The process of remote plasma nitridation of ultrathin gate dielectrics was modeled and optimized using response surface models and physical-based models. Nitrogen profiles in SiO2 were obtained by time of flight secondary ion mass spectroscopy (SIMS), and models were developed to predict the total nitrogen dose, top surface (SiO2-polysilicon) nitrogen concentration, and bottom interface (Si-SiO2) nitrogen concentration. Models were also developed to predict the resulting decrease in effective electrical oxide thickness. These models were used for gate dielectric engineering by optimizing the nitrogen profile in SiO2 dielectric for (i) maximizing the nitrogen concentration at the top surface and the total integrated nitrogen dose (for prevention of boron penetration) and (ii) minimizing nitrogen concentration at the bottom Si-SiO2 interface (for minimum degradation to metal oxide semiconductor field effect transistor transconductance).