An observational study of rotation and binarity of Galactic O-type runaway stars

TL;DR

This study compiles the largest sample of Galactic O-type runaways with kinematics, rotation, and binarity from Gaia DR3 and IACOB data to constrain runaway origins. The authors show that most runaways are slow rotators, while fast-rotating runaways are more likely BSS products, with high-velocity runaways arising from DES or two-step scenarios; SB2 systems rarely reach runaway speeds. Their analysis identifies three SB1 runaways that are HMXBs and several SB1 runaways as BH candidates, providing valuable observational benchmarks for binary evolution and compact-object formation. The results support a mixed origin for Galactic O-type runaways, with DES dominating the slow-velocity regime and BSS contributing to the fast-velocity tail, offering concrete constraints for population-synthesis models and cluster dynamics. The work highlights the utility of combining kinematics, rotation, and binarity in testing runaway-ejection theories and sets the stage for Gaia DR4-era refinement.

Abstract

Gaia DR3 data have revealed new massive runaway stars, while spectroscopic surveys enable detailed characterization. The relative contributions of binary supernova (BSS) and dynamical ejection (DES) scenarios to explain their runaway origin remain poorly constrained, particularly in the Milky Way. We aim to characterize the largest sample of Galactic O-type runaway stars ever investigated through their kinematics, rotation, and binarity to shed light into their origins. We use the GOSC-Gaia DR3 catalog, and IACOB spectroscopic information to build a sample with 214 O-type stars with projected rotational velocities ($v \sin{i}$), and a subsample of 168 O-type stars with additional information about their likely single (LS) or single-lined (SB1) spectroscopic binary nature. We also consider an additional sample of 65 double-lined (SB2) spectroscopic binaries. We find that among our sample of Galactic O-type runaways, most (74%) have $v \sin{i}<200$ km/s, whereas for normal stars this fraction is slightly higher (82%). There are no fast-moving runaways being fast rotators, except for HD 124 979. Runaways show lower SB1 fractions than normal stars, with no runaway SB1 fast-rotating systems; on average, runaways rotate faster than normal stars; and their runaway fraction is higher among fast rotators (44%) vs. the slow rotators (34%). This is consistent with BSS dominance for fast rotators. We also found that SB2 systems hardly reach runaway velocities with a low runaway fraction (10%). Runaways with 2D velocities > 60 km/s are mostly single and interpreted as DES products, while runaways with 2D velocities > 85 km/s are also interpreted as two-step products. Three of 12 runaway SB1 systems are HMXBs. Our study reveals that most Galactic O-type runaways are slow rotators, suggests a dominance of BSS among fast-rotating runaways, and of DES and two-step among the high-velocity ones. (Abridged)

Figures (8)

• Figure 1: Projected rotational velocity as a function of the 2D peculiar velocity for the 214 O-type normal (black) and runaway (blue) stars in the rotation sample. The error bars in $V_\text{PEC}^\text{2D}$ are the individual uncertainties of the stars computed in MCC2023. The uncertainties in $v \sin{i}$, not shown for clarity, are 10% of the values. The vertical dotted and solid lines indicate the usual 2D runaway threshold and the fast runaway threshold, at 25 and 85 km s$^{-1}$, respectively. The horizontal dashed line indicates the fast rotation threshold at 200 km s$^{-1}$. Top: $V_\text{PEC}^\text{2D}$ distribution for the normal and runaway stars with a bin size of 5 km s$^{-1}$. Right: $v \sin{i}$ distribution for the normal and runaway stars with a bin size of 20 km s$^{-1}$.
• Figure 2: Cumulative distribution functions of $v \sin{i}$ for the 136 O normal (black) and 78 runaway stars (blue) in the rotation sample with bins of $\sim$20 km s$^{-1}$. The curves in light colors show the corresponding empirical cumulative distribution functions. $p_{\rm KS}$ is the corresponding $p$-value resulting from the KS test.
• Figure 3: Projected rotational velocity as a function of the 2D peculiar velocity for the 168 O-type normal and runaway stars in the rotation LS-SB1 subsample. LS systems are shown as green circles, while SB1 systems are shown as red squares. Normal and runaway stars are represented with empty and filled symbols, respectively. The three known HMXBs identified among SB1 systems are indicated with red crosses. The uncertainties of the data points, and vertical and horizontal lines are the same as in Fig. \ref{['Fig:2D_rotsample']}.
• Figure 4: Interpretation of the possible runaway ejection scenarios in the $v \sin{i}$ vs. $V_\text{PEC}^\text{2D}$ plane. The labels related to the scenarios indicate the following. BSS: BSS ejection, resulting either in a bound or disrupted binary. BSS disrupted: BSS ejection of a former binary that becomes disrupted. DES disrupted: DES ejection of a former binary (note the high $v \sin{i}$) that becomes disrupted. DES-single: after DES ejection the runaway is identified as single. DES-binary: after DES ejection the runaway is identified as a binary. Two step: DES + BSS. Avoidance region: region with virtually no runaways. The colored regions are discussed in the text.
• Figure 5: Lower-left part of the projected rotational velocity as a function of the 2D peculiar velocity plot presented in Fig. \ref{['Fig:2D_rotLS_SB1']}. HMXBs are indicated with red crosses and circles, and binaries with BH candidates are indicated with black triangles and dashed circles.