Paramagnetic phases of strongly correlated ultracold fermions coupled to an optical cavity

Using real-space dynamical mean-field theory, this study numerically investigates the paramagnetic phases of ultracold fermions in a cavity-coupled optical lattice, revealing a reentrant density-wave transition at quarter filling and a cavity-mediated destabilization of the half-filled system into a density-wave phase that coexists with Fermi liquid and Mott insulating states.

The Gist

Imagine you have a giant, invisible dance floor made of a grid of tiny squares. On this floor, you place thousands of tiny, energetic dancers. These aren't just any dancers; they are ultracold fermions (a type of atom that behaves like a quantum particle). They are so cold they barely move, and they are "strongly correlated," meaning they are extremely sensitive to what their neighbors are doing. If one dancer stops, everyone else feels it.

Now, imagine this dance floor is inside a special room with mirrors on all sides—an optical cavity. You shine a laser beam across the room (the "pump").

Here is the magic trick: When the dancers move, they scatter the laser light. This light bounces off the mirrors and comes back to hit other dancers. It's like the dancers are talking to each other through a walkie-talkie that connects everyone in the room instantly, regardless of how far apart they are. This creates a long-range interaction.

The scientists in this paper wanted to see what happens when these dancers try to organize themselves under these conditions. Do they stay in a chaotic, free-flowing crowd? Or do they snap into a rigid, organized pattern?

The Three Main "Dances" (Phases)

The researchers found that depending on how crowded the floor is and how strong the "laser connection" is, the dancers fall into three distinct patterns:

1. The Free-Flowing Crowd (Fermi Liquid):

   • What it is: The dancers are moving around freely, mixing with everyone. There is no specific pattern.
   • The Analogy: Think of a mosh pit at a concert where everyone is jostling around but no one is standing in a specific spot. It's a "metallic" state where things flow easily.

2. The Solo Dancers (Mott Insulator):

   • What it is: The dancers are so repelled by each other (due to their natural "personal space" rules) that they refuse to share a square. Every square gets exactly one dancer, and they stop moving.
   • The Analogy: Imagine a crowded subway car where everyone is so uncomfortable they freeze in place, refusing to let anyone else squeeze in. The flow stops completely. This is an "insulator."

3. The Checkerboard Pattern (Density Wave):

   • What it is: The dancers spontaneously organize into a perfect checkerboard. They all decide to stand on the "white" squares and leave the "black" squares empty (or vice versa).
   • The Analogy: This is like a synchronized flash mob where everyone suddenly jumps into a perfect grid formation. The laser light helps them "see" this pattern and lock into it. This is called a Density Wave.

The Surprising Discoveries

The paper reveals some counter-intuitive and fascinating behaviors:

1. The "Hotter is Better" Surprise (at Quarter Filling)
Usually, if you heat something up, it melts and becomes disordered. But at a specific crowd density (quarter filling), the researchers found that heating the dancers actually made them organize!

• The Analogy: Imagine a messy room. Usually, if you shake it, it gets messier. But here, shaking it (adding heat) actually helped the items snap into a neat pile. Why? Because the organized "checkerboard" pattern actually has more "freedom" (entropy) for the dancers' internal spins at higher temperatures. It's a rare case where order is the more comfortable state when things get hot.

2. The "Perfect Nesting" Instability (at Half Filling)
When the floor is half-full, the dancers have a special geometric alignment (perfect Fermi surface nesting).

• The Analogy: Imagine a row of dominoes. If you push the first one, they all fall. Here, because of the geometry of the dance floor, even the weakest whisper from the laser (the smallest long-range interaction) is enough to make the entire crowd snap into the checkerboard pattern instantly. You don't need a strong push; a tiny nudge is enough to trigger a massive change.

3. The "Stuck in the Middle" Zone (Coexistence)
The researchers found a tricky zone where the system can't decide what to be.

• The Analogy: Imagine a ball sitting on a hilltop between two valleys. It could roll down into the "Free-Flow" valley or the "Checkerboard" valley. Depending on which way you nudge it (or how you start the simulation), it rolls one way or the other. This is called metastability. The system can exist in a state where it looks like it's trying to be both a crowd and a pattern at the same time, leading to a "first-order" phase transition (a sudden, dramatic jump from one state to another).

Why Does This Matter?

This isn't just about cold atoms in a lab. It's a simulator for real-world materials.

• Superconductors: Understanding how electrons organize in these patterns helps us understand how electricity can flow without resistance (superconductivity) in complex materials.
• New Materials: By tuning the laser and the "dance floor," scientists can design materials with specific properties that don't exist in nature yet.
• Quantum Computing: These systems are highly controllable, making them a testbed for understanding how quantum information behaves in complex, interacting environments.

In a nutshell: The paper shows that when you give quantum particles a way to talk to each other instantly across a room (via a laser cavity), they can spontaneously organize into beautiful, complex patterns. Sometimes, heating them up helps them organize, and sometimes, the tiniest nudge is enough to change the entire system's state. It's a dance between chaos and order, choreographed by light.

Technical Summary

1. Problem Statement

The paper investigates the interplay between strongly correlated short-range interactions (Hubbard $U$) and cavity-mediated infinite-range interactions in a two-component ultracold Fermi gas confined to a 2D square optical lattice.

While the competition between onsite and long-range interactions has been extensively studied for bosons (leading to phenomena like superradiant Mott insulators), the fermionic case remains less explored. Specifically, the authors aim to understand:

• How cavity-mediated long-range interactions affect the phase diagram of fermions at different fillings (quarter vs. half).
• The nature of the transition between the Mott Insulator (MI) and the Density Wave (DW) phase.
• Whether the transition between these phases is first-order, characterized by metastability and coexistence, similar to the bosonic case.
• The role of Fermi surface nesting in stabilizing density waves.

2. Methodology

The authors employ Real-Space Dynamical Mean-Field Theory (RDMFT) to solve the extended Fermi-Hubbard model derived from the atom-cavity system.

• Model Derivation:

  • The system consists of a balanced two-component Fermi gas ($^6$Li) in a 2D square lattice, transversely pumped by a laser and coupled to a single-mode optical cavity.
  • In the dispersive regime (large atom-pump detuning), the system is described by an extended Fermi-Hubbard Hamiltonian:
    $$\hat{H} = \hat{H}_{HB} - \frac{U_l}{2} \left( \sum_{i\sigma} (-1)^{|i|} \hat{n}_{i\sigma} \right)^2$$
    where $\hat{H}_{HB}$ is the standard Hubbard model (tunneling $t$, onsite repulsion $U_s$, chemical potential $\mu$), and the second term represents the cavity-mediated infinite-range density-density interaction with strength $U_l$. The operator $\hat{\Theta} = \frac{1}{N_s}\sum (-1)^{|i|}\hat{n}_{i\sigma}$ acts as the order parameter for checkerboard density waves.
  • The interaction is assumed to be red-detuned ($U_l > 0$).

• RDMFT Implementation:

  • The infinite-range interaction is treated via a static mean-field decoupling ($\hat{\Theta}^2 \approx 2\langle\hat{\Theta}\rangle\hat{\Theta} - \langle\hat{\Theta}\rangle^2$), reducing the problem to a set of coupled single-impurity Anderson models.
  • The self-consistency loop involves:
    1. Solving the impurity problem for each lattice site (or sublattice) using exact diagonalization (with a finite bath, typically 5 orbitals).
    2. Calculating the local Green's function and self-energy.
    3. Updating the lattice Green's function via the lattice Dyson equation.
    4. Recalculating the order parameter $\langle\hat{\Theta}\rangle$ until convergence.
  • Symmetry Breaking: To handle the checkerboard phase efficiently, the authors utilize a unit-cell procedure assuming a bipartite lattice structure (even/odd sublattices), reducing computational cost while capturing $Z_2$ parity symmetry breaking.
  • Phase Identification:
    • Fermi Liquid (FL): Finite quasiparticle residue ($Z_{e,o} > 0$) and zero imbalance ($\langle\hat{\Theta}\rangle = 0$).
    • Mott Insulator (MI): Vanishing quasiparticle residue ($Z_{e,o} = 0$) and zero imbalance.
    • Density Wave (DW): Finite imbalance ($\langle\hat{\Theta}\rangle \neq 0$). At half-filling, this phase is insulating due to the gap opened by the staggered potential.

• Thermodynamic Stability: To resolve coexistence regions (metastability), the authors compare the grand potential (approximated by the mean-field energy $\langle \hat{H}_{MF} \rangle$ at low $T$) of different converged solutions (FL, MI, DW) to determine the true thermodynamic phase boundary.

3. Key Contributions and Results

A. Quarter Filling ($n=1/2$)

• Reentrant Behavior: The authors observe a unique reentrant phase transition as temperature increases.
  • At low $T$, the system is in a homogeneous Fermi Liquid (FL).
  • As $T$ increases, it transitions into a Density Wave (DW) phase.
  • At even higher $T$, it melts back into the FL phase.
• Mechanism: This is driven by entropy. In the deep DW phase at quarter filling, occupied sites have spin degeneracy, yielding an entropy contribution of $\ln 2$ per site. The FL entropy scales linearly with $T$. The higher entropy of the ordered DW phase lowers its free energy relative to the FL at intermediate temperatures, stabilizing the order against thermal fluctuations before thermal melting occurs at high $T$.

B. Half Filling ($n=1$)

• Instability to DW: Due to perfect Fermi surface nesting at half-filling on a square lattice, the system is unstable towards the DW phase for arbitrarily small long-range interactions ($U_l$) when short-range interactions ($U_s$) are zero.
• Phase Diagram ($T=0.02$):
  • Small $U_s, U_l$: Transition from FL to DW.
  • Large $U_s$, Small $U_l$: Transition from FL to Mott Insulator (MI). The critical $U_s^c \approx 10.3$ is independent of $U_l$ in this regime.
  • Large $U_s, U_l$: Transition from MI to DW.
• First-Order Transition and Coexistence:
  • The transition between the MI and DW phases (and FL and DW) is first-order.
  • The authors identify metastable regions where solutions for different phases coexist.
  • Hysteresis: The converged solution depends on the initialization of the self-consistency loop (starting from MI vs. DW), leading to hysteresis in the order parameter $\langle\hat{\Theta}\rangle$.
  • Thermodynamic Boundary: By comparing energies, the true phase boundary is found to follow the line $N_s U_l \approx U_s$, consistent with an atomic limit analysis where the energy cost of double occupancy ($U_s$) is balanced by the gain from long-range ordering ($U_l$).

C. Comparison with Bosons

The results mirror findings in bosonic cavity systems (e.g., superradiant Mott insulators), confirming that the competition between short-range repulsion and infinite-range attraction leads to metastable coexistence regions and first-order transitions, even in fermionic systems where Pauli blocking plays a role.

4. Significance

• Theoretical Validation: The study provides a rigorous theoretical framework for understanding cavity-mediated interactions in fermionic systems, bridging the gap between standard Hubbard physics and cavity QED.
• Experimental Relevance: The predicted phase diagrams and reentrant behaviors offer specific targets for ongoing experiments with ultracold $^6$Li atoms in optical cavities. The identification of the $N_s U_l \approx U_s$ boundary provides a quantitative guide for tuning lattice depth and pump-cavity detuning.
• Methodological Advance: The successful application of RDMFT to this extended model demonstrates its capability to handle competing orders (charge order vs. Mott localization) and first-order transitions in 2D fermionic systems, including the calculation of kinetic energy and thermodynamic potentials directly from impurity solvers.
• Entropy-Driven Ordering: The discovery of entropy-driven reentrant ordering at quarter filling highlights a non-trivial thermodynamic mechanism where thermal fluctuations can induce order, a counter-intuitive phenomenon relevant to strongly correlated materials.

Conclusion

The paper establishes that cavity-mediated long-range interactions in 2D fermionic lattices induce rich phase diagrams characterized by the competition between Fermi liquid, Mott insulating, and density-wave phases. Key findings include the stabilization of density waves via Fermi surface nesting at half-filling, entropy-driven reentrant transitions at quarter-filling, and a first-order transition between Mott and density-wave insulators characterized by metastability and hysteresis. These results deepen the understanding of nonlocal interactions in quantum matter and provide a roadmap for observing complex quantum phases in ultracold atom experiments.