[Paper Review] Classical scattering at low energies

Jan Dereziński, Erik Skibsted|ArXiv.org|Apr 18, 2006

Spectral Theory in Mathematical Physics22 references3 citations

TL;DR

This paper develops a classical scattering theory for negative, slowly decaying potentials—including the attractive Coulomb potential—at low energies. It constructs a continuous family of scattering orbits parametrized by initial position, final escape direction, and energy, proving these orbits form smooth, Lagrangian manifolds, which provides a foundation for low-energy quantum scattering theory via Fourier integral operators.

ABSTRACT

For a class of negative slowly decaying potentials including the attractive Coulombic one we study the classical scattering theory in the low-energy regime. We construct a (continuous) family of classical orbits parametrized by initial position $x\in \R^d$, final direction $ω\in S^{d-1}$ of escape (to infinity) and the energy $λ\geq 0$, yielding a complete classification of the set of outgoing scattering orbits. The construction is given in the outgoing part of phase-space (a similar construction may be done in the incoming part of phase-space). For fixed $ω\in S^{d-1}$ and $λ\geq 0$ the collection of constructed orbits constitutes a smooth manifold that we show is Lagrangian. The family of those Lagrangians can be used to study the quantum mechanical scattering theory in the low-energy regime for the class of potentials considered here. We devote this study to a subsequent paper.

Motivation & Objective

• To establish a rigorous classical scattering framework for slowly decaying negative potentials at low energies, including the Coulomb potential.
• To classify all outgoing scattering orbits via a mixed initial-boundary value problem involving energy, direction, and initial position.
• To prove that the constructed family of orbits forms smooth, Lagrangian submanifolds in phase space, enabling quantization.
• To lay the groundwork for subsequent quantum mechanical low-energy scattering analysis using the classical solutions as oscillatory integral phases.

Proposed method

• Formulates a mixed problem (1.3) coupling Newton's equation, energy conservation, and asymptotic direction conditions for orbits escaping to infinity.
• Uses the implicit function theorem for spherically symmetric potentials and a perturbation/Taylor expansion approach for non-symmetric cases.
• Applies a fixed-point argument to solve the nonlinear mixed problem in the non-symmetric case, relying on a priori estimates and the uncertainty principle (Hardy inequality).
• Constructs phases $\phi^{\pm}(x,\omega,\lambda)$ via $\nabla_x \phi^{\pm} = \dot{y}(\pm 1)$, which are continuous but not smooth in $\lambda$ at $\lambda=0$.
• Employs asymptotic analysis and weighted $L^2$ estimates to control the behavior of solutions at infinity, particularly for $\lambda=0$.
• Uses limiting absorption principles and matrix estimates (e.g., $q$-matrix) to verify solution regularity in weighted $L^2$ spaces for $t \geq 1$.

Experimental results

Research questions

• RQ1 How can classical scattering orbits be parametrized continuously in initial position, final direction, and energy $\lambda \geq 0$ for slowly decaying potentials?
• RQ2 What is the regularity and geometric structure (e.g., Lagrangian) of the set of scattering orbits in phase space at low energy?
• RQ3 How do the solutions to the mixed problem (1.3) behave at $\lambda = 0$, and what is the nature of the singularity in the $\lambda$-dependence?
• RQ4 Can the classical solutions be used to construct oscillatory integral operators for quantum scattering at threshold energy $\lambda = 0$?
• RQ5 What conditions on the potential ensure existence of asymptotic normalized velocities and well-defined scattering at zero energy?

Key findings

• For potentials $V(x) = -\gamma |x|^{-\mu} + O(|x|^{-\mu-\epsilon})$ with $\mu \in (0,2)$, classical scattering orbits exist with well-defined asymptotic directions $\omega^{\pm}$ for all $\lambda \geq 0$.
• The constructed family of orbits forms a smooth, Lagrangian submanifold in phase space for each fixed $\omega \in S^{d-1}$ and $\lambda \geq 0$.
• The phase functions $\phi^{\pm}(x,\omega,\lambda)$ are jointly continuous in $x, \omega, \lambda$, but fail to be $C^1$ in $\lambda$ at $\lambda = 0$.
• The solution to the mixed problem (1.3) is shown to exist and be regular in a weighted $L^2$ sense for large $t$, with $\tilde{z} \in L^2_{-\tilde{s}}(1,\infty)$ for some $\tilde{s} < 1 + \bar{\epsilon}_1/2$.
• For $\lambda = 0$, the solution $\tilde{z}$ satisfies $\tilde{z} = O(t^{\kappa})$ with $\kappa < 1/2$, ensuring integrability in the required weighted space.
• The fixed-point argument and estimates confirm that the solution $\tilde{z}$ to the eikonal-type equation (7.10) is well-defined and lies in the appropriate function space, validating the construction.