Correlation between x-ray reflectivity measurements and surface roughness of AXAF coated witness samples
One of the specifications used to polish the AXAF witness samples was that the rms surface roughness be <EQ 5 angstrom as measured by optical profilometry. This specification was chosen based on the cost of polishing and the necessity to keep scatter to a minimum. However, it is not necessarily the best indication of the expected performance of the soft x-ray reflectivity of the surfaces. In particular, the reflectivity data from the AXAF flight optic witness samples indicate sample to sample differences of a few percent which do not correlate with the optical profilometry results for these samples. Further investigations were carried out to measure rms surface roughness using atomic force microscopy (AFM). The differences shown by AFM surface roughness measurements correlates to differences found in reflectivity for these same samples. One-dimensional power spectral density data is presented from both AFM and WYKO measurements along with the reflectivity results at 8 keV for the AXAF witness samples. The results indicate that to obtain accurate prediction of x-ray performance it is necessary to look at the scanning probe metrology data provided by the AFM, in addition to the optical profilometry data.