MUSE IFU observations of galaxies hosting Tidal Disruption Events

TL;DR

The study analyzes 20 TDE host galaxies with MUSE IFU to map extended emission line regions (EELRs) and to constrain near-nucleus stellar populations. EELRs are detected in 5 hosts and are exclusively found in nearby post-merger systems, implying a strong merger–TDE connection and suggesting past AGN activity that ionized the gas, with ionizing events occurring $2\times10^{3}$ to $7.5\ times10^{4}$ years before the TDEs. Stellar-population modelling (Starlight + BPASS) indicates youngest nuclear populations around $\sim1~\mathrm{Gyr}$ and predominance of subsolar-mass stars ($M_\star \lesssim 0.3$–$1~M_\odot$), such that most TDEs would arise from very low-mass stars given Hills-mass constraints. A tension is found between disrupted-star masses inferred from light-curve modelling (often $0.1$–$1~M_\odot$) and the SMD-predicted distributions, suggesting either biases in TDE mass inferences or missing physics, while reinforcing the role of post-merger environments in elevating TDE rates and highlighting the importance of integral-field spectroscopy for decoding TDE progenitor channels.

Abstract

We present an analysis of twenty tidal disruption event (TDE) host galaxies observed with the MUSE integral-field spectrograph on ESO VLT. We investigate the presence of extended emission line regions (EELRs) and study stellar populations mostly at sub-kpc scale around the host nuclei. EELRs are detected in 5/20 hosts, including two unreported systems. All EELRs are found at z<0.045, suggesting a distance bias and faint EELRs may be missed at higher redshift. EELRs only appear in post-merger systems and all such hosts at z<0.045 show them. Thus, we conclude that TDEs and galaxy mergers have a strong relation, and >45% of post-merger hosts in the sample exhibit EELRs. Furthermore, we constrained the distributions of stellar masses near the central black holes (BHs), using the spectral synthesis code Starlight and BPASS stellar evolution models. The youngest nuclear populations have typical ages of 1 Gyr and stellar masses below 2.5MSun. The populations that can produce observable TDEs around non-rotating BHs are dominated by subsolar-mass stars. 3/4 TDEs requiring larger stellar masses exhibit multi-peaked light curves, possibly implying relation to repeated partial disruptions of high-mass stars. The found distributions are in tension with the masses of the stars derived using light curve models. Mass segregation of the disrupted stars can enhance the rate of TDEs from supersolar-mass stars but our study implies that low-mass TDEs should still be abundant and even dominate the distribution, unless there is a mechanism that prohibits low-mass TDEs or their detection.

Figures (14)

• Figure 1: Nuclear spectra of the 20 TDE host galaxies in our sample (lighter shade) under spectra binned to 10 Å (darker shade). The spectra were extracted with an aperture equal to the FWHM measured for stars in the MUSE cubes (see Table \ref{['tab:fwhms']}). The galaxy types are Star-Forming (SF, first from the top), Quiescent (Q, second), Quiescent, Balmer Strong (QBS, third) and Post-Starburst (PSB, fourth) following French2020. Note that AT 2019ahk is classfied as a shocked PSB galaxy (SPOG) despite prominent emission lines.
• Figure 2: Bulge mass and redshift comparsion between the MUSE sample and sample of Ramsden2025 (R2025). The MUSE sample concentrates on lower bulge masses and lower redshifts than R2025. The samplesa are nominally consistent, as the MUSE sample is less biased towards more distant brighter galaxies possibly explaining the minor discrepancy. The median values ($\widetilde{M}_\mathrm{Bulge}$, $\widetilde{z}$) of are shown with a solid lines for the MUSE sample and dashed lines for R2025.
• Figure 3: The [OIII] $\lambda5007$ luminosity of the EELR bins identified in the five TDE host galaxies (coloured markers), and the $5\sigma$ detection limits of each MUSE cube against redshift (open triangles). The brightest EELRs bins should have been seen in all the host galaxies regardless of redshift, but there may be bias against galaxies with lower luminosity EELRs.
• Figure 4: The BPT diagrams for the EELRs and the nuclei of the five TDE host galaxies with detected EELRs. The values are derived from spectra after subtracting the best-fitting Starlight model (see Section \ref{['sec:SMDs']} for details). Below the solid line, the dominant powering mechanism is assumed to be star-formation Kauffmann2003, while above the dashed line the harder continuum emission (i.e. AGN) is required Kewley2001. Between the two lines, in the composite region, both mechanisms are viable. Bins where both ratios were estimated are marked with solid markers, while limiting values are shown with open markers. In the nuclear spectrum of ASASSN-14ae H$\alpha$ and H$\beta$ were not detected and [NII] falls on a strong telluric so the value is not shown on the diagram. Note that the few bins where [NII]/H$\alpha$ could not be estimated are not shown. Except a few bins in ASASSN-14ko, all EELRs require hard AGN continuum.
• Figure 5: The WHAN diagrams CidFernandes2011AAGN of the five TDE hosts that exhibit EELRs. The required ionising source is divided into star-formation (SF), strong or weak AGN (sAGN, wAGN), or LINER-like based on their H$\alpha$ EW and [NII] to H$\alpha$ ratio. All hosts exhibit some EELRs that require AGN-like ionising continuum. The nuclear spectra (red cross) show that ASASSN-14ko and AT 2019ahk are consistent with sAGN, possibly due to the presence of TDE emission in the MUSE cube, while iPTF16fnl and ASASSN-14li are LINER-like. The $3\sigma$ upper limit of H$\alpha$$\mathrm{EW}$ for ASASSN-14ae ($3.7$) effectively places it in LINER-like region, implying that no significant AGN activity is present, but no marker is shown as a strong telluric is located over the [NII] emission.