JPS63128205A - Reduced projection type alignment system - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a reduction projection alignment method for aligning a circuit pattern on a reticle when exposing the circuit pattern onto a wafer through a reduction projection lens.

[Conventional technology]

As the miniaturization of semiconductor integrated circuits progresses, the alignment precision between a reticle and a wafer during exposure using a reduction projection exposure apparatus is required to be more precise. Therefore, through-the-lens (TTL) technology is used, which uses a reduction projection lens that can perform alignment for each chip and deal with errors in the arrangement of chips within a wafer.
The e-Lens alignment method is becoming mainstream in the production of highly integrated circuits in the future.

FIG. 16 shows an example of the TTL alignment method. The circuit pattern on the reticle 1 is exposed onto the wafer 3 one to several chips at a time through a reduction projection lens 2 . 4 is a wafer stage, and 16 is a chip. first,
The reticle alignment optical system 5, 5' detects the position of the reticle initial setting pattern 15, 15' and sets the reticle 1 at the initial position. Next, the alignment pattern 14.14' on the wafer is imaged onto the alignment pattern (window pattern) 13.13' on the reticle 1 through the reduction projection lens 2, and both patterns are detected by the wafer alignment detection optical system. .

The wafer alignment detection optical system includes a mirror 6°6', a relay lens 7.7', magnifying lenses 8, 8', a movable slit 9.9', photomultiplier tubes 10, 10', and an alignment unit with the same wavelength as the exposure light. It consists of an optical fiber 11.11' that emits illumination light. If the detected positions of the wafer alignment pattern 14° 14' and the reticle alignment pattern 13° 13' do not match, move the wafer stage 4 on which the wafer 3 is mounted in the X and Y directions so that both patterns 14 .Match the positions of 14' and 13.13'. When alignment 1~ is completed in this way, the exposure system 12 irradiates exposure light. A related alignment method of this type is JP-A-55-41739.

[Problem that the invention seeks to solve]

In the above TTL alignment method, a problem that has been pointed out in the past but still remains unsolved is a decrease in alignment accuracy due to uneven coating of photoresist on the wafer. This problem has become extremely serious in recent years as semiconductor integrated circuits have become smaller and the required alignment precision has become higher.

As shown in FIG. 17, the photoresist is coated by rotating the wafer 3 at high speed (in the R direction) using a spin coater (rotary coating machine).
Due to the centrifugal force, the resist is applied to a thickness of 1 to 2 μm (in the case of a single layer resist) over the entire surface of the wafer. Therefore, the direction in which the photoresist flows is the direction in which it spreads radially from the center of the wafer as shown by arrows A, B, C, and D. FIG. 18 is an enlarged view of the r-direction alignment pattern 19 of the chip 17.

Usually, the alignment pattern is composed of a concave or convex step pattern as shown in FIG. 18, and the thickness of the photoresist 23 applied thereon draws a gentle curve depending on the shape of the step. In addition, here 21 is S↓ board,
22 is the SJO□ layer. Now, as shown in FIG. 17, the unidirectional alignment pattern 19 of the chip 17 is parallel to the flow direction B of the resist, but the directional alignment pattern 20 of the chip 18 is perpendicular to the flow direction C of the resist. . As a result, the photoresist film thickness distribution in the pattern position detection direction near the alignment pattern 19 becomes symmetrical as shown in FIG. The film thickness distribution becomes asymmetrical because the flow of the photoresist 1- is disturbed at the step portion of the pattern. In addition, the same figure (ct
) and (b), B and C indicate the direction in which the resist flows. The alignment patterns 19 and 20 are illuminated by pattern illumination lights 24tL and 25oL, but the reflected light from the patterns is approximately the same as the reflected light 24b, 25b from the photoresist surface and the layer surface 2 surface of the SL substrate 21 or S=O□ layer 22. It is obtained as interference light with reflected lights 24G and 25G. Therefore, as shown in FIG. 21, the intensity of the interference light changes periodically depending on the thickness of the photoresist 23. As a result, the intensity distribution of the reflected light from the alignment pattern 19 is symmetrical as shown in FIG. 20(C), but the intensity distribution of the reflected light from the alignment pattern 20 is asymmetrical as shown in FIG. 20(b). becomes. Therefore, in the conventional alignment system 1, which utilizes the symmetry of the detection signal waveform and uses the center of symmetry of the waveform as the center position of the alignment pattern, as shown in FIG. , ) cd was regarded as the pattern center position, resulting in an error e and a decrease in alignment accuracy. Furthermore, in the TTL alignment method, it has been pointed out that not only coating unevenness but also the problem that the contrast of the pattern detection signal changes greatly depending on the coating thickness of the photoresist has been pointed out.

In view of the above-mentioned problems of the prior art, it is an object of the present invention to provide a reduction projection type alignment that eliminates the deterioration in alignment accuracy caused by the thickness and unevenness of photoresist coating, and enables stable and highly accurate alignment of a reticle and a wafer. The goal is to provide a method.

[Means for solving problems]

In order to achieve the above object, the present invention first focused on the reflectance at the interface between the photoresist coated on the wafer and air, and the reflectance at the interface between the photoresist and the step pattern constituting the wafer alignment pattern. In other words, when considering the step pattern of SL as an alignment pattern, the reflectance at the photoresist/air interface is usually about 5.5% (air refractive index 1, O, photoresist refractive index 1.61 (wavelength 514 nm)). In contrast, the reflectance at the photoresist alignment pattern (step pattern) interface is 24% (S- refractive index 4.75 (wavelength 514 nm))
shows a high value. Therefore, we came up with the invention of the optical system shown in FIG. In the figure, illumination light 41 from a coherent light source 40 such as a laser light source is transmitted by a beam splitter 42.
Separated in two directions.

One side illuminates the alignment pattern 48 on the wafer through the reduction projection lens 2 as alignment illumination light 43 . The alignment pattern 48 is formed of a 5=45 step pattern, and a photoresist 46 is applied thereon. The other beam enters a semi-transparent mirror 51 (transmittance O in the optical system shown in FIG. 1). Now, beam splitter 42
If the optical path length Qr from the to the semi-transparent mirror 51 is made almost equal to the optical path length from the beam splitter 42 to the surface of the photoresist 46, and the optical path length difference is set within the coherence distance of the coherent light source 40, the photoresist 46 Step pattern of reflected light 49, s, z 45 from the surface and photoresist 4
The reflected light 50 from the interface 47 with the mirror 6 and the reflected light 52 from the semi-transparent mirror 51 interfere with each other at the beam splitter 42.

Interference light 53 is obtained. The details will be described in "Operation" on the next page, but as shown in FIGS. The striped patterns 95c and 9 are formed according to the relative inclination of the wavefront of the reflected light 52 in the optical axis direction.
It becomes 5b.

Note that the arrows indicate the direction of inclination of both wavefronts. Here, the striped patterns 95cL and 95b are mainly formed by the step pattern and the photoresist 4, considering the reflectance of each interface.
The influence of interference between the reflected light 50 from the interface 47 with the alignment pattern 48 and the reflected light from the semi-transparent mirror 51 is the strongest, and a large amount of interference occurs due to the phase change by the step depth a at the stepped portion of the alignment pattern 48. A signal change is obtained. Therefore,
The influence of interference caused by the film thickness d (z) of the photoresist 46 is relatively small. Therefore, the present invention detects the striped patterns 95cL and 95b using an alignment optical system (not shown) equipped with a magnifying lens and an imaging device, and extracts the signal change position of the striped pattern corresponding to the stepped portion of the alignment pattern 48. Then, the center position of the alignment pattern 48 is found, and the amount of alignment is found from the center position of the separately detected reticle alignment pattern, and relative alignment is performed to achieve the above object. In addition, FIG. 2 shows alignment pattern 4.
8 is an enlarged view showing illumination light 43 consisting of a parallel beam incident on FIG.

In addition, in the present invention, a semitransparent mirror is a mirror in a broad sense,
A case where the transmittance is 0 as in the optical system shown in FIG. 1 is also included. This semi-transparent mirror can be provided outside the on-wafer alignment pattern illumination optical path that includes the reduction projection lens, or can be provided within the illumination optical path.

In addition, in the present invention, in the optical path provided with a semi-transparent mirror,
A wavefront correction optical system shall be provided.

Further, in the present invention, the coherent light used as the alignment pattern illumination light is temporally and spatially coherent.

[Effect]

In the optical system shown in FIG. 1, the intensity of the reflected light 49 from the surface of the photoresist 46 is I□, the intensity of the reflected light 50 from the interface 47 between the step pattern of SJ: 45 and the photoresist 46 is 2, and the semi-transparent mirror 51 If the intensity of the reflected light 52 from the
χ) is approximately determined by equation (1).

However, input: wavelength of alignment illumination light 7ZtL = refractive index of air nr = refractive index of photoresist The first term gives the DC component of the interference intensity, and the second term gives the interference intensity of the reflected light 49 and the reflected light 50. The third term gives a modulation component of the interference intensity between the reflected light 49 and the reflected light 52, and the fourth term gives a modulation component of the interference intensity between the reflected light 50 and the reflected light 52. Here, the intensity of the illumination light 41 is 1, and the ratio of reflectance to transmittance of the beam splitter 42 is 1:1.
, 71L=1, nr=1.61 (input=436nm
), the reflectance of the semi-transparent mirror 51 is 95%, the photoresist 46
・If the reflectance of the air interface is 5.5% and the reflectance of the stepped pattern/photoresist 46 interface 47 is 24%, (1)
The first term of the equation becomes a straight line shown at 60 in FIG. 3, the second term becomes a curved line shown at 61, the third term at 62, and the fourth term at 63.

From FIG. 3, the interference light 53, that is, FIG. 8 (α), (b)
The intensity distribution of the striped pattern shown in FIG. At the stepped portion, a large signal change is obtained due to the phase change by the step depth C. This is because the reflectance at the interface 47 is higher than that of the photoresist 46.
This is because it is higher than that at the surface. Therefore, interference ((
1) The influence of the second and third terms in the equation becomes smaller. That is, as shown in FIG. 4(α), when the thickness distribution of the photoresist 46 is asymmetrical at the stepped portion of the alignment pattern 48, in the conventional alignment method,
As is clear from the interference light intensity curve in FIG. 21, the reflected light intensity distribution from the alignment pattern 48 is as shown in FIG.
), there is left-right asymmetry, and χd is regarded as the pattern center position compared to the true pattern center position χ, resulting in a detection error e-1.

On the other hand, FIG. 8(c) shows the striped pattern A-A of FIG. 8(b).
This figure shows the interference light intensity distribution in the area '.

That is, according to the present invention, it is possible to obtain a high-contrast interference light intensity distribution that faithfully reflects the step information of the alignment pattern 48 without being significantly influenced by the film thickness distribution of the photoresist 46. As a result, it is possible to ideally detect the true pattern center position X5, and it is possible to eliminate the deterioration in alignment accuracy caused by the photoresist film thickness and coating unevenness.

The effect of the present invention is also effective when transparent layers of 5ue267, Su, and N468 are formed on a step pattern of 5i66, as shown in FIG. That is,
With respect to the incident light 74, reflected light 75, 76, 77, 78 from the photoresist 69 surface and each layer interface 72, 71.70
Of these, the highest strength is the base 5j66 and 5402.
This is reflected light 78 from the interface 70 with 67. Therefore, the intensity distribution of the obtained interference light, that is, the intensity distribution of the striped pattern, changes greatly at the step S↓66, and the photoresist 69
The same effect as above can be obtained without being affected much by the film thickness.

〔Example〕

A first embodiment of the present invention will be described below with reference to FIGS. 6 to 12.

FIG. 6 is a diagram showing an alignment optical system in the first embodiment of the present invention. This alignment optical system includes an alignment illumination system, an alignment pattern detection optical system,
It consists of a reference optical path and an alignment pattern detection signal processing system. First, a beam 81 emitted from an Ar laser 80 (wavelength 514.5 nm) used as a coherent light source passes through a beam splitter 82 and a ray lens 83, and is separated into beams 86 and 87 by a beam splitter 84.

The beam 87 is transmitted through mirrors 85b to 85f and the wavefront correction optical system 8.
8 and a reflecting mirror 89 (reference surface), and is reflected by the reflecting mirror 89. At the same time, a wavefront correction optical system adjusts the beam diameter and adjusts the phase of the wavefront corresponding to the imaging characteristics of the reduction projection lens. Adjustments will be made. On the other hand, the beam 86 is
The light is reflected by the mirror 85a, the mirror surface 1m on the reticle, enters the center of the entrance pupil 2p of the reduction projection lens 2, and refers to the alignment pattern 48 on the wafer. FIG. 7 is an enlarged view showing the situation, in which 45 is a generation substrate, 46 is a photoresist, and 86 is an alignment illumination light formed by a parallel beam. Now, the optical path length difference between the optical path length from the beam splitter 84 to the reflecting mirror 89 and the optical path length from the beam splitter 84 to the alignment pattern 48 on the wafer can be adjusted by slightly moving the reflecting mirror 89 in the optical axis direction. It is set within the coherence distance of the Ar laser 80. As a result 1, the reflected light from the reflecting mirror 89 and the reflected light from the alignment pattern 48 surface (photoresist alignment pattern interface 47) interfere, and the beam splitter 8
4, interference light is obtained. This interference light is shown in Figure 8 (,2)
As shown in (b), striped patterns 95α and 95b are formed according to the relative inclination of the wavefront of the reflected light from the reflecting mirror 89 in the beam splitter 84 and the wavefront from the surface of the alignment pattern 48 in the optical axis direction. The interval between the striped patterns becomes smaller as the relative inclination of both wavefronts increases. Conversely, if both wavefronts are completely parallel, there will be no striped pattern;
Light and dark interference intensity appears due to the optical path difference between the inside and outside of the pattern. Note that the arrows indicate the inclination direction of both wavefronts, and 94 indicates the detection area of the alignment optical system. Here, this striped pattern 954.95b includes alignment pattern 4.
At the step portion No. 8, a large signal change appears due to the change in the step depth total phase. One of the major features of the present invention is that the signal change position of this striped pattern corresponding to the stepped portion of the alignment pattern 48 is extracted, and the center position of the alignment pattern 48 is determined from that position. Interference light obtained by the beam splitter 84,
That is, the striped patterns 95α and 95b shown in FIGS. 8(a, 8b) are imaged by the two-dimensional surface solid imaging device 91 via the relay lens 83, beam splitter 82, mirror 85, and magnifying lens 90. The photoelectrically converted striped patterns 95cL and 95b are sent to a computer 93 after being subjected to noise removal and AD conversion in a preprocessing circuit 92.

In this embodiment, in order to simplify the signal processing within the computer and shorten the alignment time, the direction of the striped pattern is set in advance in the same direction as the horizontal scanning line of the two-dimensional surface image sensor 91, that is, as shown in FIG. 8(b). Finely adjust the reflecting mirror 89 so that it points in the same direction as shown in . If one chip on a wafer is 20 mm x 20 mm, the parallelism of the wafer within one chip is about 1 μm, so once the reflecting mirror 8
If you fine-tune 9, you don't need to constantly monitor it. In the computer, signal processing is performed according to the flowchart shown in FIG. 9, and the center position of the alignment pattern 48 is determined. First, FIGS. 10(α) to (d) and AA' in FIG. 8(b).

Horizontal scanning signals 964 to 96d corresponding to portions BB', CC', and DD' are shown. In the same figure (required) to (c), a large signal change appears due to the phase change at the step part of the pattern, but in the same figure (d), that is, D-D' corresponds to the bright area between the cotton and the stripes. However, there is no signal change. Differential processing is performed on all horizontal scanning signals of these two-dimensional solid-state sensors according to the flowchart shown in FIG. As a result, the signals 97α to 97d shown in FIG. 11(α) to (d)
is obtained. When absolute value processing is applied to these signals, signals 98 to 98d shown in Fig. 12 (1 to (d)) are obtained. Therefore, appropriate threshold values are set for these signals, and the threshold level Pixel number α□~tL corresponding to
4, b□~b11. C1 to C4, etc. are determined for all horizontal scanning signals and stored in memory. In this case, the same figure (
For horizontal scanning signals that do not have a corresponding pixel number, such as the signal shown in d), subsequent processing is not performed in order to reduce calculation time. For the stored pixel number,
The following process is performed to find the center position (tc, be, Cc, etc.) of the alignment pattern for each horizontal scanning line.

Cc = 4 Cc = 4 Now, calculate the center position of each horizontal scanning line by C (-L)
, ·1 to m), the final center position χr of the wafer alignment pattern is given by equation (5).

Here, m is a pixel number corresponding to an appropriate threshold level set in the horizontal scanning signal subjected to absolute value processing, as shown in FIG. 12 (4) to (c), in other words. For example, it is the total number of horizontal scanning lines in which a signal change corresponding to a pattern step portion exists. Next, the center position X of the aramend pattern (not shown) on the reticle obtained separately
- Find the alignment amount △ based on equation (6) from r,
Stage Δ=χr−xur
(6) Move 4 slightly in the (direction). The situation is exactly the same for alignment in the direction. When the alignment is completed, exposure light is irradiated from the exposure system (not shown), and the circuit pattern on the reticle 1 is placed on the wafer. It is burned into the upper chip.The above operation is repeated for each chip.

As described above, according to this embodiment, as described in the previous section "Operation", a highly symmetrical and high contrast alignment pattern detection signal can be obtained without being affected by the photoresist film thickness distribution. Deterioration in alignment accuracy caused by uneven application of photoresist can be eliminated. In addition, in the case of the conventional one-way alignment system, not only coating unevenness but also the thickness of the photoresist coating itself causes
The contrast 1- of the alignment pattern detection signal was changing greatly, but according to this embodiment, by slightly moving the reflecting mirror 89 in the optical axis direction, the interference intensity is always kept at the slope part of the curve 63 shown in FIG. Setting, Fig. 8(a),
It is possible to maintain an even higher contrast of the striped pattern shown in (b).

Further, according to this embodiment, the rotation error of the chip can be measured based on the center position χcj of the pattern obtained from each horizontal scanning signal. That is, it becomes possible to easily realize chip rotation correction, which will be indispensable in the future exposure of high-quality integrated circuit patterns.

In this embodiment, the differential absolute value of the signal was used to extract the signal change position of the pattern step portion in each horizontal scanning signal, but in another embodiment, the horizontal scanning signal 96tI- to 9
It is also possible to directly set a threshold value in 6d (FIG. 10), detect the pixel number corresponding to the threshold level, and determine the center position of the alignment pattern from that value.

Furthermore, as another embodiment, it is also possible to obtain the pattern center position χcl for each horizontal scanning signal by utilizing the symmetry of the signal waveform. At this time, the final center position if of the wafer alignment pattern is determined from equation (5).

Another embodiment of the present invention will be described with reference to FIGS. 13 to 15.

FIG. 13 is a diagram showing an alignment optical system in a second embodiment of the present invention. This alignment optical system is
The difference from the alignment optical system shown in FIG. 6 is that the reference optical path is removed, a semi-transparent mirror (dichroic mirror) 100 is added to the lower end of the reduction projection lens 2, and the alignment illumination optical path itself is used as the reference optical path. As in the previous embodiment, a beam 81 emitted from an Ar laser 80 (wavelength: 514.5 ohm) passes through a beam splitter 82° relay lens 83, is reflected by a mirror 85, and is reflected by a mirror surface 1 m above the reticle, and is reduced and projected. Entrance pupil 2p of lens 2
incident on the center of the semi-transparent mirror (dichroic mirror) 1
After transmitting 00, alignment pattern 4 on the wafer
Light up 8. This semi-transparent mirror (dichroic mirror)
100 is coated with an antireflection film (wavelength: 514.5 om) on its upper surface (on the reduction projection lens side), and with a thin film having the reflectance/transmittance spectral characteristics shown in FIG. 15 on its lower surface 101. By slightly moving the semitransparent mirror 100 in the optical axis direction, the optical path length from its lower surface 101 to the surface of the alignment pattern 48 is set within the coherence distance of the Ar laser 8o. The reflected light from the lower surface 101 of the mirror 100 and the reflected light from the surface of the alignment pattern 48 interfere, and interference light is obtained at the semi-transparent mirror 100. As in the previous embodiment, this interference light is aligned with the wavefront of the reflected light from the lower surface 101 of the semi-transparent mirror 100 as shown in FIGS. 8(0-) and (b).
According to the relative inclination of the wavefront from the 8 surface in the optical axis direction, striped patterns 9561 and 95b are formed.

After passing through the reduction projection lens 2 and the same optical path again, this striped pattern is passed through the beam splitter 82. An image is captured by a two-dimensional surface image sensor 91 via a mirror 85% and a magnifying lens 90. The subsequent signal processing, method of detecting the center position of the alignment pattern, and alignment procedure are exactly the same as in the previous embodiment, so their explanation will be omitted. FIG. 14 shows the fine movement mechanism of the semi-transparent mirror 100. semi-transparent mirror 1
00 is assumed to be supported by a leaf spring or the like (not shown) with respect to the reduction projection lens 2. As a fine movement mechanism, semi-transparent mirror 1
Three wedge-shaped bases 102 fixed on the top surface of 00
Wedge-shaped movable spacers 102b, 103b for α, 103c, 104ω.

104 and b are three piezo elements 102c, respectively.

This is realized by making slight movements independently in the directions of the arrows at 103c and 104c. The thin film on the lower surface 101 of the semi-transparent mirror 100 has an exposure wavelength of 436 nm as shown in FIG.
It shows an extremely high transmittance for m (g-line of a mercury lamp), and there are no particular problems during exposure.

According to this embodiment, not only can the same effects as those of the previous embodiments be obtained, but also the alignment optical path itself is used as the reference optical path, thereby preventing disturbances in the reference optical path or the imaging characteristics of the reduction projection lens. The influence on the interference fringe pattern can be removed. Also, by removing the reference optical path,
This has the effect of simplifying the alignment optical system.

〔Effect of the invention〕

As explained below, the alignment method of the present invention can solve the problem of photoresist alignment, which has been pointed out in the past but still remains unsolved, and which is becoming a serious problem with the miniaturization of semiconductor integrated circuits. This eliminates the decrease in alignment accuracy caused by coating film thickness and coating unevenness, making it possible to achieve stable and highly accurate alignment.

This has the effect of providing high productivity and reliability.

[Brief explanation of the drawing]

Figure 1 is a front view showing the principle of this alignment method, Figure 2 is a front view showing the principle of this alignment method.
The figure is a perspective view showing alignment irradiation light incident on the alignment pattern, Figure 3 is a diagram showing the relationship between optical path length difference and interference intensity, and Figure 4 (=2) is an alignment pattern showing the state of uneven coating of photoresist. Sectional view of the same figure (b
) and (C) are diagrams showing the difference in reflected light intensity distribution between the conventional method and the present alignment method, FIG. 5 is a cross-sectional view of an alignment pattern portion having a transparent layer, and FIG. 6 is a diagram showing the difference in reflected light intensity distribution between the conventional method and the present alignment method. FIG. 7 is a front view showing the alignment optical system, FIG. 7 is a perspective view showing the alignment irradiation light incident on the Flymen 1 pattern, FIGS. 8(a) and (b) are plan views showing the interference fringe pattern, and FIG. Flowchart for determining center position of alignment pattern from interference fringe pattern, 10th
Figures (ω), (b), (c). (d) is a diagram showing detection signals in the longitudinal direction of the interference fringe pattern, and FIG.
A diagram showing signals after differential processing of the detection signals shown in Figures (cL), (b), (c), and (d).
. (c), (d) are shown in Figure 11 (d), (b), (c), (
Figure 13 showing the signal after absolute value processing of the signal shown in d)
The figure is a front view showing an alignment optical system in another embodiment of the present invention, FIG. 14 is a perspective view and an enlarged view showing a fine movement mechanism of a semi-transparent mirror, and FIG. 15 is a thin film coated on the lower surface of a semi-transparent mirror. A diagram showing reflectance/transmittance spectral characteristics, FIG. 16 is a perspective view showing an example of a conventional TTL alignment method,
Fig. 17 is a perspective view showing the flow direction of the photoresist spin-coated onto the wafer, Fig. 18 is an enlarged perspective view of the wafer alignment pattern, and Figs. 19 (Q) and (b) show the photoresist film in the direction of the alignment pattern. Figures 20 (L) and (b) are diagrams showing differences in thickness distribution, and Figure 20 (L) and (b) are diagrams showing differences in intensity distribution of reflected light from alignment patterns. Figure 21 is a diagram showing the relationship between photoresist film thickness and interference intensity. It is a diagram. 1... Reticle, 2... Reduction projection lens, 3...
Wafer, 14.14', 19,20,48.73...Wafer alignment pattern, 51.89...Reflector. 100...Semi-transparent mirror, 80...Ar laser, 91...
・Two-dimensional surface image sensor, 93... Computer + 95 (L
+95b...Interference fringe pattern.婉Deaf-C) Q C)H d) Figure 12 (C) (L:L) CC-χ -χ Figure 75 SilL length [nrr+] Figure 16 Figure t'r Figure 18 Figure 19 (α) (b) Shiru 20
Figure (OL) (b) χwχd

Claims (1)

[Claims] 1. When exposing a circuit pattern on a reticle onto a wafer through a reduction projection lens, the alignment pattern on the wafer is illuminated with coherent illumination light, and the alignment pattern on the wafer is The reflected light from the pattern and the illumination light are optically interfered with each other, and the two-dimensionally distributed interference pattern obtained by the interference and the alignment pattern on the reticle illuminated by the alignment illumination light are displayed using a magnifying lens and an imaging device. In an alignment method that relatively aligns the reticle and wafer by detecting the alignment using an alignment optical system equipped with a 1. A reduction projection type alignment method characterized by detecting a shape change position of an interference pattern that appears in an interference pattern distributed in the wafer, and determining the center position of the alignment pattern on the wafer from the detected position. 2. The reduction projection alignment method according to claim 1, wherein the illumination light to be interfered with is reflected light from a semi-transparent mirror. 3. The reduction projection alignment method according to claim 2, wherein the semi-transparent mirror is provided outside the optical path for illuminating the on-wafer alignment pattern including the reduction projection lens. 4. The reduction projection alignment method according to claim 2, wherein the semi-transparent mirror is provided in an optical path for illuminating an on-wafer alignment pattern that includes a reduction projection lens. 5. The reduction projection type alignment method according to claim 3, characterized in that a wavefront correction optical system is provided in the optical path provided with the semi-transparent mirror. 6. The reduction projection alignment method according to claim 1, wherein the alignment illumination light has temporal and spatial coherence.