No cosmological constraints on dark photon dark matter from resonant conversion: Impact of nonlinear plasma dynamics

TL;DR

Dark photon dark matter can resonantly convert to SM photons in the early Universe, but linear analyses overestimate the transfer of energy into the plasma. The authors show that nonlinear plasma dynamics, via the ponderomotive force, trigger density inhomogeneities and excite higher-k modes, rapidly shutting off resonant conversion before substantial heating occurs. 1+1D PIC simulations reveal two nonlinear regimes and demonstrate that the deposited energy is orders of magnitude below cosmological thresholds, weakening the DPDM constraints by factors of 3×10^3–10^7 across a wide mass range. These results emphasize the crucial role of nonlinear plasma effects in constraining light dark matter scenarios from early-universe resonant processes.

Abstract

We revisit and invalidate all dark photon dark matter constraints from resonant conversion of dark photons into photons (plasmons) in the early universe. These constraints rely on the resonant transfer of a substantial portion of the dark photon energy density into the SM plasma, heating the plasma in the process. We demonstrate that this resonant transfer saturates because of plasma nonlinearities. Dark photon dark matter resonantly converts into $k \simeq 0$ Langmuir waves in the early universe electron-ion plasma. Once the Langmuir-wave energy approaches the thermal energy of the plasma, nonlinear effects driven by the ponderomotive force become significant. In particular, we show using dedicated Particle-in-Cell simulations that large-amplitude $k = 0$ Langmuir waves excite higher-k Langmuir and ion acoustic waves, producing strong spatial variations in density and plasma frequency. These inhomogeneities suppress further resonant conversion, limiting the deposited energy to about the thermal energy of the electrons at the time of conversion, orders of magnitude below observable cosmological thresholds. Consequently, the dark photon dark matter constraints are weaker by factors of $3000$ to $10^7$ across ten orders of magnitude in dark photon mass.

Figures (11)

• Figure 1: The important time scales (Left axis) and velocities (Right axis) in our study as a function of scale factor $a$. The red and orange solid lines show inverse electron and ion plasma frequency, the blue solid line show the electron ion energy exchange time, while the black dot-dashed line shows the duration of the resonance $\varepsilon/H$ for $\varepsilon = 10^{-8}$. The blue dashed line shows the electron thermal speed, while the red dashed shows the DPDM quiver velocity $v_q^D$ also for $\varepsilon=10^{-8}$.
• Figure 2: Numerical results for resonant conversion with $\overline{\omega_p} = m_{A'}$. The solid lines (left axis) show the various energy densities ($\Delta E$) as a function of time, while the dashed lines (right axies) show the various thermal speeds as a function of time. The dotted gray line shows the initial thermal energy in the system as a comparison. The purple solid line shows the naive expectation for the growth of electron energy in linear theory of $t^2$, while the red solid line shows the actual evolution in PIC simulations. The upper panel is for a strong DPDM field with $v_q^D/v_{\rm th}^e = 0.03$ while the lower panel is for a weak DPDM field with $v_q^D/v_{\rm th}^e = 10^{-3}$.
• Figure 3: Numerical results for a Landau-Zener transition with $v_q^D/v_{\rm th}^e = 1$ (top) and $v_q^D/v_{\rm th}^e = 0.1$ (bottom). The purple solid line shows the naive expectation for the growth of electron energy in linear theory (see Eq. \ref{['eq:LZ']}), while the red solid line shows the actual evolution in a PIC simulation. Color coding is the same as in Fig. \ref{['fig:Efield']}.
• Figure 4: Updated Dark Photon Dark Matter Limits. The gray shaded regions are constraints from a variety of astrophysical and lab searches Caputo:2021eaa, while the color shaded regions are the cosmological constraints from early universe considerations (spectral distortion and $N_{\rm eff}$ in red) and late universe considerations from Dark Ages (orange) and Lyman-$\alpha$ forest (blue). Constraints on dark photons also arise from vector superradiance with gravitational wave measurements LIGOScientific:2025csrAswathi:2025nxa. The previous invalidated constraints are shown as gray dashed lines Arias:2012azCaputo:2020rnxCaputo:2020bdyMcDermott:2019lchWitte:2020rvb.
• Figure 5: Evolution of the shot noise energy density (in units of $n_e m_e c^2$; solid lines) in undriven ($A_0 = 0$) PIC simulations with different numbers of particles per cell. The analytically expected shot noise level, computed using Eq. \ref{['eq:PE_noise']}, is shown with dashed lines. We note that for the typical parameters used in our simulations, the thermal energy density (in units of $n_e m_e$) given by Eq. \ref{['eq:Eth']} is $5 \times 10^{-4}$, which is much higher than the noise level in all presented simulations.