Finite-frequency conductivity of nonlinear Luttinger liquid in smooth random potential

Finite-frequency conductivity of nonlinear Luttinger liquid in smooth random potential analyzes how one-dimensional degenerate fermion systems behave under smooth random disorder, revealing that finite-frequency conductivity arises from two distinct contributions: an elastic part largely independent of temperature and interaction, and an inelastic part strongly dependent on temperature, frequency, curvature, disorder, and interaction.

The gist: The finite frequency conductivity consists of two parts: first one is the elastic contribution, largely independent of temperature and interaction. Second one is the inelastic contribution, strongly dependent on temperature and frequency and appearing upon simultaneous presence of curvature, disorder and interaction.

Model Setup

The analysis considers a one-dimensional degenerate liquid system with a finite mass in a sample characterized by smooth disorder with scale 'd', where the potential is assumed to be large-scaled such that 'd' is much larger than the interaction range 'a' and the Fermi wavelength, i.e., d >> a > λF. The system is described by a Hamiltonian involving terms for kinetic energy, random potential, and interactions, specifically considering nonlinear fermionic dispersion which corresponds to the decay of plasmons in bosonization.

The key length scales are defined as:

1. Thermal length: lT = vF /T

2. Frequency length: lω = ˜v/ω

The conditions for the liquid to be degenerate are assumed to be λF > a >> λF; lT >> λF. The problem is formulated using bosonization technique, where cubic terms in fermionic densities appear due to the quadratic curvature of the fermionic spectrum.

Conductivity Calculation

The electric conductivity is calculated using the Kubo formalism as the response function of the current, j(x, t), defined by Eq. (5). After calculating ∂tj(x, t) and taking the uniform limit (q = 0), linear-in-densities terms are omitted because they give a full derivative with respect to x. The non-vanishing part of the second term in ∂tj(x, t) is approximated as:

∂tj(x, t) = −1/mK2 ρ(x, t)∂xU(x).

This leads to the expression for conductivity:

σ(ω) = −e squared K-4 / (ħ m squared ω 3) Im⟨⟨(∂xUρ), (∂xUρ)⟩⟩q=0.

In the absence of random potential, there is no conductivity at ω ≠ 0. A nonuniform random potential breaks this property and leads to a nonzero real part of the conductivity, as shown by Eq. (7). The calculation in the lowest order in disorder potential magnitude yields:

σ(ω) = e squared K-4 / (ħ m squared ω 3) Z ∫dq / 2πq squared Wq Dq,ω, where Dq,ω = −Im⟨⟨ρ, ρ⟩⟩q,ω and Wq = Z∫dx e(-iqx)W(x).

Non-Interacting Gas Results

For a non-interacting Fermi gas in the limit of linear dispersion (1/m → 0), the density correlator is given by Eq. (12): Dq,ω = 1/2q [δ(ω - qvF) − δ(ω + qvF)]. In the case of nonlinear dispersion and zero temperature, the correlator is finite at all q and has a rectangular shape (Eq. 13).

When using the simplified delta function approximation for Dq,ω in Eq. (8), one obtains the quantum conductivity:

σ1(ω) = e squared Wq=ω/vF / 2πħ m squared K4 / ω vF 4.

This expression coincides with the classical result (Eq. 11) at ω ≠ 0, and it is independent of temperature when the condition d >> λF is satisfied. The one-boson contribution does not contain an explicit 1/ω2 frequency dependence but decreases with frequency since Wq decreases exponentially when ω/v >> 1/d.

Interacting Liquid Results

For the interacting case, the density spectral weight is divided into two parts: Dq,ω ≃ D(0) q,ω + D(1) q,ω. The Tomonaga-Luttinger part, D(0)q,ω (single-boson contribution), approximates the peak at qc ≃ ω/v by a delta function.

The correction term D(1)q,ω (double-boson contribution), which describes the long tails stretching far away from the point qc = ω/v, is crucial for conductivity. The total spectral density correction is given by Eq. (23).

The two main contributions to conductivity are:

As a fastidious researcher, I have thoroughly analyzed the provided scientific paper concerning the finite-frequency conductivity of a nonlinear Luttinger liquid in smooth random potential. The core findings revolve around decomposing conductivity into elastic and inelastic contributions, governed by temperature, frequency, and disorder strength.

Here are the specific improvements to AI systems based on this research:

1. Development of Novel Quantum Transport Models (Theory-to-Hardware Mapping)

The paper provides a rigorous framework for calculating transport properties in complex 1D quantum systems that go beyond simple linear response theory (like standard Drude models).

Specific Improvements:

The AI system can be trained on the mathematical structure derived from the bosonization technique, particularly Equation (23) and its subsequent approximations. This involves learning how to map microscopic interaction parameters (like the Luttinger parameter, curvature effects, and interaction range) onto macroscopic transport coefficients.

What the Improved AI System Can Do:

1. Predictive Material Simulation: The AI can simulate the finite-frequency conductivity of novel 1D electronic devices (e.g., carbon nanotubes or quantum wires) under realistic conditions involving both electron-electron interactions and smooth, large-scale disorder potentials.

2. Design of Dissipation Management Structures: By understanding the dominance of the inelastic contribution at high frequencies (Equation 26), the AI can design nanoscale structures that maximize dissipation in specific frequency bands or minimize it to achieve ballistic transport for desired signals.

3. Parameter Estimation: Given experimental measurements of optical conductivity spectra, the AI can invert Equation (49) to estimate hidden material parameters such as the disorder strength scale, interaction strength, and curvature effects within the system.

4. Advanced Signal Processing for Quantum Information Systems

The paper distinguishes between two distinct contributions to conductivity: the temperature-independent elastic part and the strongly temperature-dependent inelastic part. This distinction is crucial for characterizing noise in quantum devices.

Specific Improvements:

The AI system can be specialized in spectral decomposition of transport data. Instead of treating all frequency responses uniformly, it can automatically separate the measured signal into its elastic component (related to the zero-frequency delta peak) and its inelastic component (related to scattering processes).

What the Improved AI System Can Do:

1. Quantum Noise Filtering: In quantum computing or sensing applications, noise is a major enemy. The AI can analyze transport data (like current fluctuations) and identify whether the observed noise is due to elastic scattering (which is weakly temperature-dependent) or inelastic scattering (which scales strongly with temperature), allowing for real-time filtering of thermal vs. interaction-driven decoherence.

2. Characterization of Quantum Phase Transitions: Since the transition between the regimes dominated by elastic transport and those dominated by inelastic transport occurs at specific frequency scales, the AI can use this as a signature to detect subtle changes in material properties or phase transitions within 1D materials.

3. Enhanced Modeling of Non-Equilibrium Dynamics

The paper extensively discusses thermalization, relaxation, and the role of temperature in determining which physical mechanism (single quasiparticle motion vs. two-quasiparticle scattering) dominates transport.

Specific Improvements:

The AI system can be augmented with knowledge from kinetic equations (like Eq. 9 and 10) to model time-dependent transport, moving beyond the steady-state Kubo formalism used in Section V. It can learn to predict the relaxation timescales based on the ratio of characteristic lengths like the thermal length (Equation 1) versus other system scales.

What the Improved AI System Can Do:

1. Predictive Thermalization Modeling: The AI can simulate how a quantum system relaxes after a perturbation, predicting whether it will thermalize via single-particle processes or through collective, many-body inelastic scattering channels, providing insights into the fundamental limits of thermalization in low-dimensional systems.

2. Optimization of Dynamic Control Protocols: For systems where external fields are modulated at specific frequencies (e.g., in quantum gates), the AI can determine the optimal driving frequency to induce desired transport behavior by matching it to the characteristic relaxation timescales derived from Equations (26) and (29).

In summary, this research allows for an AI system that moves beyond simple pattern recognition into a realm of predictive, physics-informed modeling of low-dimensional quantum transport.

We analyze the uniform conductivity of a one-dimensional degenerate fermion system placed in a random disorder potential so smooth that backward scattering can be neglected. We use the nonlinear Luttinger liquid model to consider effects of both interaction and the curvature of fermionic dispersion. The finite frequency conductivity, calculated in the lowest order of disorder potential, consists of two parts. First one is the elastic contribution, largely independent of temperature and interaction. Second one is the inelastic contribution, strongly dependent on temperature and frequency and appearing upon simultaneous presence of curvature, disorder and interaction. We argue that apart from such finite frequency conductivity, there should always remain the δ-function peak of conductivity at zero frequency, whose weight is weakly dependent on the disorder.