Composition quantification of electron-transparent samples by backscattered electron imaging in scanning electron microscopy

The contrast of backscattered electron (BSE)images in scanning electron microscopy (SEM) depends on material parameters which can be exploited for composition quantification if some information on the material system is available. As an example, the In-concentration in thin InxGa1-xAs layers embedded in a GaAs matrix is analyzed in this work. The spatial resolution of the technique is improved by using thin electron-transparent specimens instead of bulk samples. Although the BSEs are detected in a comparably small angular range by an annular semiconductor detector, the image intensity can be evaluated to determine the composition and local thickness of the specimen. The measured intensities are calibrated within one single image to eliminate the influence of the detection and amplification system. Quantification is performed by comparison of experimental and calculated data. Instead of using time-consuming Monte-Carlo simulations, an analytical model is applied for BSE-intensity calculations which considers single electron scattering and electron diffusion.