AVS 46th International Symposium
    Applied Surface Science Division Thursday Sessions
       Session AS-ThM

Paper AS-ThM8
A 300mm SAM, with EDX and FIB for Full Wafer Defect and Thin Film Characterization

Thursday, October 28, 1999, 10:40 am, Room 6A

Session: Real World Surface Analysis
Presenter: Y. Uritsky, Applied Materials, Inc.
Authors: Y. Uritsky, Applied Materials, Inc.
C.R. Brundle, Applied materials, Inc.
Correspondent: Click to Email
As design rules shrink and thin film stacks get thinner, the semiconductor equipment manufacturing industry is forced to move to more sophisticated approaches for its particle defect and thin film characterization needs, including surface and thin film analysis. In the past we have occasionally supplemented our full wafer (200 mm) SEM/EDX small particle analysis work by SAM, using small cut up pieces of the wafer. We have now installed the first full 300 mm wafer SAM (Smart 300 from PHI), on which we also have traditional EDX and also FIB. The capabilities of this instrument are briefly described here and examples are given of its use to a) find small defects based on navigation from light scattering files, b) comparatively analyze small particles using Auger, EDX, and FIB sectioning, and c) profile films to examine interfaces. With respect to a) above, since light scattering files are usually quite inaccurate (the predicted coordinates can easily be in error by 100's of µm's), it can be very time-consuming, if not impossible, to re-find very small particles (0.1 µm) with low SEM or Auger contrast. Use of a 300 mm capable dark field optical bench/wafer marker (MicroMark 5000) system to update the particle coordinates with +/- 5 µm) accuracy and/or to create laser-made fiducial marks, can accelerate the subsequent particle re-detection/analysis SAM procedure by a factor of 10. Another particular concern to us, since it is a widely used element in semiconductor processing and is very aggressive, is the relative ability of Auger, EDX, and FIB sectioning Auger and EDX, to reliably detect F as opposed to removing it under the probe beam. This is discussed.