The dependence of the CL‐signal on the diffusion length L, the absorption coefficient α, the dead‐layer thickness zT, and the surface recombination velocity vs is studied in detail. The existence of a maximum in the CL‐signal in dependence on the beam voltage Ub is discussed. The position of this maximum in dependence on α and L is evaluated. The quantitative interpretation of experiments on n‐GaAs is explained by means of the theoretical model developed here.

EBIC contrast measurements of individual surface‐parallel dislocations lying in the p‐region of a planar silicon diode and TEM analysis are correlated in order to study the EBIC contrast behaviour of dislocations of different type. The experimental results are in good agreement with theoretical predictions. In particular, it is possible to classify the dislocations with respect to their type by the real strength of the recombination activity. The surface recombination velocity and the minority carrier diffusion length of the bulk are determined using a simple theoretical model.

It is shown that it is possible to describe the currents induced by electron and laser beams (EBIC, LBIC), and the corresponding luminescence signals (CL, PL) in a unified formalism. This concept has certain advantages in the case of combined investigations with these methods. This formalism is used for the discussion of the information depth of the photoluminescence signal.

A new procedure of calculating the electron-beam-induced current contrast of a crystal defect is presented which yields an exact solution for the contrast of a surface-perpendicular dislocation both in a Schottky diode and in a planar p-n diode. The calculations are based on the bulk recombination model describing a dislocation as a cylindrical inhomogeneity (with radius rd) where the total lifetime (τ′) is lower than that outside the cylinder. Evaluations of the exact solution for the dependence of the contrast on the electron range show that (i) the input parameters of the model, rd and τ′, can be consistently determined from the characteristic of the contrast versus the electron range, and (ii) the traditionally used first-order recombination strength is not a generally valid measure of the recombination activity of a surface-perpendicular dislocation.

No abstract