1. Introduction
The
$\Lambda$
CDM model posits that structures in the Universe formed hierarchically – smaller structures emerged first and later merged into larger ones. This scenario is commonly referred to as a bottom-up scenario in the literature. Massive elliptical galaxies constitute one specific example within this broader model. Specifically, the hierarchical development of high-mass elliptical galaxies at
$z = 0$
occurred in two stages. First, gas-rich (wet) mergers formed the galaxy’s core and facilitated the creation of its central supermassive black hole (SMBH). The dense gas content of these wet mergers triggered an intense phase of star formation. This rapid, dissipative process was largely completed by
$z \sim 2$
. The compact galaxies formed during this stage are referred to as red nuggets (Oser et al. Reference Oser, Ostriker, Naab, Johansson and Burkert2010; Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015; Buote & Barth Reference Buote and Barth2018; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018; Buote & Barth Reference Buote and Barth2019; Schnorr-Müller et al. Reference Schnorr-Müller2021).
A slower, long-term process followed the first stage. In this second stage, red nuggets began to merge with adjacent lower-mass systems. However, since much of the dense gas in those systems had already been converted into stellar mass during the initial stage, the gas ratio was significantly reduced. As a result, these mergers are classified as dry-mergers that do not trigger star formation and are relatively collisionless. Although their impact on the centre of the galaxy is generally negligible, they do contribute to its overall size and stellar mass. This slow accretion phase continued until
$z = 0$
. This two-stage evolution is supported by many observations, semi-analytic models and simulations (Spolaor et al. Reference Spolaor, Kobayashi, Forbes, Couch and Hau2010; Oser et al. Reference Oser, Ostriker, Naab, Johansson and Burkert2010; Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015; Buote & Barth Reference Buote and Barth2018; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018; Buote & Barth Reference Buote and Barth2019, and references therein).
Nevertheless, the stochastic nature of mergers prevents some galaxies from progressing to the dry-merger phase, leading to an interrupted evolutionary pathway (Quilis & Trujillo Reference Quilis and Trujillo2013; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018). As a result, these galaxies bypass the second stage and have largely preserved the properties of red nuggets formed around
$z \sim 2$
.
NGC 1277 is the first confirmed relic at low redshift (Trujillo et al. Reference Trujillo, Ferré-Mateu, Balcells, Vazdekis and Sánchez-Blázquez2013). It shows high rotational velocity, large central velocity dispersion, and an over-massive black hole (BH) at its centre – features that reveal its relic nature. Subsequently, PGC032873 and Mrk1216 have also been identified as relics based on similar structural and dynamical properties. Analyses of their stellar populations and star formation histories further support their classification as red-nugget-like relics (Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015; Walsh et al. Reference Walsh, Van den Bosch, Gebhardt, Yıldırım, Gültekin, Husemann and Richstone2017). Notably, Ferré-Mateu et al. (Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015) also introduced the concept of a “degree of relic” to describe variations within this class of galaxies.
Consequently, these compact systems with old stellar populations (
$\gtrsim10$
Gyr) are considered relics that have undergone little structural evolution since formation (Ferré-Mateu et al. Reference Ferré-Mateu, Trujillo, Martín-Navarro, Vazdekis, Mezcua, Balcells and Domnguez2017). Extensive studies have modelled their mass, metallicity, velocity fields, luminosity, size, stellar populations, and dark matter (DM) fractions, highlighting clear distinctions between red-nugget-like relics, other compact galaxies, and present-day massive ellipticals (Yıldırım et al. Reference Yıldırım, van den Bosch, van de Ven, Husemann and Lyubenova2015, Reference Yıldırım, van den Bosch, van de Ven, Martín-Navarro, Walsh, Husemann, Gültekin and Gebhardt2017; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018; Buote & Barth Reference Buote and Barth2019).
Building on these findings, the INSPIRE (INvestigating Stellar Population In RElics) project conducted the most comprehensive, systematic study to date on relic galaxies between
$0.1\lt z\lt0.4$
(Spiniello et al. Reference Spiniello2021). The project introduced a method to identify relic galaxies, quantified their degree of relicness (DoR), and measured their spectral and morphological properties across a broad wavelength range (NIR–VIS–UVB) (Spiniello et al. Reference Spiniello2021; D’Ago et al. Reference D’Ago2023; Martín-Navarro et al. Reference Martín-Navarro2023; Spiniello et al. Reference Spiniello2024; Maksymowicz-Maciata et al. Reference Maksymowicz-Maciata2024; Scognamiglio et al. Reference Scognamiglio2024). Among the main results, relics were found to have higher stellar velocity dispersion than non-relic galaxies of the same stellar mass and to require a bottom-heavy IMF with an enhanced dwarf-to-giant ratio (Maksymowicz-Maciata et al. Reference Maksymowicz-Maciata2024). The project also showed that [Mg/Fe] and metallicity correlate with relicness (Spiniello et al. Reference Spiniello2024) and confirmed that relics can exist both in dense environments, such as clusters, and in isolation (Scognamiglio et al. Reference Scognamiglio2024).
Lisiecki et al. (Reference Lisiecki, Małek, Siudek, Pollo, Krywult, Karska and Junais2023) identified 77 relics at intermediate redshifts (
$0.5 \lt z \lt 1.0$
) using the VIPERS survey. Expanding on this work, Siudek et al. (Reference Siudek2023) investigated the physical properties and environmental densities of 42 relic galaxies, confirming their diversity of environments.
While these studies provided extensive insights into the stellar and structural properties of relic galaxies, they lacked coverage of their X-ray characteristics. Yet, observations of the hot, X-ray-emitting interstellar medium (ISM) are also crucial for understanding the nature of relics. Since observing the ISM of red nuggets at
$z \sim 2$
remains exceedingly challenging with current X-ray instrumentation due to their distance and compactness, these confirmed relics present a rare opportunity to study such environments – allowing us to probe them in the X-ray domain as well.
In this context, Werner et al. (Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018) studied the extended, hot ISM in Mrk1216 and PGC032873 with Chandra observations. Likewise, Buote & Barth (Reference Buote and Barth2019) applied hydrostatic equilibrium models to their X-ray data to estimate the dark matter content, baryon fraction, and thermodynamic structure of the ISM. These works demonstrated the diagnostic power of X-ray morphology and thermodynamic properties.
Moreover, the hot ISM can be studied in the framework of chemical evolution and enrichment scenarios (e.g. early enrichment) (Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018; Mantz et al. Reference Mantz, Allen, Glenn Morris, Simionescu, Urban, Werner and Zhuravleva2017; Biffi et al. Reference Biffi2017). Furthermore, the fact that relics have avoided the dry-merger phase offers a unique opportunity to isolate the effects of dynamical processes, such as mergers, on galaxy evolution.
In this study, we investigate the ISM of the isolated relic galaxy Mrk1216 to gain insights into its chemical enrichment history. We use archival XMM-Newton observations, which offer improved spectral resolution compared to Chandra data, allowing more precise abundance measurements. We compare our results with the radial abundance profiles of Mernier et al. (Reference Mernier2017), based on a sample of 44 nearby cool-core galaxy clusters, groups, and ellipticals from the CHEERS (CHEmical Enrichment Rgs Sample) catalogue (De Plaa et al. Reference De Plaa2017).
This paper is organised as follows: Section 2 provides background information on Mrk1216. Section 3 describes the data reduction, background treatment, and spectral fitting methods. Section 4 reports our findings, followed by Section 5, which explores their implications. Finally, Section 6 summarises the main results. Throughout this paper, we adopt the
$\Lambda$
CDM cosmology with a Hubble constant of
$H_{0} = 70$
km s
$^{-1}$
Mpc
$^{-1}$
, a matter density of
$\Omega_{M} = 0.27$
, and a cosmological constant of
$\Omega_{\Lambda} = 0.73$
. All quoted uncertainties are at the
$1\sigma$
confidence level unless otherwise stated.
2. Mrk1216
Mrk1216 has been the subject of extensive multi-wavelength studies. Most notably, it exhibits structural peculiarities that distinguish it from other nearby massive Early-Type Galaxies (ETGs). In particular, Mrk1216’s halo density is a positive outlier in the
$\Lambda$
CDM
$c_{200}/M_{200}$
relation, suggesting an unusually early formation time (Buote & Barth Reference Buote and Barth2019). Mrk1216 is likewise considered an isolated galaxy, with only two neighbouring galaxies identified within a 1-Mpc radius – both of which are more than two magnitudes fainter on the K-band (Yıldırım et al. Reference Yıldırım, van den Bosch, van de Ven, Husemann and Lyubenova2015; Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015; Buote & Barth Reference Buote and Barth2019).
Mrk1216’s total mass within
$R_{500}$
is estimated to be
$M_{500} = 4.5 \pm 0.5 \times 10^{12}\,{\rm M}_{\odot}$
, comprising the combined mass of its BH, stars, gas, and DM, with a gas fraction of
$f_{gas,500} = 0.058 \pm 0.008$
(Buote & Barth Reference Buote and Barth2019). The stellar, gas, and DM properties of Mrk1216 exhibit striking similarities to those of the nearby fossil group NGC6482 (Yıldırım et al. Reference Yıldırım, van den Bosch, van de Ven, Martín-Navarro, Walsh, Husemann, Gültekin and Gebhardt2017; Buote & Barth Reference Buote and Barth2019). This resemblance, together with its extended X-ray emission, suggests that Mrk1216 may represent the central galaxy of a fossil group that subsequently formed around it (Buote & Barth Reference Buote and Barth2019).
The effective radius (
$R_e$
) of Mrk1216 is
$2.3 \pm 0.1$
kpc, with a stellar velocity dispersion of
$368 \pm 3$
km/s (Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015). Additionally, Mrk1216’s mean mass-weighted stellar age is
$12.8 \pm 1.5$
Gyr, with a metallicity of
$0.259 \pm 0.052$
dex (Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015). Its stellar mass within
$4R_e$
is
$2.0 \pm 0.8 \times 10^{11} {\rm M}_{\odot}$
(Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015). Mrk1216’s morphology, dynamics, and density profile closely resemble those of massive galaxies at
$z\gt2$
, distinguishing it from massive ETGs at
$z \sim 0$
and corroborating its relic nature (Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018). As such, Mrk1216, in addition to being classified as both a Massive Relic Galaxy (MRG) and an Isolated Compact Elliptical Galaxy (IsoCEG), appears to be the dominant constituent of the possible fossil group surrounding it.
Buote & Barth (Reference Buote and Barth2019) found no evidence of strong internal AGN outbursts in Mrk1216. In light of other findings (e.g. temperature, entropy,
$t_c/t_{ff}$
profiles), they concluded that the galaxy exhibits gentle, precipitation-regulated AGN feedback. These characteristics make Mrk1216 an ideal candidate for inspecting the effects of passive AGN evolution.
3. Observation and data reduction
We analysed archival XMM-Newton (Jansen et al. Reference Jansen2001) observations of Mrk1216 (RA:
$08^{\rm h}28^{\rm m}47.11^{\rm s}$
, DEC:
$-06^{\rm d}56^{\rm m}24.5^{\rm s}$
), conducted on 01 November 2018 (Observation ID 0822960201) with a total exposure of 109.7 ks.
We used XMM-Newton EPIC (MOS1, MOS2, and pn) and RGS instruments to analyse Mrk1216. While the EPIC observation was used to conduct imaging analyses and to extract spectra of selected annuli, as shown in Figure 1, the RGS was used to extract a high-resolution spectrum of the central area. The analysis procedure, methods employed, and software used are described in detail in the following sections.
Adaptively smoothed and combined image of Mrk1216. The solid white annuli mark the regions selected for analysis, extending out to
$R_{500}$
. The central region is coloured black for visual clarity. The green dashed annulus indicates the background region.
3.1 EPIC analysis
We used the Extended Source Analysis Software (ESAS), integrated into the Science Analysis System (SAS) version 19, to process the XMM-Newton data. Following the procedures outlined by Snowden & Kuntz (Reference Snowden and Kuntz2014), we ran emchain on MOS event files and epchain on pn event files for initial filtering and calibration. Faulty pixels were removed by filtering with the FLAG == 0 condition. For MOS, we selected events with one to four pixels (PATTERN
$\leq 12$
), while for pn, only single- and double-pixel events (PATTERN
$\leq 4$
) were retained. We then performed mos-filter and pn-filter to identify intervals contaminated by soft protons (SPs), obtaining uncontaminated good time intervals (GTIs).
Point sources were detected using the wavdetect tool in the CIAO software package (Fruscione et al. Reference Fruscione2006). A mask file incorporating the detected sources was created and applied in ESAS tasks to exclude them from the data. After cleaning, spectra were extracted using mos-spectra, mos_back, pn-spectra, and pn_back.
3.2 Surface brightness profile
Following the reduction steps, we extracted the surface brightness profile of Mrk1216. Such profiles describe how the ISM brightness changes radially, from the galaxy centre to the outskirts, allowing radial variations to be modelled. They also identify the radius where the ISM emission drops to the background level, providing the spatial extent of detectable emission. We fitted a
$\beta$
model (Cavaliere & Fusco-Femiano Reference Cavaliere and Fusco-Femiano1976) to represent the ISM emission and a constant to represent background emission. Fit results are presented in Section 4.
3.3 Spectral analysis
The observed spectra include multiple components of different origins and characteristics. To properly model the galaxy emission, each component must be carefully treated. These components can be grouped as: Non-X-ray Background (NXB), Cosmic X-ray Background (CXB), and Source Emission (SE), described below.
3.3.1 NXB (Non-X-ray Background)
NXB arises when energetic particles interact with the detector or its surroundings, producing fluorescent X-rays that subsequently strike the detector (Kuntz & Snowden Reference Kuntz and Snowden2008). NXB consists of three components: fluorescent instrumental lines in MOS and pn detectors, quiescent particle background (QPB), and residual soft-proton (SP) contamination. Fluorescent lines were modelled as Gaussian functions with given energies, as described in Snowden & Kuntz (Reference Snowden and Kuntz2014).
QPB spectra were generated with ESAS extraction tasks and modelled with multiple power laws plus additional Gaussians as required. For each spectral region, we calculated the fit parameters and kept them fixed in the main model.
Although SP contamination is mitigated by light curve filtering during data reduction, as mentioned in Section 3.1, we included a broken power law component to account for any potential residual SP contamination as described in Snowden & Kuntz (Reference Snowden and Kuntz2014). During the fitting process, we initially set the normalisations of the residual SP components to zero. Once a stable fit was achieved, we set them free to account for any residual contamination.
Since these NXB components did not originate from X-ray photons, standard response files are not suitable for their modelling. Instead, we used diagonal response files provided in the ESAS Calibration Database (CALDB) as recommended by Snowden & Kuntz (Reference Snowden and Kuntz2014).
3.3.2 CXB (Cosmic X-ray Background)
CXB has three components – LHB (Local Hot Bubble), GH (Galactic Halo), and UPS (Unresolved Point Sources) (De Luca & Molendi Reference De Luca and Molendi2004; Kuntz & Snowden Reference Kuntz and Snowden2008). In order to model these three components, we extracted spectrum for a 600–900 arcsec region as background area, shown in Figure 1 as white dashed annuli. We also used the ROSAT All-Sky Survey (RASS) spectrum to fit our background spectra simultaneously with it. The RASS spectrum was obtained from the X-ray background tool provided by HEASARC’s (High Energy Astrophysics Science Archive Research Center) website.Footnote
a
We chose an annulus with radii of 1
$^\circ$
and 2
$^\circ$
, centring the Mrk1216.
LHB is represented by a thermal APEC model with a fixed temperature value of 0.11 keV. GH is described by two APEC thermal emission models to account for hot and cold phases. UPS is modelled with a power law with a fixed index of (
$\Gamma=1.45$
). All thermal models’ metal abundance values are fixed at 1.0 and the redshift value is set to zero. Finally, additional absorption model is added to the GH and UPS components.
After simultaneously fitting all background components to the RASS and EPIC spectra, the resulting CXB parameter values were used in the main model for the source regions. These values are listed in Table 1.
The Cosmic X-ray background (CXB) fit parameters.
3.3.3 Source emission
Source emission comprises ISM and LMXB (low-mass X-ray binary) contributions. LMXBs create a hard continuum, typically as point sources, with nearly universal cumulative spectra (Irwin, Athey, & Bregman Reference Irwin, Athey and Bregman2003; Ji et al. Reference Ji, Irwin, Athey, Bregman and Lloyd-Davies2009). While detected point sources were removed, unresolved LMXB emission can remain (Kim Reference Kim2011). Although minor for Mrk1216 (Buote & Barth Reference Buote and Barth2019), it was included for completeness. The source emission was therefore modelled with a thermal APEC for ISM and a power law with fixed
$\Gamma = 1.6$
for LMXBs (Irwin et al. Reference Irwin, Athey and Bregman2003), both absorbed by Galactic
$n_H$
.
Accurate ISM modelling is crucial for reliable abundance measurements. A single-temperature plasma assumption can bias results, particularly producing the Fe-bias effect (Buote & Fabian Reference Buote and Fabian1998). To mitigate this, we adopted the VGADEM multi-temperature model in XSPEC (Arnaud, Dorman, & Gordon Reference Arnaud, Dorman and Gordon1999), which assumes a Gaussian distribution of emission measures. The model yields mean (
$\mu_{kT}$
) and width (
$\sigma_{kT}$
) of the temperature distribution, as well as abundances of multiple elements relative to Solar (He, C, N, O, Ne, Na, Mg, Al, Si, S, Ar, Ca, Fe, Ni).
3.4 RGS analysis
High-resolution spectra were extracted from RGS following the official guidelines.Footnote b The extraction region was set to 40 arcsec width, matching the 20-arcsec radius central region used in EPIC analysis.
3.5 Spectral fitting
3.5.1 EPIC
Spectral fitting was performed using XSPEC 12.11.1 (Arnaud et al. Reference Arnaud, Dorman and Gordon1999) and AtomDB 3.0.9 (Smith et al. Reference Smith, Brickhouse, Liedahl and Raymond2001; Foster et al. Reference Foster, Ji, Smith and Brickhouse2012). MOS spectra were restricted to 0.3–6.5 keV, and pn to 0.4–6.5 keV. We used C-statistics (Cash Reference Cash1979), meaning no binning or background subtraction was applied, as required. Elemental abundances were calculated relative to Solar values using the LPGS table (Lodders, Palme, & Gail Reference Lodders, Palme and Gail2009).
The EPIC spectral model is given in Eq. (1). The FL term represents MOS and pn fluorescent lines, modelled with Gaussian components (Snowden & Kuntz Reference Snowden and Kuntz2014). The first Const accounts for cross-calibration offsets among MOS1, MOS2, and pn, and the second for differing extraction areas, scaled to match RASS following Snowden & Kuntz (Reference Snowden and Kuntz2014). Galactic absorption was modelled with PHABS using
$n_H$
from Ben Bekhti et al. (Reference Ben Bekhti2016). Since NXB is particle-based, it was modelled separately with ESAS diagonal responses (Snowden & Kuntz Reference Snowden and Kuntz2014).
\begin{equation} S = [FL + Const \times Const \times Abs(CXB + Source)] + [NXB] \end{equation}
3.5.2 RGS
The same tools and statistics were used for RGS fitting, restricting spectra to 7–22 Å. O, Ne, and Fe abundances were left free in the VGADEM component. Compared to EPIC, RGS fitting was straightforward, requiring only the plasma emission plus Galactic absorption (TBABS). The RGS spectral region and model are shown in Figure 2.
Left: Adaptively smoothed and combined image of Mrk1216. White lines show the RGS spectral region. Right: Spectral fit of the RGS region.
3.6 SNe ratio
We used the SNeRatio code (Erdim et al. Reference Erdim, Ezer, Ünver, Hazar and Hudaverdi2021) to estimate the relative contributions of different supernova types to ISM enrichment. The code models observed ISM abundances with yield sets from SNIa and SNcc progenitors (summarised in Table 2). The output is the SNIa fraction relative to the total supernova population (SNIa + SNcc), obtained by comparing X-ray derived ISM abundances with the theoretical yields.
\begin{equation}R_{(Ia)} = \frac{SNIa}{SNIa + SNcc}.\end{equation}
4. Results
In this work, we used X-ray spectral analyses to investigate the chemical properties of the isolated elliptical relic galaxy Mrk1216. The spectral analysis covered five concentric circular annuli and a single circular region encompassing the area within
$R_{500}$
, together with an additional background region beyond
$R_{500}$
. After calculating the background emission parameters, the ISM emission was modelled for each region as outlined in Section 3.
During spectral modelling, an unidentified excess emission was detected at
$\sim$
1.2 keV in the central two regions. Curiously, Buote & Barth (Reference Buote and Barth2019) reported the same excess emission in Chandra observations of Mrk1216 and proposed possible explanations for this phenomenon. That two different observatories detected the same excess suggests that this is not instrument-related incident. Nevertheless, identifying the origin of this excess emission requires further investigation, which is beyond our scope. To mitigate potential biases introduced by this excess, the 1.17–1.25 keV energy range was removed from the spectra for the regions in question. This exclusion was applied only after careful consideration, and we verified that it did not affect the modelling of the continuum or elemental abundances.
We also noted the presence of slightly S-shaped residuals in the modelled spectra of the
$0-R_{500}$
region (0–248 kpc). Each annulus yielded statistically reliable results independently, suggesting that the issue most likely arises from a substantial mixture of multi-phase plasma – that is, plasma encompassing a wide range of temperatures and metal abundances – probably due to the relatively large spatial coverage. A similar trend was reported by Kim (Reference Kim2011). Therefore, spectral analysis results for
$0-R_{500}$
– originally intended to represent the overall averages of the galaxy – are likely affected and should be interpreted with caution. That said, we did not observe this trend in any of the other regions.
The analysis of the selected regions includes radial profiles of X-ray luminosity, plasma temperatures, elemental abundances and SNIa ratios. Several of these properties have previously been examined using Chandra X-ray observations of Mrk1216 (Buote & Barth Reference Buote and Barth2018; Buote & Barth Reference Buote and Barth2019; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018). Among these, we primarily compared our results with Buote & Barth (Reference Buote and Barth2019), as their study used the most recent and deepest Chandra observations.
While the higher spatial resolution of the Chandra satellite allows the central areas to be examined in greater detail, the higher spectral resolution of the XMM-Newton satellite afforded us the ability to measure element abundance values with more precision. Moreover, XMM’s wider field of view made it possible to extend the analysis up to
$R_{500}$
for a subset of the parameters. However, due to decreasing data quality at larger radii, not all parameters could be constrained out to
$R_{500}$
– some could be measured up to this radius, while others were only accessible within smaller radii.
4.1 Surface brightness profile
To study the ISM emission, we followed the procedure described in Section 3.1, using a mask file to remove point sources. We derived a
$\beta$
model (Cavaliere & Fusco-Femiano Reference Cavaliere and Fusco-Femiano1976) to model the ISM emission and a constant model to account for any background emissions. However, the single-
$\beta$
model was failed to adequately reproduce the data. The fit was greatly improved by adding an additional
$\beta$
model, yielding a double-
$\beta$
model plus a constant background. Whereas the first
$\beta$
model corresponds to the main emission component of the galaxy and has a brighter peak with a slightly steeper decline, the second
$\beta$
model is fainter in the centre but begins to dominate from approximately the fourth radial bin onwards (
$\sim 65$
kpc). The
$r_{c1}$
and
$\beta_1$
values (
$3.41\pm0.34$
kpc and
$0.55\pm0.02$
) are consistent with Buote & Barth (Reference Buote and Barth2019)’s analysis results of
$6.2_{-1.3}^{+2.0}$
arcsecs (
$2.8_{-0.6}^{+0.9}$
kpc) and
$0.52_{-0.02}^{+0.03}$
, respectively. Fit parameters are listed in Table 3 and the modelled surface brightness profile is shown in Figure 3.
To define the spectral regions, we set a hard radial limit at the point where the brightness drops to the modelled background level within
$3\sigma$
uncertainty. Since we noticed that the
$R_{500}$
value of Mrk1216 lies with this boundary, we set the radial limit at
$R_{500}$
(
$\sim 248$
kpc) to facilitate comparisons with other studies.
We divided the
$R_{500}$
region into five concentric annuli for spectral analysis. We also defined a background annulus beyond the
$R_{500}$
, with inner and outer radii of
$268.8$
and
$403.2\,\mathrm{kpc}$
, respectively, in order to model the background emission. This region was placed as far from the centre as possible to produce a more accurate spectral representation of the background emission. Source regions are shown as solid white circles and the background region as green-dashed annuli (Figure 1). For reference, the short black vertical lines at the bottom right of Figure 3 represent Mrk1216’s radial values corresponding to
$R_{2500}$
(
$\sim 130$
kpc),
$R_{500}$
(
$\sim 248$
kpc), and
$R_{200}$
(
$\sim 358$
kpc), as reported by (Buote & Barth Reference Buote and Barth2019).
4.2 Temperature profile
We observed a slight radial decline in
$\mu_{kT}$
from
$\sim$
0.77 keV to
$\sim$
0.61 keV out to
$R_{500}$
. This decrease is significant in the first three regions, while the last two are within uncertainties (
$\sigma_{kT}$
). The central region has the broadest temperature distribution (
$\sigma_{kT} \sim 0.16$
). Although uncertainties increase with radius, the
$\sigma_{kT}$
values generally decrease, reaching near zero in the outermost region. This suggests that a single-temperature approximation may be valid there, although large uncertainties make this conclusion uncertain. Elevated background and low source counts likely contributed to these uncertainties.
Buote & Barth (Reference Buote and Barth2019) used Chandra to measure temperatures out to
$\sim$
110 kpc, finding a central peak followed by a decline between
$\sim$
0.6–1.0 keV (Table 6 of the reference paper). This corresponds to our innermost four bins (
$\sim$
103 kpc). Although the two datasets used different instruments and models (single- vs. multi-temperature), the average temperatures are broadly consistent.
Adopted Yields in SNeRatio tool.
a This column refers to cited table numbers.
b This column is given in units of Z
$_{\odot}$
.
c Available at: http://star.herts.ac.uk/chiaki/works/YIELDCK13.DAT
Fit parameters of surface brightness profile. (Two
$\beta$
models and a constant background model.)
Black data points show the surface brightness profile of Mrk1216. The red line shows the best-fit model, consisting of
$\beta_1$
,
$\beta_2$
, and constant background terms. The constant background is plotted as a green line with light-green
$3\sigma$
uncertainty. Six annuli were selected for spectra, with the sixth for background. Alternating grey areas mark annuli, the pink region marks the background annulus. Black lines at bottom right mark
$R_{2500}$
,
$R_{500}$
, and
$R_{200}$
from left to right.
4.3 Abundance profile
We successfully measured Fe abundances out to
$R_{500}$
. Mg and Si were measured up to the fourth annulus (
$\sim0.42R_{500}$
), and S in the two central regions (
$\sim0.11R_{500}$
). The resulting abundance profiles are presented in Figure 4. RGS allowed additional measurements of O, Ne, and Fe in the central
$\sim$
9 kpc. Table 4 lists all abundances.
Si-K fluorescence contamination was detected in annuli three and four. As noted by Snowden & Kuntz (Reference Snowden and Kuntz2014), this line emission is present only in the MOS1 and MOS2, but not in the pn detector. To avoid biases, Si abundances in these regions were taken exclusively from pn data.
With respect to the EPIC measurements, the Mg, Fe, and Si profiles all appear to peak centrally and then flatten out towards the outer regions. However, given the large error bars, it is difficult to reliably quantify this trend. The Fe profile also drops sharply in the final region, beyond the flattening.
Because abundance ratios are more robust tracers of enrichment (Kim Reference Kim2011), we calculated Mg/Fe, Si/Fe, and S/Fe. Mg/Fe starts above Solar, decreasing toward unity. Si/Fe is consistent with Solar within uncertainties, while S/Fe is poorly constrained but also consistent with Solar. Ratios are shown in Figure 5.
Radial abundance profiles (Mg, Si, S, and Fe) of Mrk1216 and groups average from Mernier et al. (Reference Mernier2017).
With a large sample size of 44 nearby objects, Mernier et al. (Reference Mernier2017) is one of the most comprehensive and up-to-date studies on the chemical enrichment of elliptical galaxies, groups, and clusters. Said study divided the sample into two categories based on temperature, given its correlation with mass. The first category consists of galaxy clusters above 1.7 keV and the second of elliptical galaxies and galaxy groups below 1.7 keV. The analysis results of Mrk1216 can be compared with the second category (
$\lt1.7$
keV) to ascertain whether it follows the expected average behaviour of the aforementioned sample or exhibits distinct features linked to its unique evolutionary history. We should note that both studies adopt the same solar table (LPGS, Lodders, Palme, & Gail Reference Lodders, Palme and Gail2009), which eliminates potential systematic differences between solar tables when comparing abundance values.
We compared the abundance values of Mg, Si, S, and Fe with those of Mernier et al. (Reference Mernier2017) (Figure 4). Our Mg profile shows a central peak, unlike their nearly flat distribution, with higher values in the innermost two regions, while the outer bins are consistent. The Si profiles are in good agreement. Our S profile, limited to the two central bins, is also consistent. For Fe, both studies reveal a similar overall trend, though our values are higher in the third and fourth regions and drop sharply in the outermost bin.
In terms of abundance ratios (Figure 5), the main difference lies in Mg/Fe: our profile starts with a super-solar value (significant at
$\sim1\sigma$
) and decreases with radius, whereas theirs begins at a lower ratio (
$\sim0.5$
) and increases outwards. The Si/Fe and S/Fe ratios, however, are in good agreement between the two studies. The effect of the Fe drop in the outermost bin could not be assessed since other elements were not measurable at that radius.
4.4 SNIa ratio profile
SNIa ratio profiles are valuable tools for examining the chemical evolution of galaxies, groups, and clusters, as both the ratio itself and its radial variation provide important insights into their evolutionary history. We used the SNeRatio code (Erdim et al. Reference Erdim, Ezer, Ünver, Hazar and Hudaverdi2021) to estimate the relative SNIa contribution up to
$\sim0.42 \, R_{500}$
. Fits included Mg/Fe, Si/Fe, and S/Fe in the central regions, and Mg/Fe, Si/Fe in the outer regions. O/Fe and Ne/Fe from RGS were also fitted. Results are shown in Figure 5.
Systematic differences can arise with different yield models (Erdim et al. Reference Erdim, Ezer, Ünver, Hazar and Hudaverdi2021). To ensure consistency, we adopted the same parameters as Mernier et al. (Reference Mernier2017): the N100H SNIa yields (Seitenzahl et al. Reference Seitenzahl2013),
$Z0.008$
SNcc yields (Nomoto, Kobayashi, & Tominaga Reference Nomoto, Kobayashi and Tominaga2013), a Salpeter IMF (10–40 M
$_\odot$
) (Salpeter Reference Salpeter1955), and the LPGS solar table (Lodders, Palme, & Gail Reference Lodders, Palme and Gail2009). Fit results are in Table 5 and Figure 7.
As illustrated in Figure 5, the yield model fit for regions one and two underestimates the Mg/Fe ratio and overestimates the Si/Fe ratio. To assess the impact of these discrepancies on the estimated SNIa ratio, we repeated the fits for both regions excluding the Mg/Fe. This resulted in slightly higher SNIa ratios of
$0.233_{-0.041}^{+0.056}$
and
$0.250_{-0.055}^{+0.081}$
along with improved fit statistics (
$\chi^2/dof$
= 1.73/1 and 0.56/1, respectively). Excluding Si/Fe from the fits further improved the statistics and significantly lowered the SNIa ratios to
$0.103_{-0.018}^{+0.023}$
and
$0.102_{-0.019}^{+0.025}$
, with corresponding
$\chi^2/dof$
values of
$0.72/1$
and
$0.45/1$
. Possible explanations for these under- and overestimations are discussed in Section 5. Region three and four, by contrast, exhibit better fit statistics and align more closely to model expectations.
To measure radial behaviour, we fitted a linear model to the SNIa ratio profile. The resulting fit was nearly flat with a small slope of
$0.046\pm0.106$
and an intercept of
$0.164\pm0.10$
, as shown in Figure 7. For comparison, we repeated the fit with a constant model – a line with zero slope – to quantify the average SNIa ratio value for
$\sim0.42 \, R_{500}$
radius, which we calculated to be
$0.167\pm0.006$
.
Figure 7 depicts the SNIa ratio profile of Mrk1216 from this work alongside the average profile of groups and ellipticals from Mernier et al. (Reference Mernier2017). Our results are represented by black data points, whereas those of Mernier et al. (Reference Mernier2017) are shown in blue. The horizontal blue dotted line indicates the average central SNIa ratio for groups and ellipticals and the shaded blue area represents associated scatters (see the paper for details).
Spectral fit results of Mrk1216.
Radial abundance ratio profiles ([Mg/Fe], [Si/Fe], and [S/Fe]) of Mrk1216 and groups average from Mernier et al. (Reference Mernier2017).
5. Discussion
5.1 Temperature structure
Previous studies using Chandra observations have reported the temperature profile of Mrk1216 (Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018; Buote & Barth Reference Buote and Barth2019). Thanks to its higher spatial resolution, Chandra provides a more detailed view of the central temperature structure by employing narrower annuli. The results are consistent, revealing a centrally peaked radial temperature profile characterised by a negative gradient (see Figure 8).
Kim et al. (Reference Kim2020) analysed 60 ETGs and proposed a classification of temperature profiles into six categories: hybrid-bump, hybrid-dip, double-break, positive, negative, or irregular (see their paper for detailed definitions). Based on its clear negative gradient, Mrk1216 falls into the negative class. Typically, negative-type ETGs are dynamically disturbed and resemble non-cool-core (NCC) clusters. However, Mrk1216 is a highly relaxed system with no signs of dynamical disturbance (Yıldırım et al. Reference Yıldırım, van den Bosch, van de Ven, Husemann and Lyubenova2015; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018; Buote & Barth Reference Buote and Barth2019),
Interestingly, the sample in that study includes another exception: NGC6482. Despite being a relaxed fossil system (Khosroshahi, Jones, & Ponman Reference Khosroshahi, Jones and Ponman2004), it also shows a negative temperature gradient. The similarity between these two fossil systems (Mrk1216 and NGC6482) compared with normal ETGs may provide valuable insight into heating and cooling mechanisms, highlighting the role of galaxies’ evolutionary histories. Werner et al. (Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018) proposed radio-mechanical AGN feedback as the main heating source for Mrk1216, identifying it as an outlier in both the BH – bulge mass relation and chaotic cold accretion (CCA) (Gaspari, Ruszkowski, & Oh Reference Gaspari, Ruszkowski and Oh2013). Further discussion of heating mechanisms is beyond the scope of this paper.
5.2 Metallicity structure
As noted above (Goulding et al. Reference Goulding2016, and references therein), the two-phase galaxy formation scenario posits that structures are formed via gas-rich (wet) and dissipative mergers (
$z \gtrsim 2$
) followed by a phase of passive evolution through gas-poor (dry), non-dissipative and collisionless mergers (
$2 \gtrsim z \gtrsim 0$
). The wet mergers characteristic of the first phase induce rapid and efficient episodes of star formation (Naab, Johansson, & Ostriker Reference Naab, Johansson and Ostriker2009; Oser et al. Reference Oser, Ostriker, Naab, Johansson and Burkert2010; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018; Buote & Barth Reference Buote and Barth2019). During this process,
$\alpha$
-elements – such as Mg – synthesised in the cores of massive stars are expelled through core-collapse supernovae (SNcc). The short lifespan of these massive stars results in the rapid release and recycling of Mg, making it a reliable tracer for overall enrichment.
In contrast, type Ia supernovae (SNIa), which originate from the remnants of long-lived, low-mass stars, produce mainly Fe-peak elements (Beverage et al. Reference Beverage, Kriek, Conroy, Bezanson, Franx and Van derWel2021). Given the time delay between SNcc and SNIa, rapid star formation leads to an overabundance of
$\alpha$
-elements (from SNcc) relative to Fe-peak elements (from SNIa). As a result, enhanced
$[\alpha/Fe]$
elemental abundance ratios serve as indicators of the intensity and timescale of star formation episodes (Thomas et al. Reference Thomas, Maraston, Bender and De Oliveira2005; Thomas et al. Reference Thomas, Maraston, Schawinski, Sarzi and Silk2010; Spolaor et al. Reference Spolaor, Kobayashi, Forbes, Couch and Hau2010).
Red nuggets formed through these wet mergers, and their relics, are therefore expected to show high
$[\alpha/Fe]$
. Several optical spectroscopic studies have demonstrated this (Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015; Spiniello et al. Reference Spiniello2021; Spiniello et al. Reference Spiniello2024). The abundance ratio derived from the X-ray-emitting plasma of Mrk1216 exhibits the same behaviour, showing a notably high [Mg/Fe] ratio.
Numerous studies have examined the relationship between mass and metallicity, commonly referred to as the mass–metallicity relation (MZR), finding that metallicity tends to increase with galactic mass (Lequeux et al. Reference Lequeux, Peimbert, Rayo, Serrano and Torres-Peimbert1979; Tremonti et al. Reference Tremonti2004). This trend is often attributed to the greater efficiency with which massive systems enrich accreted gas (Spolaor et al. Reference Spolaor, Kobayashi, Forbes, Couch and Hau2010). However, subsequent research has suggested that compactness may play a more significant role in this regard. Specifically, the depth of a galaxy’s gravitational potential governs its metallicity, such that, at fixed mass, more compact galaxies have steeper potential wells and consequently exhibit higher metallicities (Hoopes et al. Reference Hoopes2007; McDermid et al. Reference McDermid2015; Ellison, Catinella, & Cortese Reference Ellison, Catinella and Cortese2018; Maiolino & Mannucci Reference Maiolino and Mannucci2019; Barone et al. Reference Barone2018; Beverage et al. Reference Beverage, Kriek, Conroy, Bezanson, Franx and Van derWel2021.
The galaxies and groups in the CHEERS sample (Mernier et al. Reference Mernier2017), against which our findings were compared, exhibit temperatures ranging from 0.5 to 1.7 keV, with a mean of approximately 0.95 keV and a scatter of 0.32 keV (De Plaa et al. Reference De Plaa2017). These temperatures served as a proxy for total mass. At
$kT_{\mu,500}\approx0.71$
, Mrk1216 lies below this average (Table 4). Given its compact nature as a relic, measuring elevated metal abundances in the X-ray band would lend support to the idea that compactness, rather than mass alone, drives higher metallicity. Only central Mg differs significantly, while other abundances agree within errors, yet this pattern appears broadly consistent with the proposed scenario.
SNeRatio fit results.
SNeRatio fit results for the four central EPIC regions and for the RGS region. The black data points represent the measured [X/Fe] ratios with their associated uncertainties. The vertical bars show the model prediction for the [X/Fe] ratios, with red segments corresponding to the SNIa contribution and green segments to the SNcc contribution. The black horizontal dashed lines indicate the solar ratios.
Radial SNIa ratio profile of Mrk1216 (black data points) compared with the groups’ average from Mernier et al. (Reference Mernier2017) (blue data points). The solid green line shows the line fit whereas the shaded green area shows the uncertainty of this fit. The intercept and slope parameters of this line fit are written in the label. The dashed blue line shows the average central value (
$0.05r_{500}$
or
$0.2r_{500}$
) and the shaded blue area shows the scatter of the uncertainties of the group sample from Mernier et al. (Reference Mernier2017).
Radial temperature profile of Mrk1216. The derived multi-temperature plasma emission model (VGADEM) calculates the mean (
$\mu_{kT}$
) and standard (
$\sigma_{kT}$
) deviation of temperature values for a specific region. Temperature values from Buote & Barth (Reference Buote and Barth2019) are shown in magenta for comparison.
Furthermore, the high central [Mg/Fe] value is in line with expectations for relic galaxies, where elevated [Mg/Fe] ratios reflect rapid early star formation and enrichment. These results support the classification of Mrk1216 as a relic system. To strengthen the generality of these findings, studying a larger sample of galaxies would be valuable.
The sharp drop in Fe abundance in the outermost region is, given the potential for systematic or statistical uncertainties, difficult to interpret. These include limited photon counts and statistical challenges associated with low-temperature systems, where Fe abundance must be inferred from only L-shell emissions (Kim Reference Kim2011). Assuming accurate measurements, one plausible explanation for this drop is that the observed X-rays originate not only from the ISM but also from the less enriched IGrM of the fossil group dominated by Mrk1216. This hypothesis is supported by the fact that Mrk1216’s surface brightness profile is better described as a double-
$\beta$
model typical to cool-core systems. However, Mrk1216 lacks a cool core, and both models in the fit are broad, with the second dominating beyond the fourth radial bin (
$\sim 103$
kpc). This radius coincides with onset of the decline in Fe abundance, suggesting that the second
$\beta$
component may represent IGrM emissions. That said, the absence of measurements for other elements severely limits our ability to make a reliable assessment regarding this hypothesis.
5.3 Relative SNe contributions
The abundance ratios observed in Mrk1216 suggest that SNcc contribute more to the total abundances than do SNIa. The SNeRatio fit results show the calculated ratios (see Table 5 and Figure 5). The [Mg/Fe] ratio measured in the central two regions exceeds the value predicted by the model. Since a higher [Mg/Fe] ratio is characteristic of a relic galaxy, our ISM abundance measurements align with these expectations.
SNe ratio estimates become more reliable when a wider range of elements are included in calculations. However, observational limitations and instrumental constraints led us to use a small number of ratios – [Mg/Fe], [Si/Fe], and [S/Fe] – which increases uncertainty. This challenge is particularly pronounced in low-temperature systems like galaxies, where lower emissivity limits the ability to detect the emission lines used to determine element abundance (i.e. Ar, Ca, Ni) compared to clusters. Next generation satellites equipped with higher-resolution instruments will enable more comprehensive measurements. In the meantime, to make the most of the available data, we also used the spectrum obtained from the RGS instrument to measure the SNe ratio, which allowed us to include additional elements measurable by RGS but not EPIC (i.e. [O/Fe], [Ne/Fe]). Although the spectral regions differ, the nearly flat radial SNe profile calculated from EPIC spectra allows us to expect similar results from the RGS. Despite being somewhat higher at
$0.21_{-0.06}^{+0.09}$
, we obtained a consistent SNIa ratio within uncertainties, as was expected.
Apart from this, evaluating the theoretical background of our model suggests that the fit may be improved by adopting a more realistic IMF. We employed a simple and uniform IMF – the Salpeter IMF (Salpeter Reference Salpeter1955) – to calculate SNcc yields, in order to avoid systematic differences when comparing with other results. However, in the specific case of Mrk1216, it is well established that the stellar population of the galaxy exhibits a bottom-heavy IMF (Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015). To reflect this, we refitted the abundance ratios in the innermost region using a steeper IMF slope (2.8). This adjustment led to a slight reduction in the quality of the model fit, with the stat./dof value increasing from
$11.14/2$
to
$11.40/2$
and the SNIa contribution ratio decreasing from 16.7 to 14.2, which did not contribute to resolving the discrepancy between the observed abundance patterns and the model predictions. Conversely, using a top-heavy IMF (slope of 2.0) led to an increase in [Mg/Fe] (
$\approx1.3\%$
), slightly reducing this discrepancy, with stat./dof =
$10.96/2$
and an SNIa ratio of 18.8; still, these changes are too minor to draw any definitive conclusions.
It should be noted that the IMF used for modelling the ISM represents the cumulative stellar population responsible for enriching the hot plasma since its formation, whereas the IMF inferred from stellar analyses reflects the currently surviving stellar populations. Therefore, it is plausible that the IMF derived from the ISM differs from that of the stellar population, as it is more top-heavy in our case. Investigating such IMF variations would require more comprehensive studies with additional elemental abundance measurements, which is beyond the scope of this work.
Finally, canonical IMFs are often known to fall short in reproducing galactic enrichment trends, as they fail to capture the full complexity and variability inherent in the chemical evolution of galaxies (Cappellari et al. Reference Cappellari2012; Van Dokkum & Conroy Reference Van Dokkum and Conroy2012; Martín-Navarro et al. Reference Martín-Navarro, Vazdekis, Falcón-Barroso, La Barbera, Yıldırım and van de Ven2018; Yan, Jeřábková, & Kroupa Reference Yan, Jeřábková and Kroupa2021; Dib Reference Dib2022; Barbosa et al. Reference Barbosa, Spiniello, Arnaboldi, Coccato, Hilker and Richtler2021). In this regard, the Integrated Galaxy-wide Initial Mass Function (IGIMF) framework provides a more flexible and physically motivated alternative. By allowing the IMF to vary with the star formation rate, the IGIMF offers a more realistic modelling of chemical enrichment and star formation histories (Kroupa & Weidner Reference Kroupa and Weidner2003; Weidner & Kroupa Reference Weidner and Kroupa2005; Recchi, Calura, & Kroupa Reference Recchi, Calura and Kroupa2009; Marks et al. Reference Marks, Kroupa, Dabringhausen and Pawlowski2012; Jeřábková et al. Reference Jeřábková, Hasani Zonoozi, Kroupa, Beccari, Yan, Vazdekis and Zhang2018; Yan, Jeřábková, & Kroupa Reference Yan, Jeřábková and Kroupa2019; Hosek et al. Reference Hosek, Lu, Anderson, Najarro, Ghez, Morris, Clarkson and Albers2019; Yan et al. Reference Yan, Jeřábková and Kroupa2021; Dib Reference Dib2022). These types of methodological advancements in upcoming research may contribute significantly to resolving the ongoing discrepancies between observations and models, while also encompassing different evolutionary pathways that shape chemical enrichment of galaxies.
5.4 Enrichment scenarios
Another noteworthy aspect is the nearly flat radial SNIa ratio profile of Mrk1216, with an average ratio of
$\sim0.17$
up to
$\sim0.42R_{500}$
. These results are consistent with the average profile of groups and ellipticals in Mernier et al. (Reference Mernier2017). A flat relative SNIa contribution is often regarded as strong evidence for the early-enrichment scenario (Yates, Thomas, & Henriques Reference Yates, Thomas and Henriques2017; Ezer et al. Reference Ezer, Bulbul, Ercan, Smith and Bautz2017; Mernier et al. Reference Mernier2017; Erdim et al. Reference Erdim, Ezer, Ünver, Hazar and Hudaverdi2021; Gastaldello et al. Reference Gastaldello, Simionescu, Mernier, Biffi, Gaspari, Sato and Matsushita2021). According to this view, a significant portion of enrichment occurred prior to the formation of large systems such as galaxies or galaxy clusters (
$z\sim2$
) (Oppenheimer & Davé Reference Oppenheimer and Davé2006). Consequently, both central and outskirt plasmas, despite having evolved under distinct physical conditions, exhibit a similar relative SNe contribution.
Furthermore, the pre-enrichment scenario implies that the gas enriched within proto-groups and proto-cluster galaxies may have been expelled beyond their shallower potential wells at even earlier times (at
$z \gt= 3$
) (Gastaldello et al. Reference Gastaldello, Simionescu, Mernier, Biffi, Gaspari, Sato and Matsushita2021; Mernier et al. Reference Mernier2022, and references therein). In our case, although the uncertainties increase in the outskirts, the indication of a flat SNIa profile becomes more pronounced because of the lack of efficient mixing mechanisms in Mrk1216 (Buote & Barth Reference Buote and Barth2019), as mentioned in Section 2. Consequently, plasma from the outskirts predominantly represents the most recently accreted matter, which is distinct from plasma accreted earlier (Gastaldello et al. Reference Gastaldello, Simionescu, Mernier, Biffi, Gaspari, Sato and Matsushita2021).
6. Conclusions
In this paper, we presented the results of X-ray analyses of the relic galaxy Mrk1216 using XMM-Newton observations. We compared our findings with properties of massive ellipticals reported in previous studies and interpreted them in the context of galaxy formation and chemical enrichment scenarios. Our main conclusions are summarised as follows:
-
• The radial temperature profile of Mrk1216 exhibits a negative gradient – a feature generally associated with dynamically disturbed ETGs according to the classification proposed by Kim et al. (Reference Kim2020). Interestingly, the fossil group NGC6482 (Khosroshahi et al. Reference Khosroshahi, Jones and Ponman2004), another outlier in the same sample, also shows a similar negative profile. The presence of such gradients in relaxed fossil systems suggests a possible connection between fossil nature and the development of negative profiles, offering insights into heating–cooling processes and AGN evolution (Gaspari et al. Reference Gaspari, Ruszkowski and Oh2013; Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018).
-
• In the two-phase formation model, galaxies initially assemble through gas-rich mergers that drive rapid star formation.
$\alpha$
-elements such as Mg, produced in massive stars and expelled by SNcc, yield elevated
$[\alpha/Fe]$
ratios – signatures of short and intense star formation. In Mrk1216, we measured a significantly high [Mg/Fe] ratio in the X-ray emitting ISM, consistent with earlier stellar abundance results (Ferré-Mateu et al. Reference Ferré-Mateu, Mezcua, Trujillo, Balcells and van den Bosch2015). This agreement reinforces the view that Mrk1216 still preserves its red-nugget-like nature and has not transitioned into the second phase. -
• Abundance ratios in Mrk1216 indicate that SNcc dominated over SNIa in chemical enrichment (
$R_{Ia} \approx 0.17 \pm 0.01$
). However, the measured [Mg/Fe] in the central regions exceeds model predictions, highlighting the need for further investigation. Our adoption of a simple, uniform IMF (Salpeter) may contribute to this discrepancy. More realistic IMF prescriptions, such as the IGIMF, are recommended for improved chemical evolution modelling. -
• Mrk1216 exhibits a notably flat radial SNIa ratio profile up to
$\sim0.42R_{500}$
, consistent with an early-enrichment scenario (Werner et al. Reference Werner, Lakhchaura, Canning, Gaspari and Simionescu2018; Mantz et al. Reference Mantz, Allen, Glenn Morris, Simionescu, Urban, Werner and Zhuravleva2017; Biffi et al. Reference Biffi2017). This suggests that a significant portion of enrichment occurred before system assembly, with relatively similar SNe contributions in both the central and outer plasmas. The lack of efficient mixing mechanisms in Mrk1216 further amplifies the flat profile, implying that the outer plasma may represent the most recently accreted matter.
This study demonstrates that the structural and chemical properties of Mrk1216 are consistent with the predictions of the hierarchical formation model. Although conclusions based on a single object limit the strength of interpretations, this study emphasises how future research adopting a similar perspective can yield valuable insights. Identifying comparable systems and conducting in-depth observations will be key. Nevertheless, many unanswered questions may be addressed with the high-resolution capabilities of upcoming missions like Athena X-IFU (X-ray Integral Field Unit) (Barret et al. Reference Barret2016) or XRISM (X-ray Imaging and Spectroscopy Mission), the successor to ASTO-H/HITOMI (Tashiro Reference Tashiro2022). Finally, there is a pressing need to refine existing theoretical frameworks by incorporating more realistic IMF models, including the dust depletion effect in chemical modelling, and the resolution of uncertainties related to SNIa progenitors. Certainly, this list is not exhaustive. All advancements in these areas will contribute significantly to our understanding of chemical enrichment, structural formation, and related fields.
Acknowledgement
The authors thank the anonymous referees for their constructive comments and suggestions. We would like to express our sincerest gratitude to Dr. François Mernier, Dr. Turgay Çağlar, and Fatih Hazar for their support and insightful contributions throughout the course of this research. This work received financial support from the Scientific and Technological Research Council of Turkey (TÜBİTAK), project number 121F263. Lastly, we acknowledge that this paper forms part of M. Kyami Erdim’s PhD dissertation.
Data availability statement
The XMM-Newton raw data used in this article are available to download at the HEASARC Data Archive website (https://heasarc.gsfc.nasa.gov/docs/archive.html). The reduced data underlying this article will be shared on reasonable request to the corresponding author.
Competing interests
None.
Open access
