The final goal of diffusion studies is to determine the electrical active impurity profiles needed for the investigation of the device characteristics. Therefore, it is necessary to describe the movement of the dopants with proper theories. Diffusion theories have been developed from two major approaches: (1) the continuum theory, which describes the diffusion process by Fick's diffusion laws and (2) the atomistic theory, which involves point-defect interactions and rate equations for the exchange between the diffusion species [Fah89]. For silicon at intrinsic doping conditions the continuum theory is quite successful, the details of atomic interactions do not have to be known. With extended diffusion coefficient models it is even possible to apply the continuum theory to high doping conditions. Unfortunately, there are additional phenomena like precipitation or clustering effects, which are limiting the applicability of the simple Fickian approach for extrinsic doped regions [Sch71] [Nob83]. Various atomistic diffusion models including point-defects in equilibrium or non-equilibrium conditions have been published in the last decade to model the anomalous diffusion behavior in silicon [Tan85] [Bro87] [Nic89]. One of the important prerequisites for all these models is the initial setup of the point-defect concentrations. Most of the models are developed under surface oxidation conditions, because thereby the amount of injected point-defects is better known as e.g. the number of point-defects generated by ion implantation. Extensive measurement methods using dopant markers have been developed to verify the basic diffusion enhancement or retardation mechanisms for the different diffusion species [Fah83]. Atomistic diffusion modeling is still undergoing active development and will become more important for the next generation of devices as the dimensions are shrinking. Point-defects are not only side effects influencing the main dopant stream, they are also responsible for the electrical activation of impurity profiles.