X-ray flares from the stellar tidal disruption by a candidate supermassive black hole binary

Most massive galaxies host central supermassive black holes (SMBHs), yet many are quiescent and hard to study via accretion signatures. When a star is tidally disrupted by a SMBH, roughly half the debris remains bound and can accrete, producing a luminous multiwavelength flare (a tidal disruption event, TDE). Classical theory predicts soft X-ray thermal emission from a compact, newly formed accretion disk, with temperatures scaling as T_eff ≈ 4×10^5 M^(1/4) K for M = M_BH/10^6 M☉. However, many TDEs discovered in UV/optical show much cooler blackbody temperatures (~1–4×10^4 K) and large emitting radii (~10^14–10^15 cm), challenging standard disk-emission interpretations. Alternative ideas include emission powered by stream–stream shocks during circularization, or thermal reprocessing of inner accretion power by gas at larger radii. Early X-ray observations are crucial to distinguish these mechanisms, but most optically selected TDEs lack strong early X-ray detections. This study investigates OGLE16aaa, focusing on the timing and spectral evolution of its X-ray emission relative to UV/optical, to probe the origin of the emission, test for obscuration/reprocessing, and assess the possibility of a supermassive black hole binary (SMBHB) influencing the accretion and light curve.

Early work predicted bright, soft X-ray emission from TDE accretion disks. Observations of optically selected TDEs instead revealed cooler UV/optical blackbodies and large radii, suggesting alternate power sources: (i) stream self-intersections and shocks during debris circularization, or (ii) reprocessing of inner-disk radiation by extended gas. Only a subset of optical TDEs have shown X-ray detections, leaving the UV/optical–X-ray connection unresolved. Models propose that dense, radiation-driven outflows can initially obscure X-rays, with an ionization breakout occurring months later for ~10^6 M☉ SMBHs, allowing X-rays to escape and power declining UV/optical via reprocessing. Variable absorption by patchy ejecta could also produce X-ray variability. Additionally, disruptions in SMBHB systems can yield interrupted fallback and accretion, producing dips/gaps in the X-ray light curve; such behavior has been reported in SDSS J120136.02+300305.5. The present work places OGLE16aaa in this context, comparing its light curve, spectra, and SED to these scenarios.

Observations and data reduction: The transient OGLE16aaa (z=0.1655) was monitored in OGLE I-band and with Swift UVOT (UVW2, UVM2, UVW1) and XRT, and observed twice with XMM-Newton (PN and MOS). Swift data span from 19 Jan 2016 onward; XMM-Newton observations occurred on 9 June 2016 (XMM1) and 30 Nov 2016 (XMM2). XMM-Newton data were processed with SAS v16.0.0, high-background intervals removed, and PN spectra extracted (35″ radius) with backgrounds from nearby source-free regions; spectra were grouped to ≥5 counts/bin and fit in 0.3–2 keV using XSPEC (v12). Swift XRT data (PC mode) were reprocessed with xrtpipeline (v0.13.2); source spectra were extracted in a 40″ circle with background from 60–110″ annulus. For non-detections, 3σ upper limits were derived from background counts; count rates were converted to fluxes using WebPIMMS assuming a 60 eV blackbody and Galactic NH=2.71×10^20 cm^-2. UVOT photometry was obtained with UVOTSOURCE (5″ aperture), background from 40″ region; AB magnitudes were derived and corrected for Galactic extinction E(B−V)=0.018. Host subtraction and SED fitting: Host UV–MIR photometry (GALEX, APM b_J, 2MASS JHK, WISE 3.6/4.5 μm, and late-time UVOT U/B/V) was fit with FAST to model the host SED and generate synthetic host UV magnitudes (UVW1=20.52±0.18, UVM2=20.75±0.21, UVW2=21.03±0.18 AB), which were subtracted from UVOT measurements. UV/optical transient SEDs were fit with single-temperature blackbodies at multiple epochs to derive temperature, radius, and integrated blackbody luminosity L_BB; uncertainties were obtained via Monte Carlo perturbations of photometry (1000 realizations). X-ray spectral analysis: XMM1 PN spectrum was fit with a single redshifted blackbody (zbbody) plus Galactic absorption (NH=2.7×10^20 cm^-2). XMM2 PN required an additional component to account for excess above ~0.7 keV: either a steep power law (Γ≈6.4±0.5) or a second blackbody (kT_BB2≈90 eV) in addition to the primary soft blackbody (kT_BB1≈51 eV). Fit improvements were quantified via χ^2 and F-test (Δχ^2=46 for two extra parameters; F-test probability 3.6×10^-9). Stacked Swift peak-epoch spectrum was fit with a soft blackbody (kT_BB=73±3 eV), consistent with XMM results. Light-curve analysis and modeling: UV/optical light curves were compared to t^−5/3 and exponential decays; X-ray light curve compiled from Swift and XMM detections and upper limits. Significance of non-detections was evaluated via Poisson statistics (e.g., probability of <1 photon given expected counts). SMBHB tidal disruption models (previously applied to J1201+3003) were fit to the X-ray light curve of OGLE16aaa to reproduce the delayed rise, dips, and flares; parameters explored include primary BH mass, binary mass ratio, eccentricity, penetration factor β, orbital period, and initial phase, with a ~40-day shift between simulated fallback and observed radiation applied to match timing.

- Strongly delayed X-ray emission: No X-ray detected for ~140 days after the optical peak (Lx,0.3–2 keV < 8.9×10^41 erg s^−1, 3σ), followed by a rapid rise within ~2 weeks to a peak Lx ≈ 7×10^43 erg s^−1 (Swift, days 149–153), i.e., a brightening by a factor >60–80 relative to early limits. First detection by XMM-Newton at day 141 had Lx ≈ 2.9×10^42 erg s^−1. - Complex X-ray variability: After the peak, the source exhibited pronounced dips and rebrightenings, including a non-detection by Swift on day 302 (3σ flux limit 9.4×10^−14 erg cm^−2 s^−1) followed by a high-flux detection with XMM-Newton on day 316 (flux 3.18×10^−13 erg cm^−2 s^−1), then another dip and later low-level detections/non-detections. - X-ray spectra are quasi-soft: XMM1 is well fit by a single blackbody with kT_BB = 58±5 eV and only Galactic absorption. XMM2 requires an additional component: either a very steep power law (Γ = 6.4±0.5) or a hotter blackbody (kT_BB2 ≈ 90^{+20}_{−13} eV), with the primary blackbody at kT_BB1 ≈ 51^{+5}_{−4} eV. Stacked Swift peak spectrum has kT_BB = 73±3 eV. Emission above ~2 keV is absent. - Distinct emission regions: UV/optical SEDs are blackbodies with temperatures ~1.6–2.5×10^4 K. The UV/optical emitting radius was ~2×10^15 cm for the first ~40 days and then decreased by a factor of ~3. The X-ray blackbody radius inferred from XMM is ~10^12 cm, comparable to the Schwarzschild radius of a ~1.6×10^6 M☉ BH, indicating an inner-disk origin for X-rays. The X-ray blackbody severely underpredicts the UV/optical flux, implying separate components. - Luminosities and ratios: Peak NUV luminosity increased by a factor ~7.7 compared to GALEX 2003. UV/optical blackbody luminosity reached ~10^44 erg s^−1. The ratio Lx/Lopt was ≈1 near the X-ray peak. - Timing constraints on obscuration: For an intervening absorber at ~1–2×10^15 cm and MBH≈10^6.2 M☉, the estimated Keplerian crossing time is ~50–140 days, comparable to the X-ray dark phase and spacing between flares/dips. - SMBHB modeling: The disrupted-star fallback and accretion modulated by a SMBHB can reproduce the delayed rise and interruptions. Best-fit parameters (illustrative): primary BH mass 10^6 M☉, eccentricity e ≈ 0.4 (0.4–0.6), penetration factor β ≈ 4.5 (3.0–6.0), mass ratio q ≈ 0.25 (0.05–0.9), orbital period Torb ≈ 150 days (140–160), initial phase φ ≈ 1.7π. A ~40-day shift between simulated fallback and observed radiation was applied.

The delayed, rapid X-ray brightening and subsequent dips/flares in OGLE16aaa are atypical for standard single-SMBH TDEs. While a simple 'delayed disk formation' scenario could, in principle, produce an X-ray delay, recent simulations suggest that self-intersection heating near apocenter is insufficient to power the bright UV/optical emission, and efficient circularization near pericenter should yield prompt X-rays—contrary to early non-detections. Instead, the observations favor an obscuration/reprocessing picture: prompt inner-disk X-rays are initially absorbed by a dense, radiation-driven outflow and reprocessed into UV/optical light; after several months, an ionization breakout reduces opacity and allows X-rays to escape, consistent with the ~140-day delay for a ~10^6 M☉ BH and with Lx/Lopt ≈ 1 at peak. However, reprocessing alone does not naturally explain the later strong dips and rapid rebrightenings. Patchy or variable absorption at ~10^15 cm could modulate the observed X-rays on ~tens–hundreds of days, matching the dark phase duration and flare spacing, but would require sharp column density transitions and sustained high opacity with minimal spectral changes, which is challenging. The SMBHB scenario provides a natural mechanism for interruptions in the accretion stream: the secondary BH perturbs fallback trajectories, causing the stream to miss the primary’s accretion radius, producing dips, followed by resumed accretion and flares. SMBHB modeling reproduces the gross features of the X-ray light curve with Torb ~150 days and plausible dynamical parameters. The X-ray spectra—soft thermal with an extra component—are consistent with inner-disk emission plus a transient coronal or additional thermal component, seen in several X-ray-bright TDEs. The optical rebrightening ~20–30 days after the first peak can be explained by fluctuations in accretion power reprocessed by the ejecta, and is compatible with the SMBHB-induced episodic accretion islands early on. Overall, a hybrid picture—early-time reprocessing/obscuration followed by SMBHB-modulated accretion—best matches the multi-epoch, multi-band data.

OGLE16aaa exhibits a striking delayed (~140 days) X-ray brightening by >60× relative to early limits, rapid rise to peak within ~2 weeks, and multiple dips/flares during decay, alongside soft X-ray spectra and cool, extended UV/optical blackbody emission. The data favor an interpretation in which prompt inner accretion disk emission is initially obscured and reprocessed by a dense outflow, with ionization breakout enabling late-time X-ray escape. The later interruptions and rebrightenings are well reproduced by a tidal disruption in a candidate SMBHB system with Torb ~150 days and plausible mass ratio and eccentricity. These results imply that reprocessing plays an important role in early TDE evolution and that X-ray monitoring of optical TDEs can reveal milliparsec-scale SMBHBs. Future densely sampled UV/X-ray observations, improved modeling of extra soft X-ray components, and systematic surveys (eROSITA, Einstein Probe) in synergy with LSST will enable population studies and better discrimination between obscuration and binary-induced variability. Although the estimated GW inspiral time (~1.9×10^7 years) precludes imminent detection, such systems are key targets for future space-based GW missions.

- Sparse early and mid-phase UV coverage limits assessment of UV evolution during X-ray dips/rebrightenings and constrains tests of variable absorption. - Swift XRT spectra have low S/N; spectral parameters (e.g., temperature, additional components) are uncertain outside XMM epochs. - SMBHB model fits involve many free parameters and degeneracies (e.g., mass ratio constraints are broad, q ~0.05–0.9). A 40-day shift between simulated fallback and observed radiation was assumed. - The SMBHB model predicts ~90 days of episodic accretion after the first interruption with Lx ~10^43–10^44 erg s^−1, somewhat at odds with Swift non-detections over that period. - Host internal extinction is uncertain; SED fits did not correct for it due to large uncertainties, potentially affecting UV luminosities. - Assumptions of single-temperature blackbodies for UV/optical and X-ray components are simplifications; the origin of the additional soft X-ray component remains unconstrained.