JPH0744138B2 - Alignment device - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mask-substrate alignment device in a device for exposing a mask pattern onto a photosensitive substrate such as a semiconductor wafer.

(Background of the Invention) In recent years, a large number of reduction projection type exposure apparatuses (steppers) have been used for manufacturing semiconductor elements such as VLSI. This type of exposure apparatus projects a circuit pattern on a reticle (synonymous with a mask) onto a wafer via a projection lens. Generally, in order to obtain a high transfer resolution, the reticle is illuminated with monochromatic light, and the projection lens is chromatic aberration-corrected so that the aberration of the monochromatic light is minimized. Further, in order to improve the overlay accuracy of the reticle and the wafer, it is necessary to perform highly accurate alignment (alignment). Several methods of this alignment have been proposed and put into practical use. Among them, a so-called through-the-lens (TTL) type die-by-die alignment method for directly aligning a reticle and a wafer via a projection lens has been established. In this method, the alignment mark provided on the reticle and the back-projected image of the alignment mark attached to the exposed area on the wafer by the projection lens are simultaneously detected by the alignment microscope, and the amount of deviation between the two marks is obtained. The deviation between the projected image of the circuit pattern and the exposed region on the wafer is detected, and the reticle and the wafer are relatively finely moved so as to correct the deviation. In this case, when the reticle and the wafer are illuminated with monochromatic light for exposure, it is desirable to align the reticle and the wafer in a state where they are optically conjugate with respect to the projection lens, and then immediately expose the reticle. This means that the mark illumination light for detecting the alignment marks of the reticle and the wafer must have the same wavelength as the exposure monochromatic light. Further, since the photoresist layer is formed on the wafer, the micro area including the mark on the wafer is exposed when the mark is detected. This is inconvenient when the mark on the wafer is stored in consideration of the later process. Also, illuminate the wafer with light of a photosensitive wavelength,
When observing the bright-field image of the mark under the photoresist layer, unevenness in the thickness of the photoresist layer often causes interference fringes, resulting in erroneous detection of the mark.

Therefore, a method of irradiating the mark on the wafer with non-photosensitive light to detect the mark image has been considered. When the mark on the wafer is irradiated with the non-photosensitive light in this way, the mark image back-projected to the reticle side by the projection lens is conjugate with the pattern surface on which the reticle mark is formed due to the chromatic aberration of the projection lens. However, the image is formed at a shifted position.
This shift amount greatly exceeds the depth of focus of a high-magnification alignment microscope, and it is impossible to clearly observe the mark on the reticle and the mark on the wafer at the same time. Therefore, an optical system that corrects the optical path length corresponding to the chromatic aberration of the projection lens is provided in the projection optical path between the reticle mark and the wafer mark between the reticle and the projection lens, and the image of the wafer mark is displayed on the pattern surface of the reticle. Devices have been proposed for imaging. However, if such a correction optical system is provided between the reticle and the projection lens,
Due to the setting accuracy of the correction optical system, long-term drift, etc., there is a problem that the image of the wafer mark formed on the pattern surface of the reticle is shifted by a minute amount and the alignment accuracy is lowered.

In addition, the non-photosensitive light is used when the wafer is positioned above or below the position where the conjugate of the monochromatic light for exposure can be taken, by a certain amount (the chromatic aberration of the projection lens) without providing the correction optical system. Another method is proposed in which TTL alignment is performed, and after the alignment is completed, the wafer is returned to the original height position and exposed. However, in this method, it is necessary to move the wafer up and down at the time of alignment, and it is unavoidable that the alignment accuracy is lowered due to a mechanical error. In addition to the alignment time, the vertical movement time of the wafer is required, so that the overall time for exposure processing of the wafer becomes long and the throughput is lowered.

Further, in recent years, an X-ray exposure apparatus has been developed to transfer a finer pattern. In this apparatus, the mask and the wafer are opposed to each other with a certain proximity gap, and the mask is irradiated with soft X-rays. The target system, which is often used as an X-ray source, apparently works as a point light source, so that the soft X-ray has a certain spread angle. Therefore, it is necessary to bring the mask and the wafer as close as possible (about 10 to 30 μm) to keep the gap accurately.
Because of this limitation, synchrotron orbit synchrotron radiation (SOR), which generates soft X-rays with good parallelism, is attracting attention. The soft X-rays from the SOR have very good parallelism, and the accuracy of the proximity gap between the mask and the wafer is considerably rough, and the gap itself is 50-300μ.
It can be about m. It is clear that, due to the device configuration, the larger the gap, the easier various controls (mask / wafer handling, stage driving, etc.) and the throughput improvement. However, when trying to observe the mark on the mask and the mark on the wafer at the same time in order to align the mask and the wafer, only the mark on the mask or on the wafer is observed because the depth of focus is small in a normal alignment microscope. There is a problem that you can not do it. Of course, it is conceivable to reduce the numerical aperture (NA) of the microscope objective lens to increase the depth of focus, but this reduces the resolution of the mark image,
A major drawback arises in that the alignment accuracy is extremely reduced.

(Object of the Invention) It is an object of the present invention to solve these drawbacks and to obtain a positioning device capable of preventing a decrease in alignment accuracy due to an aberration or the like of a projection optical system.

(Outline of the Invention) The present invention is back projected through the projection optical system (1) with the image of the first marks (RM ₁ , RM ₂ ) formed on the original plate (R) serving as a reference for alignment. Has a detection system (9, 10) for optically detecting the image of the second mark (WM) on the wafer (W), and detects the relative displacement between the first mark and the second mark. In the alignment device, the portion of the original plate on which the first mark is formed is the light of the first wavelength (IL
_First illumination means (6; 60, 61) configured so that the light illuminating the portion is not directed to the wafer through the projection optical system while being illuminated by 1), and the second mark on the wafer is formed. Second illumination means (8; 80) for illuminating the illuminated portion with light (IL ₂ ) having a second wavelength different from the first wavelength emitted from a light source (7) different from the light source of the first illumination means. , 81), and the projection optical system (1) receives light of the second wavelength (BW) generated from the second mark on the wafer, and the light of the second wavelength is applied to the first mark on the original plate. First wavelength light (BR) generated from
The image of the second mark (WM ') is formed on a surface (FP) different from the original plate once without overlapping with the original plate, and the detection system detects the first wavelength generated from the first mark on the original plate. Of the first mark (RM ₁ ′, RM ₂ ′) is formed on the predetermined surface (DP) by the incident light (BR) and the second wavelength is formed on the surface (FP) different from the original plate. Of the first mark so that the second mark image (WM ′) is formed on a portion of the predetermined plane which does not overlap the first mark image (RM ₁ ′, RM ₂ ′). It has an image forming optical system (10; 30, 31) in which the aberration generated corresponding to the difference between the wavelength and the second wavelength is adjusted, and further, the image of the first mark and the image formed in the predetermined plane are Image pickup means (11) for picking up an image of two marks and outputting an image signal corresponding to each image
And the technical point is to provide an image processing circuit (12) for detecting the relative displacement between the original plate and the wafer based on the image signal.

(Embodiment) FIG. 1 is a view showing the schematic arrangement of a projection exposure apparatus according to the first embodiment of the present invention. Reticle R pattern surface
A mark RM for alignment and a circuit pattern are formed on the PS, the mark RM itself is a light-shielding thin film such as chrome, and a minute region around the mark RM is a transparent portion.
The image of the circuit pattern and the mark RM formed on the reticle R is projected onto the wafer W via the projection lens 1. On the wafer W, alignment marks WM that can be aligned with the marks RM are formed as projections or depressions, and the entire surface thereof is coated with a photoresist layer. The wafer W is held on a two-dimensionally movable stage 2, the stage 2 is moved by a drive controller 3, and its position is detected by a length measuring device 4 such as a laser interferometer. Here, the pattern surface PS of the reticle R and the surface of the wafer W are arranged so as to be conjugate with the projection lens 1 when the entire reticle R is illuminated by exposure illumination light, for example, g-line light having a wavelength of 436 nm. Be present. The projection lens 1 is made so that various aberrations with respect to g-ray light are minimized. Therefore, when the reticle R and the wafer W are conjugate with each other by illuminating the g-ray light, when the reticle R is illuminated with another wavelength, for example, e-ray light or d-ray light,
Due to the chromatic aberration of the projection lens 1, the conjugate relationship is broken.

Now, the argon (Ar) laser 5 as the first illumination light source
The illumination light IL _{1 from} is made into a parallel light flux by the lenses 6a and 6b, and irradiates the formation area of the mark RM of the reticle R obliquely from the opposite directions. The illumination light IL ₁ is set to a wavelength (for example, 514.5 nm) that makes it difficult for the photoresist layer to be exposed to light. Further, when the mark RM has a linear elongated pattern, the illumination light IL ₁ is determined so as to enter from a direction perpendicular to the edge in the longitudinal direction. On the other hand, the illumination light IL ₂ from the helium-neon (He-Ne) laser 7 as the second illumination light source is collimated into parallel light flux by the lenses 8a and 8b, and the formation area of the mark WM on the wafer W (accurately The projection area of the mark RM) is irradiated obliquely from the opposite directions.
When the illumination light IL ₂ is set to have a wavelength that makes it difficult for the photoresist layer to be exposed to light and a wavelength (for example, 632.8 nm) different from that of the illumination light IL _1, and the mark WM is a linear elongated pattern parallel to the mark RM, the illumination light is illuminated. The light IL ₂ is defined so as to be incident from a direction perpendicular to the longitudinal edge of the pattern. The oblique illumination of the reticle R or the wafer W with the illumination light IL ₁ or IL ₂ is for obtaining the dark field image of the mark RM or the mark WM. This dark field image is formed by the mirror 9 arranged above the mark RM of the reticle R and the objective lens 10 as the optical lens system of the present invention. Mirror 9
The optical axis of the objective lens 10 is bent so as to be perpendicular to the pattern surface PS of the reticle R. Then, the objective lens 10 is set so that a predetermined amount of chromatic aberration is generated between both wavelengths of the illumination lights IL ₁ and IL ₂ . That is, if the image light flux of the light from the mark WM on the wafer W that passes through the projection lens 1 and is directed to the mark RM of the reticle R is BW, this image light flux
Since BW is a wavelength different from the chromatic aberration-corrected wavelength (g-line) of the projection lens 1, the back-projected image of the mark WM by the projection lens 1 is on the surface FP separated from the pattern surface PS of the reticle R by a certain distance. Image on. The distance between the surface FP and the pattern surface PS corresponds to the amount of chromatic aberration in the illumination light IL ₂ of the projection lens 1 (the amount of deviation of the image plane in the optical axis direction).

Therefore, the amount of chromatic aberration between the illumination lights IL ₁ and IL ₂ of the objective lens 10 is made to be substantially equal to the distance between the surface FP and the pattern surface PS. In such an objective lens 10,
When the image light beam BR from the mark RM of the reticle R and the image light beam BW from the mark WM are incident, the marks on the exit side are in the same plane.
Both images of RM and mark WM (dark field image) will be formed. Therefore, the image sensor 11 such as an industrial television (ITV) camera is arranged so that the light receiving surface is located on the same surface, and the mark R
An image signal corresponding to the lightness and darkness of the images of M and WM is digitized through the image processing circuit 12 and read into the waveform memory 13. Then, the image processing circuit 12 detects the relative shift amount on the image surface of the mark RM and the mark WM based on the waveform data stored in the waveform memory 13, and outputs the shift amount information to the main calculation control circuit 14. send. Based on the current position information of the stage 2 and the displacement amount information from the length measuring device 4, the main arithmetic control circuit 14 causes the drive control unit 3 to perform a predetermined movement so as to finely move the stage 2 so that the displacement becomes zero. Generate a command. As a result, the projected image of the circuit pattern on the reticle R to be exposed and the area to be exposed on the wafer W exactly match.

Now, FIG. 2 is an enlarged view showing the arrangement of the objective lens 10, the mirror 9, the mark RM and the like in an enlarged manner. Practically, it is desirable that the mark RM be two parallel linear marks RM ₁ and RM ₂ . This is because it is possible to align the mark WM on the wafer W by pinching it in the middle of the marks RM ₁ and RM ₂ . When the optical axis of the objective lens 10 is AX, the optical axis AX is perpendicular to the pattern surface PS of the reticle R and is also perpendicular to the surface FP. Now, assuming that the aerial image WMi of the mark WM on the wafer is formed on the optical axis AX, the aerial image WMi is re-formed on the exit side surface DP of the objective lens 10 at the position where the optical axis AX passes. It is imaged and formed as a dark field image WM '.
The images of the marks RM ₁ and RM ₂ are dark field images RM ₁ ′ and R on the surface DP.
Formed as M ₂ ′. The objective lens 10 is an aerial image WMi.
And marks RM ₁ and RM ₂ are magnified at high magnification to obtain images WM ′, RM ₁ ′ and R
M ₂ ′ is formed, and preferably both the object side (reticle R side) and the image side (plane DP side) are telecentric systems.

Now, on the surface DP, three dark field images W of marks WM, RM ₁ and RM ₂
Since M ′, RM ₁ ′ and RM ₂ ′ are arranged in parallel, the image sensor 11
The third scanning line is photoelectrically detected so that the scanning lines are almost perpendicular to these dark field images WM ′, RM ₁ ′, RM ₂ ′.
An image signal having a waveform as shown is obtained. In FIG. 3, the horizontal axis represents the scanning position x on the scanning line, and the vertical axis represents the intensity I of the image signal. Peak P ₁ and dark field image R due to dark field image RM ₁ ′
The interval between the peak P ₂ and M ₂ ′ is constant. However, the generation positions x ₁ and x ₂ of the peaks P ₁ and P ₂ may shift in the x direction on the scanning line depending on the alignment accuracy of the reticle R with respect to the apparatus main body. Therefore, the image processing circuit 12 uses the dark field image W
The peaks P ₁ and P ₂ are detected together with the peak P ₃ due to M ′, and the positions x ₃ , x ₁ and x ₂ of each peak are obtained. Then the distance d ₁ position x ₁ and x _3, and calculates the distance d ₂ of the position x ₂ and x _3, the difference (d
−d ₂ ) is calculated. This difference corresponds to the amount of deviation between the reticle R and the wafer W.

Now, in the above embodiment, the illumination light for illuminating the reticle R
IL ₁ is light (green) having a wavelength of 514.5 nm from the argon laser 5, but some photoresists have some sensitivity to this wavelength. Therefore, a krypton laser with a slightly longer wavelength should be used instead of the argon laser. Light from a krypton laser (wavelength 530.9n
m) has a short wavelength for He-Ne laser light (red). Alternatively, krypton lasers include light with a wavelength of 647.1 nm, which is slightly longer than He-Ne.

Next, a second embodiment of the present invention will be described with reference to FIGS. FIG. 4 shows the overall arrangement of the optical system,
The drawing shows a specific arrangement of the objective lens 10 and the mark RM of the reticle R, and is a partially enlarged view of FIG. This example is 2
The two illumination lights IL ₁ and IL ₂ and the objective lens 10 are coaxial, and the marks
It detects and aligns the bright field images of RM and WM. In FIG. 4, the illumination light IL from the argon laser 5
₁ is collimated by an optical system 60 such as a beam expander, reflected by a mirror 61, and then half mirrors 81, 90.
Incident on the pattern surface PS of the reticle R perpendicularly. Also, the illumination light IL from the helium / neon laser 7
₂ is a convergent light with the beam diameter expanded by the optical system 80, and after being reflected by the half mirror 81, the illumination light
It passes through the half mirror 90 coaxially with IL ₁ and enters the projection lens 1 through the transparent portion of the reticle R. The illumination light IL ₂ becomes a parallel light flux on the exit side (wafer W side) of the projection lens 1 and uniformly illuminates the mark WM. The converged illumination light IL ₂ incident on the reticle R is set so as not to irradiate the marks RM ₁ and RM ₂ . The projection lens 1 has a telecentric system on both the object side and the image side.

Now, of the light generated from the mark WM by the illumination light IL ₂ , the image light beam BW which is incident on the projection lens backward and advances toward the mark RM of the reticle R is the pattern surface of the reticle R as in the first embodiment. The image is formed at a position away from the PS. This image light beam BW is reflected by the half mirror 90 and enters the objective lens 10, and an image (bright field image) WM ′ of the mark WM is formed on the surface DP. Also in FIG. 4, the reticle R and the wafer W are
When the illumination light for exposure (g-line or the like) is used, it is arranged so as to be conjugate with the projection lens 1.

Here, the relationship between the respective light rays will be described in detail with reference to FIG. First, the illumination light IL ₂ passes through an intermediate portion between the _two marks RM ₁ and RM ₂ and converges (images) on the surface QP which is separated from the pattern surface PS by a certain distance. The surface QP is the position of the focus (f) of the projection lens 1 at the wavelength of the illumination light IL ₂ . Therefore, the illumination light IL ₂ that reaches the wafer W through the projection lens 1 becomes parallel light. The spot shape of the illumination light IL ₂ formed on the surface QP may be a simple circle, or may be a sheet shape (ellipse having an extremely large major axis to minor axis ratio) that matches the shape of the mark WM.

The back projection image of the mark WM by the projection lens 1 is formed on the surface FP, and the image light beam BW is reflected by the half mirror 90 and enters the objective lens 10. Furthermore, the image rays BR from the marks RM ₁ and RM ₂
Is also reflected by the half mirror 90 and enters the objective lens 10.
Then, depending on the amount of chromatic aberration of the objective lens 10, the marks RM ₁ , R
The formation of the images of M ₂ and WM on the surface DP is exactly the same as in the first embodiment. Incidentally, as is clear from FIG.
After passing through the reticle R, the illumination light IL ₁ of the reticle R enters the projection lens 1 and irradiates the surface of the wafer W. At this time, since the illumination light IL ₁ has a wavelength different from that of the light for exposure, the images of the marks RM ₁ and RM ₂ are not formed on the wafer W, and the image light rays are reflected by the wafer W and are projected onto the projection lens 1. Even if it proceeds in the opposite direction and enters the objective lens 10 through the half mirror 90, it does not form an image on the surface DP due to the chromatic aberration of both the projection lens 1 and the objective lens 10. The illumination light IL ₁ becomes possible to illuminate the mark WM on the wafer W, the illumination light IL ₁ mark WM
Is formed on the reticle R side. However, since the image forming surface can be located at a position different from both the pattern surface PS and the surface FP, the image of the illumination light IL ₁ of the mark WM should not be formed on the surface DP (largely defocused) after all. become.

Next, a third embodiment of the present invention will be described with reference to FIG. The present embodiment is the same as the second embodiment shown in FIGS. 4 and 5 except that the observation system side of the mark has a dark field. In FIG. 6, the lens system 30 functions as a so-called Fourier exchange lens, the lens system 31 functions as an inverse Fourier transform lens, the spatial filter 20 is arranged in the Fourier plane with respect to the surface (plane FP) of the wafer W, and the spatial filter 21 is the reticle R. The pattern plane PS is arranged on the Fourier plane. Each of the spatial filters 20 and 21 blocks the O-th order light, and as a result, the image WM ′ of the mark WM and the images RM ₁ ′ and RM ₂ ′ of the marks RM ₁ and RM _2.
Is imaged as a dark field image on the surface DP. The image signal obtained in this case has substantially the same waveform as that shown in FIG. Generally, a system including a Fourier transform lens, a spatial filter arranged on the Fourier plane, and an inverse Fourier transform lens is called a re-diffraction optical system. In this embodiment,
This re-diffraction optical system is the optical lens system of the present invention, and has a predetermined amount of chromatic aberration between both wavelengths of the illumination light IL ₁ and IL ₂ .
Since the illumination lights IL ₁ and IL ₂ have different wavelengths, the Fourier planes for the lights of the respective wavelengths can be separated according to the aberration. For this reason, two spatial filters 20 and 21 are required. However, when the amount of aberration is small, the two Fourier planes come close to each other, so one spatial filter may be used for both.

Next, a fourth embodiment of the present invention will be described with reference to FIG. The present embodiment is an alignment apparatus suitable for a proximity type exposure apparatus in which the mask M and the wafer W are opposed to each other with a minute gap g of about ten to several hundreds of microns. In FIG. 7, the illumination light IL ₁ from the light source 5 for illuminating the mask M is transmitted through the lens 53 and the half mirrors 51 and 50, enters the objective lens 10a, is reflected by the mirror 9, and is a mask M (wafer W). Irradiate. The illumination light from the light source 7
IL ₂ (light having a wavelength different from that of the illumination light IL ₁ ) is reflected by the half mirror 51 through the lens 52, passes through the half mirror 50 coaxially with the illumination light IL ₁ , enters the objective lens 10 a, and is reflected by the mirror 9. The wafer W (mask M) is reflected and irradiated. Illumination light IL
Image beam BM from mark MM on mask M by irradiation of ₁
Is a half mirror 50 through the mirror 9 and the objective lens 10a.
Is reflected by, enters the lens system 10b, and forms an image on the surface DP.
Similarly, the image light beam BW from the mark WM on the wafer W due to the irradiation of the illumination light IL ₂ is reflected by the half mirror 50 via the mirror 9 and the objective lens 10a, enters the lens system 10b, and enters the surface DP.
Image on. An afocal system is provided between the objective lens 10a and the lens system 10b, and the two lenses 10a and 10b constitute an optical lens system of the present invention. In the present embodiment, the chromatic aberration amount generated by both the objective lens 10a and the lens system 10b between the wavelengths of the illumination light IL ₁ and IL ₂ is
It is set to be equal to a predetermined gap g between the mask M and the wafer W. Therefore, if the numerical aperture (NA) of the objective lens 10a is increased and the depth of focus is made sufficiently smaller than the gap g, both the images of the marks MM and WM are formed on the surface DP. By disposing an image pickup device such as an ITV or a scanning slit light receiving device on the surface DP, the positional deviation between the mask M and the wafer W can be detected.

As described above, in each embodiment of the present invention, the light receiving surface of the photoelectric detection means is arranged on the common image plane DP formed by the objective lens 10, the re-diffraction optical system, or the pair of the objective lens 10a and the lens system 10b. However, a surface conjugate with the surface DP may be formed using a relay lens system that takes chromatic aberration into consideration, and the light receiving surface may be disposed there.

Further, the illumination lights IL ₁ and IL ₂ are merely uniform illumination lights, but may be high-brightness spot lights that are focused on the reticle (mask) and the wafer, respectively. In this case, a spot light scanning system (vibration mirror, polygon mirror, etc.) is provided in the illumination means, and the mark RM (M) on the reticle R (mark M) is provided.
Even if the scattered light from M) and the scattered light from the mark WM on the wafer W are condensed on the same plane through an optical lens system having a predetermined amount of chromatic aberration and a light receiving element is arranged there. The effect of is obtained.

Although the wafer W is exemplified as the substrate for projecting the pattern of the reticle or the mask, the same applies to other fixed reference mark plates on the stage 2 used for baseline measurement of the alignment microscope.

(Effect of the Invention) As described above, according to the present invention, the projection optical system enters the light of the second wavelength generated from the second mark on the wafer to form the image of the second mark on a surface different from the original plate. Then, the light of the first wavelength generated from the first mark on the original plate is made incident, and the image of the first mark is formed on a predetermined surface. Light is incident on the first surface within a predetermined plane
Since the image forming optical system in which the aberration generated corresponding to the difference between the first wavelength and the second wavelength is adjusted is provided so as to form the image of the second mark on the portion that does not overlap the image of the mark, Since there is no need to mechanically move the original plate or Ukaha and the numerical aperture (NA) of the alignment optical system is increased to increase the resolution of the mark image, the S / N of the signal during photoelectric detection is increased.
Ratio (especially rise and fall at mark edge)
Is improved, and highly accurate alignment is possible.

[Brief description of drawings]

FIG. 1 is a plan view showing a schematic structure of a projection type exposure apparatus according to a first embodiment of the present invention, FIG. 2 is an enlarged view showing a structure of an alignment optical system portion in FIG. 1, and FIG. FIG. 4 is a waveform diagram showing an example of the waveform of an image signal, FIG. 4 is a diagram showing a schematic configuration of a projection type exposure apparatus according to a second embodiment of the present invention, and FIG. 5 is a view of the alignment optical system portion in FIG. FIG. 6 is an enlarged view showing a configuration, FIG. 6 is an enlarged view showing a configuration of an alignment optical system according to a third embodiment of the present invention, and FIG. 7 is an enlarged view showing a configuration of an alignment optical system according to a fourth embodiment of the present invention. It is a figure. [Description of symbols of main parts] 1 ... Projection lens, 2 ... Stage, 5 ... First illumination light source, 7 ... Second illumination light source, 10, 10a ... Objective lens,
IL ₁ , IL ₂ ...... Illumination light, BW, BR, BM image rays, DP ... common image plane, W ... wafer, R ... reticle, M ... mask, RM, WM, MM ... mark

Claims (3)

[Claims] 1. A detection for optically detecting an image of a first mark formed on an original plate which serves as a reference for alignment and an image of a second mark on a wafer which is back projected through a projection optical system. In a positioning device having a system for detecting a relative positional deviation between the first mark and the second mark, a portion of the original plate on which the first mark is formed is irradiated with light of a first wavelength from a light source. First illuminating means configured to illuminate and not to direct light illuminating the portion toward the wafer through the projection optical system; and a portion on the wafer where a second mark is formed, Second illuminating means for illuminating with a light of a second wavelength different from the first wavelength emitted from a light source different from the light source of the first illuminating means; and the projection optical system from a second mark on the wafer. When the generated light of the second wavelength is incident, the light of the second wavelength is The light of the first wavelength generated from the first mark on the original plate is guided to the detection system without overlapping, and the image of the second mark is formed on a surface different from that of the original plate; The light of the first wavelength generated from the first mark on the original plate is made incident to form an image of the first mark on a predetermined surface, and the light of the second wavelength formed on a surface different from the original plate is formed. Aberration that occurs corresponding to the difference between the first wavelength and the second wavelength so that the image of the second mark is formed on a portion of the predetermined surface that does not overlap the image of the first mark upon incidence. An image forming optical system in which the image of the first mark and the image of the second mark formed on the predetermined surface are captured and an image signal is output according to each image. And detecting a relative positional deviation between the original plate and the wafer based on the image signal. Aligning apparatus characterized by comprising an image processing circuit. 2. The apparatus according to claim 1, wherein the first wavelength is set shorter than the second wavelength. 3. The apparatus according to claim 1 or 2, wherein the light source of the second illuminating means is a He-Ne laser light source.