JPH09145626A - Device for detecting defect of mask - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect a very small foreign matter by providing a gain adjusting circuit for adjusting gain of output signals from multiple photo-detection optical system, a space filter, and a photoelectric conversion means, for outputting, and inspecting defect based on relative intensity difference between the output signals sum-adjusted with a sum gain adjustment circuit. SOLUTION: Objective lenses L3, L4 and L5, L6 are assigned with Fourier conversion surfaces FI2 and FI3 in between, for constituting double diffraction optical system. The conversion surface FI3 is provided with a space filter S2, and a reflection detection system is constituted by the lenses L3-L6 and a photoelectric conversion element ID2. The reflection detection system is assigned, and the lenses L1 and L2 are assigned on the opposite side, with a mask R in between, for constituting the double diffraction optical system. The illuminating light that the mask R transmits is made incident to a photoelectric conversion element ID1, and a transmission detection system is constituted by the lenses L1 and L2 and the element ID1. The transmission image signal and the reflection image signal outputted from the elements ID1 and ID2 are inputted to a signal process circuit. The signal process circuit adjusts signal intensity based on the information from a computer.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mask defect inspection apparatus, and more particularly to an apparatus for optically detecting a defect of a mask having a circuit pattern such as an IC (including a defect such as a foreign substance attached to the mask). It is a thing.

[0002]

2. Description of the Related Art As a conventional mask defect inspection apparatus or foreign matter inspection apparatus, those disclosed in, for example, JP-A-58-62543 and JP-A-61-82147 are known.
FIG. 48 is a block diagram showing a conventional mask defect inspection apparatus. This is an example of an apparatus that detects a mask defect, a foreign substance, or the like by performing a comparison calculation process on the transmitted light 310 and the reflected light 303. In the figure, the light 301 emitted from the light source La is the lens groups L306 and L305.
Through. It is reflected by the mirror M301 and becomes a light ray 102.
The light ray (illumination light) 302 passes through the lens groups L304 and L303, is reflected by the half mirror HM, passes through the objective lens L301, and illuminates the circuit drawing surface 307 on the mask 305. There is a circuit pattern 306 on the mask. Of the reflected light and scattered light generated by the mask 305, the light ray 303 directed in the reflection direction is incident on the objective lens L301. The light detection surface of the one-dimensional photoelectric conversion element DR on the object surface which is conjugate with the drawing surface of the mask on the object surface 31
It is arranged so that 1 is located. The light ray 303 forms an image on the photoelectric conversion element DR to form a reflected bright field image (reflected image). The photoelectric conversion element DR outputs the image signal IRO. Similarly, of the light rays generated from the mask 305, the light ray 310 directed in the transmission direction is incident on the objective lens L302. A one-dimensional photoelectric conversion element DT is placed on the object plane conjugate with the mask drawing surface 307 on the object plane.
Is arranged such that the light detection surface 312 thereof is located. Ray 3
10 forms an image on the photoelectric conversion element DT to form a transmitted bright field image (transmission image). Ray 310 is photoelectric conversion element DT is image signal I
Output TO. The two image signals ITO and IRO are input to the image processing system SC, the relationship between the front image and the back image is electrically corrected, each signal is multiplied by an appropriate gain, and the coordinates on the mask become the same. To be added together. The summed signals are compared with a predetermined threshold value to detect the presence or absence of defects, foreign matter, and the like.

[0003]

However, in such an apparatus, there are cases where a minute mask defect, a foreign substance, or the like cannot be sufficiently detected by a simple binarization process using one or two threshold values. It was The present invention has been made in view of the conventional problems, and minute mask defects existing on the mask (
It is an object of the present invention to provide a mask defect inspection apparatus capable of detecting flat mask defects, etc.) and semi-transparent mask defects.

[0004]

In order to solve the above-mentioned problems, according to the present invention, in a mask defect inspection apparatus for optically detecting a defect on a mask having a circuit pattern, the mask is illuminated with inspection light. An illumination system; a first light receiving optical system for receiving the inspection light reflected by the mask; a second light receiving optical system for receiving the inspection light transmitted through the mask; and a first light receiving optical system A first spatial filter for blocking the inspection light passing through a central region including an optical axis of the first light receiving optical system in an optical Fourier transform plane for the circuit pattern in the second light receiving optical system; A second spatial filter for blocking the inspection light passing through a central region including the optical axis of the second light receiving optical system in an optical Fourier transform plane for the circuit pattern in the system; and the first space. Through the filter First for photoelectrically converting the inspection light
A second photoelectric conversion means for photoelectrically converting the inspection light that has passed through the second spatial filter; a gain of a first output signal from the first photoelectric conversion means, A gain adjusting circuit for outputting a third output signal, adjusting the gain of the second output signal from the second photoelectric conversion means, and outputting as a fourth output signal. The gain is adjusted by the gain adjusting circuit. And the third output signal and the fourth
The defect is detected based on either the relative intensity difference from the output signal or the intensity ratio.

Further, according to the present invention, a defect of a mask having a circuit pattern on a first surface of a light-transmissive plate-like substrate and not having a circuit pattern on a second surface opposite to the first surface. For inspecting a mask, the first inspection target being a mask.
Illuminating means for illuminating an inspection area, which is an area in the plane of
Two light receiving means separately arranged in two spaces obtained by dividing the space including the mask by a plane including the first surface, the light receiving means being arranged in a second space including the second surface. Second light receiving means and first light receiving means disposed in the first space not including the second surface; a light beam generated from the inspection region of the mask, the first light receiving means First photoelectric conversion means for photoelectrically converting a light ray incident on the means; second photoelectric conversion means for photoelectrically converting a light ray generated from the inspection region of the mask and incident on the second light receiving means A first output signal output from the first photoelectric conversion means is gain-adjusted and output as a third output signal, and a second output signal output from the second photoelectric conversion means is adjusted in gain; A gain adjustment circuit for outputting four output signals, It was to detect the defects based on one of a relative intensity difference or intensity ratio of the third output signal of the gain is adjusted and said fourth output signal by the circuit.

Further, according to the present invention, in a defect inspection apparatus for inspecting a defect of a mask, first illuminating means for epi-illuminating the inside of a first region of the inspected surface of the mask, and first of the inspected surface.
Second illuminating means for transmitting and illuminating the area from a surface different from the surface to be inspected of the mask, and collecting only scattered light generated from the second area at the first position in the first area, First
A ray selecting means for making a ray, a ray separating means for separating the first ray into a second ray and a third ray, a first photoelectric converting means for photoelectrically converting the second ray, and a third ray A second photoelectric conversion unit that performs photoelectric conversion and a first output signal output from the first photoelectric conversion unit that is gain-adjusted and output as a third output signal, and a second output signal output from the second photoelectric conversion unit. A gain adjusting circuit for adjusting the gain of the output signal and outputting the gain as a fourth output signal is provided, and a relative intensity difference between the third output signal and the fourth output signal whose gains are adjusted by the gain adjusting circuit, or The defect is detected based on either one of the intensity ratios.

Further, the present invention is a plate-shaped light-transmitting substrate having a first surface and a second surface facing each other, wherein a pattern is drawn on the first surface and a pattern is drawn on the second surface. A defect inspection apparatus for inspecting defects of a substrate which is not, a laser light source for emitting a laser beam, and the laser beam is focused on a first region of the first surface of the substrate, and irradiated,
In the pattern drawing surface of the substrate, the first region is 1
An optical scanning means for continuously moving in a dimensional direction and one of two space areas divided by a first plane including the first surface of the substrate and arranged in a first space area not including the second surface. And a first light receiving means having a photoelectric conversion element and a second space area which is one of two space areas divided by a first plane including the first surface of the substrate and which includes the second surface. And a second light receiving means having a photoelectric conversion element, and the first light receiving means.
The gain adjustment of adjusting the gain of the first output signal from the light receiving means, outputting the third output signal, adjusting the gain of the second output signal from the second light receiving means, and outputting the fourth output signal. A circuit for detecting the defect based on either a relative intensity difference or intensity ratio between the third output signal and the fourth output signal whose gain is adjusted by the gain adjustment circuit. And

[0008]

BEST MODE FOR CARRYING OUT THE INVENTION

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a view showing the schematic arrangement of a mask defect inspection apparatus according to this embodiment, which is an apparatus having both a reflection detection system and a transmission detection system. The beam diameter of the illumination light IL emitted from the light source 1 (Ar laser, etc.) is expanded by the beam expander (lens L7, lens L8), and the objective lens L3
The light enters the vibrating mirror (half mirror) 4 located at the rear focal plane FI2. Illumination light IL reflected by the vibrating mirror 4
Enters the objective lens L3.

The position of the mask R is adjusted so that the front focal plane of the objective lens L3 coincides with the pattern PA surface of the mask R. The objective lens L3 focuses the illumination light IL on the mask R, An illumination area on the spot is formed on the mask (in FIG. 1, the optical path of the illumination light IL from the objective lens L3 to the mask R,
The optical path of the specularly transmitted light that has passed through the mask R from the mask R to the lens L1 and the lens L5 of the illumination light specularly reflected by the mask R
The optical path up to is omitted).

The vibrating mirror 4 has a rotation axis parallel to the x axis (axis perpendicular to the plane of the drawing), and has a drive system M composed of a motor or the like.
O is the optical axis AX centered on this rotation axis by the drive system MO
The vibrating mirror 4 is rotated. The drive system MO rotates the vibrating mirror 4 and scans the illumination light IL in the y direction. Objective lens L3 is a telecentric system and vibrating mirror 4
Even if the irradiation position of the illumination light IL on the mask R is changed by the, the illumination light IL is incident on the mask substantially parallel to the optical axis AX. In FIG. 1, illumination light IL from scanning point PL to scanning point PR
Be scanned in the y direction.

The apparatus of FIG. 1 is provided with a light receiving optical system (L3, L4, L5, L6) for receiving the reflected light from the mask R. These lenses L3, L4, L5, L6 are arranged along the optical axis AX. Objective lens L3 (lens with focal length f3) and lens L4 (lens with focal length f4) are arranged so that the rear focal plane of the objective lens and the front focal plane of lens L4 coincide with each other. L4 constitutes a double diffraction optical system. Therefore, the Fourier transform of the amplitude distribution of the light wave on the object-facing surface ob is obtained on the Fourier transform surface FI2. This means that the object plane is on the Fourier transform plane FI2.
This means that a Fourier transform of the amplitude distribution of the diffracted light (reflected light) from the mask pattern PA arranged in ob is obtained.

An image plane IM2 conjugate with the object is formed on the rear focal plane of the lens L4. The Fourier transform of the amplitude distribution of the light wave on the image IM2 is obtained on the Fourier transform plane FI2. Similarly, the lenses L5 and L6 are also double diffractive optical systems, and a spectrum (Fourier diffraction image) conjugate with the Fourier transform plane FI2 is obtained on the Fourier transform plane FI3.
, An image conjugate with the image formed on the image plane IM2 is obtained.
A spatial filter S2 is provided on the Fourier transform plane FI3 (details will be described later).

In this embodiment, the two sets of double diffractive optical systems are connected in series in this manner, in order to optically scan the beam spot on the object-facing surface ob, the vibrating mirror is placed on the Fourier transform surface FI2. This is because the MO is arranged and it is difficult to arrange the spatial filter on this surface. The light receiving surface of the one-dimensional photoelectric conversion element ID2 is arranged on the image plane IM3, and the signal IDD2 from the photoelectric conversion element ID2 is sent to the main control system 100 (FIG. 2).

As described above, the light receiving optical system (L3, L4, L5,
A reflection detection system is composed of L6) and photoelectric conversion element ID2.
Next, the transmission detection system will be described. On the side opposite to the reflection detection system across the mask R, the lens L1 (
A lens having a focal length f ₁ ) and a lens L ₂ (lens having a focal length f ₂ ) are arranged. The illumination light IL that has passed through the circle R enters the photoelectric conversion element ID1 via the lenses L1 and L2. lens
The lens L1 is arranged so that the front focal plane of L1 substantially matches the pattern surface PA of the mask. An image plane IM1 is formed on the back focal plane of the lens L2, and the photoelectric conversion element ID1 is arranged so that the image plane IM1 and the image surface IM1 of the photoelectric conversion element ID1 coincide with each other.
Is arranged.

Therefore, the lenses L1 and L2 also form a double diffraction optical system, and the Fourier transform of the amplitude distribution of the light wave on the object plane ob is obtained on the Fourier transform plane FI1, and an image conjugate with the object plane ob is formed. Obtained on plane IM1. The Fourier transform plane FI1 is provided with a spatial filter S1 (details will be described later). As described above, the light receiving optical system (L1, L2) and the photoelectric conversion element ID1 constitute a transmission detection system.

The objective lenses L1 and L3 are telecentric.
The illuminating light IL is
Always make an angle with the normal to the object-facing ob (the line parallel to the optical axis AX).
Degree θ ₀ And then enters the mask R. In addition, the optical axis is shown in FIG.
Regular transmission through the scan area PR-PL of the mask R along AX
Overlight (0th-order diffracted light) 11 and illumination light IL illuminate scanning point PL
At this time, the ray 12 generated by passing through the mask R and the mask R
The emitted light beam 16 and the illumination light IL illuminate the scanning point PR.
Light rays 13 that pass through the mask R when it strikes and the mask
Ray 15 is shown, which is reflected and produced by R. these
Rays 12, 13, 15 and 16 of are normals to the object-facing ob (parallel to the optical axis AX
Ray) which forms an angle θm with respect to

Further, the mask R is held by the mask stage RST, and the moving mechanism DV is arranged so that the pattern drawing surface PA is always positioned within the object plane ob of the objective lenses L3 and L1 and the optical axis AX (optical scanning line in the y direction). ), And the mask stage RST is moved in the y direction parallel to the x and y planes. A predetermined inspection area (two-dimensional area) of the mask R can be inspected by optical scanning by the movement of the vibrating mirror 4 and movement of the mask R by the movement mechanism DV. The main control system 100 controls the movement of the vibrating mirror 4, the movement of the mask R by the drive mechanism DV, the processing of signals from the photoelectric conversion elements ID1 and ID2, and the like, and comprehensively controls the entire apparatus.

Next, the main control system 100 of FIG. 1 will be described with reference to FIG.
Description will be made with reference to (a) and (b). 2 (a),
In (b), the main control system 100 has a photoelectric conversion element ID1.
Signal IDD1 output from the photoelectric conversion element ID2 and the signal IDD1 output from the photoelectric conversion element ID2 are input.
A signal processing circuit SC including an amplifier for amplifying each of the DD2 and adjusting circuits GA1 and GA2 for adjusting the gain of the amplifier.
And a main computer COM, an input device IF for inputting information such as an inspection area, and a display D for displaying an inspection result of a mask defect and the like. Here, FIG. 2 (a)
In FIG. 2B, the difference between the gain-adjusted signals is obtained by the differential amplifier DA, and the mask defect is inspected based on the difference. In FIG. 2B, the gain adjustment is performed. The ratio of each signal is obtained by the divider D, and the mask defect is inspected based on this.

The input device IF is a keyboard, a mouse, a light pen, a floppy disk drive, a bar code reader or the like. For example, an operator inputs information such as the setting value of the amplification gain, the inspection sensitivity, and the inspection area for the signals IDD1 and IDD2 based on the information of the transmittance of the mask and the reflectance of the pattern on the mask from the keyboard (input device IF) to the computer COM. To be done. Computer CO
M uses the signal IDD1 based on these input information.
And the set value of the amplification gain for the IDD2 are output to the signal processing circuit SC.

A transmission image signal IDD1 and a reflection image signal IDD2 output from the two photoelectric conversion elements ID1 and ID2.
Is input to the signal processing circuit SC. The signal processing circuit SC adjusts the signal strength of the two signals with a gain set based on the information from the computer COM. This gain adjustment amount is adjusted according to the principle described later such that the difference signal between the photoelectric signal of the reflected light from the circuit pattern and the photoelectric signal of the transmitted light becomes zero. Next, the signal processing circuit SC subtracts or ratios the two signals after the gain adjustment, and obtains the signal IDD3. The signal processing circuit SC binarizes this signal IDD3 by at least one of the two threshold values THH and THL by the window comparator circuit in the signal processing circuit SC. That is, the signal IDD3 is the threshold
When it is larger than THH, "1" is output to the computer COM, and when signal IDD3 is smaller than the threshold value THH, "0".
Is output to the computer COM. Or signal ID
When D3 is larger than the threshold THL, "0" is output to the computer COM, and when the signal IDD3 is smaller than the threshold THL, "1" is output to the computer COM. The computer COM determines that a mask defect has been detected when the logic value “1” is output from the signal processing circuit SC (when the binarized result is output from at least one of the thresholds THH and THL). When the logic value "0" is output from the signal processing circuit, it is determined that no mask defect has been detected (only the signal from the circuit pattern or the glass surface of the mask R can be obtained).

The computer COM maps the detection result of the mask defect on the display D based on the information of the detection result of the mask defect, the position of the detection light IL in the X direction and the position information of the mask R in the Y direction. . Further, the signal processing circuit SC can also correct the gain setting value by obtaining the difference between the reflectance and the transmittance of the mask defect from the two signal signal intensities.

Next, the spatial filters S1 and S2 will be described with reference to FIGS. 3 (a) and 3 (b). FIGS. 3A and 3B are views showing a Fourier transform surface formed in each light receiving system and a spatial filter provided therein.
The area Ff1 in FIG. 3A is defined by the lenses L1 and L2, and shows the circular area Ff1 on the Fourier transform plane FI1.
The area Ff2 in (b) is defined by the lenses L3, L4, L5, and L6, and indicates the circular area Ff2 on the Fourier transform plane FI3 (FI2 is the same). A region Ff1 indicates a region through which light from the mask R that can reach the photoelectric conversion element ID1 passes, and the region Ff1
f2 indicates a region through which light from the mask that can reach the photoelectric conversion element ID2 passes. In addition, FIG.
The xy coordinate system in (b) corresponds to the xy coordinate system in FIG.

In this embodiment, the Fourier transform plane of FIG.
Area Ff1 on FI1 and Fourier transform plane FL of FIG. 3 (b)
The Fourier spectra (Fourier diffraction images) transmitted through the area Ff2 above 2 are matched. Therefore, the Fourier transform plane
The lens L3, so that the maximum angle θm of the ray emitted from the object-facing surface ob corresponding to the maximum spatial frequency that can pass through FI1 and FI3 with respect to the optical axis AX in the transmitted light and the reflected light is the same.
Numerical aperture and magnification of L4, L5, and L6, numerical aperture of lenses L1 and L2,
The magnification is fixed.

As shown in FIG. 3A, in the area Ff1 on the Fourier transform surface, a spatial filter S1 for blocking the 0th-order diffracted light (transmitted light 11) of the illumination light IL in the area Ff1 is provided, As shown in FIG. 3B, a spatial filter S2 that blocks the 0th-order diffracted light (reflected light 14) of the illumination light IL in the area Ff2 is provided in the area Ff2 on the Fourier transform surface. Uv in FIGS. 3A and 3B represents a coordinate system on the Fourier transform plane.

Diffracted lights 12 and 13 generated from the mask R (generated by a circuit pattern on the mask and a mask defect) are Tm1 and Tm on the Fourier transform plane FI1 (on the area Ff1).
Diffracted lights 15 and 16 that pass through the area indicated by 2 and are reflected by the mask R (generated by being reflected by a circuit pattern or a mask defect on the mask) are on the Fourier transform surface FL2 (on the area Ff2). It passes through the areas indicated by Rm1 and Rm2. The points Rm1, Rm2, Tm1 and Tm2 are located at the limits (outermost area) of the regions Ff1 and Ff2.

Therefore, the transmitted light ray component from the mask R forming an angle θm with the optical axis AX (the angle between the principal ray of the transmitted light ray and the optical axis AX is θm) is the area F on the Fourier transform plane FL1.
This is the limit ray that can pass through f1. This means that the angle θm represents the size of the region Ff1. Similarly, a reflected ray component from the mask R forming an angle θm with the optical axis AX (the principal ray of the reflected ray and the optical axis AX
The angle formed by and is a limit light ray that can pass through the area Ff2 on the Fourier transform surface FL2, and the angle θm represents the size of the area Ff2.

Further, in FIG. 1, Fourier transform planes FI1 and FI
The intersection of 3 and the 0th-order diffracted light (transmitted, reflected) is shown by a point, but in reality, the illumination light IL is incident on the mask R at a predetermined angle (θ ₀ ), that is, the NA of the light beam. Therefore, the intersection on the Fourier transform plane has a size corresponding to the angle (θ ₀ ). In the above, the spatial filters S1 and S2 are 0.
Although it is described that the 0th order diffracted light is shielded, in other words, this means that the 0th order diffracted light has a spread of ± angle θ ₀ with respect to the optical axis AX (corresponding to the size of the 0th order diffracted light on the Fourier transform surface). Assuming that the spatial filter S1 has the object surface o
Among the light rays generated from b that travel in the transmission direction, the light rays having an angle component of an angle θ ₀ or less with respect to the normal to the object plane ob are blocked, and a light ray having an angle component of angles θ _{0 to} θ _m is passed. It is to let. Similarly, the spatial filter S2 blocks light rays having an angle component equal to or smaller than the angle θ _{0 with} respect to the normal line of the object surface ob from the light rays traveling in the reflection direction generated from the object surface ob, and the angles θ ₀ to θ _m . Allows a ray having a component to pass through.

In FIGS. 3A and 3B, the Fourier shown is
Let u be the diameter of the circular spatial filters S1, S2 on the conversion surface.
2u in v coordinate system₀ (= 2v₀ ), The angle
2θ ₀ Is this diameter 2u₀ (= 2v₀ ) Is supported.
At points Ti1 and Ti1, the 0th-order diffracted light 11 (transmitted light) is the Fourier of Fig. 1.
D) Indicates the point of collision on the conversion surface FI1. Also, the points Ri1 and Ri
2 is the 0th-order diffracted light 14 (reflected light: light with the same angle component as light ray 11
Line) is generated from the object plane ob, the Fourier transform plane FI3
Indicates the point of collision with.

Two photoelectric conversion elements ID1 and I shown in FIG.
The difference between the signals IDD1 and IDD2 output from D2 and the adjustment of the gain will be described in detail below. Object plane ob
The amplitude reflectance at the circuit pattern portion on the mask R that produces the reflected scattered light above is α, and the function of the amplitude distribution of the reflected light is f ₁
Assuming that (x, y), the reflectance distribution is αf ₁ (x, y)
, Β is the amplitude transmissivity of a portion without a circuit pattern that produces transmitted scattered light, and f ₂ (x,
y), the transmittance distribution is represented by βf ₂ (x, y). These have the relationship of equation (1).

1-f ₁ (x, y) = f ₂ (x, y) (1) Further, the amplitude distribution of the light wave of the incident beam spot is f _0.
Assuming that (x, y), the Fourier spectrum (Fourier diffraction image) is expressed by the equations (2) and (3). Fourier transform plane FI3
Spectrum above _{^{FT - [αf 1 (x,}} y) × f 0 (x, y)] = FT - [αf 1 '(x, y)] = αF 1 (u, v) * F 0 (u, v) = αF ₁ ′ (u, v) (2) FT ⁻ [αf ₁ ac (x, y) × f ₀ (x, y)] = αF ₁ ac (u, v) * F ₀ (u, v) ) = ΑF ₁ 'ac (u, v) (3), and the spectrum on the Fourier transform plane FI 1 is FT ⁻ [βf ₂ (x, y) × f ₀ (x, y)] = FT ⁻ [βf ₂ ' (X, y)] = βF ₂ (u, v) * F ₀ (u, v) = βF ₂ '(u, v) (4) FT ⁻ [βf ₂ ac (x, y) × f ₀ (x , Y)] = βF ₁ ac (u, v) * F ₀ (u, v) = − βF ₁ ′ ac (u, v) (5).

Where F ₀ (u, v), F ₁ (u, v),
F ₂ (u, v) has a relationship of (6), (7), (8).

[0033]

(Equation 1)

[0034] _{^{FT - [f 1 (x,}} y)] = F 1 (u, v) (7) FT - [f 2 (x, y)] = F 2 (u, v) (8) Babinet principle of Thus, equation (9) holds. F ₁ ′ (u, v) + F ₂ ′ (u, v) = F ₀ (u, v) (9) Here, the functions F ₁ ′ ac (u, v) and F ₂ ′ ac (u, v) are newly added. )
Is expressed by equations (10) and (11).

[0035]

(Equation 2)

From equations (6) and (9), equation (12) is obtained. F ₁ ′ ac (u, v) + F ₂ ′ ac (u, v) = 0 (12) Therefore, the Fourier spectrum that passes in this embodiment is the spectrum that passes FI 1 = βF ₂ ′ ac (u, v) ( 13) Spectra passing through FI3 = αF ₁ 'ac (u, v) (14) Equations (13) and (14) are shown, and they are the same except the coefficient from Equation (12). In FIG. 29, f ₁ (x), f ₂ (x), F ₁ '(u), and 1-dimensional case
The state of F ₂ '(u) was shown. The circuit of FIG. 29 is a case where the line: pattern is 1: 1.

The images generated on the image planes, IM3 and IM1, are (1
It is shown by the equations (5) and (16). Image = FT on the image plane _{^{IM1 - [βF 2 'ac (}} u, v)] = βf 2' ac (x, y) = -βf 1 'ac (x, y) (15) the image on the image plane IM3 _{^{= FT - [αF 1 'ac}} (u, v)] = αf 1' ac (x, y) (16) which is the same except coefficients.

Since the image is output in intensity, the electrical output IDD
1 and IDD2 are k · IDD1 = k [βf ₂ ″ (x, y)] ² = kβ ² [(f ₁ ″ (x, y)] ² (17) l · IDD2 = 1 (αf ₁ ″) (X, y)] ² = lα ² [f ₁ ″ (x, y)] ² (18) where k and l are the gains of the amplifier (17),
The same output can be obtained if kβ ² = 1α ² (19) (19) is established from Eq. (18).

The output signals IDD1 and IDD2 are input to the signal processing circuit SC in FIG. 2, but mask defects can be easily detected by setting the sensor gains k and l so as to satisfy the equation (19). Equation (19) indicates that when optimized for the reflectance α of the circuit pattern and the transmittance β of the glass portion of the mask R, k · IDD1−l · IDD2 = 0 (20). In other words, if either one of the reflectance of the mask defect is different from α or β (20)
The formula no longer holds. Therefore, the signal IDD3 such as (21)
It is sufficient to create and binarize this with an appropriate threshold (THH, THL by the window comparator).

IDD3 = k · IDD1−l · IDD2 (21) As is apparent from the equation (42), the present invention erases the image of the defect-free circuit pattern by subtracting the transmitted dark field image and the reflected dark field image. , A mask defect in which either the amplitude reflectance or the amplitude transmittance on the photomask is different from that of the defect-free circuit pattern can be depicted.

Next, a second embodiment of the present invention will be described. The apparatus configuration of this embodiment is almost the same as the inspection apparatus of the first embodiment shown in FIG. The difference from the first embodiment is that, as shown in FIG. 4, in addition to the spatial filters S1 and S2, spatial filters S11 and S22 (added to the spatial filters S1 and S2) are added to the Fourier transform planes FI1 and FI3. That is the point.

In FIG. 4, spatial filters S11 and S2
2 is when the illumination light IL is incident on the straight line portion of the pattern forming an angle of 0 °, 45 °, 95 °, 135 ° with respect to the x axis, which is the main component of the circuit pattern on the object plane ob. The generated Fourier transform pattern is blocked (blocked). In each of the spatial filters S11 and S22, three bands are mixed at an angle of 45 ° about the optical axis AX. The width of the three bands is the size of the Fourier transform spectrum of the illumination light generated when there is no circuit pattern on the object plane ob (mask R
The diameter (= 2u ₀ = 2v ₀ ) of the spatial filter S1 corresponding to the size of the passing region on the Fourier transform surface FI1 of the 0th-order diffracted light (transmitted light) of the illumination light IL transmitted through the object surface ob is uniform. Of the Fourier transform spectrum of the illumination light generated when it is covered with a highly reflective circuit pattern drawing material (at the Fourier transform surface FI3 of the 0th order diffracted light (reflected light) of the illumination light IL reflected by the mask R). The diameter of S2 corresponding to the size of the passage area (= 2u ₀ = 2v)
_By matching with ₀ ), the Fourier transform pattern (diffraction image) of the diffracted light generated from the straight line portion can be blocked.

The spatial filters S11 and S22 block the Fourier transform pattern generated from the linear portion of the circuit pattern, but pass the Fourier transform spectrum (diffraction image) of other spatial frequencies. The Fourier transform spectrum passing through is, for example, a Fourier transform spectrum of a mask defect composed of a random curve rather than a straight line portion, or 2
It is a part of the Fourier spectrum of a cell-shaped circuit pattern having a dimensional periodic structure.

According to the inspection apparatus having the spatial filter of FIG. 4 as in the second embodiment, most of the scattered light generated from the circuit pattern can be blocked, so that the image of the circuit pattern incident on the photoelectric conversion block is blocked. The amount of light can be reduced. Therefore, as compared with the inspection apparatus of FIG. 1 including only the spatial filters S1 and S2 shown in FIG. 3, the contrast of the image of the mask defect with respect to the circuit pattern image can be improved, and the minute low step mask defect can be detected more easily. .

Next, a third embodiment of the present invention will be described. FIGS. 5A, 5B, and 5C are diagrams showing a light shield and a spatial filter used in the inspection apparatus of the third embodiment of the present invention. The difference between the inspection apparatus of this embodiment and the apparatus of FIG. 1 is that the spatial filters S13 and S23 and the illumination light IL are used.
Is a light-shielding band A1 provided in the optical path of. Since the other parts are the same as those of the apparatus shown in FIG. The light-shielding band A1 is located in the optical path of the illumination light IL incident on the mask R near the Fourier transform surface FI2. FIG. 5A is a diagram showing the light shield A1, and as shown in the figure, the light shield A1 is composed of two light shields that intersect the illumination light IL in a cross shape in a substantially vertical plane. The illumination light IL that has passed through the light shield A1 is divided into four, is focused on the object plane ob, and illuminates one point. The Fourier transform spectrum of the transmitted illumination light flux generated when there is no circuit pattern of the mask R is shown by S01 in FIG. 5B, and the circuit pattern drawing surface of the photomask 1 is covered with a uniform reflective circuit pattern drawing material. The Fourier-exchange spectrum of the reflected illumination generated during the operation is shown by S02 in FIG.

The spatial filters S13 and S23 are shown in FIG.
As shown in (b) and (c), the regions where the Fourier spectra S01 and S02 (diffraction images on the Fourier transform plane) of the illumination light flux do not exist, and along the u axis and the v axis,
The light flux of the spatial frequency in the cross-shaped slit having the width v ₁ is transmitted, and the light flux of the other spatial frequencies is blocked. The spatial filters S13 and S23 block most of the Fourier transform pattern generated by the circuit pattern and transmit a part of the Fourier transform pattern generated by the mask defect. Therefore, the intensity of the circuit pattern image can be reduced and the contrast of the mask defect image can be improved. The spatial filters S13, S23 are the spatial filters S11, S23 shown in FIG.
Since the Fourier transform pattern of the cellular circuit pattern of the two-dimensional periodic structure transmitted by S22 is also blocked, the contrast of mask defects in such a circuit is also improved.

Next, a fourth embodiment of the present invention will be described. The difference between the device of the present embodiment and the device of FIG. 1 is that the device of the present embodiment does not include the lenses L2 and L6 and the photoelectric conversion element ID
1 and ID2 are arranged on the Fourier transform planes FI1 and FI3, and the photoelectric conversion elements ID1 and ID2 are single pixel sensors. Others are the same as those of the apparatus of FIG. 1, and therefore detailed description of the apparatus configuration using the drawings will be omitted.

Also in this embodiment, the Fourier transform plane FI
1, spatial filters S1 and S2 for blocking the Fourier transform pattern of the illumination light flux are provided on the FI3 similarly to FIG. Therefore, the Fourier spectra on the Fourier transform planes FL1 and FL3 are the same except for the coefficients according to the equations (13) and (14). Therefore, the photoelectric conversion elements ID1 and ID2 provided on the Fourier transform planes FL1 and FL3 are as shown in FIG.
In the spatial frequency regions Ff1 and Ff2 shown in (b), spatial filters S1 and S2 are provided, and the light amount obtained by integrating the entire spectrum of the light passing through the region excluding the region is photoelectrically converted, and outputs IDD1 and IDD2 are output. These are expressed by equations (22) and (23).

K · IDD1 = k | βF ₂ ″ (u, v) | ² (22) l · IDD2 = l | αF ₁ ″ (u, v) | ² (23) where the gain of the amplifier is represented by k, l is k as in Eq. (19)
If β ² = lα ² , the same output can be obtained in the circuit pattern having no mask defect. Here, the signal IDD3 is created in the same manner as in the equation (21) and binarized with an appropriate threshold value (THH, THL). For example, mask defects can be detected. As described above, in this embodiment, on the Fourier transform surface, the amount of transmitted light and the amount of reflected light are subtracted to generate a signal IDD3 having a signal value of zero when there is no mask defect, and the mask defect can be easily detected. .

In the present embodiment, ID1 and ID2 output outputs IDD1 and IDD2, which represent the amount of light determined by the illumination position of the illumination luminous flux IL as one value. The signal processing unit SC performs the above-described signal subtraction processing and binarization processing, and the computer COM controls a series of inspection operations. Computer COM translates photomask R,
The scan of the vibration mirror 4 is also controlled, and the adhesion position of the mask defect is obtained based on the position information and the detection information of the mask defect.

FIG. 6 shows an inspection apparatus according to the fifth embodiment of the present invention. The difference from the device of the fourth embodiment is only the number of photoelectric conversion elements and the signal processing system associated therewith. Photoelectric conversion elements SA1, SA2, SA for transmitted light such that the light detection surface (light receiving surface) is located on the Fourier transform surfaces FI1, FI3.
3, SA4 and photoelectric conversion elements SB1, SB2, S for reflected light
B3 and SB4 are installed. These photoelectric conversion elements include SA1 and SB1, SA2 and SB2, SA3 and SB3, SA
4 and SB4 are paired respectively and are arranged so that the expression (12) is established in the area of each light receiving surface.

FIG. 7 (a) is a diagram showing the arrangement of the photoelectric conversion elements SA1, SA2, SA3, SA4 on the Fourier transform plane, and FIG. 7 (b) is the photoelectric conversion elements SB1, SB2, SB3, on the Fourier transform plane. It is a figure which shows arrangement | positioning of SB4.
That is, in the uv Cartesian coordinate system in FIGS. 7A and 7B, the regions in the same quadrant are paired. FIG. 8 is a diagram showing a signal processing system in this embodiment. In FIG. 8, four pairs of photoelectric conversion outputs are input to four signal processing circuits SC1 to SC4 and 17 to 20 to perform addition and subtraction processing. Photoelectric conversion element SA1 ~
Output signals from SA4, SB1 to SB4 are SA01 to SA04,
It was set to SB01 to SB04. The output signals from SC1 to SC4 are SC01 to SC04. The output signal SC01 _-
～SC04 _- subscript - mark indicates the subtraction output, the output signal S
The ₊ symbol as a subscript of C01 _{+ to} SC04 ₊ indicates subtraction output.

Since each pair of four photoelectric conversion elements satisfies the equation (12), if S _c1 to _{C 4} are defined as follows, the four signals are binarized with appropriate threshold values (THH, THL). Then, the mask defect can be detected from each signal. _{SC01 - = k · SA01-l} · SB01 (24) SC02 - = k · SA02-l · SB01 (25) SC03 - = k · SA03-l · SB03 (26) SC04 - = k · SA04-l · SB04 ( 27) By the way, the Fourier transform spectrum of the circuit pattern is often biased on the Fourier transform plane. Utilizing this fact, in the present embodiment, it becomes possible to detect a finer low-step mask defect by selecting a pair of photoelectric conversion elements having a smaller amount of integrated light with each light receiving surface of the Fourier transform spectrum of the circuit pattern. ing.

This fifth embodiment is also similar to the above-mentioned embodiment.
Indicates the amount of transmitted light and reflected light. A first photoelectric conversion signal,
The second photoelectric conversion signal is subtracted to create a third photoelectric conversion signal, and the third photoelectric conversion signal is converted to an appropriate slice level (T
HH, THL) and binarize the signal of the mask defect. In carrying out such a method, it is necessary to subtract the signal within the dynamic range of the photoelectric conversion element. The photoelectric conversion element outputs a signal proportional to the amount of light, but the output signal has a minimum output value due to electrical noise and a maximum output value determined by the maximum rating. Usually, the dynamic range is determined by the maximum output / minimum output, and has a value unique to each photoelectric conversion element. (The resolution of the dynamic range is a ratio, but the area is conventionally expressed.) To increase the amount of scattered light of a minute mask defect above the minimum output value, increase the brightness of the illumination light IL. Good. In this case, the amount of scattered light from the circuit pattern also increases. If the amount of scattered light from the circuit pattern exceeds the maximum output value, (21)
The third photoelectric conversion signal according to the formula differs from the ideal value, and normal mask defect inspection cannot be performed.

In order to avoid the above phenomenon, in this embodiment,
On the Fourier transform plane, a region in which the amount of scattered light from the circuit pattern is small can be selected from four regions. As described above, in FIG. 8, four pairs of photoelectric conversion outputs are input to the four signal processing circuits SC1 to SC4, and addition and subtraction processing is performed in each. In the figure, the + symbols represent addition outputs SC01 ₊ , SC02 ₊ , SC03 ₊ , SC04 ₊ .

Signs − indicate subtraction outputs SC01 ₋ , SC02 ₋ , S
C03 _-, SC04 _- represents a. SC01 ₊ = k SA01 + l SB01 (28) SC02 ₊ = k SA02 + l SB02 (29) SC03 ₊ = k SA03 + l SB03 (30) SC04 ₊ = k SA04 + l SB04 (31) SC01 ₊ ~ SC04 ₊ is (28) to (31) Indicated by. The four addition outputs are input to the minimum value selection unit k ₁ and output the output selection signal MS. The selector k ₂ is the output selection signal MS
Specified by. The minimum subtraction output in the addition output is selected and input to the signal processing circuit SC. The signal processing circuit SC carries out binarization processing with appropriate threshold values (THH, THL) and inputs the mask defect detection signal to the computer COM. Also in this case, as described with reference to FIG. 2B, instead of taking the difference between the gain-adjusted signals, the ratio may be obtained and used.

By the above operation, the dynamic range can be improved, and finer mask defects can be detected. Also in this embodiment, the computer CO
M controls the inspection operation. The computer COM also controls parallel movement of the mask R, scanning of the vibrating mirror 4, and the like. The computer COM maps the inspection result and displays it on the display D. The input device IF receives input of inspection content from the operator, such as inspection area and inspection sensitivity, and transfers it to the computer COM.

The inspection apparatus of this embodiment has a Fourier transform plane F
Each of I1 and FI3 has four photoelectric conversion elements. The area occupied by the photoelectric conversion element is not limited to this example. Also, the number of photoelectric conversion elements is not limited to four. Next, a sixth embodiment of the present invention will be described. FIG. 9 is a diagram showing an inspection apparatus according to the sixth embodiment of the present invention.

The illumination light IL emitted from the light source 1 is the lens group L.
A light spot that passes through 7, L8 and L9 and is focused on the mirror M near the rear focal plane FI2 of the objective lens L3 is formed. The illumination light IL is reflected toward the objective lens L3, and the objective lens L3
Is refracted by and becomes a parallel ray, and illuminates the mask R. The lenses L3, L4, L1 and L2 are arranged along the optical axis AX. The lenses L3 and L4 constitute the double diffractive optical system, and the objective lens L3 is arranged at a position approximately 1 times the focal length f3 from the object plane ob. Fourier transform plane on the back focal plane of objective lens L3
FI2 is formed. In addition, an image plane IM2 that is common to the object plane ob is formed on the rear focal plane of the lens L4. The lenses L1 and L2 also constitute a double diffractive optical system, the objective lens L1 is located at a distance about 1 times the focal length f1 from the object plane ob, and the Fourier transform plane FI is on the rear focal plane of the lens L1. ₁ is formed. lens
An image plane IM1 conjugate with the object plane ob is also formed on the rear focal plane of L2.

On the object plane ob, the mask R is held by the mask stage RST so that the circuit pattern drawing surface PA, which is the surface to be inspected of the mask R, is located. Illumination light IL
Illuminates the visual field of the objective lenses L3 and L1. When the drawing surface PA with the pattern of the mask R is covered with a uniform and reflective circuit pattern drawing material, the illumination light IL
Is reflected on the object-facing ob and traces the original optical path,
A Fourier transform pattern of the illumination light IL is formed on the Fourier transform plane FI2. Things the amplitude distribution of the light wave of illumination light IL on the face-to-face _{f 0 (x, y),} F 0 this Fourier transform pattern
When represented by (u, v), the Fourier transform pattern F ₀
All of (u, v) are reflected by the mirror M in the direction of the lens L3. That is, the mirror M plays the role of the 0th-order light blocking spatial filter S2 of the inspection apparatus of FIG. The light ray passes through the area Ff2 on the Fourier transform plane and reaches the lens L4. When there is no circuit pattern on the pattern drawing surface PA of the mask R and the amplitude distribution of the light wave of the illumination light IL is f ₀ (x, y), the Fourier transform pattern is formed on FI ₁ and F ₀
(U, v).

On the Fourier transform plane FI1, a 0th order light shielding filter S1 is installed. The filter S1 is F ₀ (u,
v) is blocked and rays in other spatial frequency region Ff1 are transmitted. The spatial frequency domains Ff1 and Ff2 are designed to transmit light rays in the same frequency domain. Spatial frequency domain
Rays 13 and 1 corresponding to the highest spatial frequency in Ff1 and Ff2
4, 15, 16 etc. all make the same angle with respect to the optical axis AX.

The reflectance of the circuit portion is α, the reflectance distribution is αf ₁ (x, y), the transmittance is β, and the transmittance distribution is βf.
₂ (x, y), the equations (1) to (16) are established, and the same images having different amplitudes are formed on the image planes IM1 and IM2. Therefore, the output IDD1 of the photoelectric conversion elements ID1 and ID2,
IDD2 is expressed by equations (17) and (18), the gains k and l of the amplifier are determined as in equation (19), and the third output IDD3 is (2
The mask defect signal can be detected by generating it by the equation (1) and binarizing it with an appropriate binarizing circuit (window comparator).

In the sixth embodiment, the photoelectric conversion element is a one-dimensional linear sensor, and optical scanning by a scanner mirror or AOD is unnecessary. The photoelectric conversion element may be a two-dimensional area sensor. The control of the inspection operation of the computer COM is the same as that of the other embodiments. In the first to sixth embodiments described above, in order to adjust the outputs from the photoelectric conversion elements ID1 and ID2, the gain of the amplifier that amplifies the two outputs is adjusted. The ND filter or the like may be provided in front of the photoelectric conversion elements ID1 and ID2 (on the mask side) so as to be able to be taken in and out from the optical path, and the two outputs may be adjusted.

Further, the present invention can be applied to all mask defect inspection apparatuses for laser scan type photomasks, and if the incidence is an oblique incidence illumination system, the arrangement of a light receiver causes a spatial filter or the like. It is possible to equivalently configure the same effect as the effect of blocking the scattering and diffracted light from the edge of the circuit pattern. In addition, the incident system of the first to sixth embodiments described above,
The oblique incidence system may be such that the illumination light IL is obliquely incident on the normal line of the mask R, and the optical axis of the light receiving system is set to the mask R.
You may incline and arrange with respect to the normal line. Further, both the incident system and the light receiving system may be tilted with respect to the mask R. Further, in the device configuration shown in FIG. 6, the arrangement of photoelectric conversion elements arranged on the Fourier transform plane is described in JP-A-4-1202.
No. 42, JP-A-5-165196, JP-A-6-43111 or JP-A-6-94633, the spatial angle inside the photoelectric conversion elements located at both ends (
The space inside the photoelectric conversion element on the Fourier transform plane) is 0
It may be almost equal to the angle of the second light (2θ ₀ : corresponding to the magnitude of the 0th light on the Fourier transform plane). As a result, the diffracted light from the circuit pattern does not enter at least one of the plurality of photoelectric conversion elements, and the accuracy of detecting mask defects is improved.

Next, the principle of the inspection method used in the first to sixth embodiments will be described. FIG. 10 shows the intensity distribution of the light wave when the mask is illuminated from the pattern side.
The light intensity is 30% of the incident light intensity only in the part with the circuit pattern. This is the case when the circuit pattern is two-layer chrome. FIG. 11 shows the intensity distribution of the light wave that illuminates the reticle and transmits. The reticle's unpatterned intensity transmission is 96.
Since it is%, such a light intensity distribution is obtained. 10 and 11, the vertical axis represents intensity and the horizontal axis represents position. The light intensity distribution of FIG. 10 is obtained by using the amplitude reflectance α and the function f ₁ (x) that becomes 1 in the portion with the pattern as shown in FIG.
It is shown by α · f ₁ (x) | ² . Similarly, the light intensity distribution of FIG. 11 has a width transmittance β and a function f ₂ (x) = 1-f as shown in FIG.
_{If 1} (x) is used, then | β · f | ² . 12 and 1
In 3, the vertical axis represents the amplitude and the horizontal axis represents the position.

[0066] Incidentally, f ₁ (x), as shown in FIG. 14 if the vector representation, the AC component f ₁ ac of f ₁ shown in FIG. 15 (x)
And the AC component f ₁ ac (x), which is the sum of the DC component f ₁ dc (x) of f ₁ (x) shown in Fig. 16, is hatched by the hatched area and points in the figure with the x axis as the boundary. This is a function in which the areas of the two areas and the area marked with are equal. At this time, h ₁ is automatically determined.

Similarly, when f ₂ (x) is displayed as a vector,
7, which is the AC component f ₂ ac (x) shown in FIG.
And the DC component f ₂ dc (x) shown in FIG. The AC component f ₂ is a function in which the areas of the two areas, the area shaded in the figure with the x axis as the boundary and the area hatched by points, are equal in area. At this time, h ₁ is automatically determined. As is clear from FIG. 18, f ₁ ac (x) = -f ₂ ac (x). Therefore, considering these intensities, │f ₁ ac (x) │ ² = ｜ f ₂ ac (x) ｜ ² (32), and the intensity of the AC component f ₁ ac (x) and the AC component f ₂ ac (x ) Is exactly the same as. As described above, obtaining the AC component from the object can be achieved by using the method shown in FIG.
Only the 0th-order diffracted light needs to be removed by using a double-diffraction optical system and a spatial filter.

Consider two image outputs, a transmission image output IDD1 and a reflection image output IDD2 obtained by the optical system of FIG. FIG. 20 is a diagram showing FIG. 10 in consideration of the intensity reflectance of 4% of the glass portion of the reticle. The vertical axis represents intensity and the horizontal axis represents position. And like this, two intensity distributions
As a case of (intensity distribution with reflectance of 30% and reflectance of 4%), the intensity reflectance distribution of a halftone reticle or the intensity transmittance distribution of transmitted light is also equivalent.

FIG. 21 is a vector representation of an object having two intensity distributions as shown in FIG. 20 with a function f ₁ (x) g ₁ (x) proportional to the amplitude. Here, it is assumed that the function f ₁ (x) has the intensity 1 with a pattern, and the function g ₁ (x) has the intensity of (4% / 30%) ^{1/2 in} the glass part. Function f ₁ (x) and function g ₁ (x)
There is a phase difference φ ₁ between and. The composite function f ₁₀ (x) of these two functions is given by the following equation (33).

F ₁₀ (x) = f ₁ (x) + g ₁ (x) (33) FIG. 22 is a rewritten version of f ₁₀ (x), which has the relationship of the following equation (34). f ₁₀ (x) = f ₁ (x) + g ₁ (x) = f ₁₁ (x) + g ₁₁ (x) = m '・ f ₁ (x) + g ₁₁ (x) (34) (34) From the equation, the coefficient m'is a constant that does not depend on the position x, and f ₁₀
It can be seen that (x) is obtained by adding the DC component g ₁₁ (x) to the coefficient of f ₁ (x).

[0071] f ₁₁ AC component f ₁₁ ac in (x) (x) is the AC component f ₁₀ ac of f ₁ (x) and the gain only difference f ₁₀ (x) as shown in FIG. 23 (x)
Matches. DC component f ₁₀ dc of f ₁₀ (x) (x) two straight min f ₁₁ dc (x) having a phase difference phi ₃ as shown in FIG. 24, a composite function of the g ₁₁ (x). The DC component f ₁₀ dc (x) is shown in FIG.
Therefore, the AC component of a function with two intensity values or two amplitude values can be treated in the same way as f ₁ (x).

In the present invention, for example, as shown in FIG. 9 in the above-mentioned embodiment, two sets of double diffraction gratings for transmission and reflection are arranged above and below the reticle respectively, and the Fourier transform of each of them is performed. A filter that removes the 0th-order diffracted light is placed on the surface. An image of the AC component of the transmitted light and the AC component of the reflected light can be obtained from the two sensors ID1 and ID2, respectively. These images correspond to so-called dark field images or Schlieren images. Consider a Cr pattern as a circuit pattern. Coherent illumination with a line pattern of 5 μm and a space of 5 μm in a one-dimensional repeating pattern with incident light of 488 nm
When (NA of objective lens = 1), the signal IDD2 corresponding to the reflected image is as shown in FIG. 26 (a), and the signal IDD1 corresponding to the transmitted image is as shown in FIG. 26 (b). These images correspond to [α ・ f ₁₀ ac (x] ² , [β ・ f ₂ ac (x)] ² but the finite spatial frequency (0 to 1 Modulation is seen because it is formed by a sine wave of (/ 488 Cycle / μm) .Since it is the calculation in the case of one dimension, NA of the objective lens, these images from the wavelength λ, * is the convolution,

[0073]

(Equation 3)

It is obtained by In the two-dimensional case, v (x) is different from Eq. (36), but in any case, the same modulation is applied to the reflected image and transmitted image. Since f ₁₀ ac (x) = m '・ f ₁ (x), k (α ・ m') ² = Lβ ² and two difference images are created, as shown in Fig. 26 (c). So it is completely zero. 26 (a), 26 (b), and 26 (c), the vertical axis represents intensity and the horizontal axis represents position. In the examples described below, f ₁₀ ac (x) = f ₁ (x) is set without considering the reflection of glass for the sake of simplicity.

FIG. 27 (a) shows a reflection image in the case of a stain having an intensity reflectance of 20% and an intensity transmittance of 80% and a size of 0.6 μm.
7 (b) is a transmission image, and FIG. 27 (c) is a difference image obtained by multiplying these by k and l and taking the difference. The vertical axis represents intensity and the horizontal axis represents position. The intensity distribution of the object has a sharp shape due to the convolution of the amplitude distribution of the point image in Eq. (36). In addition, the images in FIGS. 27A and 27B are completely the same. This is because unlike metal such as chrome, when the object (stain) is a dielectric, energy conservation is established with intensity transmittance + intensity reflectance = 100%. Chromium is absorbed and therefore has a strong transmittance + a strong reflectance ≈ 30%.

FIGS. 28 (a), 28 (b) and 28 (c) show a reflection image and a transmission image in the case of a stain having a size of 0.6 μm and an intensity reflectance of 10% and an intensity transmittance of 90%. And FIG.
(A) and (b) show a reflection image and a transmission image, respectively. Further, the difference image obtained by multiplying the above k and l is as shown in FIG.
The vertical axis represents intensity and the horizontal axis represents position. As described above, in the present invention, if the gains k and l are appropriately determined, the image of the circuit pattern in the difference image between the reflection and the transmission is completely removed without being affected by the diffraction of the lens, and only the image of the mask defect is obtained. It becomes possible to leave the image of the mask defect as a threshold THH.
And at least one of the threshold THL,
Mask defects can be detected. For actual mask defect detection,
As shown in FIGS. 26 (c), 27 (c), and 28 (c), two window comparators (two thresholds) THH, THL
Can be sufficiently close to the ground level (zero), and finer mask defects can be detected.

In principle, even a substance such as metal that absorbs light can be detected if the intensity ratio of the reflected image and the transmitted image is slightly different from that of the circuit pattern. Next, consider a case where the light receiving surface of the photoelectric conversion element is arranged on the Fourier transform surface. The Fourier transform is expressed by the following equation (38).

[0078]

(Equation 4)

[0079] More to the Fourier transform plane of the objective lens _{^{FT - {f 10 ac (x}} ) * v (x)} = m'F 1 ac (u) · V (u) (43) FT - {f 2 ac (x) * v (x)} =-F ₁ ac (u) ・ V (u) (44) From (43) and (44), the same spectrum (diffraction image) except the coefficient m ' It is observed on the Fourier transform plane of each of the objective lens for receiving light and the objective lens for receiving reflected light. Therefore, even if the corresponding spatial frequency region (Fourier diffraction image) is converted photoelectrically by transmission and reflection, the same result (action) as when receiving the transmitted and reflected light by the image sensor arranged on the image plane of the circuit pattern is obtained. can get.

Up to now, the so-called coherent illumination, which has a uniform amplitude distribution by a plane wave having a sufficiently large illumination area, has been described. Next, the amplitude distribution of the illumination light when the illumination light is condensed in the form of a beam spot having an aperture angle and is incident on the reticle will be described. 29A and 29B, FIG. 29A shows an amplitude reflectance distribution αf ₁ (x) and an illumination light amplitude distribution f ₀ . FIG. 29B shows the amplitude transmittance distribution β f ₂ (x) and the illumination light amplitude distribution f ₀ (x. F ₀ (x).
Is expressed by Eq. (45), and the Fourier transform of f ₀ (x) is F ₀ (u)

[0081]

(Equation 5)

Regarding the AC component

[0083]

(Equation 6)

Equations (46) and (47) and FIGS. 29C and 29D.
As is clearer, the spectrum of the 0th order diffracted light has a width ± u ₀ , but it is basically the same as in the case of plane wave illumination. That is, while frequency components of the two as (Fourier diffraction image component) AC component _{F '1 ac (u) =} - F' 2 ac) relation holds. These two functions F _'1 ac (u) and -F' ₂ same amplitude image even if Fourier transform (u) again is obtained. these
This is shown in equations (51) and (52).

[0085] _{^{FT - [F '1 ac (}} u)] = f' 1 ac (x) = f 0 (x) × f 1 ac (x) (51) FT - [F '2 ac (u)] = f ′ ₂ ac (x) = f ₀ (x) × f ₂ ac (x) = -f ₀ (x) × f ₁ ac (x) (52) As described above, in the present invention, the reflective illumination method and the transmission method are used. It is generated from a mask defect with random reflectance and transmittance by removing only the information based on the component of the ray from the circuit pattern having the existing reflectance and transmittance distribution in the mask from the ray obtained by the illumination method. The spatial frequency region (the angular range of the light beam that can pass on the Fourier transform plane) where the light receiving optical system focuses is determined so that only the information of the light beam can be extracted.
The present invention corresponds to the dark-field method or the schlieren method, as opposed to the conventional inspection apparatus corresponding to the bright-field method. In the present invention, by applying the Babinet principle and the method of zero-order light cut (for example, a spatial filter), the intensity ratio of the light wave traveling in the transmitting direction and the light wave traveling in the reflecting direction can be obtained without any mask defect. In some cases it was kept constant. Therefore, in principle, no pseudo defect is generated, and a mask defect can be easily detected.

That is, the optical system for receiving the reflected light has a spatial filter for blocking the 0th-order diffracted light component of the reflected light from the mask having the circuit pattern, and the optical system for receiving the transmitted light is the excessive 0th-order diffracted light. It has a spatial filter that shields the components. Therefore, the difference or ratio between the intensity of the reflected light from the mask having the circuit pattern and the intensity of the transmitted light becomes constant without depending on the shape and type of the circuit pattern. That is, the reflected light that is reflected by the circuit pattern where the mask defect is not attached and that passes through the spatial filter described above, and the transmitted light that passes through the glass portion of the mask where the mask defect is not attached and that passes through the spatial filter described above. The difference or ratio with is constant. On the other hand, the difference or ratio between the reflected light from the mask defect and the transmitted light transmitted through the mask defect and the glass portion of the mask is not constant.

At the time of actually inspecting a mask defect, the reflected light and the transmitted light which have passed through the respective spatial filters are received by the respective detectors, and the signal processing system receiving the signal from the detector determines the intensity of the reflected light. The difference or ratio with the intensity of the transmitted light is calculated, and the signal processing system discriminates the mask defect from the circuit pattern by comparing this difference with a predetermined threshold value (two window comparator levels).

With the above-described structure and principle, it is possible to detect a low-step mask defect or a semi-transparent mask defect. Also, the gain between the intensity of reflected light and the intensity of transmitted light can be adjusted electrically or optically, so that the intensity of reflected light incident on the detector and the intensity of transmitted light can be made almost equal. , The difference between the intensity of reflected light and the intensity of transmitted light can be made almost zero. Therefore, the SN ratio (ratio between the intensity of scattered light from the mask defect and the intensity of scattered light from the circuit pattern) is improved.

Next, the seventh embodiment of the present invention will be described with reference to FIG.
Explanation will be made using 0. The light source 111 is a mercury lamp, and the emitted light beam passes through the interference filter 112 and the light beam r0,
Becomes The ray r0 is refracted by the lens 113, and an analyzer 137 that transmits linearly polarized light with a plane of polarization perpendicular to the paper surface
And becomes a linearly polarized light r1 having a polarization plane perpendicular to the paper surface. The light ray r1 passes through the ring stop 114, passes through the relay lenses 115 and 116, and is reflected by the ring reflection mirror 117 toward the objective lens 118. The light ray r1 is refracted by the objective lens 118 and illuminates the circuit drawing surface 109 on the reticle 108 by the epi-illumination dark field illumination method.

The light source 101 is a mercury lamp, and the emitted light beam passes through the interference filter 102 and becomes a light beam t0. The light ray t0 is refracted by the lens 103, passes through the analyzer 136 which transmits the linearly polarized light having the plane of polarization parallel to the paper surface, and becomes the linearly polarized light t1 having the plane of polarization perpendicular to the paper surface. The light ray t1 passes through the ring stop 104, passes through relay lenses 105 and 106, and is reflected by the ring reflection mirror 107 toward the objective lens 110. The light ray t1 is refracted by the objective lens 110 and illuminates the circuit drawing surface 109 on the reticle 108 by the transmission dark field illumination method.

The pinhole 139 provided on the optical axis AX1 of the objective lens 118 is conjugate with the circuit drawing surface 109 of the reticle 108, and the image of the pinhole 139 is the lens 138, the one-dimensional deflecting means 121, and the objective lens 118. by that it could be projected on a circuit drawing surface 109 of the reticle 108. Pinhole projected
The image of 139 is converted by the one-dimensional deflecting means 121 into the objective lens 118.
It can be located at any point on a straight line parallel to the X direction in the field of view. That is, the image of the pinhole 139 is considered to be a point-like field, and this optical system enables one-dimensional field-of-view scanning with the point-like field.

The scattered lights r2 and t2 generated from one object point in the circuit drawing surface 109 within the field of view of the objective lens 118 illuminated by the rays r1 and t1 are one-dimensional scanning means.
Deflected by 121, lens 138, pinhole 139,
The light passes through the lens 140 and reaches the polarization beam splitter 122.
The one-dimensional scanning means 121 selects a position of one specific object point from a plurality of object points in the field of view of the objective lens 118, deflects scattered light r2, t2 generated from this object point, and outputs a polarized beam. Lead to splitter 122.

Linearly polarized light r which is scattered light generated from a point-like field conjugate with the pinhole 139 and having a plane of polarization perpendicular to the paper surface
2 is reflected by the polarization beam splitter 122, refracted by the lens 125, enters the photoelectric conversion element 126, and is photoelectrically converted. The scattered light generated from the point-like field conjugate with the pinhole 139 and the linearly polarized light t of the plane of polarization parallel to the paper surface
2 passes through the polarization beam splitter 122 and the lens 123
The light is refracted by and is incident on the photoelectric conversion element 124 to be photoelectrically converted.

The attenuator 127 adjusts the signal strength ratio of the two video signals input to the differential amplifier 128 according to the command of the computer 131. Photoelectric conversion element 126, photoelectric conversion element 124
The video signal Sr and the video signal St output by the attenuator 127 have two signal intensity ratios such that the output of the differential amplifier 128 becomes zero when there is no defect in the circuit drawing surface 109 of the reticle 108. The signals (Sr ', St') are set and input to the differential amplifier 128.

The differential amplifier 128 outputs an error signal proportional to the difference between the input signal Sr 'and the input signal St'. The error signal output from the differential amplifier 128 is input to the signal processing circuit 129. The signal processing circuit 129 has a window comparator circuit which is a binarization circuit having two slice levels on the plus side and the minus side. The signal processing circuit 129 outputs the intensity value of the error signal, the output of the binarization circuit indicating the presence or absence of a defect, and the like to the synchronizer 130.

The two slice levels on the plus side and the minus side of the window comparator circuit of the signal processing circuit 129 are set to a level at which a pseudo defect does not occur due to optical noise or electrical noise. The slice level can be set externally via the interface 132 and the computer 131. The synchronization device 130 controls the synchronization of the one-dimensional scanning means 121 and the XY stage 137 during the inspection. The one-dimensional scanning means 121 is driven via an actuator 135. The XY stage 137 is driven via the actuator 134.

By the one-dimensional scanning of the visual field in the X direction by the one-dimensional scanning means 121, it is possible to inspect a line-segment-shaped region whose longitudinal direction is the X direction and the scanning line whose length is limited by the size of the visual field of the objective lens 118. Is. When the reticle 108 is moved in the Y direction at an appropriate speed at a constant speed while repeating the visual field scanning in the X direction by the one-dimensional scanning means 121,
A plurality of line-segment-shaped regions (scanning lines) are arranged at regular intervals while maintaining proper overlap in a rectangular region having a size determined by the length of the scanning line in the X direction and the moving distance of the reticle in the Y direction. Since it is spread over the circuit drawing surface, it is possible to inspect the entire surface in this rectangular area.

The synchronizer 130 adds information (XY coordinates) indicating the existence position of the defect to information such as the intensity value of the error signal sent from the signal processing circuit 129 and the output of the binarizing circuit indicating the presence or absence of the defect. Send to computer 131. Computer 13
1 generates a map showing the position of the defect in the reticle and the intensity of the error signal at the defect position, and displays it on the display unit 133.

The computer 131 can automatically set up the gain of the attenuator 127. This setup uses a defect-free reticle or a defect-free reticle. Also, the gain of the attenuator depends on the type of reticle,
For example, the optimum value differs depending on the reticle having a chrome pattern, the reticle having a halftone pattern, and the like. If the optimum gain value for each type of reticle is measured once and stored in the memory in the computer 131, it is possible to easily set up using the data in the memory for subsequent inspections.

The external operator uses the interface 13
2 to the computer 131 through the computer 131, type of reticle to be inspected (for example, reticle of chrome pattern, reticle of halftone pattern, etc.), inspection mode, inspection sensitivity, inspection area, initial setting of equipment, inspection Enter the execution etc. In this embodiment, the 0th-order spectral component of the illumination light (regular reflection light and transmitted light that goes straight) travels in a direction different from the photoelectrically converted light rays t2 and r2 and is not photoelectrically converted. It goes without saying that it is not necessary to provide a spatial filter at the position.

FIG. 31 is a diagram showing an inspection apparatus according to the eighth embodiment of the present invention. The light source 1 is a light source that emits a laser. The light beam emitted from the light source 1 is expanded into parallel rays by the beam expander 202, reflected by the vibrating mirror 203, condensed by the scanning lens 204, and reticulated.
Illuminate the upper part of R with oblique incidence. The vibrating mirror 203 deflects the light beam by being driven by the one-dimensional scanning means SC and the driver SCA, and scan lines LC parallel to the X direction on the reticle.
Optical scanning is possible on R. The scattered light generated from each point on the scanning line LCR is composed of four objective lenses L101, L201 and LL.
1, LL2 collects them, and these pupil conjugate planes (PP plan
e1, PP plane2, PP plane1L, PP plane2L) 16 photoelectric conversion elements d1, d2, d whose photo-sensing surface is located above
3, d4, d5, d6, d7, d8, dL1, dL2,
photoelectrically converted by dL3, dL4, dL5, dL6, dL7, dL8, and 16 output signals D1, D2, D3,
D4, D5, D6, D7, D8, DL1, DL2, DL
3, DL4, DL5, DL6, DL7, DL8 are input to the signal processing circuit (see FIG. 32).

Four objective lenses L101, L201, LL
1, LL2 optical axes AX101, AX201, AX
1L and AX2L are provided at positions where a regular reflection light beam generated when an illuminating light beam is reflected by a reticle R and a transmitted light beam that goes straight through without being diffracted and scattered by the reticle R enters any of the four objective lenses. ing. Reticle R
Is mounted on the reticle stage RST, and the reticle stage RST can be moved in the Y direction by the reticle slider 205. The synchronizer TIM (see FIG. 35) performs synchronous control of the one-dimensional scanning means SC and the Y stage reticle slider during inspection. One-dimensional scanning means SC is actuator SCA
Is driven through. Reticle stage RST is driven via actuator ACT.

The scanning line LC, which is a line segment-shaped region whose longitudinal direction is in the X direction, whose length is limited by the size of the angle of view of the scanning lens 204 by one field scanning in the X direction by the one-dimensional scanning means SC
It is possible to illuminate R and inspect this area. When the reticle slider 205 moves the reticle R in the Y direction at a uniform speed at a constant speed while repeating the optical scanning in the X direction by the one-dimensional scanning means SC, the length of the scanning line LCR in the X direction becomes
A plurality of line segment areas (scan lines LCR) are placed in a rectangular area of a size determined by the reticle movement distance in the Y direction at regular intervals while maintaining proper overlap.
Since it is spread on the d surface, it is possible to inspect the entire surface in this rectangular area.

FIG. 32 is a block diagram of the control system of this embodiment. 16 output signals D1, D2, D3, D4, D
5, D6, D7, D8, DL1, DL2, DL3, DL
4, DL5, DL6, DL7, DL8 are signal processing circuits S
Input to P. The signal processing circuit SP performs a predetermined calculation from the input signal, and as a result, a signal indicating a true / false state with a defect.
Det and defect signal strength indicating the signal strength at the defect part ss out
Is output. These are input to the synchronization circuit TIM.

The synchronizing circuit TIM communicates with the actuator SCA of the one-dimensional scanning means and the actuator ACT of the reticle slider 205 to control the operations thereof. That is,
The synchronizing circuit TIM performs optical scanning in a predetermined area based on a command from the computer COM, and inspects defects in this area. The synchronizing circuit TIM is based on the signal det sent from the signal processing circuit SP and indicating the presence or absence of a defect, and in the information of the defect signal strength ss out indicating the signal strength at the defective portion, the information indicating the position where the defect exists ( Add (XY coordinates) and send to computer COM.

The computer COM generates a map showing the position of the defect in the reticle and the defect signal strength ss out at the defect position, and displays it on the display DIS. The computer COM communicates with the signal processing circuit SP so that the gain in the signal processing circuit SP can be automatically set up. This setup uses a defect-free reticle or a defect-free reticle. The optimum value of the gain in the signal processing circuit SP varies depending on the type of reticle, for example, a chrome pattern reticle or a halftone pattern reticle. Once the optimum gain value for each type of reticle is measured and stored in the memory in the computer COM, it is possible to easily set up using the data in the memory for subsequent inspections.

The external operator uses the interface IF
Input the type of reticle to be inspected (chrome pattern reticle, halftone pattern reticle, etc.), inspection sensitivity, inspection area, device initial setting execution, inspection execution, etc. to the computer COM via . FIG. 33 shows a first example of the signal processing circuit SP of this embodiment. Eight pairs of photoelectric conversion signals (D1, DL1), (D2, DL
2), (D3, DL3), (D4, DL4), (D5,
DL5), (D6, DL6), (D7, DL7), (D
8, DL8) are attenuators K1, K2, K3, K4, K5, K6,
Input to K7, K8, and by the command of computer COM,
It is adjusted to a predetermined signal strength ratio. Attenuator K1, K2, K3, K
The attenuation rates of 4, K5, K6, K7, and K8 can be arbitrarily set by the computer COM via the attenuation rate setting device K0. Attenuator K1,
The outputs of K2, K3, K4, K5, K6, K7 and K8 are differential amplifiers Sa1 and S
Input to a2, Sa3, Sa4, Sa5, Sa6, Sa7, Sa8, these differential amplifiers output a differential output, which is a signal proportional to the difference between the input signals.

These differential outputs correspond to the corresponding comparators.
C1, C2, C3, C4, C5, C6, C7, and C8 are compared with the reference level of the reference level generator Vref to be a binarized signal indicating true / false and input to the AND circuit A. The eight comparators output a true value (high level) when the differential output of the differential amplifier exceeds the reference level of the reference level generator Vref, and outputs a false value (low level) otherwise. The AND circuit A outputs a true value only when all eight inputs are true values, and unexpectedly outputs a false value.

The reference level of the reference level generator Vref can be set to an arbitrary level according to a command from the computer COM, and the defect detection sensitivity can be changed by changing this level. On the other hand, the eight differential outputs are also input to the signal selector Ss. The signal selector Ss outputs the value selected from the values such as the minimum value, the maximum value and the average value among the input signals by the operator through the computer COM in advance.
Output as ss out. The minimum output is usually sufficient. This is because the minimum value has the least background noise and faithfully reflects the intensity of scattered light from the defect. These signals are updated at a rate sufficient to change the strength of the input signal and output real-time values.

FIG. 34 shows a second example of the signal processing circuit SP of this embodiment. A pair of eight photoelectric conversion signals (D1, D
L1), (D2, DL2), (D3, DL3), (D
4, DL4), (D5, DL5), (D6, DL6),
(D7, DL7) and (D8, DL8) are attenuators K, respectively.
Input to 1, K2, K3, K4, K5, K6, K7, K8, adjusted to a predetermined signal strength ratio by a command from the computer COM, and input to the differential amplifier.

Attenuators K1, K2, K3, K4, K5, K6, K7, K8
The output of the differential amplifier is Sa1, Sa2, Sa3, Sa4, Sa5, Sa6,
Input to Sa7 and Sa8, these differential amplifiers output a differential output, which is a signal proportional to the difference between the input signals. It is input to the signal selector Ss, and the signal selector Ss outputs the minimum value of the input signals as the output signal ss out. output
The signal ss out is updated at a rate sufficient to change the strength of the input signal and outputs a real-time value.

The output signal ss out is compared with the reference level of the reference level generator Vref by the comparator C1 to become a binarized signal indicating true / false, and is output as the true / false signal det. The comparator C1 outputs a true value (high level) when the output signal ss out exceeds the reference level of the reference level generator Vref, and outputs a false value (low level) otherwise.

The reference level of the reference level generator Vref can be set to an arbitrary level by a command from the computer COM, and the defect detection sensitivity can be changed by changing this level. On the other hand, the output signal ss out is also output to the synchronization circuit TIM. FIG. 35 shows a third example of the signal processing circuit SP of this embodiment.
This is an example. A pair of eight photoelectric conversion signals (D1, DL
1), (D2, DL2), (D3, DL3), (D4,
DL4), (D5, DL5), (D6, DL6), (D
7, DL7) and (D8, DL8) are attenuators K1 and K, respectively.
It is input to 2, K3, K4, K5, K6, K7, K8, and is adjusted to a predetermined signal strength ratio by the command of the computer COM.

Attenuators K1, K2, K3, K4, K5, K6, K7, K8
Output is differential amplifier Sa1, Sa2, Sa3, Sa4, Sa5, Sa6, Sa
7, is inputted to Sa8, these differential amplifiers are signal proportional to the difference between the input signal and outputs a differential output. These differential outputs are the corresponding comparators C1, C2, C3, C4, C5,
C6, C7, and C8 compare with the reference level of the reference level generator Vref to form a binary signal indicating true / false, and the outputs of the comparators C1, C2, C3, and C4 are input to the AND circuit A1. The outputs of the comparators C5, C6, C7, C8 are input to the AND circuit A2. The outputs of the AND circuits A1 and A2 are input to the OR circuit O. The OR circuit O outputs an output signal det indicating the presence / absence of a defect.

The eight comparators output a true value (high level) when the differential output of the differential amplifier exceeds the reference level of the reference level generator Vref, and outputs a false value (low level) otherwise. The AND circuits A1 and A2 output the true value only when all four inputs are true values, and unexpectedly output the false value. The OR circuit O outputs a true value if either one of the two inputs has a true value. The purpose of such a logic is that even if the scattering directivity is significantly biased depending on the shape of the defect to be detected, among the plurality of light receiving means including the objective lenses arranged in multiple directions, the received light intensity of scattered light is large. This is because if there is at least one large light receiving means, it can be considered as a defect to detect defects of any shape.

The reference level of the reference level generator Vref can be set to an arbitrary level according to a command from the computer COM, and the defect detection sensitivity can be changed by changing this level. On the other hand, four differential amplifiers Sa1, Sa2, Sa3, S
The differential output of a4, is also input to the signal selector Ss1. The differential outputs of the four differential amplifiers Sa5, Sa6, Sa7, Sa8 are also input to the signal selector Ss2. These signal selectors Ss1 and Ss2
Outputs a value selected in advance by the operator through the computer COM from the values such as the minimum value, the maximum value and the average value among the input signals.

For the signals output from these signal selectors Ss1 and Ss2, it is usually sufficient to output the minimum value. This is because the minimum value has the least background noise and faithfully reflects the intensity of scattered light from the defect. These signals are updated at a rate sufficient to change the strength of the input signal and output real-time values. The signal selector Ss3 outputs, as an output signal ssout, a value selected in advance by the operator through the computer COM from the values such as the minimum value, the maximum value and the average value among the input signals.

As for the signal output from the signal selector Ss3, it is usually sufficient to output the maximum value. Because the scattering directivity is significantly biased depending on the shape of the defect to be detected,
Of the light receiving means including the objective lenses arranged in multiple directions,
This is because the intensity of scattered light from a defect is faithfully reflected by giving priority to a signal having a large received intensity of scattered light. These signals are updated at a rate sufficient to change the strength of the input signal and output real-time values as the output signal ss out.

FIG. 36 shows the first embodiment of the signal processing circuit SP of this embodiment.
Here is an example of 4. A pair of eight photoelectric conversion signals (D1, DL
1), (D2, DL2), (D3, DL3), (D4,
DL4), (D5, DL5), (D6, DL6), (D
7, DL7) and (D8, DL8) are attenuators K1 and K, respectively.
It is input to 2, K3, K4, K5, K6, K7, K8, and is adjusted to a predetermined signal strength ratio by the command of the computer COM.

Attenuators K1, K2, K3, K4, K5, K6, K7, K8
The output of the differential amplifier is Sa1, Sa2, Sa3, Sa4, Sa5, Sa6,
Input to Sa7 and Sa8, these differential amplifiers output a differential output, which is a signal proportional to the difference between the input signals. The differential outputs of the differential amplifiers Sa1, Sa2, Sa3, Sa4 are input to the signal selector Ss1, and the differential outputs of the differential amplifiers Sa5, Sa6, Sa7, Sa8 are input to the signal selector Ss2 to select two signals. Bowl Ss1, Ss
Each 2 outputs the minimum value. This is because by outputting the minimum value, the background noise is minimized and the intensity of scattered light from the defect is faithfully reflected. The two signal selectors Ss1 and Ss2 are updated at a speed sufficient to change the intensity of the input signal and output a real-time value.

The outputs of the two signal selectors Ss1 and Ss2 are input to the signal selector Ss3. The signal output from the signal selector Ss3 is usually sufficient to output the maximum value of the two signals. This is because the scattering directivity is significantly biased depending on the shape of the defect to be detected. This is because the intensity of the scattered light from the defect is reflected in. These signals are updated at a rate sufficient to change the strength of the input signal and output real-time values as the output signal ss out.

The output signal ss out is compared with the reference level of the reference level generator Vref by the comparator C1 to become a binarized signal indicating true / false, and is output as the true / false signal det. The comparator C1 outputs a true value (high level) when the output signal ss out exceeds the reference level of the reference level generator Vref, and outputs a false value (low level) otherwise.

The reference level of the reference level generator Vref can be set to an arbitrary level by a command from the computer COM (not shown), and by changing this level, the defect detection sensitivity can be changed. On the other hand, the output signal ss out is also output to the synchronization circuit TIM. In this embodiment, the 0th-order spectral component of the illumination light (regular reflection light and transmitted light that goes straight) travels in a direction different from the photoelectrically converted light rays t2 and r2 and is not photoelectrically converted. It goes without saying that it is not necessary to provide a spatial filter at the position.

FIG. 37 is a perspective view of this embodiment. FIG.
According to FIG. 31, it is easy to understand the positional relationship of the light detection surfaces of the 16 photoelectric conversion elements arranged on the four pupil conjugate planes of FIG. Further, it shows how incident light optically scans the scanning line LCR on the circuit pattern surface. FIG. 38 is an explanatory view of what relationship the spectra observable on the four pupil conjugate planes have on the Fourier transform plane, and is a bird's-eye view from the same viewpoint as FIG. 37, but those unnecessary for the explanation are omitted. There is.

Consider a virtual sphere in which the circuit pattern surface of the reticle is located on the XY plane and is sufficiently larger than the laser spot formed by the illumination light. The equatorial plane of the sphere coincides with the circuit pattern plane. The position of the laser spot coincides with the center of the sphere. It is known that the direction of scattered light generated from the position of the laser spot can be represented by a vector equal to the radius of a sphere, and a Fourier spectrum can be obtained by orthogonally projecting the intersection point of the vector and the spherical surface onto a plane parallel to the equatorial plane (for example, JP-A-5-165196, JP-A-6-43111, and JP-A-6-94633). The coordinates of each orthographically projected point are usually represented by orthogonal coordinates parallel to the XY coordinates with the origin of the 0th-order spectrum.

Assuming that the size of the sphere is 1, the value obtained by subtracting the direction cosine of the 0th-order light from the direction cosine of the arbitrary vector from the laser spot with respect to the X and Y axes is the coordinate on the Fourier transform plane. . At this time, the unit of the Fourier transform plane is the spatial frequency (cycle), and dividing this by the incident wavelength indicates the frequency per unit length of the object (eg cycle / micron),
Intuitive and easy to understand.

Among the vectors from the position of the laser spot, the intersection point on the + Z hemisphere side is the first Fourier transform plane FT plane1.
Shown as orthographic projection above. The intersection point on the -Z hemisphere side of the vector is orthographically projected on the second Fourier transform plane FT plane 1L.
If the objective lenses L1 and LL1 have the same NA, the vectors r and rL indicating the directions of the optical axes AX101 and AX1L on which they are installed are indicated by the points r and rL on the spherical surface, and on the two Fourier transform planes. Then, the points are r1 and rL1. The optical axes AX101 and AX1L are determined so that the values of these coordinates match in the two orthogonal coordinate systems, the UV coordinates, and the UL-VL coordinates. In other words, the vectors r and rL are arranged so that their orthogonal projections on the equatorial plane are the same. With this, the transmission direction,
It is possible to receive the spatial spectrum of scattered light that is simultaneously generated in the reflection direction in a target range.

The same applies to the objective lenses L201 and LL2. If the NAs are the same, the vectors s and sL indicating the directions of the optical axes AX201 and AX2L on which they are installed are points s and sL on the spherical surface. The points are s1 and sL1 on the two Fourier transform planes. The optical axes AX201 and AX2L are determined so that the values of these coordinates match in the two orthogonal coordinate systems, the UV coordinate, and the UL-VL coordinate. In other words, the vectors s and sL are placed so that their orthographic projections on the equatorial plane match.
In this way, it is possible to add an unlimited number of optical systems to target positions across the equatorial plane.

Objective lenses L101, L201, LL1 and LL
2 pupil conjugate plane PP plane1, PP plane2, PP plane1L, PP p
Consider the relationship between lane2L and the Fourier transform planes FT plane1 and FT plane1L. The vector q is refracted by the objective lens L1 and becomes a point q2 on the pupil conjugate plane PP plane1. Also, this vector becomes a point q1 on the Fourier transform plane FT plane1, and any point on the pupil conjugate plane PP plane1 is a Fourier transform plane FT plan.
It can be seen that there is a one-to-one correspondence with one point on e1. Therefore, with the point r1 on the Fourier transform plane FT plane1 as the origin, U-
Considering the orthogonal coordinate u'-v 'parallel to the V coordinate, this is U'- with the intersection point r2 with the optical axis AX1 as the origin on the pupil conjugate plane PP plane1.
It becomes the V'coordinate. Strictly speaking, the U′-V ′ coordinates are not orthogonal coordinates, and each axis is not a straight line, but if the NA of the objective lens is approximately 0.1 or less, each axis can be considered to be a straight line.

Similarly, the point s1 on the Fourier transform plane FT plane1
Considering the orthogonal coordinate u "-v" parallel to the UV coordinate with the origin as the origin, this is the intersection point s2 with the optical axis AX1 on the pupil conjugate plane PP plane2.
It becomes the U "-V" coordinate with the origin as. Similarly, the Fourier transform plane F
Considering the orthogonal coordinates uL'-vL 'parallel to the UL-VL coordinates with the origin of the point rL1 on the T plane1L, this is the pupil conjugate plane PP plane1.
At L, UL'-VL 'coordinates are set with the intersection point rL2 with the optical axis AX1L as the origin.

Similarly, the point sL on the Fourier transform plane FT plane1L
Considering the orthogonal coordinate uL "-vL" parallel to the UL-VL coordinate with 1 being the origin, this is the UL "-VL" coordinate having the origin at the intersection point sL2 with the optical axis AX2L on the pupil conjugate plane PP plane2L. Also, the vector
A vector qL that is a target across q and the equatorial plane, that is, a vector qL in which the sign of the Y component of the direction cosine is inverted forms an intersection qL with the spherical surface, and the point qL2 and the Fourier transform plane FT plane1L on the pupil conjugate plane PP plane1L.
Above is the point qL1.

The vector p indicating the incident light, the vector t indicating the specularly reflected light rays, and the vector w indicating the straight transmitted light rays are points p ′, t ′, w ′ when orthographically projected on the equatorial plane of The points are p1, t1, and wL1 on the Fourier transform plane. Point t ', w'
Coincide with each other, and points t1 and wL1 indicate the zero-order spectrum, which is the origin of two Cartesian coordinates. FIG. 39 is a bird's-eye view showing the spatial frequency region of the light received by the light detection surface of the photoelectric conversion element. In this embodiment, two pairs of objective lenses (L1, LL1), which are plane-symmetrically arranged with respect to the XY plane in the + Z and -Z directions, are used.
(L2, LL2) exists. These pupil conjugate planes PPplane1, PP
There are four light detection surfaces on each of plane2, PP plane1L, and PP plane2L. (For example, on the pupil conjugate plane PP plane1, the light detection planes d1, d2, d3, d4) These light detection planes are each one set of pupil conjugate planes (for example, one of the symmetrically arranged objective lenses). They are arranged symmetrically with respect to the XY plane on the pupil conjugate planes PP plane1, PP plane1L) and form a pair of four light detection planes.

The shape and arrangement of the two photo-sensing surfaces on the pupil conjugate plane, which are paired with each other, are the spatial frequency domain (Fourier
Spectral domain) is Fourier transform plane FT plane1, FT plan
It is adapted to match on the e1L. As described in FIG. 38, from the bird's-eye view, these spatial frequency regions (Fourier / spectral regions) can be easily transformed into the Fourier transform plane FT.
Drawing can be easily performed on the plane 1 and the FT plane 1L by the method using the above-mentioned orthogonal projection, and the shape and position of the light detection surface on the pupil conjugate plane may be determined so that these positions and shapes are congruent.

For example, since the light detection surface d1 is on the + Z side, its Fourier spectrum is dd1 observed on the Fourier transform surface FT plane1. The region congruent to this is the Fourier transform plane F
This is ddL1 when drawn on the T plane1L, and the shape of the light detection surface dL1 on the pupil conjugate plane PPplane1L of the objective lens L101 and the objective lens LL1 arranged as the target is determined so as to receive this Fourier spectrum region. To do.

It should be noted that all the photo-sensing surfaces are strips each having a longitudinal direction in the U and UL directions on each Fourier transform surface, and are arranged at equal intervals. It is optimal for filtering the Fourier spectrum of a three-dimensional periodic pattern, and a detailed design method is disclosed in, for example, Japanese Patent Laid-Open Nos. 5-165196 and 6-94633. In the above-mentioned embodiment, a plurality of photoelectric conversion elements are used for one objective lens. This is because the photoelectric conversion elements are electrically driven by the relatively high intensity diffracted light of the circuit pattern entering the elements. Once saturated, a linear differential output cannot be obtained, and the upper limit of the dynamic range is determined. Since the lower limit of the dynamic range is of course determined by electrical and optical noise, it is necessary to expand the upper limit in order to obtain a larger dynamic range.
A plurality of photoelectric conversion elements are used for one objective lens.

The lower limit of the dynamic range can be set, of course, by reducing the beam spot size of the illuminating light beam, but the beam spot size of the illuminating light beam is a strong function such as the inspection time and cannot be easily reduced. However, even if one photoelectric conversion element is used for each objective lens, a practical performance inspection device can be configured.

FIG. 40 shows the ninth embodiment of the present invention, and the eighth embodiment.
This is an example in which the number of photoelectric conversion elements of the example is reduced and one photoelectric conversion element is provided for each objective lens. Members similar to those in FIG. 31 are described with the same reference numerals, and detailed description thereof will be omitted. The objective lens is arranged at a position where the direct reflected light and the direct transmitted light do not enter, and for example, the objective lens is arranged in a plane-symmetrical position with the pattern drawing surface of the reticle R 1 interposed therebetween.

The optimum signal processing circuit for this embodiment is shown in FIG. This figure is a modification of FIG. AND logic is not necessary, and the presence or absence of foreign matter is determined only by OR logic O. That is, when the output of either of the differential amplifiers Sa1 and Sa2 exceeds the reference level set by the computer COM (see FIG. 32), it is determined that there is a foreign substance. The signal selection circuit S is a maximum value selection circuit that always selects the maximum signal. Further, the reticle transport system and the control system may be those shown in FIG. 32 similar to the eighth embodiment.

FIG. 42 shows the tenth embodiment of the present invention, which is the ninth embodiment.
It is an example in which the light receiving optical system of the embodiment is configured by a cylindrical lens. The same members as those in FIG. 41 are described with the same reference numerals, and the details thereof will be omitted. In Fig. 42, a pair of cylindrical lenses (SLN1, SLN1L, (SLN2, SLN2L)
39 is arranged at a position where the directly reflected light (regularly reflected light) and the directly transmitted light do not enter, and is installed at a position symmetrical with respect to the pattern drawing surface of the reticle R as in the second embodiment.
As described with reference to, the design is made so that the Fourier spectra of the light rays received by both agree with each other. The light rays condensed by these cylindrical lenses are the light incident ends d1, d2, d1L, and fh2L of the optical fiber assemblies fh1, fh2, fh1L, and fh2L.
Incident from d2L, optical fiber part fb1, fb2, fb1L, fb2L
Depending on the photoelectric conversion element ph1, such as photomultiplier
Photoelectric conversion is performed by ph2, ph1L, ph2L. The signal processing circuit may be the same as that of the ninth embodiment, and the reticle transport system and control system may be the same as that of the eighth embodiment shown in FIG. FIG.
FIG. 43 is an explanatory diagram of a control system viewed from the + X direction in FIG. 42.

In the above-described embodiment, the entire surface of the reticle can be inspected by the one-dimensional scanning means for incident light (for example, vibrating mirror and objective lens) and the one-dimensional scanning means for reticle (for example, reticle stage drive system). Alternatively, the entire surface of the reticle may be inspected by two-dimensional scanning of incident light. In the above embodiment, it is important to observe the image of the detected defect after the inspection,
Or, it is desired to grasp the degree concretely. Here, the optical system according to the above-mentioned embodiment is used for the object of the dark field illumination.
A three-dimensional optical image can be obtained by the coordinated operation of laser scanning and stage movement. In addition, if the laser scanning optical system is configured to perform two-dimensional scanning in advance, inspection can be performed without moving the stage. The two-dimensional scanning can be performed by combining a conventionally known polygon mirror and the like. At this time, if the defect is displayed on the display by using the output of the differential amplifier, that is, the output of Ss out shown in FIG. 34 or the luminance signal proportional to this, the circuit pattern is not imaged. Only the detected foreign matter can be observed with high contrast, and the size of the foreign matter can be easily grasped.

Further, in the embodiment shown in FIG. 31, the scattered light from the vertical portion of the pattern edge may become optical noise, and in order to eliminate this, as shown in FIGS. It is preferable to provide analyzers An1-An4 on one or both of the optical paths of the reflected light so as to remove optical noise. This is because the polarized light of the scattered light from the vertical portion of the pattern edge has a different polarized light from that of the flat end face portion (plane parallel to the mask). As a result, it is possible to eliminate the optical noise and ideally erase the circuit pattern image on the mask. Incidentally, as shown in FIGS. 44 to 46, it is preferable that the transmission axes of the upper and lower analyzers sandwiching the mask are symmetrical with respect to the upper and lower sides.

Further, such an analyzer is shown in FIG.
The present invention is not limited to the embodiment shown in FIGS. 44 to 46, and it is preferable to apply it to other embodiments. In addition, regarding the incident direction of the irradiation light, as shown in FIG. 47 (a), when irradiation is performed from the side where the circuit pattern is formed, the shaded portion of the circuit pattern becomes a glass portion due to the vertical portion of the circuit pattern edge. The intensity ratio of the formed light wave traveling in the transmitting direction and the light wave traveling in the reflecting direction may change from an ideal value.
In order to eliminate this, as shown in FIG. 47 (b), it is preferable to make the light incident from the side opposite to the circuit pattern forming surface of the mask.

[0143]

As described above, according to the present invention, it is possible to detect a foreign material having a low step or a semitransparent foreign material. Also,
A highly sensitive foreign matter inspection device can be configured without depending on advanced image processing technology. Further, the present invention is suitable when the circuit pattern on the mask has a distribution function of one reflectance and one transmittance. Therefore, as a circuit drawing material, in addition to a light shielding member such as chrome, a light called a halftone is used. There is no problem with transparent materials. That is , according to the present invention, it is possible to deal with a conventional reticle having a circuit pattern made of a chrome light-shielding film and a halftone reticle having a circuit pattern drawn only by a phase shifter made of a light-transmitting thin film.

Further, the beam spot size can be enlarged and the inspection time can be shortened. In addition, the degree of freedom in the arrangement of the light receiving optical system is increased as compared with the conventional method of blocking the scattered light and the diffracted light from the circuit pattern only with the spatial filter, and the light receiving optical system is provided near the 0th spatial frequency spectrum, for example. be able to. Therefore, foreign matter that generates scattered light only in the vicinity of the 0th spatial frequency spectrum, which was difficult to detect in the past,
For example, it is possible to provide an inspection apparatus which is a flat foreign matter, has a gently thin thickness at the contour portion of the foreign matter, and is capable of simultaneously inspecting the presence / absence of foreign matter having an unclear contour.

[Brief description of the drawings]

FIG. 1 is a diagram showing an outline of a mask defect inspection apparatus suitable for a first embodiment of the present invention.

2 (a) and 2 (b) are diagrams showing a signal processing system of the apparatus of FIG.

3 (a) and 3 (b) are diagrams showing spatial filters S1 and S2 used in the apparatus of FIG.

4A and 4B are diagrams showing a spatial filter according to a second embodiment of the present invention.

FIG. 5 is a diagram illustrating a third embodiment of the present invention.

FIG. 6 is a diagram showing an outline of a mask defect inspection apparatus suitable for a fifth embodiment of the present invention.

7 is a diagram showing a photoelectric conversion element used in the device of FIG.

8 is a diagram showing a signal processing system used in the apparatus of FIG.

FIG. 9 is a diagram illustrating a sixth embodiment of the present invention.

FIG.

FIG. 11:

FIG.

FIG.

FIG.

FIG.

FIG.

FIG.

FIG.

FIG. 19:

FIG.

FIG. 21:

FIG. 22:

FIG. 23:

FIG. 24:

FIG. 25:

FIG. 26:

FIG. 27:

FIG. 28:

FIG. 29 is a diagram for explaining the inspection principle in the embodiment according to the present invention.

FIG. 30 is a diagram for explaining the seventh embodiment of the present invention.

FIG. 31 is a diagram illustrating an eighth embodiment of the present invention.

32 is a diagram showing a signal processing system of the apparatus of FIG. 34.

FIG. 33,

FIG. 34,

FIG. 35,

FIG. 36 is a diagram for explaining the details of the signal processing circuit SP in FIG. 31.

37 is a perspective view of the device of FIG. 31. FIG.

38 is a diagram showing a relationship on a Fourier transform plane of a spectrum observable by the device of FIG. 31.

39 is a diagram for explaining the spatial frequency received by the light detection surface of the photoelectric conversion element of the apparatus of FIG. 31.

FIG. 40 is a diagram for explaining the ninth embodiment of the present invention.

41 is a diagram illustrating the signal processing system of FIG. 40. FIG.

FIG. 42 is a diagram for explaining the tenth embodiment of the present invention.

43 is a diagram illustrating a control system of the device in FIG. 42.

FIG. 44:

FIG. 45:

FIG. 46 is a diagram illustrating a modified example of the eighth embodiment.

47 (a) and 47 (b) are views for explaining the improvement principle by the incident direction of light with respect to the mask.

FIG. 48 is a configuration diagram of a conventional mask defect inspection apparatus.

[Explanation of symbols]

1 light source R mask 4, M mirror L1-L9 lens ID1, ID2, 124, 125, D1-D8, DL1-DL8 photoelectric conversion element S1, S2 spatial filter

Claims (34)

[Claims] 1. A defect inspection apparatus for a mask which optically detects a defect on a mask having a circuit pattern; an illumination system for illuminating the mask with inspection light; and receiving the inspection light reflected by the mask. A first light receiving optical system;
A second light receiving optical system for receiving the inspection light transmitted through the mask; and an optical axis of the first light receiving optical system in an optical Fourier transform plane for the circuit pattern in the first light receiving optical system. A first spatial filter that shields the inspection light passing through a central region including: a light of the second light receiving optical system in an optical Fourier transform plane with respect to the circuit pattern in the second light receiving optical system. A second spatial filter that blocks the inspection light passing through a central region including an axis; and the first spatial filter.
First photoelectric conversion means for photoelectrically converting the inspection light that has passed through the spatial filter; second photoelectric conversion means for photoelectrically converting the inspection light that has passed through the second spatial filter; and the first photoelectric conversion means. Gain adjustment for adjusting the gain of the first output signal from the converting means and outputting as the third output signal, adjusting the gain of the second output signal from the second photoelectric converting means, and outputting as the fourth output signal. A circuit for detecting the defect based on either a relative intensity difference or intensity ratio between the third output signal and the fourth output signal whose gain is adjusted by the gain adjusting circuit. Defect inspection equipment. 2. A gain for optically adjusting a relative gain of the first output signal and the second output signal in at least one of the first light receiving optical system and the second light receiving optical system. 2. The apparatus according to claim 1, further comprising an optical system, wherein the signal processing system detects the defect based on the first output signal and the second output signal after adjusting the gain. . 3. The gain setting of the gain adjusting circuit is the third setting.
The apparatus according to claim 1, wherein the gain is adjusted such that the intensity difference of the fourth output signal is substantially zero or the intensity ratio is substantially one. 4. In the mask defect inspection, when forming an image for observing a detected defect, the same optical system as in the defect inspection is used, and the intensity difference or intensity ratio of the third and fourth output signals. The device according to claim 1, comprising a display device capable of displaying an image of a luminance signal proportional to 5. The apparatus according to claim 1, wherein an analyzer for transmitting polarized light of a predetermined plane of polarization is provided near at least one of the first and second spatial filters. 6. The device according to claim 1, wherein the inspection light is optically arranged so as to enter from the circuit drawing side of the mask. 7. The apparatus according to claim 1, wherein the inspection light is optically arranged so as to enter from a side opposite to a circuit drawing surface of the mask. 8. A defect of a mask having a circuit pattern on a first surface of a light-transmissive plate-like substrate and not having a circuit pattern on a second surface opposite to the first surface is inspected. An illumination device for illuminating an inspection area, which is an area in the first surface to be an inspection target of a mask; and a space including the mask divided into two parts by a plane including the first surface 2 Two light receiving means separately disposed in one space, the second light receiving means disposed in a second space including the second surface, and the first light receiving means not including the second surface. A first light receiving means arranged in space; a first photoelectric conversion means for photoelectrically converting a light ray generated from within the inspection region of the mask and incident on the first light receiving means; the mask Of the light rays that are generated from within the inspection area of and that are incident on the second light receiving means. A second photoelectric conversion means for photoelectric conversion; the first
The first output signal output from the photoelectric conversion means is adjusted in gain and output as a third output signal, and the second output signal output from the second photoelectric conversion means is adjusted in gain and output as a fourth output signal. A gain adjusting circuit for detecting the defect based on either a relative intensity difference or intensity ratio between the third output signal and the fourth output signal whose gains are adjusted by the gain adjusting circuit. Mask inspection device. 9. The apparatus according to claim 8, wherein the first light receiving unit and the second light receiving unit include an objective lens having a visual field in the inspection region. 10. The apparatus according to claim 8, further comprising a spatial filter near the objective lens. 11. The first photoelectric conversion means and the second photoelectric conversion means.
9. The apparatus according to claim 8, wherein the photoelectric conversion means includes an image pickup element positioned in an image plane conjugate with the inspection area. 12. The first photoelectric conversion means and the second photoelectric conversion means.
9. The apparatus according to claim 8, wherein the photoelectric conversion means is arranged near the pupil conjugate plane of the objective lens. 13. The first and second photoelectric conversion means,
9. The device according to claim 8, comprising a plurality of photoelectric conversion surfaces that output a plurality of independent photoelectric conversion signals. 14. The apparatus according to claim 8, wherein the illumination unit has an optical scanning mechanism. 15. The apparatus according to claim 8, wherein the gain setting of the gain adjusting circuit adjusts the gain so that the intensity difference between the third and fourth output signals is substantially zero or the intensity ratio is substantially one. . 16. A mask defect inspection method uses the same optical system as that used for defect inspection when imaging for observing a detected defect, and the intensity difference or intensity ratio of the third and fourth output signals. It is possible to display an image of a luminance signal proportional to
The device according to claim 8, comprising a display device. 17. The apparatus according to claim 8, wherein an analyzer is provided near at least one of the first and second spatial filters for transmitting polarized light having a predetermined plane of polarization. 18. The apparatus according to claim 8, wherein the inspection light is optically arranged so as to enter from the circuit drawing side of the mask. 19. The apparatus according to claim 8, wherein the inspection light is optically arranged so as to enter from a side opposite to a circuit drawing surface of the mask. 20. A defect inspection apparatus for inspecting a defect of a mask, comprising:
An illuminating means, a second illuminating means for illuminating the inside of the first area of the inspection surface from a surface different from the inspection surface of the mask, and a second area at a first position in the first area. Ray-selecting means for collecting only scattered light generated from the first ray as a first ray, ray-separating means for separating the first ray into a second ray and a third ray, and photoelectrically converting the second ray. A first photoelectric conversion means, a second photoelectric conversion means for photoelectrically converting the third light beam, and a first output signal output from the first photoelectric conversion means is gain-adjusted and output as a third output signal, A second output signal output from the second photoelectric conversion means, and a gain adjustment circuit for adjusting the gain to output as a fourth output signal; and the third output signal whose gain is adjusted by the gain adjustment circuit. None of the relative intensity difference or intensity ratio with the fourth output signal A mask inspection device for detecting the defect based on either one of them. 21. The apparatus according to claim 20, wherein a predetermined circuit pattern is drawn on the surface to be inspected. 22. The illumination light of each of the first illuminating means and the second illuminating means is linearly polarized light whose polarization planes are orthogonal to each other on the surface to be inspected.
The described device. 23. The device of claim 20, wherein the second area is substantially smaller than the first area. 24. The apparatus of claim 20, wherein the beam splitting means is a polarizing beam splitter. 25. The light beam selecting means comprises an objective lens and
21. Apparatus according to claim 20, characterized in that it is a combination of dimensional light deflection means. 26. The beam selecting means comprises an objective lens and
21. Apparatus according to claim 20, characterized in that it is a combination of dimensional light deflection means. 27. The gain setting of the gain adjusting circuit adjusts the gain so that the intensity difference between the third and fourth output signals is substantially zero or the intensity ratio is substantially one. apparatus. 28. In the mask defect inspection, when imaging is performed for observing a detected defect, the same optical system as in the defect inspection is used, and the intensity difference or intensity ratio of the third and fourth output signals. It is possible to display an image of a luminance signal proportional to
21. The device of claim 20, comprising a display device. 29. A plate-like light transmissive substrate having opposed first and second surfaces, wherein a pattern is drawn on the first surface and a pattern is not drawn on the second surface. A defect inspection apparatus for inspecting defects of a laser light source for emitting a laser beam, the laser beam is focused on a first region of the first surface of the substrate, and irradiated, and the first region of the substrate is irradiated. An optical scanning means for continuously moving in a one-dimensional direction within a pattern drawing surface, and one of two spatial regions divided by a first plane including the first surface of the substrate, including the second surface. A first light receiving means having a photoelectric conversion element, which is arranged in a first space area that is not present, and one of two space areas divided by a first plane including the first surface of the substrate, It is arranged in the second space region including and has a photoelectric conversion element. Two light receiving means and the gain of the first output signal from the first light receiving means is adjusted and output as a third output signal, and the gain of the second output signal from the second light receiving means is adjusted, A gain adjusting circuit for outputting four output signals, and the gain adjusting circuit adjusts the gain to one of the relative intensity difference or intensity ratio between the third output signal and the fourth output signal. A mask defect inspection apparatus for detecting the defect based on the above. 30. The first light ray generated from the first region and photoelectrically converted by the first light receiving means and the second light ray photoelectrically converted by the second light receiving means are with respect to the first surface. 30. The device according to claim 29, which is in a plane-symmetrical relationship. 31. The apparatus according to claim 29, wherein each of the first light receiving unit and the second light receiving unit has a plurality of photoelectric conversion elements of the same number. 32. Any one light ray of the plurality of first light rays has a plane symmetry relationship with any one light ray of the plurality of second light rays with respect to the first surface. One to one
30. The apparatus of claim 29, wherein the apparatus is a set of 33. The apparatus according to claim 29, wherein the gain setting of the gain adjusting circuit adjusts the gain so that the intensity difference between the third and fourth output signals is substantially zero or the intensity ratio is substantially one. . 34. In the mask defect inspection, the same optical system as that used in the defect inspection is used to form an image for observing the detected defect, and the intensity difference or intensity ratio of the third and fourth output signals is used. It is possible to display an image of a luminance signal proportional to
30. The device of claim 29, comprising a display device.