Tales of stellar and binary co-evolution, told by stellar oscillations -- Binary demographics and their impact on stellar mass, orbits, and age estimates in main-sequence and red-giant stars

TL;DR

The paper investigates how binarity biases mass, orbital properties, and age estimates for red giants by integrating asteroseismic constraints from the APOKASC3 and APO-K2 catalogs with Gaia DR3 non-single-star solutions. It quantifies binary fractions, mass distributions, and orbital parameters across evolutionary stages (MS → RGB → RC/2RC), and reveals strong, mass-dependent binary attrition during the red-giant phase, with substantial disruption of $M \leq 1.8\,M_\odot$ systems and short-period RC/2RC binaries likely shaped by past interactions such as mass transfer or common-envelope events. The study finds attrition of about $69\%$ to the low-luminosity RGB and $\sim 81\%$ to the high-luminosity RGB relative to the MS, plus roughly $38\%$ additional attrition to the RC, and highlights that RC systems with $P_\mathrm{orb} \lesssim 800$--$1000$ days are particularly influenced by prior interactions, potentially biasing ages if treated as single stars. These findings underscore the need to identify binarity when using red giants as population tracers and demonstrate the powerful synergy between asteroseismology and Gaia in constraining stellar and binary co-evolution, with future Gaia DR4, PLATO, and Roman data expected to substantially enhance the census.

Abstract

Red giants are increasingly used as stellar population tracers due to their well-understood evolution and the availability of asteroseismic observables. However, stellar binarity can alter observable properties and introduce strong biases. We aim to provide a holistic picture of the binary population and its evolution in the red giant phase by characterizing a sample of binaries hosting oscillating red giants from a combination of extensive asteroseismic, spectroscopic, and astrometric surveys. We investigate the binary properties of evolved stars in the APOKASC3 and APO-K2 catalogs, leveraging asteroseismic constraints and Gaia DR3 non-single-star solutions. We explore the mass distribution of red-giant binary systems and analyze the evolution of their binary fraction. For stars with M$\leq$1.8M$_\odot$, we find binary fractions $\sim$31% and $\sim$41% for oscillating and non-oscillating solar-like stars on the main-sequence (MS). By the power excess ($ν_\mathrm{max}$) as luminosity proxy, we detect a binary attrition of $\sim$69% and $\sim$81% on the low- and high-luminosity red-giant branch (RGB) and an additional $\sim$38% to the red clump (RC), with respect to the MS. Binaries hosting RC and secondary clump stars (2RC) stars are largely depleted at $P_\mathrm{orb}\lesssim$500 and $\lesssim$200 days, respectively. Mass-dependent differences in binary fractions and orbital properties point to more substantial binary attrition for stars with M $\leq$1.8 M$_\odot$. The distinct mass distributions and the depletion of short-period binaries during the red-giant phase underscore the impact of stellar expansion and binary interactions on stellar evolution. RC systems with $P_\mathrm{orb}\lesssim$800 to 1,000 days are likely shaped by past interactions, such as mass transfer or loss, which can lead to significantly biased age estimates if not accounted for.

Figures (9)

• Figure 1: Seismic Hertzsprung-Russell Diagrams of the investigated catalogs and binary samples. The two panels from left to right depict the targets in the APOKASC 3 catalog, with solutions in the Gaia catalogs of Two-Body Orbits (TBO) as dots and Non-Linear/Acceleration Solutions (NLAC) from NSS as squares. Additionally, the second panel shows the Binary Union Flag (x), which RUWE dominates. The third panel shows the stars with rapid surface rotation ($v\sin i$$\geq$5 km/s) as triangles. Rapid-rotating targets found in binary systems are depicted as dots and squares or marked with crosses, depending on their solution type. The right panel presents TBO, NLAC, and RUWE binary indicators for targets in the APO-K2 catalog. The marker's surface color indicates its seismically determined evolutionary state, as indicated by the legend in the left panel; RGB: red-giant branch, RC: red clump, 2RC: secondary clump, No EvSt: no evolutionary state determined but mass reported, 'No M' indicates lower quality data that did not allow for a seismically inferred stellar mass. Stars with NLAC solutions are further marked with black edges.
• Figure 2: Normalized kernel density estimates of the distributions for the stellar mass (left) and metallicity (right) for the APOKASC 3 (light) APO-K2 (dark) samples. The dots and vertical bars mark the respective sample's median and the interquartile range (25$^\mathrm{th}$–75$^\mathrm{th}$ percentiles). The dashed magenta line indicates the mass threshold of 1.8 M$_\odot$.
• Figure 3: Binary fraction as a function of their evolutionary states. Pentagons mark stars that undergo core-degeneration on the RGB (M/M$_\odot$$\leq$ 1.8), while star symbols indicate the binary fraction for stars that will ignite He under non-degenerate conditions (M/M$_\odot$$>$ 1.8). The style of the connecting lines depicts the significance of the variations as discussed in the text and specified in Table \ref{['tab:binReductionRate']}. The yellow bar indicates the binary fraction among MS stars without oscillations. The marker color depicts the evolutionary state as indicated on the horizontal axis.
• Figure 4: Normalized kernel density estimates of the stellar mass distributions for red giant stars in the APOKASC 3 (top) and APO-K2 (bottom) samples. Each KDE is split by evolutionary stage, with the left side (orange) showing RGB stars and the right side displaying core-helium burning stars. The latter are further divided into RC (teal) and 2RC (purple) stars, separated at 1.8 M$_\odot$. Distributions are shown for the full sample and several binary candidate subsamples identified by the TBO, NLAC, and Binary Union Flag (BUF) for APOKASC 3 and RUWE for APO-K2, as well as the rapid rotator subset in APOKASC 3. Dots and vertical bars indicate the median and interquartile range (25$^\mathrm{th}$–75$^\mathrm{th}$ percentiles) for each population. Horizontal orange and teal lines mark the median RGB and RC stellar masses of the respective full samples.
• Figure 5: Orbital parameters of binary systems with seismically inferred evolutionary states. In addition to the color code for the evolutionary states of red giants (Fig. \ref{['fig:seismicHRD']}), oscillating MS+SG primaries are shown in yellow. Stars found to be rapid rotators are marked as triangles. The background color map represents the KDE of the full SB9 catalog, with black lines outlining regions where the density exceeds seven times the median value. Vertical dashed and dotted lines represent the 1034 d timebase of Gaia DR3 and $\sim$500 d limit for RGB stars with degenerate cores to evolve as single stars. Likely artifact solutions for giants (P$_\mathrm{orb,TBO}$$\leq$ 10 days) are shown with a lower opacity.