JPS60100004A - Regulating device of space - Google Patents

Abstract

PURPOSE:To enable the maintenance of a space between a wafer and a mask at a fixed value in any process by a construction wherein a rotary and non-symmetrical means of altering a luminous flux width is provided in optical paths through which luminous fluxes from two substances spaced from each other reach an interference fringe forming means. CONSTITUTION:A space between a central peak Fc and a side peak Fs detected by a photodetector 30 such as a photodiode is compared with a reference value in a signal processing circuit 36, and based on the result of this comparison, an actuator control circuit 37 drives actuator rods 35, 35' and 35'' vertically. Transferring the relative positions of a mask and a measuring optical system, or measuring a space between a plurality of points, the result of the measurement and a reference value are compared with each other in a signal processing circuit 30. Based on the result of this comparison, the rods 35, 35' and 35'' are controlled vertically through the intermediary of the actuator control circuit 37, and thereby a space between a mask and a wafer can be controlled at a value tobe prescribed.

Description

DETAILED DESCRIPTION OF THE INVENTION The invention relates to a device for adjusting the spacing between a transparent object capable of reflecting at least some light and a transparent or opaque object also capable of reflecting light to a desired extent.

For example, in an X-ray exposure device, which is a type of pattern printing method, or in a gloximiry printing device that uses ultraviolet rays, the minute distance between the mask and the wafer is precisely adjusted to a predetermined value, and the distance between the two is Must be kept fully parallel.

As a method for this adjustment, a dummy mask that protrudes by the required distance apart is used in place of the actual mask, and this and the wafer are connected. Mechanical methods have been used in which the upper photoresist surface is directly contacted with Ueno/Ren, but this not only damages the photoresist surface and lowers the yield of semiconductor devices. There were problems such as errors in parallel alignment and spacing due to photoresist adhering to the dummy surface.

In order to avoid these problems, the mask and the photoresist surface of the Ueno Σ should face each other in a non-contact state, and then alignment can be performed directly. In 1975, a method using multiple white interference needles was introduced.
I applied on October 27th. (Public Patent Publication No. 52-5257
No. 9) Prior to explaining the contents of the present invention, the contents of this earlier patent will be explained. 0 Figure 1 is based on another earlier patent filed by the same applicant (Published Patent Publication No. 11852-42, filed on June 28, 1975).
This is an example of a white interference needle used in the present invention using a Wollaston prism, as shown in No. 60).

In the figure, 1 is a light source having a wavelength width such as a tungsten lamp, and a light beam 2 from this light source is transmitted to an object to be measured (for example, a plastic film, a mask, and a Ueno ivy) with a refractive index of n and a thickness of d. (air spacing, etc.) 3. A part of the light beam 2 is reflected by the first surface 31 of the object to be measured 3 and becomes a light beam 5, and another part is refracted by the first surface 31 of the object to be measured 3 and reflected by the second surface. The light flux becomes 4, and the rear surface 3
1 and becomes a light beam 6. Here, for convenience of description, the light beam 5 is a single wavefront 51, and the light beam 6 is a single wavefront 52.
Let it be represented by If the wavefront 51 and the wavefront 52 are selected to be at positions with equal optical path lengths from the light source l, the optical path length difference between them will be 2 ndcosφ. Here, the angle φ is the angle of incidence of the object 3 to the second object 32. In the figure,
Reference numeral 50 denotes an interference fringe forming section, which is arranged to receive wavefronts 51.degree.

The Wollaston prism 60 consists of a prism 54 which is cut out of a birefringent crystal material such as quartz or calcite so that its optical axis is perpendicular to the plane of the paper, and a prism 55 which is cut out so that its optical axis is vertical to the plane of the paper. 1 combined. The polarizer 53 is arranged so that the optical axis of the crystal 54, 55 and its optical axis are 45', and the analyzer 56 is arranged so that it is parallel or perpendicular to the polarizer 53. There is.

With the above configuration, the light beam that has become linearly polarized light vibrating in the 45' direction on the plane of the paper by the polarizer 53 enters the prism 54 and is divided into an ordinary ray and an extraordinary ray, and both light beams enter the prism 55, Until then, the ordinary rays became extraordinary rays,
Moreover, the extraordinary ray travels as an ordinary ray. Due to the difference in refractive index between the ordinary ray and the extraordinary ray of the birefringent crystal, the light beams emitted from the prism have mutually inclined wavefronts 9, and as a result, the wavefronts 51 and 52 are respectively wavefronts 5r.

5r, 52', and 52', and the action of the analyzer 56 results in interference fringes having an intensity distribution. The optical path difference imparted to these wavefronts makes the refractive index of the birefringent material for the ordinary ray n. , when the refractive index for the extraordinary ray is 06, it is given by 2(rla −no ) ·Y tan θ. Here, the value of Y is a coordinate value measured in a direction perpendicular to the optical axis with the point where the two prisms have the same thickness as the origin, and θ is the apex angle of the prism.

Wave surface s f To wave r15r, [Uki52'To wave Ef
The point where fi52'(!: interferes to form interference fringes is near point 58 in the figure, near Y=0, and the peak of white interference fringes occurs here.

On the other hand, wavefront 51' and wavefront 5! Interference occurs near point 57 in the figure, and wavefront 5z and wavefront 5r interfere near point 59 in the figure. The points where these peaks occur are
This is the point where 2ndcosφ and 2(n, no) 'Y-tanθ are equal, and at these points, Wollaston
The prism P creates an optical path length difference. From the above relationship, Y=±ndcosφ/1(na po)
A side peak appears near tanθ). When the resulting white interference fringes are projected onto the photodetector 30 through the lens 69, as shown in FIG. .

Fa' is raining on the side peaks. Now, using a Wollaston prism with a constant apex angle θ,
When the refractive index of the object to be measured is constant, the interval between Fc and F8 or the interval between Fc and Fs' is proportional to the thickness of the object to be measured. Alternatively, it can be replaced with interval measurement.

FIG. 3 is an example in which white interference fringes are used to adjust the distance between the mask and the mask, as shown in Japanese Patent Publication No. 52-52579. The light flux emitted from the white light source 1 is condensed by the condenser lens 23, 24, reflected by the semi-transparent mirror 26, and then becomes a parallel light flux by the objective lens 27. The photoresist 140 on the mask 11 and Uninor-12 is
to illuminate. Reference numeral 25 denotes a color filter inserted into the optical path of the illumination light flux in order to prevent the photoresist 14 used at this time from being exposed to light, and is usually a yellow filter that absorbs light flux with a wavelength of 500 nm or less. Now, when the optical system is set to a position where it illuminates the transparent part of the mask 11, the mask surface 11 and the photoresist surface 201 are exposed.
The light flux reflected by
′ makes it a parallel beam of light. Although the transparent part of the mask 11 is used here, this method uses invisible light
It is also applicable to pattern printing using lines. This is because the masks used for X-ray exposure are very thin.
This is because it is considered to be approximately transparent in the long wavelength side of the visible region and in the infrared region.

A polarizer 53. 2nd stone prism 6
0. By arranging the interference fringe forming section 50 consisting of the analyzer 56, white interference fringes having information regarding the spacing between the mask and the wafer at the currently illuminated photomask position are formed on the Wollaston prism 60. .

The shape of this interference fringe is imaged onto the photodiode array or image pickup tube 30 via the projection lens 69,
At the same time, the light is reflected by the nof mirror 28 and observed by the observer 14 through the eyepiece lenses 32 and 31.

To automatically adjust the spacing to a constant size, proceed as follows. The interval between the center peak Fc and the side peak F8 detected by the photodetector 30 such as a photodiode array is compared with a reference value in the signal processing circuit 36, and the actuator control circuit 37 controls the actuators 35 and 35' according to this result. , 35' up and down to move the relative position of the mask and the measurement optical system, or alternatively, measure the distance between multiple points by arranging multiple measurement optical systems and compare the measurement results with the The reference value and the signal processing circuit 3
Compare in 6.

From this result, via the actuator control circuit 37, the actuator rod 35.35'.

By driving the 3D vertically, the mask and wafer can be moved up and down.
As shown in Figure 3, if there is a thin layer of photoret 14 outside the air gap, the color will also have a special effect. The principle that allows distance measurement to be performed without difficulty is explained in the fourth section.
This will be explained using FIG. 181 fbl. FIG. 4 181 is
The white interference fringes obtained when measuring an object having a thickness att and a refractive index n are plotted with the horizontal axis representing the spatial coordinates and the vertical axis representing the light intensity.

FIG. 4 1bl shows the shape of interference fringes when a measured object with thickness dlj refractive index n and a measured object with thickness aZ + refractive index n2 exist at the same time, and the value of z×at is 0. It is less than .5 μm, which is smaller than the value of n1Xdl. Usually, the thickness of photoresist used for patterning semiconductors etc. is 0.2 μm or less, and on the other hand, X? The air spacing used for fa exposure and proximity baking is 5 μm to 10 μm, which is more than 20 times the thickness of the photoresist.

In this case, by regarding the shape of the center beak itself as a slightly complicated and wide shape as shown in FIG. 4, 1b+, the predetermined air spacing can be measured without any inconvenience.

The spacing adjustment method described above is an excellent method when the dimensions of the sea urchin and the mask are relatively small and the flatness of both is sufficiently good, but when the dimensions of the wafer become larger, Problems arise.

For example, the diameter of the wafer has changed from the conventional 3 inches,
When the size of the wafer increases to 6 inches or finches, it becomes difficult to maintain the flatness of the wafer used at that time unless special correction means is used from the outside. For example, in X-ray printing, it is necessary to maintain an air gap of 10 μm±1 μm over the entire surface of the mask and wafer. However, when the wafer size is 6 inches or more and various processing is performed during the semiconductor process, the wafer has a complicated uneven shape, so only the actuators 35, 35', and 35'' shown in Fig. '743 are used. This is insufficient.

In order to correct this complicated unevenness and achieve a practically sufficient half-plane, we divided the holder that vacuum-adsorbs Ueno into multiple parts, and applied a load to each actuator to correct the flatness. is used. As a prerequisite for effectively using this method, it is desirable that the distances at many points on the sea urchin be measured simultaneously. The easiest way to achieve this is to enlarge the prototype system of the spacing measuring device shown in FIG. 3 so that the entire surface of the wafer can be observed at once. However, this structure has the following drawbacks.

First, the first drawback is that if the optical system shown in FIG. 3 were expanded as it is, the optical system would become too large, too heavy, and also expensive.

For example, if the distance measuring optical system is configured so that the wafer and Wollaston prism are imaged at the same magnification, if the wafer diameter is 5 inches, the diameter of the crystal plate will also be 5 inches.
If you make it into inches or finches, it becomes a very large one.

A second drawback of increasing the size is that the area required for the measurement point increases. For example, if the field of view of the white interference fringes shown in Fig. 2 is the entire 6-inch Ueno surface, then
The spacing is actually measured only along the side beaks Fi and Fs'. For this reason, in a wafer that has large unevenness in some parts, the shape of the side beak becomes complicated. Also, the side beak Fs
It is assumed that irregularities in areas other than , F, / are not detected.

An object of the present invention is to provide a measuring device that can measure the distance between a wafer and a mask over a large area using a small crystal, and at the same time control the distance between the mask and the wafer over the entire large area. It is an object of the present invention to provide an interval adjustment device that can perform the following steps.

The present invention will be explained in detail using a first embodiment of the present invention shown in FIGS. 4 and 5. This is an application of the present invention to a proxy-mirror type optical printing apparatus. A light beam emitted from a white light source such as a tungsten lamp passes through a yellow filter 25, becomes a parallel light beam by a collimation lens 37, and enters the mask 11 and the wafer 13 at right angles.

Here, since the performance of the lens 37 is not a problem, it can also be used as a lens for printing. The wafer 13 is held in a vacuum suction holder 4 divided into many parts.
It is held by 1, 4f, . The respective holders 41, 41', etc. are connected to the actuator rods 35, 35', etc. by spherical bearings.

After the light beam from the white light source passes through the half mirror 26, a part of it is reflected by the surface 101 of the mask 11.

Further, another part of the luminous flux is reflected by the surface 201 of the wafer 13.

The optical path length difference between these three beams, that is, the air gap between the mask and the wafer, is measured by a white interferometer 50. In the optical path from the mask and wafer to the interferometer, there is a cylindrical lens 38.40 whose generatrix is perpendicular to the plane of the paper.
A cylindrical lens 39 is placed with its generatrix aligned with the plane of the paper. Due to the function of these cylindrical lenses, the images of the mask 11 and wafer 13 are
Images are formed on the Wollaston prism 60 at different magnifications in the vertical and horizontal directions. The interference fringes generated on the Wollaston prism 60 are imaged onto the photodetector 30 by a relay lens 69. The photodetector used here is a two-dimensional photodetector such as a TV image pickup tube, CCD, or MOS.
From one scanning line signal, the position of the side beak, that is, the air gap between the mask and the sea urchin... Detected.

Now, since the imaging optical system formed by the cylindrical lenses 38 and 40 shown in FIG. 5 is a reduction optical system, the numerical values obtained from different scanning lines are This corresponds to points that are far apart.

The signal obtained from the photodetector 30 is processed by the signal processing circuit 36 for each scanning line, and becomes a numerical value of the interval. This numerical value and the reference value are compared, and the control circuit 3
7 drives actuator rondos 35, 35', . . . As a result, all the signals from the respective scanning lines finally obtained from the photodetector 30 become signals corresponding to the reference air spacing.

Figure 6 is a diagram of the optical system in Figure 5 observed from a direction rotated by 90 degrees with respect to the optical axis, and the optical system for measuring the distance is
The diagram shows the system arranged in parallel. As is clear from the figure, the imaging magnification is high in the in-plane direction of the paper, that is, in the direction in which the side peaks occur in the Wollaston prism, and as a result, the measurement area on the mask and the wafer becomes small.

As described above, by forming the measurement region into a slit-shaped region, the following advantages arise. FIG. 7 is a diagram showing an example of a notch on a semiconductor substrate. Typically, a semiconductor element has rectangular patterns of the same size arranged regularly in the vertical and horizontal directions. In the figure, the solid line indicates the portion to be cut out by dicing during the final cutting, but normally no special semiconductor processing is performed on this portion. By aligning the slit-shaped measurement area used in the present invention with, for example, 102°103.104 in the figure, regardless of the semiconductor pattern,
Stable spacing measurements can be made at any stage of the manufacturing process.

A second embodiment of the invention is shown in FIG. This shows an example in which the present invention is applied to an X-ray exposure apparatus. A target 71 in a vacuum container 70 is struck by an electron beam from an electron mirror 72 and generates X-rays. The generated X-rays pass through the window 73 of the beam, then pass through the mask 11 and enter the wafer 13 for exposure. Xl5t
/ With IF light, as shown in the figure, exposure is performed using diverging light (X-rays) while maintaining a constant distance between the mask and the wafer, so the distance must be maintained at a constant value with high precision. There is. The required accuracy of this spacing is 10μm ± 1μ
However, if the spacing is inaccurate, a dimensional error will occur in the pattern on the sea urchin.

The light beam from the white light source 1 is condensed by the condenser lens 23, passes through the opening of the mask 74, and is transmitted to the beam splitter 2.
It is reflected at 6. This luminous flux is transmitted through the collimation lens 3
The light beam passes through 8 and becomes a parallel beam of light, is reflected by reflection @26, and enters the mask 11° and the wafer 13 perpendicularly. The reflected light flux travels backward along the same optical path, passes through the beam splitter 26, and then enters the cylindrical lens 40. The cylindrical lens 40 has the function of enlarging the image formed by the lens 38 in one direction. Strictly speaking, if such a configuration is adopted, astigmatism will occur, but since the light source spread is small, the depth will be within the necessary and sufficient depth.

The interference fringes formed on the Wollaston prism are focused on the photodetector 30 by a lens 69, and the distance is measured. At the time of X-ray exposure, only the reflecting mirror 26 is removed from the center. The interval adjustment is performed in the same manner as in the previous embodiment using the signal processing circuit (not shown) shown in FIG.

However, by using the present invention, it has become possible to maintain the distance between the wafer 1 and the mask at a constant value in every step of the semiconductor process. In addition, the spacing between many points can be adjusted using a small number of spacing measuring devices, and the benefits are large.・It goes without saying that outside prisms, rotisserie prisms, Senarmont prisms, etc. can be used, but it is also clear that a Michelson interference needle in which one mirror of the interference needle is tilted and fixed can also be used. . Regarding the arrangement, in this text we have shown examples of three-system child rows or a crossed T-array, but it is clear that the present invention includes the use of a larger number of optical systems and two-dimensional arrangement. Dew.

[Brief explanation of the drawing]

Figure 1 is a diagram explaining the principle by which the thickness of an object can be measured using the white interferometer used in the present invention, and Figure 2 is a diagram showing the entire two-dimensional structure of the white interference fringes used in the present invention. Figure 3 is a diagram illustrating the principle by which the gap can be adjusted according to the prior patent application of the present invention, and Figure 4 shows how to accurately measure the air gap even in the presence of a thin layer such as photoresist. Figures 5 and 46 are diagrams depicting the first embodiment of the present invention, Figure WJ7 is a diagram depicting the relationship between the silicon wafer and the measurement points used in the present invention, and Figure 8 FIG. 2 is a diagram showing a second embodiment of the present invention. In the figure, 1 is a white light source, 53 is an analyzer, 60 is a Wollaston prism, 56 is an analyzer, 50 is an interferometer, 30 is a photodetector, 11 is a mask, 13 is a wafer, and 26 is a
-fmirror, ax, +f,... are wafer holding parts, 3
5. 35'... is an actuator rod, 36 is a signal processing circuit, and 37 is an actuator control circuit. Applicant Canon Co., Ltd.

Claims (1)

Scope of Claims: (1) Two objects with a distance between them are illuminated with light beams having a wavelength width, and two light beams from the objects having an optical path difference corresponding to the distance are used to form an interference pattern. and adjusting the distance between the two objects using the interference fringes, further comprising providing a rotationally asymmetrical beam width changing means in the optical path of the light beam from the object to the interference fringe forming means. Features a spacing adjustment device. (2. In the spacing adjustment device according to claim i11, the interference fringe forming means splits the light beam from the object and overlaps each other by tilting each other. (3) Claim 2 In the interval adjusting device according to item 2, the beam width changing means has a relatively small change magnification in the direction in which no similarity occurs compared to a change magnification in a direction orthogonal thereto. (4) In the spacing adjustment device according to claim (1), the interference fringe forming means has a function of independently measuring the spacing between a plurality of points, and A space adjustment device characterized in that the space between a plurality of points is set to a desired value using space adjustment means corresponding to the points.