Mean-field phase diagrams of spinor bosons in an optical cavity

This paper employs a grand-canonical mean-field approach to map out the ground-state phase diagrams of spinor bosons in an optical cavity, revealing the emergence of novel magnetic phases—including antiferromagnetic Mott insulators, ferromagnetic density waves, and distinct supersolid regimes—under both homogeneous and harmonically trapped conditions.

The Gist

Imagine a giant, high-tech dance floor made of light (an optical lattice) where thousands of tiny, invisible dancers (ultracold atoms) are performing. These aren't just any dancers; they are "spinor bosons," which means each dancer has a secret internal state, like wearing either a Red Hat or a Blue Hat.

Now, imagine this dance floor is inside a mirrored room (an optical cavity). When the dancers move, they don't just bump into their immediate neighbors; they shout across the room, and the mirrors bounce those shouts back instantly. This creates a "long-range" conversation where every dancer feels the presence of every other dancer, no matter how far away they are.

This paper is a theoretical map of the different "dance routines" (phases) these atoms can perform under these conditions. The authors, Maksym Prodius, Mateusz Łącki, and Jakub Zakrzewski, used computer simulations to predict exactly what happens when you tweak the music (laser light) and the room's acoustics (cavity strength).

Here is the breakdown of their findings in simple terms:

1. The Setup: The Dance Floor and the Echo

The atoms are trapped in a grid. The researchers are looking at two main scenarios:

• The "Free-for-All" (Unconstrained): The dancers can wear as many Red or Blue hats as they want. The total number of hats doesn't matter.
• The "Balanced Team" (Zero Magnetization): The researchers force the team to have an equal number of Red and Blue hats. This is more like real experiments where you start with a fixed mix of atoms.

2. The "Atomic Limit": When the Dancers Freeze

First, the authors imagined the dancers couldn't move at all (no tunneling). They just stood in their spots. Even without moving, the "shouts" from the mirrors forced them into specific patterns:

• The Checkerboard (AFM): Imagine a chessboard where Red hats sit on black squares and Blue hats sit on white squares. They alternate perfectly. This is called an Antiferromagnetic Mott Insulator. It's rigid and orderly.
• The Spin-Flip Wave (FDW): In some cases, the dancers don't just alternate colors; they create a wave where one side of the room is mostly Red and the other is mostly Blue, but the density of people changes too. This is a Ferromagnetic Density Wave.
• The Entangled Wave (EDW): When the researchers forced an equal number of Red and Blue hats, something magical happened. Instead of picking a side, the dancers in the same spot became a "superposition." Imagine a dancer who is simultaneously wearing a Red hat and a Blue hat at the same time, creating a new, entangled state. This is the Entangled Density Wave.

3. Turning on the Music: Letting Them Dance (Tunneling)

Next, they let the dancers move between the grid spots (tunneling). This is where things get wild. The rigid patterns start to melt into fluid, flowing states, but they don't just become a messy soup. They form Supersolids.

Think of a Supersolid as a paradox:

• It's a Solid because the dancers are still arranged in a specific, rigid pattern (like a crystal).
• It's a Superfluid because they can flow without friction, like water.

The paper found three distinct types of these supersolid dances:

1. The Standard Supersolid: A mix of order and flow.
2. The "Spin-Imbalanced" Supersolid: A flow where the Red and Blue hats are unevenly distributed in a specific pattern.
3. The "Double-Imbalanced" Supersolid: A complex dance where both the number of people and their hat colors are unevenly distributed in a flowing pattern.

4. The "Real World" Test: The Trap

In real experiments, you can't have an infinite dance floor. You have a bowl-shaped trap (a harmonic potential) that holds the atoms in the center. The density of dancers is high in the middle and low at the edges.

The authors simulated this "bowl" and found a "Wedding Cake" structure.
Imagine looking at the dance floor from the side.

• The Center: Might be a dense, rigid block of Red/Blue alternating dancers (AFM).
• The Middle Ring: Might be a flowing, wobbly supersolid.
• The Outer Edge: Might be a sparse, flowing superfluid.

Just like a wedding cake has distinct layers, the trapped atoms form distinct "shells" of different phases. The paper maps out exactly which "layer" you get based on how strong the laser is and how much the atoms want to move.

The Big Takeaway

The most exciting discovery is that when you force the system to be balanced (equal Red and Blue hats), the "Ferromagnetic" phases (where one color dominates) disappear. They are replaced by these new Entangled Density Waves (EDW).

In a metaphor:
If you have a crowd of people and you tell them to pick a side (Team Red or Team Blue), they might split into two distinct groups. But if you tell them, "You must have exactly 50% Red and 50% Blue," they can't split. Instead, they might hold hands and spin together, creating a new, unified group identity that is neither purely Red nor purely Blue, but a quantum superposition of both.

Why Does This Matter?

This paper provides a "roadmap" for experimentalists. If they build this specific setup in a lab, they now know exactly what patterns to look for. They can tune their lasers to see if they can create these exotic "Entangled Density Waves" or "Supersolids," which are states of matter that don't exist in our everyday world but are crucial for understanding quantum mechanics and potentially building future quantum computers.

Technical Summary

1. Problem Statement

The paper investigates the ground-state properties of spinor bosons (specifically two-component bosons, e.g., hyperfine states of $^{87}\text{Rb}$) confined in a two-dimensional optical lattice placed inside a high-finesse optical cavity. The system is driven by a transverse laser with a tunable polarization angle.

Key physical features include:

• Long-range interactions: Mediated by cavity photons, leading to infinite-range interactions between atoms.
• Spin degree of freedom: Unlike previous studies on scalar bosons, this work focuses on the interplay between density and spin degrees of freedom.
• Experimental relevance: The study addresses the specific experimental configuration where the optical lattice nodes coincide with the antinodes of the cavity field.
• Constraints: The authors analyze both the unconstrained grand-canonical ensemble and the constrained case where the total magnetization is fixed (zero total magnetization), which is more representative of experimental conditions where spin-changing processes are negligible.

2. Methodology

The authors employ a grand-canonical mean-field approach based on the Gutzwiller ansatz.

• Hamiltonian: The system is modeled using an extended Bose-Hubbard Hamiltonian (Eq. 1) derived from the adiabatic elimination of the cavity field. It includes:
  • Nearest-neighbor hopping ($t$).
  • On-site repulsive interactions ($U$ and $U_{12}$).
  • Cavity-mediated scalar ($U_s$) and vectorial ($U_v$) interactions. These terms drive density imbalances ($\theta_s$) and spin imbalances ($\theta_v$), respectively.
  • A magnetization penalty term ($P$) to enforce fixed total magnetization ($M_{tot}$).
• Approaches:
  1. Atomic Limit: Tunneling ($t$) is set to zero. The problem is solved analytically by minimizing the energy density over Fock states in a two-site unit cell.
  2. Homogeneous Mean-Field: Tunneling is reintroduced. The system is treated using a two-site unit-cell Gutzwiller ansatz. Both perturbative (second-order) and non-perturbative (numerical minimization) methods are used to map phase boundaries.
  3. Inhomogeneous (Trapped) System: A harmonic trapping potential is added. Due to the infinite-range nature of cavity interactions, the Local Density Approximation (LDA) fails. Instead, the authors use a non-uniform Gutzwiller ansatz where independent coefficients are determined for every lattice site on a $35 \times 35$ grid.
  4. Phase Identification: For the trapped system, a Convolutional Neural Network (CNN) is trained to identify phases based on local observables (local magnetization and superfluid order parameter) to generate the phase diagram.

3. Key Contributions and Results

A. Atomic Limit (Zero Tunneling)

• Unconstrained Case ($P=0$):
  • Vectorial Dominance ($U_v > U_s$): The system forms Antiferromagnetic Mott Insulators (AFM) where opposite spins occupy alternating sites.
  • Scalar Dominance ($U_s > U_v$):
    • If $U_s < U/2$: Intermediate phases appear with density imbalance but no spin imbalance, identified as Ferromagnetic Density Waves (FDW).
    • If $U_s > U/2$: FDW phases persist but with fully occupied and empty sites alternating.
• Zero-Magnetization Constraint ($M_{tot}=0, P \to \infty$):
  • The AFM phases remain unchanged.
  • The FDW phases are replaced by Entangled Density Waves (EDW). In EDW, sites with equal superpositions of spin-up and spin-down states ($| \uparrow \rangle \pm | \downarrow \rangle$) alternate with empty sites. This represents a density wave of entangled pairs.

B. Homogeneous System (Finite Tunneling)

The introduction of tunneling leads to the emergence of supersolid (SS) phases surrounding the insulating lobes.

• Three Distinct Supersolid Phases:
  1. Lattice Supersolid (SS): Standard supersolid with superfluidity and density order.
  2. Antiferromagnetic Supersolid (AF-SS): Superfluidity coexisting with antiferromagnetic spin order.
  3. Type-2 Lattice Supersolid (SS2): A complex phase with simultaneous superfluidity, scalar density imbalance, and vectorial spin imbalance.
• Phase Transitions: The paper distinguishes between continuous and discontinuous transitions. Notably, transitions from insulating lobes to the superfluid phase are often discontinuous (first-order) in the non-perturbative regime, differing from perturbative predictions.
• Effect of Magnetization Constraint:
  • Under zero magnetization, the SS2 phase disappears.
  • The FDW lobes transform into EDW lobes.
  • In the scalar-dominant regime, the system transitions directly from EDW to a standard SS phase or an AF-SS phase, bypassing the complex SS2 regime.

C. Trapped System (Harmonic Potential)

The authors present the first comprehensive phase diagram for the trapped system, which is directly relevant for experiments.

• Method: The system size is fixed, and the total particle number is determined self-consistently.
• Vectorial Dominance ($U_v > U_s$):
  • Reveals a "wedding-cake" structure.
  • Phases include: Superfluid (SF) $\to$ AF-SS $\to$ AF-SS core surrounded by AFM shells $\to$ Concentric AFM shells ($\rho=1$ and $\rho=2$).
• Scalar Dominance ($U_s > U_v$):
  • Phases include: AFM $\to$ Trapped EDW (shells of density-modulated checkerboard phases with entangled spins) $\to$ Partial SS (superfluid core) $\to$ Full SS.
  • The EDW states in the trap exhibit spatially varying superposition coefficients, leading to complex phase separation patterns.

4. Significance

• New Quantum Phases: The paper identifies and characterizes novel ground states, specifically the Entangled Density Wave (EDW) and the Type-2 Supersolid (SS2), which arise solely due to the interplay of spin degrees of freedom and long-range cavity interactions.
• Experimental Guidance: By including a harmonic trap and fixing the total magnetization, the results provide a direct roadmap for future experiments with ultracold spinor gases in cavities. The predicted "wedding-cake" structures of mixed phases (AFM, AF-SS, EDW, SS) are observable signatures.
• Methodological Advancement: The work demonstrates the necessity of moving beyond the Local Density Approximation for systems with infinite-range interactions and validates the use of non-uniform Gutzwiller ansatz combined with machine learning (CNN) for phase identification in trapped systems.
• Magnetization Physics: It highlights how fixing the total magnetization (a realistic experimental constraint) fundamentally alters the phase diagram, suppressing certain supersolid regimes and transforming density waves into entangled states.

In conclusion, this paper provides a comprehensive theoretical framework for understanding the rich phase diagram of spinor bosons in optical cavities, bridging the gap between idealized homogeneous models and realistic trapped experimental setups.