AGN Jets and a Fanciful Trio of Black Holes in the Abell 85 Brightest Cluster Galaxy

A85 is a galaxy cluster that is in the process of merging with two subclusters (Ichinohe et al. 2015). As a bright X-ray cluster with a complex structure, A85 has been intensively studied with all modern X-ray observatories: Chandra, XMM-Newton and Suzaku (e.g., Kempner et al. 2002; Durret et al. 2003; Schenck et al. 2014).

The galaxy population of A85 was recently analyzed at optical wavelengths by Agulli et al. (2016) who determined the spectroscopic luminosity function with 460 confirmed cluster members.

Owen & Ledlow (1997) and Schenck et al. (2014) made clear detections of radio emission emanating from the A85 cluster and its BCG. Radio relics located ∼320 kpc from the core of A85 were discovered by Slee et al. (2001)—see also Schenck et al. (2014) and Bagchi et al. (1998). These relics are likely the result from shocks of past cluster mergers.

In the optical and X-rays, SDSS J004150.75–091824.3 is a bright (g = 14.03 mag), point-like source located ∼14'' away from the nucleus of the A85 BCG—see Figure 1. SDSS J004150.75–091824.3 is also the closest X-ray point source to the nucleus of the BCG. The core of the BCG shows a diffuse X-ray emission while SDSS J004150.75–091824.3 is bright and point-like.

Zoom In Zoom Out Reset image size

SDSS J004150.75–091824.3 was included in the photometric selection of quasars from the SDSS by Richards et al. (2004, 2009). SDSS J004150.75–091824.3 was found to have a photometric redshift of z = 0.925 in Richards et al. (2004). This value was later revised to z = 0.675 in Richards et al. (2009).

SDSS J004150.75–091824.3 was identified as a possible active galactic nucleus (AGN) in the detailed X-ray study carried out by Durret et al. (2005). These authors identify SDSS J004150.75–091824.3 as an X-ray point source on their Chandra data and also identify an optical counterpart on a SDSS image (see their Figure 2). SDSS J004150.75–091824.3 is located within an X-ray cavity (Durret et al. 2005; Ichinohe et al. 2015).

Given the projected proximity of SDSS J004150.75–091824.3 to the center of the A85 BCG, López-Cruz et al. (2014) surmised that stellar light could have biased the redshift and classification given by Richards et al. (2009). López-Cruz et al. (2014) speculated that SDSS J004150.75–091824.3 could be "a third" supermassive black hole (SMBH) associated with the A85 brightest cluster galaxy. The other two black holes of A85 would have been a hypothetical binary pair in the core of the BCG. López-Cruz et al. (2014) claimed the presence of a substantial light deficit on the nuclear region of the A85 BCG. López-Cruz et al. (2014) interpreted this light deficit as evidence for the presence of an "ultramassive" SMBH.

When galaxies merge, their central black holes are thought to consolidate into one entity. During the process of black hole coalescence, SMBH binaries are thus created (Begelman et al. 1980). Stars that come in close proximity to a binary SMBH can be slingshot away through gravitational interactions (Begelman et al. 1980; Ebisuzaki et al. 1991; Milosavljević & Merritt 2001). Over time, as this process repeats itself, a SMBH creates a deficit of stars in its vicinity.

The presence of a binary SMBH can then be inferred by the detection of stellar light deficits on the optical surface brightness profiles of, for instance, elliptical galaxies (e.g., Graham 2004; Dullo 2019). When light deficits are present in surface brightness profiles, the steep slope of the profile in the outer parts of the galaxy becomes flatter, sometimes even close to constant, with decreasing radii (e.g., Graham 2004).

Given the small spatial scales involved, the accurate characterization of the properties of stellar cores and, among them, their light deficits, require Hubble Space Telescope data (Ferrarese et al. 2006). Light deficits associated with SMBH occur in spatial scales of a few hundred parsecs in galaxies located megaparsecs away. For instance, cores in galaxies belonging to the Virgo cluster have spatial scales of 1''–4'' (Ferrarese et al. 1994).

With few exceptions (e.g., 3C 75; Owen et al. 1985), unequivocal detections of close pairs of binary AGN remain rare. Binary black holes are elusive, even when dedicated observing campaigns are carried out (Tingay & Wayth 2011).

In the following sections a new radio map of the A85 BCG with the highest resolution ever achieved is introduced. This radio map is used to search for the presence of a binary AGN. A new optical spectrum of SDSS J004150.75–091824.3 is also presented. This spectrum is used to make a precise determination of the redshift for this source. A new Gemini spectrum of the A85 BCG is also shown. High-resolution imaging obtained with Gemini and Chandra are used as ancillary data to place the new observations in a wider context. Finally, by presenting optical data obtained under different atmospheric conditions we illustrate the effect of seeing on optical surface brightness profiles.

At the distance of A85 (D = 233.9 Mpc) 1'' corresponds to 1.075 kpc.

The Chandra data shown in Figure 1 was analyzed and presented by Ichinohe et al. (2015). Details of the observations and data reduction for this data set are given in that reference.

Ichinohe et al. (2015) created a Chandra image that is a factor of ∼5 deeper than the initial Chandra image of ∼37 ks analyzed by Kempner et al. (2002). Ichinohe et al. (2015) also included XMM-Newton and Suzaku data out to the virial radius of the cluster. Ichinohe et al. (2015) found evidence that the A85 cluster is undergoing two mergers and has gas sloshing out to a radius of r ∼ 600 kpc from its center.

Figure 1 also shows a Gemini image of the A85 BCG. The Gemini r-band image was obtained with the Gemini Multi-Object Spectrograph (GMOS). This data set was discussed in depth in Madrid & Donzelli (2016). A new Gemini image obtained under poor seeing conditions is also presented in Section 5 (Gemini program GS-2016B-Q-87). The data reduction for this new image is identical to the one described in Madrid & Donzelli (2016).

The Gemini spectrum of the A85 BCG was obtained with GMOS-South on spectroscopic mode. Five exposures of 900 s were obtained on 2016 September 14. The grating used was R400 which has its blaze wavelength at 764 nanometers. The slit width was 15.

The first task of the data reduction process is to apply gprepare to the raw science data obtained through the Gemini archive. gprepare applies a flatfield and performs overscan subtraction. The spectrum is extracted from the fits file by using the task gsextract. This tasks extracts the intensity values along the spectral dimension of the CCD chip. Calibration is carried out with the task gswavelength. The details of the reduction procedure used here were described in Madrid & Donzelli (2013).

The spectroscopic data for SDSS J004150.75–091824.3 was obtained with the Optical System for Imaging and low Intermediate Resolution Integrated Spectroscopy (OSIRIS) on the Gran Telescopio Canarias. OSIRIS is an optical imager and spectrograph. A 2 × 2 binning was used for the CCD resulting in a spatial scale of 026 per pixel. Likewise, the resulting spectral resolution is 2.12 Å pixel⁻¹. The slit had a width of 1'' and the grating in use was R1000B.

The spectrum of SDSS J004150.75–091824.3 presented here was obtained on 2017 August 17. Raw science data, bias, flats and arcs were retrieved from the Grantecan public archive.

Both science data extraction and sky subtraction were done using the pyraf task apall. In the same way, arcs, bias and flats are processed with apall. The science spectra had the bias removed and were normalized by the flat. The HgAr and Ne calibration arcs were used for wavelength calibration. The emission lines given on the reference arcs provided by the observatory were used as part of the wavelength calibration. The spectra were extracted using a spatial width of 7 pixels, or 18. The spectrophotometric standard star Feige 110 is used to flux calibrate the spectra using the tasks standard, sensfunction and calibrate.

Karl G. Jansky Very Large Array (VLA) data was obtained from the National Radio Astronomy Observatory public archive. A85 was observed with the VLA in its A configuration on 2018 March 25. The total observation time was 4 hr, resulting in 2.8 hr on-source after overheads. The observations used the X-band receiver with 4 GHz of correlated bandwidth centered at 10 GHz.

The VLA data were processed with a series of tasks found within the Common Astronomy Software Applications (CASA). Standard interferometric data reduction procedures were followed to calibrate the delay, bandpass, phase, amplitude and flux. One round of phase-only self-calibration was also included. During imaging, a Briggs robust weighting of +0.5 was used to suppress point-spread function (PSF) sidelobes and achieve a resolution of 029 × 019 at a position angle of 32°. The final image had a measured sensitivity of 3 μJy/beam.

The presence of a binary black hole in the A85 BCG was postulated based on the analysis of its surface brightness profile (López-Cruz et al. 2014). However, the obvious effect of poor seeing has been ignored in recent studies that attempt to use ground-based data without sufficient resolution to study the surface brightness profile of the A85 BCG. This section highlights the risks of inferring the presence of a light deficit, or any other nuclear property of a distant galaxy, using ground-based data.

The BCG of A85 has been reported during recent years as having both a light deficit and a light excess in its core (López-Cruz et al. 2014; Bonfini et al. 2015; Madrid & Donzelli 2016; Mehrgan et al. 2019). These apparently contradicting results are a clear example of the effects of data quality on optical surface brightness profiles.

The effect of seeing on surface brightness profiles can be easily modeled, as we do in this section. Atmospheric turbulence blurs and smears images obtained with ground based telescopes. Poor seeing degrades and erases the innermost structure of surface brightness profiles. This effect is not only intuitive but was well defined over four decades ago by Schweizer (1979), among others. More recently, Côté et al. (2006) showed that data with poor seeing miss the presence of structures in galactic nuclei.

Let's consider a galaxy whose light profile can be described by two main components: a nuclear PSF and an outer spheroid. The nuclear component can be accurately modeled as a Gaussian and the main component can be represented by a Sérsic function (Sérsic 1968).

The surface brightness profile of the above galaxy observed with a ground based telescope can be evaluated by convolving its profile with the seeing. Here, seeing is considered as the distorting effects of both atmospheric turbulence and telescope optics.

Figure 4 shows the effect of seeing on the innermost regions of a surface brightness profile. Observing a galaxy light profile with excellent seeing recovers the actual profile with no distortions, as shown in the top panel of Figure 4. Both the nuclear and spheroidal components are detected when the light profile is obtained under ideal seeing conditions.

Zoom In Zoom Out Reset image size

On the other hand, when the same light profile is observed with poor seeing, its central part is entirely misrepresented, as shown in the bottom panel of Figure 4. The central nuclear component is fully blurred by seeing effects. Under bad seeing conditions, instead of detecting a central Gaussian component the slope of the profile becomes flat toward the nucleus of the galaxy.

A surface brightness profile observed with bad seeing could give the illusion of a light deficit. Poor seeing also alters the light profile creating a break on its slope. The above can mislead observers to believe on the detection of a depleted core and its associated SMBH when galaxies are observed with poor seeing conditions.

López-Cruz et al. (2014) used data obtained with the Kitt Peak National Observatory (KPNO) 0.9 m telescope to analyze the central surface brightness profile of the A85 BCG. The seeing for the KPNO 0.9 m data is reported to be 167. Bonfini et al. (2015) published a detailed analysis of the surface brightness profile of A85 using data obtained with the Canada–France–Hawaii telescope (CFHT). The CFHT data analyzed by Bonfini et al. (2015) has subarcsecond resolution: 074. Bonfini et al. (2015) found that the light deficit claimed by López-Cruz et al. (2014) did not exist. Moreover, contrary to a light deficit, Bonfini et al. (2015) find a modest light excess in the core of the A85 BCG.

Madrid & Donzelli (2016) obtained Gemini data of the A85 BCG with a seeing of 056. Madrid & Donzelli (2016) confirmed the presence of a light excess in the core of this galaxy initially found by Bonfini et al. (2015). The superior seeing of the Gemini data allow Madrid & Donzelli (2016) to detect an additional nuclear component within the inner 2 kpc of the center of the A85 BCG. This nuclear component is well resolved, with a FWHM of 085. At the distance of A85 this nuclear component has a size of ∼0.9 kpc (Madrid & Donzelli 2016).

The empirical effect of seeing on the central surface brightness profile of the A85 BCG is shown in Figure 5. This figure shows the light profile derived with Gemini data obtained during two different epochs with different seeing conditions. Figure 5 also shows the KPNO 0.9 m data used by López-Cruz et al. (2014). Note that the true inner structure is lost to both Gemini and KPNO 0.9 m under poor seeing conditions.

Zoom In Zoom Out Reset image size

The Gemini data presented in Figure 5 allow us to quantify the impact of seeing. The surface brightness profile with excellent seeing and the Gemini profile with poor seeing have a small, but measurable, difference of 0.01 mag arcsec⁻¹ at r = 4.5 kpc. This difference increases to 0.02 mag arcsec⁻¹ at r = 2.0 kpc, and to 0.04 mag arcsec⁻¹ at r = 1.0 kpc. As shown in Figure 5, the two Gemini profiles begin to diverge at about ∼1.5 kpc from the nucleus of the galaxy.

The fact that the profile obtained with bad seeing fails to accurately record the true profile of the galaxy at such large radii is crucial. López-Cruz et al. (2014) claims that the A85 BCG has a cusp radius of r_γ  = 4.57 ± 0.06 kpc. This value is within the range where the surface brightness profile is affected by seeing effects. It is precisely this cusp radius that is used to infer the presence of a SMBH.