Possible time-variable iron-K$α$ emission in the circum-nuclear region of the Circinus galaxy

TL;DR

We address how extended neutral Fe Kalpha emission traces circumnuclear gas around Circinus and whether its flux varies with AGN luminosity. We develop a pipeline combining nuclear PSF modeling with MARX simulations to isolate in-situ Fe Kalpha emission at 20–60 pc and map its distribution. We find that the eastern and western regions show significant time variability and large EWs up to ~2 keV, consistent with AGN X-ray photoionization and an X-ray echo origin. Variability timescales constrain cloud sizes to sub-parsec to tens of parsecs, and ALMA comparisons reveal X-ray feedback creating CO-deficient zones, highlighting the AGN’s impact on the circumnuclear ISM.

Abstract

We present imaging and spatially resolved spectral analyses of eight Chandra data taken for the Circinus galaxy in $\approx$ 22 years to reveal neutral iron-K$α$ emission on a circumnuclear scale ($\sim$ 10--100 pc) and search for time variability in the emission. By simulating and taking account of point-source emission from the active galactic nucleus (AGN), we detect iron-line emission $\sim$ 20--60 pc away from the nucleus, particularly in the eastern and western regions. In the two regions, possible time variability in the line flux was also detected. Our spectral analysis then finds that the observed equivalent widths can reach $\sim$ 2 keV and the slopes of underlying continua are rather inverted with $Γ< 0$. These are consistent with a scenario in which the iron emission originates from clouds illuminated by AGN X-rays; our result could provide the first extragalactic example of AGN X-ray echoes. In this scenario, we estimated the physical sizes of the illuminated clouds based on the timescale of variability to be less than 6 pc. Furthermore, we compared the iron emission distribution with the cold molecular distribution inferred by Atacama Large Millimeter/submillimeter Array (ALMA) observation of CO($J$=3--2), revealing that in the region of bright iron-line emission, the molecular emission seems to be weak. This might suggest that the AGN X-ray emission affects the chemical composition in the form of AGN feedback.

Figures (8)

• Figure 1: Nuclear spectra: 0.5--8.0>keV spectra of the observation (black) and MARX simulation (red) extracted from the central 1$.\!\!{}^{\prime\prime}$0 radius region on 2000 June 16. The orange dash-dot line, the blue dashed line, and the green dotted line the $\mathtt{TBabs2 \times powerlaw1}$ , $\mathtt{powerlaw2}$, and neutral Fe K$\alpha$ line components of the model shown in equation (1), respectively. The ratio of the observation to MARX simulation is shown in the lower panel.
• Figure 2: Images of the observation (left) and MARX simulation (right) on 2000 June 16 in the 6.2--6.5>keV band at 1/8 sub-pixel binning (smoothed with a Gaussian kernel of radius = 6>pixel and $\sigma =$ 3>pixel). Images were created on a logarithmic scale, with colors corresponding to the number of counts per pixel.
• Figure 3: Radial profiles of the observation (blue) and MARX simulation (red) in the 6.2--6.5>keV band for each observation.
• Figure 4: Neutral iron K$\alpha$ emission line maps for each observation. They were created in the 6.2--6.5>keV band by dividing the number of counts per pixel in the observation image by those in MARX simulation image. Each pixel has a size of $0.0615\times0.0615\>\mathrm{arcsec^2}$, which is 1/8 of the original pixel size, and smoothing was performed with a Gaussian kernel of radius = 6>pixel and $\sigma =$ 3>pixel. The color bar represents the ratio of the observation to the simulation. The inner 1$.\!\!{}^{\prime\prime}$0 radius circle and the outer 3$.\!\!{}^{\prime\prime}$0 radius circle define the region for extracting spectra in subsection \ref{['ssec:cnrspec']}.
• Figure 5: Spectra and models of the four circumnuclear regions (CNRs) on 2004 June 2; in-situ radiation from the CNR (orange), contamination from the center to the CNR (blue), and the sum of them (black).