KIC 8410637: a 408-day period eclipsing binary containing a pulsating giant star

Context. Detached eclipsing binaries (dEBs) are ideal targets for accurately measuring the masses and radii of their component stars. If at least one of the stars has evolved off the main sequence (MS), the masses and radii give a strict constraint on the age of the stars. Several dEBs containing a bright K giant and a fainter MS star have been discovered by the Kepler satellite. The mass and radius of a red giant (RG) star can also be derived from its asteroseismic signal. The parameters determined in this way depend on stellar models and may contain systematic errors. It is important to validate the asteroseismically determined mass and radius with independent methods. This can be done when stars are members of stellar clusters or members of dEBs. Aims. This paper presents an analysis of the dEB system KIC 8410637, which consists of an RG and an MS star. The aim is to derive accurate masses and radii for both components and provide the foundation for a strong test of the asteroseismic method and the accuracy of the deduced mass, radius, and age. Methods. We analysed high-resolution, high-signal-to-noise spectra from three different spectrographs. We also calculated a fit to the Kepler light curve and used ground-based photometry to determine the flux ratios between the component stars in the BVRI passbands. Results. We measured the masses and radii of the stars in the dEB, and the classical parameters Teff, log g, and [Fe/H] from the spectra and ground-based photometry. The RG component of KIC 8410637 is most likely in the core helium-burning red clump phase of evolution and has an age and composition that are very similar to the stars in the open cluster NGC 6819. The mass of the RG in KIC 8410637 should therefore be similar to the mass of RGs in NGC 6819, thus lending support to the latest version of the asteroseismic scaling relations. This is the first direct measurement of both mass and radius for an RG to be compared with values for RGs from asteroseismic scaling relations thereby providing an accurate comparison. We find excellent agreement between log g values derived from the binary analysis and asteroseismic scaling relations. Conclusions. We have determined the masses and radii of the two stars in the binary accurately. A detailed asteroseismic analysis will be presented in a forthcoming paper, allowing an informative comparison between the parameters determined for the dEB and from asteroseismology