ABSTRACT
The International Technology Roadmap for Semiconductors ranks dopant profiling as one of the most difficult challenges for analysis of semiconductors. Dopant mapping in the scanning electron microscope (SEM) has the potential to provide a solution. This technique has not yet found widespread application, however, mainly due to the lack of a comprehensive theoretical model, uncertain quantification, and its inability to differentiate doping levels in n-type silicon. Although a Monte Carlo model was recently published that closely matched experimental data obtained in p-doped silicon to data obtained from the theoretical model, a large discrepancy between experimental data obtained for n-type silicon was found. Here we present a Monte Carlo model that provides close matches between experimental and calculated data in both n- and p-type silicon, paving the way for a widespread application of SEM dopant contrast.
