Suppression of local magnetic moment formation and paramagnetic exchange interactions in monolayer Fe$_3$GeTe$_2$

Using the DFT+DMFT approach, this study reveals that monolayer Fe$_3$GeTe$_2$ exhibits a partially itinerant magnetic behavior where non-equivalent iron atoms display distinct local moment formation and crucial RKKY-type exchange interactions that stabilize long-range ferromagnetic order, highlighting the necessity of dynamical electron correlation treatments to accurately describe its magnetic properties.

The Gist

Imagine a tiny, flat world made of a special material called Fe3GeTe2 (let's call it "Iron-Ge-Te" for short). This material is a magnet, but it's a very strange one. It's a "metallic magnet," meaning it conducts electricity like a wire and sticks to your fridge like a magnet at the same time.

Scientists have been trying to figure out exactly how this material works, especially when it's stripped down to a single layer (a "monolayer"). This paper is like a detective story where the authors use a super-powerful computer simulation to peek inside the atomic world and solve a mystery about how the magnetism is formed.

Here is the story in simple terms:

1. The Cast of Characters: Three Types of Iron

Inside this flat material, there are three iron atoms for every unit of the structure. Think of them as three siblings living in a house:

• The "Up" and "Down" Siblings (Fe1 & Fe2): These two live on the top floor and the basement. They are very active and emotional.
• The "Middle" Sibling (Fe3): This one lives right in the middle, sandwiched between the others. This sibling is much more laid-back and chill.

2. The Mystery: Are They Localized or Itinerant?

In the world of magnets, there are two main ways atoms behave:

• The "Sticky Note" Style (Localized): The magnetic power is stuck firmly to one atom, like a sticky note on a fridge. It doesn't move.
• The "Swarm of Bees" Style (Itinerant): The magnetic power is like a swarm of bees buzzing around; it flows freely through the material.

Most scientists thought Fe3GeTe2 was a mix of both, but they weren't sure how much of each. The authors of this paper argue that it's mostly the "Swarm of Bees" style, but with a twist: the "Up/Down" siblings are acting like sticky notes, while the "Middle" sibling is definitely a swarm of bees.

3. The Investigation: The Computer Simulation

The authors used a high-tech method called DFT+DMFT.

• DFT is like taking a high-resolution photo of the material's structure.
• DMFT is like adding a time-lapse camera to see how the electrons (the tiny particles carrying the magnetism) dance and interact over time.

They found that the "Up/Down" siblings are so busy interacting with their neighbors that they develop strong, local magnetic moments (like little magnets). However, the "Middle" sibling is so connected to the flowing electrons that it doesn't really form a distinct little magnet at all. It's just part of the flow.

4. The "RKKY" Handshake

Here is the coolest part of the discovery. Even though the "Middle" sibling doesn't have its own strong magnet, it acts like a messenger.

Imagine the "Up" and "Down" siblings are two people who want to hold hands to stay together (this is what keeps the material magnetic). But they are too far apart to reach. The "Middle" sibling acts like a relay runner. It passes the "handshake" signal from one to the other.

In physics, this is called RKKY interaction. It's like the middle atom saying, "Hey, I'm not magnetic myself, but I'll help the other two connect!" Without this middle guy acting as a bridge, the whole material would lose its magnetism.

5. The Temperature Test

The scientists tested what happens when they heat up the material.

• In a normal magnet, if you heat it, the little magnets just get shaky and eventually stop pointing in the same direction.
• In this material, the relationship is more complex. The "Up/Down" siblings start to lose their grip, but the "Middle" sibling keeps the flow going. The math shows that the magnetism doesn't just disappear; it changes shape in a non-linear way (it's not a straight line on a graph).

6. The Verdict: Why It Matters

The authors calculated how "stiff" the magnetic waves are (how hard it is to wiggle the magnet) and at what temperature the material stops being magnetic (the Curie temperature).

• The Result: Their calculations matched real-world experiments almost perfectly.
• The Lesson: Previous theories thought the material was more "stuck" (localized) than it actually is. This paper proves that the material is partially flowing. The magnetism is a team effort where the "chill" middle atom is just as important as the "active" outer atoms because it keeps the team connected.

The Big Picture Analogy

Think of Fe3GeTe2 as a dance floor.

• The Outer Iron atoms are the lead dancers who are very focused and intense (strong local magnets).
• The Middle Iron atom is the DJ. The DJ isn't dancing wildly, but without the DJ playing the beat and connecting the dancers, the whole party falls apart.
• The Electrons are the music flowing through the room.

The paper shows us that you can't understand the party just by looking at the lead dancers; you have to realize that the DJ (the middle atom) and the flow of music (the itinerant electrons) are what actually keep the magnetism alive. This helps scientists design better, smaller, and more efficient electronic devices in the future.

Technical Summary

Here is a detailed technical summary of the paper "Suppression of local magnetic moment formation and paramagnetic exchange interactions in monolayer Fe3GeTe2" by Katanin et al.

1. Problem Statement

The paper addresses the complex magnetic nature of monolayer Fe$_3$GeTe$_2$ (FGT), a prototypical 2D itinerant ferromagnet. While FGT exhibits a relatively high Curie temperature ($T_C \approx 130$ K in monolayers), its magnetic behavior sits in a challenging regime between localized moments and itinerant electrons.

• The Conflict: Previous studies using Density Functional Theory plus Dynamical Mean-Field Theory (DFT+DMFT) often assumed a "Hund metal" behavior with fully formed local moments or used perturbation solvers (like SPTF) that may not be accurate for this specific compound.
• The Gap: There is a lack of understanding regarding how electron correlations affect the formation of local magnetic moments in the paramagnetic phase (above $T_C$) and how these correlations influence exchange interactions when local moments are only partially formed. Specifically, the role of the three distinct iron sites (Fe1, Fe2 above/below the Ge plane, and Fe3 within the Ge plane) requires clarification.
• Methodological Issues: Previous calculations often used unrealistically large Coulomb interaction parameters ($U \approx 5$ eV) and approximations for double-counting corrections that fixed the $d$-electron concentration to $n_d=6$, which may not be justified for FGT.

2. Methodology

The authors employ a systematic DFT+DMFT approach to study the electronic and magnetic properties of monolayer FGT in the paramagnetic phase.

• DFT Setup: Calculations were performed using the PBE functional (VASP) with a vacuum layer to simulate the monolayer. Maximally localized Wannier functions were constructed projecting onto Fe 3$d$ and Te 5$p$ states to generate the tight-binding Hamiltonian.
• DMFT Implementation:
  • Solver: Continuous-time hybridization expansion Quantum Monte Carlo (CT-QMC) via the iQIST package.
  • Parameters: Moderate Coulomb interaction $U = 2 - 4$ eV (screened) and Hund's coupling $J_H = 0.9$ eV. The double-counting correction was applied in the "around mean-field" form.
  • Exchange Interactions: Instead of the magnetic force theorem (which requires a magnetically ordered state), the authors used a formalism based on non-uniform magnetic susceptibility in the paramagnetic phase. Exchange interactions ($J_q$) were derived from the inverse of the local and non-local static susceptibilities:
    $$J_q = \chi^{-1}_{loc} - \chi^{-1}_q$$
  • Spin-Wave Analysis: The effective Heisenberg model was constructed to calculate spin-wave dispersions and stiffness ($D$), accounting for the Mermin-Wagner theorem by including magnetic anisotropy to estimate the physical Curie temperature.

3. Key Contributions

1. Site-Dependent Correlation Effects: The study reveals a strong differentiation between iron sites. Fe atoms located above and below the Ge plane (Fe1, Fe2) exhibit strong correlations and partial local moment formation, while the Fe atom within the Ge plane (Fe3) behaves more itinerantly with no pronounced local moment.
2. Non-Linear Susceptibility: The authors demonstrate that the inverse local and uniform magnetic susceptibilities show non-linear temperature dependencies over a broad range. This indicates that FGT is far from the localized limit and that local moments are only partially formed, challenging the standard Curie-Weiss law applicability in this regime.
3. RKKY-like Stabilization: The paper identifies that the long-range ferromagnetic order is stabilized by RKKY-type exchange interactions between the strongly correlated Fe1/Fe2 sites and the itinerant Fe3 sites. This mechanism compensates for the antiferromagnetic interactions between Fe1/Fe2 neighbors.
4. Revised Interaction Parameters: The study suggests that previous estimates of $U$ were too high. Using $U = 3-4$ eV yields results consistent with experimental data, whereas higher values distort the electronic structure.

4. Key Results

Electronic Properties

• Occupation: Correlations shift the Fe $d$-band peak above the Fermi level, reducing the occupation number of Fe1/Fe2 to $\approx 6.5-6.9$ (closer to half-filling) and increasing Te occupation.
• Self-Energy: Fe1/Fe2 atoms show a non-quasiparticle self-energy (indicative of local moment formation), while Fe3 atoms retain a quasiparticle form.

Magnetic Susceptibilities

• Local Moments: At $U=3-4$ eV, Fe1/Fe2 carry a substantial local moment ($\mu \gtrsim 4.5 \mu_B$), while Fe3 has a negligible moment.
• Temperature Dependence: The inverse susceptibilities are non-linear, confirming the "partially itinerant" nature of the magnetism.
• Extrapolation: Extrapolating the calculated moments to low temperatures ($T \sim T_C^{exp}$) yields a total magnetic moment $\mu^2 \approx 21 \mu_B^2$/Fe, which aligns well with experimental values derived from the Curie-Weiss law.

Exchange Interactions and Spin Dynamics

• Ferromagnetic Dominance: Most exchange interactions are ferromagnetic. The only significant antiferromagnetic interaction is between nearest-neighbor Fe1/Fe2 sites ($J_{11}, J_{22}$).
• Compensation: The antiferromagnetic $J_{11}$ is compensated by strong ferromagnetic interactions involving Fe3 ($J_{13}, J_{23}$), which act as RKKY mediators.
• Spin-Wave Stiffness ($D$):
  • Calculated $D \approx 235$ meV$\cdot\mathring{A}^2$ (using $U=4$ eV).
  • When rescaling magnetic moments to match the experimental saturation moment ($\mu_s \approx 1.6 \mu_B$/Fe), $D$ reduces to 190 meV$\cdot\mathring{A}^2$, in excellent agreement with inelastic neutron scattering data ($\approx 200$ meV$\cdot\mathring{A}^2$).
• Curie Temperature ($T_C$):
  • Using the calculated stiffness and anisotropy, the estimated $T_C$ is 140–170 K.
  • This is in good agreement with the experimental monolayer $T_C$ ($\approx 130$ K) and significantly lower than bulk values ($\approx 220$ K), correctly capturing the 2D nature of the system.

5. Significance

This work provides a crucial theoretical framework for understanding itinerant ferromagnetism in 2D materials where local moments are not fully formed.

• Validation of Method: It demonstrates that the non-uniform susceptibility approach in the paramagnetic phase is superior to magnetic force methods for systems with partially formed moments.
• Physical Insight: It resolves the "site differentiation" puzzle in FGT, showing that the Ge-plane iron atoms are not merely passive but play a critical itinerant role in mediating ferromagnetism via RKKY interactions.
• Quantitative Accuracy: By correcting the Coulomb interaction parameters and accounting for the itinerant suppression of moments, the study achieves quantitative agreement with experimental spin stiffness and Curie temperatures, offering a reliable model for designing future 2D magnetic devices.