The nature of mid-infrared excesses from hot dust around Sun-like stars

Aims. Studies of the debris disk phenomenon have shown that most systems are analogous to the Edgeworth-Kuiper Belt (EKB). However a rare subset of sun-like stars possess dust which lies, in contrast, in the terrestrial planet region. In this study we aim to determine how many sources with apparent mid-infrared excess are truly hosts of warm dust, and investigate where the dust in these systems must lie. Methods. We observed using ground-based mid-infrared imaging with TIMMI2, VISIR and MICHELLE a sample of FGK main sequence stars previously reported to have hot dust. A new modelling approach was developed to determine the constraints that can be set on the radial extent of excess emission in such observations by demonstrating how the detectability of a disk of a given flux as a fraction of the total flux from the system (F-disk/F-total) depends primarily on the ratio of disk radius to PSF width and on the uncertainty on that PSF width. Results. We confirm the presence of warm dust around three of the candidates; eta Corvi, HD 145263 and HD 202406. For eta Corvi modelling constrains the dust to lie in regions smaller than similar to 3.5 AU. The modelling constrains the dust to regions smaller than 80-100 AU for HD 145263 and HD 202406, with SED fitting suggesting the dust lies at a few tens of AU. Of two alternative models for the eta Corvi excess emission, we find that a model with one hot dust component at less than 0.'' 164 (<3 AU) (combined with the known submm dust population at similar to 150 AU) fits all the data better at the 2.6 sigma level than an alternative model with two populations of dust emitting in the mid-infrared: hot dust at less than 0.'' 19 (<3.5 AU) and a mid- temperature component at similar to 0.'' 66 (12 AU). We identify several systems which have a companion (HD 65277 and HD 79873) or background object (HD 53246, HD 123356 and HD 128400) responsible for their mid- infrared excess, and for three other systems we were able to rule out a point-like mid- infrared source near the star at the level of excess observed in lower resolution observations (HD 12039, HD 69830 and HD 191089). Conclusions. Hot dust sources are either young and possibly primordial or transitional in their emission, or have relatively small radius steady-state planetesimal belts, or they are old and luminous with transient emission. High resolution imaging can be used to constrain the location of the disk and help to discriminate between different models of disk emission. For some small disks, interferometry is needed to resolve the disk location.