WO2010006197A1 - Small defect detection sensitive, low cost specimen inspection system - Google Patents

Small defect detection sensitive, low cost specimen inspection system Download PDF

Info

Publication number
WO2010006197A1
WO2010006197A1 PCT/US2009/050151 US2009050151W WO2010006197A1 WO 2010006197 A1 WO2010006197 A1 WO 2010006197A1 US 2009050151 W US2009050151 W US 2009050151W WO 2010006197 A1 WO2010006197 A1 WO 2010006197A1
Authority
WO
WIPO (PCT)
Prior art keywords
light
unpatterned surface
illuminating
defects
unpatterned
Prior art date
Application number
PCT/US2009/050151
Other languages
French (fr)
Inventor
Michael E. Fossey
Original Assignee
Motion Optics Corporation
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Motion Optics Corporation filed Critical Motion Optics Corporation
Publication of WO2010006197A1 publication Critical patent/WO2010006197A1/en

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/95Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
    • G01N21/9501Semiconductor wafers
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/8806Specially adapted optical and illumination features
    • G01N2021/8822Dark field detection

Definitions

  • This disclosure relates to inspection of unpattemed specimen surfaces and, in particular, to a low cost specimen inspection system that is implemented with spatially selective imaging optics to reduce the effect of background light on detection of small defects.
  • Unpattemed silicon wafers both with and without deposited metal or dielectric films, have historically been inspected by systems that are constructed with a laser for illumination; optics to focus the laser to a relatively small spot (about 50 ⁇ m diameter) on the wafer surface; some combination of mechanical, electo-mechanical, or electo- optic devices to scan the laser spot across the wafer; wide angle light collection optics (or multiple light collection optics); and a photomultiplier tube or tubes to detect light scattered from particles or other defects on the surface.
  • the model numbers of inspection systems produced by Tencor and KLA- Tencor include the 6200, SP1 , SP1-TBI, and SP2.
  • Inspection systems produced by Aeronca, Estek, and ADE Optical Systems include the WIS-100, WIS600, WIS-800, WIS-900, WIS-CR80, WIS-CR81 , and AWIS.
  • Different laser wavelengths, incident angles, and collection geometries were all incorporated at various times in an effort to reduce the amount of detected light that was scattered by the surface and thereby increase the amount of detected light that was scattered by particles or other defects. Defect detection sensitivity limits characteristic of the early systems were on the order of one micron.
  • a highly sensitive defect detecting, low cost wafer inspection system achieves, with use of relatively modest power density in the illumination beam, sufficiently small defect detection sensitivity to comply with the roadmap for semiconductors.
  • a high intensity light emitting diode (LED) or a laser source illuminates a larger area of an unpatterned wafer specimen than the relatively small area illuminated in the systems described above.
  • An advantageous consequence of the larger illuminated area is that the preferred wafer inspection system disclosed is not bound by a fundamental limitation of power in the illumination beam.
  • a method of detecting defects in or on an unpatterned surface of a specimen is practiced with use of a specimen inspection system that scans a light beam across the unpatterned surface and collects with a light sensor light scattered by defects in or on the unpatterned surface.
  • defects include particles, pits, scratches, surface contamination, or surface imperfections.
  • the method entails directing the light beam to illuminate the unpatterned surface of the specimen. Defects in or on the unpatterned surface produce scattered light propagating from the unpatterned surface in response to the illuminating light beam.
  • a camera functioning as the light sensor cooperates with imaging optics and includes an array of light sensitive sensor elements, each of which has a sensor element area. The imaging optics define imaged sensor element areas on the unpatterned surface.
  • the array of light sensitive sensor elements is a component of an electron multiplying charge-coupled device (EMCCD).
  • ECCD electron multiplying charge-coupled device
  • the light beam illuminating the unpatterned surface has a beam spot area that is substantially larger than any one of the imaged sensor element areas but is substantially smaller than the unpatterned surface area.
  • the beam spot area is established by a 2 mm - 20 mm diameter beam
  • the imaged sensor element area is established by a 3 ⁇ m - 30 ⁇ m square imaged sensor element of a 1.974 mm x 1.488 mm - 19.74 mm x 14.88 mm imaged sensor array
  • the unpattemed surface area is established by a 150 mm - 450 mm diameter wafer.
  • the method also entails scanning the illuminating light beam across the unpattemed surface and directing for incidence on the light sensitive sensor elements of the camera measurement light corresponding to the scattered light from the imaged sensor element areas. Portions of the measurement light propagating from the imaging optics and impinging on the light sensitive sensor elements correspond to the imaged sensor element areas of different light scattering regions of the unpattemed surface in or on which are present the defects producing the scattered light. The portions of measurement light are substantially free from contributions of background light scattered by the imaged sensor element areas of neighboring ones of the light scattering regions encompassed within the beam spot area at any instant as the light beam is scanned across the unpattemed surface.
  • the result is a higher degree of defect detection sensitivity than that achievable by a light sensor having a sensor area that is about the same as the beam spot area.
  • the method further entails correlating intensities of measurement light impinging on the light sensitive sensor elements to sizes of defects present in the corresponding different light scattering regions of the unpattemed surface.
  • An inspection system implemented with wide area illumination, spatially selective imaging optics, and an image array light sensor greatly reduces the effect of background light on detection of small defects in the practice of the disclosed method.
  • Conventional CCD light sensors exhibit noise properties that would dominate small particle signals.
  • FIGS. 1 and 2 are respective top plan and side elevation diagramatic views of a specimen inspection system for use in practicing the disclosed defect detection method.
  • FIGS. 3A and 3B show the amounts of background light detected from an illuminated surface of a specimen undergoing inspection by, respectively, a typical prior art specimen inspection system and the specimen inspection system of FIGS. 1 and 2.
  • FIGS. 1 and 2 A preferred highly sensitive defect detecting, low cost wafer inspection system 10 is illustrated in FIGS. 1 and 2.
  • a high power LED or laser diode light source 12 emits a light beam 14 that propagates through a beam-forming lens or lens assembly 18.
  • Lens assembly 18 collimates light beam 14 to form an incident beam 20 that illuminates a region of, and thereafter propagates as a reflected beam 24 from, an unpatterned major surface 26 of a semiconductor wafer specimen 28 under inspection.
  • Wafer 28 is secured to a rotary stage 34, which is mounted on a linear stage 36 that positions wafer 28 selectively to locations in an X-Y plane.
  • Imaging optics 40 preferably implemented as a video lens, and an electron- multiplying charge-coupled device (EMCCD) camera 42 are positioned over unpatterned surface 26 of wafer 28.
  • Video lens 40 images onto light sensitive sensor elements in camera 42 the illuminated region of unpatterned surface 26.
  • a beam trap 44 is positioned to prevent reflected beam 24 from scattering into camera 42.
  • Measurement light corresponding to scattered light from the imaged light sensor element areas of different light scattering regions of unpatterned surface 26 propagates from video lens 40 and carries to camera 42 information about sizes of defects in the illuminated regions.
  • a processor 50 correlates intensities of measurement light impinging on the light sensitive sensor elements of camera 42 to sizes of defects present in corresponding different light scattering illuminated regions of unpatterned surface 26.
  • Processor 50 includes stored information relating defect sizes to corresponding scattered light intensities.
  • Table 1 lists examples of basic EMCCD sensor devices at the integrated circuit level that are suitable for use in camera 42.
  • Table 2 lists examples of complete video cameras 42 that incorporate the sensor devices of the types listed in Table 1.
  • EMCCD electronic multi-dimensional digital camera
  • CCD charge-coupled device
  • These CCD elements operate at a higher voltage than that at which the usual CCD elements operate and through an electron multiplying effect that provides nearly noise-free gain.
  • the result is a video camera that can simultaneously detect light at nearly single photon levels and operate at full video frame rates on the order of 30 Hz.
  • a preferred embodiment of wafer inspection system 10 is implemented with the Andor luca DL-658M camera because of its low cost, high video frame rate, and relatively high light detection sensitivity in the 400 nm to 500 nm range.
  • Table 3 lists examples of candidate high intensity LED sources 12 of single LED and multiple LED array types, for use in wafer system 10.
  • a candidate laser diode system available from RgBLase LLC is the FB-445 series laser diode system, which exhibits an 800 mW output at 445 nm.
  • a preferred embodiment of wafer inspection system 10 is implemented with the OD-405-99-110 LED device. Since light scatters from small particles in proportion to the inverse fourth power of the wavelength, the shorter wavelength of 405 nm gives an overall advantage, even against the higher power of the OD-470-99-110 470 nm-LED device. For an ultimate low cost system, the LXL-LR5C LED device may be optimal.
  • a typical laser-based system may be constructed with a 2000 mW, 532 nm laser source or an 800 mW, 445 nm laser diode module. Light source 12 can be easily collimated using lens assembly 18 composed of standard off-the-shelf lenses to illuminate the imaging region of unpatterned surface 26 of wafer 28.
  • a tradeoff in the design of an inspection system of this type is between defect detection sensitivity and throughput. For highest defect detection sensitivity, a smaller illuminated region with corresponding small imaging area would be chosen. For highest throughput, a larger illuminated region and corresponding large imaging area would be chosen.
  • a reasonable tradeoff entails use of an illuminated imaging area of 4.4 mm x 3.3 mm. Since the luca DL-658M camera has an array size of 6.58 mm x 4.96 mm, the selected imaging area represents a video lens magnification of 1.5.
  • video lens 40 located approximately 40 mm from the imaging array of camera 42 and 26.67 mm from wafer surface 26, and then adjusting for focus.
  • the effective pixel size at wafer surface 26 is about 6.67 ⁇ m square.
  • imaging area dictates that 2160 images be captured to cover the entire surface of a 200 mm wafer. With the 37 frames per second capability of the luca DL-658M camera, throughput of 40 wafers per hour can be achieved with commercially available linear and rotary stages capable of conventional performance. [0024] To make the system more flexible, it may be advantageous to replace the beam-forming lens assembly 18 with a variable beam expander and implement imaging optics 40 as a zoom lens. With this combination, system 10 can then be adjusted for higher specimen throughput (larger imaged area) or better surface contamination or imperfection detection sensitivity (smaller imaged area and higher intensity illumination) to better suit the application.
  • Rotary stage 34 is the higher performance stage because it moves much more often than does linear stage 36, assuming wafer 28 is scanned in a sequence of circles, rotating wafer 28 one complete rotation before moving linear stage 36.
  • a preferred embodiment of wafer inspection system 10 is implemented with a Model No. RTH-6 rotary stage available from Intellidrives, Inc. This rotary stage is implemented with direct drive and integral air bearings, which provide for very rapid move and settle times when operated with a high performance servo controller.
  • the linear stage performance requirements are more modest and can be met using the ILS200CC linear stage available from Newport Corporation.
  • inspection system 10 Much of the semiconductor industry has adopted standard front opening universal pod (FOUP)-based wafer handling systems.
  • the described preferred embodiment of inspection system 10 is designed for initial implementation on an industry standard box opener/loader-to-tool standard interface (BOLTS) configuration platform, which allows easy integration with most present day wafer handling platforms.
  • BOLTS box opener/loader-to-tool standard interface
  • a more detailed analysis that takes into account the much narrower noise bandwidth and smaller amount of scattered light of inspection system 10 makes apparent its performance advantage. Since a system has no a priori knowledge of the actual location of a defect to be detected, the collection optics of video lens 40 collects light from the entire illuminated region of water surface 26.
  • FIGS. 3A and 3B show a comparative relationship between amounts of scattered background light propagating from illuminated surface 26 of wafer 28 and detected by, respectively, a typical prior art inspection system and inspection system 10.
  • a 50 ⁇ m diameter laser spot illuminates a region 58 of wafer surface 26 in which a particle 60 resides and produces a scattered light signal.
  • the scattered light detected includes the signal light scattered by particle 60 and background light scattered by the entire 50 ⁇ m diameter illuminated region 58 of wafer surface 26.
  • scattered light detection performed by inspection 10 as shown in FIG.
  • FIG. 3B a portion 62 of an illuminated imaging area of 4.4 mm x 3.3 mm on wafer surface 26 contains a particle 60 that produces a scattered light signal.
  • FIG. 3B shows fifteen of imaged array pixels or elements 64 neighboring the imaged array element 64 in which particle 60 resides. Because inspection system 10 images the illuminated region onto a detector array of camera 42, each imaged array element 64 of about 6.67 ⁇ m square detects light scattered by and received from only a small portion of wafer surface 26. Since the imaged pixel in inspection system 10 is only 6.67 ⁇ m square, the amount of surface scatter is 44 times smaller than that from a 50 ⁇ m diameter spot detected by prior art systems. This difference in detected scattered light is illustrated in FIGS. 3A and 3B. This advantage in surface scatter tips the balance well in favor of inspection system 10.
  • the detected current due to surface scatter is:
  • I n V(2e * I surf * B) .
  • a 2 represents a wavelength at which the power of scattered light, P surf (A 2 ), has been measured from a given sample surface
  • the power of scattered light expected when the same sample surface is illuminated by a wavelength X x can be calculated as follows:
  • P ps i (P,/A bea mr(1-cos 2 ( ⁇ )/2R 2 )*(2 ⁇ /2 ) 4 *((n 2 -1)/(n 2 +2)) * (d/2) 6 *( ⁇ D 2 /4).
  • One of the difficulties encountered with a system of this type is achieving the necessary dynamic range to size larger (> 125 nm) particles. This may be accomplished by capturing two frames at each location, one of the frames captured with either a very short integration time or low illumination power level for the purpose of sizing those particles that are saturated when captured under set up conditions for maximum sensitivity.
  • inspection system 10 does not rely on expensive high power lasers and complex high speed scanning systems, it can be produced for a fraction of the cost of present day laser- and photomultiplier-based systems.
  • Inspection system 10 is also scalable to much higher power levels (and better sensitivity) without danger of wafer damage. With higher power levels, the sensitivity can reach 22 nm defect size levels.

Abstract

A method of detecting defects (60) in or on an unpatterned surface (26) of a specimen (28) entails scanning a light beam (20) across the unpatterned surface and collecting with a camera (42) measurement light scattered by defects in or on the unpatterned surface. The camera cooperates with imaging optics (40) and includes an array of light sensitive sensor elements. Portions of the measurement light propagating from the imaging optics and impinging on the light sensitive sensor elements correspond to imaged sensor element (64) areas of different light scattering regions of the unpatterned surface where defects are present. The portions of measurement light are substantially free from contributions of background light scattered by the imaged sensor element areas of neighboring ones of the light scattering regions encompassed within the beam spot area at any instant as the light beam is scanned. The result is a higher degree of defect detection sensitivity.

Description

SMALL DEFECT DETECTION SENSITIVE, LOW COST SPECIMEN INSPECTION SYSTEM
Related Application
[0001] This application claims benefit of U.S. Provisional Patent Application No. 61/134,682, filed July 11 , 2008.
Copyright Notice
[0002] © 2009 Motion Optics Corporation. A portion of the disclosure of this patent document contains material that is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever. 37 CFR § 1.71 (d).
Technical Field
[0003] This disclosure relates to inspection of unpattemed specimen surfaces and, in particular, to a low cost specimen inspection system that is implemented with spatially selective imaging optics to reduce the effect of background light on detection of small defects.
Background Information
[0004] Unpattemed silicon wafers, both with and without deposited metal or dielectric films, have historically been inspected by systems that are constructed with a laser for illumination; optics to focus the laser to a relatively small spot (about 50 μm diameter) on the wafer surface; some combination of mechanical, electo-mechanical, or electo- optic devices to scan the laser spot across the wafer; wide angle light collection optics (or multiple light collection optics); and a photomultiplier tube or tubes to detect light scattered from particles or other defects on the surface. In the United States, such systems have been predominately manufactured by Tencor, which later became KLA- Tencor, and by Aeronca, which later became Estek, a division of Kodak, and thereafter became ADE Optical Systems, a division of ADE Corporation. More recently, ADE has become part of KLA-Tencor.
[0005] The model numbers of inspection systems produced by Tencor and KLA- Tencor include the 6200, SP1 , SP1-TBI, and SP2. Inspection systems produced by Aeronca, Estek, and ADE Optical Systems include the WIS-100, WIS600, WIS-800, WIS-900, WIS-CR80, WIS-CR81 , and AWIS. Different laser wavelengths, incident angles, and collection geometries were all incorporated at various times in an effort to reduce the amount of detected light that was scattered by the surface and thereby increase the amount of detected light that was scattered by particles or other defects. Defect detection sensitivity limits characteristic of the early systems were on the order of one micron. Later systems, such as the SP1-TBI and AWIS, achieved defect detection sensitivities on the order of 50 nm. The SP2 claims 32 nm defect detection sensitivity but operates at throughput levels far below those required for production applications. [0006] Although each of these systems was unique in its own way, all of them suffered from the complications inherent with high-powered laser systems. They suffered from laser eye safety issues that complicated alignment and maintenance. To achieve high throughput, the systems required either a very high speed rotating stage (SP1) or the complexity of an acousto-optic deflector (AWIS). These complications drove high product costs and hampered the inherent reliability of the systems. [0007] Despite all the expense and complexity of these systems, their ability to detect small particles and other defects on unpatterned wafers has fallen well behind the National Technology Roadmap for Semiconductors. The best current production level for semiconductor devices utilizes minimum feature size of 45 nm. According to the roadmap, wafer inspection systems of this generation need to be capable of detecting 32 nm particles and other defects on unpatterned surfaces, with and without blanket films of metal or oxide. Currently available systems based on the technology described above can barely detect 50 nm particles and other defects on the smoothest silicon wafers at production throughput levels, let alone on wafers with the rougher films deposited, in which their best defect detection sensitivity is on the order of 100 nm. These systems can be scaled to achieve better small defect detection by increasing the power of the laser. However, because of the small spot size, laser output incident on the substrate at a level only two to four times the highest power used in any of the current systems results in an onset of damage to the substrate. This represents a fundamental limitation of the technology.
Summary of the Disclosure
[0008] In a preferred embodiment, a highly sensitive defect detecting, low cost wafer inspection system achieves, with use of relatively modest power density in the illumination beam, sufficiently small defect detection sensitivity to comply with the roadmap for semiconductors. A high intensity light emitting diode (LED) or a laser source illuminates a larger area of an unpatterned wafer specimen than the relatively small area illuminated in the systems described above. An advantageous consequence of the larger illuminated area is that the preferred wafer inspection system disclosed is not bound by a fundamental limitation of power in the illumination beam. [0009] Stated in more general terms, a method of detecting defects in or on an unpatterned surface of a specimen is practiced with use of a specimen inspection system that scans a light beam across the unpatterned surface and collects with a light sensor light scattered by defects in or on the unpatterned surface. Such defects include particles, pits, scratches, surface contamination, or surface imperfections. The method entails directing the light beam to illuminate the unpatterned surface of the specimen. Defects in or on the unpatterned surface produce scattered light propagating from the unpatterned surface in response to the illuminating light beam. A camera functioning as the light sensor cooperates with imaging optics and includes an array of light sensitive sensor elements, each of which has a sensor element area. The imaging optics define imaged sensor element areas on the unpatterned surface. In preferred embodiments, the array of light sensitive sensor elements is a component of an electron multiplying charge-coupled device (EMCCD). The light beam illuminating the unpatterned surface has a beam spot area that is substantially larger than any one of the imaged sensor element areas but is substantially smaller than the unpatterned surface area. In a preferred implementation, the beam spot area is established by a 2 mm - 20 mm diameter beam, the imaged sensor element area is established by a 3 μm - 30 μm square imaged sensor element of a 1.974 mm x 1.488 mm - 19.74 mm x 14.88 mm imaged sensor array, and the unpattemed surface area is established by a 150 mm - 450 mm diameter wafer.
[0010] The method also entails scanning the illuminating light beam across the unpattemed surface and directing for incidence on the light sensitive sensor elements of the camera measurement light corresponding to the scattered light from the imaged sensor element areas. Portions of the measurement light propagating from the imaging optics and impinging on the light sensitive sensor elements correspond to the imaged sensor element areas of different light scattering regions of the unpattemed surface in or on which are present the defects producing the scattered light. The portions of measurement light are substantially free from contributions of background light scattered by the imaged sensor element areas of neighboring ones of the light scattering regions encompassed within the beam spot area at any instant as the light beam is scanned across the unpattemed surface. The result is a higher degree of defect detection sensitivity than that achievable by a light sensor having a sensor area that is about the same as the beam spot area. The method further entails correlating intensities of measurement light impinging on the light sensitive sensor elements to sizes of defects present in the corresponding different light scattering regions of the unpattemed surface.
[0011] An inspection system implemented with wide area illumination, spatially selective imaging optics, and an image array light sensor greatly reduces the effect of background light on detection of small defects in the practice of the disclosed method. Conventional CCD light sensors exhibit noise properties that would dominate small particle signals.
[0012] Additional aspects and advantages will be apparent from the following detailed description of preferred embodiments, which proceeds with reference to the accompanying drawings.
Brief Description of the Drawings
[0013] FIGS. 1 and 2 are respective top plan and side elevation diagramatic views of a specimen inspection system for use in practicing the disclosed defect detection method. [0014] FIGS. 3A and 3B show the amounts of background light detected from an illuminated surface of a specimen undergoing inspection by, respectively, a typical prior art specimen inspection system and the specimen inspection system of FIGS. 1 and 2.
Detailed Description of Preferred Embodiments
[0015] A preferred highly sensitive defect detecting, low cost wafer inspection system 10 is illustrated in FIGS. 1 and 2. With reference to FIGS. 1 and 2, a high power LED or laser diode light source 12 emits a light beam 14 that propagates through a beam-forming lens or lens assembly 18. Lens assembly 18 collimates light beam 14 to form an incident beam 20 that illuminates a region of, and thereafter propagates as a reflected beam 24 from, an unpatterned major surface 26 of a semiconductor wafer specimen 28 under inspection. Wafer 28 is secured to a rotary stage 34, which is mounted on a linear stage 36 that positions wafer 28 selectively to locations in an X-Y plane.
[0016] Imaging optics 40, preferably implemented as a video lens, and an electron- multiplying charge-coupled device (EMCCD) camera 42 are positioned over unpatterned surface 26 of wafer 28. Video lens 40 images onto light sensitive sensor elements in camera 42 the illuminated region of unpatterned surface 26. A beam trap 44 is positioned to prevent reflected beam 24 from scattering into camera 42. Measurement light corresponding to scattered light from the imaged light sensor element areas of different light scattering regions of unpatterned surface 26 propagates from video lens 40 and carries to camera 42 information about sizes of defects in the illuminated regions. A processor 50 correlates intensities of measurement light impinging on the light sensitive sensor elements of camera 42 to sizes of defects present in corresponding different light scattering illuminated regions of unpatterned surface 26. Processor 50 includes stored information relating defect sizes to corresponding scattered light intensities.
[0017] Table 1 lists examples of basic EMCCD sensor devices at the integrated circuit level that are suitable for use in camera 42. Manufacturer Part Number
Texas Instruments TC247SPD
Texas Instruments TX285SPD
Texas Instruments TC253SPD
E2V CCD60
E2V CCD65
E2V CCD97
E2V CCD201
Table 1
[0018] Table 2 lists examples of complete video cameras 42 that incorporate the sensor devices of the types listed in Table 1.
Manufacturer Part Number
Andor luca DL-658M
Andor iXon DU-897
Andor iXon DU-860
Andor iXon DU-888
E2V L3C60
E2V L3C65
E2V L3C65P
E2V L3C90
Salvador Imaging SI-VGA60-EM
Salvador Imaging SI-1 M30-EM
Qimaging Rolera-MGi Plus
Table 2
[0019] An advantage of EMCCD and cameras incorporating EMCCDs over earlier generations of video cameras is that the EMCCD includes an additional series of charge-coupled device (CCD) elements. These CCD elements operate at a higher voltage than that at which the usual CCD elements operate and through an electron multiplying effect that provides nearly noise-free gain. The result is a video camera that can simultaneously detect light at nearly single photon levels and operate at full video frame rates on the order of 30 Hz. A preferred embodiment of wafer inspection system 10 is implemented with the Andor luca DL-658M camera because of its low cost, high video frame rate, and relatively high light detection sensitivity in the 400 nm to 500 nm range. [0020] Table 3 lists examples of candidate high intensity LED sources 12 of single LED and multiple LED array types, for use in wafer system 10.
Manufacturer Part Number Wavelength Power # LEDs
Quadica Luxeon LXL-LR5C 455 nm 630 mW 1
Opto Diode Corp OD-405-99-110 405 nm 170O mW 99
Opto Diode Corp OD-470-99-110 470 nm 210O mW 99
Table 3
A candidate laser diode system available from RgBLase LLC is the FB-445 series laser diode system, which exhibits an 800 mW output at 445 nm.
[0021] A preferred embodiment of wafer inspection system 10 is implemented with the OD-405-99-110 LED device. Since light scatters from small particles in proportion to the inverse fourth power of the wavelength, the shorter wavelength of 405 nm gives an overall advantage, even against the higher power of the OD-470-99-110 470 nm-LED device. For an ultimate low cost system, the LXL-LR5C LED device may be optimal. A typical laser-based system may be constructed with a 2000 mW, 532 nm laser source or an 800 mW, 445 nm laser diode module. Light source 12 can be easily collimated using lens assembly 18 composed of standard off-the-shelf lenses to illuminate the imaging region of unpatterned surface 26 of wafer 28. [0022] A tradeoff in the design of an inspection system of this type is between defect detection sensitivity and throughput. For highest defect detection sensitivity, a smaller illuminated region with corresponding small imaging area would be chosen. For highest throughput, a larger illuminated region and corresponding large imaging area would be chosen. A reasonable tradeoff entails use of an illuminated imaging area of 4.4 mm x 3.3 mm. Since the luca DL-658M camera has an array size of 6.58 mm x 4.96 mm, the selected imaging area represents a video lens magnification of 1.5. This can be accomplished using a very low cost video lens 40, such as the Fujinon HF16HA-1 B, mounting it backwards with video lens 40 located approximately 40 mm from the imaging array of camera 42 and 26.67 mm from wafer surface 26, and then adjusting for focus. With this arrangement, the effective pixel size at wafer surface 26 is about 6.67 μm square.
[0023] The above choice of imaging area dictates that 2160 images be captured to cover the entire surface of a 200 mm wafer. With the 37 frames per second capability of the luca DL-658M camera, throughput of 40 wafers per hour can be achieved with commercially available linear and rotary stages capable of conventional performance. [0024] To make the system more flexible, it may be advantageous to replace the beam-forming lens assembly 18 with a variable beam expander and implement imaging optics 40 as a zoom lens. With this combination, system 10 can then be adjusted for higher specimen throughput (larger imaged area) or better surface contamination or imperfection detection sensitivity (smaller imaged area and higher intensity illumination) to better suit the application. The practical range of sizes for the imaged area on wafer 28 and corresponding illuminated area ranges from about 2 mm on a side to 20 mm on a side. Below 2 mm, the throughput drops to impracticable levels for the applications described; and above 20 mm, the sizes of individual pixels are sufficiently large that the advantage of lower surface scatter begins to diminish. [0025] Rotary stage 34 is the higher performance stage because it moves much more often than does linear stage 36, assuming wafer 28 is scanned in a sequence of circles, rotating wafer 28 one complete rotation before moving linear stage 36. A preferred embodiment of wafer inspection system 10 is implemented with a Model No. RTH-6 rotary stage available from Intellidrives, Inc. This rotary stage is implemented with direct drive and integral air bearings, which provide for very rapid move and settle times when operated with a high performance servo controller. The linear stage performance requirements are more modest and can be met using the ILS200CC linear stage available from Newport Corporation.
[0026] Much of the semiconductor industry has adopted standard front opening universal pod (FOUP)-based wafer handling systems. The described preferred embodiment of inspection system 10 is designed for initial implementation on an industry standard box opener/loader-to-tool standard interface (BOLTS) configuration platform, which allows easy integration with most present day wafer handling platforms. [0027] A simple comparison based on the amount of light scattered from a particle in the above-described inspection system 10 and from a particle in a conventional laser- and photomultiplier-based inspection system would show that the conventional system produces a higher signal level. One might conclude from such comparison that the conventional system would achieve higher defect detection sensitivity. A more detailed analysis that takes into account the much narrower noise bandwidth and smaller amount of scattered light of inspection system 10 makes apparent its performance advantage. Since a system has no a priori knowledge of the actual location of a defect to be detected, the collection optics of video lens 40 collects light from the entire illuminated region of water surface 26.
[0028] FIGS. 3A and 3B show a comparative relationship between amounts of scattered background light propagating from illuminated surface 26 of wafer 28 and detected by, respectively, a typical prior art inspection system and inspection system 10. With reference to scattered light detection performed by prior art inspection systems, as shown in FIG. 3A, a 50 μm diameter laser spot illuminates a region 58 of wafer surface 26 in which a particle 60 resides and produces a scattered light signal. The scattered light detected includes the signal light scattered by particle 60 and background light scattered by the entire 50 μm diameter illuminated region 58 of wafer surface 26. With reference to scattered light detection performed by inspection 10, as shown in FIG. 3B, a portion 62 of an illuminated imaging area of 4.4 mm x 3.3 mm on wafer surface 26 contains a particle 60 that produces a scattered light signal. FIG. 3B shows fifteen of imaged array pixels or elements 64 neighboring the imaged array element 64 in which particle 60 resides. Because inspection system 10 images the illuminated region onto a detector array of camera 42, each imaged array element 64 of about 6.67 μm square detects light scattered by and received from only a small portion of wafer surface 26. Since the imaged pixel in inspection system 10 is only 6.67 μm square, the amount of surface scatter is 44 times smaller than that from a 50 μm diameter spot detected by prior art systems. This difference in detected scattered light is illustrated in FIGS. 3A and 3B. This advantage in surface scatter tips the balance well in favor of inspection system 10.
[0029] To compare the defect detection sensitivity of a conventional inspection system and inspection system 10, one first calculates the amount light scattered from the surface and collected by the system: _ BRDF * πD2Pi * cos θ surf - ^2 The light scattered into one pixel of the EMCCD is:
Psurf = Psurf / N .
The detected current due to surface scatter is:
'surf = P ' surf .
The noise caused by Poisson fluctuations of this current is:
In = V(2e * Isurf * B) .
Light scattered by microroughness scales to the fourth power of the wavelength.
If A2 represents a wavelength at which the power of scattered light, Psurf (A2 ), has been measured from a given sample surface, the power of scattered light expected when the same sample surface is illuminated by a wavelength Xx can be calculated as follows:
Psu,f (Λ ) = Psurf (^2) * (V Λ )4-
[0030] Although not rigorously correct for scattering from particles on a dielectric surface, one can use Rayleigh scattering to estimate the amount of scattered light available from a small spherical particle (in this case a 50 nm poly-styrene latex (PSL) sphere). More complicated numerical approaches exist that take into account interaction of the particle with the surface. The additional accuracy, while needed to predict absolute defect detection sensitivity, such calculations are not necessary for a comparison because the inaccuracies from the Rayleigh approximation affect both systems equally.
[0031] According to the Rayleigh approximation:
Ppsi = (P,/Abeamr(1-cos2(θ )/2R2)*(2π/2 )4*((n2-1)/(n2+2))*(d/2)6*(πD2/4).
CALCULATION PARAMETERS
Conventional System Disclosed System
Parameter Description Value Value
BRDF Scatter from silicon wafer 1.0Oe-7 2.98e-7
D Lens diameter 0.05 0.05 meters
R Lens distance 0.05 0.05 meters θ Scatter angle 65 65 degrees
P, Incident power 2 1.7 Watts
N Illuminated pixels 1 326,368 λ Wavelength 5.32e 7 4.05e-7 microns
Abeam Illumination beam area 1 .96e-9 3.26e-5 m2 n Particle dielectric constant 1.6 1.6 d Particle diameter 5e-8 5e-β meters
BW System bandwidth 5.0e6 13 Hertz e Electron charge 1 .6e-19 1.6e-19 Coulombs
P Detector sensitivity 0.5 0.5 Amps/Watt
CALCULATION RESULTS
Ppsl Collected scatter from PSL 5.34e-10 8.13e-14 Watts lpsl Current from PSL scatter 2.67e-10 4.06e-14 Amps
Psurf Collected surface scatter 6.64e-8 5.15e-13 Watts
'surf Current from surface scatter 3.32e-s 2.57e-13 Amps
In Poisson noise from lSUrf 2.3e-10 1.03e-15 Amps
Figure imgf000012_0001
[0032] One of the difficulties encountered with a system of this type is achieving the necessary dynamic range to size larger (> 125 nm) particles. This may be accomplished by capturing two frames at each location, one of the frames captured with either a very short integration time or low illumination power level for the purpose of sizing those particles that are saturated when captured under set up conditions for maximum sensitivity.
[0033] Because inspection system 10 does not rely on expensive high power lasers and complex high speed scanning systems, it can be produced for a fraction of the cost of present day laser- and photomultiplier-based systems. One can conclude that, based on the signal to noise analysis presented and the performance of conventional inspection systems, the minimum defect size the disclosed system can detect reaches
32 nm while operating at production level throughput levels, which performance compares favorably with the 50 nm sensitivity of conventional laser- and photomultiplier- based systems.
[0034] Inspection system 10 is also scalable to much higher power levels (and better sensitivity) without danger of wafer damage. With higher power levels, the sensitivity can reach 22 nm defect size levels.
[0035] It will be obvious to those having skill in the art that many changes may be made to the details of the above-described embodiments without departing from the underlying principles of the invention. The scope of the present invention should, therefore, be determined only by the following claims.

Claims

Claims
1. In a method of detecting defects in or on an unpatterned surface of a specimen, in which the method entails scanning a light beam across the unpatterned surface and collecting with a light sensor light scattered by defects in or on the unpatterned surface, the improvement comprising: directing a light beam to illuminate an unpatterned surface of a specimen, the unpatterned surface having an unpatterned surface area, and defects in or on the unpatterned surface producing scattered light propagating from the unpatterned surface in response to the illuminating light beam; providing as a light sensor a camera cooperating with imaging optics, the camera including an array of light sensitive sensor elements each of which having a sensor element area, and the imaging optics defining imaged sensor element areas on the unpatterned surface; establishing for the light beam illuminating the unpatterned surface a beam spot area that is substantially larger than any one of the imaged sensor element areas but is substantially smaller than the unpatterned surface area; scanning the illuminating light beam across the unpatterned surface; directing for incidence on the light sensitive sensor elements of the camera measurement light corresponding to the scattered light from the imaged sensor element areas, portions of the measurement light propagating from the imaging optics and impinging on the light sensitive sensor elements corresponding to the imaged sensor element areas of different light scattering regions of the unpatterned surface in or on which are present the defects producing the scattered light, the portions of measurement light substantially free from contributions of background light scattered by the imaged sensor element areas of neighboring ones of the light scattering regions encompassed within the beam spot area at any instant as the light beam is scanned across the unpatterned surface, and thereby resulting in a higher degree of defect detection sensitivity than that achievable by a light sensor having a sensor area that is about the same as the beam spot area; and correlating intensities of measurement light impinging on the light sensitive sensor elements to sizes of defects present in the corresponding different light scattering regions of the unpatterned surface.
2. The method of claim 1 , in which the array of light sensitive sensor elements is a component of an electron multiplying charge coupled device (EMCCD).
3. The method of claim 1 , in which the imaging optics includes a video lens.
4. The method of claim 1 , in which the imaging optics includes a zoom lens and, in which the light beam illuminating the unpatterned surface propagates from a variable beam-forming lens and the measurement light directed for incidence on the light sensor propagates from the zoom lens, further comprising adjusting the variable beam-forming lens and the zoom lens in combination to carry out a performance tradeoff between higher specimen throughput and surface contamination or imperfection detection sensitivity.
5. The method of claim 1 , in which a laser diode is a source of the light beam illuminating the unpatterned surface of the specimen.
6. The method of claim 1 , in which a light emitting diode (LED) laser is a source of the light beam illuminating the unpatterned surface of the specimen.
7. The method of claim 1 , in which the specimen is a semiconductor wafer.
8. The method of claim 1 , in which the beam spot area is established by a light beam diameter of between about 2 mm and about 20 mm.
9. The method of claim 1 , in which the beam spot area established by the illuminating light beam provides a small illumination region for high defect detection sensitivity, and in which for a location of a light scattering region, the camera captures multiple frames of measurement light impinging on the light sensitive sensor elements, one of the frames representing intensities of captured measurement light produced under a short integration time or a low power illuminating light beam condition to enable the correlating of intensities of measurement light scattered by defects that are saturated by intensity of the light beam illuminating the small illumination region.
10. The method of claim 9, in which the multiple frames captured achieve a dynamic range for correlating the intensities of measurement light to defect sizes of greater than 125 nm in diameter.
11. A system for detecting defects in or on an unpatterned surface of a specimen, comprising: a light source from which an illuminating light beam propagates to illuminate with a beam spot area an unpatterned surface of a specimen, the unpatterned surface having an unpatterned surface area that is substantially larger than the beam spot area, and defects in or on the unpatterned surface producing scattered light propagating from the unpatterned surface in response to the illuminating light beam; a scanning mechanism for scanning the illuminating light beam across the unpatterned surface; a camera including an array of light sensitive sensor elements each of which having a sensor element area; imaging optics cooperating with the camera to define imaged sensor element areas on the unpatterned surface, each of the imaged sensor element areas being substantially smaller than the beam spot area, and to direct for incidence on the light sensitive sensor elements of the camera measurement light corresponding to the scattered light from the imaged sensor element areas, portions of the measurement light propagating from the imaging optics and impinging on the light sensitive sensor elements corresponding to the imaged sensor element areas of different light scattering regions of the unpatterned surface in or on which are present the defects producing the scattered light, the portions of measurement light substantially free from contributions of background light scattered by the imaged sensor element areas of neighboring ones of the light scattering regions encompassed within the beam spot area at any instant as the light beam is scanned across the unpatterned surface; and a processor for correlating intensities of measurement light impinging on the light sensitive sensor elements to sizes of defects present in the corresponding different light scattering regions of the unpatterned surface.
12. The system of claim 11 , in which the array of light sensitive sensor elements is a component of an electron multiplying charge coupled device (EMCCD).
13. The system of claim 11 , in which the imaging optics includes a video lens.
14. The system of claim 11 , in which the light source includes a variable beam- forming lens from which the illuminating light beam propagates, and in which the imaging optics includes a zoom lens from which the measurement light directed for incidence on the light sensitive sensor elements propagates, and the variable beam- forming lens and the zoom lens being adjustable in combination to carry out a performance tradeoff between higher specimen throughput and surface contamination or imperfection detection sensitivity.
15. The system of claim 11 , in which the light source includes a laser diode.
16. The system of claim 11 , in which the light source includes a light emitting diode (LED) laser.
17. The system of claim 11 , in which the specimen is a semiconductor wafer.
18. The system of claim 11 , in which the beam spot area is established by a light beam diameter of between about 2 mm and about 20 mm.
19. The system of claim 11 , in which the beam spot area established by the illuminating light beam provides a small illumination region for high defect detection sensitivity, and in which for a location of a light scattering region, the camera captures multiple frames of measurement light impinging on the light sensitive sensor elements, one of the frames representing intensities of captured measurement light produced under a short integration time or a low power illuminating light beam condition to enable the correlating of intensities of measurement light scattered by defects that are saturated by intensity of the light beam illuminating the small illumination region.
20. The system of claim 19, in which the multiple frames captured achieve a dynamic range for correlating the intensities of measurement light to defect sizes of greater than 125 nm in diameter.
PCT/US2009/050151 2008-07-11 2009-07-09 Small defect detection sensitive, low cost specimen inspection system WO2010006197A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US13468208P 2008-07-11 2008-07-11
US61/134,682 2008-07-11

Publications (1)

Publication Number Publication Date
WO2010006197A1 true WO2010006197A1 (en) 2010-01-14

Family

ID=41507446

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2009/050151 WO2010006197A1 (en) 2008-07-11 2009-07-09 Small defect detection sensitive, low cost specimen inspection system

Country Status (1)

Country Link
WO (1) WO2010006197A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9399562B2 (en) 2010-07-12 2016-07-26 Otis Elevator Company Elevator speed and position detection system using an optical sensor
US10060861B2 (en) 2017-01-10 2018-08-28 Sunspring America, Inc. Technologies for identifying defects

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020162979A1 (en) * 2001-05-02 2002-11-07 Lasertec Corporation Optically scanning apparatus and defect inspection system
US20030137318A1 (en) * 2002-01-23 2003-07-24 Marian Enachescu Methods and systems employing infrared thermography for defect detection and analysis
US6636302B2 (en) * 1994-12-08 2003-10-21 Kla-Tencor Corporation Scanning system for inspecting anamolies on surfaces
US20040207836A1 (en) * 2002-09-27 2004-10-21 Rajeshwar Chhibber High dynamic range optical inspection system and method
US6809809B2 (en) * 2000-11-15 2004-10-26 Real Time Metrology, Inc. Optical method and apparatus for inspecting large area planar objects
US20070081151A1 (en) * 2005-10-06 2007-04-12 David Shortt Methods and systems for inspection of a wafer
US20080144921A1 (en) * 2005-04-26 2008-06-19 Lizotte Todd E Inspection method and apparatus for partially drilled microvias

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6636302B2 (en) * 1994-12-08 2003-10-21 Kla-Tencor Corporation Scanning system for inspecting anamolies on surfaces
US6809809B2 (en) * 2000-11-15 2004-10-26 Real Time Metrology, Inc. Optical method and apparatus for inspecting large area planar objects
US20020162979A1 (en) * 2001-05-02 2002-11-07 Lasertec Corporation Optically scanning apparatus and defect inspection system
US20030137318A1 (en) * 2002-01-23 2003-07-24 Marian Enachescu Methods and systems employing infrared thermography for defect detection and analysis
US20040207836A1 (en) * 2002-09-27 2004-10-21 Rajeshwar Chhibber High dynamic range optical inspection system and method
US20080144921A1 (en) * 2005-04-26 2008-06-19 Lizotte Todd E Inspection method and apparatus for partially drilled microvias
US20070081151A1 (en) * 2005-10-06 2007-04-12 David Shortt Methods and systems for inspection of a wafer

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9399562B2 (en) 2010-07-12 2016-07-26 Otis Elevator Company Elevator speed and position detection system using an optical sensor
US10060861B2 (en) 2017-01-10 2018-08-28 Sunspring America, Inc. Technologies for identifying defects
US10768120B2 (en) 2017-01-10 2020-09-08 Sunspring America, Inc. Technologies for identifying defects

Similar Documents

Publication Publication Date Title
KR102302641B1 (en) Tdi sensor in a darkfield system
US9915622B2 (en) Wafer inspection
JP4751617B2 (en) Defect inspection method and apparatus
EP0935747B1 (en) Automated inspection system with bright field and dark field illumination
JP6131250B2 (en) Method and apparatus for inspection of light emitting semiconductor devices using photoluminescence imaging
US8604447B2 (en) Solar metrology methods and apparatus
US8885037B2 (en) Defect inspection method and apparatus therefor
US9182351B2 (en) Optical metrology system for spectral imaging of a sample
TW200839227A (en) Automatic inspection system for flat panel substrate
WO2008110061A1 (en) Plane substrate auto-test system and the method thereof
US8416292B2 (en) Defect inspection apparatus and method
US9041921B2 (en) Defect inspection device and defect inspection method
JP2004264287A (en) Method and apparatus for identifying defect in substrate surface using dithering for reconstructing image of insufficient sampling
JP2005283190A (en) Foreign matter inspection method and device therefor
JP5497495B2 (en) High-speed inspection method and apparatus
US7924517B2 (en) Spatial filter, a system and method for collecting light from an object
US7952701B2 (en) Surface inspection method and inspecting device using the same
WO2015011968A1 (en) Inspection device
EP2386059A1 (en) A system and method for thin film quality assurance
US20120262566A1 (en) Apparatus for illuminating substrates in order to image micro cracks, pinholes and inclusions in monocrystalline and polycrystalline substrates and method therefore
CN111610197B (en) Defect detection device and defect detection method
WO2010006197A1 (en) Small defect detection sensitive, low cost specimen inspection system
US11047805B2 (en) Inspection device and detector
JP2002116155A (en) Apparatus and method for inspecting foreign matter and defect
JP2004093209A (en) Nondestructive inspection system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 09795209

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 09795209

Country of ref document: EP

Kind code of ref document: A1