Mean-field theory of the DNLS equation at positive and negative absolute temperatures

This paper presents a mean-field theory for the Discrete Nonlinear Schrödinger (DNLS) equation that utilizes a factorizable grandcanonical partition function approximation to accurately describe both the equilibrium homogeneous phase at positive temperatures and the metastable localized phase at negative temperatures, demonstrating excellent agreement with numerically exact results across the entire phase diagram.

The Gist

Imagine a crowded dance floor where everyone is dancing to a specific beat. In the world of physics, this dance floor is a grid of points (sites), and the dancers are waves of energy. The rules of their dance are governed by an equation called the Discrete Non-Linear Schrödinger (DNLS) equation.

For decades, physicists have been trying to understand how these dancers behave when the "temperature" of the room changes. Usually, heat makes things chaotic and spread out. But in this specific dance, there's a weird trick: if you crank the temperature up too high, the dancers suddenly stop spreading out and instead, they all clump together in one spot, forming a giant, intense "breather" (a localized wave).

Here is the problem: Standard physics tools (math) break down when trying to describe this clumping, especially when the temperature is "negative" (a concept where adding energy makes the system colder, leading to this clumping). The math gets messy, and the integrals (the tools used to sum up all possibilities) explode to infinity.

This paper introduces a new, clever way to solve this puzzle using Mean-Field Theory. Here is the breakdown in simple terms:

1. The Problem: The "Too Many Neighbors" Issue

Imagine you are trying to predict the mood of a party.

• The Hard Way: You look at every single person and how they are interacting with their immediate neighbors. If Person A is loud, it makes Person B loud, which makes Person C loud, and so on. This is the "exact" calculation. It's incredibly accurate but mathematically impossible to solve for a huge crowd because everyone is connected to everyone else in a complex web.
• The Old Shortcut (C2C Model): Previous scientists tried to simplify this by saying, "Let's just pretend everyone is dancing alone and ignore the neighbors." This was easy to calculate, but it failed miserably when the dancers actually started interacting. It was like predicting traffic flow by pretending cars don't bump into each other.

2. The Solution: The "Average Neighbor" Trick

The authors of this paper came up with a middle ground. They said:

"Instead of tracking exactly how Person A influences Person B, let's assume Person A is influenced by the average mood of the whole room."

In physics terms, they replaced the specific interaction between two neighbors with an interaction between a neighbor and the statistical average of the system.

• The Metaphor: Imagine you are at a party. Instead of worrying about exactly what your best friend is whispering to you, you just assume your friend is acting like the "typical" person at the party.
• Why it works: This trick breaks the complex web of connections. Suddenly, the math becomes "factorizable," meaning you can solve the problem for one person and then just multiply it by the number of people. It turns a tangled knot into a straight line.

3. The Results: A Map for the Whole Dance Floor

Using this "Average Neighbor" trick, the authors created a map (a phase diagram) that predicts exactly what the dancers will do.

• Positive Temperatures (The Hot, Chaotic Room): When the room is hot, the dancers are spread out evenly. The new theory predicts this perfectly, matching computer simulations almost exactly. It's like a perfect weather forecast for a sunny day.
• Negative Temperatures (The Clumping Room): When the room gets "negatively hot" (a weird state where energy is so high it forces order), the dancers usually collapse into a single giant clump (a breather).
  • The Catch: In reality, this clump takes a very long time to form. For a while, the dancers stay spread out in a "metastable" state (like a ball balanced on top of a hill; it wants to roll down, but it's stuck there for a long time).
  • The Breakthrough: The authors' theory is great at describing this "stuck" state. It gives a clear, simple formula for what the system looks like before it collapses. It's like having a perfect model of a ball sitting on a hill, even if we can't easily predict exactly when it will fall.

4. Why This Matters

Before this paper, we had two choices:

1. Exact Math: Too hard to solve.
2. Old Approximation: Too simple, ignored the neighbors, and gave wrong answers.

This new method is the "Goldilocks" solution. It's simple enough to write down on a napkin (giving explicit formulas) but accurate enough to match super-computer simulations.

The Big Takeaway:
The authors showed that even in a system where things are wildly complex and interacting, you can often get a near-perfect understanding by simply asking, "What is the average behavior of my neighbors?" This allows us to understand the transition from a chaotic, spread-out state to a clumped, ordered state, even in the strange realm of "negative temperatures."

In short: They found a way to simplify a chaotic dance floor without losing the rhythm, allowing us to predict exactly how the dancers will move, whether they are spread out or clumped together.

Technical Summary

1. Problem Statement

The Discrete Nonlinear Schrödinger (DNLS) equation is a fundamental model in statistical mechanics and solid-state physics, characterized by two conserved quantities: energy and mass (norm). A unique feature of the DNLS model is the existence of a phase transition between:

• A homogeneous phase at positive absolute temperatures ($T > 0$).
• A localized phase (characterized by the spontaneous emergence of "breathers" or high-amplitude peaks) at negative absolute temperatures ($T < 0$).

The Core Challenge:
While the thermodynamics of the DNLS model are well-understood in the microcanonical ensemble, the grandcanonical ensemble presents significant difficulties:

1. Positive $T$: The partition function involves non-factorizable integrals due to nearest-neighbor interactions, making analytical solutions unwieldy.
2. Negative $T$: The standard grandcanonical partition function is formally ill-defined because the integral over the mass distribution diverges (due to the unbounded growth of energy density).
3. Existing Approximations: Previous simplified models (like the C2C model) neglect inter-site interactions, failing to accurately describe the system away from the critical transition line ($T = \infty$).

The authors aim to develop a Mean-Field (MF) theory that provides explicit, analytical expressions for equilibrium observables across the entire phase diagram, including the metastable negative-temperature regime.

2. Methodology

The Mean-Field Approximation

The authors propose a specific MF approximation applied exclusively to the mass (amplitude) variables, while treating the phase variables exactly and retaining nearest-neighbor interactions.

• Hamiltonian: The original Hamiltonian is $H = \sum_n [c_n^2 + 2J\sqrt{c_n c_{n+1}} \cos(\phi_n - \phi_{n+1})]$, where $c_n$ is the local mass and $\phi_n$ is the phase.
• Approximation: The product of masses on neighboring sites, $\sqrt{c_n c_{n+1}}$, is replaced by the product of the local mass and the system-wide statistical average:
  $$\sqrt{c_n c_{n+1}} \approx \sqrt{c_n} \langle \sqrt{c} \rangle = \sqrt{c_n} q$$
  where $q \equiv \langle \sqrt{c} \rangle$ is a self-consistent order parameter.
• Resulting Hamiltonian: This transforms the Hamiltonian into a sum of single-site terms:
  $$H_{MF} = \sum_n [c_n^2 + 2Jq\sqrt{c_n} \cos(\phi_n - \phi_{n+1})]$$
  This factorization allows the grandcanonical partition function $Z$ to be written as $Z_{MF} = z^N$, where $z$ is the single-site partition function.

Handling Negative Temperatures

For $T < 0$ (where $\beta < 0$), the integral for the partition function diverges. The authors adopt a regularized grandcanonical approach based on the concept of metastable states:

• They introduce a cutoff $c^*$ in the mass integration, determined by the potential barrier of the effective thermodynamic potential.
• This allows the description of the homogeneous metastable state, which is stable on very long time scales before the system eventually collapses into a localized breather.

Analytical Expansion

The authors derive explicit formulas using a small parameter expansion $w = \beta/\mu^2$ (valid near the infinite temperature line $T=\infty$). They utilize asymptotic expansions of the Tricomi confluent hypergeometric function to obtain closed-form expressions for observables like mass density ($a$), energy densities ($h_{nl}, h_{int}$), and the order parameter ($q$).

3. Key Contributions

1. Factorizable Partition Function: The derivation of a factorizable grandcanonical partition function for the DNLS model, enabling analytical treatment of the interacting system.
2. Unified Description: A single theoretical framework that works for both positive temperatures (equilibrium) and negative temperatures (metastable homogeneous states).
3. Explicit Analytical Expressions: Provision of closed-form approximations for observables in the high-temperature limit (small $|w|$), valid on both sides of the critical line.
4. Resolution of the Chemical Potential Shift: Identification and rigorous explanation of a systematic shift in the chemical potential ($\mu \to \mu - 1$) in the MF theory at low temperatures, which is shown to be a rigorous property at $T=0$.
5. Energy Distribution Analysis: A quantitative breakdown of how total energy is distributed between the nonlinear term ($h_{nl}$) and the interaction term ($h_{int}$), introducing a ratio parameter $R$ to characterize this distribution.

4. Key Results

• Accuracy at Positive $T$:

  • The MF theory reproduces exact numerical results (obtained via grandcanonical Monte Carlo simulations) with "good-to-excellent" accuracy across the entire phase diagram.
  • It correctly predicts the ground state energy ($T=0$) and the infinite temperature line ($T=\infty$).
  • It significantly outperforms the C2C model (which neglects interactions), especially as temperature decreases.

• Accuracy at Negative $T$:

  • The theory successfully describes the metastable homogeneous state for small negative $\beta$.
  • It captures the transition line and the behavior of observables better than the C2C model, which fails to predict the correct energy-density relationships and exhibits unphysical re-entrant behavior at moderate negative temperatures.
  • The theory confirms that the interaction energy, while small near the critical line, is crucial for maintaining the stability of the homogeneous phase description.

• Observables and Scaling:

  • The ratio $R = (h_c - h_{nl}) / (h_c - h)$ is found to depend primarily on the parameter $m = \beta\mu$.
  • The analytical expression $R \approx 4 / (4 + \pi m^2/2)$ matches numerical simulations remarkably well.
  • At $T=0$, the MF theory predicts a specific relation between mass and chemical potential ($a = \mu + 1/2$), differing from the full DNLS model by a constant shift of 1 in $\mu$.

5. Significance

• Theoretical Advancement: This work represents a clear step forward from models that neglect site interactions (like C2C). It demonstrates that a mean-field approximation applied only to amplitudes (while keeping phases exact) is sufficient to capture the complex thermodynamics of the DNLS model.
• Metastability Insight: By successfully applying the grandcanonical ensemble to negative temperatures via regularization, the paper provides a consistent thermodynamic description of the metastable homogeneous phase, a state that is physically relevant because the relaxation to the true equilibrium (localized breather) is exponentially slow.
• Practical Utility: The derived explicit formulas allow researchers to quickly estimate equilibrium properties without resorting to computationally expensive numerical simulations, particularly useful for analyzing the transition region between positive and negative temperatures.
• Broader Implications: The methodology offers a template for studying other non-integrable systems with multiple conserved quantities and bounded energy spectra, bridging the gap between exact numerical methods and tractable analytical theories.