Surface roughness evaluation on mandrels and mirror shells for future X-ray telescopes

Abstract

More X-ray missions that will be operating in near future, like particular SIMBOL-X, e-Rosita, Con-X/HXT, SVOM/XIAO and Polar-X, will be based on focusing optics manufactured by means of the Ni electroforming replication technique. This production method has already been successfully exploited for SAX, XMM and Swift-XRT. Optical surfaces for X-ray reflection have to be as smooth as possible also at high spatial frequencies. Hence it will be crucial to take under control microroughness in order to reduce the scattering effects. A high rms microroughness would cause the degradation of the angular resolution and loss of effective area. Stringent requirements have therefore to be fixed for mirror shells surface roughness depending on the specific energy range investigated, and roughness evolution has to be carefully monitored during the subsequent steps of the mirror-shells realization. This means to study the roughness evolution in the chain mandrel, mirror shells, multilayer deposition and also the degradation of mandrel roughness following iterated replicas. Such a study allows inferring which phases of production are the major responsible of the roughness growth and could help to find solutions optimizing the involved processes. The exposed study is carried out in the context of the technological consolidation related to SIMBOL-X, along with a systematic metrological study of mandrels and mirror shells. To monitor the roughness increase following each replica, a multiinstrumental approach was adopted: microprofiles were analysed by means of their Power Spectral Density (PSD) in the spatial frequency range 1000-0.01 µm. This enables the direct comparison of roughness data taken with instruments characterized by different operative ranges of frequencies, and in particular optical interferometers and Atomic Force Microscopes. The performed analysis allowed us to set realistic specifications on the mandrel roughness to be achieved, and to suggest a limit for the maximum number of a replica a mandrel can undergo before being refurbished.