A perturbation theory for the Coulomb-phase infrared divergence

Technical Summary: A Perturbation Theory for the Coulomb-Phase Infrared Divergence

Problem Statement
Infrared (IR) divergences in quantum field theories with massless particles (such as QED and gravity) in four or fewer dimensions pose significant challenges for the definition of scattering amplitudes, causality, and unitarity. While inclusive cross-sections can often be rendered finite by summing over soft radiative processes, the amplitudes themselves remain divergent or ambiguous in standard perturbation theory. A specific class of these divergences, the "Coulomb-phase" (or imaginary) divergence, arises because the asymptotic states of particles interacting via long-range $1/r$ potentials do not converge to free-particle plane waves. Instead, they follow hyperbolic trajectories with logarithmic corrections. Standard perturbation theory, which assumes asymptotically non-interacting free states (plane waves), fails to account for this mismatch, leading to IR divergences at every order of the coupling expansion. While the Faddeev-Kulish (FK) framework offers a solution by modifying evolution operators (Møller operators) to include "dressed" states, explicit calculations of IR-finite amplitudes within this framework are rare and technically challenging.

Methodology
The author proposes a novel perturbation theory designed to be manifestly free of Coulomb-phase IR divergences. The core methodology involves a fundamental shift in the basis of wavefunctions used for perturbative expansion:

1. Basis Replacement: Instead of expanding around plane waves ($e^{-ip\cdot x}$), the theory expands around exact Coulomb wavefunctions. These wavefunctions incorporate the long-range $1/r$ potential to all orders in the coupling, ensuring their asymptotics match the exact in/out states of the system. The replacement is formalized as:
   $$e^{-ip\cdot x} \to e^{-ip\cdot x} {}_1F_1\left(-i\frac{e_1 e_2}{4\pi}\frac{E}{|\vec{p}|}, 1, i(|\vec{p}||\vec{x}| - \vec{p}\cdot\vec{x})\right)$$
   where ${}_1F_1$ is the confluent hypergeometric function.
2. Propagator Modification: The free-field propagator is replaced by a relativistic Coulomb Green's function. This Green's function is constructed via a spectral decomposition over the eigenstates of the "semi-free" Hamiltonian (containing the linear $A_\mu \partial^\mu$ term but excluding the quartic $A^2 \phi^2$ term).
3. Retarded Propagator: To isolate the Coulomb-phase divergence from real-emission divergences, the author utilizes a retarded propagator for the background gauge field. This effectively removes the real-photon production term ($R$ in the soft exponentiation theorem) from the calculation, allowing a focused study of the imaginary phase divergence.
4. Framework: The theory is applied to the semiclassical scattering of a spinless boson (Klein-Gordon field) on a fixed background Coulomb field generated by a source particle. The scattering amplitude is defined as the inner product of exact in/out solutions to the Klein-Gordon equation in this background.

Key Contributions and Results
The paper computes the scattering amplitude up to Next-to-Leading Order (NLO) within this new framework:

• Leading Order (LO): The LO amplitude is calculated as the inner product of the relativistic Coulomb wavefunctions. The result is an IR-finite expression that exhibits a power-law behavior $t^{-(1+i\gamma)}$ (where $t$ is the Mandelstam variable and $\gamma$ is the cusp anomalous dimension). The author demonstrates that this amplitude is free of any implicit or explicit IR cutoffs or dimensional regularization scales.
• Next-to-Leading Order (NLO): The NLO amplitude includes corrections from the quartic vertex ($A^2 \phi^2$) via the Coulomb Green's function. The calculation involves integrating over intermediate continuum and bound states. The resulting expression is also IR-finite and is expressed in terms of harmonic numbers and hypergeometric functions.
• Agreement with Exact Solution: The author compares their perturbative LO and NLO results against the known exact non-perturbative amplitude (derived via partial-wave expansion). They find perfect agreement up to the computed orders ($O(\alpha^0)$ and $O(\alpha^2)$), validating the conjecture that this perturbation theory is IR-finite.
• Runge-Lenz Symmetry: The LO amplitude is shown to possess a Runge-Lenz symmetry. Under this symmetry, the scattering states furnish a principal-series representation of the Euclidean conformal group $SO(1,3)$ on the 2-sphere. The amplitude transforms as a 2-point correlation function of scalar conformal primary operators with conformal dimension $\Delta = 1 + i\gamma$.

Significance and Claims
The paper claims to provide a manifestly Lorentz-invariant generalization of the distorted-wave Born approximation that resolves the Coulomb-phase IR divergence without introducing cutoffs.

• Alternative to Faddeev-Kulish: The author argues that their approach is complementary to the Faddeev-Kulish framework. While FK modifies the evolution operators acting on free states, this approach modifies the states themselves to match the asymptotics of the exact solution. They argue that the FK Møller operators effectively act as projection operators mapping free plane waves to these Coulomb wavefunctions.
• Simplicity and Rigor: The method yields IR-finite amplitudes that are unambiguously defined and independent of the perturbative framework used to compute them. The author notes that the exact amplitude is surprisingly simple compared to standard perturbative expansions, and they hope this simplicity extends to more complex relativistic settings.
• Future Directions: The paper suggests that while the current work focuses on a fixed background, the methodology could be extended to full QED. However, the author modestly notes that the interplay between the Coulomb background and the cloud of soft photons in a fully quantized theory remains an open question requiring further research. They also highlight the potential for this framework to shed light on the "infrared triangle" (soft theorems, asymptotic symmetries, and memory effects) from a perspective distinct from the standard FK approach.