Luminous Red Novae: Stellar Merger Transients

• Luminous Red Novae are optical/IR transients that bridge the gap between novae and supernovae, characterized by two-stage light curves and evolving spectra.
• They originate from dynamical binary mergers or common envelope ejection, as evidenced by precursor brightening and a shift from hot to red, molecular-dominated spectra.
• Observational and modeling studies reveal ejecta masses ranging from 10⁻³ to 10 M☉, impacting our understanding of binary evolution and the cosmic dust budget.

Luminous Red Novae (LRNe) are a distinct class of optical/infrared transients observed to occupy the intermediate luminosity regime between classical novae and supernovae. These events are now interpreted as the electromagnetic signatures of dynamical binary coalescence or common envelope (CE) ejection in interacting non-compact stellar binaries. LRNe provide a unique window into the physics of unstable mass transfer, CE evolution, stellar mergers, and their roles in the assembly of compact binaries and the cosmic dust budget.

1. Physical Definition, Phenomenology, and Classification

LRNe are characterized by peak absolute magnitudes ranging from $M_r \simeq -3$ to $-16$ and typical bolometric luminosities $L_{\mathrm{peak}}\sim10^4$–$10^8\,L_\odot$ (Kaminski et al., 16 May 2026). These transients exhibit slow, red-evolving ejecta (velocities $v\sim100$–$1000$ km s$^{-1}$) and a pronounced metamorphosis from hot, blue early-time spectra to very red, molecular-dominated late-time spectra (Pastorello et al., 2019, Pastorello et al., 2022). Their durations span weeks to years, with canonical multi-stage light curves comprising (I) a gradual pre-outburst brightening; (II) a rapid blue peak (rise timescales of days to a week, $T_{\mathrm{eff}}\sim7,000$–$11,000$ K); (III) an extended red plateau or secondary maximum ($T_{\mathrm{eff}}\sim4,000$–$-16$0 K); and (IV) a dust-enshrouded decline and infrared-luminous remnant stage (Kaminski et al., 16 May 2026, Pastorello et al., 2019).

The two-peaked (blue then red) or plateau-dominated light curve is a hallmark distinguishing LRNe from both classical novae and core-collapse SNe, as well as from other “intermediate luminosity red transients” (ILRTs) and LBV eruptions (Karambelkar et al., 2022, Matsumoto et al., 2022). The detection of molecular bands (TiO, VO, CO, H$-16$1O, SiO) in late spectra and the consistent transition to very red colors (V–I $-16$2 1 by $-16$3200 d after peak) are strong identifiers of the class (Pastorello et al., 2019, Pastorello et al., 2022).

2. Physical Origin: Binary Mass Transfer, Common Envelope, and Merger

The physical engine driving LRNe is the dynamical merger or CE ejection in a close binary system. This scenario is strongly supported by pre-outburst imaging (e.g. decaying binary periods and slow photometric brightening in V1309 Sco and others) (Addison et al., 2022, Pastorello et al., 2022), progenitor identification (in the Hertzsprung gap, yellow supergiant, or contact configuration) (Kaminski et al., 16 May 2026, Blagorodnova et al., 2021), and hydrodynamic modeling (Hatfull et al., 2024, Kirilov et al., 12 Aug 2025).

CE formation typically proceeds when the more massive (“donor”) star evolves off the main sequence and unstable Roche-lobe overflow (RLOF) occurs, enveloping the companion (often a lower-mass main sequence or compact star) (Addison et al., 2022, Howitt et al., 2019). The inspiral deposits orbital energy in the donor’s envelope:

$-16$4

where $-16$5 and $-16$6 are the component masses, $-16$7 and $-16$8 the initial/final separations. Envelope ejection requires that some fraction $-16$9 of $L_{\mathrm{peak}}\sim10^4$0 unbinds the envelope against its binding energy $L_{\mathrm{peak}}\sim10^4$1 (Addison et al., 2022, Twum et al., 10 Feb 2026). As the system evolves, several outcomes are possible: (a) successful envelope ejection, yielding a tighter binary; (b) runaway inspiral and merger producing a single remnant (Twum et al., 10 Feb 2026, Howitt et al., 2019); or exotic transient configurations (e.g. Thorne–Żytkow–like objects with embedded white dwarfs) (Twum et al., 10 Feb 2026).

3. Observational Properties: Light Curves, Spectra, and Progenitors

LRNe characteristically show:

• Precursor phase: Multi-year, $L_{\mathrm{peak}}\sim10^4$2–1 mag slow rise, attributed to increasing mass loss or luminosity from pre-CE L$L_{\mathrm{peak}}\sim10^4$3/L$L_{\mathrm{peak}}\sim10^4$4 outflows and circumbinary disk formation (Addison et al., 2022, Pastorello et al., 2022).
• Double-peaked morphology: A short blue peak from the fastest, lowest-mass ejecta layers, followed by a more luminous, redder, and longer plateau powered by recombination or shock interaction in the bulk ejecta (Matsumoto et al., 2022, Pastorello et al., 2019, Pastorello et al., 2022).
• Spectroscopic progression: Early-time spectra show hot continuum with strong, narrow H$L_{\mathrm{peak}}\sim10^4$5 (Lorentzian with $L_{\mathrm{peak}}\sim10^4$6–$L_{\mathrm{peak}}\sim10^4$7 km s$L_{\mathrm{peak}}\sim10^4$8), Balmer decrement H$L_{\mathrm{peak}}\sim10^4$9/H$10^8\,L_\odot$0, and P Cygni Fe II features (Williams et al., 2015, Pastorello et al., 2022). Spectral evolution proceeds to increasingly red continua, metal line forests (Fe II, Ba II, Ca II), and finally to strong molecular bands (TiO, VO, CO) as $10^8\,L_\odot$1 drops below 4000 K (Pastorello et al., 2019, Blagorodnova et al., 2021).
• Polarimetry and morphology: Low continuum polarization but strong line depolarization, with signatures of asymmetric/bipolar ejecta, especially in resolved remnants (Kaminski et al., 16 May 2026).
• Dust formation: Mid-infrared (MIR) excess, 10 $10^8\,L_\odot$2m silicate features, dust masses $10^8\,L_\odot$3–$10^8\,L_\odot$4 forming on timescales of months to years post-outburst (Reguitti et al., 20 Apr 2025, Karambelkar et al., 5 Aug 2025).
• Remnant properties: Survivors emerge as inflated M-type or red supergiant-like stars, with luminosities $10^8\,L_\odot$5–$10^8\,L_\odot$6, photospheric radii $10^8\,L_\odot$7–$10^8\,L_\odot$8, and decreasing IR brightness as dust shell fragments (Reguitti et al., 20 Apr 2025, Blagorodnova et al., 2021).

Table: Typical Light Curve and Spectral Properties

Phase	Duration	$10^8\,L_\odot$9 (K)	Spectral Features
Precursor	yrs–decades	6,000–8,000	Brightening, A/B SED, H$v\sim100$0 emission, IR excess
Blue peak	days–weeks	7,000–11,000	Balmer lines, Fe II, blue SED
Red plateau	weeks–months	4,000–5,500	Metal absorptions, weaker H lines, IR excess
Decline (dust)	months–yrs	<3,000	Molecular bands, MIR excess

[References: (Addison et al., 2022, Pastorello et al., 2019, Blagorodnova et al., 2021, Reguitti et al., 20 Apr 2025, Pastorello et al., 2022)]

4. Mass Ejection Physics, Energetics, and Diversity

The photometric and spectroscopic evolution of LRNe is interpreted in terms of the outflow energetics and geometry. Ejecta masses span $v\sim100$1–$v\sim100$2, with kinetic energies $v\sim100$3–$v\sim100$4 erg (Kaminski et al., 16 May 2026, Matsumoto et al., 2022). The plateau luminosity and duration can be analytically estimated via recombination-powered models (cf. Popov 1993):

$v\sim100$5

$v\sim100$6

(Twum et al., 10 Feb 2026, Howitt et al., 2019, Matsumoto et al., 2022). These scalings reproduce the observed $v\sim100$7–$v\sim100$8 distribution, though the brightest, longest-plateau events often exceed the recombination-powered luminosity ceiling, requiring an additional energy source.

Single-zone and multi-shell models (e.g. (Matsumoto et al., 2022)) show that the two-peaked structure arises from different ejecta components: an early peak powered by the fastest, hot layers (thermal energy) and a red plateau from recombination in slower, more massive ejecta. Observed diversity in durations and peak luminosities is set by the interplay of progenitor mass, envelope structure, ejection geometry, and pre-dynamical outflows.

Increasing evidence favors shock interaction with a pre-existing circumbinary medium (CBM) or aspherical mass loss (e.g., from L$v\sim100$9/L$1000$0) as a critical contributor to the highest-luminosity and longest-duration LRNe (Kirilov et al., 12 Aug 2025, Pastorello et al., 2019). Simulations confirm that the collision of merger ejecta ($1000$11–3 $1000$2 at 400–500 km/s) with a dense CBM of comparable mass can raise plateau luminosities to $1000$3erg/s and durations to $1000$4200 days (Kirilov et al., 12 Aug 2025).

The role of jets and “grazing envelope evolution” (GEE) has been identified as essential for explaining peculiar LRN features such as discrete bumps/dips in the plateau, rapid late-time photospheric radius evolution, and aspherical remnants in certain events (Soker, 2024).

5. Progenitor Demography, Event Rates, and Population Synthesis

Archival imaging and population synthesis indicate that LRN progenitors span a broad mass range: in the Milky Way, systems like V1309 Sco ($1000$5), V838 Mon ($1000$6–$1000$7) (Williams et al., 2015, Kaminski et al., 16 May 2026), and extragalactic examples traced to yellow supergiants of $1000$8–$1000$9 (Blagorodnova et al., 2021, Pastorello et al., 2022). Donors are predominantly post-main-sequence, often in the Hertzsprung gap or giant branch. Binary population synthesis models (e.g., COMPAS) find a Galactic LRN event rate of $^{-1}$0 yr$^{-1}$1, consistent with observed rates, and a bimodal luminosity function: dim, short-lived transients from mergers and brighter, long-plateau events from successful CE ejection (Howitt et al., 2019, Twum et al., 10 Feb 2026).

ZTF-CLU data yield a systematic volumetric LRN rate:

$^{-1}$2

for $^{-1}$3 (Karambelkar et al., 2022). The luminosity function steepens at high $^{-1}$4 ($^{-1}$5), in agreement with massive merger predictions, and the implied rate of luminous events matches the birthrate of double compact object binaries inferred from gravitational-wave data (Karambelkar et al., 2022, Jain et al., 24 Nov 2025).

However, only $^{-1}$6 of LRNe form compact-object binaries that merge in a Hubble time. The bulk culminate in single stellar remnants (Jain et al., 24 Nov 2025, Kaminski et al., 16 May 2026).

6. Late-Time Evolution, Remnants, and Dust Synthesis

At late epochs, LRNe transition to the IR/MIR, with the remnant often fading in optical but remaining bright at 3–8 $^{-1}$7m for years or decades (Reguitti et al., 20 Apr 2025, Karambelkar et al., 5 Aug 2025). SED modeling requires multiple blackbodies: a warm/hot surviving stellar photosphere ($^{-1}$8–$^{-1}$9 K), a cooler dust shell ($T_{\mathrm{eff}}\sim7,000$0–$T_{\mathrm{eff}}\sim7,000$1 K), and sometimes a cold extended echo component ($T_{\mathrm{eff}}\sim7,000$2 K) (Reguitti et al., 20 Apr 2025, Karambelkar et al., 5 Aug 2025).

Dust production is significant: dust masses of $T_{\mathrm{eff}}\sim7,000$3–$T_{\mathrm{eff}}\sim7,000$4 are inferred per event, with dust-to-gas mass ratios up to $T_{\mathrm{eff}}\sim7,000$5 and slow ($T_{\mathrm{eff}}\sim7,000$6100–250 km/s) expansion facilitating efficient survival against reverse-shock destruction (Karambelkar et al., 5 Aug 2025). JWST MIR spectroscopy reveals rich oxygen-rich molecular chemistry (H$T_{\mathrm{eff}}\sim7,000$7O, CO, SiO; “water fountain” signatures) and confirms that LRNe can contribute $T_{\mathrm{eff}}\sim7,000$825% of the core-collapse SN cosmic dust budget (Karambelkar et al., 5 Aug 2025).

The long-term stellar remnant is invariably an inflated, red supergiant-like object—never a true collapse or disappearance—confirming merger scenarios over explosive terminal events (Reguitti et al., 20 Apr 2025).

7. Open Questions, Impact, and Future Directions

Open theoretical problems remain regarding the relative roles of recombination, radiative shocks, and jets in powering LRN plateaus (Soker, 2024, Kirilov et al., 12 Aug 2025, Pastorello et al., 2019). Outstanding observational frontiers include the precise mapping of LRN luminosity functions, the prevalence of rare exotic outcomes (e.g. Thorne–Żytkow-like objects, hydrogen-rich Ca SNe), and the detailed characterization of precursor outflows and their link to binary evolutionary state (Kaminski et al., 16 May 2026, Addison et al., 2022).

Forthcoming time-domain surveys (LSST, Roman, ULTRASAT), multi-epoch IR campaigns (JWST, ELTs), and ALMA submillimeter follow-up will dramatically expand the LRN sample and enable population-scale constraints on binary assembly, dust formation, and the energetics of stellar mergers (Kaminski et al., 16 May 2026, Howitt et al., 2019). LSST is forecast to discover $T_{\mathrm{eff}}\sim7,000$9–1,500 LRNe per year, allowing tests of bimodality, light-curve morphology, and statistical progenitor-remnant mapping (Twum et al., 10 Feb 2026, Howitt et al., 2019).

LRNe are now recognized as critical tracers of binary evolution physics, mass transfer instability, the formation pathway of compact-object mergers, and significant dust sources in the ISM. Their study has opened a new era for the empirical calibration of CE theory, multi-messenger astrophysics, and the integrated lifecycle of stellar and circumstellar matter in galaxies.