Correlations in backscattering experiments - investigation by means of a new developed 180 degree backscattering facility

We describe the development of a facility that allows us to measure projectiles scattered in an arbitrarily small solid angle around 180 degree. A combination of electric and magnetic fields deflects the backscattered projectiles out of the incoming beam in a horizontal plane. A 300 mm2 PIPS-detector detects the spatially dispersed backscattered particles. To keep the detector at constant low temperature, we use a mixture of liquid and solid CS2 which coexist at 162 K. As the projectiles are also dispersed according to their charge state we are able to distinguish between, e.g., singly and doubly charged He-projectiles. By moving the detector up and down, we can continuously vary the backscattering angle between true 180 degree and 178.5 degree. Such a device is ideally suited for investigating single crystals by uniaxial channelling-blocking. Prior to these investigations, we study the enhanced backscattering yield found near 180 degree. We show that this effect can be related to a correlation of the flux distribution and of the local solid angle of the detector within the target. To put it differently, most of the projectiles are guided on their way into the target and out of the target by soft distant collisions with the same target atoms. For the first time, we show that in contrast to theoretical suppositions these projectiles are subject to random stopping power. To determine the magnitude of the enhancement undistorted by the detector system we were confronted with resolution correction. We review the problems associated with resolution correction and discuss some of the most promising procedures to solve them (parametrization, variable binning, modified Landweber iterations, and the use of mollifiers). We find that in our case only parametrization of the problem yields fairly satisfying results. Finally, we report first results obtained with single crystals. (author)