In classical condensed matter physics, we usually categorize electrons into one of two camps. Either they are itinerant, moving freely through a crystal lattice like a gas to conduct electricity (the Stoner model), or they are localized, tethered to specific atomic sites where they act as stationary magnetic moments (the Heisenberg model).
However, a growing body of research into iron-based superconductors, Hund’s metals, and moiré materials suggests that this binary is a false choice. Instead, many of the most exotic states of matter—including high-temperature superconductivity—emerge from a "dual nature" where electrons behave as both mobile charge carriers and localized spins simultaneously.
The Failure of Pure Models
The tension between these two descriptions is most visible in iron-based superconductors (FeSCs). Early in the study of iron pnictides, G. M. Zhang, Tao Xiang, and colleagues noted a puzzling "universal" linear-temperature dependence of magnetic susceptibility in Universal linear-temperature dependence of static magnetic susceptibility in iron pnictides.
Standard metallic theories couldn't explain why the magnetism increased so steadily with heat. If the electrons were purely itinerant, the magnetism should be relatively temperature-independent (Pauli paramagnetism). If they were purely localized, the susceptibility should follow a Curie-Weiss law, dropping as temperature rises. The linear increase suggested a system caught between two worlds, where itinerant electrons and local moments are in a constant, temperature-sensitive feedback loop.
The Two-Fluid Framework
To resolve this, researchers like Yi-Zhuang You and Zheng-Yu Weng proposed a "Two-Fluid" or hybrid model. In Coexisting Itinerant and Localized Electrons in Iron-Based Superconductors, they argue that the $d$-orbitals of iron atoms split the difference: some orbitals provide the "itinerant fluid" that conducts electricity, while others host the "localized fluid" of magnetic moments.
The glue holding these two fluids together is Hund’s rule coupling. This intra-atomic force encourages the spins of the mobile electrons to align with the spins of the stationary ones. This coupling creates a "schizophrenic" electronic environment:
- Magnetic Order: The local moments want to order themselves (often in antiferromagnetic patterns).
- Conductivity: The itinerant electrons want to move, but their path is influenced by the magnetic landscape created by the local moments.
- Superconductivity: When the magnetic order is suppressed (via doping or pressure), the fluctuations of these local moments can act as the "pairing glue" that allows itinerant electrons to form Cooper pairs.
Intuition from the Pomeranchuk Effect
Perhaps the most intuitive evidence for this duality comes from "Magic Angle" Twisted Bilayer Graphene (MATBG). In Entropic evidence for a Pomeranchuk effect in magic-angle graphene, Asaf Rozen, Shahal Ilani, and their team observed an electronic version of the Pomeranchuk effect—a phenomenon originally found in Helium-3 where a liquid freezes into a solid upon heating.
In MATBG, heating the material drives it from a metallic state into a magnetically ordered state. This happens because the "localized" moments in the system have a high degree of spin entropy (they can point in many directions). By moving into a state where these moments are more distinct, the system actually increases its total entropy, which is thermodynamically favorable at higher temperatures. This "frozen" magnetic state only makes sense if the electrons have a dual character: they are mobile enough to sense the temperature but localized enough to carry independent spin degrees of freedom.
Beyond Iron: A Universal Duality
This dual nature isn't limited to iron. It appears across a wide spectrum of materials:
- Hund's Metals: In materials like MnSi, Yuan Fang and others found that some electronic bands split due to magnetic order while others emerge through the screening of local moments, a behavior reminiscent of the Kondo effect in heavy-fermion systems (Quasiparticle characteristics of the weakly ferromagnetic Hund's metal MnSi).
- Van der Waals Magnets: In $Fe_3GeTe_2$, Han Wu and Shan Wu showed that record-high Curie temperatures are driven by localized moments rather than the movement of conducting electrons, challenging the standard Stoner model for metallic magnets (Spectral Evidence for Local-Moment Ferromagnetism in van der Waals Metals Fe3GaTe2 and Fe3GeTe2).
- Transition Metal Oxides: Even in $SrRuO_3$, B. I. Min demonstrated that "localized magnetic ghosts" persist even when the material is no longer globally magnetic (Nature of itinerant ferromagnetism of SrRuO3: A DFT+DMFT study).
Why It Works
The reason this research direction is so powerful is that it moves away from "either/or" physics. By treating the itinerant and localized aspects as coupled components of a single system, researchers can explain why a material can be a good metal and a strong magnet at the same time.
As Jian-Huang She and Eun-Ah Kim argue in their work on FeSe (Quantum Spin Liquid Intertwining Nematic and Superconducting Order in Fese), materials at the boundary of itinerancy require a unified treatment. When we stop trying to force electrons into a single box, we gain the ability to predict and control the complex interactions—like spin-charge-lattice coupling—that define the next generation of quantum materials.