Discovery of a Low-Luminosity, Tight Substellar Binary at the T/Y Transition

TL;DR

This paper reports the discovery of a unique, low-luminosity T/Y transition substellar binary with an unusually cold primary, providing insights into spectral features and evolution at the boundary of brown dwarf types.

Contribution

It presents the first detailed characterization of a tight, low-luminosity T/Y transition binary, including spectral decomposition and orbital constraints, advancing understanding of substellar atmospheres.

Findings

Binary components are less luminous than any known substellar binary.

The primary's temperature is about 345 K, cooler than other late-T dwarfs.

Orbital period constrained to less than 10 years, enabling future dynamical mass measurements.

Abstract

We have discovered that the brown dwarf WISEJ014656.66+423410.0 is a close binary (0.0875$\pm$0.0021 arcsec, 0.93$^{+0.12}_{-0.16}$ AU) from Keck laser guide star adaptive optics imaging. Our photometry for this system reveals that both components are less luminous than those in any known substellar binary. Combining a new integrated-light spectrum (T9p) and resolved YJH-band photometry from Keck allows us to perform spectral decomposition and assign component types of T9 and Y0. Many of the unusual features in the spectrum might be explained by high surface gravity: Y-band peak broadened to the blue; J-band peak broadened to the red; H-band peak shifted slightly to the red; and red Y-J colors. Interestingly, the very low component luminosities imply that the T9 primary is unexpectedly cold ($T_{\rm eff}$ = 345$\pm$45 K assuming an age of 10 Gyr), making it $\approx$100 K cooler than any other late-T dwarf and comparable to Y dwarfs. One intriguing explanation for this apparent discrepancy is that the J- and H-band spectral features that trigger the transition from T to Y spectral types are highly gravity-dependent. This can be tested directly in the very near future by orbit monitoring. We constrain the orbital period to be $\lesssim$10 yr by combining evolutionary model-based mass estimates for the components ($\approx$12$-$21 $M_{\rm Jup}$, 1$\sigma$ at 10 Gyr) with a statistical constraint on the semimajor axis ($\lesssim$1.3 AU). Such a period is shorter than any other known T/Y transition binary, meaning that WISEJ0146+4234AB will likely yield a dynamical mass within the next few years.