Light Particles with Spin in Inflation

TL;DR

This work shows that the Higuchi bound on light spinning fields during inflation can be evaded if those fields couple sizably to the inflaton foliation, breaking de Sitter isometries and enabling a controlled EFTI/CCWZ description. Focusing on a light spin-2 multiplet, the authors derive its diffeomorphism-covariant action, establish consistency relations for couplings to inflaton perturbations, and analyze rich phenomenology arising from mixing with the graviton. They compute explicit corrections to scalar and tensor power spectra, and derive a spectrum of non-Gaussian signatures, including squeezed-limit bispectra with characteristic angular dependence and a trispectrum with quadrupolar modulation, all enhanced in the small-sound-speed regime. The results provide testable predictions for future CMB and gravitational-wave probes and open avenues to explore higher-spin fields during inflation within a well-defined EFT framework.

Abstract

The existence of light particles with spin during inflation is prohibited by the Higuchi bound. This conclusion can be evaded if one considers states with a sizeable coupling with the inflaton foliation, since this breaks the de Sitter isometries. The action for these states can be constructed within the Effective Field Theory of Inflation, or using a CCWZ procedure. Light particles with spin have prescribed couplings with soft inflaton perturbations, which are encoded in consistency relations. We study the phenomenology of light states with spin 2. These mix with the graviton changing the tensor power spectrum and can lead to sizeable tensor non-Gaussianities. They also give rise to a scalar bispectrum and trispectrum with a characteristic angle-dependent non-Gaussianity.

Figures (12)

• Figure 1: Contributions to the scalar and tensor power spectra due to the exchange of a $\sigma$ field. Solid lines correspond to $\pi$, wavy ones to $\gamma$ and curly ones to $\sigma$. The dots indicate a contraction between a pair of free fields, i.e. the insertion of a power spectrum. One has to put the minimal number of dots in such a way that external lines cannot be connected without going through a dot (contraction), and each dot is connected to external lines from both sides, see for instance Mirbabayi:2014zpa.
• Figure 2: Leading contributions to the $\left\langle \zeta \zeta \zeta \right\rangle$ 3-point function. Diagrams where the dots are located in different places are subdominant for small $c_0$.
• Figure 3: Leading contributions to the $\left\langle \gamma \zeta \zeta \right\rangle$ 3-point function.
• Figure 4: Leading contribution to the $\left\langle \zeta\zeta \zeta \zeta \right\rangle$.
• Figure 5: $\mathcal{C}_\zeta$ as a function of $\nu = \sqrt{\frac{9}{4} - \left( \frac{m}{H} \right)^2}$, in the range of masses below the Higuchi bound: $\nu \in [\frac{1}{2}, \frac{3}{2}]$.