Domain-wall melting and entanglement in free-fermion chains with a band structure

Basic Information

• Paper ID: 2507.03780
• Title: Domain-wall melting and entanglement in free-fermion chains with a band structure
• Author: Viktor Eisler (Graz University of Technology & University of Graz)
• Classification: cond-mat.stat-mech, cond-mat.quant-gas, quant-ph
• Publication Date: September 18, 2025 (SciPost Physics Submission)
• Paper Link: https://arxiv.org/abs/2507.03780

Abstract

This paper investigates domain-wall melting phenomena in free-fermion chains with band structure, where periodic modulation of hopping amplitudes generates the band structure. The study demonstrates that entanglement entropy grows logarithmically with time, with the prefactor proportional to the number of filled quasiparticle bands in the initial state. For dimerized chains, particle density and current exhibit the same expressions as in the uniform case, requiring only velocity rescaling. The universal contribution to entropy distribution is doubled, while the non-universal part can be extracted numerically through block Toeplitz matrix analysis.

Research Background and Motivation

Problem Background

1. Domain-wall melting problem: Domain-wall melting is a classical setup for studying far-from-equilibrium dynamics in one-dimensional quantum chains, where a step-like initial density distribution is prepared and its time evolution is observed to understand transport properties.
2. Limitations of existing research:
   • Most studies focus on translationally invariant uniform chains
   • Limited understanding of cases where the time evolution operator itself breaks translational invariance
   • Lack of systematic understanding of entanglement propagation in systems with band structure
3. Research motivation:
   • Explore how periodic hopping amplitudes affect domain-wall melting dynamics
   • Understand the growth law of entanglement entropy in multi-band systems
   • Establish connections between systems with band structure and uniform systems

Core Contributions

1. Theoretical framework establishment: First systematic study of domain-wall melting in free-fermion chains with p-site translational invariance
2. Entanglement growth law: Demonstrates that entanglement entropy maintains logarithmic growth, with the prefactor proportional to the number of filled quasiparticle bands p
3. Hydrodynamic limit: For dimerized chains (p=2), density and current distributions can be obtained from the uniform case through simple velocity rescaling
4. Separation of universal and non-universal contributions:
   • Universal contributions can be obtained analytically and are doubled compared to the uniform case
   • Non-universal contributions require numerical calculation through block Toeplitz matrices
5. Analysis of oscillatory phenomena: Discovers and explains oscillations in entropy growth, with frequency determined by the band gap size

Detailed Methodology

Task Definition

Study the time evolution of domain-wall initial states in free-fermion chains with periodic hopping amplitudes, analyzing the dynamical behavior of particle density, current, and entanglement entropy.

Model Architecture

Dimerized hopping chain Hamiltonian

Ĥ = -∑_{m=-N/2+1}^{N/2} [(1+δ)/2 c†_{2m-1}c_{2m} + (1-δ)/2 c†_{2m}c_{2m+1} + h.c.]

Diagonalization process

1. Sublattice transformation: Introduce fermionic operators for two sublattices a_m = c_{2m-1}, b_m = c_{2m}
2. Fourier transformation:

   (a_m, b_m)^T = (1/√N) ∑_k e^{ikm} (a_k, b_k)^T
   

3. Band diagonalization: Obtain diagonal form through unitary transformation

   Ĥ = -∑_k ω_k(α†_k α_k - β†_k β_k)
   

   
   where the dispersion relation is:

   ω_k = √[cos²(k/2) + δ²sin²(k/2)]
   

Correlation Function Calculation

Hydrodynamic limit

In the limit m,n,t → ∞ with m/t, n/t fixed, the correlation functions are obtained through stationary phase approximation:

C_{m,n}(t) = ∫_{q_-}^{q_+} dq/(2π) e^{iq(2n-2m)} [(1, e^{iφ_q}); (e^{-iφ_q}, 1)]
           + ∫_{-q_+}^{-q_-} dq/(2π) e^{iq(2n-2m)} [(1, -e^{iφ_q}); (-e^{-iφ_q}, 1)]

where the Fermi points q_±(m,n,t) are determined by v_{q_±} = (m+n-1)/t.

Technical Innovations

1. Block matrix representation: Utilizing two-site translational invariance, correlation functions are described using 2×2 block matrices
2. Quasiparticle picture: The two integral terms correspond to independent contributions from upper and lower band quasiparticles
3. CFT-GHD combination: Combines conformal field theory methods with generalized hydrodynamics to handle entanglement propagation

Experimental Setup

Numerical computation parameters

• Chain length: N = 300 (open boundary conditions to avoid doubling in the front region)
• Time evolution: t = 200, 300
• Dimerization parameters: δ = 0.02, 0.05, 0.1, 0.2, 0.3, 0.5

Initial state setup

Domain-wall initial state: left half-chain completely filled, right half-chain completely empty:

C(0) = [1 0; 0 0]

Evaluation methods

1. Density and current: Comparison with analytical hydrodynamic results
2. Entanglement entropy: Calculated through eigenvalues of the correlation matrix
3. Scaling behavior: Verification of logarithmic growth law

Experimental Results

Main Results

1. Density and current distribution

• Density: ρ(n,t) = (1/π)arccos(n/(v_max t)), where v_max = 1-δ
• Current: J(n,t) = (v_max/π)√1-(n/(v_max t))²
• Verification: Numerical results perfectly match analytical expressions

2. Entanglement entropy growth

• Half-chain entropy: S ~ (1/3)ln t + const (doubled compared to the uniform case's 1/6 ln t)
• Universal contribution: S_u = (1/3)ln2v_max t(1-ζ²)
• Non-universal contribution: Requires numerical calculation through block Toeplitz matrices

3. Oscillatory phenomena

Discovers damped oscillations in entropy growth:

S - (1/3)ln t ≈ α + β cos(2δt + φ)/(δt)

The oscillation frequency exactly equals the band gap size 2δ.

Ablation Studies

δ → 1 limit

In the completely dimerized limit, analytically obtained:

lim_{δ→1} S_0(ζ) = (1/3)ln√(1-ζ²) + s_0

Three-band model

For three-site periodicity, verified that:

• Half-chain entropy: S = (1/2)ln t + const
• Entropy distribution exhibits multiple scaling regions, reflecting velocity differences of different bands

Case Analysis

Time scaling behavior

• t ≪ δ^{-1}: Behaves identically to uniform chains
• t ≫ δ^{-1}: Transitions to doubled logarithmic growth
• Crossover timescale determined by inverse band gap

Related Work

Domain-wall melting research

1. Uniform chains: Antal et al. first studied domain-wall melting in XX chains 10
2. Interacting systems: Extended to interacting cases through GHD methods 11
3. Entanglement studies: CFT methods for understanding entanglement propagation 21,22

Non-uniform systems

1. Local defects: Single impurities leading to linear entropy growth 38-44
2. Periodic modulation: First systematic study of periodic hopping amplitude effects in this work

Conclusions and Discussion

Main Conclusions

1. Universal scaling: Entanglement entropy maintains logarithmic growth with prefactor proportional to band number
2. Velocity rescaling: Hydrodynamic quantities can be obtained through simple velocity rescaling
3. Band gap effects: Produce characteristic oscillations with frequency determined by band gap

Limitations

1. Non-universal part: Requires numerical calculation, lacking analytical expressions
2. Edge structures: Difficult to analyze fine structures beyond stationary phase approximation
3. Interacting extensions: Dimerization breaks integrability, making GHD methods inapplicable

Future Directions

1. Arbitrary filling: Study non-fully-filled initial states
2. Edge physics: Analyze universal behavior in the t^{1/3} scaling region
3. Interacting systems: Explore whether effective hydrodynamic descriptions exist

In-depth Evaluation

Strengths

1. Theoretical completeness: Complete theoretical framework from Hamiltonian diagonalization to entanglement calculation
2. Methodological innovation: Cleverly combines CFT, GHD, and matrix analysis methods
3. Result universality: Establishes general laws for multi-band systems
4. Numerical verification: Analytical results verified with exact numerics

Weaknesses

1. Technical limitations: Lack of analytical results for block Toeplitz matrix constants
2. Physical interpretation: Deeper physical mechanisms of oscillatory phenomena need further clarification
3. Experimental accessibility: Experimental verification schemes for theoretical results not discussed

Impact

1. Theoretical contribution: Provides new perspective on dynamics of non-uniform quantum systems
2. Method value: Block Toeplitz methods generalizable to other systems
3. Application prospects: Provides theoretical guidance for quantum simulation experiments

Applicable Scenarios

1. Cold atom systems: Periodic hopping modulation realizable in optical lattices
2. Solid-state physics: Exciton dynamics in dimerized materials
3. Quantum information: Understanding entanglement propagation in multi-band systems

References

This paper cites 76 related references, covering:

• Generalized hydrodynamics theory 3-8
• Classical domain-wall melting work 9-17
• Entanglement entropy theory 18-27
• Conformal field theory methods 47-49
• Toeplitz matrix theory 53-55,72-76