Breaking Conformal Invariance: Mechanisms & Implications

• Breaking Conformal Invariance is the phenomenon where classical or quantum conformal symmetry is lost due to the introduction of explicit mass scales, spontaneous VEVs, or quantum anomalies.
• The mechanisms include explicit breaking through mass terms, spontaneous symmetry breaking that generates a dilaton, and quantum anomalies from renormalization which lead to trace anomalies.
• Applications span conformal field theory, cosmology, and holographic dualities, with insights provided through modified Ward identities, anomaly matching, and effective field theory constructions.

The breaking of conformal invariance encompasses a spectrum of physical and mathematical mechanisms through which theories with classical or quantum conformal symmetry acquire states or dynamics that inherently lack this symmetry, often by introducing explicit mass scales, spontaneous symmetry breaking, or quantum anomalies. As conformal invariance is central to the structure of critical phenomena, conformal field theory, and early-universe cosmology, the study of its possible breakings—partial, explicit, or spontaneous—illuminates both foundational aspects and phenomenological predictions of modern theoretical physics.

1. Mechanisms of Conformal Invariance Breaking

Conformal invariance can be broken through several distinct mechanisms:

• Explicit breaking: Adding dimensionful parameters (mass terms, relevant couplings) to a Lagrangian immediately destroys the exact scale and conformal invariance present in the massless system, as in large-$N$ Chern–Simons theories with massive fundamental fermions (Frishman et al., 2013). The appearance of new scales in propagators, partition functions, or bound-state spectra signals the loss of conformality.
• Spontaneous breaking: If a scalar primary operator in a CFT acquires a nonzero vacuum expectation value (VEV), the vacuum fails to be invariant under scale transformations and special conformal transformations. This triggers the emergence of a gapless Nambu–Goldstone mode, the dilaton, associated with the broken symmetries (Hinterbichler et al., 2022, Schwimmer et al., 2010, Boels et al., 2015, Karananas et al., 2017). The symmetry breaking pattern, for example $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$ or $\mathrm{ISO}(3,1)$, governs the structure of the Goldstone sector and the transformation properties of the fields.
• Quantum anomalies and renormalization effects: Even in the absence of explicit breaking, the process of quantization typically introduces a renormalization scale through regularization procedures, thereby generating a trace anomaly in the energy–momentum tensor and breaking classical conformality at the quantum level (Alhaidari, 2022, Armillis et al., 2013). Anomalous dimensions and running couplings encode this breaking.

2. Spontaneous Breaking: Effective Theories and the Dilaton

In systems exhibiting spontaneous breaking of conformal invariance, the low-energy effective theory is dominated by the dynamics of the dilaton, the Goldstone boson for the nonlinearly realized broken symmetries (Hinterbichler et al., 2022, Schwimmer et al., 2010, Boels et al., 2015). The conformal action constructed about a symmetry-breaking VEV (constant or defect-induced) takes the general form of a derivative expansion:

• The leading two-derivative term is canonically normalized for the dilaton field.
• Four-derivative and higher terms are fixed by symmetry: in D=4, two independent conformal invariants can be constructed using $(\Box\Phi)^2/\Phi^2$ and $(\partial\Phi)^4/\Phi^4$, with coefficients determined by requiring invariance under special conformal transformations.
• Nonlinear realization of the conformal group yields inhomogeneous transformations for the dilaton, e.g.,

$$\delta_D\varphi(x) = -1 - x^\mu \partial_\mu \varphi(x) + \mathcal{O}(f^{-2}),\quad \delta_{K_\mu}\varphi(x) = -2x_\mu - (2x_\mu x^\nu \partial_\nu - x^2\partial_\mu)\varphi(x)+\mathcal{O}(f^{-2})$$

(Hinterbichler et al., 2022, Boels et al., 2015).

• The associated Noether currents display non-conservation according to modified Ward identities, reflecting the breaking pattern.

The loop and derivative expansion of the effective action is organized by the scale set by the VEV or defect strength parameter, with each loop suppressed by powers of $p/f$ (for constant VEV) or by the physical time scale inverse for time-dependent VEVs (Hinterbichler et al., 2022). The resulting EFT provides systematic control over both classical and quantum corrections to observables in the broken phase.

3. Ward Identities, Anomaly Matching, and Consistency Constraints

The transition from conformal to symmetry-broken phases imposes stringent requirements on the structure of correlation functions and anomalies:

• Generalized Ward identities: In the spontaneously broken phase, the classical conformal Ward identities are modified to include source terms proportional to the dilaton one-point function (Hinterbichler et al., 2022, Karananas et al., 2017). For example, for $n$-point correlation functions of primaries,

$$\sum_{i=1}^n (\Delta_i + x_i\cdot\partial_{x_i})\langle \mathcal{O}_1(x_1)\cdots\mathcal{O}_n(x_n)\rangle = \int d^4y \langle [D, \pi(y)]\cdots \rangle$$

with analogous relations for special conformal generators (Karananas et al., 2017).

• Anomaly matching: The trace anomalies (type-A and type-B, parameterized by $a$ and $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$0) must match between the unbroken and spontaneously broken phases. The requirement that the low-energy dilaton effective action reproduce the correct anomaly structure uniquely fixes its couplings to the Euler density $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$1 and Weyl tensor squared $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$2, as well as the necessary four-derivative Wess–Zumino terms (Schwimmer et al., 2010).
• Consistency conditions on CFT data: The spectrum of operator dimensions, vacuum overlaps, and OPE coefficients in the symmetry-broken phase are constrained by matching conditions derived from the algebraic structure of the broken symmetry and the Goldstone theorem (Karananas et al., 2017).

These principles extend to systems with explicit symmetry-breaking or in the presence of defects, prescribing the structure of physical observables, anomaly-mediated couplings, and the relations between IR and UV data.

4. Explicit Breaking: Mass Terms, Defects, and Anomalies

Explicit breaking of conformal invariance introduces new scales that significantly alter the dynamics:

• Mass terms and relevant operators: The addition of mass terms (e.g., $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$3 in Chern–Simons-fermion systems) or other relevant operators violates scale invariance at the Lagrangian level and propagates through all observables. Restoration of conservation properties or high-spin symmetry may occur only in limiting procedures or at leading order in large $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$4 (Frishman et al., 2013).
• Defect-induced breaking: Space-like defects (e.g., $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$5 in the pseudo-conformal universe) or background sources can force scalar primary fields to acquire time-dependent VEVs, resulting in the breaking $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$6 or lower (Hinterbichler et al., 2022, Hinterbichler et al., 2011). The effective action for fluctuations around these profiles systematically encodes the symmetry-breaking pattern and allows the computation of corrections in powers of $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$7.
• Renormalization effects and quantum anomalies: Standard renormalization introduces a scale $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$8 into correlation functions, transforming scale (and hence conformal) invariance violations into trace anomalies. The explicit calculation in classical conformal field theories, e.g., for improved energy–momentum tensors, exhibits $\mathrm{SO}(4,2)\rightarrow\mathrm{SO}(4,1)$9 (Alhaidari, 2022). Only by either abandoning strict renormalizability or by introducing compensating fields and modified renormalization prescriptions can exact conformality be restored at the quantum level (Armillis et al., 2013).

5. Selective and Sectoral Breaking: Models and Lattice Examples

Conformal invariance can be broken in a selective or sectoral fashion:

• Sector-selective breaking: In statistical-mechanical models such as the fully packed loop (FPL) ensembles for lozenge tilings, external drives (KPZ nonlinearity) can break conformality in one sector (e.g., the geometric height field $\mathrm{ISO}(3,1)$0) while preserving it in another (e.g., the "hidden" field $\mathrm{ISO}(3,1)$1 associated with percolation loops), as shown by field-theoretic and precise numerical analysis (Cao et al., 2015).
• Approximation-induced breaking: In the nonperturbative renormalization group, leading approximations (LPA) often preserve conformal invariance, but higher-order corrections (e.g., $\mathrm{ISO}(3,1)$2 in the derivative expansion) can introduce spurious breaking, detected by modified Ward identities. Criteria such as the principle of maximal conformality (PMC) or minimal sensitivity (PMS) are used to minimize these artifacts and ensure physical results track conformal sum rules (Cabrera et al., 2024).

Such examples provide both insight into the robustness of conformal invariance in physical systems and practical diagnostics for symmetry breaking induced by modeling or computational truncations.

6. Cosmology and the Pseudo-Conformal Scenario

Conformal invariance breaking plays a central role in early-universe cosmology:

• Pseudo-conformal universe: In this alternative to inflationary cosmology, the early universe is modeled by a conformal field theory with a time-dependent expectation value $\mathrm{ISO}(3,1)$3, breaking $\mathrm{ISO}(3,1)$4 to $\mathrm{ISO}(3,1)$5. This pattern is responsible for generating a nearly scale-invariant spectrum of perturbations in spectator fields, with dilution of pre-existing curvature and anisotropies achieved via slow contraction rather than exponential expansion (Hinterbichler et al., 2011).
• Explicit weak breaking, scale tilt: Small deviations from exact conformal invariance (e.g., by deforming a quartic potential to $\mathrm{ISO}(3,1)$6 with $\mathrm{ISO}(3,1)$7) generate a corresponding small tilt $\mathrm{ISO}(3,1)$8 in the scalar power spectrum, matching CMB observations (Rubakov et al., 2010).

The transformation laws and low-energy EFT of the dilaton are essential for organizing loop corrections and for predicting non-Gaussian signatures in the power spectrum (Hinterbichler et al., 2022).

7. Holography, D-brane Constructions, and Quantum Field Theory

Holographic dualities and string-theoretic constructions provide concrete realizations and tests of conformal invariance breaking:

• Probe flavor branes in AdS/CFT: Embedding probe D-branes (e.g., D7–anti-D7 in the Klebanov–Witten background) along nontrivial cycles can spontaneously break conformal invariance in the dual field theory, with the effective action for the brane modulus capturing the dynamics of the dilaton (Ben-Ami et al., 2013).
• Exceptional anomalies in Wilson loops: Null polygonal Wilson loops in $\mathrm{ISO}(3,1)$9 SYM exhibit anomalous breaking of conformal invariance in exceptional kinematic configurations, leading to new universal functions of the coupling constant with distinct scaling from standard cusp anomalies (Dorn, 2013).
• Fishnet CFTs: The non-supersymmetric fishnet CFT demonstrates spontaneous breaking of conformal symmetry along flat directions, with vanishing vacuum energy and protected massless dilaton, even in the planar quantum theory (Karananas et al., 2019, Karananas, 2020).

These constructions elucidate both the universality of conformal anomalies and the role of symmetry breaking in determining physical phases and effective actions.

---

References:

(Cao et al., 2015, Hinterbichler et al., 2022, Rubakov et al., 2010, Okui, 2018, Karananas et al., 2017, Boels et al., 2015, Schwimmer et al., 2010, Armillis et al., 2013, Frishman et al., 2013, Karananas et al., 2019, Karananas, 2020, Alhaidari, 2022, Cabrera et al., 2024, Arbuzov et al., 2014, Ben-Ami et al., 2013, Amelino-Camelia et al., 2015, Dorn, 2013, Hinterbichler et al., 2011)