JP2008020374A - Defect inspection method and device therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem a defect signal is missed and sensitivity is degraded by light scattered by a pattern in the irregular circuit pattern portion. <P>SOLUTION: An inspection device comprises: a stage for mounting a substrate sample and freely moving in each direction of X-Y-Z-θ; a lighting system for irradiating the substrate sample form an oblique position; and an image forming optical system for forming an inspection region that is lighted on a photoreciever, and the device condenses reflection scattered light produced on the substrate sample by irradiating the light in the lighting system. Further, the device comprises a polarized-light detector for simultaneously detecting a plurality of polarized components that are different from each other. Further, a plurality of the polarized-component signals different from each other that are detected by the polarized-light detector are compared in a plurality of chips or in an image in a predetermined region, and then a statistical outlier is inspected as a defect in the sample. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention detects and analyzes defects such as foreign matter generated in a manufacturing process of forming an object by forming a pattern on a substrate, such as a semiconductor manufacturing process, a liquid crystal display element manufacturing process, a printed circuit board manufacturing process, etc. The present invention relates to a defect inspection method and apparatus for inspecting the occurrence state of defects such as foreign matters in a manufacturing process for which measures are taken.

  In a conventional semiconductor manufacturing process, if foreign matter exists on a semiconductor substrate (substrate to be inspected), it may cause a failure such as a wiring insulation failure or a short circuit. Further, the semiconductor element is miniaturized and fine foreign matter is generated in the semiconductor substrate. If present, this foreign substance also causes a failure of insulation of the capacitor and a breakdown of the gate oxide film. These foreign substances are various due to various causes such as those generated from the moving part of the transfer device, those generated from the human body, those generated by reaction in the processing apparatus by the process gas, those mixed in chemicals and materials, etc. It is mixed in the state of.

Even in the same liquid crystal display element manufacturing process, if any defect occurs such as foreign matter mixed in the pattern, it cannot be used as a display element. The situation is the same in the manufacturing process of the printed circuit board, and mixing of foreign matters causes short-circuiting of the pattern and defective connection.
As one of the conventional techniques for detecting foreign matters on a semiconductor substrate of this type, as described in Japanese Patent Application Laid-Open No. 62-89336 (conventional technique 1), a semiconductor is irradiated with a laser to produce a semiconductor. By detecting scattered light from foreign matter generated when foreign matter is attached to the substrate and comparing it with the inspection result of the same type semiconductor substrate inspected immediately before, there is no false information due to the pattern, and high sensitivity and high reliability There are those that allow frequent foreign object and defect inspection. Further, as disclosed in Japanese Patent Application Laid-Open No. 63-135848 (Prior Art 2), the scattering from the foreign matter generated when the semiconductor substrate is irradiated with a laser and the foreign matter adheres to the semiconductor substrate. There is one that detects light and analyzes the detected foreign matter by an analysis technique such as laser photoluminescence or secondary X-ray analysis (XMR).

  In addition, as a technique for inspecting the above foreign matter, the substrate to be inspected is irradiated with coherent light, and the light emitted from the repetitive pattern on the substrate to be inspected is removed by a spatial filter to emphasize foreign matters and defects having no repeatability. There are known methods for detecting them.

  Further, the circuit pattern formed on the substrate to be inspected is irradiated from a direction inclined by 45 degrees with respect to the main line group of the circuit pattern, and the 0th-order diffracted light from the main line group is opened in the objective lens. A foreign substance inspection apparatus which is not allowed to be input to the inside is known from Japanese Patent Laid-Open No. 1-117024 (prior art 3). This prior art 3 also describes shielding other straight line groups that are not the main straight line group with a spatial filter.

  Further, as prior art relating to a defect inspection apparatus and method for foreign matters, etc., Japanese Patent Laid-Open No. 1-250847 (prior art 4), Japanese Patent Laid-Open No. 6-258239 (prior art 5), and Japanese Patent Laid-Open No. 6-324003. (Prior Art 6), JP-A-8-210989 (Prior Art 7), and JP-A-8-271437 (Prior Art 8) are known.

  Further, a surface inspection apparatus that simultaneously detects a plurality of polarization components and obtains the thickness of a thin film formed on a substrate to be inspected and the physical properties of the thin film is known in Japanese Patent Laid-Open No. 2006-145305 (prior art 9). Yes.

  Further, as a technique for simultaneously detecting a plurality of polarization components, a polarization measurement method using a channel spectrum is disclosed in Non-Patent Document 1, and a polarization measurement method using a birefringence wedge is disclosed in Non-Patent Document 1 and Non-Patent Document 2. Non-Patent Document 3 discloses a polarization measurement method using a split prism and a polarization measurement method using a micro polarization element array.

JP-A-62-89336 JP-A 63-135848 Japanese Patent Laid-Open No. 1-117024 JP-A-1-250847 JP-A-6-258239 JP-A-6-324003 JP-A-8-210989 JP-A-8-271437 JP 2006-145305 A Kazuhiko Oka, “Spectropolarimetry using channel spectrum”, O plus E,, Vol. 25, No. 11, p1248 (2003) K. Oka, "Compact complete imaging polarimeter using birefringent wedge prisms", Optics Express, Vol. 11, No. 13, p1510 (2003) Hisao Kikuta et al., "Polarization Image Measurement System", O plus E, Vol. 25, No. 11, p1241 (2003)

  However, in the prior arts 1 to 8, there is a problem that in the irregular circuit pattern portion, the defect signal is overlooked by the scattered light from the pattern, and the sensitivity is lowered.

  The prior art 9 is for obtaining the thickness of the thin film and the physical properties of the thin film, and does not directly contribute to improving the sensitivity of defect detection.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a defect inspection apparatus capable of inspecting a defect on a substrate to be inspected having a pattern that emits scattered light having the same intensity as that of a defect at high speed and with high accuracy in order to solve the above-described problems. And providing a method thereof.

  In order to achieve the above object, the present invention provides an illumination optical unit that guides light emitted from a light source to a predetermined region on a substrate to be inspected with a predetermined polarization state, a predetermined azimuth angle range and a predetermined In a defect inspection apparatus comprising: a detection optical unit that guides reflected and scattered light propagating to an elevation angle range of the light to a light receiver and converts the light into an electrical signal; and a defect determination unit that extracts a signal indicating a defect based on the electrical signal. The detection optical unit has polarization detection means for independently detecting a plurality of different polarization components and obtaining a plurality of signals corresponding to the respective polarization components, and the defect determination unit obtains each of the polarization components or their calculation. A signal indicating a defect is extracted based on a distribution of end points of vectors corresponding to the plurality of signals in a space spanned by a physical quantity.

  As described above, according to the above configuration, a minute defect can be inspected at high speed and with high accuracy on a substrate to be inspected having a pattern that emits scattered light.

  A first embodiment of the present invention will be described with reference to FIGS. Hereinafter, a defect inspection on a semiconductor wafer will be described as an example.

  FIG. 1 shows the configuration of the defect inspection apparatus according to the first embodiment. Light source 1, illumination optical system 100, inspection target substrate W, objective lens 3a, spatial filter 4a, imaging lens 5a, polarization detection unit 200a, signal processing unit 300, overall control unit 6, display unit 7, calculation unit 8, storage The unit 9 includes an XYZ-θ stage driver 10, an XYZ-θ stage 11, and a light source driver 12. The light source 1, the illumination optical system 100, the objective lens 3a, the spatial filter 4a, the imaging lens 5a, and the polarization detection unit 200a are combined to form an optical system 1000.

  An outline of the operation will be described. Light emitted from the light source 1 is applied to the inspection target substrate W by the illumination optical system 100. The reflected and scattered light emitted from the inspection target substrate W is collected by the objective lens 3a, passes through the detection system optical path 14 via the spatial filter 4a and the imaging lens 5a, and is converted into an electrical signal by the polarization detection unit 200a. The Based on the obtained electrical signal, the signal processing unit 300 determines a defect on the inspection target substrate. The determined result is stored in the storage unit 9 via the overall control unit 6 and displayed on the display unit 7.

  The spatial filter 4a is arranged at the exit pupil position of the objective lens 3a or a conjugate position thereof, and diffracted light generated by illuminating a repetitive pattern with a minute pitch formed on the inspection target substrate W. The light-shielding pattern is provided with a plurality of linear light-shielding patterns with variable pitches as described in, for example, Japanese Patent Laid-Open No. 2000-105203.

  In order to illuminate the inspection target substrate W with high illuminance, the light source 1 is suitably a laser light source. In order to increase the scattering efficiency of minute defects, short-wavelength light such as deep ultraviolet laser (DUV (Deep Ultraviolet) light), vacuum ultraviolet light laser, YAG laser third or fourth harmonic, mercury lamp, xenon lamp, etc. A light source is suitable. As a light source satisfying these requirements, a light source having a visible light wavelength such as a second harmonic of a YAG laser, a halogen lamp, a mercury lamp, or a xenon lamp is suitable for suppressing the cost of components constituting the optical system and the maintenance cost. A laser light source with a high degree of polarization is suitable for generating illumination light in a specific polarization state with high efficiency.

  FIG. 2A shows the configuration of the illumination optical system 100. The intensity of the illumination light of the light emitted from the light source 1 is controlled by the attenuator 101. A polarizing plate 102 is installed as necessary, and the polarization of illumination light emitted from the light source is aligned with linearly polarized light. The polarization state of the illumination light is arbitrarily set by the phase shifters 103 and 104. The phase shifters 103 and 104 are composed of λ / 2 plates or λ / 4 plates that can rotate around the optical axis, or phase shifters that can control the phase amount. The light transmitted through the phase shifters 103 and 104 is expanded in beam diameter by the beam expander 105. The illumination light whose beam diameter has been expanded by the beam expander 105 is guided onto the inspection target substrate W by the mirror groups M1 to M9 and the cylindrical lenses 109, 110, and 111. In FIG. 2A, since the mirror M4, the cylindrical lens 109, and the mirror M7 are in an overlapping state, the display of the cylindrical lens 109 and the mirror M7 is omitted. Since the relationship between the mirror M5 and the cylindrical lens 109 and the mirror M8 and the relationship between the mirror M6 and the cylindrical lens 110 and the mirror M9 are the same, the display of the cylindrical lenses 109 and 110 and the mirrors M8 and M9 is omitted.

  Hereinafter, a case where the optical path 106 is taken will be described as an example. By retracting the mirrors M1 and M2 from the optical path, the illumination light is reflected by the mirrors M3 and M4 and takes the optical path 106. FIG. 2B is a side view showing the configuration from the mirror M4 to the inspection target substrate W. FIG. The light is condensed on an elliptical or linear region F1 on the inspection target substrate W by the cylindrical lens 109. By translating and rotating the mirror M7 in the direction of the arrow, the angle formed with the surface of the inspection target substrate W (incident angle of illumination light to the inspection target substrate: elevation angle) can be changed.

  Similarly, regarding the optical path 107, the mirror M8 and the cylindrical lens 110 are provided between the mirror M5 and the inspection target substrate W. Similarly, regarding the optical path 108, the mirror M9 and the cylindrical lens 111 are provided between the mirror M6 and the inspection target substrate W. Is installed. The cylindrical lenses 110 and 111 are arranged around the tilt and the optical axis so that the illumination light passing through each of the cylindrical lenses 110 and 111 has the same central position and longitudinal direction of the condensing region as the region F1 irradiated by the cylindrical lens 109 on the inspection target substrate W. The rotation is added. With the above configuration, it is possible to selectively irradiate illumination light from a plurality of different azimuths and elevation angles, and to irradiate the same position on the inspection target substrate W. In addition, by configuring the mirror M1 and / or M2 with a half mirror, the region F1 on the inspection target substrate W can be illuminated simultaneously from a plurality of directions and elevation directions.

  In the optical path of the illumination optical system 100, a means for changing the optical condition of the illumination light at high speed is provided, and the optical condition of the illumination light is changed during a time shorter than the accumulation time of the light receiver of the polarization detector described later. By accumulating signals according to illumination optical conditions that have changed in the instrument, it is possible to increase the number of defect types that can be detected and improve the inspection S / N. As means for changing the optical condition of the illumination light at high speed, means for scanning a light beam position on a pupil described in JP-A-2000-193443, means for rotating a diffusion plate described in JP-A-2003-177102, etc. Is mentioned.

  An enlarged image of the illumination area F1 on the surface of the inspection target substrate W is formed by the objective lens 3a and the imaging lens 5a. Since the diffracted light of the periodic pattern formed on the inspection target substrate W is condensed at the pupil conjugate position of the objective lens 3a, it is formed on the inspection target substrate W by shielding it with the spatial filter 4a. The periodic pattern image is removed.

  The polarization detection unit 200a will be described with reference to FIGS.

  FIG. 3A shows a configuration of a polarization detection unit 200a ′ that detects two different types of polarization components as a configuration realized by the amplitude division method of the polarization detection unit 200a. The polarization detection unit 200a ′ includes a non-polarization beam splitter (half mirror) 201, a polarizing plate or a phase plate, and polarization selection means 210 and 211 that can adjust the polarization state of transmitted light, and a light receiver 220, 221. Each of the light receivers 220 and 221 is installed so as to detect an image of the surface of the inspection target substrate W that is enlarged and formed by the objective lens 3a and the imaging lens 5a. The image plane conjugate position with respect to the image on the surface of the inspection target substrate W by the objective lens 3a and the imaging lens 5a is indicated by an alternate long and short dash line as the image plane 230 (the one-dot chain line on the front surface of the light receiver 221 also represents the image plane conjugate position). By using an area sensor, linear sensor, or TDI (Time Delay Integration) sensor as the light receivers 220 and 221, an image corresponding to each polarization component can be obtained.

  As an area sensor, linear sensor, or TDI sensor system, a time integration type light receiver such as a CCD system or a CMOS system is used. In the illumination optical system 100, the optical conditions are set at high speed within a time shorter than the integration time of the light receivers 220 and 221. By changing, it is possible to integrate and detect scattered light by illumination light under a plurality of optical conditions.

In addition, when a photomultiplier tube is used as the light receivers 220 and 221, high sensitivity detection is realized.
Here, the linearly polarized light component in the direction parallel to the main wiring pattern on the inspection target substrate W is detected by the light receiver 220, and the linearly polarized light component in the direction perpendicular to the main wiring pattern on the inspection target substrate W is detected by the light receiver 221. This will be described as an example.

  Of the light components that have passed through the non-polarizing beam splitter 201, the light component that has passed through the polarization selecting means 210 constituted by a polarizing plate that transmits a linearly polarized light component parallel to the main wiring pattern on the inspection target substrate W is: , Detected by the light receiver 220. On the other hand, among the light components reflected by the non-polarizing beam splitter 201, the light transmitted through the polarization selection unit 211 constituted by the polarizing plate that transmits the linearly polarized component in the direction perpendicular to the main wiring pattern on the inspection target substrate W. The component is detected by the light receiver 221.

  As another configuration example that realizes an equivalent function, a polarization beam splitter that transmits a linearly polarized component having an orientation parallel to the main wiring pattern on the inspection target substrate W is disposed instead of the non-polarization beam splitter 201, As the selection means 211, it is possible to install a polarizing plate that transmits a linearly polarized light component having an orientation perpendicular to the main wiring pattern on the inspection target substrate W. The former has an advantage that it is possible to cope with the change of the polarization direction to be detected only by changing the polarization selection means 210 and 211. The latter has the advantage that it is not necessary to consider the polarization characteristics remaining in the non-polarizing beam splitter 201, and the polarization measurement can be performed with higher accuracy than the former. In addition, by detecting linearly polarized light components orthogonal to each other as described above, the main wiring pattern on the inspection target substrate W of the total light intensity and components not depending on the polarized light components is calculated based on the obtained measurement values. The physical quantity relating to the polarization, such as the linear polarization degree in the direction parallel to the long axis and the major axis azimuth angle of the (elliptical) polarization, can be calculated.

  FIG. 3B shows a configuration of a polarization detection unit 200a ″ that detects four different types of polarization components as a configuration realized by the amplitude division method of the polarization detection unit 200a. The polarization detection unit 200a ″ is configured by non-polarization beam splitters 202 to 204, polarization selection means 212 to 215, and light receivers 222 to 225. The light that has entered the polarization detection unit 200a ″ along the detection system optical path is branched by the non-polarization beam splitters 202 to 204, and each enters the separate light receivers 222 to 225. Each of the polarization selection means 212 to 215 is configured by a combination of a polarizing plate or a phase plate, and is set so that the polarization state of transmitted light can be adjusted independently.

  Each of the light receivers 222 to 225 is installed so as to detect an image of the surface of the inspection target substrate W that is enlarged and formed by the objective lens 3a and the imaging lens 5a. Similarly to the case of FIG. 3A, a one-dot chain line in front of each of the light receivers 222 to 225 indicates an image plane conjugate position with respect to an image on the surface of the inspection target substrate W by the objective lens 3a and the imaging lens 5a. By using an area sensor, a linear sensor, or a TDI (Time Delay Integration) sensor as the light receivers 222 to 225, an image corresponding to each polarization component is obtained.

  As an area sensor, linear sensor, or TDI sensor system, a time integration type light receiver such as a CCD system or a CMOS system is used. In the illumination optical system 100, the optical conditions are set at high speed within a time shorter than the integration time of the light receivers 220 and 221. By changing, it is possible to integrate and detect scattered light by illumination light under a plurality of optical conditions.

  In addition, when a photomultiplier tube is used as the light receivers 220 and 221, high sensitivity detection is realized.

  Here, a linearly polarized light component having a predetermined azimuth (α) around the detection system optical path 14 is received by the light receiver 222, a linearly polarized light component having an azimuth α + 90 degrees is received by the light receiver 223, and a straight line having an azimuth direction of α + 45 degrees is received by the light receiver 224 The polarization component will be described by taking as an example a case where the counterclockwise circular polarization component is detected by the light receiver 225.

  The light component transmitted through the non-polarizing beam splitter 202 is further branched by the non-polarizing beam splitter 203. The light component reflected by the non-polarizing beam splitter 203 is transmitted through the polarization selection means 212 constituted by a polarizing plate that transmits the linearly polarized light component having the predetermined azimuth α, and is detected by the light receiver 222. The light component that has passed through the non-polarizing beam splitter 203 passes through the polarization selecting means 213 constituted by a polarizing plate that transmits the linearly polarized light component having the azimuth α + 90 degrees, and is detected by the light receiver 223. The light component reflected by the non-polarizing beam splitter 202 is further branched by the non-polarizing beam splitter 204. The light component that has passed through the non-polarizing beam splitter 204 passes through the polarization selection unit 214 configured by a polarizing plate that transmits the linearly polarized light component having the azimuth α + 45 degrees, and is detected by the light receiver 224. The light component reflected by the non-polarizing beam splitter 204 is transmitted through the polarization selecting means 215 constituted by a λ / 4 plate set at 0 ° azimuth and a polarizing plate that transmits the linearly polarized light component at azimuth α + 45 °. 225.

Assuming that the intensities of the light components detected by the light receivers 222, 223, 224, and 225 are I1, I2, I3, and I4, respectively, the Stokes parameter that represents the polarization state of the light component incident on the polarization detection unit 200a is represented by the following equation. S0 to S3 can be obtained, and the polarization state can be completely determined. Further, based on the Stokes parameters S0 to S3, the degree of polarization, ellipticity, and the like can be calculated in addition to the physical quantity related to the polarization.
S0 = I1 + I2
S1 = I1-I2
S2 = 2 * I3- (I1 + I2)
S3 = 2 * I4- (I1 + I2)
A configuration for detecting three different types of polarized light components can be easily inferred from FIGS. 3 (a) and 3 (b). As three different types of polarized light components, a linearly polarized light component with a predetermined azimuth α around the detection system optical path 14, a linearly polarized light component with an azimuth α + 45 degrees, and a counterclockwise circularly polarized light component are detected. By assuming that the incident light component is completely polarized, it is possible to determine the polarization state of the light component incident on the polarization detector 200a.

  4 and 5 show a configuration example of the polarization detector 200a different from the configuration shown in FIG.

  FIG. 4A shows a configuration of a polarization detecting unit 200a ′ ″ using a birefringent wedge. The polarization detection unit 200a ′ ″ includes a frequency modulation image acquisition unit 250 and a Fourier analysis unit 255. FIG. 4B shows the configuration of the frequency modulation image acquisition unit 250. The frequency-modulated image acquisition unit 250 uses a predetermined azimuth (around 0 degree azimuth) around the detection optical axis 14 as a fast axis and 90 degrees azimuth as a slow axis, and the phase amount linearly changes in the azimuth 90 degrees direction. The prism element 251, the prism element 252 having a 45-degree azimuth as the fast axis and the 135-degree azimuth as the slow axis and the phase amount linearly changing in the 0-degree direction, and the polarized light transmitting the linearly polarized light component having the 0-degree azimuth It consists of a plate 253 and an image sensor 254. The image sensor 254 is installed so as to detect an image that is transmitted through the prism elements 251 and 252 and the polarizing plate 253 and enlarged and formed by the objective lens 3a and the imaging lens 5a. With the above configuration, the image sensor 254 outputs an image signal spatially modulated at a different frequency for each polarization component. The output image signal is subjected to FFT processing in the frequency analysis unit 255 and subjected to frequency analysis, whereby a plurality of parameters corresponding to the polarization state are obtained for each position on the image.

  FIG. 5A shows a configuration of a polarization detection unit 200a ″ ″ using a micro polarization element array. The polarization detection unit 200a ″ ″ includes an image sensor 261 and a micro polarization element array 262 installed on the light receiving surface thereof. The structure of the micropolarizing element array 262 is shown in FIG. It is configured to transmit different polarization components for each pixel. In the example of FIG. 5B, the pixel 263 that transmits linearly polarized light (azimuth 0 degree) parallel to the horizontal direction of the image sensor 261 pixel array and the pixel 264 that transmits linearly polarized light parallel to the vertical direction (azimuth 0 degree). , A pixel 265 that transmits linearly polarized light with a 45 degree azimuth, a pixel 266 that transmits a linearly polarized light with a 45 degree azimuth after giving a phase delay of 90 degrees to the 0 degree and 0 degree azimuth, and 1 pixel as one unit 267 Two polarization states are obtained.

  As a method for producing such a micropolarizing element array 262, a thin-film polarizing plate having a thickness of micron order to submicron order is placed on an image sensor or a substrate, and unnecessary portions are removed by etching according to the size of the pixel. In addition, there is a method of repeating similar patterning by placing thin film polarizing plates or wave plates having different principal axis orientations. As another method, there is a method of giving optical anisotropy to each pixel by forming a fine grating having a period shorter than the wavelength of light to be used by patterning. Further, the optical resolution (diameter of the circle of confusion) determined by the imaging performance of the objective lens 3a and the imaging lens 5a is set to be equal to or greater than the combined width of four pixels as one unit for determining the polarization state. It is possible to reduce the influence of the change in image plane intensity between pixels and perform highly accurate polarization measurement.

  By scanning the XYZ-θ stage 10 in the X and Y directions, the visual field on the inspection target substrate W determined by the objective lens 3a, the imaging lens 5a, and the polarization detection unit 200a ″ ″ is inspected. Scanning is performed relative to the target substrate W. By sequentially repeating the X-direction scanning and the Y-direction scanning, a polarization component detection signal can be obtained from the entire inspection target substrate W or a part thereof.

  The configuration of the signal processing unit 300 is shown in FIG. FIG. 6A shows a defect based on the difference in signal output from the polarization detection unit 200a ′ (FIG. 3A) between adjacent chips on the inspection target substrate W on which the same pattern is formed by design. The structure of signal processing part 300 'which implement | achieves the system to determine is shown.

The signal processing unit 300 ′ illustrated in FIG. 6A includes delay memories 301 and 302, difference calculation units 303 and 304, a defect determination unit 305, and a defect determination reference calculation unit 306, and is output from the polarization detection unit 200a ′. The defect information 307 is output in response to the input of the signal I _k (signals I ₁ and I ₂ corresponding to two polarization components in FIG. 6A). Next, the operation will be described. The signal I ₁ is input to the difference calculation unit 303 and the delay memory 301. The delay memory 301 accumulates the signal and outputs it after delaying the processing time of one chip. The difference calculation unit 303 receives the signal I ₁ and the signal at the corresponding position of the adjacent chip output from the delay memory 301 and outputs the difference ΔI ₁ between them. The obtained difference signal of the signal I _{1 from} the adjacent chip is input to the defect determination unit 305 and the defect determination reference calculation unit 306. The difference between the same adjacent chips for the signals I ₂ corresponding to the polarization components to the polarization component different corresponding to the signal I ₁ is taken (ΔI _2), are input to the defect determination unit 305 and the defect judgment criterion calculating unit 306 . The defect determination reference calculation unit 306 outputs a defect determination reference 308 based on the adjacent chip difference signals ΔI ₁ and ΔI ₂ . The defect determination unit 305 performs defect determination based on the defect determination reference 308 with respect to the input of the inter-adjacent-chip difference signals ΔI ₁ and ΔI ₂ , and outputs defect information 307.

FIG. 6B corresponds to a predetermined area on the inspection target substrate W, and a defect is determined based on a series of signals (image signals) output from the polarization detector 200a ′ (FIG. 3A). The structure of signal processing part 300 '' which implement | achieves the system to perform is shown. Image signal I ₁ output from the polarization detecting unit 200a 'has a predetermined polarization component, which corresponds to an image signal detected for each position of a predetermined region on the substrate to be inspected W. Image signal I ₂ outputted from the polarizing detector 200a 'also includes the image signal I ₁ and the different polarization components, corresponding to the image signal detected for each position of a predetermined region on the substrate to be inspected W. The image signals I ₁ and I ₂ are input to the defect determination reference calculation unit 311 and the defect determination unit 312. The defect determination criterion calculation unit 311 outputs a defect determination criterion 314 based on the image signals I ₁ and I ₂ . The defect determination unit 312 performs defect determination based on the image signals I ₁ and I ₂ and the defect determination standard 314, and outputs defect information 313. The defect determination standard calculation unit 306 or 311 includes a memory, and a plurality of previously detected plural The defect criterion 308 or 314 can also be calculated based on the polarization component detection signal at the corresponding position in the chip. The defect information 307 or 313 output from the signal processing unit 300 ′ or 300 ″ includes the defect position, the defect difference image, the defect difference image for each polarization component, the defect feature amount calculated from the defect difference image, and the defect. Includes classification results. The defect classification can be performed in the defect determination unit 305 or 312, or can be performed in the calculation unit 8 based on the defect information 307 or 313.

Further, here, a case where the two polarization components I ₁ and I ₂ output from the polarization detection unit 200a ′ described in FIG. 3A are processed as an example of the configuration of the signal processing unit 300 ′ or 300 ″ will be described. However, it is output from each of the polarization detection unit 200a ″ described in FIG. 3B, the polarization detection unit 200a ′ ″ described in FIG. 4, and the polarization detection unit 200a ″ ″ described in FIG. The configuration described with reference to FIGS. 6A and 6B can also be applied when detecting four polarization components. That is, when the image signals I _{1 to} I ₄ composed of four polarization components are input and processed, the circuit configuration for processing the image signals I ₁ and I ₂ described with reference to FIGS. This can be easily realized by adopting a configuration for processing one input signal.

  A defect determination standard calculation method in the defect determination reference calculation unit 306 or 311 (hereinafter referred to as a defect determination reference calculation unit) and a defect determination method in the defect determination unit 305 or 312 (hereinafter referred to as a defect determination unit) will be described with reference to FIGS. explain.

  A method for determining a defect using a signal obtained by detecting two different polarization components will be described with reference to FIG.

FIGS. 7A-1 and 7A-2 show examples of the prior art for determining defects using only a single polarization component. Figure 7 (a-1) shows the distribution of the polarization component signal _{I 1.} Each mark A~F plotted shows the polarization component signals I ₁ corresponding to each of the plurality of chips stored in the defect judgment criterion calculating unit. From the distribution of the symbols A to F corresponding to each of the marked marks ◯ Δ ×, many are included in the I ₁ value range 401, and only the signal of A is outside the range 401. A range 401 corresponds to a defect determination criterion calculated from statistics such as an average value and standard deviation of the distribution of each plotted mark. By determining the inside of the range 401 as a normal part and the outside as a defective part, it is correctly determined that A is a defective part and C to F are normal parts, but B is erroneously determined as a normal part. End up.

  On the other hand, FIG. 7A-2 shows the distribution of the polarization component signal I2, and when the defect determination is performed by calculating the range 402 corresponding to the defect determination reference by the same method as FIG. Although B is a defective part and C to F are correctly determined to be normal parts, A is erroneously determined to be a normal part.

Figure 7 (a-3) is a diagram illustrating a method of determining a defect by utilizing the two polarization components different from each other according to the present invention, the horizontal axis polarization component signal I _1, the vertical polarization component signal I2 taken as the axis is a plot of polarization component signals (I _{1, I 2)} corresponding to each of the plurality of chips stored in the defect judgment criterion calculating unit. In the defect determination criterion calculation unit 306 or 311, a rectangular region 404 including a large number of distributions is calculated as a defect determination criterion using a statistical amount such as an average value or standard deviation of the distribution of each plotted point, The determination unit determines that the inside of the rectangular area 404 is a normal part and the outside of the rectangular area 404 that is an outlier is a defective part, so that A and B are both defective parts, and C to F are normal parts. Defect determination can be performed correctly without erroneous determination. Even if the circular region 403 is calculated using a statistical value such as an average value or standard deviation of the distribution of each plotted point as a defect determination criterion, the correct determination can be similarly performed.

As shown in FIG. 7B, the signal plotted outside the range 401 with respect to the polarization component signal I ₁ (I ₁ <Th1-OR I ₁ > Th1 +: where TH1 + is the upper limit value of the range 401 , TH1- indicates a lower limit value of the range 401), a criterion (J1) that is defective is applied, and the signal (I ₂ <Th2-OR I ₂₎ plotted outside the range 402 with respect to the polarization component signal I2 > Th2 +: where TH2 + represents the upper limit value of the range 402 and TH2- represents the lower limit value of the range 402), and applies the criterion (J2) that satisfies either criterion (J1) OR Even when (J2) is applied as the final defect determination criterion, the correct determination can be made as described above. This is equivalent to a defect determination condition that a signal plotted outside the rectangular area 404 is a defect.

Further, as shown in FIG. 7 (c), the value obtained by a predetermined arithmetic processing to the polarization component signal I ₁ and the polarization component signal I2 is plotted, delimits a determining value and the normal portion or the defective portion relative to this It is also possible to determine defects. FIG. 7C shows the calculation for the polarization component signal I ₁ and the polarization component signal I ₂ .
f (I ₁ , I ₂ ) = (I ₁ -a) ² + (I ₂ -b) ²
This is an example of plotting A to F and determining a signal (f (I ₁ , I ₂ )> Th) plotted outside the range 405 as a defect. Even with this method, a correct determination can be made as described above. This is equivalent to a defect determination condition that a signal plotted outside the circular area 403 is a defect. In general, when a defect determination condition is expressed by N the following formula I ₁ and I _2, I ₁ and the I ₂ plots the signals on the plane mapped as an axis, the normal portion or defect range included in N order curve The problem can be determined as part.

  It is also possible to determine the defect by plotting the polarization component signal with the physical quantity obtained based on the plurality of polarization component signals as an axis. As shown in FIG. 8 (a), a physical quantity obtained by an arbitrary calculation on a plurality of polarization component signals is taken on each axis.

  FIG. 8B shows an example in which the horizontal axis represents the total intensity of light and the vertical axis represents the ellipticity of polarization as a physical quantity. When scattering due to particulate defects such as foreign matter is considered, in the Rayleigh scattering region where the particle size is small with respect to the wavelength of light, the scattered light is also linearly polarized with respect to linearly polarized illumination, whereas the wavelength of light is On the other hand, it is known that the scattered light becomes elliptically polarized light with respect to the linearly polarized illumination in the Mie scattering region where the particle diameter is equal or larger. Therefore, the ellipticity of the polarization component of the scattered light to be detected tends to increase as the defect size increases. Accordingly, it is possible to estimate the size of the defect based on the detected ellipticity of the polarization component of the defect portion.

  FIG. 8C shows an example in which the horizontal axis represents the total intensity of light and the vertical axis represents the major axis azimuth of the polarization as physical quantities. It is known that the polarization direction of the reflected scattered light with respect to the polarization direction of the illumination light may differ depending on the type of defect or pattern. FIG. 8C shows an example in which foreign substances and scratches are classified based on the major axis azimuth angle of the detected polarization component of the defect.

  FIG. 8 (d) calculates the amplitude reflectance ratio Ψ and phase difference Δ used in ellipsometry from the polarization state of illumination light, which is known as a physical quantity, and a plurality of detected polarization component signals. This is an example taken on the axis. Information on the film thickness and refractive index of the thin film at each position can be obtained from these physical quantities, and therefore processing based on the information can be performed.

  FIG. 9 is an example in which values based on three physical quantities obtained based on a plurality of polarization component signals are plotted, and polarization component signals or values obtained by their calculation are plotted. FIG. 9A shows an example in which the polarization detector 200a acquires four different polarization component signals to obtain the Stokes parameters S0 to S3, and the values obtained by normalizing the Stokes parameters S1 to S3 with the Stokes parameter S0 are used as axes. is there. This corresponds to the polarization state represented by a Poincare sphere with a radius of 1, and the point corresponding to the polarization state is plotted on the Poincare sphere if it is completely polarized, or inside the Poincare sphere if it is partially polarized.

  As a defect determination criterion, an area in the three-dimensional space that is determined to be a defective part or a normal part is calculated, and defect determination is performed. Since normalization is performed with the light intensity S0, defect detection based on a distribution reflecting the difference in polarization state is possible without being affected by variations in the brightness of the original scattered light. 9B, in the Poincare sphere display, when the S1-S2 plane is the equator plane, the latitude is the ellipticity angle of polarization (the arctangent of the ellipticity), and the longitude is the major axis azimuth of the polarization. Therefore, the defect size can be estimated and the defect type can be classified in the same manner as the method shown in FIGS. 8B and 8C. Further, as shown in FIG. 9C, defect detection based on a distribution in which light intensity is taken into account in addition to the polarization state can be performed by plotting S1 to S3 without being normalized by S0. Is possible.

  A first modification of the first embodiment will be described with reference to FIGS.

  FIG. 10 shows the configuration of the first modification. The objective lens 3b, the spatial filter 4b, the imaging lens 5b, and the polarization detector 200b are added to the first embodiment. The oblique detection system 500b is configured by the objective lens 3b, the spatial filter 4b, and the imaging lens 5b, and a light component reflected and scattered at an elevation angle and an azimuth angle different from those of the objective lens 3a is guided to the polarization detection unit 200b. The configuration of the polarization detector 200b can employ any of the configurations 200a ′ to 200a ″ ″ described as the configuration of the polarization detector 200a in the first embodiment. The configuration of the polarization detection unit 200b is preferably the same as the configuration of the polarization detection unit 200a, but the configuration is not necessarily the same. A plurality of polarization component signals detected by the polarization detection unit 200b are input to the signal processing unit 300 in the same manner as those detected by the polarization detection unit 200a. The signal processing unit 300 makes a defect determination based on the plurality of polarization component signals detected by the polarization detection unit 200a and the plurality of polarization component signals detected by the polarization detection unit 200b. In addition, a signal processing unit 300b (not shown) may be provided separately from the signal processing unit 300, and defect determination based on a plurality of polarization component signals detected by the detection unit 200b may be performed independently of the signal processing unit 300. Is possible.

  FIG. 11 shows the relationship between the detection direction of the oblique detection system 500b and the illumination direction. The oblique detection system 500b is installed so that the detection direction is the same as the stage X direction. The illumination azimuth can be selected as described in FIG. Here, the illumination traveling azimuth angle relative to the stage X direction is θ (when θ is 0 °, the same azimuth angle as the detection direction), and θ is between 0 ° and −90 °, and − An example of between 90 degrees and -180 degrees is shown. The arrangement in which θ is between 0 ° and −90 ° is suitable for detecting a relatively large defect with respect to the wavelength because forward scattered light due to the defect is incident on the oblique detection system 500b. On the other hand, an arrangement in which θ is between −90 degrees and −180 degrees is suitable for detecting defects that are relatively small with respect to the wavelength because backscattered light due to defects enters the oblique detection system 500b.

  FIG. 12 shows the relationship between the detection direction of the oblique detection system 500b, the main scanning direction St1 and sub-scanning direction St2 of the XYZ-θ stage 10, and the longitudinal direction of the illumination area F1. By making the longitudinal direction of the illumination area F1 perpendicular to the main scanning direction, the entire surface of the inspection target substrate W can be efficiently scanned. By making the detection direction parallel to the main scanning direction S1 and perpendicular to the illumination area F1, the inspection target substrate W can be inspected at a high throughput when a linear sensor is used as the light receiver of the oblique detection system 500b. it can.

  FIG. 13 shows a configuration example different from FIG. 11 with respect to the illumination direction and detection direction. Illumination light is incident in a direction perpendicular to the stage main scanning direction St1, the cylindrical surface rotation axis of the cylindrical lens SL is parallel to the stage main scanning direction St1, and the longitudinal direction of the illumination area F1 is perpendicular to the stage main scanning direction St1. A cylindrical lens SL is arranged. As a result, the short direction of the illumination area F1 can be narrowed narrowly compared with the case where the illumination elevation angle is inclined in the aperture direction of the illumination light, and high illumination efficiency can be obtained, and XYZ-θ stage 10 scanning is performed. Since the positional variation in the short direction of the illumination area F1 due to the minute variation in the Z direction can be suppressed, the detection sensitivity is stabilized.

  FIG. 14 shows a configuration of an optical system 1000 in a second modification of the first embodiment. The illumination light emitted from the light source 1 is guided to the illumination area F1 on the inspection target substrate W by the half mirror 150 and the objective lens 3a through the illumination optical system 100 '. According to this configuration, the bright field image is detected by the polarization detection unit 200. The detection signal output from the polarization detection unit 200 is processed in the signal processing unit 300, and a defect is detected.

  FIG. 15 shows a configuration of an optical system 1000 in a third modification of the first embodiment. The illumination light emitted from the light source 1 is guided to the illumination area F1 on the inspection target substrate W by the dark field objective lens 3a 'via the illumination optical system 100'. According to this configuration, the annular illumination dark field image is detected by the polarization detection unit 200.

The configuration of the illumination optical system 100 ′ is shown in FIG. The intensity of the illumination light is controlled by the attenuator 101 ′. A polarizing plate 102 ′ is provided as necessary, and the polarization of the illumination light emitted from the light source is aligned with linearly polarized light. The polarization state of the illumination light is arbitrarily set by the λ / 2 plate 103 ′ and the λ / 4 plate 104 ′ that can rotate around the optical axis. In addition, when using a laser light source as the light source 1, speckle noise generation | occurrence | production can be suppressed by installing speckle reduction means 111 '. As speckle reduction means, a method of generating and superimposing a plurality of light beams having different optical path lengths using a plurality of optical fibers or quartz plates or glass plates having different optical path lengths, or a rotating diffusion plate is used. FIG. 17 shows the configuration of the optical system 1000 and the XYZ-θ stage 10 in the third modification of the first embodiment. A strobe light source that emits light intermittently is used as the light source 1 ′. Specifically, a pulse laser, an LD excitation Q-switch pulse laser, a lamp excitation Q-switch pulse laser, a flash lamp, and the like are suitable. In addition, an area sensor is used as a light receiver of the polarization detection unit 200. With this configuration, a stroboscopic image can be obtained by synchronizing the light emission of the light source 1 ′, the scanning of the XYZ-θ stage 10, and the signal accumulation of the light receiver, thereby obtaining a distortion-free two-dimensional image. High precision chip comparison becomes possible. Further, by using the r-θ rotation stage 10 ′ instead of the XYZ-θ stage 10, it is possible to scan the entire surface of the inspection target substrate W at a higher speed than XY scanning.

  FIG. 18 shows the configuration of a fourth modification of the first embodiment. Inspected substrates W1 and W2 are designed to have the same pattern. The light emitted from the light source 1 ′ is branched and applied to each of the inspection target substrates W <b> 1 and W <b> 2 placed on the r-θ rotation stage 10 ′. The reflected and scattered light from the inspection target substrate W1 enters the polarization detection unit 200 via the objective lens 3a and the imaging lens 5a. The reflected and scattered light from the inspection target substrate W2 enters the polarization detection unit 200-2 via the objective lens 3a-2 and the imaging lens 5a-2. Strobe imaging is performed in synchronization with light emission of the light source 1 ′, scanning of the XYZ-θ stage 10, signal accumulation of the light receiver in the polarization detector 200, and signal accumulation of the light receiver in the polarization detector 200-2. Do.

    When rotational scanning is performed using the r-θ rotation stage 10 ′, as shown in FIG. 19A, the pattern formed on the inspection target substrate depends on the visual field position on the inspection target substrates W, W1, and W2. The polarization state of illumination and the orientation of the polarization detector 200 or 200-2 are different. Inspection by changing the polarization state of the illumination light to circularly polarized light or non-polarization symmetric around the optical axis, or rotating the major axis azimuth angle of the polarization state of the illumination light according to the stage rotation angle corresponding to the visual field position It is possible to keep the polarization state of illumination with respect to the pattern formed on the target substrate constant. Further, as shown in FIG. 19B, the pattern formed on the inspection target substrate is corrected by rotating the azimuth angle of the detected polarization component according to the stage rotation angle corresponding to the visual field position. It is possible to remove the influence of the rotation of the polarization detection unit orientation on the entire surface of the inspection target substrate with a constant sensitivity regardless of the rotation of the polarization detection unit orientation with respect to the pattern formed on the inspection target substrate.

Next, as a second embodiment, an embodiment of an optical system that changes to the illumination optical system 100 when a pulsed UV laser light source 2001 is used instead of the light source 1 in the configuration described in FIGS. 20 to 23 will be used for explanation.

  When a pulsed UV laser is used as the light source 2001, for example, in order to obtain scattered light having a sufficient intensity to detect from a very small particle defect having a size of about 10 nm, it is necessary to increase the amount of the pulse laser to be irradiated. As a result, the peak value (maximum output) becomes very large with respect to the required average output of the pulsed laser. For example, in the case of a laser with an average output of 2 [W], a light emission frequency of 100 MHz, a pulse interval of 10 [ns], and a pulse width of 10 [ps], the peak value (maximum output) is 2 [kW], causing damage to the sample. There is a risk of giving. For this reason, it is desirable to reduce the peak value (maximum output) while maintaining the average output.

  As a method of reducing the peak value while maintaining this average output, in this embodiment, as shown in FIG. 20, the laser beam L0 emitted from the light source 2001 is expanded by a beam expanding optical system 2016, and pulse branching is performed. Splitting the laser beam for one pulse emitted from the light source into a plurality of pulses with reduced peak values by combining the optical paths after entering the optical system 2017 and branching into a plurality of optical paths having different optical path lengths, The plurality of divided pulse lasers are incident on the branching optical element 2018 (corresponding to the optical system composed of the mirrors M1 to M9 and the cylindrical lenses 109, 110, and 111 in FIG. 2) and any one of the optical paths L1, 2, 3 It is guided in the direction (corresponding to the optical paths 106, 107, and 108 in FIG. 2) to form a slit beam and irradiate the slit region 2100 of the wafer W. Configured.

  In addition, by irradiating the pulse laser beam in a plurality of portions, for example, the movement speed in the X direction (see FIG. 11) of the XYZ-θ stage 11 on which the substrate W to be inspected is placed is changed per second. When the detector 220 or 221 of the polarization detection unit 200a ′ shown in FIG. 3A is composed of a time accumulation type linear image sensor such as a CCD method or a CMOS method, the detection visual field per pixel is 1 μm. When a UV pulse laser beam with an emission frequency of 100 MHz is divided into a plurality of parts under the above-described conditions and irradiated, a laser beam exceeding several hundred pulses is irradiated on the region detected by one pixel of the detector 220 or 221. Speckle noise generated by the beam can be averaged over time and imaged, and an image with reduced noise can be obtained.

An example of the pulsed light splitting optical system 2017 is shown in FIG. In this example, the pulsed light splitting optical system 2017 is configured by a combination of a quarter wavelength plate 1711, PBS (polarization beam splitter) 1712a, 1712b, and mirrors 1713a, 1713b. The laser beam expanded by the beam expanding optical system 2016 and incident as linearly polarized light (P-polarized light in this example) is converted into elliptically polarized light by the quarter-wave plate 1711a, and is separated into P-polarized light and S-polarized light by the polarizing beam splitter 1712a. One separated P-polarized component passes through the polarization beam splitter 1712a and the polarization beam splitter 1712b. The other separated S-polarized component is reflected by the polarized beam splitter 1712a, mirror 1713a, mirror 1713b, and polarized beam splitter 1712b, respectively, and returns to the same optical axis as the P-polarized component that has passed through the polarized beam splitters 1712a and 1712b. At this time, if the distance between the polarizing beam splitter 1712a and the mirror 1713a and the distance between the polarizing beam splitter 1712b and the mirror 1713b is L / 2 [m], there is an optical path difference of L [m] between the S-polarized component and the P-polarized component. it can. If the speed of light is c [m / s], t [s] = L [m] / c [m / s] (Equation 1) between the S polarization component and the P polarization component
Time difference is generated, and two beams of pulses emitted from the laser light source 2001 as shown in FIG. 21 (b) and oscillated at the time interval T are time-divided. The pulse laser can be divided into two pulses, one each for P-polarized light and S-polarized light at time intervals t, and the peak value can be reduced to ½.

For example, the pulse interval 10 ns (10-8 sec), using a laser pulse width 10 ps ^{(10 -11} seconds), the polarization beam splitter 1712a and the mirror 1713a, and the polarization beam splitter 1712b respectively 15cm spacing mirror 1713b (0. 15m), the time difference between the S-polarized component and the P-polarized component is 1 ns (10-9 seconds). That is, the wafer surface is irradiated with a laser beam whose peak value is halved twice in a pulsed manner at intervals of 1 ns for 10 ns each time, each pulse of P-polarized light and S-polarized light.

  When the rotation angle of the quarter wave plate 1711a is adjusted so that the ratio of the S-polarization component and the P-polarization component of the incident beam of the polarization beam splitter 1712a is 1: 1 (circular polarization), the optical component to be used (polarization beam splitter) 1712a, 1712b and the loss (reflectance, transmittance) of the mirrors 1713a, 1713b), the peak values of the pulsed light of the S-polarized component and the P-polarized component in the outgoing beam of the polarizing beam splitter 1712b are different. In order to reduce the maximum value of the peak value of each pulsed light, the peak value of each pulsed light needs to be set to substantially the same size.

In the configuration of the pulse splitting optical system 2017 shown in FIG. 21 (a), the P-polarized component is only affected by the P-polarized light transmittance (Tp) of the polarizing beam splitters 1712a and 1712b. The S-polarized reflectance (Rs) of 1712a and 1712b and the S-polarized reflectance (Rm) of the mirrors 1713a and 1713b are affected. The loss ratio (Pl) is P1 = Ls / Lp = Rm ² × Rs ² / Tp ² (Equation ² ), where Ss polarization component loss is Ls and P polarization component loss is Lp.
It becomes.

  Therefore, by adjusting the rotation angle of the quarter-wave plate 1711a so that the ellipticity of the incident beam polarization of the polarization beam splitter 1712a is equal to the loss ratio, the S polarization component in the output beam of the polarization beam splitter 1712b is adjusted. And the peak value of the pulsed light of the P-polarized component can be made substantially equal. The P-polarized component and S-polarized component pulse lights divided so that the peak values are substantially equal have optical paths 106, 107,... Shown in FIG. The wafer W is irradiated through any of 108.

  In the above description, the method of dividing the pulsed light into two using the pulse division optical system 2017 has been described. However, as a modification of the pulse division optical system 2017 for further increasing the number of divisions, a method of dividing into four is illustrated in FIG. And it demonstrates using (b). The configuration of the pulse division optical system 2217 shown in FIG. 22A is a two-stage configuration of the pulse division optical system 2017 shown in FIG. The distance between the second-stage polarization beam splitter 1732c and the mirror 1733c and the distance between the polarization beam splitter 1732d and the mirror 1733d are respectively set to the first-stage polarization beam splitter 1732a and the mirror 1733a, and the polarization beam splitter 1732b and the mirror 1733b. Set to twice the interval.

  The outgoing beam from the first-stage polarization beam splitter 1732b is S-polarized pulse light having a time delay with respect to the P-polarized pulse light. By making this pulse light train circularly polarized by the ¼ wavelength plate 1731b, the half intensity of the pulse light train transmitted through the ¼ wavelength plate 1731b becomes P-polarized light, and the polarization beam splitters 1732c and 1732d are turned on. The light is transmitted, becomes half-polarized light, is reflected by the polarizing beam splitter 1732c and the mirrors 1733c and 1733d, is reflected by the polarizing beam splitter 1732d, and returns to the same optical axis as the P-polarized light. As a result, the pulsed light is divided into four, and each peak value is reduced to ¼ of the pulsed laser beam emitted from the light source 2001. Strictly speaking, since there is a loss of optical parts as described above, it is reduced from 1/4.

  In the configuration of FIG. 22A, the P-polarized pulse laser that passes through the polarizing beam splitter 1732c and passes through the polarizing beam splitter 1732d is the same as the S-polarized pulse laser that is reflected by the mirror 1733d and reflected by the polarizing beam splitter 1732d. The light passes through the optical axis and is circularly polarized by the quarter-wave plate 1731c, enters the polarization beam splitter 1734, and the optical path is separated into P-polarized light and S-polarized light. The laser beam of one P-polarized component from which the optical path is separated enters the optical path L1 and is shaped by the cylindrical lens 1735 (corresponding to the cylindrical lens 109, 110, or 111 in FIG. 2B) to be linear on the wafer 1. Illuminate the area 2110.

  On the other hand, the S-polarized beam which is reflected by the polarization beam splitter 1734 and whose optical path is bent by 90 ° enters the optical path L2, is reflected by the mirrors 1736 and 1737, is changed by the optical path, and is molded by the cylindrical lens 1738 and is formed on the wafer W. The linear region 2110 is illuminated from the direction perpendicular to the surface of the wafer W with respect to the laser beam of the P-changing component that is illuminated from the direction of the optical path L2.

At this time, the optical path L1 and the optical path L2 are designed to have different optical path lengths. The P-polarized laser beam and the S-polarized laser beam applied to the linear region 2110 on the wafer W are shown in FIG. As shown in FIG. 5, a time difference t ₀ is generated by an optical path length difference, and irradiation is performed with a timing shift. Thereby, the occurrence of interference between the P-polarized laser beam and the S-polarized laser beam irradiated on the linear region 2110 can be prevented.

  The photodetectors 220 and 221 that detect reflected and scattered light from illumination from the laser light source 2001 detect reflected and scattered light from illumination from a direction shifted by 90 ° within the time for detecting one pixel. Therefore, it becomes possible to reduce the variation in detection sensitivity due to the difference in illumination direction, and it becomes possible to detect more minute foreign matter defects stably. At this time, when the oblique detection system 500b as shown in FIG. 10 is also used, the detection direction of the oblique detection system 500b detects the reflected scattered light in the direction of the arrow 1740.

  FIG. 23 shows a configuration in which the optical path L2 in the configuration shown in FIG. 22A is changed to the optical path L3. In this configuration, the S-polarized beam reflected by the polarizing beam splitter 1734 and bent by 90 ° in the optical path is reflected by the mirrors 1736, 1747 and 1748 in the optical path L3, and the optical path is changed to be molded by the cylindrical lens 1749. The linear region 2110 on the wafer W is illuminated from the direction facing the optical path L1.

  In the configuration shown in FIG. 22A and FIG. 23, the configuration using the ¼ wavelength plate 1731c and the polarization beam splitter 1734 is shown, but the ¼ wavelength plate 1731c is deleted and replaced with the polarization beam splitter 1734. By using a non-polarizing beam splitter (not shown), the wafer W is illuminated at different timings for the P-polarized light and the S-polarized light from the directions of the optical paths L1, L2, and L3. At this time, each of the photodetectors 220 and 221 detects reflected / scattered light by sequentially illuminating with P-polarized light and S-polarized light from a direction shifted by 90 ° or 180 ° within the time for detecting one pixel. It will be. As a result, reflected / scattered light from the wafer W illuminated under a plurality of illumination conditions is detected within a time period for detecting one pixel, and the detection sensitivity is higher than that when illuminated under a single illumination condition. It becomes possible to improve, and it becomes possible to detect more minute foreign matter defects stably.

  Signals detected by the photodetectors 220 and 221 are processed by the signal processing unit 300 in the same manner as described in the first embodiment, and defects are detected.

  In the second embodiment, the polarization detection unit 200a has been described using the configuration in FIG. 3A. However, the polarization detection unit described in FIG. 3B, FIG. 4 or FIG. 5 may be used. .

  According to the present embodiment, the peak of the UV pulse laser beam can be reduced and the wafer can be irradiated, so that an extremely small defect of about 0.1 μm or smaller is not damaged to the wafer. It became possible to detect with high sensitivity.

It is a schematic block diagram which shows 1st embodiment of the defect inspection apparatus which concerns on this invention. It is a schematic block diagram which shows the structure of the illumination optical system which concerns on this invention. It is a schematic block diagram which shows the structural example using the amplitude division method of the polarization | polarized-light detection part which concerns on this invention. It is a schematic block diagram which shows the structural example using the birefringent wedge of the polarization | polarized-light detection part which concerns on this invention. It is a schematic block diagram which shows the structural example using the polarizing optical element array of the polarization detection part which concerns on this invention. It is a schematic block diagram which shows the signal processing part which concerns on this invention. It is a conceptual diagram which shows the defect determination method based on two mutually different polarization component signals in the signal processing part which concerns on this invention. It is a conceptual diagram which shows the defect determination method based on two physical quantities computed from the mutually different several polarization component signal in the signal processing part which concerns on this invention. It is a conceptual diagram which shows the defect determination method based on the three physical quantities calculated from the mutually different several polarization component signal in the signal processing part which concerns on this invention. It is a schematic block diagram which shows the optical system in the 1st modification of the defect inspection apparatus which concerns on this invention. It is the schematic which shows the detection direction of the oblique detection system in the 1st modification of the defect inspection apparatus which concerns on this invention. It is the schematic which shows the relationship between the detection direction of an oblique detection system, a stage scanning direction, and an illumination area longitudinal direction in the 1st modification of the defect inspection apparatus which concerns on this invention. It is a schematic block diagram of the structural example from which the formation method of the illumination area in the 1st modification of the defect inspection apparatus which concerns on this invention differs from a 1st Example. It is a schematic block diagram which shows the optical system in the 2nd modification of the defect inspection apparatus which concerns on this invention. It is a schematic block diagram which shows the optical system in the 3rd modification of the defect inspection apparatus which concerns on this invention. It is a schematic block diagram which shows the illumination optical system in the 2nd, 3rd, 4th, 5th modification of the defect inspection apparatus which concerns on this invention. It is a schematic block diagram which shows the optical system and stage in the 4th modification of the defect inspection apparatus which concerns on this invention. It is a schematic block diagram which shows the optical system and stage in the 5th modification of the defect inspection apparatus which concerns on this invention. It is a conceptual diagram which shows the relationship between the rotation of the visual field with respect to a test object board | substrate and rotation of a detection polarization component in the 4th, 5th modification of the defect inspection apparatus which concerns on this invention. . It is a side view of the beam expansion optical system in the 2nd example. (A) A block diagram showing a schematic configuration of the optical path branching optical system in the second embodiment, (b) a waveform signal diagram of a pels laser emitted from a laser light source, and (c) a one-pulse laser emitted from the laser light source. It is a pulse waveform signal diagram which shows the state divided | segmented into 2 pulses. (A) It is a block diagram which shows schematic structure in the modification of the pulse division | segmentation optical system in 2nd Example, (b) It is a pulse waveform signal figure which shows the state of a pulse division | segmentation. It is a block diagram which shows schematic structure in another modification of the pulse division | segmentation optical system in a 2nd Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Light source, 3a ... Objective lens, 4a ... Spatial filter, 5a ... Imaging lens, 200a ... Polarization detection part, 300 ... Signal processing part, 6 ... Overall control part, 7 ... Display part, 8 ... Calculation part, 9 ... Storage unit, 10 ... XYZ-θ stage, 100 ... Illumination optical system, 1000 ... Optical system, W ... Substrate to be inspected

Claims (14)

Light source means for emitting a laser;
Illumination optical system means for controlling the polarization state of the laser emitted from the light source means and irradiating the sample having a pattern formed on the surface from a direction inclined with respect to the surface;
Detecting means for separating and detecting reflected and scattered light from the sample irradiated with laser by the illumination optical system means for each polarization component of the reflected and scattered light through a spatial filter;
A signal processing means for detecting a defect on the sample by processing a detection signal for each polarization component separated and detected by the detection means for each polarization component;
A defect inspection apparatus comprising: output means for outputting information on defects obtained by processing by the signal processing means.   The signal processing means uses information obtained by processing the detection signal of the P polarization component among the detection signals detected separately for each polarization component and information obtained by processing the detection signal of the S polarization component. The defect inspection apparatus according to claim 1, wherein defects on the sample are detected and the detected defects are classified.   The signal processing means processes a detection signal for each of the polarization components detected separately and extracts a plurality of defect candidates, compares the feature quantities of the extracted defect candidates, and statistically deviates features. The defect inspection apparatus according to claim 1, wherein a defect candidate having a quantity is detected as a defect.   The illumination optical system means has a cylindrical lens, and irradiates the elliptical or linear region of the sample surface from a direction inclined with respect to the sample surface via the cylindrical lens. The defect inspection apparatus according to any one of claims 1 to 3.   The illumination optical system means has an optical path switching unit, and the switching unit can switch the azimuth angle and / or elevation angle for irradiating the surface of the sample with the laser whose polarization state is controlled. The defect inspection apparatus according to claim 1.   The detection means includes: a first detection optical system unit that detects light reflected and scattered in a first elevation angle direction with respect to the surface of the sample out of the reflected and scattered light from the sample irradiated with the laser; 4. The defect inspection apparatus according to claim 1, further comprising: a second detection optical system unit that detects light reflected and scattered in a second elevation angle direction with respect to the surface.   The light source means emits a pulsed laser, and the illumination optical system means has a plurality of optical paths having different optical path lengths and introduces one pulse of the pulsed laser emitted from the light source means into the plurality of optical paths. 4. The defect inspection apparatus according to claim 1, wherein the surface of the sample is irradiated by being divided into a plurality of pulses. Irradiate the sample whose pattern is formed on the surface with a laser emitted from a light source and whose polarization state is controlled.
The reflected and scattered light from the sample irradiated with the laser is separated and detected for each polarization component of the reflected and scattered light through a spatial filter,
Processing the detection signal for each polarization component detected separately for each polarization component to detect defects on the sample;
A defect inspection method characterized by outputting information on the detected defect.   Among the detection signals detected separately for each polarization component, information obtained by processing the detection signal of the P polarization component and information obtained by processing the detection signal of the S polarization component are used on the sample. 8. The defect inspection method according to claim 7, wherein a defect is detected and the detected defect is classified.   A plurality of defect candidates are extracted by processing a detection signal for each polarization component detected by separation, and feature quantities of the extracted defect candidates are compared, and defect candidates having statistically deviated feature quantities are detected. 8. The defect inspection method according to claim 7, wherein the defect is detected as a defect.   A laser whose polarization state is controlled by being emitted from the light source is irradiated onto an elliptical or linear region of the sample surface from a direction inclined with respect to the sample surface via a cylindrical lens. The defect inspection method according to any one of claims 8 to 10.   11. The defect inspection method according to claim 8, wherein the laser whose polarization state is controlled is irradiated while switching the azimuth angle and / or elevation angle with respect to the surface of the sample.   Of the reflected and scattered light from the sample irradiated with the laser, a signal obtained by detecting light reflected and scattered in the first elevation direction with respect to the surface of the sample and a second with respect to the surface of the sample 11. The defect inspection method according to claim 8, wherein a defect is detected using a signal obtained by detecting light reflected and scattered in an elevation angle direction. The laser is a pulsed laser, and one pulse of the pulsed laser is introduced into a plurality of optical paths having different optical path lengths and divided into a plurality of pulses, and the laser divided into the plurality of pulses is applied to the surface of the sample. 11. The defect inspection method according to claim 8, wherein irradiation is performed.

Images