Many-body electronic structure, self-doped double-exchange, and Hund metallicity in 1T-CrTe2 bulk and monolayer

Using DFT+DMFT, this study identifies bulk and monolayer 1T-CrTe2 as a self-doped double-exchange Hund metal characterized by coexisting itinerant and localized Cr-d orbitals, revealing that structural deformation rather than dimensionality reduction is the primary factor limiting its Curie temperature.

The Gist

Imagine a tiny, magical city built from layers of atoms. This city is called 1T-CrTe2. It's made of Chromium (Cr) and Tellurium (Te) atoms stacked like pancakes. What makes this city special is that it's a magnet that stays magnetic even at room temperature (above 300 Kelvin), which is rare for materials this thin. Scientists are very excited about it because it could be the key to building faster, smaller, and smarter computers (spintronics).

However, for a long time, scientists were confused about how this city stays so magnetic. Is it a crowd of free-roaming electrons? Or is it a group of stubborn, stationary magnets?

This paper acts like a detective story, using a powerful computer simulation (a mix of two advanced theories called DFT and DMFT) to solve the mystery. Here is the story in simple terms:

1. The Two Types of Residents (The "Dual Nature")

Inside this atomic city, the Chromium atoms have electrons living in their "apartments" (orbitals). The researchers discovered that these electrons aren't all the same. They are split into two distinct groups with very different personalities:

• The Commuters (Itinerant $e_g$ electrons): These electrons are like busy commuters. They zip around the city, moving freely from one building to another. They are the "traffic" of the city.
• The Locals (Localized $t_{2g}$ moments): These electrons are like grumpy old neighbors who never leave their houses. They stay put, but they have a strong "personality" (a magnetic spin) that they keep pointing in a specific direction.

The Analogy: Imagine a dance party. The "Locals" are standing in one spot, holding a sign that says "Spin Up!" The "Commuters" are dancing wildly around them.

2. The Magic Glue: The "Double-Exchange" Dance

How do these two groups work together to make the whole city magnetic?

In the past, scientists thought the "Locals" just talked to each other through the walls (Superexchange) or that the whole crowd moved together (Stoner model). But this paper says: No, they are dancing together!

• The "Commuters" (moving electrons) act as messengers. They hop from one "Local" to another.
• As they hop, they carry a message: "Hey, everyone, keep your signs pointing the same way!"
• This creates a chain reaction where all the "Locals" align their spins in the same direction. This is called Double-Exchange.

The Twist: Usually, you need to add extra people (doping) to a party to make this happen. But in 1T-CrTe2, the city is self-doped. The "Commuters" are already there naturally because of the chemistry of the Tellurium atoms. The city organizes itself without needing outside help.

3. The "Hund Metal" Effect: The Strict Bouncer

The paper also reveals that this city is a "Hund Metal." This is a fancy term for a material where a specific rule (Hund's Rule) acts like a strict bouncer at the club.

• The Rule: The bouncer says, "If you are in the same building, you must all face the same direction to get in!"
• The Result: This rule forces the electrons to form large, strong magnetic groups (high-spin states). It makes the "Locals" even more stubborn and the "Commuters" a bit more chaotic.
• Why it matters: This creates a unique state where the material is a metal (conducts electricity) but acts like a magnet in a very complex, "frustrated" way. It's like a traffic jam where the cars are moving, but the drivers are all arguing about which way to turn, creating a very specific kind of energy.

4. The Thin Layer Problem: Why does it get weaker when peeled?

Scientists can peel this material down to a single layer (a monolayer), like taking one pancake off the stack. Usually, when you make a magnet thinner, it gets weaker.

• The Surprise: The researchers expected the magnetism to drop because the material got thinner (less dimension).
• The Reality: They found that the thickness wasn't the main problem. The real culprit was structural deformation.
• The Analogy: Imagine the "Commuters" (electrons) need a specific bridge to hop between "Locals." When you peel the layer down to a single pancake, the floor bends slightly. The bridge gets twisted and shorter. The commuters can't hop as easily anymore.
• The Result: Because the bridge is broken, the "Locals" can't talk to each other as well, and the magnetic order weakens (the Curie temperature drops).

But here is the cool part: Even though the order got weaker, the individual magnets (the "Locals") actually got stronger! Because the "Commuters" are struggling to move, the "Locals" get even more stubborn and hold their ground tighter. This explains why experiments show that while the whole layer is less magnetic, the spin polarization (how strong the individual magnetic signals are) actually increases.

Summary

This paper tells us that 1T-CrTe2 is a unique material where:

1. Free-moving electrons and stationary magnets live together in a perfect dance (Self-doped Double-Exchange).
2. A strict rule (Hund's coupling) forces them to form strong magnetic groups, making it a "Hund Metal."
3. When you make it super thin, the floor bends, breaking the bridges the electrons use to talk to each other, which lowers the overall magnetism, even though the individual magnets get stronger.

This discovery helps scientists understand how to build better magnetic computers by knowing exactly how to tune the "bridges" (structure) and the "dance" (electron interactions).

Technical Summary

1. Problem Statement

The van der Waals (vdW) ferromagnet 1T-CrTe2 is a promising candidate for next-generation spintronics due to its high Curie temperature ($T_C > 300$ K in bulk) and robust magnetic properties down to the monolayer limit. However, the fundamental mechanism driving its high-$T_C$ ferromagnetism has remained elusive and controversial. Previous theoretical proposals have included:

• Stoner ferromagnetism (itinerant electron model).
• Superexchange (anion-mediated).
• Double Exchange (DE) (typically requiring external doping).
• RKKY interactions.

Furthermore, while the material is metallic, the interplay between electron correlations and magnetism, particularly regarding the nature of Cr-$d$ electrons (localized vs. itinerant) and the role of Hund's coupling ($J_H$), has not been fully characterized. Additionally, experimental observations show a significant reduction in $T_C$ in monolayers, but the origin (dimensionality reduction vs. structural deformation) remains debated.

2. Methodology

The authors employed a sophisticated Many-Body Electronic Structure approach combining Density Functional Theory (DFT) with Dynamical Mean-Field Theory (DMFT) (DFT+DMFT).

• Computational Framework:
  • DFT: Performed using the WIEN2k code (LDA functional) and VASP (for DFT+U checks).
  • DMFT: Implemented via the EDMFTF package with a Continuous-Time Quantum Monte Carlo (CT-QMC) impurity solver.
  • Correlated Subspace: The calculation explicitly included both $t_{2g}$ ($a_{1g}$ and $e'_g$) and $e_g$ Cr-$3d$ orbitals as the correlated subspace, acknowledging significant $d$-$p$ hybridization.
  • Interaction Parameters: Used a fully rotationally invariant Coulomb interaction ($U = 5$ eV, $J_H = 0.85$ eV) to accurately capture spin-flip and pair-hopping processes crucial for Hund's physics.
• Systems Studied:
  • Bulk 1T-CrTe2: Paramagnetic (PM) and Ferromagnetic (FM) phases.
  • Monolayer (1ML): Investigated both the fully relaxed monolayer structure and a "frozen" monolayer (1ML-b) with bulk lattice parameters to decouple structural effects from dimensional effects.

3. Key Contributions

1. Identification of Self-Doped Double Exchange: The paper establishes 1T-CrTe2 as a self-doped double-exchange ferromagnet. Unlike traditional DE systems (e.g., manganites) requiring external doping, the "doping" here is intrinsic, arising from the coexistence of itinerant and localized electrons within the same material.
2. Discovery of Dual Electronic Nature: The study reveals a distinct dual nature of Cr-$d$ electrons:
   • $e_g$ orbitals: Exhibit itinerant, Fermi-liquid-like behavior.
   • $t_{2g}$ orbitals: Exhibit localized, Curie-Weiss-like magnetic moments.
3. Hund Metallicity: The authors demonstrate that 1T-CrTe2 is a Hund metal, characterized by strong Hund's coupling ($J_H$) driving spin-freezing, orbital differentiation, and spin-orbital separation (SOS).
4. Mechanism of $T_C$ Reduction in Monolayers: The study conclusively proves that the reduction in $T_C$ in the monolayer limit is primarily driven by structural deformation (changes in Te height and Cr-Te-Cr bond angles) rather than the reduction in dimensionality itself.

4. Key Results

A. Many-Body Electronic Structure

• Spectral Functions: DFT+DMFT calculations show a narrow quasiparticle peak near the Fermi level ($E_F$) in the PM phase, which disappears in the FM phase, replaced by incoherent spectral weight transfer to higher energies.
• Dual Nature Evidence:
  • Local Spin Susceptibility ($\chi_{loc}^{spin}$): The $t_{2g}$ component shows Curie-Weiss behavior (localized), while the $e_g$ component is closer to Pauli-like (itinerant).
  • First Matsubara Rule: Analysis of the imaginary self-energy ($\text{Im}\Sigma$) confirms that $e_g$ orbitals follow Fermi-liquid scaling ($\text{Im}\Sigma \propto T$), whereas $t_{2g}$ orbitals deviate significantly, indicating localization.
• Self-Doped DE Mechanism: The itinerant $e_g$ electrons mediate the alignment of localized $t_{2g}$ moments via strong Hund's coupling. This is facilitated by the weak electronegativity of Te, which enhances $d$-$p$ hybridization and charge transfer, effectively "self-doping" the system without external impurities.

B. Hund Metal Characteristics

• Spin-Freezing: As $J_H$ increases, the $t_{2g}$ self-energy enters a "spin-frozen" regime with a finite scattering rate, a hallmark of Hund metals.
• Orbital Differentiation & SOS: There is a large separation between the onset temperatures for spin screening ($T_{spin}^{onset} \approx 2000$ K) and orbital screening ($T_{orb}^{onset} > 30000$ K) for the $t_{2g}$ manifold.
• Valence Fluctuations: The system exhibits significant charge fluctuations favoring high-spin multiplets ($t^2_{2g}$ and $t^3_{2g}$), with the $t^3_{2g}$ (half-filled) configuration having a higher probability than the nominal $t^2_{2g}$, distinguishing it from standard Hund metals.

C. Monolayer Limit and Structural Effects

• $T_C$ Reduction: The calculated $T_C$ drops from ~401 K (bulk) to ~214 K (relaxed monolayer), a ~53% reduction.
• Structural vs. Dimensional: When the monolayer structure is "frozen" to bulk parameters (1ML-b), $T_C$ remains high (~375 K, ~94% of bulk). This proves that structural relaxation (specifically the increase in Te height $h_{Te}$ and decrease in Cr-Te-Cr bond angle $\theta$) suppresses the FM coupling by reducing the Cr-$d$ to Te-$p$ hopping parameters ($V_{pd\sigma/\pi}$).
• Enhanced Local Moment: Paradoxically, the local magnetic moment ($M_s$) increases in the monolayer (from 2.99 $\mu_B$ to 3.34 $\mu_B$). This is attributed to enhanced correlations in the reduced dimension, which favor higher-spin configurations (Hund's physics), explaining the experimentally observed increase in spin polarization despite the lower $T_C$.

5. Significance

• Theoretical Framework: This work provides a unified picture connecting Double Exchange, Hund's physics, and Orbital-Selective Mott Physics in a single 2D material. It moves beyond simple Stoner or superexchange models to explain high-$T_C$ ferromagnetism in correlated metals.
• Design Principles for Spintronics: By identifying structural deformation as the key lever for tuning $T_C$, the paper suggests strain engineering as a viable strategy to control magnetic properties in 2D vdW magnets.
• Hund Metals in 2D: It establishes 1T-CrTe2 as a prototypical 2D Hund metal, expanding the known class of these materials beyond bulk iron-based superconductors and ruthenates, and offering a platform to study spin-orbital separation and non-Fermi liquid behavior in reduced dimensions.
• Resolution of Controversies: The findings reconcile conflicting experimental reports on thickness-dependent magnetism by distinguishing between the suppression of ordering temperature ($T_C$) and the enhancement of local moments/spin polarization.