Electroweak baryogenesis and dark matter from a singlet Higgs

TL;DR

This work links electroweak baryogenesis and dark matter through a Higgs-portal real singlet S that couples to the Higgs and can generate a strong first-order electroweak phase transition via a tree-level barrier. Baryogenesis arises from a dimension-6 CP-violating operator that induces a spatially varying complex top mass across the bubble wall, while S serves as a subdominant dark matter candidate with relic density f_rel ≲ 0.03 but potentially detectable Higgs-mediated scattering near current direct-detection limits. A two-field finite-temperature potential is analyzed with analytic control enabled by the barrier, complemented by numerical transport equations to compute the baryon asymmetry; a Monte Carlo scan identifies viable regions where v_c/T_c>1 and direct-detection constraints are satisfied. The mass range m_S ≈ 80–160 GeV emerges as preferred, with direct detection prospects close to XENON100 sensitivity, and the model remains amenable to UV completion and extensions to multi-Higgs scenarios, offering a concrete, testable path to connect dark matter with the origin of baryon asymmetry.

Abstract

If the Higgs boson H couples to a singlet scalar S via lambda_m |H|^2 S^2, a strong electroweak phase transition can be induced through a large potential barrier that exists already at zero temperature. In this case properties of the phase transition can be computed analytically. We show that electroweak baryogenesis can be achieved using CP violation from a dimension-6 operator that couples S to the top-quark mass, suppressed by a new physics scale that can be well above 1 TeV. Moreover the singlet is a dark matter candidate whose relic density is < 3% of the total dark matter density, but which nevertheless interacts strongly enough with nuclei (through Higgs exchange) to be just below the current XENON100 limits. The DM mass is predicted to be in the range 80-160 GeV.

Figures (9)

• Figure 1: Comparison of the $SS$ annihilation cross section (in units of the standard relic density value $\langle\sigma v\rangle_0$) as a function of $m_{ S}$ using the approximations (\ref{['twobody']}) (dashed, blue) and (\ref{['better']}) (solid, red), for the case $\lambda_m=0.5$.
• Figure 2: Scatter plot $f_{\rm rel}$ vs. $m_{S}$ from a random scan of parameter space with input parameters varying in the ranges $\lambda_m=0.1-1$, $v_0/v_c = 1.1-10$, $\log_{10} v_c/w_c = (-1)-(+1)$. Different groups of points are distinguished by their relation to XENON bound: lightest gray points (filled circles) are already excluded, orange diamonds ("marginal") are uncertain and blue circles ("allowed") are still allowed even by strongest XENON limit. Yellow plus signs show the extension of the allowed region when the upper bound on $\lambda_m$ is pushed to 1.5.
• Figure 3: Plot of $f_{\rm rel}$ as a function of $m_{ S}$ in the case $\lambda_m=0.5$. Blue dashed line shows the calculation using $\langle \sigma v \rangle_{s=4m_{ S}^2}$ and red solid line the one using the accurate thermally averaged annhilation cross section.
• Figure 4: Distributions of parameters satisfying the constraints (\ref{['eq:consistency']}, \ref{['eq:sphaleronbound']}, \ref{['BRconstraint']}) and the nominal DM direct detection bound (\ref{['Xeconst']}). Top row shows input parameters, bottom two rows are derived. Dimensionful quantities are in GeV units.
• Figure 5: Scatter plot of the expected cross section $\sigma_{\rm eff} \equiv f_{\rm rel}\, \sigma_{ SI}$ in Xenon experiment vs. $m_{S}$ corresponding to the data shown in figure \ref{['fpan']}, also with the same color coding. Solid blue ("Nominal bound") line shows the nominal XENON100 bound Aprile:2012nq and the dashed orange lines the more conservative (upper curve) and more optimistic (lower curve) bounds reflecting uncertainties due to local DM distribution.