Zach J. LewisBrett H. AndrewsRachel BezansonMichael MasedaEric F. BellRomeel DavéFrancesco D’EugenioMarijn FranxAnna GallazziAnna de GraaffYasha KaushalAngelos NersesianJeffrey A. NewmanArjen van der WelPo-Feng Wu2024-10-292024-10-292024-03-01https://scholars.lib.ntu.edu.tw/handle/123456789/722593The massive end of the gas-phase mass-metallicity relation (MZR) is a sensitive probe of active galactic nuclei (AGN) feedback that is a crucial but highly uncertain component of galaxy evolution models. In this paper, we extend the z ∼ 0.7 MZR by ∼0.5 dex up to log (M ⋆/M ⊙) ∼ 11.1. We use extremely deep VLT VIMOS spectra from the Large Early Galaxy Astrophysics Census (LEGA-C) survey to measure metallicities for 145 galaxies. The LEGA-C MZR matches the normalization of the z ∼ 0.8 DEEP2 MZR where they overlap, so we combine the two to create an MZR spanning from 9.3 to 11.1 log (M ⋆/M ⊙). The LEGA-C+DEEP2 MZR at z ∼ 0.7 is offset to slightly lower metallicities (0.05-0.13 dex) than the z ∼ 0 MZR, but it otherwise mirrors the established power-law rise at low/intermediate stellar masses and asymptotic flattening at high stellar masses. We compare the LEGA-C+DEEP2 MZR to the MZR from two cosmological simulations (IllustrisTNG and SIMBA), which predict qualitatively different metallicity trends for high-mass galaxies. This comparison highlights that our extended MZR provides a crucial observational constraint for galaxy evolution models in a mass regime where the MZR is very sensitive to choices about the implementation of AGN feedback.journal article10.3847/1538-4357/ad250c