Low open-area endpoint detection using a PCA-based T2statistic and Q statistic on optical emission spectroscopy measurements

摘要

Examines an approach for automatically identifying endpoint (thencompletion in etch of a thin film) during plasma etching of low opennarea wafers. Because many end-pointing techniques use a few manuallynselected wavelengths or simply time the etch, the resulting endpointndetection determination may only be valid for a very short number ofnruns before process drift and noise render them ineffective. Onlynrecently have researchers begun to examine methods to automaticallynselect and weight spectral channels for estimation and diagnosis ofnprocess behavior. This paper will explore the use of principal componentnanalysis (PCA)-based T2 formulation to filter out noisynspectral channels and characterize spectral variation of opticalnemission spectroscopy (OES) correlated with endpoint. This approach isnapplied and demonstrated for patterned contact and via etching usingndigital semiconductor's CMOS6 (0.35-Μm) production process