JPS6049886B2 - microscope - Google Patents

Description

[Detailed description of the invention] The present invention relates to a microscope with a deep depth of focus.

It is known that when the numerical aperture of a microscope is increased to increase imagination, the depth of focus decreases in inverse proportion to the square of the numerical aperture.

For this reason, the depth of an objective lens with a large numerical aperture is only a few microns, and what can be observed is only a small section of the target object. For example, when tracking phenomena that change from moment to moment, such as during biological observation, this kind of limitation becomes an extremely serious defect. On the other hand, in addition to such biological observation, the development of the semiconductor industry is also behind the recent demand for optical systems with a deep depth of focus.

That is, in the integrated circuit (IC) (LSI, etc.) technology brought about by the semiconductor industry, the minimum line width has already reached the micron order. IC manufacturing is done by layering multiple patterns on a semiconductor substrate such as silicon, but since the minimum line width is only microns, the overlay accuracy must be on the order of microns or submicrons. Ru. On the other hand I
C inspection also requires microscopic imagination, so the use of a microscope objective is inevitable, but since the measurement target, such as a silicon wafer, has irregularities that are deeper than the depth of the objective lens,
As I continued to measure it, I discovered that it had some drawbacks, such as the pit becoming erroneous. The fact that the microscope has a shallow depth of focus is particularly serious in the former positioning process. This is because in contact baking, which is commonly used for IC printing, or a printing method called the proximity method, the first step is to adjust the position by setting the distance between the mask with the printing pattern and the wafer to be printed to a distance of several tens of microns. Because you do it. If the mask and wafer come into contact with each other, there is no way to adjust the position, so it is essential to adjust the position at a distance of several tens of microns. It is clear that this value is considerably larger than the depth of . Therefore, for example, if pits are placed on the mask, the image of the wafer will be blurred, and conversely, if pits are placed on the wafer, the pits on the mask will be misaligned, resulting in this vicious cycle. The drawback was that alignment accuracy could not be expected. The object of the invention is to provide a microscope with a deep depth of focus to overcome the above-mentioned drawbacks.

This objective can be achieved by forming images of a plurality of observation planes that are different in the longitudinal direction with respect to the optical axis of the microscope onto the same plane using the optical path length correction means. Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a first embodiment, in which the microscope of the present invention is applied to an alignment scope of a semiconductor printing machine. 1 is a mask, and 2 is a wafer.

These masks and wafers 1 and 2 are illuminated by a known suitable illumination optical system (not shown). 3 is an objective lens.

The light beam from the objective lens 3 is split by a beam splitter 4 into light beams 5 and 6. 7 and 8 are mirrors.

9 is a half mirror for re-integrating the divided light beams 5 and 6.

Reference numerals 10 and 11 denote optical path length correction glass blocks disposed in the optical paths of the divided light beams 5 and 6, respectively.

The glass blanks 10 and 11 are used to form a mask image 13 and a wafer image 14 on the same plane 12. That is, in order to form the wafer image 14 through the optical path 6 and the mask image 13 through the optical path 5, for example, if glass blocks with the same refractive index are used, the glass blocks, What is the thickness of the block 11 relative to the thickness of the glass block 10? ■Just make it thicker. However, t is the air distance between the mask 1 and the wafer 2, α is the vertical magnification of the objective lens, and n is 3, the refractive index of the glass block. 15 is an eyepiece lens.

Since this embodiment is configured in this way, the eyepiece 15
Images 13 and 14 of the mask 1 and the wafer 2, which are spaced apart, can be observed simultaneously through the images. Therefore, a microscope with a deep depth of focus is obtained. Incidentally, not only the image 14 of the wafer but also the image of the mask 1 is formed through the optical path 6, and not only the image 13 of the mask but also the image 13 of the mask is formed through the optical path 5.
Images of the wafer 2 are also formed, but these images are defocus images and are therefore not very noticeable.
4. In this embodiment, the optical path length correction members 10 and 11 are placed in both the divided light beams 5 and 6, but they may be placed in one of the light beams. Further, although the case where the light beam is divided into two beams and two surfaces are observed has been described, it is also possible to divide the beam into three or more beams and observe three or more surfaces. FIG. 2 shows a second embodiment of the invention.

This embodiment is almost the same as the first embodiment, but the arrangement is slightly different, and the half mirror is used both for splitting the beam and for reintegrating the beam. That is, the light beams 5 and 6 divided by the half mirror 16 pass through the optical path length correction members 10 and 11 and head towards the mirrors 17 and 18. This mirror 17,1
The light beam reflected by the optical path length correction member 10,
11 and then toward a half mirror 16 where they are reintegrated. This reintegrated beam is directed by mirror 19 towards surface 12 and forms images 13, 14 of the wafer and mask on this surface. In addition, in the case of such an arrangement, considering that the light beams 5 and 6 pass through the optical path length correction members 10 and 11 twice, the members 10 and 11 are
11 thickness must be set. FIG. 3 shows a third embodiment of the invention.

In this embodiment, a mirror is used as the optical path length correction member. That is, the images 13 and 14 of the wafer and mask are formed on the same surface 12 by the mirror 18 which is arranged at a position shorter than the distance between the mirror 17 and the half mirror 16. FIG. 4 shows a fourth embodiment having a compact structure. This embodiment uses two half mirrors 20.
, 21 are used. And this half mirror 20,
21, when there is air between them, the interval is R.
In the drawing, these two half mirrors 20, 21 are supported by glass blocks, respectively.
The space is occupied by glass blocks. Since this fourth embodiment has such a configuration, an image 14 of the wafer 2 is formed on the plane by the light flux that has passed through the half mirrors 20 and 21, and is further reflected by the half mirror 21. An image 13 of the mask 1 is formed by the light beam transmitted through the half mirror 20. Although the glass block for optical path length correction may be fixed in advance, FIG. 5 shows a variable optical path length correction member applicable to the first, second, and fourth embodiments. That is, (a) is formed by two glass blocks 22 and 23 having the same refractive index,
By moving one block 23 in the direction of the arrow, a desired optical path length correction member can be obtained. In addition, (b) shows a glass plate 24 whose wall thickness varies in stages, and this glass plate 24
A desired optical path length correction member can be obtained by introducing the light beam into the light beam step by step. In addition, (c) has a wall thickness of category 25 to 2.
By rotating different glass disks every 8,
A desired optical path length correction member is obtained. Incidentally, when these optical path length correction members are used in the fourth embodiment, half mirrors may be provided on each of the front and back surfaces. In the example of FIG. 3, the mirror 17 or 18 may be movable so that the length of the optical paths 15, 6 can be adjusted. This method is highly effective when just observing with a normal microscope, but when using this method for further measurements, more detailed considerations are required.

In other words, when measuring the relative position of two objects, such as the alignment of the mask and wafer that has been explained so far, if the magnification of the mask image and the wafer image are different, a large error will result. invite. This situation is shown in FIG. 29 and 30 are the same size object and lens 31
The images are focused on 32 and 33.

29 and 30 are slightly shifted on the optical axis, and this relationship is similar to the relationship between the mask 1 and the wafer 2.

In the present invention, the positions of the images 32 and 33 are made to coincide with each other by the method described above, but the sizes of the images are not changed, so an error occurs in observing the relative positions of the two objects. This must be taken into account in the case of precision measurements. In order to eliminate such errors, it is desirable to use a telecentric microscope objective lens.

In a telecentric objective lens, the pupil position is at the front focal point of the lens, so the principal ray at the center of the luminous flux is parallel to the optical axis. This situation is shown in Figure 7.
34 pupils are placed at the focal position of the objective lens 3.
By using such a lens, object points that differ in the front and rear directions with respect to the optical axis can enter the imaging system without being laterally shifted in the direction perpendicular to the optical axis. However, the accuracy is not actually sufficient just by using a telecentric objective lens. In FIG. 7, a relay lens 35 is arranged, but unless the relay lens is also telecentric, an error in magnification will occur.
Since the relay in Figure 7 is not telecentric, the chief ray after exiting the relay lens is tilted with respect to the optical axis. Images 13 and 14 of objects 1 and 2 are formed at the positions shown in the figure, but after A and B, which are in a positional relationship parallel to the optical axis, are aligned with each other by the method of the present invention, they become A''.
and B'', the positions do not match, and it is observed as if there is a positional deviation of Δ. Figure 8 is an example where the relay lens is telecentric up to the relay lens 36. The principal ray, which is the center of the latter luminous flux, is parallel to the optical axis as in the case of entering the objective lens, so A″B
'' is observed without any lateral deviation.In other words, if the relationship between AB and A is correctly
It can be reflected in "B". The degree of telecentricity of the relay lens varies depending on the measurement accuracy, 13 and 14
It is important to select an appropriate one because it depends on the degree of displacement of the position, and telecentricity may be broken as long as the lateral displacement is allowable. In semiconductor-printed alignment scopes, the position of the objective lens moves, so the telecentricity of the relay lens is slightly disrupted, but it is possible to design it to an extent that does not cause any problems. If the distance between the observation surfaces 1 and 2 is very close, the usual microscope systems shown in FIGS. 1 to 4 may be sufficient. In addition, as shown in FIG. 9, in a system that requires precision, when both the objective relay and the objective relay are made telecentric, the optical system part 37 of the present invention using a half mirror is used.
Of course, it is placed between the relay and the imaging plane. As described above, the present invention effectively uses a half mirror to increase the depth of focus, and by adding the idea of a telecentric optical system to this, it is possible to perform precise measurements. Great effects are expected in image observation and detection.

Although the embodiments have been explained using a wafer and a mask as an example, it is well understood that if you consider the wafer and the mask as observation surfaces for observing each with a microscope, it can also be applied to ordinary microscopic observation. .

Furthermore, in the description of the embodiment, two objective lenses are required when used as an alignment scope, but here, for simplicity, only one objective lens is taken as an example.

In the θ proximity method and contact method, the distance between the mask and wafer is shortened or brought into contact after alignment, but the optical path length correction unit is linked depending on the amount of change in the distance at that time. By moving it, you can always keep both the mask and wafer in focus. Therefore, for example, if using the proximity method, if the gap is changed from 50μ during alignment to 20μ during baking, by moving the adjustment member in response to this change, the distance between the mask and the wafer can be adjusted after alignment and for baking. It is possible to clearly observe whether any positional deviation occurs when it is shortened. In addition, by reversing this method, it is also possible to calculate the distance between any two surfaces based on the amount of movement of the optical path length correction means.

[Brief explanation of the drawing]

FIG. 1 is a diagram for explaining the first embodiment of the microscope of the present invention, FIGS. 2, 3, and 4 are diagrams for explaining the second, third, and fourth embodiments, respectively. FIG. 5 is a diagram for explaining the variable optical path. A diagram showing the length correction member, Figure 6 is a diagram explaining a non-telecentric imaging system, Figure 7 is a diagram explaining an imaging system consisting of a telecentric objective lens and a non-telecentric relay lens, and Figure 8 is a diagram explaining the telecentric objective lens. FIG. 9 is a diagram illustrating the fifth embodiment. In the figure, 1 is a mask, 2 is a wafer, 3 is an objective lens, 4
, 9 are half mirrors, 7 and 8 are mirrors, 10 and 11 are optical path length correction members, 13 and 14 are masks and wafer images, and 15
is the eyepiece.

Claims (1)

[Claims] 1. A means for splitting the light beam from a telecentric microscope objective lens into a plurality of light beams, a means for re-integrating the light beam divided into the plurality of light beams, and a telecentric microscope relay lens for receiving the reintegrated light beam. In order to form images on the same plane from a plurality of observation planes that are located at distances apart on the optical axis, the optical axis between the observation planes is set in at least one optical path of the light beam divided into a plurality of pieces. A microscope capable of observing a plurality of observation planes at different positions on an optical axis, characterized in that an optical path length correcting means for correcting the optical path length by the vertical magnification of an upper distance X microscope objective lens is disposed.