Equipment fault detection with fitted wafer surfaces

This paper describes a new methodology for equipment fault detection. This methodology consists of fitting a thin-plate spline to post-process spatial data in order to construct a virtual wafer surface. The virtual wafer surface is then compared to an established baseline process surface, and the resulting spatial signature is used to detect equipment faults. Statistical distributional studies of signature metrics using a parametric bootstrapping technique provide the justification of determining the significance of the signature. Data collected from a Rapid Thermal Chemical Vapor Deposition (RTCVD) process is used to illustrate the procedures. This method detected equipment faults for all 11 wafers that were subjected to induced equipment faults.