Cosmic Rays Masquerading as Cool Cores: An Inverse-Compton Origin for Cool Core Cluster Emission

Abstract

X-ray bright cool-core (CC) clusters contain luminous radio sources accelerating cosmic ray (CR) leptons at prodigious rates. Near the acceleration region, high-energy leptons produce synchrotron (mini)halos and sometimes observable gamma rays, but these leptons have short lifetimes and so cannot propagate far from sources without some rejuvenation. However, low-energy (~0.1-1 GeV) CRs should survive for >Gyr, potentially reaching ~100 kpc before losing energy via inverse-Compton (IC) scattering of CMB photons to keV X-ray energies, with remarkably thermal X-ray spectra. In groups/clusters, this will appear similar to relatively 'cool' gas in cluster cores (i.e. CCs). In lower-mass (e.g. Milky Way/M31) halos, analogous CR IC emission will appear as hot (super-virial) gas at outer CGM radii, explaining recent diffuse X-ray observations. We show that for plausible (radio/gamma-ray observed) lepton injection rates, the CR-IC emission could contribute significantly to the X-ray surface brightness (SB) in CCs, implying that CC gas densities may have been overestimated and alleviating the cooling flow problem. A significant IC contribution to diffuse X-ray emission in CC clusters also explains the tight correlation between the X-ray 'cooling luminosity' and AGN/cavity/jet power, because the apparent CC emission is itself driven by the radio source. Comparing observed Sunyaev Zeldovich to X-ray inferred pressures at $\ll 100$ kpc in CCs represents a clean test of this scenario, and existing data appears to favor significant CR-IC. A significant IC contribution also implies that X-ray inferred gas-phase metallicities have been underestimated in CCs, potentially explaining the discrepancy between X-ray (sub-Solar) and optical/UV (super-Solar) observed metallicities in the central ~10 kpc of nearby CCs. We also discuss the model's connection to observations of multiphase gas in clusters.

Figures (8)

• Figure 1: CR emission flux for an extended ancient CR halo at three different radii (labeled), for the simple model in § \ref{['sec:model']} of CRs with a constant diffusivity+streaming speed in a galaxy cluster of $M_{\rm vir}=10^{15}\,M_{\odot}$ at $z=0.05$, with steady-state injection rate $\dot{E}_{\rm cr,\,\ell} \sim 10^{44}\,{\rm erg\,s^{-1}}$ ($\dot{E}_{43}=10$) from a central source (e.g. bright radio galaxy). We compare metagalactic background light cooray:2016.extragalactic.background.light.compilations.reviewkhaire:2019.extragalactic.background.light.spectratompkins:2023.cosmic.radio.background in a $100\,$kpc aperture, and show (inset) the flux (arbitrarily normalized in ${\rm counts\,s^{-1}\,keV^{-1}}$) at $0.3-10\,$keV X-rays, compared to thermal spectra of primordial gas. Most of the CR emission comes from CR-IC scattering CMB photons, producing a thermal continuum-like ($\sim$ keV) soft X-ray peak. The secondary peaks with low-surface brightness at $10-100\,$MeV $\gamma$-rays (from relativistic bremsstrahlung) and $\sim 0.5-10\,$MHz radio (from synchrotron) are unobservable at present, though the radio is the natural evolution of known ultra-steep and GHz sources on smaller ($\lesssim 10\,$kpc) scales.
• Figure 2: Halo mass ($M_{\rm vir}$) dependence of the qualitative properties of the extended CR X-ray emission around galaxies. For each halo we assume the self-similar model in § \ref{['sec:model']} (Fig. \ref{['fig:spectrum']}). Top: CR emission "effective" (best-fit thermal from $0.1-10\,$keV) X-ray temperature $T_{\rm X}^{\rm IC}$, and actual gas virial temperature $T_{\rm vir}$, at either $0.1\,R_{500}$ or $100\,$kpc, versus $M_{\rm vir}$. Bottom: Radii where losses truncate the CR emission sharply (§ \ref{['sec:mass']}; assuming a constant CR streaming speed), and $R_{500}$, versus $M_{\rm vir}$. Because of the relative scalings of these quantities, in intermediate/low-mass halos (Milky Way/M31, or $\lesssim 10^{13}\,M_{\odot}$), CR-IC will generically mimic "hot" (super-virial) gas in the outer CGM around galaxies (near or beyond the virial radius). In high-mass halos (group/cluster, $\gtrsim 10^{14}\,M_{\odot}$), CR-IC will mimic "cool" (sub-virial) gas in the cluster cores. The crossover point ($\sim 10^{13}-10^{14}\,M_{\odot}$) depends on details of the profiles and redshift.
• Figure 3: Projected radial profiles of X-ray derived quantities for the same model of a galaxy cluster with a luminous CR injection source as Fig. \ref{['fig:spectrum']}. The "true" gas properties are taken to follow typical non cool-core (NCC) clusters (§ \ref{['sec:profiles']}), to illustrate an extremal case where all the "cool core" arises because of CR-IC. We compare the projected surface brightness (versus 2D radius), gas density $n$, temperature $T$, entropy $K$, thermal pressure $P=n k_{B} T$, and "mass deposition" rate $\int_{0}^{r} (2\,\mu\,m_{p}/k T)\,dL_{\rm X,\,cool}$ versus (3D) radius $r$. For each we show the "true" profile from the assumed NCC profile, and the estimated "apparent" profile accounting (very simply) for the contribution of CR-IC to the X-ray spectrum from $0.1-10$ keV (§ \ref{['sec:profiles']}). For a modest CR injection luminosity expected in typical bright radio galaxies, CR-IC could easily contribute much of the observed X-ray surface brightness at $\lesssim 100\,$kpc, leading to an over-estimate of density, pressure, and apparent mass deposition rate (cooling flow strength), and under-estimate of temperature and entropy, in the cluster core (when one assumes the emission is all thermal). The toy model profiles here are remarkably similar to strong cool cores observed, and the range of CC profiles observed can be reproduced by varying $\dot{E}_{\rm cr,\,\ell}$, despite there being (by construction) little or no "true" cool thermal core in the model.
• Figure 4: Cartoon illustration of speculative time-series of events linking different cluster and radio source categories (§ \ref{['sec:cartoon']}). Clockwise from bottom-left: An initial NCC develops some cooling and briefly resembles an "uncontaminated" CC, with inflows leading to some AGN activity which drives jets/cavities/feedback and injects CRs that form a growing central radio source and diffuse out to produce more extended emission similar to the CC profiles in Fig. \ref{['fig:profiles']}. This boosts the apparent CC luminosity by the (potentially much larger) CR-IC X-ray luminosity, producing an "apparent" strong cool core (which is, in fact, an ancient cosmic-ray halo of low energy $\sim$ GeV leptons). While the central AGN is strong it will contain an active radio galaxy, and potential radio mini-halo in the center surrounded by a larger radio-invisible (MHz) multi-phase CR-IC halo. Even after the central radio source weakens, the CR-IC halo can persist, slowly fading, for $\gtrsim$ Gyr until the system again resembles a NCC. We qualitatively estimate timescales for each phase but stress this is just a cartoon, and the "order of events" could be more complicated than illustrated here.
• Figure 5: Inferred "cavity power" $P_{\rm cav} \sim 4\,P_{\rm eff}\,V/t_{\rm buoy}$ (§ \ref{['sec:agn.XR']}) versus X-ray cooling flow/cool-core luminosity $L_{\rm X,\,cool}$, which would be inferred for cavities in the CR-IC dominated models shown in Fig. \ref{['fig:profiles']} (with different $\dot{E}_{\rm cr,\,\ell}$). We show the median and $\sim 90\%$ spread (driven by the range of observed volume-filling factors of cavities). We stress the predicted scaling (Eq. \ref{['eqn:pcav']}) has no fitted/adjusted parameter here. The correlation arises in the CR-IC model because both the X-ray inferred $P_{\rm cav}$ and $L_{\rm X,\,cool}$ are functions of the same leptonic CR energy density in the CC. We compare to X-ray measurements in strong CC clusters (references shown). In the CR-IC scenario this correlation and other radio-galaxy-CC correlations arise because we are effectively measuring two different functions of the CR energy density and plotting them against one another.