Improvement of the plasma‐induced defect generation model in Si substrates and the optimization design framework

Defect generation in Si substrates during plasma processing—plasma‐induced physical damage (PPD)—has been principally modeled on the basis of the projection range of incident ions. In this paper, first, the ion dose ( D ion ) dependence of the thickness of a damaged layer ( d dam ) was implemented in the conventional PPD range model. An improved PPD range model was proposed to describe the dose evolution of d dam in addition to the ion energy ( E ion ) dependence of d dam . Then, the model prediction results were compared with the experimentally obtained d dam after Ar plasma exposure. A good agreement was found, which implied that plasma process designs should be performed by taking into account both the D ion and E ion dependences of d dam . Finally, a methodology of plasma process design based on the improved PPD range model was presented. The methodology was regarded as an optimization problem under constraints imposed by PPD criteria. Two scenarios—optimization problems—were demonstrated, where the etch rate was maximized under the following two constraints: one is by considering the acceptable d dam , and the other, the recessed Si depth and latent defect density with its trade‐off relation. The results suggested that minimizing E ion rather than lowering the flux of incident ions is essential for the above scenarios.