2025-11-12T11:52:10.501194

Large deviation principles for the Gross Pitaevskii Gibbs measure at low temperature

Packer, Seong, Sosoe
We prove the large deviation principle for the conditional Gibbs measure associated with the focusing Gross Pitaevskii equation in the low temperature regime. This conditional measure is of mixed type, being canonical in energy and microcanonical in particle number. In particular, our result extends the large deviation principle for the mixed ensemble studied by Ellis, Jordan, Otto, and Turkington to a more singular setting, where the interaction potential is unbounded and the conditional event involves diverging renormalization constants. As a consequence of the large deviation principle, the Gibbs measure concentrates along the soliton manifold in the low temperature limit.
academic

Large deviation principles for the Gross Pitaevskii Gibbs measure at low temperature

Basic Information

  • Paper ID: 2510.13206
  • Title: Large deviation principles for the Gross Pitaevskii Gibbs measure at low temperature
  • Authors: Liam Packer, Kihoon Seong, Philippe Sosoe (Cornell University)
  • Classification: math.PR (Probability Theory), math-ph (Mathematical Physics), math.AP (Analysis of Partial Differential Equations), math.MP (Mathematical Physics)
  • Publication Date: October 15, 2025
  • Paper Link: https://arxiv.org/abs/2510.13206

Abstract

This paper establishes a large deviation principle for the conditional Gibbs measure associated with the focusing Gross-Pitaevskii equation in the low-temperature limit. The conditional measure is of mixed type, canonical in energy and microcanonical in particle number. In particular, our results extend the large deviation principle for mixed ensembles studied by Ellis, Jordan, Otto, and Turkington to a more singular setting where the interaction potential is unbounded and the conditioning event involves divergent renormalization constants. As a consequence of the large deviation principle, the Gibbs measure concentrates along the soliton manifold in the low-temperature limit.

Research Background and Motivation

Problem Background

  1. Importance of the Gross-Pitaevskii equation: The Gross-Pitaevskii equation is a fundamental model describing Bose-Einstein condensation with significant physical importance. The equation takes the form: i∂tψ+∂x2ψ−∣x∣2ψ−λ∣ψ∣2ψ=0i\partial_t\psi + \partial^2_x\psi - |x|^2\psi - \lambda|\psi|^2\psi = 0 where ∣x∣2|x|^2 is the harmonic potential and λ∣ψ∣2ψ\lambda|\psi|^2\psi is the focusing interaction.
  2. Challenges in constructing Gibbs measures: In the focusing case (λ>0\lambda > 0), the standard canonical Gibbs ensemble is not normalizable because the partition function diverges. Therefore, one must employ a mixed ensemble, namely a conditional probability measure that is canonical in energy and microcanonical in particle number.
  3. Technical difficulties: Since the Gaussian measure of the harmonic oscillator is not supported on L2(R)L^2(\mathbb{R}) but rather on H−η(R)∩Lp(R)H^{-\eta}(\mathbb{R}) \cap L^p(\mathbb{R}) (p > 2), the L2L^2 mass is almost surely infinite, necessitating Wick renormalization.

Research Motivation

  1. Extension of existing theory: The work of Ellis et al. only addresses bounded interaction potentials and non-singular conditioning; this paper extends to unbounded quartic interactions and singular conditioning involving divergent renormalization constants.
  2. Physical significance: Understanding the macroscopic behavior of the system in the low-temperature limit, particularly how the measure concentrates near the soliton manifold, is crucial for understanding the long-time dynamical behavior of Bose-Einstein condensates.

Core Contributions

  1. Establishment of large deviation principle: Successfully establishes a large deviation principle for the mixed Gibbs ensemble in the singular setting where the interaction potential is unbounded and the conditioning event involves divergent renormalization constants.
  2. Proof of measure concentration: Demonstrates that in the low-temperature limit, the Gibbs measure concentrates at an exponential rate near the soliton manifold, meaning that ground state solitons are the most probable macroscopic states.
  3. Treatment of renormalization effects: A key technical contribution is proving that divergent renormalization constants vanish at the level of the large deviation rate function, so that macroscopic behavior is determined by a clean variational problem without renormalization effects.
  4. Extension of theoretical framework: Extends large deviation theory from bounded to unbounded potentials, and from non-singular to singular conditioning.

Methodology Details

Problem Formulation

Study the large deviation behavior of the mixed Gibbs measure: dρε,rD(ϕ)=ρε,A{ϕ∈B∣Mw(ϕ)∈[D−r,D+r]}d\rho^D_{\varepsilon,r}(\phi) = \rho_{\varepsilon,A}\{\phi \in B | M_w(\phi) \in [D-r, D+r]\} where Mw(ϕ)=∫R:∣ϕ∣2:dxM_w(\phi) = \int_{\mathbb{R}} :|\phi|^2: dx is the Wick-renormalized L2L^2 mass.

Model Architecture

1. Construction of the mixed ensemble

  • Grand canonical ensemble: First construct the grand canonical Gibbs measure dρε,A(ϕ)=Zε,A−1exp⁡{−1εHG(ϕ)}με(dϕ)d\rho_{\varepsilon,A}(\phi) = Z^{-1}_{\varepsilon,A} \exp\left\{-\frac{1}{\varepsilon}H^G(\phi)\right\}\mu_\varepsilon(d\phi) where HGH^G is the grand canonical Hamiltonian including chemical potential AA.
  • Conditional probability: Define the mixed ensemble via conditional probability under the Wick-renormalized mass constraint.

2. Large deviation rate function

H(\phi) - \inf_{M(\phi)=D} H(\phi) & \text{if } \phi \in H^1(\mathbb{R}) \text{ and } M(\phi) = D \\ \infty & \text{otherwise} \end{cases}$$ ### Technical Innovations #### 1. Gamma convergence method Employ Gamma convergence theory to handle the limit of variational problems: $$F_{V+f,\varepsilon}(u) \to F_{V+f,0}(u)$$ where renormalization terms vanish in the limiting process. #### 2. Asymptotic analysis of thermodynamic functions - **Microcanonical entropy**: $\lim_{r \to 0} \lim_{\varepsilon \to 0} \varepsilon \log \rho_{\varepsilon,A}(\{M_w(\phi) \in [D-r, D+r]\}) = -\inf_{M(\phi)=D} H^G(\phi)$ - **Free energy**: $\lim_{\varepsilon \to 0} \varepsilon \log Z_{\varepsilon,A} = -\inf_{\phi \in H^1} H^G(\phi)$ #### 3. Variational representation Utilize the Boué-Dupuis variational formula to represent the Laplace transform of the Gibbs measure as a stochastic control problem: $$-\varepsilon \log E_{\mu}[e^{-\frac{1}{\varepsilon}F(\sqrt{\varepsilon}\phi)}] = \inf_{u \in H_a} E[F(\varepsilon^{1/2}W + Z(u)) + \frac{1}{2}\int_0^1 \|u(t)\|^2_{L^2}dt]$$ ## Experimental Setup ### Theoretical Verification Framework This is a purely theoretical work, with results verified primarily through rigorous mathematical proofs: 1. **Function space setting**: Establish large deviation principle on $S = H^{-\eta}(\mathbb{R})$ or $L^p(\mathbb{R})$ (p > 2) 2. **Parameter conditions**: Require $D \geq D_*$, where $D_*$ ensures the minimum energy is negative 3. **Chemical potential choice**: $A \geq A_0$ sufficiently large to ensure coercivity of the grand canonical Hamiltonian ### Main Theorems **Theorem 1.2 (Large Deviation Principle)**: There exists $D_* > 0$ such that for all $D \geq D_*$, the mixed ensemble $\rho^D_{\varepsilon,r}$ satisfies a large deviation principle: - Upper bound: $\limsup_{r \to 0} \limsup_{\varepsilon \to 0} \varepsilon \log \rho^D_{\varepsilon,r}(C) \leq -\inf_{\phi \in C} J^D(\phi)$ - Lower bound: $\liminf_{r \to 0} \liminf_{\varepsilon \to 0} \varepsilon \log \rho^D_{\varepsilon,r}(O) \geq -\inf_{\phi \in O} J^D(\phi)$ **Theorem 1.3 (Measure Concentration)**: For any $\delta > 0$, there exists $c(\delta) > 0$ such that $$\rho^D_{\varepsilon,r}\left(\left\{\inf_{Q \in M^D} \|\phi - Q\|_{L^p} \geq \delta\right\}\right) \lesssim e^{-\frac{c(\delta)}{\varepsilon}}$$ ## Experimental Results ### Main Results 1. **Establishment of large deviation principle**: Successfully proves the large deviation principle for the mixed Gibbs ensemble in the singular setting, with rate function $J^D(\phi)$ and speed $\varepsilon > 0$. 2. **Vanishing of renormalization effects**: Demonstrates that although divergent renormalization constants appear at the measure level, these effects vanish at the level of the large deviation rate function, with macroscopic behavior determined by the clean variational problem $\inf_{M(\phi)=D} H(\phi)$. 3. **Exponential concentration phenomenon**: Proves that the measure concentrates at exponential rate $e^{-c(\delta)/\varepsilon}$ near the soliton manifold $M^D$, where $M^D$ is the family of minimizers of the Hamiltonian $H$ under the constraint $M(\phi) = D$. ### Technical Achievements 1. **Application of Gamma convergence**: Successfully applies Gamma convergence methods to the unbounded potential case, overcoming technical difficulties. 2. **Control of thermodynamic limits**: Precisely controls the order of limits $\varepsilon \to 0$ and $r \to 0$, establishing the asymptotic behavior of thermodynamic functions. 3. **Equivalence of variational problems**: Proves that the complex stochastic control problem is equivalent to the classical variational problem in the limit sense. ## Related Work ### Main Related Research 1. **Ellis-Jordan-Otto-Turkington (2004)**: Established the large deviation principle for mixed ensembles with bounded potentials, but required $\sum \frac{1}{\lambda_k} < \infty$, which does not apply to the harmonic oscillator case. 2. **Burq-Thomann-Tzvetkov (2013)**: Constructed Gibbs measures for the focusing Gross-Pitaevskii equation and proved their invariance under the deterministic flow. 3. **Bourgain (1994, 1997)**: Pioneering work on invariant measures for nonlinear Schrödinger equations. 4. **Lebowitz-Rose-Speer (1988)**: Foundational work on the statistical mechanics of nonlinear Schrödinger equations. ### Innovations of This Paper 1. **Breakthrough of technical limitations**: Handles the case $\sum \frac{1}{\lambda_n} = \infty$, which is inevitable in the harmonic oscillator setting. 2. **Treatment of singular conditioning**: Successfully addresses conditioning events involving divergent renormalization constants. 3. **Extension to unbounded potentials**: Extends the theory from bounded to unbounded quartic potentials. ## Conclusions and Discussion ### Main Conclusions 1. **Large deviation principle**: In the low-temperature limit, the mixed Gibbs measure satisfies a large deviation principle with rate function given by a constrained variational problem. 2. **Physical significance**: Ground state solitons are the most probable macroscopic states, with other configurations decaying at exponential rate. 3. **Dynamical implications**: Since the Gibbs measure is invariant under the Gross-Pitaevskii flow, the concentration phenomenon implies that long-time dynamical behavior can be approximated as ground state solitons plus small fluctuations. ### Limitations 1. **Large mass condition**: Requires $D \geq D_*$ to ensure negative minimum energy; treatment of small mass cases remains an open problem. 2. **One-dimensional restriction**: Results apply only to the one-dimensional case; higher-dimensional extensions face additional difficulties. 3. **Harmonic potential dependence**: The presence of the harmonic potential is essential for the results; in the translation-invariant case, the measure becomes the trivial $\delta_0$. ### Future Directions 1. **Higher-dimensional extensions**: Extend results to higher dimensions. 2. **Small mass regime**: Study behavior when $D < D_*$. 3. **Dynamical applications**: Utilize concentration results to study effective descriptions of long-time dynamics. ## In-Depth Evaluation ### Strengths 1. **Technical innovation**: Successfully extends large deviation theory to new singular settings with advanced techniques. 2. **Mathematical rigor**: Complete and rigorous proofs that address multiple technical challenges. 3. **Physical relevance**: Results have important physical significance for understanding Bose-Einstein condensation. 4. **Method generality**: Developed techniques may apply to other similar singular probability measures. ### Weaknesses 1. **Limited applicability**: Results require restrictive conditions, necessitating large mass and harmonic potential. 2. **Computational complexity**: Proof techniques are complex and may be difficult to generalize to broader settings. 3. **Lack of numerical verification**: Absence of numerical experiments supporting the theoretical results. ### Impact 1. **Theoretical contribution**: Makes important contributions to large deviation theory for singular Gibbs measures. 2. **Interdisciplinary value**: Connects probability theory, partial differential equations, and mathematical physics. 3. **Methodological significance**: Provides new technical frameworks for handling unbounded potentials and singular conditioning. ### Applicable Scenarios 1. **Mathematical physics**: Study of Gibbs measures in quantum field theory. 2. **Probability theory**: Large deviation theory for singular probability measures. 3. **Partial differential equations**: Investigation of statistical properties of nonlinear Schrödinger equations. ## References The paper cites 37 important references, including: - Large deviation theory for mixed ensembles by Ellis, Jordan, Otto, and Turkington - Gibbs measure construction by Burq, Thomann, and Tzvetkov - Invariant measure theory by Bourgain - Related literature on variational methods and Gamma convergence