Detecting gravitational waves from cosmological phase transitions with LISA: an update

TL;DR

The paper evaluates the prospects for detecting gravitational waves from cosmological phase transitions with LISA, updating previous analyses by emphasizing sound-wave–driven production and adopting conservative, simulation-informed predictions. It provides a detailed pipeline from particle physics models to GW spectra, including parameter extraction ($H_*$, $\beta$, $\alpha$, $v_w$), energy budgets, and wall dynamics, and packages these into the PTPlot web tool for real-time, up-to-date assessments. The study surveys a wide range of models (singlet extensions, EW multiplets, SUSY, SMEFT, warped extra dimensions, composite Higgs, and dark sectors) and discusses both perturbative and nonperturbative approaches, highlighting where LISA is likely to observe signals and where collider or DM experiments offer complementary constraints. Overall, the work clarifies current uncertainties (notably in wall friction, turbulence, and strong supercooling regimes) and provides a practical framework for interpreting potential LISA detections in the context of beyond-Standard-Model physics.

Abstract

We investigate the potential for observing gravitational waves from cosmological phase transitions with LISA in light of recent theoretical and experimental developments. Our analysis is based on current state-of-the-art simulations of sound waves in the cosmic fluid after the phase transition completes. We discuss the various sources of gravitational radiation, the underlying parameters describing the phase transition and a variety of viable particle physics models in this context, clarifying common misconceptions that appear in the literature and identifying open questions requiring future study. We also present a web-based tool, PTPlot, that allows users to obtain up-to-date detection prospects for a given set of phase transition parameters at LISA.

Figures (16)

• Figure 1: Blueprint for analyzing cosmological PTs in the context of LISA. See text for details.
• Figure 2: Example output of the PTPlot tool. The colored lines show the SNR that depends on $T_*$, $g_*(T_*)$ and $v_{\rm w}$. The two remaining parameters can either be $\alpha$ vs $\beta/H_*$ or $\bar{U}_{\rm f}$ vs $R_* H_*$. The dotted straight lines are the contours of the fluid turnover time $H_{\rm n} R_*/\overline{U}_{\rm f}$ quantifying the effect of turbulence. In the gray shaded region the decay of sound waves into turbulence is less important than the Hubble damping and the SNR curve reflects this effect. The model parameters are given in the text.
• Figure 3: Example output of the PTPlot tool. The plot shows an example of the GW power spectrum from a first-order PT, along with the LISA sensitivity curve ($h^2\Omega_{\rm Sens}(f)$ taken from the LISA Science Requirements Document LISA_docs). The parameters of the example model are $v_{\rm w}=0.9$, $\alpha=0.1$, $\beta/H_*=50$, $T_*=200$ GeV, $g_*=100$.
• Figure 4: LEFT: Predicted values of $\alpha$ and $\beta/H$ for $m_2 = 170$ GeV and 240 GeV (combined in one plot) in the general singlet model obtained by linearly varying the free parameters of the model and imposing requirements as described in the text. The mixing angles considered were $\sin \theta = 0.01$ (blue points) and $0.1$ (orange points). The models in blue (orange) are unlikely (likely) to be probed by the high-luminosity LHC. The expected LISA sensitivities correspond to $T_*=50$ GeV. RIGHT: Sensitivity curve for the $Z_2$ symmetric singlet extension. In both the left and right panels we have taken $v_\text{w}=1$.
• Figure 5: Results in the 2HDM. LEFT: Strength of the EWPT as given by $\alpha$ in the ($\mathrm{tan} \beta$, $m_{A_0}$) plane, for $m_{H_0} = 200$ GeV (solid red) and $m_{H_0} = 400$ GeV (dashed blue). RIGHT: 2HDM benchmark points (from a scan in $m_{H_0}$, $m_{A_0}$, see text for details) in the ($\alpha$, $\beta/H$) plane. The points currently excluded by the LHC for the Type-II 2HDM (but not for the Type-I 2HDM) are shown in orange. The expected LISA sensitivities correspond to $T_*=50$ GeV and assume $v_\text{w}=0.7$.