The W-subtype active contact binary PZ UMa with a possible more massive tertiary component

Abstract

Two sets of multiple-color (B, V, R_C, I_C) light curves of PZ UMa were observed independently with the 2.4 m telescope at the Thai National Observatory and the 1 m telescope at Yunnan Observatories. The light curves were analyzed with the Wilson–Devinney program and the two sets of light curves produced consistent results, which shows that PZ UMa is a W-subtype contact binary with an extreme mass ratio (M₁/M₂ = 0.18). The basic physical parameters of PZ UMa were determined to be M₂ = 0.77(2) M_⊙, M₁ = 0.14(1) M_⊙, R₂ = 0.92(1) R_⊙, R₁ = 0.43(1) R_⊙, L₂ = 0.46(2) L_⊙, and L₁ = 0.15(3) L_⊙. The orbital period analysis of PZ UMa revealed a 13.22 yr periodicity, which implies that there may be a tertiary component orbiting around the binary system. The mass and orbital radius of the tertiary component were calculated to be M₃ = 0.88 M_⊙ and a₃ = 3.67 au, if the orbit was coplanar with the central binary system. It is interesting that the minimum mass of the tertiary was calculated to be M_3min = 0.84 M_⊙, which means the tertiary component is even larger than the primary star and the secondary one of PZ UMa. PZ UMa is a late-type contact binary with stellar activity. The O’Connell effect appeared on its light curves when it was observed in 2016 April. However, the O’Connell effect reversed when the target was observed again in 2016 December. The changes of the O’Connell effect on such a short time-scale strongly support the occurrence of rapidly changing magnetic activity in this W UMa binary.

1 Introduction

Photometric observations of eclipsing binaries have a long history. In 1955, Kopal (1955) divided close binaries into three categories, which were detached binaries, semi-detached binaries and contact binaries, depending on whether the component stars had filled their critical Roche lobes. They usually have EA, EB, or EW type light curves correspondingly. As for contact binaries, both of the component stars have filled their critical Roche lobes. Contact binary systems share a common envelope and mass transfer occurs between the two components. This peculiar structure makes the evolution of contact binaries very complex and interesting. In 1970, Binnendijk (1970) found that there were two subtypes of contact binaries, which were A-subtype and W-subtype. For the A-subtype system, the more massive component star has a higher surface temperature, while the less-massive star is the hotter one in a W-subtype system. It is really striking to researchers that a less-massive star can be the hotter one in a binary system. The formation theory of W-subtype contact binaries is still unclear. Yildiz and Doğan (2013) pointed out that W-subtype contact binaries have initial masses larger than 1.8 M_⊙, and subsequent research by Zhou et al. (2016c) also supported their conclusions. More photometric/spectroscopic observations and analyses of W-subtype contact binaries are needed to further characterize them.

The light variations of PZ UMa (other names: GSC 03428−00212, NSVS 4861481) were first reported by Nicholson et al. (2005) during their search for new variable stars in the public data of the Northern Sky Variability Survey (Woźniak et al. 2004). It was identified as an EW type contact binary with an orbital period of 0.2627 d. Photometric observations of PZ UMa were carried out using the optical telescopes at the Thai National Observatory and Yunnan Observatories. Two sets of multiple-color light curves were obtained to study its light variations and period changes. The effective temperature given in the Gaia Data Release 2 (Gaia Collaboration 2016, 2018) is T = 5430(260) K, which means PZ UMa is a G7 type contact binary (Cox 2000).

2 Observations and data reduction

Observations of PZ UMa were carried out on 2016 April 19 and 20 with the 2.4 m telescope at Thai National Observatory, National Astronomical Research Institute of Thailand (hereafter TNO 2.4 m). The TNO 2.4 m was equipped with a 4 K × 4 K CCD camera, and its field of view was 16 square arcmin (Soonthornthum 2018). Johnson–Cousins filters were used during the observations.

PZ UMa was observed again on 2016 December 10 with the 1 m telescope at the Yunnan Observatories, Chinese Academy of Sciences (hereafter YNOs 1 m). The YNOs 1 m was equipped with an Andor DW436 2 K × 2 K CCD camera, and the field of view was 7.3 square arcmin. The same filters were used with the TNO 2.4 m. On 2018 November 29 PZ UMa was observed with the 60 cm telescope at Yunnan Observatories (hereafter YNOs 60 cm) in order to determine the mid eclipse time and the nature of the eclipses.

The Image Reduction and Analysis Facility (IRAF) (Tody 1986) was used to reduce the observational images. UCAC4 700-051550 and UCAC4 700-051554 in the same field of view were selected as the comparison (C) and check stars (Ch) respectively, in order to determine the light variations of PZ UMa. Their coordinates and V-band magnitudes are listed in table 1. The image from the Aladin Sky Atlas (Bonnarel et al. 2000) that we used to identify the targets is given in figure 1, in which PZ UMa (V), UCAC4 700-051550 (C), and UCAC4 700-051554 (Ch) are indicated. We performed parabola fits on the observed light minima in B, V, R_C, and I_C bands and obtained the averaged mid-eclipse times. In all, two primary (p) and three secondary (s) minima were detected, and these are listed in table 2. The phased light curves of PZ UMa are displayed in figure 2.

3 Investigation of the period variation

The CCD image enables very high precision measurements of the mid-eclipse times (the errors are less than 17 s in our observations), and it is a very powerful tool for detecting the orbital variations of contact binaries. According to recent research, the period of a contact binary is not constant in many cases (Wang et al. 2017; Zhang et al. 2018; Zhao et al. 2018b; Hu et al. 2018; Li et al. 2018). The mass transfer between the two component stars and the presence of a tertiary object in orbit about the binary system, as well as magnetic braking, will cause period variations, which will eventually affect the evolution of the binary systems. To understand the interior/exterior dynamic interactions of PZ UMa, the O − C method was used, and all available times of light minima are listed in the first column of table 3. All light minima were observed with the CCD method.

As shown in figure 3, the dots in the upper panel are the O − C values based on the initial ephemeris (Min. I = 2451337.7141 + 0.262674 × E), and the red line shows the calculated ones based on equation 1. The dashed line represents corrections to the initial linear ephemeris. The sine-like curve seems more intuitive when the corrections are removed, which are shown in the middle part of figure 3. The fitted residuals are shown in the lowest part. The sine-like part in the O − C curve reveals that the period of PZ UMa contains a periodic variation, with a period of 13.22 yr and an amplitude of 0.0198 d.

4 Modeling of light curves

To determine the physical parameters of PZ UMa, the two sets of light curves derived from observations made with the TNO 2.4 m and YNOs 1 m were modeled with the Wilson–Devinney program (Van Hamme & Wilson 2007; Wilson & Van Hamme 2014). The effective temperature of the primary star (when eclipsed at primary minima) was fixed at T₁ = 5430 K. The well known q-search method was used to search for a proper initial mass ratio (q = M₂/M₁). The q-search diagram is displayed in figure 4. It is obvious that PZ UMa is a W-subtype contact binary system. Then, the mass ratio is set to be a free parameter, the same as the orbital inclination (i), the dimensionless surface potential of the two stars (Ω₁ = Ω₂ for contact mode), the effective temperature of the secondary star (T₂), the monochromatic luminosity of the primary star (L₁), and the latitude (θ) and longitude (ψ) of the spots. As a third light (l₃) has been detected in many close binaries (Qian et al. 2008a, 2013; Zhou et al. 2015, 2016d), we try to set it as a free parameter. However, we can not get a converged solution with a positive parameter value of third light while running the Wilson–Devinney program, which may imply that there is no third light in PZ UMa. First, we run the light curves in different bands separately, and the solutions are listed in columns 2 to 5 in table 4 and table 5. Then, the light curves are fitted simultaneously to get the final solutions of PZ UMa, which are listed in the last column of table 4 and table 5. The errors from columns 2 to 5 are used to calculated the errors in column 6. The theoretical light curves are plotted in figure 5. The geometrical configurations at phase = 0, phase = 0.25, phase = 0.5, and phase = 0.75 are plotted in figure 6. The positions of spots are also marked. It is evident that PZ UMa is a totally eclipsed binary system, and the smaller star is totally eclipsed at the primary minimum (phase = 0).

5 Discussion and conclusion

The photometric solutions show that PZ UMa is a W-subtype contact binary with an extreme mass ratio (1/q = 0.18). Its totally eclipsed nature means that the photometric solutions are very reliable (Zhou et al. 2017; Zhao et al. 2018c). Normally, W UMa type contact binaries share a surface and the effective temperatures of the component stars are almost equal. As for PZ UMa, the temperature difference between the primary star and the secondary one is as large as several hundred Kelvin. However, its fill-out factor (f) is about 40%. This may indicate that energy transfer rate in PZ UMa is less efficient than in other contact binaries and PZ UMa may have some similarities with the B-type systems or poor thermal contact binaries (Csizmadia & Klagyivik 2004). PZ UMa may evolve into a extreme mass ratio overcontact binary system (Qian et al. 2008b; Li et al. 2017; Sarotsakulchai et al. 2018), and merge eventually as a new luminous red nova phenomenon (Hayashi et al. 1994; Soker & Tylenda 2003; Tylenda et al. 2011, 2013; Zhu et al. 2016).

The effective temperature of the secondary is 4972(18) K, which corresponds to a K1 spectral-type main-sequence star. Therefore, the mass of the secondary star is estimated to be M₂ = 0.77(2) M_⊙ (Cox 2000). The mass of the primary star is calculated to be M₁ = 0.14(1) M_⊙ while a photometric mass ratio of q = 5.62(16) is considered. Based on Kepler’s third law, the orbital semi-major axis is determined as a = 1.67(2) R_⊙. The radii and luminosities of the two stars are listed as follows: R₂ = 0.92(1) R_⊙, R₁ = 0.43(1) R_⊙, L₂ = 0.46(2) L_⊙, and L₁ = 0.15(3) L_⊙. However, radial velocity curves are needed to determine the absolute parameters with much higher credibility.

The O − C curve reveals a periodic variation with a period of 13.22 yr and an amplitude of 0.0198 d, which may be explained by the light travel time effect caused by a tertiary component. Assuming that the tertiary component was coplanar with the orbital plane of PZ UMa, the distance from the binary to the common mass center of the triple system was a₁₂ = 3.43 au and the mass function was f(m) = 0.23 M_⊙. The mass of the tertiary component was calculated to be M₃ = 0.88 M_⊙, and its orbital radius was a₃ = 3.67 au. It is reported that many close binaries are actually triple systems (Qian et al. 2007; Zhu et al. 2011; Li et al. 2014), or even quadruple systems (Zhu et al. 2013; Li et al. 2015; Zhao et al. 2018a). What is more, extraneous eclipses are observed in some eclipsing multiple stars (Zhang et al. 2017, 2018). Naoz and Fabrycky (2014) claimed that the “eccentric Kozai–Lidov” mechanism would result in a tight inner binary in a multiple star system. The more interesting thing is that the minimum mass of the tertiary is calculated to be M_3min = 0.84 M_⊙ with the largest orbital radius a_3max = 3.72 au. This means that the mass of the tertiary component is even larger than the two component stars of PZ UMa. However, we have not detected third light (l₃) through photometric analysis, which implies that the tertiary may be a compact object such as a white dwarf or a neutron star. Qian et al. (2017) and Qian et al. (2018) proposed that a compact tertiary component may even contaminate the inner binary during the early evolutionary stage of a multiple star system, which can explain the formation of a contact binary with a very high metallicity.

The O’Connell effect is commonly reported in W UMa type binaries (Zhou et al. 2016a, 2016b, 2018). It is widely accepted that the formation of the O’Connell effect is closely related to stellar magnetic activities and circumbinary dust (Liu & Yang 2003). Research on the O’Connell effect will help us to understand the magnetic activities of binary stars and the characteristics of circumbinary dust. Both of our two sets of light curves show the O’Connell effect. The light variations (Max I–Max II) between the two maxima are recorded and listed in table 6, where Max I refers to the light maximum at phase 0.25 following the primary light minimum and Max II is the maximum at phase 0.75. The O’Connell effect observed by the TNO 2.4 m was a positive one. However, it changed to be a negative one when it was observed with the YNOs 1 m just 10 months later. It is quite striking that the O’Connell effect changes over such a short time-scale. Considering that both of the components are late-type stars, we believe that magnetic activity is the most possible explanation in this case. Spot modelling was used to fit the observed light curves, and the positions of spots are marked in figure 6. It is easy to see that the migration of the spot from the “South pole” to “North pole” accounts for the changes in the O’Connell effect. The O’Connell effect appears with late-type contact binaries quite frequently. Late-type contact binaries with the O’Connell effect are ideal targets for the study of stellar magnetic activities. They may help us to understand the formation, migration, and disappearance of spots.

Acknowledgements

We thank Mr. Wayne Orchiston for improving the manuscript. This research was supported by the Chinese Natural Science Foundation (Grant No. 11703080 and 11703082) and the Yunnan Natural Science Foundation (No. 2018FB006). It was part of the research activities at the National Astronomical Research Institute of Thailand (Public Organization). This work has made use of data from the European Space Agency (ESA) mission Gaia (https://www.cosmos.esa.int/gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement. This research has made use of “Aladin Sky Atlas” developed at CDS, Strasbourg Observatory, France.

Footnotes

References