A DIM CANDIDATE COMPANION TO CEPHEI

High contrast imaging at small angles is a very useful tool for accessing inner regions around stars. While instrumental limits are constantly improved by the development of exoplanet imaging and characterization techniques, such capabilities can be used in stellar astrophysics to discover new systems or put constraint on known ones. Moreover, this small-angle capability allows a significant reduction in the size of potential space telescopes aimed at detecting and characterizing exoplanets, which is a cost-effective way of building exoplanet missions.

We have used a 1.5 m subaperture of the 5.1 m Hale telescope at Palomar with a small inner working angle (IWA) phase-mask coronagraph to detect a new candidate stellar companion to the δ Scuti star Cephei. The result is interesting for two reasons. First, Cephei was claimed to have an infrared excess measured by IRAS and recently detected by near-infrared ground-based interferometry. Such infrared excesses are generally attributed to dusty debris disks (Absil et al. 2006). We show here with an image that the origin of the near-infrared excess is a second object that is likely a late-type stellar companion. Moreover, we demonstrate that the previously reported 60 μm IRAS excess is attributable to an unrelated field object. Second, the potential companion was imaged at 1.1 diffraction beamwidths (λ/D, with the working wavelength λ ≃ 2.16 μm, and the telescope diameter D = 1.5 m) from the star. Such a small angle detection was only possible because of our phase-mask vector vortex coronagraph (VVC), demonstrating that such coronagraphs can effectively provide very small IWA.

Cephei is a bright (V = 4.2, H = 3.7, K = 3.5) and nearby (d = 26.2 ± 0.3 pc; see van Leeuwen 2007) F0 IV star with a temperature T = 7350 K, of the δ Scuti type (Table 1). Privileged targets in asteroseismologic studies (Bruntt et al. 2007), δ Scutis are multi-periodic variable stars lying at the base of the Cepheid instability strip, where it crosses the main sequence. Their small amplitude (0.001–1 mag) variability is due to simultaneous radial and non-radial pulsation modes with periods between 0.25 and 5 hr. The spectral type of δ Scuti stars ranges from A0 to F5, with nominal mass and temperature around 2 M_Sun and 7500 K, respectively. For Cephei, Kennelly et al. (1999) derived a mass of 1.8 ± 0.2 M_Sun.

Table 1.  Cephei Fundamental Stellar Properties

Properties	Cephei
Coordinates (hms)	R.A. = 22 15 02.19, decl. = +57 02 36.91
Galactic coordinates (deg)	l = 102.87, b = +00.39
Proper motion ('')	δR.A. = 0.47645, δdecl. = 0.04999
Radial velocity (km s−1)	−0.6
Parallax ('')	0.3886
Spectral type	F0 IV
V mag	4.2
K mag	3.5
Age	∼600 Myr
Distance	26.2 ± 0.3 pc
Mass	1.8 MSun
Radius	2 RSun
Temperature	7350 K

Download table as:  ASCIITypeset image

Cephei was claimed to be detected by IRAS at 12, 25, and 60 μm as PSC 22132+5647. The reported 60 μm flux density of 1.20 ± 0.08 Jy is > 10 times the expected photospheric value (the ratio of excess infrared luminosity divided by the total energy output from the photosphere was reported to be τ ≃ 1.56 × 10⁻⁴). On this basis, Oudmaijer et al. (1992), Moor et al. (2006), and Rhee et al. (2007) identified the system as hosting a bright debris disk. In Rhee et al. (2007), this excess was fitted with a single blackbody giving a putative ring of dust at 62 AU with a temperature of 65 K. Using space velocities UVW, lithium abundance, and location on an H-R diagram, Rhee et al. (2007) also estimated the age of Cephei to be a very uncertain ∼600 Myr.

In the course of an on-going survey for bright exozodiacal disks around main-sequence stars, led by one of us, the CHARA-FLUOR interferometer detected a K'-band excess for Cephei (O. Absil et al. 2011, in preparation). One possible interpretation of such near-infrared excesses is emission from hot dust close to the star (Absil et al. 2006, 2008, 2009; Di Folco et al. 2007; Akeson et al. 2009). The grain populations usually derived from such observations are quite intriguing, as they point toward very high dust replenishment rates, high cometary activity, or unlikely major collisional events. Note that due to sparse (u, v) coverage, CHARA-FLUOR poorly constrains the nature of the excess or its location inside the FLUOR 1'' FWHM field of view, suggesting that a high-resolution image might be able to distinguish between a close companion and dust.

We observed Cephei in the K_s band on 2010 June 23, as part of our program of adaptive optics (AO) high contrast imaging of nearby young bright stars at Palomar. We took 50 frames of 4.238 s each, for a total of 212.4 s on the target. We used the Palomar well-corrected subaperture (WCS) and the VVC, which have both been described in detail earlier (Serabyn et al. 2007, 2009; Mawet et al. 2010b).

Briefly, a clear, off-axis WCS is provided by a set of relay optics upstream of the Palomar AO system. The off-axis relay magnifies and shifts an off-axis subaperture pupil onto the deformable mirror, yielding roughly 10 cm actuator spacing in the selected pupil. This configuration reduces the aperture size from 5 m to 1.5 m, but allows operation in the extreme AO regime. Providing unprecedented image quality on a ground-based telescope (K_s-band Strehl ratios S ⩾ 90%), this setup is ideal for use with the VVC (Mawet et al. 2010b; Serabyn et al. 2010).

The VVC is based on a transparent phase mask, with an azimuthal phase ramp centered on the image of the central star. Upon propagation to the Lyot plane, the central starlight is redistributed outside the geometric pupil area. There, a simple diaphragm known as a Lyot stop is used to reject the starlight, leaving any off-axis feature mostly unaltered in the process: the 50% off-axis throughput, which is the commonly adopted definition of the IWA, is reached at 0.9 λ/D for a second-order VVC such as the one we used at Palomar (Mawet et al. 2010a).

The VVC allowed us to detect a second object near Cephei at 330 ± 50 mas from the host star, which corresponds to ∼1.1 times the width of the diffraction beam (λ/D ≃ 300 mas at K_s), making it the smallest angle detection ever reported with a coronagraph in λ/D units. The projected separation of the potential companion places it at 8.6 ± 1.4 AU. Figure 1(a) shows the raw image after sky subtraction, flat fielding, bad pixel replacement, and cosmic-ray removal. A faint secondary point source is clearly detected on this image to the east (P.A. = 90° ± 10°). A reference star (HD 213558) was also imaged right after Cephei, with no companion present (Figure 1(b)). The reference star was chosen with a V − K color and brightness as similar as possible to the target. Matching V magnitudes is important to ensure similar AO corrections. K magnitudes and integration time also need to be matched to ensure a proper subtraction. Also, to reduce thermal, seeing, and flexure-induced variations as much as possible, the calibration star was chosen and observed as close as possible to the hour angle and elevation of the target star, and just a few minutes after the latter.

Zoom In Zoom Out Reset image size

Finely scaling the reference star image to exactly match Cephei's apparent magnitude, and subtracting it from Cephei, leaves a clean detection of the second object (Figure 1(c)). The subtracted image was used to derive the astrometry and photometry with minimal contamination from the residual errors left after imperfect wavefront correction. Note that the 5σ detection level after subtraction is calculated to be ∼2 × 10⁻³ at ∼1 λ/D, mostly limited by residual centering errors on the VVC.

4.1. K_s-band Photometry

4.2. Probability of Association

Assuming the candidate companion is bound to Cephei, we will derive its physical properties based on an evolutionary model. We will also discuss the consequence of this discovery on the interpretation of near-infrared and far-infrared excesses. An interesting question one could then ask is why this second source has largely remained undetected so far.

5.1. Possible Nature of the Candidate Companion

According to the BCAH98 evolutionary model (Baraffe et al. 1998), and following the age estimation in Rhee et al. (2007), the candidate companion would have a mass of ∼0.5 M_Sun, and a temperature of 3650 K (Table 2). The stellar type of the candidate companion is thus likely to be at the transition between late-type K stars and early-M stars. This result is fairly independent of the age between 200 Myr up to well above 1 Gyr, making this mass determination quite robust even with big error bars (Figure 2).

Zoom In Zoom Out Reset image size

Table 2. Candidate Companion Fundamental Properties

Properties	Cephei b
Spectral type	K8-M2
K mag	7.8
Age	∼600 Myr
Distance	⩾8.6 AU from primary
Mass	0.5 MSun
Radius	∼0.7 RSun
Temperature	3650 K

Download table as:  ASCIITypeset image

The interferometrically detected near-infrared excess is likely due to the companion, and not hot dust. However, the candidate companion would only contribute ∼5% additional flux at 60 μm. It thus cannot account for the large excess detected in the IRAS beam, which was interpreted as thermal emission from cold dust at a separation of ∼62 AU (Rhee et al. 2007). In this picture, the cold dust belt would thus likely be circumbinary which is not unprecedented, see for instance the case of GG Tau (Krist et al. 2005). However, as mentioned earlier, Cephei lies away from the galactic bulge but still on the galactic plane (b = +04, see Table 1), which raises the possibility of source confusion given the large beam size of IRAS (∼25'' at 60 μm). To confirm the excess emission, the star was observed with Spitzer/MIPS on 2009 February 20 (PI: George Rieke).

The MIPS 70 μm Basic Calibrated Data (BCD) image retrieved from the Spitzer Heritage Archive (see Figure 3) shows only a very weak detection at the stellar position, and a much stronger far-infrared source located 76'' to its NNE. Photometry in a 16'' circular aperture centered on the stellar position using aperture corrections as given in the MIPS instrument handbook finds a stellar 70 μm flux density of ∼80 mJy. This value is consistent with the expected stellar photospheric emission (as extrapolated from the IRAS 12 and 25 μm flux densities). We therefore conclude that the IRAS 60 μm measurement was very likely confused by the bright adjacent source. Indeed, the reported position of the IRAS source PSC 22132+5647 is shifted 33'' E from the known position of Cephei at the time of the measurement. Thus, it was never right to associate Cephei and PSC 22132+5647, which is likely a virtual combination of two real sources: Cephei, detected at 12 and 25 μm by IRAS, and the far-infrared source to its NNE, dominating at 60 μm. In conclusion, Cephei lacks any far-infrared excess and thus does not possess a debris disk sculpted by the putative companion.

Zoom In Zoom Out Reset image size

5.2. Visible Speckle Interferometry

5.3. Absence of Indirect Detections

This Letter reports the detection of a candidate dim companion to Cephei, a δ Scuti star. This is the first image of a candidate companion, seen at 330 ± 50 mas separation, only 1.1 resolution elements from the central host star (∼8.6 AU). If physically associated, the companion has a mass of ∼0.5 M_Sun, and a temperature of 3650 K, likely placing it at the transition between late-type K stars and early M stars. The interferometrically detected near-infrared excess is likely due to the companion, and not hot dust, while the previously reported IRAS 60 μm excess was due to source confusion. Finally, the most plausible reason why this candidate companion remained hidden so far is that Hipparcos failed to detect it because of its long period, highly eccentric orbit, while the stellar pulsations of the δ Scuti primary precluded an RV detection.

This detection was made possible by state-of-the-art high contrast imaging techniques (wavefront control and vortex coronagraphy) envisioned for upcoming next-generation ground-based extreme AO systems and future space-based coronagraphs. We have demonstrated a 5σ detection capability of the order of ∼2 × 10⁻³ at ∼1λ/D, which is sufficient to directly image secondary stars and brown dwarfs to an order of magnitude fainter than speckle interferometry. At 2λ/D, we routinely reach 10⁻⁴ contrast levels, and with careful calibration ∼10⁻⁵ (Mawet et al. 2010b; Serabyn et al. 2010), two to three orders of magnitude better than current interferometric techniques. Extreme AO coupled to vector vortex coronagraphy has thus great potential for the characterization of spectroscopic binary and brown dwarf orbits, especially with a larger telescope. It can also dramatically improve imaging capabilities even with rather small telescopes. Such techniques thus open the door to the direct imaging of spectroscopic binary companions, as well as brown dwarfs and exoplanets.

This work was carried out at the European Southern Observatory (ESO) site of Vitacura (Santiago, Chile), and the Jet Propulsion Laboratory (JPL), California Institute of Technology (Caltech), under contract with the National Aeronautics and Space Administration (NASA). The data presented in this Letter are based on observations obtained at the Hale Telescope, Palomar Observatory, as part of a continuing collaboration between Caltech, NASA/JPL, and Cornell University. This work is also based (in part) on observations made with the Spitzer Space Telescope, which is operated by the JPL, Caltech, under a contract with NASA. This research has made use of the NASA/IPAC/NExScI Star and Exoplanet Database, which is operated by the JPL, Caltech, under contract with NASA, and NASA's Astrophysics Data System and of the SIMBAD database, operated at CDS (Strasbourg, France).