Optics: measuring and testing – Inspection of flaws or impurities – Surface condition
Reexamination Certificate
2000-11-28
2003-02-04
Stafira, Michael P. (Department: 2877)
Optics: measuring and testing
Inspection of flaws or impurities
Surface condition
C356S394000
Reexamination Certificate
active
06515742
ABSTRACT:
BACKGROUND OF THE INVENTION
The invention relates generally to the identification and classification of surface defects on semiconductor wafers. In particular, the invention relates to a method and system for detecting and characterizing defects on silicon wafers with epitaxially grown films.
In the electronics industry, it is desirable to use defect free silicon wafers to manufacture electronic devices. It is now known to use surface inspection systems and equipment to inspect wafer surfaces to detect the presence of defects. One example of such an inspection system is the SP 1 wafer inspection system, available from KLA-Tencor.
Williams et al. describe one way to inspect silicon wafers with epitaxially grown films using the SP1 system. See Randy Williams et al. “Evaluation of the Yield Impact of Epitaxial Defects on Advanced Semiconductor Technologies,” IEEE 1999 International Symposium on Semiconductor Manufacturing at 107-110, the entire disclosure of which is incorporated herein by reference. The SP 1 system sequentially illuminates the entire surface of a silicon wafer with a laser beam directed normal to the wafer surface. Typically, if the portion of the surface being illuminated is defect free, the light is reflected directly back and out of the system. Defects on the wafer surface, however, cause the beam to scatter and reflect at one or more angles. Any portion of the beam that is scattered back at an angle between 25 and 70 degrees is collected in a first photomultiplier tube (PMT
1
), referred to also as the “wide channel.” Any portion of the beam that is scattered back at an angle between 5 and 20 degrees is collected in a second photomultiplier tube (PMT
2
), referred to also as the “narrow channel.” Thus, the SP 1 collects information in two channels. Each channel is calibrated to a reference standard by using polystyrene latex (PSL) spheres/particles of known sizes, sometimes referred to as “PSL calibration.” The intensity of the collected light in each channel provides a basis for estimating defect sizes by comparing the collected intensity with the calibrated intensity. Therefore, each channel provides an independent defect size estimation. The general operation of the SP 1 is described below with respect to
FIGS. 1
,
2
, and
3
.
As described by Williams et al., an algorithm may be used to compare the ratio of the defect sizes detected in each of the two channels. See Williams et al. at 108. Further, Williams et al. note that because both channels are similarly calibrated, particle defects result in a ratio approximately equal to 1. In other words, as shown by equation 1:
narrow channel size/wide channel size=1 (for particle defects) [1]
the ratio of the estimated defect size as detected by the wide channel to the estimated defect size as detected by narrow channel is expected to be roughly 1. This results from the fact that the geometry of particle defects typically approximates a sphere—scattering light equally in both the narrow and wide directions.
Williams et al. also disclose that for epitaxial defects, the size estimations generated in the narrow and wide channels will typically differ because geometry of typical epitaxial defects does not approximate a sphere. As such, light scattered off of such defects will not ordinarily result in a size ratio of 1. Accordingly, a simple ratio may be used to classify detected defects as being either particles or non-particle defects. See Williams et al. at 108-09.
One important shortcoming of the simple ratio-based process disclosed by Williams is that it is limited to differentiating between particle and non-particle defects. In effect, Williams et al. classify all non-particle defects as epitaxial faults. In addition to particle defects and epitaxial faults, however, silicon wafers having epitaxially grown films may also have substrate related defects (e.g., defects located in the substrate that are not necessarily caused by the crystal pulling process). Thus, a wafer that is inspected using the inspection process described by Williams et al. may exhibit a significant defect that is not properly classified by that system and method.
In order to improve efficiency and throughput, it is important to know not only the presence of defects, but also the source and type of the defect. Cost conscious semiconductor wafer manufacturing requires attention to such information. See John Baliga, Advanced Process Control: Soon to be a Must,
Semiconductor International,
July 1999, available at http://www.semiconductor.net/semiconductor/issues/issues/1999/jul99/docs/feature1.asp, the entire disclosure of which is incorporated herein by reference. For example, particle defects may be removed by cleaning or re-cleaning the wafer. Defects associated with the epitaxial process may be the result of the recipe used. Thus, knowledge of the presence, size, and location of epitaxial defects can be used to improve and/or adjust the recipe to eliminate or reduce such defects. Finally, information regarding substrate related defects may be used to improve and/or adjust the crystal growth or wafering processes. It is therefore seen as to desirable to provide a semiconductor wafer inspection system that detects defects on the surface of the wafer, and classifies those defects into at least three categories—particle, epitaxially-related, or substrate related.
For these reasons, an improved method and system for identifying and classifying defects associated with semiconductor wafers is desired. Such a method and system improves manufacturing efficiencies by providing an automated source of data that may be used to identify defects and defect trend information.
SUMMARY OF THE INVENTION
The invention meets the above needs and overcomes the deficiencies of the prior art by providing an improved method and system for identifying and classifying defects associated with semiconductor wafers having epitaxial grown films. This is accomplished by identifying defect size information in a plurality of channels and comparing the identified defect size information to two or more curves. The curves define regions in which certain defect types tend to fall, thereby allowing defects to be classified on the basis of the size information obtained in the plurality of channels. The advantages of the present invention may be realized with minor software modifications to presently available inspection equipment.
Briefly described, a method of inspecting a workpiece surface embodying aspects of the invention includes providing a workpiece having a surface to be inspected. The workpiece surface is illuminated with an energy beam. The energy beam illuminates the workpiece surface at an angle of incidence and reflects off of the workpiece surface. A first collector collects a first portion of the energy beam reflected off of the workpiece surface at a first collection angle. A second collector collects a second portion of the energy beam reflected off of the workpiece surface at a second collection angle. A first size characterization of a defect associated with the workpiece surface is determined. A second size characterization of the defect associated with the workpiece surface is determined. A defect type characterization is determined by comparing the first and second size characterizations to a plurality of functions. The functions are selected to identify at least three defect types.
Another embodiment of the invention includes a method for inspecting a workpiece having a workpiece surface. The method includes placing the workpiece in an inspection chamber. The inspection chamber comprises a light generator for generating a light beam, a first beam collector, and a second beam collector. The light beam illuminates the surface of the workpiece at an angle of incidence. The light beam is reflected off of a defect associated with the surface of the workpiece. A first beam collector collects a portion of the reflected light beam that is reflected in a first collection angle. A second beam collector collects a portion of the reflec
MEMC Electronic Materials , Inc.
Senniger Powers Leavitt & Roedel
Stafira Michael P.
LandOfFree
Defect classification using scattered light intensities does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Defect classification using scattered light intensities, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Defect classification using scattered light intensities will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3144139