Revisiting Kepler Transiting Systems: Unvetting Planets and Constraining Relationships among Harmonics in Phase Curves

Abstract: Space-based photometric missions widely use statistical validation tools for vetting transiting planetary candidates, particularly when other traditional methods of planet confirmation are unviable. In this paper, we refute the planetary nature of three previously validated planets -- Kepler-854~b, Kepler-840~b, and Kepler-699~b -- and possibly a fourth, Kepler-747~b, using updated stellar parameters from Gaia and phase-curve analysis. In all four cases, the inferred physical radii rule out their planetary nature given the stellar radiation the companions receive. For Kepler-854~b, the mass derived from the host star's ellipsoidal variation, which had not been part of the original vetting procedure, similarly points to a non-planetary value. To contextualize our understanding of the phase curve for stellar mass companions in particular and extend our understanding of high-order harmonics, we examine Kepler eclipsing binaries with periods between 1.5 and 10 days. Using a sample of 20 systems, we report a strong power-law relation between the second cosine harmonic of the phase-curve signal and the higher cosine harmonics, which supports the hypothesis that those signals arise from the tidal interaction between the binary components. We find that the ratio between the second and third-harmonic amplitudes is $2.24 \pm 0.48$, in good agreement with the expected value of 2.4 from the classical formalism for the ellipsoidal distortion.