M31-2014-DS1: Failed Supernova Candidate

• M31-2014-DS1 is a heavily obscured stellar remnant in Andromeda and a prime candidate for studying failed supernova and black hole formation.
• Multiwavelength observations reveal a significant mid-IR brightening with a prolonged plateau, alongside strong molecular and dust signatures, while optical/NIR emissions vanish.
• Comparative analyses debate fallback accretion versus binary-driven transients, advancing our understanding of massive star evolution and the physics of envelope ejection.

M31-2014-DS1 is a heavily obscured, optically vanished stellar remnant in the Andromeda galaxy that has become the primary extragalactic candidate for observing black hole (BH) formation via the "failed supernova" mechanism. This source is associated with the abrupt disappearance of a yellow supergiant between 2014 and 2022, followed by persistent mid-infrared (mid-IR) emission and stringent non-detections in optical, near-infrared (NIR), millimeter, and X-ray bands. Its status as a failed supernova, a merger, or an intermediate-luminosity optical transient (ILOT) is the subject of active debate, with recent JWST and Chandra observations at the core of this controversy. The source is pivotal to discussions of black hole birth, fallback accretion, neutrino emission, dust obscuration, and non-terminal stellar interactions.

1. Multiwavelength Observational Properties

Since 2014, M31-2014-DS1 has transitioned from an optically visible yellow supergiant with HST+Spitzer pre-collapse photometry (luminosity $L_0 \approx 10^5 L_\odot$, $T_\mathrm{eff} \approx 4500$--$5300$ K, $R_* \approx 516$ $R_\odot$) to complete optical disappearance and emergence as an extremely red, dust-enshrouded mid-IR source. Key time-domain photometry includes:

• NEOWISE W1/W2 and JWST/MIRI data show a $\sim$50% mid-IR brightening over $\sim$900 days, followed by a plateau ($L_\mathrm{plateau}\approx4\times10^{38}$ erg s$^{-1}$) lasting nearly 1000 d, and then a $\sim 10\times$ decline over 1000 d (De et al., 2024). Fluxes at $T_\mathrm{eff} \approx 4500$0--$T_\mathrm{eff} \approx 4500$1 μm in late 2024 are $T_\mathrm{eff} \approx 4500$2–$T_\mathrm{eff} \approx 4500$3 mJy (MIRI F1500W–F2550W) (Beasor et al., 8 Jan 2026).
• Optical/NIR photometry from Gaia, PS1, PTF, and ZTF exhibits a fade by $T_\mathrm{eff} \approx 4500$4 (with limits $T_\mathrm{eff} \approx 4500$5, $T_\mathrm{eff} \approx 4500$6), ruling out both classical core-collapse supernovae and luminous optical outbursts (De et al., 2024, Beasor et al., 8 Jan 2026).
• Spectroscopy with JWST/NIRSpec and MIRI reveals strong molecular absorption: blueshifted lines of CO, CO$T_\mathrm{eff} \approx 4500$7, H$T_\mathrm{eff} \approx 4500$8O, SO$T_\mathrm{eff} \approx 4500$9 at 4–8 μm, and deep silicate absorption at 8–13 μm. Dust continuum models require dense, warm silicate grains ($5300$0 K, $5300$1 mag, $5300$2) (De et al., 9 Jan 2026, Beasor et al., 8 Jan 2026).
• Chandra X-ray Observatory: no counts within a $5300$3 aperture (10.7 ks), yielding $5300$4 erg s$5300$5 (0.5–7 keV), well below fallback accretion expectations for an unobscured BH (Beasor et al., 8 Jan 2026, De et al., 9 Jan 2026).
• SMA 1.3\,mm continuum: rms $5300$6 mJy, undetected down to $5300$7 mJy, excluding cold gas/dust masses $5300$8 (Beasor et al., 8 Jan 2026).

2. Progenitor Characteristics and Stellar Evolution

Archival photometry and detailed DUSTY and MESA modeling constrain the progenitor to a yellow supergiant of $5300$9–$R_* \approx 516$0 (alternate evolutionary scenarios extend to $R_* \approx 516$1–$R_* \approx 516$2 with enhanced winds or binary stripping (De et al., 2024)). The H envelope at collapse was low-mass ($R_* \approx 516$3–$R_* \approx 516$4), with total radius $R_* \approx 516$5 and surface dust shell at $R_* \approx 516$6–$R_* \approx 516$7 au.

MESA models yield final pre-collapse parameters as follows:

Parameter	Value	Source
Luminosity $R_* \approx 516$8	$R_* \approx 516$9	(De et al., 2024)
$R_\odot$0	4500–5300 K	(De et al., 2024, De et al., 9 Jan 2026)
H envelope mass	$R_\odot$10.3–0.6 $R_\odot$2	(De et al., 2024)
Pre-SN Mass	$R_\odot$3	(De et al., 2024, De et al., 9 Jan 2026)

Dust characteristics include a warm inner shell ($R_\odot$4 K, $R_\odot$5 au, $R_\odot$6), with a dust mass $R_\odot$7 inferred under spherical symmetry (De et al., 9 Jan 2026), although more generic non-spherical radiative transfer modeling suggests higher dust masses ($R_\odot$8) (Beasor et al., 8 Jan 2026).

3. Panchromatic Diagnostics: Fallback and Weak Ejection

The prevailing failed supernova ("fSN") interpretation ascribes the observed phenomena to a low-energy ($R_\odot$9–$\sim$0 erg) shock that ejected $\sim$10.1 $\sim$2 of H-rich envelope at $\sim$3100 km\,s$\sim$4 while the rest of the star collapsed directly into a $\sim$5–$\sim$6 black hole (De et al., 9 Jan 2026, De et al., 2024). Panchromatic evidence for this scenario includes:

• Bolometric Fading: The IR luminosity declined by $\sim$7 over 1000 d post-plateau, following $\sim$8 as expected for fallback accretion of loosely bound envelope material (De et al., 2024, De et al., 9 Jan 2026).
• Molecular Gas Expansion: Best-fit slab models indicate $\sim$9 of molecular gas in expansion at 100 km s$\sim$0, with blueshifted line profiles ($\sim$1100 km s$\sim$2) (De et al., 9 Jan 2026).
• Ejecta Geometry and Structure: Deep silicate and molecular absorption features reveal dense, warm dust, but also suggest significant deviations from spherical symmetry—implying that 1D modeling yields only lower limits for the intrinsic luminosity when poles are partially unobscured (Beasor et al., 8 Jan 2026, Soker, 20 Jan 2026).
• X-ray Non-detections: The lack of detectable X-ray emission ($\sim$3–$\sim$4 erg s$\sim$5) constrains fallback accretion efficiency ($\sim$6) and/or requires absorbing columns $\sim$7 cm$\sim$8, as expected for heavy enshrouding by slow ejecta and fallback material (De et al., 9 Jan 2026, Beasor et al., 8 Jan 2026).

A key analytic result is the accretion luminosity evolution: $\sim$9 with fallback rate $L_\mathrm{plateau}\approx4\times10^{38}$0 and radiative efficiency $L_\mathrm{plateau}\approx4\times10^{38}$1 up to $L_\mathrm{plateau}\approx4\times10^{38}$2. Dust continuum modeling and light-curve shape jointly require the actual fraction of fallback material reaching the BH ($L_\mathrm{plateau}\approx4\times10^{38}$3) to be $L_\mathrm{plateau}\approx4\times10^{38}$4 at most (De et al., 9 Jan 2026).

4. Alternative Interpretations and Model Controversies

A critical challenge to the failed SN scenario is posed by the necessity for extreme parameter tuning: most fallback must not accrete, and angular-momentum support, jets, or outflows must prevent further luminosity or radiative output for $L_\mathrm{plateau}\approx4\times10^{38}$5 years. Multiple studies (Soker, 20 Jan 2026, Beasor et al., 8 Jan 2026) argue for alternative origins:

• Binary-driven Type II ILOT: Violent binary interaction or merger (Type II ILOT) offers a plausible mechanism—ejecting a massive, non-spherical (torus or equatorial) dust shell and yielding a fading mid-IR source with no coincident X-rays or optical outburst (Soker, 20 Jan 2026, Beasor et al., 8 Jan 2026). The geometry is naturally asymmetric, matching spectroscopic and SED analyses, and high dust masses support merger models rather than SN-like mass loss.
• Fallback and Jets: Jet-driven expulsion of fallback-formed accretion disks is expected given large specific angular momentum in the pre-collapse envelope. Viscous timescales ($L_\mathrm{plateau}\approx4\times10^{38}$6 s at $L_\mathrm{plateau}\approx4\times10^{38}$7) allow prompt jet launching, predicted to quickly unbind returned material and suppress prolonged accretion (Soker, 20 Jan 2026).
• Radiative Cooling and Overluminosity: Disk-wind/ejecta interactions are expected to be extremely radiatively efficient ($L_\mathrm{plateau}\approx4\times10^{38}$8 dynamical times), producing luminosities $L_\mathrm{plateau}\approx4\times10^{38}$9, well above observed levels ($^{-1}$0), thereby challenging steady fallback-powered disk models (Soker, 20 Jan 2026).
• Empirical Prevalence: Only a tiny fraction ($^{-1}$1) of massive star deaths exhibit missing-progenitor/failed-SN signatures (Soker, 20 Jan 2026), and census studies have not found the population of vanishing red supergiants required by the failed SN framework.

5. Neutrino Constraints on Black Hole Formation

The black hole birth scenario for M31-2014-DS1 is directly testable via neutrino astronomy. SK-IV (Super-Kamiokande) performed a cluster search for $^{-1}$2 signals in the 2014–2017 window defined by JWST/NEOWISE constraints (Nakanishi et al., 5 Nov 2025). The principal features:

• Emission Profile: Standard core-collapse results in a $^{-1}$3 burst lasting $^{-1}$41–2 s, with luminosity $^{-1}$5–$^{-1}$6 erg (Suwa et al., 28 Apr 2025).
• Detector Sensitivity: Above 18 MeV, SK's background rate is $^{-1}$7 d$^{-1}$8. No $^{-1}$9-event burst clusters were found, leading to a $\sim 10\times$0 C.L. upper limit $\sim 10\times$1 erg, marginally ($\sim 10\times$2) above state-of-the-art model predictions (Shen–TM1 EOS: $\sim 10\times$3 erg) (Nakanishi et al., 5 Nov 2025, Suwa et al., 28 Apr 2025).
• Implications: Only the stiffest EOS and heaviest progenitors ($\sim 10\times$4–$\sim 10\times$5) approach or mildly exceed current exclusion limits (Suwa et al., 28 Apr 2025). Next-generation detectors (Hyper-Kamiokande, JUNO, DUNE) are predicted to achieve $\sim 10\times$6 higher sensitivity for M31 events, allowing robust discrimination among EOS and failed SN vs. alternative scenarios.

6. Synthesis: Current Status and Comparative Scenarios

The dominant interpretations for M31-2014-DS1 are summarized below:

Feature/Requirement	Failed SN (fSN)	Type II ILOT/Merger
Ejecta mass	$\sim 10\times$7 (from envelope)	$\sim 10\times$8–$\sim 10\times$9 asym.
Ejecta energy	$T_\mathrm{eff} \approx 4500$00–$T_\mathrm{eff} \approx 4500$01 erg	up to $T_\mathrm{eff} \approx 4500$02 erg
Long-term fade mechanism	Low-efficiency fallback accretion ($T_\mathrm{eff} \approx 4500$03)	Dust reprocessing, expansion/cooling
IR plateau duration	$T_\mathrm{eff} \approx 4500$04–$T_\mathrm{eff} \approx 4500$05 d, then %%%%106$T_\mathrm{eff} \approx 4500$0107%%%% fade	Years–decades, roughly constant
Dust geometry	(MODELS) Spherical or shell	Equatorial torus, asymmetric
X-ray signature	Weak (obscured/inefficient accretion)	Absent

This suggests that both paradigms can account for major aspects of the light curve and mass loss, but only merger or ILOT paradigms naturally explain asymmetry, high dust mass, persistence of IR luminosity, and severe X-ray suppression without fine-tuning fallback and averted jet formation (Soker, 20 Jan 2026, Beasor et al., 8 Jan 2026). The lack of optical outburst and the $T_\mathrm{eff} \approx 4500$08decade-long IR plateau are quantitatively matched by fallback accretion models, but only if radiative efficiency and accreted fraction are strongly suppressed.

7. Broader Implications and Future Directions

The fate of M31-2014-DS1 is critical to understanding end-states of massive stellar evolution, the origin of stellar-mass black holes, and the physics of low-energy transients. The detection (or lack thereof) of associated neutrino bursts constrains both microphysical models of neutron star matter (EOS) and the macroscopic pathways of envelope ejection, fallback accretion, and jet formation (Nakanishi et al., 5 Nov 2025, Suwa et al., 28 Apr 2025).

Ongoing JWST monitoring will clarify whether the IR luminosity continues to decline or if the optical source re-emerges, as expected in merger/ILOT scenarios but not classical failed SNe (Beasor et al., 8 Jan 2026). High spatial and spectral resolution studies may resolve dust geometry, and next-generation neutrino facilities will test the predictions of massive star collapse to black holes in distant galaxies with unprecedented precision.

The case of M31-2014-DS1 demonstrates that panchromatic, multi-messenger observations are needed to definitively distinguish between failed supernovae and non-terminal binary/merger events in massive stars, with direct implications for core-collapse physics, BH formation rates, and the role of binaries in transient astrophysics (De et al., 9 Jan 2026, Beasor et al., 8 Jan 2026, De et al., 2024, Soker, 20 Jan 2026, Nakanishi et al., 5 Nov 2025, Suwa et al., 28 Apr 2025).