Ion beam analytical methods (e.g. Rutherford Backscattering, RBS, Elastic Recoil Detection, ERD, or Nuclear Reaction Analysis, NRA) are widely used for quantitative determination of the depth distribution of elements. In porous samples however, where the diameter of the pores is a few tens of nm (i.e. unresolvable even by the most sophisticated microbeam), the measured depth distribution differs from the real one and this difference strongly depends on the pore structure.
The aim of the present work is to give a short review on the most significant effects observed so far in porous silicon samples by conventional RBS and resonant 16 O(a,a) and 15 N(p,ag) NRA measurements and point out their origin. It will be shown that due to the fluctuation of the material crossed by individual ions, the interface between the porous layer and the substrate smears out in the measured profiles and the resonance peak widens in resonant backscattering measurements. Additionally, the near surface yield in RBS spectra decreases. These effects are especially large if the incident ions fly along the pores of a ''columnar structure'' where the pores run parallel to each other. Furthermore, these effects vary with the incident angle in a way determined by the actual pore structure (porosity, average pore diameter and anisotropy of the pores). If the ions run parallel to the pores, the apparent thickness of the porous layer increases and the resonance peak sharpens, because the ions contributing to the measured depth distribution travel preferentially in the pore walls. All these effects can be well reproduced by Monte Carlo type simulations.