Absolute dimensions of solar-type eclipsing binaries. EF Aquarii: a G0 test for stellar evolution models

Recent studies have shown that stellar chromospheric activity, and its effect on convective energy transport in the envelope, is most likely the cause of significant radius and temperature discrepancies between theoretical evolution models and observations. We aim to determine absolute dimensions and abundances for the solar-type detached eclipsing binary EF Aqr, and to perform a detailed comparison with results from recent stellar evolutionary models. uvby-beta standard photometry was obtained with the Stromgren Automatic Telescope. The broadening function formalism was applied on spectra observed with HERMES at the Mercator telescope in La Palma, to obtain radial velocity curves. Masses and radii with a precision of 0.6% and 1.0% respectively have been established for both components of EF Aqr. The active 0.956 M_sol secondary shows star spots and strong Ca II H and K emission lines. The 1.224 M_sol primary shows signs of activity as well, but at a lower level. An [Fe/H] abundance of 0.00+-0.10 is derived with similar abundances for Si, Ca, Sc, Ti, V, Cr, Co, and Ni. Solar calibrated evolutionary models such as Yonsei-Yale, Victoria-Regina and BaSTI isochrones and evolutionary tracks are unable to reproduce EF Aqr, especially for the secondary, which is 9% larger and 400 K cooler than predicted. Models adopting significantly lower mixing length parameters l/H_p remove these discrepancies, as seen in other solar type binaries. For the observed metallicity, Granada models with a mixing length of l/H_p=1.30 (primary) and 1.05 (secondary) reproduce both components at a common age of 1.5+-0.6 Gyr. Observations of EF Aqr suggests that magnetic activity, and its effect on envelope convection, is likely to be the cause of discrepancies in both radius and temperature, which can be removed by adjusting the mixing length parameter of the models downwards.