Taking the temperature of quantum many-body scars

Basic Information

• Paper ID: 2503.21884
• Title: Taking the temperature of quantum many-body scars
• Authors: Phillip C. Burke (University College Dublin), Shane Dooley (Dublin Institute for Advanced Studies)
• Classification: quant-ph cond-mat.stat-mech
• Publication Date: March 31, 2025
• Paper Link: https://arxiv.org/abs/2503.21884

Abstract

Quantum many-body scars (QMBS) are eigenstates of chaotic many-body Hamiltonians that possess two seemingly contradictory properties: their energy eigenvalues correspond to high temperatures, yet their entanglement structure resembles low-temperature eigenstates (such as ground states). Conventionally, temperatures are assigned to energy eigenvalues through the canonical temperature-energy relation. This paper employs eigenstate subsystem temperature—a novel quantity that assigns temperatures to energy eigenstates based on the structure of the reduced density matrix. For thermal states, eigenstate subsystem temperature approximately equals canonical temperature. Given that QMBS possess ground-state-like entanglement structures, it is not obvious whether their eigenstate subsystem temperature would be close to canonical temperature. Surprisingly, the study finds that this is indeed the case: QMBS encode approximate "knowledge" of their position in the spectrum within their state structure.

Research Background and Motivation

Core Problem

The core problem addressed by this research is: whether the eigenstate subsystem temperature of quantum many-body scar states with ground-state-like entanglement structure is consistent with the canonical temperature based on energy eigenvalues?

Problem Significance

1. Understanding thermalization mechanisms: How quantum many-body systems thermalize and when this fails is a fundamental question in quantum many-body physics
2. Challenge to ETH: QMBS violate the eigenstate thermalization hypothesis (ETH), providing new perspectives for understanding non-thermalization phenomena
3. Extension of temperature concept: Traditional temperature definitions are based on energy, while eigenstate subsystem temperature is based on state structure; the relationship between them requires clarification

Limitations of Existing Methods

1. Traditional temperature definition: Based solely on energy eigenvalues through the canonical temperature-energy relation
2. ETH framework: For QMBS that violate ETH, the traditional framework cannot adequately describe their thermodynamic properties
3. Lack of state structure information: Traditional methods neglect the influence of internal eigenstate structure on temperature

Research Motivation

To explore whether QMBS retain information about their position in the energy spectrum within their state structure, despite possessing non-thermal entanglement structures.

Core Contributions

1. First systematic study: First systematic investigation of the relationship between eigenstate subsystem temperature and canonical temperature for QMBS
2. Important finding: Discovery that eigenstate subsystem temperature of QMBS is highly correlated with canonical temperature, despite their non-thermal entanglement structure
3. Statistical analysis: Provides statistically reliable results through ~7000 randomly generated Hamiltonians
4. Theoretical insight: Reveals the mechanism by which QMBS encode spectral position information in their state structure

Detailed Methodology

Task Definition

Input: Eigenstate |E⟩ of a chaotic many-body Hamiltonian Output: Difference δβ(E) between eigenstate subsystem temperature βS(|E⟩) and canonical temperature βC(E) Objective: Analyze whether QMBS maintain the relationship βS ≈ βC

Core Concepts

1. Eigenstate Subsystem Temperature

Eigenstate subsystem temperature βS(|E⟩) is defined as the temperature that minimizes the distance between the reduced eigenstate density matrix and the reduced canonical density matrix:

βS(|E⟩) = argminβ [d1(ρ̂S, σ̂S(β))]

Where:

• ρ̂S = TrS̄|E⟩⟨E| is the reduced eigenstate density matrix
• σ̂S(β) = TrS̄σ̂(β) is the reduced canonical density matrix
• d1(X̂,Ŷ) = ||X̂-Ŷ||1 is the trace distance

2. Canonical Temperature

Canonical temperature βC is determined through the standard canonical temperature-energy relation:

E = Tr[σ̂(βC)Ĥ] = Σ(E'e^(-βCE'))/Σ(e^(-βCE'))

Model Construction

QMBS Model

QMBS models are constructed using the Shiraishi-Mori projection embedding method:

Ĥ = Σ(n=0 to N-1) P̂n,n+1 ĥn,n+1 P̂n,n+1

Where:

• P̂ = 1̂⊗1̂ - |0⟩⟨0|⊗|0⟩⟨0| is the two-spin projection operator
• ĥ is a two-spin Hermitian matrix
• The product state |EQMBS⟩ = |0⟩⊗N is a zero-energy eigenstate

Random Model Generation

Two-spin Hamiltonian terms ĥ are randomly generated from the Gaussian unitary ensemble (GUE), ensuring the chaotic nature of the system.

Technical Innovations

1. Statistical methods: Provide statistically reliable results through numerous random realizations
2. Multi-scale analysis: Investigate the influence of system size N on results
3. Comparative analysis: Simultaneously analyze the behavior differences between QMBS and thermal states
4. Multiple metrics: Employ multiple correlation measures (variance, Pearson coefficient, spectral fraction)

Experimental Setup

Dataset

• System size: N = 12-18 spin-1/2 particles
• Random realizations: ~7000 randomly generated Hamiltonians
• Symmetry: k = 0 momentum sector
• Subsystem: Two adjacent spins selected as subsystem S

Evaluation Metrics

1. Temperature difference: δβ(E) = βS(|E⟩) - βC(E)
2. Statistical variance: ⟨(δβ)²⟩
3. Pearson correlation coefficient: corr = cov(βC,βS)/(std(βC)std(βS))
4. Trace distance: d1(ρ̂S, σ̂S(β))

Comparison Methods

• QMBS states: |EQMBS⟩ = |0⟩⊗N
• Thermal states: Neighboring thermal eigenstates with minimum positive energy
• Random models vs specific models (XXZ chain)

Implementation Details

• Exact diagonalization used to solve eigenvalues and eigenstates
• Chaotic nature of the system verified through level spacing statistics
• QMBS energy required to be located in the central half of the spectrum

Experimental Results

Main Results

1. Temperature Correlation of QMBS

• Core finding: βS of QMBS is highly correlated with βC, despite non-thermal entanglement structure
• Statistical characteristics: Mean of δβ distribution ⟨δβ⟩ ≈ 0, exhibiting Gaussian distribution
• Correlation strength: Pearson correlation coefficient indicates strong positive correlation

2. Comparison with Thermal States

• Thermal state performance: Stronger correlation between βS and βC for thermal states
• Distribution differences: δβ distribution of thermal states exhibits exponential decay; QMBS exhibits Gaussian distribution
• Convergence behavior: |δβ| of thermal states decays with system size N; QMBS remains constant

3. System Size Dependence

• QMBS: Statistical variance ⟨(δβ)²⟩ decreases slowly with N, exhibiting even-odd oscillations
• Thermal states: Variance clearly decreases with N, consistent with ETH predictions
• Distance measure: min(d1) of QMBS remains approximately 1.5; thermal states approach 0

Ablation Studies

Comparison of Different Models

• Random models: ~7000 GUE-generated Hamiltonians
• XXZ model: Specific transverse-field XXZ interactions
• Consistency: Both models display similar behavioral patterns

Diversity of Correlation Measures

1. Variance analysis: ⟨(δβ)²⟩ quantifies correlation tightness
2. Pearson coefficient: Dimensionless correlation measure
3. Spectral fraction: Dimensionless representation of eS vs eC

Key Findings

1. Unexpected correlation: Despite QMBS violating ETH and possessing ground-state-like structure, their βS remains correlated with βC
2. Encoding mechanism: QMBS encode "knowledge" of spectral position within their state structure
3. Robustness: This correlation exists across different models and system sizes
4. Distinguishing feature: Convergence behavior of δβ can distinguish QMBS from thermal states

Related Work

Major Research Directions

1. QMBS theory: First discovery of QMBS phenomena by Turner et al.
2. ETH framework: Thermalization theory established by Deutsch, Srednicki, and others
3. Eigenstate temperature: Recently proposed eigenstate subsystem temperature concept by Burke et al.

Contributions of This Work

1. First application: First application of eigenstate subsystem temperature to QMBS research
2. Systematic analysis: Provides large-scale statistical analysis
3. Theoretical extension: Extends understanding of temperature concepts for non-thermal states

Conclusions and Discussion

Main Conclusions

1. Core conclusion: QMBS maintain the relationship βS ≈ βC, indicating they encode spectral position information within their state structure
2. Mechanistic insight: Despite violating ETH, QMBS retain a certain form of "thermodynamic memory"
3. Universality: This phenomenon exists across multiple chaotic models

Limitations

1. Single QMBS type: Only studies QMBS of the form |0⟩⊗N
2. System size constraints: Limited by exact diagonalization to maximum system size N=18
3. Specific subsystem choice: Only considers subsystems of two adjacent spins

Future Directions

1. Extended QMBS types: Study QMBS in MPS form and PXP models
2. Many-body localization: Explore similar phenomena in MBL systems
3. Integrable systems: Investigate eigenstate temperature behavior in integrable systems
4. Thermodynamic entropy definition: Explore temperature definitions based on entanglement entropy

In-Depth Evaluation

Strengths

1. Strong innovation: First systematic study of eigenstate temperature of QMBS, providing new theoretical perspective
2. Rigorous methodology: Large-scale statistical analysis provides reliable results; multiple correlation measures ensure robust conclusions
3. Unexpected findings: Reveals unexpected temperature correlation in QMBS, challenging existing understanding
4. Technical refinement: Precise numerical methods and rigorous statistical analysis

Weaknesses

1. Insufficient theoretical explanation: Lacks deep theoretical explanation for why QMBS maintain temperature correlation
2. Model limitations: Primarily based on specific projection embedding models; universality requires further verification
3. Size constraints: Relatively small system sizes; thermodynamic limit behavior insufficiently clear
4. Subsystem dependence: Insufficient exploration of subsystem choice effects on results

Impact

1. Theoretical contribution: Provides new tools for understanding thermodynamic properties of QMBS
2. Methodological value: Successful application of eigenstate subsystem temperature concept
3. Inspirational significance: May inspire research on other non-thermal state systems
4. Experimental guidance: Provides theoretical predictions for experimental verification of QMBS properties

Applicable Scenarios

1. Theoretical research: Study of non-thermalization mechanisms in quantum many-body systems
2. Numerical simulation: Thermodynamic analysis of finite-size quantum systems
3. Experimental physics: Verification of QMBS properties in cold atom and ion trap platforms
4. Quantum technology: Coherent state protection in quantum simulation and quantum computing

References

46 references covering important works in QMBS theory, ETH framework, statistical mechanics, and related fields

This paper makes important contributions to the field of quantum many-body physics, revealing hidden thermodynamic structures in QMBS through ingenious numerical experiments and providing new perspectives for understanding non-thermalization phenomena. Despite certain limitations, its innovation and rigor make it an important advance in the field.