Brian Martinick

Publication Date


Document Type



Distortion of printed photoresist patterns exists in all photolithographic tools and can be generated by numerous factors. These noncorrectable overlay errors are the direct result of lens imperfections, machine inaccuracy, and reticle error. Any one of these factors can have extreme effects on pattern placement and quality. This study involved characterizing this anomaly and was the first of its kind at the Semiconductor & Microsystems Fabrication Laboratory (SMFL). Incorporating a unique test reticle, crosshairs were printed on silicon wafers. These features were measured via Manufacturing Electron Beam Exposure System (MEBES) market analysis to reveal any pattern migration. This analysis involves passing the electron stream over the beams of the crosshairs. The resulting signature from electron backscattering showed small movements of the pattern. Mathematical modeling of the raw data extracts correctable errors leaving a residual distortion map. These maps can be used as a figure of merit for the amount of pattern placement distortion within the photolithographic tool.

Included in

Engineering Commons