Combining dynamic modelling codes with medium energy ion scattering measurements to characterise plasma doping

Plasma doping ion implantation (PLAD) is becoming increasingly important in the manufacture of advanced semiconductor device structures but a fundamental understanding of PLAD is complicated. A model of PLAD into planar substrates has been constructed using the one dimensional computer code TRIDYN to predict collision cascades and hence substrate
compositional changes during implantation. Medium Energy Ion Scattering (MEIS)
measurements of dopant profiles in PLAD processed samples were used to calibrate the input ion and neutral fluxes to the model. Rules could then be proposed for how post implant profiles should be modified by a cleaning step. This learning was applied to a three dimensional TRI3DYN based model for PLAD implants into FinFET like structures. Comparison of the model to dopant profile measurements made by time of flight (TOF)-MEIS revealed the angular distributions of neutral species and doping mechanisms acting in three dimensional structures.