Micro-TDEs and Stellar Disruptions

• Micro-TDEs are stellar-mass tidal disruptions characterized by rapid fallback and hyper-Eddington accretion, distinguishing them from SMBH TDEs.
• They arise in dense star clusters, AGN disks, and nuclear environments, with dynamics driven by few-body interactions and resonant encounters.
• Recent hydrodynamic models reveal multiple regimes—including impulsive, tidal-peeling, and runaway mass transfer—that produce observable signatures across optical, X-ray, and gravitational-wave bands.

Micro–tidal disruption events (micro-TDEs) are the stellar-mass analogues of classical tidal disruption events: a stellar-mass compact object, usually a black hole and in some formulations also a neutron star, tidally disrupts or strongly strips a star, and some works extend the class to planetary or sub-stellar victims. Relative to SMBH TDEs, micro-TDEs operate at much smaller tidal radii, much shorter dynamical and fallback timescales, and far larger accretion rates in Eddington units, so their phenomenology ranges from fast blue optical/UV transients and soft X-ray flares to ultra-long gamma-ray bursts and deci-Hz gravitational-wave bursts, depending on environment, encounter geometry, and emission physics (Rastello et al., 15 Sep 2025, Perets et al., 2016, Li et al., 23 Jan 2025).

1. Definition and governing scalings

The basic disruption condition is shared with all TDEs. For a star of mass $M_*$ and radius $R_*$ encountering a compact object of mass $M_c$, the tidal radius is

$r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$

and the encounter depth is commonly parameterized by

$\beta = \frac{r_t}{r_p},$

with $r_p$ the pericenter distance. Disruption requires $r_p \lesssim r_t$, and for black-hole disruptors it must occur outside the horizon, while for neutron-star disruptors the disruption radius must exceed the stellar surface. In direct $N$-body studies of young star clusters, a micro-TDE is recorded when $r_p \le r_t$; the disrupted star is removed, and a fixed $10\%$ of its mass is accreted onto the black hole as a dynamical bookkeeping prescription rather than a hydrodynamic calculation (Rastello et al., 15 Sep 2025).

Standard fallback theory still provides the reference scaling. The tidal encounter imparts an energy spread $R_*$0, the most-bound debris returns on

$R_*$1

and, for nearly complete disruptions with approximately flat $R_*$2 near zero energy, the late-time fallback rate obeys

$R_*$3

Returning debris circularizes at roughly $R_*$4 into a thick torus or disk. Because $R_*$5 is stellar rather than supermassive, $R_*$6 is typically hours for main-sequence stars and can be far shorter in some compact-star encounters, while the accretion flow is generally hyper-Eddington or super-Eddington in mass supply even when the emergent luminosity is regulated by winds, trapping, or advection (Perets et al., 2016).

2. Dynamical production channels and astrophysical sites

Young star clusters currently provide the most detailed dynamical picture. In direct PETAR simulations of $R_*$7 realizations of clusters with total mass $R_*$8–$R_*$9, King-model initial conditions with $M_c$0, Kroupa IMF, Moe & Di Stefano binary distributions, metallicities $M_c$1 and $M_c$2, Galactic tides, and integration times of $M_c$3, micro-TDEs arise through three channels: single star–black-hole encounters ($M_c$4), binary-mediated interactions ($M_c$5), and few-body multiple encounters ($M_c$6). Nearly all encounters are highly eccentric, with $M_c$7, and the dominant mechanism is resonant few-body dynamics in dense, black-hole-rich cluster cores, where mass segregation increases encounter rates and repeated close approaches drive $M_c$8 below $M_c$9 far more efficiently than single-single passages (Rastello et al., 15 Sep 2025).

Earlier rate arguments emphasized three broader channels outside the young-cluster context: random close encounters in dense clusters, perturbations of wide binaries in the field, and natal-kick encounters in binaries after compact-object formation. In that framework, black-hole micro-TDEs in dense clusters were estimated at a few $r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$0 per Milky-Way galaxy, neutron-star micro-TDEs at a few $r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$1, and natal-kick channels were identified as potentially comparable or larger, with delays of hours to days after the associated supernova in compact systems (Perets et al., 2016).

AGN disks define a distinct low-relative-velocity regime. Gas drag and dynamical friction align stars and stellar-mass black holes with the disk, migration increases the midplane density, and black-hole–star binaries form at separations of order $r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$2 before tidal disruption as the pericenter approaches the tidal radius. Because solar-type stars cannot be tidally disrupted by non-spinning SMBHs above the Hills mass, a TDE-like flare in an AGN with $r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$3 is a particularly clean micro-TDE candidate, and the embedded-disk scenario predicts a volumetric rate of $r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$4 (Yang et al., 2021).

Nuclear environments introduce additional three-body structure. In nuclear star clusters, a stellar-mass black hole can tidally disrupt a star inside the potential of the central SMBH; some debris unbound from the stellar-mass black hole can remain bound to the SMBH and later accrete onto it, producing a delayed “encore” flare. The predicted rates for these double-flare systems are $r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$5–$r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$6 across $r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$7–$r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$8 (Ryu et al., 2024). A more specific galactic-center channel involves a white dwarf orbiting near $r_t = R_* \left(\frac{M_c}{M_*}\right)^{1/3},$9 around an SMBH and undergoing a high-speed encounter with a stellar-mass black hole on an early-EMRI orbit, yielding a $\beta = \frac{r_t}{r_p},$0–$\beta = \frac{r_t}{r_p},$1 X-ray flare from accretion onto the stellar-mass black hole and a delayed, faint optical TDE-like flare from debris later feeding the SMBH (Li et al., 23 Jan 2025).

3. Hydrodynamic regimes: impulsive disruption, tidal peeling, and runaway eccentric mass transfer

The canonical micro-TDE picture is an impulsive, highly eccentric, nearly parabolic encounter, often treated as a one-pass or few-pass disruption. Recent hydrodynamics broaden that picture by showing that low-eccentricity or moderately eccentric star–black-hole binaries can enter extended mass-transfer phases before full disruption. This is the regime of “tidal-peeling events” (TPEs), introduced for gradual, multi-orbit stripping during low-eccentricity inspiral (Xin et al., 2023).

In PHANTOM simulations of main-sequence stars with $\beta = \frac{r_t}{r_p},$2, $\beta = \frac{r_t}{r_p},$3, $\beta = \frac{r_t}{r_p},$4, and $\beta = \frac{r_t}{r_p},$5 around a $\beta = \frac{r_t}{r_p},$6 black hole, with $\beta = \frac{r_t}{r_p},$7–$\beta = \frac{r_t}{r_p},$8 and $\beta = \frac{r_t}{r_p},$9–$r_p$0, compact TPEs show periodic or quasi-periodic accretion spikes over hours to tens of hours, super-Eddington sink rates of $r_p$1–$r_p$2, rapid debris circularization into a compact optically thick disk, and strong reprocessing. At $r_p$3, total disruption occurs within $r_p$4–$r_p$5 orbits across the explored parameter space, whereas for wider pericenters there can be little or no accretion in the first few orbits despite measurable tidal stripping. The disk optical depth reaches $r_p$6, and the quoted viscous timescale is $r_p$7 for the adopted scaling, implying that observed emission can be delayed and smoothed relative to the instantaneous sink accretion history (Xin et al., 2023).

A longer-baseline hydrodynamic bridge between eccentric Roche-lobe overflow and micro-TDEs was established with SPH calculations of a $r_p$8, $r_p$9 ZAMS star interacting with a non-spinning $r_p \lesssim r_t$0 stellar-mass black hole at initial eccentricities $r_p \lesssim r_t$1–$r_p \lesssim r_t$2 and pericenter parameters $r_p \lesssim r_t$3–$r_p \lesssim r_t$4. Two pathways emerge. For $r_p \lesssim r_t$5, adiabatic expansion of the convective envelope drives runaway disruption: at $r_p \lesssim r_t$6 the star disrupts within $r_p \lesssim r_t$7–$r_p \lesssim r_t$8 orbits depending on $r_p \lesssim r_t$9, and in the fiducial case the cumulative mass loss grows as $N$0, the accretion rate peaks at $N$1, and the post-peak decay is approximately $N$2. For $N$3, orbital widening offsets envelope expansion and the system can remain in stable eccentric mass transfer for up to $N$4 simulated orbits, producing repeating, quasi-periodic flares (Chen et al., 3 Jun 2026).

This hydrodynamic structure clarifies why $N$5-body micro-TDE counts and physical disruption outcomes are not identical. In the young-cluster PETAR prescription, every crossing of $N$6 is a disruption by definition, but hydrodynamic mapping indicates a substantial partial-disruption sector. One cluster study estimates that about $N$7 of candidates lie in the full-disruption region while the remaining $N$8 are partial TDEs, with the full-disruption subset again dominated by multiple encounters (Rastello et al., 8 Sep 2025).

4. Electromagnetic, gamma-ray, and gravitational-wave phenomenology

The electromagnetic output depends on whether the observable is set by fallback, viscous processing, or wind reprocessing. In young-cluster calculations that adopt the wind-reprocessed emission models of Kremer et al. (2023), the predicted peak bolometric luminosities are $N$9–$r_p \le r_t$0, with blue optical/UV transients expected from outflow reprocessing during super-Eddington phases. The same work emphasizes that explicit $r_p \le r_t$1 fallback light curves, relativistic pericenter effects, and spin-dependent disk physics are not modeled there, so the optical/UV predictions are tied to wind physics rather than first-principles radiation hydrodynamics (Rastello et al., 15 Sep 2025).

At higher energies, micro-TDEs have long been connected to ultra-long GRBs and X-ray flashes. If a fraction $r_p \le r_t$2 of the disrupted mass is accreted with radiative efficiency $r_p \le r_t$3, the total flare energy can reach

$r_p \le r_t$4

and the characteristic engine duration is $r_p \le r_t$5–$r_p \le r_t$6, set by $r_p \le r_t$7 and/or the disk drain time. Strong outflows can reduce the accreted fraction drastically and produce much fainter events, but jet-dominated cases naturally overlap the ultra-long GRB timescale (Perets et al., 2016).

A recent case study argues that GRB 250702B is well described by a micro-TDE in which a stellar-mass black hole, and possibly a neutron star, partially or fully disrupts a main-sequence star. Three pathways are identified: a dynamical partial/repeating disruption, a natal-kick disruption, and a hybrid natal-kick plus partial case. For the stellar-mass-compact-object interpretation, the tidal scale is

$r_p \le r_t$8

the most-bound fallback time is

$r_p \le r_t$9

and the accretion time is

$10\%$0

These scalings naturally yield multi-hour prompt emission and kilosecond X-ray variability. The observed precursor–main-flare delay of roughly a day is interpreted either as the return time of a surviving core or as the ballistic delay to the first close passage after a natal kick (Beniamini et al., 26 Sep 2025).

Micro-TDEs are also prospective gravitational-wave sources. Near pericenter, the characteristic burst frequency can be approximated by

$10\%$1

so stellar-mass black-hole disruptions of main-sequence stars typically populate low frequencies, while white-dwarf disruptions shift into the deci-Hz band. In young-cluster calculations, white-dwarf disruptions by stellar-mass black holes produce deci-Hz bursts detectable up to $10\%$2 with future detectors such as LGWA and DECIGO, whereas main-sequence disruptions are much weaker and are expected to be observable only within $10\%$3 (Rastello et al., 15 Sep 2025).

5. Rates, demographics, and survey prospects

The most comprehensive present-day volumetric rate estimates come from young star clusters. Mapping the simulated production efficiency

$10\%$4

onto the cosmic star-formation history via

$10\%$5

yields a local micro-TDE rate of approximately $10\%$6–$10\%$7 and a rate of approximately $10\%$8–$10\%$9 at $R_*$00. The channel partition is robust: few-body multiples dominate at all redshifts. The expected event count rises from only a few per year within $R_*$01 to as many as $R_*$02 out to $R_*$03, although detectability is controlled by uncertain emission physics and survey selection (Rastello et al., 15 Sep 2025).

These young-cluster rates exceed earlier estimates for other environments. AGN-disk synthesis gives $R_*$04 overall, with only $R_*$05 in the especially diagnostic heavy-SMBH hosts with $R_*$06 (Yang et al., 2021). Dense-cluster and natal-kick arguments from earlier work are naturally quoted per Milky-Way-like galaxy rather than per comoving volume, at the level of a few $R_*$07 for black-hole channels and a few $R_*$08 for neutron-star channels (Perets et al., 2016). In nuclear star clusters, the subset that produces delayed SMBH encore flares occurs at $R_*$09–$R_*$10 (Ryu et al., 2024).

Survey forecasts reflect this environmental diversity. In the young-cluster calculations, LSST is the most promising facility: the abstract quotes roughly $R_*$11–$R_*$12 events per year, whereas the detectability analysis under optimistic wind-reprocessed assumptions gives “tens up to several tens of thousands” per year. ZTF and ULTRASAT also have discovery potential but with lower yields. The wide interval is attributed to uncertainties in luminosity models, cadence, limiting magnitude, and selection cuts rather than to the dynamical rates alone (Rastello et al., 15 Sep 2025). For AGNs, targeted searches in systems with $R_*$13 exploit the Hills-mass argument and therefore offer a particularly discriminating observational strategy (Yang et al., 2021).

6. Relation to classical TDEs, impostor problems, and open systematics

Micro-TDEs obey the same tidal criterion as SMBH TDEs but occupy a different parameter regime. In the recent young-cluster synthesis, their peak luminosities are $R_*$14–$R_*$15, compared with $R_*$16–$R_*$17 quoted as typical for SMBH TDEs, and their hosts are expected to be off-nuclear or associated with compact stellar systems rather than with the galactic nucleus itself. A broad literature summary further emphasizes the associated mass scalings: smaller compact-object mass shortens $R_*$18, lowers $R_*$19, and raises disk temperatures, pushing the spectral energy distribution toward X-rays and gamma rays unless strong reprocessing intervenes (Rastello et al., 15 Sep 2025, Zabludoff et al., 2021).

This creates a persistent classification problem. The fallback-like $R_*$20 decline is not unique to SMBH TDEs, and soft thermal X-rays are also not decisive in isolation. The main discriminants identified in the classification literature are accurate localization, rise and decay timescale, multi-band SED evolution, optical/UV line diagnostics, and energetics relative to the Eddington luminosity of the suspected compact object. Off-nuclear localization is a red flag for micro-TDEs, IMBH TDEs, ULXs, or GRB-related transients; conversely, in AGNs with $R_*$21, a TDE-like flare is unlikely to be a classical solar-type SMBH TDE and therefore becomes a strong micro-TDE candidate. Optical traits such as strong broad He II, He II/H$R_*$22, and the characteristic UV/optical phenomenology of SMBH TDEs are not expected to be prominent in stellar-mass black-hole micro-TDEs unless reprocessing is substantial (Zabludoff et al., 2021, Yang et al., 2021).

The main uncertainties are now well defined. On the dynamical side, cluster initial densities, black-hole natal kicks and retention, binary fractions, hardening physics, and the demographics of young star clusters dominate the rate normalization. On the microphysics side, the simplification of many $R_*$23-body prescriptions—instantaneous removal at $R_*$24, fixed accreted fraction, and no distinction between partial and full disruption—propagates directly into predicted accreted masses and $R_*$25 distributions. Hydrodynamic threshold studies add further caveats: neglect of radiative cooling, magnetic fields, and explicit radiation hydrodynamics maximizes adiabatic expansion and can shift the stability boundary between regulated eccentric mass transfer and runaway micro-TDEs. This suggests that future detections will be valuable not only as transient discoveries but also as probes of black-hole retention, natal kicks, cluster dynamical state, and the poorly constrained transition from repeated tidal stripping to catastrophic disruption (Rastello et al., 15 Sep 2025, Chen et al., 3 Jun 2026).