A Comprehensive and Computationally Efficient Modeling Strategy for Simulation of Boron Ion Implantation Into Single‐crystal Silicon With Explicit Dose and Implant Angle Dependence

Abstract

A comprehensive and computationally efficient modeling strategy for the rapid and accurate simulation of implanted impurity distribution profiles in single‐crystal silicon has been developed. This modeling strategy exploits the advantages of both Monte Carlo simulation and semi‐empirical models by combining the two approaches in a complementary manner. The dual Pearson semi‐empirical model is used to accurately and efficiently model the dose and implant angle dependence of impurity profiles as well as the dependence on energy. This new comprehensive model allows convenient and accurate simulation of implanted boron distribution profiles in single‐crystal silicon as a function of dose, tilt angle, and rotation angle, in addition to ion energy, and it has been demonstrated by implementation in the process simulation code SUPREM III.

Ion beams in silicon processing and characterization

“…This analytic function approach is much more computationally efficient than physically based models, and has been demonstrated to accurately describe the impurity profile over four to five orders of magnitude variation in the concentration. [14][15][16][17][18] In addition, this approach has been demonstrated to model the wide range of different shaped profiles resulting from channeled implants, off-axis ͑mini-mum channeling͒ implants, different doses, and low mass ͑B͒ and high mass ͑As͒ implant species. [14][15][16][17][18] The development of accurate models of this type, however, requires extensive experimental data, especially in order to correctly understand and predict the detailed channeling dependence of the profile on the various parameters.…”

“…[14][15][16][17][18] In addition, this approach has been demonstrated to model the wide range of different shaped profiles resulting from channeled implants, off-axis ͑mini-mum channeling͒ implants, different doses, and low mass ͑B͒ and high mass ͑As͒ implant species. [14][15][16][17][18] The development of accurate models of this type, however, requires extensive experimental data, especially in order to correctly understand and predict the detailed channeling dependence of the profile on the various parameters. Towards this end a very large number of wafers has been implanted, and over 2000 SIMS depth profiles have been carefully measured in order to determine values, and an interpolation scheme for the nine parameters required by the Dual-Pearson model.…”

“…The Dual-Pearson fitting procedure 14 has been highly successful in the semi-empirical model development for B, BF 2 , As, and P implants [15][16][17][18] because of its ability to account so well for the dependence of both the randomly scattered and the channeled parts of the implanted profile on all of the key implant parameters. These models have explicit dependence on energy, dose, tilt angle, and rotation angle, and examples of their excellent predictive capabilities are illustrated in Fig.…”