The Landscape of Unstable Mass Transfer in Interacting Binaries and Its Imprint on the Population of Luminous Red Novae

TL;DR

This work tackles how unstable mass transfer in interacting binaries imprints on the LRNe population by marrying binary-population synthesis with detailed stellar-structure modeling. Using COMPAS to generate a large, metallicity-specific binary sample and MESA to build donor-structure grids, the authors predict CE outcomes and ejecta properties across evolutionary stages, enabling a physically grounded LRN plateau-luminosity function. A key result is a bimodal distribution of plateau luminosities, arising from divergent mass-ejection channels—mergers with tidal disruption versus envelope ejection—and providing a framework testable with upcoming LSST data; the study also addresses gaps by proposing long-duration plateaus from non-impulsive ejections of highly extended progenitors. The work further suggests exotic CE outcomes involving white-dwarf accretors, such as Thorne-Żytkow-like objects and hydrogen-preserving calcium-rich SNe, and demonstrates how progenitor imaging can constrain the unstable mass-transfer physics driving LRNe.

Abstract

A common-envelope (CE) phase occurs when a star engulfs its companion and is widely considered the primary channel for producing Luminous Red Novae (LRNe). In this study, we combine binary-population synthesis with stellar-evolution calculations to systematically estimate the mass, velocity, and launching radius of ejecta produced during coalescence across a range of binary configurations. Our aim is to quantify how unstable mass-transfer dynamics in binaries at various evolutionary stages shape CE outcomes, enabling a predictive framework for modeling the LRN luminosity function. We find a bimodal distribution of plateau luminosities with significant implications for binary mass stability criteria that can be tested with forthcoming LSST observations. This bimodality emerges from differing mass-ejection outcomes during common-envelope interactions, which can lead either to stellar mergers, often accompanied by tidal disruption of the companion, or to successful envelope ejection. Although our predicted plateau luminosities and timescales broadly match existing observations, the models underpredict the number of LRNe with long-duration plateaus ($t_p \gtrsim 100\, \text{d}$) by about a third. We propose that these long-duration events arise from highly extended progenitors whose envelopes are ejected over multiple orbits (i.e., non-impulsively), producing relatively faint, long-lived transients. By constraining ejecta properties and incorporating pre-outburst progenitor imaging, we show how our models can clarify the physical processes that drive unstable mass transfer in these events. Finally, we argue that common-envelope interactions involving white-dwarf accretors can yield exotic outcomes, including red giants containing embedded white dwarfs that resemble Thorne-Żytków objects (TŻOs), along with calcium-rich supernovae that preserve hydrogen envelopes.

Figures (2)

• Figure 1: Initial distributions of the binary systems evolved in COMPAS. The three top panels show the distributions of semimajor axis ($a$ in units of $R_\odot$), and primary ZAMS mass ($M_1$ in units of $M_\odot$) for DEFAULT ( left), SANA ( middle), and MDS ( right) models. The differences between models are significant and clearly illustrate how model parameter choices critically influence binary population outcomes. The black-dotted ( dashed) lines denote a binary separation of 1 au (10 au). The bottom panels shows the ZAMS mass distributions for both $M_1$ (primary) and $M_2$ (secondary) stars.
• Figure 2: The population of unstable mass transfer donors in COMPAS. Top panels show the distribution of donor systems experiencing unstable mass transfer for three different initial conditions. Middle panels show the evolutionary zones based on the internal composition and core definitions derived from our MESA models (Section \ref{['sec:zones']}): Zone I ( blue), Zone II ( green) and Zone III ( pink). The dashed and dotted black lines indicate the ZAMS and upper radial limit of our MESA grid, respectively. Donors with radii beyond this limit are categorized as hyper-extended (Table \ref{['tab:1']}) and are excluded from our analysis. Bottom panels illustrate the mean value of the mass ratio, $q =M_{\text{a}}/M_\text{d}$, of the binary populations. Observed LRNe progenitors with pre-explosion imaging are plotted for comparison Matsumoto2022. From left to right (in order of increasing inferred mass), these include: V1309 Scorpii, V838 Monocerotis, M101-OT2015, SN Hunt248, and AT 2018bwo. We further classify these progenitors as either MS progenitors ( black squares; Zone I) or extended progenitors ( white dots; Zones II and III).