Mihiro YANAGIHARA, Takuya SASAKI,^*_[footnote] Minaji FURUDATE and Masaki YAMAMOTO

Research Institute for Scientific Measurements, Tohoku University, 2-1-1, Katahira, Aoba-ku, Sendai, 980-77 Japan

(Received June 5, 1995; Accepted January 31, 1996)

There exists serious inconsistency between the rms surface roughness determined from the Debye-Waller factor for the soft-x-ray reflectance analysis and that measured with an optical surface profiler. We have measured the surface profile of evaporated films using a scanning tunnelling microscope, and reproduce the profile with the Fourier components whose spatial wavelength is shorter than the coherence length of the incident soft x-rays in the reflectance measurement. The rms surface roughness derived from the high-pass filtered profile agrees well with that determined using the reflectance measurement. This result explains straightforwardly the photon-energy dependence of the surface roughness estimated by the soft-x-ray reflectance method.

Key words : surface roughness, reflectance, soft x-ray, coherence length, STM

^*Present address: Toshiba Corp., Nasu, Tochigi 329, Japan.