JPH11239952A - Alignment detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To surely discriminate a workpiece alignment mark by a camera. SOLUTION: A device is provided with a mask alignment mark 2 comprising a transparent part and an opaque part provided on a mask 1, a workpiece alignment mark 4 provided on a workpiece 3, a lighting system 9 lighting both the alignment marks, a CCD camera 8 provided to be opposed to both the alignment marks, an optical system 7 guiding an image of both the alignment marks to an image pickup part of this CCD camera, and an image processing part obtaining a relative deviation of both the alignment marks based on the image picked up in the image pickup part. An edge part is formed in the workpiece alignment mark 4, so as to be recognized as a corresponding dark part in the vicinity of the edge part in the CCD camera 8, by preventing at least partly reflecting light from in the vicinity of this edge part from passing through a lens 5 constituting an optical system 7.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of processing a metal work such as a lead frame or a printed circuit board by photolithography. The present invention also relates to an alignment detecting device for aligning a printing position of a pattern formed in a previous step with a mask pattern serving as an original.

[0002]

2. Description of the Related Art In an exposure apparatus for exposing and transferring a metal work, for example, a printed circuit board, a glass mask and a work are overlapped, and when the mark provided on the glass mask and the mark provided on the work are aligned, the work is performed. There is known a method of forming a hole in a printed circuit board as an alignment mark and aligning the hole with a microphone pattern.

[0003] The alignment apparatus is provided with an illumination optical system and a camera facing the glass mask, and performs epi-illumination on the printed circuit board via the glass mask from the illumination optical system. Then, the mask alignment mark and the work alignment mark are aligned. In this case, since the hole as the work alignment mark is clear on the printed circuit board, it can be transmitted and illuminated, and an image with good contrast can be obtained.

[0004]

However, facilities and steps for drilling holes in a printed circuit board are required.
There is a limit in the drilling precision, and there is a problem that when high-precision positioning is required, for example, when exposing a pattern in multiple layers, alignment precision is greatly affected.

Therefore, when high precision is required, there is a method of forming an alignment pattern on a printed circuit board by photolithography using a wiring layer material. In this case, the thickness of the pattern is 30 to 50 μm. However, when an alignment pattern is formed by this method, a reflection illumination system is used. However, in such an image, the contrast between the pattern and the base is small, and the pattern position may not be measured in some cases.

The present invention has been made in view of the above circumstances, and an object of the present invention is to form an edge portion on a work alignment mark so that the amount of reflected light from the vicinity of the edge portion and from other portions is obtained. An object of the present invention is to provide an alignment detecting device that can recognize a pattern of a work alignment mark by a camera as a difference between the two and can surely recognize the pattern by the camera.

[0007]

In order to achieve the above object, the present invention provides a mask alignment mark comprising a transparent portion and an opaque portion provided on a mask;
A work alignment mark provided on the work, an illumination optical system for illuminating both alignment marks, a camera provided opposite to both alignment marks, and an optical system for guiding the images of both alignment marks to an imaging unit of the camera. An image processing unit for calculating a relative shift between the two alignment marks based on an image picked up by the image pickup unit, wherein an edge portion is formed on the work alignment mark, and reflection from the vicinity of the edge portion is formed. At least a part of the light does not pass through the lens constituting the optical system, and the camera recognizes the light as a dark portion corresponding to the vicinity of the edge portion.

According to the above arrangement, the pattern of the work alignment mark can be recognized by the camera as the difference in the amount of reflected light from the vicinity of the edge of the work alignment mark and from the other. It can be reliably identified.

The image processing unit calculates a differential value of the image scanning data obtained by the imaging unit, and obtains the edge position of the work alignment mark based on the calculation result by the calculation processing unit, so that the dark portion can be obtained. The position of the edge can be obtained with high accuracy.

The work alignment mark is a combination of line-shaped patterns, and is preferably, for example, a cross, an L-shape or the like. Further, the image processing section obtains both edge portions of the line from image scanning data obtained by traversing the line-shaped pattern, and calculates and calculates a midpoint of both edge portions.

The work alignment mark is an area-shaped pattern, for example, a circle, a square, or the like. The image processing unit obtains an edge portion of the work alignment mark from data of all areas obtained by the imaging unit, binarizes the edge portion as a boundary line, and determines a pattern portion of the work alignment mark.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. 1 to 5 show a first embodiment, and FIG. 1 is a schematic configuration diagram of an alignment detection device. The glass mask 1 has a mask alignment mark 2 formed of a rectangular transparent portion 1a and an opaque portion 1b surrounding the transparent portion 1a. A metal work 3 such as a printed board or a REIT frame is provided below the glass mask 1, and a work alignment mark 4 is formed on the work 3. Here, the glass mask 1 is fixed, and the work 3 is supported by, for example, an XYθ table or the like, and is movable in the XYθ direction with respect to the glass mask 1.

Further, as a method of forming the work alignment mark 4 on the work 3, first, the work 3 is exposed to an alignment pattern, and then an etching process is performed. At this time, without etching until the pattern is half-etched on the work 3, that is, until it penetrates to the back side,
An etching time for stopping the etching up to a depth of about m is set. Then, an edge portion 4a is formed on the work alignment mark 4 as shown in FIG. 2A, and the vicinity of the edge is deflected. Note that the work alignment mark 4 is not limited to a concave portion, and may have a convex portion as shown in FIG.

On the upper side facing the glass mask 1, a CCD is provided via an optical system 7 comprising a lens 5 and a half mirror 6.
Cameras 8 are provided facing each other. The CCD camera 8 is provided with an image processing unit that images the mask alignment mark 2 and the work alignment mark 4 and calculates a relative shift between the two alignment marks based on the images. The half mirror 6 is provided with an illuminating device 9 for performing epi-illumination on the glass mask 1 and the work 3 via the half mirror 6.

Since the work alignment mark 4 is formed so that the vicinity of the edge is deflected, when the illumination device 9 receives epi-illumination, the reflected light L is obliquely reflected in the region of the etching droop from the edge 4a. You. In the optical system 7 according to the present embodiment, the main reflected light L from the region of the etching droop can be obtained by selecting an appropriate combination of the positional relationship between the elements and the numerical aperture NA of the lens 5, that is, selecting the optimal lens aperture. Has an angle equal to or larger than the opening of the lens 5, light cannot be taken in as an image of this area, and the image obtained by the CCD camera 8 has a black band shape as shown in FIG. Become.

When the differential value of the change in light amount is obtained from the pixel data of the edge image 10, the edge portion 4a is just dark, so that the contrast is emphasized, and the edge position information has a good S / N. Such measurement conditions can be obtained by stably forming the cross-sectional shape of the work 3 in the half-etched state and by using the lens 5 having a small numerical aperture.

When the numerical aperture NA of the lens 5 cannot be made sufficiently small, for example, when the selection conditions of the lens 5 are restricted, a variable stop for adjusting the lens aperture is inserted between the lens 5 and the glass mask 1. By adjusting this, the transmission of the reflected light L from the dripping region through the lens 5 may be restricted.

Conventionally, the numerical aperture NA of a lens of an optical system for detecting an alignment mark increases the resolution in order to sharpen the contrast of a pattern edge, that is, increases the depth of focus for the accuracy with which a mark can be positioned (focused). The maximum NA was set within a satisfactory range.

On the other hand, the present invention aims to ensure that the center position of the work alignment mark can be known even when the contrast between the work 3 and the work alignment mark is small. The numerical aperture of the lens 5 is reduced within the range, or a variable aperture is inserted, and the aperture is reduced by adjusting the aperture. That is, in the present invention, sufficient contrast can be obtained even if the reflection light L from the vicinity of the edge portion 4a is restricted from being incident on the CCD camera 8, that is, the aperture is reduced so that the resolution tends to decrease. The optical system 7 is set so that the depth of focus can be increased. By obtaining such an edge image 10, the center position of the work alignment mark 4 can be obtained as follows.

Next, a method of obtaining the pattern center from the edge position will be described with reference to FIG. 4. As shown in FIG. 4, for example, the light amount data 11 for each pixel of the edge image 10 on the line shown by the arrow in FIG. A difference value between adjacent data or data at equal intervals is obtained, and the difference value is set as an approximate differential value 12. The differential value data 12 has an outer peak 13a and an inner peak 13b as shown in FIG. Of these, the outer peak 13a corresponds to the actual edge portion 4a. This will be described later in detail. Since the difference value data is data for each pixel pitch, the position calculation below the pixel is performed by quadratic curve approximation calculation at several points around the difference peak value, and the X coordinate 14 corresponding to the peak 13a is calculated. 2 obtained
Since each of the X coordinates 14 indicates the edge position on both sides, the average of the two, that is, the X coordinate of the center point of the edge position on both sides is calculated, and the X coordinate of the center position of the work alignment mark 4 is obtained.

FIG. 5 shows a case where the work alignment mark 4 has a cross-shaped pattern. The X coordinates of the points A and B are obtained from the data on the predetermined broken line by the above-mentioned method, and the midpoint E is set to the mark center. Similarly, the Y coordinate of the point C and the point D is obtained from the data on the predetermined broken line, and the middle point F is set as the Y coordinate of the mark center. Normally, E and F are sufficient, but if the inclination of the work alignment mark 4 with respect to the mask alignment mark 2 cannot be ignored, the G and H points are further obtained above, and the intersection of EG and DH is set to the center of the work alignment mark 4. And

The reliability is improved by masking the change in contrast inside the dark part during data processing so that it can be ignored. It is also possible to incorporate a circuit that outputs a pulse when the image scanning data (voltage value) exceeds a certain threshold.

In a pattern in which the work alignment mark 4 has a cross shape or the like, the center position in the X-axis direction can be known by a vertical line, and the center position in the Y-axis direction can be known by a horizontal line. FIGS. 6 to 8 show a second embodiment, in which the work alignment mark 4 attached to the work 3 is an area-like pattern, and the method for finding the center of the work alignment mark 4 will be described later. This is the same as the first embodiment. FIG. 3A shows a circular pattern, and FIG. 3B shows a square pattern. The transparent portion 1a of the mask alignment mark 2 is formed in a circular shape, and the work alignment mark 4 is located in the transparent portion 1a. As in the first embodiment, the image processing unit obtains the edge portion 4 of the work alignment mark 4 from the data of the entire area obtained by the imaging unit, binarizes the edge portion 4a using the edge portion 4a as a boundary line, and obtains the pattern of the work alignment mark 4. The part is determined.

That is, the inside of the dark area bordering area is processed as a black area, and a binarized value is calculated including a plane portion where the reflected light of the etched area returns, and the center of gravity of the entire pattern is calculated.

Here, the binarization generally means, for example, that a pixel brighter than this is set to “1” and a dark pixel is set to “0” at a certain threshold. Therefore, the centroid of the pattern is determined by whether each of the pixels of the CCD arranged in a matrix is "1" or "0". Now, using a simple model shown in FIG. 7, the work alignment mark 4 indicated by the hatched portion in FIG. 7 is “1”, and the rest is “0”.
In the case of (1), the centroid of the pattern of the work alignment mark 4 is as shown in Equation 1.

[0026]

(Equation 1)

That is, the coordinates of the centroid of the pattern are (3,
1.5). Also, regarding the masking process,
Processing for intentionally determining a predetermined area to be "1" or "0". For example, in the case of FIG. 8, if the centroid of the mask alignment mark 2 is to be obtained, the work alignment mark 4 is in the way, so a predetermined area in the field of view of the camera 8 (an area indicated by a chain line; work alignment mark). 4). After masking the inside, the centroid of the mask alignment mark 2 is obtained as described above. When the centroid of the work alignment mark 2 is obtained, the outside of the dashed line may be masked.

FIGS. 9 and 10 show a third embodiment.
The same components as those in the first embodiment are denoted by the same reference numerals, and description thereof is omitted. In the present embodiment, the magnitude of the contrast between the transparent portion 1a and the non-transparent portion 1b of the mask alignment mark 2 is larger than the magnitude of the contrast between the bright portion and the dark portion of the work alignment mark 4 (when the illumination conditions are the same). If so) great. If there is a large difference between the condition for obtaining the best contrast when reading the work alignment mark 4 of the work 3 (such as the amount of illumination) and the condition for the masks 2 and 4, separate illumination devices 9 with different conditions are used. Each is best viewed.

That is, a first illumination optical system 21 facing the mask alignment mark 2 and a second illumination optical system 22 facing the work alignment mark 4 are arranged. In the case of this configuration, for each of the marks 2 and 4, the coordinates of the center position of the marks 2 and 4 with respect to the reference position of each CCD camera 8 are obtained.

The reference positions of the two CCD cameras 8 are set at a predetermined interval, and the work 3 is moved horizontally by the moving table 23 having sufficient positioning accuracy. If the reference position 8 is moved so as to overlap, the alignment is adjusted (for example, the work 3 is adjusted by the XY-θ fine movement stage).
The amount to be adjusted is determined. In the present embodiment,
When there is a large difference between the contrast condition (illumination light amount and the like) of the belt on which the alignment mark 4 of the work 3 can be taken and the respective conditions of the microphone 1, there is an advantage that images can be taken under the best conditions with separate illumination optical systems.

For example, if the center of the work alignment mark 4 is (a, b) and the center of the mask alignment 2 is (c, d) as shown in FIG.
db). The alignment may be performed on the alignment adjustment mask stage. The alignment adjustment may be performed before or after the movement of the moving table 23.

FIG. 11 shows a fourth embodiment. The same components as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted. The present embodiment is a configuration in a case where it is necessary to expose a pattern on both surfaces of the work 3 while aligning the phases of both surfaces. Since it is necessary to determine the difference between the center positions of the work alignment mark 4 on the lower surface of the work 3 and the mask alignment mark 2 of the mask 1 on the upper side of the opaque work 3, it is necessary to capture images with the respective CCD cameras 8 separately. In addition, it is necessary to match the reference positions (X and Y coordinates when the pattern surface is the XY plane) of the two CCD cameras 8 or to determine the amount of displacement if they do not match.

In each of the above embodiments, the work alignment mark 4 is provided on a part of the work 3 and the mask 1 is provided with the mask alignment mark 2 composed of the transparent portion 1a and the opaque portion 1b, as shown in FIG. As described above, the mask alignment mark 2 may be conversely placed in the area of the work alignment mark 4.

The actual image scanning data (same in other cases) obtained by traversing the alignment pattern of the work according to the present invention described in the first embodiment changes from light to dark (dark to light). Does not become a perfect rectangle, and the waveform is distorted more on the inside 24b than on the outside 24a corresponding to the edge 4a, as shown in FIG. Therefore, the peak of the differential value is higher at the portion corresponding to the edge portion 4a.

Further, since the CCD camera is a set of pixels having a finite size, data of a continuous curve is not actually obtained, but a set of data of a point which is a fractional point. Therefore, as described above, first, the scan data is stored in the memory of the image processing unit, and then, by calculating the difference between adjacent data in the calculation unit of the image processing unit, a temporary differential value is calculated and stored. A quadratic curve is approximated by taking several points around the maximum value (or the minimum value), and a position (ie, edge position) at which a more accurate peak value is obtained is calculated. If the error does not matter, the quadratic curve approximation may be omitted and the maximum value may simply be used as the peak value.

[0036]

As described above, according to the present invention, a work alignment mark formed on a work is used for aligning the mask with the work in an exposure apparatus for exposing and transferring a pattern formed on the mask onto the work. When detecting the amount of deviation between the mask and the workpiece,
Since the edge portion is formed in the work alignment mark and the vicinity of the edge portion is recognized as a dark portion by the camera, there is an effect that the work alignment mark can be surely recognized by the camera.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of an alignment detection device according to a first embodiment of the present invention.

FIG. 2 is an enlarged sectional view of the work alignment mark of the embodiment.

FIG. 3 is an operation diagram showing an edge image according to the embodiment;

FIG. 4 is an operation diagram for finding a pattern center according to the embodiment;

FIG. 5 is an operation diagram for finding a center of a linear work alignment mark of the embodiment.

FIG. 6 is a view showing an area-like work alignment mark showing a second embodiment of the present invention.

FIG. 7 is an operation diagram of the embodiment.

FIG. 8 is an operation diagram of the embodiment.

FIG. 9 is a schematic configuration diagram of an alignment detection device showing a third embodiment of the present invention.

FIG. 10 is an operation diagram of the embodiment.

FIG. 11 is a schematic configuration diagram of an alignment detection device according to a fourth embodiment of the present invention.

FIG. 12 is a diagram showing a modification of the present invention.

FIG. 13 is an explanatory diagram of image scanning data.

[Explanation of symbols]

 . DESCRIPTION OF SYMBOLS 1 ... Glass mask 2 ... Mask alignment mark 3 ... Work 4 ... Work alignment mark 4a ... Edge part 7 ... Optical system 8 ... CCD camera 9 ... Illumination device

Claims (1)

[Claims] 1. A mask alignment mark comprising a transparent portion and an opaque portion provided on a mask, a work alignment mark provided on a work, an illumination optical system for illuminating the two alignment marks, and a mask facing the two alignment marks. An alignment system comprising: a camera provided; an optical system that guides the images of both alignment marks to an imaging unit of the camera; and an image processing unit that calculates a relative shift between the alignment marks based on an image captured by the imaging unit. In the detection device, an edge portion is formed in the work alignment mark, and at least a part of reflected light from near the edge portion does not pass through a lens forming the optical system, and the camera corresponds to a portion near the edge portion. An alignment detection device characterized in that it is recognized as a dark part.