KQ Pup: A Hierarchical Triple System

• KQ Pup is defined as a southern VV Cephei-type object reinterpreted as a hierarchical triple with a red supergiant and an inner eclipsing B-type binary.
• Multi-wavelength observations, including TESS photometry, spectroscopy, and VLTI interferometry, reveal a 26-year outer orbit and a 17.26-day inner period with wind-fed interactions.
• The system offers a unique testbed for studying massive multiple evolution, wind Roche-lobe overflow, and potential progenitor channels for interacting supernovae.

Searching arXiv for papers on KQ Pup and related object identifications. First search: exact query for KQ Puppis / KQ Pup. Second search: related names V351 Pup and nova context to capture designation ambiguity. KQ Pup, or KQ Puppis, most commonly denotes a southern VV Cephei-type red-supergiant binary now understood to be a hierarchical triple system of the form $\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$, with a cool red supergiant outer tertiary and a hot inner eclipsing pair (Jadlovský et al., 29 Sep 2025). In the recent literature, it is characterized by a long outer orbit of about $26$ yr and an inner eclipsing period of $17.2596\ \mathrm{d}$, with spectroscopic, photometric, ultraviolet, and interferometric evidence indicating ongoing wind-fed interaction rather than direct Roche-lobe overflow. The designation is not entirely unambiguous, however: in nova work, KQ Pup is also used in the context of Nova Puppis 1991, V351 Pup, a physically unrelated classical nova (Wendeln et al., 2017).

1. Identification and nomenclature

KQ Puppis is identified in the red-supergiant-binary literature as KQ Pup, with cross-identifications HD 60414/60415, and was historically understood as a wide interacting system composed of an M2 Iab red supergiant, KQ Pup A, and a hot early-B companion, KQ Pup B (Jadlovský et al., 29 Sep 2025). The principal revision introduced by later work is that the “B” component is not single but an eclipsing close binary, so the system architecture is $\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$.

A recurrent source of confusion is that the designation KQ Pup also appears in the classical-nova literature, where it refers to Nova Puppis 1991, V351 Pup. That object is modeled as a classical nova with radio emission dominated by thermal free-free radiation from a spherical, clumpy, expanding shell, with an expansion-parallax distance of $5.0\pm1.5\ \mathrm{kpc}$ and ejecta mass $\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$ (Wendeln et al., 2017). This is a distinct astrophysical source from the red-supergiant system.

Additional confusion can arise from nearby or similarly named Puppis variables that are explicitly unrelated in the cited studies. A UBV survey centered on AQ Puppis and V620 Puppis does not mention KQ Pup anywhere in its title, abstract, body text, tables, or figure captions, and it is explicitly about AQ Pup and V620 Pup only (Turner et al., 2012). Likewise, PU Pup is treated as a separate object, identified as HR 2944 = HD 61429 = HIP 37173 = m Pup, and analyzed as a young interacting binary rather than as KQ Pup (Erdem et al., 2021).

2. Hierarchical-triple architecture

The current picture of KQ Pup is that it is the first clearly established hierarchical triple among Galactic red-supergiant binaries (Jadlovský et al., 29 Sep 2025). The outer orbit is the long-known $\sim 26$ yr A+B orbit, while the inner subsystem is a short-period eclipsing pair revealed by TESS. The resulting hierarchy is

$\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb}),$

with the red supergiant as the outer tertiary and Ba+Bb forming the inner binary.

The decisive new discovery was the detection of eclipses in TESS light curves with a period

$P_{\rm Ba+Bb}=17.2596~{\rm d},$

which is associated with the hot B companion. The eclipse pair is present in all four TESS sectors examined, and the secondary eclipse is displaced from phase $0.5$, implying an eccentric inner orbit. This directly overturns the earlier simplified interpretation of KQ Pup as a two-body RSG+B system.

The outer orbit remains a wide, eccentric interacting orbit with a revised period

$26$0

The global fit gives $26$1, $26$2, $26$3, $26$4, $26$5, $26$6, $26$7, $26$8, $26$9, and $17.2596\ \mathrm{d}$0, where $17.2596\ \mathrm{d}$1 (Jadlovský et al., 29 Sep 2025). The VLTI-only astrometric solution is similar, with $17.2596\ \mathrm{d}$2, $17.2596\ \mathrm{d}$3, and $17.2596\ \mathrm{d}$4, indicating that the interferometric orbit is robust even before inclusion of the full radial-velocity data set.

3. Observational basis and orbit determination

The system constraints derive from a combination of archival photometry, UV spectroscopy, newly taken optical spectra, interferometric data, and TESS photometry (Jadlovský et al., 29 Sep 2025). TESS establishes the inner eclipsing periodicity, while optical and ultraviolet spectroscopy trace the long-orbit behavior of the red supergiant and the hot component. A key observational advance is that VLTI-GRAVITY detected the hydrogen Br$17.2596\ \mathrm{d}$5 line at $17.2596\ \mathrm{d}$6, enabling astrometric tracking of the hot subsystem.

In the near-infrared, the continuum is dominated by the red supergiant, so the binary is not directly visible in continuum visibilities. Br$17.2596\ \mathrm{d}$7, however, produces a strong differential-phase signature. Using PMOIRED, the red supergiant is modeled as a uniform disk and the Br$17.2596\ \mathrm{d}$8-emitting hot component as an unresolved point source. The GRAVITY data were obtained in 2024–2025 at baselines of $17.2596\ \mathrm{d}$9 in high-resolution $\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$0-band mode with $\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$1. The fitted RSG angular diameter is about $\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$2, and the astrometrically recovered Br$\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$3 positions shift coherently by a few mas across three epochs, consistent with orbital motion of the inner pair after periastron passage.

The most stable solution gives a separation of about $\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$4 between the Br$\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$5 photocenter and the system barycenter. The measured Br$\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$6 line-flux contribution is $\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$7, while the hot component’s continuum contribution in the $\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$8 band is essentially negligible. This is described as the first use of Br$\mathrm{A}+(\mathrm{Ba}+\mathrm{Bb})$9 for relative astrometry in a red-supergiant binary (Jadlovský et al., 29 Sep 2025).

Spectroscopically, the RSG radial velocities were measured from STELLA spectra by cross-correlation against MARCS models and agree with the long-orbit solution. For the hot component, Balmer-line emission centroids in H$5.0\pm1.5\ \mathrm{kpc}$0 and H$5.0\pm1.5\ \mathrm{kpc}$1, archival IUE far-UV and near-UV spectra, and phase-dependent line-profile behavior are used as diagnostics. The Balmer emission is strongest near periastron and weaker toward apastron; the $5.0\pm1.5\ \mathrm{kpc}$2 ratio reverses near periastron, and the higher Balmer lines can nearly disappear. These observables are central to the interaction interpretation.

4. Component properties

The combined VLTI+RV orbital solution gives dynamical masses

$5.0\pm1.5\ \mathrm{kpc}$3

so the outer red supergiant is the less massive tertiary and the hot inner pair is more massive in total (Jadlovský et al., 29 Sep 2025). The paper’s abstract summarizes this as a mass of $5.0\pm1.5\ \mathrm{kpc}$4 for the RSG and $5.0\pm1.5\ \mathrm{kpc}$5 for the sum of the hot components.

For KQ Pup A, the fitted angular diameter implies a mean radius

$5.0\pm1.5\ \mathrm{kpc}$6

Independent asteroseismic estimates based on $5.0\pm1.5\ \mathrm{kpc}$7 and luminosity give a mass range $5.0\pm1.5\ \mathrm{kpc}$8, with $5.0\pm1.5\ \mathrm{kpc}$9 and $\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$0. The inferred luminosity is

$\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$1

These estimates are stated to agree well with the dynamical mass.

For the hot outer component Ba, PoWR atmosphere modeling of the IUE spectra yields approximately

$\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$2

This is consistent with a fast-rotating early-B main-sequence star near the terminal-age main sequence. For Bb, direct spectroscopy is not yet available. The minimum mass is constrained as

$\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$3

with an allowed range from MIST-based constraints of roughly

$\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$4

The favored interpretation is a low-mass main-sequence star, most likely an early A/F-type main-sequence object rather than a stripped He star, because no strong He II or stripped-star He I signatures are seen in the UV/optical spectra (Jadlovský et al., 29 Sep 2025).

5. Interaction physics and circumstellar environment

A central physical conclusion is that KQ Pup is not well described by direct Roche-lobe overflow from the red supergiant to the hot companion system (Jadlovský et al., 29 Sep 2025). The outer orbit is too wide for the red supergiant to fill its Roche lobe even at periastron, given the derived radius of the RSG and the periastron separation. Instead, the interaction is interpreted as Wind Roche Lobe Overflow, in which the dense RSG wind is gravitationally focused toward the inner binary.

The evidence is multi-pronged. First, Balmer-line emission strengthens near periastron and weakens toward apastron, consistent with enhanced mass capture when the hot subsystem moves through denser regions of the red-supergiant wind. Second, Br$\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$5 traces the hot component and shows photocenter shifts consistent with the inner pair’s motion. Third, UV $\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$6 emission, P Cygni profiles, and the Na I doublet display orbitally dependent circumstellar behavior, including a stable narrow Na I component plus an orbitally moving broader component. Fourth, the Balmer $\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$7 ratio varies on both the 17 d and $\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$8 d timescales, suggesting that the hot inner pair perturbs the accretion flow while the RSG’s own pulsation modulates the supply rate.

The paper estimates that accretion onto the inner binary is only $\log_{10}(M_{ej})=-5.2\pm0.7\ M_\odot$9 of the RSG’s mass-loss rate, so the accreted material does not significantly alter the mass budget of the hot components. It can, however, power a disk-like or envelope-like structure and account for the observed emission-line phenomenology. Interferometry also indicates that KQ Pup A has an extended molecular atmosphere, with the CO bands implying an atmospheric extension of about $\sim 26$0 above the photosphere in the near-infrared. This extended envelope is presented as making wind-RLOF plausible because the inner binary is effectively embedded in a dense, structured outflow for a significant fraction of the orbit.

6. Evolutionary interpretation and broader significance

The evolutionary interpretation is that KQ Pup is compatible with a coeval hierarchical triple with an age of roughly $\sim 26$1 (Jadlovský et al., 29 Sep 2025). In this picture, KQ Pup A is a $\sim 26$2 red supergiant approaching core collapse, Ba is an early-B star near TAMS, and Bb is a lower-mass main-sequence companion. The inner orbit is short and eccentric; from the TESS eclipse morphology and simple orbital geometry, the minimum eccentricity is estimated as

$\sim 26$3

and PHOEBE toy models suggest partial or grazing eclipses with a non-central inclination slightly below $\sim 26$4. The radial-velocity scatter of Ba indicates that the inner RV semi-amplitude is probably $\sim 26$5, favoring a relatively low-mass Bb unless the viewing geometry is unfavorable.

The absence of strong spectroscopic evidence for a stripped companion or for a major past mass-transfer episode leads to the conclusion that the system has so far evolved mainly through wind accretion from the RSG rather than through substantial inner-binary interaction. A plausible implication is that KQ Pup provides an observationally constrained example of a massive multiple system in which a red-supergiant wind feeds a compact inner binary without requiring direct Roche-lobe contact.

The broader significance extends beyond the individual system. TESS light curves are reported to show that about $\sim 26$6 of known Galactic binary RSGs may be eclipsing hierarchical triple systems, suggesting that a substantial fraction of other binary RSGs could also be triples (Jadlovský et al., 29 Sep 2025). This, in turn, bears on the census of massive multiples, on binary-interaction channels in late-stage massive-star evolution, and on possible descendants or end products. The paper explicitly suggests that such systems could be progenitors of interacting supernova environments or, after the RSG explodes, of Be/X-ray or Be+sdOB-like binaries. In that sense, KQ Pup functions not merely as an individual VV Cephei-type object but as a reference system for the study of hierarchical multiplicity, wind-fed interaction, and red-supergiant binary evolution.