Modeling SiC etching in C sub(2)F sub(6)/O sub(2) inductively coupled plasma using neural networks

The etching of SiC in a C sub(2)F sub(6)/O sub(2) inductively coupled plasma was mode led using neural networks. A 2 super(5) full factorial experiment was used to characterize the etch process. Etch rate was modeled with high prediction accuracy using neural networks with a linear function in the output layer. Effects of factor interactions on etch rate were determined by 3D plots while attempting to uncover underlying etch mechanisms.