Directly Probing Neutrino Interactions through CMB Phase Shift Measurements

1. Problem Statement

In the standard cosmological model ($\Lambda$CDM), neutrinos decouple from the primordial plasma at a redshift $z \sim 10^{10}$ and subsequently free-stream. This free-streaming nature induces a characteristic phase shift in the acoustic oscillations of the Cosmic Microwave Background (CMB). The gravitational pull of fast-moving neutrino perturbations shifts photon-baryon perturbations to larger scales, displacing the acoustic peaks in the CMB power spectra.

While this phase shift has been detected and used to confirm the existence of three free-streaming neutrino species, it remains unclear how this signature is modified if neutrinos possess non-standard interactions (e.g., self-interactions or couplings to dark matter) that delay their decoupling to later cosmic epochs. Previous analyses often relied on simplified "instantaneous decoupling" approximations or focused on amplitude changes rather than the specific phase shift signature. The authors aim to:

1. Derive the precise phase-shift templates for realistic neutrino interaction models with gradual decoupling.
2. Develop a robust analysis pipeline to directly constrain the neutrino decoupling redshift ($z_{\nu, \text{dec}}$) using CMB phase shift measurements.
3. Test these constraints against the latest CMB data (Planck, ACT, SPT) for both universal and flavor-dependent interaction scenarios.

2. Methodology

Theoretical Framework

• Gradual Decoupling: Unlike the instantaneous decoupling approximation, the authors model interactions where the scattering rate scales as a power law of the neutrino temperature: $\Gamma_\nu \propto T_\nu^n$. They focus on two physically motivated cases:
  • $n=3$: Neutrino-Dark Matter (DM) scattering.
  • $n=5$: Neutrino self-interactions (mediated by heavy particles) and certain neutrino-DM couplings.
• Phase Shift Template: The authors demonstrate that even with gradual decoupling, the induced phase shift $\delta_\ell$ retains the same functional form as the free-streaming case but with a rescaled asymptotic amplitude. The shift is parametrized as:
  $$\delta_\ell(N_{\text{eff}}) = A(N_{\text{eff}}) f_\ell$$
  where the template function $f_\ell$ is modified by an amplitude ratio $A_\infty$:
  $$f_\ell = f_\ell^{\text{SM}} \times A_\infty$$
  Here, $A_\infty = \ell_\infty / \ell_\infty^{\text{SM}}$ represents the ratio of the asymptotic phase shift amplitude in the interacting model to the Standard Model (SM) free-streaming limit.
• Amplitude Behavior:
  • $A_\infty = 1$: Corresponds to early decoupling (free-streaming limit).
  • $A_\infty \approx 0.3$: Corresponds to late decoupling (fluid-like limit). Crucially, the authors find that even fluid-like neutrinos induce a non-zero phase shift (approx. 30% of the free-streaming value) because their sound speed remains higher than the photon-baryon fluid due to baryon inertia.

Numerical Implementation

• Boltzmann Solvers: The authors use nuCLASS (a modified version of CLASS) for neutrino self-interactions and the built-in idm-idr (interacting dark matter-dark radiation) module in CLASS for neutrino-DM scattering.
• Extraction Pipeline: They adapt the methodology from Ref. [13] to extract phase shifts directly from CMB power spectra (TT, TE, EE).
  • They remove Silk damping and Integrated Sachs-Wolfe (ISW) effects to isolate acoustic peaks.
  • They apply a multipole shift $\Delta_\ell$ exclusively to the non-ISW component of the spectra.
  • They perform Markov Chain Monte Carlo (MCMC) analyses using MontePython and Cobaya to constrain the parameter $A_\infty$.

Data Sets

The analysis utilizes the latest CMB data:

• Planck 2018 (P18): High-$\ell$ (plik-lite) and low-$\ell$ (Commander/SimAll).
• Atacama Cosmology Telescope (ACT DR6): High-$\ell$ polarization and temperature.
• South Pole Telescope (SPT-3G D1): High-$\ell$ polarization and temperature.
• Combinations include P18 only, ACT+SPT, and the full combination (P18+ACT+SPT).

3. Key Contributions

1. Realistic Phase Shift Templates: The paper establishes that for realistic models with power-law scattering rates ($\Gamma \propto T^n$), the phase shift does not exhibit the sharp cutoff predicted by instantaneous decoupling. Instead, the amplitude $A_\infty$ smoothly interpolates between the free-streaming and fluid limits as a function of decoupling redshift.
2. Fluid-Like Limit Discovery: The authors explicitly quantify that fluid-like neutrinos (which never fully decouple) still induce a measurable phase shift ($A_\infty \approx 0.3$), a result previously overlooked. This is due to the fact that tightly coupled neutrinos propagate faster than the photon-baryon fluid.
3. Model-Agnostic Parameterization: By introducing $A_\infty$ as a direct observable, the authors create a signature-driven framework that maps CMB phase shift measurements directly to neutrino decoupling redshifts without needing to assume a specific interaction Lagrangian, provided the temperature scaling is known.
4. Flavor-Dependent Constraints: The framework is extended to scenarios where only a fraction ($F_{\nu, \text{int}}$) of neutrino species interact, showing how constraints weaken as the interacting fraction decreases.

4. Results

Constraints on Universal Interactions

Using the combined P18 + ACT + SPT dataset, the authors derive strong lower bounds on the neutrino decoupling redshift:

• Amplitude Constraint: $A_\infty > 0.90$ at 95% Confidence Level (C.L.).
• Decoupling Redshift ($z_{\nu, \text{dec}}$):
  • For $\Gamma \propto T^3$ (Neutrino-DM): $z_{\nu, \text{dec}} > 1.33 \times 10^4$.
  • For $\Gamma \propto T^5$ (Self-interactions): $z_{\nu, \text{dec}} > 1.71 \times 10^4$.
• Significance: These redshifts correspond to times deep within the radiation-dominated era, well before matter-radiation equality ($z_{\text{eq}} \approx 3400$). This rules out models where neutrinos remain strongly interacting until the matter-dominated epoch.

Flavor-Dependent Scenarios

When restricting interactions to only one neutrino species ($F_{\nu, \text{int}} = 1/3$):

• Constraints weaken significantly.
• Planck-only data permits fluid-like behavior extending to $z \sim 10^3$ (near matter-radiation equality).
• However, the combined ACT+SPT data still requires decoupling to occur in the radiation-dominated era, though the bounds are less stringent than in the universal case.

Data Performance

• Ground-Based Superiority: ACT and SPT data alone provide tighter constraints on the phase shift amplitude than Planck alone, particularly due to superior sensitivity to the EE polarization power spectrum at high multipoles ($\ell > 600$), where the phase shift signal-to-noise ratio peaks.
• Consistency: All dataset combinations are consistent with the Standard Model prediction ($A_\infty = 1$) at the $1\sigma$ level.

5. Significance

• Robust Probe of New Physics: The phase shift is identified as a "clean" and robust signature of free-streaming radiation. Unlike other cosmological parameters, it is difficult to mimic through modifications to initial conditions or other parameters, making it a powerful discriminator for non-standard neutrino physics.
• Complementarity: This work provides constraints that are complementary to amplitude-based analyses (which look for damping of peaks). The phase shift analysis is sensitive to the timing of decoupling, offering a unique window into the thermal history of the early universe.
• Future Prospects: The authors note that upcoming experiments like the Simons Observatory will further improve the precision of phase shift measurements, potentially distinguishing between different interaction models ($T^3$ vs. $T^5$) based on the sharpness of the transition in $A_\infty$.
• Generalizability: The methodology is not limited to neutrinos; it can be applied to any dark radiation component with similar interaction properties, broadening the scope of CMB constraints on the dark sector.

In conclusion, this paper establishes a rigorous, observationally driven framework that confirms neutrinos have been free-streaming since deep in the radiation era, placing stringent limits on any new physics that would delay this decoupling.