A Statistical Approach for Evaluating the Spatial Distribution and Local Atomic Environment of Dopants Using Atom Probe Tomography

An approach to characterize dopant spatial inhomogeneity at the nanoscale along with its local chemical environment has been developed using atom probe tomography. We achieve this through the combination of a local composition analysis using the nearest-neighbor atoms combined with a nonparametric Kolmogorov-Smirnov or Anderson-Darling statistical test. Using an in situ highly boron-doped silicon germanium layer we demonstrate that all three elements have an inhomogeneous distribution. Moreover, by subdividing the local dopant composition distribution, a correlation between the boron doping level and variations in the surrounding matrix composition was determined. These atomic-scale measurements provide new experimental insights into the dopant incorporation behavior in technologically relevant semiconductors and its relationship to the epitaxial growth processes.