Inefficient Circularization, Delayed Stream-Disk Interaction and Reprocessing: A Five-Stage Model for the Intermediate-Mass Black Hole Tidal Disruption Event EP240222a

TL;DR

This work tackles the peculiar multi-wavelength evolution of the IMBH-TDE EP240222a, where standard tidal disruption models fail due to non-negligible circularization and strong reprocessing effects. The authors introduce a five-stage semi-phenomenological framework that begins with inefficient circularization, followed by a slow-rise SSC-driven precursor disk, a fast-rise caused by delayed stream-disk interaction, a super-Eddington X-ray/optical plateau with significant reprocessing, and finally a sub-Eddington decline. A new X-ray light-curve fit using a seven-parameter MCMC approach yields a physically plausible solution, with a main-sequence star of about $0.4\,M_\sun$ and a penetration factor near unity, consistent with the observed X-ray dominance and faint early optical emission. The model clarifies how IMBH-TDEs can produce slow X-ray rises, long X-ray plateaus, and reprocessed optical emission, offering predictive signatures for identifying such events and guiding future time-domain surveys, including WD-disruption scenarios that could yield rapid starts in X-rays.

Abstract

EP240222a is the first intermediate-mass black hole (IMBH) tidal disruption event (TDE) captured in real-time with multi-wavelength observations and spectroscopic confirmation. However, its light curves deviate substantially from previous theoretical expectations. Motivated by these unique features, we have developed a novel model that successfully reproduces its peculiar evolution. Our model delineates five stages: (1) Initial Stage of inefficient circularization; (2) Slow-Rising Stage with a faint X-ray precursor disk fed by successive self-crossings; (3) Fast-Rising Stage, where delayed stream-disk interaction at momentum flux matching drives a sharp luminosity rise; (4) Plateau Stage with super-Eddington accretion, outflow, reprocessing, and a clear polar line-of-sight; and (5) Decline Stage of sub-Eddington accretion and ongoing reprocessing. Our fit indicates the disruption of a $M_* \approx 0.4~M_\odot$ main-sequence (MS) star with a penetration factor $β\approx 1.0$. Our model, which incorporates key TDE processes, establishes EP240222a-like light curves as typical IMBH-TDE signatures. The distinctive identifier is a slow rise in X-rays and a corresponding slow rise/quasi-plateau in the UV/optical, followed by a brighter, super-Eddington plateau in both bands, though other forms exist, such as the rapid rise from white dwarf (WD) disruptions over minutes to days.

Figures (4)

• Figure 1: EP240222a X-ray light curve fit. Data: eROSITA (red), LEIA (orange), EP-FXT (olive), NICER (blue), XMM-Newton (black). The observed 0.5-4 keV band corresponds to 0.516–4.130 keV in the rest frame. Overplotted: median-parameter model (red solid), point-wise median model (black dashed), and the corresponding 1$$ (68%) credible interval (gray shade). The model captures the distinct multi-stage evolution: a years-long slow rise, a sharp transition to the super-Eddington plateau (linear decay), and a subsequent power-law decline.
• Figure 2: A schematic diagram illustrating the five-stage model. (1) Initial Stage: The three main inefficient dissipation channels (nozzle shock, self-intersection, tail-main stream interaction) produce faint emission. Self-intersection primarily drives both weak inflow and outflow. Timescale: $t_{\mathrm{fb}}$. (2) Slow-Rising Stage: Within a few $t_{\mathrm{fb}}$, the main stream becomes thick from trapped radiation, while the tail remains thin. The dissipation processes continue (with the thick main stream's path shown schematically by the dashed lines), dominated by a succession of self-crossings (SSC). Inflow forms a precursor disk, which produces early X-rays. Low-luminosity X-rays lead to faint potential reprocessed UV/optical. Timescale: years. (3) Fast-Rising Stage: Triggered by momentum flux matching between the disk and stream. A runaway circularization ensues, as the main stream violently merges into the disk. Disk mass builds up rapidly ($\sim t_{\mathrm{fb}}$), driving a sharp X-ray rise. Potential strong UV/optical flares might be suppressed by high optical depths. Torques from the main stream's merger rapidly align the disk with the BH's spin, thus misaligning it with the incoming tail stream. Timescale: $t_{\mathrm{fb}}$. (4) Plateau Stage: A massive disk in a super-Eddington stage powers both outflow and Eddington-limited X-rays, which show a slow, linear decay. A polar funnel provides a clear line-of-sight to the X-rays. Weak optical arises from reprocessing by the outflow. The thin, dense tail stream penetrates the outflow to feed the disk, producing negligible light due to its low mass flux. Timescale: $t_{\mathrm{acc}}$. (5) Decline Stage: As the disk mass depletes, the accretion rate drops to sub-Eddington. Consequently, X-rays transit from a linear to a power-law decay and the outflow ceases. The reprocessing layer expands and becomes optically thin over time (not yet reached). Timescale: years. Note: Darker shades indicate higher density (for the thick stream in (2) and (3), density is represented by the dashed line's shade). Colors are for visual clarity only. Not to scale (reality: $R_{\mathrm{S}} \ll R_{\mathrm{p}} \sim R_{\mathrm{d}} \sim r_{\mathrm{s,p}} \ll a_0$).
• Figure 3: EP240222a model fit corner plot. Quoted values and black dashed lines denote the posterior medians and 1$$ (68%) credible intervals. Darker 2D regions indicate higher probability density.
• Figure 4: Early optical models vs. ZTF upper limits (purple arrows, 90% CL). Green lines show lower (persistent suppression) and upper (fully efficient radiation) bounds for $_{\mathrm{self}}(t)=0.01$, utilizing median parameters in Fig. \ref{['fig:corner']}. ZTF data are consistent with the expectation that the true light curve evolves from the lower to the upper bounds, constraining $_{\mathrm{self}}(t) \lesssim 0.01$. This implies highly inefficient self-intersection and/or inaccuracies in our simple thermal correction. In either case, we conclude inherently faint optical emission. The upper limits, all preceding fitted $t_{\mathrm{crit}}$ (red dashed), support our model as well.