Radiant energy – Photocells; circuits and apparatus – Photocell controls its own optical systems
Reexamination Certificate
1998-05-05
2001-06-19
Allen, Stephone B. (Department: 2878)
Radiant energy
Photocells; circuits and apparatus
Photocell controls its own optical systems
C250S234000, C359S385000
Reexamination Certificate
active
06248988
ABSTRACT:
TECHNICAL FIELD
The present invention relates to scanning optical microscope systems of both the conventional and confocal type, and specifically to the architecture of both the illumination and detection optics of such systems. The invention furthermore addresses the efficiency of such systems and their high speed inspection and 2- or 3-dimensional viewing and imaging capabilities.
BACKGROUND ART
Microscope systems may be used for automated inspection of patterned objects, such as photomasks or silicon wafers, for detecting defects or particles. Key requirements for automated inspection systems include high resolution, high contrast imaging for defect detection sensitivity, and high speed or throughput. The ever shrinking (now submicron) critical dimension on these objects require higher resolution to detect the critical defects or particles that are a fraction of the feature size. Shorter wavelengths and higher numerical aperture optics are needed to produce the required resolution. These shrinking feature sizes also result in a strong increase in the number of pixels per photomask that need to be processed in a short time, thus driving up the data rates at which the images have to be acquired. As a result, different systems have evolved for high speed photomask inspection, defect review, aerial imaging (i.e., the exact matching of imaging parameters used in a wafer stepper) and special applications like phase defect detection.
Imaging systems can be classified as projection systems or as scanners. Projection system architectures for high speed photomask imaging with lamp-based area illumination and either CCD or TDI (time delay integration) array detectors are limited both by the relatively low brightness of available light sources and the aberrations of the high numerical aperture (N.A.) wide band and wide field optics needed for high resolution imaging. If a laser illumination system is used, the coherence of the laser produces speckle and contributes image noise which needs to be controlled. Both lamp and laser illumination based projection systems are subject to the data rate, signal-to-noise ratio, modulation transfer function, and quantum efficiency limitations of the array detector. The object being imaged is either stepped across the field of the object until the image is acquired or moved continuously using a TDI array detector, which sacrifices some resolution due to image smearing. Moreover, these projector-type optical imaging systems make it both costly and difficult to derive multiple types of images (e.g., transmitted and reflected, brightfield and darkfield, phase contrast, and differential interference contrast (Nomarski) imaging) simultaneously from the same object by means of multiple detector arrays and electronic processing, since each new signal requires an additional expensive array detector and the maintaining of perfect pixel alignment of the multiple array detectors in the presence of vibrations and lens distortions. Laser scanning architectures using a single spot for high speed imaging are limited by the achievable data rates of acousto-optic scanners, thermal damage to the objects and the availability and reliability of cw lasers at short wavelengths.
Confocal microscope systems provide higher resolution than conventional microscopes. In U.S. Pat. No. 3,013,467, Minsky describes the basic principle of confocal microscopy. The confocal microscope includes a point light source which is focused as a spot onto an object for illumination. The emerging (e.g., reflected or transmitted) light from the object is, in turn, imaged onto a point detector and the image is obtained by scanning. The higher resolution is a result of the point source illumination as compared to uniform illumination in a conventional microscope. In addition to its greater resolution than conventional microscopes, the confocal microscope provides depth discrimination that can be used to obtain three-dimensional images from two-dimensional slices by superposition of the image data obtained at different focal depths. This makes confocal microscopy useful, not only for inspecting photomasks and other patterned objects for defects, but also for imaging biological and other low contrast or light scattering three-dimensional objects, e.g. to observe living cells. Confocal illumination and imaging requires scanning of each point in the field in order to construct a viewable image, whereby the illumination and imaging pinholes are aligned in conjugate positions and maintained in such alignment. Most confocal microscopes use sequential acquisition of single image points in a self-aligned optical scheme in order to avoid the inherent alignment problems. For example, Minsky uses a stage scanning system that moves the object relative to a fixed illumination and imaged point, while producing a synchronized identical scanning pattern on a display device that receives the point detection signal. Alternatively, U.S. Pat. No. 4,802,748 to McCarthy et al. teaches the use of a rotating pinhole (Nipkow) disk for the source and exit pinholes in order to concurrently scan the illumination and imaged spot over the object field. U.S. Pat. No. 4,927,254 to Kino et al. also uses rotating disks. Alternatively, one could raster scan a laser beam with rotating mirrors, as in U.S. Pat. No. 5,065,008 to Hakamata et al., U.S. Pat. No. 5,081,349 to Iwasaki, and as described by D. L. Dickensheets, in SPIE, vol. 2655, pages 79-85 (1996). One might also use an acousto-optic cell for spot scanning of the object. However, each of these single spot confocal scanning schemes is limited by the achievable scanning speed and possible scan aberrations. Moreover, such confocal systems either have low illumination brightness, which lengthens image acquisition times, or are limited by the potential thermal damage to an object caused by higher brightness laser sources. Laser scanning architectures are also limited by the availability of CW lasers at short (UV) wavelengths.
The above-described projection systems and high speed scanners typically do not have aerial imaging capabilities (i.e., exactly matching the imaging parameters of the wafer stepper) and defect review, and aerial imaging microscopes do not have high data rates. High resolution imaging from confocal microscopes is not available at the high data rates and high signal-to-noise ratios required for photomask inspection.
One improvement in confocal microscope systems is taught by Krause in U.S. Pat. No. 5,587,832. Krause uses an electronically addressed spatial light modulator, such as an electrostatic microshutter array or a digital mirror device (DMD) with an array of mirror-coated shutters, to produce sequential patterns of multiple, simultaneously formed, illumination spots on the object. Krause further provides means, whether it be the same spatial light modulator, or an active pixel sensor (APS) type of array detector, for masking selected pixels of an image detector array in correlation with the illumination spot pattern so that different subsets of the image corresponding to the patterns are sequentially sensed by the detector array. An image processor stores and combines the signals to form a complete frame. One limitation of such a system is provided by the digital mirror device (DMD) or other spatial light modulator. The need to produce acceptable yields of DMDs with a large number of pixel elements without a single defect limits the practical size of a DMD to about 106 pixels, each about 10 to 15 &mgr;m in size. Detector efficiency limitations require a frame time of at least 10-20 &mgr;sec, while both DMD and detector mask addressing speeds further limit the achievable data rate. Nevertheless, the multispot illumination and detection described by Krause, in principle offers higher image acquisition efficiency than single spot scanning. In order to avoid crosstalk noise in the confocal image, the illumination and detection patterns need to be selected properly. Also, the shutter arrays have to be closed for detector readout between the successive pattern
Allen Stephone B.
Conley Rose & Tayon PC
Kla-Tencor Corporation
Meyertons Eric B.
LandOfFree
Conventional and confocal multi-spot scanning optical... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Conventional and confocal multi-spot scanning optical..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Conventional and confocal multi-spot scanning optical... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2451188