Effective Potential in Subleading Logarithmic Approximation in Arbitrary Non-renormalizable Scalar Field Theory

1. Problem Statement

The paper addresses the calculation of the quantum effective potential in arbitrary scalar field theories, specifically focusing on non-renormalizable models.

• Context: The effective potential determines the ground state of a field theory. While standard Renormalization Group (RG) methods work well for renormalizable models, they face formal limitations in non-renormalizable theories.
• The Challenge: In non-renormalizable models, counterterms required to cancel UV divergences do not match the original Lagrangian structure. This leads to "infinite arbitrariness" (new operators appearing at every order), making standard RG approaches difficult.
• Goal: The authors aim to extend their previous work on the Leading Logarithmic Approximation (LLA) to the Next-to-Leading Logarithmic Approximation (NLLA). They seek to construct recurrence relations and RG equations that sum up leading and subleading logarithms to all orders of perturbation theory (PT), treating the non-renormalizable model as a UV-complete theory rather than just an effective field theory.

2. Methodology

The authors employ a formalism based on the Bogoliubov-Parasiuk-Hepp-Zimmermann (BPHZ) renormalization procedure and the R-operation.

• R-Operation and Locality: The core of the method relies on the Bogoliubov-Parasiuk theorem, which ensures the locality of counterterms. The authors utilize the property that the divergent parts of Feynman diagrams (poles in dimensional regularization, $1/\epsilon$) must satisfy specific recurrence relations to maintain locality.
• Recurrence Relations:
  • They derive relations connecting the singular parts (poles) of $n$-loop diagrams to lower-loop diagrams and counterterms.
  • Specifically, they relate the coefficients of the leading ($A_n$), subleading ($B_n$), and sub-subleading ($C_n$) poles.
  • Key Insight: The calculation of poles at any order can be reduced to the calculation of 1-, 2-, and 3-loop diagrams.
• The "R-Rule": A heuristic prescription is established to construct differential equations directly from diagrammatic structures. It involves replacing the classical potential $V_0$ with sums of counterterms ($\Sigma$) and adjusting the number of derivatives ($D^k = \partial^k/\partial\phi^k$) based on the number of internal quantum lines in the diagram.
• Handling Non-Renormalizability: Instead of trying to eliminate infinite arbitrariness, the authors assume UV divergences are subtracted in a specific way. They argue that the Leading Logarithmic Approximation (LLA) is independent of the subtraction scheme, while the NLLA depends on it. They track this scheme dependence explicitly.

3. Key Contributions and Derivations

A. Leading Logarithmic Approximation (LLA) Recap

The authors briefly review their previous work where they derived a first-order differential equation for the sum of leading poles, $\Sigma_A(z, \phi)$:
$$\frac{\partial}{\partial z}\Sigma_A = -\frac{1}{4}(D^2\Sigma_A)^2$$
where $z = g/\epsilon$. This equation mirrors the structure of a one-loop diagram.

B. Next-to-Leading Logarithmic Approximation (NLLA)

The paper's primary technical contribution is the derivation of the NLLA equations.

1. New Diagrams: Unlike LLA, NLLA requires considering:
   • Two-loop diagrams (bubble and sunset types).
   • Propagator corrections involving two derivatives (denoted as $V^G$).
2. Recurrence Relations: They derive complex recurrence relations for the subleading pole coefficients ($V^B_n$) and the derivative-dependent terms ($V^G_n$). These relations involve convolutions of leading ($V^A$) and subleading ($V^B$) terms.
3. Differential Equations: By summing the recurrence relations, they obtain a system of coupled partial differential equations (PDEs) for the generating functions $\Sigma_B$ (subleading poles) and $\Sigma_G$ (propagator corrections):
   • $\frac{\partial^2}{\partial z^2}\Sigma_B = \dots$ (a complex non-linear equation involving $D^2, D^3, D^4$ derivatives of $\Sigma_A, \Sigma_B, \Sigma_G$).
   • $\frac{\partial^2}{\partial z^2}\Sigma_G = \frac{1}{24}(D^3\Sigma_A)^2$.
4. Effective Potential Equation: They derive a linear first-order differential equation for the function $\bar{\Sigma}_B$, which sums the logarithmic corrections to the effective potential:
   $$\frac{\partial}{\partial z}\bar{\Sigma}_B = \text{RHS}(\Sigma_A, \Sigma_B, \Sigma_G)$$
   The solution is obtained by integrating and substituting $z \to g \ln(\mu^2/m^2(\phi))$.

C. Subtraction Scheme Dependence

The authors analyze how the results depend on the renormalization scheme (e.g., Minimal Subtraction vs. others).

• They demonstrate that in non-renormalizable theories, scheme dependence is not absorbed by a single coupling redefinition but requires redefining emergent operators.
• They derive a specific transformation law for the subleading term under a change of scheme:
  $$\Sigma_B^{c_1} = \Sigma_B^{MS} + c_1 \epsilon z \frac{\partial}{\partial z}\Sigma_A$$
  This confirms that the scheme dependence is controlled and linear in the subtraction constant $c_1$.

4. Results and Verification

• General Power-Like Potential: The authors apply their formalism to a general potential $V_0 = \phi^p/p!$. They derive a system of ordinary differential equations (ODEs) for the functions $f_A, f_B, f_G$. They note that the order of the ODE increases with the approximation order (1st order for LLA, 2nd order for NLLA).
• Renormalizable Case ($\phi^4$ Theory): To verify the formalism, they set $p=4$ (the renormalizable quartic model).
  • The derived equations simplify significantly.
  • The analytical solutions for the effective potential in the NLLA are found to be identical to known results in the literature (specifically Ref. [20] and standard RG results).
  • The results correctly reproduce the coefficients of the beta-function ($\beta_0, \beta_1$) and the anomalous dimension ($\gamma_1$) for $\phi^4$ theory.

5. Significance and Conclusion

• Universality: The paper proves that the BPHZ R-operation and the associated recurrence relations work equally well for non-renormalizable theories as they do for renormalizable ones.
• Formalism Extension: It successfully extends the summation of logarithms from LLA to NLLA for arbitrary scalar potentials, providing a systematic way to handle non-renormalizable interactions without relying on the Effective Field Theory (EFT) truncation.
• Locality as a Guide: The work highlights that the locality of counterterms is the fundamental principle driving the recurrence relations, independent of the multiplicative nature of renormalization.
• Limitations: The authors acknowledge that while LLA is scheme-independent and NLLA is under control, the "infinite arbitrariness" inherent to non-renormalizable theories (the need for infinite counterterms) remains a conceptual challenge that is not fully solved, though the logarithmic structure is now understood.

In summary, this paper provides a robust mathematical framework for calculating high-order quantum corrections in non-renormalizable scalar theories, bridging the gap between effective field theory approaches and a more complete UV treatment by leveraging the structural constraints of the R-operation.