JPH04199505A - Alignment apparatus - Google Patents

Abstract

PURPOSE:To correct a nominal error due to resist and enable highly accurate position detection by providing an optical system and storage and image sensing devices to detect a relative direction of a wafer and a mask and by controlling the position. CONSTITUTION:A relief-like mark M is formed on a wafer W, and a grating pattern G1 drawn with constant intervals in parallel to an axial direction whose position is to be detected is provided between a light source 18 and the wafer mark M. After one of the marks is aligned, its mark image is stored in a storage device 118. Then a stage 10 is driven and the pattern is aligned with a new mark image for another mark on the wafer W, and then the mark image is sensed by an image sensing device 114. Then the picked mark image and the mark image stored in the storage device 118 are used to identify an application state of resist applied on the wafer W, and based on the identification result, a nominal error of the mark due to the resist is corrected. Thus a nominal mark image due to the resist on the wafer mark M is corrected to realize highly accurate alignment.

Description

Detailed Description of the Invention (Industrial Field of Application) The present invention is for detecting the relative position of a wafer and a mask and controlling the position in a pattern alignment device, for example, an exposure device for manufacturing semiconductor ICs and LSIs. The present invention relates to an alignment device.

[Conventional technology]

Recently, the degree of integration of semiconductor ICs and LSIs has increased more and more, and the pattern dimensions of the elements have become on the order of submicrons.

For alignment of conventional exposure apparatuses, various one-dimensional pattern match detection methods are known.

FIG. 9 is a schematic diagram of a positioning apparatus using a one-dimensional pattern match detection method proposed by the present applicant in Japanese Patent Application No. 2-127005. In the same figure, 95 is a light source of illumination light, and after the light from the light source 95 changes its direction with a mirror 94, it passes through the illumination optical system 93 and changes its direction with a half mirror 96. 692 is an objective lens, which is a half mirror. The light that has passed through passes through the lens 92 and the mirror 9.
The direction is changed at step 1, and the wafer W passes through the exposure projection optical system 11.
irradiate. An alignment mark M is formed on the wafer, and the incident light illuminates the mark M and is reflected. The reflected light passes through the exposure optical system 11 again, changes its direction by a mirror 91, passes through an objective lens 92, a half mirror 96, an imaging optical system 97, and forms an image on an imaging device 98. 1st
FIG. 0 is a diagram illustrating a two-dimensional electrical signal obtained by the imaging device. In Figure 10(a), the wafer mark is M
It is o. The two-dimensional electrical signal is converted into an A/D converter 99
The signal is converted into a two-dimensional discrete digital signal string corresponding to the XY address of the pixel.

Reference numeral 910 in FIG. 9 is an image processing device, in which a predetermined two-dimensional window 100 containing the wafer mark image M' is set as shown in FIG. 10(a). 1
01 is the screen circumference. The positional relationship between the reticle and the device (including the imaging device) is already known. Therefore, position measurement is performed on coordinates with the upper left of this screen as the origin (reference), and then converted into practical coordinates. After that, pixel integration is performed within the window 100 in the X direction, and a discrete electrical signal string t corresponding to the light intensity is generated in the X direction as shown in FIG. 10(b).
Output (x). The integrated signal obtained in this way depends on the cross-sectional shape of the mark (FIG. 11(a)) and also depends on the resist thickness and the reflection and absorption coefficients of the resist and wafer as shown in FIGS. 11(b), (c). d) It will look like (e). However, in FIG. 11(a), 1100 is a cross section of the resist;
01 is a wafer cross section.

In order to perform pattern matching, it is necessary to prepare templates shown in FIGS. 12(a), (b), (c) and (d). Each of the curves used in FIGS. 12(a), (b), (c), and (d) was approximated by a parabola. As shown in FIG. 10(C), one-dimensional pattern matching is performed at each point using this template. The point with the highest degree of correlation is taken as the center of the mark, and the correlation function obtained for several pixels around the center of the mark is interpolated after increasing the resolution using an interpolation means, and the position is outputted to the alignment control device 911. The position of the wafer is controlled and aligned based on the calculated value.

[Problem to be solved by the invention]

The mark image on the wafer is distorted due to the uneven coating of the resist applied thereon and the asymmetry of the mark on the wafer. As a result, the measurement system measures an apparent mark image, causing so-called sagging, which degrades the accuracy of alignment.

In particular, in the process of applying a resist after forming a metal compound on the wafer surface by sputtering or vapor deposition in the semiconductor manufacturing process, errors caused by the above-mentioned apparent image due to uneven coating of the resist near the wafer marks become a serious problem. ing.

The cause of the apparent error in the metal compound process mentioned above is that due to the high reflectance of the metal compound, geometrical optical phenomena due to uneven coating of the resist are dominant compared to thin film interference and the wavelength dependence of the refractive index. Are known. It is also known that the unevenness of resist coating depends on the position of the wafer. Figure 7 shows this phenomenon. Figure 7(a) shows
It is a cross-sectional view of alignment marks on a wafer, where 71 is a wafer, 70 is a resist, and 72 is air. Also, 73.
74 are observation lights, respectively. As shown in the figure, the direction of refraction of the light beam changes significantly due to local changes in the resist surface. When the reflectance of the wafer is very high, the interference effect between the light reflected from the resist surface and the light reflected from the wafer surface may be considered to be sufficiently small compared to the effect due to geometrical optical refraction. FIG. 7(b) is a distribution diagram of the light intensity of observation light corresponding to FIG. 7(a). As the resist surface changes, the image undergoes a lateral shift (shift), and in the illustrated example, the width of the extreme value of the light intensity distribution is measured to be wider than the actual mark width. For example, like a spoon in water, the appearance of this apparent image cannot be removed by whitening the illumination light or by adjusting the wavelength of the illumination light. Therefore, it is desirable to predict the apparent amount due to the surface shape of the resist surface.

An object of the present invention is to provide an alignment device that corrects an apparent mark image based on a resist on a wafer mark and enables highly accurate alignment.

[Means and operations for solving the problem] In order to achieve the above object, the present invention provides the following positioning device. On the wafer, a device consisting of a rectangular relief mark with a short side in the direction of the axis where position detection is to be performed and a long side in the direction perpendicular to that direction is drawn at regular intervals parallel to the direction of the axis to be detected. A grid pattern is provided between the light source and the wafer mark. It also includes a storage device that temporarily stores the optical image and an imaging device.

The light rays emitted from the light source illuminate the surface with a grid,
An optical system is provided so that the grating plane and the wafer mark are optically conjugate, and the incidence is parallel to the grating pattern in the direction parallel to the grating pattern and obliquely incident to the wafer surface in the perpendicular direction.

On the wafer, a grid-like light distribution parallel to the direction in which position detection is performed is formed on the wafer mark. An optical system is provided that makes the wafer and the optical image storage device conjugate and forms specularly reflected light of a light distribution created by the illumination light from the wafer on the surface of the storage device. In this case, a phase conjugate optical crystal is used as an optical image storage device. At the same time, the wafer mark image is converted into an electrical signal in the imaging system.

forming a mark image of the wafer on a storage device and an imaging device using the optical system with respect to a reference mark on the wafer;
After positioning the image on the imaging device using electrical signals, the image is stored and used as a reference image. Thereafter, the position of the wafer is changed and a wafer mark image is obtained on the imaging device in the same manner as above. After positioning is performed using the electrical signals of the captured image,
By superimposing the stored optical M patterns, the difference in surface change of the resist within the wafer is investigated. The results are used to correct alignment errors caused by changes in the surface of the resist.

That is, a means is provided for measuring the undulations of the resist surface in the vicinity of the wafer mark, which cause distortion of the wafer mark image, and using the measurement results to correct the distortion of the wafer image.

Further, according to the present invention, by increasing the integration width, the one-dimensional pattern matching method described in the conventional method can be measured simultaneously with the measurement of the undulations near the marks on the resist surface.

[Examples] Hereinafter, the present invention will be described in detail based on illustrated examples. FIG. 1 shows an embodiment of a step-and-repeat type exposure apparatus having a wafer alignment device to which the present invention is applied. In the figure, W is a semiconductor wafer on which a position detection mark M1 and another type of position detection mark M2 are formed. 10 is a wafer stage that moves the evaporator in the x direction and the X direction; 11 is an exposure optical system;
The pattern formed on the top is exposed by the exposure illumination system 125.
It is reduced to one-fold and transferred onto a wafer. Here, the Z axis refers to the optical axis direction of the exposure optical system 125 (positive upward), and the x and y axes are perpendicular to each other and refer to a direction perpendicular to the Z axis.

Reference numeral 18 denotes a light source of observation light having a wavelength different from that of the exposure light, and the light is directed by a half mirror 19 and then illuminated by the illumination optical system 16 to illuminate the grating pattern G1. As shown in FIG. 2, the grating pattern G1 has a pitch λ1 and is surrounded by shielding plates. Reference numeral 14 denotes a correction optical system, which corrects aberrations of the exposure optical system caused by the difference in wavelength between observation light and exposure light, and makes the grating pattern and the wafer surface optically conjugate. 13 is a mirror that changes the direction of incidence. The direction of incidence on the wafer is as shown in Figure 4.
It is obliquely incident on the yz plane at an angle of θ, and is perpendicularly incident on the XZ plane. Reference numeral 41 indicates the incident direction, and reference numeral 42 indicates specularly reflected light. 43 indicates the same phase plane. Further, 40 is a wafer alignment mark Ml (or M2), which is arranged to perform alignment in the X direction. The grating mark G1 forms an image parallel to the same phase plane 43 of light in the X direction. FIG. 3(a) is a wafer image seen on the wafer. The light reflected on the wafer surface, such as the reflected light 42, is directed to the objective optical system 1 after changing its direction with the mirror 12 (FIG. 1).
5, the optical path is divided by a half mirror 110. One side is half mirror 1 by the imaging optical system 112.
13 and enters an imaging device 114 to form an image. The other is controlled by shutter 31 and mirror 111
After being reflected by the image forming optical system 115, the half mirror 1
16, an image is created on an image storage device 118. In this embodiment, an optical crystal, such as a TSO, which is a phase conjugate element is used as an image storage device. For detailed characteristics of TSO, see "Optics" Vol. 18, No. 3, P132-136.
(1989).

The optical crystal was used here for the sake of resolution. Semiconductor manufacturing equipment requires a highly accurate positioning device. In this case, the high-resolution optical crystal functions well as a storage device. S2. S3°S4 are shutters. 12
0 is an illumination optical system. 121 is a mirror, and 117 is an imaging optical system, which optically conjugates the optical crystal 118 and the imaging device 114. 119 is a mirror.

122 is an A/D conversion device, 123 is an image analysis device, 12
4 is a positioning control device.

The algorithm of this embodiment will be explained based on FIG. 5(a). In this embodiment, it is assumed that the reticle and the device (including the imaging device) have already been aligned on the order of several tens of 1 μm. Therefore, the alignment in this embodiment comes down to alignment between the imaging device and the wafer mark. Further, the operation of this embodiment is as shown in the flowchart of FIG. 14, and the final purpose is alignment at the wafer mark M2 and step-and-repeat type exposure in the vicinity thereof.

At time t1 in FIG. 5(a), the mark M1 on the wafer position Pi (FIG. 5(b)) is illuminated.

However, Pl is set at the center of the wafer. At this time, shutters 81 to $4 are all closed. Half mirror 11
Mark M separated by 0 and imaged on the imaging device 114
The image of No. 1 is shown as shown in FIG. 3(a). , 30 is an image of mark M1. Positioning is performed at time t1 by the one-dimensional pattern matching method described in the prior art.

That is, the wafer mark image on the imaging device is imaged, A/D converted, and a one-dimensional pattern is created for the one-dimensional data obtained after performing A/D conversion on the window 39 in FIG. 3(a) in the X direction. Ching is performed to calculate the center of the mark. The wafer stage is driven with respect to the center value of the wafer mark from the edge of the screen so that the center of the mark becomes the center of the screen of the imaging device. At this time, it is important that the center of the mark is not damaged by the resist. After driving the wafer stage 10 and confirming that the wafer mark image has come to the above position (t2), the shutter 31, S3
is opened, the light from the mirror 121 is used as a reference beam, the image is stored in the optical crystal 118 as a reference image, and the shutter S is opened.
l, close S3.

Here, we will introduce a simple principle of image storage using this optical crystal. The wafer image pattern with optical information, together with a coherent, same-wavelength reference beam controlled by shutter 53, creates an interference pattern on the optical crystal. Optical crystals absorb light energy depending on the light intensity distribution of interference fringes, change the refractive index, and store within the crystal a refractive index distribution that depends on the light intensity distribution. In other words, it has been memorized. Therefore, when the optical crystal is irradiated with the same reference light, a light intensity distribution, which is an interference pattern, is generated again, and as a result, the stored image can be reproduced based on the principle of homography. What is done at t2 is a memory action, and what is done at t4, which will be described later, is a regeneration action.

At time t3, the wafer stage is driven and the wafer is placed on the wafer. A mark M2 of the same type as the mark M1 on P2 (FIG. 5(b)) is illuminated. At this time, in order to form a draft pattern with a pitch different from G1 on the wafer, 01 is tilted as indicated by the dotted line Gl' in FIG. As a result, the pitch of the draft pattern becomes 1'. Similarly to the above, for M2,
One-dimensional pattern matching is performed by capturing a mark image,
The wafer stage is aligned mD with respect to the center of the wafer, and it is confirmed that the wafer mark image has come to the same predetermined position (center of the screen) as the previous wafer mark M1. Then, the center position of the mark M2 is measured using a conventional method and stored as x2 (this x2 will be corrected later). Then, at t4, the shutter 32. S3.
Open S4. Using the light that has passed through the mirror 121 as a reference light, a mark M1 is placed on the optical crystal as a reference image obtained at t2.
Play the statue. The image forming optical system 117 transfers the image formed on the optical crystal to the imaging optical system with the same light intensity. The image of the mark M2 is first formed as a test image on the imaging optical system, and the image of the mark M1 is superimposed thereon as described above. As a result, a moiré image 33 created by both optics is formed on the surface of the imaging device. This moire image is caused by changes in the surface of the resist painted on the pattern M1 and the mark M2.
It represents the difference in surface changes at positions P1 and P2 of the resist coated on top.

Here, moire fringes using moire topography will be briefly explained. For simplicity, let us assume that the magnification of the observation optical system is 1. The G1 image near mark M1 is simply expressed in the form. ρ. is the initial position.

ρ(x,y) is a modulation component and is a function depending on the undulations of the resist. At this point, the Gl'' image near the mark M2 can be written as: λ1'' is the pitch of Glo, po' is the initial phase, and p'(x, y) is the modulation component. The superposition of the two waveforms is ρ. =ρ. =C, and λ1 given λl°,
Since p (x, y) and ρ' (x, y) are relaxed functions, the superposition coefficient is set to 1-1. Further, the electrically captured intensity is J(x)=1 (within one pixel range) 1(x+t)dt depending on the number of pixels of the imaging device. The moiré fringes mentioned here are (λ1)2/(λ1-λ1
It is a component with a pitch around ). However, this depends on the difference between the known λ1 and λ1° and the modulation component. The modulation component depends on the difference in change in M1 and M2 of the resist undulations.

Next, a method of filtering the sum moiré image will be explained with reference to FIG. 13.

FIG. 13(a) shows the above-mentioned electric intensity J
(x) is shown. Now, take a moving average over a wide range (approximately 10 times λ1) for J(x).

Let the value of the moving average be D (X). FIG. 13(b) is a diagram of Ryo (X). Furthermore, if 了(x)=(J(x)−Ding(x))', then T
(x) is FIG. 13(c).

Figure 13(d) is the result obtained by applying a low-pass filter to this.
j (x). Figure 3(cl) versus Figure 3(cl)
13(d) corresponds to FIG. 13(C). Below, the electrical signal j (x) in Fig. 13(d)
deals with

Here, let us assume that the magnification is normal again.

In this way, it is possible to detect a change in the resist thickness of the mark M2 when the wafer mark M1 is used as a reference mark. The details are explained below. Here, since the light reflected from the wafer is sufficiently incoherent, interference between the two optical images can be ignored.

Next, a superimposed image is captured. When interference cannot be ignored due to light rays, images may be captured individually within the capture time of the elements of the imaging device. Alternatively, change the M2 image by driving the wafer stage in the The image may be captured at the point closest to the point.

The captured moire image (FIG. 3(b)) is quantized by the A/D converter 122 and sent to the image analyzer 123. In the image analysis device 123, a processing window 38 is set,
It is integrated in the Y-axis direction. The integrated data is shown in Figure 3 (C).
It is indicated by. Regarding (C), positioning is performed again by the one-dimensional pattern matching described in the prior art.

This is to provide the reference X coordinate, for example, the mark center position xO. Here, to shorten the time, xO is
You can also use it instead.

After alignment, the resist surface undulations near the marks are determined by the following method. A processing window of 34°, 35.36.37 is set around the mark center position xO aligned as described above, and integration is performed in the X-axis direction. At this time, the waveform is shaped by applying a high frequency cut filter in the y-axis direction using the method described above. The integrated original data is shown in FIG. 3 ((1)).

FIG. 3(e) is a result of further removing the high frequency components using the method described above. The processing window 37 is a region where the undulations of the resist surface sufficiently formed from the wafer mark can be ignored. The waveforms shown in FIGS. 6(a), 6(b), 6(c) and 6(d) are obtained by filtering high frequency cut on the waveforms obtained corresponding to each of the processing windows 34, 35, 36, and 37. The change in phase [0, 2π) in FIGS. 6(a), 6(b), and 6(c) with respect to the phase in FIG. 6(d) corresponding to the processing window 37 is the change in the resist coating unevenness difference in the X direction. . In reality, the processing window is divided more finely in the vicinity of XO, and the phase [0, 2π
] and the X direction. For example, it is determined using FFT (Fast Fourier Transform). The result is shown in FIG.

However, in FIG. 3(b), there is no data in the area corresponding to the mark edge portion 32.

Based on the information near the temporary center XO in Figure 3(c) of Figure 8, M
2 Correct the mark center position x2. ×2 coincides with xO or exists sufficiently close to XO. For example, as described in Fig. 7, sagging occurs due to the curvature and differential absolute value difference in the xz plane of the resist surface, so the measured values obtained as shown in Fig. 8 are interpolated with a spline function, and v (x
). V (X) has a value of [0,2π). Second
1.5 of the wafer mark line width around the temporary center x2 of
A quadratic function approximation is performed on the value of v (x) in a region within about twice that, and the quadratic coefficient is defined as g.

Also, take the differential absolute value of V(X) and set it as d(x). d(x) is shown in FIG. 8(e). After cutting the top and bottom of d(x) appropriately in a region within about 15 times the wafer mark line width around x2 (for example, 80% or more and 20% or less of the maximum and minimum values in the region) After cutting and removing), calculate the center of gravity and let it be dO, then dx=α・g/Igl−do・・・・・・・・・・・・
... (1) or the amount of damage. This Togima 2=x2-dx・・・・・・・・・・・・・・・・・・
・・・・・・・・・・・・・・・ Let (2) be the center position of the M2 mark with no damage.

However, α is a parameter determined experimentally, and x2 is a center value with dullness using the primary pattern matching described in the conventional example for the image of mark M2 using Figure S3 (d). The qualitative image of α is as follows. In FIG. 7, the incident light is limited to the vertical direction (optical axis direction) in FIG. The angle of resist coating unevenness θ<<1 (rad
)'s Toki...corresponds to (3). θ□ is the angle of incidence in FIG.

rT is the value of the pitch λ1 of the grating G1 in terms of the effective scale on the wafer. The amount of looseness in Fig. 7 is approximately. n is the refractive index of the resist. Δ1 is the resist thickness given in FIG. It is calculated like this for each point. However, in reality, it involves many parameters such as the effect of the wafer level difference Δ2 and the NA of the observation optical system, so it is more practical to determine it through experiments. The determining method is to determine dx based on optical information from the wafer marks after resist coating and before, and to calculate dx from the relationship of equation (1) using the above do after resist coating.

In general, if the mark shapes are similar and the devices are the same, Δ 2 α−α. ×(Δ++) −・・・・・
The relationship is as follows (6) (constant). Therefore, if α is determined under certain conditions, other Δ1. The case of Δ2 can also be easily calculated.

In this way, the true mark center TT of the wafer mark M2
will be determined. After converting this into a relative position with respect to the reticle, the wafer stage is driven to complete positioning and exposure. Moreover, in this way, the apparent error caused by the position-dependent resist on the wafer can be quantified and corrected. By utilizing this fact, for example, it can be used as a correction method when positional deviation is measured at several points on the wafer before step-and-repeat type exposure on the wafer, and alignment is performed after estimating the position of each shot on the entire wafer. be.

In this example, the light (T
TL), but it is not necessarily necessary to pass through a projection lens for exposure.

Further, in this embodiment, an optical crystal is used as a storage element, but a moiré image may be electrically formed after being stored as image data in an imaging device. The wafer mark is not limited to a rectangular shape, and may have any shape that can form a moiré image parallel to the direction in which the position is to be detected, for example. Furthermore, the angle of the wafer illumination light may be perpendicular.

〔Effect of the invention〕

As described above, the present invention is equipped with means for correcting apparent errors caused by the resist, thereby making it possible to remove unnecessary noise components, thereby enabling highly accurate position detection.

[Brief explanation of the drawing]

Fig. 1 is a schematic configuration diagram of a device according to an embodiment of the present invention, Fig. 2 is an explanatory diagram of a lattice used in an embodiment of the present invention, and Fig. 3 (a), (b), (c), (d), and (e) respectively. FIG. 4 is an explanatory diagram of image analysis of a striped pattern according to the present invention. FIG. 4 is an explanatory diagram of incident light on a wafer mark. FIGS. 6(a), (b), (c), and (d) are light intensity signal waveform diagrams at image processing windows at different positions, and FIG. 7(a) and (b) are 8(a), 8(b) and 8(c) are illustrations of the apparent error correction method, respectively, FIG. 9 is a schematic diagram of the conventional device, and FIG. 10(a)(b) 11(a), 11(b), 11(b), 11(c), 2(d), and 11(e) are waveform illustrations of the conventional mark image light intensity signal, respectively, and FIG. a)(b)
), (c) and (d) are explanatory diagrams of conventional templates, respectively. FIGS. 13(a), (b), (c) and (d) are explanatory diagrams of filtering of moiré fringes in the above embodiment, respectively. FIG. FIG. 3 is a flow diagram showing the operation of the embodiment. R, reticle, M, Ml, M2: mark, Sl, S2
．． S3. S4 shutter, G1/grid, 10:
XYZ-stage, 11: Exposure optical system, 12, 13: Mirror, 14: Correction optical system, 15, Objective optical system, 16 Illumination optical system, 18. Light source, 19. Half mirror-1120 illumination optical system, 121: mirror, 118: optical crystal, 119
．． mirror, 116: half mirror, 115: imaging optical system, 117: imaging optical system, 112: imaging optical system, 113:
Half mirror, 114: Imaging device, 122: A/D conversion device, 123: Image analysis device, 124; Control device, 12
5: exposure illumination system, 30: mark image, 31: striped pattern,
34 to 38, processing window, 32: shadow of mark image, 33: beat pattern, 41, incident light, 40: mark, 42: reflected light, 43: equal phase plane, 70 resist, 71: wafer, 72: air , 73, 74: Observation light.

Claims (6)

[Claims] (1) Relief-like marks are provided at two locations on the wafer, with the short side in the direction in which the position should be detected and the long side in the direction perpendicular to this, and a constant distance between the light source that illuminates the marks and the marks. A movable stage on which a parallel striped pattern is arranged at intervals and the wafer is mounted, and a stage for optically conjugate relationship between the pattern surface and the mark and for detecting the position of the direction of the same phase plane of light from the light source. a storage device that stores an optical image formed by the reflected light from the mark;
an imaging device for capturing an optical image using reflected light from the mark; an optical system that has an optically conjugate relationship with the storage device and the imaging device and guides the reflected light to the storage device and the imaging device;
The apparatus includes an image processing device connected to the imaging device, and a control device that drives the stage according to the processing result of the image processing device, and after aligning one mark, stores the mark image in the storage device. After the stage is driven and the stage is aligned with a new mark image corresponding to another mark on the wafer, the mark image is imaged by the imaging device, and the captured mark image and the mark image stored in the storage device are What is claimed is: 1. A positioning apparatus characterized by determining the coating state of a resist applied on a wafer surface using the above-described method, and correcting an apparent error of a mark caused by the resist based on the determination result. (2) The resist coating state is determined by superimposing a mark image captured by the imaging device and a mark image stored in a storage device to form a moire image. Alignment device described in section. (3) The two mark images are formed by changing the inclination angle of the pattern with respect to the optical axis of the illumination light to make the intervals between the striped patterns on the wafer surface different. The alignment device according to item 2. (4) The alignment device according to claim 1, wherein the incident angle of the illumination light on the wafer is inclined with respect to the wafer surface. (5) The alignment apparatus according to claim 1, wherein the incident angle of the illumination light on the wafer is perpendicular to the wafer surface. (6) The alignment device according to claim 1, wherein the storage device is made of a phase conjugate optical crystal.