MNRAS 000, 1–5 (2022) Preprint 22 July 2022 Compiled using MNRAS LATEX style file v3.

Discovery of inverse-Compton X-ray emission and robust estimation of

magnetic field in a galaxy group
F. Mernier ,1,2★ † N. Werner ,3 J. Bagchi ,4 M.-L. Gendron-Marsolais ,5,6 Gopal-Krishna,7
M. Guainazzi ,1 A. Richard-Laferrière ,8 T. W. Shimwell9 and A. Simionescu ,2,10,11
1 European Space Agency (ESA), European Space Research and Technology Centre (ESTEC), Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands
2 SRON Netherlands Institute for Space Research, Niels Bohrweg 4, 2333 CA Leiden, The Netherlands
3 Department of Theoretical Physics and Astrophysics, Faculty of Science, Masaryk University, Kotlářská 2, Brno, CZ-611 37, Czech Republic
4 Department of Physics & Electronics, CHRIST (Deemed to be University), Hosur Road, Bengaluru, 560029, India
arXiv:2207.10092v1 [astro-ph.GA] 20 Jul 2022

5 European Southern Observatory, Alonso de Córdova 3107, Vitacura, Casilla, 19001, Santiago de Chile, Chile
6 Instituto de Astrofìsica de Andalucìa (IAA-CSIC), Glorieta de la Astronomìa, 18008 Granada, Spain
7 UM-DAE Centre of Excellence in Basic Sciences (CEBS), Vidyanagari, Mumbai - 400098, India
8 Institute of Astronomy, Madingley Road, Cambridge CB3 0HA, UK
9 ASTRON, Netherlands Institute for Radio Astronomy, Oude Hoogeveensedijk 4, Dwingeloo, 7991 PD, The Netherlands
10 Leiden Observatory, Leiden University, PO Box 9513, NL-2300 RA Leiden, The Netherlands
11 Kavli Institute for the Physics and Mathematics of the Universe (WPI), University of Tokyo, Kashiwa 277-8583, Japan

Accepted XXX. Received YYY; in original form 2022 July 19

ABSTRACT
Observed in a significant fraction of clusters and groups of galaxies, diffuse radio synchrotron emission reveals the presence of
relativistic electrons and magnetic fields permeating large-scale systems. Although these non-thermal electrons are expected
to upscatter cosmic microwave background photons up to hard X-ray energies, such inverse-Compton (IC) X-ray emission has
so far not been unambiguously detected on cluster/group scales. Here we report the first robust (4.6 𝜎) detection of extended
IC X-ray emission from MRC 0116+111, a group of galaxies. This unambiguous detection provides the most direct and least
model-dependent estimate of the volume-averaged magnetic field within a galaxy group. Such estimates can serve as a fulcrum
to theories of magnetic field generation within the largest gravitationally bound systems in the Universe.

Key words: magnetic fields – galaxies: clusters: individual: MRC 0116+111 – galaxies: clusters: intracluster medium – X-rays:
galaxies: clusters

1 INTRODUCTION and evolution of cluster magnetic fields, which is fragmentary at the

moment (Vazza et al. 2021). Other methods for estimating magnetic
At the dawn of the high-energy astrophysics era, the origin of the
fields in clusters and groups of galaxies are, however, afflicted by ob-
bright, extended X-ray emission seen towards clusters and groups of
servational biases. The estimates based on the well known technique
galaxies was debated (Brecher & Burbidge 1972; Lea et al. 1973).
of Faraday rotation (e.g. Böhringer et al. 2016) depend critically on
The advent of dedicated space observatories established that these
the local topography of magnetic fields along the line of sight and,
systems shine in X-rays via thermal bremsstrahlung and line emission
furthermore, remain highly sensitive to the foreground Galactic in-
from a hot (107−8 K), collisionally ionised medium permeating them
terstellar dust. The alternative method utilises the radio synchrotron
(Sarazin 1986). However, radio wavelength detections of diffuse syn-
emission solely, which depends not only on the intensity of magnetic
chrotron radiation from relativistic electrons in some clusters (e.g.
field, but also on the uncertain energy density of their relativistic
van Weeren et al. 2019) implies the existence of a wide-spread rela-
electron population. Usually, these two parameters are commonly
tivistic plasma which should also upscatter cosmic microwave back-
estimated by assuming equipartition between magnetic and particle
ground (CMB) photons up to X-ray energies via inverse-Compton
energy densities – whereby the total energy density is nearly min-
(IC) scattering.
imised (Feretti et al. 2012). The validity of this assumption on various
The importance of detecting this diffuse IC X-ray emission in clus-
physical scales, however, remains beset with formidable uncertainty
ters/groups of galaxies has been long highlighted (Petrosian et al.
(e.g. Petrosian et al. 2008). The degeneracy inherent to the equipar-
2008; Feretti et al. 2012) because in combination with radio observa-
tition assumption can be effectively broken using the detection of IC
tions, it can yield the most direct estimate of the intracluster magnetic
X-ray emission, enabling a reliable estimate of the energy density of
field. The latter is, in fact, a critical input for understanding the origin
the relativistic electron population.

★ E-mail: francois.mernier@esa.int Measuring diffuse IC emission at large scales constitutes thus a

† ESA Research Fellow. promising way to boost our understanding of cosmic magnetic fields.

© 2022 The Authors

2 F. Mernier et al.
From the earliest hunts (Rephaeli et al. 1987) to recent attempts three 2.2 Spectral analysis
decades later (Wik et al. 2012; Ota et al. 2014; Cova et al. 2019;
Aiming to analyse spectrally the entire extent of the diffuse radio
Rojas Bolivar et al. 2021), observations have yielded, however, only
emission (to maximise the statistics), we extract data from an ellipti-
upper limits for any diffuse IC X-rays from clusters. The rare detec-
cal region similar to that of Mernier et al. (2019, see their fig. 2). The
tions pertain to the Coma cluster (Rephaeli et al. 1999; Rephaeli &
redistribution matrix file (RMF) and ancillary response file (ARF)
Gruber 2002; Fusco-Femiano et al. 1999, 2004), Abell 85 (Bagchi
are obtained from the same elliptical region. Due to the fairly small
et al. 1998), the Ophiuchus cluster (Eckert et al. 2008), and the Bul-
X-ray extent of MRC 0116+111 (< 1 arcmin), the background can be
let cluster (Petrosian et al. 2006; Ajello et al. 2010); however all
directly subtracted from our spectra. We choose a background region
these claims have been found spurious (or, rendered controversial) in
(i) located on the same chip as the source region for all instruments
subsequent more sensitive observations (Rossetti & Molendi 2004;
of all observations, (ii) devoid of X-ray sources, and (iii) of the same
Fusco-Femiano et al. 2007; Eckert et al. 2007; Lutovinov et al. 2008;
size and shape as the source region. The small relative separation be-
Durret et al. 2005; Wik et al. 2009, 2014). A key drawback with these
tween this background region and the source (∼ 1.6 arcmin) ensures
objects is that the intracluster gas at such high temperatures radiates
vignetting effects to be negligible. The impact of this method has
thermal X-rays that dominate the IC component at a few keV ener-
been already discussed in Mernier et al. (2019), however it is further
gies, rendering the latter practically undetectable. This circumstance
addressed below.
often mandates complementary observations at higher energies (i.e.
We use the SPEX fitting package to fit most of our data in the
even beyond 10 keV), for example using the NuSTAR or INTEGRAL
0.55–5 keV energy band (to minimise further contamination from
space observatories, in order to confirm the putative IC X-rays. How-
the soft foreground and the hard instrumental background) with the
ever, this often gets mired in imperfect cross-calibration between the
the C-statistics fitting method. Spectral bins are grouped optimally
instruments covering different energy bands, which adds substan-
(Kaastra & Bleeker 2016), though with a gradually wider rebinning
tial uncertainty. Moreover, the hot gas pervading these dynamically
with higher energy (within factors 5–80) to ensure that the count rate
disturbed clusters often has a multi-temperature structure, particu-
of every bin remains significantly higher than zero. The latter is also
larly due to recently shocked regions (Donnert et al. 2017), which
necessary to avoid possible biases from C-statistics in low counts,
complicates spectral modelling.
background-subtracted fits (see the SPEX manual).
In this Letter, we report results from a deep XMM-Newton ob-
Our modelling consists of a collisionally ionised plasma emis-
servation of the galaxy group MRC 0116+111 (or OTL 0116+111 at
sion model with a Gaussian-shaped multi-temperature distribution
its discovery; Joshi & Singal 1980; Gopal-Krishna et al. 2002). In
of width 𝜎𝑇 fixed to 0.2 (i.e. gdem, thermal model) and a power
addition to its relatively cool gaseous medium (𝑘𝑇 ' 0.7–0.8 keV)
law model (i.e. po; non-thermal IC emission model), both redshifted
(Mernier et al. 2019), this group is known to be a source of spec-
(𝑧 = 0.131; Mernier et al. 2019) and absorbed by atomic interstellar
tacularly bright diffuse radio emission possibly related to an ancient
hydrogen (𝑛H = 3.81 × 1020 atoms cm−2 ; Kalberla et al. 2005). Fol-
pair of relativistic plasma bubbles, and potentially originating from
lowing its close to universal value, the metallicity (𝑍) of the thermal
previously intense activity in the supermassive black hole of its cen-
model is fixed to 0.3 Solar with all relevant abundances tied to the
tral brightest galaxy (Bagchi et al. 2009). The combination of these
Fe value (Mernier et al. 2018). The slope Γ of the po model can be
two extreme properties makes this group a prime target for detecting
well constrained by the radio spectral index 𝛼syn of the source below
X-ray diffuse IC emission. Specifically, the emission spectrum of
its frequency break at ∼400 MHz (𝛼syn = 0.55 ± 0.05; Bagchi et al.
its thermal plasma should peak at soft X-ray energies, with negligi-
2009), and is thus fixed to Γ = 1.55 (see also Mernier et al. 2019).
ble emission above a few keV, where the non-thermal IC emission,
Figure 1 shows our integrated spectrum. Found with an average
shaped spectrally as a power law, is hence expected to dominate.
best-fitting temperature of 𝑘𝑇 = 0.62 ± 0.04 keV, the pure ther-
Throughout this study, we assume that H0 = 70 km s−1 Mpc−1 ,
mal model (red) systematically leaves an X-ray excess in the 2–
ΩM = 0.27 and ΩΛ = 0.73. At the distance of MRC 0116+111 (𝑧 =
5 keV band, which is well reproduced by a non-thermal component
0.131), 1 arcmin corresponds to 179.0 kpc. Chemical abundances
(blue) and detected at a high 4.6 𝜎 significance level. Since the
are given in the (proto-) solar units of Lodders et al. (2009). Unless
multi-temperature structure of the gas expectedly extends the spec-
mentioned otherwise, the quoted uncertainties are 1 𝜎.
tral profile of the thermal component to higher energies, assuming
a single-temperature plasma in the group (cie model) enhances the
non-thermal detection to an even higher confidence level (> 7 𝜎).
Further multi-temperature models are discussed in Sect. 3.
2 DATA AND RESULTS
2.1 XMM-Newton data reduction
2.3 Spatial analysis
This work is based on the three XMM-Newton EPIC observations of
MRC 0116+111 available so far – taken respectively in January 2014 Motivated by the presence of a non-thermal component dominating
(ObsID: 0722900101) and December 2020 (ObsID: 0864110101, the X-ray emission beyond ∼2 keV, we now analyse its spatial distri-
0864110101) – for a total raw exposure of ∼290 ks. The data reduc- bution. With the notable exception of two bright sources ∼80 arcsec
tion process follows the pilot study of Mernier et al. (2019) and uses NE and SW away from the group (thus outside of our region of in-
the XMM-Newton SAS (Science Analysis System) software (v18.0.0) terest), the search for point-like sources (performed via the routine
with the up-to-date calibration files (December 2021). Following edetect_chain) leads to no significant detection in the vicinity of
the standard pipeline recommendations, the EPIC MOS (1 and 2) MRC 0116+111. The brightest of the two point sources is in fact
and EPIC pn data are processed using, respectively, the commands a background quasar (SDSS J011904.92 +112420.4; 𝑧 = 1.0; Lyke
emproc and epproc. For each of the 9 datasets, we carefully identify et al. 2020) that is unrelated to the group.
flaring events and discard them following the 2𝜎-clipping method Figure 2 shows that, unlike its soft thermal counterpart (left), the
described in Mernier et al. (2015), leaving us with 219.9 ks and diffuse hard X-ray emission (right) has a peak that coincides strik-
170.5 ks of clean MOS and pn exposures, respectively. ingly with the eastern peak of radio synchrotron emission imaged at

MNRAS 000, 1–5 (2022)

 Inverse-Compton X-ray emission in a galaxy group 3

10 2 Stacked data (MOS+pn, 3 observations) Table 1. Investigation of various systematic effects and their impact on the
Total model
gdem model (thermal) IC detection confidence (see text).
po model (inverse Compton)

10 3 Description C-stat/d.o.f. 𝐹IC,2 keV IC det.

Counts/s/keV

(10−15 erg/s/cm2 /keV) confid.

Baseline fit𝑎 278.1/200 1.55 ± 0.22 7.1 𝜎

10 4 Multi-T: gdem 262.5/200 1.30 ± 0.28 4.6 𝜎

Multi-T: wdem 273.9/200 1.50 ± 0.22 6.8 𝜎
Multi-T: cie+cie 255.9/198 1.18 ± 0.27 4.3 𝜎
MOS only 180.1/139 1.27 ± 0.27 4.7 𝜎
10 15 pn only 94.7/52 1.85 ± 0.34 5.4 𝜎
Modelled bkg 173.9/193 1.76 ± 0.49 3.6 𝜎
Rel. err.

0 Fit: [0.55–4] keV 274.7/200 1.43 ± 0.22 6.4 𝜎

Fit: [0.55–10] keV 332.6/209 1.37 ± 0.21 6.4 𝜎
1 SPEXACT v2.05 278.7/200 1.63 ± 0.23 7.2 𝜎
0.6 1 2 3 4 5
Energy (keV) AtomDB v3.0.9 459.2/404 1.33 ± 0.22 6.0 𝜎
Free 𝑍 and 𝑛H 260.5/198 0.92 ± 0.26 3.5 𝜎
Figure 1. Combined EPIC MOS and pn spectrum of MRC 0116+111 (stacked Free Γ, [0.8–5] keV 175.9/129 1.11 ± 0.53 2.1 𝜎
here for clarity). The additive thermal (gdem) and non-thermal components
are displayed in unbinned resolution for visibility. Spectral residuals to the
total model (i.e. [data - model] / model) are shown in the bottom panel. stand them with minimal interdependence. Table 1 summarises our
tests and their outcome.
First, in principle a hard tail seen in cluster/group spectra could be
621 MHz (Bagchi et al. 2009), while the slight offset seen between the (at least partly) explained by an over-simplified thermal modelling
radio synchrotron western peak and the secondary (southern) IC X- of its gas component (Cova et al. 2019). To test this hypothesis, in
ray lobe, conceivable if relativistic electrons over-densities decouple addition to the gdem fit discussed above, we replace the cie model
locally from regions of high magnetic field intensities, argues for no- from our baseline fit by, successively (i) a power law-distributed
table plasma weather and dynamics within this group. Although the multi-temperature model (wdem) with its slope 𝑝𝑇 fixed to 0.25 (as
origin of the group’s diffuse radio plasma has been debated (Bagchi found in e.g. the Virgo cluster; Kaastra et al. 2004); and (ii) a two-
et al. 2009), its spatial coincidence found here with the IC X-ray temperature model (cie+cie) with free temperatures and emission
emission argues for aged remnants of the synchrotron radio lobes measures. Table 1 shows that, in both cases, the additional IC emis-
arising from a past episode of nuclear activity in the central galaxy. sion remains significant at more than 4 𝜎. We note that, in the gdem
Except one (early-type) galaxy in its (fainter) southern lobe, no model, reproducing the cooler emission and the hard X-ray excess
optical counterpart coincides with the extent of the hard X-ray emis- would require the temperature distribution width to be exceedingly
sion. In addition, a short (∼18 ks) snapshot observation from Chan- broad (𝜎𝑇 ' 0.6). Not only has such an extreme value never been
dra allows to exclude any significant contribution from possible point reported in the literature, it would also imply that half of the group’s
sources within the entire region encompassed by the radio emission emission originates from a gas with 𝑘𝑇 > 5 keV, extending even
(Mernier et al. 2019). Finally, we have verified that each of the two further to 𝑘𝑇 = 10 keV and beyond. This latter consideration seems
hard X-ray lobes (25 arcsec across) largely exceeds the point spread hardly physical for a group having so few galaxies and with so faint
function (PSF) of the EPIC instruments. The apparent extent of the X-ray luminosity (Bagchi et al. 2009; Mernier et al. 2019). Similarly,
background quasar is naturally explained by its saturated brightness adding a third cie component to our cie+cie modelling results
contrast on the figures for better visibility of the group itself – even does not alter the robustness of our results. Since the gdem model is
though its emission profile is found to be perfectly consistent with the physically the most realistic multi-temperature distribution for this
shape of the EPIC PSF (and to drop to a negligible level well before group (the wdem model being suited mostly for cool-core clusters),
the group emission). Altogether, this makes us confident that the hard we adopt it as our most accurate assumption.
X-ray emission of the group is, indeed, significantly extended. Second, we must ensure that the additional power law component
is not an instrumental artefact. To do so, we re-evaluate our free
parameters for the MOS and pn spectra separately. We find that the
IC component is detected in both instruments with high significance
(4.7 𝜎 and 5.4 𝜎 for MOS and pn, respectively).
3 SYSTEMATICS AND ROBUSTNESS OF THE RESULTS
Third, it is essential to ensure that the background is correctly
In order to confirm the robustness of our spectral results, it is essential accounted for in our spectral analysis. In addition to our robust
to ensure that any systematic effects are smaller than the statistical background-subtracting strategy described above, an alternative ap-
uncertainties. This is especially true for sources like MRC 0116+111 proach consists of modelling spectrally the background instead of
where, despite the deep exposure, the number of net counts remains subtracting it (e.g. Mernier et al. 2015). The thermal foreground,
limited. These sources of systematics can all be addressed by per- originating from the local hot bubble and from the Galactic thermal
forming a series of alternative fitting tests, in which the flux density emission, is then modelled with an (unabsorbed) cie model plus
and significance of the IC emission can be directly compared to an (absorbed) cie model, with free emission measures as well as
a simple cie+po modelling case. Whilst the modelling ultimately temperatures fixed to 𝑘𝑇LHB = 0.08 keV and 𝑘𝑇GTE = 0.2 keV, re-
adopted in this work is that of a multi-temperature modelling, choos- spectively. The cosmic X-ray background is modelled with a power
ing a single-temperature approach as formal baseline modelling is law of index ΓCXB = 1.41 and an emission measure of 3.7×1049 pho-
indeed preferred here to better isolate systematic effects and under- tons s−1 keV−1 , adopted to account for the integrated emission of

MNRAS 000, 1–5 (2022)

4 F. Mernier et al.

0.3 - 2 keV 2 - 4.5 keV

24:00.0

24:00.0
Declination

Declination
30.0

30.0
11:23:00.0

11:23:00.0
22:30.0

22:30.0
30 arcsec 30 arcsec
90 kpc 90 kpc

06.0 05.0 04.0 03.0 02.0 01.0 1:19:00.0 59.0 58.0 18:57.0 06.0 05.0 04.0 03.0 02.0 01.0 1:19:00.0 59.0 58.0 18:57.0
Right ascension Right ascension

0 0.034 0.1 0.24 0.52 1.1 2.2 4.3 8.7 17 350 0.003 0.0089 0.021 0.044 0.092 0.19 0.37 0.75 1.5 3

Figure 2. Stacked, background-subtracted, vignetting-corrected XMM-Newton EPIC surface brightness images of MRC 0166+111, overlayed with radio contours
(621 MHz, GMRT). Left: Soft band (0.3–2 keV), smoothed with a Gaussian of radius 4. Right: Hard band (2–4.5 keV), smoothed with a Gaussian of radius 9.

point sources unresolved below the (2–8 keV) flux threshold limit of 4 DISCUSSION AND IMPLICATIONS
2.2 × 10−5 erg cm2 s−1 obtained from our observations. In addition,
we also model two “instrumental” components and leave them un- There are a number of claimed detections of IC X-ray emission from
folded through the ARF. While the residual soft proton background X-ray “hot spots” seen in radio-galaxies (e.g. Hardcastle et al. 2002)
is modelled with a simple power law, the hard particle background is and from collimated jets of relativistic particles in blazars (e.g. Wor-
modelled with a broken power law and a series of Gaussian compo- rall et al. 2020), although validity of such an interpretation for the
nents to account for the emission of fluorescent lines (Mernier et al. latter objects has been questioned in some cases (e.g. Breiding et al.
2015). Despite this very conservative approach resulting into many 2017). Regardless of this, the present finding differs both in the nature
more degrees of freedom than in our baseline fitting, the IC compo- of the system and the physical scale and kinematics of the IC emitting
nent remains significantly detected (3.6 𝜎) with similar best-fitting plasma. Whilst our hard X-ray image reveals a non-thermal contri-
results. bution from the vicinity of the likely remnant radio lobes associated
Fourth, whilst our adopted fitting range arguably constitutes the with the central dominant galaxy (see Fig. 2), the flux contributed
best compromise of maximising the true signal from the source while by these lobes accounts for less than half of the total IC emission
avoiding additional sources of noise, it is important to ensure that our detected from the entire volume of the group, the remainder fraction
results are not much affected by this choice – in particular by the upper originates from the intragroup gas beyond the remnant radio lobes. To
energy limit as main tracer of the IC emission. Encouragingly, the our knowledge, this is the first time that extended IC emission – asso-
IC emission remains confidently constrained at similar flux densities ciated both with relativistic electrons pervading the entire group and
when varying this upper limit from 4 keV to 10 keV. with a more concentrated, lobe-like component – has been robustly
Fifth, our results are independent on the atomic code and databases detected in a system containing multiple galaxies.
used to model the thermal emission of the gas. In particular, we find Both the diffuse radio emission and the diffuse IC X-ray emission
no notable difference in our results when using AtomDB (v3.0.9) or share a common origin from a population of relativistic electrons
an older version of SPEXACT (v2.05) in our fits. permeating large-scale structures. Whereas the flux of the former
Finally, although thawing more parameters in our fits would in- depends on both the relativistic electron density and the volume-
evitably lead to larger statistical uncertainties, it may be instructive to averaged magnetic field (hence implying a degeneracy between these
explore how doing so would affect our results. Encouragingly, we find two a priori unknown values), the flux of the latter depends solely on
that free 𝑛 𝐻 and 𝑍 still result in a comfortable IC detection (3.5 𝜎). the relativistic electron density. Consequently, the ratio between these
In this case, we note that the best-fit metallicity is found to drop two fluxes provides unique constraints on the magnetic field strength
to ∼0.07 Solar, which seems hardly physical given the &30% Solar (Feretti et al. 2012; Ota et al. 2014; Mernier et al. 2019), devoid of
levels found in virtually all clusters and groups (Mernier et al. 2018). any bias unlike other methodologies (Sect. 1). In fact, upper limits
In fact, fixing the metallicity to the most plausible values (i.e. be- to IC X-ray emission from previous work on clusters only translate
tween 0.5 and 1 Solar) results in more line emission – hence a steeper into lower limits on intracluster magnetic fields of & 0.1 − 1 𝜇G
thermal profile and, in turn, further boosting the significance of the (Bartels et al. 2015; Cova et al. 2019; Rojas Bolivar et al. 2021). Our
IC emission. Similarly, despite the reliability of the radio spectral present robust detection of IC emission allows accurate and essen-
index (Bagchi et al. 2009), we also verify whether our results remain tially model-independent estimate of the volume-averaged magnetic
essentially unchanged if we free the IC X-ray photon index Γ. Here field of the group, which is found to be (1.9 ± 0.3) 𝜇G. In addi-
we choose to restrict our fitting range to 0.8–5 keV (and fixing the tion to being close to the lower limit of & 2.6 𝜇G (90% confidence;
temperature of the thermal model to its best-fitting baseline value Mernier et al. 2019), it lies well within the lower limits estimated for
0.64 keV) to ensure the IC slope to reproduce the hard tail prefer- rich clusters. Cosmological magneto-hydrodynamical simulations do
entially. This fit leads to Γ = 1.5 ± 0.6, in good agreement with our predict 𝜇G level magnetic fields in the cores of rich clusters; however
initially adopted value. Quite remarkably, even in this extreme case they suggest fields that are smaller by an (or, sometimes, even two)
the IC emission remains detected with more than 2 𝜎 confidence. order of magnitude in the case of groups having gas densities and

MNRAS 000, 1–5 (2022)

 Inverse-Compton X-ray emission in a galaxy group 5
temperatures similar to MRC 0116+111 (e.g. Donnert et al. 2018). Bartels R., Zandanel F., Ando S., 2015, A&A, 582, A20
Interestingly, were the magnetic field in our object indeed so weak, Böhringer H., Chon G., Kronberg P. P., 2016, A&A, 596, A22
the corresponding diffuse IC X-ray emission would be a factor & 30 Brecher K., Burbidge G. R., 1972, Nature, 237, 440
stronger than even our highly significant detection reported here. Breiding P., Meyer E. T., Georganopoulos M., Keenan M. E., DeNigris N. S.,
Therefore, it may be instructive to devise suitable alternative mecha- Hewitt J., 2017, ApJ, 849, 95
Cova F., et al., 2019, A&A, 628, A83
nisms for the seeding and growth of cosmic magnetic fields, allowing
Donnert J. M. F., Beck A. M., Dolag K., Röttgering H. J. A., 2017, MNRAS,
them to be stronger in groups of galaxies. Whereas NuSTAR observa- 471, 4587
tions could in principle allow to set even stronger constraints on the Donnert J., Vazza F., Brüggen M., ZuHone J., 2018, Space Sci. Rev., 214,
IC X-ray emission of this source, targeted searches using current and 122
future observing facilities will be vital for ascertaining whether the Durret F., Lima Neto G. B., Forman W., 2005, A&A, 432, 809
present object has a magnetic field that is representative of a wider Eckert D., Neronov A., Courvoisier T. J. L., Produit N., 2007, A&A, 470, 835
population of groups. This will be attempted by taking advantage of Eckert D., Produit N., Paltani S., Neronov A., Courvoisier T. J. L., 2008,
(i) synergies between X-ray surveys (such as eROSITA) and radio A&A, 479, 27
surveys (such as LOFAR or SKA), and (ii) the outstanding spectral Feretti L., Giovannini G., Govoni F., Murgia M., 2012, A&ARv, 20, 54
and/or photon collecting capabilities offered by the next generation Fusco-Femiano R., Dal Fiume D., Feretti L., Giovannini G., Grandi P., Matt
G., Molendi S., Santangelo A., 1999, ApJ, 513, L21
of X-ray observatories (e.g. XRISM, Athena).
Fusco-Femiano R., Orlandini M., Brunetti G., Feretti L., Giovannini G.,
Grandi P., Setti G., 2004, ApJ, 602, L73
Fusco-Femiano R., Landi R., Orlandini M., 2007, ApJ, 654, L9
ACKNOWLEDGEMENTS Gopal-Krishna Kulkarni V. K., Bagchi J., Melnick J., 2002, in Pramesh Rao
A., Swarup G., Gopal-Krishna eds, IAU Symposium Vol. 199, The Uni-
F.M. thanks Soumyajit Mandal for fruitful discussions. N.W. verse at Low Radio Frequencies. p. 159
is supported by the GACR grant 21-13491X. M.-L.G.-M. ac- Hardcastle M. J., Birkinshaw M., Cameron R. A., Harris D. E., Looney L. W.,
knowledges financial support from grant RTI2018-096228-B-C31 Worrall D. M., 2002, ApJ, 581, 948
(MCIU/AEI/FEDER,UE), from the coordination of the participation Joshi M. N., Singal A. K., 1980, Memoirs of the Astronomical Society of
in SKA-SPAIN financed by the Ministry of Science and Innovation India, 1, 49
(MICIN), and from the State Agency for Research of the Spanish Kaastra J. S., Bleeker J. A. M., 2016, A&A, 587, A151
Kaastra J. S., et al., 2004, A&A, 413, 415
Ministry of Science, Innovation and Universities through the "Cen-
Kalberla P. M. W., Burton W. B., Hartmann D., Arnal E. M., Bajaja E., Morras
ter of Excellence Severo Ochoa" awarded to the Instituto de As- R., Pöppel W. G. L., 2005, A&A, 440, 775
trofìsica de Andalucìa (SEV-2017-0709). G.-K. thanks the Indian Lea S. M., Silk J., Kellogg E., Murray S., 1973, ApJ, 184, L105
National Science Academy for a Senior Scientist fellowship. A.R.L. Lodders K., Palme H., Gail H. P., 2009, Landolt B&ouml;rnstein, 4B, 712
is supported by the Gates Cambridge Scholarship, by the St John’s Lutovinov A. A., Vikhlinin A., Churazov E. M., Revnivtsev M. G., Sunyaev
College Benefactors’ Scholarships, by NSERC through the Postgrad- R. A., 2008, ApJ, 687, 968
uate Scholarship-Doctoral Program (PGS D) under grant PGSD3- Lyke B. W., et al., 2020, ApJS, 250, 8
535124-2019 and by FRQNT through the FRQNT Graduate Studies Mernier F., de Plaa J., Lovisari L., Pinto C., Zhang Y. Y., Kaastra J. S., Werner
Research Scholarship - Doctoral level under grant #274532. A.S. is N., Simionescu A., 2015, A&A, 575, A37
supported by the Women In Science Excel (WISE) programme of Mernier F., et al., 2018, Space Sci. Rev., 214, 129
Mernier F., Werner N., Bagchi J., Simionescu A., Böhringer H., Allen S. W.,
the Netherlands Organisation for Scientific Research (NWO), and ac-
Jacob J., 2019, MNRAS, 486, 5430
knowledges the World Premier Research Center Initiative (WPI) and Ota N., Nagayoshi K., Pratt G. W., Kitayama T., Oshima T., Reiprich T. H.,
the Kavli IPMU for the continued hospitality. This work is based on 2014, A&A, 562, A60
observations obtained with XMM-Newton, an ESA science mission Petrosian V., Madejski G., Luli K., 2006, ApJ, 652, 948
with instruments and contributions directly funded by ESA mem- Petrosian V., Bykov A., Rephaeli Y., 2008, Space Sci. Rev., 134, 191
ber states and the USA (NASA). SRON is supported financially by Rephaeli Y., Gruber D., 2002, ApJ, 579, 587
NWO. Rephaeli Y., Gruber D. E., Rothschild R. E., 1987, ApJ, 320, 139
Rephaeli Y., Gruber D., Blanco P., 1999, ApJ, 511, L21
Rojas Bolivar R. A., Wik D. R., Giacintucci S., Gastaldello F., Hornstrup A.,
Westergaard N.-J., Madejski G., 2021, ApJ, 906, 87
DATA AVAILABILITY Rossetti M., Molendi S., 2004, A&A, 414, L41
Sarazin C. L., 1986, Reviews of Modern Physics, 58, 1
The original data presented in this article are publicly available from
Vazza F., et al., 2021, Galaxies, 9, 109
the XMM-Newton Science Archive (https://nxsa.esac.esa.int/nxsa-
Wik D. R., Sarazin C. L., Finoguenov A., Matsushita K., Nakazawa K., Clarke
web/). The corresponding XMM-Newton data are reduced with the T. E., 2009, ApJ, 696, 1700
SAS software (https://www.cosmos.esa.int/web/xmm-newton/sas) Wik D. R., Sarazin C. L., Zhang Y.-Y., Baumgartner W. H., Mushotzky R. F.,
and further analysed spectrally with the SPEX fitting pack- Tueller J., Okajima T., Clarke T. E., 2012, ApJ, 748, 67
age (https://www.sron.nl/astrophysics-spex). Additional derived data Wik D. R., et al., 2014, ApJ, 792, 48
products can be obtained from the main author upon request. Worrall D. M., Birkinshaw M., Marshall H. L., Schwartz D. A., Siemigi-
nowska A., Wardle J. F. C., 2020, MNRAS, 497, 988
van Weeren R. J., de Gasperin F., Akamatsu H., Brüggen M., Feretti L., Kang
H., Stroe A., Zandanel F., 2019, Space Sci. Rev., 215, 16
REFERENCES
Ajello M., Rebusco P., Cappelluti N., Reimer O., Böhringer H., La Parola V., This paper has been typeset from a TEX/LATEX file prepared by the author.
Cusumano G., 2010, ApJ, 725, 1688
Bagchi J., Pislar V., Lima Neto G. B., 1998, MNRAS, 296, L23
Bagchi J., Jacob J., Gopal-Krishna Werner N., Wadnerkar N., Belapure J.,
Kumbharkhane A. C., 2009, MNRAS, 399, 601

MNRAS 000, 1–5 (2022)