Multiplicity of Massive stars in the Milky Way (M3W). II. The brightest single O stars: Variability and systemic velocities

Multiplicity of Massive stars in the Milky Way ( is a project designed to augment our knowledge of the dynamical properties of massive stars within several kiloparsecs of the Sun. An often overlooked aspect of such properties is the determination of accurate systemic velocities, as they are affected by variability effects such as pulsations, line infilling by wind and disks, and incomplete knowledge about the central wavelengths. This lack of accuracy impairs studies on the origin of runaway and walkaway stars as well as on the velocity dispersions of stellar clusters. In this second paper of the series we wanted to increase the accuracy of the O stars systemic velocities in both isolated and multiple systems. To that end, we established how different wavelength regions are affected by variability, determined the thresholds for the detection of spectroscopic binarity, and obtained accurate central wavelengths including the effects of contamination, forbidden lines, and wind infilling. Our sample consists of the 41 brightest O stars (B łe 7.0 mag) thought to be single at the beginning of the project (summer 2025). We used multi-epoch high-resolution spectroscopy from the łili project cleaned with the UNWIND code and stellar models computed with FASTWIND. The epoch spectra were cross-correlated to study their variability in the form of radial velocity (RV) dispersions. The RVs were computed from (a) cross-correlation of H and He lines with FASTWIND models, (b) profile fitting of metallic lines, and (c) (for very fast rotators) cross-correlation of the full spectrum with a template star. A comparison between the RV dispersions for wavelength regions respectively dominated by He lines and by Hβ reveals different groups of stars in terms of variability: (a) non-supergiant slow rotators, (b) fast rotators, (c) supergiants, and (d) Oe stars. This is likely a manifestation of the different role of pulsations, winds, and disks in RV variability. We used that analysis to determine RV dispersion thresholds for the detection of new spectroscopic binaries (SBs). We also used a FASTWIND grid to establish the effects of (a) the presence of lines close to Balmer lines, (b) the existence of forbidden components in two lines, and (c) wind infilling in the determination of accurate RVs. With all of that, we determined RVs for our 41 stars, checked their accuracy comparing results from our three methods, and in some cases discovered significant differences with literature values. Our detailed analysis has further revealed four SB1 systems that were unknown in the summer of 2025 (68 Cyg, HD , HD A, and HD ), one new SB2 (HD ), two new Oe stars (HD and HD ), and three new Onfp stars (ξ Per, 68 Cyg, and HD ). He ii He i 152424 209339 93521 46150 57682 93521 175876