The twin red giant branch system BD+20 5391 A case study of low-mass double-core evolution

TL;DR

BD+20 5391 is a rare SB2 system consisting of two nearly identical red giants in an $P \approx 81$ d orbit. Using 15 Ondřejov spectra and one high-resolution HERMES spectrum, the authors perform radial-velocity fitting, spectral analysis, and SED/HRD-based evolutionary inferences, anchored by Gaia DR3 parallax. They find both stars have $M \approx 1.4$–$1.5\,M_\odot$, $T_{\rm eff} \approx 4900$ K, and $M_{\rm He} \approx 0.335\,M_\odot$ at the onset of Roche lobe overflow (RLOF), with a current separation $a \approx 111\,R_\odot$ and periastron $r_{\rm peri} \approx 93\,R_\odot$. The near-simultaneous RLOF predicts a double-core CE phase, offering two main outcomes: a short-period double He-WD binary if the envelope is ejected, or a merger producing a He-rich hot subdwarf (potentially magnetic) if the cores merge. The study provides a benchmark for interpreting RGB binary interactions and their descendants in upcoming large spectroscopic surveys.

Abstract

Understanding interactions of binary systems on the red giant branch is crucial to understanding the formation of compact stellar remnants such as helium-core white dwarfs (He-WDs) and hot subdwarfs. However, the detailed evolution of such systems, particularly those with nearly identical components, remains under-explored. We aim to analyse the double-lined spectroscopic binary system BD+20 5391, composed of two red giant stars, in order to characterise its orbital and stellar parameters and to constrain its evolution. Spectroscopic data were collected between 2020 and 2025 using the Ondřejov Echelle Spectrograph and the Mercator Échelle Spectrograph. The time-resolved spectra were fitted with models to determine the radial velocity curve and derive the system's parameters. We then used the position of both stars in the Hertzsprung-Russell diagram to constrain the system's current evolutionary state, and we discuss potential outcomes of future interactions between the binary components. We find that the two stars in BD+20 5391 will likely initiate Roche lobe overflow (RLOF) simultaneously, leading to a double-core evolution scenario. The stars' helium core masses at RLOF onset will be almost identical, at 0.33 $\mathrm{M}_{\odot}$. This synchronised evolution suggests two possible outcomes: common envelope ejection, resulting in a short-period double He-WD binary, or a merger without envelope ejection. In the former case, the resulting double He-WD may merge later and form a hot subdwarf star. This study provides a valuable benchmark example for understanding the evolution of interacting red giant binaries, which will be discovered in substantial numbers in upcoming large-scale spectroscopic surveys.

Figures (6)

• Figure 1: Colour-magnitude diagram of observed red giant stars, based on Gaia DR3 GaiaDR3. The sample includes RGB stars just below the red clump within 500 pc of the Sun. Blue triangles mark spectra from OES, red circles from AISAS. BD+20 5391 is highlighted in yellow. The grey background marks all Gaia DR3 objects with parallax_over_error$>100$ within 500 pc of the Sun.
• Figure 2: Radial velocity curve of BD+20 5391 for components A (red) and B (blue). Circles show measurements obtained with the OES, and the triangles mark the HERMES measurement.
• Figure 3: Selected ranges in the double-lined HERMES spectrum of BD+20 5391. The upper section shows the observed data in light grey and the combined model fit as a dashed black line, while the lower part displays the contributions of the individual components, offset by $-$0.5 for clarity. Component A is shown in red, component B in blue; strong lines are labelled.
• Figure 4: SED of BD+20 5391. Component A is red, component B blue, and the combined model is grey. Photometric measurements are provided by Gaia/XP DeAngeli2023_XP, GaiaRiello2021, APASS Henden_apass9, and 2MASS Skrutskie2006_2MASS. Interstellar extinction was applied to the models.
• Figure 5: HRD for BD+20 5391. Component A is marked in red and component B in blue. The squares denote the parameters derived from the SED and spectral fitting, while triangles indicate the onset of RLOF. The best-fit MIST evolutionary tracks are represented by discrete dots extending from the zero-age main sequence to the tip of the RGB. To visualise the uncertainties, we also show a random subset of tracks from our MCMC calculation.