The high-redshift radio galaxy 3C 294 at low frequencies: radio detection of the X-ray Ghost

TL;DR

This paper presents the first radio detection of the X-ray ICCMB Ghost surrounding the high-redshift radio galaxy 3C 294 ($z=1.8$) by combining LOFAR 144 MHz data with archival higher-frequency radio observations and deep Chandra X-ray data. By jointly modelling synchrotron and inverse-Compton emission from the same electron population using PYSYNCH, the authors constrain a very low-energy electron distribution ($γ_{ ext{min}}\sim1$, $γ_{ ext{break}}\lesssim10^{3}$, $γ_{ ext{max}}\lesssim10^{4}$) and a magnetic field of a few nT in the lobes, revealing an enormous lobe energy around $3\times10^{64}$ erg and an age of roughly $13$ Myr. Spectral-index mapping and hotspot ageing analyses indicate restarted, precessing jets with multiple generations, while inner hotspots imply active particle acceleration in a renewed phase. The findings imply ICSMB-dominated X-ray emission in aged plasma and highlight the necessity of very low-frequency observations to detect such lobes at high redshift, with significant implications for AGN feedback and the census of IC Ghosts in the early universe.

Abstract

We report on the very first radio detection associated with the peculiar hourglass-morphology X-rays surrounding 3C 294 at z=1.8. Using International Low Frequency Array (LOFAR) data at 144 MHz and Chandra data at 0.3-6 keV, we find that the co-spatial diffuse radio and X-ray emission is well described by synchrotron and inverse-Compton processes by the same electron population. Through modelling of this rare low-energy plasma, we find that the most defining property of the electrons up-scattering CMB photons at this redshift is very low electron Lorentz factors ($γ_{\text{max}}\ll 10^{4}$ and $γ_{\text{break}}\lesssim 10^{3}$) in the lobe: deep low frequency (<150 MHz) observations are critical to the detection of radio lobes at high redshift. The physical conditions imply a total energy in the diffuse emission significantly greater than that implied by the temperature of the protocluster gas: 3C 294 is one of the most powerful known radio-loud systems in a dense protocluster environment. Through resolved spectral analysis of archival radio data up to 15 GHz, we find evidence that the inner hotspots are due to restarted activity, while the outer hotspots remain energetic, suggesting a rapid duty cycle while the jet precesses. This allowed the low-energy aged plasma driving the X-rays to remain spatially distinct from the high-energy plasma. Together, our results promise a revelation of AGN-related radio emission at high redshift using future low-frequency arrays such as SKA-LOW.

Figures (9)

• Figure 1: Chandra X-ray image of 3C 294, adapted from fabi03 and erlu06, overlaid by radio contours at different frequencies. Left: VLA 1.4 GHz (white) and VLA 5 GHz contours (blue). Right: LOFAR 144 MHz (white) and VLA 5 GHz contours (blue). Contours are 3$\sigma$ multiplied by 2$^n$ where $n=[0,1,2...10]$, except for the blue contours, which start at 80$\sigma$. Radio contours have an angular resolution $\sim0.3$ arcsec, except the VLA 1.4 GHz (left) at 1.2 arcsec. The reprocessed 5 GHz data is sensitive to the radio core, unseen in previous publications. The low-energy plasma at 144 MHz (right) coincident with the diffuse X-ray emission is clearly detected, along with knots along the northern jet between the core and the northern inner hotspot.
• Figure 2: X-ray 0.3-6 keV spectrum of the diffuse regions shown in Figure \ref{['fig:3c294_lowres']}, fitted with an absorbed (PHABS*ACISABS) non-thermal power-law model (orange line).
• Figure 3: Same as in Figure \ref{['fig:3c294']} but with the LOFAR image convolved at 1.2 arcsec resolution (white contours) using $uv-$tapering and with the 1.5 GHz VLA data (blue contours) convolved to the same resolution. The grey rectangles describe the regions used for the X-ray spectral fits by fabi03, which we also use for flux measurements of the 144 MHz emission.
• Figure 4: Modelled spectra using the parameters in Table \ref{['tab:model_params']}, fitted to the measurements in Table \ref{['tab:fluxes']}. Blue curves show the synchrotron emission, and yellow curves show the IC emission, for models with physical parameters consistent with our observational constraints. Red arrows represent upper limits at GHz frequencies, based on 3$\sigma$ of the box regions presented in Figure \ref{['fig:3c294_lowres']}. The upper limits at 5 GHz and above are for reference only, and not used to constrain the spectra.
• Figure 5: Distributions of model parameters based on the constrained spectra in Figure \ref{['fig:plot_spec']}, where blue histograms display models consistent with the radio and X-ray flux measurements and upper limits, and orange histograms display models that are not (i.e. predicting fluxes inconsistent with our measurements or higher than our radio upper limits). The $p-$values from two-sample Wilcoxon Mann-Whitney U tests are displayed above each plot (rounded to two decimal places -- the $p-$value for the $\gamma_{\text{max}}$ distributions is $\sim10^{-16}$). The medians of each distribution are shown by dashed lines. Note the skew of the Lorentz factor distributions is due to the requirement of $\gamma_{\text{min}}\leqslant\gamma_{\text{break}}\leqslant\gamma_{\text{max}}$ after drawing from log-uniform distributions.