JP2012202980A - Inspection method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress defective measurement of line width information of a pattern.SOLUTION: The surface inspection method comprises: a step 128 of irradiating a surface of a wafer, including a substrate and a foundation part which coats a surface of the substrate and where a repeated pattern is formed, with light and detecting light intensity information of reflected light from the surface of the wafer irradiated with the light, by a measurement surface of a light receiving optical system; a step 130 of obtaining a detection signal of first light intensity at a first position, on the measurement surface; and a step 140 of obtaining information of a line width of the repeated pattern using the information of the foundation part and the detection signal of the first light intensity.

Description

  The present invention relates to an inspection method and apparatus capable of detecting line width information of a pattern formed on a substrate such as a wafer in a semiconductor device manufacturing process, for example.

In the manufacturing process of micro devices such as semiconductor devices and electronic devices, a substrate is processed using a substrate processing apparatus such as an exposure apparatus.
In the manufacturing process, for example, the line width of the pattern on the substrate after exposure and development is managed. It manages the detection performance of defects generated in the manufacturing process and the CD value (critical dimension) which is the line width of the pattern.

  Patent Document 1 below discloses an example of an inspection apparatus that measures the line width of a pattern formed on a substrate.

US Patent Application Publication No. 2006/0192953

  For example, when the amount of change in polarization due to the structural birefringence effect of the pattern is extracted and the line width of the pattern formed on the substrate is measured, the polarization may change depending on the state of the substrate other than the line width. . If information on the substrate other than the line width is included in the amount of change in polarization, measurement failure may occur. As a result, there is a possibility that a problem occurs in CD management.

  An object of an aspect of the present invention is to provide an inspection method and apparatus that suppress measurement failure of pattern line width information.

  According to the first aspect of the present invention, the surface of a substrate that includes a base material and a coating material that covers the surface of the base material and that has a pattern formed thereon is irradiated with light containing a polarization component. Detecting the light intensity information of the reflected light from the surface of the substrate irradiated with light on the measurement surface of the light receiving optical system, obtaining the first light intensity at the first position on the measurement surface, Using the information on the covering material and the first light intensity, obtaining information on the line width of the pattern is provided.

  According to the second aspect, the irradiation unit that irradiates the surface of the substrate on which the pattern is formed with the light including the polarization component, and the light receiving optical device that receives the reflected light from the surface of the substrate irradiated with the light. And a detection unit for detecting light intensity information of the reflected light on the measurement surface of the light receiving optical system, and a calculation unit for obtaining line width information of the pattern from the light intensity information detected by the detection unit. And the substrate includes a base material and a coating material that covers the surface of the base material and has the pattern formed thereon, and the calculation unit has the first light at the first position on the measurement surface. There is provided an inspection apparatus that obtains the intensity, and obtains information on the line width of the pattern using the information on the covering material and the first light intensity.

  According to the present invention, it is possible to suppress the measurement failure of the line width of the pattern.

It is a figure showing the whole surface inspection device composition of an example of an embodiment. (A) is an enlarged perspective view which shows a repeating pattern, (b) is an enlarged plan view which shows a repeating pattern. It is a figure which shows the pupil image detected with the image pick-up element. It is a figure which shows the state which divided the pupil image into the area | region. (A) And (b) is a figure which shows the relationship between a repeating pattern and the polarization direction of incident light, respectively, (c) is a figure which shows an example of the measurement point in a pupil coordinate system. (A) is a figure which shows an example of a calibration curve, (b) is a figure which shows an example of the correlation with CD conversion value of a reference wafer, and CD-SEM value, (c) is CD conversion value before correction | amendment of a mass production wafer, and It is a figure which shows an example of the correlation with CD-SEM value. It is an expansion perspective view which shows the state which has the polarization direction of the light which injects into a repeating pattern in an entrance plane. (A) is a figure which shows the azimuth | direction angle of a repeating pattern, (b) is a figure which shows an example of the measurement point in the pupil image corresponding to the azimuth | direction angle of Fig.8 (a). (A) And (b) is a figure which shows the area | region which has a correction point in a pupil image, when the azimuth of a repeating pattern is 0 degree | times and 90 degree | times, respectively. (A) is a figure which shows an example of the correlation with the gradation value of a correction point, and the deviation value with respect to CD-SEM value, (b) is the correlation of CD conversion value after correcting the deviation value, and CD-SEM value. It is a figure which shows an example (A) is a figure which shows an example of the dispersion | variation in the deviation value with respect to CD-SEM value, (b) is a figure which shows an example of the dispersion | variation in the deviation value with respect to CD-SEM value after correction | amendment. (A) is a flowchart which shows an example of the method of determining the coefficient which calculates CD conversion value, (b) is a flowchart which shows an example of the method of determining the coefficient which calculates deviation value. (A) is a plan view showing an example of the azimuth angle of the reference wafer when determining the measurement point, and (b) is a plan view showing an example of the azimuth angle of the reference wafer when determining the correction point. (A) is a figure which shows an example of the correlation coefficient of a deviation value and a gradation value, (b) is a figure which shows the coordinate of a pupil image. It is a flowchart which shows an example of the surface inspection method. (A) is a top view which shows an example of the azimuth angle of a wafer when acquiring measurement point data, (b) is a top view which shows an example of the azimuth angle of a wafer when acquiring correction point data.

  Hereinafter, an exemplary embodiment of the present invention will be described with reference to the drawings. FIG. 1 shows a surface inspection apparatus 1 of the present embodiment. The surface inspection apparatus 1 includes a stage 6 on which a semiconductor wafer 5 (hereinafter simply referred to as a wafer 5) as an object to be inspected is placed, an objective lens 7, a prism 8, an illumination optical system 10, and detection optics. A system 15, an arithmetic processing unit 20, a main control system 3 composed of a computer that comprehensively controls the operation of the entire apparatus, and other driving units are provided. For example, after exposure of the wafer 5 by an exposure apparatus (not shown) and development of a photoresist by a coater / developer (not shown), or after a pattern is formed by an etching apparatus or the like (not shown), the wafer 5 is transferred by a transfer system (not shown). The pattern (repetitive pattern) is placed on the stage 6 with the formation surface facing up.

Hereinafter, the Z-axis is taken in parallel to the optical axis of the objective lens 7, the X-axis is taken in a direction perpendicular to the paper surface of FIG. 1 within a plane perpendicular to the Z-axis, and the Y-axis is taken in a direction parallel to the paper surface of FIG. I will explain. A rotation angle around an axis parallel to the Z axis in the XY plane is referred to as an azimuth angle.
The stage 6 is fixed to the upper part of the XY stage 6A that can move in the X direction and the Y direction along the surface of a base member (not shown), and can be rotated around an axis parallel to the Z axis. And a rotating stage 6B. The wafer 5 is held on the surface (XY plane) of the rotary stage 6B by vacuum suction or the like. The main control system 3 controls the position of the XY stage 6A and the rotation angle of the rotary stage 6B via the drive unit 4 based on the measurement value of a position measuring device (not shown) such as a laser interferometer.

  The illumination optical system 10 includes a light source 11, a condensing lens 12, a wavelength selection filter 13, and a first polarizing filter 14 in order of arrangement from the right side to the left side in FIG. As the light source 11, a mercury lamp or the like is used as an example. The mercury lamp has a plurality of wavelengths λ, for example, e-line (λ = 546 nm), g-line (λ = 436 nm), h-line (λ = 405 nm), j-line (λ = 313 nm), and λ = 250 nm. Generate nearby light. The wavelength selection filter 13 is used to select only light having a specific wavelength from among the light having a plurality of wavelengths.

  The light source 11 may be a light source having a wide wavelength band such as a halogen lamp or a blue excitation type LED. When a light source having a wide wavelength band is used, the wavelength selection filter 13 is, for example, blue (B: λ = 436 nm), green (G: λ = 546 nm), or red (R: λ = 700 nm). A filter that transmits light may be used. Therefore, it is preferable that the wavelength selection filter 13 has a configuration in which the wavelength and the wavelength band can be selectively changed by inserting and removing from the optical path as a plurality of filters having different transmission wavelength bands.

The light emitted from the light source 11 passes through the condenser lens 12 and the wavelength selection filter 13 and then passes through the first polarizing filter 14. The first polarizing filter 14 is arranged so that the transmitted light is polarized in the X direction, that is, the polarization direction P of linearly polarized light obtained by transmitting through the first polarizing filter 14 is the X direction.
The light transmitted through the first polarizing filter 14 is reflected downward by the prism 8. Since the light reflected by the prism 8 passes through the condenser lens 12, the incident light I directed to the wafer 5 becomes parallel light. Incident light I is applied to the wafer 5 through the objective lens 7.

  Reflected light from the wafer 5 becomes parallel reflected light J through the objective lens 7, passes through the prism 8, and reaches the detection optical system 15. The detection optical system 15 includes a second polarizing filter 16, a relay lens 17, and an imaging element 18 such as a two-dimensional CCD in the order of arrangement toward the + Z direction. The reflected light J that has passed through the prism 8 passes through the second polarizing filter 16, is extended or enlarged (or reduced) by the relay lens 17, and then enters the image sensor 18. The reflected light incident on the image sensor 18 is photoelectrically converted by a plurality of pixels of the image sensor 18 to become a detection signal (pixel signal), and this detection signal is output to the arithmetic processing unit 20 including a computer. The arithmetic processing unit 20 processes the detection signal to determine the line width of the repetitive pattern formed on the surface of the wafer 5, and outputs the calculated line width information to the main control system 3.

  The second polarizing filter 16 is arranged so that the transmitted light is polarized in the Y direction, that is, the polarization direction Q of the linearly polarized light obtained by transmitting through the second polarizing filter 16 is the Y direction. Thus, the state where the polarization direction of the light transmitted through the first polarizing filter 14 and the polarization direction of the light transmitted through the second polarizing filter 16 are orthogonal to each other is called crossed Nicol. Of the reflected light J transmitted through the objective lens 7 and the prism 8, a polarization component perpendicular to the polarization direction of the incident light I (linearly polarized light) is detected by the image sensor 18 (detection optical system 15). The first polarizing filter 14 on the incident side may be referred to as a polarizer (polarizer), and the second polarizing filter 16 on the light receiving side may be referred to as an analyzer (analyzer).

In the above-described embodiment, the first polarizing filter 14 and the second polarizing filter 16 are used to pass the light including the polarization component. However, the method of passing the light including the polarization component is not limited to this. .
For example, a linear polarizer may be used to irradiate the wafer 5 with linearly polarized light. Examples of the linear polarizer include those using anisotropic crystals and those using reflection. As such a linear polarizer, for example, one described in US Patent Publication No. 2004-0201889 can be used. Further, only one of the first polarizing filter 14 and the second polarizing filter 16 may be a linear polarizer.

Further, the arrangement of the first polarizing filter 14 is not limited to the above-described embodiment. For example, it may be provided closer to the light source 11 than the condenser lens 12. Of course, the light source 11 may be provided with a polarizing filter 14 and the light from the light source may be converted to light containing a polarization component.
In addition, you may provide the drive parts 14d and 16d which rotate the polarizing filters 14 and 16 around an optical axis by the same angle, maintaining a cross Nicol state. In this case, the relationship between the polarization direction of the incident light and the azimuth angle of the wafer 5 can be controlled by rotating the polarization filters 14 and 16 via the drive units 14d and 16d.

In the present embodiment, the first and second polarizing filters 14 and 16 are arranged so that the transmitted light is polarized in the X direction and the Y direction, respectively. However, as long as the crossed Nicols state is maintained. The direction is not limited to this. For example, the polarization direction of the light transmitted through each of the first and second polarizing filters 14 and 16 may be the Y direction and the X direction.
In addition, you may measure on conditions different from a crossed Nicol state. For example, the polarization direction of the light transmitted through the second polarization filter 16 may be different from the orthogonal direction with respect to the polarization direction of the light transmitted through the polarization filter 14. In addition, the image sensor 18 may detect light that passes through only one polarization filter.

On the + Z direction side of the relay lens 17 of the detection optical system 15, a measurement surface DP for measuring the light intensity distribution of the reflected light that has passed through the objective lens 7 is set. The light receiving surface of the image sensor 18 is disposed on the measurement surface DP.
The pupil plane of the objective lens 7 is the exit pupil of the objective lens 7. In the present embodiment, the measurement surface DP is provided at the pupil plane of the objective lens 7 or at a position conjugate with this plane. The measurement plane DP may be near the pupil plane of the objective lens 7 or a position conjugate with this plane. The vicinity of the pupil plane or the conjugate position includes a position along the optical axis until the measurement plane DP comes into contact with the relay lens 17.

Further, in the vicinity of the pupil plane or the conjugate position, the distance along the optical axis until the measurement surface DP and the relay lens 17 come into contact with each other is the same distance along the optical axis so as to be separated from the relay lens 17. Includes location. In the present embodiment, the pupil plane of the objective lens 7 is the rear focal position of the objective lens 7. In the present embodiment, the pupil plane of the objective lens 7 is a Fourier transform plane for the pattern of the wafer 5. Therefore, the angle component of the reflected light from the wafer 5 is converted into position information on the measurement surface DP.
Hereinafter, the light intensity distribution formed on the measurement surface DP by the reflected light from the wafer 5 is also referred to as a pupil image. Note that it is possible to eliminate the wavelength selection filter 13 by using a color CCD that outputs detection signals of red, green, and blue light for the image sensor 18.

In the present embodiment, as shown in FIG. 1, a part of the incident light beam I1 of the incident light I is incident on the wafer 5 at an incident angle K and reaches the image sensor 18 as a reflected light beam J1. The position 19 on the image sensor 18 is determined by the azimuth angle and the incident angle K of the incident surface of the incident light beam I1.
Here, the structural birefringence effect in the repeated pattern will be described with reference to FIG. Details of structural birefringence are described, for example, in Max Born and Emil Wolf: “Principles of Optics: Sixth Edition” (Principles of Optics: Sixth Edition), (Pergamon Press, 1980).

As shown in FIG. 2A, a wafer 5 in FIG. 1 includes a base material 34 made of a flat plate member such as silicon, and a single layer such as an oxide film provided so as to cover the surface of the base material 34. Alternatively, a plurality of base portions 32 and a repetitive pattern 30 which is a linear pattern (line and space pattern; hereinafter referred to as an L & S pattern) formed on the surface of the base portion 32 are provided. In the repeating pattern 30, the line width of the line portion 31L t _1, the refractive index of the line portion 31L and n _1, the line width of the space portion 31S t _2, the refractive index of the base portion 32 of the space portion 31S and n ₂ To do. Then, the pattern pitch is t ₁ + t ₂ . The repeating pattern 30 can be regarded as a uniaxial crystal having a first material (line portion 31L) and a second material (base portion 32 in the space portion 31S) that are different from each other. As shown in FIG. 2 (b), the refractive index n _o for light oscillating in parallel to the boundary surface between the line portion and the space portion, the refractive index for light oscillating perpendicularly to the boundary surface as n _e, repeating pattern Birefringence at 30 can be handled.

When the light 33 is incident on the repetitive pattern 30 in FIG. 2A, the polarization direction of the light 33 (vibration direction of the electric field) is the direction R (X direction) in FIG. is there. Further, the azimuth angle of the repeated pattern 30 is an angle formed by the periodic direction (repeated direction) of the repeated pattern 30 with respect to the X axis. When repeating pattern 30 is inclined with respect to the X axis in azimuth theta, direction R can be decomposed into components R _e perpendicular periodic direction component R _o and the boundary surface parallel to the boundary surface. Then, the refractive index n _o, n _e, respectively, (the angle between the incident surface forms a periodic direction and the light 33 of the repeating pattern 30) incident direction of light 33, the angle of incidence of the light 33, refractive indexes n _1, n ₂ And a function of the line widths t ₁ and t ₂ .

When the phase difference between the component parallel to the boundary surface of the reflected light due to structural birefringence and the component perpendicular to the boundary surface is δ and the height of the line portion 31L is h, the phase difference δ is high. is h and the refractive index n _o, using a n _e is given by the following equation.
_{δ = h (n o -n e} ) ... (1)
After the exposure and development, the line portion 31L of the repeated pattern 30 is a photoresist, and the base portion 32 of the space portion 31S is, for example, a metal film or an oxide film. Although the thickness of the photoresist depends on the exposure wavelength, the thickness of the resist for ArF laser light source used in the most advanced exposure apparatus is about 100 nm or less. As another example of the repetitive pattern 30, for example, after etching, the line portion 31L is, for example, a metal film.

In the present embodiment, the position 19 (position on the pupil image) of the reflected light beam J1 on the light receiving surface (measurement surface DP) of the reflected light beam J1 is determined by the azimuth angle and the incident angle K of the incident light beam I1 in FIG. Is done. When the light 33 is incident on the repetitive pattern 30, the polarization component (direction R) of the light 33 incident on the repetitive pattern 30 is divided into two polarization components Re and R0 shown in FIG. One polarization component Re is a component perpendicular to the repetition direction. The other polarization component R0 is a component perpendicular to the repetition direction. The two polarization components R0 and Re are independently subjected to different amplitude changes and phase changes. The reason why the amplitude change and the phase change are different is that the complex reflectivity (that is, the complex amplitude reflectivity) is different due to the anisotropy of the repetitive pattern 30 and is called structural birefringence. As a result, the reflected lights of the two polarization components Re and Ro have different amplitudes and phases, and the reflected light obtained by combining these becomes elliptically polarized light. That is, the azimuth angle of the wafer 5 (repeat pattern 30), the direction of the incident surface of the incident light, and the refractive index n _o, is n _e changed according to the angle of incidence K (polarization angle with respect to the incident plane) polarization direction of the incident light Thus, the phase difference δ in the equation (1) changes. Therefore, the polarization state of the reflected light beam J1 changes, and the intensity of the light that passes through the second polarizing filter 16 changes. Therefore, depending on the position on the measurement surface DP (pupil image), a portion where the correlation between the line width of the line portion 31L of the repetitive pattern 30 and the detection signal is high and a portion where the correlation is low are generated. The line width can be calculated.

  FIG. 3 shows a pupil image 36 formed on the light receiving surface (measurement surface DP) of the image sensor 18. The Px axis on the measurement surface DP corresponds to the X axis parallel to the polarization direction of the light transmitted through the polarizing filter 14 (polarizer) in FIG. 1, and the Py axis is the light transmitted through the polarizing filter 16 (analyzer). Corresponds to the Y axis parallel to the polarization direction. The Px axis and the Py axis are orthogonal to each other, and the origin (intersection) O between the Px axis and the Py axis is the optical axis of the objective lens 7. Hereinafter, the coordinates (Px, Py) are also referred to as pupil coordinates, and the orthogonal coordinate system represented by the pupil coordinates is also referred to as pupil coordinate system (Px, Py). In each portion 36c of the pupil image 36, a portion having a high light intensity gradation value (a portion having a large detection signal obtained from that portion) is represented by a dark color (black) in accordance with the gradation legend 38. A portion having a low gradation value (a portion having a small detection signal obtained from the portion) is represented by a light color (white color).

  The arithmetic processing unit 20 of the present embodiment collectively processes signals for each pixel of the image sensor 18 in predetermined pixel units as shown in FIG. In FIG. 4, a region (pupil image) in the pupil coordinate system on the measurement plane DP is divided into M × N sections of M columns in the Px direction and N rows in the Py direction. The sum of the pixel signals is used as a detection signal obtained by photoelectrically converting light incident on the section 40. The detection signal is represented by a gradation value. M and N are natural numbers, for example, about 10 to 100. Since the pupil image is circular, the natural numbers M and N are determined so that the section 40 is square as an example. It is possible to select a section at an arbitrary position from the pupil coordinate system (pupil image). When the light source 11 is a white light source and the image sensor 18 is a color CCD, it is possible to selectively extract detection signals of light of three colors of red, green, and blue (RGB) in each section. It is.

  Hereinafter, in this embodiment, the principle for detecting the line width of the repetitive pattern using the detection signal corresponding to the light intensity of each section 40 from the image sensor 18 will be described. This detection is performed by the arithmetic processing unit 20. In this embodiment, as shown in FIG. 5A, the polarization direction of incident light is a direction R parallel to the X axis. In this case, as disclosed in, for example, US Patent Application Publication No. 2006/0192953, when the azimuth angle of the repeated pattern 30 is 45 degrees, it is obtained from a predetermined section in the pupil coordinate system. When the sensitivity or correlation of the detection signal with respect to the line width of the repetitive pattern 30 is maximized and the azimuth angles of the repetitive pattern 30 are 0 degrees and 90 degrees, the line width of the detection signal obtained from a predetermined section in the pupil coordinate system There is almost no sensitivity or correlation to. Therefore, in this embodiment, when the azimuth angle of the repetitive pattern 30 is an arbitrary angle within the range of 0 degrees to 90 degrees, a detection signal of a section having a high correlation with the line width is acquired, and the azimuth of the repetitive pattern 30 When the angle is 0 degree or 90 degrees, a detection signal of a section having little correlation with the line width is acquired. Of the detection signals obtained from the sections having little correlation with the line width, the detection signals obtained from the sections having a high correlation with the state of the ground portion 32 of the repetitive pattern 30 are caused by the state of the ground portion 32. Correct measurement error.

At this time, in the pupil coordinate system (Px, Py), two sections having a correlation with the line width of the repetitive pattern 30 are set as the first and second sections. The second section is a position that is different from the first section, or is a portion that receives light of a different wavelength even at the same position as the first section.
In this case, a surface inspection is performed on a wafer serving as a line width reference (hereinafter referred to as a reference wafer) on which a repetitive pattern of various line widths whose line widths are measured in advance by, for example, a CD-SEM (scanning electron microscope). Place on stage 6 of device 1. Then, by picking up the pupil image by the reflected light from the reference wafer with the image pickup device 18, the line width of the repetitive pattern from the detection signal (gradation value) corresponding to the received light amount in the first and second sections. Is calculated (hereinafter referred to as a calibration curve). A mass-produced wafer whose line width is measured may be used instead of the reference wafer. In addition, when there is a detection signal corresponding to the amount of light received in the first and second sections by imaging a pupil image of reflected light from the reference wafer with the image sensor 18 in advance, the detection signal May be used to determine the calibration curve and calculate the line width of the repetitive pattern.
The method for measuring the line width is not limited to the CD-SEM, and a method such as AFM may be used. Of course, an SEM that is not specialized for the length measurement function, such as a CD-SEM, may be used.

  When determining the calibration curve, the deviation (difference) between the value actually measured by the CD-SEM (CD-SEM value) and the line width calculated from the detection signals of the two sections is minimized. The combination of the positions of the two sections in the pupil coordinate system and the coefficient of the calibration curve are determined. When the detection signals from the two sections are S1 and S2, and k1, k2, and k3 are coefficients, the detection signal is converted into a CD (critical dimension) that is the line width of the repeated pattern (hereinafter referred to as a CD conversion value). ) Is expressed as follows.

CD conversion value = k1 × S1 + k2 × S2 + k3 (2)
In this description, the combination of two sections in the pupil image has been described. However, at least one section using the detection signal may be in the pupil coordinate system. For example, when combining detection signals from three sections in the pupil coordinate system, the third section is different from the first and second sections, or is the same as the first or second section and has a different wavelength. This is the part that receives light. Furthermore, a combination of n sections (n is an integer of 3 or more) in the pupil coordinate system may be used. When the detection signals of the n sections are Si (i = 1 to n) and k1, k2,..., K (n + 1) are coefficients, the CD conversion value is referred to as the next calibration curve (hereinafter referred to as Expression α). ).

CD conversion value = k1 × S1 + k2 × S2 +... + Kn × Sn + k (n + 1)
= Α (3)
The detection signals S1 to Sn (gradation values) corresponding to the positions in the pupil coordinate system used in Expression α include components (background components) correlated with the background portion of the repetitive pattern. Therefore, in this embodiment, in order to extract the background component, the calibration curve obtained earlier using the detection signal of a predetermined section in the pupil coordinate system with the azimuth angle of the repeated pattern set to 0 degree or 90 degrees. Is corrected. That is, the base component is removed from the formula α. The calibration curve at this stage is represented by equation β. If the divergence value between the equation α and the CD-SEM value is Δ, the relationship between the equation α and the equation β can be expressed by the following equation.

β = α−Δ (4)
The divergence value Δ is obtained by receiving light of a predetermined wavelength in a section at a predetermined position in the pupil coordinate system with respect to a repetitive pattern set to a predetermined azimuth angle (0 degree or 90 degrees) in order to extract a background component. Using the detection signal X (tone value) and the coefficients e and f obtained in this way, the following calculation formula can be used.

Δ = e × X + f (5)
The following calculation formula (calibration curve) is obtained from Formula (4) and Formula (5).
β = α− (e × X + f) (6)
Therefore, by using Expression (6), it is possible to obtain the line width of the repeated pattern with high accuracy by removing the influence of the base component. In addition, although Formula (6) was represented with the linear linear equation, a nonlinear high-order polynomial may be sufficient. Further, when obtaining the divergence value Δ in the equation (5), a high-order polynomial may be used, and further using a linear equation or a high-order polynomial of detection signals of the pupil images of a plurality of sections in the pupil coordinate system. Also good.

Next, a method for obtaining a calculation formula corresponding to formula (6) will be specifically described. In the following description, for the sake of simplicity, it is assumed that there is one section in the pupil coordinate system for calculating the CD conversion value (formula α).
In FIG. 4, the pupil coordinate system has two diagonal lines Py = Px and Py = −Px, but has the highest correlation with the line width change without being influenced by other factors other than the line width. The position exists on the diagonal of either of these two. In this case, the incident angle of each light beam incident on the repetitive pattern 30, the azimuth angle of the incident surface of each light beam, and the polarization direction (polarization angle) with respect to the incident surface of each light beam are symmetric with respect to the origin O (optical axis). Optically obvious.

  Furthermore, the polarization direction of the incident light is the X direction, and it is also clear that the polarization characteristics of each light beam are symmetric with respect to the diagonal lines Py = Px and Py = −Px. Therefore, the polarization characteristics are symmetric with respect to the optical axis O, and also with respect to the two diagonal lines. Which of the diagonal lines Py = Px and Py = −Px is used is determined by the relationship between the polarization direction of the incident light and the periodic direction of the repeated pattern 30.

  As shown in FIG. 5A, when the azimuth angle of the repetitive pattern 30 is 45 degrees clockwise, there is a correlation with the line width change on the diagonal line indicated by Py = −Px in the pupil coordinate system. The highest position exists. As shown in FIG. 5B, when the azimuth angle of the repetitive pattern 30 is 45 degrees counterclockwise, there is a correlation with the line width change on the diagonal line indicated by Py = Px in the pupil coordinate system. The highest position exists. However, it is known that when approaching the origin O (optical axis) of the pupil coordinate system, the state other than the line width is affected, and the correlation with the line width tends to decrease.

  First, a method for creating a calibration curve for calculating the CD converted value of the equation α will be described. In this creation, a reference wafer whose line width is measured in advance by, for example, a CD-SEM as described above is used. Since the CD-SEM has a narrow field of view at one measurement point, it is preferable to measure a plurality of points in the wide field of view of the surface inspection apparatus 1 of the present embodiment to obtain an average value. Thus, the correlation between the line width value (CD-SEM value) actually measured by the CD-SEM and the detection signal (gradation value) of the section at a predetermined position in the pupil coordinate system formed by the surface inspection apparatus 1 is obtained. Obtained for a large number of repetitive patterns having different line widths.

  As an example, the azimuth angle of the repetitive pattern is set as shown in FIG. 5A, and the CD-SEM value and, for example, the detection signal S (gradation) of the section 42 at the position in the pupil coordinate system shown in FIG. The coefficients k1 and k2 of the calibration curve are determined so that the deviation from the line width (CD converted value) calculated from the (value) is minimized. When the relationship between the CD-SEM value and the detection signal S is as shown in FIG. 6A, the calibration curve is the following linear expression representing the straight line 39. The following formula corresponds to formula (3) (formula α).

CD conversion value = k1 × S + k2 = α (7)
FIG. 6B shows the line width (CD conversion value) (vertical axis) calculated using the equation (7) from the detection signal S of the section 42 in FIG. It represents the relationship with the CD-SEM value (horizontal axis) which is a value. In addition, when using the detection signals of a plurality of sections in the pupil coordinate system, equation (3) may be used instead of equation (7).

Note that the crossed Nicols transmitted light amount due to structural birefringence does not always change in gradation with a linear expression with respect to the line width of the repetitive pattern. In particular, when the line width change range is wide, there is a tendency to deviate from a straight line. In that case, it is preferable to convert by a higher order expression such as a second order expression instead of the expression (7).
In FIG. 6B, a straight line 44 indicates that the CD converted value is the same value as the CD-SEM value, and outer dotted straight lines 45 and 46 indicate the allowable range of the CD converted value with respect to the CD-SEM value. Show. Therefore, FIG. 6B shows that the CD conversion value is within the standard value.

  On the other hand, the CD converted value (vertical axis) in FIG. 6C is obtained from the detection signal S of the section 42 in the pupil coordinate system for a mass-produced wafer (lots 1 to 6) different from the reference wafer used for creating the calibration curve. It is the line width calculated using the equation (7). The CD conversion value in FIG. 6C is out of the range (standard) of the straight lines 45 and 46 in some lots. This is because a deviation error from the CD-SEM value (an error after subtracting the CD-SEM value from the CD converted value) is large. This is because the degree of influence of the change differs between the CD-SEM value and the CD conversion value while the wafer surface state other than the line width changes in the wafer plane.

  That is, the CD-SEM irradiates and scans an electron beam, collects secondary electrons generated from the pattern edge, extracts a predetermined threshold value portion from the intensity signal of the secondary electrons, and sets the pattern line width. In the CD-SEM, the arrangement of the secondary electron collector and the setting of the threshold value are optimized and set to a condition having a good correlation with the width of the pattern cross section, but is affected by the subtle shape of the pattern. Basically, the electron beam does not reach the base (lower layer) of the repeated pattern of the wafer.

  On the other hand, in the case of the method of replacing the polarization state change amount with the pattern line width as in this embodiment, the phase change due to structural birefringence is not only the pattern line width but also the shape and height of the line portion of the repetitive pattern, etc. For example, visible light as incident light may reach the lower layer. Further, the phase change of the reflected light and the influence of the lower layer on the reflected light also vary depending on the wavelength, and also vary depending on the incident angle, the azimuth angle between the incident surface and the repetitive pattern, and the polarization direction. Thus, since the detection method using structural birefringence is a method based on a detection principle different from that of CD-SEM, it is inevitable that the detection method is affected by states other than the line width.

  However, in a production line that has managed the manufacturing process of a semiconductor device with a CD-SEM, a line width measurement performance in accordance with the CD-SEM management is required. Therefore, a deviation error from the CD-SEM value needs to be reduced. . Therefore, in the present embodiment, the conditions such as the incident angle of the incident light, the azimuth angle between the incident surface of the incident light and the repetitive pattern, the polarization direction of the incident light, and the wavelength of the detected light described above are combined. That is, among these conditions, the line width measurement accuracy is improved by appropriately combining the line width of the repetitive pattern and / or conditions affected by elements other than the line width.

First, the conditions for not being affected by the line width of the repeated pattern will be described with reference to FIG. In FIG. 7, the relationship between the incident light 33 and the repetitive pattern 30 is the following orientation relationship A.
(Directional relationship A): The incident light 33 is p-polarized light whose polarization direction 49 coincides with the incident surface 48. Further, the incident surface 48 is parallel to the line portion of the repeated pattern 30.

  Therefore, the electric vector in the polarization direction 49 of the incident light 33 has only a component in the direction parallel to the line portion of the repetitive pattern 30, and the component in the periodic direction of the repetitive pattern 30 is zero. Structural birefringence is caused by the difference in refractive index between the periodic direction of the repeated pattern 30 and the direction perpendicular thereto as a phase difference, but the polarization direction of incident light is only a component in a direction parallel to the line portion of the repeated pattern. If not, there is no phase difference between the two directions in the reflected light. That is, in the arrangement of the polarization direction 49 and the direction of the repetitive pattern 30 as shown in FIG. 7, even if the line width of the repetitive pattern 30 changes, the phase difference between the two directions does not change in the reflected light. Only a change in reflectance occurs. Accordingly, the reflected light detects a lot of information such as a state other than the line width rather than a change in the line width, that is, the film thickness variation of the base portion 32. Therefore, the reflected light (the transmitted light amount of the second polarizing filter 16) is detected. ) Detection signal can be used as background information.

Also in the case of the following azimuth relationships B, C, and D, the reflected light detects a lot of information other than the line width rather than the line width, that is, information such as the film thickness variation of the underlying portion 32. The detection signal of the reflected light can be used as background information.
(Directional relationship B): The polarization direction of incident light is perpendicular to the incident surface (s-polarized state). The incident surface is parallel to the line part of the repeated pattern.

(Orientation relationship C): The polarization direction of incident light coincides with the incident surface (p-polarized state). The incident surface is perpendicular to the line portion of the repetitive pattern (parallel to the periodic direction).
(Directional relationship D): The polarization direction of incident light is perpendicular to the incident surface (s-polarized state). The incident surface is perpendicular to the line part of the repeating pattern.
Therefore, the influence of the background can be corrected by the following equation.

Correction CD conversion value = measurement point data (CD information and background information) −correction point data (background information) (8)
In the equation (8), the measurement point data is, for example, the section 42 in the pupil coordinate system of FIG. 5C as the measurement point, and the detection signal obtained from the pupil image at this measurement point is substituted into the equation (7). It is a CD conversion value obtained as described above. The correction point data refers to the pupil image of the correction point with a predetermined section in the pupil coordinate system as the correction point when the relationship between the incident light and the repetitive pattern is in the above azimuth relationships A to D. This is a value obtained by substituting the detection signal into a conversion equation corresponding to equation (5).

When the correction point is used, the relationship between the incident light and the repetitive pattern is that the incident surface is parallel or perpendicular to the line portion of the repetitive pattern, and the polarization state of the incident light is p-polarized light or s-polarized light. Equation (8) is equivalent to Equation (6).
Specifically, when Expression (8) is applied, as shown in FIG. 8A, the azimuth angle θ of the repetitive pattern 30 has a correlation between the polarization state change amount between 0 degrees and 90 degrees and the line width. An arbitrary azimuth angle (for example, 45 degrees) is set, and an equation (7) is obtained from a detection signal obtained from a measurement point 50 as a predetermined section in the pupil coordinate system shown in FIG. 8B and a corresponding CD-SEM value. ) K1 and k2 are determined.

  Next, the repeated pattern 30 is set to an azimuth angle of 0 degrees in FIG. 9A or 90 degrees in FIG. 9B, that is, an azimuth angle intended to extract only background information. 9A, the incident light incident surface 48 is set parallel or perpendicular to the line portion of the pattern 30, and the incident light polarization direction 49 is set to s-polarized light (state 0s) or p-polarized light (state 0s). Set to state 0p). In state 0s, a predetermined section in the partition area 54 along the Py axis of the pupil coordinate system in which the pupil image 52 is formed is used as a correction point. In state 0p, the section along the Px axis of the pupil coordinate system is used. Using a predetermined section in the region 56 as a correction point, background information is obtained from the detection signal of the correction point.

  9B, the incident light incident surface 48 is set to be perpendicular or parallel to the line portion of the pattern 30 repeatedly, and the incident light polarization direction 49 is set to s-polarized light (state 90s) or p-polarized light (state 90s). State 90p). In the state 90s, a predetermined section in the partitioned area 54 of the pupil coordinate system is used as a correction point. In the state 90p, a predetermined section in the partitioned area 56 of the pupil coordinate system is used as a correction point, and a detection signal of the correction point is detected. Get ground information from Here, for the sake of easy understanding, the specimen used when determining the measurement point 50 (section having CD information and background information) in FIG. 8B is referred to as the reference wafer A, and those in FIGS. 9A and 9B. A sample used when determining a correction point (section having background information) will be described as a reference wafer B.

  FIG. 10A shows the result of obtaining the detection signal X (tone value) (background information) of the correction point in the pupil coordinate system from several reference wafers B. In FIG. 10A, the horizontal axis represents the detection signal X, and the vertical axis represents the CD-converted value and the CD-SEM value calculated by substituting the detection signal (gradation value) at the above measurement point into Equation (7). The deviation value Δ. The correction point is a partition at a position where the correlation between the detection signal X and the deviation value Δ is the highest among the plurality of partitions in the partition regions 54 and 56 in FIG. 9A or 9B. As an example, assume that one section in the section area 54 of FIG. A linear expression (Expression (5)) that minimizes the error between the detection signal X and the deviation value Δ in FIG. From the slope and offset of the straight line 58, the coefficients e and f of the equation (5) are obtained. Thereafter, the deviation value Δ is calculated by substituting the detection signal X (tone value) of the correction point 59 into the equation (5).

  Therefore, the arithmetic processing unit 20 uses the detection signal X of the correction point 59 from the CD conversion value (CD information and background information) obtained from the expression (7) using the detection signal S of the measurement point 50. By subtracting the divergence value Δ (background information) obtained from (), the corrected CD conversion value corresponding to the equations (6) and (8) can be obtained as follows. An error caused by the background component is removed from the corrected CD converted value. Expression (9) (Expression (6) or Expression (8)) is also referred to as a calibration curve for obtaining a corrected CD conversion value.

Correction CD conversion value = CD conversion value−deviation value Δ
= (K1 * S + k2)-(e * X + f) (9)
FIG. 10B shows a corrected CD converted value and a CD-SEM value obtained by substituting the detection signal S at the measurement point and the detection signal X at the correction point into the equation (9) for the mass production wafer (lots 1 to 6). Shows the relationship. Compared with FIG. 6C, which is a result obtained only by the detection signal (CD information and background information) at the measurement point, FIG. 10B is clearly improved, and the corrected CD conversion value is a straight line 45, 46. It can be seen that it falls within the standards.

  FIG. 11A shows a deviation value Δ between the CD converted value and the CD-SEM value obtained from only the measurement point detection signals (CD information and ground information) for the mass-produced wafers of the lots 1 to 6. . FIG. 11B shows the deviation value between the corrected CD conversion value and the CD-SEM value obtained from the measurement point detection signal and the correction point detection signal for these mass-produced wafers. In FIG. 11A, the average value indicated by the dotted line of the deviation value Δ of each measurement point for all lots is greater than 0, and it can be seen that the variation of the deviation value Δ within each lot is also large. On the other hand, in FIG. 11B, the average value of the divergence values of all the lots overlaps with the position of almost 0, and there are many cases where the variation of the divergence values in each lot is improved. Therefore, it can be seen from Equation (9) that the background component is corrected and the line width of the repeated pattern can be measured more accurately.

  Next, in the present embodiment, a flowchart of FIG. 12A is shown for an example of a method for determining the measurement points in the pupil coordinate system and the coefficients k1 and k2 of Expression (9) (Expression (6) or Expression (8)). The description will be given with reference. An example of a method of determining the correction point in the pupil coordinate system and the coefficients e and f of the equation (9) and determining the equation (9) (calibration curve) will be described with reference to the flowchart of FIG. To do. In addition, although the case where there is one measurement point and correction point will be described, a plurality of measurement points and correction points may be provided.

  First, in step 102 in FIG. 12A, one or a plurality of reference wafers A which are samples for determining measurement points (sections having CD information and background information) in the pupil coordinate system are determined. Specifically, the surface inspection apparatus 1 forms a repeated pattern having various line widths including the range of the repeated pattern line width to be measured, and the repeated pattern line width (CD-SEM value) is CD−. The reference wafer 5A in FIG. 13A measured by the SEM is the reference wafer A. In FIG. 13A, L & S patterns having different line widths in a plurality of shot areas in a number of shot areas SAi (i = 1 to N: N is an integer of several tens to several hundreds) of the reference wafer 5A. A repetitive pattern LSi is formed. The reference wafer 5A is placed on the stage 6 in FIG.

  Next, in step 104, the rotary stage 6B is driven, the azimuth angle θ of the reference wafer 5A in FIG. 13A is set to an arbitrary value within 0 to 90 degrees, and the light from the light source 11 is directed in the X direction. The reference wafer 5A is irradiated in a state of being linearly polarized in (direction R), and a pupil image of reflected light from the reference wafer 5A is picked up by the image pickup device 18. Then, for each of a plurality of sections (section 40 in FIG. 4) in the pupil coordinate system on the measurement surface DP, a calculation formula (CD conversion) using a detection signal S (tone value) obtained from a repeated pattern having a different line width. The coefficients k1 and k2 are determined by the least square method so that the square sum (deviation error) of the error between the CD converted value obtained from the value = k1 × S + k2) and the corresponding CD-SEM value is minimized. Further, among the plurality of sections, a section having the smallest deviation error is set as a measurement point MP1 (for example, the measurement point 50 in FIG. 8A).

  Since the position where the deviation error is minimized also varies depending on the wavelength, the CD converted value and the deviation error are calculated using the detection signals of the light having a plurality of wavelengths, and the wavelength of the light where the deviation error is minimized is measured. The wavelength of light used when obtaining the detection signal at the point MP1 is used. Furthermore, if necessary, the azimuth angle θ when the deviation error is minimized may be determined by changing the azimuth angle θ of the reference wafer 5A within a range of 0 degrees to 90 degrees.

  Next, at step 106, the coordinates of the measurement point MP1 having the smallest deviation value in the pupil coordinate system are determined, and the wavelength of the light used at the measurement point MP1 and the coefficients k1 and k2 are calculated by the equation (9). It is determined as the wavelength and coefficient of light used at the time. The azimuth angle of the reference wafer 5A, the coordinates of the measurement point MP1, the wavelength, and the coefficients k1 and k2 are stored in a storage unit in the arithmetic processing unit 20. Further, for each repetitive pattern whose line width information is measured on the reference wafer 5A, a deviation value Δ between the CD-SEM value and the CD conversion value calculated using the coefficients k1 and k2 is calculated, and the calculation result is stored in the storage unit. To remember.

  Next, in step 112 of FIG. 12B, one or a plurality of reference wafers B which are samples for determining correction points (sections having background information) in the pupil coordinate system are determined. In the reference wafer 5B, for example, a repeated pattern having the same various line width as that of the reference wafer 5A is formed on the same base as the base of the mass production wafer to be measured by the surface inspection apparatus 1. The reference wafer 5B may be the same as the reference wafer 5A. Assume that the reference wafer 5B in FIG. The reference wafer 5B is also placed on the stage 6 in FIG.

  Next, in step 114, the rotary stage 6B is driven to set the azimuth angle θ of the reference wafer 5B in FIG. 13B to 0 degree or 90 degrees, and the light from the light source 11 is directed in the X direction (direction R). The reference wafer 5B is irradiated in a linearly polarized state, and a pupil image of the reflected light from the reference wafer 5B is picked up by the image sensor 18. Then, for each of a plurality of sections in the partitioned area 54 (s-polarized area) and the partitioned area 56 (p-polarized area) in the pupil coordinate system on the measurement plane DP in FIG. The sum of squares of errors between the deviation value calculated from the calculation formula (e × X + f) using the detection signal X (tone value) obtained from the reflected light and the corresponding deviation value Δ stored in step 106. The coefficients e and f are determined by the least square method so that is minimized. Further, among the plurality of sections, a section where the sum of squares of the error is minimum is set as a correction point AP1 (for example, a correction point 59 in FIG. 9A). The azimuth angle θ of the reference wafer 5B set in step 114 may be approximately 0 degrees or approximately 90 degrees. Substantially 0 degrees or 90 degrees means that the coefficients e and f may deviate from 0 degrees or 90 degrees as long as the coefficients e and f can be determined.

  Since the position where the sum of squares of the error is minimized also varies depending on the wavelength, a divergence error is calculated using detection signals of light of a plurality of wavelengths, and the light of the wavelength where the sum of squares of the error is minimized. It is assumed that the light for obtaining the detection signal of the correction point AP1. Further, the azimuth angle of the reference wafer 5B is set to 0 degree or 90 degrees, and the azimuth angle when the sum of squares of the error becomes smaller is set as the azimuth angle of the wafer when acquiring the base information.

Then, the coordinates of the correction point AP1 where the sum of squares of the error is the smallest in the pupil coordinate system are determined, and the wavelength of the light used at the correction point AP1 and the coefficients e and f are calculated by Equation (9). It is determined as the wavelength and coefficient of light used at the time. The azimuth angle of the reference wafer 5B, the coordinates of the correction point AP1, the wavelength, and the coefficients e and f are stored in the storage unit in the arithmetic processing unit 20.
In the next step 116, the CD converted value calculated using the coefficients e and f obtained in step 114 and the CD are calculated from the equation α for calculating the CD converted value using the coefficients k1 and k2 obtained in step 106. The arithmetic processing unit 20 uses the equation β (the equation (6) or the equation (9)) in the form of subtracting the deviation value Δ from the SEM value as a calibration curve for correcting the influence of the background and obtaining the line width of the repeated pattern. Is stored in the storage unit.

Here, another example of a method for determining a correction point in the pupil coordinate system in step 114 will be described with reference to FIGS. 14 (a) and 14 (b). In this example, paying attention to the pupil coordinates in the second quadrant of FIG. 4, the pupil coordinates of the measurement point and the correction point are obtained.
The horizontal axis of FIG. 14A represents the polarization direction of incident light (p-polarized light and s-polarized light) described in FIGS. 9A and 9B, the azimuth angle (0 degree, 90 degrees) of the reference wafer, It represents the wavelengths of light to be detected (red R, green G, blue B). (PxPy) on the horizontal axis in FIG. 14A is a continuous notation of the value of the Px axis and the value of the Py axis in the pupil coordinate system. Here, the pupil coordinate system (Px, Py) on the measurement surface DP is set as shown in FIG.

  In FIG. 14B, the pupil coordinate system is divided into 45 × 45 sections having 45 columns in the Px direction and 45 rows in the Py direction, and partitioned areas 60 and 62 passing through the quadrant and the optical axis where the measurement point 64 is located. Only used. (PxPy) of the partition region 60 used when the incident light is p-polarized is (123) to (2323), and (PxPy) of the partition region 62 used when the incident light is s-polarized is (2301) to (2323). ) And (2323) is on the optical axis. In the present embodiment, the correction points are in the partitioned areas 60 and 62.

  Further, the vertical axis of FIG. 14A represents the detection signal X of each partition in the partition regions 60 and 64 of FIG. 14B and the coefficients k1 and k2 determined in step 106 of FIG. A correlation coefficient between a CD conversion value calculated using the detection signal S of the measurement point 64 in the pupil coordinate system and a deviation value Δ between the CD-SEM value is shown. In this example, the correlation is high in the region 66 when the incident light is p-polarized light, the azimuth angle of the reference wafer 5B is 90 degrees, and the wavelength of the light to be used is red (R). Therefore, the correction point 68 in the partitioned area 60 in FIG. 14B corresponding to the area 66 is determined as the correction point AP1 in step 114 in FIG.

In this example, the wavelength of light used at the measurement point 64 tends to be higher in green (G) and / or blue (B) than in red, with a correlation with the line width.
Therefore, the wavelength of light having a high correlation with the deviation value Δ used at the correction point 68 (here, red) tends to be longer than the wavelength of light used at the measurement point 64. This is because the light used at the measurement point includes ground information of a repetitive pattern, and therefore, light in a wavelength range that transmits the ground to some extent, that is, light in a wavelength range such as red longer than blue and green is preferable. Conceivable. Therefore, in this case, it is preferable that the wavelength of light used at the measurement point 64 is green and / or blue, and the wavelength of light used at the correction point 68 is red. Thereby, the measurement accuracy of the pitch of the repetitive pattern can be improved.

  Next, in the surface inspection apparatus 1 of the present embodiment, the flowchart of FIG. 15 is shown for one example of a method for measuring the line width of the repetitive pattern of the mass production wafer using the calibration curve determined in step 116 of FIG. The description will be given with reference. This measurement operation is controlled by the main control system 3. This measuring method is performed, for example, to check whether the line width of the repeated pattern is within an allowable range after the exposure and development steps or after the etching step. Further, this measurement method is executed, for example, on a mass production wafer at the head of a predetermined lot in which the state of the base portion of the wafer has changed from the previous lot.

  First, in step 122 of FIG. 15, a mass production wafer as a test object is placed on the stage 6 of the surface inspection apparatus 1. Assume that the mass-produced wafer is the wafer 5 in FIG. In many shot regions SBj (j = 1 to M: M is an integer of several 10 to several 100) on the surface of the wafer 5, a repetitive pattern 30 composed of an L & S pattern having the same line width is formed. Note that repeated patterns having different line widths may be formed in the shot region SBj. Actually, patterns having various line widths are formed in the shot area SBj, and the repetitive pattern 30 indicates a pattern to be measured.

  In the next step 124, the azimuth angle θ of the wafer 5 is set to the azimuth angle of the reference wafer A determined in step 104 in FIG. 12A via the rotary stage 6B. In the next step 126, the light from the light source 11 is sequentially irradiated onto the plurality of measurement areas MA of the shot area SBj to be measured on the wafer 5 in a state of being linearly polarized in the X direction (direction R). In the next step 128, a pupil image of the reflected light from the wafer 5 is imaged by the image sensor 18. The pupil image is captured for each of the plurality of measurement areas MA.

In the next step 130, the arithmetic processing unit 20 acquires the detection signal S (tone value) of the light having the wavelength determined from each pupil image at the measurement point MP1 in the pupil coordinate system determined in step 104. . At this time, the CD conversion value (measurement point data) of Expression (9) may be calculated using the detection signal S.
In the next step 132, the azimuth angle of the wafer 5 is set to the azimuth angle (0 degree or 90 degrees) of the reference wafer B determined in step 114 of FIG. 12B via the rotary stage 6B. If the azimuth angle is set to approximately 0 degrees or approximately 90 degrees in step 114, the azimuth angle of the reference wafer B is also set to approximately 0 degrees or approximately 90 degrees in step 132. Is done. As an example, the azimuth angle of the wafer 5 is set as shown in FIG. In the next step 134, the light from the light source 11 is sequentially irradiated onto the plurality of measurement areas MB of the shot area SBj to be measured on the wafer 5 in a state where the light is linearly polarized in the X direction (direction R). In the next step 136, a pupil image of reflected light from the wafer 5 is picked up by the image pickup device 18. The pupil image is captured for each of the plurality of measurement areas MB.

In the next step 138, the arithmetic processing unit 20 acquires the detection signal X (tone value) of the light having the wavelength determined from each pupil image at the correction point AP1 in the pupil coordinate system determined in step 114. . At this time, the deviation value Δ (correction point data) of Expression (9) may be calculated using the detection signal X.
In the next step 140, the arithmetic processing unit 20 repeats the wafer 5 from the calibration curve (formula (9)) determined in step 116 using the detection signal S of the measurement point MP1 and the detection signal X of the correction point AP1. A corrected CD conversion value that is the line width of the pattern 30 is calculated. In the next step 142, the arithmetic processing unit 20 outputs the calculated corrected CD converted value to the main control system 3. The main control system 3 determines whether or not the corrected CD conversion value (measured value of the line width) is within the allowable range. If it is within the allowable range, the main control system 3 proceeds to step 146 and inspects the next specimen. I do. On the other hand, if the measured value of the line width is outside the allowable range in step 142, the process proceeds to the countermeasure process in step 144. For example, when the line width measured on the mass production wafer is the line width of the resist pattern, the mass production wafer is returned to the rework process. The rework process is, for example, a process in which a resist pattern is removed and a resist is applied again and exposed. In addition, if the measured value of the line width is outside the allowable range, the wafer may be discarded. Further, the wafer processing in the device manufacturing process may be corrected so as to improve the cause (for example, the exposure conditions of the exposure apparatus) outside the allowable range and to be within the allowable range.

In this way, it is possible to efficiently measure the line width of the repetitive pattern in an arbitrary measurement region of an arbitrary shot region of the measurement target with respect to the mass-produced wafer to be measured using the surface inspection apparatus 1.
When acquiring the measurement point MP1 and the correction point AP1, the measurement accuracy is improved if the number of measurement areas MA and the number of measurement areas MB in the shot area of the wafer 5 are set to the same number. However, since the acquired pupil image (data amount) increases, the measurement throughput decreases. Accordingly, for the correction point AP1, for example, only one measurement region MBC at the center in the shot region SBj in FIG. 16B is obtained without acquiring a pupil image in the measurement region MB corresponding one-to-one with the measurement point MP1. The pupil image is acquired by using the divergence value Δ calculated from the detection signal of the correction point AP1 of the pupil image in common with respect to the CD conversion values acquired for the plurality of measurement areas MA in FIG. May be.

In addition, the average value of the divergence values Δ acquired with respect to some of the plurality of measurement areas MB in the shot area SBj is commonly used to reduce the number of pupil images to be acquired, thereby reducing the measurement throughput. It can be improved.
As described above, the surface inspection apparatus 1 according to this embodiment includes the illumination optical system 10 that irradiates the surface of the wafer 5 (substrate) on which the repetitive pattern LSB is formed with linearly polarized light, and the reflected light from the surface of the wafer 5. Light intensity information of a polarization component perpendicular to the direction of the linearly polarized light (which may be substantially perpendicular) is detected on the objective lens 7 that receives light and the measurement surface DP that measures the light intensity distribution of the reflected light that passes through the objective lens 7. And an arithmetic processing unit 20 that obtains information on the line width of the repetitive pattern LSB from the detection signal of the light intensity information detected by the image sensor 18 in the detection optical system 15. Then, the arithmetic processing unit 20 obtains a detection signal S corresponding to the light intensity at the measurement point MP1 having a correlation with the line width on the measurement surface DP, and the correlation with the line width is measured on the measurement surface DP. A detection signal X corresponding to the light intensity at the correction point AP1 lower than the point MP1 is obtained, and information on the line width of the repetitive pattern LSB (corrected CD converted value of Expression (9)) is obtained using the detection signals S and X. (Steps 130, 138, 140).

As described above, according to the present embodiment, the light intensity at the measurement point having a high correlation with the line width of the repetitive pattern and the light intensity at the correction point having a correlation with the line width lower than the measurement point are used for repetition. Since errors due to factors other than the line width of the pattern (for example, the background state) can be reduced, the measurement accuracy of the line width information can be improved. Therefore, it is possible to suppress the measurement failure of the line width information.
Further, according to the present embodiment, in step 130, the detection signal S (CD information and background information) is obtained from the pupil image of the measurement point in the pupil coordinate system, and in step 138, the detection signal X is calculated from the pupil image of the correction point. (Background information). In step 140, the CD information from which the background information is removed is obtained by subtracting the deviation value Δ calculated from the detection signal X from the CD conversion value calculated from the detection signal S.

As a result, the deviation error from the line width value of the repetitive pattern measured by another line width measuring apparatus (surface inspection apparatus) such as a CD-SEM can be greatly reduced.
Further, according to the present embodiment, the error is based on the difference between the line width value calculated from the conversion formula and the line width value measured in advance by a CD-SEM or the like depending on the position in the plane of the reference wafer. A certain in-plane error can be reduced.

Furthermore, using the conversion formula determined by the reference wafer, for example, a line width value converted when measuring another wafer such as a mass production wafer used in the manufacturing process of a semiconductor device, and an actual measurement using a CD-SEM or the like It is possible to reduce an inter-wafer error, which is an error based on a deviation generated between the line width value.
Further, in the CD-SEM, as the acceleration voltage of the electron beam is increased to increase the measurement accuracy, the damage to the wafer as the test object increases. In the surface inspection apparatus 1 according to the present embodiment, since an electron beam is not used and, for example, visible light (which may be ultraviolet light or infrared light) is used, damage to the wafer is small. Wafer damage means, for example, that a pattern on a wafer is destroyed and does not have a predetermined performance.

Moreover, the measurement visual field of the surface inspection apparatus 1 can be set tens of thousands of times larger than that of the CD-SEM, and one measurement area can be measured within approximately one second. Therefore, a wide range of the wafer can be measured at high speed.
In the above-described embodiment, the measurement surface DP for measuring the light intensity distribution of the reflected light from the test wafer is provided on the pupil plane of the objective lens 7 or this conjugate plane. However, the measurement plane DP may be arranged in the vicinity of the pupil plane of the objective lens 7 or in the vicinity of a plane conjugate with the pupil plane. The surface in the vicinity thereof is a surface where the positions where the reflected light from the wafer passes are substantially different from each other when the incident angle K of the light incident on the wafer is different.

Further, the measurement points for obtaining the CD information and the background information may be a plurality of points, and the correction points for obtaining the background information may be a plurality of points. Furthermore, a high-order polynomial, an exponential function, or the like may be used instead of the equation (9).
In the above embodiment, the ground information of the repetitive pattern is obtained from the detection signal of the correction point in the pupil coordinate system detected in steps 132, 134, and 136. However, for example, the relationship between the various underlayer states of the wafer (in accordance with the semiconductor device manufacturing process) and the deviation value Δ in the equation (5) may be obtained and stored. In this case, when measuring the line width of the repetitive pattern of the wafer, the divergence value Δ corresponding to the state of the ground is read out from various stored divergence values instead of steps 132, 134, and 136. Good. This eliminates the step of changing the azimuth angle of the wafer and acquiring a detection signal of a predetermined correction point in the pupil coordinate system.

  In this embodiment, the repeated pattern on the wafer is measured, but an object different from the wafer may be used. For example, a printed board may be used. Furthermore, in the present embodiment, the repeated pattern is the measurement target, but the measurement target may be an isolated line in which the pattern is not repeated. Further, the interval between the repeated patterns is not single and may be different for each repetition. Further, the pattern shape may be, for example, an L shape other than the line shape, or a contact hole shape.

In the present embodiment, the linearly polarized light is irradiated onto the wafer 5. However, the light is not limited to linearly polarized light as long as it includes a polarized light component, and may be elliptically polarized light or circularly polarized light. Moreover, the light irradiated to the wafer 5 may not be completely polarized light including only specific polarized light but may be partially polarized light in a mixed state with non-polarized light including polarized light other than specific light.
Moreover, according to the surface inspection apparatus 1, when the light containing a polarization component changes with the patterns on a wafer, it is possible to reduce the error caused by the ground state. In this case, it is possible to irradiate light including a polarization component separately for each wavelength, and detect a change in polarization due to a pattern for each wavelength. Further, light including a plurality of wavelengths may be irradiated on the wafer, and a change in polarization due to the pattern may be detected for each wavelength. Even in this case, it is possible to reduce errors caused by the state of the ground. These methods include a scatterometry method or a scatterometer. References describing these methods include, for example, US Patent Application Publication No. 2008/0074666, US Patent Application Publication No. 2008/259343, and US Patent Application Publication No. 2006/0274310.

  Moreover, as the surface inspection apparatus 1, for example, an inspection apparatus that detects the diffracted light of the light irradiated on the wafer 5 can be used. As the surface inspection apparatus 1, for example, as described in U.S. Patent Application Publication No. 2006/0192953, light generated by irradiating a wafer with linearly polarized light in a regular reflection direction from the wafer. It is also possible to use an inspection apparatus that extracts defects and detects defects in the repeated pattern of the wafer. Even in the inspection apparatus described above, errors due to the state of the ground can be reduced.

As described above, the present invention is not limited to the above-described embodiment, and can have various configurations without departing from the gist of the present invention.
Note that the requirements of the above-described embodiments can be combined as appropriate. Some components may not be used. In addition, as long as it is permitted by law, the disclosures of all published publications and US patents related to the inspection devices and the like cited in the above embodiments and modifications are incorporated herein by reference.

  DESCRIPTION OF SYMBOLS 1 ... Surface inspection apparatus, 5 ... Wafer, 5A, 5B ... Reference wafer, 7 ... Objective lens, 8 ... Prism, 10 ... Illumination optical system, 14 ... 1st polarizing filter, 15 ... Detection optical system, 16 ... 2nd Polarizing filter, 18 ... image sensor, 20 ... arithmetic processing unit, 50 ... measurement point, 59 ... correction point

Claims (17)

Reflecting from the surface of the substrate irradiated with the light, irradiating the surface of the substrate including the base material and the coating material covering the surface of the base material and having a pattern formed thereon with light containing a polarization component Detecting the light intensity information of the light on the measurement surface of the light receiving optical system;
Obtaining a first light intensity at a first position on the measurement surface;
Using the information on the covering material and the first light intensity to obtain line width information of the pattern;
Including inspection methods. Detecting light intensity information of the reflected light on the measurement surface of the light receiving optical system,
Irradiating the surface of the substrate with linearly polarized light, and out of the reflected light from the surface of the substrate irradiated with the light, information on the light intensity of the polarization component substantially perpendicular to the direction of the linearly polarized light is used as the light receiving optics. The inspection method according to claim 1, comprising detecting on the measurement surface of a system. Determining a second light intensity at a second position on the measurement surface;
The inspection method according to claim 2, wherein information on the covering material is obtained using the second light intensity.   The inspection method according to claim 3, wherein measurement at the second position is performed based on a change in the state of the covering material.   The said measurement surface WHEREIN: The correlation with the light intensity in the said 2nd position and the state of the said coating | covering material is higher than the correlation with the light intensity in the said 1st position, and the said line | wire width. Inspection method.   In the measurement surface, the first position is a position corresponding to a direction intersecting with an arbitrary angle within a range of 0 degrees to 90 degrees with respect to a polarization direction of the linearly polarized light, and the second position is The inspection method according to claim 3, wherein the inspection method is located at a position corresponding to a vicinity of 0 degree or 90 degrees with respect to a polarization direction of the linearly polarized light. The light applied to the surface of the substrate includes blue light, green light and red light,
The first light intensity obtained at the first position is light intensity of at least one of blue light and green light,
The inspection method according to claim 3, wherein the second light intensity obtained at the second position is a light intensity of red light.   The inspection method according to any one of claims 3 to 7, wherein at least one of the first position and the second position is set at a plurality of locations on the measurement surface. The measurement surface includes a plurality of set positions,
From the plurality of positions,
In order to identify the first position,
Using a first reference substrate on which a pattern having a known line width is formed, obtaining correlation information between the light intensity of the polarization component and the known line width at each of the plurality of positions;
Obtaining deviation information between the line width obtained from the light intensity at the plurality of positions and the known line width;
In order to identify the second position from the plurality of positions,
Using a second reference substrate on which a pattern having a known line width is formed to obtain correlation information between the light intensity of the polarization component and the deviation information at the plurality of positions,
The inspection method according to any one of claims 3 to 8, further comprising: identifying a position having a high correlation between the light intensity of the polarization component and the deviation information from the plurality of positions.   The inspection method according to any one of claims 1 to 9, wherein the measurement surface of the light receiving optical system is a pupil surface of the light receiving optical system or a conjugate surface thereof, or a surface in the vicinity of these surfaces. An irradiation unit for irradiating the surface of the substrate on which the pattern is formed with light containing a polarization component;
A light receiving optical system for receiving reflected light from the surface of the substrate irradiated with the light;
On the measurement surface of the light receiving optical system, a detection unit that detects light intensity information of the reflected light;
A calculation unit for obtaining line width information of the pattern from the light intensity information detected by the detection unit,
The substrate includes a base material, and a coating material that covers the surface of the base material and on which the pattern is formed,
The said calculating part is a test | inspection apparatus which calculates | requires the 1st light intensity in a 1st position in the said measurement surface, and calculates | requires the information of the line width of the said pattern using the information of the said covering material, and the said 1st light intensity.   The inspection apparatus according to claim 11, wherein the irradiation unit irradiates the surface of the substrate with linearly polarized light, and the detection unit detects light intensity information of a polarization component substantially perpendicular to the direction of the linearly polarized light.   The inspection apparatus according to claim 12, further comprising a relative rotation mechanism that controls a relative angle between a polarization direction of the linearly polarized light applied to the substrate and a periodic direction of the pattern of the substrate.   The said calculating part calculates | requires the 2nd light intensity in a 2nd position in the said measurement surface, and calculates | requires the information of the said covering material using the said 2nd light intensity. Inspection equipment.   The inspection device according to any one of claims 11 to 14, wherein the detection unit includes an imaging device having a light receiving surface disposed on the measurement surface. The irradiation unit irradiates the surface of the substrate with light including blue light, green light, and red light,
The first light intensity obtained at the first position is light intensity of at least one of blue light and green light,
The inspection apparatus according to claim 14, wherein the second light intensity obtained at the second position is a light intensity of red light.   The inspection apparatus according to any one of claims 11 to 16, wherein the measurement surface of the light receiving optical system is a pupil surface of the light receiving optical system or a conjugate surface thereof, or a surface in the vicinity of these surfaces.

Images