Luttinger's Theorem Violation and Green's Function Topological Invariants in a Fractional Chern Insulator

Using exact diagonalization of the fermionic Harper-Hofstadter-Hubbard model, this study demonstrates the violation of Luttinger's theorem in fractional Chern insulators and elucidates how the fractional many-body Chern number and integer Green's function invariants are encoded in the Středa responses of the Luttinger integral and count, respectively, while proposing an experimental protocol to extract these topological invariants via local density-of-states measurements.

The Gist

Imagine you are trying to understand a massive, chaotic city by looking at a single map. In the world of physics, this "city" is a material made of electrons, and the "map" is a mathematical tool called the Green's function.

For decades, physicists have relied on a famous rule called Luttinger's Theorem. Think of this theorem as a strict census taker. It says: "If you count the number of 'poles' (peaks) in your map below a certain energy line, that count must exactly equal the number of people (electrons) living in the city."

In normal materials, this rule works perfectly. The census taker is accurate.

However, this paper investigates a very special, exotic type of city called a Fractional Chern Insulator (FCI). In these cities, the "citizens" aren't just normal electrons; they are fractionalized quasiparticles. Imagine if a single electron could split into three ghostly fragments, each carrying only 1/3 of an electron's charge. These fragments don't behave like normal people; they are a collective, quantum dance that has no direct connection to the individual electrons that started the dance.

The authors of this paper asked a big question: Does the census taker (Luttinger's Theorem) still work in this exotic city?

The Discovery: The Census Taker is Confused

Using powerful supercomputers to simulate these materials (a method called "Exact Diagonalization"), the team found that Luttinger's Theorem breaks down.

Here is the analogy:

• The Census Count (Luttinger Count): The map shows a specific number of peaks. The census taker counts them and says, "There are 1.0 citizens here."
• The Real Population: But if you actually count the electrons, there are only 0.33 (one-third) of a citizen per spot.
• The Result: The census taker is wrong! The map says "1," but the reality is "1/3."

This violation is a huge deal. It proves that in these exotic states, the simple relationship between the "map" (Green's function) and the "reality" (particle density) has snapped.

The Solution: Splitting the Bill

So, if the census is wrong, how do we fix the map? The authors realized the "error" isn't random; it's structured. They split the total count into two parts, like splitting a restaurant bill:

1. The Integer Part (The Luttinger Count): This part of the map still looks like a normal, whole number (like 1). It represents the "skeleton" of the material—the underlying grid or lattice structure.
2. The Fractional Part (The Luttinger Integral): This is the "tip" or the "correction." It's a negative number that cancels out the extra count from the first part.

The Magic Formula:
$$\text{Total Electrons} = (\text{Integer Count}) + (\text{Fractional Correction})$$

In their exotic city:

• The Integer Count says: 1
• The Fractional Correction says: -0.66
• Total: $1 - 0.66 = 0.33$ (The correct fractional charge!)

The "Středa Response": How the City Reacts to a Magnetic Storm

To prove this, the team didn't just count; they watched how the city reacted when they turned on a magnetic field (like a storm).

• The Integer Part: When the magnetic storm hit, the Integer Count didn't change its "whole number" nature. It stayed stubbornly at 1. This tells us that the "skeleton" of the material is still topologically integer-based.
• The Fractional Part: The Correction part, however, reacted beautifully. It shifted by exactly -2/3.

This shift is the key. The "fractional" nature of the exotic material (the 1/3 charge) is hidden entirely inside this Correction Term. The "Integer" part of the map is a red herring; the real magic is in the correction.

The Big Picture: Why This Matters

1. New Topology: For a long time, physicists thought the "topological number" (which tells us how the material conducts electricity) was just a simple integer. This paper shows that in fractional states, the topological number is actually a sum of an integer and a fractional correction.
2. Experimental Proof: The authors didn't just do math; they proposed a way for real-world scientists to see this. They suggested using a microscope (Scanning Tunneling Microscopy) to look at the "local density of states" (essentially, taking a photo of the electron energy levels). By analyzing the peaks and valleys in these photos, experimentalists can now extract these "Luttinger counts" and "corrections" without needing to solve the impossible math of the whole system.

Summary in a Nutshell

Imagine you are trying to weigh a bag of marbles.

• Old Rule: "Count the marbles, and that's the weight."
• New Reality: In this special bag, the marbles are made of a weird material. If you count them, you get 1. But the bag actually weighs 1/3.
• The Breakthrough: The authors realized the scale is broken. It has a "Base Weight" (1) and a "Mystery Discount" (-2/3).
• The Conclusion: The "Base Weight" tells us about the shape of the bag (the lattice), but the "Mystery Discount" tells us about the magic inside (the fractional charge).

This paper successfully identifies where the "magic" lives in the math, proving that while the old rules break, a new, more complex set of rules (involving these corrections) perfectly describes the exotic world of fractional quantum matter.

Technical Summary

1. Problem Statement

The paper addresses a fundamental conflict in the theory of strongly correlated topological phases, specifically Fractional Chern Insulators (FCIs) and Fractional Quantum Hall (FQH) states.

• The Conflict: In weakly interacting systems, the quantized Hall conductivity is linked to a topological invariant of the single-particle Green's function, the Ishikawa-Matsuyama invariant ($N_3[G]$), via the relation $\sigma_H = (e^2/h)N_3[G]$. This invariant is an integer. However, in FQH/FCI states, the Hall conductivity is fractionally quantized ($\sigma_H = (e^2/h)C_{MB}$), where $C_{MB}$ is the many-body Chern number.
• The Question: How does the integer-valued Green's function invariant $N_3[G]$ relate to the fractional many-body Chern number $C_{MB}$ in the presence of strong correlations? Does Luttinger's theorem (which relates particle density to the volume enclosed by the Fermi surface) hold in these phases?
• The Gap: Previous studies relied on edge theories or perturbative assumptions. A direct, non-perturbative evaluation of Green's function invariants deep within the bulk of a fractionalized phase has been elusive.

2. Methodology

The authors employ a combination of exact diagonalization (ED) and analytical field theory to investigate the fermionic Harper-Hofstadter-Hubbard model, a lattice model known to host FCI phases.

• Numerical Approach:

  • Model: Spinless fermions on a lattice with a uniform magnetic field (flux $\phi$) and nearest-neighbor interactions ($V$), projected onto the lowest Hofstadter band (LHB).
  • Parameters: Filling factor $\nu = 1/3$, flux $\phi = 1/5$, and interaction strength $V/t = 10$.
  • Techniques:
    • Lanczos Exact Diagonalization: Used to compute ground states and excited states for systems with open boundaries (box) and periodic boundaries (torus).
    • Green's Function Calculation: Computed via the Lehmann representation using the ground state and $N \pm 1$ particle sectors.
    • Contour Integration: The Luttinger count ($N_1$) and Luttinger integral ($\Delta N_1$) are evaluated by integrating the Green's function and self-energy along contours in the complex frequency plane.
  • Středa Response: The authors calculate the derivative of particle densities with respect to magnetic flux ($\Phi_0 \partial N / \partial \Phi$) to extract topological invariants.

• Analytical Approach:

  • Derivation of an explicit expression for $N_3[G]$ assuming no Bloch band mixing (diagonal Green's function in the band basis).
  • Utilization of the Středa-Widom relation and Luttinger's theorem decomposition to link $N_3[G]$, $C_{MB}$, and the Luttinger integral.

3. Key Contributions

1. Direct Numerical Demonstration of Luttinger's Theorem Violation:
   The paper provides the first exact numerical computation showing that Luttinger's theorem breaks down in the bulk of an FCI. The particle density $n$ is not equal to the Luttinger count $n_1$; instead, $n = n_1 + \Delta n_1$, where the Luttinger integral $\Delta n_1$ is non-zero and negative.

2. Decomposition of the Hall Response:
   The authors demonstrate that the fractional Hall response is entirely encoded in the Středa response of the Luttinger integral, while the Green's function invariant $N_3[G]$ remains an integer.

   • $N_3[G] \approx 1$ (Integer).
   • $C_{MB} = 1/3$ (Fractional).
   • The discrepancy is captured by $\Delta N_3 = C_{MB} - N_3[G] = \Phi_0 \partial (\Delta N_1) / \partial \Phi \approx -2/3$.

3. Analytical Proof of $N_3[G]$ Determinants:
   They analytically prove that in the absence of Bloch band mixing, $N_3[G]$ is fully determined by the Luttinger count and the Chern number of the occupied Bloch band. Specifically, $N_3[G]$ reflects the single-particle topology inherited from the underlying band structure, not the fractional many-body topology.

4. Experimental Protocol Proposal:
   The paper proposes a practical scheme to extract these topological invariants ($N_1$, $\Delta N_1$, $N_3$, $\Delta N_3$) from local density of states (LDOS) measurements, which are accessible via scanning tunneling microscopy (STM) in moiré materials and graphene.

4. Key Results

• Spectral Structure: The bulk Green's function exhibits sharp, coherent hole-like and electron-like peaks (long-lived quasiparticles) separated by a charge gap. Crucially, a zero of the Green's function (a pole in the self-energy) emerges within the gap.
• Violation Mechanism: The violation of Luttinger's theorem arises because the spectral weight of the hole peak is less than unity (due to interaction-induced redistribution), but the Luttinger count assigns a unit weight to the pole below the chemical potential. The Luttinger integral $\Delta N_1$ compensates for this difference.
• Topological Invariants:
  • $N_3[G]$: Calculated as $\approx 1$. This integer value corresponds to the Chern number of the underlying single-particle band (which is 1 for the lowest Hofstadter band) and the specific choice of chemical potential relative to the Green's function zero.
  • $C_{MB}$: Calculated as $1/3$, consistent with the Laughlin state.
  • Středa Responses:
    • $\partial n_1 / \partial B \approx 1$ (Integer).
    • $\partial \Delta n_1 / \partial B \approx -2/3$ (Fractional).
    • Total density response $\partial n / \partial B \approx 1/3$.
• Chemical Potential Dependence: The value of $N_3[G]$ is sensitive to the position of the chemical potential relative to the Green's function zero. If the chemical potential crosses the zero, $N_3[G]$ changes (e.g., from 1 to 0), while $C_{MB}$ remains fixed as long as the many-body gap persists.

5. Significance

• Resolution of the Integer-Fractional Paradox: The paper clarifies that the Ishikawa-Matsuyama invariant $N_3[G]$ is not the many-body Chern number in fractionalized phases. Instead, $N_3[G]$ captures the single-particle topology, while the fractional nature of the Hall effect is carried by the Luttinger integral (which represents the breakdown of the adiabatic connection between electrons and quasiparticles).
• Role of Green's Function Zeros: It reinforces the theoretical insight that the existence of zeros in the single-particle Green's function is the mechanism responsible for the violation of Luttinger's theorem and the deviation of $N_3[G]$ from $C_{MB}$.
• Experimental Relevance: By linking these abstract topological invariants to the local density of states (LDOS), the authors provide a roadmap for experimentalists to verify these theoretical predictions in real materials (e.g., twisted bilayer graphene or moiré heterostructures) using STM.
• Theoretical Framework: The work establishes a rigorous connection between the Středa response, Luttinger's theorem, and topological invariants in strongly correlated systems, offering a new perspective on how to characterize topological order beyond the single-particle picture.

In summary, this work demonstrates that in Fractional Chern Insulators, the "topology" seen by a single-particle probe (Green's function) is integer-valued and distinct from the fractional many-body topology. The fractionalization is hidden in the Luttinger integral, which accounts for the non-adiabatic nature of the quasiparticles.