The use of X-ray photoelectron spectroscopy (XPS) as a technique for n
on-destructive depth profiling of technical samples is often hindered
by their roughness. In this paper, we elaborate on earlier work where
we reported on the apparent existence of a 'magic' angle for the deter
mination of the thickness of uniform overlayers on rough substrates, S
imple calculations for fully three-dimensional model rough surfaces st
rongly suggest that the thickness of an overlayer on a rough substrate
can be determined accurately without taking the roughness into accoun
t in the analysis of the XPS intensities, the neglect of roughness eff
ects leading to an average error less than 10% in the determined thick
ness. (C) 1997 Elsevier Science B.V.