Large N_f Expansion in Gauge Theories

• Large N_f expansion is a systematic method in quantum field theories that uses an inverse fermion flavor expansion to reveal UV and IR fixed points.
• This technique resums infinite subsets of Feynman diagrams, yielding analytic expressions for beta functions and delineating conformal window boundaries.
• It underpins studies of renormalization schemes, phase diagrams, and asymptotically safe behavior in both gauge and Yukawa theories.

The large $N_f$ expansion refers to the systematic analytic expansion in inverse powers of the number of fermion flavors, $N_f$, as an organizing principle in quantum field theories, particularly in gauge-fermion and related models. This expansion enables resummation of infinite subsets of Feynman diagrams and produces controlled expressions for critical observables well beyond fixed-order perturbation theory. Its importance extends across renormalization group studies, phase diagrams of gauge theories, conformal window determinations, and the search for ultraviolet (UV) fixed points and asymptotically safe behaviors in non-asymptotically free regimes.

1. Beta Function Structure and Singularities at Large $N_f$

The renormalization group $\beta$-function for the gauge coupling in both abelian and non-abelian cases is constructed as a $1/N_f$ expansion while holding $A \equiv N_f\alpha/\pi$ (“’t Hooft coupling”) fixed: $$\frac{3}{2} \frac{\beta(\alpha)}{A} = 1 + \sum_{i=1}^\infty \frac{F_i(A)}{N_f^i}$$ for abelian theories, and with $H_i(A)$ for non-abelian cases (1006.2119). The leading term is determined by the fermion vacuum polarization, while subleading $F_i(A)$ encode resummed classes of diagrams with increasing complexity.

The first nontrivial order is known exactly: $$F_1(A) = \int_{0}^{A/3} I_1(x)\, dx,\qquad I_1(x) = \frac{(1+x)(2x-1)^2(2x-3)^2 \sin^3(\pi x) \Gamma(x-1)^2 \Gamma(-2x)}{(x-2) \pi^3}$$ The expansion exhibits a finite (nonzero) radius of convergence, with singularities at

$x = \frac{5}{2} + n,\quad n = 0,1,2,\ldots$

translating to logarithmic singularities at $A = 15/2 + 3n$, which are critical to the nontrivial UV/IR fixed point structure (1006.2119, Dondi et al., 2019).

For nonabelian (e.g., SU($N$)) theories, additional group theory and gluon self-interaction contributions introduce further singularities, notably a pole at $A = 3$: $$H_1(A) = -\frac{11}{4}\frac{C_G}{T_R} + \int_{0}^{A/3} I_1(x) I_2(x) dx$$ with $I_2(x)$ contributing new structural features to the singularity pattern (1006.2119, Antipin et al., 2017).

2. Nontrivial Fixed Points and the Conformal Window

The singular structure of $F_1(A)$ implies that the corrected $\beta$-function,

$1 + \frac{F_1(A)}{N_f},$

has zeros near each $A_0 = 15/2 + 3n$, with the first such pair given as

$A = \frac{15}{2} \pm 0.0117\, e^{-15\pi^2 N_f/7}$

interpreted as UV and IR fixed points, respectively (1006.2119, Dondi et al., 2019).

This architecture is generic: in QED-like theories, as $N_f$ rises above a critical value, the one-loop Landau pole is replaced by a nontrivial fixed point. In the non-abelian case, the analogous leading singularity is at $A=3$, leading to a UV fixed point for $A^* = 3 - \exp[-a(N_f/N_c) + b]$ (with representation-dependent constants $a$, $b$) (Antipin et al., 2017). This demarcates a so-called “ultraviolet conformal window” in the $N_f$ vs $N_c$ phase diagram, wherein theories—termed “Safe QCD”—possess a fundamental UV completion via an interacting fixed point, even when asymptotic freedom is lost (Antipin et al., 2017, Cacciapaglia et al., 2020).

3. Impact of Higher-Order $1/N_f$ Corrections and Scheme Dependence

Although the leading $1/N_f$ result is analytically controlled and finite in radius of convergence, higher-order corrections (e.g., $F_2(A)$, $F_3(A)$) introduce singularities of increasing strength. For example, $F_2(A)$ manifests a sequence of simple poles aligned with the singularities of $F_1(A)$, and in general, $F_i(A)$ can feature $(i{-}1)$-th order poles (1006.2119). This structure can potentially conspire to form essential singularities in the summed $\beta$-function, challenging strict perturbative control and the reliability of any low-order truncation near critical couplings.

The renormalization scheme dependence compounds this issue. A general finite redefinition of the coupling: $$K = \widetilde{K}\left[1 + t_1 \frac{\widetilde{K}}{N_f} + t_2 \left(\frac{\widetilde{K}}{N_f}\right)^2 + \ldots\right]$$ modifies higher-order terms such that, unless all $t_i$ vanish, the subleading terms acquire increasingly singular derivatives of $F_1$ (e.g., $F_1'(A)$, $F_1''(A)$, etc.), thereby invalidating the subleading nature of the $1/N_f$ expansion except in a unique “baseline” scheme (Pinoy et al., 22 Jul 2025). This restricts the physical interpretability of the fixed point unless the leading behavior is preserved under scheme transformations.

4. Applications to Conformal Phases, Critical Flavor Number, and Sphere Free Energy

Large $N_f$ techniques are extensively employed in mapping conformal windows:

• For asymptotically free gauge theories, the conformal window—where a Banks–Zaks IR fixed point is operative—is bounded above by the loss of asymptotic freedom and below by the critical flavor number $N_f^{\rm cr}$ where the conformal phase ends due to chiral symmetry breaking (Lee, 2020).
• The lower boundary is set by demanding, via a Banks–Zaks conformal expansion (scheme-independent, up to fourth order in $\Delta_{N_f} = N_f^{AF} - N_f$), that the anomalous dimension $\gamma_{\bar{\psi}\psi}$ saturates a physical criticality criterion, e.g., $\gamma_{\bar{\psi}\psi}=1$ or $$\gamma_{\bar{\psi}\psi}(2-\gamma_{\bar{\psi}\psi})=1$$, with the latter criterion producing better interpolation across the entire window (Lee, 2020).
• Uncertainty quantification is addressed both via Padé approximants (for convergent expansions) and via Borel-plane analysis (for asymptotic series). The inferred $N_f^{\rm cr}$ values for the vector representation of $\mathrm{SU}(3)$ and $\mathrm{SU}(2)$ are $\sim10$ and $\sim6$, respectively (Lee, 2020).

Large $N_f$ expansion is also central in computations of universal quantities such as sphere free energies $F$ in conformal gauge theories. For U(1) gauge theory coupled to $N_f$ massless fermions, the leading large-$N_f$ result for the $d$-sphere free energy can be matched against resummed $4-\epsilon$ expansions, establishing critical flavor thresholds for the persistence of the conformal phase (e.g., $N_{\rm crit} \leq 4$ in QED$_3$) (Giombi et al., 2015, Tarnopolsky, 2016).

5. Universal Structure and Cross-Theory Comparison

A remarkable universality emerges in the analytic structure of large $N_f$ expansions:

• For gauge theories (QED, QCD) and their supersymmetric counterparts, the $1/N_f$-expanded $\beta$-functions admit closed-form expressions whose integrand singularities (e.g., isolated simple poles in Gamma function factors) determine the radius of convergence and branch point structure (Dondi et al., 2019).
• The location of these singularities ($K=15/2$ for QED, $K=3$ for QCD) governs where UV fixed points may arise and thus stipulate the boundaries of the nonperturbatively controlled regime.
• In non-supersymmetric QED and QCD, the coefficient of the logarithmic singularity is such that a nontrivial UV fixed point exists; in the supersymmetric case, the reversed sign prevents a “safe” fixed point from emerging (Dondi et al., 2019).
• In Yukawa and scalar QED models, analogous bubble-chain resummations show analytical behavior with finite radius and singularities at model-specific locations (e.g., $K=5$ for simple Yukawa) (Alanne et al., 2018, Zheng et al., 2018).

6. Implications, Limitations, and Future Directions

These findings have direct implications for the ultraviolet fate of gauge-fermion theories, suggesting that “asymptotic safety”—existence of a UV interacting fixed point—can arise through the resummed large $N_f$ structure, especially in non-abelian gauge theories (“Safe QCD”) (Antipin et al., 2017, Cacciapaglia et al., 2020). However, the increasing singularity of higher-order corrections and profound scheme dependence imply that robust confirmation requires either full series resummation or nonperturbative methods such as the functional renormalization group or lattice simulations (Pinoy et al., 22 Jul 2025).

The large $N_f$ expansion remains a foundational tool for nonperturbative analysis. It provides insight into both the detailed critical behavior at the conformal edge (e.g., operator scaling dimensions, universal free energies (Tarnopolsky, 2016, Benvenuti et al., 2019)) and the global structure of the phase diagram (e.g., chiral symmetry restoration, deconfinement transition, QCD thermodynamics at high flavor number (Ahmad, 21 Mar 2025, Ahmad et al., 20 Jan 2024, Fejos et al., 31 Mar 2024)).

Further advances will rely on the analytical continuation and resummation techniques to handle the singularities and on cross-validation with fully non-perturbative computational frameworks, especially for theories with several competing scales or in near-conformal regimes.