United States Patent (19)
`Falk
`
`(54)
`
`METHOD AND APPARATUS FOR MAGING
`SEMCONDUCTOR DEVICE PROPERTIES
`
`Inventor: Robert Aaron Falk, Renton, Wash.
`(75)
`73) Assignee: OptoMetrix, Inc., Renton, Wash.
`
`21
`22
`51
`52
`58
`
`56
`
`Appl. No.:711A20
`Filed:
`Sep. 5, 1996
`Int. Cl. ................ G01N 21/00
`U.S. C. .............
`... 356/390; 356/414; 356/343
`Field of Search ................................... 356/340, 414,
`356/10, 343, 390
`
`References Cited
`U.S. PATENT DOCUMENTS
`
`4,758,092
`5,028,135
`5,280.272
`5,540,494
`
`7/1988 Heinrich et al..
`7/1991 Cheung et al. ......................... 356/340
`1/1994 Nagashima et al. .....
`... 356/340
`7/1996 Purvis, Jr. et al. ..................... 356/340
`OTHER PUBLICATIONS
`Brunfeld, A. et al., "High Resolution Optical Profilometer.”
`SPIE 680:118-123, 1986.
`Heinrich, H.K. et al., "Noninvasive Sheet Charge Density
`Probe for Integrated Silicon Devices." Appl. Phys. Lett,
`48(16):1066-1068. Apr. 1986.
`Corle, T.R. et al., "Distance Measurements by Differential
`Confocal
`Optical
`Ranging."
`Applied
`Optics,
`26(12):2416-2420, Jun. 15, 1987.
`Falk, R.A. et al. "Optical Probe Techniques for Avalanching
`Photoconductors." 8th IEEE Pulsed Power Conference,
`29-32, 1991.
`Schoenbach, K.H. et al. "Temporal Development of Electric
`Field Structures in Photoconductive GaAs Switches." Appl.
`Phys. Lett, 63(15):2100-2102, 1993.
`Chen, T. et al., "Measurement Principle and Error Analysis
`for an Optical Heterodyne Profilometer." SPIE,
`201:800-803, 1993.
`
`USOO5754298A
`Patent Number:
`11
`45) Date of Patent:
`
`5,754.298
`May 19, 1998
`
`Goldstein. M. et al., "Hetrodyn Interferometerfor the Detec
`tion of Electric and Thermal Signals in Integrated Circuits
`Through
`the
`Substrate."
`Rev.
`Sci.
`Instrum.,
`64(10):3009-3013, 1993.
`Falk, R.A. et al., "Dynamic Optical Probing of High-Power
`Photoconductors." 9th IEEE Pulsed Power Conference, pp.
`88-91, 1993.
`Falk, R.A. et al., "Electro-Optic Imagery of High-Voltage
`GaAs Photoconductive Switches." IEEE Transactions on
`Electron Devices, 42(1):43–49, 1995.
`Adams, J.C., "Electro-Optic Imaging of Internal Fields in
`(111) GaAs Photoconductors." IEEE Transactions on Elec
`tron Devices, 42(6):1081-1085, 1995.
`
`Primary Examiner-Frank G. Font
`Assistant Examiner-Reginald A. Ratliff
`Attorney, Agent, or Firm-Christensen O'Connor Johnson
`& Kindness PLLC
`ABSTRACT
`57
`A radiant energy point source (10) generates radiant energy,
`and a mechanism (12, 15, 16) focuses the radiant energy
`generated by the point source onto a target (18) and scans the
`target with the focused radiant energy. A collector (16. 14)
`collects the focused radiant energy that is scattered from the
`target and a splitter (22) splits the collected radiant energy
`into two paths. Each of the two paths of the collected radiant
`energy is focused onto separate focal spots by a focusing
`mechanism (24). A pair of spatial filters (26. 28) filter the
`collected radiant energy at the focal spots. The spatial filters
`are offset from each other along the path of the focused
`radiant energy. Detectors (30, 32) separately detect the
`focused radiant energy which passes through each of the
`spatial filters and produce signals proportional to the quan
`tity of detected focused radiant energy present. Finally, the
`produced signals are combined into an image signal related
`to the distance traveled by the radiant energy from the
`focusing mechanism to the target and back to the collector.
`
`15 Claims, 7 Drawing Sheets
`
`
`
`
`
`--- Pinhole Offset, d.
`3
`
`3SHAPE EXHIBIT 1011
`3Shape v. Align
`IPR2019-00160
`
`

`

`U.S. Patent
`US. Patent
`
`May 19, 1998
`May 19, 1993
`
`Sheet 1 of 7
`Sheet 1 of 7
`
`5,754.298
`5,754,298
`
`
`
`
`
`
`
`
`
` PinholeOffset,d
` 36
`
`f
`C
`
`Figure1
`
`g
`
`34
`
`

`

`U.S. Patent
`
`May 19, 1998
`
`Sheet 2 of 7
`
`5,754.298
`
`
`
`u"euf Sole
`
`

`

`U.S. Patent
`
`May 19, 1998
`
`Sheet 3 of 7
`
`5,754.298
`
`
`
`&
`
`x
`Xxxxx xxxx xxxx
`
`XXXX S.
`N: &S XXXXXXXXXXXXXXXXXX
`
`

`

`U.S. Patent
`
`May 19, 1998
`
`Sheet 4 of 7
`
`5,754.298
`
`
`
`89
`
`#7 ?un61
`
`| | | | | | | | | | | |
`
`

`

`U.S. Patent
`
`May 19, 1998
`
`Sheet 5 of 7
`
`5,754.298
`
`/8
`
`88
`
`JOSS0001g
`
`G ?un61–
`
`
`
`
`
`
`
`

`

`U.S. Patent
`US. Patent
`
`May 19, 1998
`May 19, 1998
`
`Sheet 6 of 7
`Sheet 6 of 7
`
`5,754.298
`5,754,298
`
`Metal
`
`Contacts
`
`Silicide
`
`co
`
`
`
`16
`
`5.
`
`E is Cl
`
` llllllltllllllli
`illlllllllllllllIHIHH
`
`Cl
`
`S.
`94
`
`p-Substrate [“l:! Figure6
` 90
`
`

`

`U.S. Patent
`
`May 19, 1998
`
`Sheet 7 of 7
`
`5,754.298
`
`
`
`Z ?un61–
`
`

`

`1
`METHOD AND APPARATUS FOR MAGNG
`SEMCONDUCTOR DEVICE PROPERTIES
`
`5,754.298
`
`2
`Falk et al. in "Electro-Optic Imagery of High-Voltage GaAs
`Photoconductive Switches." IEEE Trans. Electron Devices
`42, 43-9 (1995) and by J.C. Adams et al., "Electro-Optic
`Imaging of Internal Fields in (111) GaAs Photoconductors."
`IEEE Trans. Elect. Devices 42, 1081-85, (1995). These
`techniques involve analyzing the polarization rotation of
`light passing through a GaAs sample, wherein the optical
`wavelength is well below the absorption bandedge of the
`GaAs sample. In order to compensate for the multi-valued
`nature of the polarization rotation, specialized algorithms
`were utilized for processing the images. Although a remark
`able measurement, the techniques employed fail to work in
`semiconductors such as Si and are only useful for electric
`fields, i.e., they are not applicable to temperature or carrier
`density measurements.
`Heinrich et al. and Goldstein et al. have demonstrated
`optical, high-speed sampling of carrier density and thermal
`effects in semiconductors at a single spatial point. They
`reveal that electric field could be sensed indirectly through
`the change in carrier density which occurs in the depletion
`region of reversed biased junctions. The work of Heinrich et
`al. is described in "Noninvasive Sheet Charge Density Probe
`for Integrated Silicon Devices." Appl. Phys. Lett. 48, 1066
`8, (1986). The work of Goldstein et al. is described in
`“Heterodyne Interferometer for the Detection of Electric and
`Thermal Signals in Integrated Circuits through the
`Substrate." Rev. Sci. Instrum. 64. 3009-13 (1993). Both
`groups utilize interferometric means to extract a signal from
`the changes in refractive index caused by the two optical
`effects. In both cases, a pair of optical beams is brought
`through the back side of the semiconductor device. One
`beam, used as a reference. is reflected off a convenient point
`on the upper surface of the device and brought back into an
`optical detector. The second beam is positioned onto the
`point of interest, reflected off of the upper surface and
`combined with the reference beam to form the interferomet
`ric signal. In the case of Heinrich et al... a modified Nomarski
`interference microscope was utilized as the interferometric
`system. Goldstein et al. utilized a variant on a heterodyne.
`interference microscope.
`The detection schemes of Heinrich et al. and Goldstein et
`al. were performed at a single point. A seemingly obvious
`extension of their work would be to scan the optical beam(s)
`in order to assemble an image of the target. However, this
`presumption proves false in the face of actual semiconductor
`devices. In both techniques, the detected light returns to the
`optical system after reflection from either the metalization or
`the refractive index changes at the upper surface of the
`semiconductor. Reflections from the back surface are
`assumed to be rejected by the optical system. In a typical
`semiconductor, the optical path length to the reflective
`interface is not constant with position. Path length differs
`due to changes in the physical height of the upper surface
`and the refractive index of intervening layers. The optical
`path length variations are of the order of the optical probe
`wavelength in the semiconductor, i.e., large compared to the
`expected signals. In addition, large changes in the reflection
`coefficient occur as the optical probe is moved around the
`upper surface. Thus, unlike the previous single point detec
`tion methods. imaging requires the sensing of a small phase
`signal on top of large phase background phase signals and
`reflection amplitude variations.
`The method of Heinrich et al. is particularly ill suited to
`deal with these issues. Heinrich et al. requires the use of a
`convenient reference spot. As described, the reference spot
`is at a fixed distance from the probe and it is clear that not
`all points on the device are convenient. More specifically, as
`
`This invention was made with Government support
`under Contract No. F29601-C-0096 awarded by the Depart
`ment of the Air Force. The Government has certain rights in
`the invention.
`
`FELD OF THE INVENTION
`This invention relates to methods and apparatus for imag
`ing characteristics of a target, more specifically, methods
`and apparatus for accurately imaging internal characteristics
`of semiconductor circuits.
`
`BACKGROUND OF THE INVENTION
`
`Semiconductor Imagery
`The internal characteristics of a semiconductor affect the
`optical properties, the optical absorption coefficient and the
`refractive index, within the semiconductor. Specifically, the
`internal characteristics of electric field, temperature, and
`carrier density cause changes in both optical properties.
`Absorption shifts are typically strongest at optical wave
`lengths near the semiconductor bandedge, whereas the
`refractive index shifts occur over a broad range of wave
`lengths. In Table 1, theoretical estimates for the magnitudes
`of the optical properties in terms of small signal variations
`for a gallium arsenide (GaAs) semiconductor and the effects
`of electric field (E) or hole density, temperature (T), and
`electron density (N) on refractive index (n) and absorption
`coefficient (O), in GaAs are shown. Other semiconductor
`materials, such as silicon (Si) show similar results.
`
`O
`
`5
`
`25
`
`30
`
`TABLE 1.
`
`Refractive Index
`
`Absorption
`
`"** -Si--3 x 10 (env) -- = 3 x 10-(1,v)
`Temperature -- = 3 x 10 (1/K) -- = 7 (1/K'cm)
`
`Carrier Density s
`
`= 4x 102 (cm)
`
`s
`
`=3 x 10-19 (cm2)
`
`Absorption imagery in GaAs has been described by
`Schoenbach et al., "Temporal Development of Electric Field
`Structures in Photoconductive GaAs Switches." Appl. Phys.
`Lett. 63. 2100-2 (1993), and Falk et al. "Dynamic Optical
`Probing of High-Power Photoconductors." 9th IEEE Pulsed
`Power Conference, 88-91 (1993). In general, a wavelength
`near the material bandedge is utilized to illuminate the
`sample of interest and an image of the sample is obtained on
`an appropriate detector, e.g., a charged coupled device
`(CCD) camera. These techniques perform adequately for
`large changes in absorption, but have limited sensitivity
`when compared to interferometric techniques. Further, cali
`bration for absorption imagery can prove difficult due to the
`rapid changes in absorption with changes in the internal
`properties, as described in "Optical Probe Techniques for
`Avalanching Photoconductors." Eighth IEEE International
`Pulsed Power Conference, by R.A. Falket al., 29-32 (1991).
`Electric fields only affect the refractive index of semicon
`ductors which are noncentrosymmetric. For example, strong
`electro-optic effects occur in GaAs but not in Si. Techniques
`for obtaining electro-optic images in noncentrosymmetric
`semiconductors, specifically GaAs, are described by R.A.
`
`35
`
`40
`
`45
`
`50
`
`55
`
`65
`
`

`

`3
`the beams are scanned, the relative phase and amplitude
`between the two optical beams will shift, The phase shifts
`bring about difficulty in maintaining optimal signal-to-noise
`condition. In fact, a relative phase shift of TU2 can cause the
`small signal response to be zero. Amplitude shifts cause
`degradation in the signal-to-noise due to reduced fringe
`contrast. The temporal variations in amplitude and phase
`shifts resulting from the scanning process can also produce
`false signals. Even if means to scan the probe beam while
`fixing the position of the reference beam were introduced,
`these defects would still be present. Heinrich et al. neither
`discuss the option of imaging nor describe means of over
`coming the above difficulties.
`Goldstein et al. utilize heterodyne techniques, which
`convert the DC signals of Heinrich et al. into an AC signal
`at the heterodyne frequency. At first glance, the heterodyne
`methodology would appear to remove several of the above
`objectionable defects. However, the "spectral analysis"
`given by Goldstein et al. obscures several points and in one
`case yields incorrect conclusions. An analysis of a hetero
`dyne interferometer, which uses the standard starting points
`of time varying amplitude and phase, is given in Chen et al.
`"Measurement Principle and Error Analysis for an Optical
`Heterodyne Profilometer." T. Chen, Z. Li, J.B. Chen. Proc.
`SPIE 2101, 800-3 (1993). The Chen analysis clearly shows
`that the heterodyne signal is subject to the same defects as
`the Heinrich approach. However, as this point is rather
`important, the equations relating to phase detection in Gold
`stein et al. will be reviewed in some detail.
`It is first noted that the reference field (second equation in
`Equation 3 of Goldstein et al.) and the probe field (Equation
`7 of Goldstein et al.) are given the same time dependent
`amplitude coefficient, A(t). An independent term describing
`relative amplitude variations in the two fields is not
`included. The conclusion in Goldstein et al., that the final
`phase signal shown in their Equation 10 is independent of
`amplitude modulations, is therefore incorrect. Their Equa
`tion 10 contains no amplitude modulations simply because
`none were included in the starting fields. Allowing for
`independent variations in the amplitudes, A(t), clearly intro
`duces amplitude noise into Equation 10. Chen et al. shows
`the correct dependence on relative amplitudes and their
`effect on fringe contrast, etc.
`The second point about the Goldstein et al. analysis is
`somewhat more subtle due to the "spectral analysis"
`approach used. In their analysis. they assume that the signal
`is given by a single frequency phase modulation along with
`a fixed relative phase angle. d. General time dependencies
`are presumably analyzed by Fourier decomposition. Equa
`tion 10 therefore indicates that do is an unimportant additive
`phase, which can presumably be removed by AC coupling
`techniques. If d is not assumed to be constant, however, this
`is no longer the case. The spectral components of the d term
`will mix with the single frequency term in the cosine
`creating numerous sidebands and amplitude dependencies.
`The exact effect depends on the amplitudes of various
`frequency components and appears complex in the Gold
`stein et al. analysis. The Chen et al. analysis avoids the
`"spectral analysis" approach and clearly shows that the two
`phase effects (phase signal and background variations) con
`tribute equally to the final signal.
`Both the Heinrich et al. and Goldstein et al. techniques
`suffer reduced transverse resolution due to defocusing,
`which occurs as the beam is scanned over an uneven surface.
`Additionally, optical returns from intervening layers and
`objects (e.g. defects) can cause signal degradation in the
`Heinrich et al. technique. The Goldstein etal technique may
`
`4
`be able to reject out of focus returns. However, this conclu
`sion depends upon details of the optical system which are
`not given.
`
`Scanning Confocal Imagery
`Scanning confocal imagery differs from standard imaging
`systems in that individual image pixels are gathered
`sequentially, one at a time, from a spot focused onto the
`object being imaged. The main disadvantages of this system
`are the inability to capture images of single-shot events and
`the need for either a mechanical or optical scanning system.
`The former is fundamental, where as the latter simply adds
`system complexity in some areas while potentially decreas
`ing it in others. The advantages of the scanning system are
`numerous and include: 1) extended wavelength range while
`maintaining high detection efficiency, 2) improved trans
`verse resolution. 3) improved depth discrimination, 4)
`extended field of view, and 5) extended on-line processing.
`Confocal systems come in two arrangements Type 1 and
`Type 2 (for details of confocal systems, see Confocal
`Microscopy, T. Wilson, Ed. Academic Press (1990)). The
`distinction between the two types is that the Type 2 system
`focuses the light returned from the target through a pinhole
`prior to detection. Although a seemingly small difference,
`the inclusion of the pinhole causes profound differences in
`the optical performance of the two types of systems. Of
`specific interest herein is the difference in optical response
`with target distance.
`Assume that the target is moved back and forth along the
`optical axis. For small changes in the target position the
`detector in the Type 1 system will record no variation in
`optical intensity. This result derives from the detector being
`underfilled. For large variations in the target position, the
`detector will eventually become overfilled. At this point, a
`relatively slow 1/z variation in the optical intensity will
`occur, where z is the target displacement from the focused
`position. Brunfeld et al., "High Resolution Optical
`Profilometer.” SPIE 680, 118-23 (1986), have used this
`effect to produce a depth of focus sensor with resolution on
`the order of a tenth micron.
`Variation in the optical intensity with longitudinal dis
`placement of the target is quite different for the Type 2
`confocal system. The coherent nature of the Type 2 system
`yields a response of the form
`1(a)-ling)- with a -w--- (£) 2.
`where I is the peak optical response, C. is defined as the
`modified optical coordinate, u is the standardly defined
`optical coordinate, is the optical wavelength. O. is the lens
`diameter, f is the collection lens focal length and Z is the
`longitudinal target displacement from the lens focal spot.
`The squared part of the equation after the Io term is defined
`as the longitudinal transfer function of the system. This
`functionality falls of much faster than for the Type 1 system.
`For typical values of the variables, the intensity will fall off
`to half the peak value in less than 1 micrometer of target
`displacement.
`It is important to note that the symmetry of the optical
`system does not distinguish between where in the optical
`path from the target and the point detector (pinhole
`photodetector combination) that the optical path length, Z,
`varies. Specifically, changes in the point detector position or
`in the refractive index along the optical path produce equiva
`lent results to changing the target position.
`The strong intensity variation with target distance in a
`Type 2 confocal system has been used for years to obtain
`
`(1)
`
`2
`
`5,754.298
`
`10
`
`15
`
`25
`
`30
`
`35
`
`45
`
`55
`
`65
`
`

`

`5,754.298
`
`15
`
`25
`
`6
`pinholes, wherein the offset of the pinholes is mechanically
`set to a predetermined value, and the detectors are a pair of
`photodetectors. The system further includes an electronic
`circuit for producing an image signal, wherein the image
`signal is the difference in the signals generated by the pair
`of photodetectors divided by the sum of the signals gener
`ated by the pair of photodetectors. The image signal is
`produced according to the optical path length between the
`second lens and the target with the exact nature of the
`relationship being determined by the particular values of the
`pinhole offsets and the image signal being independent of
`the return optical intensity.
`In accordance with still other aspects of this invention, the
`imaging system further includes a controller for electroni
`cally controlling the spatial filter offsets, wherein the con
`troller includes an electro-optic substrate placed between the
`third lens and the two pinholes. The electro-optic substrate
`has a transparent electrode on one side which is common to
`both optical paths and a pair of transparent electrodes on the
`opposing side with each of the pair arranged to intersect only
`one of the pair of optical paths produced by the combination
`of the Wollastan prism and the third lens. Application of a
`voltage to one of the pair of transparent electrodes and the
`common electrode changes the refractive index of the part of
`the electro-optic substrate which intersects, changing one of
`the pair of optical paths and thereby adjusting the image
`signal.
`In accordance with still further aspects of this invention.
`an electronic feedback circuit changes the common pinhole
`offset according to changes in the optical path length
`between the second lens and the target. The electronic
`feedback circuit further includes a feedback amplifier which
`applies a common voltage between the pair of transparent
`electrodes and the common transparent electrode of the
`electro-optic substrate for maintaining a constant value of
`the image signal with the applied common voltage becoming
`a new image signal, thereby obtaining a closed loop system,
`wherein signal linearity is determined by the linearity of the
`electro-optic substrate and the closed loop system is able to
`track changes in the optical path length between the second
`lens and the target,
`In accordance with still yet other aspects of this invention,
`the imaging system further includes a changing mechanism
`for changing the intensity of the optical radiation point
`source to maintain a constant total intensity at the pair of
`photodetectors. The changing mechanism further includes a
`feedback amplifier which has as an input a signal propor
`tional to the sum of the intensities at the pair of photode
`tectors and applies a signal to a light source whose intensity
`is proportional to the signal so as to maintain a constant
`value of the sum of the outputs from the photodetectors,
`thereby allowing maintenance of a signal-to-noise ratio in
`the image signal which is independent of the targets optical
`characteristics.
`As will be readily appreciated form the foregoing
`summary, the invention provides a new and improved
`method and apparatus for imaging internal characteristics of
`a target, such as the internal characterisics of semiconductor
`circuits.
`
`5
`rough distance information and to produce 3-D images
`(Confocal Microscopy, T. Wilson, Ed., Academic Press
`(1990)).T.R. Corle et al. "Distance Measurements by Dif
`ferential Confocal Optical Ranging." Applied Optics 26,
`2416-20 (1987), were able to obtain interferometric accu
`racies and resolutions by using a technique that produced a
`signal proportional to the derivative of the intensity response
`in Equation 1 above. In their system, the target was vibrated
`at frequency, (), which results in a signal at that frequency
`which is proportional to the derivative. The main zero
`crossing of this derivative occurs at the peak response of the
`system and can be used to accurately determine the target
`distance. Resolution better than 0.01 nanometer was dem
`onstrated. There exist several difficulties with this technique:
`1) it is often impractical to vibrate the target of interest; 2)
`temporal sampling resolution is limited by the vibration
`frequency; 3) the system is open loop and if made closed
`loop, the loop closing time would be limited by the vibration
`frequency; 4) varying optical intensity can produce signal
`errors; and 5) ultra-high speed optical pulse sampling
`techniques, such as used by Heinrich et al. would prove
`difficult.
`The present invention is directed to overcoming the
`foregoing and other disadvantages. More specifically, the
`present invention is directed to providing a method and
`apparatus suitable for confocal imaging internal character
`istics of a target, such as the internal characterisics of
`semiconductor circuits.
`SUMMARY OF THE INVENTION
`In accordance with this invention, a method and apparatus
`for confocal imaging internal characteristics of a target, such
`as semiconductor circuits, is provided. An imaging system in
`accordance with this invention includes a radiant energy
`point source for generating radiant energy, and a mechanism
`for focusing the radiant energy generated by the point source
`onto a target and scanning the target with the focused radiant
`energy. The system also includes a collector for collecting
`the focused radiant energy that is scattered from the target
`and a splitter for splitting the collected radiant energy into
`two paths. Each of the two paths of the collected radiant
`energy is focused onto separate focal spots by a focusing
`mechanism. A pair of spatial filters are provided for filtering
`the collected radiant energy, one spatial filter for each of the
`two paths of the collected radiant energy. Each of the spatial
`filters is placed approximately at one of the focal spots. The
`spatial filters are offset from each other along the path of the
`focused radiant energy. The system further includes detec
`tors for separately detecting the focused radiant energy that
`passes through each of the spatial filters and produces
`signals proportional to the quantity of detected focused
`radiant energy present. Finally, the system includes a mecha
`nism for combining the produced signals into an image
`signal related to the distance traveled by the radiant energy
`from the focusing mechanism to the target and back to the
`collector.
`In accordance with other aspects of this invention, the
`radiant energy is optical radiation.
`In accordance with further aspects of this invention, the
`focusing mechanism includes a first lens for collimating the
`optical radiation, and a second lens for focusing the colli
`mated optical radiation onto the target. The collector
`includes the second lens for collimating the scattered optical
`radiation from the target along the optical path to the first
`lens, and a beam splitter for redirecting the optical radiation
`to the splitter, wherein the splitter is a Wollastan prism. The
`focusing mechanism is a third lens. the spatial filters are
`
`35
`
`45
`
`50
`
`55
`
`BRIEF DESCRIPTION OF THE DRAWTINGS
`The foregoing aspects and many of the attendant advan
`tages of this invention will become more readily appreciated
`as the same becomes better understood by reference to the
`following detailed description, when taken in conjunction
`with the accompanying drawings, wherein:
`
`65
`
`

`

`7
`FIG. 1 is a schematic diagram of a dual detector confocal
`imaging system;
`FIG. 2 is a graph of various signal response curves of the
`system of FIG. 1;
`FIG. 3 is a shematic diagram of the dual detector confocal
`imaging system of FIG. 1 with an electro-optic substrate;
`FIG. 4 is a functional block diagram of the dual detector
`confocal imaging system of FIG. 3 with an electronic
`feedback control circuit;
`FIG. 5 is a functional block diagram of the imaging
`system of FIG. 4 coupled to a processor and a display
`device;
`FIG. 6 is a functional diagram of the dual detector
`confocal imaging system of FIG. 4 used to image a semi
`conductor circuit; and
`FIG. 7 is a flow diagram of a method performed by the
`dual detector confocal imaging system of FIG. 4 used to
`image a semiconductor circuit.
`DETALED DESCRIPTION OF THE
`PREFERRED EMBODIMENT
`FIG. 1 sets forth a preferred embodiment of an optical
`system for imaging small refractive index changes in accor
`dance with the present invention, specifically as applied to
`imaging refractive index changes which occur in semicon
`ductor devices due to electric field, temperature, carrier
`density, and other properties.
`The imaging system includes an illumination portion and
`a light collection and detection portion. The illumination
`portion includes a point light source 10, such as a laser
`focused through a pinhole, a first lens 12 which collimates
`light from the point source, a beam splitter 14 and a second
`lens 16 which focuses collimated light on a focal spot or
`point 20 on a target 18. Between beam splitter 14 and second
`lens 16 is a scanner 15 which can be any one of a number
`of optical scanners typically used in confocal imagery, such
`as mechanical scanners, acousto-optic scanners and
`mechanical target motion scanners. The light collection and
`detection portion includes a target 18 which reflects focused
`light from second lens 16, with the second lens 16 and the
`beam splitter 14 as common elements with the illumination
`portion. In addition, the light collection and detection por
`tion includes a Wollastan prism 22 or equivalent device for
`splitting an optical beam into two parts, a third lens 24, a first
`pinhole 26, a second pinhole 28, a first detector 30 with a
`first detector output 34 and a second detector 32 with a
`second detector output 36. Finally, an intensity control 38
`modulates the optical intensity of point light source 10. Point
`light source modulation can take on many forms depending
`on the nature of the optical source. For example, an external
`electro-optic or acousto-optic modulator might be
`employed. Alternatively, the current applied to an optical
`source such as a laser diode or light emitting diode may
`supply the modulator.
`Operation of Invention
`The imaging system shown in FIG. 1 is a modified
`confocal imaging system. The configuration shown in FIGS.
`1 and 3 and described below is one example of an imple
`mentation of the confocal imaging system of the present
`invention. Other implementations, e.g. acoustic, electron
`beam, can also effectively operate within the confocal imag
`ing system of the present invention. The configuration
`shown performs reflection imagery.
`The optical paths in the system are indicated by the lines
`with arrows, a pair of lines indicating an optical beam
`
`45
`
`50
`
`55
`
`65
`
`5,754.298
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`8
`between components of the system. The arrows indicate the
`direction of the optical beam in each section of the system.
`As shown in FIG. 1, light emerging from point light source
`10 is collimated by first lens 12 and passes through beam
`splitter 14 and scanner 15 to second lens 16. Some light
`losses are expected at beam splitter 14. The use of suitable
`polarization optics minimizes light losses for polarized light.
`Second lens 16 focuses the light of point light source 10 onto
`target 18 at focal spot 20. Scanner 15 scans the focal spot
`transversely over the target. Conversely the optical light
`beam can remain stationary and the target moved trans
`versely to scan the section of interest. Confocal imagery
`utilizes the scanning process to produce an image of the
`target. For best operation of the confocal system, focal spot
`20is diffraction limited. Generally, diffraction limited opera
`tion requires that first lens 12 and second lens 16 are
`corrected for aberrations, such as spherical aberration. These
`corrections imply the use of compound lenses. In addition.
`the physical size of point light source 10 must be less than
`the diffraction limit of first lens 12. Relaxation of the
`diffraction limit requirement reduces the sensitivity of the
`system and the transverse resolution of the images. This
`reduction is acceptable for certain operations.
`The light beam reflected from target 18 is collected by
`second lens 16 and recollimated. Beam splitter 14 receives
`the reflected collimated light beam from second lens 16 and
`reflects the light beam received from second lens 16 to the
`Wollastan prism 22. The Wollastan prism 22 receives the
`reflected light beam from the beam splitter 14 and divides
`the reflected light beam into two divergent light beams.
`Beam splitters with properties similar to the Wollastan prism
`may be used in its place. The angle separation of the divided
`light beams is small enough to allow a single lens 24 to
`process them. The separation angle between the two output
`optical beams is determined by the type of the Wollastan
`prism 22.
`The two sets of two angularly separated beams produced
`by the Wollastan prism 22 pass through third lens 24. The
`third lens 24 focuses a first beam from the set of separated
`beams at a first pinhole 26. Also, the third lens 24 focuses a
`second beam from the set of separated beams at a second pin
`hole 28. The third lens causes the angular separation of the
`beams produced by the Wollastan prism 22 to produce a pair
`of focal spots separated transversely by a distance given by
`(for small angles) the product of the separation angle in
`radians and the focal length of the third lens. The first beam
`passes through pinhole 26 and is received by first detector 30
`and the second beam passes through pinhole 28 and is
`received b