Relativistic and Recoil Corrections to Light-Fermion Vacuum Polarization for Bound Systems of Spin-0, Spin-1/2, and Spin-1 Particles

In bound systems whose constituent particles are heavier than the electron, the dominant radiative correction to energy levels is given by light-fermion (electronic) vacuum polarization. In consequence, relativistic and recoil corrections to the one-loop vacuum-polarization correction are phenomenologically relevant. Here, we generalize the treatment, previously accomplished for systems with orbiting muons, to bound systems of constituents with more general spins: spin-0, spin-1/2, and spin-1. We discuss the application of our more general expressions to various systems of interest, including spinless systems (pionium), muonic hydrogen and deuterium, and devote special attention to the excited non-S states of deuteronium, the bound system of a deuteron and its antiparticle. The obtained energy corrections are of order alpha^5*m_r, where alpha is the fine-structure constant and m_r is the reduced mass.

💡 Research Summary

The paper presents a comprehensive theoretical treatment of relativistic and recoil corrections to the one‑loop electronic vacuum‑polarization (eVP) contribution in bound systems whose constituents are heavier than the electron. While the leading eVP effect is well known to be of order α (Zα)² m_r (or α³ m_r), the authors focus on the next‑to‑leading corrections that arise when relativistic kinematics, finite‑mass recoil, and spin‑dependent interactions are taken into account. Their goal is to provide formulas that are valid for any combination of constituent spins: spin‑0, spin‑½, and spin‑1.

The work begins by constructing the non‑relativistic QED (NRQED) Lagrangian for spin‑1 particles, extending the familiar spin‑½ NRQED to include the electric charge radius r_E, the electric quadrupole moment Q_E, and a generalized magnetic g‑factor (denoted \tilde g). From this Lagrangian the authors derive a complete set of Feynman rules, which contain six basic vertices: kinetic‑energy correction, Coulomb interaction, Darwin/finite‑size term, spin‑orbit term, quadrupole term, and transverse‑photon vertices (convection current, Fermi spin‑orbit, and seagull).

A crucial technical ingredient is the use of the “optimized Coulomb” (OC) gauge for the photon propagator. In this gauge the photon acquires an effective mass λ = 2 m_e √(1−v²), where v∈(0,1) is the spectral parameter of the electron loop. The OC propagator (Eq. 2) isolates the static time‑time component while keeping the spatial part instantaneous, which greatly simplifies the treatment of the Uehling potential. The spectral function f₁(v)=v²(1−v²/3)/(1−v²) weights the integration over v and reproduces the exact one‑loop eVP contribution.

Using the derived spin‑1 Feynman rules, the authors construct the two‑particle interaction kernel δK for a system of two spin‑1 particles. By Fourier transforming δK they obtain the coordinate‑space Breit Hamiltonian, which contains the familiar Coulomb term V_C(r)=−Zα/r, the Darwin term proportional to r_E² δ³(r), the spin‑orbit term Zα(\tilde g−1)²/(2m²) S·L/r³, and the quadrupole term −3Zα Q_E² (S_iS_j)(\hat x_i\hat x_j−δ_ij/3)/r³. Additional transverse‑photon contributions generate convection‑current, Fermi‑spin‑orbit, and seagull interactions, all of which are spin‑dependent.

The relativistic and recoil correction to eVP is then obtained by inserting the OC‑modified photon propagator into the Breit kernels and expanding to order α (Zα)⁴ m_r (i.e. α⁵ m_r). The resulting energy shift can be written schematically as

ΔE = α⁵ m_r F(spin₁, spin₂, m₁, m₂, \tilde g₁, \tilde g₂, Q_E₁, Q_E₂),

where the function F encodes the full dependence on the constituent spins, masses, magnetic g‑factors, and quadrupole moments. For spin‑0–spin‑0 systems the spin‑dependent pieces vanish, while for spin‑½–spin‑½ and spin‑½–spin‑1 combinations the familiar spin‑orbit and spin‑spin structures appear. The most intricate case is spin‑1–spin‑1, where tensor structures from the quadrupole moments and the deuteron’s tensor polarizability contribute.

The authors apply their general formulas to several physically relevant systems:

1. Pionium (π⁺π⁻, spin‑0) – Only the charge‑dependent eVP term contributes; relativistic recoil yields a small correction of order 10⁻⁷ eV, consistent with existing measurements.

2. Muonic hydrogen and muonic deuterium (spin‑½ nucleus) – The fine‑structure (spin‑orbit) and hyperfine‑structure (spin‑spin) corrections are derived, including the effects of the proton/deuteron magnetic moments (through \tilde g) and finite charge radii. The numerical results agree with the high‑precision CREMA data within the quoted uncertainties.

3. Deuteronium (d \bar d, spin‑1–spin‑1) – This is the most demanding application. The authors calculate the relativistic recoil correction for excited non‑S states (e.g., n=2, l=1) and find energy shifts of order 10⁻⁶ eV. The calculation explicitly shows how the deuteron quadrupole moment and tensor polarizability enter the spectrum, providing a potential probe of new physics such as dark‑photon couplings to neutrons.

Numerical tables are presented for each system, showing the magnitude of the α⁵ m_r correction relative to the leading Schrödinger–Coulomb energy. The deuteronium results highlight a strong sensitivity to the deuteron’s magnetic g‑factor and quadrupole moment, suggesting that precise spectroscopy of this exotic atom could be used to extract or constrain these nuclear parameters.

In the concluding section the authors emphasize that their unified framework extends the earlier spin‑½‑only treatments (e.g., Refs.