Cold fronts and metal anisotropies in the X-ray cool core of the galaxy cluster Zw1742+3306
S. Ettori, F. Gastaldello, M. Gitti, E. O'Sullivan, M. Gaspari, F. Brighenti, L. David, A.C. Edge
AAstronomy & Astrophysics manuscript no. zw1742 c (cid:13)
ESO 2018September 25, 2018
Cold fronts and metal anisotropiesin the X-ray cool core of the galaxy cluster Zw 1742 + S. Ettori , , F. Gastaldello , , M. Gitti , E. O’Sullivan , M. Gaspari , F. Brighenti , L. David , A.C. Edge INAF, Osservatorio Astronomico di Bologna, via Ranzani 1, I-40127 Bologna, Italye-mail: [email protected] INFN, Sezione di Bologna, viale Berti Pichat 6/2, I-40127 Bologna, Italy INAF, IASF, Via E. Bassini 15, I-20133 Milano, Italy University of California Irvine, 4129, Frederick Reines Hall, Irvine, CA, 92697-4575, USA Physics and Astronomy Department, University of Bologna, via Ranzani 1, 40127 Bologna, Italy Harvard-Smithsonian centre for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA Max Planck Institute for Astrophysics, Karl-Schwarzschild-Strasse 1, 85741 Garching, Germany Department of Physics, University of Durham, south Road, Durham, DH1 3LE, UKAccepted on 14 May 2013
ABSTRACT
Context.
In recent years, our understanding of the cool cores of galaxy clusters has changed. Once thought to be relatively simpleplaces where gas cools and flows toward the centre, now they are believed to be very dynamic places where heating from the centralActive Galactic Nucleus (AGN) and cooling, as inferred from active star formation, molecular gas, and H α nebulosity, find an uneasyenergetic balance. Aims.
We want to characterize the X-ray properties of the nearby cool-core cluster Zw 1742 + − erg s − cm − in the 0.1–2.4 keV band) and H α wavelengths (H α luminosity > erg s − ). Methods.
We used
Chandra data to analyse the spatial and spectral properties of the cool core of Zw 1742 + z = . H α and presents the brightest central galaxy located in a diffuse X-ray emission with multiple peaks insurface brightness. Results.
We show that the X-ray cool core of the galaxy cluster Zw 1742 + ×
90 kpc and is alignedwith the cold fronts and with the X-ray emission on larger scales. We suggest that all these peculiarities in the X-ray emission ofZw 1742 + Key words.
X-ray: galaxies: clusters – (galaxies:) cooling flows – galaxies: clusters: general
1. Introduction
A considerable fraction of the local population of clusters ofgalaxies ( ∼ -10 M (cid:12) yr − . In the absence of a heating source, the gaswill accrete onto the BCG and form stars (Fabian 1994). Theend products of cooling in the forms of cold molecular cloudsand star formation are observed in many BCGs (e.g. Edge 2001,O’Dea et al. 2008) but at a level of at least an order of mag-nitude below those expected from uninterrupted cooling overthe age of clusters. X-ray observations with Chandra and
XMM-Newton provided crucial evidence against a simple cooling flowmodel establishing a lack of strong emission from gas coolingbelow ∼ T vir /
3, indicating much less cool gas than expected in a steady cooling flow and thus lowering the previous cooling ratesby a factor of 5-10 (e.g. Peterson & Fabian 2006). A compen-sating heat source must therefore resupply the radiative lossesand quench the cooling flow. The most likely heat source is me-chanical heating from the radio Active Galactic Nucleus (AGN)in the BCG at the centre of the cool core. The new establishedparadigm therefore involves a repetitive feedback cycle wherethe massive black hole in the central galaxy, fed by the gas alsoforming molecular clouds and stars, launches mechanically pow-erful radio jets / outflows that heat the surrounding atmosphere,slowing efficiently the rate of cooling. There is clear observa-tional evidence for the steps involved in the feedback process,since almost all cool cores harbour radio sources (Burns 1990,Sun 2009) that produce bubbles of relativistic plasma inflatedby the radio jets. These jets interact with the surrounding ICMcreating X-ray surface brightness depressions (or cavities) caus-ing heating as they rise through the ICM. Together with weakshocks associated with the outburst, the energies provided by theAGN are indeed comparable to those needed to prevent gas fromcooling. Many excellent reviews of this topic are now available(Fabian 2012, McNamara & Nulsen 2012, Gitti et al. 2012). a r X i v : . [ a s t r o - ph . C O ] M a y ttori et al.: The X-ray cool core in Zw 1742 + Fig. 1.
ACIS-I ( left ) and ACIS-S3 ( centre ) exposures of Zw 1742 + Right : a zoomed version of the ACIS-S3 image with bin-size of 0.492 arcsecand convolved with a Gaussian of 3 sigma in pixel.A distinctive feature of cool cores is the presence of an ironabundance peak, higher on average by a factor of 3 to 4 than thebulk of the ICM, consistent with being produced by stars associ-ated with the BCG (e.g. De Grandi & Molendi 2001, De Grandiet al. 2004). Several studies have shown that the iron abundanceprofile is broader than the stellar light profile of the BCG (e.g.Rebusco et al. 2006, Graham et al. 2006) indicating that met-als are diffusing outward, their distribution being broadened bymergers or AGN outflows, even though allowance should bemade for the opposite process of concentration of optical lightnaturally associated to the formation of the BCG (De Grandi& Molendi 2001). AGN bipolar outflows have recently receivedstrong observational and theoretical support: following early re-sults showing metal-rich gas along the cavities and radio jets ofsome individual clusters and groups (e.g. Simionescu et al. 2008,2009; Gitti et al. 2011; Kirkpatrick et al. 2009; O’Sullivan et al.2011), Kirkpatrick et al. (2011) found consistently in a sampleof ten clusters an anisotropic distribution of iron aligned withthe cavity and large scale radio emission axes. The radial extentof the metal outflow is found to scale with the jet power and itis greater than the extent of the inner cavities, showing this tobe a long-lasting effect sustained over multiple generations ofoutbursts. The amount of transported gas is substantial, and itis consistent with the results of simulations showing that AGNoutflows are able to advect ambient, iron-rich material from thecore to much larger radii of the order of hundreds of kpc (e.g.Gaspari et al. 2011a, 2011b).Another striking feature revealed by
Chandra and
XMM-Newton in cool cores is the almost ubiquitous presence of coldfronts (e.g. Ghizzardi et al. 2010). Cold fronts are sharp sur-face brightness discontinuities, interpreted as contact edges be-tween regions of gas with different entropies (see the review byMarkevitch & Vikhlinin 2007). In cool core clusters, cold frontsare most likely induced by minor mergers that produce a distur-bance on the gas in the core of the main cluster, displace it fromthe centre of the potential well, and decouple it from the under-lying dark matter halo through ram pressure (e.g. Ascasibar &Markevitch 2006). The oscillation of the gas of the core aroundthe minimum of the potential generates a succession of radiallypropagating cold fronts, appearing as concentric edges in the sur-face brightness distribution of the cluster. These fronts may forma spiral structure when the sloshing direction is near the plane ofthe sky and the merger has a non-zero angular momentum. Whenthe sloshing direction is not in the plane of the sky, concentric arcs are observed. The sequence of events is described in greatdetail in the simulations presented in Ascasibar & Markevitch(2006) and Roediger et al. (2011).Last but not least, observational signature of cool cores isthe fact that the BCGs at their centres have extensive opti-cal emission-line (H α ) luminosity (e.g. Crawford et al. 1999,McDonald et al. 2010). The emission can extend for tens of kpcand in the most extreme case of NGC 1275 in Perseus out to 100kpc. The ionization mechanism of these filaments is still muchdebated; however, the observational evidence increasingly pointsto their origin in the cooling ICM, as the strength of the H α emis-sion correlates with the intensity of blue continuum due to starformation and to the amount of molecular gas (e.g. Crawford etal. 1999, Edge 2001). The relation between short cooling times,star formation, and H α emission has been resolved in a sharpthreshold: high star formation rates and nebular emission arepreferentially found in BCGs in atmospheres with central cool-ing times shorter than 0.5 Gyr (or equivalently with central en-tropies lower than 30 keV cm , Rafferty et al. 2008; Cavagnoloet al. 2008). These phenomena can all be linked by the occur-rence of thermal instabilities, with warm gas condensing out ofthe hot keV phase and forming layers of H α surrounding thedense core of molecular gas (e.g. Gaspari et al. 2012 and ref-erences therein). Clearly H α emission is a tracer of cooling ac-tivity and it has been used to indirectly probe the evolution ofcool cores with time (McDonald 2011, Samuele et al. 2011).This connection can be also exploited locally to select addi-tional interesting objects to study the cooling/heating balance incool cores. This is the case of the galaxy cluster analysed in thepresent work, Zw 1742 + α emission associated with it.
2. Properties of Zw 1742 + We have selected Zw 1742 + S / N = .
12 in the
Planck
Early Sunyaev-Zeldovich catalogue (PlanckCollaboration 2011), from the ROSAT Brightest Cluster Sample(BCS) sample (203 clusters in the northern hemisphere, 90%complete at 4 . × − erg s − cm − , Ebeling et al. 1998) amongthe objects with X-ray fluxes greater than 10 − erg s − cm − inthe 0.1–2.4 keV band and H α -lumininosity > erg s − asquoted in Crawford et al. (1999). Most of the objects satisfy-ing this selection are well-known cool core clusters (i.e. ranked + frontW_SfrontW_NcavS ? cavN ?frontE BCG
Fig. 2.
Deviations from the smoothed X-ray emission. (
Left ) Unsharp mask obtained by subtracting two Gaussian convolved images(with FWHM of 2 and 20 pixels). Some of the features discussed in the text are shown. (
Right ) Exposure-corrected X-ray imageoverplotted with the best-fit 2D β − model (green contours) and the positive (red) and negative (blue) residuals at 68.3%, 90%, 95%,and 99.73% level of confidence. Fig. 3. + . (cid:48)(cid:48) in size, and correspond to circular spectral extraction regions of radius 4 − (cid:48)(cid:48) inradius containing > ∼ ∼ + + H =
70 km s − Mpc − , Ω m = − Ω Λ = . × erg s − (Ebeling et al. 1998). The brightest cluster galaxy at (RA, Dec) =(17 h m s , +32 ◦ m s ) has an H α luminosity of 9.7 × erg s − and analysis of the optical continuum implies a star for-mation rate of 0.3 M (cid:12) yr − (Crawford et al. 1999). The BCG + has a total infrared luminosity of 7.4 × erg s − derived from Spitzer
IRAC amd MIPS photometry (Quillen et al. 2008) thatcorresponds to an equivalent star formation rate of 3.7 M (cid:12) yr − (O’Dea et al. 2008). The discrepancy between the optical andMIR star formation rates can be accounted for by dust absorp-tion in the optical. Finally, the BCG contains a bright (74mJy at5GHz, Gregory & Condon 1991), flat spectrum ( α ∼ − .
5) radiosource, but is also associated with a steeper spectrum component( α ∼ − .
2) detected at 74 and 325MHz in VLSS (Cohen et al.2007) and WENSS (Rengelink et al. 1997). Without higher res-olution radio imaging, the relationship between these two com-ponents remains unclear.We present here the results obtained from our recent
Chandra
ACIS-S observation (PI: Gitti; nominal exposure of 44ksec performed in November 2009). We also discuss the 8 ksecACIS-I exposure (PI: Murray; collected in January 2007) avail-able in the archive.
3. X-ray spatial and spectral analysis
The ACIS-S exposure was reprocessed with the most recentcalibration files at the moment of the data analysis (
CALDB4.4.10 ). We applied the prescriptions recommended for ACISdata reduction using the
CIAO 4.4 package. No flares are evi-dent in the light-curve and the good-time intervals cumulate to atotal of 45.4 sec. The ACIS-I exposure was also reprocessed al-lowing us to use 8.0 ksec out of the 8.4 ksec archived. The latterdataset was used only to produce images. Hereafter, a Galacticabsorption corresponding to a column density n H of 3 . × particle cm − is assumed.In Fig. 1, we show the raw image obtained in the soft (0.5–2keV) band. The unsharp mask image, obtained by subtracting theraw image convolved with a Guassian of 2 (cid:48)(cid:48) and 20 (cid:48)(cid:48) , is shown inFig. 2 with some interesting features overplotted. We also fitted a2D β − model, with a fixed centre on the cD galaxy. We obtainedthe best-fit parameters r c = . ± . β = . ± . (cid:15) = . ± .
01, and angle of ellipticity of 0 ± β − model from theraw image are shown in Fig. 2. The central AGN is clearly de-tected as a point source in X-rays. This source seems rather sta-ble in the two Chandra observations (2007-01-28 for the ACIS-Iobservation, 2009-11-26 for the ACIS-S observation). A fit witha power law (and fixed Galactic absorption) provides a constrainton the photon index of 1 . ± . . × − erg s − cm − and 9 . × − in the 0.5–2 and2–10 keV bands, respectively, corresponding to an X-ray lumi-nosity of 6 . × erg s − and 1 . × erg s − .The X-ray centroid, estimated after the exclusion of the allthe detected point-sources (BCG included) over a region withan increasing radius up to 200 arcsec from the BCG, oscillatesaround the brightest central galaxy, moving to the east by 3 arc-sec and, then, to the west by about 4 arcsec. Over larger scales,we modelled the extended X-ray emission enclosing 60%, 70%,80%, and 90% of the X-ray light with an ellipse with elliptic-ity (equal to 1 − b / a , where b and a are the minor and majoraxis, respectively) varying from 0.29 to 0.14 and an angle from175 to 165 degrees (measured anti-clockwise from a CartesianX-axis), confirming the elliptical, irregular shape of the X-raysurface brightness.For the spectral analysis, the cluster’s X-ray emission wasmodelled with both one and two thermal models ( apec modelin XSPEC 12.7.1 ). The spectral fit was performed in 0 . − apec model fits to spectra with1500 net counts. Since the spectral extraction regions are typi-cally larger than the map pixels, individual pixel values are notindependent and the maps are analogous to adaptively smoothedimages, with more smoothing in regions of lower surface bright-ness. To investigate the 2D variation in entropy and pressure inthe ICM, we follow the method of Rossetti et al. (2007) andestimate a pseudo-density, based on the square root of the nor-malization of the apec model divided by the area of the spec-tral extraction region. This is proportional to the integrated elec-tron density along the line of sight. The maps of pseudo-density,pseudo-entropy and pseudo-pressure should thus trace any pro-jected structure in these important quantities.Combined with the unsharp mask, these maps allow us toidentify a number of interesting features, which are further char-acterized through surface brightness profiles extracted in circularannuli from the exposure-corrected 0.5-2 keV image, centreedon the BCG (see Fig. 4). As we discuss in the following sec-tions, we find robust indication for a cold front at ∼ (cid:48)(cid:48) (23kpc) west of the BCG and a front at about 80 kpc to the east (seeFig. 5).In Fig. 6, we plot the best-fit gas temperature, estimated witha single thermal (1T) model, as a function of the distance of thecentre of these regions from the brightest cluster galaxy. Thetypical radial structure of a cooling core cluster with an ambientgas temperature of about 5 keV appears.We discuss below the three most interesting regions: thefront in surface brightness at west, another possible front at east,and an anisotropic elongated structure overabundant in metalscompared to the neighbouring ICM. The map of residuals and a mild decrement in the surface brightness profile also suggest thepresence of an X-ray cavity at 10 kpc north of the BCG and ofan even weaker cavity to the south of the BCG. Given the poorcounts statistic available in this area, we can only estimate a gastemperature of about 3 keV with metallicity of 0.65 times the so-lar value (0.9 when a 2T thermal model is adopted; see Table 1).In the remaining regions, a single-phase gas provides a gooddescription of the thermal emission in the northern and southernsectors (labelled sectorN and sectorS in Table 1), with the regionat the south that presents a smoother temperature distributionaround 4.7 keV. Located at about 23 kpc from the BCG, this front shows a cleardecrement in the surface brightness (Fig. 4) and a rise in thetemperature estimate from about 2 keV to 5 keV once a singlethermal model is adopted. However, the gas inside the disconti-nuity (labelled frontW-in in Table 1; see circular region in Fig. 4that includes frontW-inN and frontW-inS ) shows a statisticallyhigh significance for the need of a second thermal component.This gas, therefore, seems multiphase, with a cool thermal com-ponent at 1.5 keV and a second component hotter than 2.6 keV(at 1 σ ). + . : : . . : . . : : . Right Ascension (2000) D ec li n a t i on ( ) frontE frontW_SsectorS cavNsectorN frontW_N Fig. 4. ( Left ) Regions of interest for the spectral analysis (see Table 1) overplotted to the unsharp mask: moving outwards, sectorN and sectorS are divided in two regions ( in and out ); frontE is divided in 2 regions across the front; frontW-in indicates the circularregion in front of the edge; frontW-inN and frontW-outN, ..., out4N are the regions across the front pointing to the north-west; frontW-inS and frontW-outS, ..., out4S are the regions pointing to south-west. ( Right ) Surface brightness profiles extracted from theexposure-corrected 0.5–2 keV image fixing the centre on the BCG. The angles that define the sectors are measured anti-clockwisefrom the positive X-axis (angle = ◦ ). The arrows indicate the positions of the two fronts where there is a sharp steepening of thegradient. At the bottom, the ratios between the surface brightness profiles extracted in sectors and the profile obtained from theremaining X-ray emission assumed to be more undisturbed are plotted. Fig. 5.
Surface brightness profiles of the fronts detected in our analysis with the best-fit with a broken-power law describing the gasdensity profile overplotted with a dashed line. (
From left to right ): to the west, northern and southern sectors, at about 12 kpc fromthe centre fixed at (RA, Dec) = (17 h m s , +32 ◦ m s ) and 30 kpc from the BCG; to the east, at 80 kpc from the centrelocated on the BCG; to the south/west at about 60 kpc from the BCG. The centre of the annular regions used to extract each profilehas been chosen to best match the curvature of the radial bins with the shape of the edges as seen in Fig. 4.We performed detailed spatial and spectral analysis in theregion around this front. In proximity of the edge in the surfacebrightness profile, we modelled the gas density with a brokenpower law n e ( r ) = n e , in (cid:16) rR f (cid:17) − β in , r < R f , n e , out (cid:16) rR f (cid:17) − β out , r > R f , (1)where R f indicates the location of the front. We projected thismodel along the line of sight and fit it to the surface bright- ness profile. We measured a ratio between the gas densities at R f of 3 . ± .
10, when the sector between 335 and 30 de-grees (Cartesian X-axis = 0 ◦ ; north is 90 ◦ ) is considered. Oncewe zoom into this region and define two additional sectors,one pointing to the north (0–80 degrees) and the other south-ward (295–0 degrees), we estimate a ratio of 3 . ± .
06 and3 . ± .
10, respectively. Analysing the corresponding spec-tra extracted from boxes aligned along the front, we measure T = . ± .
38 keV and 4 . ± .
42 in the northern and south-ern sector, with metal abundance of 0 . ± .
21 and 0 . ± . + T (keV) A ( Z (cid:12) ) C-stat/DOF 2 T (keV) A ( Z (cid:12) ) C-stat/DOF F-testBCG-noAGN 2 . + . − . . + . − . . + . − . , 0 . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . , 4 . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . , 1 . + . − . . + . − . . + . − . . + . − . . > − . , 1 . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . > − . , 1 . > − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . , 1 . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . > − . , 2 . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . . + . − . Table 1.
Spectral analysis of the regions of interest by adopting one and two absorbed apec models in the energy range 0 . − Fig. 6.
Best-fit temperatures, from the 1T model for the selectedregions listed in Table 1, as a function of the distance of theregion where the spectrum has been accumulated from the BCG.solar, respectively; when we fix these values as second compo-nents contributing to the total emission in the internal part andleave free the normalization, we obtain T = . ± .
14 keV and1 . ± .
05 in the northern and southern sector, with metal abun-dance of 0 . ± .
16 and 0 . ± .
10 solar, respectively. Thesemeasurements correspond to a ratio between the temperatures of 2 . ± .
38 and 3 . ± .
37 for the sectors to the north and south,respectively, which agrees well with the values observed in thedensity discontinuity. Combining the estimates of the jumps indensity and temperature, we can evaluate the jump in the pres-sure profile to be 0 . ± .
13 in the north, 1 . ± .
11 in the south,consistent with unity. Thus we confirm that the front is indeed acontact discontinuity and not a shock.
At about 80 kpc from the BCG, we notice a decrement in thesurface brightness profile extracted in the sector between 130and 200 degrees and an increase in the gas temperature from 3.7to 5 keV, with the metallicity that decreases by a factor of 2. Thisgas appears isothermal, statistically speaking.The fit with a broken power law in the gas density distri-bution provides a measure of the discontinuity of 1 . ± . . + . − . and 5 . ± .
35 keV inside and outside the front,respectively, with an estimate of the pressure jump P out / P in of1 . ± . ∼
60 kpc) from the BCG along the directionto south-west (between 305 and 335 degrees in our referenceframe). Along this sector, we measure an increasing temperatureup to 6 . ± .
9) keV in the annuli just before the break in thesurface brightness profile shown in Fig. 5, with a lower tem-perature (5 . ± . ∼ . ± .
4) would imply a gas density jump of about 1.3, for aMach number ≈ . . ± . +
14 4
Fig. 7.
Regions of interest for the spectral analysis plotted over (left) the exposure-corrected image and (right) the abundance mapindicating the spectroscopically estimated values of metallicity following the colour bar on the right. The regions labelled from 1 to16 (here, only 4 and 14 are indicated for the sake of clarity) have the measured metal abundance plotted in Fig. 8. The BCG (greencircle) and the two fronts (green arcs) are indicated.
The abundance map in Fig. 7 highlights an elongated structurewith metal abundance higher by about 50% than the mean valueestimated in the surrounding regions. A detailed spectral analy-sis shows that a single-phase gas is a good model for the spectraextracted from boxes with dimensions of about 30 ×
20 arcsec and indicates a clear enhancement in the metal abundance in astructure with size ≈ (276 ×
86) kpc and an orientation at anangle of 153 degrees from the Cartesian X-axis (see Fig. 7 and8). This metal-rich elongated structure has an orientation verysimilar to both the overall distribution of the X-ray light ( ∼ ∼
167 degrees; all these angles are measured anti-clockwisefrom the Cartesian X-axis). This structure also has an associatedenhancement of the X-ray surface brightness as evaluated fromthe map of the residuals.An additional abundance map has been obtained considering6000 (instead of 1500) net counts to allow the modelling witha 2T thermal model with the abundance tied between the twocomponents. This analysis allows us to appreciate the impact ofa 2T model on the estimate of the ICM metallicity in the clusterand to highlight the iron-bias due to a single thermal modellingof a multi-phase gas. In Fig. 9 we show the comparison betweenthe 2T abundance map obtained with 6000 count, the same mapfitted with a single temperature model, and the original high-resolution 1T abundance map. The only remarkable effect of the2T fits is to remove the biased low abundances in the cool core,whereas the metallicity values are unaffected elsewhere.In the right panel of Fig. 8 we show the metal abundanceprofiles extracted along two different directions, one followingthe elongated structure and the other almost orthogonal to it. Wequantify (and quote in the plot) the difference in metallicity as anexcess in iron mass associated to the metal-rich structure. Thisexcess is measured as ∆ M Fe = A Fe m H Z (cid:12) (4 π × ) . D lum / (1 + z ) (cid:80) i ∆ A i K . i V . i , where A Fe is the atomic weight of iron; m H isthe proton mass; the luminosity distance D lum of 342.7 Mpc atthe cluster redshift is adopted. The sum is done using the spec-troscopic best-fit values of abundance A i and normalization K i (from the adopted apec model) and assuming a cylindrical vol- ume V i (with height given from the extension of the region con-sidered and a radius of the base of 43 kpc; no significant changeis obtained by assuming slightly different geometries). The es-timated value of ∆ M Fe ≈ . × M (cid:12) within 150 kpc is fullyconsistent with the iron production expected from the BCG infew Gyrs (e.g. De Grandi et al. 2004, B¨oringher et al. 2004).For example, assuming a BCG B-band luminosity of 1 . × L B , (cid:12) , we find that in the last 5 Gyr the SN-Ia+stellar winds in-ject ∼ . × M (cid:12) of iron in the ICM. For this estimate weadopt a current SNIa rate of 0.15 SNu, with a time dependency S Nu ( t ) ∝ t − . and a mean stellar metallicity of 1.6 solar (seeB¨oringher et al. 2004 for further details). The SN-II are not ex-pected to be important contributors of iron in the last few Gyrs,given the estimated star formation rate ( ∼ . (cid:12) yr − , O’Dea etal. 2008).
4. Discussion and conclusions
The system Zw 1742 + ≈ . ∼ σ when the scatter around the mean is consideredand a corresponding excess in iron mass of about 2 . × M (cid:12) (see Fig. 7 and 8). A plausible cavity a few tens of kpc north ofthe BCG, and a weaker counter-part to the south, can be eitherthe product of the activity of the central AGN or the apparentproduct of the enhanced emission in the neighbouring regionscaused by the ongoing sloshing activity. If real, these cavities areprobably connected to breaks in the jets/outflows and are morelike to be destroyed from the central turbulence/sloshing, sug- + Fig. 8.
Distribution of the metal abundance of the ICM along the structure described in Fig. 7. (
Left ) Estimates along the regionslabelled from 1 to 16. The apparently coherent structure corresponds to the regions between 4 and 14 (red points) where themeasures of the metal abundance are systematically higher than the values measured in the surrounding gas by about 50%. Thedifference between the weighted averages of the abundance values along this structure and in the “off” region (a combination of“off1” and “off2” regions) used as reference is ∼ . σ when the scatter is considered (shaded region and dotted lines, respectively)and increases to 4 . σ when the error on the mean is used. The dimension of the elongated structure is about (3 . ×
1) arcmin ≈ (276 ×
86) kpc . ( Right ) Spectral measurements as function of the distance from the BCG. The measurements were done alongtwo different directions, one along the elongated structure (red dots) and the other almost orthogonally to it (blue diamonds). Thetwo values match at about 150 kpc. At the top, the excess in iron mass in unit of solar mass as a function of increasing radius isindicated. The elemental abundances are given with respect to the solar values from Anders & Grevesse (1989). c t s ,
2T 6000 c t s ,
1’ / 86.15 kpc c t s , Fig. 9.
Maps of the measurements of the metal abundance as obtained with ( left ) a 2T model and requiring 6000 net count, ( centre )a single temperature model and the same number of net counts, ( right ) the 1-T model and 1500 net counts (i.e. the same abundancemap shown in Fig. 3). The BCG (white circle) and the front to the west (white arc) are indicated.gesting that they are not the main source of the feedback actingin Zw 1742 + + ature structures. However, they also prove that the sloshing alonecannot, for instance, redistribute a metal distribution with a steeppeak originating from stellar mass loss of the central galaxy intoa flatter one (as for the case of M87). To produce that, they re-quire additional processes such as diffusion induced from turbu-lence (e.g. Rebusco et al. 2005, 2006) caused by AGN-inflatedbuoyant bubbles or repeated minor-mergers, for example. ActiveGalactic Nuclei activity, in particular, can act in a complemen-tary way to the gas sloshing, the former being more effective inthe very cluster centre, the latter at larger radii. They concludethat the observed metal distribution, even across the cold fronts,depends strongly on the initial (unknown) distribution; sloshingaffects it towards the cold fronts and broadens it on larger scales.It is, therefore, difficult to outline a picture of the formation ofthe metal elongated structure, whether it was already in situ as aproduct of some major-merger that formed Zw 1742 + R Fe = (cid:16) P jet / erg s − (cid:17) . kpc, with a rms scatter aroundthe fit of about 0.5 dex. From Fig. 7, we estimate the maxi-mum radius R Fe of ≈
150 kpc at which a significant enhance-ment in metallicity has been detected. It corresponds to a jetpower of about 10 erg s − , which is consistent with the re-sults of simulations showing that metals originating mainly fromSnIa explosions occurring in the cD galaxy are easily transportedalong the jet-axis up to 150-200 kpc when the AGN is very ac-tive (see e.g. Gaspari et al. 2011a, b). This activity would natu-rally explain the evidence that: (i) the metal distribution is highlyanisotropic, with a bipolar symmetric pattern; (ii) at larger radii,the metal distribution expands laterally, forming two or morelobes of conical geometries, since ambient pressure diminishes(and thus also collimation); (iii) eastern and western arms showsimilar metal contrast, slightly more metallic along the spine(observation that is indeed difficult to reproduce by sloshing);(iv) cold/warm dense gas uplift is very common for mechanicalfeedback, especially within tens kpc, and this could provide asimple explanation for the clear cold front near the centre (alongwith the small X-ray peak shift). Overall, we argue that the pre-vious AGN outflow event has uplifted cold dense gas in the in-ner few tens of kpc (more shifted towards west), creating thecold front. At the same time, it has enriched the hotter gas up to300 kpc. The direction of the bipolar outflow has probably beendistorted in an S-shaped geometry near the kpc base, becauseof denser clumps (possibly creating fragmented bubbles like theones marginally detected to the north and south of the BCG).For the outflow scenario to work in the presence of the elon-gated metal-enrich structure, the relatively central ICM shouldhave acquired ≈ × M (cid:12) of iron before the occurrence of theAGN outburst. The iron reservoir may pile up in the core duringa more quiescent Gyr evolution, or be the result of the complex dynamics associated with previous feedback events. Turbulenceand bulk motion may alter the metallicity structure, inducinga more isotropic distribution in iron abundance, although thisis balanced by the recurrent outflows (with typical duty cycle ≈
100 Myr). The detected weak shock could be, therefore, aby-product of a recent period of activity (bubbles are not essen-tial to produce AGN feedback). Nevertheless, it is difficult totrace the activity of previous events in X-ray surface brightnessmaps, especially at r >
100 kpc (the underlying profile steeplydecreases with radius and bubbles/shocks quickly decay). Forexample, Figs. 11 (panel h) and 14 (panel b) in Gaspari et al.(2011a) show that the X-ray surface brightness map is very reg-ular up to 200 kpc, while the metal anisotropy is still present.A follow-up in radio wavebands (3h EVLA observation at1.4 GHz, PI: Gitti, are in progress) and with H α resolved imag-ing would help to characterize the cool core emission and itsinterplay with central AGN activity, and would also better de-fine the occurrence of thermal instabilities, with layers of H α emitting gas that form from warm gas condensing out of the hotphase and that surround the dense core of molecular gas (e.g.Gaspari et al. 2012 and references therein). Acknowledgments
We thank the anonymous referee for the useful commentsthat improved the presentation of the work. We acknowledgethe financial contribution from contract ASI-INAF I/023/05/0and I/088/06/0. SE acknowledges financial support from
Chandra grant GO0-11136X and from FP7-PEOPLE-IRSES-CAFEGroups (grant agreement 247653). FG acknowledges fi-nancial contributions from the Italian Space Agency throughASI/INAF agreement I/032/10/0 for XMM-Newton operations.
References