JPS5953482B2 - Relative position deviation measurement method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a relative position deviation measurement method for measuring the relative position deviation of two members, and in particular, the effective relative position deviation θ when aligning a mask and an exposed object in an exposure process. Regarding measurement methods.

Configuring a device that detects the relative position of two members such as a mask and an exposed object in an exposure process and performs relative alignment is an important issue in photographic technology and pattern forming technology as its application. It is becoming.

However, this technique has different difficulties from the length measurement technique in that marks that serve as a reference for relative position must be formed on each member as part of the pattern. This differs from measurement that relies on a scale held as a reference by the device, such as a high-precision scale.
It is necessary to measure a different object each time, which is the pattern mark on the object. Also, in the photographic process, it must be taken into account that the shape of the mater changes as it passes through the process, becoming thinner or thicker. Due to the above conditions, such a relative position measuring method requires detecting the center of the mark and measuring the relative positional deviation of the center.

The following two are representative of such conventional techniques.

(1) As shown in Fig. 1, the center position deviation of the slit marks 1a and 2a made in two patterns 1 and 2 is measured.One pattern 2 is given a sinusoidal vibration, and it is irradiated with light to transmit the transmitted light. This is a so-called vibration method in which the light is received by the light receiving element 3.

The waveform of the light-receiving element 3 is as shown in FIG. 2, and the deviation of the center position is detected by the presence of frequency components as indicated by dotted lines.
Since this component is extracted using a bandpass filter, other noise components are also removed, allowing for high-sensitivity detection with a good S/N ratio. (2) Two patterns 1 and 2 as shown in Figure 3
If one pattern 1 is provided with a slit 1a and the other pattern 2 is provided with a mark 2b, and the light spot is scanned in the direction of the arrow, the output waveform of the light receiving element 3 will become as shown in FIG.

The difference between Tl and t2 of this waveform becomes a relative positional deviation if the scanning speed is constant. With such conventional techniques, vibrating the pattern in the former case results in pattern blurring during the photographic process, and can only be used for initial settings.

That is, measurement cannot be performed during exposure. Also, the pattern is not static, so handle it carefully. is difficult. The latter tends to be disturbed by even a small scratch, and the time measurement accuracy of Tl and t2 is not very high, and it is also difficult to reduce the size of the spot, and it is also difficult to maintain a constant scanning speed.

In view of the above points, it is an object of the present invention to provide a relative position deviation measuring method that is easy to measure and has high measurement accuracy.

To achieve this objective, the relative position deviation measuring method of the present invention
first and second alignment patterns are provided on the respective planes of the first and second members that are in contact with each other, and the size of the second alignment pattern is configured to be larger than that of the first; Light scanning is performed with a beam spot that is larger than the difference in size between the first and second patterns and smaller than the first pattern, and one end of the first pattern and one end of the second pattern are scanned. A light receiving element detects a signal corresponding to a first distance to the part and a second distance between the other end of the first pattern and the other end of the second pattern to determine the relative position deviation. In the relative position deviation measurement method, the first
and a signal corresponding to the second distance is peak-held, and a difference in magnitude of the peak-held signals corresponding to each distance is extracted to measure the relative position deviation.

The present invention will be explained below based on one embodiment. FIG. 5 is a block diagram of an embodiment of the present invention, and FIG. 6 is a waveform diagram of each part of the block diagram in FIG. It has part 1″a, and 2
'' is a second member such as a wafer and has an opening 2''a as an alignment mark.

5, 6a, 6b, and 7 are scanners, first slit, and second slit, respectively.
A slit, a light source, and a first member 1″ and a second member 2
” is optically scanned.

4 is a half mirror that passes the light from the scanner 5 and guides a part of the reflected light from the first member 1'' to the light receiving element 3; 8 is a condensing lens; 9 is an optical scanning drive circuit; 10 is a first peak hold circuit such as an integrating circuit, 11 is a sampling circuit, 12 is a sample pulse generation circuit, 13 is a second peak hold circuit, and 14 is a bandpass filter.

This operation will be explained in conjunction with the waveform diagram in FIG.

The light irradiation surface of the first member 1'' is a reflective surface, and the light irradiation surface of the second member 2'' is a non-reflection surface.
Therefore, when the light from the light source 7 is scanned and irradiated by the scanner 5, the light is reflected between BB'' and A-A'' of the first member 1'' and guided to the light receiving element 3. This light receiving element 3
The output appears as shown in FIG. 6a. In the conventional method, for example, the accuracy was determined because the optical scanning speed was required to be constant, but in this method, the A of this waveform
．． Focusing only on the difference in light intensity between the two points b, averaging this,
In addition, the extraction is performed with high precision, and the key point is to ensure that the optical scanning speed does not become an error factor.

That is, A. The heater value between b is sampled and held in the first peak hold circuit 10, and the waveform is corrected as shown in FIG. 6b. Five resets of the peak hold are performed by pulses generated from the optical scanning drive circuit 9 indicating both poles of scanning vibration (set at speed 0). Optical scanning drive circuit 9
It is necessary to keep the scanning frequency accurate based on something like a crystal oscillator.

However, even if the amplitude of the optical scanning or the vibration of the optical spot is slightly shifted from the center of the vibration with respect to the alignment mark, this should not be a cause of error.

Therefore, the waveform shown in FIG. 6b obtained in this way is generated by the sampling pulse generation circuit 12 in synchronization with the fundamental vibration, and the sample circuit 11
If sampled, a waveform like that shown in FIG. 6C can be obtained.

Then, if this sampler pulse is passed through the second peak 5 hold circuit 13 which applies a strain set,
A waveform like this is obtained. If this waveform is passed through a fundamental frequency band pass filter 14, an AC output (shown by the dotted line in FIG. 6d) proportional to the difference between the peak values A and b is obtained. Furthermore, if this is synchronously detected, the polarity of the deviation can also be determined. This output is supplied as a positioning command to a servo motor (not shown) that positions the first member 1'' or the second member 2''. In Figure 5, even if the intensity of the light beam that irradiates the mark between A and N and between B and B'' is the same, the light receiving position of the light receiving element that receives the light is at a different position if viewed minutely. It is also possible that the optical-to-electrical conversion sensitivity is different, which could be a cause of positional error.

In order to alleviate this error factor, it is desirable to expand the light in front of the light receiving element and receive the light in as wide a range as possible. In this embodiment, the light receiving element 3 is disposed through a lens 8 at a position different from the focal point of the lens to disperse the light. or,
The points indicated as AA'' and BB'' in FIG. 5 must be smaller than the scanning beam spot in order to accurately convert the distance into an amplitude difference of light.

Another condition is that the beam spot must not irradiate AB at the same time. If the beam spot is designed to be too small, even small scratches in the pattern tend to have a large effect on the signal, which is undesirable. As a countermeasure against this problem, it is preferable to make the light beam and pattern elongated in the direction perpendicular to the plane of the paper in FIG. 5. That is, as shown in FIG. 7, the light beam is shaped into an ellipse so as not to simultaneously irradiate AB. Also, the alignment mark of the first member 1'' is located at the opening 1''a.
Instead, a non-reflective surface 1''b may be provided. In FIG. 5, the wall surface of the opening 2'a of the second member 2'' is an inclined surface, but if it were a mirror surface, The reflected light is deflected to the edge, and only the reflected light from the surface of the first member 1a enters the light receiving element 3, allowing for a more accurate comparison of AN and BB''. As an alignment pattern, in FIG. 5, the non-reflection surface portion of the first member 1''a is a hole, but through holes are often problematic. In Fig. 8, instead of this, there is a recess 1"C consisting of a slope. Even though this recess is a reflective surface, the light is deflected to the side, so from the perspective of the light receiving element, it is the same as a non-reflective surface. As another embodiment, it is possible to consider a method in which the illumination sources are intermittent so as to obtain a signal as shown in FIG.

This is an improvement to avoid the signal peaks A and b appearing twice in one cycle in Figure 1, which makes signal processing somewhat difficult. The irradiation light is cut off when the reflected light decreases due to the movement of the spot and reaches the minimum level.Next, the light source is turned on again at the turning point of the scanner, and this time, on the return stroke, the irradiation light is turned off. When the reflected light passes through the heater and reaches a minimum point, the light is cut off again. In this way, by irradiating A-A'' on the outward path and irradiating B-B'' on the return path, a signal as shown in FIG. 9a can be obtained. This is peak held and reset at the vibration return point. If the signal is passed through the peak hold circuit 10 shown in FIG. 5, the result will be as shown in FIG. 9b.
By passing this through a sample gate 17 circuit synchronized with the vibration, the signal shown in FIG. 9C can be obtained. This signal can be made into an alternating current signal proportional to the difference in Ab by passing through a band-pass filter with a constant vibration frequency.
The advantage of this is that there are fewer unnecessary signals in the original signal, so the width of the sampling gate can be longer.
Compared to the figure, the power of the signal in Figure 9 is greater and the signal can be obtained even if it passes through a filter.Therefore, there is no need to pass it through the peak hold circuit 13 again as in Figure 5, It becomes possible to reduce the influence of errors. As described above, according to the present invention, it is possible to accurately measure the relative position deviation regardless of the error in the optical scanning speed, and since the alignment mark is stationary, alignment is possible even during exposure, which is extremely practical in terms of Useful.

[Brief explanation of the drawing]

1, 2, 3, and 4 are explanatory diagrams of conventional measurement methods, FIG. 5 is a configuration diagram of an embodiment of the present invention, and FIG. 6 is a waveform diagram of each part of the configuration diagram in FIG. FIG. 7 is an explanatory diagram of another embodiment of the present invention, FIG. 8 is an explanatory diagram of alignment marks used in the embodiment of the present invention, and FIG. 9 is a waveform diagram of another embodiment of the present invention. In the figure, 1 is a first member, 2 is a second member, 3 is a light receiving element, 4 is a half mirror, 5 is a scanner, 6a and 6b are slits, 7 is a light source, 8 is a lens, and 9 is a scanner drive unit, 10,
13 is a heater hold circuit, 12 is a sample pulse generation circuit, 1] is a sample circuit, and 14 is a band pass filter.

Claims (1)

[Claims] 1. For relative positioning on the respective planes of the first and second members that are in contact with each other, the central portion is a low reflective surface and the other surfaces are reflective when viewed from the light receiving element. A first alignment pattern and a second alignment pattern having a light passing portion larger in size than the center portion of the first pattern and the other portions having low reflection are formed, and these are overlapped. At the same time, optical scanning is performed with a beam spot that is larger than the difference in size between the center portions of the first and second patterns but smaller than the size of the center portion of the first pattern, so that one of the first patterns is the first pattern between the end and one end of the second pattern.
and a second distance between the other end of the first pattern and the other end of the second pattern by detecting with a light receiving element to measure the relative positional deviation. In the positional deviation measuring method, the relative positional deviation is measured by peak-holding the signals corresponding to the first and second distances, and extracting the difference in magnitude of the signals corresponding to the peak-held distances. Features: Relative position deviation measurement method. 2. Claim 1, characterized in that the peak-held signal for extracting the difference in magnitude of the signal is sampled in synchronization with optical scanning, and the difference in magnitude of the sample output is measured. relative position deviation measurement method. 3. In order to extract the difference in the magnitude of the signal, a light scanning frequency component is extracted after shaping the peak hold signal, and a deviation is obtained from the amplitude of the frequency component. Position deviation measurement method.