Using a pulsed-plasma enhanced chemical vapor deposition, silicon nitride (SiN) films were deposited in SiH(4)-NH(3) plasma at room-temperature. The bias power and duty ratio were varied in the range of 30-90W and 20-80%, respectively. Ion energy diagnostics was conducted and correlated with the SiN deposition rate. The diagnostic parameters examined include a high ion energy (E(h)), a low ion energy (E(l)), a high ion energy flux (N(h)), a low ion energy flux (N(l)), and a ratio of N(h)-N(l). As duty ratio was varied at 30 W, the deposition rate strongly depended on the E(h) (or El). For the same duty ratio variation at other high powers, it was highly correlated with the N(h)/N(l). For the variations in the bias power at a given duty ratio, the deposition rate was closely related to the ion energy flux in the power range of 30-70 W. For all the variations, the deposition rate was varied between 234 and 305 angstrom/min. Meanwhile, the experimental data were modeled by using a neural network. The R(2) value of neural network model is about 99.8% and the model was utilized to examine the effect of diagnostic variables for optimization of refractive index. (C) 2010 Elsevier B. V. All rights reserved.