy or ​​bandwidth​​ ($W$), which promotes itinerancy. Others may have poor overlap, leading to a narrow bandwidth, making them inherently more susceptible to localization.

Now, let's turn on Hund's coupling in such an unequal system. Suppose we have a wide-band orbital and a narrow-band one, and we are in a regime where the Coulomb repulsion $U$ is stronger than the narrow bandwidth but weaker than the wide bandwidth ($W_\text{wide} > U > W_\text{narrow}$).

Hund's coupling acts as a great divider. By strongly enforcing local spin alignment, it effectively suppresses charge fluctuations and hopping between the different orbitals. It tells the electrons in the wide-band orbital and the narrow-band orbital, "You solve your own problems!" The electrons in the wide-band orbital have enough kinetic energy to overcome the repulsion and remain metallic. But the electrons in the narrow-band orbital, now isolated and with little kinetic energy to fight back, succumb to the repulsion. They get stuck. They form a Mott insulating state.

This leads to the astonishing possibility of an ​​Orbital-Selective Mott Phase (OSMP)​​: a state of matter where, on the very same atom, some electrons are localized insulators forming magnetic moments, while others are itinerant, conducting metals. This is not science fiction; it is a leading concept for understanding the physics of many complex materials. This orbital differentiation also means different orbitals can have different coherence temperatures. As you heat up the material, the more correlated, narrow-band orbitals can become incoherent while the others remain coherent, leading to complex, multi-stage crossovers in the material's properties. This selectivity even extends to subtle properties like the ​​orbital angular momentum​​, which is typically "quenched" (averaged to zero) in a crystal. In an OSMP, the itinerant orbitals have their momentum strongly quenched as expected, but the localized orbitals can behave more like isolated atoms and retain a piece of their orbital character.

Hund's coupling, a simple rule governing electron etiquette on a single atom, thus blossoms within a crystal into a universe of complex behavior. It generates a new class of "bad" metals, distinct from simple band metals and Mott insulators, and paves the way for exotic phases where electrons on the same atom lead dizzyingly different lives. Understanding this principle is a crucial step on the path to unraveling the mysteries of some of the most exciting materials of our time, including the high-temperature superconductors.

Now that we have grappled with the peculiar inner workings of a Hund’s metal, we might be tempted to ask, "So what?" Is this just a curious corner of quantum theory, a playground for physicists, or does it have something to say about the world we can touch, measure, and use? The answer, it turns in out, is that this seemingly esoteric concept is a master key, unlocking the secrets of some of the most technologically important and scientifically perplexing materials of our time. The journey from the abstract rule of Friedrich Hund to the frontiers of materials science is a breathtaking illustration of how the simplest principles can orchestrate the grand symphony of the quantum world.

Long before the term "Hund's metal" was coined, the influence of Hund's coupling was shaping our understanding of a most fundamental property of matter: magnetism. The magnetism of a material is nothing more than the collective behavior of countless microscopic electron spins. Hund's rule, by dictating the preferred spin arrangement within a single atom, provides the blueprint for how these spins will ultimately cooperate—or compete—across an entire crystal.

A beautiful and direct example is found in a family of materials called manganites. In certain manganites, some manganese ions have a stable, localized "core" of three electrons with their spins locked together, pointing in the same direction, accompanied by a fourth, itinerant electron that is free to roam the crystal. Now, imagine this itinerant electron tries to hop to a neighboring manganese ion. Hund's rule on that destination ion acts like a strict bouncer at a club: it imposes a massive energy penalty unless the incoming electron's spin aligns with the destination's core spin. Hopping is easy if the neighbor’s core spin is already parallel, but nearly impossible if it's antiparallel. Since delocalizing and hopping around is what electrons do to lower their kinetic energy, the entire system finds it is most energetically favorable for all the core spins to align ferromagnetically. This allows the itinerant electrons to move freely, like travelers on a superhighway with no roadblocks. This elegant mechanism, known as ​​double exchange​​, shows Hund's rule in its most direct, powerful role: forging macroscopic ferromagnetism from a simple, on-site energy preference.

The story becomes more subtle, and perhaps more interesting, when there are no itinerant electrons to carry the message. In many insulating materials, like common metal oxides, the magnetic ions are too far apart to interact directly. How do they communicate their spin orientation? The answer is ​​superexchange​​, an indirect conversation mediated by the non-magnetic atoms (like oxygen) that sit between them. Here, the outcome depends on a delicate dance of geometry, quantum tunneling, and the Pauli exclusion principle.

Consider two magnetic ions linked by an oxygen atom in a straight line, a $180^\circ$ bond. A virtual process can occur where an oxygen electron momentarily hops to the first metal ion, and another electron hops from the second metal ion to fill the hole on the oxygen. However, the Pauli principle puts a strict condition on this process. If the two metal ions have parallel spins (ferromagnetic), the process is hindered because an electron cannot hop onto a site that already has an electron with the same spin. But if the spins are antiparallel (antiferromagnetic), the path is clear. Thus, this virtual hopping stabilizes the antiferromagnetic state.

Now, let’s bend the bond to $90^\circ$. The direct path is blocked by orbital symmetries. The dominant antiferromagnetic mechanism is quenched. In this quiet moment, a much weaker, higher-order process can take center stage. Imagine two electrons, one from each metal ion, virtually hopping onto the same oxygen atom into two different, orthogonal $p$ orbitals. What is the lowest energy state for these two electrons on the oxygen? Hund's rule! They will prefer to be in a high-spin, parallel state. This means the virtual process is most favorable if the electrons started with parallel spins back on their home metal ions. The result? A weak ferromagnetic coupling emerges. These "Goodenough-Kanamori rules" are a triumph of quantum intuition, showing how Hund's coupling, in concert with crystal geometry, acts as the supreme architect of magnetism in a vast array of materials, from simple rust to the magnetic components in modern electronics.

While Hund's coupling has long been known as a source of magnetism, its role in creating the "Hund's metal" state reveals a far stranger and more nuanced character. This physics is most prominent in multi-orbital systems like the iron-based superconductors and ruthenates. Here, Hund's coupling doesn't simply create a magnet; it fundamentally rewires the electronic properties, creating a state of matter that is neither a simple, well-behaved metal nor a clean-cut insulator.

A key feature of this world is ​​orbital selectivity​​. It turns out that in these materials, not all $d$-orbitals are created equal. Some, due to their shape and orientation, have a wider electronic bandwidth, meaning their electrons are naturally more itinerant. Others, like the $d_{xy}$ orbital in many iron pnictides, have a narrower bandwidth. For these "lazy" electrons, the kinetic energy is smaller, so the ratio of the Coulomb repulsion $U$ to the bandwidth $W$, a key measure of correlation strength, is effectively much larger. Hund's coupling then acts as an amplifier for this effect. It so strongly penalizes charge fluctuations that the electrons in the narrowest bands can be pushed to the brink of localization—their motion freezing into an incoherent soup—while their counterparts in wider bands continue to move about as reasonably well-defined quasiparticles. In the most extreme case, this leads to an orbital-selective Mott phase, a bizarre state where a single atom is simultaneously metallic and insulating depending on which orbital you look at.

How can we be sure such a strange state exists? We need to peer inside the material. One powerful tool is optical spectroscopy. By shining light on a material and measuring how it is absorbed, we can deduce the kinetic energy of its electrons. In a simple metal, this kinetic energy should be close to what standard band theory (like Density Functional Theory, or DFT) predicts. In a Hund's metal, however, experiments reveal that the electron's kinetic energy is drastically reduced, sometimes to as little as a third of what simple theories predict. This missing kinetic energy is a smoking-gun signature of the strong electronic correlations that have brought the electrons to a near standstill.

An even more direct probe is Angle-Resolved Photoemission Spectroscopy (ARPES), which acts like a quantum microscope, kicking electrons out of the material and measuring their energy and momentum. By carefully choosing the polarization of the incoming light, we can selectively "talk" to electrons in different orbitals. When we do this with a Hund's metal, we see the orbital selectivity in plain sight. Electrons from the more coherent orbitals emerge in sharp, well-defined bands. But when we tune our probe to the more correlated orbital (e.g., $d_{xy}$), the sharp band dissolves into a broad, incoherent haze. This haze contains the fingerprints of the atom-like excitations, called Hubbard bands, which are split by an energy scale directly proportional to the Hund's coupling $J_H$.

This strange behavior isn't static; it evolves dramatically with temperature. At high temperatures, a Hund's metal is a "bad metal," full of incoherent, localized magnetic moments. As it cools, a coherence scale $T_\text{coh}$ is reached, below which the electrons begin to condense into a proper, albeit strongly renormalized, Fermi liquid. We can model this as a competition between the localizing forces of spin-orbit coupling and crystal fields, which want to establish a robust atomic moment, and the delocalizing force of itinerancy, which grows upon cooling. This crossover from localized to itinerant behavior is a defining feature of the Hund's metal state.

One of the most delicate consequences of this crossover involves a subtle quantity: the orbital angular momentum, $\langle \mathbf{L} \rangle$. In most solids, the electric field of the crystal lattice "quenches" orbital motion, forcing $\langle \mathbf{L} \rangle$ to nearly zero. However, a weak relativistic effect called spin-orbit coupling (SOC) can partially "unquench" it. In a Hund's metal, the electronic motion is sluggish; the effective kinetic energy scale, proportional to $ZW$, is severely reduced. This makes the SOC, though small, relatively more powerful. It can more effectively mix the orbital states and partially revive the orbital motion. This beautiful interplay, where strong correlations unexpectedly enhance a relativistic effect, can be tracked experimentally using the circular dichroism in ARPES, a measure of how the material responds differently to left- and right-circularly polarized light.

Perhaps the most exciting and profound connection of Hund's metal physics is to the mystery of high-temperature superconductivity in the iron-based materials. These materials are a paradox. They are "bad metals," strongly correlated, and sit right next to magnetically-ordered phases—all characteristics that, according to conventional wisdom, are terrible for superconductivity. And yet, they are fantastic superconductors.

The theory of Hund's metals provides a stunning resolution to this paradox. Superconductivity requires a "glue" to bind electrons into Cooper pairs. In these materials, the glue is widely believed to be spin fluctuations—the very same magnetic tendencies that Hund's coupling so potently enhances. This is the "good" side of Hund's coupling: it strengthens the pairing interaction.

However, there is a catch. This same Hund's coupling is also responsible for the "bad metal" behavior. It makes the electrons heavy and incoherent, suppressing their quasiparticle weight $Z$. These degraded quasiparticles are the very objects that need to be glued together. This is the "bad" side of Hund's coupling: it weakens the particles that need to pair up.

Superconductivity in a Hund's metal is therefore the result of a delicate compromise, a "just right" balancing act. The pairing strength, $\lambda$, is a product of these two competing factors: the strength of the glue (enhanced by $J_H$) and the coherence of the electrons (suppressed by $J_H$). The fact that superconductivity emerges at all, and at such high temperatures, is a testament to the system finding a precarious sweet spot in the midst of this furious battle.

From a simple atomic rule, we have traveled to the heart of collective magnetism, probed the strange quantum world of orbital-selective phases, and finally arrived at the doorstep of one of the greatest unsolved problems in physics. The physics of Hund's metals is not a niche subfield; it is a central organizing principle that reveals the inherent beauty and unity of the quantum world, showing how the simplest ingredients, when mixed in the crucible of a crystal, can give rise to the most complex, surprising, and ultimately useful emergent phenomena. The journey of discovery is far from over.