The OmegaWhite Survey for Short-period Variable Stars. V. Discovery of an Ultracompact Hot Subdwarf Binary with a Compact Companion in a 44-minute Orbit

Hot subdwarfs (sdOs/sdBs) are low-mass He-stars with very thin hydrogen envelopes. They have effective temperatures similar to O and B stars but are several orders of magnitude less luminous due to their small size (Heber 1986, 2009, 2016). Maxted et al. (2001) and Napiwotzki et al. (2004) showed that $\gt 50 \%$ of sdB/sdO stars are in compact binaries with orbital periods ${P}_{\mathrm{orb}}\lt 10$ days. Formation and orbital shrinkage through a common envelope phase are the only way to form such compact binaries (Han et al. 2002, 2003; Nelemans 2010).

It has been shown that hot subdwarfs in compact sdB/sdO + white dwarf (WD) binaries with ${P}_{\mathrm{orb}}\lesssim 2$ hr on the exit of the common envelope phase still burn helium when the sdB/sdO fills its Roche Lobe (RL). Due to the emission of gravitational waves, the binary is predicted to shrink until the hot subdwarf star fills its RL at an orbital period of 16–50 minutes and starts mass-transfer. He-rich material is then transferred to the WD companion with typical mass-transfer rates of $\dot{M}\approx {10}^{-8}\,$ ${M}_{\odot }$ yr⁻¹ (e.g., Savonije et al. 1986; Tutukov & Fedorova 1989; Tutukov & Yungelson 1990; Iben & Tutukov 1991; Yungelson 2008; Piersanti et al. 2014; Brooks et al. 2015).

Most of the known compact sdB/sdO binaries reside in systems with orbital periods $\geqslant 0.1$ days where the sdB/sdO will have turned into a carbon/oxygen WD when both components come into contact (Kupfer et al. 2015). So far, only two hot subdwarf binaries with a white dwarf companion and an orbital period ${P}_{\mathrm{orb}}\lt 90\,\mathrm{minutes}$ are known (Vennes et al. 2012; Geier et al. 2013; Kupfer et al. 2017).

Just recently, Kupfer et al. (2017) discovered the ultracompact sdB+WD binary, PTF1 J082340.04+081936.5, with ${P}_{\mathrm{orb}}=87\,\mathrm{minutes}$. The object is close to the limit to start future accretion, while the sdB is still burning helium. If the sdB still burns helium when the system comes into contact, helium-rich material will be accreted onto the WD companion and the most likely outcome is a helium-accreting AM CVn-type system, because the companion is by then a low-mass white dwarf (${M}_{\mathrm{WD}}={0.46}_{-0.09}^{+0.12}$ ${M}_{\odot }$).

The most compact sdB binary with a WD companion is CD-30°11223 (${P}_{\mathrm{orb}}=70.5\,\mathrm{minutes}$, Vennes et al. 2012; Geier et al. 2013). The sdB will fill its RL in $\approx 40$ million years, well within the He-burning lifetime of the sdB star. The WD companion is massive (${M}_{\mathrm{WD}}\approx 0.75$ ${M}_{\odot }$) and after accreting 0.1 ${M}_{\odot }$, He-burning is predicted to be ignited unstably in the accreted helium layer on the surface of the white dwarf (Brooks et al. 2015; Bauer et al. 2017). This could trigger the ignition of carbon in the core, which might disrupt the WD even when the mass is significantly below the Chandrasekhar mass, a so-called double-detonation supernova type Ia (e.g., Livne 1990; Livne & Arnett 1995; Fink et al. 2010; Woosley & Kasen 2011; Wang et al. 2013; Shen & Bildsten 2014). The hot subdwarf will become unbound and ejected with the orbital velocity, which can be up to $\approx 1000$ $\mathrm{km}\,{{\rm{s}}}^{-1}$. The hypervelocity sdO star US708 (Hirsch et al. 2005) has been proposed as a candidate for such a donor remnant (Geier et al. 2013, 2015).

Bildsten et al. (2007) showed that unstable He shell-burning can detonate the He shell without disrupting the WD, which can be observed as a faint and fast Ia supernova. The detonation will increase the orbit and the binary system will lose contact. Gravitational wave radiation will decrease the orbit again and bring both objects back into contact, which can trigger several subsequent weaker flashes (Brooks et al. 2015).

The OmegaWhite survey is a high-cadence synoptic survey of the southern Galactic Plane and Galactic Bulge, the main aim of which is to identify ultracompact binaries (Macfarlane et al. 2015; Toma et al. 2016; Macfarlane et al. 2017a, 2017b). OmegaWhite observations are obtained using the VST telescope at the European Southern Observatory's (ESO) Paranal site in Chile. Each field is imaged an average of 38 times over a 2 hr period, with 39 s integrations. Light curves are extracted from image differencing and a Lomb–Scargle and AoV analysis is performed on all extracted light curves to identify short-period variables. OW J074106.0–294811.0 (OW J0741 hereafter) was discovered as a faint ($g\,=$ 20.0) blue ($u-g=-1.23$ mag, $g-r=0.18$ mag) source, photometrically variable on a period of 22.6 minutes in the OmegaWhite survey (Macfarlane et al. 2015). It was noted in Toma et al. (2016) that further observations of OW J0741 showed that it was a binary with a 44-minute orbital period. Here, we report the discovery of OW J0741 as an ultracompact sdOB + WD binary with an orbital period of ${P}_{\mathrm{orb}}=44.66\,\mathrm{minutes}$, making it the most compact hot subdwarf binary known today.

The paper is structured as follows. Section 2 describes the observations used in the analysis. The orbital and binary parameter, as well as the atmospheric parameters, are described in Sections 3 and 4. The light curve analysis and the system parameters are presented in Sections 5 and 6. Section 7 discusses the structure and evolutionary history of the system using MESA. A summary and conclusions are given in Section 8.

2.1. Optical Photometry

Optical photometric data were obtained using the South African Astronomical Observatory (SAAO) 1.9 m Telescope with the Sutherland High Speed Optical Camera (SHOC), Keck/LRIS, Gemini South/GMOS as well as the 3.5 m New Technology Telescope (NTT) with ULTRACAM and the Southern African Large Telescope (SALT) with SALTICAM.

Photometry from the SAAO was obtained using the 1.9 m telescope and SHOC (Coppejans et al. 2013), which uses an Andor E2V CCD and works in frame-transfer mode (with 1024 × 1024 active pixels) so that the readout time is effectively zero. It has been designed to run at up to 20 frames s⁻¹ with high-accuracy timing for each frame, but because this object is so faint, typical exposure times were 30–40 s (see Table 1). No filter was used in order to maximize the count rate, and the SHOC data were reduced using an in-house SAAO aperture photometry pipeline based on IRAF routines. On the 1.9 m telescope, SHOC has a field-of-view of 28 × 28, so several brighter stars were available for differential photometry. We used three photometrically stable field stars for photometric calibration. The observations were conducted under partly cloudy conditions, with an average seeing of 12.

Table 1.  Summary of the Observations of OW J0741

Date	UT	Tele./Inst.	${N}_{\exp }$	Exp. Time (s)	Coverage (Å)/Filter
Photometry
2016 Apr 14	17:47–20:29	SAAO/SHOC	325	30	clear
2016 Apr 15	17:13–19:22	SAAO/SHOC	257	30	clear
2016 Apr 18	17:04–19:29	SAAO/SHOC	216	40	clear
2016 Apr 19	18:28–19:58	SAAO/SHOC	136	40	clear
2016 Nov 03	14:47–15:39	Keck/LRIS	56	20	g'
2016 Nov 28	05:30–06:32	Gemini/GMOS	95	23	g'
2016 Dec 05	05:37–06:39	Gemini/GMOS	95	23	g'
2016 Dec 06	05:32–06:35	Gemini/GMOS	95	23	g'
2016 Dec 07	05:32–07:00	NTT/ULTRACAM	207	20/40	u'
2016 Dec 07	05:32–07:00	NTT/ULTRACAM	412	10/20	g' r'
2016 Dec 07	21:57–22:45	SALT/Salticam	1470	2	clear
Spectroscopy
2016 Sep 20	08:37–09:27	VLT/FORS2	9	300	3500–6300
2016 Nov 03	06:38–08:07	Gemini/GMOS	20	240	3500–6700
2016 Nov 03	13:22–14:08	Keck/LRIS	10	240	3400–5600
2016 Nov 04	05:39–06:23	Gemini/GMOS	10	240	3500–6700
2016 Nov 28	12:00–13:38	Keck/LRIS	20	240	3400–5600
2016 Dec 29	10:00–10:41	Keck/LRIS	9	240	3400–5600
2017 Jan 26	09:15–10:01	Keck/LRIS	10	240	3400–5600
Swift
2016 Oct 24	04:06–06:02	XRT	1	2944	0.2–10 keV
2016 Oct 24	04:06–06:02	UVOT	1	2922	1600–2260 (UVW2)

Download table as:  ASCIITypeset image

Observations of OW J0741 were made on 2016 December 7 using the high-speed photometer SALTICAM (see O'Donoghue et al. 2006) mounted on SALT at the Sutherland Observatory in South Africa. Observations were made with no filter and an exposure time of 2 s and the data set consists of 1470 images. The SALTICAM frame-transfer mode allows for exposures with 200 ms readout time.

Photometric follow-up observations were obtained with Keck and Gemini South using LRIS (Oke et al. 1995) and GMOS (Hook et al. 2004) in imaging mode, respectively. Both setups used the g' filter and a 20 s exposure time. For GMOS we used a 2 × 2 binning and a 1 × 1 arcmin window to reduce the readout time to ≈10 s. Gemini observed OW J0741 at an airmass $\approx 1.05$ with an average seeing of 07. The Keck light curve was obtained at an airmass of $\approx 1.5$ and an average seeing of 11. The readout time of the CCD was 42 s for each exposure.

High-cadence observations were obtained using ULTRACAM (Dhillon et al. 2007) on the ESO 3.58 m NTT on 2016 December 6. ULTRACAM is a high-speed photometer that observes in three color bands simultaneously and uses frame-transfer CCDs to reduce dead time between exposures to almost zero. Sloan filters u', g', and r' were used. The exposure time was in the beginning 10 s and later 20 s in g' and r', and 20 and 40 s in u' to compensate for the lower throughput of that band. These observations covered two full orbits of the system under clear conditions, with an average seeing of 1''. Each image was bias- and dark-subtracted, and divided through by a twilight flat-field.

The photometric data from Keck, Gemini, and ULTRACAM were reduced using the ULTRACAM pipeline. The count rate of OW J0741 was extracted using differential aperture photometry, using the same comparison star for Keck, Gemini, and ULTRACAM to remove atmospheric transparency effects.

An aperture of 1.7 × the FWHM of the star was used to extract the photometry in the Keck and Gemini data. For the ULTRACAM data, the aperture width was reduced to 1.5 × the FWHM of the star, in order to avoid contamination from nearby stars due to the lower resolving power of the NTT.

2.2. Spectroscopy

2.3. X-Ray and UV Observations

The photometric data that led to the discovery of OW J0741 were a 2.5 hr sequence of $44\times 39\,{\rm{s}}$ exposures in the g'-band as part of the OmegaWhite survey (Macfarlane et al. 2015). This light curve showed a strong peak in its power spectrum at 22.6 minutes and a corresponding amplitude of 0.22 mag (Toma et al. 2016). The SAAO 1.9 m telescope and SHOC in 2016 April (see Table 1) show unequal minima on top of the predominantly sinusoidal variability. The dominant modulation in the light curve is caused by ellipsoidal deformation of the sdOB and the unequal minima are caused due to gravity-darkening of the deformed sdOB. The unequal depth of the minima was confirmed when we obtained 8 m class photometric data from Keck and Gemini.

To determine an ephemeris we initially determined the time of the deepest minima in the Keck and Gemini light curves by eye. This allowed us to assign an unambiguous cycle number to each minimum and remove any secondary minima. We then determined a linear fit to these times assuming a constant period. This allowed us to determine the period of OW J0741 to better than 0.1 s. We were then able to assign a cycle number to the minima in the light curve of the SAAO data obtained in 2016 April and the ULTRACAM data taken in 2016 December. A linear fit to these minima gives an ephemeris of

$$\begin{eqnarray}{T}_{o}(\mathrm{HMJD}) & = & 57695.611284\pm 0.000166\\ & & +0.031015829\pm (7.1\times {10}^{-8})E.\end{eqnarray} \tag{ 1 }$$

We are not able to use the OmegaWhite data taken in 2012 March to refine this further because there is an ambiguity regarding which of the observed minima are the deepest minima.

The individual spectra that we obtained for OW J0741 all have a relatively low S/N (≲10) and hence are not well suited for searching for a radial velocity (RV) period. We therefore folded the spectra of OW J0741 on the ephemeris shown in Equation (1) into 14 phase-bins and co-added individual spectra observed at the same binary phase. This increased the S/N per phase-bin to ≈25.

We used the FITSB2 routine (Napiwotzki et al. 2004) to measure the velocities. FITSB2 fits Gaussians, Lorentzians, and polynomials to the hydrogen and helium lines to fit the continuum, line, and line core of the individual lines. Using a ${\chi }^{2}$-minimization we fitted the wavelength shifts compared to the rest wavelengths of all suitable spectral lines (Geier et al. 2011). Assuming circular orbits, a sine curve was fitted to the folded RV data points. We find a semi-amplitude $K=422.5\,\pm 21.5$ $\mathrm{km}\,{{\rm{s}}}^{-1}$ and a systemic velocity of $\gamma =-14.0\,\pm 11.5$ $\mathrm{km}\,{{\rm{s}}}^{-1}$ (Figure 1), making it the most compact hot subdwarf binary with the largest RV amplitude.

Zoom In Zoom Out Reset image size

The atmospheric parameters of effective temperature ${T}_{\mathrm{eff}}$, surface gravity, $\mathrm{log}g\$, and helium abundance, $\mathrm{log}y$, where $y=n(\mathrm{He})/n({\rm{H}})$, were determined for the sdOB by fitting the rest-wavelength-corrected average spectra with metal-free NLTE model spectra (Stroeer et al. 2007). To obtain the average spectrum we shifted the 14 phase-binned spectra that were used for the RV measurement to the rest-wavelength and calculated the combined average spectrum. The spectrum shows Balmer lines as well as neutral (He i) and ionized (He ii) helium lines (see Figure 2). The ionization equilibrium between the He i and He ii is most sensitive to the effective temperature of the sdO, whereas the broad He ii and the hydrogen lines in the blue are most sensitive to $\mathrm{log}g$. We find ${T}_{\mathrm{eff}}\$= 39 400 ± 500 K, $\mathrm{log}g\$= 5.74 ± 0.09, and $\mathrm{log}y$ = −0.14 ± 0.07 (Figure 3). The occurrence of both He i and He ii as well as the increased helium abundance $\mathrm{log}y\gt -1$ classifies the hot subdwarf as intermediate He-sdOB (see Heber 2016). The errors were derived using a ${\chi }^{2}$-minimization.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The LCURVE code was used to perform the light curve analysis (Copperwheat et al. 2010). LCURVE uses grids of points to model the two stars. The shape of the stars in the binary is set by a Roche potential. We assume that the orbit is circular and that the rotation periods of the stars are synchronized to the orbital period. The flux that each point on the grid emits is calculated by assuming a blackbody of a certain temperature at the bandpass wavelength, corrected for limb-darkening, gravity-darkening, Doppler-beaming, and the reflection effect.

Some additional information was used as input. The orbital period was fixed to 44.66279 minutes as derived in Section 3. From spectroscopy the ${T}_{\mathrm{eff}}$, $\mathrm{log}g$, and semi-amplitude K were fixed to the values derived in Sections 3 and 4. Additionally, as a lower limit to the radius (mass) of the white dwarf companion, we use the zero-temperature mass–radius relation by Eggleton (quoted from Verbunt & Rappaport 1988). The passband specific gravity-darkening was calculated as described in Bloemen et al. (2011). We use $\beta =0.33\pm 0.01$ in g' and $\beta =0.34\pm 0.01$ in r'. The limb-darkening coefficients were calculated with the Claret limb-darkening prescription (Claret 2004). We used ${{\rm{a}}}_{1}=0.613$, ${{\rm{a}}}_{2}=-0.645$, ${{\rm{a}}}_{3}=0.621$, and ${{\rm{a}}}_{4}=-0.239$ for g' and ${{\rm{a}}}_{1}=0.422$, ${{\rm{a}}}_{2}=-0.444$, ${{\rm{a}}}_{3}=0.475$, and ${{\rm{a}}}_{4}=-0.194$ for r'. This leaves as free parameters in the model the mass ratio $q=\tfrac{{M}_{\mathrm{sdOB}}}{{M}_{\mathrm{WD}}}$, the inclination i, secondary temperature ${T}_{\mathrm{WD}}$, the scaled radii of both components ${r}_{\mathrm{sdOB}/\mathrm{WD}}$, the velocity scale ($[K\,+{K}_{\mathrm{WD}}]/\sin i$) and the beaming parameter B (${F}_{\lambda }\,={F}_{0,\lambda }[1-B\tfrac{{v}_{r}}{c}]$, see Bloemen et al. 2011). In addition, to account for any residual airmass effect we add a third-order polynomial.

To determine the uncertainties in the parameters we combine LCURVE with emcee (Foreman-Mackey et al. 2013), an implementation of an MCMC sampler that uses a number of parallel chains to explore the solution space. We use 256 chains and let them run until the chains stabilized to a solution, which took approximately 1000 generations.

For the Keck and ULTRACAM g'-band and r'-band light curves we find two distinct solutions: one that finds the system eclipsing and the other that finds the system non-eclipsing (Figure 4). The Gemini light curves have a higher precision and we find exclusively non-eclipsing solutions (Figure 5). This becomes particularly obvious in the significant larger error bars for inclination (i), mass ratio (q), and sdOB mass (${M}_{\mathrm{sdOB}}$) for the individual fits to the Keck and ULTRACM light curves (Table 2). To obtain the final solution we fit all six light curves simultaneously. From the simultaneous fit we only find non-eclipsing solutions (Figure 6).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 2.  Results From the Individual Light Curve Fits

Light Curve	$q=\tfrac{{M}_{\mathrm{sdOB}}}{{M}_{\mathrm{WD}}}$	${M}_{\mathrm{sdOB}}$	${R}_{\mathrm{sdOB}}$	${M}_{\mathrm{WD}}$	i	a	$\tfrac{{R}_{\mathrm{sdOB}}}{a}$
		(${M}_{\odot }$)	(${R}_{\odot })$	(${M}_{\odot }$)	${(}^{\circ })$	(${R}_{\odot })$
Keck (2016 Nov 03)	0.47 ± 0.23	0.40 ± 0.25	0.13 ± 0.03	0.84 ± 0.22	62.3 ± 11.5	0.45 ± 0.04	0.30 ± 0.04
Gemini (2016 Nov 28)	0.29 ± 0.13	0.25 ± 0.15	0.12 ± 0.02	0.92 ± 0.21	50.1 ± 6.6	0.44 ± 0.04	0.27 ± 0.03
Gemini (2016 Dec 05)	0.37 ± 0.14	0.32 ± 0.18	0.13 ± 0.02	0.91 ± 0.22	53.3 ± 6.3	0.45 ± 0.04	0.29 ± 0.03
Gemini (2016 Dec 06)	0.39 ± 0.16	0.36 ± 0.20	0.13 ± 0.02	0.92 ± 0.21	55.0 ± 6.1	0.45 ± 0.04	0.29 ± 0.03
ULTRACAM g' (2016 Dec 07)	0.56 ± 0.25	0.44 ± 0.26	0.14 ± 0.03	0.81 ± 0.20	69.6 ± 11.4	0.45 ± 0.04	0.31 ± 0.03
ULTRACAM r' (2016 Dec 07)	0.50 ± 0.26	0.40 ± 0.27	0.14 ± 0.03	0.79 ± 0.18	64.4 ± 12.8	0.44 ± 0.04	0.31 ± 0.03
Combined Analysis	0.32 ± 0.10	0.23 ± 0.12	0.11 ± 0.02	0.72 ± 0.17	57.4 ± 4.7	0.41 ± 0.04	0.28 ± 0.02

Download table as:  ASCIITypeset image

From the strong light curve variability caused by ellipsoidal modulation in combination with the RV amplitude and the spectroscopic solution, we can solve the system and derive system parameters. Solutions were calculated for each individual light curve, as well as from a simultaneous fit to all six light curves. The results from the latter are taken as the final solution (Table 2).

We find that the system consists of a low-mass sdOB with a high-mass WD companion. A mass ratio $q={M}_{\mathrm{sdOB}}/{M}_{\mathrm{WD}}\,=0.32\pm 0.10$, a mass for the sdOB ${M}_{\mathrm{sdOB}}=0.23\pm 0.12$ ${M}_{\odot }$, and a WD companion mass ${M}_{\mathrm{WD}}=0.72\pm 0.17$ ${M}_{\odot }$ were derived. The inclination is found to be i = 574 ± 47 (Table 3).

Table 3.  Overview of the Derived Parameters for OW J0741

Right Ascension	R.A. [hr]	07:41:06.1
Declination	Decl. ${[}^{\circ }]$	–29:48:11.0
Visual Magnitude	${m}_{{\rm{g}}}$	20.02 ± 0.11
Atmospheric Parameters of the sdOB
Effective Temperature	${T}_{\mathrm{eff}}\$ [K]	39 400 ± 500
Surface Gravity	$\mathrm{log}g\$	5.74 ± 0.09
Helium Abundance	$\mathrm{log}y$	−0.14 ± 0.06
Orbital Parameters
	T0 [MHJD]	57695.611284 ± 1.66 × 10−4
Orbital Period	${P}_{\mathrm{orb}}\,[\min ]$	44.66279 ± 1.16 × 10−4
RV Semi-amplitude	K [$\mathrm{km}\,{{\rm{s}}}^{-1}$]	422.5 ± 21.5
System Velocity	γ [$\mathrm{km}\,{{\rm{s}}}^{-1}$]	−14.0 ± 11.5
Binary Mass Function	${f}_{{\rm{m}}}$ [${M}_{\odot }$]	0.242 ± 0.020
Derived Parameters
Mass Ratio	$q=\tfrac{{M}_{\mathrm{sdOB}}}{{M}_{\mathrm{WD}}}$	0.32 ± 0.10
sdOB Mass	${M}_{\mathrm{sdOB}}$ [${M}_{\odot }$]	0.23 ± 0.12
sdOB Radius	${R}_{\mathrm{sdOB}}$ [${R}_{\odot }]$	0.11 ± 0.02
WD Mass	${M}_{\mathrm{WD}}$ [${M}_{\odot }$]	0.72 ± 0.17
Orbital Inclination	$i\,{[}^{\circ }]$	57.4 ± 4.7
Separation	a [${R}_{\odot }]$	0.41 ± 0.04
Distance	d [kpc]	6.6 ± 2.1

Download table as:  ASCIITypeset image

We calculate the distance to OW J0741 using the visual V magnitude (${m}_{{\rm{V}}}$), the sdOB mass (${M}_{\mathrm{sdOB}}$), ${T}_{\mathrm{eff}}\$and $\mathrm{log}g\$ as described in Ramspeck et al. (2001). The visual V magnitude ${m}_{{\rm{V}}}=20.06$ mag was calculated following the conversion from SDSS colors as described in Jester et al. (2005). Because OW J0741 is located toward the Galactic Bulge, significant reddening can occur. Green et al. (2015) calculated 3D extinction maps from PanSTARRS data¹⁶ and find toward the direction of OW J0741 an extinction of $E(B-V)\,={0.34}_{-0.034}^{+0.032}$ at distances above 6 kpc, resulting in a total extinction in the V-band of A_V = 1.05 mag using R_V = 3.1. With the corrected magnitude we find a distance of $d=6.6\pm 2.1\,\mathrm{kpc}$, which for a Galactic latitude of $-3\buildrel{\circ}\over{.} 5$ gives a height below the Galactic plane of ∼400 pc.

For the X-ray observations no source was detected with a 3σ upper limit of 0.0036 ct s⁻¹. Assuming a 1 keV thermal bremsstrahlung and a Galactic absorption of $4.4\times {10}^{21}$ cm⁻² (the average to the edge of the Galaxy) we find an upper limit of $2.6\times {10}^{-13}$ erg cm⁻² s⁻¹ to the unabsorbed flux (this limit is not strongly dependent on the assumed model). Given the large distance to OW J0741, the upper limit for the X-ray luminosity ($\lt 2\times {10}^{33}$ erg s⁻¹) is not very constraining, since the upper end of X-ray luminosity of hot subdwarfs is $\sim {10}^{31}$ erg s⁻¹ (Mereghetti & La Palombara 2016). In contrast, OWJ0741 was detected in the UVW2 filter (peak effective area 1928 Å) with a corrected count rate of 0.34 ± 0.01 ct s⁻¹, implying a flux $2.04\pm 0.08\times {10}^{-16}$ erg cm⁻² s⁻¹ Å⁻¹.

7.1. Origin of the sdOB Star

Using the constraints from the observations outlined above, we use the stellar evolution code MESA (Paxton et al. 2011, 2013, 2015) to put constraints on the nature of the sdOB star. We then take the most likely model and evolve it forward in time to generate predictions about the future of this system.

7.1.1. He-star Model

We first test a model with the canonical sdOB mass of $0.462\,{M}_{\odot }$ with a hydrogen envelope mass of $2.5\times {10}^{-4}\,{M}_{\odot }$. We find that after core He-burning and during shell He-burning, this model spends $\approx 4\,\mathrm{Myr}$ in crossing the observation box (see dot-dashed purple line in Figure 7). There are a few problems with the system being in this state. First, given the log g, orbital period, and mass ratio, this model fills its RL at orbital periods $\gt 50$ minutes, meaning that the higher mass (compared to the derived mass from observations) implies a radius larger than the current RL.¹⁷ Second, the derived mass ratio from observations implies that if the He-star had a mass of $0.46\,{M}_{\odot }$, the compact companion must be super-Chandrasekhar, as well as that the derived mass from the observations is inconsistent with the canonical sdOB mass.

Zoom In Zoom Out Reset image size

We also tested He-star models between $0.325\,{M}_{\odot }$ and $0.462\,{M}_{\odot }$ that experienced He core-burning, and found that during the He core-burning stage, ${T}_{\mathrm{eff}}\$was too low and $\mathrm{log}g$ was too high to match the measurements. If the models were able to produce a He-star massive enough to experience He shell-burning, this stage reached higher ${T}_{\mathrm{eff}}\$values, but with similar $\mathrm{log}g\$ values to the He core-burning stage. Therefore, only He shell-burning models near $0.462\,{M}_{\odot }$ were able to match the ${T}_{\mathrm{eff}}\$and $\mathrm{log}g\$ measurements.

7.1.2. He White Dwarf Model

In a second scenario, we consider the sdOB star as a helium white dwarf (He WD) that did not start helium-burning. If we assume that the star is a He WD that is just coming out of a CE event, then the ${T}_{\mathrm{eff}}\$and $\mathrm{log}g\$ observation box gives a solution of a unique mass (Althaus et al. 2013). Figure 7 shows that this mass must be close to $0.320\,{M}_{\odot }$, shown by the middle solid curve. This model spends $\approx$ 220,000 years in the observation box and has a post-CE age of $\approx 1.1\,\mathrm{Myr}$.

He WDs of this mass range experience diffusion-induced H novae (Althaus et al. 2001), and the tracks of these novae pass through the same $\mathrm{log}g$ values, but at higher ${T}_{\mathrm{eff}}\$for a given stellar mass. Therefore, we can construct a lower-mass He WD model ($0.242\,{M}_{\odot }$) that has low ${T}_{\mathrm{eff}}\$just out of the CE but passes through the observation box after the first diffusion-induced H-nova. This is shown by the orange dashed curve in Figure 7. This model spends $\approx$ 66,000 years in the observation box and has a post-CE age of $\approx 11\,\mathrm{Myr}$. Since the $0.320\,{M}_{\odot }$ He WD model spends more than a factor of 3 more time in the observation box, we conclude that the $0.320\,{M}_{\odot }$ model is the more likely fit to our observations, even though the lower-mass He WD model fits the derived mass better.

7.2. Future Predictions

OW J0741 was discovered as one of bluest variable sources in the OmegaWhite survey. The VST observations revealed variability with a period of 22.6 minutes. Follow-up observations show that OW J0741 is an ultracompact sdOB binary with a compact companion with ${P}_{\mathrm{orb}}=44.66279\,\pm 1.16\times {10}^{-4}$ minutes, making it the most compact hot subdwarf binary known today.

High-S/N photometry obtained with Gemini exclude an eclipse that allows us to put tight constraints on the system parameters. Combining Gemini, Keck, and ULTRACAM light curves with spectroscopy, we find a mass ratio $q={M}_{\mathrm{sdOB}}/{M}_{\mathrm{WD}}=0.32\,\pm 0.10$, a mass for the sdOB ${M}_{\mathrm{sdOB}}=0.23\pm 0.12$ ${M}_{\odot }$ and a WD companion mass ${M}_{\mathrm{WD}}=0.72\pm 0.17$ ${M}_{\odot }$. The derived sdOB mass is inconsistent with the canonical mass for hot subdwarfs of $\approx 0.47$ ${M}_{\odot }$ and therefore the sdOB has not evolved from the standard hot subdwarf channel where the envelope of the subdwarf progenitor gets stripped at the tip of the red-giant branch.

To put constraints on the nature of the sdOB star, we compared the derived ${T}_{\mathrm{eff}}\$and $\mathrm{log}g\$ to evolutionary tracks for He-stars and He white dwarfs computed with MESA. For the He-star scenario, only a He shell-burning star with a mass around $0.462\,{M}_{\odot }$ is consistent with the derived ${T}_{\mathrm{eff}}\$and log g. However, such a high mass is inconsistent with the derived sdOB mass. For the white dwarf scenario, a He white dwarf with a mass of 0.320 ${M}_{\odot }$ and a common envelope age of $\approx 1.1\,\mathrm{Myr}$ is fully consistent with the derived system parameters. Although the evolutionary timescale is about 10 times faster for a He white dwarf, we conclude that the most likely nature of the sdOB is a He white dwarf with a mass of 0.320 ${M}_{\odot }$ and a common envelope age of $\approx 1.1\,\mathrm{Myr}$.

To study the future evolution of the system we have constructed MESA models, assuming a 0.320 ${M}_{\odot }$ He white dwarf with a massive white dwarf companion of 0.85 ${M}_{\odot }$, consistent with the derived mass ratio. In 17.6 Myr, the He WD will start RLOF at an orbital period of ≈5 minutes. Depending on the spin–orbit synchronization timescale the object will either merge to form an R CrB star or end up as a stable accreting AM CVn-type system with a He WD donor. Better understandings of the spin–orbit synchronization timescales are required to decide whether the system will merge or prevent the merger. As an AM CVn, the system will show weak He shell flashes, but none strong enough to trigger a double-detonation SN Ia. Therefore, whether the system merges on contact or not, we conclude that this binary is not an SN Ia progenitor.

This work was supported by the GROWTH project funded by the National Science Foundation under grant No. 1545949. J.v.R. acknowledges support by the Netherlands Research School of Astronomy (NOVA) and the foundation for Fundamental Research on Matter (FOM). This research is funded in part by the Gordon and Betty Moore Foundation through Grant GBMF5076 and was also supported by the National Science Foundation under grant PHY 11-25915. We acknowledge stimulating workshops at Sky House where these ideas germinated.

T.R.M. acknowledge the support from the Science and Technology Facilities Council (STFC), ST/L00733. S.G. is supported by the Deutsche Forschungsgemeinschaft (DFG) through grant GE2506/9-1. D.K. acknowledges financial support from the National Research Foundation of South Africa. P.J.G. acknowledges support from NOVA for the original OmegaWhite observations and hospitality of the Kavli Institute for Theoretical Physics. Armagh Observatory is core funded by the Northern Ireland Executive through the Department for Communities.

This paper uses observations made at the South African Astronomical Observatory. Based on observations obtained at the European Southern Observatory, proposal 297.D-5010. Some results presented in this paper are based on observations obtained at the Gemini Observatory, proposal ID GS-2016B-FT-9, which is operated by the Association of Universities for Research in Astronomy, Inc., under a cooperative agreement with the NSF on behalf of the Gemini partnership: the National Science Foundation (United States), the National Research Council (Canada), CONICYT (Chile), Ministerio de Ciencia, Tecnología e Innovación Productiva (Argentina), and Ministério da Ciência, Tecnologia e Inovação (Brazil). The authors thank the staff at the ESO Paranal and Gemini South observatories for performing the observations in service mode. Some of the data presented herein were obtained at the W.M. Keck Observatory, which is operated as a scientific partnership among the California Institute of Technology, the University of California and the National Aeronautics and Space Administration. The Observatory was made possible by the generous financial support of the W.M. Keck Foundation. The authors wish to recognize and acknowledge the very significant cultural role and reverence that the summit of Maunakea has always had within the indigenous Hawaiian community. We are most fortunate to have the opportunity to conduct observations from this mountain. We thank the Swift team for the granting our DDT request.

Facilities: Gemini:South (GMOS-S) - , Keck:I (LRIS) - , VLT:Antu (FORS2) - , NTT (ULTRACAM) - , SALT (SALTICAM) - , Radcliffe (SHOC) - .

Software: LCURVE (Copperwheat et al. 2010), emcee (Foreman-Mackey et al. 2013), MESA (Paxton et al. 2011, 2013, 2015), FITSB2 (Napiwotzki et al. 2004).