JPH0547634A - Position detector - Google Patents

Abstract

PURPOSE:To provide a very accurate and noise-resistant position detector. CONSTITUTION:A slit pattern 104 of a reference member 103 and a slit pattern 106 of a reticule M are line-and-space patterns (periodic pattern). These slit patterns, as specific patterns for changing the intensity of a photoelectric signal partially in the direction nearly vertical to the cycle (pitch) direction, are rectangular patterns which have a partially extended part. A first signal processing circuit 301 detects a relative position of the slit patterns 104 and 106 based on the periodicity of the photoelectric signal and a second signal processing circuit 302 detects the position of a peak based on the intensity of the photoelectric signal. The relative position of these slit patterns are detected by specifying the position of the center of a signal wave form within + or -half the cycle of the periodic pattern based on the information about the detected peak position.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for detecting the relative position of two objects, and more particularly to a mask or reticle and a photosensitive substrate (photosensitive substrate suitable for an exposure apparatus used in a lithography process for manufacturing a semiconductor element or a liquid crystal display element). The present invention relates to a device for detecting a relative positional relationship with a semiconductor wafer or a glass plate).

[0002]

2. Description of the Related Art Conventionally, various methods have been proposed as means for measuring the position by knowing the positional relationship between two objects. Among them, two objects are provided with slits, respectively, while irradiating one slit with the illumination light, the two objects are moved relative to each other, and enter the photoelectric detector (PIN photodiode or the like) through the other slit. A method of finding the relative positional relationship between two objects by obtaining a change in the light amount (light intensity) of illumination light is widely known, as used in encoders and the like.

Among them, an example of application to an exposure apparatus for manufacturing a semiconductor integrated circuit which requires particularly high precision is disclosed in, for example, Japanese Patent Application Laid-Open No. 64-10105. According to this publication, a slit formed on a reference member on a wafer stage is illuminated from below, and an image of this slit is formed on a pattern surface of a reticle via a projection lens. Further, the wafer stage is finely moved to relatively move the slit image and the slit on the reticle, and the light transmitted through the slit of the reticle is received by the photoelectric detector. Then, the signal from the photoelectric detector is transferred to the unit movement amount (0.0
Sampling is performed in synchronization with an up / down pulse generated every 2 μm), the sampled value is converted into a digital value, and the relative position is detected by a predetermined calculation process. Here, the position where the signal waveform has a peak is obtained from the change of the photoelectric signal according to the overlapping degree of the two slits.

In addition to the above method, as a method of reducing a measurement error by, for example, an averaging method, a signal waveform having a plurality of peaks generated by relative scanning of a plurality of slits having one or the same shape is known. There was a method of detecting the peak based on the slit interval and measuring the entire position from the average value of the individual positions. Further, a method of applying the method disclosed in Japanese Patent Laid-Open No. 1-299402 to the above method has been considered. This is, by using a plurality of slits of equal shape arranged at equal intervals, to extract the fundamental wave component and its higher-order components from the signal waveform having periodicity by Fourier series expansion, and with them, It was a method of calculating the relative position by the phase difference between the fundamental wave component and its higher-order component, which were introduced by design.

[0005]

However, in the prior art as described above, for example, in the method of detecting the position by detecting the peak of the signal waveform, when the waveform is disturbed due to the influence of noise or a change in the light amount of the light source itself. However, there is a problem that an incorrect position is detected. In addition, the method of detection using a signal waveform having a plurality of peaks has not been able to achieve essential improvement, although there is a difference in degree. Furthermore, when using a light source that emits light in pulses, such as an excimer laser, it is necessary to detect the amount of light in time with the timing of light emission, and when trying to detect a long range,
There is a problem in that it takes a long time in order to obtain a practically sufficient measurement accuracy due to the light emission interval.

Further, the method of detecting the relative position using the phase difference between the waveforms is not easily affected by noise or a change in the light amount of the light source itself, and even if the change in the light amount is detected at relatively coarse intervals, it is considerably accurate. However, if the phase difference between the fundamental wave component guided by the design and the fundamental wave component of the actually detected waveform deviates by more than half a cycle, there was a problem that erroneous detection would occur due to that principle. ..

To solve this problem, Japanese Patent Laid-Open No. 1-2994
In Japanese Patent Laid-Open No. 02-202, pre-alignment is preliminarily performed so as not to shift by more than a half cycle. However,
Pre-alignment requires time, a separate pre-alignment mark is required, and it is necessary to calibrate in advance the offset (offset) between the pre-alignment mark and the slit used for actual measurement due to manufacturing error. There were problems such as being there.

The present invention has been proposed in order to solve the above problems, and an object of the present invention is to obtain a position detecting device capable of highly accurate and noise-resistant measurement in a short time.

[0009]

To achieve the above object, in the present invention, a second object having a second pattern is relatively moved in a predetermined first direction with respect to a first object having a first pattern. Moving means, means for detecting the position of the second object, illuminating means for illuminating the first pattern and the second pattern with illumination light, and the first pattern and the second pattern for illuminating with the illumination light. And a photoelectric detection means for receiving light generated from the pattern, and a detection signal output from the photoelectric detection means when the first object and the second object are relatively moved in the first direction, and the detection means. In the device that detects the relative position of the first object and the second object in the first direction based on the detection signal of, the first pattern and the second pattern are:
A specific pattern that has periodicity in the first direction and that partially changes the intensity of the detection signal is formed in the second direction that intersects the first direction; First signal processing means for generating first position information based on periodicity; and second signal processing means for generating second position information based on intensity of a detection signal from the photoelectric detection means are provided. , The first position information and the second
It was decided to detect the relative position based on the position information. In the following embodiments, the intensities of photoelectric signals are partially changed by forming the patterns 11, 13, 15 and the like as shown in FIGS. 1 and 2 on each of two objects.

[0010]

In the present invention, the fundamental wave component and its components are obtained from the signal waveform obtained when the method disclosed in the above-mentioned Japanese Patent Laid-Open No. 1-299402, that is, the periodic pattern formed on each of two objects is relatively scanned. A method of extracting the higher-order components by Fourier series expansion and calculating the relative position from the phase difference between them and the fundamental wave component and the higher-order components derived by design (hereinafter simply referred to as the phase difference measurement method) is adopted. To do. Therefore, in the present invention, when a slit pattern is considered as an example of a pattern to be formed on each of two objects, when one slit pattern moves relative to the other slit pattern, the light amount P moves. It becomes a function f (x) for the quantity x, and attention is paid to the fact that the function f (x) is proportional to the overlapping degree of the two slit patterns, that is, the area of the overlapping openings. Causes a waveform that partially changes). That is, in the present invention, the signal intensity is partially changed by adding a specific pattern to a periodic pattern formed on each of two objects in a direction intersecting the period (pitch) direction, in other words, It was decided to generate a waveform (peak or bottom) that can be distinguished from other peaks in the signal. Therefore, even if pre-alignment or the like is not performed in the phase difference measuring method, the center position of the signal waveform is set to the periodic pattern by using the waveform of the photoelectric signal in which the signal intensity partially changes corresponding to the specific pattern as a guide. It can be specified within ± half cycle. Therefore, by relatively moving the two objects only once, the center position of the signal waveform can be specified within ± half period and the relative position can be calculated by the phase difference measuring method. As a result, only one relative movement is required for position measurement, it is possible to prevent a decrease in throughput, and it is possible to realize position measurement with high accuracy and resistance to noise.

Here, the definition of terms is first started. First, the patterns used in the present invention are classified into two types: slit patterns and mark patterns. The slit pattern is a kind of pattern which transmits the measuring light, and the mark pattern is a kind of pattern which reflects the measuring light. This classification is for use in the following description of the basic principles. This is because the pattern shape is important in the present invention, and the distinction between transmission and reflection of light is not so important. This is because, as described in the explanation of the basic principle below, the signal waveforms obtained by various patterns are equivalent to the measurement in the present invention and cannot be distinguished at all. Further, the shape of each pattern is a two-dimensional shape seen on the optical axis of the measurement light, and even if the actual shape is three-dimensionally complicated, it is sufficient that the conditions of the present invention are eventually satisfied.

More specifically defined, the slit is an opening that transmits the measurement light, and the periphery of the slit blocks the measurement light at least over the scanning range. Generally, it is sufficient if there is a minimum signal-detectable difference between the opening and the peripheral portion thereof, that is, a difference in light transmittance, and it is not always necessary to obtain complete transmission and blocking. Further, as this modification, it is assumed that the same structure is obtained by inverting the structure in which the measurement light is blocked at the pattern portion and is transmitted at the peripheral portion.
Similarly, it suffices that the mark pattern and its surroundings have different reflectances in a signal detectable range, and it is meaningless to specify that one mark pattern is either bright or dark.

Regarding the most important pattern shape in the present invention, in the drawing of the present invention, for example, in FIG. 1, a rectangle extending in a direction substantially perpendicular to the scanning (measurement) direction of two objects (patterns). Although the shapes and patterns thereof are arranged evenly (predetermined pitch) in the scanning direction, they are not necessarily restricted to these shapes. For example, as shown in FIGS. 5A and 5B, shapes such as a lattice pattern and a diagonal pattern are also conceivable. However, for position detection by Fourier series expansion, the width and interval of each pattern are uniform with respect to the scanning direction as disclosed in Japanese Patent Laid-Open No. 1-299402.
It goes without saying that the generated waveform needs to have periodicity.

Further, the scanning direction is not particularly limited as long as the conditions of the above pattern are satisfied. Although the present invention has been described as having a means for detecting the position of the second object, in the case where the position x of the second object and the time t are represented by the relationship x = f (t), the function f (t) Obviously, the means for detecting the position can be replaced by a means for detecting the time, if is known. For example, when an object that rotates at a constant rotation speed is used, it is easy to measure by replacing the angle with time.

In the present invention, the measurement using light (continuous light or pulsed light may be used) is mainly described, but this light can be applied from infrared rays to ultraviolet rays. In addition, X-rays, electron beams, particle beams, radiation,
Sound waves or radio waves can be used. In other words, even if it is a wave or a particle, it is basically applicable as long as it has a straight traveling property and can form a pattern having different transmittance or reflectance with respect to it and can measure its intensity. ..

Next, the basic principle will be described. First,
It is assumed that one rectangular pattern (single slit) 41 as shown in FIG. 13A is formed on each of the two objects. Here, it is assumed that the longitudinal directions of the slit patterns are substantially perpendicular to the relative movement direction of the two objects, and the movement direction is the width and the direction perpendicular to the movement direction is the two slit patterns. Are assumed to have the same width. In this case, the lengths of the slit patterns provided on the two objects may be different from each other, but as a result, the length is regulated to be shorter. It is assumed that the distance (distance) between the two objects is zero and the thickness of the objects is also zero. Further, it is assumed that the measurement light is a parallel plane wave, and the influence of diffraction, refraction, etc. is minute and can be ignored. Further, it is assumed that the peripheral portion of the slit pattern completely blocks the measurement light.

Now, FIG. 15A shows a change in light quantity (light intensity) obtained when the two single slits 41 as described above are relatively scanned. In FIG. 15A, the horizontal axis represents the movement amount (position) x, and the vertical axis represents the light amount p (that is, the level (voltage value) of the photoelectric signal according to the intensity of the light transmitted through the two single slits). Here, as is clear from the figure, a triangular waveform 51 having one peak is obtained. The position x _p represents a position where the light amount p reaches a peak (light amount p ₁ ), that is, a position when two single slits exactly overlap each other.

Next, under the same conditions as above, one of the two objects has n pieces (7 pieces in the figure) as shown in FIG.
Consider a case in which a pattern (multi-slit) 42 including a slit pattern of (book) is formed. Here, each slit pattern has the same shape and the same size as the single slit 41, and the interval between the slits is twice the width, that is, the line and space is 1: 1. Now, the change in the amount of light (waveform of the photoelectric signal) obtained when the single slit 41 and the multi slit 42 are relatively scanned is shown in FIG.
The waveform 52 has n peaks (triangular waves) as shown in (b).

Further, consider the case where the multi-slit 42 shown in FIG. 13B is formed in each of two objects under the same conditions as above. At this time, both have the same shape, the same size, and the same number n. The signal waveform obtained at this time is shown in FIG. As is clear from the figure, here, an amplitude-modulated triangular waveform 53 having (2n-1) peaks is obtained. At this time, the envelope Ep connecting the peaks becomes a triangle. Further, when the two multi-slits 42 exactly overlap, the light amount p reaches a peak (light amount p _n ≈n · p ₁ ), and its position is represented by x _p . By the way, if the number of multi-slits of one object is m under the same conditions as above, the signal waveform will be as shown in FIG.
An amplitude-modulated triangular waveform 54 having (n + m-1) peaks as shown in FIG. 5 (d) is obtained. At this time,
The envelope Ep connecting the peaks has a trapezoidal shape.

Up to now, the change in the amount of transmitted light between slit patterns has been described, but it is also possible to consider one slit pattern by replacing it with a mark pattern. Even in such a case, the irradiation of the plane wave from the slit pattern side is the same, but FIG.
It is necessary to have a configuration as shown in each of (f). Since the configurations shown in FIGS. 16A to 16F are generally well known, only FIG. 16A will be described here. In FIG. 16A, the illumination light from the light source 65 is applied to the mark pattern 63 via the slit pattern 64, and the return light from both patterns is passed through the beam splitter (or half mirror) 67 to the photoelectric detector 66.
It is designed to receive light. 16A, it is assumed that the two objects 61 and 62 are arranged in close proximity to each other, or that the two objects 61 and 62 are arranged in an optically conjugate relationship. Further, it is assumed that the photoelectric detector 66 and the mark pattern 63 and the pattern 64 are set in a Fourier transform relationship by a lens system (not shown). The above is shown in FIGS. 16 (b) to 16 (f).
Is exactly the same. The surface of the object 61 on the mark pattern side is completely reflected by the measurement light over the scanning range or more, and the mark pattern 63
Is non-reflecting or completely transmitting the measuring light, and has the same shape and the same size as the slit pattern.

In the case of adopting the above structure, for example, the single slit 41 shown in FIG. 13A as the pattern 64 and the slit 41 as the mark pattern 63 are used.
If the same one is used, the intensity of the photoelectric signal from the photoelectric detector 66 obtained when these are relatively scanned is such that the reflected light decreases in proportion to the overlapping area of the two patterns. , It becomes zero when they completely overlap. Therefore, the signal waveform obtained at this time is shown in FIG.
An inverted triangular waveform 55 as shown in (e) is obtained. This is a waveform that is completely opposite to the waveform 51 shown in FIG. 15A, and similar results (waveforms between slit patterns are completely opposite) for other combinations (for example, multi-slit and multi-mark). Waveform) is obtained.

If the surface of the object 61 on the side of the mark pattern is non-reflective or is completely transmissive and the mark pattern is replaced with perfect reflection in the same configuration as in FIG. A waveform similar to that of the patterns (waveform 51 shown in FIG. 15A) is obtained. Further, as shown in FIGS. 15 (c) and 15 (d), the same consideration can be given to the case where the measurement light is once reflected by the mark pattern and the reflected light is received by the back surface of the slit pattern. The waveforms in this case are also the waveforms 51 to 5 shown in FIGS. 15A to 15D, respectively.
4 or vice versa. Similarly,
It can be said that the same result can be obtained when the reflected lights from the mark patterns as shown in FIGS. 16E and 16F are measured.

In the following embodiments, in order to shorten the description, the above-mentioned combinations are respectively of transmission / transmission type (slit patterns), transmission / reflection type (FIG. 16 (a),
(B)), reflection / transmission type (FIGS. 16C and 16D), reflection / reflection type (FIGS. 16E and 16F). By the way, it is easy to detect the relative position between two objects by peak detection for an ideal signal waveform as shown in FIG. For example, in the case of the waveform 51 shown in FIG. 15A, the position with the largest amount of light may be searched for and the position x _p at that time may be calculated. Further, in the case of the waveform 52 shown in FIG. 15B, if the interval of each slit of the multi-slit 42 and the number n are known, it can be easily obtained by performing the same processing as above. FIG. 15 (c),
The same applies to the case of the waveforms 53 and 54 shown in (d). The waveform 55 shown in FIG. 15E, and FIG.
The same applies to the case of the waveforms having shapes opposite to those of the waveforms 52 to 54 shown in (d), if the position with the smallest light amount is searched for.

The characteristics of position detection for these waveforms will be described. In the case of the waveform 51 shown in FIG. 15A, the signal processing is simple. FIG. 15 (b)-
It can be said that the averaging effect can be obtained in the case of the waveforms 52 to 54 shown in (d). FIG. 15 (c),
In the case of the waveforms 53 and 54 shown in FIG.
Assuming that the light quantity of the waveform 51 shown in (a) is p ₁ (however, n <m in FIG. 15D), the light quantity p _n at the position where the light quantity is greatest becomes n × p ₁ , and S / It is advantageous in N ratio. Further, the manufacturing error between the slit patterns is also averaged. Furthermore, the waveform 5 shown in FIG.
In the case of 3, it can be said that the position x _p with the largest light amount is the center of the pattern, which is easy to calculate.

By the way, in the case of the waveforms 52 to 54 shown in FIGS. 15 (b) to 15 (d), it is assumed that they are periodic waveforms, so that they are disclosed in Japanese Patent Laid-Open No. 1-299402. There is also a method of calculating the relative position by extracting the fundamental wave component and its higher-order components from the signal waveform by Fourier series expansion, and calculating the phase difference between them and the fundamental wave component and its higher-order components that are derived by design. Applicable. This method is characterized by being less susceptible to noise. In the case of this method, as shown in the above publication, it is not possible to measure more than ± half period of the fundamental wave component in principle, so it is necessary to pre-align it so that it fits in the vicinity of the reference position in advance,
In the case of the waveform 53 shown in FIG. 15C, the center position is obtained by the peak detection of the signal waveform as described above,
This can be handled by shifting the reference position in calculation.

The above can be said only when an ideal waveform as shown in FIG. 15 is obtained, and it is almost impossible to obtain such a waveform in practice. Actually, the influence of diffraction and diffusion of the measurement light, the change in the light amount of the measurement light itself, the conversion error when converting into an electric signal by the photoelectric detector, the manufacturing error of each slit pattern, and the noise generated in various places, etc. Accordingly, for example, the change in light amount (photoelectric signal) obtained when the single slits shown in FIG.
The waveform 43 is as shown in FIG. In addition, FIG.
The light amount change (photoelectric signal) obtained when the multi-slits shown in (b) are relatively scanned has a waveform 44 as shown in FIG. 14 (b). In particular, when the influence of the noise or the like is large, a pseudo peak may appear, and in FIG.
In some cases, the place with the largest amount of light, such as the waveform 45 shown in (c), is not necessarily the center of the pattern. The waveform 45 appears when an excimer laser or the like is used as the measurement light. In such a case, if peak detection is performed with the waveform 43 shown in FIG. 14A, the possibility of false detection increases. 14
If the phase difference is detected with the waveform 44 shown in (b), the measurement accuracy is relatively high, but a pseudo peak may appear as shown in FIG. 14 (c), and the pattern center may not be specified. It is necessary to pre-align two objects (patterns) in advance. Normally, position detection is performed using the above-mentioned two types of slit patterns due to the relationship of the measurement system.

Now, let us consider a pattern 11 as shown in FIG. The feature of the pattern 11 is that the length of the central pattern of the multi-slit 42 shown in FIG. 13B is extended in the longitudinal direction (direction intersecting the cycle (pitch) direction). When relative scanning is performed with such slit patterns, when the two slit patterns substantially overlap each other, the extended central patterns also coincide with each other, and the amount of transmitted light increases dramatically. The waveform of the photoelectric signal (change in light amount) obtained at this time is a waveform 12 as shown in FIG.
The extended portion of the pattern 11 corresponds to the specific pattern of the present invention.

With such a waveform 12, even if there is noise or other influence, if the extension length is determined so that the amount of light in the vicinity of the center of the waveform 12 is maximized, this is used as a guide to detect the phase difference. Is easy to do. That is, the slit pattern 11 is obtained by obtaining the position where the intensity of the photoelectric signal changes greatly partially by simple peak detection.
Even if it is a measurement system where the pattern center can always be specified from the design value of (that is, the position of the pattern where the length is extended), and the waveform is disturbed accordingly (the effect of noise is large),
It becomes possible to omit the pre-alignment. It is obvious that the waveform 12 at this time has no influence on the fundamental wave component for performing the phase difference detection. Also,
In order to calculate the actually required pattern length, it is necessary to consider the amount of diffracted light due to the type of measurement light, the effect of diffusion, the level of noise that may exist in the measurement system, etc.
It can be calculated sufficiently.

It should be noted that, as in the pattern 11 shown in FIG. 1A, the length of the central pattern of the plurality of rectangular patterns is not extended, but the pattern other than the central pattern is shortened, or The same effect can be obtained by shortening only the central pattern, but the S / N ratio may be deteriorated. Therefore, it is desirable to extend the length of a part of the pattern like the pattern 11.

When the patterns at both ends are extended as in the pattern 13 shown in FIG. 1 (b), there are horn-shaped protrusions at both ends and the center as shown by the waveform 14 in FIG. 2 (b). Waveform is generated. Further, when the patterns at both ends and the center are extended as in the pattern 15 shown in FIG. 1C, the horn-shaped protrusions are formed at both ends and the center and in the middle as shown by the waveform 16 in FIG. 2C. A waveform with is generated. This is obvious when considering the area of the portion where the two slit patterns actually overlap. Incidentally, the waveforms 12, 1
If the maximum amount of light in each of 4 and 16 is the same, the length of extension of 13 from pattern 11 and 15 from 13 may be shorter.

Furthermore, instead of extending the pattern,
Like a pattern 17 shown in FIG. 1D, a pattern having the same width beside a normal periodic pattern (that is, in a direction intersecting with the periodic direction, or a direction orthogonal to the periodic direction) (corresponding to a specific pattern of the present invention). ) Is added, a similar waveform is obtained.
Further, as shown in FIG. 6A, if the patterns at both ends and the pattern at the center are extended in opposite directions, mutual interference at the extended portions is reduced, so that only the peak at the center of the waveform can be particularly increased. Further, when the number of periodic patterns can be increased, the peak at the center of the waveform can be particularly increased by using the patterns shown in FIGS. 6B and 6C. At this time, it is possible to select the optimum pattern by calculating and simulating which pattern should be extended. In the patterns shown in FIGS. 1 and 6, the extending lengths are all shown to be equal, but they do not necessarily have to be the same length. Further, it goes without saying that any of the patterns is a periodic pattern arranged at a predetermined pitch, and further has a specific pattern (corresponding to the above pattern extension portion or the like) of the present invention in a direction intersecting the periodic direction. It is clear that

By the way, as another way of thinking, if a waveform serving as a guide is generated before and after the waveform for detecting the phase difference,
The same effect as the above can be obtained. An example of this kind of pattern is shown in FIG. 3, and a signal waveform obtained when these patterns are relatively scanned is shown in FIG. It is easily understood that when the patterns 21 and 22 as shown in FIG. 3A are formed on each of the two objects, the waveform 23 as shown in FIG. 4A is obtained. Similarly, when patterns 22 and 24 as shown in FIG. 3B are used, a signal waveform 25 as shown in FIG. 4B is obtained. At this time, as shown in FIG.
The waveforms (peaks) shown at both ends in (b) are used only for specifying the center position of the signal waveform within ± half cycle of the periodic pattern, and such a waveform serving as a guide is a position. Since it is not used for phase difference detection, the width of the guide forming pattern arranged adjacent to the periodic pattern does not need to be particularly equal to the width of the periodic pattern.
Therefore, when the pattern shown in FIG. 3 is used, the width of the pattern for creating the guide can be increased, in other words, the pattern area can be increased without extending the pattern too much. As compared with the patterns shown in FIGS. 1 and 6, there is an advantage that the length of the pattern for creating the guide to be extended in the direction perpendicular to the periodic direction can be short.

Further, when the periodic pattern of the present invention is formed on a cylinder as shown in FIG. 7 (a), and as shown in FIG. 7 (b), a pattern is formed on the disc so as to extend radially from the center of rotation. When forming the, it is also possible to detect the angle.

[0034]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, a case where the position detecting device according to the present invention is applied to an exposure apparatus for manufacturing a semiconductor integrated circuit which requires highly accurate position detection will be described. FIG. 8 is a diagram showing a schematic configuration of a projection exposure apparatus having a position detection device according to the first embodiment of the present invention. In this embodiment, a case will be described in which the relative positional relationship between the reticle (or mask) and the wafer stage is detected by the transmission / transmission method.
It should be noted that this type of measurement is performed during fine alignment of the reticle or during baseline measurement.

As shown in FIG. 8, the reticle M is an exposure illumination light (i-line, KrF excimer laser, etc.) whose light flux is uniformized by a main illumination system (not shown) including an optical integrator and the like. Is illuminated with a substantially uniform illuminance.
A light transmissive pattern 106 is formed in the vicinity of the pattern area of the reticle M, and in this embodiment, the pattern 106 is, for example, the slit pattern 11 shown in FIG. 1A. To do. The illumination light that has passed through the pattern area of the reticle M is on both sides (or one side).
The light enters the telecentric projection lens 105, and the projection lens 105 reduces the image of the circuit pattern formed on the lower surface (pattern surface) of the reticle M to 1/5 or 1/10, and a resist layer is formed on the surface. The image is projected on the wafer W. The wafer holder 110 is provided so as to suck the wafer W in a vacuum and is capable of minute rotation with respect to the wafer stage 101 that moves two-dimensionally in the X and Y directions. The wafer stage 101 is two-dimensionally moved by a step-and-repeat method by a motor 112, and when transfer exposure of the reticle M onto one shot area on the wafer W is completed, stepping is performed to the next shot position. Interferometer 1
Reference numeral 11 always detects the two-dimensional position of the wafer stage 101 with a resolution of, for example, about 0.02 μm.

On the wafer stage 101, the fiducial mark 1 used for baseline measurement or the like is also used.
A reference member (glass substrate) 103 including 04 is provided so as to substantially coincide with the surface position of the wafer W. As the fiducial mark 104, for example, two sets of the light transmissive slit patterns 11 shown in FIG. 1A are formed on the reference member 103 so as to extend in the X and Y directions, respectively. The slit pattern 11 has a duty ratio of 1: 1.
Line and space pattern, in which only the central rectangular pattern is in the cycle (pitch) direction (first
Direction) and is formed to extend in a direction (second direction) substantially perpendicular to the (direction). This is the waveform (light intensity) of the photoelectric signal obtained when the slit patterns 104 and 106 are relatively scanned.
Is partially changed, and the center position of the signal waveform is specified within ± half period in the phase difference detection described later.

Illumination light (exposure light) L which is partly branched from the main illumination system is guided to the inside of the wafer stage 103 by a light transmission system (not shown) such as an optical fiber, and then reflected by the mirror 102 to be a reference member. Transmitted down to 103,
The reference member 103 is configured to irradiate from the back surface thereof. As a result, the slit pattern 104 is illuminated from below (inside the wafer stage 101), and the illumination light L transmitted through this slit pattern 104 is transmitted.
Forms a projected image of the slit pattern 104 on the back surface (pattern surface) of the reticle M via the projection lens 105. When the wafer stage 101 is further finely moved to relatively scan the projected image of the slit pattern 104 and the slit pattern 106 on the reticle M in the periodic direction (measurement direction), the illumination light passing through the slit pattern 106 is reflected by the mirror 107. After being reflected, the light passes through the condenser lens 108 and is received by the photoelectric detector (photodetector) 109 arranged near the pupil conjugate plane of the projection lens 105.

Further, the photoelectric signal (waveform 12 shown in FIG. 2A) output from the photodetector 109 is input to the signal processing system 300, and the signal processing system 300 also inputs the position signal from the interferometer 111. .. In this embodiment, the signal processing system 300 has two sets of signal processing circuits, and the first signal processing circuit 301 is the same as the above-mentioned Japanese Patent Laid-Open No. 1-2994.
In accordance with the method disclosed in Japanese Patent Publication No. 02-202, the photoelectric signal is Fourier-series expanded to extract the fundamental wave component and its higher-order components, and the extracted fundamental components and their higher-order components are derived by design. The relative position is calculated from the phase difference with the component. Since the details are disclosed in the above publication, the description thereof is omitted here. On the other hand, the second signal processing circuit 302 detects the peak of the photoelectric signal by a predetermined signal processing (for example, slice level) (especially, in the present embodiment, at a position where the peak appears at approximately the center of the waveform 12 shown in FIG. 2A). The detection result is used to identify the center position of the signal waveform within ± half period (180 °). The main control system 500 is the signal processing system 30.
0, that is, the detection results (first and second position information) of each of the first and second signal processing circuits 301 and 302 are input,
Based on these two slit patterns 104, 10
The relative position of 6 is calculated. That is, even if the relative position (first position information) calculated by the first signal processing circuit 301 according to the phase difference measurement method is deviated by ± half period or more, the main control system 500 detects the second signal processing circuit 302. By specifying the center position of the signal waveform within ± half cycle based on the obtained peak position (second position information), the relative position can be calculated with high accuracy. Further, the main control system 500 gives a predetermined drive command to the motor 112 on the basis of the position information from the interferometer 111, and also controls the entire device.

The operation of the apparatus having the above structure will be briefly described below. In FIG. 8, the main control system 500 drives the wafer stage 101 with the motor 112 to drive the reference member 103.
Is sent under the projection lens 105. And the illumination light L
Is irradiated from the back surface of the reference member 103 to form a projected image of the slit pattern 104 on the pattern surface of the reticle M.
Then, the wafer stage 101 is slightly moved to relatively scan the projected image and the slit pattern 106, and the transmitted light is received by the photodetector 109. Signal processing system 30
0 samples the photoelectric signal from the photodetector 109 in synchronism with the position of the wafer stage 101 measured by the interferometer 111, converts the sampled value into a digital value, and stores it in a memory (not shown) in the order of addresses. .. After that, the waveform processing of the photoelectric signal is performed in each of the first and second signal processing circuits 301 and 302 as described above, and the main control system 500 further determines the two slit patterns 104 and 106 based on the detection result. That is, reticle M and wafer stage 101 (reference member 103)
Calculate the relative position with.

As described above, in the present embodiment, the structure as shown in FIG. 8 is adopted, but conversely, illumination light is irradiated from the upper surface of the reticle M to project the projected image of the slit pattern 106 on the reference member 10.
It is also possible to adopt a configuration in which an image is formed on the image 3 and the amount of transmitted light is detected via the slit pattern 104. Also,
Although the exposure light partially branched from the main illumination system is used as the illumination light L, it goes without saying that light from another light source (not necessarily the same wavelength as the exposure light, as long as it has a wavelength close to the exposure wavelength) , For example, it may be a harmonic)
May be used as the illumination light L. Furthermore, Reticle M
Needless to say, if at least one of the reference member 103 and the reference member 103 has a reflective mark pattern, the reflective / transmissive type or the reflective / reflective type as shown in FIGS. 16C and 16E can be applied. Further, in FIG. 8, the light receiving system (photodetector 109 and the like) is arranged on the reticle M. The slit pattern 106 may be arranged so as to receive the transmitted light of the slit pattern 106 via this beam splitter. In this case, the arrangement (position) of the slit pattern 106 on the reticle M can be freely selected. The advantage that it can be obtained. With the configuration as shown in FIG.
The slits 106 on the mask M can be arranged freely.

Next, a second embodiment of the present invention will be described with reference to FIG. In this embodiment, a case where the present invention is applied to an alignment system of a projection exposure apparatus, and particularly a case where a relative positional relationship between a reticle and a wafer is detected via a projection lens by a transmission / reflection type system will be described. Incidentally, FIG.
Then, members having the same functions and actions as those shown in FIG. 8 are designated by the same reference numerals. Further, since the waveform processing of the photoelectric signal in this embodiment is exactly the same as that in the first embodiment,
Then, the signal processing system 300 and the main control system 500 are omitted. In this embodiment, only points different from the first embodiment will be described.

In FIG. 9, the illumination light L partially branched from the main illumination system (not shown) is the slit pattern 1 of the reticle M.
06, and the projected image is formed on the wafer W via the projection lens 105. Even if the exposure light is not branched from the main illumination system, the exposure light is applied only to a partial area including the slit pattern 106 by a reticle blind (not shown) so that the area other than the slit pattern 106 is not irradiated with the exposure light. You may do it. Also, the wafer W
It is assumed that an alignment mark pattern 120 (for example, one having the same shape as the slit pattern 11 shown in FIG. 1A) is formed in each of the plurality of shot areas above. Here, the mark pattern 120 may have a pattern with respect to the illumination light L that has a reflectance different from that of a region other than the mark region. As a method, a method in which the surface is roughened by etching or the like and the measurement light is scattered is conceivable, but the mark forming method is not particularly limited.

Now, by slightly moving the wafer stage 101,
The projected image of the slit pattern 106 and the mark pattern 12
When 0 is relatively scanned, the illumination light L is reflected via the projection lens 105 according to the reflectance due to the mark shape. The reflected light is partially branched by the half mirror 121 and condensed on the light receiving surface of the photo detector 109 by the condenser lens 108. At this time, the photoelectric signal output from the photodetector 109 is waveform-processed by the same operation as that of the first embodiment, so that the reticle M and the wafer W are processed.
The relative position with can be obtained. Although the half mirror 121 is arranged between the reticle M and the projection lens 105 in this embodiment, it should be arranged between the wafer W and the projection lens 105 in consideration of the transmittance of the projection lens 105. May be. Of course, it may be arranged above the reticle M.

Next, a third embodiment of the present invention will be described with reference to FIG. In this embodiment, when the present invention is applied to a TTL alignment system of a projection exposure apparatus,
Particularly, the case where the position of the wafer is exclusively detected through the projection lens by the transmissive / reflective method will be described. Incidentally, FIG.
In the alignment system shown in FIG. 1, the reference plate 131 arranged in a plane substantially conjugate with the wafer W with respect to the projection lens 105.
The relative position between the upper reference pattern (slit pattern) 131 and the mark pattern attached to the shot area on the wafer W is detected. Further, in FIG. 10, members having the same functions and actions as those shown in FIG. 8 are denoted by the same reference numerals, and the waveform processing of the photoelectric signal in this embodiment is exactly the same as that in the first embodiment. Therefore, the signal processing system 30
0 and the main control system 500 are omitted. Further, in the present embodiment, only the points different from the first and second embodiments will be described.

In FIG. 10, the illumination light L is reflected by the half mirror 132 to illuminate the reference plate 130, and the light transmitted through the slit pattern 131 on the reference plate 130 is reflected by the mirror 133 and passes through the projection lens 105. Then, a projected image of the slit pattern 131 is formed on the wafer W. The measurement operation, the waveform processing operation, and the like are the same as those in the first and second embodiments, and thus the description thereof is omitted here. As described above, the present invention can be easily applied to a TTL alignment system, and its measurement accuracy can be improved. Note that if the positional relationship between the reticle M (slit pattern) and the slit pattern 131 on the reference plate 130 is accurately calibrated in advance, the mark pattern 120 becomes the slit pattern 131.
It goes without saying that the projection image of the reticle pattern and the shot area are accurately aligned by aligning with each other. Note that the configuration shown in FIG. 10 is combined with the configuration shown in FIG. 8, that is, in FIG.
When the mirror 133 is arranged as a half mirror between the projection lens 105 and the projection lens 105 and the alignment system shown in FIG. 10 is incorporated, the illumination light L from the reference member 103 also irradiates the reference plate 130 via the half mirror 133. As a result, the position of the reference plate 130 can be detected at the same time as the position of the reticle M is detected.

Next, a fourth embodiment of the present invention will be described with reference to FIG. In the present embodiment, a case where the present invention is applied to an off-axis type alignment system of a projection exposure apparatus, particularly a case where only the position of a wafer is detected by a transmission / reflection type method will be described. In the alignment system shown in FIG. 11, the reference pattern (slit pattern) 141 on the reference plate 140 arranged in the plane substantially conjugate with the wafer W with respect to the objective lens 142 and the shot area on the wafer W are attached. The relative position to the mark pattern will be detected. Further, in FIG. 11, members having the same functions and actions as those shown in FIG. 8 are designated by the same reference numerals, and the waveform processing of the photoelectric signal in this embodiment is exactly the same as that in the first embodiment. The signal processing system 300 and the main control system 500 are omitted. Further, the measurement operation in this embodiment is exactly the same as that in the third embodiment, so that it will be briefly described here. In the off-axis alignment system as shown in FIG. 11, since the mark pattern 120 is not measured via the projection lens 105, the wavelength range of the illumination light L to be used can be arbitrarily set, and the design There is an advantage that the degree of freedom is large, and broadband illumination light can be used in consideration of, for example, thin film interference in the resist layer.

In FIG. 11, the illumination light L passes through the half mirror 143 and the slit pattern 1 on the reference plate 140.
41, and the projection image of the slit pattern 141 is formed on the wafer W by the objective lens 143. The reflected light from the mark pattern 120 on the wafer W is reflected by the objective lens 14
2, the reference plate 140, the half mirror 143, and the condenser lens 108 to receive light by the photodetector 109. Therefore, when the projected image of the slit pattern 141 and the mark pattern 120 are relatively scanned, the photoelectric signal from the photodetector 109 has a waveform as shown in FIG. Also in this embodiment, the position of the mark pattern 120 with respect to the slit pattern 141 can be accurately detected by performing the waveform processing in the signal processing system 300, exactly as in the first to third embodiments. become.

By the way, in the above second to fourth embodiments, the slit pattern 106 on the reticle M and the slit patterns 131, 141 on the reference plates 130, 140.
Is replaced with a reflective mark pattern, and
It is obvious that the reflection / reflection type detection can be performed by using the optical system as shown in FIG. 6 (e). Next, a fifth embodiment of the present invention will be described with reference to FIG. In this embodiment, a case where the present invention is applied to a rotary encoder which detects an angle by a transmission / reflection type method will be described.

In FIG. 12, the reference plate (disc) 201 is configured to be rotatable by a motor 205, and the reference plate 201
A reflective mark pattern 202 (for example, FIG. 7B) according to the present invention is formed on the surface facing the angle detection plate 203. The angle detection plate 203 is a shaft 206
Is connected to the measured object 200 via. Needless to say, the measurement object 200, the angle detection plate 203, and the reference plate 201 need to have their centers of rotation substantially aligned. The method may be arbitrary. Also,
A transmissive slit pattern 204 according to the present invention is formed on the angle detection plate 203 so as to face the mark pattern 202. An optical system for detecting the relative angle between the mark pattern 202 and the slit pattern 204 is arranged on the surface of the angle detection plate 203 opposite to the slit pattern 204. The illumination light from the light source 211 is
Collimating lens 212, half mirror 213, and 1
After passing through the quarter wave plate 214, the slit pattern 204
And the mark pattern 202 is irradiated. The reflected light from the mark pattern 202 is returned to the slit pattern 2 again.
After passing through the 04, 1/4 wavelength plate 214 and reflected by the half mirror, the light is condensed by the condenser lens 215 on the light receiving surface of the photodetector 216.

If the relationship between the angular displacement of the reference plate 201 and the time t is clear, for example, as the motor 205 rotates at a constant number of revolutions, the above-mentioned operation is performed according to the waveform generated from the relationship of the light amount with respect to the time t. It is easy to detect the angle by the same waveform processing as the example. Although the angle detecting device has been described in the present embodiment, it can be applied to a linear position detecting device such as a linear encoder by replacing it with a linear movement. Further, in this embodiment, it rotates 2
Although one disk is used, the relative angle between the two cylinders can be detected in exactly the same manner by using two cylinders as shown in FIG. 7A. In the system, the light source 211 is a semiconductor laser and the photodetector 216
If it is a phototransistor and the half mirror 213 is a polarization beam splitter, the structure is almost the same as that of a pickup of a compact disc player, so that a small optical system can be realized, and it can be manufactured at low cost by utilizing parts. is there.

As described above, in the first to fifth embodiments, an example in which the position detecting device according to the present invention is applied to a projection exposure apparatus or a rotary encoder has been described. However, the present invention is an alignment system of a proximity type exposure apparatus. Needless to say, the present invention can be applied to a processing device such as an inspection device used in a semiconductor manufacturing process, and it can be applied to all devices that measure relative displacement of two objects. Needless to say, as schematically shown in FIG. 16, the apparatus configuration (in particular, the detection optical system) to which the present invention can be applied may be arbitrary.

[0052]

As described above, according to the present invention, highly accurate position detection resistant to noise can be performed with one set of patterns and by only one relative scan, which is necessary for prealignment. The time required for detection is not needed, and the detection time can be shortened. Further, since the pattern for pre-alignment and the device are not required and the adjustment thereof is not required, there are effects such as a reduction in the number of parts, a shortening of the manufacturing process of the device, and a cost reduction. In addition, since the combination of transmission and reflection can be freely selected according to the measurement system, there is an effect that it can be applied to a wide range of fields. Furthermore, the present invention can be applied to the angle detection device as shown in the fifth embodiment.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a pattern shape suitable for the present invention.

FIG. 2 is a diagram showing an example of a signal waveform obtained when the pattern shown in FIG. 1 is used.

FIG. 3 is a diagram showing another example of a pattern shape suitable for the present invention.

FIG. 4 is a diagram showing an example of a signal waveform obtained when the pattern shown in FIG. 3 is used.

FIG. 5 is a diagram showing another example of a pattern shape suitable for the present invention.

FIG. 6 is a diagram showing another example of a pattern shape suitable for the present invention.

FIG. 7 is a diagram showing another example of a pattern shape suitable when the present invention is applied to an angle detection device.

FIG. 8 is a diagram showing a schematic configuration of a projection exposure apparatus having a position detection device according to the first embodiment of the present invention.

FIG. 9 is a diagram for explaining a second embodiment of the present invention.

FIG. 10 is a diagram for explaining a third embodiment of the present invention.
..

FIG. 11 is a diagram for explaining a fourth embodiment of the present invention.
..

FIG. 12 is a diagram for explaining a fifth embodiment of the present invention.
..

FIG. 13 is a diagram showing an example of a pattern shape used in a conventional position detecting device.

FIG. 14 is a diagram showing an example of a signal waveform actually obtained when the pattern shown in FIG. 13 is used.

FIG. 15 is a diagram showing an example of an ideal signal waveform obtained when the pattern shown in FIG. 13 is used.

FIG. 16 is a diagram schematically showing an example of the configuration of a position detection device according to the present invention, particularly a pattern detection optical system.

[Explanation of symbols]

 11, 13, 15, 17 Slit pattern 101 Wafer stage 103 Reference member 104 Fiducial mark (slit pattern) 105 Projection lens 109 Photodetector 111 Interferometer 112 Motor 300 Signal processing system 301 First signal processing circuit 302 Second signal processing circuit 500 Main control system L Measuring light M Reticle W Wafer

Claims (1)

[Claims] 1. A means for relatively moving a second object having a second pattern in a predetermined first direction with respect to a first object having a first pattern, and detecting the position of the second object. Means, illuminating means for irradiating the first pattern and the second pattern with illumination light, and photoelectric detection for receiving light generated from the first pattern and the second pattern by the illumination light irradiation. Means for providing the first object and the second object
Based on a detection signal output from the photoelectric detection unit when the object is relatively moved in the first direction and a detection signal from the detection unit, the first and second objects are detected. In a device for detecting a relative position in one direction, the first pattern and the second pattern have a periodicity in the first direction and a specific pattern for partially changing the intensity of the detection signal. Is the first
Is formed in a second direction intersecting with the direction; and first signal processing means for generating first position information based on the periodicity of the detection signal from the photoelectric detection means; A second signal processing means for generating second position information based on the intensity of the detection signal of 1., and detecting the relative position based on the first position information and the second position information. Characteristic position detection device.