Artificial intelligence approach to analyze SIMS profiles of 11B, 31P and 75As in n- and p-type silicon substrates: experimental investigation

Abstract

Abstract In this work, we report an effective approach based on an artificial intelligence technique to investigate the secondary ions mass spectroscopy (SIMS) profiles of boron, phosphorus and arsenic ions. Those dopant ions were implanted into n- and p-type (100) Silicon substrate using the ion implantation technique with energy of 100 and 180 keV. Annealing treatment was conducted at various temperatures ranging from 900 to 1030 °C for 30 min. The doping profile parameters such as the activation energy, diffusion coefficient, junction depth, implant dose, projected range and standard deviation were determined using particle swarm optimization (PSO) algorithm. The efficiency of this strategy was experimentally verified by the fitting between both real measured SIMS profile and predicted ones. In addition, a set of simulated doping profiles was generated for different annealing time to prove the ability of this approach to accurately estimate the above parameters even when changing the experimental conditions.