JP3252526B2 - Position detecting device and method of manufacturing semiconductor device using the same - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a position detecting device and a method for manufacturing a semiconductor device using the same, for example, a projection exposure apparatus (stepper) for manufacturing a semiconductor device formed on a first object surface such as a reticle. This is suitable for a case in which an electronic circuit pattern is projected onto a second object surface such as a wafer by a projection lens, and relative positioning (position detection) between the reticle and the wafer during exposure transfer is performed.

[0002]

2. Description of the Related Art In a projection exposure apparatus for manufacturing semiconductor devices, relative positioning between a reticle and a wafer is an important factor for achieving high precision.

In particular, recently, high integration of semiconductor devices has required alignment accuracy of submicron or less.

One method for detecting the relative position (alignment) between a reticle and a wafer is a TTL method using a projection lens.

[0005] Of these, a reference position (reference mark) on a predetermined surface and an alignment mark provided on the wafer surface using a light beam having a wavelength different from that of exposure light when illuminating the pattern on the reticle surface and printing it on the wafer surface. There have been proposed various position detecting devices for detecting a relative position with respect to the position.

The applicant of the present invention has disclosed, for example, Japanese Patent Application No. 1-198261.
The publication proposes an observation method and an observation apparatus capable of performing high-accuracy position detection while observing an alignment mark image on a predetermined surface, for example, an imaging means surface using an imaging means such as a CCD camera.

Further, if the wavelength width of the light beam used in the position detecting device is narrow, for example, when observing an alignment mark on a wafer surface coated with a resist, many interference fringes are generated due to light reflected from the resist surface and the substrate surface. However, this causes a detection error.

In order to reduce interference fringes at this time, there has been proposed a projection exposure apparatus which observes an alignment mark by using a light source which emits a polychromatic light beam having a wide full width at half maximum of several tens nm and a wide spectral width. I have.

[0009]

A problem with the conventional position detecting device is that it is difficult to distinguish which position of the alignment mark on the wafer surface is being detected due to the relationship between the optical depth of the detecting optical system and the process. There was a point. This problem has been a cause of deteriorating the position detection accuracy.

Next, the reason will be described with reference to the drawings.

FIG. 7 is a schematic sectional view of an alignment mark.

FIG. 1 shows a case where an alignment mark is formed by etching silicon (Si) for simplicity, and a photoresist (PR) is applied on the surface.

When an image of the wafer surface is formed on the surface of an image pickup means such as a CCD by the detection optical system, for example, the wavelength λ of the used light beam is set to 6
32.8 nm, N.F. If A = 0.5, the optical depth is ± λ / 2N. If A ² , the depth is ± 1.3 μm (range 2.5 μm).

In many current semiconductor processes, the step d of the alignment mark and the resist thickness T are both 1 μm.
Or more.

As shown in FIG. 7, the illumination light is reflected light L1 reflected on the photoresist surface PRF, refracted light L2 refracted and bent by the inclined portion of the photoresist PR, and scattered at the edge of the alignment mark. There are various types of reflected light such as the scattered light L3, and an alignment mark image is created on the image pickup device surface of the CCD camera by all the reflected light.

Assuming that the refractive index of the photoresist PR is N, the geometric optical path length is (T + d) / N, which is on the order of the optical depth.

The alignment mark image on the image sensor surface is composed of the reflected light from the photoresist surface PRF and the reflected light from the wafer surface, and it is not possible to distinguish which surface is being detected. .

FIG. 8 shows a case where the cross-sectional shape of the alignment mark becomes asymmetric due to the wafer process.

As described above, the detection optical system detects reflected light from each surface without discrimination between the photoresist surface PRF, the wafer surface (Top) WF, and the wafer bottom surface (Bottom) WB. If you want to match WB, you can't.

FIG. 9 shows a process in which the transparent film on the wafer surface has a multilayer structure.

An alignment mark is formed by Locos, and a PSG is provided thereon. At this time, if the coverage of the PSG is poor and the cross-sectional shape becomes asymmetric, the image signal becomes asymmetric, the Locos is symmetric (no error), and even if it is desired to match the Locos, the detection accuracy deteriorates due to the asymmetric PSG. Come.

If the application of the photoresist PR is poor, FIG.
Even if the lower part of the alignment mark is symmetrical as shown in FIG. 10 (A), the waveform of the video signal from the imaging means is changed because the photoresist PR is asymmetric near the alignment mark.
It becomes asymmetric as shown in FIG.

As a result, the accuracy of detecting the position of the alignment mark decreases as described above. The causes of the asymmetrical video signal waveforms include the underlying pattern of the wafer,
There are various factors such as a transparent film, a translucent film, a resist, and the like provided on the surface, and when high alignment accuracy is required, these factors cannot be ignored.

According to the present invention, it is possible to detect only the position information on the wafer surface on which the transparent film is formed, for example, the bottom surface of the wafer.
Alternatively, it is possible to provide a position detection device that improves the detection accuracy of position information due to process variations by detecting only the position information of Locos, which is a transparent film on the upper portion of a wafer, and a method of manufacturing a semiconductor element using the same. Aim.

[0025]

According to a first aspect of the present invention, there is provided a position detecting apparatus for detecting a position of a wafer coated with a transparent film, the alignment mark being covered by the transparent film on the wafer. Confocal microscope, detecting means for detecting the position of the surface of the transparent film on the wafer in the optical axis direction, the position of the surface of the transparent film detected by the detecting means in the optical axis direction, and the film of the transparent film Means for moving the wafer in the optical axis direction based on the thickness and the refractive index so that the alignment mark comes to a focus position of the confocal microscope. According to a second aspect of the present invention, in the first aspect of the present invention, the alignment mark has a step, and the moving means moves the wafer so that an upper part of the step of the alignment mark is at a focus position of the confocal microscope. It is characterized in that it is moved in the optical axis direction. The invention of claim 3 is claim 1
In the invention, the alignment mark has a step,
The moving unit is configured to detect a step of the alignment mark based on an optical axis direction position of the surface of the transparent film detected by the detection unit, a thickness and a refractive index of the transparent film, and a step amount of the alignment mark. The wafer is moved in the optical axis direction such that a bottom portion is at a focus position of the confocal microscope. The position detecting device of the invention according to claim 4 is a position detecting device for detecting a position of a wafer coated with a transparent film, wherein the position detecting device has a confocal microscope for detecting an alignment mark of the wafer covered with the transparent film. The confocal microscope is provided with a lattice member in which a number of slit-shaped pinholes are arranged in the vicinity of a plane optically conjugate with the wafer in parallel with the plane, and moves the lattice member obliquely with respect to the optical axis. While the light from the alignment mark is photoelectrically converted through each of the slit-shaped pinholes, based on the obtained electric signal, when the lattice member is at a position optically conjugate with the alignment mark, The position of the alignment mark is detected. According to a fifth aspect of the present invention, in the position detecting device according to any one of the first to fourth aspects, the transparent film is a resist.

According to a sixth aspect of the present invention, there is provided an exposure apparatus for aligning a wafer with respect to a mask or a reticle using the position detecting device according to any one of the first to fifth aspects.
A projection optical system for projecting the mask or reticle pattern onto the wafer.

According to a seventh aspect of the present invention, there is provided a method of manufacturing a semiconductor device, comprising the step of exposing a wafer using the exposure apparatus according to the sixth aspect.

[0028]

FIG. 1 is a schematic view of a main part of a first embodiment when the present invention is applied to a projection exposure apparatus for manufacturing a semiconductor device.

In the drawing, reference numeral 23 denotes a projection optical system, and an illumination system 28.
2 as the first object illuminated with exposure light from
The pattern on one surface is projected on the surface of the wafer 22 as the second object. Then, semiconductor elements are manufactured from the wafer 22 through a known developing process.

The wafer 22 is coated with a transparent resist at the alignment wavelength of the non-photosensitive wavelength as a photoreceptor, and has a tilt disposed on an XY stage 24 with an interferometer and a tilt movable in the optical axis direction. It is fixedly arranged on the stage 25.

The tilt stage 25 is controlled with submicron drive accuracy using, for example, piezo as a drive source.

XY stage 24 and tilt stage 25
Are respectively monitored by a laser length measuring device (not shown), and the moving amounts are respectively controlled.

Reference numeral 29 denotes an AF detection system for wafer exposure.

Reference numeral 1 denotes a confocal microscope, which is a so-called non-TTL-type which does not pass through the projection optical system 23.
It is arranged as an off Axis microscope. The confocal microscope 1 is provided with a focus detection system (AF detection system) 2 for detecting positional information on the resist surface.

The focus detection system 2 detects position information on the resist surface by using an Air method, an optical oblique incidence method, or the like.

The confocal microscope 1 is, for example, shown in FIG.
The pinhole 32 is arranged on a plane optically conjugate with the detection surface 22 as shown in FIG. L1 is an illumination system.

When the wafer surface 22, which is the detection surface, is defocused, the light beam spreads on the pinhole surface 32 as shown in FIG. 12, and the light beam passing through the pinhole decreases.

Based on this optical principle, a detection optical system having a shallower depth of focus than a conventional optical system is constructed.

FIG. 13 is a schematic view of a main part when detecting positional information of the wafer surface 22 two-dimensionally by a confocal microscope. In FIG. 13, reference numeral 3 denotes a Nippou disk, and FIG.
As shown in FIG. 4, the disk is provided with a large number of pinholes and rotates about a shaft 3a.

Referring back to FIG. 1, reference numeral 26 denotes a control system which detects and controls positional information of the wafer 22 such as the wafer surface and the resist surface. Various information of the process is input to the control system 26 in advance.

For example, in FIG. 7, the thickness T of the photoresist PR provided on the wafer surface, the refractive index N of the photoresist PR, the step amount d of the alignment mark, and the alignment surface (for example, the surface WT of the wafer or the surface of the wafer) (Bottom WB)
And so on.

Next, a flow of detecting the position information on the wafer surface in this embodiment will be described.

Step 1. Wafer 22 to be detected
Alignment mark on the surface is confocal microscope 1
The XY stage 24 is moved so as to be located below the.

Step 2. A photoresist surface PRF is detected by the AF detection system 2 . The drive amount Z ₀ in the Z direction (optical axis direction) up to the focal plane of the confocal microscope 1 is calculated from the value at this time.

Step 3. When the input mode, for example, the confocal microscope 1 is adjusted to the wafer surface WT, the wafer 22 is driven by the tilt stage 25 in the Z direction by Z ₀ + T / N.

Step 4. The position of the wafer surface WT is detected by the confocal microscope 1.

When detecting the bottom surface WB of the wafer, the driving amount in the Z direction in step 3 is Z ₀ + (T + d) / N.

Since such a value is a known value in advance, the degree of the detection position below the surface PRF of the photoresist (PR) is given in advance by linking with the job for processing the wafer, so that the alignment signal is given. The focus position at the time of detection is determined.

Conventionally, there has been only a low-precision method of simply observing the contrast of an image and identifying its peak. However, instead of the contrast as in the present invention, the position to be grasped is set in advance. Detects reasonable and highly accurate signals.

The information of the portion above and below the step of the alignment mark is separated and detected, and the difference between the two is detected to determine which information is a desired value (for example, another means such as a wafer before resist application, The value to be detected may be determined by knowing whether or not the scaling value is close to the desired value) by analyzing a scaling that is known in advance or a variation from a uniform expansion and contraction component.

FIG. 2 is a schematic view of a main part of a second embodiment when the present invention is applied to a projection exposure apparatus for manufacturing a semiconductor device.

In this embodiment, the TTL-off A for detecting position information on the wafer surface 22 using the confocal microscope 6 via the projection optical system 23 is different from the first embodiment shown in FIG.
The difference is that it is used as an xi microscope, and other configurations are substantially the same.

The configuration of the confocal microscope 6 is basically the same as that shown in FIG.

The AF detection system 29 used at the time of exposure of the wafer 22 is also used at the time of alignment.

In the present embodiment, when a pinfall is used instead of the “Nippou disk” to achieve “confocal”, the position information is detected by moving or scanning the XY stage 24. Is required. At this time, the photoelectric conversion element does not need to be the imaging means 4 such as the CCD camera in FIG. 13, but may be, for example, the Si detector 31 in FIG. 11, and the arrangement position does not need to be located on the image plane.

FIG. 3 is a schematic view of an alignment mark according to this embodiment.

In the present embodiment, when detecting an alignment mark in the XY direction, the XY stage 24
Scanning in the direction detects position information in the X direction, and scanning in the Y direction detects position information in the Y direction. The depth of focus is determined by the size of the pinfall for “confocal”. Generally, a pinhole diameter about the limit resolution of the imaging optical system is employed. However, when the pinfall diameter becomes small, the following two problems occur. (1-1) The amount of light decreases. (1-2) The detection time becomes longer.

In particular, the detection time (1-2) will be described focusing on the X direction of the alignment mark in FIG.

The pin fall diameter is set to, for example, φ0.5 on the wafer.
When the distance is set to μm, the XY stage 24 is driven and scanned so that a pinfall forms an imaging relationship with the scanning line S1.
In order to perform high-precision position detection, the scanning lines are moved not only in S1 but also in the scanning lines S2 to S
4 scans need to be performed. Therefore, for example, scanning is performed four times in total for the scanning lines S1 to S4, which takes time.

Therefore, the present invention solves the above-mentioned problems (1-1) and (1-2) by employing a slit-shaped pinfall which is in alignment with the alignment mark. The narrow dimension of the slit is 0.5 μm, which is the same as the diameter of the pin fall, and the long dimension is in accordance with the length of the alignment mark. In order to remove the influence of the edge, the long dimension is slightly shorter than the length of the alignment mark. Specifically, in the case of the alignment mark of FIG.
The distance from 1 to S4 corresponds. The length of the long dimension should be 5 times longer than the length of the alignment mark to remove edge effects.
It is preferably shorter than μm.

FIG. 4 is a schematic view of a main part of a confocal microscope constructed for the alignment mark of FIG.

In the figure, the slit-shaped pinfall is P
_1, P ₂ and there are two Ke corresponds to 90 ° perpendicular to the alignment mark of the X-Y-direction, respectively, in FIG 3. Si detectors 41 and 42 are arranged for slit-shaped pinfalls P ₁ and P ₂ , respectively. First, X-
The Y stage 24 is scanned once, the light is received by the Si detector 41, and the position in the X direction is detected.

Next, the XY stage 24 is moved, and the position of the alignment mark in the Y direction is detected by the Si detector 42.

At this time, the Y direction in FIG. 4 is a direction perpendicular to the paper surface. The focus setting at this time is the same as the method of the first embodiment.

The slit-shaped pinfalls P ₁ and P ₂ shown in FIG.
Is vibrated in the X direction and the Y direction, and the Si detector 4
By using a CCD camera instead of 1,42, X-
A confocal microscope that achieves the object of the present invention without scanning the Y stage 24 is enabled.

When a chip and an alignment mark as shown in FIG. 5A are used, a microscope 7 for measuring in the X direction and a microscope 8 for measuring in the Y direction can be arranged as shown in FIG. 5B.

At this time, each of the microscopes 7 and 8 has one slit-shaped pinfall, and the vibration direction is also one direction. Therefore, the microscopes 7 and 8 are the same and the angle is 90 °.
What is necessary is just to arrange them orthogonally.

Instead of vibrating the slit, it is also possible to make a confocal microscope by driving the transparent portion of the liquid crystal array shutter as if the slit vibrated.

FIGS. 6A, 6B and 6C show driving examples of the liquid crystal array shutter. In FIG. 6A, the transparent portion T1 is on the left, and progresses over time like the transparent portion T2 in FIG. 6B, and is driven to the right like the transparent portion T3 in FIG. 6C.
The driving time is set to be several times during the accumulation time of the CCD camera. By using a liquid crystal array shutter, scanning can be performed at high speed and a detection optical system free from vibration can be formed, so that the stability of the system is improved.

Next, the method of determining the Focus Offset of the AF detection systems 2 and 29 for alignment in Embodiments 1 and 2 will be described.

FIG. 15 shows marks for focus measurement. This is shown as a mark on the reticle, where the black portions are Cr and the white portions are transparent portions without Cr.
When a reticle pattern is transferred and developed on a wafer using a positive resist, only the Cr portion remains, and when it is etched using the resist, only the portion without the Cr portion (white portion in FIG. 15) is dented.

Each measurement of three windows A, B, and C as shown in FIG. 16 is performed on this mark by the confocal microscope described above. Each measured value at this time is referred to as TOC _A , TOC
Assuming C _B and TOC _C ,

[0073]

(Equation 1) The value Δ is measured by changing the focus of the wafer, and the focus value where Δ = φ is set to Focus off.
set.

If there is a predetermined offset set in the measurement system, the value Δ of the Focus change is measured in the same manner as in FIG.
Focus that matches with is Focus off
set.

If the distortion of the process is as shown in FIG.
In the cases shown in (B) and (C), the value Δ is Defoc
Us does not change.

FIG. 18 (B) shows the design value (line width W ₀ ), FIG. 18 (B) shows a concave mark, and skews to the right by ε. FIG. 18 (C) shows a convex mark and ε to the left by ε Suppose it is distorted. At this time, both marks are shifted to the right by ε / 2, and the value Δ does not appear. Therefore, at this time, the line width W is measured, and the defocus amount whose value matches the predetermined amount of the line width W ₀ is defined as the defocus amount.
ff set may be used.

FIG. 18D shows an example in which measured values of the defocus versus the line width W are graphed. When the defocus is raised, the line width W of the convex mark becomes shorter, and the concave mark becomes longer.
In this case, the defocus value (bottom in the example of FIG. 18) that coincides with W = W ₀ and the unevenness is off s.
et.

If the process is complicated and the line width W ₀ is not determined, as shown in FIG.
The mark of hoe can be set to a predetermined value. Of course, the line width W ₀ of the Tokitotsu mark and a concave mark will be different, it may be off set processing.

As described above, in the first and second embodiments, by using a confocal microscope, a surface deviated by a predetermined amount from the photoresist surface is detected, thereby enabling alignment with a desired process.

In the case of an object to be measured as shown in FIGS. 8 and 9, the thickness and the refractive index of PR, PSG, and LOCOS and the surface to be matched are input.

As a result, it is possible to perform position detection while preventing a decrease in alignment accuracy due to a process error.

In the present invention, a diffraction grating (hereinafter, referred to as a “grating”) having a large number of slit-shaped pinholes (grating member) instead of a confocal slit-shaped pinhole is used in the measurement direction. The same effect as described above can be obtained even if the vibration is applied.

FIG. 19 is a schematic view of the grating G.

[0084] The shape of the grating G opening (transparent) portion W ₀ is the same as the case of one month of the slit-shaped pinholes.
If the pitch P is small, the light in the adjacent opening wraps around and becomes DC component noise light. For example, if the narrow size of the opening is 0.5 μm and the pitch is 0.6 μm (all dimensions on the wafer), the opaque portion between the openings W is 0.1 μm, and the light transmitted through the opening defocuses the wafer. If it is reflected when it is, it will enter the adjacent opening.

For this purpose, it is necessary to increase the ratio of the transparent portion and the opaque portion (duty) to 1: 1 or more. However, increasing this ratio will decrease the amount of light.

Assume that the opening is fixed at 0.5 μm. Then, an increase in the ratio of the opaque portions is equivalent to an increase in the pitch. In the case of receiving light with a CCD camera or the like, if the pitch is large, the measurement accuracy is deteriorated due to the posture accuracy of vibrating the grating.

Referring to FIG. 20, a case where the opening of the slit overlaps with one light receiving cell of the CCD camera only once during the accumulation time will be described. One opening W ₁ is receiving cells from the light receiving cell D1 between the CCD storage time D5
Suppose you move to

At this time, if the pitch P of the openings is the same as that of the five light receiving cells, the other opening W ₂ moves from the light receiving cell D 6 to the light receiving cell D 11 at that time.

When there is unevenness in the moving speed, the amount of light is inversely proportional, causing distortion of the signal waveform after photoelectric conversion, leading to deterioration of alignment accuracy.

Therefore, in the case where irregularities occur randomly as many apertures pass through one light receiving cell,
By the averaging effect, deterioration of measurement accuracy due to uneven scanning due to the posture accuracy of the opening W is eliminated.

Next, a specific example will be described using numerical examples.

The following values are converted values on the wafer surface unless otherwise specified.

[0093] accumulation time of the pitch P _c 0.2 [mu] m CCD camera between CCD camera light receiving cells T _c 30 msec grating pitch P _g 2 [mu] m grating openings W _a 0.5 [mu] m grating moving velocity V 0.5 mm of / Sec It should be noted that the time T _P for processing after photoelectric conversion is first set to “0Sec”.
Let us consider a numerical example as "c".

The number of times N through which the aperture of the grating passes through one light receiving cell within the storage time is N = V * T _c / P _g
= 500 * 0.03 / 2 = 7.5 times.

The amount of light received by the CCD camera is proportional to the number N.

In order to eliminate the “unevenness” of the speed by averaging, it is preferable that the number N is large. To increase the number N, it is necessary to set the speed V and the accumulation time _Tc to a large value, or to set the pitch _Pg to a small value. Amount is also proportional to the Deyuti D and the opening W _a of the grating G.

In the above case, D = W _a / P _g = 4 Next, a description will be given of a confocal detection system having an autofocus function at the same time.

This is achieved by moving the confocal pinhole (pinhole plate) also in the optical axis direction.

FIG. 21 is a schematic view of a main part of a third embodiment of the present invention. The figure shows a case where the present invention is applied to a grating movement type.

FIG. 21 shows a state where the wafer 106 is projected onto the projection optical system 105.
This is a case in which the detection is performed through. The illumination light 102 emitted from the fiber 101 is reflected only by the S component via the polarization beam splitter 103 and passes through the grating 104. Further, through the projection optical system 105, the wafer 10
6. Illuminate the alignment mark on 6. The reflected light is
By transmitting the light again through the projection optical system 105 and the grating 104, "confocal" detection is enabled.

FIG. 22 is a schematic view seen from the arrow side shown in FIG. The grating 104 is moved obliquely with respect to the optical axis 107 as shown by the arrow in the figure.

Since the object moves in the measurement direction, "confocal" is the same as described above. The image is received while moving in the optical axis direction, and the grating 104 is received.
Is a position optically conjugate with the wafer 106, and the total amount of received light becomes maximum, thereby enabling the autofocus function.

The grating 1 at the speed V shown in FIG.
04 When moving at an angle between the optical axis theta, and speed V _z which moves in the optical axis direction, the speed V _x of moving in a direction to measure the wafer position is as follows.

V _z = V Cos θ V _x = V Sin θ When light is received by the CCD camera 114 via the objective lens LENS, the change in the optical axis direction during the accumulation time T _c is
Assuming that the depth DOF is within the optical depth, if the depth DOF is 0.2 μm, then DOF ≧ V _z * T _c , so that 0.2 / 0.03 = 6.67 ≧ V _z (μm / Sec).

The numerical values up to now have been calculated on the wafer 106, but it is assumed that the number on the grating 104 will be calculated from now.

Assuming that the grating 104 is placed at an optical magnification of 20 times from the position of the wafer 106 for position detection, the speed V _z is 0.2 * 20 ² /0.03=2.67≧V _z (mm / S
ec).

Here, V _z = 2.67, V = 10 mm / S
Assuming that ec, from V _z = V Cos θ, θ = Cos ⁻¹ (2.67 / 10) = 74.5 (deg)
Becomes

From V _x = V Sin _θ , V _x = 10 * Sin (74.5) = 9.64 (mm / S)
ec).

[0109] variation of the measurement direction between the storage time T _c S _c
S _c = T _c * V _x = 0.03 * 9.64 = 0.289
(Mm) on the wafer 106, which is 14.5 (μm) divided by 20 times the optical magnification.

Assuming that the pitch of the grating 104 is 1 μm, the number N of times that one aperture of the grating 104 passes through one light receiving cell of the CCD camera 114 within the accumulation time is 14.5 times.

In the optical axis direction, the speed V _z (= VCos θ)
In so grating 104 moves, the shortest time T _R can be measured continuously, when _{T R = T C + T P} , detection, or location measurements for auto focus distance Z = V _z * T _R Becomes possible.

Assuming that T _R = 50 mSec, (T _P = 20
mSec) Z = 2.67 * 0.05 / 20 ² = 0.33 μm.

In this numerical example, measurement can be performed at a pitch of 0.33 μm.

When performing autofocus, it is not necessary to have a uniform moving speed of the grating 104. For this reason, focus is measured without allowing uniform movement in the direction of the white arrow 109 in FIG. 22, and after the focus is determined, the wafer 106 is moved by that value to adjust the focus.
By moving uniformly in the direction of the opposite arrow 108,
A sequence for detecting the position of the wafer 106 is also conceivable.

Of course, if measurements can be made simultaneously,
Focusing and position detection can be performed in a single movement.

As described above, a confocal detection method having an autofocus function at the same time is possible.

When the grating 104 is moved in the direction of the optical axis, the focusing relationship between the grating 104 and the CCD camera 114 is shifted, so that autofocusing is possible. However, defocusing lowers the image contrast.

As a countermeasure, (2-1) the optical system after the grating 104 moves by the same amount as the grating.

(2-2) The optical path length is inversely corrected by the amount of movement of the grating 104.

If the posture accuracy of the correction is poor, the accuracy of the detection is deteriorated.

Therefore, the grating 104 and the wafer 10
By arranging a reference mark between the position and the reference mark 6 and simultaneously detecting the reference mark during the correction, it is possible to prevent the detection accuracy from deteriorating even if the posture accuracy is poor.

In the case where the attitude accuracy is low but reproducibility is obtained, the amount of deviation can be obtained before the wafer is detected and processed as an offset.

FIG. 22 shows a correction system 1 in which the optical path length is inversely corrected.
10 shows a schematic diagram using FIG. In the figure, the correction system 110 velocity V _z / 2 of the optical path length may move as the optical path is constant when the grating 104 is moved at a speed V _z.

FIG. 21 shows an example in which the reference mark 112 is arranged and illuminated by the illumination system 113.

A reference mark 112 is illuminated from an illumination system 113 such as an LED that emits light having a wavelength different from that of the alignment light, reflected by a dichroic mirror 111, and reflected by a CCD camera 114.
Is received at.

FIG. 23 shows the grating 104 at this time.
FIG.

In the figure, the center is the shape of a grating and the windows W _s1 and W _s2 for reference marks are transparent around the center. At this time, the widths of the windows W _s1 and W _s2 are determined so that the reference mark 112 can be received by the CCD camera 114 even if the grating 104 moves sideways.

In this embodiment, as described above, the wafer 10
Gratings 104 and C that can move position information on six surfaces
The detection is performed using the CD camera 114 or the like.

[0129]

As described above, according to the present invention, only the positional information on the wafer surface on which the transparent film is formed, for example, on the bottom surface of the wafer, or the transparent film on the wafer, such as L, is detected.
By detecting only the position information of ocos, it is possible to achieve a position detection device that improves the detection accuracy of the position information due to process variations and a method of manufacturing a semiconductor element using the same.

[Brief description of the drawings]

FIG. 1 is a schematic view of a main part of a first embodiment of the present invention.

FIG. 2 is a schematic diagram of a main part of a second embodiment of the present invention.

FIG. 3 is an explanatory view of an alignment mark in FIG. 2;

FIG. 4 is an explanatory view of a confocal microscope according to the present invention.

FIG. 5 is an explanatory view of a part of FIG. 4 according to the present invention;

FIG. 6 is an explanatory diagram on an imaging element surface when a liquid crystal array shutter is used.

FIG. 7 is an explanatory sectional view of a photoresist and a wafer mark.

FIG. 8 is an explanatory sectional view of a photoresist and a wafer mark.

FIG. 9 is an explanatory sectional view of a photoresist and a wafer mark.

FIG. 10 is an explanatory sectional view of a photoresist and a wafer mark.

FIG. 11 is a schematic diagram of a main part of a confocal microscope according to the present invention.

FIG. 12 is a schematic diagram of a main part of a confocal microscope according to the present invention.

FIG. 13 is a schematic diagram of a main part of a confocal microscope according to the present invention.

FIG. 14 is an explanatory view of a Nippou disk.

FIG. 15 is a mark for focus measurement according to the present invention.

FIG. 16 is a mark for focus measurement according to the present invention.

FIG. 17 is a mark for focus measurement according to the present invention.

FIG. 18 is an explanatory sectional view of a focus measurement mark according to the present invention.

FIG. 19 is an explanatory diagram of a grating according to the present invention.

FIG. 20 is an explanatory diagram of an image sensor of a CCD camera according to the present invention.

FIG. 21 is a schematic view of a main part of a third embodiment of the present invention.

FIG. 22 is an explanatory view of a part of FIG. 21;

FIG. 23 is an explanatory view of a part of FIG. 21;

[Explanation of symbols]

 1,6 confocal microscope 2,29 focus detection system 21 reticle (mask) 22,106 wafer 23,105 projection optical system 24 XY stage 25 tilt stage 26 control system 32 pinhole 104 diffraction grating

Continuation of the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H01L 27/027 G03F 7/22 G03F 9/00

Claims (7)

(57) [Claims] 1. A position of a wafer coated with a transparent film is detected.
Position detecting device, the transparent film of the wafer
Confocal microscope to detect covered alignment marks
And the position of the surface of the transparent film on the wafer in the optical axis direction.
Detecting means for detecting, and the transparency detected by the detecting means.
The position of the surface of the bright film in the optical axis direction, the thickness of the transparent film,
Based on the folding ratio, the alignment mark
The wafer is moved to the focus position of the point microscope with the light
Means for moving in the axial direction.
Position detection device. 2. The alignment mark has a step.
The moving means may include a step of the alignment mark.
Of the confocal microscope so that the top of the
Moving the wafer in the optical axis direction.
The position detecting device according to claim 1. 3. The alignment mark has a step,
The moving means may be before the detection by the detection means.
The position of the surface of the transparent film in the optical axis direction and the thickness of the transparent film and
Based on the refractive index and the step amount of the alignment mark
The bottom of the step of the alignment mark is the confocal
Place the wafer on the optical axis so that it is in the focus position of the microscope.
The position according to claim 1, wherein the position is moved in the direction.
Position detection device. 4. A position of a wafer coated with a transparent film is detected.
Position detecting device, the transparent film of the wafer
Confocal microscope to detect covered alignment marks
The confocal microscope is optically conjugate with the wafer.
A grid with many slit-shaped pinholes arranged near a plane
A member is provided parallel to the plane, and the grating member is
While moving in an oblique direction with respect to
Light from the laser through the slit pinholes
Converted, based on the obtained electrical signal, the grid member is
The alignment mark is at an optically conjugate position
Detecting the position of the alignment mark at the time.
A position detecting device to be used as an indicator. 5. The method according to claim 1, wherein the transparent film is a resist.
The position detection according to any one of claims 1 to 4,
apparatus. 6. The position according to claim 1, wherein:
Using the position detection device Wafer to mask or reticle
Means for aligning the mask or reticle
A projection optical system for projecting the pattern onto the wafer.
An exposure apparatus, comprising: 7. A wafer using the exposure apparatus according to claim 6.
Exposing a semiconductor device.
Production method.