Prospects for Cosmological Collider Physics

TL;DR

This work investigates the detectability of heavy fields present during inflation via cosmological collider signatures in 21-cm cosmology. It develops model-independent clock-signal templates for both $m/H>3/2$ (oscillatory) and $m/H<3/2$ (intermediate) regimes, and assesses their observability with a cosmic-variance-limited 21-cm survey of the dark ages $z\sim 30-100$ using a Fisher formalism. The analysis accounts for secondary non-Gaussianities and windowed templates to isolate the clock signal, showing that oscillatory (clock) signatures can yield mass measurements with plausible baselines, while gravitational-only couplings are likely undetectable; self-interactions and direct couplings offer the strongest detection prospects. The results indicate that cosmological collider physics could, in principle, reveal the mass spectrum of fields at inflationary energies, providing a unique window into high-energy physics inaccessible to terrestrial experiments, assuming future 21-cm instrumentation and foreground removal can reach the necessary $k$-space coverage.

Abstract

It is generally expected that heavy fields are present during inflation, which can leave their imprint in late-time cosmological observables. The main signature of these fields is a small amount of distinctly shaped non-Gaussianity, which if detected, would provide a wealth of information about the particle spectrum of the inflationary Universe. Here we investigate to what extent these signatures can be detected or constrained using futuristic 21-cm surveys. We construct model-independent templates that extract the squeezed-limit behavior of the bispectrum, and examine their overlap with standard inflationary shapes and secondary non-Gaussianities. We then use these templates to forecast detection thresholds for different masses and couplings using a 3D reconstruction of modes during the dark ages ($z\sim 30-100$). We consider interactions of several broad classes of models and quantify their detectability as a function of the baseline of a dark ages interferometer. Our analysis shows that there exists the tantalizing possibility of discovering new particles with different masses and interactions with future 21-cm surveys.

Figures (12)

• Figure 1: Simplest tree-level bispectrum Feynman diagrams for QSFI. Type A diagrams, where one leg is taken to be the background field, were computed in Arkani-Hamed:2015bzaChen:2014cwa. Type B diagrams were computed in Chen:2014cwa. Diagrams of type C were included in the general EFT studies in Noumi:2012vrBaumann:2011nk.
• Figure 2: Bispectrum shape function for different values of $\mu$ for two different couplings, $\dot{\zeta}^2 \sigma$ and $(\nabla \zeta)^2 \sigma$, normalized to unity at $k_1/k_3=1$. The plot contains the full bispectrum, including all terms and permutations. Our results for $\dot{\zeta}^2 \sigma$ are in exact agreement with the results in Ref. Chen:2015lza Fig. 6. Note that the full shape includes both the clock signal (non-analytic in momentum) and non-clock signal (analytic in momentum). The former is Boltzmann-suppressed at large $\mu$, while the latter is power-law-suppressed. In the last two figures, the normalization at the equilateral point is dominated by the latter.
• Figure 3: Primordial shape function on the axis $k_1=k_2$, $k_3=1$ Mpc$^{-1}$, normalized to $1$ for $k_1=1$ Mpc$^{-1}$.
• Figure 4: Top: Primordial bispectrum correlation Eq. \ref{['eq:innerprod']} between the full bispectrum shape of Chen:2015lza and the standard inflation shapes as a function of $\mu$. Bottom: Primordial shape auto-correlation of the full bispectrum shape of Chen:2015lza as a function of $\mu$. All correlators are very large, showing the small difference between different frequencies.
• Figure 5: Template Eq. \ref{['eq:template1']} compared to the full numerical result for $\mu=2$. All permutations are included.