MONOS Project: Northern O-Star Multiplicity

• MONOS is a dedicated project that homogeneously characterizes Galactic O-type spectroscopic binaries and multiple systems in the northern hemisphere.
• It integrates high-resolution multi-epoch spectroscopy, classification from GOSSS, and lucky imaging from AstraLux to resolve hierarchical multiplicity and refine orbital parameters.
• The project combines Gaia DR3 and TESS photometry with spectroscopic data to jointly constrain periods, orbital geometry, and key stellar parameters in massive binary systems.

Searching arXiv for the MONOS astrophysical project and its supporting LiLiMaRlin context. Searching for "MONOS Multiplicity Of Northern O-type Spectroscopic systems". MONOS, short for Multiplicity Of Northern O-type Spectroscopic systems, is a project devoted to the homogeneous characterization of Galactic O-type spectroscopic binaries and multiple systems in the northern hemisphere, defined by $\delta>-20^\circ$. Its program combines classification spectroscopy, high-resolution multi-epoch spectroscopy, and high-angular-resolution imaging to establish spectral types, resolve hierarchical multiplicity, test published orbital solutions, and derive new spectroscopic orbits. In later work, the project extends this framework with Gaia DR3 epoch photometry and TESS light curves, allowing periods, orbital geometry, and stellar parameters to be constrained jointly with radial velocities (Apellániz et al., 2019, Holgado et al., 6 Aug 2025).

1. Scientific scope and astrophysical rationale

MONOS is motivated by the astrophysical centrality of O stars and by the fact that their multiplicity is not a secondary detail but a primary determinant of their evolution. O stars dominate the feedback budget of galaxies through their ionizing radiation, winds, and supernovae, and their high multiplicity implies that binary and multiple-star interactions are essential for understanding their evolution. In particular, close massive binaries can interact during the main sequence, exchange mass, and alter both the stellar population and the chemical and dynamical evolution of galaxies (Apellániz et al., 2019).

The project is explicitly northern in scope. The boundary between the northern and southern programs is set at $\delta=-20^\circ$, chosen so that the number of OB stars in each hemisphere is roughly balanced given the increasing abundance of OB stars in the southern sky toward the Galactic Center and at higher extinction. In this division, MONOS is the northern counterpart to southern efforts such as OWN; LiLiMaRlin was constructed in part to support both hemispheres through a common high-resolution spectral archive (Apellániz et al., 2018).

From the outset, MONOS was framed not as a one-time cataloging exercise but as a staged program. The first paper concentrates on sample definition, homogeneous spectral classifications, and visual multiplicity for systems with previously published spectroscopic or eclipsing orbits. The later orbital papers move from classification and multiplicity vetting to period determination, orbit refinement, and combined spectroscopic-photometric modeling (Apellániz et al., 2019, Holgado et al., 6 Aug 2025).

2. Sample definition and multiplicity nomenclature

The initial MONOS sample was built from the Galactic O-Star Catalog. From the private GOSC version, the authors identified 520 O-type systems with $\delta>-20^\circ$. From that parent set, they selected 92 systems with previously published spectroscopic and/or eclipsing orbits, restricting companion types to O, B, or compact objects. Five of the selected systems had only eclipsing orbits in the literature and no published spectroscopic orbit. The paper also included ten optical companions observed with the primary targets (Apellániz et al., 2019).

A deliberate exclusion criterion was applied to systems suspected of radial-velocity variability but lacking a published orbit. Such variability can be caused by pulsations or other non-orbital effects, so those systems were not admitted to the first MONOS sample (Apellániz et al., 2019). This selection choice is methodologically important: MONOS initially prioritized orbit-bearing systems for which a homogeneous reassessment could be grounded in prior literature and later improved with new data.

The project also formalized a richer multiplicity nomenclature than the standard SB1/SB2 dichotomy. SB1 and SB2 retain their standard meaning as single-lined and double-lined spectroscopic binaries. E marks eclipsing systems, a indicates spatially resolved astrometric or visual components, s indicates spatial separation in the GOSSS long-slit data, C denotes a constant outer companion in the spectroscopic aperture, and S denotes a spectroscopic outer component. Compound labels such as SB2E+Ca, SB2a+Sa, or SB2E+SB1as therefore encode both spectroscopic and visual hierarchy (Apellániz et al., 2019).

This terminology reflects the project’s central premise: many O-type systems are hierarchical, and orbital interpretation frequently requires simultaneous handling of spectroscopic structure, visual multiplicity, and aperture contamination. A plausible implication is that MONOS treats multiplicity as a system-level inference problem rather than as a binary-labeling problem.

3. Observational infrastructure and data homogenization

MONOS rests on three principal observational resources: GOSSS for homogeneous spectral classification, LiLiMaRlin for high-resolution multi-epoch spectroscopy, and AstraLux plus WDS for visual multiplicity constraints.

Resource	Role in MONOS	Key characteristics
GOSSS	Homogeneous spectral classification	Blue-violet spectra at $R\sim2500$
LiLiMaRlin	Multi-epoch high-resolution spectroscopy	18,077 spectra of 1,665 stars
AstraLux + WDS	Visual multiplicity and astrometry	Lucky imaging and catalog support

GOSSS provides blue-violet spectroscopy at $R\sim2500$ with high S/N for optically accessible Galactic O stars. In MONOS it is used for homogeneous classifications and, in favorable cases, for spatial deconvolution of visual components in long-slit data. For close visual pairs, the project also uses lucky spectroscopy (Apellániz et al., 2019).

LiLiMaRlin is the high-resolution spectroscopic backbone. It is a “library of libraries” assembled from the surveys OWN, IACOB, NoMaDS, and CAFÉ-BEANS, together with spectra from additional programs and public archives. The version described in the 2018 paper contains 18,077 spectra of 1,665 stars obtained with seven telescope/instrument combinations: HET/HRS, NOT/FIES, CAHA/CAFÉ, MPG-ESO/FEROS, OHP/ELODIE+SOPHIE, Mercator/HERMES, and Stella/SES (Apellániz et al., 2018).

The importance of LiLiMaRlin to MONOS is not mere accumulation but standardization. Construction begins from GOSC, after which spectra are gathered and subjected to a strict curation workflow. The authors explicitly emphasize source-identification control because archives may contain wrong or missing IDs and even coordinate errors of order $\sim 2^\prime$, especially dangerous in clusters. Source confusion in multiple systems is resolved using GOSC, the WDS catalog, lucky imaging with AstraLux Norte and Sur, and lucky spectroscopy. Spectra are discarded if they are noisy, lamp-contaminated, or have poor order stitching. Post-processing then checks header errors in coordinates and time, verifies wavelength calibration and heliocentric correction, rectifies the spectra, corrects for telluric lines, and converts each spectrum into a binary FITS file in a uniform format indexed in a MySQL archive with a GOSC-like interface (Apellániz et al., 2018).

In MONOS itself, LiLiMaRlin epochs with the best radial-velocity separation are smoothed to $R=2500$ and classified using the same MGB and OB2500 v3.0 standards as GOSSS. This cross-resolution harmonization allows classification consistency while preserving the multi-epoch radial-velocity leverage of the original high-resolution data (Apellániz et al., 2019).

For visual multiplicity, MONOS relies on AstraLux lucky imaging at Calar Alto and on the WDS catalog. AstraLux provides near-diffraction-limited images in the $z$, $i$, $zn$, and $\delta=-20^\circ$0 bands. It is used to measure separation $\delta=-20^\circ$1, position angle $\delta=-20^\circ$2, and magnitude difference $\delta=-20^\circ$3 through a custom PSF-fitting pipeline, with some older measurements recalibrated using Gaia DR2 astrometry (Apellániz et al., 2019).

4. First catalog and classification results

The first MONOS paper presents homogeneous spectral classifications for 92 O-type spectroscopic multiple systems and 10 optical companions (Apellániz et al., 2019). Within the 92-system sample, the authors report 17 O-type objects receiving their first GOSSS spectral classification and six B-type objects receiving their first GOSSS classifications. The paper also revises several previously known systems because new epochs reveal larger velocity separations or better spatial disentangling.

A major outcome is the re-identification of multiplicity state in systems formerly treated more simply. Some systems previously classified as SB1 are shown to be SB2 for the first time, and separate classifications are obtained for both components in cases such as BD $\delta=-20^\circ$416 4826, HD 168 112 AB, and $\delta=-20^\circ$5 Ori A. HD 170 097 A is identified as SB2E for the first time, and V747 Cep is confirmed as SB1E from LiLiMaRlin data (Apellániz et al., 2019).

The project also contributes new information on visual companions. In total, the paper reports nine new astrometric companions: one each around HD 190 967, ALS 15 133, Cyg OB2-A11, Cyg OB2-1, ALS 12 502, DN Cas, and MY Cam A, plus two around IU Aur AB (Apellániz et al., 2019). Several complex hierarchies are also clarified with AstraLux and Gaia-assisted analysis.

Two systems are especially prominent in the classification results. For $\delta=-20^\circ$6 Ori AaAbB, lucky spectroscopy obtained near the Aa,Ab periastron allows the first single-observation spectral classifications of Aa, Ab, and B. For $\delta=-20^\circ$7 Ori CaCb, MONOS adds the system to the class of Galactic Of?p stars, making it the sixth member of that class and classifying it as O7 f?p var (Apellániz et al., 2019).

The multiplicity statistics derived in the first paper underscore the project’s emphasis on hierarchy. The sample contains more triple- or higher-order systems than double systems. Using the MONOS multiplicity counts, about three quarters of the sample have only one massive nearby companion in the strict spectroscopic sense, but once wider and lower-mass companions are included only about two fifths are double systems, about one third are triple systems, and the remainder are higher-order (Apellániz et al., 2019). The paper therefore argues that massive stars prefer high-order multiplicity over simple binarity within this sample.

5. Orbit-recovery strategy and combined photometric-spectroscopic modeling

The first MONOS paper states that LiLiMaRlin spectra are being compared with published orbital solutions in the north, after which new orbits will be derived (Apellániz et al., 2018). This establishes the core MONOS workflow: first validate or challenge the literature orbit using a homogenized high-resolution archive, then derive improved solutions where necessary. The paper identifies LS III +46 11 and $\delta=-20^\circ$8 Ori AaAb as prototypes for this style of analysis, including phased radial-velocity curves for $\delta=-20^\circ$9 and $\delta>-20^\circ$0 constructed from LiLiMaRlin and GOSSS data (Apellániz et al., 2018).

By the time of the third MONOS paper, this workflow had expanded to incorporate space-based photometry. That study begins from 154 Galactic O-type stars in the ALS/Gaia cross-match with Gaia DR3 epoch photometry. The average Gaia time series contains about 50 epochs over 34 months, while TESS typically provides 250–300 points per star across multiple sectors. Periods were found for 70 systems, most of them short-period binaries with $\delta>-20^\circ$1 d, after which the analysis was narrowed to 10 key systems observable from the north and meeting MONOS criteria (Holgado et al., 6 Aug 2025).

Period determination is intentionally redundant. MONOS uses three independent methods: the Lomb–Scargle periodogram, phase dispersion minimization (PDM), and the String Minimization Method, described as a 2D generalization of PDM (Holgado et al., 6 Aug 2025). This redundancy is required because automated pipelines can produce half-period aliases or overlook additional signals.

The spectroscopic workflow is iterative and explicitly multi-stage. Preliminary radial velocities are measured manually in IRAF, usually from He I or He II lines; a first orbit is fit using gbart; that orbit is then used in UNWIND spectral disentangling; radial velocities are remeasured from the disentangled profiles; and fastwind synthetic spectra are used to refine systemic velocities $\delta>-20^\circ$2, especially because wind infilling can shift line centers in O stars. Disentangled components are then classified at $\delta>-20^\circ$3 by comparison with the GOSSS/ALS spectral standards using MGB (Holgado et al., 6 Aug 2025).

The final orbital and stellar parameters are derived with the legacy version of PHOEBE, using Gaia and TESS photometry together with radial-velocity curves, approximate masses and temperatures from spectral types, assumptions about limb darkening and gravity darkening, and, where needed, third light. The analysis also computes the fillout factor following Mochnacki & Doughty (1972) to diagnose Roche-lobe filling and contact degree (Holgado et al., 6 Aug 2025). In this phase of MONOS, the project becomes a full joint-inference pipeline for period, geometry, masses, radii, and evolutionary configuration.

6. New orbital solutions, physical interpretation, and methodological limits

The third MONOS paper presents eight previously unpublished orbits and two refined benchmark systems for 10 SB2E systems (Holgado et al., 6 Aug 2025). Several results are singled out as especially important. BD +61 487 is reported as the first known Oe+O spectroscopic binary. HD 169 727 is identified as a system of overcontact O-type supergiants in an eccentric orbit with evidence of past mass transfer. Across the sample, the derived solutions are consistent with spectral classifications and theory, including short periods ($\delta>-20^\circ$4 days), high mass ratios, and semi-detached or overcontact configurations (Holgado et al., 6 Aug 2025).

These systems reinforce a recurrent MONOS theme: the short-period regime is where tidal effects, synchronization, and mass transfer dominate massive-binary evolution. The prominence of contact and overcontact solutions among the shortest-period systems is therefore astrophysically consequential rather than merely taxonomic. In the specific case of HD 169 727, the mass and luminosity inversion are interpreted in the paper as evidence for advanced interaction and likely mass transfer (Holgado et al., 6 Aug 2025).

The project also identifies pitfalls in automated variability and multiplicity analysis. One system has a period exactly twice the half-period value found in Gaia-based automated catalogs, and another exhibits a hidden second eclipsing binary within a stronger short-period signal (Holgado et al., 6 Aug 2025). In the earlier catalog paper, the authors likewise cautioned that the sample is biased toward bright systems with known orbits and is not adequate for a full multiplicity census; some SB1? objects may eventually be removed if their radial-velocity variability is shown to be non-orbital (Apellániz et al., 2019). These cautions are central to interpreting MONOS results: the project prioritizes homogeneous characterization and orbit reliability over immediate statistical completeness.

The broader outlook is explicitly survey-oriented. The combined use of Gaia, TESS, and high-resolution spectroscopy is described as effective and scalable for future surveys of O-type binaries (Holgado et al., 6 Aug 2025). On the spectroscopic side, LiLiMaRlin is expected to continue growing through ongoing projects and archive mining, including future exploration of HARPS@ESO 3.6 m, UVES@VLT, ESPaDOnS@CFHT, and NARVAL@Bernard Lyot (Apellániz et al., 2018). This suggests an evolving MONOS architecture in which increasing time baseline, spectral coverage, and instrumental diversity are not auxiliary conveniences but primary enablers of orbit recovery and multiplicity disambiguation.

In sum, MONOS has developed from a northern catalog-and-classification effort into a coordinated program for hierarchical multiplicity analysis and orbit determination in Galactic O-type systems. Its distinctiveness lies in the combination of homogeneous classification standards, rigorously curated multi-epoch spectroscopy, explicit treatment of visual and higher-order multiplicity, and later integration of survey photometry for scalable orbital modeling (Apellániz et al., 2019, Holgado et al., 6 Aug 2025).