JPS60163432A - Pattern position detecting device - Google Patents

Abstract

PURPOSE:To enable the high-accuracy detection by performing sampling of signals at a high speed by a method wherein an object having at least one slit is moved at a high speed and at a constant speed and a pattern signal is synchronized with a reference clock pulse signal and is AD-converted for sampling. CONSTITUTION:A pattern for positioning 2 on a water 1 and a reference pattern 4 on a reticle 3 which functions as a reference are projected with magnification on a scanning plane of a slit 6 by magnifying optical systems 5 and 8 and an image is formed. Meanwhile, the plural slits 6 are formed on a rotary drum 11 and the magnified projection image of positioning pattern is scanned at a constant speed by the scanning means which is composed of a motor control circuit 12 for performing rotary control of a motor 7 by phase comparison servo method. Brightness and darkness of the light passing slits generated by slit scanning are photoelectrically transferred by a photoelectric transfer device 9 and these are inputted in an AD converting circuit 14. Then these are synchronized with reference clock pulse signals 18 outputted from a clock pulse generator 13 and photo-digitallized to be inputted in a memory circuit 15.

Description

Detailed Description of the Invention [Field of Application of the Invention] The present invention is applicable to an apparatus for exposing a circuit pattern such as L8Ie on a wafer or an apparatus for inspecting the dimensions of a line-exposed circuit pattern. The present invention relates to a pattern position detection device that detects a position.

[Background of invention]

Conventionally, in a pattern position detection device used in an exposure apparatus, for example, a pattern 2 to be detected on a wafer l and a reference pattern 4 on a reticle 3 are slit by magnifying optical systems 5 and 8, as shown in FIG. The enlarged projected image is scanned by a slit, and the light passing through the slit is converted into light 11L by a photoelectric conversion gain 9.
Convert to obtain pattern signal. After this pattern signal is stored as a digital signal, the pattern position is calculated. The slit scanning means consists of the slit 6 and the moving table 19.
The slit is located at the origin! As the length measuring device moves from (not shown) to the moving table 19, pulses are output from the length measuring device.

On the other hand, the light passing through the slit is subjected to photoelectric conversion, and then AD converted and sampled by the AD conversion circuit 14 in synchronization with the pulse signal from the length measuring device.

The AD-converted signal is stored in the memory circuit 15, and the pattern position is calculated by the arithmetic processing device (see Japanese Patent Publication No. 57-48853). However, with this sampling method, there is a limit to the response frequency of the length measuring device, and it is not possible to detect signals at high speed.'Furthermore, if the pattern signal has many noise components and the 8N ratio is small, the detection accuracy may be affected. This method has the drawback that the pattern position cannot be detected with high precision due to poor performance.

[Purpose of the invention]

An object of the present invention is to perform signal sampling at high speed when detecting the pattern position described above, and to perform 8N
It is an object of the present invention to provide a pattern position detection device that can detect a pattern position with high precision by obtaining a signal with a large ratio.

[Summary of the invention]

In order to achieve this object, the present invention moves an object having one or more slits at high speed and at a constant speed, and performs AD conversion and sampling of pattern signals in synchronization with a reference clock pulse signal. In addition, a method is used to remove noise components and increase the signal-to-noise ratio of the signal by integrating the signal sampled by scanning a slit once and the signal sampled by scanning a new slit. Eliminate the high speed, high 11th pattern! It is possible to detect gold.

[Embodiments of the invention]

Hereinafter, the present invention will be explained in detail using Examples.

FIG. 2 shows an example in which the present invention is implemented in a reduction projection exposure apparatus. The reduction projection exposure device transfers the IC, L8I# circuit pattern drawn on the reticle 3 to the reduction lens 5.
The reduced pattern is projected onto the p-wafer 1, and the reduced pattern is formed by miscellaneous light. In this reduction projection exposure apparatus, when aligning the pattern on the wafer that has already been formed in the previous process with the pattern to be newly formed, it is necessary to detect the position of the alignment pattern 2 on the wafer accurately and at high speed. There is.

In this alignment pattern detector, the wafer 1
The upper positioning pattern 2 and the reference pattern 4 on the reticle 3 serving as a reference are enlarged and projected onto the scanning surface of the slit 6 by an enlarging optical system 5.8 as shown in FIG. 3 and formed into an image. On the other hand, a plurality of slits 6 are formed in a rotating drum 11, and a scanning means consisting of a motor 7 and a motor control circuit 12 that controls the motor 7 on the rotation side by, for example, a phase ratio servo system, is used to form a positioning hole. The enlarged projected image of the turn is scanned at a constant speed. The brightness and darkness of the light passing through the slit caused by the slit scanning is photoelectrically converted by a photoelectric converter 9, such as a photomultiplier, and inputted into an AD conversion circuit 141'c. The AD switching circuit digitizes the output from the photoelectric converter and inputs it to the storage circuit 15 in synchronization with the reference clock pulse signal 1B output from the clock pulse generator 13.

On the other hand, for example, a reference point is set such that a reference signal is generated when the light beam 27 on the optical axis produced by the light emitting element 25 and the light receiving element 26 passes through a slit that scans the enlarged image of the alignment pattern. It is made up of mechanisms. When the reference signal A generated by passing through the reference point setting mechanism before the slit scans the enlarged image of the alignment pattern is input to the timing circuit 22, the timing circuit sends the ADR nip circuit 14 and the tag A to counter 16
It commands the start of D conversion and the start of the 100↑row of the reference clock pulse signal. Therefore, the storage circuit 15 inputs and stores the AD replacement value from the computer 24 up to a preset count value in proportion to the count value of the digital counter 160 that counts the reference clock pulse signal from the time when the slit passes the reference point. do.

FIG. 4 shows an example of signals stored in the storage circuit 15. The horizontal axis is the count of P-scan pulses relative to the reference, but since the slit is scanned at a constant speed by a scanning device, it can be converted into the position coordinate X of the slit. The relationship between the count value 1 of the reference clock pulse and the position coordinate XI of the slit is determined by the frequency Rf of the reference clock pulse, the scanning speed V of the slit, and the magnification M of the enlarged projection optical system. . is the sampling start point according to the reference signal. Moreover, the vertical axis is the AD conversion value Y of the photoelectric converter, which indicates the brightness of the light passing through the slit.

As described above, the memory circuit 15 stores the pattern signal obtained by the slit 6 scanning the enlarged image of the alignment pattern once, and the next slit 61 further enlarges the same alignment pattern. Scan the image and perform sampling. From the time when the slit 61 passes the reference point, the contents of the position corresponding to the callan MOL are transferred from the memory circuit 15 to the buffer circuit 21 for each count of the reference clock pulse.
This content and the ADi conversion value are added by the adder 19 and stored in the storage circuit 15 at a position corresponding to the count value. By performing this operation up to a predetermined count value, the memory circuit 15 stores the integrated result of two signals obtained by scanning the same enlarged image with the slits 6 and 61. This operation is performed the number of times set in the cumulative number counter 23 from the computer 24. From the signals stored in the memory circuit 15, the relative position Δ between the center coordinates XR and Xw of patterns 2 and 4 is calculated by the calculation processing circuit 17 using a statistical method.

By scanning the enlarged projected image of the pattern to be detected with a plurality of slits and integrating the respective pattern detection signals, noise components in the signal can be removed and a signal with a high S/N ratio can be obtained. The pattern position detection accuracy can be improved. In addition, when scanning a projected image with multiple different slits, the origin of the signal may be missed by the different slits due to the slits scanning the projected image passing a light beam that serves as a reference point indicating the start of signal sampling. It has become possible to eliminate positional changes and integrate four arrow signals obtained from different slits without changing their positions.

In addition, since there is no element that limits the scanning speed of the slit in the slit length device, which was a problem in the past, the slit can be scanned at cylinder speed, and the plurality 11 as described above is replaced by a plurality of slits. Even if signal sampling is performed twice, the time required for detection is 1/10 of the conventional method in which sampling is performed while measuring the amount of slit movement.

〔Effect of the invention〕

As described in detail above, according to the present invention, it is possible to perform high-speed sampling of the signal for detecting the turn position, obtain a signal with a large S/N ratio, and detect the pattern position with high precision. Make it familiar.

[Brief explanation of the drawing]

Figure 1 is a diagram explaining the conventional pattern detection method, Figure 2 is a diagram explaining the conventional pattern detection method.
FIG. 3 is a diagram illustrating an example of a pattern projected onto a slit scanning surface, and FIG. 4 is a diagram illustrating an example of a signal stored in a memory circuit. It is. 6...Slit, 61, 62, 63, 64, 65°6
6.67...Slit, 7...Slit scanning motor, 8...Enlarged projection optical system, 9...Photoelectric converter, 1
0...Mirror, 11...Drum, 12...Motor control circuit, 20...Amplifier, 25...Light emitting element, 2
6... Photodetector, continued from page 1 [Phase] Inventor Yoshio Kawamura 1-28 Hata, Higashikoigakubo, Kokubunji City, Hitachi, Ltd. Central Research Laboratory

Claims (1)

[Claims] 1. A device for scanning an object with a chromatic projection image of a pattern X-a on an object and photoelectrically detecting the position of the pattern, comprising at least one slit and ,
a scanning means for scanning the slit; the slit is installed to pass through the projected image before scanning;
The reference point setting mechanism operated by the slit itself and the light 1 of the pattern! A pattern position detection device comprising a circuit for sampling a converted signal and an arithmetic processing circuit for the pattern position. 2. The slit is composed of a plurality of slits, and the detection signal obtained by scanning any one of the slits is in addition to the detection signal obtained by scanning any one of the slits. 2. The pattern position detection device according to claim 1, wherein the pattern position detection device performs the integration.