Can we really go for direct exo-planet detection from the ground?

We are currently investigating the possibilities for a high-contrast, adaptive optics assisted instrument to be placed as a 2nd-generation instrument on ESO's VLT. This instrument will consist of an "extreme-ao" system capable of producing very high Strehl ratios, a contrast-enhancing device and two differential imaging detection systems. It will be designed to collect photons directly coming from the surface of substellar companions - ideally down to planetary masses - to bright, nearby stars and disentangle them from the stellar photons. We will present our current design study for such an instrument and discuss the various ways to tell stellar from companion photons. These ways include the use of polarimetric and/or spectroscopic information a's well as making use of knowledge about photon statistics. Results of our latest simulations regarding the instrument will be presented and the expected performance discussed. Derived from the simulated performance we will also give details about the expected science impact of the planet finder. This will comprise the chances of finding different types of exo-planets - notably the dilemma of going for hot planets marginally separated from their parent stars or cold, far-away plamnets delivering very little radiation, the scientific return of such detections and follow-up examinations, as well as other topics like star-formation, debris disks, and planetary nebulae where a high-resolution, high-contrast system will trigger new break-throughs.