QUANTITATIVE COMPARISON OF HIGH-RESOLUTION TEM IMAGES WITH IMAGE SIMULATIONS

The quantitative comparison of a simulation with an experimental image can be divided into three parts: the comparison of their mean intensity, contrast and pattern. The theoretical basis for such an approach has been discussed elsewhere and here we apply it to the evaluation of a typical through-focal series of images of a representative III-V specimen. Particular attention will be payed to the independent determination of the experimental parameters. A quantitative assessment of the effect on the matching process of noise and errors in the simulation parameters will also be made. The application of the approach is then used to demonstrate the inadequacy of current image simulations and the reasons for the mismatch which is typically found between the experimental and simulated image intensities are discussed.