JPS59101827A - Detecting optical system - Google Patents

Abstract

PURPOSE:To interrupt unnecessary diffracted light by the longitudinal line and lateral line of an actual element, to detect only desired luminous flux and to improve the accuracy of alignment by adequately combining a way of transmission of lighting luminous flux and a diaphragm of a light-receiving system. CONSTITUTION:The diaphragms C1, C2 of an illumination system Sc are arranged on a diaphragm plate Pc so that images C'1, C'2 on the imaging surface Pi surface of the diaphragms are formed to a section except a region surrounded by two pairs of parallel lines M, N. The diffraction patterns of the longitudinal and lateral lines often observed in the actual element of a wafer are not observed on the imaging surface because they are distributed in the lateral direction and the longitudinal direction while using each point in C'1, C'2 as the positions of 0 order beams and do not pass through an opening Co. When a line group pattern extending in the direction of 45 deg. is formed to a body surface 0 as a mask for alignment, the diffracted patterns pass through the opening Co of the diaphragm plate Pc because the distribution of diffracted light by the pattern extends in the direction of 45 deg. as shown in crosses, are detected in a photoelectric manner by an imaging surface I, and can be observed with eyes.

Description

DETAILED DESCRIPTION OF THE INVENTION Or, it relates to a device for visual observation or photoelectric detection, or both using a television receiver, which can clearly detect patterns written separately from existing butter/patterns. The present invention is suitable for use in devices for detecting alignment marks in semiconductor integrated circuit manufacturing equipment.

In recent years, as the integration of semiconductor integrated circuits has progressed, the width of circuits has become increasingly narrower. In response to this trend, in order to print fine electronic circuit (actual element) patterns onto wafers, an optical method is being developed that uses a high-resolution projection imaging lens or reflection system to transfer the circuit patterns on the mask. Projection exposure equipment and equipment that transfers images by bringing the mask and wafer into contact or in close proximity (proximity) are used.

When transferring, it is essential to align the mask and wafer in a specified positional relationship with a high precision of 1 μm or less prior to the exposure process, but alignment work is necessary to process a large number of wafers per unit time. Since automation is required, consideration must be given accordingly. Conventionally, when aligning a mask and a wafer, alignment marks are placed on each of the mask and wafer, and a microscope system is used to look into these marks and the two are aligned so that they overlap or are aligned in a fixed relationship. method is adopted.

At that time, the alignment mark is photoelectrically detected and automatically aligned, but it is possible to visually observe it for confirmation.)
In special cases, manual alignment can also be performed with visual observation.

Japanese Patent Application Laid-Open No. 54-5406 by the present applicant shows an example of automatic alignment), for example, the reflected light from the alignment mark is subjected to spatial frequency filtering at the pupil position in the microscope, and the diffracted light is detected. .

On the other hand, the location on the wafer for placing the alignment mark is when one of the chips is crushed and placed there,
It may be provided at the scribe line S (FIG. 1) between chips. However, since the real element pattern has the same optical effect as the elements constituting the alignment mark, if there is an alignment mark near the real element, the diffracted reflected light from the real element will mix with the diffracted reflected light from the alignment mark, causing alignment problems. make it difficult.

In order to avoid this difficulty, it is optimal to separate the alignment mark from the actual device by a sufficient distance, but this causes the problem that the effective area of the actual device decreases.

In addition to the alignment of the mask and wafer as described above, the alignment operation involves first measuring the coordinate position of the wafer with relatively rough precision, and then moving it to a predetermined stage for accurate positioning. There is an operation called alignment. In this case as well, if the position of the wafer is to be detected photoelectrically, it will be similarly affected by the actual device. However, actual semiconductor devices have the peculiarity that patterns are constructed based on orthogonal lines as required by design rules.

An object of the present invention is to clearly detect a pattern to be detected even if the pattern to be detected and other patterns are placed close to each other. In the embodiment of the present invention, in order to perform dark-field illumination on a wafer having an alignment mark formed by a group of inclined lines near the actual device pattern mainly consisting of vertical and horizontal lines, the aperture PC is used for illumination. The wafer illumination system installed in the optical path and the flat surface of the wafer
An imaging system incorporating another diaphragm P1 provided at a position optically conjugate with the diaphragm PC is arranged with the diaphragm P1 interposed as a mirror surface, and the shape of the aperture of the aperture of the illumination system diaphragm P is shaped into an image formation shape. aperture fi
For No. 11, the diffracted reflection light of the alignment mark is selected, and the other shapes are determined to be cut quickly. Further details of the embodiment will be explained with reference to FIG.

In the figure, O is the object plane, which is the surface of the wafer or mask. Also, Sc is the illumination system, La is the illumination light source, Ll is the first condenser lens, Pc is the illumination diaphragm plate (details will be described later), LO is the second condenser lens, Be is the beam splitter-1L, and is the microscope objective lens. be. Light source La
The emitted light flux illuminates the aperture of the aperture plate Pc via the 1:2/denser lens IJ. The light beam emitted from one point on this aperture undergoes an imaging action by the second condenser lens LO, is reflected by the beam splitter Be, and is reflected by the objective lens L.
.

The image is once formed at the pupil position, and then the object plane O is further illuminated via the objective lens TJO. The imaging relationship with is drawn as a solid line.

Next, Sl is an imaging system, LO is a microscope objective lens, Be is a beam splitter, 1Lr is a relay lens, and Le is an erector lens for image magnification. The light beam emitted from a point on the object plane 0 is converged by the objective lens Lo, and then focused on the intermediate imaging position 0 by the relay lens 'Lr. An image is formed on the final surface 1. This final surface is a photoelectric detection surface, a viewing surface visually observed through an eyepiece, or a television imaging surface. The imaging relationship of this system is drawn with a dashed line.

In addition, the light ray drawn as a solid line in this imaging system Si is a light ray that is reflected and returned when specular reflection is performed on the vertically epi-illuminated object O. After passing through the objective lens LO, the light ray is returned to the pupil lens.
The image is formed on the second image plane (Pl) via the relay lens Lr.

However, in this example, the second imaging plane is located inside the erector lens Lθ, but it may be located outside. A diaphragm plate P1 is arranged on the second imaging plane to cut unnecessary light. Se is a well-known photoelectric detection device, T is a photoelectric converter, and ■ is an electrical processing circuit. Since this part is not related to the purpose of the invention, the explanation will be omitted.

When dark-field illumination is performed using the above configuration, the direction of the striations forming the alignment mark written on the object and the distribution of diffracted light due to the striations at the pupil or pupil conjugate position within the microscope are shown below. explain.

First, a line pattern in an arbitrary direction is placed on the object plane 0, a pinhole is made in the center of the illumination diaphragm plate Pc, and the line pattern is illuminated. Then, on the aperture plane P1 in the imaging system, the diffracted light due to the pattern is lined up in a direction perpendicular to the lines of the pattern. For example, if the line pattern lv is placed in the vertical direction (this is the direction perpendicular to the front panel of the semiconductor printing apparatus) as shown in FIG.
If the line pattern h is placed in the horizontal direction as shown in FIG. 4, the diffracted light beams will be arranged in the vertical direction. When the position of the pinhole on the aperture plate Pc is shifted, the position of the O-order and lower diffracted lights on the aperture surface P1 of the imaging system also moves correspondingly. Therefore, if the aperture of the pinhole is widened, then all that is required is to integrate the diffracted light of each pinhole. That is, the distribution of the diffracted light on the imaging system aperture plane P1 is determined by the directionality of the pattern carved on the object and the aperture shape of the aperture plate PC of the illumination system. Note that the opening portion may be replaced with a mirror surface of the same shape.

Below, we will explain in more detail a dark-field illumination system that was realized based on this principle and can selectively detect patterns in a specific direction with high accuracy. Figure 5 shows the aperture position (Pc) in the illumination system.
) Preferred shape of opening provided in 0. ;02 is shown.

FIG. 8 shows an image plane when the apertures C1, 02 of the aperture plate Pc are projected onto the aperture plate P1 of the imaging system. 0' and 02' correspond to the images of the apertures C1 and 02, respectively. Figure 9 depicts the same arrangement as Figure 8, but also includes two sets of parallel lines, M,! : Add N. The direction of the parallel lines corresponds to the directions of the vertical lines and horizontal lines constituting the actual element, and each filament circumscribes the opening (C0) and is orthogonal to each other. As shown in the figure, the aperture of the illumination system is 10. t0
2) on the PC screen so that the image (a;, a;) on the 21 plane is outside the area surrounded by each of the two sets of parallel lines M and N and is symmetrical with respect to the optical axis. Place on top. Then, in accordance with the above-mentioned principle, the diffraction patterns of vertical and horizontal lines, which are common in actual elements on the wafer, are distributed in the horizontal and vertical directions, respectively, with each point in 0: , C4 as the position of the zero-order light. As a result, the pattern in these directions is not observed on the imaging plane (1) because the light does not pass through the aperture CO.

On the other hand, if a line group pattern extending in an oblique direction (particularly in a 45° direction) is provided on the object plane 0 as an alignment mark, the distribution of diffracted light due to this pattern will be the 8th
As shown by the X mark in the figure, it spreads in the 45° direction of the reverse inclination, so it passes through the opening Co of the diaphragm plate Pc. As a result, by appropriately combining the way the illumination light flux is given and the aperture of the light receiving system, unnecessary diffracted light due to the vertical and horizontal lines of the actual element can be blocked, only the desired light flux can be detected, and alignment can be achieved. Accuracy can be improved.

In addition, when performing bright field illumination as necessary, an aperture plate having a circular aperture shown in FIG. 6 is provided as the aperture plate Pc of the illumination system to illuminate the object plane 0.

! The image on the object plane is formed by the objective lens Lo, I) Ray lens Lr.

Then, it passes through the erector lens Lθ and forms an image on the final imaging plane (2). In other words, switching between bright field and dark field is performed at the aperture plane (P
This can be achieved by simply replacing the aperture plates with different opening shapes in c).

As described above, an optical system having the above-mentioned characteristics is suitable as an optical system for aligning a mask or a reticle with a camera in semiconductor manufacturing equipment. That is, if a mark in a direction different from the vertical and horizontal lines of the actual element, for example, in the 45c' direction, is used as the alignment mark, the optical system of the present invention can be applied as is. There are various ways to align the mask or reticle and the unit. There are methods of observing both of them at the same time and methods of observing them individually, but the present invention is suitable for both of them. Applicable.

The method of observing both at the same time is gloximity, of course in the case of contact burning, and the so-called T in the case of projection imaging.
This will be the TL method. When the projection optical system is a mirror system, white light can be used, but when it is a lens system, the optical system of the present invention is required to limit the wavelength range due to chromatic aberration. Specifically, it is essential to insert an interference filter or a sharp cut filter.

The arrangement of the unit As shown in FIG. 9(A) indicates the arrangement for simultaneous observation. However, U is the projection lens, q
is an illumination system), M is a mask, and W is a wafer.

In this arrangement, the image detected by AB is signal-processed through an electric processing system, and positioning is performed by a drive system (not shown).

On the other hand, the optical system of the present invention is also suitable as an optical system for observing and detecting a mask or reticle and a wafer individually. The most typical example is the off-axis alignment method used in step-and-repeat printing devices called steppers.

In the off-axis method, a reference position is set at a predetermined distance from a position where printing is performed using a projection optical system. The basic position is the observation position of the microscope, and after aligning the Unino with high precision, the wafer is transported to the position where it is printed using the projection optical system, relying on the precision of the stage. The optical system of the present invention is suitable as a microscope for setting this reference position. Since it is not limited by the projection optical system, it is possible to use a high-resolution objective lens, and it is also possible to use white light. FIG. 10 shows the arrangement of the off-axis alignment method.

The system depicted in FIG. 10 relates to pre-alignment of a projection type semiconductor printing apparatus, and FIG. 11 relates to alignment of wafer W and mask M. Note that ■ is, for example, a reduction projection lens, Q is a mask illumination system, PAS is a pre-alignment detector, ST is a moving stage on which a wafer is placed, and AS is a mask/wafer alignment detector.

Mask (reticle) and wafer alignment systems generally require high accuracy on the order of submicrons, but on the other hand, increasing the accuracy has the disadvantage of having to narrow the detection range (field of view). Therefore, when performing alignment through the projection lens (TTL) as shown in Figure 11, the Uni-No W (
It is necessary to have a function to measure the position shown in Fig. 10) and feed it. In other words, the moving stage S is
T is moved by a predetermined amount via a servo mechanism. The detection function for this purpose is called sea urchin/fly pre-alignment. In particular, Unino usually involves printing work using a different model of printing equipment, so it is often not possible to create a wafer shape that matches a specific model, and it is difficult to guarantee accuracy with mechanical pre-alignment. The UNINO 1 pre-alignment microscope system must have a wide detection range for the above-mentioned purpose.

For this reason, more vertical and horizontal lines of the actual device intrude into the field of view than with the Unino-Mask alignment microscope. Even when attempting to perform pre-alignment under such conditions, For example, as shown in FIG. 1, place an alignment mark for pre-alignment on the scribe line between the chips, and mark 4. ．． If the illumination system of the present invention is used (in the case of Fig. 1, illumination is provided by fan-shaped openings at the upper left and lower right; using the aperture shown in Fig. 5), the above-mentioned effect can be achieved. The position of the alignment mark can be detected with high precision within the field of view of the alignment microscope without being affected by the actual element. In this case, the fact that the alignment mark can be accommodated within the scribe line without providing a region for pre-alignment has a significant yield effect on manufacturing.

By adjusting the amount by which the wafer is fed to the vicinity of the printing position W' based on the mark position signal obtained in this way, the wafer can be fed with high precision to the location where it is aligned with the mask. I can do it.

The microscope system after feeding (As in FIG. 11) may adopt this method, or may use another method since the pre-alignment combination has been completed.

As another embodiment of the aperture shape according to the present invention, a case where the aperture of the imaging system in the microscope has a rectangular aperture or a polygonal shape as shown in FIG. 12 can be considered.

In this case, the aperture shape of the aperture of the illumination system within the microscope can be determined in accordance with the shape of the aperture υ in the imaging system. That is, when observing a state in which the aperture of the illumination system is imaged at the aperture position of the imaging system, the aperture of the illumination system aperture can be obtained by extrapolating each side of the rectangular aperture of the imaging system aperture. It is preferable to make it exist in a region that is not sandwiched between two sets of parallel lines that are orthogonal to each other. This situation is shown in FIG. 8, where the illumination system has an aperture as shown in FIG. 5, the imaging system has an aperture having the shape shown in FIG. 12, and the alignment mark is shown in FIG. 1.

By using a rectangular aperture as the aperture in the imaging system, there are the following advantages. The first advantage is that the ability to detect diagonal patterns with respect to vertical and horizontal lines can be further improved compared to a circular aperture (the part indicated by co in FIG. 8). This is because, among the diffracted lights (x marks) of the diagonal pattern in Fig. 8, the light that comes to the position P is
In the case of a circular aperture, it is blocked, but in the case of a rectangular aperture, it is possible to pass through the diaphragm.As a result, vertical,
While the horizontal line is the same as for the circular opening,
Regarding the diffracted light of the diagonal pattern to be detected, the amount of light received on the image plane 1 increases by an amount of P, which leads to an improvement in the detection accuracy of this pattern.

The second advantage is that when used as a replacement for the bright field microscope system, it not only increases the amount of light but also increases the resolution of diagonal patterns.

The aperture inserted into the illumination system can be modified in various ways depending on the shape of the pattern. For example, the aperture shape of the illumination system is
As shown in FIG. 15, it may be preferable to have apertures provided at four locations symmetrically and obliquely (for example, in the ±45° direction) with respect to the optical axis. By using the cross pattern shown in FIG. 14 as an alignment mark and using this aperture diaphragm, based on the principle described so far, the selectivity of the alignment mark with respect to the actual elements on the wafer is increased.
This leads to improved alignment accuracy.

As explained above, by adopting the illumination method according to the present invention that can selectively detect only oblique directions (optimally 45° direction) in the mask/wafer alignment microscope system of semiconductor printing equipment, it is possible to It has become possible to reduce the influence of diffracted light from vertical and horizontal lines and improve alignment ability. Furthermore, since the alignment mark can be placed close to the actual element, the restrictions on the placement location of the mark are significantly reduced compared to conventional techniques, and it is also possible to expand the effective area on the mask.

Further, by adopting this method to a general microscope length measuring mechanism, such as a pre-alignment mechanism of a semiconductor printing device, it is possible to improve the φ ratio and increase the signal detection accuracy.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a plan view showing an example of arrangement of alignment marks on a wafer. FIG. 2 is an optical cross-sectional view of an embodiment of the present invention. FIG. 3 and FIG. 4 are diagrams showing the relationship between line patterns and diffracted light, respectively. Fifth
6 and 7 are plan views showing the constituent members of the embodiment, respectively, and FIG. 8 is an explanatory diagram of the optical function. FIG. 9 is a conceptual diagram of a semiconductor printing device. FIG. 10 is an explanatory diagram of pre-alignment. FIG. 11 is an explanatory diagram of mask-wafer alignment. 12th
FIG. 15 and FIG. 15 are plan views showing constituent members of another embodiment, respectively. FIG. 14 is a diagram showing another example of alignment marks. In the figure, W+Fist/Wafer AM...Alignment mark 0...Object surface So...Microscope illumination system ■Ja...Light source Pc...Aperture plate Lc...Second condenser lens Be...・Beam splitter Ep...pupil LO...microscope objective lens Sl...microscope imaging system Lr...relay lens Le...erector lens PL...diaphragm platework...final Imaging plane CC...Illumination aperture 92 CO...Aperture opening Rono V, to...Line pattern Q...Illumination system U...Fist with the projection lens. Applicant Canon Co., Ltd. area 10 units=====Go” 811 layers

Claims (1)

[Scope of Claims] (1) An illumination system incorporating a first aperture means for dark-field illuminating an object, and a second aperture means disposed through the object at a position substantially optically conjugate with the first aperture means. An imaging system is provided with a diaphragm, and the image of the first diaphragm is projected onto the second diaphragm, and the image does not have a common aperture area, and the shape of the first diaphragm and the second diaphragm are different from each other. 1. A detection optical system characterized in that the shape of the means is related to block diffracted light from a pattern in a specific direction of an object. (2) The first aperture means is an area that is not sandwiched by mutually orthogonal parallel lines circumscribing the aperture of the second aperture means when projected onto the second aperture means, and is symmetrical to the optical axis. 2. The detection optical system according to claim 1, which is a light-shielding plate having an aperture. (6) The detection optical system according to claim 1, wherein the object is a wafer or a mask for semiconductor manufacturing, and the pattern in a specific direction is a real device pattern.