Highly Sensitive Large Area Bolometers For Scintillation Studies Below 100 Mk (From Near Ir To Soft X-Rays).

At very low temperature large area bolometers may show a better sensitivity than photomultipliers or semiconductor diodes, while allowing fluorescence measurements of cool targets with no window, no infrared background. good optical couplings and a flat response on a large absorption bandpass. The optical absorber of these composite bolometers can be matched to the desired bandpass. Here we present the design, the performances and calibration tests of a new generation of large area (5 cm(2)) optical bolometers with a pure germanium disk absorbing on a wide spectral band from near-IR to X-rays. Performances obtained at 25 mK are very promising : Noise Equivalent Power as low as 4x10(-17) W/rootHZ in the photometry mode, energy threshold about 50 eV in the single photon detection mode, and time constant tau-3 ms. These detectors of low mass (0.25 g) have been recently successfully used for detecting the fluorescence emitted by much more massive bolometers, having for example a BGO (92 g), or a CaWO(4) (54 g) target. The simultaneous detection of heat and light in these << double bolometers>> permits the identification of each event in the massive target (alpha decay, gamma or cosmic ray interaction, neutron recoil...). Thanks to the consecutive excellent subtraction of the radioactive and cosmic rays background, it is a powerful tool developed by several groups for fundamental resarch : study of very rare decays of atoms, measurement of internal very low radioactivity content in single crystals. direct detection of dark matter recoils in massive fluorescence targets, detection of solar neutrino fluorescence events in liquid (4)He... Recently obtained results which support this new promising field are reminded: the first detection of the rare alpha decay of (209)Bi, and new scintillation data on Al(2)O(3) (sapphire), LiF or TeO(2) at 20 mK. We discuss the ultimate performances at 12 mK of the optical bolometers as a function of their area, as well as the optimisation of their absorbing part to the desired bandpass, and finally, we estimate achievable improvements of our current technology.