An XMM-Newton View of the ANdromeda Galaxy as Explored in a Legacy Survey (New-ANGELS) II: Luminosity Function of X-ray Sources

TL;DR

We perform a regionally resolved X-ray luminosity function study of 4,506 X-ray sources around M31 using the New-ANGELS XMM-Newton survey, correcting incompleteness with analytical sensitivity maps and forward-modeling the background and stellar-population-linked components. The analysis reveals LMXBs as the dominant population in the disk, a systematically lower α_LMXB than in other galaxies, and notable regional variation linked to the star-formation history within M31. By combining disk-center decomposition with SFH-informed X-ray binary evolution, we find evidence for a rapid fading of LMXB luminosity around ~1 Gyr after star-formation events, which helps explain the dimness of Inner Arm regions and the galaxy-wide under-luminosity. These results suggest that the commonly cited universal LMXB–stellar mass relation exhibits substantial regional scatter and that nuanced stellar-age and metallicity histories must be incorporated into population synthesis models of X-ray binaries.

Abstract

As part of the New-ANGELS program, we systematically investigate the X-ray luminosity functions (XLFs) of 4506 X-ray sources projected within a radius of 2.5 deg centering on M31. We construct XLFs for different regions in the disk and halo of M31, accounting for the incompleteness with an effective sensitivity map. Assuming that the halo regions contain (mostly) foreground stars and background active galactic nuclei, they are taken as "background" for deriving the XLFs of the sources in the disk. Through modeling XLFs, we decompose the X-ray sources into distinct populations for each region. We find that low-mass X-ray binaries are the dominant X-ray population throughout the disk of M31. The XLFs of M31 reveal a consistently lower integrated LMXB luminosity per stellar mass ($α_\mathrm{LMXB}$) compared to other galaxies, likely due to M31's prolonged period of quiescent star formation. Variations in the XLF shape and $α_\mathrm{LMXB}$ across different regions of M31 suggest that the relationship between integrated luminosity and stellar mass may vary within the galaxy. Additionally, the relatively low integrated luminosity observed in the inner-arm region provides crucial evidence for a rapid fading of M31's LMXBs around 1 Gyr, a finding consistent with recent observations of other nearby galaxies.

Figures (17)

• Figure 1: (a): The sensitivity map in the energy range of 2.0-4.5 keV; (b): The sky coverage function generated from the sensitivity maps in the energy ranges of 0.2-0.5 keV, 0.5-1.0 keV, 1.0-2.0 keV, 2.0-4.5 keV, and 4.5-12.0 keV.
• Figure 2: (a): A false-color composite image of the M31 survey area, combining X-ray data from XMM-Newton and infrared data from WISE; (b): The observation coverage of M31. The covered area is divided into four regions, as indicated in different colors. The central region is defined as within a radius of $8^{\prime}$ from galactic center, the disk region as outside the central region and along galactic disk, with a height of less than $30^{\prime}$ from galactic center along the minor axis. The north and south regions are the left region in the halo which are $30^{\prime}$ away from the major axis of the galaxy.
• Figure 3: X-ray luminosity functions, for the energy ranges of (a) 0.2-0.5 keV, (b) 0.5-1.0 keV, (c) 1.0-2.0 keV, (d) 2.0-4.5 keV, and (e) 4.5-12.0 keV. For comparison, the results are shown for different regions separately.
• Figure 4: X-ray luminosity functions, with the best-fit models (in dashed lines) between 2.0 and 4.5 keV. Left: the south (in orange) and north regions (in blue). For comparison, a AGN model (luoCHANDRADEEPFIELDSOUTH2016) is shown in grey dotted line. The black line represents the difference between the XLF of the north and that of the south, obtained by subtraction. Right: XLF of the disk region, central $8^{\prime}$ and the central $3^{\prime}-8^{\prime}$ region, all with the contribution of the sources in the background region (north and south combined) subtracted. Also shown in the latter panels are the XLFs of HMXB (in green dotted line) and LMXB (in red dotted line) in the Milky Way (2002AA...391..923G), for comparison.
• Figure 5: The XLF fitting results for the background region (left), the center region (middle), and the disk region (right) between 2.0 and 4.5 keV. The background region combines sources from the north and south regions. In the top panel of each region panel, the black line represents the incompleteness-corrected cumulative XLF, while the grey shaded area indicates uncertainties based on Gehrels' variance function 1986ApJ...303..336G. The green cross corresponds to the AGN XLF in CDF-S luoCHANDRADEEPFIELDSOUTH2016. The blue and red dashed lines represent the fitted background and LMXB components, respectively. The purple solid line combines the background and LMXB components. In the middle panel of the figures, the uncorrected distribution of the sources and the corresponding folded model are presented using the same color scheme as the top panel. The third panel is the ratio between model and observed data in middle panel. The bottom panel displays the residuals in C-statistic($C=2 \sum m_\mathrm{i}-\mathrm{N_i}+\mathrm{N_i} \mathrm{ln}(\mathrm{N_i}/m_\mathrm{i})$ with m and N as the model and the data,Kaastra2017). Additionally, in the top panel for center region, the thin black stairs represent the XLF between $0-15^{\prime}$ for comparison.