🔬 condensed matter

Interplay of ferromagnetism, nematicity and Fermi surface nesting in kagome flat band

Motivated by recent experiments on Fe-doped CoSn, this study employs Hartree-Fock calculations on a kagome model to demonstrate that while on-site interactions favor ferromagnetism, sizable inter-sublattice interactions stabilize a nematic phase as a generic outcome of partially filled kagome flat bands.

Original authors: Yuman He, Wentao Jiang, Siqi Wu, Xuzhe Ying, Berthold Jack, Xi Dai, Hoi Chun Po

Published 2026-03-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Yuman He, Wentao Jiang, Siqi Wu, Xuzhe Ying, Berthold Jack, Xi Dai, Hoi Chun Po

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A Dance on a Triangular Floor

Imagine a dance floor made of a Kagome lattice. This isn't a square grid; it's a pattern of triangles sharing corners, looking like a woven basket or a honeycomb with holes in the middle.

In this paper, scientists are studying what happens when you put electrons (the dancers) on this specific floor, specifically when the "flat band" is partially filled. In physics terms, a "flat band" means the electrons have very little energy to move around; they are stuck in a low-energy state, like dancers who are tired and just standing in one spot.

Recently, experiments on a material called CoSn (doped with Iron) showed something weird: when they added a few "holes" (removed a few dancers), the material didn't just become magnetic or stay normal. It developed a Nematic Phase.

What is Nematicity?
Think of a crowd of people in a square room.

Normal state: Everyone is facing random directions, or the room looks the same no matter which way you turn it (6-fold symmetry).
Nematic state: Suddenly, everyone decides to face North-South. The room still looks like a square, but it looks different if you rotate it 60 degrees vs. 90 degrees. The symmetry is broken. The crowd has "chosen a direction."

The paper asks: Why does this happen? Is it because the dancers want to be magnetic (Ferromagnetism), or because they want to align in a specific pattern (Nematicity)?

The Two Competing Forces

The researchers used a computer model (a "self-consistent Hartree-Fock" calculation, which is just a fancy way of saying "let's simulate the rules of the game") to see who wins the fight between two types of interactions:

1. The "On-Site" Rule (Ferromagnetism)

Imagine a rule that says: "If two dancers are on the exact same spot, they must spin in the same direction."

The Analogy: This is like a strict bouncer at a club. If two people try to stand on the same tiny tile, they must both raise their right hands (spin up) or both raise their left hands (spin down).
The Result: This creates Ferromagnetism. Everyone spins the same way. It's a "magnetic" crowd.

2. The "Neighbor" Rule (Nematicity)

Now imagine a second rule: "If you are standing next to someone, you must coordinate your position with them to avoid crowding."

The Analogy: This is like a game of musical chairs where the dancers realize that if they all cluster in one corner, they get squished. So, they spread out in a specific pattern to make the most space. They might all lean to the left, or all lean to the right.
The Result: This creates Nematicity. The electrons don't necessarily spin the same way, but they rearrange their positions to break the symmetry of the room.

The Discovery: Who Wins?

The paper finds a fascinating competition:

If the "On-Site" rule is too strong: The electrons just all spin the same way (Ferromagnetism).
If the "Neighbor" rule is strong (which it is in these materials): The electrons decide to break the symmetry and align in a specific direction (Nematicity).

The Twist: In the real material (CoSn), the "Neighbor" rule is surprisingly powerful. Even though the electrons are stuck in a "flat band" (which usually makes them want to be magnetic), the fact that they interact with their neighbors across the lattice forces them into a Nematic phase.

The authors show that this Nematic phase is very stable. It survives over a wide range of "doping" (how many dancers are on the floor) and temperature. This explains why the recent experiments saw this weird symmetry-breaking phase so often.

The "Van Hove" Singularity: The Traffic Jam

There is a second concept mentioned: Fermi Surface Nesting.

The Analogy: Imagine the dancers are moving around a track. Usually, the track is curvy. But at a specific point (called the Van Hove Singularity), the track becomes perfectly straight and parallel.
The Effect: If the track is straight, the dancers can easily "nest" or line up perfectly with each other, like cars in a traffic jam. This usually causes a "Charge Density Wave" (a ripple in the crowd).
The Paper's Finding: In this specific Kagome material, the track is almost straight, but slightly curved. Because of this curve, the "traffic jam" (nesting) isn't strong enough to win. The "Neighbor" rule (Nematicity) is the real boss here. The curvature of the track prevents the electrons from forming the ripples seen in other materials, leaving the Nematic phase as the winner.

Summary: The Takeaway

The Setup: Electrons are dancing on a triangular (Kagome) floor.
The Conflict: Do they want to all spin the same way (Magnetism) or rearrange their positions to break the room's symmetry (Nematicity)?
The Winner: In this specific material, the interaction between neighbors is so strong that it forces the electrons to break symmetry and become Nematic, even though they are in a "flat" energy state that usually favors magnetism.
Why it matters: This gives scientists a simple, minimal framework to understand why these weird, correlated phases happen in flat-band materials. It's not magic; it's just the electrons playing a game of "don't crowd the neighbors."

In short: The paper explains that in these special materials, the electrons are so sensitive to their neighbors that they give up on being perfectly magnetic and instead choose to organize themselves into a specific, directional pattern. This explains the strange experimental results seen in Iron-doped Cobalt-Tin.

1. Problem Statement

Recent experiments on Fe-doped CoSn have revealed a series of correlated phases upon hole-doping the kagome flat bands (FBs). Most notably, a nematic phase (breaking six-fold rotational symmetry down to two-fold) prevails over a wide range of doping and temperatures.

Theoretical Gap: While ferromagnetism (FM) is theoretically established for on-site interactions in flat bands, and Van Hove singularity (VHS) driven instabilities explain phenomena in dispersive bands (like in AV $_3$ Sb $_5$ ), the mechanism driving nematicity specifically within partially filled, weakly dispersive kagome flat bands remains unclear.
Key Question: How do on-site repulsion, inter-sublattice interactions, and Fermi surface nesting compete to determine the ground state in this specific regime? Why does nematicity dominate over FM in certain doping windows, and why are translational symmetry-breaking phases (like Charge Density Waves) less prominent here compared to dispersive band systems?

2. Methodology

The authors employ a Self-Consistent Hartree-Fock (SCHF) approach on a kagome lattice model to investigate interaction-driven phases.

Model Hamiltonian:
- Non-interacting part: A tight-binding (TB) model including nearest-neighbor (NN) and next-nearest-neighbor (NNN) hoppings ( $t_1, t_2$ ). The NNN hopping ( $t_2$ ) introduces weak dispersion to the ideally flat band, creating a VHS at the $M$ point.
- Interactions: Density-density interactions up to second-nearest neighbors are included:
  - $U$ : On-site repulsion.
  - $V_1, V_2$ : Inter-sublattice interactions (nearest and next-nearest neighbors).
- Orbital Basis: The model uses locally rotated $d_{x^2-y^2}$ -type orbitals forming the upper flat band, consistent with CoSn-family compounds.
Computational Approach:
- SCHF Calculations: The authors perform mean-field calculations scanning the flat-band filling from empty to full.
- Symmetry Breaking: They use random combinations of basis functions (derived from $D_6$ point group irreducible representations) as perturbations to seed symmetry-breaking orders (FM, nematic, stripe, $2\times2$ ).
- Translation Symmetry: Both translation-invariant (unit cell) and translation-symmetry-breaking ( $2\times2$ supercell) cases are analyzed to check for nesting-driven instabilities.
- Finite Temperature: A pseudo-temperature is introduced to smooth the Fermi surface near the VHS, avoiding numerical issues with degeneracy.

3. Key Contributions and Mechanisms

A. Competition between Ferromagnetism and Nematicity

The core finding is a direct competition between two mechanisms driven by different interaction ranges:

Stoner Ferromagnetism: Driven primarily by on-site interaction ( $U$ ). This favors a uniform spin polarization (FM).
Nematicity: Driven by sizable inter-sublattice interactions ( $V_1, V_2$ ).
- Mechanism: The extended nature of the flat-band orbitals (hexagon states) enhances inter-sublattice Coulomb repulsion. To minimize this energy, the system redistributes charge unevenly among the sublattices (A, B, C), breaking the six-fold rotational symmetry ( $C_6 \to C_2$ ) while maintaining translational symmetry.
- Analytical Insight: A mean-field analysis shows that while $U$ favors FM, the condition for nematic instability is roughly $(4(V_1+V_2) - U)D > 4/3$ . For flat bands with large $D$ (density of states) and significant $V$ , nematicity becomes energetically favorable.

B. Role of Weak Dispersion and Nesting

Weak Nesting: Unlike ideal flat bands or dispersive bands with parallel Fermi surface edges, the weakly dispersive flat band in this model has a heavily curved Fermi surface near the VHS.
Consequence: This curvature significantly suppresses the nesting susceptibility ( $\chi_m$ ) compared to the uniform susceptibility ( $\chi_0$ ). The ratio $\chi_m/\chi_0$ remains finite rather than diverging.
Result: This explains why translational symmetry-breaking phases (like CDW) are less robust and less dominant in Fe-doped CoSn compared to systems driven by dispersive-band VHSs.

4. Key Results

Phase Diagram (Translation Invariant):
- Low Filling / High $U$ : Ferromagnetism dominates.
- Intermediate to High Filling / High $V$ : Nematic phases (both symmetric and nematic-FM) prevail over a wide doping window.
- Nematic-FM: At large interaction strengths, a phase with both ferromagnetic order and nematic symmetry breaking emerges. The magnetic moment in this phase continues to grow with interaction strength, unlike symmetric FM which saturates.
- Peak Nematicity: The strongest nematic order ( $\langle \Delta \hat{C}_6 \rangle \approx 0.52$ ) occurs at filling $\sim 0.8$ with strong interactions, consistent with the experimental observation of nematicity in hole-doped CoSn.
Phase Diagram (Translation Symmetry Breaking):
- When allowing for $2\times2$ unit cells, translational symmetry-breaking phases (stripes, $2\times2$ orders) appear at low interaction strengths near the VHS filling.
- However, as interaction strength increases, nematic phases dominate. The system prefers rotational symmetry breaking over translational symmetry breaking, rationalizing the absence of strong CDW orders in the experimental flat-band regime.
Band Structure Signatures:
- In the nematic phase, the band structure shows a lifting of bands near one $M$ point (e.g., $M_3$ ) and lowering near others ( $M_1, M_2$ ), accompanied by a redistribution of sublattice weights (e.g., sublattice B becomes heavily occupied while A and C are depleted). This matches the spectral weight splitting observed in experiments.

5. Significance and Conclusion

Unified Framework: The paper establishes a minimal framework where nematicity is a generic outcome of partially filled kagome flat bands when inter-sublattice interactions are significant. It bridges the gap between flat-band ferromagnetism and VHS-driven instabilities.
Explanation of Experiments: The results provide a theoretical explanation for the prevalence of the nematic phase in Fe-doped CoSn, attributing it to the competition between on-site $U$ and extended $V$ interactions within the flat band manifold, rather than a collaboration between different band sectors.
Suppression of CDW: The study clarifies why charge density waves are less prominent in this specific regime: the weak dispersion of the flat band leads to poor Fermi surface nesting, making translational symmetry breaking less favorable than rotational symmetry breaking (nematicity) at realistic interaction strengths.
Future Directions: The authors suggest that while their single-orbital model captures the essence, future work should explore the interplay with spin-orbit coupling and the hybridization of in-plane and out-of-plane orbitals to fully bridge the gap with phenomenological descriptions.

In summary, the paper demonstrates that extended Coulomb interactions in weakly dispersive kagome flat bands naturally stabilize nematic order over ferromagnetism and charge density waves, offering a robust theoretical basis for the correlated phases observed in recent experiments on CoSn.