Nebular Spectroscopy of Kepler's Brightest Supernova

We present late-time ($\sim$240-260 days after peak brightness) optical photometry and nebular (+236 and +264 days) spectroscopy of SN 2018oh, the brightest Type Ia supernova (SN Ia) observed by the Kepler telescope. The Kepler/K2 30-minute cadence observations started days before explosion and continued past peak brightness. For several days after explosion, SN 2018oh had blue "excess" flux in addition to a normal SN rise. The flux excess can be explained by the interaction between the SN and a Roche-lobe filling non-degenerate companion star. Such a scenario should also strip material from the companion star, that would emit once the SN ejecta become optically thin, imprinting relatively narrow emission features in its nebular spectrum. We search our nebular spectra for signs of this interaction, including close examination of wavelengths of hydrogen and helium transitions, finding no significant narrow emission. We place upper limits on the luminosity of these features of $2.6,\ 2.9\ \mathrm{and}\ 2.1\times10^{37}\ \mathrm{erg\ s^{-1}}$ for H$\alpha$, He I $\lambda$5875, and He I $\lambda$6678, respectively. Assuming a simple model for the amount of swept-up material, we estimate upper mass limits for hydrogen of $5.4\times10^{-4}\ \mathrm{M_{\odot}}$ and helium of $4.7\times10^{-4}\ \mathrm{M_{\odot}}$. Such stringent limits are unexpected for the companion-interaction scenario consistent with the early data. No known model can explain the excess flux, its blue color, and the lack of late-time narrow emission features.