Submitted Manuscript: Confidential Direct Imaging Discovery of a Low-Mass Exoplanet Between Three Stars Kevin Wagner1,7,8*, Dániel Apai1,2,8, Markus Kasper3, Kaitlin Kratter1, Melissa McClure3, Massimo Robberto4, Jean-Luc Beuzit5,6 1Department of Astronomy and Steward Observatory, The University of Arizona, 933 N. Cherry Avenue, Tucson, AZ 85721, USA. 2Lunar and Planetary Laboratory, The University of Arizona, 1640 E. University Boulevard, Tucson, AZ 85718, USA. 3European Southern Observatory, Karl-Schwarzschild-Strasse 2, D-85748 Garching, Germany. 4Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218, USA. 5Université Grenoble Alpes, IPAG, F-38000 Grenoble, France!. 6CNRS, IPAG, F-38000 Grenoble, France!. 7National Science Foundation Graduate Research Fellow. 8Earths in Other Solar Systems Team, NASA Nexus for Exoplanet System Science. *Correspondence to: [email protected] Abstract: We present the discovery of a young giant planet in the HD131399 triple system through imaging and spectroscopy with VLT/SPHERE. HD131399Ab is one of the lowest mass and coldest directly imaged planets with a mass of 4±1 times that of Jupiter and an effective temperature of 850±50 kelvin. The planet is seen at 82 au projected separation from the primary star, which is surprisingly comparable to the ~300–400 au semi-major axis of the hierarchical triple star system and dynamically unlike any other known exoplanet. Through N-body simulations we confirm that a range of stable orbits exist for the planet, and that the gravitational influence of the binary excites an orbital evolution that is compatible with a wide range of plausible formation scenarios and evolutionary histories. The location of HD131399Ab on a wide orbit in a triple system is unexpected and demonstrates that massive planets may form or migrate to long and possibly unstable orbits in multi-star systems. One Sentence Summary: This article presents the discovery and characterization of a dynamically active young exoplanet within a triple star system. Main Text: Thousands of planets around other stars have recently been discovered (e.g. 1, 2), revealing a greater diversity than predicted by traditional planet formation models based on the solar system. Extreme examples of this are planets within binary and multiple star systems, which form and evolve in time-evolving radiation and gravitational fields. Direct imaging allows for the detection and characterization through spectroscopy of long-period giant planets – enabling constraints to be placed on planet formation models via predictions of planet population statistics and atmospheric properties. However, most direct imaging surveys have traditionally excluded visual binary or multiple systems whose separations are less than a few hundred astronomical units1 due to the assumption that such planetary systems would either be disrupted or never form. As a result of this observational bias, most directly imaged exoplanets have been found around single stars. A surprising result from the handful of confirmed directly imaged planets is that long- period giant planets appear to be more common around A-type stars (M~1.6–2.4 M⊙, where M⊙ 1Where one astronomical unit, or au, stands for the mean Earth–Sun distance. = 1 solar mass) than around solar-type and lower mass stars (3–14). Since roughly half of A-stars belong to binary or multiple systems (15, 16), to build a complete census of long-period giant planets requires investigation of both single and multiple systems. In the latter, a variety of long- period orbits might arise as a result of planet-planet or planet-star scattering events (17) and even some with chaotic orbits that wander between the stellar components (18, 19) – all of which may be readily detected by direct imaging. In our ongoing adaptive optics imaging campaign using VLT/SPHERE (20) we are sampling a coeval population of ~100 young single and multiple A- type stars in the Upper Scorpius-Centaurus-Lupus association to place constraints on the primordial occurrence rate of long-period giant planets. HD131399ABC is a 16±1 Myr old triple system in Upper Centaurus Lupus (UCL; 21, 22) at a distance of 98±7 pc (23) whose basic properties are given in Table 1. Despite its young age, the system shows no evidence of infrared excess and thus its primordial disk has likely been depleted to beneath detectable levels (24). We observed HD131399 on June 12, 2015 with SPHERE, obtaining a wide range of near-infrared spectral coverage from Y– to K-band (0.95– 2.25 µm) and diffraction-limited imaging with an 8.2-meter telescope aperture. Our observations, described in detail in the supplementary online materials (SOM), resulted in the discovery of HD131399Ab, a point source with a contrast to HD131399A of ΔK1(2.1 µm)=12.5±0.1 magnitudes and projected separation of 0.84 arcsec, or 82±6 au (see Fig. 1 and Table 1). After the initial discovery, we obtained follow-up observations with SPHERE to verify common proper motion and to improve the quality of the near-infrared spectrum, enabling the characterization of the planet’s atmospheric properties. HD131399A HD131399Ab HD131399B HD131399C Spectral Type A1V T3±1 G3 K5 Mass 1.82 M⊙ 4±1 MJup 0.96 M⊙ 0.57 M⊙ Temperature 9300 K 850±50 K 5700 K 4400 K (TEff) Distance (pc) 98±7 Projected 0.839±0.004 2015 June 3.149±0.006 3.215±0.006 separation from 2016 March A (arcsec) 0.834±0.004 3.150±0.006 3.220±0.006 Projected 82±6 309±22 315±22 separation from A (au) Position angle 194.2±0.3 2015 June 221.9±0.3 222.0±0.3 (Degrees E of N 2016 March from A) 194.1±0.3 221.5±0.3 221.9±0.3 J magnitude 6.772±0.018 20.0±0.2 H magnitude 6.708±0.034 19.7±0.2 K-band K=6.643±0.026 K1=19.1±0.1 K1=8.47±0.04 K1=10.49±0.06 magnitude Table 1. Basic parameters of the stars and directly imaged planet in HD131399. The masses, effective temperatures, and spectral types were found through comparison of their K1 luminosity to evolutionary models (23). J, H, and K magnitudes for HD131399A were obtained from the 2MASS All Sky Catalog of point sources. A description of our astrometric calibrations and associated uncertainties can be found in the SOM. Fig. 1. VLT/SPHERE images revealing HD131399Ab and the stellar components of the hierarchical triple system. The central regions that are affected by the coronagraph and residual scattered starlight are blocked by a mask, with the location of HD131399A indicated by the crosshairs. The right image shows a composite of the PSF subtracted aperture around the planet (dashed) superposed on the wide field image showing the stellar components of the system whose luminosities are suppressed to the level of the planet for clarity. We detect HD131399Ab with a signal to noise ratio in Y– (1.04 µm), J– (1.25 µm), and H-band (1.62 µm) of 8–10 and in K1 and K2 (2.11 µm and 2.25 µm) of 23.5 and 11.3, respectively. We extract the position of the planet via injection of a negative Gaussian PSF over a grid of 0.01-pixel spacing, from which we determine its position as the location with the minimum square of residual intensity in an aperture of one FWHM diameter around the planet. This method is typically accurate to determine the center of a well-sampled PSF to within ~0.1 pixels (25), which enables precision astrometry for the planet’s confirmation and orbital analysis. Following astrometric calibrations (see SOM for details), we measure a positional displacement to HD131399A of Δα = 2.7±8.7 mas and Δδ = 5.0±8.0 mas over the nine-month baseline, where the uncertainties are dominated by the calibration of the instrument orientation across the two epochs. This allows us to reject the hypothesis of a background object with 95% confidence, which would have moved relative to HD131399A by Δα = 22.3±0.6 mas and Δδ = 23.6±0.6 mas due to relatively high proper motion of the system (23). Assuming a Keplerian orbit for the planet with a semi-major axis of 82 au yields a period of roughly 550 years, which for a face-on circular orbit over nine months is expected to produce ~7 mas of relative motion or less for eccentric or inclined orbits, which is consistent with the observations. We convert the object’s absolute K1 magnitude to a mass estimate via comparison to widely used evolutionary tracks calculated for “hot-start” initial conditions (26), in which the planet retains its initial entropy of formation. Systematic interpolation between evolutionary tracks argues for a mass of M=4±1 MJup, which places HD131399Ab firmly in the planetary mass regime. Since at a given luminosity age and mass are degenerate, the well-constrained distance and age of the system (98±7 pc and 16±1 Myr, respectively) make HD131399Ab an important example as a low-mass directly imaged exoplanet. The closest low-mass analogue to HD131399Ab is 51 Eridani b, which has an even lower possible minimum mass of M=2–12 MJup. Unlike other more luminous directly imaged planets, HD131399Ab and 51 Eridani b are also consistent with less-widely used cold-start core-accretion models (noted for 51 Eri b in 14), in which an arbitrary cooling parameter is introduced to allow the planet to lose some fraction of its initial entropy throughout its formation due to inefficient accretion (27). These models are less sensitive to mass since the initial energy loss is currently unconstrained, but within these systematic uncertainties HD131399Ab and 51 Eridani b could both be as massive as 12 MJup in the most extreme scenario.