Evidence for the merger hypothesis in V4332 Sgr: a low $^{12}$C/$^{13}$C ratio and multiple outbursts

TL;DR

This study tests whether the 1994 eruption of V4332 Sgr was caused by a stellar merger or planetary ingestion. It combines a revised light curve showing multiple outbursts with high-resolution CO spectroscopy that yields a very low $^{12}$C/$^{13}$C ratio of $3.5\pm1.0$, compatible with deep mixing from a merger. The energy budget and spectral diagnostics, including strong Na/K features and a pre-existing edge-on disk, favor a merger with surface enrichment from CNO-processed material over planet ingestion. The findings provide direct support for the merger hypothesis in V4332 Sgr and extend the case for LRNe as merger-driven events, while highlighting the role of disks and multiple outbursts in such systems.

Abstract

Following the detections of the first extragalactic ``Luminous Red Nova'' (LRN) M31 RV in 1989, and its first Galactic counterpart V4332~Sgr in 1994, there have been many discoveries of similar, or closely related, objects. They are important because they bridge the luminosity gap between the brightest novae and supernovae, a largely unexplored parameter space. The cause of eruptions in LRNe is still unclear, a stellar merger being the most favored mechanism. However, barring V1309~Sco, there has been no direct evidence for a merger in the other objects. Here we present strong evidence that V4332~Sgr was a merger event. High resolution infrared observations of the CO fundamental band show an unusually small $^{12}$C/$^{13}$C ratio of $3.5\pm1$. This indicates that a violent event had occurred, whose effects penetrated deep enough to allow CNO cycle processed $^{13}$C in the inner H burning shell to be brought to the surface. We rule out planetary ingestion, and propose that the eruption was due to a merger between V4332~Sgr and a companion star. It is shown that V4332~Sgr was likely surrounded by an edge-on disk before its eruption. If this disk was a flattened common envelope containing V4332~Sgr and a companion star, then a merger scenario would not be inconsistent. Furthermore, V4332~Sgr had multiple outbursts, previously unreported but an important piece of information, since multiple outbursts are a trait shared by many LRNe.

Figures (3)

• Figure 1: Top: light curve of V4332 Sgr, showing three clear and two possible peaks. Data from AAVSO and AFOEV. See text for discussion of the three points enclosed in the red square. Middle and bottom: V838 Mon and CK Vul, similar objects displaying multiple outbursts; the data for CK Vul are from Table 1 of shara85
• Figure 2: Observed echelle spectra of V4332 Sgr (black) with the model fit (green). The $^{12}$CO and $^{13}$CO rotational lines are marked with blue and red ticks respectively. Details are given in Table \ref{['log']} and in the text.
• Figure 3: Top left: strong emission in the sodium and potassium resonance lines in a small section of the V4332 Sgr spectrum taken in 2005 from the Very Large Telescope by one of us (DPKB; PI of program ESO-VLT 075.D-0511). These data are analyzed in detail by tylenda15. The bottom panel shows a LTE model plot to show that the unusual strength of the K and Na lines can be explained without using enhanced abundances (see text). Right column: H$\alpha$ profile of V4332 Sgr (red) on 1994 March 7 comparing its deep central absorption with the shell profiles of three Be stars silaj10 to show the similarity in profiles and hence establish that V4332 Sgr had a disk around it even before its 1994 eruption. See text for details.