Emergent magnetic order in the antiferromagnetic Kitaev model with a [111] field

Using hierarchical mean-field theory with 24-site clusters, this study shows that the antiferromagnetic Kitaev model under a [111] magnetic field transitions from a spin liquid through two intermediate phases characterized by stripe and chiral orders into a trivial partially polarized phase, a finding validated by exact diagonalization and topological observables.

The Gist

Imagine a tiny, magical dance floor made of a honeycomb pattern (like a beehive). On this floor, tiny dancers (electrons with "spin") try to decide how to move. In a special setup called the Kitaev model, these dancers are forced to interact in a very specific, frustrating way: they only speak to their neighbors when looking in a particular direction.

Normally, when these dancers interact this way, they choose neither a single leader nor a rigid formation. Instead, they enter a chaotic, fluid state known as a quantum spin liquid. In this state, the dancers constantly switch partners and refuse to settle down. This is a "topological" state, meaning the entire system possesses a hidden, global shape that is very hard to break, similar to a knot that cannot be untied without cutting the rope.

The Experiment: A Magnetic Wind
The researchers in this work asked: "What happens if we blow a strong, steady wind (a magnetic field) over this dance floor?" Specifically, they let the wind blow from a [111] direction (a specific angle in 3D space).

Previous studies suggested that as the wind grew stronger, the dancers would slowly begin to align with the wind, transforming the chaotic fluid into a calm, ordered line (a "partially polarized" phase). They believed there might be a brief, disordered intermediate phase, but were unsure what it looked like.

The New Discovery: Two Hidden Intermediate Stages
Using a powerful new simulation method called Hierarchical Mean-Field Theory (HMFT)—which acts like zooming in on small groups of dancers to see how they influence their neighbors—the authors found that the story is much more complex. The dancers do not simply transition from "chaos" to "order." They pass through two distinct intermediate phases before finally settling down.

Here is the dancers' journey, simply explained:

1. The Starting Point (The Spin Liquid): At low wind speeds, the dancers are in their famous fluid, chaotic state. They are "topological," meaning they share a special, unbreakable connection with one another.
2. The First Stop: The "Striped" Phase: As the wind strengthens, the dancers suddenly decide to form stripes. Imagine the dancers suddenly organizing themselves into rows, with everyone in one row looking in one direction and the next row looking in the opposite direction. This is a big deal because it breaks the perfect symmetry of the dance floor. The dancers are no longer in a fluid state; they have developed a rigid, long-range pattern (like a striped shirt).
3. The Second Stop: The "Chiral" Phase: As the wind becomes even stronger, the dancers do not immediately align with the wind. Instead, they enter a "twisted" state. Imagine the dancers still looking partially toward the wind, but also turning in a specific direction (clockwise or counterclockwise) relative to their neighbors. The authors call this the Chiral partially polarized phase. It is a mixture of order imposed by the wind and a specific "handedness" or rotation.
4. The Final Destination: The Polarized Phase: Finally, at very high wind speeds, the dancers abandon their patterns and simply all look toward the wind, becoming a simple, aligned line.

Why This Matters
The researchers compared their results with other computer simulations (such as Exact Diagonalization). They found that their new method (HMFT) could clearly identify these two hidden intermediate stages, whereas earlier methods missed them or interpreted them differently.

• The "Striped" Phase was a surprise because it showed that the system develops a genuine, physical order (stripes) rather than remaining a "featureless" fluid.
• The "Chiral" Phase was a new discovery that no one had clearly identified before. It possesses a specific rotation (chirality) that persists even as the system becomes more ordered.

The Conclusion
Imagine the magnetic field as a conductor trying to get a jazz band (the spin liquid) to play a simple march. The work shows that the band does not simply switch from jazz to marching immediately. First, they play a strange, structured blues song (the Striped Phase), then a complex, swirling waltz (the Chiral Phase), and then they finally begin to march in step.

The authors used a method that examines small groups of dancers to predict how the entire crowd behaves, and they proved that this method is excellently suited for investigating these complex, topological quantum systems. They confirmed their results by performing smaller, exact calculations to ensure their "big picture" was correct.

Technical Summary

Problem Statement
The Kitaev honeycomb model (KHM) is a paradigmatic, exactly solvable model that hosts a topological $\mathbb{Z}_2$ quantum spin liquid ground state (KSL). While the properties at zero field are well understood, the fate of the KSL under an external magnetic field remains the subject of intense debate, particularly regarding the nature of the intermediate phases that emerge before the system transitions into a trivial, partially polarized (PP) state. Previous numerical studies using exact diagonalization (ED) and density matrix renormalization group (DMRG) have largely suggested the existence of a gapless $U(1)$ quantum spin liquid or a gapped topological phase in the intermediate regime. However, these methods reach their limits when simultaneously capturing long-range order and behavior in the thermodynamic limit. This work investigates the quantum phase diagram of the antiferromagnetic KHM under a magnetic field applied in the [111] direction, aiming to elucidate the nature of these intermediate phases.

Methodology
The authors apply the Hierarchical Mean-Field Theory (HMFT), an algebraic and numerical framework that utilizes spin clusters to preserve relevant symmetries and short-range quantum correlations.

• Cluster Approach: The study employs clusters with $N_c=6$ and $N_c=24$ lattice sites. These clusters tile the lattice and preserve the $C_6$ rotational symmetry of the honeycomb lattice (combined with a spin rotation $C_3^S$).
• Variational Approach: The method uses a uniform cluster Gutzwiller ansatz (CGA), where the wavefunction is a product of identical cluster states. The variational parameters are optimized to minimize the energy density in the thermodynamic limit.
• Validation and Support: To assess the validity of the HMFT predictions, the authors perform exact diagonalization (ED) on clusters with 18 and 24 lattice sites using periodic boundary conditions (PBC).
• Observables: The study calculates magnetization, sublattice scalar chirality, plaquette flux, topological entanglement entropy, and the many-body Chern number. Finite-size scaling is performed to extrapolate the results to the thermodynamic limit.

Main Contributions and Results
The application of HMFT to the KHM in a [111] field reveals a phase diagram that differs significantly from previous ED and DMRG results. The system transitions from the KSL phase through two distinct intermediate phases before reaching the trivial PP phase:

1. Intermediate Phase (Stripe Order):

   • At a critical field $h_1 \approx 0.14$, the KSL undergoes a first-order transition into an intermediate phase.
   • Contrary to earlier claims of a structureless quantum spin liquid (QSL), this phase exhibits spontaneous symmetry breaking (SSB) of the $C_6 \times C_3^S$ symmetry, reduced to $C_2 \times C_1^S$.
   • This phase is characterized by magnetic stripe order, evidenced by a non-vanishing staggered magnetization of the stripe order and a non-vanishing symmetry-breaking parameter $O_i$ within the cluster.
   • ED results on 24-site clusters support this finding, showing enhanced susceptibility and static structure factors corresponding to stripe order.
   • The many-body Chern number in this phase is ill-defined and jumps between integer values, which the authors interpret as evidence for a gapless spectrum rather than a gapped topological state.

2. Chiral Partially Polarized ($\chi$-PP) Phase:

   • At a second critical field $h_2 \approx 0.25$, another transition occurs (identified by level crossings between competing mean-field solutions).
   • This phase, termed the chiral partially polarized ($\chi$-PP) phase, was overlooked in previous ED and DMRG studies.
   • It is characterized by the simultaneous presence of partial polarization in the [111] direction and finite sublattice scalar chirality ($\chi \approx 0.25$).
   • Unlike the intermediate phase, the scalar chirality in the $\chi$-PP phase permeates the interior of the cluster, not just the edges.
   • This phase is separated from the trivial PP phase by a second-order transition at $h_3 \approx 0.51$.
   • The many-body Chern number in the $\chi$-PP phase is determined to be $C=0$, suggesting that a chiral phase with a zero Chern number can exist.

3. Trivial Partially Polarized (PP) Phase:

   • At high fields ($h > h_3$), the system enters a trivial PP phase with nearly saturated magnetization and $C=0$.

Significance and Claims
The work claims to establish HMFT as a robust method for investigating topological quantum spin liquids, particularly in regimes where long-range order competes with topological features.

• Resolution of Intermediate Phases: The work challenges the consensus of a structureless intermediate QSL and instead proposes a phase with long-range stripe order followed by a chiral phase.
• Methodological Validation: By comparing HMFT with the exact solution at $h=0$ and reproducing known topological properties (such as $S_{topo} = -\log 2$ and $C=1$ in the KSL), the authors validate the method's ability to capture complex quantum phases.
• Discovery of the $\chi$-PP: The identification of the $\chi$-PP phase underscores HMFT's capability to detect symmetry-breaking transitions and chiral orders that may be obscured in finite-size ED or DMRG calculations due to boundary effects or system size limitations.
• Topological Characterization: The study demonstrates that the intermediate phase lacks a well-defined Chern number, supporting the hypothesis of a gapless spectrum, while the $\chi$-PP phase is shown to be a chiral but topologically trivial phase ($C=0$).

The authors conclude that while their results provide a comprehensive picture of the phase diagram, a rigorous confirmation of the behavior in the thermodynamic limit for the chiral and stripe orders may require larger clusters or new computational approaches beyond current classical methods.