$δ$ Circini: A massive hierarchical triple system with an eclipsing binary

TL;DR

This work presents a comprehensive, multi-technique study of the massive hierarchical triple δ Circini, combining TESS photometry, high-resolution spectroscopy, archival data, and VLTI interferometry to derive a precise orbital architecture and fundamental parameters for all three stars. The outer orbit is quantified as $P_{\mathrm{outer}}=1603.24\pm0.19$ d with a coplanar configuration relative to the inner $P_{\mathrm{inner}}=3.90244719\pm0.00000052$ d eclipsing binary, yielding a total system mass of $M_{\mathrm{tot}}=53.04\pm0.29\ M_\odot$ and a distance of $d=809.9\pm1.8$ pc. Spectral disentangling and spectral fitting provide Teff and log g for all components, enabling robust evolutionary modelling with MESA that predicts coeval ages around $4\pm1$ Myr and a dynamically unstable mass-transfer phase in the inner binary leading to a merger into a $\sim36\ M_\odot$ Wolf–Rayet remnant, ultimately collapsing to a black hole. The precise orbital solution and age-dating, together with the system’s likely membership in the ASCC 79 subgroup, make δ Cir a key benchmark for testing massive-star formation and binary/evolutionary theory in hierarchical triples.

Abstract

$δ$ Circini is known to be a massive multiple system containing a 3.9 d inner eclipsing binary in a slightly elliptical orbit exhibiting slow apsidal motion and a distant tertiary with a probable period of 1644 d. All three components of the system are O- or B-type stars. We carried out a comprehensive study of the system, based on light curves from TESS and other instruments, a new series of echelle spectra, older spectra from the ESO archive, and several VLTI interferometric observations. Due to the large amount of different types of data covering both orbits in the system, we obtained a more precise value of the long orbital period ($1603.24\pm0.19$ d) and fully determined all other orbital parameters. Although both orbits are eccentric, their period ratio is large enough for the system to be dynamically stable. The inner and outer orbits are in the same plane, which means that no Kozai-Lidov mechanism is acting in the system. Assuming solar metallicity in our MESA models, we found ages of $(4.4\pm 0.1)$, $(4.7\pm 0.2)$, and $(3.8\pm1.3)$ Myr for the primary, the secondary, and the tertiary, respectively. Our evolutionary scenario predicts that the inner eclipsing binary will merge within approximately 1.7 Myr and eventually evolve into a black hole. The distance to the system, estimated from the angular size of the outer orbit is $(809.9 \pm 1.8)$ pc, which implies that $δ$ Cir might be located close to the centre of a stellar population ASCC 79, a subgroup of the young Circinus complex. With a total mass of $(53.04\pm0.29)$ M$_{\odot}$, $δ$ Cir can contribute a significant fraction of the total mass of the population.

Figures (10)

• Figure 1: Some of the normalised disentangled line profiles. Primary is always on the bottom, secondary in the middle and tertiary on the top. The continuum level of the secondary and the tertiary is shifted up by 0.2 and 0.4 respectively.
• Figure 2: Upper: RV curve of the outer orbit. Black x-signs with grey error bars correspond to the RVs of the inner eclipsing binary centre of mass, while the red x-signs with grey error bars (in some cases too small to be seen) are the RVs of the outer component. The systemic velocity, $v_\gamma$, has been subtracted from the RVs. Lower:$O-C$ residuals in the units of $\sigma$ for each individual point.
• Figure 3: The relative astrometric orbit of the outer component (dashed line) relative to the centre of mass of the inner binary (black plus sign). The black error ellipses (5$\sigma$) are our measurements (in some cases too small to be seen), and the black x-signs are the corresponding points on the computed orbit, which in each case fall within the error ellipses. The grey line is the line of nodes, the turquoise plus sign is the ascending node, the magenta plus sign is the periastron, and the arrow shows the direction of the outer orbit.
• Figure 4: Upper: RV curve of the inner orbit. Black x-signs denote the RVs of the primary, and the red x-signs denote the RVs of the secondary. The error bars are too small to be visible. The systemic velocity, $v_\gamma$, has been subtracted from the RVs. Lower:$O-C$ residuals in the units of $\sigma$ for each individual point.
• Figure 5: Final phase plot of the TESS light curves from sector 12 and sector 65. Left: light curve from TESS sector 12 and below its residuals in the units of $\sigma$. Right: light curve and its residuals from TESS sector 65.