Raining in MKW 3 s: A Chandra-MUSE Analysis of X-Ray Cold Filaments around 3CR 318.1

The Third Cambridge Catalog of radio sources and its revised versions (3C, 3CR, 3CRR; Edge et al. 1959; Bennett 1962; Spinrad et al. 1985; Laing et al. 1983) constitute one of the most valuable samples of radio-loud active galactic nuclei (AGN). In 2018, Balmaverde et al. (2018) started the MUse RAdio Loud Emission line Snapshot (MURALES) survey to observe all southern 3CR sources with the Multi-Unit Spectroscopic Explorer (MUSE; Bacon et al. 2010) and enrich the multifrequency coverage of the 3CR catalog.

During the MURALES survey, an optical filamentary structure was detected in 3CR 318.1 (Balmaverde et al. 2019), the brightest cluster galaxy (BCG) of the galaxy cluster MKW 3 s (Morgan et al. 1975), in agreement with observations by Edwards et al. (2009). This structure consists of two ionized gas filaments extending ∼25 kpc in the south and the southwest directions, both with almost constant velocities along their projected lengths (∼230 km s⁻¹ and ∼280 km s⁻¹, respectively). The filament that points to the south presents a bright knot at its southern end. In contrast with the typical narrow-line regions seen in other radio galaxies, these filaments extend beyond tens of kiloparsecs and present drastically different emission line ratios (e.g., [O iii]/H_β ∼ 0.1 instead of the typical value of ∼10).

Similar filamentary structures were discovered in other BCGs, as NGC 1275 in the center of the Perseus galaxy cluster (Lynds 1970; Conselice et al. 2001 and Fabian et al. 2008), spatially associated with an X-ray excess (Fabian et al. 2011). The same situation occurs for other BCGs, where a tight connection between X-ray and optical filament emission was found (McDonald et al. 2010).

However, the physical processes responsible for the ionization in these optical filaments is still debated. While Voit et al. (1994) proposed ionization due to hot cooling intracluster medium (ICM) as the excitation mechanism behind these filaments, McDonald et al. (2010), who analyzed a sample of 23 filaments around BGCs, suggested that these types of filaments are due to ICM thermal conduction. In the case of the Perseus cluster, Fabian et al. (2011) rejected this scenario, due to the lack of a thick interface between the cold and hot gas, proposing that the origin of cold X-ray filaments is due to penetration of cold gas by the hot surrounding gas through reconnection diffusion.

Here we present a comparison between optical and X-ray images of 3CR 318.1 to investigate the main ionization processes underlying the origin of these optical filaments, thus focusing only on the first ∼30 kpc at the center of MKW 3 s (for works on the large-scale structures see, e.g., Mazzotta et al. 2002, 2004; Bîrzan et al. 2020).

This paper is organized as follows. A brief description of 3CR 318.1 is given in Section 2. Optical and X-ray data analyses are reported in Section 3. Results are presented in Section 4, while Section 5 is devoted to our conclusions.

Unless otherwise stated, we adopted cgs units for numerical results and assumed a flat cosmology with H₀ = 69.6 km s⁻¹ Mpc⁻¹, Ω_M = 0.286, and Ω_Λ = 0.714 (Bennett et al. 2014). Spectral indices, α, are defined by flux density, S_ν ∝ ν^−α .

3CR 318.1 is the radio source associated with the nearby (z = 0.0453, which corresponds to 0.896 kpc arcsec⁻¹) galaxy NGC 5920, the BCG of the galaxy cluster MKW 3 s. Based on its optical line emission, it can be classified as an extremely low-excitation galaxy (see Capetti et al. 2013). The central source presents a steep spectrum at low radio frequencies (i.e., ${\alpha }_{1.4\ \mathrm{GHz}}^{150\ \mathrm{MHz}}=0.63$) with a luminosity of $\mathrm{log}{L}_{150\mathrm{MHz}}=23.16$ W Hz⁻¹ (Capetti et al. 2020). The surrounding diffuse radio emission also shows an extremely steep radio spectrum (${\alpha }_{235\ \mathrm{MHz}}^{1.28\ \mathrm{GHz}}=2.42$), leading to a classification of relic radio galaxy by Giacintucci et al. (2007). The host galaxy of 3CR 318.1 presents an AB UV magnitude at 2600 Å of 19.04 ± 0.04, corrected for Galactic reddening (Cardelli et al. 1989), with E(B − V) = 0.0356 ± 0.0010. ²² This magnitude yields a UV luminosity at 2600 Å of L_UV = (4.96 ± 0.17) × 10⁴² erg s⁻¹. An overview of 3CR 318.1 and its environment is shown in Figure 1.

Zoom In Zoom Out Reset image size

Peres et al. (1998) found that MKW 3 s presents a moderate "cooling flow" ($\dot{M}\sim 170\ {M}_{\odot }$ yr⁻¹). Using deep (∼57 ks) Chandra X-ray observations of the central 200 kpc, Mazzotta et al. (2002) identified the presence of a cavity, located at ∼90 kpc south of the X-ray nucleus, hotter than the ∼3 keV average cluster temperature and filled by radio emission arising from the southern lobe (Mazzotta et al. 2004), as recently confirmed by Bîrzan et al. (2020) using Low Frequency Array observations.

X-ray data reduction was performed following standard procedures of the Chandra Interactive Analysis of Observations v4.10 (CIAO; Fruscione et al. 2006) threads ²³ , adopting the Chandra Calibration Database v4.8.4.1, and using the ∼57 ks Chandra observation (ObsID 900). The MUSE data set was obtained from the MURALES survey (ID 099.B-0137(A); two exposures of 10 minutes each), with a mean seeing in the V band at zenith of 138 (see Balmaverde et al. 2019 for details on the data reduction and analysis). Here we focus on the analysis of the [N ii]λ6584 emission since it is more luminous and extended than the Hα emission. The [N ii]/Hα ratio is constant along filaments with [N ii]/Hα ∼ 1.9.

Astrometric registration to align radio, optical, and X-ray images was performed by measuring the radio centroid at 1.4 GHz (beam size of 45), and the X-ray centroid in the 0.5–7 keV energy range and then aligning them following the procedure from Massaro et al. (2011a). We shifted the X-ray image 0.568'' to the northeast; then, we aligned the position of the optical host galaxy in the MUSE continuum image with the radio centroid shifting it by 162. These shifts are consistent with those previously reported for the 3CR Chandra Snapshot Survey (Massaro et al. 2011a, 2013, 2015, 2018; Jimenez-Gallardo et al. 2020). Registration was then verified by checking the alignment of other pointlike field sources in the Sloan Digital Sky Survey (SDSS) and in the MUSE continuum images. The comparison between registered X-ray and [N ii]λ6584 optical images is shown in Figure 2.

Zoom In Zoom Out Reset image size

In the same figure, we show the regions selected to perform X-ray spectral analysis, based on the morphology of the [N ii]λ6584 emission. These, labeled according to the nomenclature of Massaro et al. (2011a), are

• 1.  
  two circles of 2'' radius, centered on (i) the radio core and (ii) the [N ii]λ6584 emission line knot at the end of the south filament (core and s14, respectively);
• 2.  
  two polygons along the southwest filament, namely, sw4 (inner 4 kpc) and sw14 (remaining 13 kpc).

We chose, as background for the spectral analysis of the filament regions, a larger circle located ∼45'' to the northeast where no pointlike and/or extended X-ray sources are detected, to obtain the spectrum of the excess emission only instead of the line-of-sight average. We also performed the spectral analysis of the galaxy cluster emission, considering two polygonal regions, one at each side of the filaments and a background region at ∼100'' to compare the inner and outer galaxy cluster emission.

Background-subtracted X-ray spectra in the 0.5–4 keV energy range were extracted using the CIAO routine specextract and the background region containing neighboring ICM emission to the filament and analyzed using Sherpa v4.12.1 (Freeman et al. 2001). To ensure the validity of Gaussian statistics, spectral data were binned to at least 25 photons per bin.

We tried fitting the final spectra using an absorbed bremsstrahlung model; however, it failed to reproduce the soft X-ray excess below 2 keV. Thus, we adopted a collisional ionization gas model (xsapec ²⁴ ), absorbed by Galactic hydrogen column density (xswabs ²⁵ ; N_H,Gal = 3.8 × 10²⁰ cm⁻²; Kalberla et al. 2005). The xsapec model had three free parameters: normalization, plasma temperature kT, and metal abundance Z.

We confirmed the presence of a soft X-ray (i.e., 0.5–3 keV) excess spatially coincident with the [N ii]λ6584 filament (regions sw4 and sw14) at a level of confidence >5σ with respect to the surrounding ICM emission. Detection significance, reported in Gaussian equivalent, was computed assuming a Poisson distribution of photons in the inner background region.

Assuming the presence of emission lines below 2 keV (see Section 4), we created narrowband X-ray images, shown in Figure 3 (see, e.g., Massaro et al. 2013, 2015, 2018). We chose three energy ranges mainly attributable to (i) Fe (0.9–1.2 keV), (ii) Mg (1.2–1.5 keV), and (iii) S+Si (1.6–2.1 keV) ionized elements. X-ray flux maps were created by using monochromatic exposure maps set to the nominal energies of 1, 1.3, and 1.8 keV for Fe, Mg, and S+Si, respectively, to take into account the detector effective area at different energies.

Zoom In Zoom Out Reset image size

From the optical perspective, we modeled the main emission lines present in the MUSE spectra, i.e., Hβ, Hα, [O iii]λ5007, [O i]λ6300, [N ii]λ6584, and the [S ii] doublet at λ λ 6716, 6731 using Gaussian functions at different locations across the filament and the knot. The top panel of Figure 4 shows the MUSE spectrum of the optical knot sw14, as an example. The H β line was modeled using an additional broad component only present in the core and in region s4.

Zoom In Zoom Out Reset image size

4.1. Optical

4.2. X-Rays

We performed the X-ray spectral analysis in all regions described in Section 3, along optical filaments and in the galaxy cluster (excluding filaments, core, and knots). Best-fit results, obtained with the absorbed xsapec model, are reported in Figure 2. We also fitted regions sw4 and sw14 together and obtained a gas temperature of ${2.4}_{-0.3}^{+0.3}$ keV (with $Z={0.5}_{-0.2}^{+0.3}\ {Z}_{\odot }$ and reduced ${\chi }_{\nu }^{2}=0.857$ with ν = 59 degrees of freedom) for the southwestern filament. This is colder than the surrounding ICM, which has ${kT}={3.2}_{-0.1}^{+0.1}$ keV (reduced ${\chi }_{\nu }^{2}=0.906;\nu =207$). The ICM temperature is consistent with the average temperature found by Mazzotta et al. (2002) for the inner regions of MKW 3 s.

Metal abundance of the optical filament, $Z={0.3}_{-0.2}^{+0.2}$ for sw14 (with reduced ${\chi }_{\nu }^{2}=0.588;\ \nu =36$), appears to be marginally lower than that of the surrounding ICM ($Z={0.72}_{-0.07}^{+0.08}$). This is in agreement with literature results on other filaments (see, e.g., Fabian et al. 2011; McDonald et al. 2010).

Assuming that narrowband images trace the X-ray emission lines, the Mg emission has a better spatial association with the optical filaments, while Fe and S+Si show X-ray emission more extended than the filaments, as well as strong emission at the core and the s14 (knot) regions (see Figure 3). We computed the X-ray flux along the filament (regions sw4 and sw14) in the Fe, Mg, and S+Si band and compared them with optical and X-ray fluxes. The X-ray fluxes along the filament (regions sw4 and sw14) are ${F}_{\mathrm{Fe}}={2.9}_{-0.3}^{+0.2}\times {10}^{-14}$ erg s⁻¹ cm⁻² in the Fe band, ${F}_{\mathrm{Mg}}={1.9}_{-0.1}^{+0.1}\times {10}^{-14}$ erg s⁻¹ cm⁻² in the Mg band, ${F}_{{\rm{S}}+\mathrm{Si}}={2.3}_{-0.2}^{+0.1}\times {10}^{-14}$ erg s⁻¹ cm⁻² in the S+Si band, and ${F}_{0.5-3\mathrm{keV}}={1.32}_{-0.08}^{+0.05}\times {10}^{-13}$ erg s⁻¹ cm⁻² in the soft band (0.5–3 keV). Detection significance of the excess X-ray emission along optical filaments is above 5σ confidence level for both Mg and Fe, while it is not significant for the S+Si narrowband image, thus the S+Si emission along the filament is consistent with the ICM emission.

We also obtained F_Hα /F_Fe = ${2.9}_{-0.4}^{+0.4}\times {10}^{-2}$ and F_Hα /F_{0.5−3 keV} = ${6.5}_{-0.8}^{+0.7}\times {10}^{-3}$. In contrast with the filaments of 3CR 318.1, Sanders & Fabian (2007) and Fabian et al. (2011) found that those in Perseus have soft X-ray emission of the same order as the Hα emission, while the Hα emission is an order of magnitude above the Fe emission, which highlights the much lower ionization state here. Additionally, we obtained F_{[O III]}/F_{0.5−3 keV} = ${1.15}_{-0.14}^{+0.12}\times {10}^{-3}$, which is three orders of magnitude below the value obtained by Balmaverde et al. (2012) for the emission line regions in nine 3CR radio galaxies.

Finally, we derived the 0.5–3 keV, exposure-corrected, X-ray surface brightness profile with azimuthal bins of 10° centered on 3CR 318.1 core, excluding the inner 2'' and extending up to ∼30 kpc (see Jimenez-Gallardo et al. 2021 for additional details). We chose to make the bin containing the [N ii]λ6584 filament correspond to the 180° bin. The resulting X-ray surface brightness profile is shown in Figure 5. The background level is two orders of magnitude below the filament emission. The smooth increase of the surface brightness toward the [N ii]λ6584 filament supports the hypothesis that the filament could have originated due to outflows in the direction of the radio jets.

Zoom In Zoom Out Reset image size

The MURALES survey revealed the presence of optical filaments in 3CR 318.1, the BCG of MKW 3 s. Here we compared optical VLT/MUSE and X-ray Chandra observations to shed light on their physical origin.

From an optical perspective, intensity ratios of various rest-frame emission lines were used to distinguish between different ionization mechanisms. Possible explanations of filamentary emission include: ionization due to hot cooling ICM (Voit et al. 1994), ICM thermal conduction (McDonald et al. 2010), and reconnection diffusion (Fabian et al. 2011). However, individually, none of the ionization mechanisms listed can account for all line ratios simultaneously. As shown in Figure 4, line ratios measured in 3CR 318.1 show extremely low values of [O iii]/Hβ in the diagnostic diagrams; similarly, McDonald et al. (2012) found systematically lower [O iii]/Hβ values in filaments of cool core galaxy clusters and higher values of [N ii]/Hα, [O i]/Hα, and [S ii]/Hα than those found in galaxies in the SDSS. Although the diagnostic diagrams are used to discriminate between star-forming and AGN-dominated galaxies, they tend to fail when the ionization is due to more complex situations and/or different ionization mechanisms (see, e.g., Stasińska et al. 2008; Capetti & Baldi 2011; Balmaverde et al. 2018) and therefore they may have a limited validity in assessing the ionization conditions of BCGs. Thus, the line ratios in 3CR 318.1 imply that the emission from the filaments is due to a combination of ionization mechanisms. McDonald et al. (2012) suggested that line ratios are due to a combination of star formation and ionization from slow shocks (∼100–400 km s⁻¹; see also Allen et al. 2008). Since our results are similar to those of McDonald et al. (2012), we argue that their conclusion applies also to the case of 3CR 318.1. An additional contribution from self-ionizing cooling ICM is suggested by the low velocity dispersion (∼60 km s⁻¹) of the optical filaments in 3CR 318.1, which implies a small contribution of shocks to the total ionization. Therefore, we conclude that the underlying ionization mechanisms include a combination of photoionization due to star formation, self-ionizing cooling ICM, and a small contribution of ionization due to slow shocks. Additionally, although McDonald et al. (2012) already found a decrease of the emission line width with radius, thanks to the MUSE data, it was discovered that this decrease occurs sharply in the case of 3CR 318.1 (from ∼200 km s⁻¹ in the core to ∼60 km s⁻¹ along the filaments; see Balmaverde et al. 2019).

We detected an excess of X-ray emission above the ICM along the [N ii]λ6584 filaments. This X-ray filament is colder and despite the large uncertainties, appears to have a lower metal abundance than the surrounding ICM, in agreement with literature results (McDonald et al. 2010). The association between the X-ray and optical filamentary morphologies, together with the radio structure, suggests that the filaments could have originated from AGN-driven outflows in the direction of the radio jet. This scenario is in agreement with the smooth 0.5–3 keV X-ray surface brightness profile at the location of the [N ii]λ6584 filament and with simulations carried out by Qiu et al. (2019, 2020), in which cold filaments originate from warm outflows (10⁴–10⁷ K). Additionally, works such as Gaspari et al. (2018) and Voit (2021) predict that optical emission line nebulae would present velocity dispersions of ∼100–200 km s⁻¹ in the cases where the nebulae originated from compression and catastrophic cooling (see also Gaspari et al. 2012, 2013, 2015, 2017; Voit et al. 2017 for previous works on multiphase condensation). However, the [N ii]λ6584 filaments in 3CR 318.1 present very low velocity dispersions (∼60 km s⁻¹; see Figure 8 in Balmaverde et al. 2019), so the AGN uplift scenario could be favored.

We thank the anonymous referee for their useful comments that led to the improvement of the paper. This work is supported by the "Departments of Excellence 2018-2022 Grant" awarded by the Italian Ministry of Education, University and Research (MIUR) (L. 232/2016). This research has made use of resources provided by the Ministry of Education, Universities and Research for the grant MASF_FFABR_17_01. This investigation is supported by the National Aeronautics and Space Administration (NASA) grants GO9-20083X and GO0-21110X. F.M. is indebted to S. Bianchi for his valuable input on X-ray photoionizaton scenarios. A.J. acknowledges the financial support (MASF_CONTR_FIN_18_01) from the Italian National Institute of Astrophysics under the agreement with the Instituto de Astrofisica de Canarias for the "Becas Internacionales para Licenciados y/o Graduados Convocatoria de 2017." A.P. acknowledges financial support from the Consorzio Interuniversitario per la fisica Spaziale (CIFS) under the agreement related to the grant MASF_CONTR_FIN_18_02. W.F. and R.K. acknowledge support from the Smithsonian Institution and the Chandra High Resolution Camera Project through NASA contract NAS8-03060. G.V. acknowledges support from ANID programs FONDECYT Postdoctorado 3200802 and Basal-CATA AFB-170002.