A Candidate Low-mass Disk-eclipsing Binary in the ~316 Myr Open Cluster UPK 13

Abstract

UPK~13-c2 is a candidate member of the $\sim$316~Myr open cluster UPK~13 and was previously classified as a white dwarf + main-sequence (WD$+$MS) binary with 99.44\% confidence. We present multi-band photometric evidence that it is instead more plausibly a late-K/early-M binary with a misaligned circumbinary disk. The photometry reveals a flat-bottomed eclipse at $P=36.71$~days with an approximately achromatic $\sim$40\% flux decrement from the optical through $W1$, a reduced $W2$ depth, and a prominent mid-infrared excess at $W3/W4$. Two independent diagnostics strongly disfavor the WD$+$MS interpretation. First, the flat eclipse floor and 2.5-day ingress require complete occultation of an extended stellar component; a white dwarf would cross the disk edge in $\lesssim$2~hr and cannot naturally reproduce the observed multi-day trapezoid. Second, the difference spectrum is well fit by a single $\sim$4000~K thermal spectral energy distribution, favoring a late-K/early-M dwarf over a white dwarf as the occulted source. A decoupled SED decomposition yields a template-based late-K/early-M binary estimate of M1V+K9V with $M_{\rm tot}\approx1.4\,M_\odot$ and an overall systematic uncertainty of about 20\%. Under a sharp-edge eclipse model, forward modeling of the light curve favors an eccentric, spatially localized occulting structure. The light-curve morphology places UPK~13-c2 in the same geometric class as KH~15D and Bernhard-2. If cluster membership is confirmed, UPK~13-c2 may be the oldest known main-sequence disk-eclipsing binary.