Ambiguities in the generation of CFJ-terms in a QED with dimension-5 operators in one loop

1. Problem Statement

The paper investigates the radiative generation of Carroll-Field-Jackiw (CFJ) terms in a Lorentz-violating extension of Quantum Electrodynamics (QED). Specifically, the authors focus on a model incorporating dimension-5, CPT-odd operators in the fermion Lagrangian.

• The Core Issue: Previous studies (notably Petrov et al. [5]) have shown that the generation of the CFJ term (a CPT-odd, Lorentz-violating photon mass term) is highly sensitive to the regularization scheme used. Different schemes yield different results for the CFJ coefficient.
• The Gap: While the Ward-Takahashi identity (gauge invariance) is often used to constrain such theories, it remains unclear whether gauge invariance alone is sufficient to fix the ambiguities arising from surface terms in the one-loop calculation of this specific dimension-5 model.
• Goal: To systematically separate regularization-dependent ambiguities from intrinsic divergences and determine if the CFJ term can be uniquely predicted or if it remains scheme-dependent even after imposing gauge symmetry.

2. Methodology: Implicit Regularization (IR)

The authors employ Implicit Regularization (IR), a framework that preserves the dimensionality of spacetime (crucial for handling $\gamma_5$ matrices and Levi-Civita tensors) and isolates ambiguities without modifying the theory's dimension.

• Separation of Divergences: IR allows the authors to split loop integrals into:
  1. Regularization-Independent Terms: Intrinsic divergences (logarithmic $I_{log}$ and quadratic $I_{quad}$) that are independent of the specific regulator.
  2. Regularization-Dependent Terms: Surface terms ($c_1, c_2, c_3, c_4, c_5$), which arise from the difference between integrals with shifted momenta. These terms vanish in Dimensional Regularization (DR) but are non-zero and arbitrary in other schemes.
• Calculation Strategy:
  • The authors calculate the one-loop effective action for the gauge field up to second order in the photon field ($A_\mu$) and first order in the Lorentz-violating coefficients ($a^{(5)}$ and $b^{(5)}$).
  • They compute six distinct polarization tensor amplitudes ($\Pi^{\mu\nu}_n$) corresponding to different interaction vertices derived from the dimension-5 Lagrangian.
  • They explicitly avoid symmetric integration to prevent masking the surface term ambiguities.

3. Key Results

A. Vanishing and Non-Vanishing Contributions

• Parity Constraints: The amplitude $\Pi^{\mu\nu}_1$ vanishes due to parity.
• Antisymmetric Tensors: Contributions involving the antisymmetric tensors $a^{(5)}_{\mu\alpha\beta F}$ and $b^{(5)}_{\mu\alpha\beta F}$ (specifically $\Pi^{\mu\nu}_{5,bF}$ and $\Pi^{\mu\nu}_{6,bF}$) vanish identically at one-loop, regardless of the regularization scheme.
• CFJ Generation: The terms capable of generating the CFJ structure (proportional to the Levi-Civita tensor $\epsilon^{\mu\nu\alpha\beta}$) arise from $\Pi^{\mu\nu}_{4,b}$, $\Pi^{\mu\nu}_{5,aF}$, and $\Pi^{\mu\nu}_{6,aF}$.

B. The Role of Surface Terms

The calculation reveals that the CFJ term is not just a function of intrinsic divergences ($I_{log}, I_{quad}$) but is heavily contaminated by surface terms ($c_i$). The general expression for the CFJ polarization tensor is:
$$\Pi^{\mu\nu}_{CFJ} \propto c_1 [\dots] + I_{log} [\dots] + I_{quad} [\dots]$$
where $c_1$ is a surface term defined by the difference between integrals of the form $\int \frac{p_\mu p_\nu}{(p^2-m^2)^2}$ and $\frac{1}{4}g_{\mu\nu}\int \frac{p^2}{(p^2-m^2)^2}$.

C. Symmetry Constraints (Ward-Takahashi Identity)

The authors apply the Ward-Takahashi identity ($k_\mu \Pi^{\mu\nu} = 0$) to constrain the surface terms:

• Partial Fixing: The identity successfully fixes $c_2 = 0$ and establishes linear relationships between $c_3, c_4, c_5$ and $c_1$.
• Residual Ambiguity: Crucially, the identity fails to fix $c_1$. The parameter $c_1$ remains undetermined.
• Conclusion on Gauge Invariance: Gauge invariance alone is insufficient to uniquely determine the radiatively induced CFJ coefficient in this model.

D. Comparison with Dimensional Regularization (DR)

• When the authors map their IR results to Dimensional Regularization (where $c_i = 0$ and $I_{log}, I_{quad}$ appear as poles in $\epsilon$), their general expression reduces exactly to the result found in previous literature (Petrov et al. [5]).
• This confirms that the "standard" DR result is a specific case of a more general, ambiguous structure where surface terms are set to zero by the scheme.

E. Critique of Tensor Ansatz

The paper critiques a previous proposal (in [5]) that imposed a specific relation between the Lorentz-violating tensors ($2a^\alpha_F = b^\alpha$) to cancel logarithmic divergences.

• Finding: In the IR framework, this ansatz only cancels the logarithmic divergence ($I_{log}$) but leaves the quadratic divergence ($I_{quad}$) and the surface term ($c_1$) untouched.
• Implication: This "cancellation" is an artifact of the specific scheme and does not resolve the fundamental ambiguity; it merely shifts the dependence to the remaining surface term.

4. Significance and Conclusions

1. Inherent Ambiguity: The generation of the CFJ term in dimension-5 Lorentz-violating QED is not universal. It inherently carries an ambiguity proportional to the surface term $c_1$, which depends on the regularization prescription.
2. Limitations of Gauge Symmetry: The study demonstrates that for theories with higher-dimensional operators, the Ward-Takahashi identity is not a strong enough constraint to fix all regularization ambiguities.
3. Methodological Clarity: Implicit Regularization provides a transparent tool to expose these hidden ambiguities, which are often obscured in Dimensional Regularization.
4. Future Directions: To physically fix the CFJ coefficient, additional constraints beyond gauge invariance are required. The authors suggest potential paths such as:
   • Momentum Routing Invariance: Requiring physical observables to be independent of how internal momenta are routed in Feynman diagrams.
   • Renormalization Conditions: Imposing specific physical renormalization schemes.
   • Phenomenology: Using experimental bounds on Lorentz violation to constrain the remaining free parameter.

In summary, the paper establishes that the radiative generation of CPT-odd operators in Lorentz-violating extensions of QED is a scheme-dependent process that cannot be fully resolved by standard symmetry arguments alone, necessitating a careful, case-by-case physical justification for the chosen regularization parameters.