JPH1012530A - Apparatus for alignment - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an apparatus for alignment wherein beat signal obtained from a reticle mark is not affected, even by precisely adjusting the difference in interference fringe of two wavelengths, and by adjusting difference of interfer ence fringe on a wafer, when alignment is performed by heterodyne interference method, using two wavelengths of light. SOLUTION: Reticle mark and wafer mark are illuminated by two color lights of laser light sources 11, 12. Wavelength components of the laser light source 11, included in diffraction light from the reticle mark, are guided to a photoelectric detecting element 30, and two color diffraction lights from the wafer mark are guided to a photoelectric detecting element 31. Beat signals SR, SW are then obtained from the photoelectric detecting elements 30, 31 respectively. The phase difference between the beat signals SR and SW when the laser light source 11 is imaginary turned off is obtained, on the basis of information about the beat signal SW when the laser light source 12 is turned off, and information of the beat signal SW when both the laser light sources 11, 12 is turned in. Difference in the interference fringe is adjusted, based on the phase difference obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a photolithography process for manufacturing, for example, a semiconductor device, a liquid crystal display device, an image pickup device (CCD), or a thin film magnetic head.
The present invention relates to an alignment (positioning) device provided in an exposure device that transfers a pattern formed on a mask onto a photosensitive substrate via a projection optical system. In particular, using alignment light of a plurality of wavelengths for the photosensitive substrate, The present invention is suitably applied to an alignment apparatus of a heterodyne interference system for performing relative positioning between a mask and a photosensitive substrate.

[0002]

2. Description of the Related Art In a projection exposure apparatus used for manufacturing a semiconductor device or the like, it is necessary to maintain high accuracy of superimposition of a plurality of circuit patterns formed on a wafer (or a glass plate or the like). In particular, when exposing a pattern image of a reticle as a mask on the second and subsequent layers on the wafer, it is important to align the reticle and the wafer with high accuracy. For that purpose, an alignment system for detecting the positional relationship between the reticle and the wafer with high accuracy is required. At present, an alignment system that can be realized with the highest accuracy in principle is a TTR (through-the-track) that detects the relative positional relationship between a reticle and a wafer via a projection optical system.
(Reticle) method. By the way, as the detection light (alignment light) used in the alignment system, it is necessary to use a light flux having a wavelength different from the wavelength of the exposure light so that the photoresist applied to the wafer is not exposed.

On the other hand, there is a problem that chromatic aberration occurs in the projection optical system with respect to alignment light having a wavelength different from the exposure light, and a measure against this chromatic aberration is disclosed in Japanese Patent Laid-Open Publication No. Hei 3-3224 (hereinafter referred to as "Document 1"). ") And Tokuhei 1-404
No. 90 (hereinafter, referred to as “Document 2”). The method disclosed in Document 1 corrects chromatic aberration between an exposure wavelength and an alignment wavelength by arranging one chromatic aberration correcting lens around an optical axis at an entrance pupil position of a projection optical system. Alignment is performed by detecting ± 1st-order diffracted light from the wafer mark.

In the method disclosed in Reference 2,
A correction optical system is arranged outside the exposure optical path between the reticle and the projection optical system or inside the exposure optical path to correct the chromatic aberration caused by the alignment light passing through the projection optical system. Then, alignment is performed by detecting a reticle mark image formed on the wafer and the wafer mark via the projection optical system.

Further, with respect to correction of chromatic aberration with respect to alignment light, the present applicant disposes a chromatic aberration control member such as a phase grating near the pupil plane of the projection optical system to deflect the alignment light so that the reticle and the wafer can be moved. Has proposed a heterodyne interference type alignment device that aligns the position of the laser beam with high accuracy. In addition to the problem of chromatic aberration, if the position detection of a wafer mark is performed by monochromatic light, return light becomes weak due to conditions such as the thickness of the photoresist or the step of the mark, and it is difficult to detect the mark. It may be. Accordingly, the present applicant uses light of a plurality of wavelengths (colors) as alignment light and arranges a phase grating for controlling chromatic aberration for each light flux of each wavelength, thereby achieving more accurate alignment. A device that can do this is also proposed.

[0006]

In the prior art as described above, an alignment apparatus using light beams of a plurality of wavelengths as alignment light and arranging a phase grating so as to correspond to the vicinity of a pupil plane of a projection optical system. According to the above, since the chromatic aberration of the projection optical system itself and the chromatic aberration of the phase grating arranged near the pupil of the projection optical system cancel each other, it is possible to satisfy the condition that the chromatic aberration can be almost ignored at a plurality of alignment light wavelengths. it can.

However, strictly, some chromatic aberration remains, so that there is an inconvenience that interference fringes of a plurality of colors cannot be exactly matched on the reticle and on the wafer at the same time. Further, when the observation system (microscope) of the alignment apparatus is moved within the exposure area, the amount of fringe deviation (the amount of interference fringe deviation) on the wafer with respect to alignment light of a plurality of wavelengths changes due to the characteristics of the projection optical system. In this regard, since the characteristics of the reflectance with respect to wavelength are different depending on various layers on the wafer on which a multi-layer circuit pattern is formed, one layer hardly reflects a light beam having one wavelength, and another layer does not. The light beam having the other wavelength may not be reflected at all.

For this reason, when the fringe shift amount changes as described above, for example, a layer having different reflectances at two wavelengths has a smaller fringe shift amount than a layer which reflects light beams of two different wavelengths evenly. There is a disadvantage that the alignment detection position is shifted by an amount. In this case, there is a method (contrast method) in which the optical system is adjusted so that the contrast of a beat signal obtained from a wafer mark by light beams having two different wavelengths is maximized, thereby reducing the amount of fringe displacement. However, in order to make an adjustment so that the error due to the influence of the fringe shift can be sufficiently ignored, it is necessary to suppress the fringe shift amount within a range from several nm to several tens nm, and the contrast method is extremely insufficient.

Further, even if an adjustment can be made to reduce the amount of fringe shift generated on the wafer due to light beams of two wavelengths, the adjustment automatically causes a fringe shift on the reticle. The light amount ratio of the light beam of the wavelength is 1: 1.
When the fringe shift amount becomes a half pitch, the amplitude of the beat signal obtained from the reticle mark becomes almost zero.

In view of the above, the present invention can measure and adjust the fringe deviation with respect to the alignment light of two wavelengths with high accuracy, and even if the fringe deviation on the wafer is adjusted, it can be obtained from the reticle mark. It is an object of the present invention to provide a heterodyne interference type alignment apparatus in which a beat signal to be received is not affected.

[0011]

According to the alignment apparatus of the present invention, a pattern formed on a mask (4) under illumination light for exposure is exposed through a projection optical system (5) to a photosensitive substrate (6). An alignment apparatus, which is provided in an exposure apparatus for projecting the light onto the photosensitive substrate (6) and aligns the photosensitive substrate (6) based on a diffraction grating mark (48A) formed on the photosensitive substrate (6), has a first wavelength. The two light beams (B ₁ , B ₂ ) having a predetermined frequency difference at (λ ₁ ) are converted into a predetermined reference grating (3
5A) and a first light irradiating means (11, 13 to 26) for irradiating the diffraction grating mark (48A) via the projection optical system (5) with a first wavelength different from the first wavelength. Two beams (B ₃ , B ₄ ) having the same frequency difference as the predetermined frequency difference at the second wavelength (λ ₂ ) are irradiated on the diffraction grating mark (48A) via the projection optical system (5). The second light irradiating means (12, 13 to 26) and one or more sets of heterodyne beams (RB ₁ ⁺¹ and RB ₂ ^-1 ) generated by diffraction at the reference grating (35A). First photoelectric detection means (30) for generating a reference beat signal (S _R ) and one or more sets of heterodyne beams (W) generated by diffraction at the diffraction grating mark (48A).
Having ¹⁾ a photoelectric conversion to test the beat signal (second photoelectric detecting means for generating S _W) (31), the ^{_{- B 1 +1, WB 2 -1}} , WB 3 +1, WB 4.

Further, according to the present invention, the reference beat signal (S
_R ) and the reference grating (35A) and the diffraction grating mark (4) based on the phase difference between the beat signal (S _W ) and the test beat signal (S _W ).
8A), and a first wavelength (λ ₁ ) on a diffraction grating mark (48A) on the photosensitive substrate (6). Irradiation positions of the two light beams (B ₁ , B ₂ ) and the second wavelength (λ ₂ )
Relative position adjusting means (25) for adjusting the relative position of the two light beams (B ₃ , B ₄ ) with respect to the measurement direction with respect to the measurement direction, and the light intensity of the light beam (B ₃ , B ₄ ) of the second wavelength. The phase difference Δφ ₁ between the reference beat signal (S _R ) and the test beat signal (S _W ) and the amplitude A _{1 of the} test beat signal (S _W ), The phase difference Δφ between the reference beat signal (S _R ) and the test beat signal (S _W ) when both the light intensities of the light beams of the wavelengths are increased.
₀ and on the basis of the amplitude A ₀ of the the test beat signal (S _W), virtually the first light flux of the wavelength (B _1, B ₂₎ the reference beat signal when the light intensity is reduced in the ( S _R )
And its the test beat signal calculation means for calculating a phase difference [Delta] [phi ₂ between the (S _W) (64a), relative to the phase difference [Delta] [phi ₂ calculated by the calculation means and the phase difference [Delta] [phi ₁ equals Control means (64) for driving the position adjusting means (25).

According to the alignment apparatus of the present invention, the amount of fringe shift caused by the two alignment lights is adjusted with high accuracy. Therefore, since the position of the diffraction grating mark (48A) on the photosensitive substrate (6) is measured simultaneously with the alignment light having two different wavelengths, the thickness of the photosensitive material such as a photoresist on the photosensitive substrate (4) is reduced. Even if the spectral reflectance is not uniform under various conditions such as the film thickness, film quality, and film level of the product, highly accurate positioning can be performed.

The position of the reference grid (35A) is detected by
Since the alignment is performed using only one wavelength of alignment light, even when the positional relationship between the two wavelengths of alignment light is optimally corrected on the photosensitive substrate (6), the reference grating (35) is used.
There is no adverse effect such as a decrease in contrast of the alignment signal in A). In this case, it is desirable that the calculating means (64a) obtains the phase difference Δφ ₂ by the following calculation, for example.

Δφ ₂ = (Δφ ₀ + Δφ ₁ ) / 2 + arctan [{(A ₀ + A ₁ ) / (A ₀ −A ₁ )} tan {(Δφ ₀ −Δφ ₁ ) / 2}] (1) When the diffraction grating formed on the mask (4) is used as the reference grating (35A), the first and second light irradiation means respectively operate on the photosensitive substrate (6) via the mask (4). It is desirable to irradiate the diffraction grating mark (48A) with two light beams of the first wavelength and two light beams of the second wavelength. Thereby, alignment by the TTR method is performed.

When the first and second wavelengths are different from the wavelength of the illumination light for exposure, respectively, a Fourier transform plane for the mask (4) in the projection optical system (5) or the Fourier transform plane is used. Of the first and second
It is desirable to dispose an irradiation light control element (10) for controlling the chromatic aberration of the projection optical system (5) with respect to the light beams of the first and second wavelengths to predetermined values, respectively, in a region through which the light beam of the wavelength passes. Thereby, for example, even under the alignment light, the mask (4) and the photosensitive substrate (6) become conjugate, and the detection optical system is simplified.

[0017]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an alignment apparatus according to the present invention will be described below with reference to the drawings. FIG. 1 is a schematic configuration diagram showing a projection exposure apparatus including an alignment apparatus according to the present embodiment.
The reticle 4 is held on a reticle stage 9, and a wafer 6 coated with a photoresist as a photosensitive substrate is held on a wafer stage 61.
The illumination luminous flux of wavelength λ ₀ for exposure from irradiates the reticle 4 via the dichroic mirror 3, and a pattern image of the reticle 4 is projected through the projection optical system 5 on the wafer 6 under the illumination light. The shot area is exposed. The reticle stage 9 and the wafer stage 61 perform positioning (alignment) of the reticle 4 and the wafer 6 on a plane perpendicular to the optical axis of the projection optical system 5, respectively. The two-dimensional coordinates of the reticle stage 9 and the wafer stage 61 correspond to the interferometer 62 on the reticle 4 side and the interferometer 6 on the wafer 6 side, respectively.
3, the detection result is supplied to a central control system 64, and the central control system 64 includes a reticle stage control system 65,
The operation of the reticle stage 9 and the operation of the wafer stage 61 are controlled via the wafer stage control system 66. The central control system 64 is connected to an arithmetic unit 64a including an auxiliary computer for numerical calculation.

A reference mark member 41 on which a reference mark or the like is formed is fixed on the wafer stage 61. When the reticle 4 is aligned, the reference mark member 41 is moved into the exposure field of the projection optical system 5, and The mark is illuminated from the bottom side with a light beam having the same wavelength λ ₀ as the illumination light for exposure. The illumination light passing through the reference mark forms a reference mark image on a reticle alignment mark provided on the lower surface of the reticle 4 via the projection optical system 5, and the reticle alignment microscopes 39 and 40 above the reticle 4.
Thereby, the images of the reticle alignment mark and the reference mark are picked up, and an image pickup signal is supplied to the central control system 64. The central control system 64 processes the supplied image signal to determine the amount of displacement of the reticle alignment mark with respect to the reference mark.

Further, in order to perform a final alignment,
Alignment of TTR system and heterodyne interference system
System is provided. This alignment system is
Control system 69, alignment light transmission system 1a, beam splitter
, An objective lens 2, an alignment light receiving system 1b,
And an alignment signal processing system 70 and the like.
You. When performing alignment, the central control system 64 uses a laser
Via the control system 69, two in the alignment light transmission system 1a
Exposure wavelength λ₀Wavelength λ different from ₁of
Laser beam and wavelength λ₁Wavelength λ close to_TwoLaser Bee
And cause. Laser light in alignment light transmission system 1a
The laser beam emitted from the source undergoes a predetermined frequency modulation.
Beam alignment splitter 2 as alignment illumination light
6, through the objective lens 2 and the dichroic mirror 3,
The reticle mark on the reticle 4 is irradiated,
Part of the alignment illumination light that has passed through
Reference mark 41 (or wafer marker on wafer 6)
H).

Then, the heterodyne beam generated by diffraction at the reference mark (or wafer mark) and the heterodyne beam generated by diffraction at the reticle mark are transmitted through the dichroic mirror 3, the objective lens 2, and the beam splitter 26. Are incident on the alignment light receiving system 1b, and two beat signals are generated in the alignment light receiving system 1b. These beat signals are supplied to the alignment signal processing system 70, where the phase difference between the two beat signals is detected, and the detected phase difference is supplied to the central control system 64. Based on the detected phase difference, the central control system 64
Calculates a target phase difference at the time of alignment or performs final alignment. Further, the alignment signal processing system 70 of the present example also has a function of detecting the amplitude of the beat signal.

Next, the reticle alignment microscope of this embodiment and the heterodyne type alignment system will be described in detail. FIG. 2 is a schematic configuration diagram of a stage system and an alignment optical system of the projection exposure apparatus of the present embodiment. In FIG. 2, the Z axis is taken in parallel with the optical axis AX of the projection optical system 5, and
An orthogonal coordinate system on a plane perpendicular to the axis is defined as an X axis and a Y axis.
FIG. 2A is a front view of the projection exposure apparatus viewed in the Y direction, and FIG. 2B is a side view of FIG. 2A. The alignment optical system 1 corresponds to the alignment light transmission system 1a, the beam splitter 26, and the alignment light receiving system 1b in FIG. 1, and includes an X stage 8X and a Y stage 8Y for positioning the wafer 6 in the X direction and the Y direction, respectively. This corresponds to one wafer stage 61.

In FIG. 2, a wafer holder 7 for holding the wafer 6 is provided between the wafer 6 and the X stage 8X and the Y stage 8Y. Actually, a Z stage for positioning the wafer 6 in the Z direction is mounted on the X stage 8X, and the reference mark member 41 is provided on the Z stage. The alignment optical system 1 according to the present embodiment includes a reticle alignment illumination light RB _{1 having} an average wavelength λ ₁ different from the exposure wavelength λ ₀ and a frequency difference Δf (50 kHz in this example).
RB ₂ and wafer alignment illumination light WB ₁ , WB ₂
And a frequency difference Δf (= 5) at an average wavelength λ ₂ close to the wavelength λ _1.
0 kHz) reticle alignment illumination light RB ₃ , RB
₄ and the wafer alignment illumination light WB ₃ , WB ₄ .

As shown in FIG. 2A, the reticle alignment illumination lights RB ₁ and RB ₂ are condensed on the reticle 4 by the objective lens 2 and are respectively formed on diffraction grating-like reticle marks 35A on the lower surface of the reticle 4. Incident angle − θ _R1 , θ _R1
Incident. FIG. 3 shows the pattern arrangement of the reticle 4 of the present example. In FIG. 3, a light-shielding band 33 is formed around a transfer pattern region 32 at the center of the reticle 4, and extends in the X direction in the light-shielding band 33. reticle mark 3 for the Y axis composed of the diffraction grating formed at a pitch P _R in the Y direction during the sides
6A and 36B are formed, the reticle mark 35A and 35B for the X-axis made of the diffraction grating formed at a pitch P _R in the X direction during a side extending in the Y direction is formed. Transmissive windows (hereinafter, referred to as “reticle windows”) 37A and 3 for passing alignment illumination light toward the wafer side inside reticle marks 35A and 35B, respectively.
7B are formed, and reticle windows 38A and 38B are formed inside the reticle marks 36A and 36B for passing the alignment illumination light directed toward the wafer 6 side. Further, cross-shaped reticle alignment marks 34A and 34B are formed so as to sandwich the light shielding band 33 in the X direction.

FIG. 9 is an enlarged view of reticle mark 35A and reticle window 37A in FIG. 3. In FIG. 9, reticle mark 35A has reticle alignment illumination light RB.
_{1, RB 2, RB 3,} RB 4 aligned illuminated illumination light 50 is made of, the reticle window 37A wafer alignment illumination light _{WB 1, WB 2, WB 3} , WB alignment illumination light 51 consisting of ₄ is passing .

Referring back to FIG. 2, the incident angles -θ _R1 , θ _R1 and the grating pitch P _R of the reticle mark 35A have a relationship represented by the following equation (2), and the reticle alignment illumination light RB ₁ +
First order diffracted light RB ₁ ⁺¹ and reticle alignment illumination light RB
_{2 and} the -1st order diffracted light RB ₂ ^-1 are generated directly above, respectively.
The light returns to the alignment optical system 1 via the objective lens 2 as a heterodyne beam (reticle alignment detection light).

Sin (θ _R1 ) = λ ₁ / P _R (2) Similarly, the reticle alignment illumination lights RB ₃ and RB ₄ are also condensed on the reticle 4 by the objective lens 2 and are focused on the reticle mark 35 A on the reticle 4. Irradiation is performed at incident angles −θ _R2 and θ _R2 . At this time, the reticle alignment illumination light R
The + 1st-order diffracted light RB ₃ ^{+1 of} B ₃ and the −1st-order diffracted light RB ₄ ^{-1 of the} reticle alignment illumination light RB ₄ are generated directly above, and return to the alignment optical system 1 via the objective lens 2.

On the other hand, the wafer alignment illumination light WB₁,
WB_TwoPasses through the reticle window 37A on the reticle 4 and
The light reaches the chromatic aberration control plate 10 in the shadow optical system 5. Chromatic aberration control
Illumination light WB of plate 10₁, WB_TwoIn the area where
On-axis chromatic aberration control element G1 each in the form of a diffraction grating_A, G1_B
(See FIG. 8), and the illumination light WB₁, WB _Two
Is the angle -θ_G1, Θ_G1Only bent, diffraction grating
Angle -θ with respect to the wafer mark 48A_W1,
θ_W1Irradiation.

FIG. 4 shows a part of a shot arrangement on the wafer 6. In FIG. 4, a diffraction grating formed at a pitch P _W in the Y direction so as to sandwich the shot area 47 to be exposed in the X direction. The wafer marks 49A and 49B for the Y-axis are formed, and the wafer marks 48A and 48B for the X-axis are formed by a diffraction grating formed at a pitch _PW in the X-direction so as to sandwich the shot area 47 in the Y-direction. ing. Each of the other shot areas is also provided with a wafer mark.

FIG. 10 shows the wafer mark 48A shown in FIG.
FIG. 10 is an enlarged view of FIG.
A, wafer alignment illumination light WB₁, WB_Two , WB
_Three, WB_FourIllumination light 51 composed of
ing. Returning to FIG. 2A, the incident angle −θ_W1, Θ_W1And Ue
Lattice pitch P of HAMARK 48A _WIs the following equation (3)
And the wafer alignment illumination light WB₁+1
Next order diffracted light WB₁ ⁺¹ And wafer alignment illumination light WB_Twoof
-1st order diffracted light WB_Two ^-1 Each occur directly above,
These two diffracted lights are heterodyne beams (wafer alignment
(Light detection light).

Sin (θ _W1 ) = λ ₁ / P _W (3) Similarly, since the wavelengths of the wafer alignment illumination lights WB ₃ and WB ₄ are close to the wafer alignment illumination bundles WB ₁ and WB _2, they are on the chromatic aberration control plate 10. Can be regarded as substantially on the axial chromatic aberration control elements G1 _A and G1 _B , respectively. Therefore, the illumination lights WB ₃ and WB ₄ are respectively at angles −θ _G2 ,
The light beam is bent by θ _G2 and irradiated onto the wafer mark 48A at incident angles −θ _W2 and θ _W2 , respectively. And the illumination light W
Order diffracted light WB ₃ ⁺¹ of B ₃ and the -1-order diffracted light WB ₄ ^-1 of the illumination light WB ₄ occurs directly above each a heterodyne beam.

In this case, as shown in FIG. 2B, the alignment illumination light is deflected by the chromatic aberration control plate 10,
Angle θ with respect to wafer 6 in the non-measurement direction (Y direction)
_{Since the} light enters at an angle of _m , the position at which each heterodyne beam passes on the chromatic aberration control plate 10 is different from the position at the time of incidence. The heterodyne beam from the wafer mark 48A passes through the axial chromatic aberration control element G1 _C (see FIG. 8) on the chromatic aberration control plate 10 so that the chromatic aberration in the horizontal direction is corrected and goes to the reticle window 37A. Thereafter, each detection light returns to the alignment optical system 1 again via the reticle window 37A and the objective lens 2. Further, the wafer alignment illumination light illuminates a position shifted by Δβ in the Y direction on the surface of the wafer 6 as compared with the case where the chromatic aberration control plate 10 is not provided.

FIG. 8 is a block diagram showing the chromatic aberration control plate 10 according to the present embodiment. In FIG. 8, 12 axial chromatic aberration control elements are provided on a transmissive chromatic aberration control plate 10 made of a glass substrate or the like. Is arranged. Then, the three-axis chromatic aberration control elements G1 _A, G1 _B, G1 _C is used for the deflection of the wafer alignment illumination light, and a heterodyne beam alignment optical system 1 FIG. Actually, since there is another three-axis alignment optical system, the chromatic aberration control plate 1
0 throughout On 12 (= 3 × 4) pieces of the axial chromatic aberration control elements _{G1 A, G1 B, G1 C} ~G4 A, G4 B, G4 C are formed.

Here, the alignment optical system 1 of this embodiment will be described in detail with reference to FIG. 6 and FIG. FIG.
FIG. 6B is a view of the alignment optical system 1 viewed from the same direction as FIG. 2B, and FIG.
FIG. 6C is a diagram viewed from the same direction as FIG.
FIG. 7 is a perspective view showing a front part of the alignment optical system.

In FIG. 6, a laser beam having a wavelength λ ₁ emitted from a first laser light source 11 composed of a laser diode, for example, and a wavelength λ emitted from a second laser light source 12 composed of a He—Ne laser light source, for example. the _second laser beams are combined by the beam combining prism 13 in alignment illumination light B, acousto-optic modulation device that is driven by the respective frequency F ₁ (hereinafter, referred to as "AOM")
14 is incident. Most AOMs are of the type in which a light beam is incident at a Bragg angle (so-called Bragg-type AOM) in order to increase the intensity of the + 1st-order diffracted light. What is obtained evenly, that is, Raman-nasal AOM. Since the two laser beams in the alignment illumination light B have slightly different wavelengths, the two laser beams are diffracted by the AOM 14 at slightly different angles. The alignment illumination light diffracted in different directions for each wavelength is shown by a solid line and a broken line in FIG. These +1 order and -1 then diffracted alignment illumination light are respectively the original frequency + F _1, and -F ₁
Frequency difference.

In FIG. 7, the laser beam emitted from the AOM 14 passes through a lens 15 and passes through a spatial filter 1.
6 and only the light beam diffracted by ± 1 order by the spatial filter 16 is selected as the alignment illumination light, and is then disposed via the lens 17 at a 45 ° rotation with respect to the AOM 14 and at the frequency F AO driven by ₂
The light enters so as to intersect M18. The AOM 18 is a Bragg-type AOM, and the incident angle is 1 / sin 45 of the Bragg angle.
° is set. Therefore, 45 around the optical axis.
The angle of incidence of the alignment illumination light is the Bragg angle when viewed from a plane rotated by °. In this case, the laser beam that has been subjected to + F ₁ frequency modulation by the AOM 14 generates strongly −1st-order diffracted light at the AOM 18, and at the same time receives the −F ₂ frequency modulation by the AOM 18, and is emitted from the laser light sources 11 and 12. On the other hand, it becomes the alignment illumination lights B ₁ and B ₃ that have been subjected to + (F ₁ −F ₂ ) frequency modulation. Similarly, the laser beam subjected to frequency modulation of -F ₁ in AOM 14 AOM18
+ 1st order diffracted light is strongly generated at the same time,
_The alignment illumination lights B ₂ and B ₄ are subjected to frequency modulation of −2 and subjected to frequency modulation of − (F ₁ −F ₂ ) with respect to the light emitted from the laser light sources 11 and 12. as a result,
Alignment illumination light B ₁ , B ₃ and alignment illumination light B
The frequency difference Δf between ₂ and B ₄ is represented by 2 (F ₁ −F ₂ ). In this example, it is assumed that Δf is set to 50 kHz. The alignment light beam that has passed through the AOM 18 enters the spatial filter 20 through the lens 19,
The alignment illumination lights B _{1 to} B ₄ selected at 0 are guided to the subsequent stage.

Referring back to FIG. 6, the alignment illumination lights B _{1 to} B
₄ is condensed on a field stop 22 by a lens 21 and a beam shape on a reticle 4 or a wafer 6 is determined, and then a reticle / wafer beam separating prism 23 causes reticle alignment illumination light RB _{1 to} RB ₄ and wafer alignment. It is divided into illumination light WB ₁ ~WB _4. After that, the alignment illumination light reaches the direct-view prism 25 via the lens 24. The direct-view prism 25 is rotatable about the optical axis, and is driven by a motor 80 in accordance with an instruction from the central control system 64 in FIG. Direct-view prism 25
Rotate, the relative angles of the two colors of illumination light of wavelengths λ ₁ and λ ₂ change, and the alignment illumination light RB ₁ and WB of wavelength λ ₁
The alignment illumination lights RB ₃ , WB ₃ , RB ₄ , and WB ₄ having the wavelength λ ₂ are separated from ₁ , RB ₂ , and WB ₂ , respectively. The alignment illumination light whose relative angle has changed in this way goes to the objective lens 2 in FIG. 2 via the beam splitter 26. Due to the change in the relative angle, the relative relationship between the irradiation positions of the alignment illumination light of two colors of wavelengths λ ₁ and λ ₂ on the reticle 4 and the wafer 6 also changes.

On the other hand, the reticle mark 37A shown in FIG.
The heterodyne beam from Ehamark 48A is shown in FIG.
After returning to the alignment optical system 1, the beam splitter 2
6, the reticle 4, via the lens 27,
Light separating prism 2 at a position conjugate with wafer 6
8, heterodyne with reticle mark 37A
Beam and heterodyne beam from wafer mark 48A
And separated. Diffracted light RB from reticle mark
₁ ⁺¹ , RB_Two ^-1 And RB_Three ⁺¹ , RB_Four ^-1 Heterogeneous
The Dyne beam passes through the detection light separation prism 28 and has a wavelength
λ₁Diffracted by the color filter 29 that transmits only the light beam of
Light RB₁ ⁺¹ , RB_Two ^-1 Only the photodetection element 3
0 is received. Diffracted light WB from wafer mark
₁ ⁺¹ , WB_Two ^{- 1} And WB_Three ⁺¹, WB_Four ^-1 Heterogeneous
The Dyne beam is reflected by the detection light separation prism 28.
Thus, the light is received by the photoelectric detection element 31. Photoelectric detector
Reticle corresponding to the position of reticle mark 35A from child 30
Curve beat signal S_RIs output from the photoelectric detection element 31.
Wafer beat signal corresponding to the position of wafer mark 45A
S _WIs output.

Reticle beat signal S _R is diffracted light RB
₁ ^+1, a sinusoidal beat signal of a frequency Δf due to RB ₂ ^-1, wafer beat signal S _W diffracted light WB ₁ ^+1, W
This is a sinusoidal beat signal having a frequency Δf based on B ₂ ⁻¹ , WB ₃ ⁺¹ and WB ₄ ⁻¹ . The phase difference Δφ [ra
d] changes depending on the relative movement amount of the reticle 4 and the wafer 6 in the X direction, and the relative movement amount Δx is given by the following (4):
And (5).

Δx (on reticle) = P _R · Δφ / (4π) (4) Δx (on wafer) = P _W · Δφ / (4π) (5) Also, for the Y axis, The beat signal corresponding to the reticle mark and the wafer mark of
The phase difference between these beat signals is obtained in the same manner as in the X direction.

Although the wafer mark is positioned below the projection optical system 5 in FIG. 2, the reference mark member 41 is moved below the projection optical system 5 when the alignment optical system 1 is adjusted (calibrated). And reference mark member 4
1 is illuminated with the wafer alignment illumination light. FIG. 5 is a plan view for explaining the configuration on the reference mark member 41. In FIG. 5, a frame is provided at a predetermined interval in the X direction on the upper surface of the transparent reference mark member 41 made of a glass substrate or the like. Reference marks 44A and 44B are formed, and reference diffraction grating marks 46A and 46B are formed between the reference marks at a predetermined pitch in the Y direction. The reference diffraction marks are formed at a predetermined pitch in the X direction. Grid marks 45A and 45B are formed. The pitch of the reference diffraction grating marks 45A to 46B is the same as the wafer mark 4 in FIG.
Equal to the pitch of 8A to 49B.

As described above, at the time of reticle alignment, the reference mark member 41 is moved into the exposure field of the projection optical system 5, and the reference marks 44A and 44B are illuminated from the bottom side with a light beam having the same wavelength λ ₀ as the exposure light. . When the alignment optical system 1 is calibrated, the reference mark member 41 is also moved into the exposure field of the projection optical system 5 and the grid mark 45 on the reference mark member 41 is moved.
A is illuminated with the wafer alignment illumination light similarly to the wafer mark.

As described above, in the alignment optical system 1 used in the alignment apparatus of this embodiment, in order to reduce the influence of the thin film interference and the like, the two wavelengths λ ₁ and λ ₂ are applied to the wafer mark. In addition to using alignment illumination light, a color filter 29 that transmits only alignment illumination light of wavelength λ ₁ is provided in front of the reticle mark photoelectric detection element 30 in order to eliminate fringe deviation on the reticle mark side. ₂ to block the alignment illumination light, wavelength λ ₁
Reticle beat signal S using only the alignment illumination light
_R is output. Furthermore, a rotatable direct-view prism 2
5 is provided to change the relative relationship between the irradiation positions on the wafer 6 of the alignment illumination light of the _two wavelengths λ ₁ and λ ₂ .

Next, the alignment operation in this embodiment is divided into two steps, namely, a step consisting of adjustment of the alignment optical system 1 and a step consisting of alignment and exposure of the wafer 6, and mainly with reference to FIG. This will be specifically described. In the following, the positioning operation in the X direction will be described, but the positioning in the Y direction is performed in the same manner.

First, the adjustment of the alignment optical system 1, which is the first step, will be described. First, reticle 4
Is transported onto the reticle stage 9 by a reticle autoloader (not shown). Here, the reticle 4 is first subjected to reticle alignment. The reticle alignment is performed by using exposure light of wavelength λ ₀ , and reticle alignment marks 34 A and 34 B shown in FIG. 3 formed on reticle 4 and wafer stage 61 shown in FIG.
The reference marks 44A and 44B formed on the reference mark member 41 mounted on the reticle alignment microscope 3
Observed according to 9, 40. At the time when the reticle alignment is completed, the alignment light transmitting system 1a
The reference diffraction grating mark 45A on the reference mark member 41 is irradiated with wafer alignment illumination light. At this time, the reticle mark 35A on the reticle R is irradiated with reticle alignment illumination light.

The wafer alignment illumination light applied to the reference diffraction grating mark 45A is diffracted by the reference diffraction grating mark, and the diffracted light beam is received by the alignment light receiving system 1b. The photoelectric detection element 31 of FIG.
Beat signal (hereinafter, this signal is also referred to as a "wafer beat signal S _W") is outputted from. The amplitude information of the beat signal is sent to the central control system 6 via the alignment signal processing system 70.
Is supplied to the 4, the central control system 64 so that the amplitude of the wafer beat signal S _W is maximized, adjusting the rotational angle of the direct vision prism 25 via the motor 80 alignment light sending system in 1a shown in FIG. 6 . This completes the coarse adjustment.

After the coarse adjustment step is completed, the process proceeds to the fine adjustment step.
This fine adjustment step will be described separately in two steps, a first fine adjustment step and a second fine adjustment step. First, in the first fine adjustment step, based on an instruction from the central control system 64, the laser control system 69 causes the laser light source 12 of the alignment illumination light having the wavelength λ _{2 in} FIG.
Is turned off. In this case, the wavelength λ from the laser light source 11
_This means that the relative position between the reticle mark 35A on the reticle 4 and the reference diffraction grating mark 45A on the reference mark member 41 is detected only by the _one alignment illumination light.

[0047] Figure 11 is a reticle beat signal S _R corresponding to the reticle marks 35A to be supplied to the alignment signal processing system 70 as an alignment signal in the first fine adjustment process, the wafer beat signal S _W corresponding to the reference diffraction grating mark 45A and the horizontal axis represents time t, and the vertical axis represents both the beat signal S _R, S _W. In this case, the reticle beat signal S _R, and the amplitude A _W1 of the phase difference [Delta] [phi ₁ and wafer beat signal S _W of the wafer beat signal S _W is determined by the alignment signal processing system 70. This phase difference Δφ ₁ is 0
Is preferable, but it is not necessarily required to be 0.
Just measure it.

Next, in the second fine adjustment step, the laser light source 12 shown in FIG. That is, detection similar to that in the first fine adjustment step is performed by the alignment illumination light of wavelength λ ₁ from the laser light source 11 and the alignment illumination light of wavelength λ ₂ from the second laser light source 12. Figure 12 shows a reticle beat signal S _R corresponding to the reticle marks 35A to be supplied to the alignment signal processing system 70 as an alignment signal in the second fine adjustment process, the wafer beat signal S _W corresponding to the reference diffraction grating mark 45A, The horizontal axis is time t,
The vertical axis represents both the beat signal S _R, S _W. In this case, the alignment signal processing system 70, the phase difference Δφ of the reticle beat signal S _R and the wafer beat signal S _W
0, is determined and the amplitude A _W0 of the wafer beat signal S _W. Incidentally, by comparing FIGS. 11 and 12, the wafer beat signal S _W detected by the laser beam of two wavelengths amplitude A
Who _W0 is than the amplitude A _W1 of the wafer beat signal S _W detected by the laser beam of one wavelength, but is smaller, not an separate special. This is because if the amount of fringe shift between the two wavelengths is large, the respective wafer beat signals corresponding to the two wavelengths reduce the contrast with each other.

Thereafter, the central control system 64 sends the alignment results Δφ ₁ , A _W1 , Δφ from the alignment signal processing system 70.
₀ , A _W0 and supplies them to the operation unit 64a.
The phase difference [Delta] [phi ₂ between the reticle beat signal _R and the wafer beat signal S _W when performing alignment in if only the laser light source 12, 4a only alignment illumination light having a wavelength lambda ₂ when the lighting of the following (6) It is obtained by the formula and supplied to the central processing system 64. However, at this time, it is assumed that only the reticle mark 35A is irradiated with the illumination light from the laser light source 11. Here, the difference between the phase differences Δφ ₁ and Δφ ₂ corresponds to the deviation of the interference fringes of the wavelengths λ ₁ and λ ₂ . The fringe shift amount is ideally 0, but if it exceeds the allowable amount, the central processing system 64 adjusts the direct-view prism 25 to the motor 80 in the alignment light transmission system 1a according to the fringe shift amount. Instruct to rotate by an angle. Then, the difference between the phase differences Δφ ₁ and Δφ ₂ is measured again by the same method, and the adjustment process is repeated until the difference becomes equal to or less than the allowable amount. The allowable amount varies depending on the required alignment accuracy, but with the recent high-density integration of LSIs, an allowable amount of 10 nm or less will be required in the future.

Δφ ₂ = (Δφ ₀ + Δφ ₁ ) / 2 + arctan [{(A _W0 + A _W1 ) / (A _W0 −A _W1 )} tan {(Δφ ₀ −Δφ ₁ ) / 2}] (6) By this method, interference fringes on the reference mark member 41 due to the alignment illumination light of two different wavelengths λ ₁ and λ ₂ are accurately corrected. That is, by the above adjustment,
Two wavelengths lambda _1, the phase difference of the optical beat signal of lambda ₂ of the alignment illumination light becomes within the allowable range, it is possible to obtain a good wafer beat signal S _W contrast during final alignment. Here, the central control system 64 illuminates the laser light source 11, 12 again two colors, measures and stores the phase difference [Delta] [phi ₀ between reticle beat signal S _R and the wafer beat signal S _W.

Here, the process of deriving equation (6) will be described. First, the reticle beat signal S _R obtained by the illumination light from the laser diode is expressed as follows using the angular frequency ω, the amplitude A _R , and the phase ε corresponding to the beat frequency. _{S R = A R sin (ωt} -ε) (B1) Similarly, the wafer beat signal S _W1 obtained when irradiated with only the illumination light from the laser diode, using the amplitudes A _W1 and phase beta, the following To

[0052] _{S W1 = A W1 sin (ωt} -β) (B2) In addition, wafer beat signal S _W2 obtained when irradiated only with the illumination light from the virtually He-Ne laser light source, the amplitude A _W2 and phase Using γ, it is expressed as follows. _{S W2 = A W2 sin (ωt} -γ) (B3) Further, the wafer beat signal S _W obtained when lit both laser diodes and He-Ne laser light source,
It is expressed as follows using the amplitude A _W0 and the phase α.

S _W = A _W0 sin (ωt−α) (B4) The following relationship exists between the expressions (B2) and (B4). A _W2 sin (ωt−γ) = A _W0 sin (ωt−α) −A _W1 sin (ωt−β) (B5) In order to obtain the phase γ from this equation, the following variables C and δ are introduced. I do.

C = ωt− (α + β) / 2, δ = (α−β) / 2 (B6) When these variables are used, the right side of the expression (B5) is as follows. A _W0 sin (C−δ) −A _W1 sin (C + δ) (B7) By transforming the equation (B7), the following equation is obtained. A _W0 sin Ccos δ−A _W0 cos Csin δ−A _W1 sin Ccos δ−A _W1 cos Csin δ = (A _W0 −A _W1 ) sin Ccos δ− (A _W0 + A _W1 ) cos Csin δ (B8) New variables G and ξ that satisfy the following equation are introduced.

(A _W0 −A _W1 ) cos δ = G cos ξ (B9) (A _W0 + A _W1 ) sin δ = G sin と (B10) Using these variables G and ξ, the equation (B8) becomes become. The fact that equation (B8) is equal to the left side of equation (B5) is used. A _W2 sin (ωt−γ) = Gsin Ccos ξ−Gcos Csin ξ = Gsin (C−ξ) = Gsin {ωt− (α + β) / 2−ξ｝ (B11) where (B9) and (B10) From the equation, the variable を 満 た す satisfies the following equation.

Tan ξ = (A _W0 + A _W1 ) sin δ / ｛(A _W0 −A _W1 ) cos δ｝ = ｛(A _W0 + A _W1 ) / (A _W0 −A _W1 )｝ tan δ = ｛(A _W0 + A _W1 ) / (A _W0 −A _W1 ) {tan {(α−β) / 2} (B12) Accordingly, from the equation (B11), the amplitude A _W2 and the phase γ are as follows. A _W2 = G (B13) γ = (α + β) / 2 + ξ = (α + β) / 2 + arctan [{(A _W0 + A _W1 ) / (A _W0 −A _W1 )} tan {(α−β) / 2}] ( B14) next, the phase difference [Delta] [phi ₁ between the reticle beat signal S _R and the wafer beat signal S _W1 obtained when turned off the He-Ne laser light source is located at the difference between the phase β and ε, as shown in the following equation The amplitude A _W1 of the wafer beat signal SW ₁ at that time is measured.

[0057] _{Δφ 1 = β-ε (B15} ) The phase difference between the reticle beat signal S _R and the wafer beat signal S _W obtained when both the light source lit [Delta] [phi ₀
Is the difference between the phase α and ε, as shown in the following equation, the amplitude A _W0 of the wafer beat signal S _W at that time is measured. Δφ ₀ = α−ε (B16) Here, the phase difference to be obtained is the phase difference Δφ ₂ between the reticle beat signal S _R and the wafer beat signal SW ₂ obtained when the laser diode is virtually turned off only on the wafer side. ,
That is, the difference between the phase γ and the phase ε is as follows, and this difference is obtained from the equations (B14) to (B16) as follows:
Matches the formula.

Δφ ₂ = γ−ε = (α−ε + β−ε) / 2 + arctan [{( _AW0 + _AW1 ) / ( _AW0 − _AW1 )} tan {(α−ε−β + ε) / 2}] = (Δφ ₀ + Δφ ₁ ) / 2 + arctan [{(A _W0 + A _W1 ) / (A _W0 −A _W1 )} tan {(Δφ ₀ -Δφ ₁ ) / 2}] This equation can be variously modified. It is.

Next, the alignment and exposure of the wafer 6, which is the second step, will be described. In FIG.
The wafer 6 is transferred onto the wafer stage 61 by a wafer autoloader (not shown). Based on the measurement result of the wafer pre-alignment system (not shown), the wafer mark 48
The shot area 47 shown in FIG. 4 of the wafer 6 is positioned at the exposure position below the projection optical system 5 with an accuracy of ± 1/4 or less of the pitch _PW of A, 49A. After the positioning, the alignment illuminating light is emitted from the alignment light transmitting system 1a to the reticle mark 35A and the wafer mark 48A, and the final alignment (fine alignment) between the reticle 4 and the shot area 47 of the wafer 6 is performed.

[0060] In this final alignment, the reticle beat signal S _R and the wafer beat signal S _W and the reticle stage 9 or the wafer stage so that the phase difference becomes a phase difference [Delta] [phi _0, which is measured at the end of the first step of the above 61
Is controlled. Shot area 4 of reticle 4 and wafer 6
7 is transmitted to the central control system 64 through the alignment signal processing system 70, the exposure light is emitted from the exposure illumination system 67, and the pattern image on the reticle 4 is transferred into the shot area 47. . In this manner, each shot area on the wafer 6 is sequentially exposed, and when the exposure of all the shot areas on the wafer 6 is completed, the wafer 6 is replaced with the next wafer and the same as performed on the wafer 6 again. Is performed and exposure is performed.

As described above, in this example, the reticle mark 35A of the reticle 4 is illuminated with only the alignment illumination light having the wavelength λ ₁ during alignment. Basically, the reticle mark of the reticle is usually chrome (Cr) on quartz glass.
Since it is patterned by a film, it is considered that an extremely ideal mark is formed as compared with a wafer mark, and it is not particularly necessary to perform alignment using alignment illumination light of _two wavelengths λ ₁ and λ ₂ . The embodiment of this example makes good use of this point. Since the reticle beat signal S _R is based on only the alignment illumination light having the wavelength λ ₁ , it is always stable even when the direct-view prism 25 of FIG. 6 is rotated. the reticle beat signal S _R that is obtained.

In the above embodiment, a laser diode is used as the first laser light source 11 and a He—Ne laser light source is used as the second laser light source 12, so that
When turning off the second laser light source 12, the shutter may block the light beam. Further, although the alignment illumination light of two wavelengths λ ₁ and λ ₂ is used as the illumination light for alignment, alignment illumination light of three or more wavelengths may be used.

Although the direct-view prism 25 is used for adjusting the interference fringes, the same effect can be obtained by using a grating such as a diffraction grating. In the above-described embodiment, the present invention is applied to the alignment apparatus of the heterodyne interference system of the TTR system.
The present invention can also be applied to an alignment apparatus of a heterodyne interference type or an off-axis type. In this case, a reference diffraction grating is used instead of the reticle mark. Further, as a reference, for example, an electric signal for driving the AOM may be used.

It should be noted that the present invention is not limited to the above-described embodiment, and it is needless to say that various configurations can be adopted without departing from the gist of the present invention.

[0065]

According to the alignment apparatus of the present invention, the phase difference Δφ ₁ between the reference beat signal and the test beat signal when the light intensity of the light beam of the second wavelength is reduced, and the amplitude of the test beat signal A ₁ , the phase difference Δφ ₀ between the reference beat signal and the test beat signal when the light intensities of the light beams of the first and second wavelengths are both increased, and the amplitude A _{0 of the} test beat signal. Based on the calculated phase difference Δφ ₂ between the reference beat signal and the test beat signal when the light intensity of the light beam of the first wavelength is reduced virtually, the phase difference Δφ ₂ and the phase difference Δφ _The relative position adjusting means is driven so that ₁ becomes equal to ₁ . Therefore, when simultaneously measuring the position of the photosensitive substrate with the alignment light beams having two different wavelengths, the amount of fringe shift between the two alignment lights can be accurately reduced to almost zero.
There is an advantage that can be adjusted. Since the reference grid (mark on the mask) is detected in one color, the signal obtained from the reference grid is not affected even if the amount of fringe deviation is adjusted.

When the calculating means obtains the phase difference Δφ ₂ by the calculation formula (1), the fringe displacement amount is obtained with high accuracy by one measurement, and the fringe displacement adjustment is performed in a short time. be able to. Also, when a diffraction grating mark on the mask is used as a reference grating, and the two light irradiating means respectively irradiate the diffraction grating mark on the photosensitive substrate with a light beam through the mask, a high precision by the TTR method is used. Another advantage is that the amount of displacement between the mask and the photosensitive substrate can be directly detected.

Further, when the first and second wavelengths are different from the wavelength of the illumination light for exposure, on the Fourier transform plane for the mask in the projection optical system or on the plane near the Fourier transform plane, the first and second wavelengths are different. When irradiating light control elements for controlling the chromatic aberration of the projection optical system with respect to the light beams of the first and second wavelengths to respective predetermined values are respectively arranged in regions through which the light beams of the first and second wavelengths pass. Has the advantage that the mask and its substrate can be easily conjugated with respect to the alignment light beam, so that alignment can be performed by the TTR method with a simple optical system using a common detection optical system for the mask and the substrate.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram showing an embodiment of an alignment apparatus according to the present invention.

2A is a front view mainly showing a stage system and an alignment optical system of the alignment apparatus of FIG. 1, and FIG. 2B is a side view of FIG. 2A.

FIG. 3 is a plan view showing a pattern arrangement of a reticle 4 of FIG. 2;

FIG. 4 is an enlarged plan view showing a shot area and a wafer mark of the wafer 6 of FIG. 2;

FIG. 5 is an enlarged plan view showing a mark arrangement on a reference mark member 41 of FIG. 2;

6A is a front view showing the configuration of the alignment optical system 1 of FIG. 2, FIG. 6B is a side view of FIG. 6A, and FIG. 6C is a bottom view of FIG. 6A.

FIG. 7 is a perspective view showing a configuration from an AOM 14 to a spatial filter 20 in FIG.

8 is a plan view showing an arrangement of a chromatic aberration control element on a chromatic aberration control plate 10 in FIG.

9 is a reticle mark 35A and reticle window 3 of FIG.
It is an enlarged plan view which shows 7A.

FIG. 10 is an enlarged plan view showing a wafer mark 48A of FIG.

FIG. 11 is a diagram illustrating a configuration of an alignment apparatus according to an embodiment of the present invention, which is obtained by an alignment system using a heterodyne interference method.
FIG. 6 is a waveform diagram showing a phase relationship when only one beat signal is illuminated with an alignment light beam having a first wavelength.

FIG. 12 is a diagram illustrating a configuration of an alignment apparatus according to an embodiment of the present invention.
FIG. 9 is a waveform diagram showing a phase relationship when one beat signal is illuminated with alignment light beams having first and second wavelengths.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Alignment optical system 2 Objective lens 3 Dichroic mirror 4 Reticle 5 Projection optical system 6 Wafer 7 Wafer holder 8X X stage 8Y Y stage 9 Reticle stage 10 Chromatic aberration control plate 11,12 Laser light source 14,18 Acoustic optical modulator (AOM) 22 Field of view Aperture 23 Reticle / wafer beam separation prism 25 Direct-view prism 26 Beam splitter 29 Color filter 30, 31 Photodetector 34A, 34B Reticle alignment mark 35A, 35B, 36A, 36B Reticle mark 39, 40 Reticle alignment microscope 48A, 48B, 49A, 49B Wafer mark 41 Reference mark member 44A, 44B Reference mark 45A, 45B, 46A, 46B Reference diffraction grating mark 64 Central control system 64a Operation unit

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification number Agency reference number FI Technical display location H01L 21/30 516B 525M

Claims (4)

[Claims] 1. A diffraction grating provided on an exposure apparatus for projecting a pattern formed on a mask under an illumination light for exposure onto a photosensitive substrate via a projection optical system, and formed on the photosensitive substrate. An alignment device for performing positioning of the photosensitive substrate based on a shape mark, irradiating a predetermined reference grating with two light beams having a predetermined frequency difference at a first wavelength, and the diffraction grating via the projection optical system. A first light irradiating means for irradiating the shape mark, and a diffraction grating through the projection optical system, the two light fluxes having the same frequency difference as the predetermined frequency difference at a second wavelength different from the first wavelength. Second light irradiating means for irradiating the mark, first photoelectric detecting means for photoelectrically converting one or more sets of heterodyne beams generated by diffraction on the reference grating to generate a reference beat signal, Second photoelectric detection means for photoelectrically converting at least one set of heterodyne beams generated by diffraction at the lattice mark to generate a test beat signal; and a phase difference between the reference beat signal and the test beat signal. A position shift amount detection unit for detecting a relative position shift amount between the reference grating and the diffraction grating mark based on the first and second wavelengths of the first wavelength on the diffraction grating mark on the photosensitive substrate. Relative position adjusting means for adjusting the relative position of the irradiation position and the irradiation position of the two light beams of the second wavelength with respect to the measurement direction; and the reference beat signal when the light intensity of the light beam of the second wavelength is reduced. Phase difference Δφ _{1 from the} test beat signal
And the phase difference Δφ ₀ between the reference beat signal and the test beat signal when both the amplitude A ₁ of the test beat signal and the light intensity of the light beams of the first and second wavelengths are increased, and The phase difference Δφ between the reference beat signal and the test beat signal when the light intensity of the light beam of the first wavelength is virtually reduced based on the amplitude A _{0 of the} detection beat signal.
₂ , and control means for driving the relative position adjusting means such that the phase difference Δφ ₂ calculated by the calculating means is equal to the phase difference Δφ _1. Alignment device. 2. The alignment apparatus according to claim 1, wherein said calculation means obtains said phase difference Δφ ₂ by the following calculation. Δφ ₂ = (Δφ ₀ + Δφ ₁ ) / 2 + arctan [{(A ₀ + A ₁ )
/ (A ₀ −A ₁ ) {tan {(Δφ ₀ −Δφ ₁ ) / 2}] 3. The alignment apparatus according to claim 1, wherein said reference grating is a diffraction grating mark formed on said mask, and said first and second light irradiation means are each said mask. And irradiating the diffraction grating mark on the photosensitive substrate with two light beams of the first wavelength and the two light beams of the second wavelength via the light emitting device. 4. The alignment apparatus according to claim 1, wherein the first and second wavelengths are different from the wavelength of the illumination light for exposure, respectively, and the first and second wavelengths are different from each other. An alignment apparatus, wherein an irradiation light control element for controlling chromatic aberration of the projection optical system with respect to the light beams of the first and second wavelengths is arranged near a Fourier transform surface with respect to a pattern surface.