Low-hard to high-soft spectral state transitions in the faintest early-X-ray-detected optical tidal disruption event TDE 2025aarm

Andrea Merloni; Arne Rau; Chichuan Jin; Elias Kyritsis; Gaurava K. Jaisawal; Hannah C. I. Wichern; Jean Somalwar; Kirpal Nandra; Lixin Dai; Panos Charalampopoulos

A nearby tidal disruption event shows the same hard-to-soft X-ray state transition that stellar-mass black holes do.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-14 15:04 UTC pith:BT32FSUL

load-bearing objection First reported low-hard o high-soft path in a thermal TDE, plus the faintest early X-ray detection near optical peak; the rising branch is real but rests on sparse early counts and unconstrained thermal parameters outside the XMM epochs. the 3 major comments →

arxiv 2607.09850 v2 pith:BT32FSUL submitted 2026-07-10 astro-ph.HE astro-ph.GA

Low-hard to high-soft spectral state transitions in the faintest early-X-ray-detected optical tidal disruption event TDE 2025aarm

Pietro Baldini , Arne Rau , Andrea Merloni , Jean Somalwar , Se\'an J. Brennan , Elias Kyritsis , Hannah C. I. Wichern , Peter Boorman

show 6 more authors

Panos Charalampopoulos Lixin Dai Gaurava K. Jaisawal Chichuan Jin Tianying Lian Kirpal Nandra

This is my paper

classification astro-ph.HE astro-ph.GA

keywords tidal disruption eventsX-ray spectral statesaccretion diskscoronablack-hole X-ray binariesscale-invariant accretionBowen fluorescenceselection effects

verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

TDE 2025aarm is the second-closest known tidal disruption event, close enough that months of deep X-ray monitoring could catch emission as faint as a few times 10^39 erg/s near optical peak. The source later brightened by nearly two orders of magnitude. Time-resolved spectra show it starting power-law hard, becoming disk soft as luminosity rose, then hardening again. That low-hard to high-soft sequence is routine in black-hole X-ray binaries but had not been reported in a thermal TDE. The authors model the change as a shifting mix of an accretion disk and a Comptonizing corona, and they confirm the optical classification with N III Bowen lines. Because the early X-rays were so faint, the paper also argues that many optically selected TDEs look X-ray dark simply because earlier follow-up was too shallow or too short.

Core claim

TDE 2025aarm is the first thermal tidal disruption event reported to evolve from a low-hard, power-law-dominated X-ray state into a high-soft, disk-dominated state as luminosity rises, then to harden again, matching the classic state sequence of stellar-mass black-hole binaries and describable by varying disk-plus-corona contributions.

What carries the argument

Time-resolved X-ray spectroscopy with a diskbb-plus-powerlaw (or simpl imes diskbb) model, tracked on a photon-index versus flux plane that serves as a multi-instrument surrogate for the hardness-intensity diagram of X-ray binaries.

Load-bearing premise

The hard nuclear X-rays at every epoch are assumed to be produced by the TDE's own disk-corona system rather than by a nuclear X-ray binary population, an ultraluminous X-ray source, or shocks between unbound debris and surrounding gas.

What would settle it

A deep X-ray observation of the host nucleus after the TDE has fully faded that either recovers a steady hard source at the early luminosity level or shows the source has disappeared, together with a joint radio-plus-X-ray test of the debris-shock model.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

Deep, long X-ray campaigns on other nearby TDEs should recover the same rising hard-to-soft branch if the transition is common rather than unique.
The optical-to-X-ray luminosity ratio can reach several thousand near optical peak, so many previously 'X-ray dark' optical TDEs may simply have been undersampled.
Viscous timescales and the Eddington-ratio threshold for the hard-to-soft switch become testable across seven orders of magnitude in black-hole mass.
Bowen N III features and extreme early X-ray faintness can be read as consistent with high-inclination, reprocessing-heavy viewing angles.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If the disk-corona picture holds, routine monitoring of the next few nearby TDEs could map how often the full Q-shaped hardness-intensity loop appears in supermassive systems.
The same data set that rules out a persistent nuclear X-ray binary would also tighten the allowed contribution of unbound-debris shocks versus coronal Comptonization.
A measured transition Eddington ratio near a few percent would give a direct calibration point for scale-invariant accretion models that previously relied only on stellar-mass binaries and AGN.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Desk Editor's Note

First reported low-hard o high-soft path in a thermal TDE, plus the faintest early X-ray detection near optical peak; the rising branch is real but rests on sparse early counts and unconstrained thermal parameters outside the XMM epochs.

read the letter

This is the first thermal TDE with a documented low-hard to high-soft X-ray path, plus the faintest early detection near optical peak (~7e39 erg/s). That combination is new and worth the community’s attention.

What they did well: six months of multi-mission coverage (EP/FXT, Chandra, XMM), unbinned C-stat BXA fits, honest model comparison (powerlaw preferred over pure diskbb), soft-excess modeling with XMM, and a clean simpl imes diskbb plus Monte-Carlo map onto the Γ–flux plane. The optical side is solid—host subtraction, N III Bowen classification, and consistent BH-mass estimates from σ* and SED. They put the early faintness in the Guolo et al. (2024) LBB/LX plane and make the selection-effect argument cleanly. The XRB-analogy discussion (timescales via Goodwin & Mummery, Eddington ratios ~0.3–2 %) is careful rather than hand-wavy.

Soft spots, in proportion. Outside XMM1/2 the diskbb parameters are unconstrained, so the early “hard” label is really “powerlaw preferred and Γ ≲ 2.3.” With CXO1/EP1 counts that low, a soft thermal component can hide; CXO1 is only ~1.8σ above the Lehmer LMXRB prediction and the ULX Poisson probability is ~12 %. They flag both, and the later soft-state + re-hardening are better constrained, but the rising branch of the claimed HID loop is the least secure piece. The Matsumoto & Piran shock/synchrotron alternative remains open without joint radio–X-ray work. None of this kills the result; it just means the “first full low-hard o high-soft TDE” claim is still conditional on the early hard emission being TDE-related corona.

Who it is for: anyone working on TDE accretion, corona formation, or scale-invariance arguments. The data tables and spectra are inspectable. It deserves a serious referee, not a desk reject. I would bring it to reading group and expect to cite the early-faintness and state-transition claims once the early-hard origin is settled.

Referee Report

3 major / 6 minor

Summary. The paper presents multi-instrument X-ray monitoring (Chandra, EP/FXT, XMM-Newton) and HET/LRS2 optical spectroscopy of the nearby TDE 2025aarm. It reports an early 0.2–10 keV luminosity of ~7e39 erg/s near optical peak (faintest early X-ray detection among optical TDEs), a subsequent rise by nearly two orders of magnitude, and a spectral evolution from hard (power-law-dominated, Γ ≲ 2.3) to soft (disk-dominated, Γ ≳ 3) and later re-hardening. Spectral fits (powerlaw, diskbb+powerlaw, simpl×diskbb) and Monte Carlo mapping onto the Γ–flux plane are used to argue that the evolution is driven by changing relative contributions of an accretion disk and a Comptonizing corona, analogous to the low-hard to high-soft branch of black-hole XRB outbursts. Optical spectra confirm a Bowen N III TDE classification; host SED and kinematics yield M_BH ~ 10^7 M_⊙. The authors interpret the extreme early L_BB/L_X as support for selection-effect explanations of the X-ray-bright/faint dichotomy and for scale-invariant accretion.

Significance. If the low-hard to high-soft path is secure, this is the first thermal TDE reported to complete the rising branch of an XRB-like HID loop, strengthening the case for disk–corona physics that is scale-invariant across stellar-mass and supermassive black holes. The unprecedented early faintness also supplies a concrete data point at the extreme end of the L_BB/L_X distribution and motivates deeper, longer X-ray campaigns. Strengths include the multi-facility cadence, unbinned C-stat BXA fitting with PCA backgrounds, explicit model comparison via Bayesian evidence, soft-excess modeling with XMM, and transparent discussion of XRB/ULX contamination and the Matsumoto & Piran shock alternative. The optical classification and host mass estimates are solid supporting material.

major comments (3)

The central claim that TDE 2025aarm is the first thermal TDE to show a low-hard-to-high-soft transition rests on the early epochs (CXO1, EP1, CXO2, EP2, EP3) being genuinely corona-dominated. Table 2 and §3.1 show that diskbb parameters are unconstrained outside XMM1/XMM2; the hard-state label is therefore only that a single powerlaw is preferred over pure diskbb and that Γ ≲ 2.3. With the low net counts of CXO1/EP1 a soft thermal component can remain hidden, so the rising branch of the claimed HID loop (Fig. 3) is the least secure part of the argument. The manuscript should either (i) quantify, via simulations or nested-model comparison, the maximum thermal fraction still allowed by the early spectra, or (ii) rephrase the claim to emphasize the well-constrained soft-state and re-hardening phases while treating the early hard state as suggestive rather than definitive.
§5.1 finds CXO1 only ~1.8σ above the Lehmer et al. (2010) LMXRB prediction inside the Chandra aperture and a ~12% Poisson probability of ≥1 ULX. The subsequent steady rise of the hard component is used to favor a TDE origin, but the early hard-state identification (and therefore the full low-hard-to-high-soft path) remains vulnerable to a nuclear XRB/ULX contribution. A quantitative upper limit on the non-TDE fraction at each epoch, or an explicit statement that the early hard state could be contaminated, is needed before the XRB-analogy claim can be considered load-bearing.
§5.3 notes that Matsumoto & Piran (2026) can reproduce the X-ray evolution via synchrotron shocks from unbound debris, and that a joint radio–X-ray test is currently unavailable. Because the corona interpretation underpins the scale-invariance conclusion, the paper should either perform a minimal joint consistency check with the existing radio detection (Christy et al. 2025) or clearly demote the corona claim to one of two viable scenarios until such data exist.

minor comments (6)

Fig. 1 middle panel and Table 2: Γ is derived from single-powerlaw fits while fluxes come from diskbb+powerlaw; a short note clarifying that the two models are used for different purposes would avoid confusion.
Table 2 notes that thermal parameters are unconstrained for most epochs; the table caption or a footnote could flag which epochs have only upper limits on kT/Norm to make the limitation immediately visible.
§3.2 simulations fix Γ_scatt = 1.8 and explore only two normalizations; a brief statement that the qualitative track is robust to modest changes in these choices would strengthen the figure.
Optical line-fitting (Fig. 5, Table B.1): the spectral-merging artefact around N III λ4641 is acknowledged; a quantitative upper limit on any residual N III λ4641 flux would make the Bowen classification more robust.
Typographical: “tha Palomar” → “the Palomar”; “Zwicki” → “Zwicky”; occasional double spaces and “ÅL Etudes” in the acknowledgements.
Eq. (1) for L_Bol omits the Kubota/Shimura–Takahara correction factors; the text already notes this, but a parenthetical range with and without the factors would help the reader.

Circularity Check

0 steps flagged

Observational spectral-state claim uses standard external models; no derivation reduces to its own inputs by construction.

full rationale

The paper's central claim is empirical: time-resolved X-ray spectra of TDE 2025aarm show a hard-to-soft-to-hard evolution in photon index and flux, interpreted with standard diskbb/powerlaw/simpl models and compared qualitatively to XRB HID tracks. Model preference is assessed via Bayesian evidence on the data (powerlaw preferred over pure diskbb; soft excess requires diskbb+pow or simpl imes diskbb on XMM). The §3.2 simulations map assumed (kT, f_scatt, Norm) combinations onto the observed Γ–flux plane for visualization only; they do not fit parameters to a subset and then 'predict' a forced related quantity. Eddington-ratio estimates use diskbb parameters from the same spectra plus external scaling relations and a cited viscous-timescale argument (Goodwin & Mummery 2026), not a self-defined uniqueness theorem. Companion notes (Simongini et al. 2026 for discovery/optical peak; Matsumoto & Piran 2026 for an alternative shock scenario) supply context and caveats, not a load-bearing uniqueness or ansatz that forces the state-transition claim. No self-definitional loop, fitted-input-as-prediction, or renaming of a known result as a new derivation is present. Score 1 reflects only ordinary, non-load-bearing self-reference to the discovery note.

Axiom & Free-Parameter Ledger

4 free parameters · 5 axioms · 0 invented entities

The central claim rests on standard X-ray spectral components, Galactic absorption and redshift fixed from external catalogs, BH-mass scaling relations, and the interpretive mapping of Γ/flux evolution onto an XRB-like disk–corona picture. Free parameters are the usual per-epoch spectral fit parameters; no new physical entity is introduced beyond applying the known corona concept to this TDE.

free parameters (4)

Per-epoch power-law photon index Γ
Fitted independently to each spectrum; used to define hard (Γ<3) vs soft (Γ>3) states and to place points on the Γ–flux plane.
diskbb kT_in and normalization
Fitted when statistics allow (especially XMM1/2, EP4–6); drive the claimed soft-state disk dominance and bolometric luminosity estimates via L_Bol ∝ R_in² T_in⁴.
simpl scattering fraction f_scatt and Compton Γ
Fitted in the physically motivated model; used to argue corona vs disk relative contribution changes across the transition.
Host-subtraction multiplicative normalization for HET spectra
Chosen to remove Ca H&K absorption; affects residual line fluxes and Bowen classification strength.

axioms (5)

domain assumption Hard X-ray power-law emission in accreting black holes can be modeled as Comptonization of disk seed photons (simpl×diskbb / corona picture).
Invoked in §3.2 and §5.3 to interpret spectral evolution as disk–corona interplay analogous to XRBs.
domain assumption Line-of-sight Galactic N_H = 4.4×10^20 cm⁻² (HI4PI) and host redshift z=0.0138 (DESI DR1) are fixed and correct.
Fixed in all spectral fits (§3.1); errors would rescale fluxes and soft-band shape.
domain assumption Γ=3 is a useful threshold separating thermal and non-thermal emission in accreting SMBHs.
Used to label hard vs soft states (Fig. 1 caption; Grotova et al. citations); the transition narrative depends on this conventional cut.
domain assumption M–σ and M–M* scaling relations give a ~10^7 M_⊙ black hole for Eddington-ratio comparisons.
§4.1–4.2 and §5.3; used to claim transition near ~0.3–2% Eddington as in XRBs.
standard math Standard nested-sampling X-ray spectral inference with PCA background models is adequate for unbinned EP/Chandra/XMM spectra.
BXA/UltraNest methodology in §3; background model for FXT borrowed from eROSITA PCA.

pith-pipeline@v1.1.0-grok45 · 33135 in / 3654 out tokens · 40025 ms · 2026-07-14T15:04:06.969579+00:00 · methodology

0 comments

read the original abstract

We report the X-ray and optical spectroscopic properties of TDE 2025aarm, the second closest tidal disruption event (TDE) discovered to date. The proximity of this source, combined with a deep and intense X-ray monitoring campaign spanning six months, allowed us to probe the source down to an unprecedented 0.2-10 keV luminosity of $\sim7\times10^{39}$ erg s$^{-1}$ close to the optical peak. This renders TDE 2025aarm the faintest early-X-ray-detected TDE to date. After the first X-ray detection, the source brightened by nearly two orders of magnitude, reaching a peak luminosity of $\sim5\times10^{41}$ erg s$^{-1}$ about four months after the optical peak. Through time-resolved X-ray spectral analysis, we find that TDE 2025aarm evolved from an initially hard, power-law-dominated X-ray state into a softer, disk-dominated state as the luminosity increased, before hardening again at later times. Such low-hard-to-high-soft state transitions are commonly observed in black hole X-ray binaries (XRBs) but have not previously been reported in thermal TDEs. We show that the spectral evolution can be described by variations in the relative contributions of an accretion disk and a Comptonizing component, qualitatively resembling the disk--corona evolution observed in XRBs. We also present the results of our optical spectroscopic follow-up campaign with HET/LRS2, confirming the TDE classification and revealing NIII Bowen fluorescence features. The extremely faint early-time X-ray emission of TDE 2025aarm further supports the idea that the historical dichotomy between X-ray-bright and X-ray-undetected TDEs is largely driven by selection effects related to the depth, cadence, and duration of X-ray follow-up observations. TDE 2025aarm therefore provides new insight into both the accretion physics of TDEs and the possible universality of accretion across several orders of magnitude in black hole mass.

Figures

Figures reproduced from arXiv: 2607.09850 by Andrea Merloni, Arne Rau, Chichuan Jin, Elias Kyritsis, Gaurava K. Jaisawal, Hannah C. I. Wichern, Jean Somalwar, Kirpal Nandra, Lixin Dai, Panos Charalampopoulos, Peter Boorman, Pietro Baldini, Se\'an J. Brennan, Tianying Lian.